MLLM_leaderboard / eval-results /Undi95 /Mistral-11B-TestBench11 /results_2023-10-28T01-59-23.177639.json
Wwwduojin's picture
Upload 1529 files
a312f2f
{
"config_general": {
"model_name": "Undi95/Mistral-11B-TestBench11",
"model_sha": "2c95dc683b5256749b0cad984f59d0af041822a6",
"model_size": "20.74 GB",
"model_dtype": "torch.bfloat16",
"lighteval_sha": "0f318ecf002208468154899217b3ba7c6ae09374",
"num_few_shot_default": 0,
"num_fewshot_seeds": 1,
"override_batch_size": 1,
"max_samples": null,
"job_id": ""
},
"results": {
"harness|drop|3": {
"em": 0.02904781879194631,
"em_stderr": 0.0017198688690203193,
"f1": 0.09573615771812093,
"f1_stderr": 0.0021674728464020697
},
"harness|gsm8k|5": {
"acc": 0.14935557240333586,
"acc_stderr": 0.00981809072372729
},
"harness|winogrande|5": {
"acc": 0.7774269928966061,
"acc_stderr": 0.011690933809712667
},
"all": {
"em": 0.02904781879194631,
"em_stderr": 0.0017198688690203193,
"f1": 0.09573615771812093,
"f1_stderr": 0.0021674728464020697,
"acc": 0.463391282649971,
"acc_stderr": 0.010754512266719978
}
},
"versions": {
"harness|drop|3": 1,
"harness|gsm8k|5": 0,
"harness|winogrande|5": 0,
"all": 0
},
"config_tasks": {
"harness|drop": "LM Harness task",
"harness|gsm8k": "LM Harness task",
"harness|winogrande": "LM Harness task"
},
"summary_tasks": {
"harness|drop|3": {
"hashes": {
"hash_examples": "1d27416e8324e9a3",
"hash_full_prompts": "a5513ff9a741b385",
"hash_input_tokens": "a4fb946366902edf",
"hash_cont_tokens": "d7ba2a20d0ecc3d2"
},
"truncated": 0,
"non-truncated": 9536,
"padded": 0,
"non-padded": 9536,
"effective_few_shots": 3.0,
"num_truncated_few_shots": 0
},
"harness|gsm8k|5": {
"hashes": {
"hash_examples": "4c0843a5d99bcfdc",
"hash_full_prompts": "41d55e83abc0e02d",
"hash_input_tokens": "6af0ae8cfe684f50",
"hash_cont_tokens": "359572aca9338c1f"
},
"truncated": 0,
"non-truncated": 1319,
"padded": 0,
"non-padded": 1319,
"effective_few_shots": 5.0,
"num_truncated_few_shots": 0
},
"harness|winogrande|5": {
"hashes": {
"hash_examples": "aada0a176fd81218",
"hash_full_prompts": "c8655cbd12de8409",
"hash_input_tokens": "6bf335f26fed6442",
"hash_cont_tokens": "618558fb93c0f288"
},
"truncated": 0,
"non-truncated": 2534,
"padded": 2397,
"non-padded": 137,
"effective_few_shots": 5.0,
"num_truncated_few_shots": 0
}
},
"summary_general": {
"hashes": {
"hash_examples": "9b4d8993161e637d",
"hash_full_prompts": "08215e527b7e60a5",
"hash_input_tokens": "7c6dc027ca91c1a8",
"hash_cont_tokens": "a046a5e185f8240b"
},
"total_evaluation_time_secondes": "14685.71157336235",
"truncated": 0,
"non-truncated": 13389,
"padded": 2397,
"non-padded": 10992,
"num_truncated_few_shots": 0
}
}