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

[generation] bring back tests on vision models (#38603)

Browse Source 
* bring back geenration tests on VLMs

* remove head mask tests overwritten
This commit is contained in:
Raushan Turganbay
2025-06-06 10:23:15 +02:00
committed by GitHub
parent 90c4b90a10
commit dbfc79c17c
14 changed files with 66 additions and 272 deletions
Download Patch File Download Diff File Expand all files Collapse all files

34
tests/models/longt5/test_modeling_longt5.py
View File

@@ -626,40 +626,6 @@ class LongT5ModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTesterMix
model = LongT5Model.from_pretrained(model_name)
self.assertIsNotNone(model)
def test_generate_with_head_masking(self):
attention_names = ["encoder_attentions", "decoder_attentions", "cross_attentions"]
config_and_inputs = self.model_tester.prepare_config_and_inputs()
config = config_and_inputs[0]
max_length = config_and_inputs[1].shape[-1] + 3
model = LongT5ForConditionalGeneration(config).eval()
model.to(torch_device)
head_masking = {
"head_mask": torch.zeros(config.num_layers, config.num_heads, device=torch_device),
"decoder_head_mask": torch.zeros(config.num_decoder_layers, config.num_heads, device=torch_device),
"cross_attn_head_mask": torch.zeros(config.num_decoder_layers, config.num_heads, device=torch_device),
}
for attn_name, (name, mask) in zip(attention_names, head_masking.items()):
head_masks = {name: mask}
# Explicitly pass decoder_head_mask as it is required from LONGT5 model when head_mask specified
if name == "head_mask":
head_masks["decoder_head_mask"] = torch.ones(
config.num_decoder_layers, config.num_heads, device=torch_device
)
out = model.generate(
config_and_inputs[1],
num_beams=1,
max_length=max_length,
output_attentions=True,
return_dict_in_generate=True,
**head_masks,
)
# We check the state of decoder_attentions and cross_attentions just from the last step
attn_weights = out[attn_name] if attn_name == attention_names[0] else out[attn_name][-1]
self.assertEqual(sum([w.sum().item() for w in attn_weights]), 0.0)
def test_attention_outputs(self):
if not self.has_attentions:
self.skipTest(reason="has_attentions is set to False")