ci : change python3 -> python

ggml-ci

.github/workflows/build.yml

@@ -665,7 +665,7 @@ jobs:
           - build: 'llvm-arm64'
             defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON'
           - build: 'msvc-arm64'
-            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-msvc.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=O'
+            defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-msvc.cmake -DGGML_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON'
           - build: 'llvm-arm64-opencl-adreno'
             defines: '-G "Ninja Multi-Config" -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DCMAKE_PREFIX_PATH="$env:RUNNER_TEMP/opencl-arm64-release" -DGGML_OPENCL=ON -DGGML_OPENCL_USE_ADRENO_KERNELS=ON'
 
@@ -1237,7 +1237,7 @@ jobs:
 
       - name: Create release
         id: create_release
-        uses: anzz1/action-create-release@v1
+        uses: ggml-org/action-create-release@v1
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         with:

.github/workflows/docker.yml

@@ -97,10 +97,9 @@ jobs:
           GITHUB_BRANCH_NAME: ${{ github.head_ref || github.ref_name }}
           GITHUB_REPOSITORY_OWNER: '${{ github.repository_owner }}'
 
-      # https://github.com/jlumbroso/free-disk-space/tree/54081f138730dfa15788a46383842cd2f914a1be#example
       - name: Free Disk Space (Ubuntu)
         if: ${{ matrix.config.free_disk_space == true }}
-        uses: jlumbroso/free-disk-space@main
+        uses: ggml-org/free-disk-space@v1.3.1
         with:
           # this might remove tools that are actually needed,
           # if set to "true" but frees about 6 GB

.github/workflows/editorconfig.yml

@@ -23,5 +23,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
-      - uses: editorconfig-checker/action-editorconfig-checker@main
+      - uses: editorconfig-checker/action-editorconfig-checker@v2
+        with:
+          version: v3.0.3
       - run: editorconfig-checker

CODEOWNERS

@@ -3,3 +3,9 @@
 /ci/ @ggerganov
 /.devops/*.Dockerfile @ngxson
 /examples/server/ @ngxson
+/ggml/src/ggml-cuda/fattn* @JohannesGaessler
+/ggml/src/ggml-cuda/mmq.* @JohannesGaessler
+/ggml/src/ggml-cuda/mmv.* @JohannesGaessler
+/ggml/src/ggml-cuda/mmvq.* @JohannesGaessler
+/ggml/src/ggml-opt.cpp @JohannesGaessler
+/ggml/src/gguf.cpp @JohannesGaessler

CONTRIBUTING.md

@@ -1,10 +1,10 @@
 # Pull requests (for contributors)
 
 - Test your changes:
-  - Execute [the full CI locally on your machine](ci/README.md) before publishing
-  - Verify that the perplexity and the performance are not affected negatively by your changes (use `llama-perplexity` and `llama-bench`)
-  - If you modified the `ggml` source, run the `test-backend-ops` tool to check whether different backend implementations of the `ggml` operators produce consistent results (this requires access to at least two different `ggml` backends)
-  - If you modified a `ggml` operator or added a new one, add the corresponding test cases to `test-backend-ops`
+    - Execute [the full CI locally on your machine](ci/README.md) before publishing
+    - Verify that the perplexity and the performance are not affected negatively by your changes (use `llama-perplexity` and `llama-bench`)
+    - If you modified the `ggml` source, run the `test-backend-ops` tool to check whether different backend implementations of the `ggml` operators produce consistent results (this requires access to at least two different `ggml` backends)
+    - If you modified a `ggml` operator or added a new one, add the corresponding test cases to `test-backend-ops`
 - Consider allowing write access to your branch for faster reviews, as reviewers can push commits directly
 - If your PR becomes stale, don't hesitate to ping the maintainers in the comments
 
@@ -20,14 +20,104 @@
 - Avoid adding third-party dependencies, extra files, extra headers, etc.
 - Always consider cross-compatibility with other operating systems and architectures
 - Avoid fancy-looking modern STL constructs, use basic `for` loops, avoid templates, keep it simple
-- There are no strict rules for the code style, but try to follow the patterns in the code (indentation, spaces, etc.). Vertical alignment makes things more readable and easier to batch edit
+- Vertical alignment makes things more readable and easier to batch edit
 - Clean-up any trailing whitespaces, use 4 spaces for indentation, brackets on the same line, `void * ptr`, `int & a`
-- Naming usually optimizes for common prefix (see https://github.com/ggerganov/ggml/pull/302#discussion_r1243240963)
+- Use sized integer types such as `int32_t` in the public API, e.g. `size_t` may also be appropriate for allocation sizes or byte offsets
+- Declare structs with `struct foo {}` instead of `typedef struct foo {} foo`
+    - In C++ code omit optional `struct` and `enum` keyword whenever they are not necessary
+    ```cpp
+    // OK
+    llama_context * ctx;
+    const llama_rope_type rope_type;
+
+    // not OK
+    struct llama_context * ctx;
+    const enum llama_rope_type rope_type;
+    ```
+
+    _(NOTE: this guideline is yet to be applied to the `llama.cpp` codebase. New code should follow this guideline.)_
+
+- Try to follow the existing patterns in the code (indentation, spaces, etc.). In case of doubt use `clang-format` to format the added code
+- For anything not covered in the current guidelines, refer to the [C++ Core Guidelines](https://isocpp.github.io/CppCoreGuidelines/CppCoreGuidelines)
 - Tensors store data in row-major order. We refer to dimension 0 as columns, 1 as rows, 2 as matrices
 - Matrix multiplication is unconventional: [`C = ggml_mul_mat(ctx, A, B)`](https://github.com/ggerganov/llama.cpp/blob/880e352277fc017df4d5794f0c21c44e1eae2b84/ggml.h#L1058-L1064) means $C^T = A B^T \Leftrightarrow C = B A^T.$
 
 ![matmul](media/matmul.png)
 
+# Naming guidelines
+
+- Use `snake_case` for function, variable and type names
+- Naming usually optimizes for longest common prefix (see https://github.com/ggerganov/ggml/pull/302#discussion_r1243240963)
+
+    ```cpp
+    // not OK
+    int small_number;
+    int big_number;
+
+    // OK
+    int number_small;
+    int number_big;
+    ```
+
+- Enum values are always in upper case and prefixed with the enum name
+
+    ```cpp
+    enum llama_vocab_type {
+        LLAMA_VOCAB_TYPE_NONE = 0,
+        LLAMA_VOCAB_TYPE_SPM  = 1,
+        LLAMA_VOCAB_TYPE_BPE  = 2,
+        LLAMA_VOCAB_TYPE_WPM  = 3,
+        LLAMA_VOCAB_TYPE_UGM  = 4,
+        LLAMA_VOCAB_TYPE_RWKV = 5,
+    };
+    ```
+
+- The general naming pattern is `<class>_<method>`, with `<method>` being `<action>_<noun>`
+
+    ```cpp
+    llama_model_init();           // class: "llama_model",         method: "init"
+    llama_sampler_chain_remove(); // class: "llama_sampler_chain", method: "remove"
+    llama_sampler_get_seed();     // class: "llama_sampler",       method: "get_seed"
+    llama_set_embeddings();       // class: "llama_context",       method: "set_embeddings"
+    llama_n_threads();            // class: "llama_context",       method: "n_threads"
+    llama_adapter_lora_free();    // class: "llama_adapter_lora",  method: "free"
+    ```
+
+    - The `get` `<action>` can be omitted
+    - The `<noun>` can be omitted if not necessary
+    - The `_context` suffix of the `<class>` is optional. Use it to disambiguate symbols when needed
+    - Use `init`/`free` for constructor/destructor `<action>`
+
+- Use the `_t` suffix when a type is supposed to be opaque to the user - it's not relevant to them if it is a struct or anything else
+
+    ```cpp
+    typedef struct llama_context * llama_context_t;
+
+    enum llama_pooling_type llama_pooling_type(const llama_context_t ctx);
+    ```
+
+    _(NOTE: this guideline is yet to be applied to the `llama.cpp` codebase. New code should follow this guideline)_
+
+- C/C++ filenames are all lowercase with dashes. Headers use the `.h` extension. Source files use the `.c` or `.cpp` extension
+- Python filenames are all lowercase with underscores
+
+- _(TODO: abbreviations usage)_
+
+# Preprocessor directives
+
+- _(TODO: add guidelines with examples and apply them to the codebase)_
+
+    ```cpp
+    #ifdef FOO
+    #endif // FOO
+    ```
+
+# Documentation
+
+- Documentation is a community effort
+- When you need to look into the source code to figure out how to use an API consider adding a short summary to the header file for future reference
+- When you notice incorrect or outdated documentation, please update it
+
 # Resources
 
 The Github issues, PRs and discussions contain a lot of information that can be useful to get familiar with the codebase. For convenience, some of the more important information is referenced from Github projects:

README.md

@@ -69,6 +69,7 @@ Instructions for adding support for new models: [HOWTO-add-model.md](docs/develo
 - [x] [Qwen models](https://huggingface.co/models?search=Qwen/Qwen)
 - [x] [PLaMo-13B](https://github.com/ggerganov/llama.cpp/pull/3557)
 - [x] [Phi models](https://huggingface.co/models?search=microsoft/phi)
+- [x] [PhiMoE](https://github.com/ggerganov/llama.cpp/pull/11003)
 - [x] [GPT-2](https://huggingface.co/gpt2)
 - [x] [Orion 14B](https://github.com/ggerganov/llama.cpp/pull/5118)
 - [x] [InternLM2](https://huggingface.co/models?search=internlm2)
@@ -98,6 +99,7 @@ Instructions for adding support for new models: [HOWTO-add-model.md](docs/develo
 - [x] [Jais](https://huggingface.co/inceptionai/jais-13b-chat)
 - [x] [Bielik-11B-v2.3](https://huggingface.co/collections/speakleash/bielik-11b-v23-66ee813238d9b526a072408a)
 - [x] [RWKV-6](https://github.com/BlinkDL/RWKV-LM)
+- [x] [QRWKV-6](https://huggingface.co/recursal/QRWKV6-32B-Instruct-Preview-v0.1)
 - [x] [GigaChat-20B-A3B](https://huggingface.co/ai-sage/GigaChat-20B-A3B-instruct)
 
 #### Multimodal
@@ -243,6 +245,8 @@ The [Hugging Face](https://huggingface.co) platform hosts a [number of LLMs](htt
 - [Trending](https://huggingface.co/models?library=gguf&sort=trending)
 - [LLaMA](https://huggingface.co/models?sort=trending&search=llama+gguf)
 
+You can either manually download the GGUF file or directly use any `llama.cpp`-compatible models from Hugging Face by using this CLI argument: `-hf <user>/<model>[:quant]`
+
 After downloading a model, use the CLI tools to run it locally - see below.
 
 `llama.cpp` requires the model to be stored in the [GGUF](https://github.com/ggerganov/ggml/blob/master/docs/gguf.md) file format. Models in other data formats can be converted to GGUF using the `convert_*.py` Python scripts in this repo.
@@ -261,21 +265,12 @@ To learn more about model quantization, [read this documentation](examples/quant
 #### A CLI tool for accessing and experimenting with most of `llama.cpp`'s functionality.
 
 - <details open>
-    <summary>Run simple text completion</summary>
-
-    ```bash
-    llama-cli -m model.gguf -p "I believe the meaning of life is" -n 128
-
-    # I believe the meaning of life is to find your own truth and to live in accordance with it. For me, this means being true to myself and following my passions, even if they don't align with societal expectations. I think that's what I love about yoga – it's not just a physical practice, but a spiritual one too. It's about connecting with yourself, listening to your inner voice, and honoring your own unique journey.
-    ```
-
-    </details>
-
-- <details>
     <summary>Run in conversation mode</summary>
 
+    Models with a built-in chat template will automatically activate conversation mode. If this doesn't occur, you can manually enable it by adding `-cnv` and specifying a suitable chat template with `--chat-template NAME`
+
     ```bash
-    llama-cli -m model.gguf -p "You are a helpful assistant" -cnv
+    llama-cli -m model.gguf
 
     # > hi, who are you?
     # Hi there! I'm your helpful assistant! I'm an AI-powered chatbot designed to assist and provide information to users like you. I'm here to help answer your questions, provide guidance, and offer support on a wide range of topics. I'm a friendly and knowledgeable AI, and I'm always happy to help with anything you need. What's on your mind, and how can I assist you today?
@@ -287,17 +282,28 @@ To learn more about model quantization, [read this documentation](examples/quant
     </details>
 
 - <details>
-    <summary>Run with custom chat template</summary>
+    <summary>Run in conversation mode with custom chat template</summary>
 
     ```bash
-    # use the "chatml" template
-    llama-cli -m model.gguf -p "You are a helpful assistant" -cnv --chat-template chatml
+    # use the "chatml" template (use -h to see the list of supported templates)
+    llama-cli -m model.gguf -cnv --chat-template chatml
 
     # use a custom template
-    llama-cli -m model.gguf -p "You are a helpful assistant" -cnv --in-prefix 'User: ' --reverse-prompt 'User:'
+    llama-cli -m model.gguf -cnv --in-prefix 'User: ' --reverse-prompt 'User:'
     ```
 
-    [Supported templates](https://github.com/ggerganov/llama.cpp/wiki/Templates-supported-by-llama_chat_apply_template)
+    </details>
+
+- <details>
+    <summary>Run simple text completion</summary>
+
+    To disable conversation mode explicitly, use `-no-cnv`
+
+    ```bash
+    llama-cli -m model.gguf -p "I believe the meaning of life is" -n 128 -no-cnv
+
+    # I believe the meaning of life is to find your own truth and to live in accordance with it. For me, this means being true to myself and following my passions, even if they don't align with societal expectations. I think that's what I love about yoga – it's not just a physical practice, but a spiritual one too. It's about connecting with yourself, listening to your inner voice, and honoring your own unique journey.
+    ```
 
     </details>
 

ci/run.sh

@@ -299,7 +299,7 @@ function gg_run_open_llama_7b_v2 {
     (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time make -j$(nproc)                                    ) 2>&1 | tee -a $OUT/${ci}-make.log
 
-    python3 ../examples/convert_legacy_llama.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
+    python ../examples/convert_legacy_llama.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
 
     model_f16="${path_models}/ggml-model-f16.gguf"
     model_q8_0="${path_models}/ggml-model-q8_0.gguf"
@@ -326,17 +326,17 @@ function gg_run_open_llama_7b_v2 {
     ./bin/llama-quantize ${model_f16} ${model_q5_k} q5_k
     ./bin/llama-quantize ${model_f16} ${model_q6_k} q6_k
 
-    (time ./bin/llama-cli --model ${model_f16}  -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-f16.log
-    (time ./bin/llama-cli --model ${model_q8_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q8_0.log
-    (time ./bin/llama-cli --model ${model_q4_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_0.log
-    (time ./bin/llama-cli --model ${model_q4_1} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_1.log
-    (time ./bin/llama-cli --model ${model_q5_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_0.log
-    (time ./bin/llama-cli --model ${model_q5_1} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_1.log
-    (time ./bin/llama-cli --model ${model_q2_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q2_k.log
-    (time ./bin/llama-cli --model ${model_q3_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q3_k.log
-    (time ./bin/llama-cli --model ${model_q4_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_k.log
-    (time ./bin/llama-cli --model ${model_q5_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_k.log
-    (time ./bin/llama-cli --model ${model_q6_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q6_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_f16}  -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-f16.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q8_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q8_0.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q4_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_0.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q4_1} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_1.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q5_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_0.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q5_1} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_1.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q2_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q2_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q3_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q3_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q4_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q5_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q6_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q6_k.log
 
     (time ./bin/llama-perplexity --model ${model_f16}  -f ${wiki_test} -t 1 -ngl 99 -c 2048 -b 512 --chunks 4 ) 2>&1 | tee -a $OUT/${ci}-tg-f16.log
     (time ./bin/llama-perplexity --model ${model_q8_0} -f ${wiki_test} -t 1 -ngl 99 -c 2048 -b 512 --chunks 4 ) 2>&1 | tee -a $OUT/${ci}-tg-q8_0.log
@@ -433,7 +433,7 @@ function gg_run_pythia_1_4b {
     (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time make -j$(nproc)                                    ) 2>&1 | tee -a $OUT/${ci}-make.log
 
-    python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
+    python ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
 
     model_f16="${path_models}/ggml-model-f16.gguf"
     model_q8_0="${path_models}/ggml-model-q8_0.gguf"
@@ -460,17 +460,17 @@ function gg_run_pythia_1_4b {
     ./bin/llama-quantize ${model_f16} ${model_q5_k} q5_k
     ./bin/llama-quantize ${model_f16} ${model_q6_k} q6_k
 
-    (time ./bin/llama-cli --model ${model_f16}  -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-f16.log
-    (time ./bin/llama-cli --model ${model_q8_0} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q8_0.log
-    (time ./bin/llama-cli --model ${model_q4_0} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_0.log
-    (time ./bin/llama-cli --model ${model_q4_1} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_1.log
-    (time ./bin/llama-cli --model ${model_q5_0} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_0.log
-    (time ./bin/llama-cli --model ${model_q5_1} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_1.log
-    (time ./bin/llama-cli --model ${model_q2_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q2_k.log
-    (time ./bin/llama-cli --model ${model_q3_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q3_k.log
-    (time ./bin/llama-cli --model ${model_q4_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_k.log
-    (time ./bin/llama-cli --model ${model_q5_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_k.log
-    (time ./bin/llama-cli --model ${model_q6_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q6_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_f16}  -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-f16.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q8_0} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q8_0.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q4_0} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_0.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q4_1} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_1.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q5_0} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_0.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q5_1} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_1.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q2_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q2_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q3_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q3_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q4_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q5_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q6_k} -ngl 99 -c 0 -s 1234 -n 64 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q6_k.log
 
     (time ./bin/llama-perplexity --model ${model_f16}  -f ${wiki_test_60} -ngl 99 -c 128 -b 128 --chunks 1 ) 2>&1 | tee -a $OUT/${ci}-tg-f16.log
     (time ./bin/llama-perplexity --model ${model_q8_0} -f ${wiki_test_60} -ngl 99 -c 128 -b 128 --chunks 1 ) 2>&1 | tee -a $OUT/${ci}-tg-q8_0.log
@@ -564,7 +564,7 @@ function gg_run_pythia_2_8b {
     (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time make -j$(nproc)                                    ) 2>&1 | tee -a $OUT/${ci}-make.log
 
-    python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
+    python ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
 
     model_f16="${path_models}/ggml-model-f16.gguf"
     model_q8_0="${path_models}/ggml-model-q8_0.gguf"
@@ -591,17 +591,17 @@ function gg_run_pythia_2_8b {
     ./bin/llama-quantize ${model_f16} ${model_q5_k} q5_k
     ./bin/llama-quantize ${model_f16} ${model_q6_k} q6_k
 
-    (time ./bin/llama-cli --model ${model_f16}  -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-f16.log
-    (time ./bin/llama-cli --model ${model_q8_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q8_0.log
-    (time ./bin/llama-cli --model ${model_q4_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_0.log
-    (time ./bin/llama-cli --model ${model_q4_1} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_1.log
-    (time ./bin/llama-cli --model ${model_q5_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_0.log
-    (time ./bin/llama-cli --model ${model_q5_1} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_1.log
-    (time ./bin/llama-cli --model ${model_q2_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q2_k.log
-    (time ./bin/llama-cli --model ${model_q3_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q3_k.log
-    (time ./bin/llama-cli --model ${model_q4_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_k.log
-    (time ./bin/llama-cli --model ${model_q5_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_k.log
-    (time ./bin/llama-cli --model ${model_q6_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q6_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_f16}  -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-f16.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q8_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q8_0.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q4_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_0.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q4_1} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_1.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q5_0} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_0.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q5_1} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_1.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q2_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q2_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q3_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q3_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q4_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q4_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q5_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q5_k.log
+    (time ./bin/llama-cli -no-cnv --model ${model_q6_k} -t 1 -ngl 99 -c 0 -s 1234 -n 256 --ignore-eos -p "I believe the meaning of life is" ) 2>&1 | tee -a $OUT/${ci}-tg-q6_k.log
 
     (time ./bin/llama-perplexity --model ${model_f16}  -f ${wiki_test} -t 1 -ngl 99 -c 2048 -b 512 --chunks 4 ) 2>&1 | tee -a $OUT/${ci}-tg-f16.log
     (time ./bin/llama-perplexity --model ${model_q8_0} -f ${wiki_test} -t 1 -ngl 99 -c 2048 -b 512 --chunks 4 ) 2>&1 | tee -a $OUT/${ci}-tg-q8_0.log
@@ -699,7 +699,7 @@ function gg_run_embd_bge_small {
     (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time make -j$(nproc)                                    ) 2>&1 | tee -a $OUT/${ci}-make.log
 
-    python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
+    python ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
 
     model_f16="${path_models}/ggml-model-f16.gguf"
     model_q8_0="${path_models}/ggml-model-q8_0.gguf"
@@ -747,7 +747,7 @@ function gg_run_rerank_tiny {
     (time cmake -DCMAKE_BUILD_TYPE=Release ${CMAKE_EXTRA} .. ) 2>&1 | tee -a $OUT/${ci}-cmake.log
     (time make -j$(nproc)                                    ) 2>&1 | tee -a $OUT/${ci}-make.log
 
-    python3 ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
+    python ../convert_hf_to_gguf.py ${path_models} --outfile ${path_models}/ggml-model-f16.gguf
 
     model_f16="${path_models}/ggml-model-f16.gguf"
 
@@ -814,8 +814,8 @@ if [ -z ${GG_BUILD_LOW_PERF} ]; then
     mkdir -p ${mnt_models}
     ln -sfn ${mnt_models} ${SRC}/models-mnt
 
-    # Create a fresh python3 venv and enter it
-    if ! python3 -m venv "$MNT/venv"; then
+    # Create a fresh python venv and enter it
+    if ! python -m venv "$MNT/venv"; then
         echo "Error: Failed to create Python virtual environment at $MNT/venv."
         exit 1
     fi

common/arg.cpp

@@ -22,6 +22,11 @@ common_arg & common_arg::set_examples(std::initializer_list<enum llama_example>
     return *this;
 }
 
+common_arg & common_arg::set_excludes(std::initializer_list<enum llama_example> excludes) {
+    this->excludes = std::move(excludes);
+    return *this;
+}
+
 common_arg & common_arg::set_env(const char * env) {
     help = help + "\n(env: " + env + ")";
     this->env = env;
@@ -37,6 +42,10 @@ bool common_arg::in_example(enum llama_example ex) {
     return examples.find(ex) != examples.end();
 }
 
+bool common_arg::is_exclude(enum llama_example ex) {
+    return excludes.find(ex) != excludes.end();
+}
+
 bool common_arg::get_value_from_env(std::string & output) {
     if (env == nullptr) return false;
     char * value = std::getenv(env);
@@ -121,17 +130,26 @@ std::string common_arg::to_string() {
 
 static void common_params_handle_model_default(
         std::string & model,
-        std::string & model_url,
+        const std::string & model_url,
         std::string & hf_repo,
-        std::string & hf_file) {
+        std::string & hf_file,
+        const std::string & hf_token) {
     if (!hf_repo.empty()) {
         // short-hand to avoid specifying --hf-file -> default it to --model
         if (hf_file.empty()) {
             if (model.empty()) {
-                throw std::invalid_argument("error: --hf-repo requires either --hf-file or --model\n");
+                auto auto_detected = common_get_hf_file(hf_repo, hf_token);
+                if (auto_detected.first.empty() || auto_detected.second.empty()) {
+                    exit(1); // built without CURL, error message already printed
+                }
+                hf_repo = auto_detected.first;
+                hf_file = auto_detected.second;
+            } else {
+                hf_file = model;
             }
-            hf_file = model;
-        } else if (model.empty()) {
+        }
+        // make sure model path is present (for caching purposes)
+        if (model.empty()) {
             // this is to avoid different repo having same file name, or same file name in different subdirs
             std::string filename = hf_repo + "_" + hf_file;
             // to make sure we don't have any slashes in the filename
@@ -281,8 +299,8 @@ static bool common_params_parse_ex(int argc, char ** argv, common_params_context
     }
 
     // TODO: refactor model params in a common struct
-    common_params_handle_model_default(params.model,         params.model_url,         params.hf_repo,         params.hf_file);
-    common_params_handle_model_default(params.vocoder.model, params.vocoder.model_url, params.vocoder.hf_repo, params.vocoder.hf_file);
+    common_params_handle_model_default(params.model,         params.model_url,         params.hf_repo,         params.hf_file,         params.hf_token);
+    common_params_handle_model_default(params.vocoder.model, params.vocoder.model_url, params.vocoder.hf_repo, params.vocoder.hf_file, params.hf_token);
 
     if (params.escape) {
         string_process_escapes(params.prompt);
@@ -420,7 +438,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
      * - if both {LLAMA_EXAMPLE_COMMON, LLAMA_EXAMPLE_*,} are set, we will prioritize the LLAMA_EXAMPLE_* matching current example
      */
     auto add_opt = [&](common_arg arg) {
-        if (arg.in_example(ex) || arg.in_example(LLAMA_EXAMPLE_COMMON)) {
+        if ((arg.in_example(ex) || arg.in_example(LLAMA_EXAMPLE_COMMON)) && !arg.is_exclude(ex)) {
             ctx_arg.options.push_back(std::move(arg));
         }
     };
@@ -649,7 +667,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
         [](common_params & params, const std::string & value) {
             params.prompt = value;
         }
-    ));
+    ).set_excludes({LLAMA_EXAMPLE_SERVER}));
     add_opt(common_arg(
         {"--no-perf"},
         string_format("disable internal libllama performance timings (default: %s)", params.no_perf ? "true" : "false"),
@@ -673,7 +691,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
                 params.prompt.pop_back();
             }
         }
-    ));
+    ).set_excludes({LLAMA_EXAMPLE_SERVER}));
     add_opt(common_arg(
         {"--in-file"}, "FNAME",
         "an input file (repeat to specify multiple files)",
@@ -700,7 +718,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
             params.prompt = ss.str();
             fprintf(stderr, "Read %zu bytes from binary file %s\n", params.prompt.size(), value.c_str());
         }
-    ));
+    ).set_excludes({LLAMA_EXAMPLE_SERVER}));
     add_opt(common_arg(
         {"-e", "--escape"},
         string_format("process escapes sequences (\\n, \\r, \\t, \\', \\\", \\\\) (default: %s)", params.escape ? "true" : "false"),
@@ -759,15 +777,19 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
     ).set_examples({LLAMA_EXAMPLE_MAIN, LLAMA_EXAMPLE_SERVER}));
     add_opt(common_arg(
         {"-cnv", "--conversation"},
-        string_format(
-            "run in conversation mode:\n"
-            "- does not print special tokens and suffix/prefix\n"
-            "- interactive mode is also enabled\n"
-            "(default: %s)",
-            params.conversation ? "true" : "false"
-        ),
+        "run in conversation mode:\n"
+        "- does not print special tokens and suffix/prefix\n"
+        "- interactive mode is also enabled\n"
+        "(default: auto enabled if chat template is available)",
         [](common_params & params) {
-            params.conversation = true;
+            params.conversation_mode = COMMON_CONVERSATION_MODE_ENABLED;
+        }
+    ).set_examples({LLAMA_EXAMPLE_MAIN}));
+    add_opt(common_arg(
+        {"-no-cnv", "--no-conversation"},
+        "force disable conversation mode (default: false)",
+        [](common_params & params) {
+            params.conversation_mode = COMMON_CONVERSATION_MODE_DISABLED;
         }
     ).set_examples({LLAMA_EXAMPLE_MAIN}));
     add_opt(common_arg(
@@ -1574,21 +1596,23 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
         }
     ).set_env("LLAMA_ARG_MODEL_URL"));
     add_opt(common_arg(
-        {"-hfr", "--hf-repo"}, "REPO",
-        "Hugging Face model repository (default: unused)",
+        {"-hf", "-hfr", "--hf-repo"}, "<user>/<model>[:quant]",
+        "Hugging Face model repository; quant is optional, case-insensitive, default to Q4_K_M, or falls back to the first file in the repo if Q4_K_M doesn't exist.\n"
+        "example: unsloth/phi-4-GGUF:q4_k_m\n"
+        "(default: unused)",
         [](common_params & params, const std::string & value) {
             params.hf_repo = value;
         }
     ).set_env("LLAMA_ARG_HF_REPO"));
     add_opt(common_arg(
         {"-hff", "--hf-file"}, "FILE",
-        "Hugging Face model file (default: unused)",
+        "Hugging Face model file. If specified, it will override the quant in --hf-repo (default: unused)",
         [](common_params & params, const std::string & value) {
             params.hf_file = value;
         }
     ).set_env("LLAMA_ARG_HF_FILE"));
     add_opt(common_arg(
-        {"-hfrv", "--hf-repo-v"}, "REPO",
+        {"-hfv", "-hfrv", "--hf-repo-v"}, "<user>/<model>[:quant]",
         "Hugging Face model repository for the vocoder model (default: unused)",
         [](common_params & params, const std::string & value) {
             params.vocoder.hf_repo = value;

common/arg.h

@@ -12,6 +12,7 @@
 
 struct common_arg {
     std::set<enum llama_example> examples = {LLAMA_EXAMPLE_COMMON};
+    std::set<enum llama_example> excludes = {};
     std::vector<const char *> args;
     const char * value_hint   = nullptr; // help text or example for arg value
     const char * value_hint_2 = nullptr; // for second arg value
@@ -53,9 +54,11 @@ struct common_arg {
     ) : args(args), value_hint(value_hint), value_hint_2(value_hint_2), help(help), handler_str_str(handler) {}
 
     common_arg & set_examples(std::initializer_list<enum llama_example> examples);
+    common_arg & set_excludes(std::initializer_list<enum llama_example> excludes);
     common_arg & set_env(const char * env);
     common_arg & set_sparam();
     bool in_example(enum llama_example ex);
+    bool is_exclude(enum llama_example ex);
     bool get_value_from_env(std::string & output);
     bool has_value_from_env();
     std::string to_string();

common/common.cpp

@@ -2,6 +2,9 @@
 #define _SILENCE_CXX17_CODECVT_HEADER_DEPRECATION_WARNING
 #endif
 
+#include "ggml.h"
+#include "gguf.h"
+
 #include "common.h"
 #include "log.h"
 // Change JSON_ASSERT from assert() to GGML_ASSERT:
@@ -70,6 +73,22 @@
 #include <sys/syslimits.h>
 #endif
 #define LLAMA_CURL_MAX_URL_LENGTH 2084 // Maximum URL Length in Chrome: 2083
+
+//
+// CURL utils
+//
+
+using curl_ptr = std::unique_ptr<CURL, decltype(&curl_easy_cleanup)>;
+
+// cannot use unique_ptr for curl_slist, because we cannot update without destroying the old one
+struct curl_slist_ptr {
+    struct curl_slist * ptr = nullptr;
+    ~curl_slist_ptr() {
+        if (ptr) {
+            curl_slist_free_all(ptr);
+        }
+    }
+};
 #endif // LLAMA_USE_CURL
 
 using json = nlohmann::ordered_json;
@@ -854,21 +873,23 @@ struct common_init_result common_init_from_params(common_params & params) {
         return iparams;
     }
 
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     if (params.reranking) {
         bool ok = true;
 
-        if (llama_token_bos(model) == LLAMA_TOKEN_NULL) {
-            LOG_WRN("%s: warning: model does not have a  BOS token, reranking will not work\n", __func__);
+        if (llama_vocab_bos(vocab) == LLAMA_TOKEN_NULL) {
+            LOG_WRN("%s: warning: vocab does not have a  BOS token, reranking will not work\n", __func__);
             ok = false;
         }
 
-        if (llama_token_eos(model) == LLAMA_TOKEN_NULL) {
-            LOG_WRN("%s: warning: model does not have an EOS token, reranking will not work\n", __func__);
+        if (llama_vocab_eos(vocab) == LLAMA_TOKEN_NULL) {
+            LOG_WRN("%s: warning: vocab does not have an EOS token, reranking will not work\n", __func__);
             ok = false;
         }
 
-        if (llama_token_sep(model) == LLAMA_TOKEN_NULL) {
-            LOG_WRN("%s: warning: model does not have a  SEP token, reranking will not work\n", __func__);
+        if (llama_vocab_sep(vocab) == LLAMA_TOKEN_NULL) {
+            LOG_WRN("%s: warning: vocab does not have a  SEP token, reranking will not work\n", __func__);
             ok = false;
         }
 
@@ -881,7 +902,7 @@ struct common_init_result common_init_from_params(common_params & params) {
 
     auto cparams = common_context_params_to_llama(params);
 
-    llama_context * lctx = llama_new_context_with_model(model, cparams);
+    llama_context * lctx = llama_init_from_model(model, cparams);
     if (lctx == NULL) {
         LOG_ERR("%s: failed to create context with model '%s'\n", __func__, params.model.c_str());
         llama_model_free(model);
@@ -895,7 +916,7 @@ struct common_init_result common_init_from_params(common_params & params) {
 
     if (!params.control_vectors.empty()) {
         if (params.control_vector_layer_start <= 0) params.control_vector_layer_start = 1;
-        if (params.control_vector_layer_end   <= 0) params.control_vector_layer_end   = llama_n_layer(model);
+        if (params.control_vector_layer_end   <= 0) params.control_vector_layer_end   = llama_model_n_layer(model);
 
         const auto cvec = common_control_vector_load(params.control_vectors);
         if (cvec.n_embd == -1) {
@@ -905,12 +926,13 @@ struct common_init_result common_init_from_params(common_params & params) {
             return iparams;
         }
 
-        int err = llama_control_vector_apply(lctx,
-                                             cvec.data.data(),
-                                             cvec.data.size(),
-                                             cvec.n_embd,
-                                             params.control_vector_layer_start,
-                                             params.control_vector_layer_end);
+        int err = llama_apply_adapter_cvec(
+                lctx,
+                cvec.data.data(),
+                cvec.data.size(),
+                cvec.n_embd,
+                params.control_vector_layer_start,
+                params.control_vector_layer_end);
         if (err) {
             llama_free(lctx);
             llama_model_free(model);
@@ -921,8 +943,8 @@ struct common_init_result common_init_from_params(common_params & params) {
 
     // load and optionally apply lora adapters
     for (auto & la : params.lora_adapters) {
-        llama_lora_adapter_ptr lora;
-        lora.reset(llama_lora_adapter_init(model, la.path.c_str()));
+        llama_adapter_lora_ptr lora;
+        lora.reset(llama_adapter_lora_init(model, la.path.c_str()));
         if (lora == nullptr) {
             LOG_ERR("%s: failed to apply lora adapter '%s'\n", __func__, la.path.c_str());
             llama_free(lctx);
@@ -935,17 +957,17 @@ struct common_init_result common_init_from_params(common_params & params) {
     }
 
     if (!params.lora_init_without_apply) {
-        common_lora_adapters_apply(lctx, params.lora_adapters);
+        common_set_adapter_lora(lctx, params.lora_adapters);
     }
 
-    if (params.sampling.ignore_eos && llama_token_eos(model) == LLAMA_TOKEN_NULL) {
-        LOG_WRN("%s: warning: model does not have an EOS token, ignoring --ignore-eos\n", __func__);
+    if (params.sampling.ignore_eos && llama_vocab_eos(vocab) == LLAMA_TOKEN_NULL) {
+        LOG_WRN("%s: warning: vocab does not have an EOS token, ignoring --ignore-eos\n", __func__);
         params.sampling.ignore_eos = false;
     }
 
     if (params.sampling.ignore_eos) {
-        for (llama_token i = 0; i < llama_n_vocab(model); i++) {
-            if (llama_token_is_eog(model, i)) {
+        for (llama_token i = 0; i < llama_vocab_n_tokens(vocab); i++) {
+            if (llama_vocab_is_eog(vocab, i)) {
                 LOG_INF("%s: added %s logit bias = %f\n", __func__, common_token_to_piece(lctx, i).c_str(), -INFINITY);
                 params.sampling.logit_bias.push_back({i, -INFINITY});
             }
@@ -966,8 +988,9 @@ struct common_init_result common_init_from_params(common_params & params) {
         LOG_WRN("%s: warming up the model with an empty run - please wait ... (--no-warmup to disable)\n", __func__);
 
         std::vector<llama_token> tmp;
-        llama_token bos = llama_token_bos(model);
-        llama_token eos = llama_token_eos(model);
+        llama_token bos = llama_vocab_bos(vocab);
+        llama_token eos = llama_vocab_eos(vocab);
+
         // some models (e.g. T5) don't have a BOS token
         if (bos != LLAMA_TOKEN_NULL) {
             tmp.push_back(bos);
@@ -1002,11 +1025,11 @@ struct common_init_result common_init_from_params(common_params & params) {
     return iparams;
 }
 
-void common_lora_adapters_apply(struct llama_context * ctx, std::vector<common_lora_adapter_info> & lora) {
-    llama_lora_adapter_clear(ctx);
+void common_set_adapter_lora(struct llama_context * ctx, std::vector<common_adapter_lora_info> & lora) {
+    llama_clear_adapter_lora(ctx);
     for (auto & la : lora) {
         if (la.scale != 0.0f) {
-            llama_lora_adapter_set(ctx, la.ptr, la.scale);
+            llama_set_adapter_lora(ctx, la.ptr, la.scale);
         }
     }
 }
@@ -1123,7 +1146,8 @@ static bool curl_perform_with_retry(const std::string & url, CURL * curl, int ma
 
 static bool common_download_file(const std::string & url, const std::string & path, const std::string & hf_token) {
     // Initialize libcurl
-    std::unique_ptr<CURL, decltype(&curl_easy_cleanup)> curl(curl_easy_init(), &curl_easy_cleanup);
+    curl_ptr       curl(curl_easy_init(), &curl_easy_cleanup);
+    curl_slist_ptr http_headers;
     if (!curl) {
         LOG_ERR("%s: error initializing libcurl\n", __func__);
         return false;
@@ -1137,11 +1161,9 @@ static bool common_download_file(const std::string & url, const std::string & pa
 
     // Check if hf-token or bearer-token was specified
     if (!hf_token.empty()) {
-      std::string auth_header = "Authorization: Bearer ";
-      auth_header += hf_token.c_str();
-      struct curl_slist *http_headers = NULL;
-      http_headers = curl_slist_append(http_headers, auth_header.c_str());
-      curl_easy_setopt(curl.get(), CURLOPT_HTTPHEADER, http_headers);
+        std::string auth_header = "Authorization: Bearer " + hf_token;
+        http_headers.ptr = curl_slist_append(http_headers.ptr, auth_header.c_str());
+        curl_easy_setopt(curl.get(), CURLOPT_HTTPHEADER, http_headers.ptr);
     }
 
 #if defined(_WIN32)
@@ -1437,6 +1459,80 @@ struct llama_model * common_load_model_from_hf(
     return common_load_model_from_url(model_url, local_path, hf_token, params);
 }
 
+/**
+ * Allow getting the HF file from the HF repo with tag (like ollama), for example:
+ * - bartowski/Llama-3.2-3B-Instruct-GGUF:q4
+ * - bartowski/Llama-3.2-3B-Instruct-GGUF:Q4_K_M
+ * - bartowski/Llama-3.2-3B-Instruct-GGUF:q5_k_s
+ * Tag is optional, default to "latest" (meaning it checks for Q4_K_M first, then Q4, then if not found, return the first GGUF file in repo)
+ *
+ * Return pair of <repo, file> (with "repo" already having tag removed)
+ *
+ * Note: we use the Ollama-compatible HF API, but not using the blobId. Instead, we use the special "ggufFile" field which returns the value for "hf_file". This is done to be backward-compatible with existing cache files.
+ */
+std::pair<std::string, std::string> common_get_hf_file(const std::string & hf_repo_with_tag, const std::string & hf_token) {
+    auto parts = string_split<std::string>(hf_repo_with_tag, ':');
+    std::string tag = parts.size() > 1 ? parts.back() : "latest";
+    std::string hf_repo = parts[0];
+    if (string_split<std::string>(hf_repo, '/').size() != 2) {
+        throw std::invalid_argument("error: invalid HF repo format, expected <user>/<model>[:quant]\n");
+    }
+
+    // fetch model info from Hugging Face Hub API
+    json model_info;
+    curl_ptr       curl(curl_easy_init(), &curl_easy_cleanup);
+    curl_slist_ptr http_headers;
+    std::string res_str;
+    std::string url = "https://huggingface.co/v2/" + hf_repo + "/manifests/" + tag;
+    curl_easy_setopt(curl.get(), CURLOPT_URL, url.c_str());
+    curl_easy_setopt(curl.get(), CURLOPT_NOPROGRESS, 1L);
+    typedef size_t(*CURLOPT_WRITEFUNCTION_PTR)(void * ptr, size_t size, size_t nmemb, void * data);
+    auto write_callback = [](void * ptr, size_t size, size_t nmemb, void * data) -> size_t {
+        static_cast<std::string *>(data)->append((char * ) ptr, size * nmemb);
+        return size * nmemb;
+    };
+    curl_easy_setopt(curl.get(), CURLOPT_WRITEFUNCTION, static_cast<CURLOPT_WRITEFUNCTION_PTR>(write_callback));
+    curl_easy_setopt(curl.get(), CURLOPT_WRITEDATA, &res_str);
+#if defined(_WIN32)
+    curl_easy_setopt(curl.get(), CURLOPT_SSL_OPTIONS, CURLSSLOPT_NATIVE_CA);
+#endif
+    if (!hf_token.empty()) {
+        std::string auth_header = "Authorization: Bearer " + hf_token;
+        http_headers.ptr = curl_slist_append(http_headers.ptr, auth_header.c_str());
+    }
+    // Important: the User-Agent must be "llama-cpp" to get the "ggufFile" field in the response
+    http_headers.ptr = curl_slist_append(http_headers.ptr, "User-Agent: llama-cpp");
+    http_headers.ptr = curl_slist_append(http_headers.ptr, "Accept: application/json");
+    curl_easy_setopt(curl.get(), CURLOPT_HTTPHEADER, http_headers.ptr);
+
+    CURLcode res = curl_easy_perform(curl.get());
+
+    if (res != CURLE_OK) {
+        throw std::runtime_error("error: cannot make GET request to HF API");
+    }
+
+    long res_code;
+    curl_easy_getinfo(curl.get(), CURLINFO_RESPONSE_CODE, &res_code);
+    if (res_code == 200) {
+        model_info = json::parse(res_str);
+    } else if (res_code == 401) {
+        throw std::runtime_error("error: model is private or does not exist; if you are accessing a gated model, please provide a valid HF token");
+    } else {
+        throw std::runtime_error(string_format("error from HF API, response code: %ld, data: %s", res_code, res_str.c_str()));
+    }
+
+    // check response
+    if (!model_info.contains("ggufFile")) {
+        throw std::runtime_error("error: model does not have ggufFile");
+    }
+    json & gguf_file = model_info.at("ggufFile");
+    if (!gguf_file.contains("rfilename")) {
+        throw std::runtime_error("error: ggufFile does not have rfilename");
+    }
+
+    return std::make_pair(hf_repo, gguf_file.at("rfilename"));
+}
+
 #else
 
 struct llama_model * common_load_model_from_url(
@@ -1458,6 +1554,11 @@ struct llama_model * common_load_model_from_hf(
     return nullptr;
 }
 
+std::pair<std::string, std::string> common_get_hf_file(const std::string &, const std::string &) {
+    LOG_WRN("%s: llama.cpp built without libcurl, downloading from Hugging Face not supported.\n", __func__);
+    return std::make_pair("", "");
+}
+
 #endif // LLAMA_USE_CURL
 
 //
@@ -1556,21 +1657,23 @@ std::vector<llama_token> common_tokenize(
            const std::string & text,
                         bool   add_special,
                         bool   parse_special) {
-    return common_tokenize(llama_get_model(ctx), text, add_special, parse_special);
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+    return common_tokenize(vocab, text, add_special, parse_special);
 }
 
 std::vector<llama_token> common_tokenize(
-    const struct llama_model * model,
+    const struct llama_vocab * vocab,
            const std::string & text,
                         bool   add_special,
                         bool   parse_special) {
     // upper limit for the number of tokens
     int n_tokens = text.length() + 2 * add_special;
     std::vector<llama_token> result(n_tokens);
-    n_tokens = llama_tokenize(model, text.data(), text.length(), result.data(), result.size(), add_special, parse_special);
+    n_tokens = llama_tokenize(vocab, text.data(), text.length(), result.data(), result.size(), add_special, parse_special);
     if (n_tokens < 0) {
         result.resize(-n_tokens);
-        int check = llama_tokenize(model, text.data(), text.length(), result.data(), result.size(), add_special, parse_special);
+        int check = llama_tokenize(vocab, text.data(), text.length(), result.data(), result.size(), add_special, parse_special);
         GGML_ASSERT(check == -n_tokens);
     } else {
         result.resize(n_tokens);
@@ -1579,12 +1682,18 @@ std::vector<llama_token> common_tokenize(
 }
 
 std::string common_token_to_piece(const struct llama_context * ctx, llama_token token, bool special) {
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+    return common_token_to_piece(vocab, token, special);
+}
+
+std::string common_token_to_piece(const struct llama_vocab * vocab, llama_token token, bool special) {
     std::string piece;
     piece.resize(piece.capacity());  // using string internal cache, 15 bytes + '\n'
-    const int n_chars = llama_token_to_piece(llama_get_model(ctx), token, &piece[0], piece.size(), 0, special);
+    const int n_chars = llama_token_to_piece(vocab, token, &piece[0], piece.size(), 0, special);
     if (n_chars < 0) {
         piece.resize(-n_chars);
-        int check = llama_token_to_piece(llama_get_model(ctx), token, &piece[0], piece.size(), 0, special);
+        int check = llama_token_to_piece(vocab, token, &piece[0], piece.size(), 0, special);
         GGML_ASSERT(check == -n_chars);
     }
     else {
@@ -1594,13 +1703,19 @@ std::string common_token_to_piece(const struct llama_context * ctx, llama_token
     return piece;
 }
 
-std::string common_detokenize(llama_context * ctx, const std::vector<llama_token> & tokens, bool special) {
+std::string common_detokenize(const struct llama_context * ctx, const std::vector<llama_token> & tokens, bool special) {
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+    return common_detokenize(vocab, tokens, special);
+}
+
+std::string common_detokenize(const struct llama_vocab * vocab, const std::vector<llama_token> & tokens, bool special) {
     std::string text;
     text.resize(std::max(text.capacity(), tokens.size()));
-    int32_t n_chars = llama_detokenize(llama_get_model(ctx), tokens.data(), (int32_t)tokens.size(), &text[0], (int32_t)text.size(), false, special);
+    int32_t n_chars = llama_detokenize(vocab, tokens.data(), (int32_t)tokens.size(), &text[0], (int32_t)text.size(), false, special);
     if (n_chars < 0) {
         text.resize(-n_chars);
-        n_chars = llama_detokenize(llama_get_model(ctx), tokens.data(), (int32_t)tokens.size(), &text[0], (int32_t)text.size(), false, special);
+        n_chars = llama_detokenize(vocab, tokens.data(), (int32_t)tokens.size(), &text[0], (int32_t)text.size(), false, special);
         GGML_ASSERT(n_chars <= (int32_t)text.size());  // whitespace trimming is performed after per-token detokenization
     }
 
@@ -1615,20 +1730,13 @@ std::string common_detokenize(llama_context * ctx, const std::vector<llama_token
 //
 
 std::string common_get_builtin_chat_template(const struct llama_model * model) {
-    static const char * template_key = "tokenizer.chat_template";
-    // call with NULL buffer to get the total size of the string
-    int32_t res = llama_model_meta_val_str(model, template_key, NULL, 0);
-    if (res > 0) {
-        std::vector<char> model_template(res + 1, 0);
-        llama_model_meta_val_str(model, template_key, model_template.data(), model_template.size());
-        return std::string(model_template.data(), model_template.size() - 1);
-    }
-    return "";
+    const char * ptr_tmpl = llama_model_chat_template(model);
+    return ptr_tmpl == nullptr ? "" : ptr_tmpl;
 }
 
 bool common_chat_verify_template(const std::string & tmpl) {
     llama_chat_message chat[] = {{"user", "test"}};
-    int res = llama_chat_apply_template(nullptr, tmpl.c_str(), chat, 1, true, nullptr, 0);
+    const int res = llama_chat_apply_template(tmpl.c_str(), chat, 1, true, nullptr, 0);
     return res >= 0;
 }
 
@@ -1639,16 +1747,16 @@ std::string common_chat_apply_template(const struct llama_model * model,
     int alloc_size = 0;
     bool fallback = false; // indicate if we must fallback to default chatml
     std::vector<llama_chat_message> chat;
-    for (auto & msg : msgs) {
+    for (const auto & msg : msgs) {
         chat.push_back({msg.role.c_str(), msg.content.c_str()});
         alloc_size += (msg.role.size() + msg.content.size()) * 1.25;
     }
 
-    const char * ptr_tmpl = tmpl.empty() ? nullptr : tmpl.c_str();
+    const char * ptr_tmpl = tmpl.empty() ? llama_model_chat_template(model) : tmpl.c_str();
     std::vector<char> buf(alloc_size);
 
     // run the first time to get the total output length
-    int32_t res = llama_chat_apply_template(model, ptr_tmpl, chat.data(), chat.size(), add_ass, buf.data(), buf.size());
+    int32_t res = llama_chat_apply_template(ptr_tmpl, chat.data(), chat.size(), add_ass, buf.data(), buf.size());
 
     // error: chat template is not supported
     if (res < 0) {
@@ -1656,18 +1764,17 @@ std::string common_chat_apply_template(const struct llama_model * model,
             // if the custom "tmpl" is not supported, we throw an error
             // this is a bit redundant (for good), since we're not sure if user validated the custom template with llama_chat_verify_template()
             throw std::runtime_error("this custom template is not supported");
-        } else {
-            // If the built-in template is not supported, we default to chatml
-            res = llama_chat_apply_template(nullptr, "chatml", chat.data(), chat.size(), add_ass, buf.data(), buf.size());
-            fallback = true;
         }
+
+        // If the built-in template is not supported, we default to chatml
+        res = llama_chat_apply_template("chatml", chat.data(), chat.size(), add_ass, buf.data(), buf.size());
+        fallback = true;
     }
 
     // if it turns out that our buffer is too small, we resize it
     if ((size_t) res > buf.size()) {
         buf.resize(res);
         res = llama_chat_apply_template(
-            fallback ? nullptr : model,
             fallback ? "chatml" : ptr_tmpl,
             chat.data(), chat.size(), add_ass, buf.data(), buf.size());
     }

common/common.h

@@ -24,11 +24,11 @@
 
 #define DEFAULT_MODEL_PATH "models/7B/ggml-model-f16.gguf"
 
-struct common_lora_adapter_info {
+struct common_adapter_lora_info {
     std::string path;
     float scale;
 
-    struct llama_lora_adapter * ptr;
+    struct llama_adapter_lora * ptr;
 };
 
 using llama_tokens = std::vector<llama_token>;
@@ -103,6 +103,12 @@ enum dimre_method {
     DIMRE_METHOD_MEAN,
 };
 
+enum common_conversation_mode {
+    COMMON_CONVERSATION_MODE_DISABLED = 0,
+    COMMON_CONVERSATION_MODE_ENABLED  = 1,
+    COMMON_CONVERSATION_MODE_AUTO     = 2,
+};
+
 // sampling parameters
 struct common_params_sampling {
     uint32_t seed = LLAMA_DEFAULT_SEED; // the seed used to initialize llama_sampler
@@ -246,8 +252,8 @@ struct common_params {
     std::vector<std::string> antiprompt; // strings upon which more user input is prompted (a.k.a. reverse prompts)
     std::vector<llama_model_kv_override> kv_overrides;
 
-    bool lora_init_without_apply = false; // only load lora to memory, but do not apply it to ctx (user can manually apply lora later using llama_lora_adapter_apply)
-    std::vector<common_lora_adapter_info> lora_adapters; // lora adapter path with user defined scale
+    bool lora_init_without_apply = false; // only load lora to memory, but do not apply it to ctx (user can manually apply lora later using llama_adapter_lora_apply)
+    std::vector<common_adapter_lora_info> lora_adapters; // lora adapter path with user defined scale
 
     std::vector<common_control_vector_load_info> control_vectors; // control vector with user defined scale
 
@@ -275,7 +281,6 @@ struct common_params {
     bool special           = false; // enable special token output
     bool interactive       = false; // interactive mode
     bool interactive_first = false; // wait for user input immediately
-    bool conversation      = false; // conversation mode (does not print special tokens and suffix/prefix)
     bool prompt_cache_all  = false; // save user input and generations to prompt cache
     bool prompt_cache_ro   = false; // open the prompt cache read-only and do not update it
 
@@ -301,6 +306,8 @@ struct common_params {
     ggml_type cache_type_k = GGML_TYPE_F16; // KV cache data type for the K
     ggml_type cache_type_v = GGML_TYPE_F16; // KV cache data type for the V
 
+    common_conversation_mode conversation_mode = COMMON_CONVERSATION_MODE_AUTO;
+
     // multimodal models (see examples/llava)
     std::string mmproj = "";        // path to multimodal projector                                         // NOLINT
     std::vector<std::string> image; // path to image file(s)
@@ -454,6 +461,11 @@ static bool string_starts_with(const std::string & str,
     return str.rfind(prefix, 0) == 0;
 }
 
+static bool string_ends_with(const std::string & str,
+                               const std::string & suffix) {  // While we wait for C++20's std::string::ends_with...
+    return str.size() >= suffix.size() && str.compare(str.size()-suffix.size(), suffix.size(), suffix) == 0;
+}
+
 bool string_parse_kv_override(const char * data, std::vector<llama_model_kv_override> & overrides);
 void string_process_escapes(std::string & input);
 
@@ -481,7 +493,7 @@ struct common_init_result {
     llama_model_ptr   model;
     llama_context_ptr context;
 
-    std::vector<llama_lora_adapter_ptr> lora;
+    std::vector<llama_adapter_lora_ptr> lora;
 };
 
 struct common_init_result     common_init_from_params(common_params & params);
@@ -501,9 +513,12 @@ struct llama_model * common_load_model_from_hf(
     const std::string & local_path,
     const std::string & hf_token,
     const struct llama_model_params & params);
+std::pair<std::string, std::string> common_get_hf_file(
+    const std::string & hf_repo_with_tag,
+    const std::string & hf_token);
 
 // clear LoRA adapters from context, then apply new list of adapters
-void common_lora_adapters_apply(struct llama_context * ctx, std::vector<common_lora_adapter_info> & lora);
+void common_set_adapter_lora(struct llama_context * ctx, std::vector<common_adapter_lora_info> & lora);
 
 //
 // Batch utils
@@ -541,7 +556,7 @@ std::vector<llama_token> common_tokenize(
                         bool   parse_special = false);
 
 std::vector<llama_token> common_tokenize(
-    const struct llama_model * model,
+    const struct llama_vocab * vocab,
            const std::string & text,
                         bool   add_special,
                         bool   parse_special = false);
@@ -553,11 +568,21 @@ std::string common_token_to_piece(
                        llama_token   token,
                        bool          special = true);
 
+std::string common_token_to_piece(
+          const struct llama_vocab * vocab,
+                       llama_token   token,
+                       bool          special = true);
+
 // detokenizes a vector of tokens into a string
 // should work similar to Python's `tokenizer.decode`
 // optionally renders special/control tokens
 std::string common_detokenize(
-                         llama_context * ctx,
+            const struct llama_context * ctx,
+        const std::vector<llama_token> & tokens,
+                                  bool   special = true);
+
+std::string common_detokenize(
+              const struct llama_vocab * vocab,
         const std::vector<llama_token> & tokens,
                                   bool   special = true);
 

common/sampling.cpp

@@ -113,7 +113,10 @@ struct common_sampler {
     void set_logits(struct llama_context * ctx, int idx) {
         const auto * logits = llama_get_logits_ith(ctx, idx);
 
-        const int n_vocab = llama_n_vocab(llama_get_model(ctx));
+        const llama_model * model = llama_get_model(ctx);
+        const llama_vocab * vocab = llama_model_get_vocab(model);
+
+        const int n_vocab = llama_vocab_n_tokens(vocab);
 
         cur.resize(n_vocab);
 
@@ -142,13 +145,15 @@ std::string common_params_sampling::print() const {
 }
 
 struct common_sampler * common_sampler_init(const struct llama_model * model, const struct common_params_sampling & params) {
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     llama_sampler_chain_params lparams = llama_sampler_chain_default_params();
 
     lparams.no_perf = params.no_perf;
 
     auto * result = new common_sampler {
         /* .params = */ params,
-        /* .grmr   = */ llama_sampler_init_grammar(model, params.grammar.c_str(), "root"),
+        /* .grmr   = */ llama_sampler_init_grammar(vocab, params.grammar.c_str(), "root"),
         /* .chain  = */ llama_sampler_chain_init(lparams),
         /* .prev   = */ ring_buffer<llama_token>(std::max(32, params.n_prev)),
         /* .cur    = */ {},
@@ -157,7 +162,7 @@ struct common_sampler * common_sampler_init(const struct llama_model * model, co
 
     llama_sampler_chain_add(result->chain,
             llama_sampler_init_logit_bias(
-                llama_n_vocab(model),
+                llama_vocab_n_tokens(vocab),
                 params.logit_bias.size(),
                 params.logit_bias.data()));
 
@@ -172,7 +177,7 @@ struct common_sampler * common_sampler_init(const struct llama_model * model, co
                             c_breakers.push_back(str.c_str());
                         }
 
-                        llama_sampler_chain_add(result->chain, llama_sampler_init_dry      (model, params.dry_multiplier, params.dry_base, params.dry_allowed_length, params.dry_penalty_last_n, c_breakers.data(), c_breakers.size()));
+                        llama_sampler_chain_add(result->chain, llama_sampler_init_dry      (vocab, llama_model_n_ctx_train(model), params.dry_multiplier, params.dry_base, params.dry_allowed_length, params.dry_penalty_last_n, c_breakers.data(), c_breakers.size()));
                     }
                     break;
                 case COMMON_SAMPLER_TYPE_TOP_K:
@@ -194,7 +199,7 @@ struct common_sampler * common_sampler_init(const struct llama_model * model, co
                     llama_sampler_chain_add(result->chain, llama_sampler_init_temp_ext (params.temp, params.dynatemp_range, params.dynatemp_exponent));
                     break;
                 case COMMON_SAMPLER_TYPE_INFILL:
-                    llama_sampler_chain_add(result->chain, llama_sampler_init_infill   (model));
+                    llama_sampler_chain_add(result->chain, llama_sampler_init_infill   (vocab));
                     break;
                 case COMMON_SAMPLER_TYPE_PENALTIES:
                     llama_sampler_chain_add(result->chain, llama_sampler_init_penalties(params.penalty_last_n, params.penalty_repeat, params.penalty_freq, params.penalty_present));
@@ -206,7 +211,7 @@ struct common_sampler * common_sampler_init(const struct llama_model * model, co
         llama_sampler_chain_add(result->chain, llama_sampler_init_dist(params.seed));
     } else if (params.mirostat == 1) {
         llama_sampler_chain_add(result->chain, llama_sampler_init_temp(params.temp));
-        llama_sampler_chain_add(result->chain, llama_sampler_init_mirostat(llama_n_vocab(model), params.seed, params.mirostat_tau, params.mirostat_eta, 100));
+        llama_sampler_chain_add(result->chain, llama_sampler_init_mirostat(llama_vocab_n_tokens(vocab), params.seed, params.mirostat_tau, params.mirostat_eta, 100));
     } else if (params.mirostat == 2) {
         llama_sampler_chain_add(result->chain, llama_sampler_init_temp(params.temp));
         llama_sampler_chain_add(result->chain, llama_sampler_init_mirostat_v2(params.seed, params.mirostat_tau, params.mirostat_eta));

common/speculative.cpp

@@ -79,10 +79,13 @@ bool common_speculative_are_compatible(
     const struct llama_model * model_tgt = llama_get_model(ctx_tgt);
     const struct llama_model * model_dft = llama_get_model(ctx_dft);
 
-    const bool vocab_type_tgt = llama_vocab_type(model_tgt);
+    const struct llama_vocab * vocab_tgt = llama_model_get_vocab(model_tgt);
+    const struct llama_vocab * vocab_dft = llama_model_get_vocab(model_dft);
+
+    const bool vocab_type_tgt = llama_vocab_type(vocab_tgt);
     LOG_DBG("%s: vocab_type tgt: %d\n", __func__, vocab_type_tgt);
 
-    const bool vocab_type_dft = llama_vocab_type(model_dft);
+    const bool vocab_type_dft = llama_vocab_type(vocab_dft);
     LOG_DBG("%s: vocab_type dft: %d\n", __func__, vocab_type_dft);
 
     if (vocab_type_tgt != vocab_type_dft) {
@@ -91,34 +94,34 @@ bool common_speculative_are_compatible(
         return false;
     }
 
-    if (llama_add_bos_token(model_tgt) != llama_add_bos_token(model_dft) ||
-        llama_add_eos_token(model_tgt) != llama_add_eos_token(model_dft) ||
-        llama_token_bos(model_tgt) != llama_token_bos(model_dft) ||
-        llama_token_eos(model_tgt) != llama_token_eos(model_dft)) {
-        LOG_ERR("%s: draft model special tokens must match target model to use speculation\n", __func__);
-        LOG_ERR("%s: tgt: bos = %d (%d), eos = %d (%d)\n", __func__, llama_token_bos(model_tgt), llama_add_bos_token(model_tgt), llama_token_eos(model_tgt), llama_add_eos_token(model_tgt));
-        LOG_ERR("%s: dft: bos = %d (%d), eos = %d (%d)\n", __func__, llama_token_bos(model_dft), llama_add_bos_token(model_dft), llama_token_eos(model_dft), llama_add_eos_token(model_dft));
+    if (llama_vocab_get_add_bos(vocab_tgt) != llama_vocab_get_add_bos(vocab_dft) ||
+        llama_vocab_get_add_eos(vocab_tgt) != llama_vocab_get_add_eos(vocab_dft) ||
+        llama_vocab_bos(vocab_tgt) != llama_vocab_bos(vocab_dft) ||
+        llama_vocab_eos(vocab_tgt) != llama_vocab_eos(vocab_dft)) {
+        LOG_ERR("%s: draft vocab special tokens must match target vocab to use speculation\n", __func__);
+        LOG_ERR("%s: tgt: bos = %d (%d), eos = %d (%d)\n", __func__, llama_vocab_bos(vocab_tgt), llama_vocab_get_add_bos(vocab_tgt), llama_vocab_eos(vocab_tgt), llama_vocab_get_add_eos(vocab_tgt));
+        LOG_ERR("%s: dft: bos = %d (%d), eos = %d (%d)\n", __func__, llama_vocab_bos(vocab_dft), llama_vocab_get_add_bos(vocab_dft), llama_vocab_eos(vocab_dft), llama_vocab_get_add_eos(vocab_dft));
         return false;
     }
 
     {
-        const int n_vocab_tgt = llama_n_vocab(model_tgt);
-        const int n_vocab_dft = llama_n_vocab(model_dft);
+        const int n_vocab_tgt = llama_vocab_n_tokens(vocab_tgt);
+        const int n_vocab_dft = llama_vocab_n_tokens(vocab_dft);
 
         const int vocab_diff = std::abs(n_vocab_tgt - n_vocab_dft);
 
         if (vocab_diff > SPEC_VOCAB_MAX_SIZE_DIFFERENCE) {
             LOG_ERR("%s: draft model vocab must closely match target model to use speculation but "
                          "target vocab size %d does not match draft vocab size %d - difference %d, max allowed %d\n",
-                    __func__, n_vocab_tgt, llama_n_vocab(model_dft), vocab_diff, SPEC_VOCAB_MAX_SIZE_DIFFERENCE);
+                    __func__, n_vocab_tgt, llama_vocab_n_tokens(vocab_dft), vocab_diff, SPEC_VOCAB_MAX_SIZE_DIFFERENCE);
             return false;
         }
 
         for (int i = SPEC_VOCAB_CHECK_START_TOKEN_ID; i < std::min(n_vocab_tgt, n_vocab_dft); ++i) {
-            const char * token_text_tgt = llama_token_get_text(model_tgt, i);
-            const char * token_text_dft = llama_token_get_text(model_dft, i);
+            const char * token_text_tgt = llama_vocab_get_text(vocab_tgt, i);
+            const char * token_text_dft = llama_vocab_get_text(vocab_dft, i);
             if (std::strcmp(token_text_tgt, token_text_dft) != 0) {
-                LOG_ERR("%s: draft model vocab must match target model to use speculation but "
+                LOG_ERR("%s: draft vocab vocab must match target vocab to use speculation but "
                              "token %d content differs - target '%s', draft '%s'\n", __func__, i,
                         common_token_to_piece(ctx_tgt, i).c_str(),
                         common_token_to_piece(ctx_dft, i).c_str());

convert_hf_to_gguf.py

@@ -326,6 +326,7 @@ class Model:
                             gguf.MODEL_TENSOR.TIME_MIX_W2,
                             gguf.MODEL_TENSOR.TIME_MIX_DECAY_W1,
                             gguf.MODEL_TENSOR.TIME_MIX_DECAY_W2,
+                            gguf.MODEL_TENSOR.TIME_MIX_LERP_FUSED,
                             gguf.MODEL_TENSOR.POSNET_NORM1,
                             gguf.MODEL_TENSOR.POSNET_NORM2,
                         )
@@ -477,6 +478,11 @@ class Model:
             return modelcls
         return func
 
+    @classmethod
+    def print_registered_models(cls):
+        for name in sorted(cls._model_classes.keys()):
+            logger.error(f"- {name}")
+
     @classmethod
     def from_model_architecture(cls, arch: str) -> type[Model]:
         try:
@@ -2562,6 +2568,63 @@ class Phi3MiniModel(Model):
         yield (self.format_tensor_name(gguf.MODEL_TENSOR.ROPE_FACTORS_SHORT), torch.tensor(short_factors, dtype=torch.float32))
 
 
+@Model.register("PhiMoEForCausalLM")
+class PhiMoeModel(Phi3MiniModel):
+    model_arch = gguf.MODEL_ARCH.PHIMOE
+
+    _experts: list[dict[str, Tensor]] | None = None
+
+    def set_gguf_parameters(self):
+        super().set_gguf_parameters()
+        self.gguf_writer.add_expert_used_count(self.hparams["num_experts_per_tok"])
+        self.gguf_writer.add_expert_count(self.hparams["num_local_experts"])
+
+    def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iterable[tuple[str, Tensor]]:
+        # process the experts separately
+        if name.find("block_sparse_moe.experts") != -1:
+            n_experts = self.hparams["num_local_experts"]
+            assert bid is not None
+
+            if self._experts is None:
+                self._experts = [{} for _ in range(self.block_count)]
+
+            self._experts[bid][name] = data_torch
+
+            if len(self._experts[bid]) >= n_experts * 3:
+                tensors: list[tuple[str, Tensor]] = []
+
+                # merge the experts into a single 3d tensor
+                for w_name in ["w1", "w2", "w3"]:
+                    datas: list[Tensor] = []
+
+                    for xid in range(n_experts):
+                        ename = f"model.layers.{bid}.block_sparse_moe.experts.{xid}.{w_name}.weight"
+                        datas.append(self._experts[bid][ename])
+                        del self._experts[bid][ename]
+
+                    data_torch = torch.stack(datas, dim=0)
+
+                    merged_name = f"model.layers.{bid}.block_sparse_moe.experts.{w_name}.weight"
+
+                    new_name = self.map_tensor_name(merged_name)
+
+                    tensors.append((new_name, data_torch))
+                return tensors
+            else:
+                return []
+
+        return [(self.map_tensor_name(name), data_torch)]
+
+    def prepare_tensors(self):
+        super().prepare_tensors()
+
+        if self._experts is not None:
+            # flatten `list[dict[str, Tensor]]` into `list[str]`
+            experts = [k for d in self._experts for k in d.keys()]
+            if len(experts) > 0:
+                raise ValueError(f"Unprocessed experts: {experts}")
+
+
 @Model.register("PlamoForCausalLM")
 class PlamoModel(Model):
     model_arch = gguf.MODEL_ARCH.PLAMO
@@ -3259,6 +3322,8 @@ class Rwkv6Model(Model):
         # required by llama.cpp, unused
         self.gguf_writer.add_head_count(0)
 
+    lerp_weights: dict[int, dict[str, Tensor]] = {}
+
     def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iterable[tuple[str, Tensor]]:
         new_name = self.map_tensor_name(name)
 
@@ -3274,14 +3339,84 @@ class Rwkv6Model(Model):
         if new_name.endswith("time_mix_decay.weight") or "lerp" in new_name:
             data_torch = data_torch.squeeze()
 
-        rescale_every_n_layers = self.hparams["rescale_every"]
-        if rescale_every_n_layers > 0:
-            if new_name.endswith("time_mix_output.weight") or new_name.endswith("channel_mix_value.weight"):
-                data_torch = data_torch.div_(2 ** int(bid // rescale_every_n_layers))
+        try:
+            rescale_every_n_layers = self.hparams["rescale_every"]
+            if rescale_every_n_layers > 0:
+                if new_name.endswith("time_mix_output.weight") or new_name.endswith("channel_mix_value.weight"):
+                    data_torch = data_torch.div_(2 ** int(bid // rescale_every_n_layers))
+        except KeyError:
+            pass
+
+        # concat time_mix_lerp weights to reduce some cpu overhead
+        # also reduces the number of tensors in the model
+        if bid is not None and "time_mix_lerp" in new_name and "time_mix_lerp_x" not in new_name:
+            try:
+                self.lerp_weights[bid][new_name] = data_torch
+            except KeyError:
+                self.lerp_weights[bid] = {new_name: data_torch}
+            if all(f"blk.{bid}.time_mix_lerp_{i}.weight" in self.lerp_weights[bid].keys() for i in ["w", "k", "v", "r", "g"]):
+                new_name = f"blk.{bid}.time_mix_lerp_fused.weight"
+                data = torch.stack([self.lerp_weights[bid][f"blk.{bid}.time_mix_lerp_{i}.weight"].unsqueeze(0) for i in ["w", "k", "v", "r", "g"]], dim=0).unsqueeze(1)
+                yield (new_name, data)
+            return
 
         yield (new_name, data_torch)
 
 
+@Model.register("RWKV6Qwen2ForCausalLM")
+class RWKV6Qwen2Model(Rwkv6Model):
+    model_arch = gguf.MODEL_ARCH.RWKV6QWEN2
+
+    def set_vocab(self):
+        try:
+            self._set_vocab_sentencepiece()
+        except FileNotFoundError:
+            self._set_vocab_gpt2()
+
+    def set_gguf_parameters(self):
+        block_count = self.hparams["num_hidden_layers"]
+        num_attention_heads = self.hparams["num_attention_heads"]
+        num_key_value_heads = self.hparams["num_key_value_heads"]
+        hidden_size = self.hparams["hidden_size"]
+        head_size = hidden_size // num_attention_heads
+        rms_norm_eps = self.hparams["rms_norm_eps"]
+        intermediate_size = self.hparams["intermediate_size"]
+        time_mix_extra_dim = 64 if hidden_size >= 4096 else 32
+        time_decay_extra_dim = 128 if hidden_size >= 4096 else 64
+
+        # RWKV isn't context limited
+        self.gguf_writer.add_context_length(1048576)
+        self.gguf_writer.add_embedding_length(hidden_size)
+        self.gguf_writer.add_block_count(block_count)
+        self.gguf_writer.add_wkv_head_size(head_size)
+        self.gguf_writer.add_time_mix_extra_dim(time_mix_extra_dim)
+        self.gguf_writer.add_time_decay_extra_dim(time_decay_extra_dim)
+        self.gguf_writer.add_feed_forward_length(intermediate_size)
+        self.gguf_writer.add_file_type(self.ftype)
+
+        # special parameters for time_mixing in RWKV6QWEN2
+        self.gguf_writer.add_layer_norm_rms_eps(rms_norm_eps)
+        self.gguf_writer.add_token_shift_count(1)
+        # RWKV6QWEN2 use grouped key/value like GQA
+        self.gguf_writer.add_head_count_kv(num_key_value_heads)
+
+        # required by llama.cpp, unused
+        self.gguf_writer.add_head_count(0)
+
+    def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iterable[tuple[str, Tensor]]:
+        for new_name, data in super().modify_tensors(data_torch, name, bid):
+            if "time_mix_w1" in new_name or "time_mix_w2" in new_name:
+                data = data.view(5, -1, data.shape[-1])
+                # rwkv6qwen2 has a different order of rkvwg instead of the original wkvrg
+                # permute them here to avoid code changes
+                data = torch.stack([data[3], data[1], data[2], data[0], data[4]], dim=0).view(-1, data.shape[-1])
+                if "w2" in new_name:
+                    data = data.view(5, -1, data.shape[-1])
+                yield (new_name, data)
+                continue
+            yield (new_name, data)
+
+
 @Model.register("MambaForCausalLM", "MambaLMHeadModel", "FalconMambaForCausalLM")
 class MambaModel(Model):
     model_arch = gguf.MODEL_ARCH.MAMBA
@@ -4799,6 +4934,7 @@ def parse_args() -> argparse.Namespace:
     parser.add_argument(
         "model", type=Path,
         help="directory containing model file",
+        nargs="?",
     )
     parser.add_argument(
         "--use-temp-file", action="store_true",
@@ -4836,8 +4972,15 @@ def parse_args() -> argparse.Namespace:
         "--metadata", type=Path,
         help="Specify the path for an authorship metadata override file"
     )
+    parser.add_argument(
+        "--print-supported-models", action="store_true",
+        help="Print the supported models"
+    )
 
-    return parser.parse_args()
+    args = parser.parse_args()
+    if not args.print_supported_models and args.model is None:
+        parser.error("the following arguments are required: model")
+    return args
 
 
 def split_str_to_n_bytes(split_str: str) -> int:
@@ -4861,6 +5004,11 @@ def split_str_to_n_bytes(split_str: str) -> int:
 def main() -> None:
     args = parse_args()
 
+    if args.print_supported_models:
+        logger.error("Supported models:")
+        Model.print_registered_models()
+        sys.exit(0)
+
     if args.verbose:
         logging.basicConfig(level=logging.DEBUG)
     else:

convert_lora_to_gguf.py

@@ -226,6 +226,9 @@ def get_base_tensor_name(lora_tensor_name: str) -> str:
     base_name = lora_tensor_name.replace("base_model.model.", "")
     base_name = base_name.replace(".lora_A.weight", ".weight")
     base_name = base_name.replace(".lora_B.weight", ".weight")
+    # models produced by mergekit-extract-lora have token embeddings in the adapter
+    base_name = base_name.replace(".lora_embedding_A", ".weight")
+    base_name = base_name.replace(".lora_embedding_B", ".weight")
     return base_name
 
 
@@ -260,6 +263,10 @@ def parse_args() -> argparse.Namespace:
         "--base", type=Path,
         help="directory containing Hugging Face model config files (config.json, tokenizer.json) for the base model that the adapter is based on - only config is needed, actual model weights are not required. If base model is unspecified, it will be loaded from Hugging Face hub based on the adapter config",
     )
+    parser.add_argument(
+        "--base-model-id", type=str,
+        help="the model ID of the base model, if it is not available locally or in the adapter config. If specified, it will ignore --base and load the base model config from the Hugging Face hub (Example: 'meta-llama/Llama-3.2-1B-Instruct')",
+    )
     parser.add_argument(
         "lora_path", type=Path,
         help="directory containing Hugging Face PEFT LoRA config (adapter_model.json) and weights (adapter_model.safetensors or adapter_model.bin)",
@@ -290,6 +297,7 @@ if __name__ == '__main__':
 
     dir_base_model: Path | None = args.base
     dir_lora: Path = args.lora_path
+    base_model_id: str | None = args.base_model_id
     lora_config = dir_lora / "adapter_config.json"
     input_model = dir_lora / "adapter_model.safetensors"
 
@@ -313,7 +321,10 @@ if __name__ == '__main__':
         lparams: dict[str, Any] = json.load(f)
 
     # load base model
-    if dir_base_model is None:
+    if base_model_id is not None:
+        logger.info(f"Loading base model from Hugging Face: {base_model_id}")
+        hparams = load_hparams_from_hf(base_model_id)
+    elif dir_base_model is None:
         if "base_model_name_or_path" in lparams:
             model_id = lparams["base_model_name_or_path"]
             logger.info(f"Loading base model from Hugging Face: {model_id}")
@@ -371,11 +382,16 @@ if __name__ == '__main__':
                     if self.lazy:
                         tensor = LazyTorchTensor.from_eager(tensor)
                     base_name = get_base_tensor_name(name)
-                    is_lora_a = ".lora_A.weight" in name
-                    is_lora_b = ".lora_B.weight" in name
+                    # note: mergekit-extract-lora also adds token embeddings to the adapter
+                    is_lora_a = ".lora_A.weight" in name or ".lora_embedding_A" in name
+                    is_lora_b = ".lora_B.weight" in name or ".lora_embedding_B" in name
                     if not is_lora_a and not is_lora_b:
                         if ".base_layer.weight" in name:
                             continue
+                        # mergekit-extract-lora add these layernorm to the adapter, we need to keep them
+                        if "_layernorm" in name or ".norm" in name:
+                            yield (base_name, tensor)
+                            continue
                         logger.error(f"Unexpected name '{name}': Not a lora_A or lora_B tensor")
                         if ".embed_tokens.weight" in name or ".lm_head.weight" in name:
                             logger.error("Embeddings is present in the adapter. This can be due to new tokens added during fine tuning")
@@ -407,9 +423,21 @@ if __name__ == '__main__':
                 if name == "lm_head.weight" and len(dest) == 0:
                     raise ValueError("lm_head is present in adapter, but is ignored in base model")
                 for dest_name, dest_data in dest:
+                    # mergekit-extract-lora add these layernorm to the adapter
+                    if "_norm" in dest_name:
+                        assert dest_data.dim() == 1
+                        yield (dest_name, dest_data)
+                        continue
+
+                    # otherwise, we must get the lora_A and lora_B tensors
                     assert isinstance(dest_data, LoraTorchTensor)
                     lora_a, lora_b = dest_data.get_lora_A_B()
 
+                    # note: mergekit-extract-lora flip and transpose A and B
+                    # here we only need to transpose token_embd.lora_a, see llm_build_inp_embd()
+                    if "token_embd.weight" in dest_name:
+                        lora_a = lora_a.T
+
                     yield (dest_name + ".lora_a", lora_a)
                     yield (dest_name + ".lora_b", lora_b)
 

docs/build.md

@@ -127,6 +127,8 @@ For detailed info, please refer to [llama.cpp for SYCL](./backend/SYCL.md).
 
 This provides GPU acceleration using an NVIDIA GPU. Make sure to have the CUDA toolkit installed. You can download it from your Linux distro's package manager (e.g. `apt install nvidia-cuda-toolkit`) or from the [NVIDIA developer site](https://developer.nvidia.com/cuda-downloads).
 
+If you are using Fedora (using Fedora Workstation, or an 'Atomic' variant such as Silverblue), or would like to set up CUDA in a toolbox, please consider our [Fedora CUDA guide](./cuda-fedora.md). Unfortunately, the process is not as simple as one might expect.
+
 - Using `CMake`:
 
   ```bash

