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

Add siqa and csqa tasks #158

Merged
merged 1 commit into from
Dec 6, 2023
Merged

Add siqa and csqa tasks #158

merged 1 commit into from
Dec 6, 2023

Conversation

yulinggu-cs
Copy link
Contributor

Add Social IQA and Commonsense QA tasks.

Test from running: PYTHONPATH="/Users/yulingg/My_Git_Repositories/catwalk" python llm_eval/run_lm_eval.py --model lm::pretrained=EleutherAI/pythia-160m,revision=step140000 --task social_iqa csqa --split validation --full_output_file predictions.jsonl --metrics_file metrics.json --model_max_length 2048 --limit 10 --max_batch_tokens 4096 --num_recorded_inputs 3 --num_shots 0:

... 'correct_choice': 1, 'unconditioned_input': [('Answer:', ' cheating'), ('Answer:', ' a bad child'), ('Answer:', ' interested in making money')], 
'metrics': {'acc_raw': 0, 'predicted_index_raw': 0, 'acc_per_token': 0, 'predicted_index_per_token': 2, 'acc_per_char': 0, 'predicted_index_per_char': 2, 'acc_uncond': 0, 'predicted_index_uncond': 2, 'probability_mass': 2.3027600944029188e-07, 'acc': 0}}, 
'model_input': [('Question: Lee sold things belonging to her mom without her permission. How would Lee feel afterwards?\nAnswer:', 
' cheating'),
 ('Question: Lee sold things belonging to her mom without her permission. How would Lee feel afterwards?\nAnswer:', 
' a bad child'),
 ('Question: Lee sold things belonging to her mom without her permission. How would Lee feel afterwards?\nAnswer:', 
' interested in making money')]} 
...
 'correct_choice': 2, 'unconditioned_input': [('Answer:', ' hospital'), ('Answer:', ' park'), ('Answer:', ' corner'), ('Answer:', ' motel'), ('Answer:', ' office')], 
'metrics': {'acc_raw': 0, 'predicted_index_raw': 1, 'acc_per_token': 0, 'predicted_index_per_token': 1, 'acc_per_char': 0, 'predicted_index_per_char': 0, 'acc_uncond': 0, 'predicted_index_uncond': 1, 'probability_mass': 9.05275873202387e-06, 'acc': 0}},
 'model_input': [('Question: Many towns and cities have trash cans where on sidewalks?\nAnswer:', 
' hospital'),  ('Question: Many towns and cities have trash cans where on sidewalks?\nAnswer:', 
' park'), ('Question: Many towns and cities have trash cans where on sidewalks?\nAnswer:', 
' corner'), ('Question: Many towns and cities have trash cans where on sidewalks?\nAnswer:', 
' motel'), ('Question: Many towns and cities have trash cans where on sidewalks?\nAnswer:', 
' office')]} 

Copy link
Contributor

@OyvindTafjord OyvindTafjord left a comment

Choose a reason for hiding this comment

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

Thanks, LGTM!

@OyvindTafjord OyvindTafjord merged commit cfca1bf into olmo-eval Dec 6, 2023
@OyvindTafjord OyvindTafjord deleted the add-siqa branch December 6, 2023 14:26
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants