| {"Muennighoff/xstory_cloze_ar": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7968232958305758}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9232296492389146}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6677696889477167}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9265387160820648}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9126406353408338}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_es": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8729318332230311}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9417604235605559}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7359364659166115}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9430840502978161}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9318332230311053}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_eu": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7054930509596293}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8663136995367307}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6320317670416943}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8689609530112509}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8524156187954997}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_hi": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.798808735936466}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8702845797485109}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6604897418927862}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8788881535407015}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.870946393117141}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_id": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8557246856386499}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9212442091330245}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7041694242223693}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9205823957643945}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9066843150231635}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_zh": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.900066181336863}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9232296492389146}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.684976836532098}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9311714096624751}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', 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='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.9199205823957644}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xwinograd_en": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6847311827956989}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5135483870967742}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6787096774193548}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5053763440860215}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.690752688172043}, "template_name": "underscore refer to"}}, "Muennighoff/xwinograd_fr": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6506024096385542}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.4939759036144578}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6867469879518072}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.46987951807228917}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6626506024096386}, "template_name": "underscore refer to"}}, "Muennighoff/xwinograd_pt": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', 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='test', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6349809885931559}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', 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='test', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.4866920152091255}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', 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='test', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6387832699619772}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', 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='test', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.49429657794676807}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', 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='test', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6425855513307985}, "template_name": "underscore refer to"}}, "Muennighoff/xwinograd_zh": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6865079365079365}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5277777777777778}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6884920634920635}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.4861111111111111}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', 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='validation', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6904761904761905}, "template_name": "underscore refer to"}}, "anli_dev_r1": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.486}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.427}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.474}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.39}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.46}, "template_name": "justified in saying"}}, "anli_dev_r2": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.441}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.406}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.426}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.36}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.419}, "template_name": "justified in saying"}}, "anli_dev_r3": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.455}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.42}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.445}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.32083333333333336}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', 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='validation', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.4266666666666667}, "template_name": "justified in saying"}}, "story_cloze_2016": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', 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='validation', target_max_length=256, template_config_name=None, template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.9567076429716729}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', 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='validation', target_max_length=256, template_config_name=None, template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.9625868519508284}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', 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='validation', target_max_length=256, template_config_name=None, template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.7814003206841261}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', 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='validation', target_max_length=256, template_config_name=None, template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.9577765900587921}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', 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='validation', target_max_length=256, template_config_name=None, template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.951362907536077}, "template_name": "Story Continuation and Options"}}, "super_glue_cb": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.8214285714285714}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.375}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.8214285714285714}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.7321428571428571}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.7678571428571429}, "template_name": "justified in saying"}}, "super_glue_copa": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='C1 or C2? premise, so/because\u2026', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.62}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.87}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.88}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.91}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.88}, "template_name": "plausible_alternatives"}}, "super_glue_rte": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.8303249097472925}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.855595667870036}, "template_name": "MNLI crowdsource"}, "does it follow that": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='does it follow that', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.7833935018050542}, "template_name": "does it follow that"}, "guaranteed true": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='guaranteed true', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.8122743682310469}, "template_name": "guaranteed true"}, "should assume": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', 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='validation', target_max_length=256, template_config_name=None, template_name='should assume', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.8194945848375451}, "template_name": "should assume"}}, "winogrande_winogrande_xl": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', 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='validation', target_max_length=256, template_config_name=None, template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.584846093133386}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', 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='validation', target_max_length=256, template_config_name=None, template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.5217048145224941}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', 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='validation', target_max_length=256, template_config_name=None, template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.5840568271507498}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', 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='validation', target_max_length=256, template_config_name=None, template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.5114443567482242}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', 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='validation', target_max_length=256, template_config_name=None, template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.5927387529597474}, "template_name": "underscore refer to"}}, "xcopa_id": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise, so/because\u2026', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.55}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.78}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.86}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.79}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.84}, "template_name": "plausible_alternatives"}}, "xcopa_sw": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise, so/because\u2026', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.57}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.6}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.6}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.64}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.62}, "template_name": "plausible_alternatives"}}, "xcopa_ta": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise, so/because\u2026', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.58}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.67}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.67}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.68}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.69}, "template_name": "plausible_alternatives"}}, "xcopa_vi": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise, so/because\u2026', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.55}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.83}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.87}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.84}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.86}, "template_name": "plausible_alternatives"}}, "xcopa_zh": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise, so/because\u2026', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.55}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.83}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.9}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.9}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', 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='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.86}, "template_name": "plausible_alternatives"}}, "xnli_ar": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5357429718875502}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.41004016064257026}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5606425702811245}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.6068273092369478}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5437751004016064}, "template_name": "justified in saying"}}, "xnli_en": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.6168674698795181}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.45502008032128516}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.6092369477911647}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.6746987951807228}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5895582329317269}, "template_name": "justified in saying"}}, "xnli_es": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.585140562248996}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4357429718875502}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5883534136546185}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.6124497991967871}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5734939759036145}, "template_name": "justified in saying"}}, "xnli_fr": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5771084337349398}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.43012048192771085}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5807228915662651}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.6136546184738956}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5694779116465863}, "template_name": "justified in saying"}}, "xnli_hi": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5248995983935743}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3795180722891566}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5506024096385542}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5682730923694779}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5353413654618474}, "template_name": "justified in saying"}}, "xnli_sw": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4795180722891566}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.39196787148594375}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5208835341365462}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5036144578313253}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5184738955823294}, "template_name": "justified in saying"}}, "xnli_ur": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.46586345381526106}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3718875502008032}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5080321285140562}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4995983935742972}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5080321285140562}, "template_name": "justified in saying"}}, "xnli_vi": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5578313253012048}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.42449799196787147}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5678714859437751}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.6100401606425703}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5538152610441767}, "template_name": "justified in saying"}}, "xnli_zh": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5526104417670683}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.38473895582329315}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5690763052208835}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5674698795180723}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', 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='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5622489959839357}, "template_name": "justified in saying"}}} |