Skip to content

Issues: pytorch/torchtune

v0.6.0 tracker
#2232 opened Jan 6, 2025 by joecummings
Open
Testing tracker
#1890 opened Oct 23, 2024 by felipemello1
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

252 Open 602 Closed
252 Open 602 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

Step based checkpointing
#2299 opened Jan 25, 2025 by xTRam1
Training never starts - stuck after Loss is intialized
#2297 opened Jan 24, 2025 by datamancerai
5
Tokens per second calculation best practice Things we should be doing but aren't triage review This issue should be discussed in weekly review
#2296 opened Jan 24, 2025 by EugenHotaj
7
Roadmap for distributed recipes using NPU as a backend
#2288 opened Jan 22, 2025 by Nicorgi
8 tasks
deepseek r1 support? enhancement New feature or request triage review This issue should be discussed in weekly review
#2287 opened Jan 21, 2025 by johnnynunez
8
Documentation for evaluation on a custom dataset for a custom task discussion Start a discussion documentation Improvements or additions to documentation triage review This issue should be discussed in weekly review
#2286 opened Jan 21, 2025 by karrtikiyer
@joecummings @RdoubleA
16
Saving multiple checkpoints per epoch enhancement New feature or request triaged This issue has been assigned an owner and appropriate label
#2285 opened Jan 21, 2025 by EugenHotaj
@joecummings
2
Inconsistent initialization of RoPE embedding across component builders best practice Things we should be doing but aren't better engineering Tasks which help improve eng productivity e.g. building tools, cleaning up code, writing docs
#2283 opened Jan 20, 2025 by Ankur-singh
[RFC] Proposal for tune cat Command discussion Start a discussion rfc Request for comments
#2281 opened Jan 19, 2025 by Ankur-singh
@Ankur-singh @joecummings
2
Roadmap for other parallelisms discussion Start a discussion triaged This issue has been assigned an owner and appropriate label
#2280 opened Jan 19, 2025 by rahul-sarvam
@acisseJZhong
5
Llama3.2 vision does not run with distributed state dict bug Something isn't working triaged This issue has been assigned an owner and appropriate label
#2277 opened Jan 17, 2025 by acisseJZhong
@mori360
1
The current instantiation does not trigger the initialization of submodules discussion Start a discussion triaged This issue has been assigned an owner and appropriate label
#2273 opened Jan 16, 2025 by dz1iang
@RdoubleA
4
DPO after / on top of LoRA tuning discussion Start a discussion triaged This issue has been assigned an owner and appropriate label
#2272 opened Jan 16, 2025 by albertbn
@joecummings
2
Don't use _get_clones best practice Things we should be doing but aren't community help wanted We would love the community's help completing this issue
#2270 opened Jan 16, 2025 by ebsmothers
8
About the CLS token for the llama3_2_vision_encoder discussion Start a discussion triaged This issue has been assigned an owner and appropriate label
#2268 opened Jan 15, 2025 by dfloreaa
@pbontrager @RdoubleA
4
Expose FSDP2 MixedPrecisionPolicy params enhancement New feature or request triaged This issue has been assigned an owner and appropriate label
#2267 opened Jan 14, 2025 by EugenHotaj
@joecummings
1
adding support for LR schedule for full distributed finetune best practice Things we should be doing but aren't better engineering Tasks which help improve eng productivity e.g. building tools, cleaning up code, writing docs triaged This issue has been assigned an owner and appropriate label
#2263 opened Jan 13, 2025 by tginart
@felipemello1
2
[RFC] Additional chat loss masking strategies community help wanted We would love the community's help completing this issue discussion Start a discussion enhancement New feature or request good first issue Good for newcomers rfc Request for comments
#2261 opened Jan 13, 2025 by RdoubleA
@supreethmanyam
2
Request: adding py.typed for type checkers better engineering Tasks which help improve eng productivity e.g. building tools, cleaning up code, writing docs triaged This issue has been assigned an owner and appropriate label
#2258 opened Jan 13, 2025 by jamesbraza
@joecummings
2
Qlora uses more memory than regular lora triaged This issue has been assigned an owner and appropriate label
#2255 opened Jan 11, 2025 by AndrewMead10
@ebsmothers
11
Very slow convergence with bf16 discussion Start a discussion triaged This issue has been assigned an owner and appropriate label
#2254 opened Jan 11, 2025 by EugenHotaj
@felipemello1
20
Lora and Dora finetuning produces identical results bug Something isn't working high-priority
#2250 opened Jan 10, 2025 by AndrewMead10
@ebsmothers
8
Finetuning Llama 3.1 8B Base Model on ChatML Format Dataset – Loss Reaches NaN After 2000 Steps triaged This issue has been assigned an owner and appropriate label
#2246 opened Jan 10, 2025 by abdul-456
@joecummings @ebsmothers
11
Overriding kv cache entries in torchtune models discussion Start a discussion triaged This issue has been assigned an owner and appropriate label
#2241 opened Jan 9, 2025 by telgamal-1
@joecummings @ebsmothers
2
Finetune meta-llama/Llama-Guard-3-1B bug Something isn't working triaged This issue has been assigned an owner and appropriate label
#2237 opened Jan 8, 2025 by jingzhaoou
@RdoubleA
11
ProTip! Follow long discussions with comments:>50.