metal : add CONV_TRANSPOSE_2D (#16542)

* initial: headers and metal-device.cpp updates

* adding conv_transpose_2d

* fix type

* fix type: int32->int64

* Update ggml/src/ggml-metal/ggml-metal.metal

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* Update ggml/src/ggml-metal/ggml-metal.metal

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* Update ggml/src/ggml-metal/ggml-metal.metal

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* add checks for src[0] and src[1]; add type checks

* Update ggml-metal.metal

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* add more tests, add optimization to threading

* add dynamic memory allocation in metal

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Ilia Ilmer
2025-10-17 02:33:58 -04:00
committed by GitHub
parent 79967ec596
commit 9ad4f1931e
8 changed files with 198 additions and 0 deletions

View File

@@ -653,6 +653,11 @@ bool ggml_metal_device_supports_op(ggml_metal_device_t dev, const struct ggml_te
case GGML_OP_SCALE:
case GGML_OP_CONV_TRANSPOSE_1D:
return true;
case GGML_OP_CONV_TRANSPOSE_2D:
return ggml_is_contiguous(op->src[0]) && ggml_is_contiguous(op->src[1]) &&
(op->src[0]->type == GGML_TYPE_F16 || op->src[0]->type == GGML_TYPE_F32) &&
op->src[1]->type == GGML_TYPE_F32 &&
op->type == GGML_TYPE_F32;
case GGML_OP_CLAMP:
return op->src[0]->type == GGML_TYPE_F32;
case GGML_OP_SQR: