
Loading weights:   0%|          | 0/311 [00:00<?, ?it/s]
Loading weights: 100%|██████████| 311/311 [00:00<00:00, 4599.37it/s]
full 3e-06 step 0 loss 1.34 exact 5/32 train_s 0
full 3e-06 step 4 loss 0.5246 exact 20/32 train_s 5.89
full 3e-06 step 8 loss 0.6869 exact 21/32 train_s 11.68
full 3e-06 step 16 loss 0.3737 exact 22/32 train_s 23.25
Saved artifacts/qwen3-06b/2026-09-08-efficiency-screen/full-lr3e-6
