benchmark_scores/result_mgsm.json

{
  "results": {
    "mgsm": {
      "acc": 0.248,
      "acc_stderr": 0.027367497504863555
    }
  },
  "versions": {
    "mgsm": 1.0
  },
  "config": {
    "model": "hf-causal-experimental",
    "model_args": "pretrained=HachiML/youri-2x7b_dev,tokenizer=HachiML/youri-2x7b_dev,use_accelerate=True,dtype=auto",
    "num_fewshot": [
      5
    ],
    "batch_size": 2,
    "device": "cuda",
    "no_cache": false,
    "limit": null,
    "bootstrap_iters": 100000,
    "description_dict": null
  }
}