ggml: bump submodule for CUDA k-quant GET_ROWS support

Browse Source

Device-side embedding and codebook lookups now cover q2_K to q6_K, so a
quantized token_embd no longer drops the graph out of the direct device
path. i-quants are left as a TODO.

...

This commit is contained in:

Pascal

2026-07-19 21:29:06 +02:00

parent 7c4fea29ce

commit e93292bee1

1 changed files with 1 additions and 1 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

ggml

+1 -1

Submodule ggml updated: 05777aac23...f9f45e143d