docs/cuda-fedora.md

@@ -0,0 +1,317 @@
+# Setting Up CUDA on Fedora
+
+In this guide we setup [Nvidia CUDA](https://docs.nvidia.com/cuda/) in a toolbox container. This guide is applicable for:
+- [Fedora Workstation](https://fedoraproject.org/workstation/)
+- [Atomic Desktops for Fedora](https://fedoraproject.org/atomic-desktops/)
+- [Fedora Spins](https://fedoraproject.org/spins)
+- [Other Distributions](https://containertoolbx.org/distros/), including `Red Hat Enterprise Linux >= 8.`, `Arch Linux`, and `Ubuntu`.
+
+
+## Table of Contents
+
+- [Prerequisites](#prerequisites)
+- [Monitoring NVIDIA CUDA Repositories](#monitoring-nvidia-cuda-repositories)
+- [Using the Fedora 39 CUDA Repository](#using-the-fedora-39-cuda-repository)
+- [Creating a Fedora Toolbox Environment](#creating-a-fedora-toolbox-environment)
+- [Installing Essential Development Tools](#installing-essential-development-tools)
+- [Adding the CUDA Repository](#adding-the-cuda-repository)
+- [Installing `nvidia-driver-libs`](#installing-nvidia-driver-libs)
+- [Manually Resolving Package Conflicts](#manually-resolving-package-conflicts)
+- [Finalizing the Installation of `nvidia-driver-libs`](#finalizing-the-installation-of-nvidia-driver-libs)
+- [Installing the CUDA Meta-Package](#installing-the-cuda-meta-package)
+- [Configuring the Environment](#configuring-the-environment)
+- [Verifying the Installation](#verifying-the-installation)
+- [Conclusion](#conclusion)
+- [Troubleshooting](#troubleshooting)
+- [Additional Notes](#additional-notes)
+- [References](#references)
+
+## Prerequisites
+
+- **Toolbox Installed on the Host System** `Fedora Silverblue` and `Fedora Workstation` both have toolbox by default, other distributions may need to install the [toolbox package](https://containertoolbx.org/install/).
+- **NVIDIA Drivers and Graphics Card installed on Host System (optional)** To run CUDA program, such as `llama.cpp`, the host should be setup to access your NVIDIA hardware. Fedora Hosts can use the [RPM Fusion Repository](https://rpmfusion.org/Howto/NVIDIA).
+- **Internet connectivity** to download packages.
+
+### Monitoring NVIDIA CUDA Repositories
+
+Before proceeding, it is advisable to check if NVIDIA has updated their CUDA repositories for your Fedora version. NVIDIA's repositories can be found at:
+
+- [Fedora 40 CUDA Repository](https://developer.download.nvidia.com/compute/cuda/repos/fedora40/x86_64/)
+- [Fedora 41 CUDA Repository](https://developer.download.nvidia.com/compute/cuda/repos/fedora41/x86_64/)
+
+As of the latest update, these repositories do not contain the `cuda` meta-package or are missing essential components.
+
+### Using the Fedora 39 CUDA Repository
+
+Since the newer repositories are incomplete, we'll use the Fedora 39 repository:
+
+- [Fedora 39 CUDA Repository](https://developer.download.nvidia.com/compute/cuda/repos/fedora39/x86_64/)
+
+**Note:** Fedora 39 is no longer maintained, so we recommend using a toolbox environment to prevent system conflicts.
+
+## Creating a Fedora Toolbox Environment
+
+This guide focuses on Fedora hosts, but with small adjustments, it can work for other hosts. Using a Fedora 39 toolbox allows us to install the necessary packages without affecting the host system.
+
+**Note:** Toolbox is available for other systems, and even without Toolbox, it is possible to use Podman or Docker.
+
+We do not recommend installing on the host system, as Fedora 39 is out-of-maintenance, and instead you should upgrade to a maintained version of Fedora for your host.
+
+1. **Create a Fedora 39 Toolbox:**
+
+   ```bash
+   toolbox create --image registry.fedoraproject.org/fedora-toolbox:39 --container fedora-toolbox-39-cuda
+   ```
+
+2. **Enter the Toolbox:**
+
+   ```bash
+   toolbox enter --container fedora-toolbox-39-cuda
+   ```
+
+   Inside the toolbox, you have root privileges and can install packages without affecting the host system.
+
+## Installing Essential Development Tools
+
+1. **Synchronize the DNF Package Manager:**
+
+   ```bash
+   sudo dnf distro-sync
+   ```
+
+2. **Install the Default Text Editor (Optional):**
+
+   ```bash
+   sudo dnf install vim-default-editor --allowerasing
+   ```
+
+   The `--allowerasing` flag resolves any package conflicts.
+
+3. **Install Development Tools and Libraries:**
+
+   ```bash
+   sudo dnf install @c-development @development-tools cmake
+   ```
+
+   This installs essential packages for compiling software, including `gcc`, `make`, and other development headers.
+
+## Adding the CUDA Repository
+
+Add the NVIDIA CUDA repository to your DNF configuration:
+
+```bash
+sudo dnf config-manager --add-repo https://developer.download.nvidia.com/compute/cuda/repos/fedora39/x86_64/cuda-fedora39.repo
+```
+
+After adding the repository, synchronize the package manager again:
+
+```bash
+sudo dnf distro-sync
+```
+
+## Installing `nvidia-driver-libs`
+
+Attempt to install `nvidia-driver-libs`:
+
+```bash
+sudo dnf install nvidia-driver-libs
+```
+
+**Explanation:**
+
+- `nvidia-driver-libs` contains necessary NVIDIA driver libraries required by CUDA.
+- This step might fail due to conflicts with existing NVIDIA drivers on the host system.
+
+## Manually Resolving Package Conflicts
+
+If the installation fails due to conflicts, we'll manually download and install the required packages, excluding conflicting files.
+
+### 1. Download the `nvidia-driver-libs` RPM
+
+```bash
+sudo dnf download --arch x86_64 nvidia-driver-libs
+```
+
+You should see a file similar to:
+
+```
+nvidia-driver-libs-560.35.05-1.fc39.x86_64.rpm
+```
+
+### 2. Attempt to Install the RPM
+
+```bash
+sudo dnf install nvidia-driver-libs-560.35.05-1.fc39.x86_64.rpm
+```
+
+**Expected Error:**
+
+Installation may fail with errors pointing to conflicts with `egl-gbm` and `egl-wayland`.
+
+**Note: It is important to carefully read the error messages to identify the exact paths that need to be excluded.**
+
+### 3. Download Dependencies
+
+```bash
+sudo dnf download --arch x86_64 egl-gbm egl-wayland
+```
+
+### 4. Install `egl-gbm` with Excluded Paths
+
+Exclude conflicting files during installation:
+
+```bash
+sudo rpm --install --verbose --hash \
+  --excludepath=/usr/lib64/libnvidia-egl-gbm.so.1.1.2 \
+  --excludepath=/usr/share/egl/egl_external_platform.d/15_nvidia_gbm.json \
+  egl-gbm-1.1.2^20240919gitb24587d-3.fc39.x86_64.rpm
+```
+
+**Explanation:**
+
+- The `--excludepath` option skips installing files that conflict with existing files.
+- Adjust the paths based on the error messages you receive.
+
+### 5. Install `egl-wayland` with Excluded Paths
+
+```bash
+sudo rpm --install --verbose --hash \
+  --excludepath=/usr/share/egl/egl_external_platform.d/10_nvidia_wayland.json \
+  egl-wayland-1.1.17^20241118giteeb29e1-5.fc39.x86_64.rpm
+```
+
+### 6. Install `nvidia-driver-libs` with Excluded Paths
+
+```bash
+sudo rpm --install --verbose --hash \
+  --excludepath=/usr/share/glvnd/egl_vendor.d/10_nvidia.json \
+  --excludepath=/usr/share/nvidia/nvoptix.bin \
+  nvidia-driver-libs-560.35.05-1.fc39.x86_64.rpm
+```
+
+**Note:**
+
+- Replace the paths with the ones causing conflicts in your installation if they differ.
+- The `--verbose` and `--hash` options provide detailed output during installation.
+
+## Finalizing the Installation of `nvidia-driver-libs`
+
+After manually installing the dependencies, run:
+
+```bash
+sudo dnf install nvidia-driver-libs
+```
+
+You should receive a message indicating the package is already installed:
+
+```
+Package nvidia-driver-libs-3:560.35.05-1.fc39.x86_64 is already installed.
+Dependencies resolved.
+Nothing to do.
+Complete!
+```
+
+## Installing the CUDA Meta-Package
+
+Now that the driver libraries are installed, proceed to install CUDA:
+
+```bash
+sudo dnf install cuda
+```
+
+This installs the CUDA toolkit and associated packages.
+
+## Configuring the Environment
+
+To use CUDA, add its binary directory to your system's `PATH`.
+
+1. **Create a Profile Script:**
+
+   ```bash
+   sudo sh -c 'echo "export PATH=\$PATH:/usr/local/cuda/bin" >> /etc/profile.d/cuda.sh'
+   ```
+
+   **Explanation:**
+
+   - We add to  `/etc/profile.d/` as the `/etc/` folder is unique to this particular container, and is not shared with other containers or the host system.
+   - The backslash `\` before `$PATH` ensures the variable is correctly written into the script.
+
+2. **Make the Script Executable:**
+
+   ```bash
+   sudo chmod +x /etc/profile.d/cuda.sh
+   ```
+
+3. **Source the Script to Update Your Environment:**
+
+   ```bash
+   source /etc/profile.d/cuda.sh
+   ```
+
+   **Note:** This command updates your current shell session with the new `PATH`. The `/etc/profile.d/cuda.sh` script ensures that the CUDA binaries are available in your `PATH` for all future sessions.
+
+## Verifying the Installation
+
+To confirm that CUDA is correctly installed and configured, check the version of the NVIDIA CUDA Compiler (`nvcc`):
+
+```bash
+nvcc --version
+```
+
+You should see output similar to:
+
+```
+nvcc: NVIDIA (R) Cuda compiler driver
+Copyright (c) 2005-2024 NVIDIA Corporation
+Built on Tue_Oct_29_23:50:19_PDT_2024
+Cuda compilation tools, release 12.6, V12.6.85
+Build cuda_12.6.r12.6/compiler.35059454_0
+```
+
+This output confirms that the CUDA compiler is accessible and indicates the installed version.
+
+## Conclusion
+
+You have successfully set up CUDA on Fedora within a toolbox environment using the Fedora 39 CUDA repository. By manually resolving package conflicts and configuring the environment, you can develop CUDA applications without affecting your host system.
+
+## Troubleshooting
+
+- **Installation Failures:**
+  - If you encounter errors during installation, carefully read the error messages. They often indicate conflicting files or missing dependencies.
+  - Use the `--excludepath` option with `rpm` to exclude conflicting files during manual installations.
+
+- **Driver Conflicts:**
+  - Since the host system may already have NVIDIA drivers installed, conflicts can arise. Using the toolbox environment helps isolate these issues.
+
+- **Environment Variables Not Set:**
+  - If `nvcc` is not found after installation, ensure that `/usr/local/cuda/bin` is in your `PATH`.
+  - Run `echo $PATH` to check if the path is included.
+  - Re-source the profile script or open a new terminal session.
+
+## Additional Notes
+
+- **Updating CUDA in the Future:**
+  - Keep an eye on the official NVIDIA repositories for updates to your Fedora version.
+  - When an updated repository becomes available, adjust your `dnf` configuration accordingly.
+
+- **Building `llama.cpp`:**
+  - With CUDA installed, you can follow these [build instructions for `llama.cpp`](https://github.com/ggerganov/llama.cpp/blob/master/docs/build.md) to compile it with CUDA support.
+  - Ensure that any CUDA-specific build flags or paths are correctly set in your build configuration.
+
+- **Using the Toolbox Environment:**
+  - The toolbox environment is isolated from your host system, which helps prevent conflicts.
+  - Remember that system files and configurations inside the toolbox are separate from the host. By default the home directory of the user is shared between the host and the toolbox.
+
+---
+
+**Disclaimer:** Manually installing and modifying system packages can lead to instability of the container. The above steps are provided as a guideline and may need adjustments based on your specific system configuration. Always back up important data before making significant system changes, especially as your home folder is writable and shared with he toolbox.
+
+**Acknowledgments:** Special thanks to the Fedora community and NVIDIA documentation for providing resources that assisted in creating this guide.
+
+## References
+
+- [Fedora Toolbox Documentation](https://docs.fedoraproject.org/en-US/fedora-silverblue/toolbox/)
+- [NVIDIA CUDA Installation Guide](https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html)
+- [Podman Documentation](https://podman.io/get-started)
+
+---

docs/development/HOWTO-add-model.md

@@ -28,7 +28,7 @@ The required steps to implement for an HF model are:
 ```python
 @Model.register("MyModelForCausalLM")
 class MyModel(Model):
-    model_arch = gguf.MODEL_ARCH.GROK
+    model_arch = gguf.MODEL_ARCH.MYMODEL
 ```
 
 2. Define the layout of the GGUF tensors in [constants.py](/gguf-py/gguf/constants.py)
@@ -79,14 +79,14 @@ Depending on the model configuration, tokenizer, code and tensors layout, you wi
 - `Model#set_vocab`
 - `Model#write_tensors`
 
-NOTE: Tensor names must end with `.weight` suffix, that is the convention and several tools like `quantize` expect this to proceed the weights.
+NOTE: Tensor names must end with `.weight` or `.bias` suffixes, that is the convention and several tools like `quantize` expect this to proceed the weights.
 
 ### 2. Define the model architecture in `llama.cpp`
 
 The model params and tensors layout must be defined in `llama.cpp`:
 1. Define a new `llm_arch`
 2. Define the tensors layout in `LLM_TENSOR_NAMES`
-3. Add any non standard metadata in `llm_load_hparams`
+3. Add any non-standard metadata in `llm_load_hparams`
 4. Create the tensors for inference in `llm_load_tensors`
 5. If the model has a RoPE operation, add the rope type in `llama_rope_type`
 
@@ -96,9 +96,9 @@ NOTE: The dimensions in `ggml` are typically in the reverse order of the `pytorc
 
 This is the funniest part, you have to provide the inference graph implementation of the new model architecture in `llama_build_graph`.
 
-Have a look at existing implementation like `build_llama`, `build_dbrx` or `build_bert`.
+Have a look at existing implementations like `build_llama`, `build_dbrx` or `build_bert`.
 
-When implementing a new graph, please note that the underlying `ggml` backends might not support them all, support for missing backend operations can be added in another PR.
+Some `ggml` backends do not support all operations. Backend implementations can be added in a separate PR.
 
 Note: to debug the inference graph: you can use [llama-eval-callback](/examples/eval-callback/).
 

examples/batched-bench/batched-bench.cpp

@@ -50,7 +50,7 @@ int main(int argc, char ** argv) {
     // ensure enough sequences are available
     ctx_params.n_seq_max = n_pl.empty() ? 1 : *std::max_element(n_pl.begin(), n_pl.end());
 
-    llama_context * ctx = llama_new_context_with_model(model, ctx_params);
+    llama_context * ctx = llama_init_from_model(model, ctx_params);
 
     if (ctx == NULL) {
         fprintf(stderr , "%s: error: failed to create the llama_context\n" , __func__);

examples/batched.swift/Sources/main.swift

@@ -23,12 +23,12 @@ defer {
 }
 
 let model_params = llama_model_default_params()
-guard let model = llama_load_model_from_file(modelPath.cString(using: .utf8), model_params) else {
+guard let model = llama_model_load_from_file(modelPath.cString(using: .utf8), model_params) else {
     print("Failed to load model")
     exit(1)
 }
 defer {
-    llama_free_model(model)
+    llama_model_free(model)
 }
 
 var tokens = tokenize(text: prompt, add_bos: true)
@@ -141,7 +141,7 @@ while n_cur <= n_len {
         let new_token_id = llama_sampler_sample(smpl, context, i_batch[i])
 
         // is it an end of stream? -> mark the stream as finished
-        if llama_token_is_eog(model, new_token_id) || n_cur == n_len {
+        if llama_vocab_is_eog(model, new_token_id) || n_cur == n_len {
             i_batch[i] = -1
             // print("")
             if n_parallel > 1 {

examples/batched/batched.cpp

@@ -48,10 +48,12 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     // tokenize the prompt
 
     std::vector<llama_token> tokens_list;
-    tokens_list = common_tokenize(model, params.prompt, true);
+    tokens_list = common_tokenize(vocab, params.prompt, true);
 
     const int n_kv_req = tokens_list.size() + (n_predict - tokens_list.size())*n_parallel;
 
@@ -62,7 +64,7 @@ int main(int argc, char ** argv) {
     ctx_params.n_ctx   = n_kv_req;
     ctx_params.n_batch = std::max(n_predict, n_parallel);
 
-    llama_context * ctx = llama_new_context_with_model(model, ctx_params);
+    llama_context * ctx = llama_init_from_model(model, ctx_params);
 
     auto sparams = llama_sampler_chain_default_params();
     sparams.no_perf = false;
@@ -121,7 +123,7 @@ int main(int argc, char ** argv) {
 
         llama_token decoder_start_token_id = llama_model_decoder_start_token(model);
         if (decoder_start_token_id == LLAMA_TOKEN_NULL) {
-            decoder_start_token_id = llama_token_bos(model);
+            decoder_start_token_id = llama_vocab_bos(vocab);
         }
 
         common_batch_clear(batch);
@@ -174,7 +176,7 @@ int main(int argc, char ** argv) {
             const llama_token new_token_id = llama_sampler_sample(smpl, ctx, i_batch[i]);
 
             // is it an end of generation? -> mark the stream as finished
-            if (llama_token_is_eog(model, new_token_id) || n_cur == n_predict) {
+            if (llama_vocab_is_eog(vocab, new_token_id) || n_cur == n_predict) {
                 i_batch[i] = -1;
                 LOG("\n");
                 if (n_parallel > 1) {

examples/convert-llama2c-to-ggml/convert-llama2c-to-ggml.cpp

@@ -1,4 +1,6 @@
 #include "ggml.h"
+#include "gguf.h"
+
 #include "llama.h"
 #include "common.h"
 #include "log.h"
@@ -909,7 +911,7 @@ int main(int argc, char ** argv) {
     load_vocab(params.fn_vocab_model, &config, &vocab);
 
     struct my_llama_model model;
-    model.hparams.n_vocab   = config.vocab_size; //llama_n_vocab(lctx);
+    model.hparams.n_vocab   = config.vocab_size; //llama_vocab_n_vocab(lctx);
     model.hparams.n_ctx     = params.n_ctx;
     model.hparams.n_embd    = config.dim; //params.n_embd;
     model.hparams.n_ff      = config.hidden_dim;

examples/cvector-generator/cvector-generator.cpp

@@ -1,7 +1,9 @@
+#include "ggml.h"
+#include "gguf.h"
+
 #include "arg.h"
 #include "common.h"
 #include "llama.h"
-#include "ggml.h"
 #include "pca.hpp"
 #include "mean.hpp"
 
@@ -271,7 +273,9 @@ struct tokenized_prompt {
     size_t max_seq_len;
 
     tokenized_prompt(llama_context * ctx, std::string pos, std::string neg) {
-        const bool add_bos = llama_add_bos_token(llama_get_model(ctx));
+        const llama_model * model = llama_get_model(ctx);
+        const llama_vocab * vocab = llama_model_get_vocab(model);
+        const bool add_bos = llama_vocab_get_add_bos(vocab);
         tokens_pos = common_tokenize(ctx, pos, add_bos, true);
         tokens_neg = common_tokenize(ctx, neg, add_bos, true);
         max_seq_len = std::max(tokens_pos.size(), tokens_neg.size());
@@ -419,8 +423,8 @@ int main(int argc, char ** argv) {
     llama_context * ctx = llama_init.context.get();
 
     // int n_ctx = llama_n_ctx(ctx);
-    int n_layers = llama_n_layer(model);
-    int n_embd = llama_n_embd(model);
+    int n_layers = llama_model_n_layer(model);
+    int n_embd = llama_model_n_embd(model);
 
     // get model hint param (a.k.a model arch name)
     char model_hint[128];

examples/embedding/embedding.cpp

@@ -105,7 +105,9 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
-    const int n_ctx_train = llama_n_ctx_train(model);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
+    const int n_ctx_train = llama_model_n_ctx_train(model);
     const int n_ctx = llama_n_ctx(ctx);
 
     const enum llama_pooling_type pooling_type = llama_pooling_type(ctx);
@@ -148,7 +150,7 @@ int main(int argc, char ** argv) {
     // check if the last token is SEP
     // it should be automatically added by the tokenizer when 'tokenizer.ggml.add_eos_token' is set to 'true'
     for (auto & inp : inputs) {
-        if (inp.empty() || inp.back() != llama_token_sep(model)) {
+        if (inp.empty() || inp.back() != llama_vocab_sep(vocab)) {
             LOG_WRN("%s: last token in the prompt is not SEP\n", __func__);
             LOG_WRN("%s: 'tokenizer.ggml.add_eos_token' should be set to 'true' in the GGUF header\n", __func__);
         }
@@ -181,7 +183,7 @@ int main(int argc, char ** argv) {
     }
 
     // allocate output
-    const int n_embd = llama_n_embd(model);
+    const int n_embd = llama_model_n_embd(model);
     std::vector<float> embeddings(n_embd_count * n_embd, 0);
     float * emb = embeddings.data();
 

examples/eval-callback/eval-callback.cpp

@@ -127,7 +127,10 @@ static bool ggml_debug(struct ggml_tensor * t, bool ask, void * user_data) {
 }
 
 static bool run(llama_context * ctx, const common_params & params) {
-    const bool add_bos = llama_add_bos_token(llama_get_model(ctx));
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
+    const bool add_bos = llama_vocab_get_add_bos(vocab);
 
     std::vector<llama_token> tokens = common_tokenize(ctx, params.prompt, add_bos);
 

examples/export-lora/export-lora.cpp

@@ -1,12 +1,13 @@
-#include "arg.h"
-#include "common.h"
 #include "ggml.h"
 #include "ggml-alloc.h"
+#include "gguf.h"
+
+#include "arg.h"
+#include "common.h"
 
 #include <map>
 #include <vector>
 #include <string>
-#include <thread>
 #include <fstream>
 
 static bool g_verbose = false;
@@ -128,7 +129,7 @@ struct lora_merge_ctx {
 
     lora_merge_ctx(
             std::string & base_fname,
-            std::vector<common_lora_adapter_info> & lora_files,
+            std::vector<common_adapter_lora_info> & lora_files,
             std::string & outfile,
             int n_threads) : base_model(base_fname, 0), n_threads(n_threads), fout(outfile, std::ios::binary) {
         fout.exceptions(std::ofstream::failbit); // fail fast on write errors

examples/gguf-hash/gguf-hash.cpp

@@ -1,4 +1,5 @@
 #include "ggml.h"
+#include "gguf.h"
 
 #include <cstdlib>   /* abort() */
 #include <cstddef>

examples/gguf-split/gguf-split.cpp

@@ -1,16 +1,18 @@
+#include "ggml.h"
+#include "gguf.h"
 #include "llama.h"
 #include "common.h"
 
 #include <algorithm>
+#include <cinttypes>
+#include <climits>
+#include <cstdio>
 #include <cstdlib>
+#include <stdexcept>
+#include <cstring>
 #include <fstream>
 #include <string>
 #include <vector>
-#include <climits>
-
-#include <cstdio>
-#include <cstring>
-#include <stdexcept>
 
 #if defined(_WIN32)
     #include <windows.h>
@@ -296,7 +298,7 @@ struct split_strategy {
                 total_size += ggml_nbytes(t);
             }
             total_size = total_size / 1000 / 1000; // convert to megabytes
-            printf("split %05d: n_tensors = %d, total_size = %zuM\n", i_split + 1, gguf_get_n_tensors(ctx_out), total_size);
+            printf("split %05d: n_tensors = %" PRIi64 ", total_size = %zuM\n", i_split + 1, gguf_get_n_tensors(ctx_out), total_size);
             i_split++;
         }
     }

examples/gguf/gguf.cpp

@@ -1,10 +1,9 @@
 #include "ggml.h"
+#include "gguf.h"
 
 #include <cstdio>
-#include <cinttypes>
 #include <string>
 #include <sstream>
-#include <fstream>
 #include <vector>
 
 #undef MIN
@@ -135,9 +134,10 @@ static bool gguf_ex_read_0(const std::string & fname) {
 
         for (int i = 0; i < n_tensors; ++i) {
             const char * name   = gguf_get_tensor_name  (ctx, i);
+            const size_t size   = gguf_get_tensor_size  (ctx, i);
             const size_t offset = gguf_get_tensor_offset(ctx, i);
 
-            printf("%s: tensor[%d]: name = %s, offset = %zu\n", __func__, i, name, offset);
+            printf("%s: tensor[%d]: name = %s, size = %zu, offset = %zu\n", __func__, i, name, size, offset);
         }
     }
 
@@ -182,9 +182,10 @@ static bool gguf_ex_read_1(const std::string & fname, bool check_data) {
 
         for (int i = 0; i < n_tensors; ++i) {
             const char * name   = gguf_get_tensor_name  (ctx, i);
+            const size_t size   = gguf_get_tensor_size  (ctx, i);
             const size_t offset = gguf_get_tensor_offset(ctx, i);
 
-            printf("%s: tensor[%d]: name = %s, offset = %zu\n", __func__, i, name, offset);
+            printf("%s: tensor[%d]: name = %s, size = %zu, offset = %zu\n", __func__, i, name, size, offset);
         }
     }
 
@@ -199,7 +200,8 @@ static bool gguf_ex_read_1(const std::string & fname, bool check_data) {
 
             struct ggml_tensor * cur = ggml_get_tensor(ctx_data, name);
 
-            printf("%s: tensor[%d]: n_dims = %d, name = %s, data = %p\n", __func__, i, ggml_n_dims(cur), cur->name, cur->data);
+            printf("%s: tensor[%d]: n_dims = %d, ne = (%d, %d, %d, %d), name = %s, data = %p\n",
+                __func__, i, ggml_n_dims(cur), int(cur->ne[0]), int(cur->ne[1]), int(cur->ne[2]), int(cur->ne[3]), cur->name, cur->data);
 
             // print first 10 elements
             const float * data = (const float *) cur->data;
@@ -215,7 +217,7 @@ static bool gguf_ex_read_1(const std::string & fname, bool check_data) {
                 const float * data = (const float *) cur->data;
                 for (int j = 0; j < ggml_nelements(cur); ++j) {
                     if (data[j] != 100 + i) {
-                        fprintf(stderr, "%s: tensor[%d]: data[%d] = %f\n", __func__, i, j, data[j]);
+                        fprintf(stderr, "%s: tensor[%d], data[%d]: found %f, expected %f\n", __func__, i, j, data[j], float(100 + i));
                         gguf_free(ctx);
                         return false;
                     }
@@ -245,6 +247,8 @@ int main(int argc, char ** argv) {
         check_data = false;
     }
 
+    srand(123456);
+
     const std::string fname(argv[1]);
     const std::string mode (argv[2]);
 

examples/gritlm/gritlm.cpp

@@ -11,6 +11,7 @@ static std::vector<std::vector<float>> encode(llama_context * ctx, const std::ve
     std::vector<std::vector<float>> result;
 
     const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
 
     llama_batch batch = llama_batch_init(llama_n_batch(ctx), 0, 1);
 
@@ -19,16 +20,16 @@ static std::vector<std::vector<float>> encode(llama_context * ctx, const std::ve
 
         const std::string input_string = instruction + sentences[i];
 
