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

Fix vipllava for generation (#29874)

Browse Source 
* fix vipllava generation

* consistent llava code

* revert llava tests changes
This commit is contained in:
Raushan Turganbay
2024-04-03 21:00:08 +05:00
committed by GitHub
parent 240e10626b
commit cc75f1ac73
3 changed files with 7 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tests/models/llava_next/test_modeling_llava_next.py
View File

@@ -423,7 +423,7 @@ class LlavaNextForConditionalGenerationIntegrationTest(unittest.TestCase):
output = model(**inputs)
expected_slice = torch.tensor(
[[-4.7695, -4.5664, -0.2786], [-10.6172, -10.8906, -2.5234], [-6.7344, -7.2422, -0.6758]],
[[-4.7695, -4.5664, -0.2786], [-10.6250, -10.8906, -2.5254], [-6.7383, -7.2461, -0.6787]],
dtype=torch.float32,
device=torch_device,
)