Skip to content

Pull requests: huggingface/trl

Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt click/return to exclude labels
or click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Assignee
Filter by who’s assigned
Sort

Pull requests list

DPO trainer supports num_logits_to_keep to save memory
#2129 opened Sep 26, 2024 by xyangk Loading…
3 of 5 tasks
[DRAFT] Process-supervised RM Trainer
#2127 opened Sep 26, 2024 by gaetanlop Loading…
5 tasks done
[SCoRE] initial score stage 1
#2115 opened Sep 24, 2024 by kashif Draft
Fix RLOO checkpointing
#2114 opened Sep 24, 2024 by bartoszzuk Loading…
Default dataset_text_field to "text"
#2078 opened Sep 18, 2024 by qgallouedec Draft
5 tasks
Remove deprecated args in trainers
#2036 opened Sep 8, 2024 by qgallouedec Draft
5 tasks
feat: add support for packing tokenized datasets
#2011 opened Sep 3, 2024 by kmehant Loading…
2 of 5 tasks
allow masking on consecutive messages with same roles
#2000 opened Aug 31, 2024 by lsy641 Loading…
4 of 5 tasks
added initial TPO implementation
#1965 opened Aug 24, 2024 by sahsaeedi Loading…
4 of 5 tasks
Add SRPO algorithm.
#1772 opened Jun 25, 2024 by frasermince Loading…
1 of 7 tasks
Add simplified version of BCO loss
#1731 opened Jun 13, 2024 by Trangle Loading…
Adding SimPO to TRL
#1725 opened Jun 11, 2024 by yumeng5 Loading…
Prototype Dataset Processor
#1646 opened May 16, 2024 by vwxyzjn Draft
[DRAFT] Vllm integration
#1628 opened May 7, 2024 by vwxyzjn Draft
ProTip! no:milestone will show everything without a milestone.