huggingface / huggingface/picotron
Why do we need to call `init_model_parameters` after `load_state_dict`
Open
- Dominant language
- Python
- Stars
- 2.3k
- Forks
- 203
- PR merge metrics
- No merged PRs in 30d
Description
In [code](https://github.com/huggingface/picotron/blob/main/picotron/checkpoint.py#L111), we call `init_model_parameters` after `load_state_dict`. It will overwrite weights from checkpoint.
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.