Skip to content

Issues: pytorch/torchtune

v0.4.0 Tracker
#1747 opened Oct 2, 2024 by joecummings
Open
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Clear current search query, filters, and sorts
26 Open 24 Closed
26 Open 24 Closed
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
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

recipe for hyperparameter sweep enhancement New feature or request
#1752 opened Oct 4, 2024 by felipemello1
2
lr scheduler is not optional enhancement New feature or request
#1751 opened Oct 4, 2024 by felipemello1
1
Can we support class weight in the CEWithChunkedOutputLoss class enhancement New feature or request
#1746 opened Oct 2, 2024 by ye-jin-shop
@felipemello1
4
mistral-nemo support community help wanted We would love the community's help completing this issue enhancement New feature or request
#1716 opened Sep 30, 2024 by win4r
mistral-small support community help wanted We would love the community's help completing this issue enhancement New feature or request
#1715 opened Sep 30, 2024 by win4r
Feature Request: Full Support for Direct Preference Optimization (DPO) enhancement New feature or request
#1532 opened Sep 10, 2024 by pipSu
3
How to disable Checkpointing for Full tuning or PEFT runs? enhancement New feature or request question Further information is requested
#1487 opened Sep 4, 2024 by premmotgi
4
Bring to torchtune memory wins for generate.py enhancement New feature or request
#1448 opened Aug 29, 2024 by felipemello1
Add configurable system prompt in generate community help wanted We would love the community's help completing this issue enhancement New feature or request
#1393 opened Aug 22, 2024 by RdoubleA
Message validation in SFTDataset community help wanted We would love the community's help completing this issue enhancement New feature or request
#1391 opened Aug 22, 2024 by RdoubleA
Streamline/better documentation for torchtune -> transformers workflow discussion Start a discussion documentation Improvements or additions to documentation enhancement New feature or request
#1388 opened Aug 21, 2024 by SalmanMohammadi
Add more coding models enhancement New feature or request
#1336 opened Aug 14, 2024 by joecummings
1
Add LR Scheduler to full finetune recipes community help wanted We would love the community's help completing this issue enhancement New feature or request
#1308 opened Aug 12, 2024 by joecummings
4
[FEATURE REQUEST] Compile model+loss_fn together enhancement New feature or request
#1228 opened Jul 26, 2024 by gau-nernst
@felipemello1
25
torch.compile support enhancement New feature or request
#1187 opened Jul 17, 2024 by youngsheen
@ebsmothers
2
[feature request] Saving / Loading packed dataset enhancement New feature or request help wanted Extra attention is needed
#1149 opened Jul 8, 2024 by ScottHoang
6
want dora and nef-tune supports! enhancement New feature or request
#1100 opened Jun 20, 2024 by jeffchy
@calvinpelletier
3
Recommendations for obtaining validation dataset loss after each epoch enhancement New feature or request
#1042 opened Jun 1, 2024 by dcsuka
3
Make objects in notebooks importable in config enhancement New feature or request
#1003 opened May 18, 2024 by RdoubleA
@RdoubleA
Apply QLoRA to output projections and token embedding enhancement New feature or request
#1000 opened May 17, 2024 by rohan-varma
4
No param to control save checkpoints every N steps ? enhancement New feature or request
#988 opened May 16, 2024 by apachemycat
4
lm harness distributed evaluation? enhancement New feature or request
#930 opened May 3, 2024 by monk1337
@joecummings
3
Validation and early stopping during training community help wanted We would love the community's help completing this issue enhancement New feature or request high-priority
#883 opened Apr 26, 2024 by kinggongzilla
@ebsmothers
5
How to save a trained model so it can be loaded with HF from_pretrained()? enhancement New feature or request
#832 opened Apr 22, 2024 by calmitchell617
@joecummings @kartikayk
29
Running LoRA + QLoRA on Colab Notebooks enhancement New feature or request
#828 opened Apr 21, 2024 by kartikayk
@ebsmothers @kartikayk
1
ProTip! Add no:assignee to see everything that’s not assigned.