ExternalVendorCode/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/whisper.cpp.git synced 2025-04-19 16:41:10 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

cuda/vulkan: specify fp32-only support for some operations in supports_op (ggml/1129)

Browse Source 
* cuda: restrict SILU_BACK to fp32, since fp16 exceeds the desired test threshold

* vulkan: specify fp32-only support for certain ops (that are now tested for fp16 as well)

* f32 sigmoid in vulkan supports op

* Revert "f32 sigmoid in vulkan supports op"

This reverts commit c6f04b3c19bf4504c2776149c6d8cd84e0b48acb.
This commit is contained in:
cmdr2 2025-02-28 15:59:55 +05:30 committed by Georgi Gerganov
parent 60d2ddebdf
commit 6ac8e6b2ce
2 changed files with 7 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
ggml/src/ggml-cuda/ggml-cuda.cu
View File

@ -3153,7 +3153,7 @@ static bool ggml_backend_cuda_device_supports_op(ggml_backend_dev_t dev, const g
return false;
} break;
case GGML_OP_SILU_BACK:
return ggml_is_contiguous(op->src[0]);
return ggml_is_contiguous(op->src[0]) && op->src[0]->type == GGML_TYPE_F32;
break;
case GGML_OP_NORM:
case GGML_OP_RMS_NORM:

11
ggml/src/ggml-vulkan/ggml-vulkan.cpp
View File

@ -8371,7 +8371,7 @@ static bool ggml_backend_vk_device_supports_op(ggml_backend_dev_t dev, const ggm
case GGML_UNARY_OP_SILU:
case GGML_UNARY_OP_RELU:
case GGML_UNARY_OP_TANH:
return ggml_is_contiguous(op->src[0]);
return ggml_is_contiguous(op->src[0]) && op->src[0]->type == GGML_TYPE_F32;
default:
return false;
}
@ -8571,17 +8571,18 @@ static bool ggml_backend_vk_device_supports_op(ggml_backend_dev_t dev, const ggm
case GGML_OP_RMS_NORM:
return ggml_is_contiguous(op->src[0]);
case GGML_OP_ADD:
case GGML_OP_ACC:
case GGML_OP_SUB:
case GGML_OP_MUL:
case GGML_OP_DIV:
case GGML_OP_CONCAT:
case GGML_OP_UPSCALE:
case GGML_OP_SCALE:
case GGML_OP_SQR:
case GGML_OP_SIN:
case GGML_OP_COS:
case GGML_OP_CLAMP:
return op->src[0]->type == GGML_TYPE_F32;
case GGML_OP_ACC:
case GGML_OP_CONCAT:
case GGML_OP_UPSCALE:
case GGML_OP_SCALE:
case GGML_OP_PAD:
case GGML_OP_DIAG_MASK_INF:
case GGML_OP_SOFT_MAX: