blazeofchi's picture
Upload folder using huggingface_hub
9fbb8cc verified
Raw
History Blame Contribute Delete
3.06 kB
{
"device": "cuda",
"history": [
{
"epoch": 0,
"loss": 3.11750134399959
},
{
"epoch": 1,
"loss": 2.809019318648747
},
{
"epoch": 2,
"loss": 2.8290022100721086
},
{
"epoch": 3,
"loss": 2.817454559462411
},
{
"epoch": 4,
"loss": 2.805349017892565
},
{
"epoch": 5,
"loss": 2.7349614586148943
},
{
"epoch": 6,
"loss": 2.7246988671166554
},
{
"epoch": 7,
"loss": 2.7136061021259854
},
{
"epoch": 8,
"loss": 2.687513675008501
},
{
"epoch": 9,
"loss": 2.7999841826302663
},
{
"epoch": 10,
"loss": 2.7054038047790527
},
{
"epoch": 11,
"loss": 2.659690465245928
},
{
"epoch": 12,
"loss": 2.848549485206604
},
{
"epoch": 13,
"loss": 2.659048242228372
},
{
"epoch": 14,
"loss": 2.6345216717038835
},
{
"epoch": 15,
"loss": 2.6133223516600474
},
{
"epoch": 16,
"loss": 2.5888123171670094
},
{
"epoch": 17,
"loss": 2.580440470150539
},
{
"epoch": 18,
"loss": 2.6477542349270413
},
{
"epoch": 19,
"loss": 2.6182081358773366
},
{
"epoch": 20,
"loss": 2.5893647926194325
},
{
"epoch": 21,
"loss": 2.591572182519095
},
{
"epoch": 22,
"loss": 2.5323841656957353
},
{
"epoch": 23,
"loss": 2.544121095112392
},
{
"epoch": 24,
"loss": 2.5288875699043274
},
{
"epoch": 25,
"loss": 2.5025893364633833
},
{
"epoch": 26,
"loss": 2.603824121611459
},
{
"epoch": 27,
"loss": 2.543979917253767
},
{
"epoch": 28,
"loss": 2.4952145644596646
},
{
"epoch": 29,
"loss": 2.539324470928737
},
{
"epoch": 30,
"loss": 2.5138033287865773
},
{
"epoch": 31,
"loss": 2.4722729665892467
},
{
"epoch": 32,
"loss": 2.408249352659498
},
{
"epoch": 33,
"loss": 2.4780009303774153
},
{
"epoch": 34,
"loss": 2.4572549717766896
},
{
"epoch": 35,
"loss": 2.393418550491333
},
{
"epoch": 36,
"loss": 2.4449552042143687
},
{
"epoch": 37,
"loss": 2.616686258997236
},
{
"epoch": 38,
"loss": 2.4797831092561995
},
{
"epoch": 39,
"loss": 2.474162220954895
}
],
"output_dir": "research/models/20260628-qwen3-0p6b-logits-orchestrator-full-gpu-l40s",
"record_count": 53,
"schema_version": "mempool.qwen_logits_orchestrator_train_report.v1",
"worker_ids": [
"ollama-cloud-deepseek-v4-pro",
"ollama-cloud-glm-5.2",
"ollama-cloud-kimi-k2.7-code",
"ollama-cloud-qwen3-coder-480b"
],
"workflow_labels": [
"direct",
"verify_then_fallback"
],
"base_model": "Qwen/Qwen3-0.6B"
}