| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.1, | |
| "eval_steps": 2000, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 13.625, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 1.2531, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 19.75, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 1.1239, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 13.6875, | |
| "learning_rate": 9e-07, | |
| "loss": 1.0691, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 15.75, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.0435, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 1.5e-06, | |
| "loss": 1.0728, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 1.8e-06, | |
| "loss": 1.1723, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 9.5, | |
| "learning_rate": 2.1000000000000002e-06, | |
| "loss": 0.9071, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.861, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.0, | |
| "learning_rate": 2.7e-06, | |
| "loss": 0.9021, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 3e-06, | |
| "loss": 0.6285, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 3.3e-06, | |
| "loss": 0.5687, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 3.6e-06, | |
| "loss": 0.5357, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 3.9e-06, | |
| "loss": 0.3229, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 0.3235, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.1306, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.043, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.09765625, | |
| "learning_rate": 5.1e-06, | |
| "loss": 0.0117, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 5.4e-06, | |
| "loss": 0.0033, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 5.7000000000000005e-06, | |
| "loss": 0.0088, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.023681640625, | |
| "learning_rate": 6e-06, | |
| "loss": 0.0046, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.0400390625, | |
| "learning_rate": 6.3e-06, | |
| "loss": 0.0049, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 6.6e-06, | |
| "loss": 0.0015, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 6.900000000000001e-06, | |
| "loss": 0.0082, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.0400390625, | |
| "learning_rate": 7.2e-06, | |
| "loss": 0.0005, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.0134, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.023681640625, | |
| "learning_rate": 7.8e-06, | |
| "loss": 0.0023, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.032958984375, | |
| "learning_rate": 8.1e-06, | |
| "loss": 0.0005, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.0218, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.02783203125, | |
| "learning_rate": 8.7e-06, | |
| "loss": 0.0005, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.011962890625, | |
| "learning_rate": 9e-06, | |
| "loss": 0.0006, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 9.3e-06, | |
| "loss": 0.0013, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.00872802734375, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.0003, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.0166015625, | |
| "learning_rate": 9.9e-06, | |
| "loss": 0.0006, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.00811767578125, | |
| "learning_rate": 1.02e-05, | |
| "loss": 0.0122, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.00921630859375, | |
| "learning_rate": 1.05e-05, | |
| "loss": 0.0005, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.0125732421875, | |
| "learning_rate": 1.08e-05, | |
| "loss": 0.0003, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.019287109375, | |
| "learning_rate": 1.11e-05, | |
| "loss": 0.0005, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.007049560546875, | |
| "learning_rate": 1.1400000000000001e-05, | |
| "loss": 0.0004, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.00799560546875, | |
| "learning_rate": 1.1700000000000001e-05, | |
| "loss": 0.0091, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.00750732421875, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.0003, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 1.2299999999999999e-05, | |
| "loss": 0.0042, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.06103515625, | |
| "learning_rate": 1.26e-05, | |
| "loss": 0.0005, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.006439208984375, | |
| "learning_rate": 1.29e-05, | |
| "loss": 0.0002, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.009033203125, | |
| "learning_rate": 1.32e-05, | |
| "loss": 0.0002, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.0069580078125, | |
| "learning_rate": 1.3500000000000001e-05, | |
| "loss": 0.0003, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.0211181640625, | |
| "learning_rate": 1.3800000000000002e-05, | |
| "loss": 0.0001, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.0067138671875, | |
| "learning_rate": 1.4099999999999999e-05, | |
| "loss": 0.0001, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.005889892578125, | |
| "learning_rate": 1.44e-05, | |
| "loss": 0.0002, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.00799560546875, | |
