mirror of
https://github.com/ggerganov/llama.cpp.git
synced 2024-12-25 05:48:47 +01:00
f4d7e54974
* iq3_xxs: quantize/dequantize RMSE seems a bit high-ish at about half-way between q2_K and q3_K, so need to check more. * iq3_xxs: CUDA dequantize works * iq2_xxs: tuning quantization * iq3_xxs: starting to look better PPL on wiki.test.raw LLaMA-v1-7B: 6.4218 LLaMA-v2-7B: 6.3560 Mistral-7B : 6.0717 This is better than Q3_K_XS, with a 5% reduction in quantized model size. * iq3_xxs: CUDA dot product We have PP-512: 5891 t/s TG-128: 143.9 t/s * iq3_xxs: scalar and AVX2 dot products * iq3_xxs: ARM_NEON and Metal Metal performance is decent, ARM_NEON is pathetic * iq3_xxs: slightly better grid points * Faster iq3_xxs and iq2_xs dot products on CUDA * iq3_xxs: add some quant mix * iq3_xxs: fix failing quantization test Dot product still fails. Is this real? * iq3_xxs: hopefully fix ROCm * iq3_xxs: failing tests This time the dot product accuracy did find an actual bug in the AVX2 implementation. * Add IQ3_XXS to test-backend-ops --------- Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com> |
||
|---|---|---|
| .. | ||
| .gitignore | ||
| CMakeLists.txt | ||
| get-model.cpp | ||
| get-model.h | ||
| test-autorelease.cpp | ||
| test-backend-ops.cpp | ||
| test-c.c | ||
| test-double-float.cpp | ||
| test-grad0.cpp | ||
| test-grammar-parser.cpp | ||
| test-llama-grammar.cpp | ||
| test-model-load-cancel.cpp | ||
| test-opt.cpp | ||
| test-quantize-fns.cpp | ||
| test-quantize-perf.cpp | ||
| test-rope.cpp | ||
| test-sampling.cpp | ||
| test-tokenizer-0-falcon.cpp | ||
| test-tokenizer-0-falcon.py | ||
| test-tokenizer-0-llama.cpp | ||
| test-tokenizer-0-llama.py | ||
| test-tokenizer-1-bpe.cpp | ||
| test-tokenizer-1-llama.cpp | ||