evaluation_bloomz/evaluation_val/selqa/answer_selection_analysis/make-sense-rand/results.json

{
  "dataset_name": "selqa",
  "dataset_config_name": "answer_selection_analysis",
  "template_name": "make-sense-rand",
  "evaluation": {
    "accuracy": 0.9426751592356688
  },
  "arguments": "Namespace(config_name=None, dataset_config_name='answer_selection_analysis', dataset_name='selqa', 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='make-sense-rand', tokenizer_name=None, use_slow_tokenizer=False)"
}