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

Revert low cpu mem tie weights (#29135)

Browse Source 
* Revert "Add tie_weights() to LM heads and set bias in set_output_embeddings() (#28948)"

This reverts commit 725f4ad1cc.

* Revert "Patch to skip failing `test_save_load_low_cpu_mem_usage` tests (#29043)"

This reverts commit 4156f517ce.
This commit is contained in:
amyeroberts
2024-02-20 12:06:46 +00:00
committed by GitHub
parent 15cfe38942
commit 0996a10077
26 changed files with 0 additions and 144 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/models/bert_generation/test_modeling_bert_generation.py
View File

@@ -305,12 +305,6 @@ class BertGenerationEncoderTest(ModelTesterMixin, GenerationTesterMixin, Pipelin
model = BertGenerationEncoder.from_pretrained("google/bert_for_seq_generation_L-24_bbc_encoder")
self.assertIsNotNone(model)
@unittest.skip(
"Not currently compatible. Fails with - NotImplementedError: Cannot copy out of meta tensor; no data!"
)
def test_save_load_low_cpu_mem_usage(self):
pass
@require_torch
class BertGenerationEncoderIntegrationTest(unittest.TestCase):