-        std::vector<llama_token> inputs = common_tokenize(model, input_string, true, false);
+        std::vector<llama_token> inputs = common_tokenize(vocab, input_string, true, false);
 
         const int32_t n_toks = inputs.size();
 
         // GritLM seems to have EOS = ""
         // https://github.com/ContextualAI/gritlm/blob/92025b16534712b31b3c4aaaf069350e222bd5f8/gritlm/gritlm.py#L18
-        // inputs.push_back(llama_token_eos(model));
+        // inputs.push_back(llama_vocab_eos(vocab));
 
         // we want to ignore instruction tokens for mean pooling
-        const int32_t n_inst = common_tokenize(model, instruction, true, false).size();
+        const int32_t n_inst = common_tokenize(vocab, instruction, true, false).size();
 
 #ifdef GRIT_DEBUG
         // debug tokens - should be matching as referenced in the GritLM sample
@@ -52,7 +53,7 @@ static std::vector<std::vector<float>> encode(llama_context * ctx, const std::ve
         llama_decode(ctx, batch);
 
         // get embedding dimensions
-        uint64_t n_embd = llama_n_embd(model);
+        uint64_t n_embd = llama_model_n_embd(model);
 
         // allocate embedding output
         std::vector<float> emb_unorm(n_embd, 0.0f);
@@ -97,7 +98,9 @@ static std::string generate(llama_context * ctx, llama_sampler * smpl, const std
     std::string result;
 
     const llama_model * model = llama_get_model(ctx);
-    llama_token eos_token = llama_token_eos(model);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
+    llama_token eos_token = llama_vocab_eos(vocab);
 
     llama_kv_cache_clear(ctx);
     llama_set_embeddings(ctx, false);
@@ -105,7 +108,7 @@ static std::string generate(llama_context * ctx, llama_sampler * smpl, const std
 
     llama_batch bat = llama_batch_init(llama_n_batch(ctx), 0, 1);
 
-    std::vector<llama_token> inputs = common_tokenize(model, prompt, false, true);
+    std::vector<llama_token> inputs = common_tokenize(vocab, prompt, false, true);
     int32_t i_current_token = 0;
 
     while (true) {
@@ -168,7 +171,7 @@ int main(int argc, char * argv[]) {
     llama_model * model = llama_model_load_from_file(params.model.c_str(), mparams);
 
     // create generation context
-    llama_context * ctx = llama_new_context_with_model(model, cparams);
+    llama_context * ctx = llama_init_from_model(model, cparams);
 
     auto sparams = llama_sampler_chain_default_params();
 
@@ -197,7 +200,7 @@ int main(int argc, char * argv[]) {
         const std::vector<std::vector<float>> d_rep = encode(ctx, documents, gritlm_instruction(""));
         const std::vector<std::vector<float>> q_rep = encode(ctx, queries,   gritlm_instruction(instruction));
 
-        const int n_embd = llama_n_embd(model);
+        const int n_embd = llama_model_n_embd(model);
 
         const float cosine_sim_q0_d0 = common_embd_similarity_cos(q_rep[0].data(), d_rep[0].data(), n_embd);
         const float cosine_sim_q0_d1 = common_embd_similarity_cos(q_rep[0].data(), d_rep[1].data(), n_embd);

examples/imatrix/imatrix.cpp

@@ -7,7 +7,6 @@
 #include <cstdio>
 #include <cstring>
 #include <ctime>
-#include <sstream>
 #include <thread>
 #include <mutex>
 #include <vector>
@@ -40,7 +39,7 @@ public:
     void set_params(common_params params) { m_params = std::move(params); }
     bool collect_imatrix(struct ggml_tensor * t, bool ask, void * user_data);
     void save_imatrix(int ncall = -1) const;
-    bool load_imatrix(const char * file_name);
+    bool load_imatrix(const char * fname);
 private:
     std::unordered_map<std::string, Stats> m_stats;
     common_params                          m_params;
@@ -429,10 +428,13 @@ static void process_logits(
 }
 
 static bool compute_imatrix(llama_context * ctx, const common_params & params) {
-    const bool add_bos = llama_add_bos_token(llama_get_model(ctx));
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
+    const bool add_bos = llama_vocab_get_add_bos(vocab);
     const int n_ctx = llama_n_ctx(ctx);
 
-    GGML_ASSERT(!llama_add_eos_token(llama_get_model(ctx)));
+    GGML_ASSERT(!llama_vocab_get_add_eos(vocab));
 
     auto tim1 = std::chrono::high_resolution_clock::now();
     LOG_INF("%s: tokenizing the input ..\n", __func__);
@@ -468,7 +470,7 @@ static bool compute_imatrix(llama_context * ctx, const common_params & params) {
     const int n_chunk_max = tokens.size() / n_ctx;
 
     const int n_chunk = params.n_chunks < 0 ? n_chunk_max : std::min(params.n_chunks, n_chunk_max);
-    const int n_vocab = llama_n_vocab(llama_get_model(ctx));
+    const int n_vocab = llama_vocab_n_tokens(vocab);
     const int n_batch = params.n_batch;
 
     int count = 0;
@@ -508,7 +510,7 @@ static bool compute_imatrix(llama_context * ctx, const common_params & params) {
 
             // add BOS token for the first batch of each chunk
             if (add_bos && j == 0) {
-                tokens[batch_start] = llama_token_bos(llama_get_model(ctx));
+                tokens[batch_start] = llama_vocab_bos(vocab);
             }
 
             common_batch_clear(batch);
@@ -627,7 +629,7 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
-    const int n_ctx_train = llama_n_ctx_train(model);
+    const int n_ctx_train = llama_model_n_ctx_train(model);
     if (params.n_ctx > n_ctx_train) {
         LOG_WRN("%s: model was trained on only %d context tokens (%d specified)\n",
                 __func__, n_ctx_train, params.n_ctx);

examples/infill/infill.cpp

@@ -139,7 +139,9 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
-    const int n_ctx_train = llama_n_ctx_train(model);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
+    const int n_ctx_train = llama_model_n_ctx_train(model);
     const int n_ctx = llama_n_ctx(ctx);
     LOG_DBG("n_ctx: %d\n", n_ctx);
 
@@ -152,28 +154,28 @@ int main(int argc, char ** argv) {
         LOG_INF("\n");
         LOG_INF("%s\n", common_params_get_system_info(params).c_str());
     }
-    const bool add_bos = llama_add_bos_token(model);
-    GGML_ASSERT(!llama_add_eos_token(model));
+    const bool add_bos = llama_vocab_get_add_bos(vocab);
+    GGML_ASSERT(!llama_vocab_get_add_eos(vocab));
 
     std::vector<llama_token> embd_inp;
     std::vector<llama_token> embd_end;
     std::vector<llama_token> inp_pfx = common_tokenize(ctx, params.input_prefix, false);
     std::vector<llama_token> inp_sfx = common_tokenize(ctx, params.input_suffix, false);
 
-    GGML_ASSERT(llama_token_fim_pre(model) >= 0);
-    GGML_ASSERT(llama_token_fim_suf(model) >= 0);
+    GGML_ASSERT(llama_vocab_fim_pre(vocab) >= 0);
+    GGML_ASSERT(llama_vocab_fim_suf(vocab) >= 0);
 
-    inp_pfx.insert(inp_pfx.begin(), llama_token_fim_pre(model));
-    inp_sfx.insert(inp_sfx.begin(), llama_token_fim_suf(model));
+    inp_pfx.insert(inp_pfx.begin(), llama_vocab_fim_pre(vocab));
+    inp_sfx.insert(inp_sfx.begin(), llama_vocab_fim_suf(vocab));
 
     embd_inp = params.spm_infill ? inp_sfx : inp_pfx;
     embd_end = params.spm_infill ? inp_pfx : inp_sfx;
     if (add_bos) {
-        embd_inp.insert(embd_inp.begin(), llama_token_bos(model));
+        embd_inp.insert(embd_inp.begin(), llama_vocab_bos(vocab));
     }
     embd_inp.insert(embd_inp.end(), embd_end.begin(), embd_end.end());
 
-    const llama_token middle_token = llama_token_fim_mid(model);
+    const llama_token middle_token = llama_vocab_fim_mid(vocab);
     if (middle_token >= 0) {
         embd_inp.push_back(middle_token);
     }
@@ -185,7 +187,7 @@ int main(int argc, char ** argv) {
 
     // Should not run without any tokens
     if (embd_inp.empty()) {
-        embd_inp.push_back(llama_token_bos(model));
+        embd_inp.push_back(llama_vocab_bos(vocab));
         LOG_WRN("embd_inp was considered empty and bos was added: %s\n", string_from(ctx, embd_inp).c_str());
     }
 
@@ -420,10 +422,10 @@ int main(int argc, char ** argv) {
         // if not currently processing queued inputs;
         if ((int) embd_inp.size() <= n_consumed) {
             // deal with eot token in infill mode
-            if ((common_sampler_last(smpl) == llama_token_eot(model) || is_interacting) && params.interactive){
+            if ((common_sampler_last(smpl) == llama_vocab_eot(vocab) || is_interacting) && params.interactive){
                 if (is_interacting && !params.interactive_first) {
                     // print an eot token
-                    LOG("%s", common_token_to_piece(ctx, llama_token_eot(model)).c_str());
+                    LOG("%s", common_token_to_piece(ctx, llama_vocab_eot(vocab)).c_str());
                 }
                 LOG("\n");
                 console::set_display(console::user_input);
@@ -463,13 +465,13 @@ int main(int argc, char ** argv) {
                 std::vector<llama_token> inp_pfx = common_tokenize(ctx, params.input_prefix, false);
                 std::vector<llama_token> inp_sfx = common_tokenize(ctx, params.input_suffix, false);
 
-                inp_pfx.insert(inp_pfx.begin(), llama_token_fim_pre(model));
-                inp_sfx.insert(inp_sfx.begin(), llama_token_fim_suf(model));
+                inp_pfx.insert(inp_pfx.begin(), llama_vocab_fim_pre(vocab));
+                inp_sfx.insert(inp_sfx.begin(), llama_vocab_fim_suf(vocab));
 
                 embd_inp = params.spm_infill ? inp_sfx : inp_pfx;
                 embd_end = params.spm_infill ? inp_pfx : inp_sfx;
                 if (add_bos) {
-                    embd_inp.insert(embd_inp.begin(), llama_token_bos(model));
+                    embd_inp.insert(embd_inp.begin(), llama_vocab_bos(vocab));
                 }
                 embd_inp.insert(embd_inp.end(), embd_end.begin(), embd_end.end());
 
@@ -484,7 +486,7 @@ int main(int argc, char ** argv) {
                 is_interacting = false;
             }
             // deal with end of generation tokens in interactive mode
-            else if (llama_token_is_eog(model, common_sampler_last(smpl))) {
+            else if (llama_vocab_is_eog(vocab, common_sampler_last(smpl))) {
                 LOG_DBG("found EOS token\n");
 
                 if (params.interactive) {
@@ -500,7 +502,7 @@ int main(int argc, char ** argv) {
 
                 if (params.input_prefix_bos) {
                     LOG_DBG("adding input prefix BOS token\n");
-                    embd_inp.push_back(llama_token_bos(model));
+                    embd_inp.push_back(llama_vocab_bos(vocab));
                 }
 
                 std::string buffer;
@@ -563,7 +565,7 @@ int main(int argc, char ** argv) {
         }
 
         // end of generation
-        if (!embd.empty() && llama_token_is_eog(model, embd.back()) && !params.interactive) {
+        if (!embd.empty() && llama_vocab_is_eog(vocab, embd.back()) && !params.interactive) {
             break;
         }
 
@@ -575,7 +577,7 @@ int main(int argc, char ** argv) {
         }
     }
     if (!params.interactive && n_remain <= 0) {
-        LOG("%s", common_token_to_piece(ctx, llama_token_eot(model)).c_str());
+        LOG("%s", common_token_to_piece(ctx, llama_vocab_eot(vocab)).c_str());
     }
 
     LOG("\n");

examples/llama-bench/llama-bench.cpp

@@ -1401,7 +1401,8 @@ static void test_prompt(llama_context * ctx, int n_prompt, int n_batch, int n_th
     llama_set_n_threads(ctx, n_threads, n_threads);
 
     const llama_model * model   = llama_get_model(ctx);
-    const int32_t       n_vocab = llama_n_vocab(model);
+    const llama_vocab * vocab   = llama_model_get_vocab(model);
+    const int32_t       n_vocab = llama_vocab_n_tokens(vocab);
 
     std::vector<llama_token> tokens(n_batch);
 
@@ -1409,7 +1410,7 @@ static void test_prompt(llama_context * ctx, int n_prompt, int n_batch, int n_th
 
     while (n_processed < n_prompt) {
         int n_tokens = std::min(n_prompt - n_processed, n_batch);
-        tokens[0]    = n_processed == 0 && llama_add_bos_token(model) ? llama_token_bos(model) : std::rand() % n_vocab;
+        tokens[0]    = n_processed == 0 && llama_vocab_get_add_bos(vocab) ? llama_vocab_bos(vocab) : std::rand() % n_vocab;
         for (int i = 1; i < n_tokens; i++) {
             tokens[i] = std::rand() % n_vocab;
         }
@@ -1424,9 +1425,10 @@ static void test_gen(llama_context * ctx, int n_gen, int n_threads) {
     llama_set_n_threads(ctx, n_threads, n_threads);
 
     const llama_model * model   = llama_get_model(ctx);
-    const int32_t       n_vocab = llama_n_vocab(model);
+    const llama_vocab * vocab   = llama_model_get_vocab(model);
+    const int32_t       n_vocab = llama_vocab_n_tokens(vocab);
 
-    llama_token token = llama_add_bos_token(model) ? llama_token_bos(model) : std::rand() % n_vocab;
+    llama_token token = llama_vocab_get_add_bos(vocab) ? llama_vocab_bos(vocab) : std::rand() % n_vocab;
 
     for (int i = 0; i < n_gen; i++) {
         llama_decode(ctx, llama_batch_get_one(&token, 1));
@@ -1537,7 +1539,7 @@ int main(int argc, char ** argv) {
             prev_inst = &inst;
         }
 
-        llama_context * ctx = llama_new_context_with_model(lmodel, inst.to_llama_cparams());
+        llama_context * ctx = llama_init_from_model(lmodel, inst.to_llama_cparams());
         if (ctx == NULL) {
             fprintf(stderr, "%s: error: failed to create context with model '%s'\n", __func__, inst.model.c_str());
             llama_model_free(lmodel);

examples/llama.android/llama/src/main/cpp/llama-android.cpp

@@ -87,7 +87,7 @@ Java_android_llama_cpp_LLamaAndroid_load_1model(JNIEnv *env, jobject, jstring fi
     auto path_to_model = env->GetStringUTFChars(filename, 0);
     LOGi("Loading model from %s", path_to_model);
 
-    auto model = llama_load_model_from_file(path_to_model, model_params);
+    auto model = llama_model_load_from_file(path_to_model, model_params);
     env->ReleaseStringUTFChars(filename, path_to_model);
 
     if (!model) {
@@ -102,7 +102,7 @@ Java_android_llama_cpp_LLamaAndroid_load_1model(JNIEnv *env, jobject, jstring fi
 extern "C"
 JNIEXPORT void JNICALL
 Java_android_llama_cpp_LLamaAndroid_free_1model(JNIEnv *, jobject, jlong model) {
-    llama_free_model(reinterpret_cast<llama_model *>(model));
+    llama_model_free(reinterpret_cast<llama_model *>(model));
 }
 
 extern "C"
@@ -405,6 +405,7 @@ Java_android_llama_cpp_LLamaAndroid_completion_1loop(
     const auto batch   = reinterpret_cast<llama_batch   *>(batch_pointer);
     const auto sampler = reinterpret_cast<llama_sampler *>(sampler_pointer);
     const auto model = llama_get_model(context);
+    const auto vocab = llama_model_get_vocab(model);
 
     if (!la_int_var) la_int_var = env->GetObjectClass(intvar_ncur);
     if (!la_int_var_value) la_int_var_value = env->GetMethodID(la_int_var, "getValue", "()I");
@@ -414,7 +415,7 @@ Java_android_llama_cpp_LLamaAndroid_completion_1loop(
     const auto new_token_id = llama_sampler_sample(sampler, context, -1);
 
     const auto n_cur = env->CallIntMethod(intvar_ncur, la_int_var_value);
-    if (llama_token_is_eog(model, new_token_id) || n_cur == n_len) {
+    if (llama_vocab_is_eog(vocab, new_token_id) || n_cur == n_len) {
         return nullptr;
     }
 

examples/llama.swiftui/llama.cpp.swift/LibLlama.swift

@@ -52,8 +52,8 @@ actor LlamaContext {
     deinit {
         llama_sampler_free(sampling)
         llama_batch_free(batch)
+        llama_model_free(model)
         llama_free(context)
-        llama_free_model(model)
         llama_backend_free()
     }
 
@@ -65,7 +65,7 @@ actor LlamaContext {
         model_params.n_gpu_layers = 0
         print("Running on simulator, force use n_gpu_layers = 0")
 #endif
-        let model = llama_load_model_from_file(path, model_params)
+        let model = llama_model_load_from_file(path, model_params)
         guard let model else {
             print("Could not load model at \(path)")
             throw LlamaError.couldNotInitializeContext
@@ -151,7 +151,7 @@ actor LlamaContext {
 
         new_token_id = llama_sampler_sample(sampling, context, batch.n_tokens - 1)
 
-        if llama_token_is_eog(model, new_token_id) || n_cur == n_len {
+        if llama_vocab_is_eog(model, new_token_id) || n_cur == n_len {
             print("\n")
             is_done = true
             let new_token_str = String(cString: temporary_invalid_cchars + [0])

examples/llava/clip.cpp

@@ -7,6 +7,7 @@
 #include "ggml-cpu.h"
 #include "ggml-alloc.h"
 #include "ggml-backend.h"
+#include "gguf.h"
 
 //#ifdef GGML_USE_CUDA
 //#include "ggml-cuda.h"
@@ -262,7 +263,7 @@ static std::string gguf_kv_to_str(const struct gguf_context * ctx_gguf, int i) {
             {
                 const enum gguf_type arr_type = gguf_get_arr_type(ctx_gguf, i);
                 int arr_n = gguf_get_arr_n(ctx_gguf, i);
-                const void * data = gguf_get_arr_data(ctx_gguf, i);
+                const void * data = arr_type == GGUF_TYPE_STRING ? nullptr : gguf_get_arr_data(ctx_gguf, i);
                 std::stringstream ss;
                 ss << "[";
                 for (int j = 0; j < arr_n; j++) {
@@ -2734,7 +2735,8 @@ bool clip_model_quantize(const char * fname_inp, const char * fname_out, const i
         total_size_org += orig_size;
         total_size_new += new_size;
         gguf_set_tensor_type(ctx_out, name.c_str(), new_type);
-        gguf_set_tensor_data(ctx_out, name.c_str(), new_data, new_size);
+        GGML_ASSERT(gguf_get_tensor_size(ctx_out, gguf_find_tensor(ctx_out, name.c_str())) == new_size);
+        gguf_set_tensor_data(ctx_out, name.c_str(), new_data);
         fout.write((const char *)new_data, new_size);
         size_t pad = GGML_PAD(new_size, gguf_get_alignment(ctx_out)) - new_size;
         for (size_t j = 0; j < pad; ++j) {

examples/llava/llava-cli.cpp

@@ -47,8 +47,12 @@ static const char * sample(struct common_sampler * smpl,
                            int * n_past) {
     const llama_token id = common_sampler_sample(smpl, ctx_llama, -1);
     common_sampler_accept(smpl, id, true);
+
+    const llama_model * model = llama_get_model(ctx_llama);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     static std::string ret;
-    if (llama_token_is_eog(llama_get_model(ctx_llama), id)) {
+    if (llama_vocab_is_eog(vocab, id)) {
         ret = "</s>";
     } else {
         ret = common_token_to_piece(ctx_llama, id);
@@ -239,11 +243,10 @@ static struct llava_context * llava_init_context(common_params * params, llama_m
 
     auto ctx_clip = clip_model_load(clip_path, /*verbosity=*/ 1);
 
-
     llama_context_params ctx_params = common_context_params_to_llama(*params);
     ctx_params.n_ctx           = params->n_ctx < 2048 ? 2048 : params->n_ctx; // we need a longer context size to process image embeddings
 
-    llama_context * ctx_llama = llama_new_context_with_model(model, ctx_params);
+    llama_context * ctx_llama = llama_init_from_model(model, ctx_params);
 
     if (ctx_llama == NULL) {
         LOG_ERR("%s: failed to create the llama_context\n" , __func__);

examples/llava/llava.cpp

@@ -384,7 +384,7 @@ static bool encode_image_with_clip(clip_ctx * ctx_clip, int n_threads, const cli
 
 bool llava_validate_embed_size(const llama_context * ctx_llama, const clip_ctx * ctx_clip) {
         // make sure that the correct mmproj was used, i.e., compare apples to apples
-    int n_llama_embd = llama_n_embd(llama_get_model(ctx_llama));
+    int n_llama_embd = llama_model_n_embd(llama_get_model(ctx_llama));
     auto n_image_embd = clip_n_mmproj_embd(ctx_clip);
     if (n_image_embd != n_llama_embd) {
         LOG_ERR("%s: embedding dim of the multimodal projector (%d) is not equal to that of LLaMA (%d). Make sure that you use the correct mmproj file.\n", __func__, n_image_embd, n_llama_embd);
@@ -456,7 +456,7 @@ struct llava_embd_batch {
 };
 
 bool llava_eval_image_embed(llama_context * ctx_llama, const struct llava_image_embed * image_embed, int n_batch, int * n_past) {
-    int n_embd  = llama_n_embd(llama_get_model(ctx_llama));
+    int n_embd  = llama_model_n_embd(llama_get_model(ctx_llama));
 
     for (int i = 0; i < image_embed->n_image_pos; i += n_batch) {
         int n_eval = image_embed->n_image_pos - i;

examples/llava/minicpmv-cli.cpp

@@ -54,7 +54,7 @@ static struct llava_context * llava_init_context(common_params * params, llama_m
         ctx_params.n_ctx = params->n_ctx;
     }
 
-    llama_context * ctx_llama = llama_new_context_with_model(model, ctx_params);
+    llama_context * ctx_llama = llama_init_from_model(model, ctx_params);
 
     if (ctx_llama == NULL) {
         LOG_ERR("%s: failed to create the llama_context\n" , __func__);
@@ -167,8 +167,12 @@ static const char * sample(struct common_sampler * smpl,
                            int * n_past) {
     const llama_token id = common_sampler_sample(smpl, ctx_llama, -1);
     common_sampler_accept(smpl, id, true);
+
+    const llama_model * model = llama_get_model(ctx_llama);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     static std::string ret;
-    if (llama_token_is_eog(llama_get_model(ctx_llama), id)) {
+    if (llama_vocab_is_eog(vocab, id)) {
         ret = "</s>";
     } else {
         ret = common_token_to_piece(ctx_llama, id);

examples/llava/qwen2vl-cli.cpp

@@ -27,7 +27,7 @@
 
 static bool qwen2vl_eval_image_embed(llama_context * ctx_llama, const struct llava_image_embed * image_embed,
                                      int n_batch, int * n_past, int * st_pos_id, struct clip_image_size * image_size) {
-    int n_embd  = llama_n_embd(llama_get_model(ctx_llama));
+    int n_embd  = llama_model_n_embd(llama_get_model(ctx_llama));
     const int patch_size = 14 * 2;
     const int ph = image_size->height / patch_size + (image_size->height % patch_size > 0);
     const int pw = image_size->width / patch_size + (image_size->width % patch_size > 0);
@@ -132,8 +132,12 @@ static const char * sample(struct common_sampler * smpl,
                            int * n_past, int * st_pos_id) {
     const llama_token id = common_sampler_sample(smpl, ctx_llama, -1);
     common_sampler_accept(smpl, id, true);
+
+    const llama_model * model = llama_get_model(ctx_llama);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     static std::string ret;
-    if (llama_token_is_eog(llama_get_model(ctx_llama), id)) {
+    if (llama_vocab_is_eog(vocab, id)) {
         ret = "</s>";
     } else {
         ret = common_token_to_piece(ctx_llama, id);
@@ -328,11 +332,10 @@ static struct llava_context * llava_init_context(common_params * params, llama_m
 
     auto ctx_clip = clip_model_load(clip_path, /*verbosity=*/ 1);
 
-
     llama_context_params ctx_params = common_context_params_to_llama(*params);
     ctx_params.n_ctx           = params->n_ctx < 2048 ? 2048 : params->n_ctx; // we need a longer context size to process image embeddings
 
-    llama_context * ctx_llama = llama_new_context_with_model(model, ctx_params);
+    llama_context * ctx_llama = llama_init_from_model(model, ctx_params);
 
     if (ctx_llama == NULL) {
         LOG_ERR("%s: failed to create the llama_context\n" , __func__);
@@ -481,7 +484,7 @@ static void debug_test_mrope_2d() {
 }
 
 static void debug_dump_img_embed(struct llava_context * ctx_llava) {
-    int n_embd  = llama_n_embd(llama_get_model(ctx_llava->ctx_llama));
+    int n_embd  = llama_model_n_embd(llama_get_model(ctx_llava->ctx_llama));
     int ne = n_embd * 4;
     float vals[56 * 56 * 3];
     // float embd[ne];

examples/lookahead/lookahead.cpp

@@ -61,6 +61,8 @@ int main(int argc, char ** argv) {
     llama_model * model = llama_init.model.get();
     llama_context * ctx = llama_init.context.get();
 
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     // Tokenize the prompt
     std::vector<llama_token> inp;
     std::vector<llama_token> all;
@@ -147,7 +149,7 @@ int main(int argc, char ** argv) {
     }
 
     // here we keep adding new n-grams as we go
-    ngram_container ngrams_observed(llama_n_vocab(model), N, G);
+    ngram_container ngrams_observed(llama_vocab_n_tokens(vocab), N, G);
 
     // debug
     struct llama_kv_cache_view kvc_view = llama_kv_cache_view_init(ctx, W + G + 1);
@@ -297,7 +299,7 @@ int main(int argc, char ** argv) {
                 }
                 fflush(stdout);
 
-                if (llama_token_is_eog(model, id)) {
+                if (llama_vocab_is_eog(vocab, id)) {
                     has_eos = true;
                 }
 

examples/lookup/lookup.cpp

@@ -36,6 +36,8 @@ int main(int argc, char ** argv){
     llama_model * model = llama_init.model.get();
     llama_context * ctx = llama_init.context.get();
 
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     // tokenize the prompt
     std::vector<llama_token> inp;
     inp = common_tokenize(ctx, params.prompt, true, true);
@@ -136,7 +138,7 @@ int main(int argc, char ** argv){
                 LOG("%s", token_str.c_str());
             }
 
-            if (llama_token_is_eog(model, id)) {
+            if (llama_vocab_is_eog(vocab, id)) {
                 has_eos = true;
             }
 

examples/main/main.cpp

@@ -5,7 +5,6 @@
 #include "sampling.h"
 #include "llama.h"
 
-#include <cassert>
 #include <cstdio>
 #include <cstring>
 #include <ctime>
@@ -31,6 +30,8 @@
 #pragma warning(disable: 4244 4267) // possible loss of data
 #endif
 
+static const char * DEFAULT_SYSTEM_MESSAGE = "You are a helpful assistant";
+
 static llama_context           ** g_ctx;
 static llama_model             ** g_model;
 static common_sampler          ** g_smpl;
@@ -163,6 +164,8 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     LOG_INF("%s: llama threadpool init, n_threads = %d\n", __func__, (int) params.cpuparams.n_threads);
 
     auto * reg = ggml_backend_dev_backend_reg(ggml_backend_dev_by_type(GGML_BACKEND_DEVICE_TYPE_CPU));
@@ -196,15 +199,31 @@ int main(int argc, char ** argv) {
 
     llama_attach_threadpool(ctx, threadpool, threadpool_batch);
 
-    const int n_ctx_train = llama_n_ctx_train(model);
+    const int n_ctx_train = llama_model_n_ctx_train(model);
     const int n_ctx = llama_n_ctx(ctx);
 
     if (n_ctx > n_ctx_train) {
         LOG_WRN("%s: model was trained on only %d context tokens (%d specified)\n", __func__, n_ctx_train, n_ctx);
     }
 
+    // auto enable conversation mode if chat template is available
+    const bool has_chat_template = !common_get_builtin_chat_template(model).empty() || !params.chat_template.empty();
+    if (params.conversation_mode == COMMON_CONVERSATION_MODE_AUTO) {
+        if (has_chat_template) {
+            LOG_INF("%s: chat template is available, enabling conversation mode (disable it with -no-cnv)\n", __func__);
+            params.conversation_mode = COMMON_CONVERSATION_MODE_ENABLED;
+        } else {
+            params.conversation_mode = COMMON_CONVERSATION_MODE_DISABLED;
+        }
+    }
+
+    // in case user force-activate conversation mode (via -cnv) without proper chat template, we show a warning
+    if (params.conversation_mode && !has_chat_template) {
+        LOG_WRN("%s: chat template is not available or is not supported. This may cause the model to output suboptimal responses\n", __func__);
+    }
+
     // print chat template example in conversation mode
-    if (params.conversation) {
+    if (params.conversation_mode) {
         if (params.enable_chat_template) {
             LOG_INF("%s: chat template example:\n%s\n", __func__, common_chat_format_example(model, params.chat_template).c_str());
         } else {
@@ -241,9 +260,9 @@ int main(int argc, char ** argv) {
         }
     }
 
-    const bool add_bos = llama_add_bos_token(model);
+    const bool add_bos = llama_vocab_get_add_bos(vocab);
     if (!llama_model_has_encoder(model)) {
-        GGML_ASSERT(!llama_add_eos_token(model));
+        GGML_ASSERT(!llama_vocab_get_add_eos(vocab));
     }
 
     LOG_DBG("n_ctx: %d, add_bos: %d\n", n_ctx, add_bos);
@@ -251,8 +270,10 @@ int main(int argc, char ** argv) {
     std::vector<llama_token> embd_inp;
 
     {
-        auto prompt = (params.conversation && params.enable_chat_template && !params.prompt.empty())
-            ? chat_add_and_format(model, chat_msgs, "system", params.prompt) // format the system prompt in conversation mode
+        auto prompt = (params.conversation_mode && params.enable_chat_template)
+            // format the system prompt in conversation mode (fallback to default if empty)
+            ? chat_add_and_format(model, chat_msgs, "system", params.prompt.empty() ? DEFAULT_SYSTEM_MESSAGE : params.prompt)
+            // otherwise use the prompt as is
             : params.prompt;
         if (params.interactive_first || !params.prompt.empty() || session_tokens.empty()) {
             LOG_DBG("tokenize the prompt\n");
@@ -269,7 +290,7 @@ int main(int argc, char ** argv) {
     // Should not run without any tokens
     if (embd_inp.empty()) {
         if (add_bos) {
-            embd_inp.push_back(llama_token_bos(model));
+            embd_inp.push_back(llama_vocab_bos(vocab));
             LOG_WRN("embd_inp was considered empty and bos was added: %s\n", string_from(ctx, embd_inp).c_str());
         } else {
             LOG_ERR("input is empty\n");
@@ -326,7 +347,7 @@ int main(int argc, char ** argv) {
         params.n_keep += add_bos; // always keep the BOS token
     }
 
-    if (params.conversation) {
+    if (params.conversation_mode) {
         params.interactive_first = true;
     }
 
@@ -450,7 +471,11 @@ int main(int argc, char ** argv) {
 #if defined (__unix__) || (defined (__APPLE__) && defined (__MACH__)) || defined (_WIN32)
         LOG_INF(       " - Press Ctrl+C to interject at any time.\n");
 #endif
-        LOG_INF(       "%s\n", control_message);
+        LOG_INF(       "%s", control_message);
+        if (params.conversation_mode && params.enable_chat_template && params.prompt.empty()) {
+            LOG_INF(   " - Using default system message. To change it, set a different value via -p PROMPT or -f FILE argument.\n");
+        }
+        LOG_INF("\n");
 
         is_interacting = params.interactive_first;
     }
@@ -495,7 +520,7 @@ int main(int argc, char ** argv) {
 
         llama_token decoder_start_token_id = llama_model_decoder_start_token(model);
         if (decoder_start_token_id == LLAMA_TOKEN_NULL) {
-            decoder_start_token_id = llama_token_bos(model);
+            decoder_start_token_id = llama_vocab_bos(vocab);
         }
 
         embd_inp.clear();
@@ -742,7 +767,7 @@ int main(int argc, char ** argv) {
             }
 
             // deal with end of generation tokens in interactive mode
-            if (llama_token_is_eog(model, common_sampler_last(smpl))) {
+            if (llama_vocab_is_eog(vocab, common_sampler_last(smpl))) {
                 LOG_DBG("found an EOG token\n");
 
                 if (params.interactive) {
@@ -762,7 +787,7 @@ int main(int argc, char ** argv) {
             }
 
             // if current token is not EOG, we add it to current assistant message
-            if (params.conversation) {
+            if (params.conversation_mode) {
                 const auto id = common_sampler_last(smpl);
                 assistant_ss << common_token_to_piece(ctx, id, false);
             }
@@ -770,17 +795,17 @@ int main(int argc, char ** argv) {
             if (n_past > 0 && is_interacting) {
                 LOG_DBG("waiting for user input\n");
 
-                if (params.conversation) {
+                if (params.conversation_mode) {
                     LOG("\n> ");
                 }
 
                 if (params.input_prefix_bos) {
                     LOG_DBG("adding input prefix BOS token\n");
-                    embd_inp.push_back(llama_token_bos(model));
+                    embd_inp.push_back(llama_vocab_bos(vocab));
                 }
 
                 std::string buffer;
-                if (!params.input_prefix.empty() && !params.conversation) {
+                if (!params.input_prefix.empty() && !params.conversation_mode) {
                     LOG_DBG("appending input prefix: '%s'\n", params.input_prefix.c_str());
                     LOG("%s", params.input_prefix.c_str());
                 }
@@ -804,7 +829,7 @@ int main(int argc, char ** argv) {
                 // Entering a empty line lets the user pass control back
                 if (buffer.length() > 1) {
                     // append input suffix if any
-                    if (!params.input_suffix.empty() && !params.conversation) {
+                    if (!params.input_suffix.empty() && !params.conversation_mode) {
                         LOG_DBG("appending input suffix: '%s'\n", params.input_suffix.c_str());
                         LOG("%s", params.input_suffix.c_str());
                     }
@@ -817,7 +842,7 @@ int main(int argc, char ** argv) {
                         string_process_escapes(buffer);
                     }
 
-                    bool format_chat = params.conversation && params.enable_chat_template;
+                    bool format_chat = params.conversation_mode && params.enable_chat_template;
                     std::string user_inp = format_chat
                         ? chat_add_and_format(model, chat_msgs, "user", std::move(buffer))
                         : std::move(buffer);
@@ -830,8 +855,8 @@ int main(int argc, char ** argv) {
 
                     // if user stop generation mid-way, we must add EOT to finish model's last response
                     if (need_insert_eot && format_chat) {
-                        llama_token eot = llama_token_eot(model);
-                        embd_inp.push_back(eot == LLAMA_TOKEN_NULL ? llama_token_eos(model) : eot);
+                        llama_token eot = llama_vocab_eot(vocab);
+                        embd_inp.push_back(eot == LLAMA_TOKEN_NULL ? llama_vocab_eos(vocab) : eot);
                         need_insert_eot = false;
                     }
 
@@ -866,7 +891,7 @@ int main(int argc, char ** argv) {
         }
 
         // end of generation
-        if (!embd.empty() && llama_token_is_eog(model, embd.back()) && !(params.interactive)) {
+        if (!embd.empty() && llama_vocab_is_eog(vocab, embd.back()) && !(params.interactive)) {
             LOG(" [end of text]\n");
             break;
         }

examples/parallel/parallel.cpp

@@ -135,6 +135,8 @@ int main(int argc, char ** argv) {
     llama_model * model = llama_init.model.get();
     llama_context * ctx = llama_init.context.get();
 
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     // load the prompts from an external file if there are any
     if (params.prompt.empty()) {
         LOG_INF("\033[32mNo new questions so proceed with build-in defaults.\033[0m\n");
@@ -358,7 +360,7 @@ int main(int argc, char ** argv) {
                 //        client.id, client.seq_id, id, client.n_decoded, client.i_batch, token_str.c_str());
 
                 if (client.n_decoded > 2 &&
-                        (llama_token_is_eog(model, id) ||
+                        (llama_vocab_is_eog(vocab, id) ||
                          (params.n_predict > 0 && client.n_decoded + client.n_prompt >= params.n_predict) ||
                          client.response.find("User:") != std::string::npos ||
                          client.response.find('\n') != std::string::npos)) {

examples/passkey/passkey.cpp

@@ -70,15 +70,17 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     // initialize the context
 
     llama_context_params ctx_params = common_context_params_to_llama(params);
 
-    ctx_params.n_ctx = llama_n_ctx_train(model)*n_grp + n_keep;
+    ctx_params.n_ctx = llama_model_n_ctx_train(model)*n_grp + n_keep;
 
     GGML_ASSERT(ctx_params.n_batch % n_grp == 0 && "n_batch must be divisible by n_grp");
 
-    llama_context * ctx = llama_new_context_with_model(model, ctx_params);
+    llama_context * ctx = llama_init_from_model(model, ctx_params);
     if (ctx == NULL) {
         LOG_ERR("%s: failed to create the llama_context\n" , __func__);
         return 1;
@@ -223,7 +225,7 @@ int main(int argc, char ** argv) {
             const llama_token new_token_id = llama_sampler_sample(smpl, ctx, batch.n_tokens - 1);
 
             // is it an end of generation?
-            if (llama_token_is_eog(model, new_token_id) || n_cur == n_len) {
+            if (llama_vocab_is_eog(vocab, new_token_id) || n_cur == n_len) {
                 LOG("\n");
 
                 break;

examples/perplexity/perplexity.cpp

@@ -296,8 +296,11 @@ static results_perplexity perplexity_v2(llama_context * ctx, const common_params
     // Output: `perplexity: 13.5106 [114/114]`
     // BOS tokens will be added for each chunk before eval
 
-    const bool add_bos = llama_add_bos_token(llama_get_model(ctx));
-    GGML_ASSERT(!llama_add_eos_token(llama_get_model(ctx)));
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
+    const bool add_bos = llama_vocab_get_add_bos(vocab);
+    GGML_ASSERT(!llama_vocab_get_add_eos(vocab));
 
     LOG_INF("%s: tokenizing the input ..\n", __func__);
 
@@ -338,7 +341,7 @@ static results_perplexity perplexity_v2(llama_context * ctx, const common_params
     const int n_chunk = params.n_chunks < 0 ? n_chunk_max : std::min(params.n_chunks, n_chunk_max);
     const int n_batch = params.n_batch;
 
-    const int n_vocab = llama_n_vocab(llama_get_model(ctx));
+    const int n_vocab = llama_vocab_n_tokens(vocab);
 
     int count = 0;
     double nll = 0.0;
@@ -382,7 +385,7 @@ static results_perplexity perplexity_v2(llama_context * ctx, const common_params
 
             // add BOS token for the first batch of each chunk
             if (add_bos && j == 0) {
-                tokens[batch_start] = llama_token_bos(llama_get_model(ctx));
+                tokens[batch_start] = llama_vocab_bos(vocab);
             }
 
             const auto * batch_logits = llama_get_logits(ctx);
@@ -444,8 +447,11 @@ static results_perplexity perplexity(llama_context * ctx, const common_params &
     // Output: `perplexity: 13.5106 [114/114]`
     // BOS tokens will be added for each chunk before eval
 
-    const bool add_bos = llama_add_bos_token(llama_get_model(ctx));
-    GGML_ASSERT(!llama_add_eos_token(llama_get_model(ctx)));
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
+    const bool add_bos = llama_vocab_get_add_bos(vocab);
+    GGML_ASSERT(!llama_vocab_get_add_eos(vocab));
 
     std::ofstream logits_stream;
     if (!params.logits_file.empty()) {
@@ -485,7 +491,7 @@ static results_perplexity perplexity(llama_context * ctx, const common_params &
     const int n_chunk = params.n_chunks < 0 ? n_chunk_max : std::min(params.n_chunks, n_chunk_max);
     const int n_batch = params.n_batch;
 
-    const int n_vocab = llama_n_vocab(llama_get_model(ctx));
+    const int n_vocab = llama_vocab_n_tokens(vocab);
 
     int count = 0;
     double nll = 0.0;
@@ -557,7 +563,7 @@ static results_perplexity perplexity(llama_context * ctx, const common_params &
 
                 // add BOS token for the first batch of each chunk
                 if (add_bos && j == 0) {
-                    tokens[seq_start] = llama_token_bos(llama_get_model(ctx));
+                    tokens[seq_start] = llama_vocab_bos(vocab);
                 }
 
                 for (int k = 0; k < batch_size; ++k) {
@@ -732,6 +738,9 @@ static void compute_logprobs(const float * batch_logits, int n_vocab, std::vecto
 }
 
 static void hellaswag_score(llama_context * ctx, const common_params & params) {
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     // Calculates hellaswag score (acc_norm) from prompt
     //
     // Data extracted from the HellaSwag validation dataset (MIT license) https://github.com/rowanz/hellaswag/blob/master/data/hellaswag_val.jsonl
@@ -765,7 +774,7 @@ static void hellaswag_score(llama_context * ctx, const common_params & params) {
     size_t hs_task_count = prompt_lines.size()/6;
     LOG_INF("%s : loaded %zu tasks from prompt.\n", __func__, hs_task_count);
 
-    const bool is_spm = llama_vocab_type(llama_get_model(ctx)) == LLAMA_VOCAB_TYPE_SPM;
+    const bool is_spm = llama_vocab_type(vocab) == LLAMA_VOCAB_TYPE_SPM;
     LOG_INF("================================= is_spm = %d\n", is_spm);
 
     // The tasks should be randomized so the score stabilizes quickly.
@@ -848,7 +857,7 @@ static void hellaswag_score(llama_context * ctx, const common_params & params) {
     const int n_ctx   = llama_n_ctx(ctx);
     const int n_batch = params.n_batch;
 
-    const int n_vocab = llama_n_vocab(llama_get_model(ctx));
+    const int n_vocab = llama_vocab_n_tokens(vocab);
 
     const int max_tasks_per_batch = 32;
     const int max_seq = std::min(4*max_tasks_per_batch, (int) llama_n_seq_max(ctx));
@@ -1072,6 +1081,8 @@ static std::vector<winogrande_entry> load_winogrande_from_csv(const std::string
  *
  */
 static void winogrande_score(llama_context * ctx, const common_params & params) {
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
 
     constexpr int k_min_trailing_ctx = 3;
 
@@ -1130,7 +1141,7 @@ static void winogrande_score(llama_context * ctx, const common_params & params)
     const int n_ctx   = llama_n_ctx(ctx);
     const int n_batch = params.n_batch;
 
-    const int n_vocab = llama_n_vocab(llama_get_model(ctx));
+    const int n_vocab = llama_vocab_n_tokens(vocab);
 
     const int max_tasks_per_batch = 128;
     const int max_seq = std::min(2*max_tasks_per_batch, (int) llama_n_seq_max(ctx));
@@ -1374,6 +1385,8 @@ static bool multiple_choice_prepare_one_task(llama_context * ctx, multiple_choic
 //     https://huggingface.co/datasets/truthful_qa
 //
 static void multiple_choice_score(llama_context * ctx, const common_params & params) {
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
 
     std::istringstream strstream(params.prompt);
     uint32_t n_task;
@@ -1482,7 +1495,7 @@ static void multiple_choice_score(llama_context * ctx, const common_params & par
     const int n_ctx   = llama_n_ctx(ctx);
     const int n_batch = params.n_batch;
 
-    const int n_vocab = llama_n_vocab(llama_get_model(ctx));
+    const int n_vocab = llama_vocab_n_tokens(vocab);
 
     const int max_tasks_per_batch = 32;
     const int max_seq = std::min(4*max_tasks_per_batch, (int) llama_n_seq_max(ctx));
@@ -1655,6 +1668,9 @@ static void multiple_choice_score(llama_context * ctx, const common_params & par
 }
 
 static void kl_divergence(llama_context * ctx, const common_params & params) {
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     if (params.logits_file.empty()) {
         LOG_ERR("%s: you must provide a name of a file containing the log probabilities of the base model\n", __func__);
         return;
@@ -1688,8 +1704,8 @@ static void kl_divergence(llama_context * ctx, const common_params & params) {
         LOG_ERR("%s: failed reading n_vocab, n_chunk from %s\n", __func__, params.logits_file.c_str());
         return;
     }
-    if (n_vocab != llama_n_vocab(llama_get_model(ctx))) {
-        LOG_ERR("%s: inconsistent vocabulary (%d vs %d)\n", __func__, n_vocab, llama_n_vocab(llama_get_model(ctx)));
+    if (n_vocab != llama_vocab_n_tokens(vocab)) {
+        LOG_ERR("%s: inconsistent vocabulary (%d vs %d)\n", __func__, n_vocab, llama_vocab_n_tokens(vocab));
     }
 
     std::vector<llama_token> tokens(size_t(n_ctx) * n_chunk);
@@ -1701,8 +1717,8 @@ static void kl_divergence(llama_context * ctx, const common_params & params) {
     const int n_batch = params.n_batch;
     const int num_batches = (n_ctx + n_batch - 1)/n_batch;
     const int nv = 2*((n_vocab + 1)/2) + 4;
-    const bool add_bos = llama_add_bos_token(llama_get_model(ctx));
-    GGML_ASSERT(!llama_add_eos_token(llama_get_model(ctx)));
+    const bool add_bos = llama_vocab_get_add_bos(vocab);
+    GGML_ASSERT(!llama_vocab_get_add_eos(vocab));
 
     std::vector<uint16_t> log_probs_uint16(size_t(n_ctx - 1 - n_ctx/2) * nv);
     std::vector<float>    kld_values(size_t(n_ctx - 1 - n_ctx/2)*n_chunk);
@@ -1761,7 +1777,7 @@ static void kl_divergence(llama_context * ctx, const common_params & params) {
 
             // add BOS token for the first batch of each chunk
             if (add_bos && j == 0) {
-                tokens[batch_start] = llama_token_bos(llama_get_model(ctx));
+                tokens[batch_start] = llama_vocab_bos(vocab);
             }
 
             common_batch_clear(batch);
@@ -1995,7 +2011,7 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
-    const int n_ctx_train = llama_n_ctx_train(model);
+    const int n_ctx_train = llama_model_n_ctx_train(model);
 
     if (params.n_ctx > n_ctx_train) {
         LOG_WRN("%s: model was trained on only %d context tokens (%d specified)\n",

examples/quantize-stats/quantize-stats.cpp

@@ -319,7 +319,7 @@ int main(int argc, char ** argv) {
         auto cparams = llama_context_default_params();
         cparams.n_ctx = 256;
 
-        ctx = llama_new_context_with_model(model, cparams);
+        ctx = llama_init_from_model(model, cparams);
 
         if (ctx == NULL) {
             fprintf(stderr, "%s: error: failed to create context with model '%s'\n", __func__, params.model.c_str());

examples/retrieval/retrieval.cpp

@@ -159,7 +159,9 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
-    const int n_ctx_train = llama_n_ctx_train(model);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
+    const int n_ctx_train = llama_model_n_ctx_train(model);
     const int n_ctx = llama_n_ctx(ctx);
 
     const enum llama_pooling_type pooling_type = llama_pooling_type(ctx);
@@ -192,8 +194,8 @@ int main(int argc, char ** argv) {
             return 1;
         }
         // add eos if not present
-        if (llama_token_eos(model) >= 0 && (inp.empty() || inp.back() != llama_token_eos(model))) {
-            inp.push_back(llama_token_eos(model));
+        if (llama_vocab_eos(vocab) >= 0 && (inp.empty() || inp.back() != llama_vocab_eos(vocab))) {
+            inp.push_back(llama_vocab_eos(vocab));
         }
         chunk.tokens = inp;
     }
@@ -215,7 +217,7 @@ int main(int argc, char ** argv) {
     struct llama_batch batch = llama_batch_init(n_batch, 0, 1);
 
     // allocate output
-    const int n_embd = llama_n_embd(model);
+    const int n_embd = llama_model_n_embd(model);
     std::vector<float> embeddings(n_chunks * n_embd, 0);
     float * emb = embeddings.data();
 

examples/run/run.cpp

@@ -11,6 +11,8 @@
 #    include <curl/curl.h>
 #endif
 
+#include <signal.h>
+
 #include <climits>
 #include <cstdarg>
 #include <cstdio>
@@ -25,6 +27,13 @@
 #include "json.hpp"
 #include "llama-cpp.h"
 
+#if defined(__unix__) || (defined(__APPLE__) && defined(__MACH__)) || defined(_WIN32)
+[[noreturn]] static void sigint_handler(int) {
+    printf("\n\033[0m");
+    exit(0);  // not ideal, but it's the only way to guarantee exit in all cases
+}
+#endif
+
 GGML_ATTRIBUTE_FORMAT(1, 2)
 static std::string fmt(const char * fmt, ...) {
     va_list ap;
@@ -83,6 +92,7 @@ class Opt {
         }
 
         ctx_params.n_batch        = context_size >= 0 ? context_size : context_size_default;
+        ctx_params.n_ctx          = ctx_params.n_batch;
         model_params.n_gpu_layers = ngl >= 0 ? ngl : ngl_default;
         temperature               = temperature >= 0 ? temperature : temperature_default;
 
@@ -675,7 +685,7 @@ class LlamaData {
 
     // Initializes the context with the specified parameters
     llama_context_ptr initialize_context(const llama_model_ptr & model, const Opt & opt) {
-        llama_context_ptr context(llama_new_context_with_model(model.get(), opt.ctx_params));
+        llama_context_ptr context(llama_init_from_model(model.get(), opt.ctx_params));
         if (!context) {
             printe("%s: error: failed to create the llama_context\n", __func__);
         }
@@ -703,11 +713,11 @@ static void add_message(const char * role, const std::string & text, LlamaData &
 // Function to apply the chat template and resize `formatted` if needed
 static int apply_chat_template(LlamaData & llama_data, const bool append) {
     int result = llama_chat_apply_template(
-        llama_data.model.get(), nullptr, llama_data.messages.data(), llama_data.messages.size(), append,
+        llama_model_chat_template(llama_data.model.get()), llama_data.messages.data(), llama_data.messages.size(), append,
         append ? llama_data.fmtted.data() : nullptr, append ? llama_data.fmtted.size() : 0);
     if (append && result > static_cast<int>(llama_data.fmtted.size())) {
         llama_data.fmtted.resize(result);
-        result = llama_chat_apply_template(llama_data.model.get(), nullptr, llama_data.messages.data(),
+        result = llama_chat_apply_template(llama_model_chat_template(llama_data.model.get()), llama_data.messages.data(),
                                            llama_data.messages.size(), append, llama_data.fmtted.data(),
                                            llama_data.fmtted.size());
     }
@@ -716,11 +726,11 @@ static int apply_chat_template(LlamaData & llama_data, const bool append) {
 }
 
 // Function to tokenize the prompt
-static int tokenize_prompt(const llama_model_ptr & model, const std::string & prompt,
+static int tokenize_prompt(const llama_vocab * vocab, const std::string & prompt,
                            std::vector<llama_token> & prompt_tokens) {
-    const int n_prompt_tokens = -llama_tokenize(model.get(), prompt.c_str(), prompt.size(), NULL, 0, true, true);
+    const int n_prompt_tokens = -llama_tokenize(vocab, prompt.c_str(), prompt.size(), NULL, 0, true, true);
     prompt_tokens.resize(n_prompt_tokens);
-    if (llama_tokenize(model.get(), prompt.c_str(), prompt.size(), prompt_tokens.data(), prompt_tokens.size(), true,
+    if (llama_tokenize(vocab, prompt.c_str(), prompt.size(), prompt_tokens.data(), prompt_tokens.size(), true,
                        true) < 0) {
         printe("failed to tokenize the prompt\n");
         return -1;
@@ -743,9 +753,9 @@ static int check_context_size(const llama_context_ptr & ctx, const llama_batch &
 }
 
 // convert the token to a string
-static int convert_token_to_string(const llama_model_ptr & model, const llama_token token_id, std::string & piece) {
+static int convert_token_to_string(const llama_vocab * vocab, const llama_token token_id, std::string & piece) {
     char buf[256];
-    int  n = llama_token_to_piece(model.get(), token_id, buf, sizeof(buf), 0, true);
+    int  n = llama_token_to_piece(vocab, token_id, buf, sizeof(buf), 0, true);
     if (n < 0) {
         printe("failed to convert token to piece\n");
         return 1;
@@ -763,8 +773,10 @@ static void print_word_and_concatenate_to_response(const std::string & piece, st
 
 // helper function to evaluate a prompt and generate a response
 static int generate(LlamaData & llama_data, const std::string & prompt, std::string & response) {
+    const llama_vocab * vocab = llama_model_get_vocab(llama_data.model.get());
+
     std::vector<llama_token> tokens;
-    if (tokenize_prompt(llama_data.model, prompt, tokens) < 0) {
+    if (tokenize_prompt(vocab, prompt, tokens) < 0) {
         return 1;
     }
 
@@ -780,12 +792,12 @@ static int generate(LlamaData & llama_data, const std::string & prompt, std::str
 
