bloomz
/
evaluation_bloomz
/evaluation_val
/pubmed_qa
/pqa_labeled
/Long_Answer_to_Final_Decision
/results.json
{ | |
"dataset_name": "pubmed_qa", | |
"dataset_config_name": "pqa_labeled", | |
"template_name": "Long Answer to Final Decision", | |
"evaluation": { | |
"accuracy": 0.704 | |
}, | |
"arguments": "Namespace(config_name=None, dataset_config_name='pqa_labeled', dataset_name='pubmed_qa', debug=False, dtype='bfloat16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/commun/experiments/muennighoff/bloomckpt/176bt0/xp3capmixnewcodelonglossseq_global_step498', output_dir='/gpfsscratch/rech/six/commun/commun/experiments/muennighoff/bloomckpt/176bt0/xp3capmixnewcodelonglossseq_global_step498/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='train', target_max_length=256, template_config_name=None, template_name='Long Answer to Final Decision', tokenizer_name=None, use_slow_tokenizer=False)" | |
} |