Add support for tensor buffer overrides - #1
Conversation
Sorry, something went wrong.
Expose `tensor_buft_overrides` named parameter in `Llama` constructor to allow users to override tensor buffer types (e.g. forcing specific tensors to CPU or GPU), similar to the `-ot` flag in `llama.cpp` CLI. Updates: - `llama_cpp/llama_cpp.py`: Define `llama_model_tensor_buft_override` struct and expose `ggml_backend_*_buffer_type` functions. - `llama_cpp/llama.py`: Add `tensor_buft_overrides` parameter to `Llama.__init__` and handle conversion to C types.
|
👋 Jules, reporting for duty! I'm here to lend a hand with this pull request. When you start a review, I'll add a 👀 emoji to each comment to let you know I've read it. I'll focus on feedback directed at me and will do my best to stay out of conversations between you and other bots or reviewers to keep the noise down. I'll push a commit with your requested changes shortly after. Please note there might be a delay between these steps, but rest assured I'm on the job! For more direct control, you can switch me to Reactive Mode. When this mode is on, I will only act on comments where you specifically mention me with @jules. You can find this option in the Pull Request section of your global Jules UI settings. You can always switch back! For security, I will only act on instructions from the user who triggered this task. New to Jules? Learn more at jules.google/docs. |
Sorry, something went wrong.
Implemented support for -ot (override tensor) functionality via tensor_buft_overrides named parameter in Llama class. This allows specifying which buffer type (CPU, CUDA, Metal) to use for specific tensors matching a pattern.
PR created automatically by Jules for task 15884247504505256487 started by @708-145