diff options
| author | Kawrakow <48489457+ikawrakow@users.noreply.github.com> | 2024-07-27 07:55:01 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-07-27 07:55:01 +0200 |
| commit | 154e0d75fccf1784fe9ff6fd76a630b66563da3d (patch) | |
| tree | 81ce6dbb5b1900c1aa78a879f0593c694cab9d27 /examples/quantize-stats | |
| parent | 0684c3e9c70d49323b4fc517128cbe222cab7f96 (diff) | |
Merge mainline llama.cpp (#3)
* Merging mainline - WIP
* Merging mainline - WIP
AVX2 and CUDA appear to work.
CUDA performance seems slightly (~1-2%) lower as it is so often
the case with llama.cpp/ggml after some "improvements" have been made.
* Merging mainline - fix Metal
* Remove check
---------
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
Diffstat (limited to 'examples/quantize-stats')
| -rw-r--r-- | examples/quantize-stats/quantize-stats.cpp | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/examples/quantize-stats/quantize-stats.cpp b/examples/quantize-stats/quantize-stats.cpp index 4eb8f953..837a17c7 100644 --- a/examples/quantize-stats/quantize-stats.cpp +++ b/examples/quantize-stats/quantize-stats.cpp @@ -154,7 +154,7 @@ static void test_roundtrip_on_chunk( } if (use_reference) { - qfns.from_float_reference(input_scratch, quantized_scratch, chunk_size); + qfns.from_float_ref(input_scratch, quantized_scratch, chunk_size); } else { qfns.from_float(input_scratch, quantized_scratch, chunk_size); } |