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

Fix deprecated PT functions (#37237)

Browse Source 
* Fix deprecated PT functions

Signed-off-by: cyy <cyyever@outlook.com>

* Revert some changes

Signed-off-by: cyy <cyyever@outlook.com>

---------

Signed-off-by: cyy <cyyever@outlook.com>
This commit is contained in:
cyyever
2025-04-04 19:31:11 +08:00
committed by GitHub
parent b016de1ae4
commit edd345b52e
3 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
tests/generation/test_utils.py
View File

@@ -2724,7 +2724,7 @@ class UtilsFunctionsTest(unittest.TestCase):
# Case 1
input_ids = torch.randint(0, 16, (2, 8), dtype=torch.int64)[:, :0]
inputs_embeds = torch.rand((2, 8), dtype=torch.float32)
cache_position = torch.range(0, 7, dtype=torch.int64)
cache_position = torch.arange(0, 8, dtype=torch.int64)
eager1, eager2 = GenerationMixin()._cache_dependant_input_preparation(input_ids, inputs_embeds, cache_position)
export1, export2 = GenerationMixin()._cache_dependant_input_preparation_exporting(
input_ids, inputs_embeds, cache_position
@@ -2735,7 +2735,7 @@ class UtilsFunctionsTest(unittest.TestCase):
# Case 2
input_ids = torch.randint(0, 16, (2, 8), dtype=torch.int64)
inputs_embeds = torch.rand((2, 8), dtype=torch.float32)
cache_position = torch.range(0, 7, dtype=torch.int64)
cache_position = torch.arange(0, 8, dtype=torch.int64)
eager1, eager2 = GenerationMixin()._cache_dependant_input_preparation(input_ids, inputs_embeds, cache_position)
export1, export2 = GenerationMixin()._cache_dependant_input_preparation_exporting(
input_ids, inputs_embeds, cache_position
@@ -2746,7 +2746,7 @@ class UtilsFunctionsTest(unittest.TestCase):
# Case 3
input_ids = torch.randint(0, 16, (2, 12), dtype=torch.int64)
inputs_embeds = None
cache_position = torch.range(0, 7, dtype=torch.int64)
cache_position = torch.arange(0, 8, dtype=torch.int64)
eager1, eager2 = GenerationMixin()._cache_dependant_input_preparation(input_ids, inputs_embeds, cache_position)
export1, export2 = GenerationMixin()._cache_dependant_input_preparation_exporting(
input_ids, inputs_embeds, cache_position
@@ -2757,7 +2757,7 @@ class UtilsFunctionsTest(unittest.TestCase):
# Case 4
input_ids = torch.randint(0, 16, (2, 8), dtype=torch.int64)
inputs_embeds = None
cache_position = torch.range(0, 7, dtype=torch.int64)
cache_position = torch.arange(0, 8, dtype=torch.int64)
eager1, eager2 = GenerationMixin()._cache_dependant_input_preparation(input_ids, inputs_embeds, cache_position)
export1, export2 = GenerationMixin()._cache_dependant_input_preparation_exporting(
input_ids, inputs_embeds, cache_position

2
tests/models/falcon_mamba/test_modeling_falcon_mamba.py
View File

@@ -231,7 +231,7 @@ class FalconMambaModelTester:
token_emb, cache, cache_position=torch.arange(0, config.conv_kernel, device=input_ids.device)
)
loss = torch.log(1 + torch.abs(outputs.sum()))
loss = torch.log1p(torch.abs(outputs.sum()))
self.parent.assertEqual(loss.shape, ())
self.parent.assertEqual(outputs.shape, (self.batch_size, self.seq_length, self.hidden_size))
loss.backward()

2
tests/models/mamba/test_modeling_mamba.py
View File

@@ -205,7 +205,7 @@ class MambaModelTester:
token_emb, cache, cache_position=torch.arange(0, config.conv_kernel, device=input_ids.device)
)
loss = torch.log(1 + torch.abs(outputs.sum()))
loss = torch.log1p(torch.abs(outputs.sum()))
self.parent.assertEqual(loss.shape, ())
self.parent.assertEqual(outputs.shape, (self.batch_size, self.seq_length, self.hidden_size))
loss.backward()