         // sample the next token, check is it an end of generation?
         new_token_id = llama_sampler_sample(llama_data.sampler.get(), llama_data.context.get(), -1);
-        if (llama_token_is_eog(llama_data.model.get(), new_token_id)) {
+        if (llama_vocab_is_eog(vocab, new_token_id)) {
             break;
         }
 
         std::string piece;
-        if (convert_token_to_string(llama_data.model, new_token_id, piece)) {
+        if (convert_token_to_string(vocab, new_token_id, piece)) {
             return 1;
         }
 
@@ -800,7 +812,20 @@ static int generate(LlamaData & llama_data, const std::string & prompt, std::str
 
 static int read_user_input(std::string & user) {
     std::getline(std::cin, user);
-    return user.empty();  // Should have data in happy path
+    if (std::cin.eof()) {
+        printf("\n");
+        return 1;
+    }
+
+    if (user == "/bye") {
+        return 1;
+    }
+
+    if (user.empty()) {
+        return 2;
+    }
+
+    return 0;  // Should have data in happy path
 }
 
 // Function to generate a response based on the prompt
@@ -867,7 +892,25 @@ static bool is_stdout_a_terminal() {
 #endif
 }
 
-// Function to tokenize the prompt
+// Function to handle user input
+static int get_user_input(std::string & user_input, const std::string & user) {
+    while (true) {
+        const int ret = handle_user_input(user_input, user);
+        if (ret == 1) {
+            return 1;
+        }
+
+        if (ret == 2) {
+            continue;
+        }
+
+        break;
+    }
+
+    return 0;
+}
+
+// Main chat loop function
 static int chat_loop(LlamaData & llama_data, const std::string & user) {
     int prev_len = 0;
     llama_data.fmtted.resize(llama_n_ctx(llama_data.context.get()));
@@ -875,7 +918,8 @@ static int chat_loop(LlamaData & llama_data, const std::string & user) {
     while (true) {
         // Get user input
         std::string user_input;
-        while (handle_user_input(user_input, user)) {
+        if (get_user_input(user_input, user) == 1) {
+            return 0;
         }
 
         add_message("user", user.empty() ? user_input : user, llama_data);
@@ -916,7 +960,23 @@ static std::string read_pipe_data() {
     return result.str();
 }
 
+static void ctrl_c_handling() {
+#if defined(__unix__) || (defined(__APPLE__) && defined(__MACH__))
+    struct sigaction sigint_action;
+    sigint_action.sa_handler = sigint_handler;
+    sigemptyset(&sigint_action.sa_mask);
+    sigint_action.sa_flags = 0;
+    sigaction(SIGINT, &sigint_action, NULL);
+#elif defined(_WIN32)
+    auto console_ctrl_handler = +[](DWORD ctrl_type) -> BOOL {
+        return (ctrl_type == CTRL_C_EVENT) ? (sigint_handler(SIGINT), true) : false;
+    };
+    SetConsoleCtrlHandler(reinterpret_cast<PHANDLER_ROUTINE>(console_ctrl_handler), true);
+#endif
+}
+
 int main(int argc, const char ** argv) {
+    ctrl_c_handling();
     Opt       opt;
     const int ret = opt.init(argc, argv);
     if (ret == 2) {

examples/save-load-state/save-load-state.cpp

@@ -97,7 +97,7 @@ int main(int argc, char ** argv) {
     printf("\n\n");
 
     // make new context
-    llama_context * ctx2 = llama_new_context_with_model(model, common_context_params_to_llama(params));
+    llama_context * ctx2 = llama_init_from_model(model, common_context_params_to_llama(params));
 
     llama_sampler * smpl2 = llama_sampler_chain_init(sparams);
 
@@ -154,7 +154,7 @@ int main(int argc, char ** argv) {
     }
 
     // make new context
-    llama_context * ctx3 = llama_new_context_with_model(model, common_context_params_to_llama(params));
+    llama_context * ctx3 = llama_init_from_model(model, common_context_params_to_llama(params));
 
     llama_sampler * smpl3 = llama_sampler_chain_init(sparams);
 

examples/server/README.md

@@ -45,10 +45,7 @@ The project is under active development, and we are [looking for feedback and co
 | `-ub, --ubatch-size N` | physical maximum batch size (default: 512)<br/>(env: LLAMA_ARG_UBATCH) |
 | `--keep N` | number of tokens to keep from the initial prompt (default: 0, -1 = all) |
 | `-fa, --flash-attn` | enable Flash Attention (default: disabled)<br/>(env: LLAMA_ARG_FLASH_ATTN) |
-| `-p, --prompt PROMPT` | prompt to start generation with |
 | `--no-perf` | disable internal libllama performance timings (default: false)<br/>(env: LLAMA_ARG_NO_PERF) |
-| `-f, --file FNAME` | a file containing the prompt (default: none) |
-| `-bf, --binary-file FNAME` | binary file containing the prompt (default: none) |
 | `-e, --escape` | process escapes sequences (\n, \r, \t, \', \", \\) (default: true) |
 | `--no-escape` | do not process escape sequences |
 | `--rope-scaling {none,linear,yarn}` | RoPE frequency scaling method, defaults to linear unless specified by the model<br/>(env: LLAMA_ARG_ROPE_SCALING_TYPE) |

examples/server/server.cpp

@@ -98,7 +98,7 @@ struct slot_params {
     int64_t t_max_prompt_ms  = -1; // TODO: implement
     int64_t t_max_predict_ms = -1; // if positive, limit the generation phase to this time limit
 
-    std::vector<common_lora_adapter_info> lora;
+    std::vector<common_adapter_lora_info> lora;
 
     std::vector<std::string> antiprompt;
     std::vector<std::string> response_fields;
@@ -198,15 +198,17 @@ struct server_task {
     bool metrics_reset_bucket = false;
 
     // used by SERVER_TASK_TYPE_SET_LORA
-    std::vector<common_lora_adapter_info> set_lora;
+    std::vector<common_adapter_lora_info> set_lora;
 
     server_task(server_task_type type) : type(type) {}
 
     static slot_params params_from_json_cmpl(
-            const llama_model * model,
             const llama_context * ctx,
             const common_params & params_base,
             const json & data) {
+        const llama_model * model = llama_get_model(ctx);
+        const llama_vocab * vocab = llama_model_get_vocab(model);
+
         slot_params params;
 
         // Sampling parameter defaults are loaded from the global server context (but individual requests can still override them)
@@ -329,7 +331,7 @@ struct server_task {
 
             const auto & logit_bias = data.find("logit_bias");
             if (logit_bias != data.end() && logit_bias->is_array()) {
-                const int n_vocab = llama_n_vocab(model);
+                const int n_vocab = llama_vocab_n_tokens(vocab);
                 for (const auto & el : *logit_bias) {
                     // TODO: we may want to throw errors here, in case "el" is incorrect
                     if (el.is_array() && el.size() == 2) {
@@ -348,7 +350,7 @@ struct server_task {
                                 params.sampling.logit_bias.push_back({tok, bias});
                             }
                         } else if (el[0].is_string()) {
-                            auto toks = common_tokenize(model, el[0].get<std::string>(), false);
+                            auto toks = common_tokenize(vocab, el[0].get<std::string>(), false);
                             for (auto tok : toks) {
                                 params.sampling.logit_bias.push_back({tok, bias});
                             }
@@ -1131,7 +1133,7 @@ struct server_slot {
 
     common_speculative * spec = nullptr;
 
-    std::vector<common_lora_adapter_info> lora;
+    std::vector<common_adapter_lora_info> lora;
 
     // the index relative to completion multi-task request
     size_t index = 0;
@@ -1633,6 +1635,8 @@ struct server_context {
     llama_model * model = nullptr;
     llama_context * ctx = nullptr;
 
+    const llama_vocab * vocab = nullptr;
+
     llama_model * model_dft = nullptr;
 
     llama_context_params cparams_dft;
@@ -1690,10 +1694,12 @@ struct server_context {
             return false;
         }
 
+        vocab = llama_model_get_vocab(model);
+
         n_ctx = llama_n_ctx(ctx);
 
-        add_bos_token = llama_add_bos_token(model);
-        has_eos_token = llama_token_eos(model) != LLAMA_TOKEN_NULL;
+        add_bos_token = llama_vocab_get_add_bos(vocab);
+        has_eos_token = llama_vocab_eos(vocab) != LLAMA_TOKEN_NULL;
 
         if (!params_base.speculative.model.empty()) {
             SRV_INF("loading draft model '%s'\n", params_base.speculative.model.c_str());
@@ -1736,7 +1742,8 @@ struct server_context {
 
     bool validate_builtin_chat_template() const {
         llama_chat_message chat[] = {{"user", "test"}};
-        int32_t chat_res = llama_chat_apply_template(model, nullptr, chat, 1, true, nullptr, 0);
+        const char * tmpl = llama_model_chat_template(model);
+        const int32_t chat_res = llama_chat_apply_template(tmpl, chat, 1, true, nullptr, 0);
         return chat_res > 0;
     }
 
@@ -1756,7 +1763,7 @@ struct server_context {
             if (model_dft) {
                 slot.batch_spec = llama_batch_init(params_base.speculative.n_max + 1, 0, 1);
 
-                slot.ctx_dft = llama_new_context_with_model(model_dft, cparams_dft);
+                slot.ctx_dft = llama_init_from_model(model_dft, cparams_dft);
                 if (slot.ctx_dft == nullptr) {
                     SRV_ERR("%s", "failed to create draft context\n");
                     return;
@@ -1891,7 +1898,7 @@ struct server_context {
         }
 
         if (slot.params.ignore_eos && has_eos_token) {
-            slot.params.sampling.logit_bias.push_back({llama_token_eos(model), -INFINITY});
+            slot.params.sampling.logit_bias.push_back({llama_vocab_eos(vocab), -INFINITY});
         }
 
         {
@@ -2047,14 +2054,14 @@ struct server_context {
                     slot.n_decoded, slot.n_prompt_tokens, slot.n_past, slot.n_ctx);
         }
 
-        if (llama_token_is_eog(model, result.tok)) {
+        if (llama_vocab_is_eog(vocab, result.tok)) {
             slot.stop           = STOP_TYPE_EOS;
             slot.has_next_token = false;
 
             SLT_DBG(slot, "%s", "stopped by EOS\n");
         }
 
-        const auto n_ctx_train = llama_n_ctx_train(model);
+        const auto n_ctx_train = llama_model_n_ctx_train(model);
 
         if (slot.params.n_predict < 1 && slot.n_predict < 1 && slot.n_prompt_tokens + slot.n_decoded >= n_ctx_train) {
             slot.truncated      = true;
@@ -2074,7 +2081,7 @@ struct server_context {
 
     void populate_token_probs(const server_slot & slot, completion_token_output & result, bool post_sampling, bool special, int idx) {
         size_t n_probs = slot.params.sampling.n_probs;
-        size_t n_vocab = llama_n_vocab(llama_get_model(ctx));
+        size_t n_vocab = llama_vocab_n_tokens(vocab);
         if (post_sampling) {
             const auto * cur_p = common_sampler_get_candidates(slot.smpl);
             const size_t max_probs = cur_p->size;
@@ -2225,7 +2232,7 @@ struct server_context {
         res->n_tokens  = slot.n_prompt_tokens;
         res->oaicompat = slot.params.oaicompat;
 
-        const int n_embd = llama_n_embd(model);
+        const int n_embd = llama_model_n_embd(model);
 
         std::vector<float> embd_res(n_embd, 0.0f);
 
@@ -2927,7 +2934,7 @@ struct server_context {
             // make sure we're in the right embedding mode
             llama_set_embeddings(ctx, slot_batched->is_non_causal());
             // apply lora, only need to do it once per batch
-            common_lora_adapters_apply(ctx, slot_batched->lora);
+            common_set_adapter_lora(ctx, slot_batched->lora);
         }
 
         // process the created batch of tokens
@@ -3129,12 +3136,12 @@ struct server_context {
 
     json model_meta() const {
         return json {
-            {"vocab_type",  llama_vocab_type    (model)},
-            {"n_vocab",     llama_n_vocab       (model)},
-            {"n_ctx_train", llama_n_ctx_train   (model)},
-            {"n_embd",      llama_n_embd        (model)},
-            {"n_params",    llama_model_n_params(model)},
-            {"size",        llama_model_size    (model)},
+            {"vocab_type",  llama_vocab_type       (vocab)},
+            {"n_vocab",     llama_vocab_n_tokens   (vocab)},
+            {"n_ctx_train", llama_model_n_ctx_train(model)},
+            {"n_embd",      llama_model_n_embd     (model)},
+            {"n_params",    llama_model_n_params   (model)},
+            {"size",        llama_model_size       (model)},
         };
     }
 };
@@ -3639,7 +3646,7 @@ int main(int argc, char ** argv) {
         std::vector<server_task> tasks;
 
         try {
-            std::vector<llama_tokens> tokenized_prompts = tokenize_input_prompts(ctx_server.ctx, data.at("prompt"), true, true);
+            std::vector<llama_tokens> tokenized_prompts = tokenize_input_prompts(ctx_server.vocab, data.at("prompt"), true, true);
             tasks.reserve(tokenized_prompts.size());
             for (size_t i = 0; i < tokenized_prompts.size(); i++) {
                 server_task task = server_task(type);
@@ -3649,7 +3656,6 @@ int main(int argc, char ** argv) {
 
                 task.prompt_tokens    = std::move(tokenized_prompts[i]);
                 task.params           = server_task::params_from_json_cmpl(
-                                            ctx_server.model,
                                             ctx_server.ctx,
                                             ctx_server.params_base,
                                             data);
@@ -3745,13 +3751,13 @@ int main(int argc, char ** argv) {
     const auto handle_infill = [&ctx_server, &res_error, &handle_completions_impl](const httplib::Request & req, httplib::Response & res) {
         // check model compatibility
         std::string err;
-        if (llama_token_fim_pre(ctx_server.model) == LLAMA_TOKEN_NULL) {
+        if (llama_vocab_fim_pre(ctx_server.vocab) == LLAMA_TOKEN_NULL) {
             err += "prefix token is missing. ";
         }
-        if (llama_token_fim_suf(ctx_server.model) == LLAMA_TOKEN_NULL) {
+        if (llama_vocab_fim_suf(ctx_server.vocab) == LLAMA_TOKEN_NULL) {
             err += "suffix token is missing. ";
         }
-        if (llama_token_fim_mid(ctx_server.model) == LLAMA_TOKEN_NULL) {
+        if (llama_vocab_fim_mid(ctx_server.vocab) == LLAMA_TOKEN_NULL) {
             err += "middle token is missing. ";
         }
         if (!err.empty()) {
@@ -3797,10 +3803,10 @@ int main(int argc, char ** argv) {
         data["input_extra"] = input_extra; // default to empty array if it's not exist
 
         std::string prompt = json_value(data, "prompt", std::string());
-        std::vector<llama_tokens> tokenized_prompts = tokenize_input_prompts(ctx_server.ctx, prompt, false, true);
+        std::vector<llama_tokens> tokenized_prompts = tokenize_input_prompts(ctx_server.vocab, prompt, false, true);
         SRV_DBG("creating infill tasks, n_prompts = %d\n", (int) tokenized_prompts.size());
         data["prompt"] = format_infill(
-            ctx_server.ctx,
+            ctx_server.vocab,
             data.at("input_prefix"),
             data.at("input_suffix"),
             data.at("input_extra"),
@@ -3857,7 +3863,7 @@ int main(int argc, char ** argv) {
             const bool add_special = json_value(body, "add_special", false);
             const bool with_pieces = json_value(body, "with_pieces", false);
 
-            llama_tokens tokens = tokenize_mixed(ctx_server.ctx, body.at("content"), add_special, true);
+            llama_tokens tokens = tokenize_mixed(ctx_server.vocab, body.at("content"), add_special, true);
 
             if (with_pieces) {
                 for (const auto& token : tokens) {
@@ -3933,7 +3939,7 @@ int main(int argc, char ** argv) {
             }
         }
 
-        std::vector<llama_tokens> tokenized_prompts = tokenize_input_prompts(ctx_server.ctx, prompt, true, true);
+        std::vector<llama_tokens> tokenized_prompts = tokenize_input_prompts(ctx_server.vocab, prompt, true, true);
         for (const auto & tokens : tokenized_prompts) {
             // this check is necessary for models that do not add BOS token to the input
             if (tokens.empty()) {
@@ -4033,20 +4039,20 @@ int main(int argc, char ** argv) {
             return;
         }
 
-        llama_tokens tokenized_query = tokenize_input_prompts(ctx_server.ctx, query, /* add_special */ false, true)[0];
+        llama_tokens tokenized_query = tokenize_input_prompts(ctx_server.vocab, query, /* add_special */ false, true)[0];
 
         // create and queue the task
         json responses = json::array();
         bool error = false;
         {
             std::vector<server_task> tasks;
-            std::vector<llama_tokens> tokenized_docs = tokenize_input_prompts(ctx_server.ctx, documents, /* add_special */ false, true);
+            std::vector<llama_tokens> tokenized_docs = tokenize_input_prompts(ctx_server.vocab, documents, /* add_special */ false, true);
             tasks.reserve(tokenized_docs.size());
             for (size_t i = 0; i < tokenized_docs.size(); i++) {
                 server_task task   = server_task(SERVER_TASK_TYPE_RERANK);
                 task.id            = ctx_server.queue_tasks.get_new_id();
                 task.index         = i;
-                task.prompt_tokens = format_rerank(ctx_server.model, tokenized_query, tokenized_docs[i]);
+                task.prompt_tokens = format_rerank(ctx_server.vocab, tokenized_query, tokenized_docs[i]);
                 tasks.push_back(task);
             }
 

examples/server/utils.hpp

@@ -118,7 +118,7 @@ static json json_get_nested_values(const std::vector<std::string> & paths, const
  * - only string, example: "string"
  * - mixed string and tokens, example: [12, 34, "string", 56, 78]
  */
-static llama_tokens tokenize_mixed(const llama_context * ctx, const json & json_prompt, bool add_special, bool parse_special) {
+static llama_tokens tokenize_mixed(const llama_vocab * vocab, const json & json_prompt, bool add_special, bool parse_special) {
     // If `add_bos` is true, we only add BOS, when json_prompt is a string,
     // or the first element of the json_prompt array is a string.
     llama_tokens prompt_tokens;
@@ -131,10 +131,10 @@ static llama_tokens tokenize_mixed(const llama_context * ctx, const json & json_
 
                 llama_tokens p;
                 if (first) {
-                    p = common_tokenize(ctx, s, add_special, parse_special);
+                    p = common_tokenize(vocab, s, add_special, parse_special);
                     first = false;
                 } else {
-                    p = common_tokenize(ctx, s, false, parse_special);
+                    p = common_tokenize(vocab, s, false, parse_special);
                 }
 
                 prompt_tokens.insert(prompt_tokens.end(), p.begin(), p.end());
@@ -148,7 +148,7 @@ static llama_tokens tokenize_mixed(const llama_context * ctx, const json & json_
         }
     } else {
         auto s = json_prompt.template get<std::string>();
-        prompt_tokens = common_tokenize(ctx, s, add_special, parse_special);
+        prompt_tokens = common_tokenize(vocab, s, add_special, parse_special);
     }
 
     return prompt_tokens;
@@ -166,11 +166,11 @@ static llama_tokens tokenize_mixed(const llama_context * ctx, const json & json_
  * - "prompt": [[12, 34, 56], [78, 90, 12]]
  * - "prompt": [[12, 34, "string", 56, 78], [12, 34, 56]]
  */
-static std::vector<llama_tokens> tokenize_input_prompts(llama_context * ctx, const json & json_prompt, bool add_special, bool parse_special) {
+static std::vector<llama_tokens> tokenize_input_prompts(const llama_vocab * vocab, const json & json_prompt, bool add_special, bool parse_special) {
     std::vector<llama_tokens> result;
     if (json_prompt.is_string() || json_is_array_of_mixed_numbers_strings(json_prompt)) {
         // string or mixed
-        result.push_back(tokenize_mixed(ctx, json_prompt, add_special, parse_special));
+        result.push_back(tokenize_mixed(vocab, json_prompt, add_special, parse_special));
     } else if (json_is_array_of_numbers(json_prompt)) {
         // array of tokens
         result.push_back(json_prompt.get<llama_tokens>());
@@ -179,7 +179,7 @@ static std::vector<llama_tokens> tokenize_input_prompts(llama_context * ctx, con
         result.reserve(json_prompt.size());
         for (const auto & p : json_prompt) {
             if (p.is_string() || json_is_array_of_mixed_numbers_strings(p)) {
-                result.push_back(tokenize_mixed(ctx, p, add_special, parse_special));
+                result.push_back(tokenize_mixed(vocab, p, add_special, parse_special));
             } else if (json_is_array_of_numbers(p)) {
                 // array of tokens
                 result.push_back(p.get<llama_tokens>());
@@ -231,21 +231,23 @@ static size_t validate_utf8(const std::string& text) {
 //
 
 // format rerank task: [BOS]query[EOS][SEP]doc[EOS]
-static llama_tokens format_rerank(const struct llama_model * model, const llama_tokens & query, const llama_tokens & doc) {
+static llama_tokens format_rerank(const struct llama_vocab * vocab, const llama_tokens & query, const llama_tokens & doc) {
     llama_tokens result;
+
     result.reserve(doc.size() + query.size() + 4);
-    result.push_back(llama_token_bos(model));
+    result.push_back(llama_vocab_bos(vocab));
     result.insert(result.end(), query.begin(), query.end());
-    result.push_back(llama_token_eos(model));
-    result.push_back(llama_token_sep(model));
+    result.push_back(llama_vocab_eos(vocab));
+    result.push_back(llama_vocab_sep(vocab));
     result.insert(result.end(), doc.begin(), doc.end());
-    result.push_back(llama_token_eos(model));
+    result.push_back(llama_vocab_eos(vocab));
+
     return result;
 }
 
 // format infill task
 static llama_tokens format_infill(
-        const llama_context * ctx,
+        const llama_vocab * vocab,
         const json & input_prefix,
         const json & input_suffix,
         const json & input_extra,
@@ -272,15 +274,14 @@ static llama_tokens format_infill(
     llama_tokens extra_tokens;
     extra_tokens.reserve(n_ctx);
 
-    auto model = llama_get_model(ctx);
-    auto tokens_prefix = tokenize_mixed(ctx, input_prefix, false, false);
-    auto tokens_suffix = tokenize_mixed(ctx, input_suffix, false, false);
+    auto tokens_prefix = tokenize_mixed(vocab, input_prefix, false, false);
+    auto tokens_suffix = tokenize_mixed(vocab, input_suffix, false, false);
 
-    if (llama_token_fim_rep(model) != LLAMA_TOKEN_NULL) {
+    if (llama_vocab_fim_rep(vocab) != LLAMA_TOKEN_NULL) {
         // TODO: make project name an input
-        static const auto k_fim_repo = common_tokenize(ctx, "myproject\n", false, false);
+        static const auto k_fim_repo = common_tokenize(vocab, "myproject\n", false, false);
 
-        extra_tokens.push_back(llama_token_fim_rep(model));
+        extra_tokens.push_back(llama_vocab_fim_rep(vocab));
         extra_tokens.insert(extra_tokens.end(), k_fim_repo.begin(), k_fim_repo.end());
     }
     for (const auto & chunk : input_extra) {
@@ -288,28 +289,28 @@ static llama_tokens format_infill(
         const std::string text     = json_value(chunk, "text",     std::string());
         const std::string filename = json_value(chunk, "filename", std::string("tmp"));
 
-        if (llama_token_fim_sep(model) != LLAMA_TOKEN_NULL) {
-            const auto k_fim_file = common_tokenize(ctx, filename + "\n", false, false);
+        if (llama_vocab_fim_sep(vocab) != LLAMA_TOKEN_NULL) {
+            const auto k_fim_file = common_tokenize(vocab, filename + "\n", false, false);
 
-            extra_tokens.insert(extra_tokens.end(), llama_token_fim_sep(model));
+            extra_tokens.insert(extra_tokens.end(), llama_vocab_fim_sep(vocab));
             extra_tokens.insert(extra_tokens.end(), k_fim_file.begin(), k_fim_file.end());
         } else {
             // chunk separator in binary form to avoid confusing the AI
             static const char k_chunk_prefix_str[] = {0x0a, 0x0a, 0x2d, 0x2d, 0x2d, 0x20, 0x73, 0x6e, 0x69, 0x70, 0x70, 0x65, 0x74, 0x20, 0x2d, 0x2d, 0x2d, 0x0a, 0x0a, 0x00};
-            static const auto k_chunk_prefix_tokens = common_tokenize(ctx, k_chunk_prefix_str, false, false);
+            static const auto k_chunk_prefix_tokens = common_tokenize(vocab, k_chunk_prefix_str, false, false);
 
             extra_tokens.insert(extra_tokens.end(), k_chunk_prefix_tokens.begin(), k_chunk_prefix_tokens.end());
         }
 
-        const auto chunk_tokens = common_tokenize(ctx, text, false, false);
+        const auto chunk_tokens = common_tokenize(vocab, text, false, false);
         extra_tokens.insert(extra_tokens.end(), chunk_tokens.begin(), chunk_tokens.end());
     }
 
-    if (llama_token_fim_sep(model) != LLAMA_TOKEN_NULL) {
+    if (llama_vocab_fim_sep(vocab) != LLAMA_TOKEN_NULL) {
         // TODO: current filename
-        static const auto k_fim_file = common_tokenize(ctx, "filename\n", false, false);
+        static const auto k_fim_file = common_tokenize(vocab, "filename\n", false, false);
 
-        extra_tokens.insert(extra_tokens.end(), llama_token_fim_sep(model));
+        extra_tokens.insert(extra_tokens.end(), llama_vocab_fim_sep(vocab));
         extra_tokens.insert(extra_tokens.end(), k_fim_file.begin(), k_fim_file.end());
     }
 
@@ -325,15 +326,15 @@ static llama_tokens format_infill(
     tokens_prefix.erase(tokens_prefix.begin(), tokens_prefix.begin() + tokens_prefix.size() - n_prefix_take);
     tokens_suffix.resize(n_suffix_take);
 
-    tokens_prefix.insert(tokens_prefix.begin(), llama_token_fim_pre(model));
+    tokens_prefix.insert(tokens_prefix.begin(), llama_vocab_fim_pre(vocab));
     tokens_prefix.insert(tokens_prefix.end(),   tokens_prompt.begin(), tokens_prompt.end());
-    tokens_suffix.insert(tokens_suffix.begin(), llama_token_fim_suf(model));
+    tokens_suffix.insert(tokens_suffix.begin(), llama_vocab_fim_suf(vocab));
 
     auto embd_inp = spm_infill ? tokens_suffix : tokens_prefix;
     auto embd_end = spm_infill ? tokens_prefix : tokens_suffix;
 
-    if (llama_add_bos_token(model)) {
-        embd_inp.insert(embd_inp.begin(), llama_token_bos(model));
+    if (llama_vocab_get_add_bos(vocab)) {
+        embd_inp.insert(embd_inp.begin(), llama_vocab_bos(vocab));
     }
 
     SRV_DBG("extra: n_ctx = %d, n_extra_take = %d, n_extra = %d\n", n_ctx, n_extra_take, (int) extra_tokens.size());
@@ -342,7 +343,7 @@ static llama_tokens format_infill(
     embd_inp.insert(embd_inp.begin(), extra_tokens.end() - n_extra_take, extra_tokens.end());
 
     embd_inp.insert(embd_inp.end(), embd_end.begin(), embd_end.end());
-    embd_inp.push_back(llama_token_fim_mid(model));
+    embd_inp.push_back(llama_vocab_fim_mid(vocab));
 
     return embd_inp;
 }
@@ -764,14 +765,18 @@ static json format_logit_bias(const std::vector<llama_logit_bias> & logit_bias)
     return data;
 }
 
-static std::string safe_json_to_str(json data) {
+static std::string safe_json_to_str(const json & data) {
     return data.dump(-1, ' ', false, json::error_handler_t::replace);
 }
 
 static std::vector<llama_token_data> get_token_probabilities(llama_context * ctx, int idx) {
     std::vector<llama_token_data> cur;
     const auto * logits = llama_get_logits_ith(ctx, idx);
-    const int n_vocab = llama_n_vocab(llama_get_model(ctx));
+
+    const llama_model * model = llama_get_model(ctx);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
+    const int n_vocab = llama_vocab_n_tokens(vocab);
 
     cur.resize(n_vocab);
     for (llama_token token_id = 0; token_id < n_vocab; token_id++) {
@@ -799,8 +804,8 @@ static std::vector<llama_token_data> get_token_probabilities(llama_context * ctx
 }
 
 static bool are_lora_equal(
-        const std::vector<common_lora_adapter_info> & l1,
-        const std::vector<common_lora_adapter_info> & l2) {
+        const std::vector<common_adapter_lora_info> & l1,
+        const std::vector<common_adapter_lora_info> & l2) {
     if (l1.size() != l2.size()) {
         return false;
     }
@@ -814,10 +819,10 @@ static bool are_lora_equal(
 }
 
 // parse lora config from JSON request, returned a copy of lora_base with updated scale
-static std::vector<common_lora_adapter_info> parse_lora_request(
-        const std::vector<common_lora_adapter_info> & lora_base,
+static std::vector<common_adapter_lora_info> parse_lora_request(
+        const std::vector<common_adapter_lora_info> & lora_base,
         const json & data) {
-    std::vector<common_lora_adapter_info> lora(lora_base);
+    std::vector<common_adapter_lora_info> lora(lora_base);
     int max_idx = lora.size();
 
     // clear existing value

examples/server/webui/index.html

@@ -37,7 +37,7 @@
           <div v-for="conv in conversations" :class="{
             'btn btn-ghost justify-start font-normal': true,
             'btn-active': conv.id === viewingConvId,
-          }" @click="setViewingConv(conv.id)">
+          }" @click="setViewingConv(conv.id)" dir="auto">
             <span class="truncate">{{ conv.messages[0].content }}</span>
           </div>
           <div class="text-center text-xs opacity-40 mt-auto mx-4">
@@ -62,53 +62,57 @@
           <!-- action buttons (top right) -->
           <div class="flex items-center">
             <div v-if="messages.length > 0" class="dropdown dropdown-end">
-              <!-- "more" button -->
+              <!-- "..." button -->
               <button tabindex="0" role="button" class="btn m-1" :disabled="isGenerating">
                 <svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" fill="currentColor" class="bi bi-three-dots-vertical" viewBox="0 0 16 16">
                   <path d="M9.5 13a1.5 1.5 0 1 1-3 0 1.5 1.5 0 0 1 3 0m0-5a1.5 1.5 0 1 1-3 0 1.5 1.5 0 0 1 3 0m0-5a1.5 1.5 0 1 1-3 0 1.5 1.5 0 0 1 3 0"/>
                 </svg>
               </button>
-              <!-- "more" dropdown menu -->
+              <!-- "delete" dropdown menu -->
               <ul tabindex="0" class="dropdown-content menu bg-base-100 rounded-box z-[1] w-52 p-2 shadow">
                 <li @click="downloadConv(viewingConvId)"><a>Download</a></li>
                 <li class="text-error" @click="deleteConv(viewingConvId)"><a>Delete</a></li>
               </ul>
             </div>
-            <button class="btn" @click="showConfigDialog = true" :disabled="isGenerating">
-              <!-- settings button -->
-              <svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" fill="currentColor" class="bi bi-gear" viewBox="0 0 16 16">
-                <path d="M8 4.754a3.246 3.246 0 1 0 0 6.492 3.246 3.246 0 0 0 0-6.492M5.754 8a2.246 2.246 0 1 1 4.492 0 2.246 2.246 0 0 1-4.492 0"/>
-                <path d="M9.796 1.343c-.527-1.79-3.065-1.79-3.592 0l-.094.319a.873.873 0 0 1-1.255.52l-.292-.16c-1.64-.892-3.433.902-2.54 2.541l.159.292a.873.873 0 0 1-.52 1.255l-.319.094c-1.79.527-1.79 3.065 0 3.592l.319.094a.873.873 0 0 1 .52 1.255l-.16.292c-.892 1.64.901 3.434 2.541 2.54l.292-.159a.873.873 0 0 1 1.255.52l.094.319c.527 1.79 3.065 1.79 3.592 0l.094-.319a.873.873 0 0 1 1.255-.52l.292.16c1.64.893 3.434-.902 2.54-2.541l-.159-.292a.873.873 0 0 1 .52-1.255l.319-.094c1.79-.527 1.79-3.065 0-3.592l-.319-.094a.873.873 0 0 1-.52-1.255l.16-.292c.893-1.64-.902-3.433-2.541-2.54l-.292.159a.873.873 0 0 1-1.255-.52zm-2.633.283c.246-.835 1.428-.835 1.674 0l.094.319a1.873 1.873 0 0 0 2.693 1.115l.291-.16c.764-.415 1.6.42 1.184 1.185l-.159.292a1.873 1.873 0 0 0 1.116 2.692l.318.094c.835.246.835 1.428 0 1.674l-.319.094a1.873 1.873 0 0 0-1.115 2.693l.16.291c.415.764-.42 1.6-1.185 1.184l-.291-.159a1.873 1.873 0 0 0-2.693 1.116l-.094.318c-.246.835-1.428.835-1.674 0l-.094-.319a1.873 1.873 0 0 0-2.692-1.115l-.292.16c-.764.415-1.6-.42-1.184-1.185l.159-.291A1.873 1.873 0 0 0 1.945 8.93l-.319-.094c-.835-.246-.835-1.428 0-1.674l.319-.094A1.873 1.873 0 0 0 3.06 4.377l-.16-.292c-.415-.764.42-1.6 1.185-1.184l.292.159a1.873 1.873 0 0 0 2.692-1.115z"/>
-              </svg>
-            </button>
+            <div class="tooltip tooltip-bottom" data-tip="Settings">
+              <button class="btn" @click="showConfigDialog = true" :disabled="isGenerating">
+                <!-- settings button -->
+                <svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" fill="currentColor" class="bi bi-gear" viewBox="0 0 16 16">
+                  <path d="M8 4.754a3.246 3.246 0 1 0 0 6.492 3.246 3.246 0 0 0 0-6.492M5.754 8a2.246 2.246 0 1 1 4.492 0 2.246 2.246 0 0 1-4.492 0"/>
+                  <path d="M9.796 1.343c-.527-1.79-3.065-1.79-3.592 0l-.094.319a.873.873 0 0 1-1.255.52l-.292-.16c-1.64-.892-3.433.902-2.54 2.541l.159.292a.873.873 0 0 1-.52 1.255l-.319.094c-1.79.527-1.79 3.065 0 3.592l.319.094a.873.873 0 0 1 .52 1.255l-.16.292c-.892 1.64.901 3.434 2.541 2.54l.292-.159a.873.873 0 0 1 1.255.52l.094.319c.527 1.79 3.065 1.79 3.592 0l.094-.319a.873.873 0 0 1 1.255-.52l.292.16c1.64.893 3.434-.902 2.54-2.541l-.159-.292a.873.873 0 0 1 .52-1.255l.319-.094c1.79-.527 1.79-3.065 0-3.592l-.319-.094a.873.873 0 0 1-.52-1.255l.16-.292c.893-1.64-.902-3.433-2.541-2.54l-.292.159a.873.873 0 0 1-1.255-.52zm-2.633.283c.246-.835 1.428-.835 1.674 0l.094.319a1.873 1.873 0 0 0 2.693 1.115l.291-.16c.764-.415 1.6.42 1.184 1.185l-.159.292a1.873 1.873 0 0 0 1.116 2.692l.318.094c.835.246.835 1.428 0 1.674l-.319.094a1.873 1.873 0 0 0-1.115 2.693l.16.291c.415.764-.42 1.6-1.185 1.184l-.291-.159a1.873 1.873 0 0 0-2.693 1.116l-.094.318c-.246.835-1.428.835-1.674 0l-.094-.319a1.873 1.873 0 0 0-2.692-1.115l-.292.16c-.764.415-1.6-.42-1.184-1.185l.159-.291A1.873 1.873 0 0 0 1.945 8.93l-.319-.094c-.835-.246-.835-1.428 0-1.674l.319-.094A1.873 1.873 0 0 0 3.06 4.377l-.16-.292c-.415-.764.42-1.6 1.185-1.184l.292.159a1.873 1.873 0 0 0 2.692-1.115z"/>
+                </svg>
+              </button>
+            </div>
 
             <!-- theme controller is copied from https://daisyui.com/components/theme-controller/ -->
-            <div class="dropdown dropdown-end dropdown-bottom">
-              <div tabindex="0" role="button" class="btn m-1">
-                <svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" fill="currentColor" class="bi bi-palette2" viewBox="0 0 16 16">
-                  <path d="M0 .5A.5.5 0 0 1 .5 0h5a.5.5 0 0 1 .5.5v5.277l4.147-4.131a.5.5 0 0 1 .707 0l3.535 3.536a.5.5 0 0 1 0 .708L10.261 10H15.5a.5.5 0 0 1 .5.5v5a.5.5 0 0 1-.5.5H3a3 3 0 0 1-2.121-.879A3 3 0 0 1 0 13.044m6-.21 7.328-7.3-2.829-2.828L6 7.188zM4.5 13a1.5 1.5 0 1 0-3 0 1.5 1.5 0 0 0 3 0M15 15v-4H9.258l-4.015 4zM0 .5v12.495zm0 12.495V13z"/>
-                </svg>
+            <div class="tooltip tooltip-bottom" data-tip="Themes">
+              <div class="dropdown dropdown-end dropdown-bottom">
+                <div tabindex="0" role="button" class="btn m-1">
+                  <svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" fill="currentColor" class="bi bi-palette2" viewBox="0 0 16 16">
+                    <path d="M0 .5A.5.5 0 0 1 .5 0h5a.5.5 0 0 1 .5.5v5.277l4.147-4.131a.5.5 0 0 1 .707 0l3.535 3.536a.5.5 0 0 1 0 .708L10.261 10H15.5a.5.5 0 0 1 .5.5v5a.5.5 0 0 1-.5.5H3a3 3 0 0 1-2.121-.879A3 3 0 0 1 0 13.044m6-.21 7.328-7.3-2.829-2.828L6 7.188zM4.5 13a1.5 1.5 0 1 0-3 0 1.5 1.5 0 0 0 3 0M15 15v-4H9.258l-4.015 4zM0 .5v12.495zm0 12.495V13z"/>
+                  </svg>
+                </div>
+                <ul tabindex="0" class="dropdown-content bg-base-300 rounded-box z-[1] w-52 p-2 shadow-2xl h-80 overflow-y-auto">
+                  <li>
+                    <button
+                      class="btn btn-sm btn-block btn-ghost justify-start"
+                      :class="{ 'btn-active': selectedTheme === 'auto' }"
+                      @click="setSelectedTheme('auto')">
+                      auto
+                    </button>
+                  </li>
+                  <li v-for="theme in themes">
+                    <input
+                      type="radio"
+                      name="theme-dropdown"
+                      class="theme-controller btn btn-sm btn-block btn-ghost justify-start"
+                      :aria-label="theme"
+                      :value="theme"
+                      :checked="selectedTheme === theme"
+                      @click="setSelectedTheme(theme)" />
+                  </li>
+                </ul>
               </div>
-              <ul tabindex="0" class="dropdown-content bg-base-300 rounded-box z-[1] w-52 p-2 shadow-2xl h-80 overflow-y-auto">
-                <li>
-                  <button
-                    class="btn btn-sm btn-block btn-ghost justify-start"
-                    :class="{ 'btn-active': selectedTheme === 'auto' }"
-                    @click="setSelectedTheme('auto')">
-                    auto
-                  </button>
-                </li>
-                <li v-for="theme in themes">
-                  <input
-                    type="radio"
-                    name="theme-dropdown"
-                    class="theme-controller btn btn-sm btn-block btn-ghost justify-start"
-                    :aria-label="theme"
-                    :value="theme"
-                    :checked="selectedTheme === theme"
-                    @click="setSelectedTheme(theme)" />
-                </li>
-              </ul>
             </div>
           </div>
         </div>
@@ -152,6 +156,7 @@
             @keydown.enter.shift.exact.prevent="inputMsg += '\n'"
             :disabled="isGenerating"
             id="msg-input"
+            dir="auto"
           ></textarea>
           <button v-if="!isGenerating" class="btn btn-primary ml-2" @click="sendMessage" :disabled="inputMsg.length === 0">Send</button>
           <button v-else class="btn btn-neutral ml-2" @click="stopGeneration">Stop</button>
@@ -244,6 +249,7 @@
         <!-- textarea for editing message -->
         <template v-if="editingContent !== null">
           <textarea
+            dir="auto"
             class="textarea textarea-bordered bg-base-100 text-base-content w-[calc(90vw-8em)] lg:w-96"
             v-model="editingContent"></textarea>
           <br/>
@@ -254,7 +260,9 @@
           <!-- show loading dots for pending message -->
           <span v-if="msg.content === null" class="loading loading-dots loading-md"></span>
           <!-- render message as markdown -->
-          <vue-markdown v-else :source="msg.content"></vue-markdown>
+          <div v-else dir="auto">
+            <vue-markdown :source="msg.content"></vue-markdown>
+          </div>
           <!-- render timings if enabled -->
           <div class="dropdown dropdown-hover dropdown-top mt-2" v-if="timings && config.showTokensPerSecond">
             <div tabindex="0" role="button" class="cursor-pointer font-semibold text-sm opacity-60">Speed: {{ timings.predicted_per_second.toFixed(1) }} t/s</div>

examples/server/webui/src/main.js

@@ -111,12 +111,12 @@ const VueMarkdown = defineComponent(
       highlight: function (str, lang) { // Add highlight.js
         if (lang && hljs.getLanguage(lang)) {
           try {
-            return '<pre><code class="hljs">' +
+            return '<pre dir="auto"><code class="hljs">' +
                    hljs.highlight(str, { language: lang, ignoreIllegals: true }).value +
                    '</code></pre>';
           } catch (__) {}
         }
-        return '<pre><code class="hljs">' + md.value.utils.escapeHtml(str) + '</code></pre>';
+        return '<pre dir="auto"><code class="hljs">' + md.value.utils.escapeHtml(str) + '</code></pre>';
       }
     }));
     // support latex with double dollar sign and square brackets

examples/simple-chat/simple-chat.cpp

@@ -75,12 +75,14 @@ int main(int argc, char ** argv) {
         return 1;
     }
 
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     // initialize the context
     llama_context_params ctx_params = llama_context_default_params();
     ctx_params.n_ctx = n_ctx;
     ctx_params.n_batch = n_ctx;
 
-    llama_context * ctx = llama_new_context_with_model(model, ctx_params);
+    llama_context * ctx = llama_init_from_model(model, ctx_params);
     if (!ctx) {
         fprintf(stderr , "%s: error: failed to create the llama_context\n" , __func__);
         return 1;
@@ -97,9 +99,9 @@ int main(int argc, char ** argv) {
         std::string response;
 
         // tokenize the prompt
-        const int n_prompt_tokens = -llama_tokenize(model, prompt.c_str(), prompt.size(), NULL, 0, true, true);
+        const int n_prompt_tokens = -llama_tokenize(vocab, prompt.c_str(), prompt.size(), NULL, 0, true, true);
         std::vector<llama_token> prompt_tokens(n_prompt_tokens);
-        if (llama_tokenize(model, prompt.c_str(), prompt.size(), prompt_tokens.data(), prompt_tokens.size(), llama_get_kv_cache_used_cells(ctx) == 0, true) < 0) {
+        if (llama_tokenize(vocab, prompt.c_str(), prompt.size(), prompt_tokens.data(), prompt_tokens.size(), llama_get_kv_cache_used_cells(ctx) == 0, true) < 0) {
             GGML_ABORT("failed to tokenize the prompt\n");
         }
 
@@ -124,13 +126,13 @@ int main(int argc, char ** argv) {
             new_token_id = llama_sampler_sample(smpl, ctx, -1);
 
             // is it an end of generation?
-            if (llama_token_is_eog(model, new_token_id)) {
+            if (llama_vocab_is_eog(vocab, new_token_id)) {
                 break;
             }
 
             // convert the token to a string, print it and add it to the response
             char buf[256];
-            int n = llama_token_to_piece(model, new_token_id, buf, sizeof(buf), 0, true);
+            int n = llama_token_to_piece(vocab, new_token_id, buf, sizeof(buf), 0, true);
             if (n < 0) {
                 GGML_ABORT("failed to convert token to piece\n");
             }
@@ -159,12 +161,14 @@ int main(int argc, char ** argv) {
             break;
         }
 
+        const char * tmpl = llama_model_chat_template(model);
+
         // add the user input to the message list and format it
         messages.push_back({"user", strdup(user.c_str())});
-        int new_len = llama_chat_apply_template(model, nullptr, messages.data(), messages.size(), true, formatted.data(), formatted.size());
+        int new_len = llama_chat_apply_template(tmpl, messages.data(), messages.size(), true, formatted.data(), formatted.size());
         if (new_len > (int)formatted.size()) {
             formatted.resize(new_len);
-            new_len = llama_chat_apply_template(model, nullptr, messages.data(), messages.size(), true, formatted.data(), formatted.size());
+            new_len = llama_chat_apply_template(tmpl, messages.data(), messages.size(), true, formatted.data(), formatted.size());
         }
         if (new_len < 0) {
             fprintf(stderr, "failed to apply the chat template\n");
@@ -181,7 +185,7 @@ int main(int argc, char ** argv) {
 
         // add the response to the messages
         messages.push_back({"assistant", strdup(response.c_str())});
-        prev_len = llama_chat_apply_template(model, nullptr, messages.data(), messages.size(), false, nullptr, 0);
+        prev_len = llama_chat_apply_template(tmpl, messages.data(), messages.size(), false, nullptr, 0);
         if (prev_len < 0) {
             fprintf(stderr, "failed to apply the chat template\n");
             return 1;

examples/simple/simple.cpp

@@ -84,6 +84,7 @@ int main(int argc, char ** argv) {
     model_params.n_gpu_layers = ngl;
 
     llama_model * model = llama_model_load_from_file(model_path.c_str(), model_params);
+    const llama_vocab * vocab = llama_model_get_vocab(model);
 
     if (model == NULL) {
         fprintf(stderr , "%s: error: unable to load model\n" , __func__);
@@ -93,11 +94,11 @@ int main(int argc, char ** argv) {
     // tokenize the prompt
 
     // find the number of tokens in the prompt
-    const int n_prompt = -llama_tokenize(model, prompt.c_str(), prompt.size(), NULL, 0, true, true);
+    const int n_prompt = -llama_tokenize(vocab, prompt.c_str(), prompt.size(), NULL, 0, true, true);
 
     // allocate space for the tokens and tokenize the prompt
     std::vector<llama_token> prompt_tokens(n_prompt);
-    if (llama_tokenize(model, prompt.c_str(), prompt.size(), prompt_tokens.data(), prompt_tokens.size(), true, true) < 0) {
+    if (llama_tokenize(vocab, prompt.c_str(), prompt.size(), prompt_tokens.data(), prompt_tokens.size(), true, true) < 0) {
         fprintf(stderr, "%s: error: failed to tokenize the prompt\n", __func__);
         return 1;
     }
@@ -112,7 +113,7 @@ int main(int argc, char ** argv) {
     // enable performance counters
     ctx_params.no_perf = false;
 
-    llama_context * ctx = llama_new_context_with_model(model, ctx_params);
+    llama_context * ctx = llama_init_from_model(model, ctx_params);
 
     if (ctx == NULL) {
         fprintf(stderr , "%s: error: failed to create the llama_context\n" , __func__);
@@ -131,7 +132,7 @@ int main(int argc, char ** argv) {
 
     for (auto id : prompt_tokens) {
         char buf[128];
-        int n = llama_token_to_piece(model, id, buf, sizeof(buf), 0, true);
+        int n = llama_token_to_piece(vocab, id, buf, sizeof(buf), 0, true);
         if (n < 0) {
             fprintf(stderr, "%s: error: failed to convert token to piece\n", __func__);
             return 1;
@@ -164,12 +165,12 @@ int main(int argc, char ** argv) {
             new_token_id = llama_sampler_sample(smpl, ctx, -1);
 
             // is it an end of generation?
-            if (llama_token_is_eog(model, new_token_id)) {
+            if (llama_vocab_is_eog(vocab, new_token_id)) {
                 break;
             }
 
             char buf[128];
-            int n = llama_token_to_piece(model, new_token_id, buf, sizeof(buf), 0, true);
+            int n = llama_token_to_piece(vocab, new_token_id, buf, sizeof(buf), 0, true);
             if (n < 0) {
                 fprintf(stderr, "%s: error: failed to convert token to piece\n", __func__);
                 return 1;

examples/speculative-simple/speculative-simple.cpp

@@ -45,6 +45,8 @@ int main(int argc, char ** argv) {
     model_tgt = llama_init_tgt.model.get();
     ctx_tgt   = llama_init_tgt.context.get();
 
+    const llama_vocab * vocab = llama_model_get_vocab(model_tgt);
+
     // load the draft model
     params.devices      = params.speculative.devices;
     params.model        = params.speculative.model;
@@ -196,7 +198,7 @@ int main(int argc, char ** argv) {
 
             id_last = ids[i];
 
-            if (llama_token_is_eog(model_tgt, id_last)) {
+            if (llama_vocab_is_eog(vocab, id_last)) {
                 has_eos = true;
                 break;
             }

examples/speculative/speculative.cpp

@@ -90,10 +90,13 @@ int main(int argc, char ** argv) {
     model_dft = llama_init_dft.model.get();
     ctx_dft   = llama_init_dft.context.get();
 
-    const bool vocab_type_tgt = llama_vocab_type(model_tgt);
+    const llama_vocab * vocab_tgt = llama_model_get_vocab(model_tgt);
+    const llama_vocab * vocab_dft = llama_model_get_vocab(model_dft);
+
+    const bool vocab_type_tgt = llama_vocab_type(vocab_tgt);
     LOG_DBG("vocab_type tgt: %d\n", vocab_type_tgt);
 
-    const bool vocab_type_dft = llama_vocab_type(model_dft);
+    const bool vocab_type_dft = llama_vocab_type(vocab_dft);
     LOG_DBG("vocab_type dft: %d\n", vocab_type_dft);
 
     if (vocab_type_tgt != vocab_type_dft) {
@@ -103,18 +106,18 @@ int main(int argc, char ** argv) {
     }
 
     if (
-        llama_add_bos_token(model_tgt) != llama_add_bos_token(model_dft) ||
-        llama_add_eos_token(model_tgt) != llama_add_eos_token(model_dft) ||
-        llama_token_bos(model_tgt) != llama_token_bos(model_dft) ||
-        llama_token_eos(model_tgt) != llama_token_eos(model_dft)
+        llama_vocab_get_add_bos(vocab_tgt) != llama_vocab_get_add_bos(vocab_dft) ||
+        llama_vocab_get_add_eos(vocab_tgt) != llama_vocab_get_add_eos(vocab_dft) ||
+        llama_vocab_bos(vocab_tgt) != llama_vocab_bos(vocab_dft) ||
+        llama_vocab_eos(vocab_tgt) != llama_vocab_eos(vocab_dft)
     ) {
         LOG_ERR("%s: draft model special tokens must match target model to use speculation\n", __func__);
         return 1;
     }
 
     {
-        const int n_vocab_tgt = llama_n_vocab(model_tgt);
-        const int n_vocab_dft = llama_n_vocab(model_dft);
+        const int n_vocab_tgt = llama_vocab_n_tokens(vocab_tgt);
+        const int n_vocab_dft = llama_vocab_n_tokens(vocab_dft);
         const int vocab_diff  = n_vocab_tgt > n_vocab_dft
             ? n_vocab_tgt - n_vocab_dft
             : n_vocab_dft - n_vocab_tgt;
@@ -122,13 +125,13 @@ int main(int argc, char ** argv) {
         if (vocab_diff > SPEC_VOCAB_MAX_SIZE_DIFFERENCE) {
             LOG_ERR("%s: draft model vocab must closely match target model to use speculation but ", __func__);
             LOG_ERR("target vocab size %d does not match draft vocab size %d - difference %d, max allowed %d\n",
-                    n_vocab_tgt, llama_n_vocab(model_dft), vocab_diff, SPEC_VOCAB_MAX_SIZE_DIFFERENCE);
+                    n_vocab_tgt, llama_vocab_n_tokens(vocab_dft), vocab_diff, SPEC_VOCAB_MAX_SIZE_DIFFERENCE);
             return 1;
         }
 
         for (int i = SPEC_VOCAB_CHECK_START_TOKEN_ID; i < std::min(n_vocab_tgt, n_vocab_dft); ++i) {
-            const char * token_text_tgt = llama_token_get_text(model_tgt, i);
-            const char * token_text_dft = llama_token_get_text(model_dft, i);
+            const char * token_text_tgt = llama_vocab_get_text(vocab_tgt, i);
+            const char * token_text_dft = llama_vocab_get_text(vocab_dft, i);
             if (std::strcmp(token_text_tgt, token_text_dft) != 0) {
                 LOG_ERR("%s: draft model vocab must match target model to use speculation but ", __func__);
                 LOG_ERR("token %d content differs - target '%s', draft '%s'\n", i,
@@ -170,7 +173,7 @@ int main(int argc, char ** argv) {
     const auto t_enc_end = ggml_time_us();
 
     // the 2 models should have the same vocab
-    //GGML_ASSERT(n_vocab == llama_n_vocab(model_dft));
+    //GGML_ASSERT(n_vocab == llama_vocab_n_tokens(model_dft));
 
