intel/auto-round

Pull Requests Commits

fix several regressions including lm-head quantization, 3bit asym torch backend,etc (#796)

wenhuach21 committed 147 days ago

Verified 08d6b9e7

add fp8 export format check (#779)

n1ck-guo committed 147 days ago

Verified 5ead4c6d

align formats packing device to API (#795)

WeiweiZhang1 committed 149 days ago

Verified dc81a878

rename llmcompressor to llm_compressor for align with other formats (#780)

WeiweiZhang1 committed 149 days ago

Verified e66652a4

fix bug and add nvfp in alg-ext with slight improvement (#794)

wenhuach21 committed 150 days ago

Verified 7e014ca3

Support auto device mapping (#781)

Kaihui-intel committed 150 days ago

Verified d7d2efad

torch backend bugfix and speedup ut (#793)

WeiweiZhang1 committed 150 days ago

Verified 4c597def

add limit arg for eval (#764)

n1ck-guo committed 150 days ago

Verified add749ac

refine nvfp4 algorithm (#790)

wenhuach21 committed 150 days ago

Verified 536810e1

mxfp/nvfp/fp8 support torch compile in tuning (#789)

wenhuach21 committed 151 days ago

Verified 5f87a06b

refine nvfp code, typofix (#777)

WeiweiZhang1 committed 151 days ago

Verified fced2ba7

fix alg_ext regression and support mxfp4 in it with slight improvement (#785)

wenhuach21 committed 151 days ago

Verified ca3f733e

hot fix for nfp4 scheme (#784)

wenhuach21 committed 152 days ago

Verified 7e52af2a

Speedup FP4 packing (#766)

yiliu30 committed 152 days ago

Verified 253b48f2

[API CHANGE] Stage 1 add quant scheme and consolidate device and device_map (#774)

wenhuach21 committed 152 days ago

Verified 8bcf7eab

gguf format supports for fp8 model (#778)

n1ck-guo committed 153 days ago

Verified 572aad1d

fix bug of auto_round exporting (#772)

n1ck-guo committed 153 days ago

Verified d01704da

fix set_layer_config bug (#768)

wenhuach21 committed 157 days ago

Verified bf58cf97

Keep the model’s buffer dtype unchanged in most cases (#770)

wenhuach21 committed 158 days ago

Verified b2c7a3de

fix gguf bug by int zp (#771)

n1ck-guo committed 159 days ago

Verified f01e8047

fix gguf ut bug (#767)

n1ck-guo committed 159 days ago

Verified 48ab0db9

refine code and fix activation quantization eval regression (#762)

wenhuach21 committed 160 days ago

Verified 305923bc

Revert "Speedup FP4 packing (#760)" (#763)

yiliu30 committed 160 days ago

Verified ec76f612

support quant lm_head for rtn w8afp8 static quant (#754)

n1ck-guo committed 160 days ago

Verified de7ecc1b

[pre-commit.ci] auto fixes from pre-commit.com hooks

pre-commit-ci[bot] committed 161 days ago

041a66d7

Merge branch 'main' into hengguo/smoothquant

n1ck-guo committed 161 days ago

Verified 09db12b4

fix_script_fp_layer_config_for_bits_checking (#756)

WeiweiZhang1 committed 161 days ago

Verified 5e406991

Speedup FP4 packing (#760)

yiliu30 committed 161 days ago

Verified 3648639e

fix some torch compile warnings (#755)

wenhuach21 committed 161 days ago

Verified 1f1d0a9c

api support for fp8 model and mllm api support load from str (#752)

n1ck-guo committed 164 days ago

Verified fcbcaa48

Newer Older