riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-08-02 19:21:31 +06:00
Code Issues Actions 3 Packages Projects Releases Wiki Activity

Optimize inference only mode memory if ipex is used (#21083)

Browse Source 
* Optimize inference only mode memory if ipex is used

Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>

* fix code style

Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>

Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>
This commit is contained in:
Wang, Yi 2023-01-12 17:01:17 +08:00 committed by GitHub
parent 6767ce71d6
commit e849e5bb4a
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 2 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/transformers/trainer.py
View File

@ -1309,8 +1309,9 @@ class Trainer:
if not training:
model.eval()
dtype = torch.bfloat16 if not self.is_in_train and self.args.bf16_full_eval else dtype
# conv_bn_folding is disabled as it fails in symbolic tracing, resulting in ipex warnings
model = ipex.optimize(model, dtype=dtype, level="O1", conv_bn_folding=False)
model = ipex.optimize(model, dtype=dtype, level="O1", conv_bn_folding=False, inplace=not self.is_in_train)
else:
if not model.training:
model.train()