     // how many tokens to draft each time
     int n_draft = params.speculative.n_max;
@@ -386,7 +389,7 @@ int main(int argc, char ** argv) {
                     }
                 }
 
-                if (llama_token_is_eog(model_tgt, token_id)) {
+                if (llama_vocab_is_eog(vocab_tgt, token_id)) {
                     has_eos = true;
                 }
                 ++n_predict;

examples/tokenize/tokenize.cpp

@@ -344,8 +344,10 @@ int main(int raw_argc, char ** raw_argv) {
         return 1;
     }
 
+    const llama_vocab * vocab = llama_model_get_vocab(model);
+
     llama_context_params ctx_params = llama_context_default_params();
-    llama_context * ctx = llama_new_context_with_model(model, ctx_params);
+    llama_context * ctx = llama_init_from_model(model, ctx_params);
     if (!ctx) {
         fprintf(stderr, "Error: could not create context.\n");
         return 1;
@@ -365,7 +367,7 @@ int main(int raw_argc, char ** raw_argv) {
         prompt = stdin_buffer.str();
     }
 
-    const bool model_wants_add_bos = llama_add_bos_token(model);
+    const bool model_wants_add_bos = llama_vocab_get_add_bos(vocab);
     const bool add_bos = model_wants_add_bos && !no_bos;
     const bool parse_special = !no_parse_special;
     const bool escape = !no_escape;
@@ -375,7 +377,7 @@ int main(int raw_argc, char ** raw_argv) {
     }
 
     std::vector<llama_token> tokens;
-    tokens = common_tokenize(model, prompt, add_bos, parse_special);
+    tokens = common_tokenize(vocab, prompt, add_bos, parse_special);
 
     if (printing_ids) {
         printf("[");

examples/tts/README.md

@@ -0,0 +1,117 @@
+# llama.cpp/example/tts
+This example demonstrates the Text To Speech feature. It uses a
+[model](https://www.outeai.com/blog/outetts-0.2-500m) from
+[outeai](https://www.outeai.com/).
+
+## Quickstart
+If you have built llama.cpp with `-DLLAMA_CURL=ON` you can simply run the
+following command and the required models will be downloaded automatically:
+```console
+$ build/bin/llama-tts --tts-oute-default -p "Hello world" && aplay output.wav
+```
+For details about the models and how to convert them to the required format
+see the following sections.
+
+### Model conversion
+Checkout or download the model that contains the LLM model:
+```console
+$ pushd models
+$ git clone --branch main --single-branch --depth 1 https://huggingface.co/OuteAI/OuteTTS-0.2-500M
+$ cd OuteTTS-0.2-500M && git lfs install && git lfs pull
+$ popd
+```
+Convert the model to .gguf format:
+```console
+(venv) python convert_hf_to_gguf.py models/OuteTTS-0.2-500M \
+    --outfile models/outetts-0.2-0.5B-f16.gguf --outtype f16
+```
+The generated model will be `models/outetts-0.2-0.5B-f16.gguf`.
+
+We can optionally quantize this to Q8_0 using the following command:
+```console
+$ build/bin/llama-quantize models/outetts-0.2-0.5B-f16.gguf \
+    models/outetts-0.2-0.5B-q8_0.gguf q8_0
+```
+The quantized model will be `models/outetts-0.2-0.5B-q8_0.gguf`.
+
+Next we do something simlar for the audio decoder. First download or checkout
+the model for the voice decoder:
+```console
+$ pushd models
+$ git clone --branch main --single-branch --depth 1 https://huggingface.co/novateur/WavTokenizer-large-speech-75token
+$ cd WavTokenizer-large-speech-75token && git lfs install && git lfs pull
+$ popd
+```
+This model file is PyTorch checkpoint (.ckpt) and we first need to convert it to
+huggingface format:
+```console
+(venv) python examples/tts/convert_pt_to_hf.py \
+    models/WavTokenizer-large-speech-75token/wavtokenizer_large_speech_320_24k.ckpt
+...
+Model has been successfully converted and saved to models/WavTokenizer-large-speech-75token/model.safetensors
+Metadata has been saved to models/WavTokenizer-large-speech-75token/index.json
+Config has been saved to models/WavTokenizer-large-speech-75tokenconfig.json
+```
+Then we can convert the huggingface format to gguf:
+```console
+(venv) python convert_hf_to_gguf.py models/WavTokenizer-large-speech-75token \
+    --outfile models/wavtokenizer-large-75-f16.gguf --outtype f16
+...
+INFO:hf-to-gguf:Model successfully exported to models/wavtokenizer-large-75-f16.gguf
+```
+
+### Running the example
+
+With both of the models generated, the LLM model and the voice decoder model,
+we can run the example:
+```console
+$ build/bin/llama-tts -m  ./models/outetts-0.2-0.5B-q8_0.gguf \
+    -mv ./models/wavtokenizer-large-75-f16.gguf \
+    -p "Hello world"
+...
+main: audio written to file 'output.wav'
+```
+The output.wav file will contain the audio of the prompt. This can be heard
+by playing the file with a media player. On Linux the following command will
+play the audio:
+```console
+$ aplay output.wav
+```
+
+### Running the example with llama-server
+Running this example with `llama-server` is also possible and requires two
+server instances to be started. One will serve the LLM model and the other
+will serve the voice decoder model.
+
+The LLM model server can be started with the following command:
+```console
+$ ./build/bin/llama-server -m ./models/outetts-0.2-0.5B-q8_0.gguf --port 8020
+```
+
+And the voice decoder model server can be started using:
+```console
+./build/bin/llama-server -m ./models/wavtokenizer-large-75-f16.gguf --port 8021 --embeddings --pooling none
+```
+
+Then we can run [tts-outetts.py](tts-outetts.py) to generate the audio.
+
+First create a virtual environment for python and install the required
+dependencies (this in only required to be done once):
+```console
+$ python3 -m venv venv
+$ source venv/bin/activate
+(venv) pip install requests numpy
+```
+
+And then run the python script using:
+```conole
+(venv) python ./examples/tts/tts-outetts.py http://localhost:8020 http://localhost:8021 "Hello world"
+spectrogram generated: n_codes: 90, n_embd: 1282
+converting to audio ...
+audio generated: 28800 samples
+audio written to file "output.wav"
+```
+And to play the audio we can again use aplay or any other media player:
+```console
+$ aplay output.wav
+```

examples/tts/tts-outetts.py

@@ -3,6 +3,121 @@ import sys
 #import struct
 import requests
 import re
+import struct
+import numpy as np
+from concurrent.futures import ThreadPoolExecutor
+
+
+def fill_hann_window(size, periodic=True):
+    if periodic:
+        return np.hanning(size + 1)[:-1]
+    return np.hanning(size)
+
+
+def irfft(n_fft, complex_input):
+    return np.fft.irfft(complex_input, n=n_fft)
+
+
+def fold(buffer, n_out, n_win, n_hop, n_pad):
+    result = np.zeros(n_out)
+    n_frames = len(buffer) // n_win
+
+    for i in range(n_frames):
+        start = i * n_hop
+        end = start + n_win
+        result[start:end] += buffer[i * n_win:(i + 1) * n_win]
+
+    return result[n_pad:-n_pad] if n_pad > 0 else result
+
+
+def process_frame(args):
+    l, n_fft, ST, hann = args
+    frame = irfft(n_fft, ST[l])
+    frame = frame * hann
+    hann2 = hann * hann
+    return frame, hann2
+
+
+def embd_to_audio(embd, n_codes, n_embd, n_thread=4):
+    embd = np.asarray(embd, dtype=np.float32).reshape(n_codes, n_embd)
+
+    n_fft = 1280
+    n_hop = 320
+    n_win = 1280
+    n_pad = (n_win - n_hop) // 2
+    n_out = (n_codes - 1) * n_hop + n_win
+
+    hann = fill_hann_window(n_fft, True)
+
+    E = np.zeros((n_embd, n_codes), dtype=np.float32)
+    for l in range(n_codes):
+        for k in range(n_embd):
+            E[k, l] = embd[l, k]
+
+    half_embd = n_embd // 2
+    S = np.zeros((n_codes, half_embd + 1), dtype=np.complex64)
+
+    for k in range(half_embd):
+        for l in range(n_codes):
+            mag = E[k, l]
+            phi = E[k + half_embd, l]
+
+            mag = np.clip(np.exp(mag), 0, 1e2)
+            S[l, k] = mag * np.exp(1j * phi)
+
+    res = np.zeros(n_codes * n_fft)
+    hann2_buffer = np.zeros(n_codes * n_fft)
+
+    with ThreadPoolExecutor(max_workers=n_thread) as executor:
+        args = [(l, n_fft, S, hann) for l in range(n_codes)]
+        results = list(executor.map(process_frame, args))
+
+        for l, (frame, hann2) in enumerate(results):
+            res[l*n_fft:(l+1)*n_fft] = frame
+            hann2_buffer[l*n_fft:(l+1)*n_fft] = hann2
+
+    audio = fold(res, n_out, n_win, n_hop, n_pad)
+    env = fold(hann2_buffer, n_out, n_win, n_hop, n_pad)
+
+    mask = env > 1e-10
+    audio[mask] /= env[mask]
+
+    return audio
+
+
+def save_wav(filename, audio_data, sample_rate):
+    num_channels = 1
+    bits_per_sample = 16
+    bytes_per_sample = bits_per_sample // 8
+    data_size = len(audio_data) * bytes_per_sample
+    byte_rate = sample_rate * num_channels * bytes_per_sample
+    block_align = num_channels * bytes_per_sample
+    chunk_size = 36 + data_size  # 36 = size of header minus first 8 bytes
+
+    header = struct.pack(
+        '<4sI4s4sIHHIIHH4sI',
+        b'RIFF',
+        chunk_size,
+        b'WAVE',
+        b'fmt ',
+        16,                # fmt chunk size
+        1,                 # audio format (PCM)
+        num_channels,
+        sample_rate,
+        byte_rate,
+        block_align,
+        bits_per_sample,
+        b'data',
+        data_size
+    )
+
+    audio_data = np.clip(audio_data * 32767, -32768, 32767)
+    pcm_data = audio_data.astype(np.int16)
+
+    with open(filename, 'wb') as f:
+        f.write(header)
+        f.write(pcm_data.tobytes())
+
 
 def process_text(text: str):
     text = re.sub(r'\d+(\.\d+)?', lambda x: x.group(), text.lower()) # TODO this needs to be fixed
@@ -170,6 +285,15 @@ n_embd = len(embd[0])
 print('spectrogram generated: n_codes: %d, n_embd: %d' % (n_codes, n_embd))
 
 # post-process the spectrogram to convert to audio
-# TODO: see the tts.cpp:embd_to_audio() and implement it in Python
 print('converting to audio ...')
-print('TODO: see the tts.cpp:embd_to_audio() and implement it in Python')
+audio = embd_to_audio(embd, n_codes, n_embd)
+print('audio generated: %d samples' % len(audio))
+
+filename = "output.wav"
+sample_rate = 24000 # sampling rate
+
+# zero out first 0.25 seconds
+audio[:24000 // 4] = 0.0
+
+save_wav(filename, audio, sample_rate)
+print('audio written to file "%s"' % filename)

examples/tts/tts.cpp

@@ -414,15 +414,15 @@ static void prompt_add(llama_tokens & prompt, const llama_tokens & tokens) {
     prompt.insert(prompt.end(), tokens.begin(), tokens.end());
 }
 
-static void prompt_add(llama_tokens & prompt, const llama_model * model, const std::string & txt, bool add_special, bool parse_special) {
-    auto tmp = common_tokenize(model, txt, add_special, parse_special);
+static void prompt_add(llama_tokens & prompt, const llama_vocab * vocab, const std::string & txt, bool add_special, bool parse_special) {
+    auto tmp = common_tokenize(vocab, txt, add_special, parse_special);
     prompt_add(prompt, tmp);
 }
 
-static void prompt_init(llama_tokens & prompt, const llama_model * model) {
+static void prompt_init(llama_tokens & prompt, const llama_vocab * vocab) {
     prompt.clear();
 
-    prompt_add(prompt, model, "<|im_start|>\n", true, true);
+    prompt_add(prompt, vocab, "<|im_start|>\n", true, true);
 }
 
 int main(int argc, char ** argv) {
@@ -462,6 +462,8 @@ int main(int argc, char ** argv) {
     model_ttc = llama_init_ttc.model.get();
     ctx_ttc   = llama_init_ttc.context.get();
 
+    const llama_vocab * vocab = llama_model_get_vocab(model_ttc);
+
     // TODO: refactor in a common struct
     params.model     = params.vocoder.model;
     params.model_url = params.vocoder.model_url;
@@ -499,9 +501,9 @@ int main(int argc, char ** argv) {
 
         std::vector<llama_token> prompt_inp;
 
-        prompt_init(prompt_inp, model_ttc);
+        prompt_init(prompt_inp, vocab);
 
-        prompt_add(prompt_inp, model_ttc, "<|text_start|>the<|text_sep|>overall<|text_sep|>package<|text_sep|>from<|text_sep|>just<|text_sep|>two<|text_sep|>people<|text_sep|>is<|text_sep|>pretty<|text_sep|>remarkable<|text_sep|>sure<|text_sep|>i<|text_sep|>have<|text_sep|>some<|text_sep|>critiques<|text_sep|>about<|text_sep|>some<|text_sep|>of<|text_sep|>the<|text_sep|>gameplay<|text_sep|>aspects<|text_sep|>but<|text_sep|>its<|text_sep|>still<|text_sep|>really<|text_sep|>enjoyable<|text_sep|>and<|text_sep|>it<|text_sep|>looks<|text_sep|>lovely<|text_sep|>", false, true);
+        prompt_add(prompt_inp, vocab, "<|text_start|>the<|text_sep|>overall<|text_sep|>package<|text_sep|>from<|text_sep|>just<|text_sep|>two<|text_sep|>people<|text_sep|>is<|text_sep|>pretty<|text_sep|>remarkable<|text_sep|>sure<|text_sep|>i<|text_sep|>have<|text_sep|>some<|text_sep|>critiques<|text_sep|>about<|text_sep|>some<|text_sep|>of<|text_sep|>the<|text_sep|>gameplay<|text_sep|>aspects<|text_sep|>but<|text_sep|>its<|text_sep|>still<|text_sep|>really<|text_sep|>enjoyable<|text_sep|>and<|text_sep|>it<|text_sep|>looks<|text_sep|>lovely<|text_sep|>", false, true);
 
         // convert the input text into the necessary format expected by OuteTTS
         {
@@ -509,10 +511,10 @@ int main(int argc, char ** argv) {
 
             LOG_INF("%s: prompt: '%s'\n", __func__, prompt_clean.c_str());
 
-            prompt_add(prompt_inp, model_ttc, prompt_clean, false, true);
+            prompt_add(prompt_inp, vocab, prompt_clean, false, true);
         }
 
-        prompt_add(prompt_inp, model_ttc, "<|text_end|>\n", false, true);
+        prompt_add(prompt_inp, vocab, "<|text_end|>\n", false, true);
 
         // disabled to save time on tokenizing each time
         // TODO: load voices from the json files
@@ -549,7 +551,7 @@ it<|t_0.09|><|code_start|><|848|><|1366|><|395|><|1601|><|1513|><|593|><|1302|><
 looks<|t_0.27|><|code_start|><|1281|><|1266|><|1755|><|572|><|248|><|1751|><|1257|><|695|><|1380|><|457|><|659|><|585|><|1315|><|1105|><|1776|><|736|><|24|><|736|><|654|><|1027|><|code_end|>
 lovely<|t_0.56|><|code_start|><|634|><|596|><|1766|><|1556|><|1306|><|1285|><|1481|><|1721|><|1123|><|438|><|1246|><|1251|><|795|><|659|><|1381|><|1658|><|217|><|1772|><|562|><|952|><|107|><|1129|><|1112|><|467|><|550|><|1079|><|840|><|1615|><|1469|><|1380|><|168|><|917|><|836|><|1827|><|437|><|583|><|67|><|595|><|1087|><|1646|><|1493|><|1677|><|code_end|>)";
 
-        auto tmp = common_tokenize(model_ttc, voice_data, false, true);
+        auto tmp = common_tokenize(vocab, voice_data, false, true);
         printf("\n\n");
         for (int i = 0; i < tmp.size(); ++i) {
             printf("%d, ", tmp[i]);
@@ -735,9 +737,9 @@ lovely<|t_0.56|><|code_start|><|634|><|596|><|1766|><|1556|><|1306|><|1285|><|14
                 const auto * cands = common_sampler_get_candidates(smpl[i]);
 
                 // is it an end of generation? -> mark the stream as finished
-                if (llama_token_is_eog(model_ttc, new_token_id) || n_decode == n_predict) {
+                if (llama_vocab_is_eog(vocab, new_token_id) || n_decode == n_predict) {
                     std::string reason;
-                    if (llama_token_is_eog(model_ttc, new_token_id)) {
+                    if (llama_vocab_is_eog(vocab, new_token_id)) {
                         reason = "eos";
                     } else {
                         reason = "n_predict";
@@ -873,7 +875,7 @@ lovely<|t_0.56|><|code_start|><|634|><|596|><|1766|><|1556|><|1306|><|1285|><|14
 
 #if 1
     // spectral operations
-    const int n_embd = llama_n_embd(model_cts);
+    const int n_embd = llama_model_n_embd(model_cts);
     const float * embd = llama_get_embeddings(ctx_cts);
 
     auto audio = embd_to_audio(embd, n_codes, n_embd, params.cpuparams.n_threads);

ggml/CMakeLists.txt

@@ -185,6 +185,9 @@ option(GGML_OPENCL_PROFILING                "ggml: use OpenCL profiling (increas
 option(GGML_OPENCL_EMBED_KERNELS            "ggml: embed kernels"                             ON)
 option(GGML_OPENCL_USE_ADRENO_KERNELS       "ggml: use optimized kernels for Adreno"          ON)
 
+# toolchain for vulkan-shaders-gen
+set   (GGML_VULKAN_SHADERS_GEN_TOOLCHAIN "" CACHE FILEPATH "ggml: toolchain file for vulkan-shaders-gen")
+
 # extra artifacts
 option(GGML_BUILD_TESTS    "ggml: build tests"    ${GGML_STANDALONE})
 option(GGML_BUILD_EXAMPLES "ggml: build examples" ${GGML_STANDALONE})
@@ -243,7 +246,8 @@ set(GGML_PUBLIC_HEADERS
     include/ggml-metal.h
     include/ggml-rpc.h
     include/ggml-sycl.h
-    include/ggml-vulkan.h)
+    include/ggml-vulkan.h
+    include/gguf.h)
 
 set_target_properties(ggml PROPERTIES PUBLIC_HEADER "${GGML_PUBLIC_HEADERS}")
 #if (GGML_METAL)

ggml/include/ggml-cpp.h

@@ -7,6 +7,7 @@
 #include "ggml.h"
 #include "ggml-alloc.h"
 #include "ggml-backend.h"
+#include "gguf.h"
 #include <memory>
 
 // Smart pointers for ggml types

ggml/include/ggml.h

@@ -241,12 +241,6 @@
 #define GGML_ROPE_TYPE_MROPE  8
 #define GGML_ROPE_TYPE_VISION 24
 
-#define GGUF_MAGIC "GGUF"
-
-#define GGUF_VERSION 3
-
-#define GGUF_DEFAULT_ALIGNMENT 32
-
 #define GGML_UNUSED(x) (void)(x)
 
 #define GGML_PAD(x, n) (((x) + (n) - 1) & ~((n) - 1))
@@ -403,12 +397,6 @@ extern "C" {
         GGML_PREC_F32,
     };
 
-    enum ggml_backend_type {
-        GGML_BACKEND_TYPE_CPU = 0,
-        GGML_BACKEND_TYPE_GPU = 10,
-        GGML_BACKEND_TYPE_GPU_SPLIT = 20,
-    };
-
     // model file types
     enum ggml_ftype {
         GGML_FTYPE_UNKNOWN        = -1,
@@ -513,6 +501,7 @@ extern "C" {
         GGML_OP_GET_REL_POS,
         GGML_OP_ADD_REL_POS,
         GGML_OP_RWKV_WKV6,
+        GGML_OP_GATED_LINEAR_ATTN,
 
         GGML_OP_UNARY,
 
@@ -587,8 +576,6 @@ extern "C" {
     struct ggml_tensor {
         enum ggml_type type;
 
-        GGML_DEPRECATED(enum ggml_backend_type backend, "use the buffer type to find the storage location of the tensor");
-
         struct ggml_backend_buffer * buffer;
 
         int64_t ne[GGML_MAX_DIMS]; // number of elements
@@ -1513,7 +1500,7 @@ extern "C" {
 
     // rotary position embedding backward, i.e compute dx from dy
     // a - dy
-    GGML_API struct ggml_tensor * ggml_rope_back(
+    GGML_API struct ggml_tensor * ggml_rope_ext_back(
             struct ggml_context * ctx,
             struct ggml_tensor  * a, // gradients of ggml_rope result
             struct ggml_tensor  * b, // positions
@@ -1528,6 +1515,23 @@ extern "C" {
             float                 beta_fast,
             float                 beta_slow);
 
+    GGML_API struct ggml_tensor * ggml_rope_multi_back(
+            struct ggml_context * ctx,
+            struct ggml_tensor  * a,
+            struct ggml_tensor  * b,
+            struct ggml_tensor  * c,
+            int                   n_dims,
+            int                   sections[4],
+            int                   mode,
+            int                   n_ctx_orig,
+            float                 freq_base,
+            float                 freq_scale,
+            float                 ext_factor,
+            float                 attn_factor,
+            float                 beta_fast,
+            float                 beta_slow);
+
+
     // clamp
     // in-place, returns view(a)
     GGML_API struct ggml_tensor * ggml_clamp(
@@ -1873,6 +1877,15 @@ extern "C" {
             struct ggml_tensor  * td,
             struct ggml_tensor  * state);
 
+    GGML_API struct ggml_tensor * ggml_gated_linear_attn(
+            struct ggml_context * ctx,
+            struct ggml_tensor  * k,
+            struct ggml_tensor  * v,
+            struct ggml_tensor  * q,
+            struct ggml_tensor  * g,
+            struct ggml_tensor  * state,
+            float scale);
+
     // custom operators
 
     typedef void (*ggml_unary_op_f32_t) (const int, float *, const float *);
@@ -2111,132 +2124,6 @@ extern "C" {
                    int64_t   n_per_row,
                const float * imatrix);
 
-    //
-    // gguf
-    //
-
-    enum gguf_type {
-        GGUF_TYPE_UINT8   = 0,
-        GGUF_TYPE_INT8    = 1,
-        GGUF_TYPE_UINT16  = 2,
-        GGUF_TYPE_INT16   = 3,
-        GGUF_TYPE_UINT32  = 4,
-        GGUF_TYPE_INT32   = 5,
-        GGUF_TYPE_FLOAT32 = 6,
-        GGUF_TYPE_BOOL    = 7,
-        GGUF_TYPE_STRING  = 8,
-        GGUF_TYPE_ARRAY   = 9,
-        GGUF_TYPE_UINT64  = 10,
-        GGUF_TYPE_INT64   = 11,
-        GGUF_TYPE_FLOAT64 = 12,
-        GGUF_TYPE_COUNT,       // marks the end of the enum
-    };
-
-    struct gguf_context;
-
-    struct gguf_init_params {
-        bool no_alloc;
-
-        // if not NULL, create a ggml_context and allocate the tensor data in it
-        struct ggml_context ** ctx;
-    };
-
-    GGML_API struct gguf_context * gguf_init_empty(void);
-    GGML_API struct gguf_context * gguf_init_from_file(const char * fname, struct gguf_init_params params);
-    //GGML_API struct gguf_context * gguf_init_from_buffer(..);
-
-    GGML_API void gguf_free(struct gguf_context * ctx);
-
-    GGML_API const char * gguf_type_name(enum gguf_type type);
-
-    GGML_API int    gguf_get_version    (const struct gguf_context * ctx);
-    GGML_API size_t gguf_get_alignment  (const struct gguf_context * ctx);
-    GGML_API size_t gguf_get_data_offset(const struct gguf_context * ctx);
-    GGML_API void * gguf_get_data       (const struct gguf_context * ctx);
-
-    GGML_API int          gguf_get_n_kv(const struct gguf_context * ctx);
-    GGML_API int          gguf_find_key(const struct gguf_context * ctx, const char * key);
-    GGML_API const char * gguf_get_key (const struct gguf_context * ctx, int key_id);
-
-    GGML_API enum gguf_type gguf_get_kv_type (const struct gguf_context * ctx, int key_id);
-    GGML_API enum gguf_type gguf_get_arr_type(const struct gguf_context * ctx, int key_id);
-
-    // will abort if the wrong type is used for the key
-    GGML_API uint8_t      gguf_get_val_u8  (const struct gguf_context * ctx, int key_id);
-    GGML_API int8_t       gguf_get_val_i8  (const struct gguf_context * ctx, int key_id);
-    GGML_API uint16_t     gguf_get_val_u16 (const struct gguf_context * ctx, int key_id);
-    GGML_API int16_t      gguf_get_val_i16 (const struct gguf_context * ctx, int key_id);
-    GGML_API uint32_t     gguf_get_val_u32 (const struct gguf_context * ctx, int key_id);
-    GGML_API int32_t      gguf_get_val_i32 (const struct gguf_context * ctx, int key_id);
-    GGML_API float        gguf_get_val_f32 (const struct gguf_context * ctx, int key_id);
-    GGML_API uint64_t     gguf_get_val_u64 (const struct gguf_context * ctx, int key_id);
-    GGML_API int64_t      gguf_get_val_i64 (const struct gguf_context * ctx, int key_id);
-    GGML_API double       gguf_get_val_f64 (const struct gguf_context * ctx, int key_id);
-    GGML_API bool         gguf_get_val_bool(const struct gguf_context * ctx, int key_id);
-    GGML_API const char * gguf_get_val_str (const struct gguf_context * ctx, int key_id);
-    GGML_API const void * gguf_get_val_data(const struct gguf_context * ctx, int key_id);
-    GGML_API int          gguf_get_arr_n   (const struct gguf_context * ctx, int key_id);
-    GGML_API const void * gguf_get_arr_data(const struct gguf_context * ctx, int key_id);
-    GGML_API const char * gguf_get_arr_str (const struct gguf_context * ctx, int key_id, int i);
-
-    GGML_API int            gguf_get_n_tensors    (const struct gguf_context * ctx);
-    GGML_API int            gguf_find_tensor      (const struct gguf_context * ctx, const char * name);
-    GGML_API size_t         gguf_get_tensor_offset(const struct gguf_context * ctx, int i);
-    GGML_API char *         gguf_get_tensor_name  (const struct gguf_context * ctx, int i);
-    GGML_API enum ggml_type gguf_get_tensor_type  (const struct gguf_context * ctx, int i);
-
-    // removes key if it exists
-    GGML_API void gguf_remove_key(struct gguf_context * ctx, const char * key);
-
-    // overrides existing values or adds a new one
-    GGML_API void gguf_set_val_u8  (struct gguf_context * ctx, const char * key, uint8_t  val);
-    GGML_API void gguf_set_val_i8  (struct gguf_context * ctx, const char * key, int8_t   val);
-    GGML_API void gguf_set_val_u16 (struct gguf_context * ctx, const char * key, uint16_t val);
-    GGML_API void gguf_set_val_i16 (struct gguf_context * ctx, const char * key, int16_t  val);
-    GGML_API void gguf_set_val_u32 (struct gguf_context * ctx, const char * key, uint32_t val);
-    GGML_API void gguf_set_val_i32 (struct gguf_context * ctx, const char * key, int32_t  val);
-    GGML_API void gguf_set_val_f32 (struct gguf_context * ctx, const char * key, float    val);
-    GGML_API void gguf_set_val_u64 (struct gguf_context * ctx, const char * key, uint64_t val);
-    GGML_API void gguf_set_val_i64 (struct gguf_context * ctx, const char * key, int64_t  val);
-    GGML_API void gguf_set_val_f64 (struct gguf_context * ctx, const char * key, double   val);
-    GGML_API void gguf_set_val_bool(struct gguf_context * ctx, const char * key, bool     val);
-    GGML_API void gguf_set_val_str (struct gguf_context * ctx, const char * key, const char * val);
-    GGML_API void gguf_set_arr_data(struct gguf_context * ctx, const char * key, enum gguf_type type, const void * data, int n);
-    GGML_API void gguf_set_arr_str (struct gguf_context * ctx, const char * key, const char ** data, int n);
-
-    // set or add KV pairs from another context
-    GGML_API void gguf_set_kv(struct gguf_context * ctx, struct gguf_context * src);
-
-    // manage tensor info
-    GGML_API void gguf_add_tensor(struct gguf_context * ctx, const struct ggml_tensor * tensor);
-    GGML_API void gguf_set_tensor_type(struct gguf_context * ctx, const char * name, enum ggml_type type);
-    GGML_API void gguf_set_tensor_data(struct gguf_context * ctx, const char * name, const void * data, size_t size);
-
-    // writing gguf files can be done in 2 ways:
-    //
-    // - write the entire gguf_context to a binary file in a single pass:
-    //
-    //   gguf_write_to_file(ctx, fname);
-    //
-    // - first prepare a file with a placeholder for the meta data, write the tensor data, then write the meta data:
-    //
-    //   FILE * f = fopen(fname, "wb");
-    //   fseek(f, gguf_get_meta_size(ctx), SEEK_SET);
-    //   fwrite(f, ...);
-    //   void * data = gguf_meta_get_meta_data(ctx);
-    //   fseek(f, 0, SEEK_SET);
-    //   fwrite(f, data, gguf_get_meta_size(ctx));
-    //   free(data);
-    //   fclose(f);
-    //
-
-    // write the entire context to a binary file
-    GGML_API void gguf_write_to_file(const struct gguf_context * ctx, const char * fname, bool only_meta);
-
-    // get the size in bytes of the meta data (header, kv pairs, tensor info) including padding
-    GGML_API size_t gguf_get_meta_size(const struct gguf_context * ctx);
-    GGML_API void   gguf_get_meta_data(const struct gguf_context * ctx, void * data);
-
 #ifdef __cplusplus
     // restrict not standard in C++
 #    if defined(__GNUC__)

ggml/include/gguf.h

@@ -0,0 +1,202 @@
+// This file contains functionality related to "GGUF" files, the binary file format used by ggml.
+// GGUF files have the following structure:
+//
+// 1. File magic "GGUF" (4 bytes).
+// 2. File version (uint32_t).
+// 3. Number of ggml tensors in file (int64_t).
+// 4. Number of key-value-pairs in file (int64_t).
+// 5. For each KV pair:
+//   1. The key (string).
+//   2. The value type (gguf_type).
+//   3a. If the value type is GGUF_TYPE_ARRAY:
+//     1. The type of the array (gguf_type).
+//     2. The number of elements in the array (uint64_t).
+//     3. The binary representation of each element in the array.
+//   3b. Otherwise:
+//     1. The binary representation of the value.
+// 6. For each ggml tensor:
+//   1. The tensor name (string).
+//   2. The number of dimensions of the tensor (uint32_t).
+//   3. For each dimension:
+//     1. The size of the tensor in the dimension (int64_t).
+//   4. The tensor data type (ggml_type).
+//   5. The tensor data offset in the tensor data binary blob (uint64_t).
+// 7. The tensor data binary blob (optional, aligned).
+//
+// Strings are serialized as the string length (uint64_t) followed by the C string without the null terminator.
+// All enums are stored as int32_t.
+// All bool values are stored as int8_t.
+// If the special key "general.alignment" (uint32_t) is defined it is used for alignment,
+//   otherwise GGUF_DEFAULT_ALIGNMENT is used.
+//
+// Module maintainer: Johannes Gäßler (@JohannesGaessler, johannesg@5d6.de)
+
+#pragma once
+
+#include "ggml.h"
+
+#include <stdbool.h>
+#include <stdint.h>
+
+#define GGUF_MAGIC   "GGUF"
+#define GGUF_VERSION 3
+
+#define GGUF_KEY_GENERAL_ALIGNMENT "general.alignment"
+
+#define GGUF_DEFAULT_ALIGNMENT 32
+
+#ifdef  __cplusplus
+extern "C" {
+#endif
+
+    // types that can be stored as GGUF KV data
+    enum gguf_type {
+        GGUF_TYPE_UINT8   = 0,
+        GGUF_TYPE_INT8    = 1,
+        GGUF_TYPE_UINT16  = 2,
+        GGUF_TYPE_INT16   = 3,
+        GGUF_TYPE_UINT32  = 4,
+        GGUF_TYPE_INT32   = 5,
+        GGUF_TYPE_FLOAT32 = 6,
+        GGUF_TYPE_BOOL    = 7,
+        GGUF_TYPE_STRING  = 8,
+        GGUF_TYPE_ARRAY   = 9,
+        GGUF_TYPE_UINT64  = 10,
+        GGUF_TYPE_INT64   = 11,
+        GGUF_TYPE_FLOAT64 = 12,
+        GGUF_TYPE_COUNT,       // marks the end of the enum
+    };
+
+    struct gguf_context;
+
+    struct gguf_init_params {
+        bool no_alloc;
+
+        // if not NULL, create a ggml_context and allocate the tensor data in it
+        struct ggml_context ** ctx;
+    };
+
+    GGML_API struct gguf_context * gguf_init_empty(void);
+    GGML_API struct gguf_context * gguf_init_from_file(const char * fname, struct gguf_init_params params);
+    //GGML_API struct gguf_context * gguf_init_from_buffer(..);
+
+    GGML_API void gguf_free(struct gguf_context * ctx);
+
+    GGML_API const char * gguf_type_name(enum gguf_type type);
+
+    GGML_API uint32_t gguf_get_version    (const struct gguf_context * ctx);
+    GGML_API size_t   gguf_get_alignment  (const struct gguf_context * ctx);
+    GGML_API size_t   gguf_get_data_offset(const struct gguf_context * ctx);
+
+    GGML_API int64_t      gguf_get_n_kv(const struct gguf_context * ctx);
+    GGML_API int64_t      gguf_find_key(const struct gguf_context * ctx, const char * key); // returns -1 if key is not found
+    GGML_API const char * gguf_get_key (const struct gguf_context * ctx, int64_t key_id);
+
+    GGML_API enum gguf_type gguf_get_kv_type (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API enum gguf_type gguf_get_arr_type(const struct gguf_context * ctx, int64_t key_id);
+
+    // will abort if the wrong type is used for the key
+    GGML_API uint8_t      gguf_get_val_u8  (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API int8_t       gguf_get_val_i8  (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API uint16_t     gguf_get_val_u16 (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API int16_t      gguf_get_val_i16 (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API uint32_t     gguf_get_val_u32 (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API int32_t      gguf_get_val_i32 (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API float        gguf_get_val_f32 (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API uint64_t     gguf_get_val_u64 (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API int64_t      gguf_get_val_i64 (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API double       gguf_get_val_f64 (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API bool         gguf_get_val_bool(const struct gguf_context * ctx, int64_t key_id);
+    GGML_API const char * gguf_get_val_str (const struct gguf_context * ctx, int64_t key_id);
+    GGML_API const void * gguf_get_val_data(const struct gguf_context * ctx, int64_t key_id);
+    GGML_API size_t       gguf_get_arr_n   (const struct gguf_context * ctx, int64_t key_id);
+
+    // get raw pointer to the first element of the array with the given key_id
+    // for bool arrays, note that they are always stored as int8 on all platforms (usually this makes no difference)
+    GGML_API const void * gguf_get_arr_data(const struct gguf_context * ctx, int64_t key_id);
+
+    // get ith C string from array with given key_id
+    GGML_API const char * gguf_get_arr_str (const struct gguf_context * ctx, int64_t key_id, size_t i);
+
+    GGML_API int64_t        gguf_get_n_tensors    (const struct gguf_context * ctx);
+    GGML_API int64_t        gguf_find_tensor      (const struct gguf_context * ctx, const char * name); // returns -1 if the tensor is not found
+    GGML_API size_t         gguf_get_tensor_offset(const struct gguf_context * ctx, int64_t tensor_id);
+    GGML_API const char *   gguf_get_tensor_name  (const struct gguf_context * ctx, int64_t tensor_id);
+    GGML_API enum ggml_type gguf_get_tensor_type  (const struct gguf_context * ctx, int64_t tensor_id);
+    GGML_API size_t         gguf_get_tensor_size  (const struct gguf_context * ctx, int64_t tensor_id);
+
+    // removes key if it exists, returns id that the key had prior to removal (-1 if it didn't exist)
+    GGML_API int64_t gguf_remove_key(struct gguf_context * ctx, const char * key);
+
+    // overrides an existing KV pair or adds a new one, the new KV pair is always at the back
+    GGML_API void gguf_set_val_u8  (struct gguf_context * ctx, const char * key, uint8_t      val);
+    GGML_API void gguf_set_val_i8  (struct gguf_context * ctx, const char * key, int8_t       val);
+    GGML_API void gguf_set_val_u16 (struct gguf_context * ctx, const char * key, uint16_t     val);
+    GGML_API void gguf_set_val_i16 (struct gguf_context * ctx, const char * key, int16_t      val);
+    GGML_API void gguf_set_val_u32 (struct gguf_context * ctx, const char * key, uint32_t     val);
+    GGML_API void gguf_set_val_i32 (struct gguf_context * ctx, const char * key, int32_t      val);
+    GGML_API void gguf_set_val_f32 (struct gguf_context * ctx, const char * key, float        val);
+    GGML_API void gguf_set_val_u64 (struct gguf_context * ctx, const char * key, uint64_t     val);
+    GGML_API void gguf_set_val_i64 (struct gguf_context * ctx, const char * key, int64_t      val);
+    GGML_API void gguf_set_val_f64 (struct gguf_context * ctx, const char * key, double       val);
+    GGML_API void gguf_set_val_bool(struct gguf_context * ctx, const char * key, bool         val);
+    GGML_API void gguf_set_val_str (struct gguf_context * ctx, const char * key, const char * val);
+
+    // creates a new array with n elements of the given type and copies the corresponding number of bytes from data
+    GGML_API void gguf_set_arr_data(struct gguf_context * ctx, const char * key, enum gguf_type type, const void * data, size_t n);
+
+    // creates a new array with n strings and copies the corresponding strings from data
+    GGML_API void gguf_set_arr_str (struct gguf_context * ctx, const char * key, const char ** data, size_t n);
+
+    // set or add KV pairs from another context
+    GGML_API void gguf_set_kv(struct gguf_context * ctx, const struct gguf_context * src);
+
+    // add tensor to GGUF context, tensor name must be unique
+    GGML_API void gguf_add_tensor(struct gguf_context * ctx, const struct ggml_tensor * tensor);
+
+    // after changing a tensor's type, the offsets of all tensors with higher indices are immediately recalculated
+    //   in such a way that the tensor data remains as one contiguous block (except for padding)
+    GGML_API void gguf_set_tensor_type(struct gguf_context * ctx, const char * name, enum ggml_type type);
+
+    // assumes that at least gguf_get_tensor_size bytes can be read from data
+    GGML_API void gguf_set_tensor_data(struct gguf_context * ctx, const char * name, const void * data);
+
+    // writing gguf files can be done in 3 ways:
+    //
+    // - write the entire gguf_context to a binary file in a single pass:
+    //
+    //   gguf_write_to_file(ctx, fname, /*only_meta =*/ false);
+    //
+    // - write only the meta data to a file, then re-open the file and append the tensor data:
+    //
+    //   gguf_write_to_file(ctx, fname, /*only_meta =*/ true);
+    //   FILE * f = fopen(fname, "ab");
+    //   fwrite(f, ...); // write tensor data
+    //   fclose(f);
+    //
+    // - first prepare a file with a placeholder for the meta data, write the tensor data, then write the meta data:
+    //
+    //   FILE * f = fopen(fname, "wb");
+    //   const size_t size_meta = gguf_get_meta_size(ctx);
+    //   fseek(f, size_meta, SEEK_SET);
+    //   fwrite(f, ...); // write tensor data
+    //   void * data = malloc(size_meta);
+    //   gguf_get_meta_data(ctx, data);
+    //   rewind(f);
+    //   fwrite(data, 1, data, f);
+    //   free(data);
+    //   fclose(f);
+    //
+
+    // write the entire context to a binary file
+    GGML_API bool gguf_write_to_file(const struct gguf_context * ctx, const char * fname, bool only_meta);
+
+    // get the size in bytes of the meta data (header, kv pairs, tensor info) including padding
+    GGML_API size_t gguf_get_meta_size(const struct gguf_context * ctx);
+
+    // writes the meta data to pointer "data"
+    GGML_API void   gguf_get_meta_data(const struct gguf_context * ctx, void * data);
+
+#ifdef  __cplusplus
+}
+#endif

ggml/src/CMakeLists.txt

@@ -208,6 +208,7 @@ add_library(ggml-base
             ../include/ggml-backend.h
             ../include/ggml-cpp.h
             ../include/ggml-opt.h
+            ../include/gguf.h
             ggml.c
             ggml-alloc.c
             ggml-backend.cpp
@@ -215,7 +216,8 @@ add_library(ggml-base
             ggml-threading.cpp
             ggml-threading.h
             ggml-quants.c
-            ggml-quants.h)
+            ggml-quants.h
+            gguf.cpp)
 
 target_include_directories(ggml-base PRIVATE .)
 

ggml/src/ggml-backend-reg.cpp

@@ -574,4 +574,9 @@ void ggml_backend_load_all_from_path(const char * dir_path) {
     ggml_backend_load_best("opencl", silent, dir_path);
     ggml_backend_load_best("musa", silent, dir_path);
     ggml_backend_load_best("cpu", silent, dir_path);
+    // check the environment variable GGML_BACKEND_PATH to load an out-of-tree backend
+    const char * backend_path = std::getenv("GGML_BACKEND_PATH");
+    if (backend_path) {
+        ggml_backend_load(backend_path);
+    }
 }

ggml/src/ggml-backend.cpp

@@ -764,7 +764,7 @@ static int ggml_backend_sched_backend_id_from_cur(ggml_backend_sched_t sched, st
         if (tensor->op != GGML_OP_ROPE && src->buffer != NULL && src->buffer->usage == GGML_BACKEND_BUFFER_USAGE_WEIGHTS) {
             int src_backend_id = ggml_backend_sched_backend_from_buffer(sched, src, tensor);
             // check if a backend with higher prio wants to offload the op
-            if (src_backend_id == sched->n_backends - 1) {
+            if (src_backend_id == sched->n_backends - 1 && ggml_backend_buffer_is_host(src->buffer)) {
                 for (int b = 0; b < src_backend_id; b++) {
                     if (ggml_backend_supports_op(sched->backends[b], tensor) && ggml_backend_offload_op(sched->backends[b], tensor)) {
                         SET_CAUSE(tensor, "1.off");

ggml/src/ggml-cpu/ggml-cpu-aarch64.cpp

@@ -4169,6 +4169,8 @@ static ggml_backend_buffer_t ggml_backend_cpu_aarch64_buffer_type_alloc_buffer(g
     buffer->buft              = buft;
     buffer->iface.init_tensor = ggml_backend_cpu_aarch64_buffer_init_tensor;
     buffer->iface.set_tensor  = ggml_backend_cpu_aarch64_buffer_set_tensor;
+    buffer->iface.get_tensor  = nullptr;
+    buffer->iface.cpy_tensor  = nullptr;
     return buffer;
 }
 

ggml/src/ggml-cpu/ggml-cpu.c

@@ -11803,9 +11803,9 @@ static void ggml_compute_forward_add_rel_pos(
 static void ggml_compute_forward_rwkv_wkv6_f32(
         const struct ggml_compute_params * params,
         struct ggml_tensor * dst) {
-    const int64_t T = dst->src[1]->ne[3];
+    const int64_t T = dst->src[1]->ne[2];
     const int64_t C = dst->ne[0];
-    const int64_t HEADS = dst->src[1]->ne[2];
+    const int64_t HEADS = dst->src[1]->ne[1];
     const int64_t n_seqs = dst->src[5]->ne[1];
     const int64_t head_size = C / HEADS;
 
@@ -12000,6 +12000,197 @@ static void ggml_compute_forward_rwkv_wkv6(
     }
 }
 
+// ggml_compute_forward_gla
+
+static void ggml_compute_forward_gla_f32(
+        const struct ggml_compute_params * params,
+        struct ggml_tensor * dst) {
+    const int64_t T = dst->src[1]->ne[2];
+    const int64_t C = dst->ne[0];
+    const int64_t HEADS = dst->src[1]->ne[1];
+    const int64_t n_seqs = dst->src[4]->ne[1];
+    const int64_t head_size = C / HEADS;
+    const float scale = ggml_get_op_params_f32(dst, 0);
+
+    float * dst_data = (float *) dst->data;
+    float * state = ((float *) dst->data) + C * T;
+
+    const int ith = params->ith;
+    const int nth = params->nth;
+
+    if (ith >= HEADS) {
+        return;
+    }
+
+    const int h_start = (HEADS * ith) / nth;
+    const int h_end = ((HEADS * (ith + 1)) / nth < HEADS) ?
+                (HEADS * (ith + 1)) / nth : HEADS;
+
+    float * k = (float *) dst->src[0]->data;
+    float * v = (float *) dst->src[1]->data;
+    float * q = (float *) dst->src[2]->data;
+    float * g = (float *) dst->src[3]->data;
+
+    size_t t_stride = HEADS * head_size; // Same to C
+
+    size_t h_stride = C / HEADS;
+    GGML_ASSERT(C % HEADS == 0); // C must be divisible by HEADS
+    size_t h_stride_2d = head_size * head_size;
+
+    if (ith == 0) {
+        memset(dst_data, 0, T * C * sizeof(float));
+    }
+    ggml_barrier(params->threadpool);
+
+
+    #if defined(__AVX__) && !defined(__AVX512F__)
+        #define GGML_F32X GGML_F32x8
+        #define GGML_F32X_SET1 GGML_F32x8_SET1
+        #define GGML_F32X_LOAD GGML_F32x8_LOAD
+        #define GGML_F32X_STORE GGML_F32x8_STORE
+        #define GGML_F32X_MUL GGML_F32x8_MUL
+        #define GGML_F32X_FMA GGML_F32x8_FMA
+        #define GLA_VECTOR_SIZE 8
+    #elif defined(__AVX512F__)
+        #define GGML_F32X GGML_F32x16
+        #define GGML_F32X_SET1 GGML_F32x16_SET1
+        #define GGML_F32X_LOAD GGML_F32x16_LOAD
+        #define GGML_F32X_STORE GGML_F32x16_STORE
+        #define GGML_F32X_MUL GGML_F32x16_MUL
+        #define GGML_F32X_FMA GGML_F32x16_FMA
+        #define GLA_VECTOR_SIZE 16
+    #elif defined(__ARM_NEON) && defined(__aarch64__)
+        #define GGML_F32X GGML_F32x4
+        #define GGML_F32X_SET1 GGML_F32x4_SET1
+        #define GGML_F32X_LOAD GGML_F32x4_LOAD
+        #define GGML_F32X_STORE GGML_F32x4_STORE
+        #define GGML_F32X_MUL GGML_F32x4_MUL
+        #define GGML_F32X_FMA GGML_F32x4_FMA
+        #define GLA_VECTOR_SIZE 4
+    #endif
+
+    #ifdef GLA_VECTOR_SIZE
+        const int64_t vec_count = head_size / GLA_VECTOR_SIZE;
+
+        for (int64_t t = 0; t < T; t++) {
+            size_t t_offset = t * t_stride;
+            size_t state_offset = head_size * C * (t / (T / n_seqs));
+            float * state_cur = state + state_offset;
+            float * state_prev = t % (T / n_seqs) ? state_cur : (float*)dst->src[4]->data + state_offset;
+
+            for (int64_t h = h_start; h < h_end; h++) {
+                size_t h_offset = h * h_stride;
+                size_t t_h_offset = t_offset + h_offset;
+                size_t h_2d_offset = h * h_stride_2d;
+
+                for (int64_t i = 0; i < head_size; i++) {
+                    size_t t_h_i_offset = t_h_offset + i;
+                    size_t h_2d_i_offset = h_2d_offset + i * h_stride;
+
+                    float k_val = k[t_h_i_offset];
+                    float q_val = q[t_h_i_offset] * scale;
+                    float g_val = g[t_h_i_offset];
+
+                    // Broadcast scalar values to vectors
+                    GGML_F32X k_vec = GGML_F32X_SET1(k_val);
+                    GGML_F32X q_vec = GGML_F32X_SET1(q_val);
+                    GGML_F32X g_vec = GGML_F32X_SET1(g_val);
+
+                    for (int64_t j = 0; j < vec_count; j++) {
+                        size_t base_j = j * GLA_VECTOR_SIZE;
+                        size_t t_h_j_offset = t_h_offset + base_j;
+                        size_t h_2d_i_j_offset = h_2d_i_offset + base_j;
+
+                        // Load x elements at once
+                        GGML_F32X v_vec = GGML_F32X_LOAD(&v[t_h_j_offset]);
+                        GGML_F32X prev_state_vec = GGML_F32X_LOAD(&state_prev[h_2d_i_j_offset]);
+                        GGML_F32X dst_vec = GGML_F32X_LOAD(&dst_data[t_h_j_offset]);
+
+                        // Compute kv = v * k
+                        GGML_F32X kv_vec = GGML_F32X_MUL(v_vec, k_vec);
+
+                        // Compute temp = prev_state * g + kv
+                        GGML_F32X temp_vec = GGML_F32X_FMA(kv_vec, prev_state_vec, g_vec);
+
+                        // Update dst: dst += temp * q
+                        dst_vec = GGML_F32X_FMA(dst_vec, temp_vec, q_vec);
+                        GGML_F32X_STORE(&dst_data[t_h_j_offset], dst_vec);
+
+                        // Update state
+                        GGML_F32X_STORE(&state_cur[h_2d_i_j_offset], temp_vec);
+                    }
+
+                    // Handle remaining elements, this will not be used.
+                    for (int64_t j = vec_count * GLA_VECTOR_SIZE; j < head_size; j++) {
+                        size_t t_h_j_offset = t_h_offset + j;
+                        size_t h_2d_i_j_offset = h_2d_i_offset + j;
+                        float v_val = v[t_h_j_offset];
+                        float kv_val = v_val * k_val;
+                        float prev_state_val = state_prev[h_2d_i_j_offset];
+                        float temp_val = kv_val + prev_state_val * g_val;
+                        dst_data[t_h_j_offset] += temp_val * q_val;
+                        state_cur[h_2d_i_j_offset] = temp_val;
+                    }
+                }
+            }
+        }
+
+    #else
+        for (int64_t t = 0; t < T; t++) {
+            size_t t_offset = t * t_stride;
+            size_t state_offset = head_size * C * (t / (T / n_seqs));
+            float * state_cur = state + state_offset;
+            float * state_prev = t % (T / n_seqs) ? state_cur : (float*)dst->src[4]->data + state_offset;
+
+            for (int64_t h = h_start; h < h_end; h++) {
+                size_t h_offset = h * h_stride;
+                size_t t_h_offset = t_offset + h_offset;
+                size_t h_2d_offset = h * h_stride_2d;
+
+                for (int64_t i = 0; i < head_size; i++) {
+                    size_t t_h_i_offset = t_h_offset + i;
+                    size_t h_2d_i_offset = h_2d_offset + i * h_stride;
+
+                    float k_val = k[t_h_i_offset];
+                    float q_val = q[t_h_i_offset] * scale;
+                    float g_val = g[t_h_i_offset];
+
+                    for (int64_t j = 0; j < head_size; j++) {
+                        size_t t_h_j_offset = t_h_offset + j;
+                        size_t h_2d_i_j_offset = h_2d_i_offset + j;
+
+                        float v_val = v[t_h_j_offset];
+                        float kv_val = v_val * k_val;
+                        float prev_state_val = state_prev[h_2d_i_j_offset];
+                        float temp_val = prev_state_val * g_val + kv_val;
+                        dst_data[t_h_j_offset] += temp_val * q_val;
+                        state_cur[h_2d_i_j_offset] = temp_val;
+                    }
+                }
+            }
+        }
+    #endif
+}
+
+
+static void ggml_compute_forward_gla(
+        const struct ggml_compute_params * params,
+        struct ggml_tensor * dst) {
+
+    const struct ggml_tensor * src0 = dst->src[0];
+
+    switch (src0->type) {
+        case GGML_TYPE_F32:
+            {
+                ggml_compute_forward_gla_f32(params, dst);
+            } break;
+        default:
+            {
+                GGML_ABORT("fatal error");
+            }
+    }
+}
+
 // ggml_compute_forward_map_unary
 
 static void ggml_compute_forward_map_unary_f32(
@@ -12749,6 +12940,10 @@ static void ggml_compute_forward(struct ggml_compute_params * params, struct ggm
             {
                 ggml_compute_forward_rwkv_wkv6(params, tensor);
             } break;
+        case GGML_OP_GATED_LINEAR_ATTN:
+            {
+                ggml_compute_forward_gla(params, tensor);
+            } break;
         case GGML_OP_MAP_UNARY:
             {
                 ggml_unary_op_f32_t fun;
@@ -13047,6 +13242,7 @@ static int ggml_get_n_tasks(struct ggml_tensor * node, int n_threads) {
         case GGML_OP_WIN_UNPART:
         case GGML_OP_GET_REL_POS:
         case GGML_OP_RWKV_WKV6:
+        case GGML_OP_GATED_LINEAR_ATTN:
         case GGML_OP_MAP_UNARY:
         case GGML_OP_MAP_BINARY:
         case GGML_OP_MAP_CUSTOM1_F32:
@@ -13472,6 +13668,7 @@ struct ggml_cplan ggml_graph_plan(
                     } break;
                 case GGML_OP_SOFT_MAX:
                 case GGML_OP_ROPE:
+                case GGML_OP_ROPE_BACK:
                     {
                         cur = ggml_type_size(GGML_TYPE_F32) * node->ne[0] * n_tasks;
                     } break;

ggml/src/ggml-cpu/ggml-cpu.cpp

@@ -403,8 +403,6 @@ static bool ggml_backend_cpu_device_supports_op(ggml_backend_dev_t dev, const st
                 op->type != GGML_TYPE_IQ1_M; // missing type_traits.from_float
         case GGML_OP_MUL_MAT:
             return src1->type == GGML_TYPE_F32 || src1->type == ggml_get_type_traits_cpu(src0->type)->vec_dot_type;
-        case GGML_OP_ROPE_BACK:
-            return op->src[2] == NULL && (op->op_params[2] & 4) == 0;
         case GGML_OP_IM2COL_BACK:
             return src0->type == GGML_TYPE_F32 && src1->type == GGML_TYPE_F32;
         case GGML_OP_OUT_PROD:

ggml/src/ggml-cpu/llamafile/sgemm.cpp

@@ -54,6 +54,7 @@
 #include "ggml-quants.h"
 