| "learning_rate": 1.47e-05, | |
| "loss": 0.0002, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.0004, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.00787353515625, | |
| "learning_rate": 1.4984210526315789e-05, | |
| "loss": 0.0002, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.019287109375, | |
| "learning_rate": 1.496842105263158e-05, | |
| "loss": 0.0002, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.0047607421875, | |
| "learning_rate": 1.4952631578947368e-05, | |
| "loss": 0.0004, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.007293701171875, | |
| "learning_rate": 1.4936842105263158e-05, | |
| "loss": 0.0003, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.0050048828125, | |
| "learning_rate": 1.4921052631578947e-05, | |
| "loss": 0.0002, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.006072998046875, | |
| "learning_rate": 1.4905263157894737e-05, | |
| "loss": 0.001, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.006805419921875, | |
| "learning_rate": 1.4889473684210526e-05, | |
| "loss": 0.0002, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.0205078125, | |
| "learning_rate": 1.4873684210526315e-05, | |
| "loss": 0.0002, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.00433349609375, | |
| "learning_rate": 1.4857894736842107e-05, | |
| "loss": 0.0002, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.015380859375, | |
| "learning_rate": 1.4842105263157895e-05, | |
| "loss": 0.0001, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.007568359375, | |
| "learning_rate": 1.4826315789473686e-05, | |
| "loss": 0.0002, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.03466796875, | |
| "learning_rate": 1.4810526315789474e-05, | |
| "loss": 0.0002, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.00872802734375, | |
| "learning_rate": 1.4794736842105265e-05, | |
| "loss": 0.0002, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.00799560546875, | |
| "learning_rate": 1.4778947368421053e-05, | |
| "loss": 0.0002, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.0179443359375, | |
| "learning_rate": 1.4763157894736842e-05, | |
| "loss": 0.0001, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.00897216796875, | |
| "learning_rate": 1.4747368421052632e-05, | |
| "loss": 0.0002, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.00396728515625, | |
| "learning_rate": 1.4731578947368421e-05, | |
| "loss": 0.0001, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.00860595703125, | |
| "learning_rate": 1.4715789473684211e-05, | |
| "loss": 0.0001, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.006072998046875, | |
| "learning_rate": 1.47e-05, | |
| "loss": 0.0002, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0042724609375, | |
| "learning_rate": 1.468421052631579e-05, | |
| "loss": 0.0002, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0125732421875, | |
| "learning_rate": 1.4668421052631579e-05, | |
| "loss": 0.0001, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0086669921875, | |
| "learning_rate": 1.4652631578947367e-05, | |
| "loss": 0.0001, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.01300048828125, | |
| "learning_rate": 1.4636842105263158e-05, | |
| "loss": 0.0002, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0128173828125, | |
| "learning_rate": 1.4621052631578946e-05, | |
| "loss": 0.0002, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.10888671875, | |
| "learning_rate": 1.4605263157894737e-05, | |
| "loss": 0.0003, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0169677734375, | |
| "learning_rate": 1.4589473684210527e-05, | |
| "loss": 0.0002, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0038604736328125, | |
| "learning_rate": 1.4573684210526317e-05, | |
| "loss": 0.0001, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0032806396484375, | |
| "learning_rate": 1.4557894736842106e-05, | |
| "loss": 0.0002, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.00433349609375, | |
| "learning_rate": 1.4542105263157895e-05, | |
| "loss": 0.0001, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.004058837890625, | |
| "learning_rate": 1.4526315789473685e-05, | |
| "loss": 0.0002, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.00299072265625, | |
| "learning_rate": 1.4510526315789474e-05, | |
| "loss": 0.0001, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.01025390625, | |
| "learning_rate": 1.4494736842105264e-05, | |
| "loss": 0.0001, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.015625, | |
| "learning_rate": 1.4478947368421053e-05, | |
| "loss": 0.0002, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0040283203125, | |
| "learning_rate": 1.4463157894736843e-05, | |
| "loss": 0.0002, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0120849609375, | |
| "learning_rate": 1.4447368421052632e-05, | |
| "loss": 0.0001, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.005096435546875, | |
| "learning_rate": 1.443157894736842e-05, | |
| "loss": 0.0001, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.00933837890625, | |
| "learning_rate": 1.441578947368421e-05, | |
| "loss": 0.0001, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.00445556640625, | |
| "learning_rate": 1.44e-05, | |
| "loss": 0.0001, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0029144287109375, | |
| "learning_rate": 1.438421052631579e-05, | |
| "loss": 0.0001, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.016357421875, | |
| "learning_rate": 1.4368421052631578e-05, | |
| "loss": 0.0001, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.003570556640625, | |
| "learning_rate": 1.4352631578947369e-05, | |
| "loss": 0.0049, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.0068359375, | |
| "learning_rate": 1.4336842105263159e-05, | |
| "loss": 0.0002, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.00335693359375, | |
| "learning_rate": 1.4321052631578948e-05, | |
| "loss": 0.0002, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.003082275390625, | |
| "learning_rate": 1.4305263157894738e-05, | |
| "loss": 0.0001, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.025634765625, | |
| "learning_rate": 1.4289473684210527e-05, | |
| "loss": 0.0001, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.00396728515625, | |
| "learning_rate": 1.4273684210526317e-05, | |
| "loss": 0.0001, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.00885009765625, | |
| "learning_rate": 1.4257894736842106e-05, | |
| "loss": 0.0001, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.005950927734375, | |
| "learning_rate": 1.4242105263157896e-05, | |
| "loss": 0.0001, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.0191650390625, | |
| "learning_rate": 1.4226315789473685e-05, | |
| "loss": 0.0002, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.0059814453125, | |
| "learning_rate": 1.4210526315789473e-05, | |
| "loss": 0.0002, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 1.4194736842105264e-05, | |
| "loss": 0.0008, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.0048828125, | |
| "learning_rate": 1.4178947368421052e-05, | |
| "loss": 0.0001, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.004241943359375, | |
| "learning_rate": 1.4163157894736843e-05, | |
| "loss": 0.0001, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.004913330078125, | |
| "learning_rate": 1.4147368421052631e-05, | |
| "loss": 0.0001, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.00640869140625, | |
| "learning_rate": 1.4131578947368422e-05, | |
| "loss": 0.0002, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.01202392578125, | |
| "learning_rate": 1.411578947368421e-05, | |
| "loss": 0.0001, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.004486083984375, | |
| "learning_rate": 1.4099999999999999e-05, | |
| "loss": 0.0001, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.003936767578125, | |
| "learning_rate": 1.408421052631579e-05, | |
| "loss": 0.0001, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.0098876953125, | |
| "learning_rate": 1.406842105263158e-05, | |
| "loss": 0.0003, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.005279541015625, | |
| "learning_rate": 1.405263157894737e-05, | |
| "loss": 0.0001, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0068359375, | |
| "learning_rate": 1.4036842105263158e-05, | |
| "loss": 0.0001, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.00537109375, | |
| "learning_rate": 1.4021052631578949e-05, | |
| "loss": 0.0001, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.007568359375, | |
| "learning_rate": 1.4005263157894737e-05, | |
| "loss": 0.0001, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.004241943359375, | |
| "learning_rate": 1.3989473684210526e-05, | |
| "loss": 0.0001, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0111083984375, | |
| "learning_rate": 1.3973684210526316e-05, | |
| "loss": 0.0001, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.01806640625, | |
| "learning_rate": 1.3957894736842105e-05, | |
| "loss": 0.0001, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.01416015625, | |
| "learning_rate": 1.3942105263157895e-05, | |
| "loss": 0.0009, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0091552734375, | |
| "learning_rate": 1.3926315789473684e-05, | |
| "loss": 0.0001, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.002899169921875, | |
| "learning_rate": 1.3910526315789474e-05, | |
| "loss": 0.0001, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.005523681640625, | |
| "learning_rate": 1.3894736842105263e-05, | |
| "loss": 0.0001, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.01214599609375, | |
| "learning_rate": 1.3878947368421052e-05, | |
| "loss": 0.0001, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0037384033203125, | |
| "learning_rate": 1.3863157894736842e-05, | |
| "loss": 0.0001, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.008056640625, | |
| "learning_rate": 1.384736842105263e-05, | |
| "loss": 0.0001, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0081787109375, | |
| "learning_rate": 1.3831578947368421e-05, | |
| "loss": 0.0001, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0035400390625, | |
| "learning_rate": 1.3815789473684211e-05, | |
| "loss": 0.0002, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.004486083984375, | |
| "learning_rate": 1.3800000000000002e-05, | |
| "loss": 0.0001, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.004150390625, | |
