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

264 Open 630 Closed
264 Open 630 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

Phi3 mini 128k Instruct Config
#2394 opened Feb 14, 2025 by sujayrittikar
1
torch._inductor.exc.LoweringException: NoValidChoicesError using torch 2.6.0
#2392 opened Feb 13, 2025 by nathan-az
2
parallelize_module() parallelism module fqn wildcard doesn't work for llama3.2
#2379 opened Feb 11, 2025 by acisseJZhong
1
pyproject.toml wrong dev deps organization
#2375 opened Feb 10, 2025 by bogdansalyp
1
2
expose max-autotune in configs for better perf community help wanted We would love the community's help completing this issue
#2373 opened Feb 10, 2025 by felipemello1
2
Disable reshard_after_forward for last transformer layer FSDP param group distributed Anything related to distributed env (multi-GPU, multi-node) triaged This issue has been assigned an owner and appropriate label
#2372 opened Feb 10, 2025 by SalmanMohammadi
@SalmanMohammadi
2
Adding new role "Tool" to support Llama 3.3 models
#2371 opened Feb 9, 2025 by init27
@Ankur-singh
4
Misleading import error message for torchao
#2369 opened Feb 9, 2025 by bogdansalyp
1
Resume from checkpoint broken with distributed optimizer-in-backward
#2360 opened Feb 7, 2025 by ebsmothers
@mori360
Incorrect Default Config File Paths for Llama 3.1 8B and Qwen 2.5 7B Models
#2352 opened Feb 6, 2025 by MaxHastings
1
Discussion: Update dataloader to skip rows that dont require training best practice Things we should be doing but aren't discussion Start a discussion triage review This issue should be discussed in weekly review
#2344 opened Feb 5, 2025 by felipemello1
@felipemello1
4
CEWithChunkedOutputLoss does not check division by zero discussion Start a discussion triaged This issue has been assigned an owner and appropriate label
#2341 opened Feb 4, 2025 by pocca2048
@joecummings
6
Feature request: GRPO support
#2340 opened Feb 4, 2025 by tikikun
5
Seed: null isn't random bug Something isn't working triaged This issue has been assigned an owner and appropriate label
#2338 opened Feb 3, 2025 by bogdansalyp
1
Apply gradient accumulation fix to DPO/PPO recipes
#2334 opened Feb 3, 2025 by bogdansalyp
@ebsmothers
2
Distributed DPO loss normalization by amount of tokens
#2333 opened Feb 3, 2025 by bogdansalyp
2
FIRE Relative Positional Encodings
#2325 opened Jan 31, 2025 by kaddu341
8
Reading TorchProfiler after run
#2323 opened Jan 31, 2025 by fabiogeraci
6
Classifiers (reward models) in torchtune
#2320 opened Jan 30, 2025 by EugenHotaj
4
How to run torchtune on AMD Instinct MI300X
#2319 opened Jan 30, 2025 by kailashg26
2
Rename and document cleanup_before_training
#2315 opened Jan 29, 2025 by joecummings
Investigate the optimal scenario in which to use torch_set_num_thread()
#2312 opened Jan 29, 2025 by joecummings
Unable to reproduce QAT results from Blog
#2310 opened Jan 29, 2025 by AbhinavDutta
9
Differing component implementation logic across recipes 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 bug Something isn't working triaged This issue has been assigned an owner and appropriate label
#2307 opened Jan 29, 2025 by EugenHotaj
4
Move from PIL to torchvision.io.decode_image best practice Things we should be doing but aren't community help wanted We would love the community's help completing this issue
#2303 opened Jan 28, 2025 by ebsmothers
@Ankur-singh
12
ProTip! Exclude everything labeled bug with -label:bug.