 #include <atomic>
+#include <array>
 
 #ifdef _MSC_VER
 #define NOINLINE __declspec(noinline)
@@ -1051,6 +1052,704 @@ class tinyBLAS_Q0_AVX {
       } \
    } \
 
+template <typename TA, typename TB, typename TC>
+class tinyBLAS_Q0_PPC {
+  public:
+    tinyBLAS_Q0_PPC(int64_t k,
+                const TA *A, int64_t lda,
+                const TB *B, int64_t ldb,
+                TC *C, int64_t ldc,
+                int ith, int nth)
+        : A(A), B(B), C(C), k(k), lda(lda), ldb(ldb), ldc(ldc), ith(ith), nth(nth) {
+    }
+
+    void matmul(int64_t m, int64_t n) {
+        mnpack(0, m, 0, n);
+    }
+
+  private:
+
+    template<int RM, int RN>
+    inline void save_res(int ii, int jj, int idx, vector float* fin_res) {
+       for (int I = 0; I < RM; I++) {
+          for (int J = 0; J < RN; J++) {
+             *((float*)(C+ii+((jj+J)*ldc)+I)) = *((float*)&fin_res[idx+I]+J);
+          }
+       }
+    }
+
+    template<int size>
+    inline void compute(acc_t* ACC, int c_idx, int s_idx, std::array<int, size>& comparray, vector float* vs, vector float* fin_res) {
+       vector signed int vec_C[4];
+       vector float CA[4] = {0};
+       vector float res[4] = {0};
+       __builtin_mma_disassemble_acc(vec_C, ACC);
+       for (int i = 0; i < 4; i++) {
+          CA[i] = vec_splats((float)(((double)comparray[c_idx+i]) * -128.0));
+          res[i] = vec_add(vec_ctf(vec_C[i], 0), CA[i]);
+          fin_res[s_idx+i] = vec_madd(res[i], vs[s_idx+i], fin_res[s_idx+i]);
+       }
+    }
+
+    template<typename VA, typename VB>
+    void packNormal(const TA* a, int64_t lda, int rows, int cols, VA* vec, bool flip) {
+        int64_t i, j;
+        TA *aoffset = NULL;
+        VA *vecOffset = NULL;
+        TA *aoffset1 = NULL, *aoffset2 = NULL, *aoffset3 = NULL, *aoffset4 = NULL;
+        TA *aoffset5 = NULL, *aoffset6 = NULL, *aoffset7 = NULL, *aoffset8 = NULL;
+        __vector_pair C1, C2, C3, C4, C5, C6, C7, C8;
+        VB c1[2] = {0}, c2[2] = {0}, c3[2] = {0}, c4[2]={0};
+        VB c5[2] = {0}, c6[2] = {0}, c7[2] = {0}, c8[2]={0};
+        VB t1, t2, t3, t4, t5, t6, t7, t8;
+        vector unsigned char xor_vector;
+        uint8_t flip_vec = 0x80;
+        xor_vector = vec_splats(flip_vec);
+        vector unsigned char swiz1 = {0, 1, 2, 3, 4, 5, 6, 7, 16, 17, 18, 19, 20, 21, 22, 23};
+        vector unsigned char swiz2 = {8, 9, 10, 11, 12, 13, 14, 15, 24, 25, 26, 27, 28, 29, 30, 31};
+        vector unsigned char swiz3 = {0, 1, 2, 3, 8, 9, 10, 11, 16, 17, 18, 19, 24, 25, 26, 27};
+        vector unsigned char swiz4 = {4, 5, 6, 7, 12, 13, 14, 15, 20, 21, 22, 23, 28, 29, 30, 31};
+
+        aoffset = const_cast<TA*>(a);
+        vecOffset = vec;
+        j = (rows >> 3);
+        if (j > 0) {
+            do {
+            aoffset1 = aoffset;
+            aoffset2 = aoffset1 + lda;
+            aoffset3 = aoffset2 + lda;
+            aoffset4 = aoffset3 + lda;
+            aoffset5 = aoffset4 + lda;
+            aoffset6 = aoffset5 + lda;
+            aoffset7 = aoffset6 + lda;
+            aoffset8 = aoffset7 + lda;
+            aoffset += 8 * lda;
+
+            i = (cols >> 3);
+            if (i > 0) {
+               do {
+                    C1 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset1->qs);
+                    C2 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset2->qs);
+                    C3 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset3->qs);
+                    C4 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset4->qs);
+                    C5 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset5->qs);
+                    C6 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset6->qs);
+                    C7 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset7->qs);
+                    C8 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset8->qs);
+
+                    __builtin_vsx_disassemble_pair(c1, &C1);
+                    __builtin_vsx_disassemble_pair(c2, &C2);
+                    __builtin_vsx_disassemble_pair(c3, &C3);
+                    __builtin_vsx_disassemble_pair(c4, &C4);
+                    __builtin_vsx_disassemble_pair(c5, &C5);
+                    __builtin_vsx_disassemble_pair(c6, &C6);
+                    __builtin_vsx_disassemble_pair(c7, &C7);
+                    __builtin_vsx_disassemble_pair(c8, &C8);
+
+                    t1 = vec_perm(c1[0], c2[0], swiz1);
+                    t2 = vec_perm(c1[0], c2[0], swiz2);
+                    t3 = vec_perm(c3[0], c4[0], swiz1);
+                    t4 = vec_perm(c3[0], c4[0], swiz2);
+                    t5 = vec_perm(t1, t3, swiz3);
+                    t6 = vec_perm(t1, t3, swiz4);
+                    t7 = vec_perm(t2, t4, swiz3);
+                    t8 = vec_perm(t2, t4, swiz4);
+                    if (flip == true) {
+                       t5 = vec_xor(t5, xor_vector);
+                       t6 = vec_xor(t6, xor_vector);
+                       t7 = vec_xor(t7, xor_vector);
+                       t8 = vec_xor(t8, xor_vector);
+                    }
+                    vec_xst(t5, 0, vecOffset);
+                    vec_xst(t6, 0, vecOffset+16);
+                    vec_xst(t7, 0, vecOffset+32);
+                    vec_xst(t8, 0, vecOffset+48);
+
+                    t1 = vec_perm(c1[1], c2[1], swiz1);
+                    t2 = vec_perm(c1[1], c2[1], swiz2);
+                    t3 = vec_perm(c3[1], c4[1], swiz1);
+                    t4 = vec_perm(c3[1], c4[1], swiz2);
+                    t5 = vec_perm(t1, t3, swiz3);
+                    t6 = vec_perm(t1, t3, swiz4);
+                    t7 = vec_perm(t2, t4, swiz3);
+                    t8 = vec_perm(t2, t4, swiz4);
+                    if (flip == true) {
+                       t5 = vec_xor(t5, xor_vector);
+                       t6 = vec_xor(t6, xor_vector);
+                       t7 = vec_xor(t7, xor_vector);
+                       t8 = vec_xor(t8, xor_vector);
+                    }
+                    vec_xst(t5, 0, vecOffset+64);
+                    vec_xst(t6, 0, vecOffset+80);
+                    vec_xst(t7, 0, vecOffset+96);
+                    vec_xst(t8, 0, vecOffset+112);
+
+                    t1 = vec_perm(c5[0], c6[0], swiz1);
+                    t2 = vec_perm(c5[0], c6[0], swiz2);
+                    t3 = vec_perm(c7[0], c8[0], swiz1);
+                    t4 = vec_perm(c7[0], c8[0], swiz2);
+                    t5 = vec_perm(t1, t3, swiz3);
+                    t6 = vec_perm(t1, t3, swiz4);
+                    t7 = vec_perm(t2, t4, swiz3);
+                    t8 = vec_perm(t2, t4, swiz4);
+                    if (flip == true) {
+                       t5 = vec_xor(t5, xor_vector);
+                       t6 = vec_xor(t6, xor_vector);
+                       t7 = vec_xor(t7, xor_vector);
+                       t8 = vec_xor(t8, xor_vector);
+                    }
+                    vec_xst(t5, 0, vecOffset+128);
+                    vec_xst(t6, 0, vecOffset+144);
+                    vec_xst(t7, 0, vecOffset+160);
+                    vec_xst(t8, 0, vecOffset+176);
+
+                    t1 = vec_perm(c5[1], c6[1], swiz1);
+                    t2 = vec_perm(c5[1], c6[1], swiz2);
+                    t3 = vec_perm(c7[1], c8[1], swiz1);
+                    t4 = vec_perm(c7[1], c8[1], swiz2);
+                    t5 = vec_perm(t1, t3, swiz3);
+                    t6 = vec_perm(t1, t3, swiz4);
+                    t7 = vec_perm(t2, t4, swiz3);
+                    t8 = vec_perm(t2, t4, swiz4);
+                    if (flip == true) {
+                       t5 = vec_xor(t5, xor_vector);
+                       t6 = vec_xor(t6, xor_vector);
+                       t7 = vec_xor(t7, xor_vector);
+                       t8 = vec_xor(t8, xor_vector);
+                    }
+                    vec_xst(t5, 0, vecOffset+192);
+                    vec_xst(t6, 0, vecOffset+208);
+                    vec_xst(t7, 0, vecOffset+224);
+                    vec_xst(t8, 0, vecOffset+240);
+
+                    aoffset1 += lda;
+                    aoffset2 += lda;
+                    aoffset3 += lda;
+                    aoffset4 += lda;
+                    aoffset5 += lda;
+                    aoffset6 += lda;
+                    aoffset7 += lda;
+                    aoffset8 += lda;
+                    vecOffset += 256;
+                    i--;
+               } while(i > 0);
+            }
+            j--;
+        } while(j > 0);
+    }
+
+    if (rows & 4) {
+            aoffset1 = aoffset;
+            aoffset2 = aoffset1 + lda;
+            aoffset3 = aoffset2 + lda;
+            aoffset4 = aoffset3 + lda;
+            aoffset += 4 * lda;
+
+        i = (cols >> 3);
+            if (i > 0) {
+               do {
+                    C1 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset1->qs);
+                    C2 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset2->qs);
+                    C3 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset3->qs);
+                    C4 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset4->qs);
+
+                    __builtin_vsx_disassemble_pair(c1, &C1);
+                    __builtin_vsx_disassemble_pair(c2, &C2);
+                    __builtin_vsx_disassemble_pair(c3, &C3);
+                    __builtin_vsx_disassemble_pair(c4, &C4);
+
+                    t1 = vec_perm(c1[0], c2[0], swiz1);
+                    t2 = vec_perm(c1[0], c2[0], swiz2);
+                    t3 = vec_perm(c3[0], c4[0], swiz1);
+                    t4 = vec_perm(c3[0], c4[0], swiz2);
+                    t5 = vec_perm(t1, t3, swiz3);
+                    t6 = vec_perm(t1, t3, swiz4);
+                    t7 = vec_perm(t2, t4, swiz3);
+                    t8 = vec_perm(t2, t4, swiz4);
+                    if (flip == true) {
+                       t5 = vec_xor(t5, xor_vector);
+                       t6 = vec_xor(t6, xor_vector);
+                       t7 = vec_xor(t7, xor_vector);
+                       t8 = vec_xor(t8, xor_vector);
+                    }
+                    vec_xst(t5, 0, vecOffset);
+                    vec_xst(t6, 0, vecOffset+16);
+                    vec_xst(t7, 0, vecOffset+32);
+                    vec_xst(t8, 0, vecOffset+48);
+
+                    t1 = vec_perm(c1[1], c2[1], swiz1);
+                    t2 = vec_perm(c1[1], c2[1], swiz2);
+                    t3 = vec_perm(c3[1], c4[1], swiz1);
+                    t4 = vec_perm(c3[1], c4[1], swiz2);
+                    t5 = vec_perm(t1, t3, swiz3);
+                    t6 = vec_perm(t1, t3, swiz4);
+                    t7 = vec_perm(t2, t4, swiz3);
+                    t8 = vec_perm(t2, t4, swiz4);
+                    if (flip == true) {
+                       t5 = vec_xor(t5, xor_vector);
+                       t6 = vec_xor(t6, xor_vector);
+                       t7 = vec_xor(t7, xor_vector);
+                       t8 = vec_xor(t8, xor_vector);
+                    }
+                    vec_xst(t5, 0, vecOffset+64);
+                    vec_xst(t6, 0, vecOffset+80);
+                    vec_xst(t7, 0, vecOffset+96);
+                    vec_xst(t8, 0, vecOffset+112);
+
+                    aoffset1 += lda;
+                    aoffset2 += lda;
+                    aoffset3 += lda;
+                    aoffset4 += lda;
+                    vecOffset += 128;
+                    i--;
+               } while(i > 0);
+            }
+        }
+        if (rows & 3) {
+            aoffset1 = aoffset;
+            aoffset2 = aoffset1 + lda;
+            aoffset3 = aoffset2 + lda;
+            i = (cols >> 3);
+        if (i > 0) {
+                do {
+                    switch(rows) {
+                        case 3: C3 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset3->qs);
+                                __builtin_vsx_disassemble_pair(c3, &C3);
+                        case 2: C2 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset2->qs);
+                                __builtin_vsx_disassemble_pair(c2, &C2);
+                        case 1: C1 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset1->qs);
+                                __builtin_vsx_disassemble_pair(c1, &C1);
+                                break;
+                    }
+                    t1 = vec_perm(c1[0], c2[0], swiz1);
+                    t2 = vec_perm(c1[0], c2[0], swiz2);
+                    t3 = vec_perm(c3[0], c4[0], swiz1);
+                    t4 = vec_perm(c3[0], c4[0], swiz2);
+                    t5 = vec_perm(t1, t3, swiz3);
+                    t6 = vec_perm(t1, t3, swiz4);
+                    t7 = vec_perm(t2, t4, swiz3);
+                    t8 = vec_perm(t2, t4, swiz4);
+                    if (flip == true) {
+                       t5 = vec_xor(t5, xor_vector);
+                       t6 = vec_xor(t6, xor_vector);
+                       t7 = vec_xor(t7, xor_vector);
+                       t8 = vec_xor(t8, xor_vector);
+                    }
+                    vec_xst(t5, 0, vecOffset);
+                    vec_xst(t6, 0, vecOffset+16);
+                    vec_xst(t7, 0, vecOffset+32);
+                    vec_xst(t8, 0, vecOffset+48);
+
+                    t1 = vec_perm(c1[1], c2[1], swiz1);
+                    t2 = vec_perm(c1[1], c2[1], swiz2);
+                    t3 = vec_perm(c3[1], c4[1], swiz1);
+                    t4 = vec_perm(c3[1], c4[1], swiz2);
+                    t5 = vec_perm(t1, t3, swiz3);
+                    t6 = vec_perm(t1, t3, swiz4);
+                    t7 = vec_perm(t2, t4, swiz3);
+                    t8 = vec_perm(t2, t4, swiz4);
+                    if (flip == true) {
+                       t5 = vec_xor(t5, xor_vector);
+                       t6 = vec_xor(t6, xor_vector);
+                       t7 = vec_xor(t7, xor_vector);
+                       t8 = vec_xor(t8, xor_vector);
+                    }
+                    vec_xst(t5, 0, vecOffset+64);
+                    vec_xst(t6, 0, vecOffset+80);
+                    vec_xst(t7, 0, vecOffset+96);
+                    vec_xst(t8, 0, vecOffset+112);
+
+                    aoffset1 += lda;
+                    aoffset2 += lda;
+                    aoffset3 += lda;
+                    vecOffset += 128;
+                    i--;
+               } while(i > 0);
+            }
+        }
+    }
+
+    void mnpack(int64_t m0, int64_t m, int64_t n0, int64_t n) {
+        int64_t mc, nc, mp, np;
+        int m_rem = MIN(m - m0, 8);
+        int n_rem = MIN(n - n0, 8);
+        // TO-DO: KERNEL_16x8 and KERNEL_8x16 are having some performance
+        // issues. After resolving them, below code will be enabled.
+        /*if (m_rem >= 16 && n_rem >= 8) {
+            mc = 16;
+            nc = 8;
+            gemm<16,8>(m0, m, n0, n);
+        } else if(m_rem >= 8 && n_rem >= 16) {
+            mc = 8;
+            nc = 16;
+            gemm<8,16>(m0, m, n0, n);
+        }*/
+        if (m_rem >= 8 && n_rem >= 8) {
+            mc = 8;
+            nc = 8;
+            gemm<8,8>(m0, m, n0, n);
+        } else if (m_rem >= 4 && n_rem >= 8) {
+            mc = 4;
+            nc = 8;
+            gemm<4,8>(m0, m, n0, n);
+        } else if (m_rem >= 8 && n_rem >= 4) {
+            mc = 8;
+            nc = 4;
+            gemm<8,4>(m0, m, n0, n);
+        } else if (m_rem >= 4 && n_rem >= 4) {
+            mc = 4;
+            nc = 4;
+            gemm_small<4, 4>(m0, m, n0, n);
+        } else if ((m_rem < 4) && (n_rem > 4)) {
+            nc = 4;
+            switch(m_rem) {
+                case 1:
+                    mc = 1;
+                    gemm_small<1, 4>(m0, m, n0, n);
+                    break;
+                case 2:
+                    mc = 2;
+                    gemm_small<2, 4>(m0, m, n0, n);
+                    break;
+                case 3:
+                    mc = 3;
+                    gemm_small<3, 4>(m0, m, n0, n);
+                    break;
+                default:
+                    return;
+            }
+        } else if ((m_rem > 4) && (n_rem < 4)) {
+            mc = 4;
+            switch(n_rem) {
+                case 1:
+                    nc = 1;
+                    gemm_small<4, 1>(m0, m, n0, n);
+                    break;
+                case 2:
+                    nc = 2;
+                    gemm_small<4, 2>(m0, m, n0, n);
+                    break;
+                case 3:
+                    nc = 3;
+                    gemm_small<4, 3>(m0, m, n0, n);
+                    break;
+                default:
+                    return;
+            }
+        } else {
+            switch((m_rem << 4) | n_rem) {
+                case 0x43:
+                    mc = 4;
+                    nc = 3;
+                    gemm_small<4, 3>(m0, m, n0, n);
+                    break;
+                case 0x42:
+                    mc = 4;
+                    nc = 2;
+                    gemm_small<4, 2>(m0, m, n0, n);
+                    break;
+                case 0x41:
+                    mc = 4;
+                    nc = 1;
+                    gemm_small<4, 1>(m0, m, n0, n);
+                    break;
+                case 0x34:
+                    mc = 3;
+                    nc = 4;
+                    gemm_small<3, 4>(m0, m, n0, n);
+                    break;
+                case 0x33:
+                    mc = 3;
+                    nc = 3;
+                    gemm_small<3, 3>(m0, m, n0, n);
+                    break;
+                case 0x32:
+                    mc = 3;
+                    nc = 2;
+                    gemm_small<3, 2>(m0, m, n0, n);
+                    break;
+                case 0x31:
+                    mc = 3;
+                    nc = 1;
+                    gemm_small<3, 1>(m0, m, n0, n);
+                    break;
+                case 0x24:
+                    mc = 2;
+                    nc = 4;
+                    gemm_small<2, 4>(m0, m, n0, n);
+                    break;
+                case 0x23:
+                    mc = 2;
+                    nc = 3;
+                    gemm_small<2, 3>(m0, m, n0, n);
+                    break;
+                case 0x22:
+                    mc = 2;
+                    nc = 2;
+                    gemm_small<2, 2>(m0, m, n0, n);
+                    break;
+                case 0x21:
+                    mc = 2;
+                    nc = 1;
+                    gemm_small<2, 1>(m0, m, n0, n);
+                    break;
+                case 0x14:
+                    mc = 1;
+                    nc = 4;
+                    gemm_small<1, 4>(m0, m, n0, n);
+                    break;
+                case 0x13:
+                    mc = 1;
+                    nc = 3;
+                    gemm_small<1, 3>(m0, m, n0, n);
+                    break;
+                case 0x12:
+                    mc = 1;
+                    nc = 2;
+                    gemm_small<1, 2>(m0, m, n0, n);
+                    break;
+                case 0x11:
+                    mc = 1;
+                    nc = 1;
+                    gemm_small<1, 1>(m0, m, n0, n);
+                    break;
+                default:
+                    return;
+            }
+        }
+        mp = m0 + (m - m0) / mc * mc;
+        np = n0 + (n - n0) / nc * nc;
+        mnpack(mp, m, n0, np);
+        mnpack(m0, m, np, n);
+    }
+
+    void KERNEL_4x8(int64_t ii, int64_t jj) {
+        vec_t vec_A[8], vec_B[16] = {0};
+        acc_t acc_0, acc_1;
+        std::array<int, 4> comparray;
+        vector float fin_res[8] = {0};
+        vector float vs[8] = {0};
+        for (int l = 0; l < k; l++) {
+            __builtin_mma_xxsetaccz(&acc_0);
+            __builtin_mma_xxsetaccz(&acc_1);
+            packNormal<int8_t, vector signed char>((A+(ii*lda)+l), lda, 4, 8, (int8_t*)vec_A, false);
+            packNormal<uint8_t, vector unsigned char>((B+(jj*ldb)+l), ldb, 8, 8, (uint8_t*)vec_B, true);
+            for(int x = 0; x < 8; x++) {
+                __builtin_mma_xvi8ger4pp(&acc_0, vec_A[x], vec_B[x]);
+                __builtin_mma_xvi8ger4pp(&acc_1, vec_A[x], vec_B[x+8]);
+            }
+            for (int I = 0; I<4; I++) {
+                for (int J = 0; J<4; J++) {
+                    *((float*)&vs[I]+J) = (unhalf((A+((ii+I)*lda)+l)->d) * unhalf((B+((jj+J)*ldb)+l)->d));
+                    *((float*)&vs[I+4]+J) = (unhalf((A+((ii+I)*lda)+l)->d) * unhalf((B+((jj+J+4)*ldb)+l)->d));
+                }
+            }
+            auto aoffset = A+(ii*lda)+l;
+            for (int i = 0; i < 4; i++) {
+                comparray[i] = 0;
+                int ca = 0;
+                const int8_t *at = aoffset->qs;
+                for (int j = 0; j < 32; j++)
+                    ca += (int)*at++;
+                comparray[i] = ca;
+                aoffset += lda;
+            }
+            compute<4>(&acc_0, 0, 0, comparray, vs, fin_res);
+            compute<4>(&acc_1, 0, 4, comparray, vs, fin_res);
+        }
+        save_res<4, 4>(ii, jj, 0, fin_res);
+        save_res<4, 4>(ii, jj+4, 4, fin_res);
+    }
+
+    void KERNEL_8x4(int64_t ii, int64_t jj) {
+        vec_t vec_A[16], vec_B[8] = {0};
+        acc_t acc_0, acc_1;
+        std::array<int, 8> comparray;
+        vector float fin_res[8] = {0};
+        vector float vs[8] = {0};
+        for (int l = 0; l < k; l++) {
+            __builtin_mma_xxsetaccz(&acc_0);
+            __builtin_mma_xxsetaccz(&acc_1);
+            packNormal<int8_t, vector signed char>((A+(ii*lda)+l), lda, 8, 8, (int8_t*)vec_A, false);
+            packNormal<uint8_t, vector unsigned char>((B+(jj*ldb)+l), ldb, 4, 8, (uint8_t*)vec_B, true);
+            for(int x = 0; x < 8; x++) {
+                __builtin_mma_xvi8ger4pp(&acc_0, vec_A[x], vec_B[x]);
+                __builtin_mma_xvi8ger4pp(&acc_1, vec_A[x+8], vec_B[x]);
+            }
+            for (int I = 0; I<8; I++) {
+                for (int J = 0; J<4; J++) {
+                    *((float*)&vs[I]+J) = (unhalf((A+((ii+I)*lda)+l)->d) * unhalf((B+((jj+J)*ldb)+l)->d));
+                }
+            }
+            auto aoffset = A+(ii*lda)+l;
+            for (int i = 0; i < 8; i++) {
+                comparray[i] = 0;
+                int ca = 0;
+                const int8_t *at = aoffset->qs;
+                for (int j = 0; j < 32; j++)
+                    ca += (int)*at++;
+                comparray[i] = ca;
+                aoffset += lda;
+            }
+            compute<8>(&acc_0, 0, 0, comparray, vs, fin_res);
+            compute<8>(&acc_1, 4, 4, comparray, vs, fin_res);
+        }
+        save_res<4, 4>(ii, jj, 0, fin_res);
+        save_res<4, 4>(ii+4, jj, 4, fin_res);
+    }
+
+    void KERNEL_8x8(int64_t ii, int64_t jj) {
+        vec_t vec_A[16], vec_B[16] = {0};
+        acc_t acc_0, acc_1, acc_2, acc_3;
+        std::array<int, 8> comparray;
+        vector float fin_res[16] = {0};
+        vector float vs[16] = {0};
+        for (int l = 0; l < k; l++) {
+            __builtin_mma_xxsetaccz(&acc_0);
+            __builtin_mma_xxsetaccz(&acc_1);
+            __builtin_mma_xxsetaccz(&acc_2);
+            __builtin_mma_xxsetaccz(&acc_3);
+            packNormal<int8_t, vector signed char>((A+(ii*lda)+l), lda, 8, 8, (int8_t*)vec_A, false);
+            packNormal<uint8_t, vector unsigned char>((B+(jj*ldb)+l), ldb, 8, 8, (uint8_t*)vec_B, true);
+            for(int x = 0; x < 8; x++) {
+                __builtin_mma_xvi8ger4pp(&acc_0, vec_A[x], vec_B[x]);
+                __builtin_mma_xvi8ger4pp(&acc_1, vec_A[x+8], vec_B[x]);
+                __builtin_mma_xvi8ger4pp(&acc_2, vec_A[x], vec_B[x+8]);
+                __builtin_mma_xvi8ger4pp(&acc_3, vec_A[x+8], vec_B[x+8]);
+            }
+            for (int I = 0; I<8; I++) {
+                for (int J = 0; J<4; J++) {
+                    *((float*)&vs[I]+J) = (unhalf((A+((ii+I)*lda)+l)->d) * unhalf((B+((jj+J)*ldb)+l)->d));
+                    *((float*)&vs[I+8]+J) = (unhalf((A+((ii+I)*lda)+l)->d) * unhalf((B+((jj+J+4)*ldb)+l)->d));
+                }
+            }
+            auto aoffset = A+(ii*lda)+l;
+            for (int i = 0; i < 8; i++) {
+                comparray[i] = 0;
+                int ca = 0;
+                const int8_t *at = aoffset->qs;
+                for (int j = 0; j < 32; j++)
+                    ca += (int)*at++;
+                comparray[i] = ca;
+                aoffset += lda;
+            }
+            compute<8>(&acc_0, 0, 0, comparray, vs, fin_res);
+            compute<8>(&acc_1, 4, 4, comparray, vs, fin_res);
+            compute<8>(&acc_2, 0, 8, comparray, vs, fin_res);
+            compute<8>(&acc_3, 4, 12, comparray, vs, fin_res);
+        }
+        save_res<4, 4>(ii, jj, 0, fin_res);
+        save_res<4, 4>(ii+4, jj, 4, fin_res);
+        save_res<4, 4>(ii, jj+4, 8, fin_res);
+        save_res<4, 4>(ii+4, jj+4, 12, fin_res);
+    }
+
+    template<int RM, int RN>
+    void gemm_small(int64_t m0, int64_t m, int64_t n0, int64_t n) {
+        int64_t ytiles = (m - m0) / RM;
+        int64_t xtiles = (n - n0) / RN;
+        int64_t tiles = xtiles * ytiles;
+        int64_t duty = (tiles + nth - 1) / nth;
+        int64_t start = duty * ith;
+        int64_t end = start + duty;
+        vec_t vec_A[8], vec_B[8] = {0};
+        vector signed int vec_C[4];
+        acc_t acc_0;
+
+        if (end > tiles)
+            end = tiles;
+        for (int64_t job = start; job < end; ++job) {
+            int64_t ii = m0 + job / xtiles * RM;
+            int64_t jj = n0 + job % xtiles * RN;
+            std::array<int, RM> comparray;
+            vector float res[4] = {0};
+            vector float fin_res[4] = {0};
+            vector float vs[4] = {0};
+            vector float CA[4] = {0};
+            __builtin_prefetch((A+(ii*lda)+0)->qs, 0, 1); // prefetch first value
+            __builtin_prefetch((B+(jj*ldb)+0)->qs, 0, 1); // prefetch first value
+            for (int l = 0; l < k; l++) {
+                __builtin_prefetch((A+(ii*lda)+(l+1))->qs, 0, 1); // prefetch one loop ahead
+                __builtin_prefetch((B+(jj*ldb)+(l+1))->qs, 0, 1); // prefetch one loop ahead
+                __builtin_mma_xxsetaccz(&acc_0);
+                packNormal<int8_t, vector signed char>((A+(ii*lda)+l), lda, RM, 8, (int8_t*)vec_A, false);
+                packNormal<uint8_t, vector unsigned char>((B+(jj*ldb)+l), ldb, RN, 8, (uint8_t*)vec_B, true);
+                for(int x = 0; x < 8; x+=4) {
+                    __builtin_mma_xvi8ger4pp(&acc_0, vec_A[x], vec_B[x]);
+                    __builtin_mma_xvi8ger4pp(&acc_0, vec_A[x+1], vec_B[x+1]);
+                    __builtin_mma_xvi8ger4pp(&acc_0, vec_A[x+2], vec_B[x+2]);
+                    __builtin_mma_xvi8ger4pp(&acc_0, vec_A[x+3], vec_B[x+3]);
+                }
+                for (int I = 0; I<RM; I++) {
+                    for (int J = 0; J<RN; J++) {
+                        *((float*)&vs[I]+J) = (unhalf((A+((ii+I)*lda)+l)->d) * unhalf((B+((jj+J)*ldb)+l)->d));
+                    }
+                }
+                __builtin_mma_disassemble_acc(vec_C, &acc_0);
+                auto aoffset = A+(ii*lda)+l;
+                for (int i = 0; i < RM; i++) {
+                    comparray[i] = 0;
+                    int ca = 0;
+                    const int8_t *at = aoffset->qs;
+                    for (int j = 0; j < 32; j++)
+                        ca += (int)*at++;
+                    comparray[i] = ca;
+                    aoffset += lda;
+                }
+
+                for (int i = 0; i < RM; i++) {
+                    CA[i] = vec_splats((float)(((double)comparray[i]) * -128.0));
+                    res[i] = vec_add(vec_ctf(vec_C[i], 0), CA[i]);
+                    fin_res[i] = vec_madd(res[i], vs[i], fin_res[i]);
+                }
+            }
+            save_res<RM, RN>(ii, jj, 0, fin_res);
+        }
+    }
+
+    template<int RM, int RN>
+    inline void kernel(int64_t ii, int64_t jj) {
+       if constexpr(RM == 4 && RN == 8) {
+          KERNEL_4x8(ii,jj);
+       } else if constexpr(RM == 8 && RN == 4) {
+          KERNEL_8x4(ii,jj);
+       } else if constexpr(RM == 8 && RN == 8) {
+          KERNEL_8x8(ii,jj);
+       } else {
+          static_assert(false, "RN/RM values not supported");
+       }
+    }
+
+    template <int RM, int RN>
+    NOINLINE void gemm(int64_t m0, int64_t m, int64_t n0, int64_t n) {
+        int64_t ytiles = (m - m0) / RM;
+        int64_t xtiles = (n - n0) / RN;
+        int64_t tiles = xtiles * ytiles;
+        int64_t duty = (tiles + nth - 1) / nth;
+        int64_t start = duty * ith;
+        int64_t end = start + duty;
+        if (end > tiles)
+            end = tiles;
+        for (int64_t job = start; job < end; ++job) {
+            int64_t ii = m0 + job / xtiles * RM;
+            int64_t jj = n0 + job % xtiles * RN;
+            kernel<RM, RN>(ii, jj);
+        }
+    }
+
+    const TA *const A;
+    const TB *const B;
+    TC *C;
+    TA *At;
+    TB *Bt;
+    const int64_t k;
+    const int64_t lda;
+    const int64_t ldb;
+    const int64_t ldc;
+    const int ith;
+    const int nth;
+};
+
 template <typename TA, typename TB, typename TC>
 class tinyBLAS_PPC {
   public:
@@ -1070,13 +1769,17 @@ class tinyBLAS_PPC {
 
     void (tinyBLAS_PPC::*kernel)(int64_t, int64_t);
 
-    void READ_BLOCK(const float* a, int64_t lda, int rows, int cols, float* vec) {
+    template<typename VA>
+    void packTranspose(const TA* a, int64_t lda, int rows, int cols, TA* vec) {
         int64_t i, j;
-        float *aoffset = NULL, *boffset = NULL;
-        float *aoffset1 = NULL, *aoffset2 = NULL, *aoffset3 = NULL, *aoffset4 = NULL;
-        float *aoffset5 = NULL, *aoffset6 = NULL, *aoffset7 = NULL, *aoffset8 = NULL;
-
-        aoffset = const_cast<float*>(a);
+        TA *aoffset = NULL, *boffset = NULL;
+        TA *aoffset1 = NULL, *aoffset2 = NULL, *aoffset3 = NULL, *aoffset4 = NULL;
+        TA *aoffset5 = NULL, *aoffset6 = NULL, *aoffset7 = NULL, *aoffset8 = NULL;
+        __vector_pair C1, C2, C3, C4, C5, C6, C7, C8;
+        VA c1[2] = {0}, c2[2] = {0}, c3[2] = {0}, c4[2] = {0};
+        VA c5[2] = {0}, c6[2] = {0}, c7[2] = {0}, c8[2] = {0};
+        VA t1, t2, t3, t4, t5, t6, t7, t8;
+        aoffset = const_cast<TA*>(a);
         boffset = vec;
         j = (rows >> 3);
         if (j > 0) {
@@ -1092,9 +1795,6 @@ class tinyBLAS_PPC {
                 aoffset += 8 * lda;
                 i = (cols >> 3);
                 if (i > 0) {
-                    __vector_pair C1, C2, C3, C4, C5, C6, C7, C8;
-                    vector float c1[2], c2[2], c3[2], c4[2], c5[2], c6[2], c7[2], c8[2];
-                    vector float t1, t2, t3, t4, t5, t6, t7, t8;
                     do {
                         C1 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset1);
                         C2 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset2);
@@ -1174,21 +1874,19 @@ class tinyBLAS_PPC {
                     } while(i > 0);
                 }
                 if (cols & 4) {
-                    vector float c1, c2, c3, c4, c5, c6, c7, c8;
-                    vector float t1, t2, t3, t4, t5, t6, t7, t8;
-                    c1 = vec_xl(0, aoffset1);
-                    c2 = vec_xl(0, aoffset2);
-                    c3 = vec_xl(0, aoffset3);
-                    c4 = vec_xl(0, aoffset4);
-                    c5 = vec_xl(0, aoffset5);
-                    c6 = vec_xl(0, aoffset6);
-                    c7 = vec_xl(0, aoffset7);
-                    c8 = vec_xl(0, aoffset8);
+                    c1[0] = vec_xl(0, aoffset1);
+                    c2[0] = vec_xl(0, aoffset2);
+                    c3[0] = vec_xl(0, aoffset3);
+                    c4[0] = vec_xl(0, aoffset4);
+                    c5[0] = vec_xl(0, aoffset5);
+                    c6[0] = vec_xl(0, aoffset6);
+                    c7[0] = vec_xl(0, aoffset7);
+                    c8[0] = vec_xl(0, aoffset8);
 
-                    t1 = vec_mergeh(c1, c2);
-                    t2 = vec_mergeh(c3, c4);
-                    t3 = vec_mergeh(c5, c6);
-                    t4 = vec_mergeh(c7, c8);
+                    t1 = vec_mergeh(c1[0], c2[0]);
+                    t2 = vec_mergeh(c3[0], c4[0]);
+                    t3 = vec_mergeh(c5[0], c6[0]);
+                    t4 = vec_mergeh(c7[0], c8[0]);
                     t5 = vec_xxpermdi(t1, t2, 0);
                     t6 = vec_xxpermdi(t3, t4, 0);
                     t7 = vec_xxpermdi(t1, t2, 3);
@@ -1198,10 +1896,10 @@ class tinyBLAS_PPC {
                     vec_xst(t7, 0, boffset+8);
                     vec_xst(t8, 0, boffset+12);
 
-                    t1 = vec_mergel(c1, c2);
-                    t2 = vec_mergel(c3, c4);
-                    t3 = vec_mergel(c5, c6);
-                    t4 = vec_mergel(c7, c8);
+                    t1 = vec_mergel(c1[0], c2[0]);
+                    t2 = vec_mergel(c3[0], c4[0]);
+                    t3 = vec_mergel(c5[0], c6[0]);
+                    t4 = vec_mergel(c7[0], c8[0]);
                     t5 = vec_xxpermdi(t1, t2, 0);
                     t6 = vec_xxpermdi(t3, t4, 0);
                     t7 = vec_xxpermdi(t1, t2, 3);
@@ -1223,9 +1921,6 @@ class tinyBLAS_PPC {
             aoffset += 4 * lda;
             i = (cols >> 3);
             if (i > 0) {
-                __vector_pair C1, C2, C3, C4;
-                vector float c1[2], c2[2], c3[2], c4[2];
-                vector float t1, t2, t3, t4, t5, t6, t7, t8;
                 do {
                     C1 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset1);
                     C2 = __builtin_vsx_lxvp(0, (__vector_pair*)aoffset2);
@@ -1272,22 +1967,20 @@ class tinyBLAS_PPC {
             }
 
             if (cols & 4) {
-                vector float c1, c2, c3, c4;
-                vector float t1, t2, t3, t4;
-                c1 = vec_xl(0, aoffset1);
-                c2 = vec_xl(0, aoffset2);
-                c3 = vec_xl(0, aoffset3);
-                c4 = vec_xl(0, aoffset4);
+                c1[0] = vec_xl(0, aoffset1);
+                c2[0] = vec_xl(0, aoffset2);
+                c3[0] = vec_xl(0, aoffset3);
+                c4[0] = vec_xl(0, aoffset4);
 
-                t1 = vec_mergeh(c1, c2);
-                t2 = vec_mergeh(c3, c4);
+                t1 = vec_mergeh(c1[0], c2[0]);
+                t2 = vec_mergeh(c3[0], c4[0]);
                 t3 = vec_xxpermdi(t1, t2, 0);
                 t4 = vec_xxpermdi(t1, t2, 3);
                 vec_xst(t3, 0, boffset);
                 vec_xst(t4, 0, boffset+4);
 
-                t1 = vec_mergel(c1, c2);
-                t2 = vec_mergel(c3, c4);
+                t1 = vec_mergel(c1[0], c2[0]);
+                t2 = vec_mergel(c3[0], c4[0]);
                 t3 = vec_xxpermdi(t1, t2, 0);
                 t4 = vec_xxpermdi(t1, t2, 3);
                 vec_xst(t3, 0, boffset+8);
@@ -1299,21 +1992,19 @@ class tinyBLAS_PPC {
             aoffset2 = aoffset1 + lda;
             aoffset3 = aoffset2 + lda;
             if (cols & 4) {
-                vector float c1, c2, c3, c4 = {0};
-                vector float t1, t2, t3, t4;
-                c1 = vec_xl(0, aoffset1);
-                c2 = vec_xl(0, aoffset2);
-                c3 = vec_xl(0, aoffset3);
+                c1[0] = vec_xl(0, aoffset1);
+                c2[0] = vec_xl(0, aoffset2);
+                c3[0] = vec_xl(0, aoffset3);
 
-                t1 = vec_mergeh(c1, c2);
-                t2 = vec_mergeh(c3, c4);
+                t1 = vec_mergeh(c1[0], c2[0]);
+                t2 = vec_mergeh(c3[0], c4[0]);
                 t3 = vec_xxpermdi(t1, t2, 0);
                 t4 = vec_xxpermdi(t1, t2, 3);
                 vec_xst(t3, 0, boffset);
                 vec_xst(t4, 0, boffset+4);
 
-                t1 = vec_mergel(c1, c2);
-                t2 = vec_mergel(c3, c4);
+                t1 = vec_mergel(c1[0], c2[0]);
+                t2 = vec_mergel(c3[0], c4[0]);
                 t3 = vec_xxpermdi(t1, t2, 0);
                 t4 = vec_xxpermdi(t1, t2, 3);
                 vec_xst(t3, 0, boffset+8);
@@ -1321,14 +2012,13 @@ class tinyBLAS_PPC {
             }
         }
     }
-
     void KERNEL_4x4(int64_t ii, int64_t jj) {
         vec_t vec_A[4], vec_B[4], vec_C[4];
         acc_t acc_0;
         __builtin_mma_xxsetaccz(&acc_0);
         for (int l = 0; l < k; l+=4) {
-            READ_BLOCK(A+(ii*lda)+l, lda, 4, 4, (float*)vec_A);
-            READ_BLOCK(B+(jj*ldb)+l, ldb, 4, 4, (float*)vec_B);
+            packTranspose<vector float>(A+(ii*lda)+l, lda, 4, 4, (TA*)vec_A);
+            packTranspose<vector float>(B+(jj*ldb)+l, ldb, 4, 4, (TA*)vec_B);
             __builtin_mma_xvf32gerpp(&acc_0, vec_A[0], vec_B[0]);
             __builtin_mma_xvf32gerpp(&acc_0, vec_A[1], vec_B[1]);
             __builtin_mma_xvf32gerpp(&acc_0, vec_A[2], vec_B[2]);
@@ -1343,8 +2033,8 @@ class tinyBLAS_PPC {
         __builtin_mma_xxsetaccz(&acc_0);
         __builtin_mma_xxsetaccz(&acc_1);
         for (int64_t l = 0; l < k; l+=4) {
-            READ_BLOCK(A+(ii*lda)+l, lda, 4, 4, (float*)vec_A);
-            READ_BLOCK(B+(jj*ldb)+l, ldb, 8, 4, (float*)vec_B);
+            packTranspose<vector float>(A+(ii*lda)+l, lda, 4, 4, (TA*)vec_A);
+            packTranspose<vector float>(B+(jj*ldb)+l, ldb, 8, 4, (TA*)vec_B);
             __builtin_mma_xvf32gerpp(&acc_0, vec_A[0], (vec_t)vec_B[0]);
             __builtin_mma_xvf32gerpp(&acc_1, vec_A[0], (vec_t)vec_B[1]);
             __builtin_mma_xvf32gerpp(&acc_0, vec_A[1], (vec_t)vec_B[2]);
@@ -1364,8 +2054,8 @@ class tinyBLAS_PPC {
         __builtin_mma_xxsetaccz(&acc_0);
         __builtin_mma_xxsetaccz(&acc_1);
         for (int64_t l = 0; l < k; l+=4) {
-            READ_BLOCK(A+(ii*lda)+l, lda, 8, 4, (float*)vec_A);
-            READ_BLOCK(B+(jj*ldb)+l, ldb, 4, 4, (float*)vec_B);
+            packTranspose<vector float>(A+(ii*lda)+l, lda, 8, 4, (TA*)vec_A);
+            packTranspose<vector float>(B+(jj*ldb)+l, ldb, 4, 4, (TA*)vec_B);
             __builtin_mma_xvf32gerpp(&acc_0, (vec_t)vec_A[0], vec_B[0]);
             __builtin_mma_xvf32gerpp(&acc_1, (vec_t)vec_A[1], vec_B[0]);
             __builtin_mma_xvf32gerpp(&acc_0, (vec_t)vec_A[2], vec_B[1]);
@@ -1387,8 +2077,8 @@ class tinyBLAS_PPC {
         __builtin_mma_xxsetaccz(&acc_2);
         __builtin_mma_xxsetaccz(&acc_3);
         for (int l = 0; l < k; l+=8) {
-            READ_BLOCK(A+(ii*lda)+l, lda, 8, 8, (float*)vec_A);
-            READ_BLOCK(B+(jj*ldb)+l, ldb, 8, 8, (float*)vec_B);
+            packTranspose<vector float>(A+(ii*lda)+l, lda, 8, 8, (TA*)vec_A);
+            packTranspose<vector float>(B+(jj*ldb)+l, ldb, 8, 8, (TA*)vec_B);
             for(int x = 0; x < 16; x+=2) {
                 __builtin_mma_xvf32gerpp(&acc_0, (vec_t)vec_A[x], vec_B[x]);
                 __builtin_mma_xvf32gerpp(&acc_1, (vec_t)vec_A[x], vec_B[x+1]);
@@ -1571,15 +2261,15 @@ class tinyBLAS_PPC {
             vec_t vec_A[4], vec_B[4];
             for (int l=0; l<k; l+=4) {
                 if (RN >= 4 && RM == 1) {
-                    float* a = const_cast<float*>(A+(ii)*lda+l);
-                    READ_BLOCK(B+(jj*ldb)+l, ldb, 4, 4, (float*)vec_B);
+                    TA* a = const_cast<TA*>(A+(ii)*lda+l);
+                    packTranspose<vector float>(B+(jj*ldb)+l, ldb, 4, 4, (TA*)vec_B);
                     vec_A[0] = (vec_t)vec_xl(0,a);
-                    vec_A[1] = (vec_t)vec_splats(*((float*)&vec_A+1));
-                    vec_A[2] = (vec_t)vec_splats(*((float*)&vec_A+2));
-                    vec_A[3] = (vec_t)vec_splats(*((float*)&vec_A+3));
+                    vec_A[1] = (vec_t)vec_splats(*((TA*)&vec_A+1));
+                    vec_A[2] = (vec_t)vec_splats(*((TA*)&vec_A+2));
+                    vec_A[3] = (vec_t)vec_splats(*((TA*)&vec_A+3));
                 } else {
-                    READ_BLOCK(A+(ii*lda)+l, lda, RM, 4, (float*)vec_A);
-                    READ_BLOCK(B+(jj*ldb)+l, ldb, RN, 4, (float*)vec_B);
+                    packTranspose<vector float>(A+(ii*lda)+l, lda, RM, 4, (TA*)vec_A);
+                    packTranspose<vector float>(B+(jj*ldb)+l, ldb, RN, 4, (TA*)vec_B);
                 }
                 __builtin_mma_xvf32gerpp(&acc_0, vec_A[0], vec_B[0]);
                 __builtin_mma_xvf32gerpp(&acc_0, vec_A[1], vec_B[1]);
@@ -1589,7 +2279,7 @@ class tinyBLAS_PPC {
             __builtin_mma_disassemble_acc(vec_C, &acc_0);
             for (int I = 0; I < RM; I++) {
                 for (int J = 0; J < RN; J++) {
-                    *((float*)(C+ii+((jj+J)*ldc)+I)) = *((float*)&vec_C[I]+J);
+                    *((TC*)(C+ii+((jj+J)*ldc)+I)) = *((TC*)&vec_C[I]+J);
                 }
             }
        }
@@ -1812,6 +2502,20 @@ bool llamafile_sgemm(const struct ggml_compute_params * params, int64_t m, int64
             params->ith, params->nth};
         tb.matmul(m, n);
         return true;
+
+#elif defined(__MMA__)
+        if (n < 8 && n != 4)
+           return false;
+        if (m < 8 && m != 4)
+           return false;
+        tinyBLAS_Q0_PPC<block_q8_0, block_q8_0, float> tb{
+            k, (const block_q8_0 *)A, lda,
+            (const block_q8_0 *)B, ldb,
+            (float *)C, ldc,
+            params->ith, params->nth};
+        tb.matmul(m, n);
+        return true;
+
 #else
         return false;
 #endif

ggml/src/ggml-cuda/concat.cu

@@ -124,7 +124,7 @@ static __global__ void __launch_bounds__(CUDA_CONCAT_BLOCK_SIZE)
           uint64_t   nb1,
           uint64_t   nb2,
           uint64_t   nb3){
-    static_assert(dim >= 0 && dim <= 3);
+    static_assert(dim >= 0 && dim <= 3, "dim must be in [0, 3]");
 
     const int64_t i3 = blockIdx.z;
     const int64_t i2 = blockIdx.y;

ggml/src/ggml-cuda/ggml-cuda.cu

@@ -37,6 +37,7 @@
 #include "ggml-cuda/unary.cuh"
 #include "ggml-cuda/upscale.cuh"
 #include "ggml-cuda/wkv6.cuh"
+#include "ggml-cuda/gla.cuh"
 
 #include <algorithm>
 #include <array>
@@ -2140,6 +2141,9 @@ static bool ggml_cuda_compute_forward(ggml_backend_cuda_context & ctx, struct gg
         case GGML_OP_ROPE:
             ggml_cuda_op_rope(ctx, dst);
             break;
+        case GGML_OP_ROPE_BACK:
+            ggml_cuda_op_rope_back(ctx, dst);
+            break;
         case GGML_OP_IM2COL:
             ggml_cuda_op_im2col(ctx, dst);
             break;
@@ -2167,6 +2171,9 @@ static bool ggml_cuda_compute_forward(ggml_backend_cuda_context & ctx, struct gg
         case GGML_OP_RWKV_WKV6:
             ggml_cuda_op_rwkv_wkv6(ctx, dst);
             break;
+        case GGML_OP_GATED_LINEAR_ATTN:
+            ggml_cuda_op_gated_linear_attn(ctx, dst);
+            break;
         case GGML_OP_CROSS_ENTROPY_LOSS_BACK:
             ggml_cuda_cross_entropy_loss_back(ctx, dst);
             break;
@@ -2285,6 +2292,66 @@ static void ggml_backend_cuda_synchronize(ggml_backend_t backend) {
 }
 
 #ifdef USE_CUDA_GRAPH
+static bool check_node_graph_compatibility_and_refresh_copy_ops(ggml_backend_cuda_context * cuda_ctx, ggml_cgraph * cgraph,
+    std::vector<void *> & ggml_cuda_cpy_fn_ptrs, bool use_cuda_graph) {
+
+    // Loop over nodes in GGML graph to obtain info needed for CUDA graph
+    cuda_ctx->cuda_graph->updated_kernel_arg.clear();
+    for (int i = 0; i < cgraph->n_nodes; i++) {
+        ggml_tensor * node = cgraph->nodes[i];
+
+        if (ggml_is_empty(node) || node->op == GGML_OP_RESHAPE || node->op == GGML_OP_TRANSPOSE || node->op == GGML_OP_VIEW || node->op == GGML_OP_PERMUTE || node->op == GGML_OP_NONE) {
+            continue;
+        }
+
+        if (node->src[0] && node->src[0]->buffer && ggml_backend_buft_is_cuda_split(node->src[0]->buffer->buft)) {
+            use_cuda_graph = false; // Split buffers are not supported by CUDA graph capture
+#ifndef NDEBUG
+            GGML_LOG_DEBUG("%s: disabling CUDA graphs due to split buffer\n", __func__);
+#endif
+        }
+
+        if (node->op == GGML_OP_MUL_MAT_ID) {
+            use_cuda_graph = false; // This node type is not supported by CUDA graph capture
+#ifndef NDEBUG
+            GGML_LOG_DEBUG("%s: disabling CUDA graphs due to mul_mat_id\n", __func__);
+#endif
+        }
+
+        if (node->op == GGML_OP_ADD && node->src[1] && node->src[1]->ne[1] > 1) {
+            // disable CUDA graphs for batch size > 1 for now.
+            // Changes in batch size or context size can cause changes to the grid size of some kernels.
+            use_cuda_graph = false;
+#ifndef NDEBUG
+            GGML_LOG_DEBUG("%s: disabling CUDA graphs due to batch size > 1 [%s] [%ld %ld %ld %ld]\n", __func__, node->name, node->ne[0], node->ne[1], node->ne[2], node->ne[3]);
+#endif
+        }
+
+        if (node->op == GGML_OP_CPY) {
+            // store the copy op parameter which changes with each token.
+            cuda_ctx->cuda_graph->updated_kernel_arg.push_back((char **) &(node->src[1]->data));
+            // store a pointer to each copy op CUDA kernel to identify it later
+            void * ptr = ggml_cuda_cpy_fn(node->src[0], node->src[1]);
+            if (!ptr) {
+                use_cuda_graph = false;
+#ifndef NDEBUG
+                GGML_LOG_DEBUG("%s: disabling CUDA graphs due to unsupported copy op\n", __func__);
+#endif
+            } else {
+                if (std::find(ggml_cuda_cpy_fn_ptrs.begin(), ggml_cuda_cpy_fn_ptrs.end(), ptr) == ggml_cuda_cpy_fn_ptrs.end()) {
+                    ggml_cuda_cpy_fn_ptrs.push_back(ptr);
+                }
+            }
+        }
+
+        if (!use_cuda_graph) {
+            break;
+        }
+    }
+
+    return use_cuda_graph;
+}
+
 static void set_ggml_graph_node_properties(ggml_tensor * node, ggml_graph_node_properties * graph_node_properties) {
     graph_node_properties->node_address = node->data;
     graph_node_properties->node_op = node->op;
@@ -2335,149 +2402,105 @@ static bool ggml_graph_node_has_matching_properties(ggml_tensor * node, ggml_gra
 
     return true;
 }
-#endif
 
-static enum ggml_status ggml_backend_cuda_graph_compute(ggml_backend_t backend, ggml_cgraph * cgraph) {
-    ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context;
+static void maintain_cuda_graph(ggml_backend_cuda_context * cuda_ctx, std::vector<void *> & ggml_cuda_cpy_fn_ptrs, bool cuda_graph_update_required) {
 
