transformers
feat: Sequential beam search(a.k.a Low-memory beam search)
#26304
Merged
Merged
feat: Sequential beam search(a.k.a Low-memory beam search) #26304
gante
merged
18 commits into
huggingface:main
from
epfl-dlab:fix_issue_22639
gante
commented
on 2023-10-13
Saibo-creator
changed the title
add sequential beam search (clean version)
4680d171
fix bug in key_value split, this solves the problem of output mismatch
98f17f5b
update the logger message with accordant weights_file_name (#28181)
04234f6c
[`Llava`] Fix llava index errors (#28032)
30493fac
[`Awq`] Enable the possibility to skip quantization for some target m…
b5d8e260
fix FA2 when using quantization (#28203)
010566e2
small typo (#28229)
8030da3e
Update docs around mixing hf scheduler with deepspeed optimizer (#28223)
8d85c889
Fix trainer saving safetensors: metadata is None (#28219)
62911a38
fix bug:divide by zero in _maybe_log_save_evaluate() (#28251)
2bbb953b
[Whisper] Fix errors with MPS backend introduced by new code on word-…
b18598dd
Remove fast tokenization warning in Data Collators (#28213)
e2181ac7
![dependabot[bot]](https://avatars.githubusercontent.com/in/29110?v=4&s=60)
Bump tj-actions/changed-files from 22.2 to 41 in /.github/workflows (…
14a86990
Merge branch 'main' into fix_issue_22639
6054c33e
gante
approved these changes
on 2024-01-10
refactor _split_model_inputs function for sequential beam search.
d78d0382
fix error with test_modeling_whisper.py
5f568c86
fix syntax error(feature only available in python3.10)
1c3fd14a
fix typo
270c1f59
ArthurZucker
approved these changes
on 2024-01-19
Saibo-creator
changed the title
Saibo-creator
changed the title
Login to write a write a comment.
Login via GitHub