Skip to content

Issues: pytorch/torchtitan

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
94 Open 580 Closed
94 Open 580 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

update _config_file to train_config*s* (fix break) CLA Signed This label is managed by the Meta Open Source bot.
#16 by lessw2020 was merged Jan 25, 2024 Loading…
1
Add metrics to collect during training
#3 by wanchaol was closed Apr 2, 2024 First OSS release of Torchtrain
4
4
Add lr scheduler for optimizer
#4 by wanchaol was closed Feb 7, 2024 training techniques enablement
3
add meta initilization for the model
#5 by wanchaol was closed Mar 5, 2024 training techniques enablement
1
2
add fsdp to the model
#6 by wanchaol was closed Feb 7, 2024 3D Parallel enablement
1
add AC/selective AC to the model
#7 by wanchaol was closed Mar 19, 2024 training techniques enablement
1
@tianyu-l
add torch.compile to the model
#8 by wanchaol was closed May 3, 2024 training techniques enablement
2
update readme with local setups
#9 by wanchaol was closed Feb 23, 2024
enable distributed checkpoint save/load
#10 by wanchaol was closed Feb 23, 2024 torchtrain infrastructure building
add Sequence Parallel fp8 support to train llama
#11 by wanchaol was closed May 3, 2024 FP8 Enablement
add pipeline parallelism to train llama enhancement New feature or request
#12 by wanchaol was closed Jun 12, 2024 3D Parallel enablement
@wconstab @kwen2501
2
setup ci and linter
#13 by wanchaol was closed Apr 26, 2024 torchtrain infrastructure building
2
Add profiler CLA Signed This label is managed by the Meta Open Source bot.
#1 by lessw2020 was merged Jan 19, 2024 Loading…
19
integrate torch.compile, controllable via config CLA Signed This label is managed by the Meta Open Source bot.
#14 by lessw2020 was merged Jan 25, 2024 Loading…
18
Add Tensorboard
#17 by wanchaol was closed Feb 23, 2024 torchtrain infrastructure building
2
@tianyu-l
correct default download dir for download_tokenizer.py CLA Signed This label is managed by the Meta Open Source bot.
#18 by lessw2020 was closed Feb 15, 2024 Loading…
2
FSDP gradient norm clipping
#19 by wanchaol was closed Feb 7, 2024 training techniques enablement
1
@tianyu-l
1
Implement depthwise truncated model init
#20 by lessw2020 was closed Feb 23, 2024 training techniques enablement
2
Add meta init CLA Signed This label is managed by the Meta Open Source bot.
#21 by lessw2020 was closed Mar 6, 2024 Loading…
13
add context parallelism when ready enhancement New feature or request
#22 by lessw2020 was closed Oct 29, 2024
@fegin
1
add linear lr warmup and lr decay scheduler CLA Signed This label is managed by the Meta Open Source bot.
#23 by lessw2020 was merged Feb 1, 2024 Loading…
12
[torchtrain] add gradient scaler CLA Signed This label is managed by the Meta Open Source bot.
#25 by tianyu-l was merged Jan 31, 2024 Loading…
1
9
Enable checkpointing with DCP CLA Signed This label is managed by the Meta Open Source bot.
#26 by fegin was merged Feb 6, 2024 Loading…
20
Add ParallelDims helper and --pp_degree option CLA Signed This label is managed by the Meta Open Source bot.
#27 by wconstab was merged Jan 31, 2024 Loading…
6
Implement sft
#673 by aniltrkkn was closed Nov 8, 2024 Loading…
1
ProTip! Adding no:label will show everything without a label.