-    ggml_cuda_set_device(cuda_ctx->device);
+    if (cuda_graph_update_required) {
+        // Extract nodes from graph
+        // First call with null argument gets number of nodes in graph
+        CUDA_CHECK(cudaGraphGetNodes(cuda_ctx->cuda_graph->graph, nullptr, &cuda_ctx->cuda_graph->num_nodes));
+        // Subsequent call with non-null argument gets nodes
+        cuda_ctx->cuda_graph->nodes.clear();
+        cuda_ctx->cuda_graph->nodes.resize(cuda_ctx->cuda_graph->num_nodes);
+        cuda_ctx->cuda_graph->params.clear();
+        cuda_ctx->cuda_graph->params.resize(cuda_ctx->cuda_graph->num_nodes);
+        if (cuda_ctx->cuda_graph->num_nodes > 0) {
+            CUDA_CHECK(cudaGraphGetNodes(cuda_ctx->cuda_graph->graph, cuda_ctx->cuda_graph->nodes.data(), &cuda_ctx->cuda_graph->num_nodes));
 
-#ifdef USE_CUDA_GRAPH
-    static const bool disable_cuda_graphs_due_to_env = (getenv("GGML_CUDA_DISABLE_GRAPHS") != nullptr);
-
-    // Objects required for CUDA Graph
-    if (cuda_ctx->cuda_graph == nullptr) {
-        cuda_ctx->cuda_graph.reset(new ggml_cuda_graph());
-    }
-
-    bool use_cuda_graph = true;
-    bool cuda_graph_update_required = false;
-    // vector of pointers to CUDA cpy kernels, which are required to identify
-    // kernel parameters which need updated in the graph for each token
-    std::vector<void *> ggml_cuda_cpy_fn_ptrs;
-
-    if (cuda_ctx->cuda_graph->graph == nullptr) {
-        if (ggml_cuda_info().devices[cuda_ctx->device].cc < GGML_CUDA_CC_AMPERE) {
-            cuda_ctx->cuda_graph->disable_due_to_gpu_arch = true;
-#ifndef NDEBUG
-            GGML_LOG_DEBUG("%s: disabling CUDA graphs due to GPU architecture\n", __func__);
-#endif
-        }
-    }
-
-    // Disable CUDA graphs in presence of env var, old GPU, use-case which is changing too rapidly,
-    // or previous graph capture failure.
-    // Also disable for multi-gpu for now. TO DO investigate
-    if (disable_cuda_graphs_due_to_env
-        || cuda_ctx->cuda_graph->disable_due_to_gpu_arch
-        || cuda_ctx->cuda_graph->disable_due_to_too_many_updates
-        || cuda_ctx->cuda_graph->disable_due_to_failed_graph_capture) {
-        use_cuda_graph = false;
-    }
-
-    if (use_cuda_graph) {
-        if (cuda_ctx->cuda_graph->instance == nullptr) {
-            cuda_graph_update_required = true;
-        }
-
-        // Check if the graph size has changed
-        if (cuda_ctx->cuda_graph->ggml_graph_properties.size() != (size_t)cgraph->n_nodes) {
-            cuda_graph_update_required = true;
-            cuda_ctx->cuda_graph->ggml_graph_properties.resize(cgraph->n_nodes);
-        }
-
-        // Loop over nodes in GGML graph to determine if CUDA graph update is required
-        // and store properties to allow this comparison for the next token
-        for (int i = 0; i < cgraph->n_nodes; i++) {
-            bool has_matching_properties = true;
-            if (!cuda_graph_update_required) {
-                has_matching_properties = ggml_graph_node_has_matching_properties(cgraph->nodes[i], &cuda_ctx->cuda_graph->ggml_graph_properties[i]);
-            }
-            if (!has_matching_properties) {
-                cuda_graph_update_required = true;
-            }
-            set_ggml_graph_node_properties(cgraph->nodes[i], &cuda_ctx->cuda_graph->ggml_graph_properties[i]);
-        }
-
-        // Loop over nodes in GGML graph to obtain info needed for CUDA graph
-        cuda_ctx->cuda_graph->updated_kernel_arg.clear();
-        for (int i = 0; i < cgraph->n_nodes; i++) {
-            ggml_tensor * node = cgraph->nodes[i];
-
-            if (ggml_is_empty(node) || node->op == GGML_OP_RESHAPE || node->op == GGML_OP_TRANSPOSE || node->op == GGML_OP_VIEW || node->op == GGML_OP_PERMUTE || node->op == GGML_OP_NONE) {
-                continue;
-            }
-
-            if (node->src[0] && node->src[0]->buffer && ggml_backend_buft_is_cuda_split(node->src[0]->buffer->buft)) {
-                use_cuda_graph = false; // Split buffers are not supported by CUDA graph capture
-#ifndef NDEBUG
-                GGML_LOG_DEBUG("%s: disabling CUDA graphs due to split buffer\n", __func__);
-#endif
-            }
-
-            if (node->op == GGML_OP_MUL_MAT_ID) {
-                use_cuda_graph = false; // This node type is not supported by CUDA graph capture
-#ifndef NDEBUG
-                GGML_LOG_DEBUG("%s: disabling CUDA graphs due to mul_mat_id\n", __func__);
-#endif
-            }
-
-            if (node->op == GGML_OP_ADD && node->src[1] && node->src[1]->ne[1] > 1) {
-                // disable CUDA graphs for batch size > 1 for now.
-                // Changes in batch size or context size can cause changes to the grid size of some kernels.
-                use_cuda_graph = false;
-#ifndef NDEBUG
-                GGML_LOG_DEBUG("%s: disabling CUDA graphs due to batch size > 1 [%s] [%ld %ld %ld %ld]\n", __func__, node->name, node->ne[0], node->ne[1], node->ne[2], node->ne[3]);
-#endif
-            }
-
-            if (node->op == GGML_OP_CPY) {
-                // store the copy op parameter which changes with each token.
-                cuda_ctx->cuda_graph->updated_kernel_arg.push_back((char **) &(node->src[1]->data));
-                // store a pointer to each copy op CUDA kernel to identify it later
-                void * ptr = ggml_cuda_cpy_fn(node->src[0], node->src[1]);
-                if (!ptr) {
-                    use_cuda_graph = false;
-#ifndef NDEBUG
-                    GGML_LOG_DEBUG("%s: disabling CUDA graphs due to unsupported copy op\n", __func__);
-#endif
-                } else {
-                    if (std::find(ggml_cuda_cpy_fn_ptrs.begin(), ggml_cuda_cpy_fn_ptrs.end(), ptr) == ggml_cuda_cpy_fn_ptrs.end()) {
-                        ggml_cuda_cpy_fn_ptrs.push_back(ptr);
+            // Loop over nodes, and extract kernel parameters from each node
+            for (size_t i = 0; i < cuda_ctx->cuda_graph->num_nodes; i++) {
+                cudaGraphNodeType node_type;
+                CUDA_CHECK(cudaGraphNodeGetType(cuda_ctx->cuda_graph->nodes[i], &node_type));
+                if (node_type == cudaGraphNodeTypeKernel) {
+                    cudaError_t stat = cudaGraphKernelNodeGetParams(cuda_ctx->cuda_graph->nodes[i], &cuda_ctx->cuda_graph->params[i]); // Get params using runtime
+                    if (stat == cudaErrorInvalidDeviceFunction) {
+                        // Fails due to incorrect handling by CUDA runtime of CUDA BLAS node.
+                        // We don't need to update blas nodes, so clear error and move on.
+                        cudaGetLastError();
+                    } else {
+                        GGML_ASSERT(stat == cudaSuccess);
                     }
                 }
             }
-
-            if (!use_cuda_graph) {
-                break;
+        }
+    } else {
+        // One of the arguments to the copy kernel is updated for each token, hence we need to
+        // replace that argument with the updated value in the CUDA graph
+        // on update steps, the live parameters will already be captured
+        int k = 0;
+        for (size_t i = 0; i < cuda_ctx->cuda_graph->num_nodes; i++) {
+            if(count(ggml_cuda_cpy_fn_ptrs.begin(), ggml_cuda_cpy_fn_ptrs.end(), cuda_ctx->cuda_graph->params[i].func) > 0) {
+                char ** updated_kernel_arg_ptr = cuda_ctx->cuda_graph->updated_kernel_arg.at(k++);
+                cuda_ctx->cuda_graph->params[i].kernelParams[1] = updated_kernel_arg_ptr;
+                CUDA_CHECK(cudaGraphKernelNodeSetParams(cuda_ctx->cuda_graph->nodes[i], &cuda_ctx->cuda_graph->params[i]));
             }
         }
-
-        // Disable CUDA graphs (from the next token) if the use-case is demanding too many consecutive graph updates.
-        if (use_cuda_graph && cuda_graph_update_required) {
-            cuda_ctx->cuda_graph->number_consecutive_updates++;
-        } else {
-            cuda_ctx->cuda_graph->number_consecutive_updates = 0;
-        }
-
-        if (cuda_ctx->cuda_graph->number_consecutive_updates >= 4) {
-            cuda_ctx->cuda_graph->disable_due_to_too_many_updates = true;
-#ifndef NDEBUG
-            GGML_LOG_DEBUG("%s: disabling CUDA graphs due to too many consecutive updates\n", __func__);
-#endif
-        }
     }
+}
 
-    if (use_cuda_graph && cuda_graph_update_required) { // Start CUDA graph capture
-        CUDA_CHECK(cudaStreamBeginCapture(cuda_ctx->stream(), cudaStreamCaptureModeRelaxed));
-    }
+static bool is_cuda_graph_update_required(ggml_backend_cuda_context * cuda_ctx, ggml_cgraph * cgraph) {
 
-#else
-    bool use_cuda_graph = false;
     bool cuda_graph_update_required = false;
-#endif // USE_CUDA_GRAPH
 
-    bool graph_evaluated_or_captured = false;
+    if (cuda_ctx->cuda_graph->instance == nullptr) {
+        cuda_graph_update_required = true;
+    }
+
+    // Check if the graph size has changed
+    if (cuda_ctx->cuda_graph->ggml_graph_properties.size() != (size_t)cgraph->n_nodes) {
+        cuda_graph_update_required = true;
+        cuda_ctx->cuda_graph->ggml_graph_properties.resize(cgraph->n_nodes);
+    }
+
+    // Loop over nodes in GGML graph to determine if CUDA graph update is required
+    // and store properties to allow this comparison for the next token
+    for (int i = 0; i < cgraph->n_nodes; i++) {
+        bool has_matching_properties = true;
+        if (!cuda_graph_update_required) {
+            has_matching_properties = ggml_graph_node_has_matching_properties(cgraph->nodes[i], &cuda_ctx->cuda_graph->ggml_graph_properties[i]);
+        }
+        if (!has_matching_properties) {
+            cuda_graph_update_required = true;
+        }
+        set_ggml_graph_node_properties(cgraph->nodes[i], &cuda_ctx->cuda_graph->ggml_graph_properties[i]);
+    }
+
+    return cuda_graph_update_required;
+}
+
+static void update_cuda_graph_executable(ggml_backend_cuda_context * cuda_ctx) {
+
+    cudaGraphExecUpdateResultInfo result_info;
+    cudaError_t stat = cudaGraphExecUpdate(cuda_ctx->cuda_graph->instance, cuda_ctx->cuda_graph->graph, &result_info);
+    if (stat == cudaErrorGraphExecUpdateFailure) {
+#ifndef NDEBUG
+        GGML_LOG_DEBUG("%s: CUDA graph update failed\n", __func__);
+#endif
+        // The pre-existing graph exec cannot be updated due to violated constraints
+        // so instead clear error and re-instantiate
+        cudaGetLastError();
+        CUDA_CHECK(cudaGraphExecDestroy(cuda_ctx->cuda_graph->instance));
+        cuda_ctx->cuda_graph->instance = nullptr;
+        CUDA_CHECK(cudaGraphInstantiate(&cuda_ctx->cuda_graph->instance, cuda_ctx->cuda_graph->graph, NULL, NULL, 0));
+    } else {
+        GGML_ASSERT(stat == cudaSuccess);
+    }
+}
+#endif
+
+static void evaluate_and_capture_cuda_graph(ggml_backend_cuda_context * cuda_ctx, ggml_cgraph * cgraph,
+   [[maybe_unused]] std::vector<void *> & ggml_cuda_cpy_fn_ptrs,  bool & graph_evaluated_or_captured, bool & use_cuda_graph,
+    bool & cuda_graph_update_required) {
 
     while (!graph_evaluated_or_captured) {
         // Only perform the graph execution if CUDA graphs are not enabled, or we are capturing the graph.
@@ -2515,19 +2538,8 @@ static enum ggml_status ggml_backend_cuda_graph_compute(ggml_backend_t backend,
                 CUDA_CHECK(cudaGraphDestroy(cuda_ctx->cuda_graph->graph));
                 cuda_ctx->cuda_graph->graph = nullptr;
             }
-            CUDA_CHECK(cudaStreamEndCapture(cuda_ctx->stream(), &cuda_ctx->cuda_graph->graph));
 
-#if 0
-            if (disable_cuda_graphs_due_to_failed_capture) {
-                use_cuda_graph = false;
-                cuda_ctx->cuda_graph->disable_due_to_failed_graph_capture = true;
-#ifndef NDEBUG
-                GGML_LOG_DEBUG("%s: disabling CUDA graphs due to failed graph capture\n", __func__);
-#endif
-            } else {
-                graph_evaluated_or_captured = true; // CUDA graph has been captured
-            }
-#endif
+            CUDA_CHECK(cudaStreamEndCapture(cuda_ctx->stream(), &cuda_ctx->cuda_graph->graph));
             graph_evaluated_or_captured = true; // CUDA graph has been captured
         } else {
             graph_evaluated_or_captured = true; // ggml graph has been directly evaluated
@@ -2540,72 +2552,91 @@ static enum ggml_status ggml_backend_cuda_graph_compute(ggml_backend_t backend,
         }
 
         // Perform update to graph (if required for this token), and change copy parameter (required for every token)
-
-        if (cuda_graph_update_required) {
-            // Extract nodes from graph
-            // First call with null argument gets number of nodes in graph
-            CUDA_CHECK(cudaGraphGetNodes(cuda_ctx->cuda_graph->graph, nullptr, &cuda_ctx->cuda_graph->num_nodes));
-            // Subsequent call with non-null argument gets nodes
-            cuda_ctx->cuda_graph->nodes.clear();
-            cuda_ctx->cuda_graph->nodes.resize(cuda_ctx->cuda_graph->num_nodes);
-            cuda_ctx->cuda_graph->params.clear();
-            cuda_ctx->cuda_graph->params.resize(cuda_ctx->cuda_graph->num_nodes);
-            if (cuda_ctx->cuda_graph->num_nodes > 0) {
-                CUDA_CHECK(cudaGraphGetNodes(cuda_ctx->cuda_graph->graph, cuda_ctx->cuda_graph->nodes.data(), &cuda_ctx->cuda_graph->num_nodes));
-
-                // Loop over nodes, and extract kernel parameters from each node
-                for (size_t i = 0; i < cuda_ctx->cuda_graph->num_nodes; i++) {
-                    cudaGraphNodeType node_type;
-                    CUDA_CHECK(cudaGraphNodeGetType(cuda_ctx->cuda_graph->nodes[i], &node_type));
-                    if (node_type == cudaGraphNodeTypeKernel) {
-                        cudaError_t stat = cudaGraphKernelNodeGetParams(cuda_ctx->cuda_graph->nodes[i], &cuda_ctx->cuda_graph->params[i]); // Get params using runtime
-                        if (stat == cudaErrorInvalidDeviceFunction) {
-                            // Fails due to incorrect handling by CUDA runtime of CUDA BLAS node.
-                            // We don't need to update blas nodes, so clear error and move on.
-                            cudaGetLastError();
-                        } else {
-                            GGML_ASSERT(stat == cudaSuccess);
-                        }
-                    }
-                }
-            }
-        }
-
-        // One of the arguments to the copy kernel is updated for each token, hence we need to
-        // replace that argument with the updated value in the CUDA graph
-        if (!cuda_graph_update_required) { // on update steps, the live parameters will already be captured
-            int k = 0;
-            for (size_t i = 0; i < cuda_ctx->cuda_graph->num_nodes; i++) {
-                if(count(ggml_cuda_cpy_fn_ptrs.begin(), ggml_cuda_cpy_fn_ptrs.end(), cuda_ctx->cuda_graph->params[i].func) > 0) {
-                    char ** updated_kernel_arg_ptr = cuda_ctx->cuda_graph->updated_kernel_arg.at(k++);
-                    cuda_ctx->cuda_graph->params[i].kernelParams[1] = updated_kernel_arg_ptr;
-                    CUDA_CHECK(cudaGraphKernelNodeSetParams(cuda_ctx->cuda_graph->nodes[i], &cuda_ctx->cuda_graph->params[i]));
-                }
-            }
-        }
+        maintain_cuda_graph(cuda_ctx, ggml_cuda_cpy_fn_ptrs, cuda_graph_update_required);
 
         // Update graph executable
-        cudaGraphExecUpdateResultInfo result_info;
-        cudaError_t stat = cudaGraphExecUpdate(cuda_ctx->cuda_graph->instance, cuda_ctx->cuda_graph->graph, &result_info);
-        if (stat == cudaErrorGraphExecUpdateFailure) {
-#ifndef NDEBUG
-            GGML_LOG_DEBUG("%s: CUDA graph update failed\n", __func__);
-#endif
-            // The pre-existing graph exec cannot be updated due to violated constraints
-            // so instead clear error and re-instantiate
-            cudaGetLastError();
-            CUDA_CHECK(cudaGraphExecDestroy(cuda_ctx->cuda_graph->instance));
-            cuda_ctx->cuda_graph->instance = nullptr;
-            CUDA_CHECK(cudaGraphInstantiate(&cuda_ctx->cuda_graph->instance, cuda_ctx->cuda_graph->graph, NULL, NULL, 0));
-        } else {
-            GGML_ASSERT(stat == cudaSuccess);
-        }
+        update_cuda_graph_executable(cuda_ctx);
+
         // Launch graph
         CUDA_CHECK(cudaGraphLaunch(cuda_ctx->cuda_graph->instance, cuda_ctx->stream()));
 #else
         graph_evaluated_or_captured = true;
-#endif // USE_CUDA_GRAPH
+#endif  // USE_CUDA_GRAPH
     }
+}
+
+static enum ggml_status ggml_backend_cuda_graph_compute(ggml_backend_t backend, ggml_cgraph * cgraph) {
+    ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context;
+
+    ggml_cuda_set_device(cuda_ctx->device);
+
+    // vector of pointers to CUDA cpy kernels, which are required to identify
+    // kernel parameters which need updated in the graph for each token
+    std::vector<void *> ggml_cuda_cpy_fn_ptrs;
+
+#ifdef USE_CUDA_GRAPH
+    static const bool disable_cuda_graphs_due_to_env = (getenv("GGML_CUDA_DISABLE_GRAPHS") != nullptr);
+
+    // Objects required for CUDA Graph
+    if (cuda_ctx->cuda_graph == nullptr) {
+        cuda_ctx->cuda_graph.reset(new ggml_cuda_graph());
+    }
+
+    bool use_cuda_graph = true;
+    bool cuda_graph_update_required = false;
+
+    if (cuda_ctx->cuda_graph->graph == nullptr) {
+        if (ggml_cuda_info().devices[cuda_ctx->device].cc < GGML_CUDA_CC_AMPERE) {
+            cuda_ctx->cuda_graph->disable_due_to_gpu_arch = true;
+#ifndef NDEBUG
+            GGML_LOG_DEBUG("%s: disabling CUDA graphs due to GPU architecture\n", __func__);
+#endif
+        }
+    }
+
+    // Disable CUDA graphs in presence of env var, old GPU, use-case which is changing too rapidly,
+    // or previous graph capture failure.
+    // Also disable for multi-gpu for now. TO DO investigate
+    if (disable_cuda_graphs_due_to_env
+        || cuda_ctx->cuda_graph->disable_due_to_gpu_arch
+        || cuda_ctx->cuda_graph->disable_due_to_too_many_updates
+        || cuda_ctx->cuda_graph->disable_due_to_failed_graph_capture) {
+        use_cuda_graph = false;
+    }
+
+    if (use_cuda_graph) {
+        cuda_graph_update_required = is_cuda_graph_update_required(cuda_ctx, cgraph);
+
+        use_cuda_graph = check_node_graph_compatibility_and_refresh_copy_ops(cuda_ctx, cgraph,
+                             ggml_cuda_cpy_fn_ptrs, use_cuda_graph);
+
+        // Disable CUDA graphs (from the next token) if the use-case is demanding too many consecutive graph updates.
+        if (use_cuda_graph && cuda_graph_update_required) {
+            cuda_ctx->cuda_graph->number_consecutive_updates++;
+        } else {
+            cuda_ctx->cuda_graph->number_consecutive_updates = 0;
+        }
+
+        if (cuda_ctx->cuda_graph->number_consecutive_updates >= 4) {
+            cuda_ctx->cuda_graph->disable_due_to_too_many_updates = true;
+#ifndef NDEBUG
+            GGML_LOG_DEBUG("%s: disabling CUDA graphs due to too many consecutive updates\n", __func__);
+#endif
+        }
+    }
+
+    if (use_cuda_graph && cuda_graph_update_required) { // Start CUDA graph capture
+        CUDA_CHECK(cudaStreamBeginCapture(cuda_ctx->stream(), cudaStreamCaptureModeRelaxed));
+    }
+
+#else
+    bool use_cuda_graph = false;
+    bool cuda_graph_update_required = false;
+#endif // USE_CUDA_GRAPH
+
+    bool graph_evaluated_or_captured = false;
+
+    evaluate_and_capture_cuda_graph(cuda_ctx, cgraph, ggml_cuda_cpy_fn_ptrs, graph_evaluated_or_captured, use_cuda_graph, cuda_graph_update_required);
 
     return GGML_STATUS_SUCCESS;
 }
@@ -2997,7 +3028,11 @@ static bool ggml_backend_cuda_device_supports_op(ggml_backend_dev_t dev, const g
         case GGML_OP_SOFT_MAX:
             return true;
         case GGML_OP_ROPE:
-            return ggml_is_contiguous(op->src[0]);
+        case GGML_OP_ROPE_BACK: {
+            const size_t ts = ggml_type_size(op->src[0]->type);
+            const int64_t ne0_012 = op->src[0]->ne[0] * op->src[0]->ne[1] * op->src[0]->ne[2];
+            return op->src[0]->nb[0] == ts && op->src[0]->nb[3] == ne0_012*ts;
+        }
         case GGML_OP_IM2COL:
         case GGML_OP_POOL_2D:
         case GGML_OP_SUM:
@@ -3011,6 +3046,7 @@ static bool ggml_backend_cuda_device_supports_op(ggml_backend_dev_t dev, const g
         case GGML_OP_TIMESTEP_EMBEDDING:
         case GGML_OP_LEAKY_RELU:
         case GGML_OP_RWKV_WKV6:
+        case GGML_OP_GATED_LINEAR_ATTN:
             return true;
         case GGML_OP_FLASH_ATTN_EXT: {
 #ifndef FLASH_ATTN_AVAILABLE
@@ -3052,6 +3088,7 @@ static int64_t get_op_batch_size(const ggml_tensor * op) {
             return op->ne[1];
         case GGML_OP_MUL_MAT_ID:
         case GGML_OP_ROPE:
+        case GGML_OP_ROPE_BACK:
             return op->ne[2];
         default:
             return ggml_nrows(op);

ggml/src/ggml-cuda/gla.cu

@@ -0,0 +1,93 @@
+#include "common.cuh"
+#include "gla.cuh"
+
+template<int HEAD_SIZE>
+static __global__ void gated_linear_attn_f32(const int B, const int T, const int C, const int H, const float scale,
+     const float * k, const float * v, const float * r, const float * td, const float * s, float * dst) {
+    const int tid = threadIdx.x;
+    const int bid = blockIdx.x;
+
+    const int head_size = HEAD_SIZE;
+    const int batch_i = bid / H;
+    const int head_i = bid % H;
+    const int state_size = C * head_size;
+    const int n_seq_tokens = T / B;
+
+    float state[head_size];
+    __shared__ float _k[head_size], _r[head_size], _td[head_size];
+
+    #pragma unroll
+    for (int i = 0; i < head_size; i++) {
+        state[i] = s[batch_i * state_size + head_i * head_size * head_size + i * head_size + tid];
+    }
+
+    for (int t = batch_i * n_seq_tokens * C + head_i * head_size + tid; t < (batch_i + 1) * n_seq_tokens * C + head_i * head_size + tid; t += C) {
+        __syncthreads();
+        _k[tid] = k[t];
+        _r[tid] = r[t];
+        _td[tid] = td[t];
+        __syncthreads();
+
+        const float _v = v[t];
+        float y = 0;
+        for (int j = 0; j < head_size; j += 4) {
+            const float4 & k = (float4 &)(_k[j]);
+            const float4 & r = (float4 &)(_r[j]);
+            const float4 & td = (float4 &)(_td[j]);
+            float4 & s = (float4 &)(state[j]);
+            float4 kv;
+
+            kv.x = k.x * _v;
+            kv.y = k.y * _v;
+            kv.z = k.z * _v;
+            kv.w = k.w * _v;
+
+            s.x = s.x * td.x + kv.x;
+            s.y = s.y * td.y + kv.y;
+            s.z = s.z * td.z + kv.z;
+            s.w = s.w * td.w + kv.w;
+
+            y += r.x * s.x;
+            y += r.y * s.y;
+            y += r.z * s.z;
+            y += r.w * s.w;
+        }
+        dst[t] = y * scale;
+    }
+
+    #pragma unroll
+    for (int i = 0; i < head_size; i++) {
+        dst[T * C + batch_i * state_size + head_i * head_size * head_size + i * head_size + tid] = state[i];
+    }
+}
+
+void ggml_cuda_op_gated_linear_attn(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
+    const float * k_d  = (const float *)dst->src[0]->data;
+    const float * v_d  = (const float *)dst->src[1]->data;
+    const float * r_d  = (const float *)dst->src[2]->data;
+    const float * td_d = (const float *)dst->src[3]->data;
+    const float * s_d  = (const float *)dst->src[4]->data;
+
+    const int64_t B = dst->src[4]->ne[1];
+    const int64_t T = dst->src[0]->ne[2];
+    const int64_t C = dst->ne[0];
+    const int64_t H = dst->src[0]->ne[1];
+
+    float scale;
+    memcpy(&scale, (float*)dst->op_params, sizeof(float));
+
+    float * dst_d = (float *)dst->data;
+
+    cudaStream_t stream = ctx.stream();
+
+    GGML_ASSERT(dst->src[4]->type == GGML_TYPE_F32);
+    GGML_ASSERT(C % H == 0);
+    GGML_ASSERT(C / H == 64 || C / H == 128);
+
+
+    if (C / H == 64) {
+        gated_linear_attn_f32<64><<<B * H, C / H, 0, stream>>>(B, T, C, H, scale, k_d, v_d, r_d, td_d, s_d, dst_d);
+    } else {
+        gated_linear_attn_f32<128><<<B * H, C / H, 0, stream>>>(B, T, C, H, scale, k_d, v_d, r_d, td_d, s_d, dst_d);
+    }
+}

ggml/src/ggml-cuda/gla.cuh

@@ -0,0 +1,3 @@
+#include "common.cuh"
+
+void ggml_cuda_op_gated_linear_attn(ggml_backend_cuda_context & ctx, ggml_tensor * dst);

ggml/src/ggml-cuda/rope.cu

@@ -16,9 +16,10 @@ static __device__ float rope_yarn_ramp(const float low, const float high, const
 
 // YaRN algorithm based on LlamaYaRNScaledRotaryEmbedding.py from https://github.com/jquesnelle/yarn
 // MIT licensed. Copyright (c) 2023 Jeffrey Quesnelle and Bowen Peng.
+template<bool forward>
 static __device__ void rope_yarn(
-    float theta_extrap, float freq_scale, rope_corr_dims corr_dims, int64_t i0, float ext_factor, float mscale,
-    float * cos_theta, float * sin_theta) {
+        const float theta_extrap, const float freq_scale, const rope_corr_dims corr_dims, const int64_t i0, const float ext_factor,
+        float mscale, float & cos_theta, float & sin_theta) {
     // Get n-d rotational scaling corrected for extrapolation
     float theta_interp = freq_scale * theta_extrap;
     float theta = theta_interp;
@@ -29,24 +30,28 @@ static __device__ void rope_yarn(
         // Get n-d magnitude scaling corrected for interpolation
         mscale *= 1.0f + 0.1f * logf(1.0f / freq_scale);
     }
-    *cos_theta = cosf(theta) * mscale;
-    *sin_theta = sinf(theta) * mscale;
+    cos_theta = cosf(theta) * mscale;
+    sin_theta = sinf(theta) * mscale;
+    if (!forward) {
+        sin_theta *= -1.0f;
+    }
 }
 
-template<typename T, bool has_ff>
+template<bool forward, bool has_ff, typename T>
 static __global__ void rope_norm(
-    const T * x, T * dst, int ne0, int n_dims, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float ext_factor, float attn_factor, rope_corr_dims corr_dims, float theta_scale, const float * freq_factors) {
+        const T * __restrict__ x, T * __restrict__ dst, const int ne0, const int ne1, const int s1, const int s2, const int n_dims,
+        const int32_t * __restrict__ pos, const float freq_scale, const float ext_factor, const float attn_factor,
+        const rope_corr_dims corr_dims, const float theta_scale, const float * __restrict__ freq_factors) {
     const int i0 = 2*(blockDim.y*blockIdx.y + threadIdx.y);
 
     if (i0 >= ne0) {
         return;
     }
 
-    const int row = blockDim.x*blockIdx.x + threadIdx.x;
+    const int row_dst = blockDim.x*blockIdx.x + threadIdx.x;
 
     if (i0 >= n_dims) {
-        const int i = row*ne0 + i0;
+        const int i = row_dst*ne0 + i0;
 
         dst[i + 0] = x[i + 0];
         dst[i + 1] = x[i + 1];
@@ -54,39 +59,43 @@ static __global__ void rope_norm(
         return;
     }
 
-    const int i  = row*ne0 + i0;
-    const int i2 = row/p_delta_rows;
+    const int row_x     = row_dst % ne1;
+    const int channel_x = row_dst / ne1;
 
-    const float theta_base = pos[i2]*powf(theta_scale, i0/2.0f);
+    const int idst = row_dst*ne0 + i0;
+    const int ix   = channel_x*s2 + row_x*s1 + i0;
+
+    const float theta_base = pos[channel_x]*powf(theta_scale, i0/2.0f);
 
     const float freq_factor = has_ff ? freq_factors[i0/2] : 1.0f;
 
     float cos_theta;
     float sin_theta;
 
-    rope_yarn(theta_base/freq_factor, freq_scale, corr_dims, i0, ext_factor, attn_factor, &cos_theta, &sin_theta);
+    rope_yarn<forward>(theta_base/freq_factor, freq_scale, corr_dims, i0, ext_factor, attn_factor, cos_theta, sin_theta);
 
-    const float x0 = x[i + 0];
-    const float x1 = x[i + 1];
+    const float x0 = x[ix + 0];
+    const float x1 = x[ix + 1];
 
-    dst[i + 0] = x0*cos_theta - x1*sin_theta;
-    dst[i + 1] = x0*sin_theta + x1*cos_theta;
+    dst[idst + 0] = x0*cos_theta - x1*sin_theta;
+    dst[idst + 1] = x0*sin_theta + x1*cos_theta;
 }
 
-template<typename T, bool has_ff>
+template<bool forward, bool has_ff, typename T>
 static __global__ void rope_neox(
-    const T * x, T * dst, int ne0, int n_dims, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float ext_factor, float attn_factor, rope_corr_dims corr_dims, float theta_scale, const float * freq_factors) {
+        const T * __restrict__ x, T * __restrict__ dst, const int ne0, const int ne1, const int s1, const int s2, const int n_dims,
+        const int32_t * __restrict__ pos, const float freq_scale, const float ext_factor, const float attn_factor,
+        const rope_corr_dims corr_dims, const float theta_scale, const float * __restrict__ freq_factors) {
     const int i0 = 2*(blockDim.y*blockIdx.y + threadIdx.y);
 
     if (i0 >= ne0) {
         return;
     }
 
-    const int row = blockDim.x*blockIdx.x + threadIdx.x;
+    const int row_dst = blockDim.x*blockIdx.x + threadIdx.x;
 
     if (i0 >= n_dims) {
-        const int i = row*ne0 + i0;
+        const int i = row_dst*ne0 + i0;
 
         dst[i + 0] = x[i + 0];
         dst[i + 1] = x[i + 1];
@@ -94,39 +103,43 @@ static __global__ void rope_neox(
         return;
     }
 
-    const int i  = row*ne0 + i0/2;
-    const int i2 = row/p_delta_rows;
+    const int row_x     = row_dst % ne1;
+    const int channel_x = row_dst / ne1;
 
-    const float theta_base = pos[i2]*powf(theta_scale, i0/2.0f);
+    const int idst = row_dst*ne0 + i0/2;
+    const int ix   = channel_x*s2 + row_x*s1 + i0/2;
+
+    const float theta_base = pos[channel_x]*powf(theta_scale, i0/2.0f);
 
     const float freq_factor = has_ff ? freq_factors[i0/2] : 1.0f;
 
     float cos_theta;
     float sin_theta;
 
-    rope_yarn(theta_base/freq_factor, freq_scale, corr_dims, i0, ext_factor, attn_factor, &cos_theta, &sin_theta);
+    rope_yarn<forward>(theta_base/freq_factor, freq_scale, corr_dims, i0, ext_factor, attn_factor, cos_theta, sin_theta);
 
-    const float x0 = x[i + 0];
-    const float x1 = x[i + n_dims/2];
+    const float x0 = x[ix + 0];
+    const float x1 = x[ix + n_dims/2];
 
-    dst[i + 0]        = x0*cos_theta - x1*sin_theta;
-    dst[i + n_dims/2] = x0*sin_theta + x1*cos_theta;
+    dst[idst + 0]        = x0*cos_theta - x1*sin_theta;
+    dst[idst + n_dims/2] = x0*sin_theta + x1*cos_theta;
 }
 
-template<typename T, bool has_ff>
+template<bool forward, bool has_ff, typename T>
 static __global__ void rope_multi(
-    const T * x, T * dst, int ne0, int ne2, int n_dims, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float ext_factor, float attn_factor, rope_corr_dims corr_dims, float theta_scale, const float * freq_factors, mrope_sections sections) {
+        const T * __restrict__ x, T * __restrict__ dst, const int ne0, const int ne1, const int ne2, const int s1, const int s2,
+        const int n_dims, const int32_t * __restrict__ pos, const float freq_scale, const float ext_factor, const float attn_factor,
+        const rope_corr_dims corr_dims, const float theta_scale, const float * __restrict__ freq_factors, const mrope_sections sections) {
     const int i0 = 2*(blockDim.y*blockIdx.y + threadIdx.y);
 
     if (i0 >= ne0) {
         return;
     }
 
-    const int row = blockDim.x*blockIdx.x + threadIdx.x;
+    const int row_dst = blockDim.x*blockIdx.x + threadIdx.x;
 
     if (i0 >= n_dims) {
-        const int i = row*ne0 + i0;
+        const int i = row_dst*ne0 + i0;
 
         dst[i + 0] = x[i + 0];
         dst[i + 1] = x[i + 1];
@@ -134,25 +147,28 @@ static __global__ void rope_multi(
         return;
     }
 
-    const int i  = row*ne0 + i0/2;
-    const int i2 = row/p_delta_rows;
+    const int row_x     = row_dst % ne1;
+    const int channel_x = row_dst / ne1;
 
-    int sect_dims = sections.v[0] + sections.v[1] + sections.v[2] + sections.v[3];
-    int sec_w = sections.v[1] + sections.v[0];
-    int sector = (i0 / 2) % sect_dims;
+    const int idst = row_dst*ne0 + i0/2;
+    const int ix   = channel_x*s2 + row_x*s1 + i0/2;
+
+    const int sect_dims = sections.v[0] + sections.v[1] + sections.v[2] + sections.v[3];
+    const int sec_w = sections.v[1] + sections.v[0];
+    const int sector = (i0 / 2) % sect_dims;
 
     float theta_base = 0.0;
     if (sector < sections.v[0]) {
-        theta_base = pos[i2]*powf(theta_scale, i0/2.0f);
+        theta_base = pos[channel_x]*powf(theta_scale, i0/2.0f);
     }
     else if (sector >= sections.v[0] && sector < sec_w) {
-        theta_base = pos[i2 + ne2 * 1]*powf(theta_scale, i0/2.0f);
+        theta_base = pos[channel_x + ne2 * 1]*powf(theta_scale, i0/2.0f);
     }
     else if (sector >= sec_w && sector < sec_w + sections.v[2]) {
-        theta_base = pos[i2 + ne2 * 2]*powf(theta_scale, i0/2.0f);
+        theta_base = pos[channel_x + ne2 * 2]*powf(theta_scale, i0/2.0f);
     }
     else if (sector >= sec_w + sections.v[2]) {
-        theta_base = pos[i2 + ne2 * 3]*powf(theta_scale, i0/2.0f);
+        theta_base = pos[channel_x + ne2 * 3]*powf(theta_scale, i0/2.0f);
     }
 
     const float freq_factor = has_ff ? freq_factors[i0/2] : 1.0f;
@@ -160,42 +176,46 @@ static __global__ void rope_multi(
     float cos_theta;
     float sin_theta;
 
-    rope_yarn(theta_base/freq_factor, freq_scale, corr_dims, i0, ext_factor, attn_factor, &cos_theta, &sin_theta);
+    rope_yarn<forward>(theta_base/freq_factor, freq_scale, corr_dims, i0, ext_factor, attn_factor, cos_theta, sin_theta);
 
-    const float x0 = x[i + 0];
-    const float x1 = x[i + n_dims/2];
+    const float x0 = x[ix + 0];
+    const float x1 = x[ix + n_dims/2];
 
-    dst[i + 0]        = x0*cos_theta - x1*sin_theta;
-    dst[i + n_dims/2] = x0*sin_theta + x1*cos_theta;
+    dst[idst + 0]        = x0*cos_theta - x1*sin_theta;
+    dst[idst + n_dims/2] = x0*sin_theta + x1*cos_theta;
 }
 
-template<typename T, bool has_ff>
+template<bool forward, bool has_ff, typename T>
 static __global__ void rope_vision(
-    const T * x, T * dst, int ne0, int ne2, int n_dims, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float ext_factor, float attn_factor, rope_corr_dims corr_dims, float theta_scale, const float * freq_factors, mrope_sections sections) {
+        const T * __restrict__ x, T * __restrict__ dst, const int ne0, const int ne1, const int ne2, const int s1, const int s2, const int n_dims,
+        const int32_t * __restrict__ pos, const float freq_scale, const float ext_factor, const float attn_factor, const rope_corr_dims corr_dims,
+        const float theta_scale, const float * __restrict__ freq_factors, const mrope_sections sections) {
     const int i0 = 2*(blockDim.y*blockIdx.y + threadIdx.y);
 
     if (i0 >= ne0) {
         return;
     }
 
-    const int row = blockDim.x*blockIdx.x + threadIdx.x;
+    const int row_dst = blockDim.x*blockIdx.x + threadIdx.x;
 
-    const int i  = row*ne0 + i0/2;
-    const int i2 = row/p_delta_rows; // i2-th tokens
+    const int row_x     = row_dst % ne1;
+    const int channel_x = row_dst / ne1;
 
-    int sect_dims = sections.v[0] + sections.v[1];
-    int sec_w = sections.v[1] + sections.v[0];
-    int sector = (i0 / 2) % sect_dims;
+    const int idst = row_dst*ne0 + i0/2;
+    const int ix   = channel_x*s2 + row_x*s1 + i0/2;
+
+    const int sect_dims = sections.v[0] + sections.v[1];
+    const int sec_w = sections.v[1] + sections.v[0];
+    const int sector = (i0 / 2) % sect_dims;
 
     float theta_base = 0.0;
     if (sector < sections.v[0]) {
         const int p = sector;
-        theta_base = pos[i2]*powf(theta_scale, p);
+        theta_base = pos[channel_x]*powf(theta_scale, p);
     }
     else if (sector >= sections.v[0] && sector < sec_w) {
         const int p = sector - sections.v[0];
-        theta_base = pos[i2 + ne2]*powf(theta_scale, p);
+        theta_base = pos[channel_x + ne2]*powf(theta_scale, p);
     }
 
     const float freq_factor = has_ff ? freq_factors[i0/2] : 1.0f;
@@ -203,19 +223,20 @@ static __global__ void rope_vision(
     float cos_theta;
     float sin_theta;
 
-    rope_yarn(theta_base/freq_factor, freq_scale, corr_dims, i0, ext_factor, attn_factor, &cos_theta, &sin_theta);
+    rope_yarn<forward>(theta_base/freq_factor, freq_scale, corr_dims, i0, ext_factor, attn_factor, cos_theta, sin_theta);
 
-    const float x0 = x[i + 0];
-    const float x1 = x[i + n_dims];
+    const float x0 = x[ix + 0];
+    const float x1 = x[ix + n_dims];
 
-    dst[i + 0]      = x0*cos_theta - x1*sin_theta;
-    dst[i + n_dims] = x0*sin_theta + x1*cos_theta;
+    dst[idst + 0]      = x0*cos_theta - x1*sin_theta;
+    dst[idst + n_dims] = x0*sin_theta + x1*cos_theta;
 }
 
-template<typename T>
+template<bool forward, typename T>
 static void rope_norm_cuda(
-    const T * x, T * dst, int ne0, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, cudaStream_t stream) {
+        const T * __restrict__ x, T * __restrict__ dst, const int ne0, const int ne1, const int s1, const int s2, const int n_dims, const int nr,
+        const int32_t * __restrict__ pos, const float freq_scale, const float freq_base, const float ext_factor, const float attn_factor,
+        const rope_corr_dims corr_dims, const float * __restrict__ freq_factors, cudaStream_t stream) {
     GGML_ASSERT(ne0 % 2 == 0);
     const dim3 block_dims(1, CUDA_ROPE_BLOCK_SIZE, 1);
     const int n_blocks_x = (ne0 + 2*CUDA_ROPE_BLOCK_SIZE - 1) / (2*CUDA_ROPE_BLOCK_SIZE);
@@ -224,22 +245,21 @@ static void rope_norm_cuda(
     const float theta_scale = powf(freq_base, -2.0f/n_dims);
 
     if (freq_factors == nullptr) {
-        rope_norm<T, false><<<block_nums, block_dims, 0, stream>>>(
-                x, dst, ne0, n_dims, pos, freq_scale, p_delta_rows, ext_factor, attn_factor, corr_dims,
-                theta_scale, freq_factors
-                );
+        rope_norm<forward, false><<<block_nums, block_dims, 0, stream>>>(
+            x, dst, ne0, ne1, s1, s2, n_dims, pos, freq_scale, ext_factor,
+            attn_factor, corr_dims, theta_scale, freq_factors);
     } else {
-        rope_norm<T, true><<<block_nums, block_dims, 0, stream>>>(
-                x, dst, ne0, n_dims, pos, freq_scale, p_delta_rows, ext_factor, attn_factor, corr_dims,
-                theta_scale, freq_factors
-                );
+        rope_norm<forward, true><<<block_nums, block_dims, 0, stream>>>(
+            x, dst, ne0, ne1, s1, s2, n_dims, pos, freq_scale, ext_factor,
+            attn_factor, corr_dims, theta_scale, freq_factors);
     }
 }
 
-template<typename T>
+template<bool forward, typename T>
 static void rope_neox_cuda(
-    const T * x, T * dst, int ne0, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, cudaStream_t stream) {
+        const T * __restrict__ x, T * __restrict__ dst, const int ne0, const int ne1, const int s1, const int s2, const int n_dims, const int nr,
+        const int32_t * __restrict__ pos, const float freq_scale, const float freq_base, const float ext_factor, const float attn_factor,
+        const rope_corr_dims corr_dims, const float * __restrict__ freq_factors, cudaStream_t stream) {
     GGML_ASSERT(ne0 % 2 == 0);
     const dim3 block_dims(1, CUDA_ROPE_BLOCK_SIZE, 1);
     const int n_blocks_x = (ne0 + 2*CUDA_ROPE_BLOCK_SIZE - 1) / (2*CUDA_ROPE_BLOCK_SIZE);
@@ -248,22 +268,21 @@ static void rope_neox_cuda(
     const float theta_scale = powf(freq_base, -2.0f/n_dims);
 
     if (freq_factors == nullptr) {
-        rope_neox<T, false><<<block_nums, block_dims, 0, stream>>>(
-                x, dst, ne0, n_dims, pos, freq_scale, p_delta_rows, ext_factor, attn_factor, corr_dims,
-                theta_scale, freq_factors
-                );
+        rope_neox<forward, false, T><<<block_nums, block_dims, 0, stream>>>(
+            x, dst, ne0, ne1, s1, s2, n_dims, pos, freq_scale, ext_factor,
+            attn_factor, corr_dims, theta_scale, freq_factors);
     } else {
-        rope_neox<T, true><<<block_nums, block_dims, 0, stream>>>(
-                x, dst, ne0, n_dims, pos, freq_scale, p_delta_rows, ext_factor, attn_factor, corr_dims,
-                theta_scale, freq_factors
-                );
+        rope_neox<forward, true, T><<<block_nums, block_dims, 0, stream>>>(
+            x, dst, ne0, ne1, s1, s2, n_dims, pos, freq_scale, ext_factor,
+            attn_factor, corr_dims, theta_scale, freq_factors);
     }
 }
 
-template<typename T>
+template<bool forward, typename T>
 static void rope_multi_cuda(
-    const T * x, T * dst, int ne0, int ne2, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, mrope_sections sections, cudaStream_t stream) {
+        const T * __restrict__ x, T * __restrict__ dst, const int ne0, const int ne1, const int ne2, const int s1, const int s2, const int n_dims, const int nr,
+        const int32_t * __restrict__ pos, const float freq_scale, const float freq_base, const float ext_factor, const float attn_factor,
+        const rope_corr_dims corr_dims, const float * __restrict__ freq_factors, const mrope_sections sections, cudaStream_t stream) {
     GGML_ASSERT(ne0 % 2 == 0);
     const dim3 block_dims(1, CUDA_ROPE_BLOCK_SIZE, 1);
     const int n_blocks_x = (ne0 + 2*CUDA_ROPE_BLOCK_SIZE - 1) / (2*CUDA_ROPE_BLOCK_SIZE);
@@ -272,22 +291,21 @@ static void rope_multi_cuda(
     const float theta_scale = powf(freq_base, -2.0f/n_dims);
 
     if (freq_factors == nullptr) {
-        rope_multi<T, false><<<block_nums, block_dims, 0, stream>>>(
-                x, dst, ne0, ne2, n_dims, pos, freq_scale, p_delta_rows, ext_factor, attn_factor, corr_dims,
-                theta_scale, freq_factors, sections
-                );
+        rope_multi<forward, false, T><<<block_nums, block_dims, 0, stream>>>(
+            x, dst, ne0, ne1, ne2, s1, s2, n_dims, pos, freq_scale, ext_factor,
+            attn_factor, corr_dims, theta_scale, freq_factors, sections);
     } else {
-        rope_multi<T, true><<<block_nums, block_dims, 0, stream>>>(
-                x, dst, ne0, ne2, n_dims, pos, freq_scale, p_delta_rows, ext_factor, attn_factor, corr_dims,
-                theta_scale, freq_factors, sections
-                );
+        rope_multi<forward, true, T><<<block_nums, block_dims, 0, stream>>>(
+            x, dst, ne0, ne1, ne2, s1, s2, n_dims, pos, freq_scale, ext_factor,
+            attn_factor, corr_dims, theta_scale, freq_factors, sections);
     }
 }
 
-template<typename T>
+template<bool forward, typename T>
 static void rope_vision_cuda(
-    const T * x, T * dst, int ne0, int ne2, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, mrope_sections sections, cudaStream_t stream) {
+        const T * __restrict__ x, T * __restrict__ dst, const int ne0, const int ne1, const int ne2, const int s1, const int s2, const int n_dims, const int nr,
+        const int32_t * __restrict__ pos, const float freq_scale, const float freq_base, const float ext_factor, const float attn_factor,
+        const rope_corr_dims corr_dims, const float * __restrict__ freq_factors, const mrope_sections sections, cudaStream_t stream) {
     GGML_ASSERT(ne0 % 2 == 0);
     const dim3 block_dims(1, CUDA_ROPE_BLOCK_SIZE, 1);
     const int n_blocks_x = (ne0 + 2*CUDA_ROPE_BLOCK_SIZE - 1) / (2*CUDA_ROPE_BLOCK_SIZE);
@@ -298,80 +316,18 @@ static void rope_vision_cuda(
     const float theta_scale = powf(freq_base, -2.0f/n_dims);
 
     if (freq_factors == nullptr) {
-        rope_vision<T, false><<<block_nums, block_dims, 0, stream>>>(
-                x, dst, ne0, ne2, n_dims, pos, freq_scale, p_delta_rows, ext_factor, attn_factor, corr_dims,
-                theta_scale, freq_factors, sections
-                );
+        rope_vision<forward, false, T><<<block_nums, block_dims, 0, stream>>>(
+            x, dst, ne0, ne1, ne2, s1, s2, n_dims, pos, freq_scale, ext_factor,
+            attn_factor, corr_dims, theta_scale, freq_factors, sections);
     } else {
-        rope_vision<T, true><<<block_nums, block_dims, 0, stream>>>(
-                x, dst, ne0, ne2, n_dims, pos, freq_scale, p_delta_rows, ext_factor, attn_factor, corr_dims,
-                theta_scale, freq_factors, sections
-                );
+        rope_vision<forward, true, T><<<block_nums, block_dims, 0, stream>>>(
+            x, dst, ne0, ne1, ne2, s1, s2, n_dims, pos, freq_scale, ext_factor,
+            attn_factor, corr_dims, theta_scale, freq_factors, sections);
     }
 }
 
-static void rope_norm_cuda_f16(
-    const half * x, half * dst, int ne0, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, cudaStream_t stream) {
-
-    rope_norm_cuda<half>(x, dst, ne0, n_dims, nr, pos, freq_scale, p_delta_rows, freq_base, ext_factor, attn_factor, corr_dims, freq_factors, stream);
-}
-
-static void rope_norm_cuda_f32(
-    const float * x, float * dst, int ne0, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, cudaStream_t stream) {
-
-    rope_norm_cuda<float>(x, dst, ne0, n_dims, nr, pos, freq_scale, p_delta_rows, freq_base, ext_factor, attn_factor, corr_dims, freq_factors, stream);
-}
-
-static void rope_neox_cuda_f16(
-    const half * x, half * dst, int ne0, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, cudaStream_t stream) {
-
-    rope_neox_cuda<half>(x, dst, ne0, n_dims, nr, pos, freq_scale, p_delta_rows, freq_base, ext_factor, attn_factor, corr_dims, freq_factors, stream);
-}
-
-static void rope_neox_cuda_f32(
-    const float * x, float * dst, int ne0, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, cudaStream_t stream
-) {
-
-    rope_neox_cuda<float>(x, dst, ne0, n_dims, nr, pos, freq_scale, p_delta_rows, freq_base, ext_factor, attn_factor, corr_dims, freq_factors, stream);
-}
-
-static void rope_multi_cuda_f16(
-    const half * x, half * dst, int ne0, int ne2, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, mrope_sections sections, cudaStream_t stream
-) {
-
-    rope_multi_cuda<half>(x, dst, ne0, ne2, n_dims, nr, pos, freq_scale, p_delta_rows, freq_base, ext_factor, attn_factor, corr_dims, freq_factors, sections, stream);
-}
-
-static void rope_multi_cuda_f32(
-    const float * x, float * dst, int ne0, int ne2, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, mrope_sections sections, cudaStream_t stream
-) {
-
-    rope_multi_cuda<float>(x, dst, ne0, ne2, n_dims, nr, pos, freq_scale, p_delta_rows, freq_base, ext_factor, attn_factor, corr_dims, freq_factors, sections, stream);
-}
-
-static void rope_vision_cuda_f16(
-    const half * x, half * dst, int ne0, int ne2, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, mrope_sections sections, cudaStream_t stream
-) {
-
-    rope_vision_cuda<half>(x, dst, ne0, ne2, n_dims, nr, pos, freq_scale, p_delta_rows, freq_base, ext_factor, attn_factor, corr_dims, freq_factors, sections, stream);
-}
-
-static void rope_vision_cuda_f32(
-    const float * x, float * dst, int ne0, int ne2, int n_dims, int nr, const int32_t * pos, float freq_scale, int p_delta_rows,
-    float freq_base, float ext_factor, float attn_factor, rope_corr_dims corr_dims, const float * freq_factors, mrope_sections sections, cudaStream_t stream
-) {
-
-    rope_vision_cuda<float>(x, dst, ne0, ne2, n_dims, nr, pos, freq_scale, p_delta_rows, freq_base, ext_factor, attn_factor, corr_dims, freq_factors, sections, stream);
-}
-
-void ggml_cuda_op_rope(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
+template <bool forward>
+void ggml_cuda_op_rope_impl(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
     const ggml_tensor * src0 = dst->src[0];
     const ggml_tensor * src1 = dst->src[1];
     const ggml_tensor * src2 = dst->src[2];
@@ -382,7 +338,6 @@ void ggml_cuda_op_rope(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
     float * dst_d = (float *)dst->data;
     cudaStream_t stream = ctx.stream();
 
