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

Remove padding and multiple D2D copies for MTP (#24086)

Browse Source 
* Make ggml_gated_delta_net take only the initial recurrent state (D, 1, n_seqs) and passes the snapshot count K as an op parameter instead of inferring it from state->ne[1].

Remove the padding hack and copy all emitted snapshots into the recurrent cache with a single strided ggml_cpy

* Make GDN changes in all backends. Address review comments.

* Fix CI build errors
This commit is contained in:
Gaurav Garg
2026-06-10 23:21:16 +05:30
committed by GitHub
parent d2462f8f7a
commit e95dae18d6
20 changed files with 118 additions and 106 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-metal/ggml-metal-device.cpp
+2 -2
View File
@@ -590,8 +590,8 @@ ggml_metal_pipeline_with_params ggml_metal_library_get_pipeline_gated_delta_net(
const int ne20 = op->src[2]->ne[0]; // S_v
const int ne21 = op->src[2]->ne[1]; // H
const int ne30 = op->src[3]->ne[0]; // G
// state is src[5], 3D (S_v*S_v*H, K, n_seqs); K is the snapshot slot count.
const int K = op->src[5]->ne[1];
// state is src[5], 4D [S_v, S_v, H_v, n_seqs] (s0 only); K is op param 0.
const int K = ggml_get_op_params_i32(op, 0);
const int nsg = op->src[2]->ne[0]/32;