devils-agent / reports /qwen-baseline.json
devildasdf's picture
Upload experimental BAIM code, research checkpoints and measured evaluations
795f737 verified
Raw
History Blame Contribute Delete
2.64 kB
{
"model": "Qwen/Qwen2.5-0.5B-Instruct",
"revision": "7ae557604adf67be50417f59c2c2f167def9a775",
"parameters": 494032768,
"dtype": "float32",
"threads": 2,
"load_seconds": 3.181326099904254,
"samples": 8,
"valid_rate": 0.0,
"joint_accuracy": 0.0,
"median_wall_ms": 6813.391300034709,
"results": [
{
"index": 0,
"valid": false,
"joint_action_target_correct": false,
"target_in_candidates": true,
"input_tokens": 859,
"output_tokens": 7,
"wall_ms": 8279.813800007105,
"observed_rss_bytes": 2485579776
},
{
"index": 1,
"valid": false,
"joint_action_target_correct": false,
"target_in_candidates": false,
"input_tokens": 744,
"output_tokens": 5,
"wall_ms": 6579.704500036314,
"observed_rss_bytes": 2485760000
},
{
"index": 2,
"valid": false,
"joint_action_target_correct": false,
"target_in_candidates": true,
"input_tokens": 760,
"output_tokens": 7,
"wall_ms": 7047.078100033104,
"observed_rss_bytes": 2483740672
},
{
"index": 3,
"valid": false,
"joint_action_target_correct": false,
"target_in_candidates": true,
"input_tokens": 659,
"output_tokens": 7,
"wall_ms": 6190.024700015783,
"observed_rss_bytes": 2495209472
},
{
"index": 4,
"valid": false,
"joint_action_target_correct": false,
"target_in_candidates": true,
"input_tokens": 656,
"output_tokens": 9,
"wall_ms": 6563.73789999634,
"observed_rss_bytes": 2494586880
},
{
"index": 5,
"valid": false,
"joint_action_target_correct": false,
"target_in_candidates": false,
"input_tokens": 887,
"output_tokens": 7,
"wall_ms": 7589.049200061709,
"observed_rss_bytes": 2498666496
},
{
"index": 6,
"valid": false,
"joint_action_target_correct": false,
"target_in_candidates": true,
"input_tokens": 677,
"output_tokens": 7,
"wall_ms": 6104.094199952669,
"observed_rss_bytes": 2494570496
},
{
"index": 7,
"valid": false,
"joint_action_target_correct": false,
"target_in_candidates": false,
"input_tokens": 865,
"output_tokens": 7,
"wall_ms": 7474.766599945724,
"observed_rss_bytes": 2498916352
}
],
"scope": "Small ordered training-shard smoke test, previous human actions supplied. No execution; values not scored. Not target VPS."
}