evaluation_xnlihtmt/xnliht/merged.json

{"xnli_ar": {"GPT-3 style_arht": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='GPT-3 style_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4457831325301205}, "template_name": "GPT-3 style_arht"}, "MNLI crowdsource_arht": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='MNLI crowdsource_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3461847389558233}, "template_name": "MNLI crowdsource_arht"}, "can we infer_arht": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='can we infer_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4807228915662651}, "template_name": "can we infer_arht"}, "guaranteed/possible/impossible_arht": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='guaranteed/possible/impossible_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.41847389558232934}, "template_name": "guaranteed/possible/impossible_arht"}, "justified in saying_arht": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='justified in saying_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40200803212851405}, "template_name": "justified in saying_arht"}}, "xnli_es": {"GPT-3 style_esht": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='GPT-3 style_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4875502008032129}, "template_name": "GPT-3 style_esht"}, "MNLI crowdsource_esht": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='MNLI crowdsource_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "MNLI crowdsource_esht"}, "can we infer_esht": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='can we infer_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "can we infer_esht"}, "guaranteed/possible/impossible_esht": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='guaranteed/possible/impossible_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.48714859437751}, "template_name": "guaranteed/possible/impossible_esht"}, "justified in saying_esht": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='justified in saying_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "justified in saying_esht"}}, "xnli_fr": {"GPT-3 style_frht": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='GPT-3 style_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5028112449799197}, "template_name": "GPT-3 style_frht"}, "MNLI crowdsource_frht": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='MNLI crowdsource_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3542168674698795}, "template_name": "MNLI crowdsource_frht"}, "can we infer_frht": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='can we infer_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.46546184738955826}, "template_name": "can we infer_frht"}, "guaranteed/possible/impossible_frht": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='guaranteed/possible/impossible_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3887550200803213}, "template_name": "guaranteed/possible/impossible_frht"}, "justified in saying_frht": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='justified in saying_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5321285140562249}, "template_name": "justified in saying_frht"}}, "xnli_hi": {"GPT-3 style_hiht": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='GPT-3 style_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "GPT-3 style_hiht"}, "MNLI crowdsource_hiht": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='MNLI crowdsource_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.42610441767068274}, "template_name": "MNLI crowdsource_hiht"}, "can we infer_hiht": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='can we infer_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40562248995983935}, "template_name": "can we infer_hiht"}, "guaranteed/possible/impossible_hiht": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='guaranteed/possible/impossible_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3485943775100402}, "template_name": "guaranteed/possible/impossible_hiht"}, "justified in saying_hiht": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='justified in saying_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.41767068273092367}, "template_name": "justified in saying_hiht"}}, "xnli_sw": {"GPT-3 style_swht": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='GPT-3 style_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.348995983935743}, "template_name": "GPT-3 style_swht"}, "MNLI crowdsource_swht": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='MNLI crowdsource_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "MNLI crowdsource_swht"}, "can we infer_swht": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='can we infer_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.35943775100401604}, "template_name": "can we infer_swht"}, "guaranteed/possible/impossible_swht": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='guaranteed/possible/impossible_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3586345381526104}, "template_name": "guaranteed/possible/impossible_swht"}, "justified in saying_swht": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='justified in saying_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3570281124497992}, "template_name": "justified in saying_swht"}}, "xnli_ur": {"GPT-3 style_urht": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='GPT-3 style_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.41767068273092367}, "template_name": "GPT-3 style_urht"}, "MNLI crowdsource_urht": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='MNLI crowdsource_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.342570281124498}, "template_name": "MNLI crowdsource_urht"}, "can we infer_urht": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='can we infer_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3337349397590361}, "template_name": "can we infer_urht"}, "guaranteed/possible/impossible_urht": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='guaranteed/possible/impossible_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3963855421686747}, "template_name": "guaranteed/possible/impossible_urht"}, "justified in saying_urht": {"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='justified in saying_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3365461847389558}, "template_name": "justified in saying_urht"}}, "xnli_vi": {"GPT-3 style_viht": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='GPT-3 style_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4759036144578313}, "template_name": "GPT-3 style_viht"}, "MNLI crowdsource_viht": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='MNLI crowdsource_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.38835341365461845}, "template_name": "MNLI crowdsource_viht"}, "can we infer_viht": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='can we infer_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.41244979919678715}, "template_name": "can we infer_viht"}, "guaranteed/possible/impossible_viht": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='guaranteed/possible/impossible_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.41445783132530123}, "template_name": "guaranteed/possible/impossible_viht"}, "justified in saying_viht": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='justified in saying_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40401606425702813}, "template_name": "justified in saying_viht"}}, "xnli_zh": {"GPT-3 style_zhht": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='GPT-3 style_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.44096385542168676}, "template_name": "GPT-3 style_zhht"}, "MNLI crowdsource_zhht": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='MNLI crowdsource_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.35943775100401604}, "template_name": "MNLI crowdsource_zhht"}, "can we infer_zhht": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='can we infer_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.351004016064257}, "template_name": "can we infer_zhht"}, "guaranteed/possible/impossible_zhht": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='guaranteed/possible/impossible_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3526104417670683}, "template_name": "guaranteed/possible/impossible_zhht"}, "justified in saying_zhht": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b2-xp3capmixnewcodelonglossseq/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='justified in saying_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3538152610441767}, "template_name": "justified in saying_zhht"}}}