evaluation_val/riddle_sense/most_suitable_answer/results.json

{
  "dataset_name": "riddle_sense",
  "dataset_config_name": null,
  "template_name": "most_suitable_answer",
  "evaluation": {
    "accuracy": 0.24877571008814886
  },
  "arguments": "Namespace(config_name=None, dataset_config_name=None, dataset_name='riddle_sense', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/commun/experiments/muennighoff/bloomckpt/760mt0/xp3capmixlossseq_global_step250', output_dir='/gpfsscratch/rech/six/commun/commun/experiments/muennighoff/bloomckpt/760mt0/xp3capmixlossseq_global_step250/evaluation', pad_to_max_length=False, per_device_eval_batch_size=4, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='most_suitable_answer', tokenizer_name=None, use_slow_tokenizer=False)"
}