Experiments
Track runs, compare metrics, and reproduce results
Run comparison · med-7b LR sweep
Best runs
run-a17
lr=5e-5, r=16
0.284
bestrun-b04
lr=3e-5, r=16
0.289
run-c11
lr=1e-4, r=32
0.302
run-d22
lr=5e-5, r=8
0.294
All experiments
| ID | Name | Metric | Best | Runs | Status |
|---|---|---|---|---|---|
| exp_2401 | med-7b LR sweep | eval/loss | 0.284 | 24 | Completed |
| exp_2402 | radiology augmentations | auc | 0.941 | 12 | Running |
| exp_2403 | voice-triage RLHF reward | reward | 4.72 | 8 | Running |
| exp_2404 | claims-nlp domain sweep | f1 | 0.887 | 32 | Completed |
| exp_2405 | kyc-classifier distillation | accuracy | 0.958 | 6 | Completed |