-    GGML_ASSERT(ggml_is_contiguous(src0));
     GGML_ASSERT(src0->type == GGML_TYPE_F32 || src0->type == GGML_TYPE_F16);
     GGML_ASSERT( dst->type == GGML_TYPE_F32 ||  dst->type == GGML_TYPE_F16);
     GGML_ASSERT(src0->type == dst->type);
@@ -392,6 +347,9 @@ void ggml_cuda_op_rope(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
     const int64_t ne02 = src0->ne[2]; // num heads
     const int64_t nr = ggml_nrows(src0);
 
+    const size_t s01 = src0->nb[1] / ggml_type_size(src0->type);
+    const size_t s02 = src0->nb[2] / ggml_type_size(src0->type);
+
     //const int n_past     = ((int32_t *) dst->op_params)[0];
     const int n_dims     = ((int32_t *) dst->op_params)[1];
     const int mode       = ((int32_t *) dst->op_params)[2];
@@ -440,59 +398,59 @@ void ggml_cuda_op_rope(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
     // compute
     if (is_neox) {
         if (src0->type == GGML_TYPE_F32) {
-            rope_neox_cuda_f32(
-                (const float *)src0_d, (float *)dst_d, ne00, n_dims, nr, pos, freq_scale, ne01, freq_base, ext_factor,
-                attn_factor, corr_dims, freq_factors, stream
-            );
+            rope_neox_cuda<forward>(
+                (const float *) src0_d, (float *) dst_d, ne00, ne01, s01, s02, n_dims, nr, pos, freq_scale,
+                freq_base, ext_factor, attn_factor, corr_dims, freq_factors, stream);
         } else if (src0->type == GGML_TYPE_F16) {
-            rope_neox_cuda_f16(
-                (const half *)src0_d, (half *)dst_d, ne00, n_dims, nr, pos, freq_scale, ne01, freq_base, ext_factor,
-                attn_factor, corr_dims, freq_factors, stream
-            );
+            rope_neox_cuda<forward>(
+                (const half *) src0_d, (half *) dst_d, ne00, ne01, s01, s02, n_dims, nr, pos, freq_scale,
+                freq_base, ext_factor, attn_factor, corr_dims, freq_factors, stream);
         } else {
             GGML_ABORT("fatal error");
         }
     } else if (is_mrope && !is_vision) {
         if (src0->type == GGML_TYPE_F32) {
-            rope_multi_cuda_f32(
-                (const float *)src0_d, (float *)dst_d, ne00, ne02, n_dims, nr, pos, freq_scale, ne01, freq_base, ext_factor,
-                attn_factor, corr_dims, freq_factors, sections, stream
-            );
+            rope_multi_cuda<forward>(
+                (const float *) src0_d, (float *) dst_d, ne00, ne01, ne02, s01, s02, n_dims, nr, pos, freq_scale,
+                freq_base, ext_factor, attn_factor, corr_dims, freq_factors, sections, stream);
         } else if (src0->type == GGML_TYPE_F16) {
-            rope_multi_cuda_f16(
-                (const half *)src0_d, (half *)dst_d, ne00, ne02, n_dims, nr, pos, freq_scale, ne01, freq_base, ext_factor,
-                attn_factor, corr_dims, freq_factors, sections, stream
-            );
+            rope_multi_cuda<forward>(
+                (const half *) src0_d, (half *) dst_d, ne00, ne01, ne02, s01, s02, n_dims, nr, pos, freq_scale,
+                freq_base, ext_factor, attn_factor, corr_dims, freq_factors, sections, stream);
         } else {
             GGML_ABORT("fatal error");
         }
     } else if (is_vision) {
         if (src0->type == GGML_TYPE_F32) {
-            rope_vision_cuda_f32(
-                (const float *)src0_d, (float *)dst_d, ne00, ne02, n_dims, nr, pos, freq_scale, ne01, freq_base, ext_factor,
-                attn_factor, corr_dims, freq_factors, sections, stream
-            );
+            rope_vision_cuda<forward>(
+                (const float *) src0_d, (float *) dst_d, ne00, ne01, ne02, s01, s02, n_dims, nr, pos, freq_scale,
+                freq_base, ext_factor, attn_factor, corr_dims, freq_factors, sections, stream);
         } else if (src0->type == GGML_TYPE_F16) {
-            rope_vision_cuda_f16(
-                (const half *)src0_d, (half *)dst_d, ne00, ne02, n_dims, nr, pos, freq_scale, ne01, freq_base, ext_factor,
-                attn_factor, corr_dims, freq_factors, sections, stream
-            );
+            rope_vision_cuda<forward>(
+                (const half *) src0_d, (half *) dst_d, ne00, ne01, ne02, s01, s02, n_dims, nr, pos, freq_scale,
+                freq_base, ext_factor, attn_factor, corr_dims, freq_factors, sections, stream);
         } else {
             GGML_ABORT("fatal error");
         }
     } else {
         if (src0->type == GGML_TYPE_F32) {
-            rope_norm_cuda_f32(
-                (const float *)src0_d, (float *)dst_d, ne00, n_dims, nr, pos, freq_scale, ne01, freq_base, ext_factor,
-                attn_factor, corr_dims, freq_factors, stream
-            );
+            rope_norm_cuda<forward>(
+                (const float *) src0_d, (float *) dst_d, ne00, ne01, s01, s02, n_dims, nr, pos, freq_scale,
+                freq_base, ext_factor, attn_factor, corr_dims, freq_factors, stream);
         } else if (src0->type == GGML_TYPE_F16) {
-            rope_norm_cuda_f16(
-                (const half *)src0_d, (half *)dst_d, ne00, n_dims, nr, pos, freq_scale, ne01, freq_base, ext_factor,
-                attn_factor, corr_dims, freq_factors, stream
-            );
+            rope_norm_cuda<forward>(
+                (const half *) src0_d, (half *) dst_d, ne00, ne01, s01, s02, n_dims, nr, pos, freq_scale,
+                freq_base, ext_factor, attn_factor, corr_dims, freq_factors, stream);
         } else {
             GGML_ABORT("fatal error");
         }
     }
 }
+
+void ggml_cuda_op_rope(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
+    ggml_cuda_op_rope_impl<true>(ctx, dst);
+}
+
+void ggml_cuda_op_rope_back(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
+    ggml_cuda_op_rope_impl<false>(ctx, dst);
+}

ggml/src/ggml-cuda/rope.cuh

@@ -3,3 +3,5 @@
 #define CUDA_ROPE_BLOCK_SIZE 256
 
 void ggml_cuda_op_rope(ggml_backend_cuda_context & ctx, ggml_tensor * dst);
+
+void ggml_cuda_op_rope_back(ggml_backend_cuda_context & ctx, ggml_tensor * dst);

ggml/src/ggml-cuda/wkv6.cu

@@ -73,9 +73,9 @@ void ggml_cuda_op_rwkv_wkv6(ggml_backend_cuda_context & ctx, ggml_tensor * dst)
     const float * s_d  = (const float *)dst->src[5]->data;
 
     const int64_t B = dst->src[5]->ne[1];
-    const int64_t T = dst->src[0]->ne[3];
+    const int64_t T = dst->src[0]->ne[2];
     const int64_t C = dst->ne[0];
-    const int64_t H = dst->src[0]->ne[2];
+    const int64_t H = dst->src[0]->ne[1];
 
     float * dst_d = (float *)dst->data;
 

ggml/src/ggml-hip/CMakeLists.txt

@@ -70,7 +70,9 @@ ggml_add_backend_library(ggml-hip
                         )
 
 # TODO: do not use CUDA definitions for HIP
-target_compile_definitions(ggml PUBLIC GGML_USE_CUDA)
+if (NOT GGML_BACKEND_DL)
+    target_compile_definitions(ggml PUBLIC GGML_USE_CUDA)
+endif()
 
 add_compile_definitions(GGML_USE_HIP)
 

ggml/src/ggml-impl.h

@@ -3,6 +3,8 @@
 // GGML internal header
 
 #include "ggml.h"
+#include "gguf.h"
+
 #include <assert.h>
 #include <math.h>
 #include <stdlib.h> // load `stdlib.h` before other headers to work around MinGW bug: https://sourceforge.net/p/mingw-w64/bugs/192/
@@ -551,22 +553,15 @@ static inline ggml_bf16_t ggml_compute_fp32_to_bf16(float s) {
 #define GGML_FP32_TO_BF16(x) ggml_compute_fp32_to_bf16(x)
 #define GGML_BF16_TO_FP32(x) ggml_compute_bf16_to_fp32(x)
 
-// expose GGUF internals for test code
-
-GGML_API size_t gguf_type_size(enum gguf_type type);
-
-GGML_API struct gguf_context * gguf_init_from_file_impl(FILE * file, struct gguf_init_params params);
-
-struct gguf_buf {
-    void * data;
-    size_t size;
-    size_t offset;
-};
-GGML_API struct gguf_buf gguf_buf_init(size_t size);
-GGML_API void gguf_buf_free(struct gguf_buf buf);
-
-GGML_API void gguf_write_to_buf(const struct gguf_context * ctx, struct gguf_buf * buf, bool only_meta);
-
 #ifdef __cplusplus
 }
 #endif
+
+#ifdef __cplusplus
+#include <vector>
+
+// expose GGUF internals for test code
+GGML_API size_t gguf_type_size(enum gguf_type type);
+GGML_API struct gguf_context * gguf_init_from_file_impl(FILE * file, struct gguf_init_params params);
+GGML_API void gguf_write_to_buf(const struct gguf_context * ctx, std::vector<int8_t> & buf, bool only_meta);
+#endif // __cplusplus

ggml/src/ggml-rpc/ggml-rpc.cpp

@@ -27,15 +27,6 @@
 #endif
 #include <cstring>
 
-#define UNUSED GGML_UNUSED
-
-#define GGML_DEBUG 0
-#if (GGML_DEBUG >= 1)
-#define GGML_PRINT_DEBUG(...) printf(__VA_ARGS__)
-#else
-#define GGML_PRINT_DEBUG(...)
-#endif
-
 #ifdef _WIN32
 typedef SOCKET sockfd_t;
 using ssize_t = __int64;
@@ -411,7 +402,7 @@ static std::shared_ptr<socket_t> get_socket(const std::string & endpoint) {
         initialized = true;
     }
 #else
-    UNUSED(initialized);
+    GGML_UNUSED(initialized);
 #endif
     auto sock = socket_connect(host.c_str(), port);
     if (sock == nullptr) {
@@ -640,7 +631,7 @@ static void ggml_backend_rpc_free(ggml_backend_t backend) {
 }
 
 static void ggml_backend_rpc_synchronize(ggml_backend_t backend) {
-    UNUSED(backend);
+    GGML_UNUSED(backend);
     // this is no-op because we don't have any async operations
 }
 
@@ -850,7 +841,7 @@ void rpc_server::alloc_buffer(const rpc_msg_alloc_buffer_req & request, rpc_msg_
         GGML_PRINT_DEBUG("[%s] size: %" PRIu64 " -> remote_ptr: %" PRIx64 ", remote_size: %" PRIu64 "\n", __func__, request.size, response.remote_ptr, response.remote_size);
         buffers.insert(buffer);
     } else {
-        GGML_PRINT_DEBUG("[%s] size: %" PRIu64 " -> failed\n", __func__, request.size);
+        GGML_LOG_ERROR("[%s] size: %" PRIu64 " -> failed\n", __func__, request.size);
     }
 }
 
@@ -872,7 +863,7 @@ bool rpc_server::buffer_get_base(const rpc_msg_buffer_get_base_req & request, rp
     GGML_PRINT_DEBUG("[%s] remote_ptr: %" PRIx64 "\n", __func__, request.remote_ptr);
     ggml_backend_buffer_t buffer = reinterpret_cast<ggml_backend_buffer_t>(request.remote_ptr);
     if (buffers.find(buffer) == buffers.end()) {
-        GGML_PRINT_DEBUG("[%s] buffer not found\n", __func__);
+        GGML_LOG_ERROR("[%s] buffer not found\n", __func__);
         return false;
     }
     void * base = ggml_backend_buffer_get_base(buffer);
@@ -884,7 +875,7 @@ bool rpc_server::free_buffer(const rpc_msg_free_buffer_req & request) {
     GGML_PRINT_DEBUG("[%s] remote_ptr: %" PRIx64 "\n", __func__, request.remote_ptr);
     ggml_backend_buffer_t buffer = reinterpret_cast<ggml_backend_buffer_t>(request.remote_ptr);
     if (buffers.find(buffer) == buffers.end()) {
-        GGML_PRINT_DEBUG("[%s] buffer not found\n", __func__);
+        GGML_LOG_ERROR("[%s] buffer not found\n", __func__);
         return false;
     }
     ggml_backend_buffer_free(buffer);
@@ -896,7 +887,7 @@ bool rpc_server::buffer_clear(const rpc_msg_buffer_clear_req & request) {
     GGML_PRINT_DEBUG("[%s] remote_ptr: %" PRIx64 ", value: %u\n", __func__, request.remote_ptr, request.value);
     ggml_backend_buffer_t buffer = reinterpret_cast<ggml_backend_buffer_t>(request.remote_ptr);
     if (buffers.find(buffer) == buffers.end()) {
-        GGML_PRINT_DEBUG("[%s] buffer not found\n", __func__);
+        GGML_LOG_ERROR("[%s] buffer not found\n", __func__);
         return false;
     }
     ggml_backend_buffer_clear(buffer, request.value);
@@ -952,7 +943,7 @@ bool rpc_server::set_tensor(const std::vector<uint8_t> & input) {
     struct ggml_context * ctx = ggml_init(params);
     ggml_tensor * tensor = deserialize_tensor(ctx, in_tensor);
     if (tensor == nullptr) {
-        GGML_PRINT_DEBUG("[%s] error deserializing tensor\n", __func__);
+        GGML_LOG_ERROR("[%s] error deserializing tensor\n", __func__);
         ggml_free(ctx);
         return false;
     }
@@ -1017,7 +1008,7 @@ bool rpc_server::get_tensor(const rpc_msg_get_tensor_req & request, std::vector<
     struct ggml_context * ctx = ggml_init(params);
     ggml_tensor * tensor = deserialize_tensor(ctx, &request.tensor);
     if (tensor == nullptr) {
-        GGML_PRINT_DEBUG("[%s] error deserializing tensor\n", __func__);
+        GGML_LOG_ERROR("[%s] error deserializing tensor\n", __func__);
         ggml_free(ctx);
         return false;
     }
@@ -1051,7 +1042,7 @@ bool rpc_server::copy_tensor(const rpc_msg_copy_tensor_req & request, rpc_msg_co
     ggml_tensor * src = deserialize_tensor(ctx, &request.src);
     ggml_tensor * dst = deserialize_tensor(ctx, &request.dst);
     if (src == nullptr || dst == nullptr) {
-        GGML_PRINT_DEBUG("[%s] error deserializing tensors\n", __func__);
+        GGML_LOG_ERROR("[%s] error deserializing tensors\n", __func__);
         ggml_free(ctx);
         return false;
     }
@@ -1385,14 +1376,14 @@ static void ggml_backend_rpc_device_get_memory(ggml_backend_dev_t dev, size_t *
 
     ggml_backend_rpc_get_device_memory(ctx->endpoint.c_str(), free, total);
 
-    UNUSED(dev);
+    GGML_UNUSED(dev);
 }
 
 static enum ggml_backend_dev_type ggml_backend_rpc_device_get_type(ggml_backend_dev_t dev) {
     // TODO: obtain value from the server
     return GGML_BACKEND_DEVICE_TYPE_GPU;
 
-    UNUSED(dev);
+    GGML_UNUSED(dev);
 }
 
 static void ggml_backend_rpc_device_get_props(ggml_backend_dev_t dev, struct ggml_backend_dev_props * props) {
@@ -1413,7 +1404,7 @@ static ggml_backend_t ggml_backend_rpc_device_init(ggml_backend_dev_t dev, const
 
     return ggml_backend_rpc_init(ctx->endpoint.c_str());
 
-    UNUSED(params);
+    GGML_UNUSED(params);
 }
 
 static ggml_backend_buffer_type_t ggml_backend_rpc_device_get_buffer_type(ggml_backend_dev_t dev) {
@@ -1421,12 +1412,12 @@ static ggml_backend_buffer_type_t ggml_backend_rpc_device_get_buffer_type(ggml_b
 
     return ggml_backend_rpc_buffer_type(ctx->endpoint.c_str());
 
-    UNUSED(dev);
+    GGML_UNUSED(dev);
 }
 
 static bool ggml_backend_rpc_device_supports_op(ggml_backend_dev_t dev, const struct ggml_tensor * op) {
-    UNUSED(dev);
-    UNUSED(op);
+    GGML_UNUSED(dev);
+    GGML_UNUSED(op);
     //TODO: call the remote backend and cache the results
     return true;
 }
@@ -1463,20 +1454,20 @@ static const struct ggml_backend_device_i ggml_backend_rpc_device_i = {
 static const char * ggml_backend_rpc_reg_get_name(ggml_backend_reg_t reg) {
     return "RPC";
 
-    UNUSED(reg);
+    GGML_UNUSED(reg);
 }
 
 static size_t ggml_backend_rpc_reg_get_device_count(ggml_backend_reg_t reg) {
     return 0;
 
-    UNUSED(reg);
+    GGML_UNUSED(reg);
 }
 
 static ggml_backend_dev_t ggml_backend_rpc_reg_get_device(ggml_backend_reg_t reg, size_t index) {
     GGML_ABORT("The RPC backend does not have enumerated devices - use ggml_backend_add_device instead");
 
-    UNUSED(reg);
-    UNUSED(index);
+    GGML_UNUSED(reg);
+    GGML_UNUSED(index);
 }
 
 static void * ggml_backend_rpc_get_proc_address(ggml_backend_reg_t reg, const char * name) {
@@ -1485,7 +1476,7 @@ static void * ggml_backend_rpc_get_proc_address(ggml_backend_reg_t reg, const ch
     }
     return NULL;
 
-    UNUSED(reg);
+    GGML_UNUSED(reg);
 }
 
 static const struct ggml_backend_reg_i ggml_backend_rpc_reg_i = {

ggml/src/ggml-sycl/backend.hpp

@@ -29,5 +29,6 @@
 #include "wkv6.hpp"
 #include "outprod.hpp"
 #include "element_wise.hpp"
+#include "gla.hpp"
 
 #endif // GGML_SYCL_BACKEND_HPP

ggml/src/ggml-sycl/common.cpp

@@ -51,6 +51,10 @@ void ggml_sycl_host_free(void* ptr) try {
   std::exit(1);
 }
 
+bool gpu_has_xmx(sycl::device &dev) {
+    return dev.has(sycl::aspect::ext_intel_matrix);
+}
+
 int64_t downsample_sycl_global_range(int64_t accumulate_block_num, int64_t block_size) {
   const int64_t max_range = std::numeric_limits<int>::max();
   int64_t sycl_down_blk_size = block_size;

ggml/src/ggml-sycl/common.hpp

@@ -662,6 +662,7 @@ inline void ggml_sycl_op_bin_bcast(ggml_backend_sycl_context & ctx, const ggml_t
     }
 }
 
+bool gpu_has_xmx(sycl::device &dev);
 
 void ggml_sycl_op_flatten(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
                                  const ggml_tensor *src1, ggml_tensor *dst,

ggml/src/ggml-sycl/concat.cpp

@@ -158,8 +158,9 @@ static void concat_f32_sycl_non_cont(
       });
 }
 
-void ggml_sycl_op_concat(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
-                                const ggml_tensor *src1, ggml_tensor *dst) {
+void ggml_sycl_op_concat(ggml_backend_sycl_context & ctx, ggml_tensor *dst) {
+  const ggml_tensor *src0 = dst->src[0];
+  const ggml_tensor *src1 = dst->src[1];
   queue_ptr stream = ctx.stream();
 
   const int32_t dim = ((int32_t *)dst->op_params)[0];

ggml/src/ggml-sycl/concat.hpp

@@ -15,7 +15,6 @@
 
 #include "common.hpp"
 
-void ggml_sycl_op_concat(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
-                                const ggml_tensor *src1, ggml_tensor *dst);
+void ggml_sycl_op_concat(ggml_backend_sycl_context & ctx, ggml_tensor *dst);
 
 #endif // GGML_SYCL_CONCAT_HPP

ggml/src/ggml-sycl/conv.cpp

@@ -71,8 +71,9 @@ static void conv_transpose_1d_f32_f32_sycl(
         });
 }
 
-void ggml_sycl_op_conv_transpose_1d(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
-    const ggml_tensor *src1, ggml_tensor *dst) {
+void ggml_sycl_op_conv_transpose_1d(ggml_backend_sycl_context & ctx, ggml_tensor *dst) {
+    const ggml_tensor *src0 = dst->src[0];
+    const ggml_tensor *src1 = dst->src[1];
     const float * src0_d = (const float *)src0->data;
     const float * src1_d = (const float *)src1->data;
 

ggml/src/ggml-sycl/conv.hpp

@@ -15,7 +15,6 @@
 
 #include "common.hpp"
 
-void ggml_sycl_op_conv_transpose_1d(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
-  const ggml_tensor *src1, ggml_tensor *dst);
+void ggml_sycl_op_conv_transpose_1d(ggml_backend_sycl_context & ctx, ggml_tensor *dst);
 
 #endif // GGML_SYCL_CONV_HPP

ggml/src/ggml-sycl/element_wise.cpp

@@ -882,149 +882,149 @@ inline void ggml_sycl_op_div(ggml_backend_sycl_context & ctx, const ggml_tensor
 }
 
 
-void ggml_sycl_sqrt(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_sqrt(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_sqrt);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_sqrt);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_sin(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_sin(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_sin);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_sin);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_cos(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_cos(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_cos);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_cos);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_acc(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_acc(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_acc);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_acc);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_gelu(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_gelu(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_gelu);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_gelu);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_silu(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_silu(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_silu);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_silu);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_gelu_quick(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_gelu_quick(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_gelu_quick);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_gelu_quick);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_tanh(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_tanh(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_tanh);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_tanh);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_relu(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_relu(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_relu);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_relu);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_sigmoid(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_sigmoid(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_sigmoid);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_sigmoid);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_hardsigmoid(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_hardsigmoid(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_hardsigmoid);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_hardsigmoid);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_hardswish(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_hardswish(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_hardswish);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_hardswish);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
 
-void ggml_sycl_exp(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_exp(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_exp);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_exp);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_log(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_log(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_log);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_log);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_neg(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_neg(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_neg);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_neg);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_step(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_step(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_step);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_step);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_leaky_relu(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_leaky_relu(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_leaky_relu);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_leaky_relu);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_sqr(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_sqr(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_sqr);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_sqr);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_upscale(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_upscale(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_upscale);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_upscale);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_pad(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_pad(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_pad);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_pad);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
 
 
-void ggml_sycl_add(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_add(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_add);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_add);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_sub(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_sub(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_sub);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_sub);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_mul(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_mul(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_mul);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_mul);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-void ggml_sycl_div(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+void ggml_sycl_div(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_div);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_div);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }

ggml/src/ggml-sycl/element_wise.hpp

@@ -25,52 +25,52 @@ static __dpct_inline__ float op_div(const float a, const float b) {
 }
 
 
-void ggml_sycl_sqrt(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_sqrt(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_sin(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_sin(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_cos(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_cos(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_acc(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_acc(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_gelu(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_gelu(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_silu(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_silu(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_gelu_quick(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_gelu_quick(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_tanh(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_tanh(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_relu(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_relu(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_sigmoid(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_sigmoid(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_hardsigmoid(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_hardsigmoid(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_hardswish(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_hardswish(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_exp(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_exp(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_log(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_log(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_neg(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_neg(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_step(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_step(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_leaky_relu(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_leaky_relu(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_sqr(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_sqr(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_upscale(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_upscale(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_pad(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_pad(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_add(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_add(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_sub(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_sub(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_mul(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_mul(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
-void ggml_sycl_div(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
+void ggml_sycl_div(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
 
 #endif // GGML_SYCL_ELEMENTWISE_HPP

ggml/src/ggml-sycl/ggml-sycl.cpp

@@ -54,18 +54,12 @@ static ggml_sycl_device_info ggml_sycl_init() {
     GGML_ASSERT(info.device_count <= GGML_SYCL_MAX_DEVICES);
 
     int64_t total_vram = 0;
-#if defined(GGML_SYCL_FORCE_MMQ)
-    GGML_LOG_INFO("%s: GGML_SYCL_FORCE_MMQ:   yes\n", __func__);
-#else
-    GGML_LOG_INFO("%s: GGML_SYCL_FORCE_MMQ:   no\n", __func__);
-#endif
-#if defined(SYCL_USE_XMX)
-    GGML_LOG_INFO("%s: SYCL_USE_XMX: yes\n", __func__);
-#else
-    GGML_LOG_INFO("%s: SYCL_USE_XMX: no\n", __func__);
-#endif
-    GGML_LOG_INFO("%s: found %d %s devices:\n", __func__, info.device_count, GGML_SYCL_NAME);
-
+/* This is a bit misleading;  reserved for later */
+// #if defined(SYCL_USE_XMX)
+//     GGML_LOG_INFO("%s: SYCL_USE_XMX: yes\n", __func__);
+// #else
+//     GGML_LOG_INFO("%s: SYCL_USE_XMX: no\n", __func__);
+// #endif
     for (int i = 0; i < info.device_count; ++i) {
         info.devices[i].vmm = 0;
         dpct::device_info prop;
@@ -109,11 +103,11 @@ void print_device_detail(int id, sycl::device &device, std::string device_type)
     name = std::regex_replace(name, std::regex("\\(TM\\)"), "");
 
     auto global_mem_size = prop.get_global_mem_size()/1000000;
-
-    GGML_LOG_INFO("|%2d|%19s|%39s|%7s|%7d|%8d|%5d|%6luM|%21s|\n", id, device_type.c_str(),
+    std::string xmx = gpu_has_xmx(device) ? "yes" : "no";
+    GGML_LOG_INFO("|%2d|%19s|%39s|%7s|%7d|%8d|%5d|%6luM|%21s|%14s|\n", id, device_type.c_str(),
             name.c_str(), version.c_str(), prop.get_max_compute_units(),
             prop.get_max_work_group_size(), prop.get_max_sub_group_size(),
-            global_mem_size, device.get_info<sycl::info::device::driver_version>().c_str());
+            global_mem_size, device.get_info<sycl::info::device::driver_version>().c_str(), xmx.c_str());
 }
 
 void ggml_backend_sycl_print_sycl_devices() {
@@ -124,16 +118,16 @@ void ggml_backend_sycl_print_sycl_devices() {
 
     GGML_LOG_INFO(
         "|  |                   |                                       |      "
-        " |Max    |        |Max  |Global |                     |\n");
+        " |Max    |        |Max  |Global |                     |         XMX  |\n");
     GGML_LOG_INFO(
         "|  |                   |                                       |      "
-        " |compute|Max work|sub  |mem    |                     |\n");
+        " |compute|Max work|sub  |mem    |                     |          or  |\n");
     GGML_LOG_INFO(
         "|ID|        Device Type|                                   "
-        "Name|Version|units  |group   |group|size   |       Driver version|\n");
+        "Name|Version|units  |group   |group|size   |       Driver version| Tensor Cores |\n");
     GGML_LOG_INFO(
         "|--|-------------------|---------------------------------------|------"
-        "-|-------|--------|-----|-------|---------------------|\n");
+        "-|-------|--------|-----|-------|---------------------|--------------|\n");
 
     for (int id = 0; id < device_count; ++id) {
       sycl::device device = dpct::dev_mgr::instance().get_device(id);
@@ -164,14 +158,18 @@ static void ggml_check_sycl() try {
     static bool initialized = false;
 
     if (!initialized) {
-        GGML_LOG_INFO("[SYCL] call ggml_check_sycl\n");
+        GGML_SYCL_DEBUG("[SYCL] call ggml_check_sycl\n");
         g_ggml_sycl_debug = get_sycl_env("GGML_SYCL_DEBUG", 0);
-        GGML_LOG_INFO("%s: GGML_SYCL_DEBUG: %d\n", __func__, g_ggml_sycl_debug);
-
-#if defined(GGML_SYCL_F16)
-        GGML_LOG_INFO("%s: GGML_SYCL_F16: yes\n", __func__);
+        GGML_LOG_INFO("GGML_SYCL_DEBUG: %d\n", g_ggml_sycl_debug);
+#if defined(GGML_SYCL_FORCE_MMQ)
+        GGML_LOG_INFO("GGML_SYCL_FORCE_MMQ:   yes\n");
 #else
-        GGML_LOG_INFO("%s: GGML_SYCL_F16: no\n", __func__);
+        GGML_LOG_INFO("GGML_SYCL_FORCE_MMQ:   no\n");
+#endif
+#if defined(GGML_SYCL_F16)
+        GGML_LOG_INFO("GGML_SYCL_F16: yes\n");
+#else
+        GGML_LOG_INFO("GGML_SYCL_F16: no\n");
 #endif
 
 /* NOT REMOVE, keep it for next optimize for XMX.
@@ -1189,7 +1187,6 @@ std::unique_ptr<ggml_sycl_pool> ggml_backend_sycl_context::new_pool_for_device(q
 /// kernels
 
 typedef void (*cpy_kernel_t)(const char * cx, char * cdst);
-typedef void (*ggml_sycl_func_t)(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst);
 typedef void (*ggml_sycl_op_mul_mat_t)(
     ggml_backend_sycl_context & ctx,
     const ggml_tensor *src0, const ggml_tensor *src1, ggml_tensor *dst,
@@ -3171,33 +3168,33 @@ catch (sycl::exception const &exc) {
 }
 
 
-static void ggml_sycl_repeat(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+static void ggml_sycl_repeat(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_repeat);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_repeat);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-static void ggml_sycl_get_rows(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+static void ggml_sycl_get_rows(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_get_rows);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_get_rows);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-static void ggml_sycl_norm(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+static void ggml_sycl_norm(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_norm);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_norm);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-static void ggml_sycl_rms_norm(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+static void ggml_sycl_rms_norm(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_rms_norm);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_rms_norm);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
-static void ggml_sycl_group_norm(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+static void ggml_sycl_group_norm(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     GGML_SYCL_DEBUG("call %s\n", __func__);
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_group_norm);
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_group_norm);
     GGML_SYCL_DEBUG("call %s done\n", __func__);
 }
 
@@ -3572,9 +3569,10 @@ __dpct_inline__ static void k_copy_dst_from_contiguous(
     }
 }
 
-static void ggml_sycl_mul_mat_id(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
-                                 const ggml_tensor *src1,
+static void ggml_sycl_mul_mat_id(ggml_backend_sycl_context & ctx,
                                  ggml_tensor *dst) try {
+    const ggml_tensor *src0 = dst->src[0];
+    const ggml_tensor *src1 = dst->src[1];
     GGML_ASSERT(!ggml_backend_buffer_is_sycl_split(src0->buffer) && "mul_mat_id does not support split buffers");
 
     const ggml_tensor *ids = dst->src[2];
@@ -3740,12 +3738,12 @@ catch (sycl::exception const &exc) {
   std::exit(1);
 }
 
-static void ggml_sycl_scale(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_scale);
+static void ggml_sycl_scale(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_scale);
 }
 
-static void ggml_sycl_clamp(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_clamp);
+static void ggml_sycl_clamp(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_clamp);
 }
 
 static void ggml_sycl_cpy(ggml_backend_sycl_context & ctx, const ggml_tensor *src0, const ggml_tensor *src1,
@@ -3787,7 +3785,6 @@ static void ggml_sycl_cpy(ggml_backend_sycl_context & ctx, const ggml_tensor *sr
                 ggml_type_name(src0->type), ggml_type_name(src1->type));
         GGML_ABORT("fatal error");
     }
-
     GGML_UNUSED(dst);
 }
 catch (sycl::exception const &exc) {
@@ -3796,59 +3793,52 @@ catch (sycl::exception const &exc) {
   std::exit(1);
 }
 
-static void ggml_sycl_dup(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
+static void ggml_sycl_dup(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
     // TODO: why do we pass dst as src1 here?
-    ggml_sycl_cpy(ctx, src0, dst, nullptr);
-    GGML_UNUSED(src1);
+    ggml_sycl_cpy(ctx, dst->src[0], dst, nullptr);
 }
 
-static void ggml_sycl_diag_mask_inf(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_diag_mask_inf);
+static void ggml_sycl_diag_mask_inf(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_diag_mask_inf);
 }
 
-static void ggml_sycl_soft_max(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_soft_max);
+static void ggml_sycl_soft_max(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_soft_max);
 }
 
-static void ggml_sycl_rope(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    GGML_ASSERT(ggml_is_contiguous(src0)); // TODO: this restriction is temporary until non-cont support is implemented
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_rope);
+static void ggml_sycl_rope(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    GGML_ASSERT(ggml_is_contiguous(dst->src[0])); // TODO: this restriction is temporary until non-cont support is implemented
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_rope);
 }
 
-static void ggml_sycl_pool2d(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_pool2d);
+static void ggml_sycl_pool2d(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_pool2d);
 }
 
-static void ggml_sycl_im2col(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_im2col);
+static void ggml_sycl_im2col(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_im2col);
 }
 
-static void ggml_sycl_sum(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    GGML_ASSERT(ggml_is_contiguous(src0));
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_sum);
+static void ggml_sycl_sum(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    GGML_ASSERT(ggml_is_contiguous(dst->src[0]));
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_sum);
 }
 
-static void ggml_sycl_sum_rows(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    GGML_ASSERT(ggml_is_contiguous(src0));
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_sum_rows);
+static void ggml_sycl_sum_rows(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    GGML_ASSERT(ggml_is_contiguous(dst->src[0]));
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_sum_rows);
 }
 
-static void ggml_sycl_argsort(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    GGML_ASSERT(ggml_is_contiguous(src0));
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_argsort);
+static void ggml_sycl_argsort(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    GGML_ASSERT(ggml_is_contiguous(dst->src[0]));
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_argsort);
 }
 
-static void ggml_sycl_argmax(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    GGML_ASSERT(ggml_is_contiguous(src0));
-    ggml_sycl_op_flatten(ctx, src0, src1, dst, ggml_sycl_op_argmax);
+static void ggml_sycl_argmax(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    GGML_ASSERT(ggml_is_contiguous(dst->src[0]));
+    ggml_sycl_op_flatten(ctx, dst->src[0], dst->src[1], dst, ggml_sycl_op_argmax);
 }
 
-static void ggml_sycl_nop(ggml_backend_sycl_context & ctx, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
-    GGML_UNUSED(src0);
-    GGML_UNUSED(src1);
-    GGML_UNUSED(dst);
-    GGML_UNUSED(ctx);
-}
 
 void ggml_sycl_set_main_device(const int main_device) try {
     if (dpct::get_current_device_id() == static_cast<unsigned int> (main_device)) {
@@ -3871,191 +3861,192 @@ catch (sycl::exception const &exc) {
   std::exit(1);
 }
 
-bool ggml_sycl_compute_forward(ggml_backend_sycl_context & ctx, struct ggml_tensor * tensor) {
+bool ggml_sycl_compute_forward(ggml_backend_sycl_context & ctx, struct ggml_tensor * dst) {
     if (!g_sycl_loaded) return false;
 
-    ggml_sycl_func_t func;
+    if (dst->src[0] != nullptr && ggml_backend_buffer_is_sycl_split(dst->src[0]->buffer)) {
+        ggml_sycl_set_peer_access(dst->src[1]->ne[1], ctx.device);
+    }
 
-    switch (tensor->op) {
+    switch (dst->op) {
         case GGML_OP_ARGMAX:
-            func = ggml_sycl_argmax;
+            ggml_sycl_argmax(ctx, dst);
             break;
         case GGML_OP_CONV_TRANSPOSE_1D:
-            func = ggml_sycl_op_conv_transpose_1d;
+            ggml_sycl_op_conv_transpose_1d(ctx, dst);
             break;
         case GGML_OP_REPEAT:
-            func = ggml_sycl_repeat;
+            ggml_sycl_repeat(ctx, dst);
             break;
         case GGML_OP_GET_ROWS:
-            func = ggml_sycl_get_rows;
+            ggml_sycl_get_rows(ctx, dst);
             break;
         case GGML_OP_DUP:
-            func = ggml_sycl_dup;
+            ggml_sycl_dup(ctx, dst);
             break;
         case GGML_OP_ADD:
         case GGML_OP_ADD1: // TODO: more efficient implementation
-            func = ggml_sycl_add;
+            ggml_sycl_add(ctx, dst);
             break;
         case GGML_OP_SUB:
-            func = ggml_sycl_sub;
+            ggml_sycl_sub(ctx, dst);
             break;
         case GGML_OP_ACC:
-            func = ggml_sycl_acc;
+            ggml_sycl_acc(ctx, dst);
             break;
         case GGML_OP_MUL:
-            func = ggml_sycl_mul;
+            ggml_sycl_mul(ctx, dst);
             break;
         case GGML_OP_LOG:
-            func = ggml_sycl_log;
+            ggml_sycl_log(ctx, dst);
             break;
         case GGML_OP_DIV:
-            func = ggml_sycl_div;
+            ggml_sycl_div(ctx, dst);
             break;
         case GGML_OP_UNARY:
-            switch (ggml_get_unary_op(tensor)) {
+            switch (ggml_get_unary_op(dst)) {
                 case GGML_UNARY_OP_NEG:
-                    func = ggml_sycl_neg;
+                    ggml_sycl_neg(ctx, dst);
                     break;
                 case GGML_UNARY_OP_STEP:
-                    func = ggml_sycl_step;
+                    ggml_sycl_step(ctx, dst);
                     break;
                 case GGML_UNARY_OP_GELU:
-                    func = ggml_sycl_gelu;
+                    ggml_sycl_gelu(ctx, dst);
                     break;
                 case GGML_UNARY_OP_SILU:
-                    func = ggml_sycl_silu;
+                    ggml_sycl_silu(ctx, dst);
                     break;
                 case GGML_UNARY_OP_GELU_QUICK:
-                    func = ggml_sycl_gelu_quick;
+                    ggml_sycl_gelu_quick(ctx, dst);
                     break;
                 case GGML_UNARY_OP_TANH:
-                    func = ggml_sycl_tanh;
+                    ggml_sycl_tanh(ctx, dst);
                     break;
                 case GGML_UNARY_OP_RELU:
-                    func = ggml_sycl_relu;
+                    ggml_sycl_relu(ctx, dst);
                     break;
                 case GGML_UNARY_OP_SIGMOID:
-                    func = ggml_sycl_sigmoid;
+                    ggml_sycl_sigmoid(ctx, dst);
                     break;
                 case GGML_UNARY_OP_HARDSIGMOID:
-                    func = ggml_sycl_hardsigmoid;
+                    ggml_sycl_hardsigmoid(ctx, dst);
                     break;
                 case GGML_UNARY_OP_HARDSWISH:
-                    func = ggml_sycl_hardswish;
+                    ggml_sycl_hardswish(ctx, dst);
                     break;
                 case GGML_UNARY_OP_EXP:
-                    func = ggml_sycl_exp;
+                    ggml_sycl_exp(ctx, dst);
                     break;
                 default:
                     return false;
             }
             break;
         case GGML_OP_NORM:
-            func = ggml_sycl_norm;
+            ggml_sycl_norm(ctx, dst);
             break;
         case GGML_OP_GROUP_NORM:
-            func = ggml_sycl_group_norm;
+            ggml_sycl_group_norm(ctx, dst);
             break;
         case GGML_OP_CONCAT:
-            func = ggml_sycl_op_concat;
+            ggml_sycl_op_concat(ctx, dst);
             break;
         case GGML_OP_UPSCALE:
-            func = ggml_sycl_upscale;
+            ggml_sycl_upscale(ctx, dst);
             break;
         case GGML_OP_PAD:
-            func = ggml_sycl_pad;
+            ggml_sycl_pad(ctx, dst);
             break;
         case GGML_OP_LEAKY_RELU:
-            func = ggml_sycl_leaky_relu;
+            ggml_sycl_leaky_relu(ctx, dst);
             break;
         case GGML_OP_RMS_NORM:
-            func = ggml_sycl_rms_norm;
+            ggml_sycl_rms_norm(ctx, dst);
             break;
         case GGML_OP_MUL_MAT:
-            if (tensor->src[0]->ne[3] != tensor->src[1]->ne[3]) {
+            if (dst->src[0]->ne[3] != dst->src[1]->ne[3]) {
                 return false;
             }
-            func = ggml_sycl_mul_mat;
+            /* ggml_sycl_mul_mat_id is dependent on ggml_sycl_mul_mat */
+            ggml_sycl_mul_mat(ctx, dst->src[0], dst->src[1], dst);
             break;
         case GGML_OP_MUL_MAT_ID:
-            if (tensor->src[0]->ne[3] != tensor->src[1]->ne[3]) {
+            if (dst->src[0]->ne[3] != dst->src[1]->ne[3]) {
                 return false;
             }
-            func = ggml_sycl_mul_mat_id;
+            ggml_sycl_mul_mat_id(ctx, dst);
             break;
         case GGML_OP_OUT_PROD:
-            func = ggml_sycl_op_out_prod;
+            ggml_sycl_op_out_prod(ctx, dst);
             break;
         case GGML_OP_SCALE:
-            func = ggml_sycl_scale;
+            ggml_sycl_scale(ctx, dst);
             break;
         case GGML_OP_SQR:
-            func = ggml_sycl_sqr;
+            ggml_sycl_sqr(ctx, dst);
             break;
         case GGML_OP_SQRT:
-            func = ggml_sycl_sqrt;
+            ggml_sycl_sqrt(ctx, dst);
             break;
         case GGML_OP_SIN:
-            func = ggml_sycl_sin;
+            ggml_sycl_sin(ctx, dst);
             break;
         case GGML_OP_COS:
-            func = ggml_sycl_cos;
+            ggml_sycl_cos(ctx, dst);
             break;
         case GGML_OP_CLAMP:
-            func = ggml_sycl_clamp;
+            ggml_sycl_clamp(ctx, dst);
             break;
         case GGML_OP_CPY:
-            func = ggml_sycl_cpy;
+            ggml_sycl_cpy(ctx, dst->src[0], dst->src[1], dst);
             break;
         case GGML_OP_CONT:
-            func = ggml_sycl_dup;
+            ggml_sycl_dup(ctx, dst);
             break;
         case GGML_OP_NONE:
         case GGML_OP_RESHAPE:
         case GGML_OP_VIEW:
         case GGML_OP_PERMUTE:
         case GGML_OP_TRANSPOSE:
-            func = ggml_sycl_nop;
+            GGML_SYCL_DEBUG("%s: Tensor NO-OP\n", __func__);
             break;
         case GGML_OP_DIAG_MASK_INF:
-            func = ggml_sycl_diag_mask_inf;
+            ggml_sycl_diag_mask_inf(ctx, dst);
             break;
         case GGML_OP_SOFT_MAX:
-            func = ggml_sycl_soft_max;
+            ggml_sycl_soft_max(ctx, dst);
             break;
         case GGML_OP_ROPE:
-            func = ggml_sycl_rope;
+            ggml_sycl_rope(ctx, dst);
             break;
         case GGML_OP_IM2COL:
-            func = ggml_sycl_im2col;
+            ggml_sycl_im2col(ctx, dst);
             break;
         case GGML_OP_POOL_2D:
-            func = ggml_sycl_pool2d;
+            ggml_sycl_pool2d(ctx, dst);
             break;
         case GGML_OP_SUM:
-            func = ggml_sycl_sum;
+            ggml_sycl_sum(ctx, dst);
             break;
         case GGML_OP_SUM_ROWS:
-            func = ggml_sycl_sum_rows;
+            ggml_sycl_sum_rows(ctx, dst);
             break;
         case GGML_OP_ARGSORT:
-            func = ggml_sycl_argsort;
+            ggml_sycl_argsort(ctx, dst);
             break;
         case GGML_OP_TIMESTEP_EMBEDDING:
-            func = ggml_sycl_op_timestep_embedding;
+            ggml_sycl_op_timestep_embedding(ctx, dst);
             break;
         case GGML_OP_RWKV_WKV6:
-            func = ggml_sycl_op_rwkv_wkv6;
+            ggml_sycl_op_rwkv_wkv6(ctx, dst);
+            break;
+        case GGML_OP_GATED_LINEAR_ATTN:
+            ggml_sycl_op_gated_linear_attn(ctx, dst);
             break;
         default:
             return false;
     }
 
-    if (tensor->src[0] != nullptr && ggml_backend_buffer_is_sycl_split(tensor->src[0]->buffer)) {
-        ggml_sycl_set_peer_access(tensor->src[1]->ne[1], ctx.device);
-    }
-
-    func(ctx, tensor->src[0], tensor->src[1], tensor);
     return true;
 }
 
@@ -4519,6 +4510,7 @@ static bool ggml_backend_sycl_device_supports_op(ggml_backend_dev_t dev, const g
         case GGML_OP_LEAKY_RELU:
         case GGML_OP_TIMESTEP_EMBEDDING:
         case GGML_OP_RWKV_WKV6:
+        case GGML_OP_GATED_LINEAR_ATTN:
             return true;
         default:
             return false;

ggml/src/ggml-sycl/gla.cpp

@@ -0,0 +1,105 @@
+#include <sycl/sycl.hpp>
+
+#include "common.hpp"
+
+template <u_int HEAD_SIZE>
+static void gated_linear_attn_f32_kernel(const dpct::queue_ptr stream, u_int B, u_int T, u_int C, u_int H, float scale,
+                                         const float * k, const float * v, const float * r, const float * td,
+                                         const float * s, float * dst) {
+    const u_int head_size    = HEAD_SIZE;
+    const u_int state_size   = C * head_size;
+    const u_int n_seq_tokens = T / B;
+    sycl::range<1> block_dims((C / H));
+    sycl::range<1> grid_dims((B * H));
+    stream->submit([&](sycl::handler & cgh) {
+        /* local memory accessors*/
+        auto _k  = sycl::local_accessor<float, 1>(sycl::range<1>(head_size), cgh);
+        auto _r  = sycl::local_accessor<float, 1>(sycl::range<1>(head_size), cgh);
+        auto _td = sycl::local_accessor<float, 1>(sycl::range<1>(head_size), cgh);
+
+        cgh.parallel_for(sycl::nd_range<1>(grid_dims * block_dims, block_dims), [=](sycl::nd_item<1> item) {
+            u_int tid = item.get_local_id(0);
+            u_int bid = item.get_group(0);
+
+            u_int batch_i = bid / H;
+            u_int head_i  = bid % H;
+
+            float state[head_size];
+
+#pragma unroll
+            for (u_int i = 0; i < head_size; i++) {
+                state[i] = s[batch_i * state_size + head_i * head_size * head_size + i * head_size + tid];
+            }
+
+            for (u_int t = batch_i * n_seq_tokens * C + head_i * head_size + tid;
+                 t < (batch_i + 1) * n_seq_tokens * C + head_i * head_size + tid; t += C) {
+
+                item.barrier(sycl::access::fence_space::local_space);  //sync threads
+                _k[tid]  = k[t];
+                _r[tid]  = r[t];
+                _td[tid] = td[t];
+                item.barrier(sycl::access::fence_space::local_space);  //sync threads
+
+                const float _v = v[t];
+                float       y  = 0;
+
+                for (u_int j = 0; j < head_size; j += 4) {
+                    const sycl::float4 & k  = (sycl::float4 &) (_k[j]);
+                    const sycl::float4 & r  = (sycl::float4 &) (_r[j]);
+                    const sycl::float4 & td = (sycl::float4 &) (_td[j]);
+                    sycl::float4 &       s  = (sycl::float4 &) (state[j]);
+                    sycl::float4         kv;
+
+                    kv.x() = k.x() * _v;
+                    kv.y() = k.y() * _v;
+                    kv.z() = k.z() * _v;
+                    kv.w() = k.w() * _v;
+
+                    s.x() = s.x() * td.x() + kv.x();
+                    s.y() = s.y() * td.y() + kv.y();
+                    s.z() = s.z() * td.z() + kv.z();
+                    s.w() = s.w() * td.w() + kv.w();
+
+                    y += r.x() * s.x();
+                    y += r.y() * s.y();
+                    y += r.z() * s.z();
+                    y += r.w() * s.w();
+                }
+                dst[t] = y * scale;
+            }
+#pragma unroll
+            for (u_int i = 0; i < head_size; i++) {
+                dst[T * C + batch_i * state_size + head_i * head_size * head_size + i * head_size + tid] = state[i];
+            }
+        });
+    });
+}
+
+void ggml_sycl_op_gated_linear_attn(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    const float * k_d  = static_cast<const float *>(dst->src[0]->data);
+    const float * v_d  = static_cast<const float *>(dst->src[1]->data);
+    const float * r_d  = static_cast<const float *>(dst->src[2]->data);
+    const float * td_d = static_cast<const float *>(dst->src[3]->data);
+    const float * s_d  = static_cast<const float *>(dst->src[4]->data);
+
+    const int64_t B = dst->src[4]->ne[1];
+    const int64_t T = dst->src[0]->ne[2];
+    const int64_t C = dst->ne[0];
+    const int64_t H = dst->src[0]->ne[1];
+
+    dpct::queue_ptr stream = ctx.stream();
+    GGML_ASSERT(dst->src[4]->type == GGML_TYPE_F32);
+    GGML_ASSERT(C % H == 0);
+    GGML_ASSERT(C / H == 64 || C / H == 128);
+
+    float scale;
+    memcpy(&scale, dst->op_params, sizeof(float));
+
+    float * dst_d = (float *) dst->data;
+
+    if (C / H == 64) {
+        gated_linear_attn_f32_kernel<64>(stream, B, T, C, H, scale, k_d, v_d, r_d, td_d, s_d, dst_d);
+    } else {
+        gated_linear_attn_f32_kernel<128>(stream, B, T, C, H, scale, k_d, v_d, r_d, td_d, s_d, dst_d);
+    }
+}

ggml/src/ggml-sycl/gla.hpp

@@ -0,0 +1,8 @@
+#ifndef GGML_SYCL_GLA_HPP
+#define GGML_SYCL_GLA_HPP
+
+#include "common.hpp"
+
+void ggml_sycl_op_gated_linear_attn(ggml_backend_sycl_context & ctx, ggml_tensor * dst);
+
+#endif  // GGML_SYCL_GLA_HPP

ggml/src/ggml-sycl/outprod.cpp

@@ -3,9 +3,9 @@
 #include "outprod.hpp"
 
 
-void ggml_sycl_op_out_prod(ggml_backend_sycl_context& ctx, const ggml_tensor* src0,
-    const ggml_tensor* src1, ggml_tensor* dst) {
-
+void ggml_sycl_op_out_prod(ggml_backend_sycl_context& ctx, ggml_tensor* dst) {
+    const ggml_tensor *src0 = dst->src[0];
+    const ggml_tensor *src1 = dst->src[1];
 
     GGML_ASSERT(src0->type == GGML_TYPE_F32);
     GGML_ASSERT(src1->type == GGML_TYPE_F32);

ggml/src/ggml-sycl/outprod.hpp

@@ -3,8 +3,7 @@
 
 #include "common.hpp"
 
-void ggml_sycl_op_out_prod(ggml_backend_sycl_context& ctx, const ggml_tensor* src0,
-    const ggml_tensor* src1, ggml_tensor* dst);
+void ggml_sycl_op_out_prod(ggml_backend_sycl_context& ctx, ggml_tensor* dst);
 
 
 #endif // GGML_SYCL_OUTPROD_HPP

ggml/src/ggml-sycl/tsembd.cpp

@@ -55,8 +55,9 @@ static void timestep_embedding_f32_sycl(
         });
 }
 
-void ggml_sycl_op_timestep_embedding(ggml_backend_sycl_context & ctx, const ggml_tensor *src0,
-    const ggml_tensor *src1, ggml_tensor * dst) {
+void ggml_sycl_op_timestep_embedding(ggml_backend_sycl_context & ctx, ggml_tensor * dst) {
+    const ggml_tensor *src0 = dst->src[0];
+    const ggml_tensor *src1 = dst->src[1];
     const float * src0_d = (const float *)src0->data;
     float * dst_d = (float *)dst->data;
     dpct::queue_ptr stream = ctx.stream();