Skip to content

Improve mason.py by allowing to pass local models for GCP and bubble up dataset cache errors #797

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
mnoukhov merged 2 commits into from
Jul 17, 2025
+9 −2
Merged

Improve mason.py by allowing to pass local models for GCP and bubble up dataset cache errors #797

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
d2f6104
allow passing local models, bubble up dataset cache errors
mnoukhov Jul 17, 2025
09741f7
Merge branch 'main' into fix-mason-cacheing-error
mnoukhov Jul 17, 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
11 changes: 9 additions & 2 deletions mason.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -635,6 +635,8 @@ def remove_arg_from_list(lst: List[str], item: str, remove_value: bool = False):
dataset_config_hashes.append(dataset_config_hash)
stderr = result.stderr
return_code = result.returncode
if return_code != 0:
raise Exception(f"Error code {return_code} when creating cached dataset")
console.log("✅✅✅ Finished running the caching command")

if file in OPEN_INSTRUCT_RESUMABLES and idx != -1 and len(args.auto_checkpoint_state_dir) > 0:
Expand Down Expand Up @@ -706,8 +708,13 @@ def remove_arg_from_list(lst: List[str], item: str, remove_value: bool = False):
break

commit_hash = get_commit_hash(model_name_or_path, model_revision, "config.json", "model")
download_from_hf(model_name_or_path, model_revision) # first download the model
path = download_from_hf(model_name_or_path, model_revision) # then get the path
if os.path.exists(model_name_or_path):
path = model_name_or_path
model_name_or_path = os.path.basename(model_name_or_path)
console.log(f"Local model is already downloaded, using path basename as model name {model_name_or_path}, note that commit hash is {commit_hash}")
else:
download_from_hf(model_name_or_path, model_revision) # first download the model
path = download_from_hf(model_name_or_path, model_revision) # then get the path
gs_saved_path = f"gs://ai2-llm/post-training/deletable_cache_models/{model_name_or_path}/{commit_hash}"
gs_folder = gs_folder_exists(gs_saved_path) # race condition exists, but it's fine since we are launching mason sequentially
if not gs_folder:
Expand Down