Skip to content
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.

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

VHELM improvements #2825

Merged
merged 6 commits into from
Aug 1, 2024
Merged

VHELM improvements #2825

merged 6 commits into from
Aug 1, 2024

Conversation

teetone
Copy link
Member

@teetone teetone commented Jul 22, 2024

Changes

@teetone teetone added the VHELM Holistic Evaluation of Vision-Language Models (VLM) label Jul 22, 2024
src/helm/benchmark/run_specs/vlm_run_specs.py Show resolved Hide resolved
src/helm/benchmark/run_specs/vlm_run_specs.py Show resolved Hide resolved
src/helm/benchmark/run_specs/vlm_run_specs.py Show resolved Hide resolved

# Save the image to disk
image = row["image"]
image_file_name: str = generate_hash(image) + ".jpg"
Copy link
Collaborator

@yifanmai yifanmai Jul 29, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why not just name this f"{split}_{row_index}.jpg"? Then we won't need hashes.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I want a unique hash per image because I worried the underlying huggingface dataset could get reshuffled.

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You can deal with this by passing the githash e.g. revision="e9488045cbad16c973f031c7a8f7466b5dcc3794" to load_dataset(), then you don't have to worry about mutations. We should probably do this for all other usages of load_dataset(), also.

@teetone teetone requested a review from yifanmai July 31, 2024 04:36
Copy link
Collaborator

@yifanmai yifanmai left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

OK to land, remaining comments are optional.

@teetone teetone merged commit d3397d0 into main Aug 1, 2024
9 checks passed
@teetone teetone deleted the image2structfix branch August 1, 2024 09:48
This was referenced Aug 5, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
VHELM Holistic Evaluation of Vision-Language Models (VLM)
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants