Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

metal : add opt_step_adamw and op_sum (#16529)

Browse Source 
* scaffold to support opt step adamw on metal (not written so far)

* add opt-step-adamw kernel for metal

* pass op->src[4] as a separate buffer to the pipeline

* add bounds check to opt-step-adamw kernel

* complete scaffold for GGML_OP_SUM

* naive GGML_OP_SUM kernel

* remove unwanted comment

* change OP_SUM capability gate

* Add has_simdgroup_reduction to both ops to pass CI
This commit is contained in:
Sam/Samuel
2025-10-12 21:43:14 +03:00
committed by GitHub
parent 81d54bbfd5
commit a31cf36ad9
7 changed files with 172 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-metal/ggml-metal-ops.h
+2
View File
@@ -50,6 +50,7 @@ int ggml_metal_op_scale (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_clamp (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_unary (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_glu (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_sum (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_sum_rows (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_get_rows (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_set_rows (ggml_metal_op_t ctx, int idx);
@@ -78,6 +79,7 @@ int ggml_metal_op_timestep_embedding(ggml_metal_op_t ctx, int idx);
int ggml_metal_op_argmax (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_argsort (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_leaky_relu (ggml_metal_op_t ctx, int idx);
int ggml_metal_op_opt_step_adamw (ggml_metal_op_t ctx, int idx);
#ifdef __cplusplus
}