diff options
| author | Georgi Gerganov <ggerganov@gmail.com> | 2024-05-11 10:32:41 +0300 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-05-11 10:32:41 +0300 |
| commit | 9cb317f77e53067f7a138cc89ef7657148eae8e6 (patch) | |
| tree | 3ba1d2d80d1d7c8b4ab01f6396a3febaae26e91b /ggml-cuda.cu | |
| parent | e849648888a11de13aaaa4cb2eda3f5a9c7b444d (diff) | |
ggml : full ALiBi support (#7192)
* ggml : full ALiBi support
* ggml : update ggml_soft_max_ext() CUDA, SYCL
* ggml : ggml_flash_attn_ext() support ALiBi (CPU)
* ggml : ggml_flash_attn_ext() support ALiBi (Metal)
* ggml : fix warning
* ggml : ggml_flash_attn_ext() support ALiBi (CUDA)
ggml-ci
* ggml : fix assert message
* vulkan : add dev notes
* ggml : require mask when using ALiBi
ggml-ci
* convert : fix convert for refact models
Diffstat (limited to 'ggml-cuda.cu')
| -rw-r--r-- | ggml-cuda.cu | 5 |
1 files changed, 0 insertions, 5 deletions
diff --git a/ggml-cuda.cu b/ggml-cuda.cu index 6f89a7cc..c5c77879 100644 --- a/ggml-cuda.cu +++ b/ggml-cuda.cu @@ -4,7 +4,6 @@ #include "ggml-cuda/common.cuh" #include "ggml-cuda/acc.cuh" -#include "ggml-cuda/alibi.cuh" #include "ggml-cuda/arange.cuh" #include "ggml-cuda/argsort.cuh" #include "ggml-cuda/binbcast.cuh" @@ -2277,9 +2276,6 @@ 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_ALIBI: - ggml_cuda_op_alibi(ctx, dst); - break; case GGML_OP_IM2COL: ggml_cuda_op_im2col(ctx, dst); break; @@ -2829,7 +2825,6 @@ GGML_CALL static bool ggml_backend_cuda_supports_op(ggml_backend_t backend, cons case GGML_OP_DIAG_MASK_INF: case GGML_OP_SOFT_MAX: case GGML_OP_ROPE: - case GGML_OP_ALIBI: case GGML_OP_IM2COL: case GGML_OP_POOL_2D: case GGML_OP_SUM_ROWS: |