Skip to content

Update mast launcher info and adjust for latest torchtitan #229

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

Merged
merged 3 commits into from
Nov 5, 2025
Merged

Update mast launcher info and adjust for latest torchtitan #229

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
adc49e6
Update mast launcher info
wconstab Nov 4, 2025
c385791
remove print_args option deleted from upstream torchtitan
wconstab Nov 5, 2025
ad9c003
Fix llama3_auto_parallel name to match upstream torchtitan/autoparallel
wconstab Nov 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions mast/.torchxconfig
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,13 +3,14 @@ conda_path_in_fbpkg = conda
activate_conda = False
fbpkg_ids = fb-py-spy:prod
hpcIdentity = pytorch_distributed
rmAttribution = pytorch4all_clients_approved
rmAttribution = msl_infra_pytorch_dev
workspace_fbpkg_name = torchtitan_workspace
conda_pack_ignore_missing_files = True
git = False
hpcJobOncall = meta_conda
modelTypeName = gen_ai_conda
hpcClusterUuid = MastProdCluster
hpcClusterUuid = MastGenAICluster
localityConstraints = region;gtn
forceSingleRegion = False
use_caf = False

Expand Down
1 change: 0 additions & 1 deletion mast/run_torchtitan.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -78,5 +78,4 @@ python torchtitan/train.py \
--validation.dataset_path "${dataset_path}" \
--metrics.save_tb_folder "${save_tb_folder}" \
--metrics.disable_color_printing \
--job.print_args \
$overrides
12 changes: 6 additions & 6 deletions mast/sweep.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -104,12 +104,12 @@ def maybe_find_pulp(maybe_path: Optional[str] = None) -> Optional[str]:
],
"llama3_autop_1d_compile": llama3_1d_common_opts
+ [
"--model.name=llama3_auto_parallel",
"--model.name=auto_parallel.llama3",
"--compile.enable",
],
"llama3_autop_1d_compile_bucket_reorder": llama3_1d_common_opts
+ [
"--model.name=llama3_auto_parallel",
"--model.name=auto_parallel.llama3",
"--compile.enable",
"--experimental.bucket_all_gathers_fx=fsdp",
"--experimental.bucket_reduce_scatters_fx=fsdp",
Expand All @@ -125,12 +125,12 @@ def maybe_find_pulp(maybe_path: Optional[str] = None) -> Optional[str]:
],
"llama3_autop_2d_compile": llama3_2d_common_opts
+ [
"--model.name=llama3_auto_parallel",
"--model.name=auto_parallel.llama3",
"--compile.enable",
],
"llama3_autop_2d_compile_bucket_reorder": llama3_2d_common_opts
+ [
"--model.name=llama3_auto_parallel",
"--model.name=auto_parallel.llama3",
"--compile.enable",
"--experimental.bucket_all_gathers_fx=fsdp",
"--experimental.bucket_reduce_scatters_fx=fsdp",
Expand All @@ -153,13 +153,13 @@ def maybe_find_pulp(maybe_path: Optional[str] = None) -> Optional[str]:
| {
"llama3_autop_1d_compile_ruisi_bucket_reorder": llama3_1d_common_opts
+ [
"--model.name=llama3_auto_parallel",
"--model.name=auto_parallel.llama3",
"--compile.enable",
"--experimental.enable_simplefsdp_passes",
],
"llama3_autop_2d_compile_ruisi_bucket_reorder": llama3_2d_common_opts
+ [
"--model.name=llama3_auto_parallel",
"--model.name=auto_parallel.llama3",
"--compile.enable",
"--experimental.enable_simplefsdp_passes",
],
Expand Down
Loading