diff options
| author | Iwan Kawrakow <iwan.kawrakow@gmail.com> | 2024-08-08 16:27:43 +0200 |
|---|---|---|
| committer | Kawrakow <48489457+ikawrakow@users.noreply.github.com> | 2024-08-09 16:00:31 +0200 |
| commit | a829cb7794996b2cedccce242ecc08917ce9ce7a (patch) | |
| tree | 79b42ff425fc5a1cc2d1d9ec94b77700e8f67cf2 /examples/pydantic_models_to_grammar_examples.py | |
| parent | 48c4389e3d616cda898ad4c12612b99c22f45e0d (diff) | |
iq6_k: Metal
About 4% slower than Q6_K for PP-512, but 10% faster for TG-128.
Someone has screwed up Q6_K TG performance on Metal? With the
cobntinuous "improvements" in ggml I wouldn't be surprised.
Need to look into it later.
Diffstat (limited to 'examples/pydantic_models_to_grammar_examples.py')
0 files changed, 0 insertions, 0 deletions