evaluation_bloomz-7b1-p3/evaluation_val/selqa/answer_selection_analysis/would-make-sense-qu-rand/results.json

{
  "dataset_name": "selqa",
  "dataset_config_name": "answer_selection_analysis",
  "template_name": "would-make-sense-qu-rand",
  "evaluation": {
    "accuracy": 0.8993630573248408
  },
  "arguments": "Namespace(config_name=None, dataset_config_name='answer_selection_analysis', dataset_name='selqa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/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='would-make-sense-qu-rand', tokenizer_name=None, use_slow_tokenizer=False)"
}