SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

[Trainer] memory tracker metrics (#10225)

Browse Source 
* memory tracker metrics

* go back to eval for somewhat consistency

* handle no-gpu case

* deal with stackable eval calls

* restore callback order

* style

* simplify the API

* add test

* docs

* consistently use eval_ prefix

* improve docs

* Update src/transformers/trainer_utils.py

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* rename method

* style

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
This commit is contained in:
Stas Bekman
2021-02-18 09:27:32 -08:00
committed by GitHub
parent d7f38c5d1d
commit 97e688bc22
7 changed files with 294 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
examples/tests/deepspeed/test_deepspeed.py
View File

@@ -88,8 +88,8 @@ class TestDeepSpeed(TestCasePlus):
extra_args_str="--do_eval",
remove_args_str="--do_train",
)
val_metrics = load_json(os.path.join(output_dir, "val_results.json"))
assert "val_bleu" in val_metrics
val_metrics = load_json(os.path.join(output_dir, "eval_results.json"))
assert "eval_bleu" in val_metrics
# XXX: need to do better validation beyond just that the run was successful
def run_quick(self, distributed=True, extra_args_str=None, remove_args_str=None):