Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-24 13:28:50 +01:00
Code Issues Packages Projects Releases Wiki Activity

cuda : mark I16 and I32 ops as unsupported

Browse Source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2024-01-03 13:01:44 +02:00
parent 75e3fd8581
commit d55356d3ba
1 changed files with 10 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
ggml-cuda.cu
View File

@ -10039,14 +10039,22 @@ static bool ggml_backend_cuda_supports_op(ggml_backend_t backend, const ggml_ten
}
return false;
} break;
case GGML_OP_DUP:
case GGML_OP_REPEAT:
case GGML_OP_CONCAT:
{
ggml_type src0_type = op->src[0]->type;
if (src0_type != GGML_TYPE_I32 && src0_type != GGML_TYPE_I16) {
return true;
}
return false;
} break;
case GGML_OP_NONE:
case GGML_OP_RESHAPE:
case GGML_OP_VIEW:
case GGML_OP_PERMUTE:
case GGML_OP_TRANSPOSE:
case GGML_OP_NORM:
case GGML_OP_REPEAT:
case GGML_OP_DUP:
case GGML_OP_ADD:
case GGML_OP_MUL:
case GGML_OP_DIV:
@ -10063,7 +10071,6 @@ static bool ggml_backend_cuda_supports_op(ggml_backend_t backend, const ggml_ten
case GGML_OP_SUM_ROWS:
case GGML_OP_ARGSORT:
case GGML_OP_ACC:
case GGML_OP_CONCAT:
case GGML_OP_GROUP_NORM:
case GGML_OP_UPSCALE:
case GGML_OP_PAD: