ayo/ipex-llm
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

LLM: Fix bigdl_ipex_int8 warning (#10890)

Browse source
This commit is contained in:
Wang, Jian4 2024-04-26 11:18:44 +08:00 committed by GitHub
parent fb3c268d13
commit 3e8ed54270
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
python/llm/src/ipex_llm/transformers/convert_ipex.py
View file

@ -138,7 +138,7 @@ def _ipex_optimize_model(model, rms_classes, qtype):
} }
qconfig = ipex.quantization.get_weight_only_quant_qconfig_mapping( qconfig = ipex.quantization.get_weight_only_quant_qconfig_mapping(
weight_dtype=torch.qint8, # INT8 weight_dtype=torch.qint8, # INT8
lowp_mode=ipex.quantization.WoqLowpMode.INT8, lowp_mode=ipex.quantization.WoqLowpMode.BF16,
act_quant_mode=act_quant_mode_dict["PER_IC_BLOCK"], act_quant_mode=act_quant_mode_dict["PER_IC_BLOCK"],
group_size=-1, group_size=-1,
) )