riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-07-31 18:22:34 +06:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

Specifying torch dtype in Qwen2VLForConditionalGeneration (#33953)

Browse Source 
* Specifying torch dtype

* Reverting change & changing fallback _from_config() dtype
This commit is contained in:
Hamza Tahboub 2024-10-10 05:39:33 -07:00 committed by GitHub
parent f8a260e2a4
commit dda3f91d06
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/transformers/modeling_utils.py
View File

@ -1499,7 +1499,7 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
torch_dtype (`torch.dtype`, *optional*): torch_dtype (`torch.dtype`, *optional*):
Override the default `torch.dtype` and load the model under this dtype. Override the default `torch.dtype` and load the model under this dtype.
""" """
torch_dtype = kwargs.pop("torch_dtype", None) torch_dtype = kwargs.pop("torch_dtype", torch.get_default_dtype())
use_flash_attention_2 = kwargs.pop("use_flash_attention_2", False) use_flash_attention_2 = kwargs.pop("use_flash_attention_2", False)
# override default dtype if needed # override default dtype if needed