Skip to content

Issues: pytorch/torchtune

v0.5.0 tracker
#2008 opened Nov 14, 2024 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

225 Open 564 Closed
225 Open 564 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

How is "Full finetune 8B on 24GB card" achieved?
#2168 opened Dec 18, 2024 by fzyzcjy
Adding an export GH label
#2166 opened Dec 17, 2024 by Jack-Khuu
1
GPU Middle Class? discussion Start a discussion distributed Anything related to distributed env (multi-GPU, multi-node) triaged This issue has been assigned an owner and appropriate label
#2161 opened Dec 16, 2024 by EugenHotaj
3
DPO not saving adapter config bug Something isn't working
#2159 opened Dec 15, 2024 by SalmanMohammadi
1
@SalmanMohammadi
1
Move update_recipe_state to its own util best practice Things we should be doing but aren't triaged This issue has been assigned an owner and appropriate label
#2158 opened Dec 13, 2024 by joecummings
Add Multi-Dataset and streaming documentation
#2156 opened Dec 12, 2024 by andrewkho
@andrewkho
what should I do if I want to improve the performance of hellaswag? discussion Start a discussion
#2154 opened Dec 12, 2024 by mathCrazyy
@lindawangg @joecummings
3
Qwen2.5 does not support Qlora? discussion Start a discussion enhancement New feature or request
#2153 opened Dec 12, 2024 by mathCrazyy
@joecummings
1
Invalid kwarg fused passed to bitsandbytes AdamW8bit better engineering Tasks which help improve eng productivity e.g. building tools, cleaning up code, writing docs
#2152 opened Dec 12, 2024 by mlazos
@felipemello1
5
How to retrieve the distilled model in a manner similar to the OpenAI API interface ? discussion Start a discussion
#2148 opened Dec 11, 2024 by lingq1
@joecummings
12
Loss becomes NaN during finetuning when turning on optimizer_in_bwd=True
#2145 opened Dec 10, 2024 by acisseJZhong
2
70B Fine-tuning GPUs Utilization discussion Start a discussion distributed Anything related to distributed env (multi-GPU, multi-node)
#2142 opened Dec 10, 2024 by fabiogeraci
@joecummings
4
Are there any plans to support context parallel? enhancement New feature or request
#2141 opened Dec 10, 2024 by dz1iang
@felipemello1
2
[small bug + generalization] saving config.yaml to output_dir better engineering Tasks which help improve eng productivity e.g. building tools, cleaning up code, writing docs bug Something isn't working
#2137 opened Dec 9, 2024 by felipemello1
1
Query on Gradient accumulation discussion Start a discussion
#2134 opened Dec 9, 2024 by Vattikondadheeraj
@ebsmothers
1
want to fine-tuned llama3.2.1b on MMLU and Arc_challenge and gsm8k(maths) discussion Start a discussion
#2132 opened Dec 8, 2024 by sorobedio
@joecummings
1
Make it possible to distill into a full finetune model enhancement New feature or request
#2122 opened Dec 6, 2024 by joecummings
2
Does it support distillation for large models like Qwen2-72B and LLaMA 3.1-70B? discussion Start a discussion
#2117 opened Dec 6, 2024 by lingq1
@lindawangg
9
[RFC] Remove automatic weight merging when training LoRA discussion Start a discussion
#2115 opened Dec 5, 2024 by felipemello1
@joecummings
3
[RFC] Unify activation checkpointing APIs rfc Request for comments
#2114 opened Dec 5, 2024 by ebsmothers
@ebsmothers
3
Flex attention error bug Something isn't working
#2113 opened Dec 5, 2024 by ye-jin-shop
@joecummings
2
Make it easier to support chunked-style losses
#2106 opened Dec 3, 2024 by felipemello1
[Feature Request]: Support logging MFU as a metrics
#2100 opened Dec 2, 2024 by xingyaoww
2
Multi GPU timeout on save checkpoint (WorkNCCL, Watchdog, timeout) distributed Anything related to distributed env (multi-GPU, multi-node) inference Anything related to our inference capabilities
#2093 opened Nov 29, 2024 by albertbn
@joecummings @felipemello1
13
TensorBoardLogger: AttributeError: module 'tensorflow' has no attribute 'io' bug Something isn't working
#2090 opened Nov 28, 2024 by fabiogeraci
@joecummings
11
ProTip! Exclude everything labeled bug with -label:bug.