intel/auto-round

Pull Requests Commits

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

wenhuach21 committed 224 days ago

Verified ca3f733e

hot fix for nfp4 scheme (#784)

wenhuach21 committed 224 days ago

Verified 7e52af2a

Speedup FP4 packing (#766)

yiliu30 committed 224 days ago

Verified 253b48f2

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

wenhuach21 committed 224 days ago

Verified 8bcf7eab

gguf format supports for fp8 model (#778)

n1ck-guo committed 225 days ago

Verified 572aad1d

fix bug of auto_round exporting (#772)

n1ck-guo committed 225 days ago

Verified d01704da

fix set_layer_config bug (#768)

wenhuach21 committed 230 days ago

Verified bf58cf97

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

wenhuach21 committed 230 days ago

Verified b2c7a3de

fix gguf bug by int zp (#771)

n1ck-guo committed 231 days ago

Verified f01e8047

fix gguf ut bug (#767)

n1ck-guo committed 231 days ago

Verified 48ab0db9

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

wenhuach21 committed 232 days ago

Verified 305923bc

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

yiliu30 committed 232 days ago

Verified ec76f612

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

n1ck-guo committed 232 days ago

Verified de7ecc1b

fix_script_fp_layer_config_for_bits_checking (#756)

WeiweiZhang1 committed 233 days ago

Verified 5e406991

Speedup FP4 packing (#760)

yiliu30 committed 233 days ago

Verified 3648639e

fix some torch compile warnings (#755)

wenhuach21 committed 233 days ago

Verified 1f1d0a9c

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

n1ck-guo committed 236 days ago

Verified fcbcaa48

fix alg_ext moe and model str input bug (#751)

wenhuach21 committed 236 days ago

Verified 8f84bcb7

support deepspeed LinearLayer and LinearAllreduce (#698)

xin3he committed 236 days ago

Verified 8c2df83a

mllm supports loading fp8 model and fix bug of loading fp8 model (#750)

n1ck-guo committed 237 days ago

Verified c1e719b9

refine code (#749)

wenhuach21 committed 237 days ago

Verified e9b2b83b

Add Static FP8 KV Support (#737)

yiliu30 committed 237 days ago

Verified fe1db091

improved algorithm for int2 (#748)

wenhuach21 committed 238 days ago

Verified 64e76116

fix bug of loading fp8 model (#747)

n1ck-guo committed 238 days ago

Verified 00293227

support llama4 quant (#744)

mengniwang95 committed 238 days ago

Verified 2df63f27

fix UT check for int zp (#745)

WeiweiZhang1 committed 238 days ago

Verified 1f79db9b

Optimize pack zeros for int sym (#743)

WeiweiZhang1 committed 239 days ago

Verified 8888dda7

support auto-round-mllm command (#742)

n1ck-guo committed 240 days ago

Verified cd980c71

[WIP] support for glm4.5 gguf (#735)

n1ck-guo committed 243 days ago

Verified 4f740ccb

Enable mxfp exporting (#649)

WeiweiZhang1 committed 243 days ago

Verified 64b8afee

Older