| "learning_rate": 1.378421052631579e-05, | |
| "loss": 0.0001, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0023651123046875, | |
| "learning_rate": 1.3768421052631579e-05, | |
| "loss": 0.0001, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0029449462890625, | |
| "learning_rate": 1.375263157894737e-05, | |
| "loss": 0.0001, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0032196044921875, | |
| "learning_rate": 1.3736842105263158e-05, | |
| "loss": 0.0001, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0027923583984375, | |
| "learning_rate": 1.3721052631578948e-05, | |
| "loss": 0.0001, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.003173828125, | |
| "learning_rate": 1.3705263157894737e-05, | |
| "loss": 0.0001, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0027923583984375, | |
| "learning_rate": 1.3689473684210527e-05, | |
| "loss": 0.0001, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.004150390625, | |
| "learning_rate": 1.3673684210526316e-05, | |
| "loss": 0.0001, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.00604248046875, | |
| "learning_rate": 1.3657894736842106e-05, | |
| "loss": 0.0001, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.005645751953125, | |
| "learning_rate": 1.3642105263157895e-05, | |
| "loss": 0.0001, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.00787353515625, | |
| "learning_rate": 1.3626315789473684e-05, | |
| "loss": 0.0001, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.003997802734375, | |
| "learning_rate": 1.3610526315789474e-05, | |
| "loss": 0.0001, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0024261474609375, | |
| "learning_rate": 1.3594736842105263e-05, | |
| "loss": 0.0002, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0022430419921875, | |
| "learning_rate": 1.3578947368421053e-05, | |
| "loss": 0.0001, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.00518798828125, | |
| "learning_rate": 1.3563157894736842e-05, | |
| "loss": 0.0001, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0035858154296875, | |
| "learning_rate": 1.3547368421052634e-05, | |
| "loss": 0.0005, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0027008056640625, | |
| "learning_rate": 1.3531578947368422e-05, | |
| "loss": 0.0001, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0023193359375, | |
| "learning_rate": 1.3515789473684211e-05, | |
| "loss": 0.0002, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0025177001953125, | |
| "learning_rate": 1.3500000000000001e-05, | |
| "loss": 0.0003, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0040283203125, | |
| "learning_rate": 1.348421052631579e-05, | |
| "loss": 0.0001, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.005645751953125, | |
| "learning_rate": 1.346842105263158e-05, | |
| "loss": 0.0001, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0859375, | |
| "learning_rate": 1.3452631578947369e-05, | |
| "loss": 0.0004, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.005615234375, | |
| "learning_rate": 1.343684210526316e-05, | |
| "loss": 0.0001, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.005767822265625, | |
| "learning_rate": 1.3421052631578948e-05, | |
| "loss": 0.0001, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0052490234375, | |
| "learning_rate": 1.3405263157894736e-05, | |
| "loss": 0.0001, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.005615234375, | |
| "learning_rate": 1.3389473684210527e-05, | |
| "loss": 0.0001, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.003570556640625, | |
| "learning_rate": 1.3373684210526315e-05, | |
| "loss": 0.0021, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.002471923828125, | |
| "learning_rate": 1.3357894736842106e-05, | |
| "loss": 0.0001, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.004241943359375, | |
| "learning_rate": 1.3342105263157894e-05, | |
| "loss": 0.0011, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.00958251953125, | |
| "learning_rate": 1.3326315789473685e-05, | |
| "loss": 0.0001, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.004241943359375, | |
| "learning_rate": 1.3310526315789473e-05, | |
| "loss": 0.0003, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0030059814453125, | |
| "learning_rate": 1.3294736842105262e-05, | |
| "loss": 0.0001, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.00653076171875, | |
| "learning_rate": 1.3278947368421054e-05, | |
| "loss": 0.0001, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0086669921875, | |
| "learning_rate": 1.3263157894736843e-05, | |
| "loss": 0.0001, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0034332275390625, | |
| "learning_rate": 1.3247368421052633e-05, | |
| "loss": 0.0001, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0027008056640625, | |
| "learning_rate": 1.3231578947368422e-05, | |
| "loss": 0.0001, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0023956298828125, | |
| "learning_rate": 1.3215789473684212e-05, | |
| "loss": 0.0001, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.005889892578125, | |
| "learning_rate": 1.32e-05, | |
| "loss": 0.0001, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.00191497802734375, | |
| "learning_rate": 1.318421052631579e-05, | |
| "loss": 0.0001, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.00335693359375, | |
| "learning_rate": 1.316842105263158e-05, | |
| "loss": 0.0001, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0047607421875, | |
| "learning_rate": 1.3152631578947368e-05, | |
| "loss": 0.0001, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.00193023681640625, | |
| "learning_rate": 1.3136842105263159e-05, | |
| "loss": 0.0001, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.005401611328125, | |
| "learning_rate": 1.3121052631578947e-05, | |
| "loss": 0.0001, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.008056640625, | |
| "learning_rate": 1.3105263157894738e-05, | |
| "loss": 0.0001, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0036773681640625, | |
| "learning_rate": 1.3089473684210526e-05, | |
| "loss": 0.0001, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.00457763671875, | |
| "learning_rate": 1.3073684210526315e-05, | |
| "loss": 0.0001, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.005401611328125, | |
| "learning_rate": 1.3057894736842105e-05, | |
| "loss": 0.0001, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.00185394287109375, | |
| "learning_rate": 1.3042105263157894e-05, | |
| "loss": 0.0001, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0037078857421875, | |
| "learning_rate": 1.3026315789473684e-05, | |
| "loss": 0.0001, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.002410888671875, | |
| "learning_rate": 1.3010526315789475e-05, | |
| "loss": 0.0001, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0086669921875, | |
| "learning_rate": 1.2994736842105265e-05, | |
| "loss": 0.0001, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0024871826171875, | |
| "learning_rate": 1.2978947368421054e-05, | |
| "loss": 0.0001, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0028839111328125, | |
| "learning_rate": 1.2963157894736842e-05, | |
| "loss": 0.0001, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.010009765625, | |
| "learning_rate": 1.2947368421052633e-05, | |
| "loss": 0.0001, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.00909423828125, | |
| "learning_rate": 1.2931578947368421e-05, | |
| "loss": 0.0001, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.003692626953125, | |
| "learning_rate": 1.2915789473684212e-05, | |
| "loss": 0.0002, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0140380859375, | |
| "learning_rate": 1.29e-05, | |
| "loss": 0.0001, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.006866455078125, | |
| "learning_rate": 1.288421052631579e-05, | |
| "loss": 0.0001, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0024566650390625, | |
| "learning_rate": 1.2868421052631579e-05, | |
| "loss": 0.0001, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0027313232421875, | |
| "learning_rate": 1.2852631578947368e-05, | |
| "loss": 0.0001, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.002105712890625, | |
| "learning_rate": 1.2836842105263158e-05, | |
| "loss": 0.0001, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.004241943359375, | |
| "learning_rate": 1.2821052631578947e-05, | |
| "loss": 0.0001, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.005706787109375, | |
| "learning_rate": 1.2805263157894737e-05, | |
| "loss": 0.0001, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.0079345703125, | |
| "learning_rate": 1.2789473684210526e-05, | |
| "loss": 0.0002, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.0037994384765625, | |
| "learning_rate": 1.2773684210526316e-05, | |
| "loss": 0.0001, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.0052490234375, | |
| "learning_rate": 1.2757894736842106e-05, | |
| "loss": 0.0001, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.0067138671875, | |
| "learning_rate": 1.2742105263157895e-05, | |
| "loss": 0.0001, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.006622314453125, | |
| "learning_rate": 1.2726315789473685e-05, | |
| "loss": 0.0001, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.00933837890625, | |
| "learning_rate": 1.2710526315789474e-05, | |
| "loss": 0.0001, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.00421142578125, | |
| "learning_rate": 1.2694736842105264e-05, | |
| "loss": 0.0001, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.00194549560546875, | |
| "learning_rate": 1.2678947368421053e-05, | |
| "loss": 0.0001, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.0035552978515625, | |
| "learning_rate": 1.2663157894736843e-05, | |
| "loss": 0.0001, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.0024261474609375, | |
| "learning_rate": 1.2647368421052632e-05, | |
| "loss": 0.0001, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.0033111572265625, | |
| "learning_rate": 1.263157894736842e-05, | |
| "loss": 0.0001, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_loss": 9.133540152106434e-05, | |
| "eval_runtime": 30.1215, | |
| "eval_samples_per_second": 33.199, | |
| "eval_steps_per_second": 33.199, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 2000, | |
| "total_flos": 1.613922041856e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |