MeedoSam's picture
Uploaded checkpoint-8000
a942271 verified
raw
history blame contribute delete
No virus
127 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.4,
"eval_steps": 2000,
"global_step": 8000,
"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
},
{
"epoch": 0.1,
"grad_norm": 0.00262451171875,
"learning_rate": 1.2615789473684211e-05,
"loss": 0.0004,
"step": 2010
},
{
"epoch": 0.1,
"grad_norm": 0.00457763671875,
"learning_rate": 1.26e-05,
"loss": 0.0001,
"step": 2020
},
{
"epoch": 0.1,
"grad_norm": 0.006500244140625,
"learning_rate": 1.258421052631579e-05,
"loss": 0.0001,
"step": 2030
},
{
"epoch": 0.1,
"grad_norm": 0.00640869140625,
"learning_rate": 1.2568421052631579e-05,
"loss": 0.0001,
"step": 2040
},
{
"epoch": 0.1,
"grad_norm": 0.004852294921875,
"learning_rate": 1.2552631578947369e-05,
"loss": 0.0001,
"step": 2050
},
{
"epoch": 0.1,
"grad_norm": 0.00177764892578125,
"learning_rate": 1.2536842105263158e-05,
"loss": 0.0001,
"step": 2060
},
{
"epoch": 0.1,
"grad_norm": 0.011474609375,
"learning_rate": 1.2521052631578946e-05,
"loss": 0.0001,
"step": 2070
},
{
"epoch": 0.1,
"grad_norm": 0.0042724609375,
"learning_rate": 1.2505263157894737e-05,
"loss": 0.0001,
"step": 2080
},
{
"epoch": 0.1,
"grad_norm": 0.003753662109375,
"learning_rate": 1.2489473684210527e-05,
"loss": 0.0001,
"step": 2090
},
{
"epoch": 0.1,
"grad_norm": 0.00872802734375,
"learning_rate": 1.2473684210526317e-05,
"loss": 0.0007,
"step": 2100
},
{
"epoch": 0.11,
"grad_norm": 0.005096435546875,
"learning_rate": 1.2457894736842106e-05,
"loss": 0.0001,
"step": 2110
},
{
"epoch": 0.11,
"grad_norm": 0.002655029296875,
"learning_rate": 1.2442105263157896e-05,
"loss": 0.0001,
"step": 2120
},
{
"epoch": 0.11,
"grad_norm": 0.01092529296875,
"learning_rate": 1.2426315789473685e-05,
"loss": 0.0001,
"step": 2130
},
{
"epoch": 0.11,
"grad_norm": 0.0037994384765625,
"learning_rate": 1.2410526315789474e-05,
"loss": 0.0003,
"step": 2140
},
{
"epoch": 0.11,
"grad_norm": 0.0030670166015625,
"learning_rate": 1.2394736842105264e-05,
"loss": 0.0001,
"step": 2150
},
{
"epoch": 0.11,
"grad_norm": 0.00640869140625,
"learning_rate": 1.2378947368421053e-05,
"loss": 0.0001,
"step": 2160
},
{
"epoch": 0.11,
"grad_norm": 0.0021820068359375,
"learning_rate": 1.2363157894736843e-05,
"loss": 0.0002,
"step": 2170
},
{
"epoch": 0.11,
"grad_norm": 0.0047607421875,
"learning_rate": 1.2347368421052631e-05,
"loss": 0.0001,
"step": 2180
},
{
"epoch": 0.11,
"grad_norm": 0.00494384765625,
"learning_rate": 1.2331578947368422e-05,
"loss": 0.0001,
"step": 2190
},
{
"epoch": 0.11,
"grad_norm": 0.002593994140625,
"learning_rate": 1.231578947368421e-05,
"loss": 0.0001,
"step": 2200
},
{
"epoch": 0.11,
"grad_norm": 0.00555419921875,
"learning_rate": 1.2299999999999999e-05,
"loss": 0.0001,
"step": 2210
},
{
"epoch": 0.11,
"grad_norm": 0.0023345947265625,
"learning_rate": 1.228421052631579e-05,
"loss": 0.0001,
"step": 2220
},
{
"epoch": 0.11,
"grad_norm": 0.007110595703125,
"learning_rate": 1.2268421052631578e-05,
"loss": 0.0001,
"step": 2230
},
{
"epoch": 0.11,
"grad_norm": 0.0032196044921875,
"learning_rate": 1.2252631578947368e-05,
"loss": 0.0001,
"step": 2240
},
{
"epoch": 0.11,
"grad_norm": 0.004486083984375,
"learning_rate": 1.2236842105263159e-05,
"loss": 0.0001,
"step": 2250
},
{
"epoch": 0.11,
"grad_norm": 0.0026397705078125,
"learning_rate": 1.2221052631578949e-05,
"loss": 0.0001,
"step": 2260
},
{
"epoch": 0.11,
"grad_norm": 0.003173828125,
"learning_rate": 1.2205263157894738e-05,
"loss": 0.0001,
"step": 2270
},
{
"epoch": 0.11,
"grad_norm": 0.002105712890625,
"learning_rate": 1.2189473684210526e-05,
"loss": 0.0001,
"step": 2280
},
{
"epoch": 0.11,
"grad_norm": 0.0042724609375,
"learning_rate": 1.2173684210526317e-05,
"loss": 0.0001,
"step": 2290
},
{
"epoch": 0.12,
"grad_norm": 0.003692626953125,
"learning_rate": 1.2157894736842105e-05,
"loss": 0.0001,
"step": 2300
},
{
"epoch": 0.12,
"grad_norm": 0.0020599365234375,
"learning_rate": 1.2142105263157896e-05,
"loss": 0.0,
"step": 2310
},
{
"epoch": 0.12,
"grad_norm": 0.0025177001953125,
"learning_rate": 1.2126315789473684e-05,
"loss": 0.0001,
"step": 2320
},
{
"epoch": 0.12,
"grad_norm": 0.0026092529296875,
"learning_rate": 1.2110526315789475e-05,
"loss": 0.0001,
"step": 2330
},
{
"epoch": 0.12,
"grad_norm": 0.0026702880859375,
"learning_rate": 1.2094736842105263e-05,
"loss": 0.0001,
"step": 2340
},
{
"epoch": 0.12,
"grad_norm": 0.00180816650390625,
"learning_rate": 1.2078947368421052e-05,
"loss": 0.0001,
"step": 2350
},
{
"epoch": 0.12,
"grad_norm": 0.0025787353515625,
"learning_rate": 1.2063157894736842e-05,
"loss": 0.0001,
"step": 2360
},
{
"epoch": 0.12,
"grad_norm": 0.006988525390625,
"learning_rate": 1.2047368421052631e-05,
"loss": 0.0001,
"step": 2370
},
{
"epoch": 0.12,
"grad_norm": 0.0035552978515625,
"learning_rate": 1.2031578947368421e-05,
"loss": 0.0001,
"step": 2380
},
{
"epoch": 0.12,
"grad_norm": 0.005645751953125,
"learning_rate": 1.201578947368421e-05,
"loss": 0.0001,
"step": 2390
},
{
"epoch": 0.12,
"grad_norm": 0.00677490234375,
"learning_rate": 1.2e-05,
"loss": 0.0001,
"step": 2400
},
{
"epoch": 0.12,
"grad_norm": 0.0086669921875,
"learning_rate": 1.1984210526315789e-05,
"loss": 0.0001,
"step": 2410
},
{
"epoch": 0.12,
"grad_norm": 0.0023345947265625,
"learning_rate": 1.196842105263158e-05,
"loss": 0.0001,
"step": 2420
},
{
"epoch": 0.12,
"grad_norm": 0.0022735595703125,
"learning_rate": 1.195263157894737e-05,
"loss": 0.0001,
"step": 2430
},
{
"epoch": 0.12,
"grad_norm": 0.004852294921875,
"learning_rate": 1.1936842105263158e-05,
"loss": 0.0001,
"step": 2440
},
{
"epoch": 0.12,
"grad_norm": 0.00182342529296875,
"learning_rate": 1.1921052631578949e-05,
"loss": 0.0001,
"step": 2450
},
{
"epoch": 0.12,
"grad_norm": 0.00408935546875,
"learning_rate": 1.1905263157894737e-05,
"loss": 0.0001,
"step": 2460
},
{
"epoch": 0.12,
"grad_norm": 0.01123046875,
"learning_rate": 1.1889473684210528e-05,
"loss": 0.0001,
"step": 2470
},
{
"epoch": 0.12,
"grad_norm": 0.0023040771484375,
"learning_rate": 1.1873684210526316e-05,
"loss": 0.0001,
"step": 2480
},
{
"epoch": 0.12,
"grad_norm": 0.005126953125,
"learning_rate": 1.1857894736842105e-05,
"loss": 0.0001,
"step": 2490
},
{
"epoch": 0.12,
"grad_norm": 0.0142822265625,
"learning_rate": 1.1842105263157895e-05,
"loss": 0.0001,
"step": 2500
},
{
"epoch": 0.13,
"grad_norm": 0.05126953125,
"learning_rate": 1.1826315789473684e-05,
"loss": 0.0002,
"step": 2510
},
{
"epoch": 0.13,
"grad_norm": 0.002044677734375,
"learning_rate": 1.1810526315789474e-05,
"loss": 0.0001,
"step": 2520
},
{
"epoch": 0.13,
"grad_norm": 0.0023040771484375,
"learning_rate": 1.1794736842105263e-05,
"loss": 0.0001,
"step": 2530
},
{
"epoch": 0.13,
"grad_norm": 0.0032958984375,
"learning_rate": 1.1778947368421053e-05,
"loss": 0.0001,
"step": 2540
},
{
"epoch": 0.13,
"grad_norm": 0.00189208984375,
"learning_rate": 1.1763157894736842e-05,
"loss": 0.0001,
"step": 2550
},
{
"epoch": 0.13,
"grad_norm": 0.005340576171875,
"learning_rate": 1.174736842105263e-05,
"loss": 0.0001,
"step": 2560
},
{
"epoch": 0.13,
"grad_norm": 0.0017547607421875,
"learning_rate": 1.173157894736842e-05,
"loss": 0.0001,
"step": 2570
},
{
"epoch": 0.13,
"grad_norm": 0.003814697265625,
"learning_rate": 1.171578947368421e-05,
"loss": 0.0001,
"step": 2580
},
{
"epoch": 0.13,
"grad_norm": 0.003570556640625,
"learning_rate": 1.1700000000000001e-05,
"loss": 0.0001,
"step": 2590
},
{
"epoch": 0.13,
"grad_norm": 0.01275634765625,
"learning_rate": 1.168421052631579e-05,
"loss": 0.0001,
"step": 2600
},
{
"epoch": 0.13,
"grad_norm": 0.003387451171875,
"learning_rate": 1.166842105263158e-05,
"loss": 0.0003,
"step": 2610
},
{
"epoch": 0.13,
"grad_norm": 0.0038909912109375,
"learning_rate": 1.1652631578947369e-05,
"loss": 0.0001,
"step": 2620
},
{
"epoch": 0.13,
"grad_norm": 0.0042724609375,
"learning_rate": 1.1636842105263158e-05,
"loss": 0.0001,
"step": 2630
},
{
"epoch": 0.13,
"grad_norm": 0.00421142578125,
"learning_rate": 1.1621052631578948e-05,
"loss": 0.0001,
"step": 2640
},
{
"epoch": 0.13,
"grad_norm": 0.01104736328125,
"learning_rate": 1.1605263157894737e-05,
"loss": 0.0001,
"step": 2650
},
{
"epoch": 0.13,
"grad_norm": 0.0091552734375,
"learning_rate": 1.1589473684210527e-05,
"loss": 0.0001,
"step": 2660
},
{
"epoch": 0.13,
"grad_norm": 0.006072998046875,
"learning_rate": 1.1573684210526316e-05,
"loss": 0.0001,
"step": 2670
},
{
"epoch": 0.13,
"grad_norm": 0.0023193359375,
"learning_rate": 1.1557894736842106e-05,
"loss": 0.0001,
"step": 2680
},
{
"epoch": 0.13,
"grad_norm": 0.003936767578125,
"learning_rate": 1.1542105263157895e-05,
"loss": 0.0001,
"step": 2690
},
{
"epoch": 0.14,
"grad_norm": 0.0027008056640625,
"learning_rate": 1.1526315789473683e-05,
"loss": 0.0001,
"step": 2700
},
{
"epoch": 0.14,
"grad_norm": 0.002532958984375,
"learning_rate": 1.1510526315789474e-05,
"loss": 0.0001,
"step": 2710
},
{
"epoch": 0.14,
"grad_norm": 0.005889892578125,
"learning_rate": 1.1494736842105262e-05,
"loss": 0.0001,
"step": 2720
},
{
"epoch": 0.14,
"grad_norm": 0.0037078857421875,
"learning_rate": 1.1478947368421053e-05,
"loss": 0.0001,
"step": 2730
},
{
"epoch": 0.14,
"grad_norm": 0.00174713134765625,
"learning_rate": 1.1463157894736841e-05,
"loss": 0.0001,
"step": 2740
},
{
"epoch": 0.14,
"grad_norm": 0.015625,
"learning_rate": 1.1447368421052632e-05,
"loss": 0.0001,
"step": 2750
},
{
"epoch": 0.14,
"grad_norm": 0.0048828125,
"learning_rate": 1.1431578947368422e-05,
"loss": 0.0001,
"step": 2760
},
{
"epoch": 0.14,
"grad_norm": 0.00555419921875,
"learning_rate": 1.141578947368421e-05,
"loss": 0.0001,
"step": 2770
},
{
"epoch": 0.14,
"grad_norm": 0.00555419921875,
"learning_rate": 1.1400000000000001e-05,
"loss": 0.0003,
"step": 2780
},
{
"epoch": 0.14,
"grad_norm": 0.0050048828125,
"learning_rate": 1.138421052631579e-05,
"loss": 0.0001,
"step": 2790
},
{
"epoch": 0.14,
"grad_norm": 0.003448486328125,
"learning_rate": 1.136842105263158e-05,
"loss": 0.0001,
"step": 2800
},
{
"epoch": 0.14,
"grad_norm": 0.00174713134765625,
"learning_rate": 1.1352631578947369e-05,
"loss": 0.0001,
"step": 2810
},
{
"epoch": 0.14,
"grad_norm": 0.0086669921875,
"learning_rate": 1.1336842105263159e-05,
"loss": 0.0001,
"step": 2820
},
{
"epoch": 0.14,
"grad_norm": 0.003570556640625,
"learning_rate": 1.1321052631578948e-05,
"loss": 0.0001,
"step": 2830
},
{
"epoch": 0.14,
"grad_norm": 0.0036468505859375,
"learning_rate": 1.1305263157894736e-05,
"loss": 0.0001,
"step": 2840
},
{
"epoch": 0.14,
"grad_norm": 0.00372314453125,
"learning_rate": 1.1289473684210527e-05,
"loss": 0.0001,
"step": 2850
},
{
"epoch": 0.14,
"grad_norm": 0.003753662109375,
"learning_rate": 1.1273684210526315e-05,
"loss": 0.0001,
"step": 2860
},
{
"epoch": 0.14,
"grad_norm": 0.0034332275390625,
"learning_rate": 1.1257894736842106e-05,
"loss": 0.0001,
"step": 2870
},
{
"epoch": 0.14,
"grad_norm": 0.00179290771484375,
"learning_rate": 1.1242105263157894e-05,
"loss": 0.0001,
"step": 2880
},
{
"epoch": 0.14,
"grad_norm": 0.002288818359375,
"learning_rate": 1.1226315789473685e-05,
"loss": 0.0001,
"step": 2890
},
{
"epoch": 0.14,
"grad_norm": 0.00927734375,
"learning_rate": 1.1210526315789473e-05,
"loss": 0.0001,
"step": 2900
},
{
"epoch": 0.15,
"grad_norm": 0.00482177734375,
"learning_rate": 1.1194736842105264e-05,
"loss": 0.0001,
"step": 2910
},
{
"epoch": 0.15,
"grad_norm": 0.0021820068359375,
"learning_rate": 1.1178947368421054e-05,
"loss": 0.0001,
"step": 2920
},
{
"epoch": 0.15,
"grad_norm": 0.0017852783203125,
"learning_rate": 1.1163157894736842e-05,
"loss": 0.0001,
"step": 2930
},
{
"epoch": 0.15,
"grad_norm": 0.0022430419921875,
"learning_rate": 1.1147368421052633e-05,
"loss": 0.0001,
"step": 2940
},
{
"epoch": 0.15,
"grad_norm": 0.0023193359375,
"learning_rate": 1.1131578947368421e-05,
"loss": 0.0001,
"step": 2950
},
{
"epoch": 0.15,
"grad_norm": 0.0023956298828125,
"learning_rate": 1.1115789473684212e-05,
"loss": 0.0001,
"step": 2960
},
{
"epoch": 0.15,
"grad_norm": 0.004730224609375,
"learning_rate": 1.11e-05,
"loss": 0.0001,
"step": 2970
},
{
"epoch": 0.15,
"grad_norm": 0.007659912109375,
"learning_rate": 1.108421052631579e-05,
"loss": 0.0001,
"step": 2980
},
{
"epoch": 0.15,
"grad_norm": 0.0079345703125,
"learning_rate": 1.106842105263158e-05,
"loss": 0.0001,
"step": 2990
},
{
"epoch": 0.15,
"grad_norm": 0.002471923828125,
"learning_rate": 1.1052631578947368e-05,
"loss": 0.0001,
"step": 3000
},
{
"epoch": 0.15,
"grad_norm": 0.0057373046875,
"learning_rate": 1.1036842105263158e-05,
"loss": 0.0001,
"step": 3010
},
{
"epoch": 0.15,
"grad_norm": 0.0020904541015625,
"learning_rate": 1.1021052631578947e-05,
"loss": 0.0001,
"step": 3020
},
{
"epoch": 0.15,
"grad_norm": 0.0040283203125,
"learning_rate": 1.1005263157894737e-05,
"loss": 0.0001,
"step": 3030
},
{
"epoch": 0.15,
"grad_norm": 0.0072021484375,
"learning_rate": 1.0989473684210526e-05,
"loss": 0.0001,
"step": 3040
},
{
"epoch": 0.15,
"grad_norm": 0.00179290771484375,
"learning_rate": 1.0973684210526316e-05,
"loss": 0.0002,
"step": 3050
},
{
"epoch": 0.15,
"grad_norm": 0.0022430419921875,
"learning_rate": 1.0957894736842105e-05,
"loss": 0.0001,
"step": 3060
},
{
"epoch": 0.15,
"grad_norm": 0.0030517578125,
"learning_rate": 1.0942105263157894e-05,
"loss": 0.0001,
"step": 3070
},
{
"epoch": 0.15,
"grad_norm": 0.004425048828125,
"learning_rate": 1.0926315789473684e-05,
"loss": 0.0001,
"step": 3080
},
{
"epoch": 0.15,
"grad_norm": 0.0020294189453125,
"learning_rate": 1.0910526315789474e-05,
"loss": 0.0001,
"step": 3090
},
{
"epoch": 0.15,
"grad_norm": 0.00823974609375,
"learning_rate": 1.0894736842105265e-05,
"loss": 0.0001,
"step": 3100
},
{
"epoch": 0.16,
"grad_norm": 0.0017547607421875,
"learning_rate": 1.0878947368421053e-05,
"loss": 0.0001,
"step": 3110
},
{
"epoch": 0.16,
"grad_norm": 0.0017242431640625,
"learning_rate": 1.0863157894736844e-05,
"loss": 0.0001,
"step": 3120
},
{
"epoch": 0.16,
"grad_norm": 0.0018768310546875,
"learning_rate": 1.0847368421052632e-05,
"loss": 0.0001,
"step": 3130
},
{
"epoch": 0.16,
"grad_norm": 0.002593994140625,
"learning_rate": 1.0831578947368421e-05,
"loss": 0.0001,
"step": 3140
},
{
"epoch": 0.16,
"grad_norm": 0.00482177734375,
"learning_rate": 1.0815789473684211e-05,
"loss": 0.0001,
"step": 3150
},
{
"epoch": 0.16,
"grad_norm": 0.00213623046875,
"learning_rate": 1.08e-05,
"loss": 0.0001,
"step": 3160
},
{
"epoch": 0.16,
"grad_norm": 0.0019989013671875,
"learning_rate": 1.078421052631579e-05,
"loss": 0.0001,
"step": 3170
},
{
"epoch": 0.16,
"grad_norm": 0.00390625,
"learning_rate": 1.0768421052631579e-05,
"loss": 0.0001,
"step": 3180
},
{
"epoch": 0.16,
"grad_norm": 0.004730224609375,
"learning_rate": 1.075263157894737e-05,
"loss": 0.0001,
"step": 3190
},
{
"epoch": 0.16,
"grad_norm": 0.0093994140625,
"learning_rate": 1.0736842105263158e-05,
"loss": 0.0001,
"step": 3200
},
{
"epoch": 0.16,
"grad_norm": 0.00494384765625,
"learning_rate": 1.0721052631578947e-05,
"loss": 0.0002,
"step": 3210
},
{
"epoch": 0.16,
"grad_norm": 0.0026702880859375,
"learning_rate": 1.0705263157894737e-05,
"loss": 0.0002,
"step": 3220
},
{
"epoch": 0.16,
"grad_norm": 0.0034332275390625,
"learning_rate": 1.0689473684210526e-05,
"loss": 0.0001,
"step": 3230
},
{
"epoch": 0.16,
"grad_norm": 0.00189208984375,
"learning_rate": 1.0673684210526316e-05,
"loss": 0.0001,
"step": 3240
},
{
"epoch": 0.16,
"grad_norm": 0.1455078125,
"learning_rate": 1.0657894736842106e-05,
"loss": 0.0003,
"step": 3250
},
{
"epoch": 0.16,
"grad_norm": 0.001739501953125,
"learning_rate": 1.0642105263157897e-05,
"loss": 0.0001,
"step": 3260
},
{
"epoch": 0.16,
"grad_norm": 0.003631591796875,
"learning_rate": 1.0626315789473685e-05,
"loss": 0.0001,
"step": 3270
},
{
"epoch": 0.16,
"grad_norm": 0.00726318359375,
"learning_rate": 1.0610526315789474e-05,
"loss": 0.0001,
"step": 3280
},
{
"epoch": 0.16,
"grad_norm": 0.006256103515625,
"learning_rate": 1.0594736842105264e-05,
"loss": 0.0001,
"step": 3290
},
{
"epoch": 0.17,
"grad_norm": 0.0057373046875,
"learning_rate": 1.0578947368421053e-05,
"loss": 0.0,
"step": 3300
},
{
"epoch": 0.17,
"grad_norm": 0.0079345703125,
"learning_rate": 1.0563157894736843e-05,
"loss": 0.0001,
"step": 3310
},
{
"epoch": 0.17,
"grad_norm": 0.003326416015625,
"learning_rate": 1.0547368421052632e-05,
"loss": 0.0001,
"step": 3320
},
{
"epoch": 0.17,
"grad_norm": 0.0040283203125,
"learning_rate": 1.0531578947368422e-05,
"loss": 0.0001,
"step": 3330
},
{
"epoch": 0.17,
"grad_norm": 0.002410888671875,
"learning_rate": 1.051578947368421e-05,
"loss": 0.0001,
"step": 3340
},
{
"epoch": 0.17,
"grad_norm": 0.00244140625,
"learning_rate": 1.05e-05,
"loss": 0.0001,
"step": 3350
},
{
"epoch": 0.17,
"grad_norm": 0.003448486328125,
"learning_rate": 1.048421052631579e-05,
"loss": 0.0001,
"step": 3360
},
{
"epoch": 0.17,
"grad_norm": 0.002105712890625,
"learning_rate": 1.0468421052631578e-05,
"loss": 0.0001,
"step": 3370
},
{
"epoch": 0.17,
"grad_norm": 0.00171661376953125,
"learning_rate": 1.0452631578947369e-05,
"loss": 0.0001,
"step": 3380
},
{
"epoch": 0.17,
"grad_norm": 0.00323486328125,
"learning_rate": 1.0436842105263157e-05,
"loss": 0.0001,
"step": 3390
},
{
"epoch": 0.17,
"grad_norm": 0.002593994140625,
"learning_rate": 1.0421052631578948e-05,
"loss": 0.0001,
"step": 3400
},
{
"epoch": 0.17,
"grad_norm": 0.0029754638671875,
"learning_rate": 1.0405263157894736e-05,
"loss": 0.0001,
"step": 3410
},
{
"epoch": 0.17,
"grad_norm": 0.002288818359375,
"learning_rate": 1.0389473684210527e-05,
"loss": 0.0001,
"step": 3420
},
{
"epoch": 0.17,
"grad_norm": 0.00311279296875,
"learning_rate": 1.0373684210526317e-05,
"loss": 0.0001,
"step": 3430
},
{
"epoch": 0.17,
"grad_norm": 0.005645751953125,
"learning_rate": 1.0357894736842106e-05,
"loss": 0.0001,
"step": 3440
},
{
"epoch": 0.17,
"grad_norm": 0.004119873046875,
"learning_rate": 1.0342105263157896e-05,
"loss": 0.0001,
"step": 3450
},
{
"epoch": 0.17,
"grad_norm": 0.002197265625,
"learning_rate": 1.0326315789473685e-05,
"loss": 0.0001,
"step": 3460
},
{
"epoch": 0.17,
"grad_norm": 0.003173828125,
"learning_rate": 1.0310526315789475e-05,
"loss": 0.0001,
"step": 3470
},
{
"epoch": 0.17,
"grad_norm": 0.0040283203125,
"learning_rate": 1.0294736842105264e-05,
"loss": 0.0001,
"step": 3480
},
{
"epoch": 0.17,
"grad_norm": 0.0030670166015625,
"learning_rate": 1.0278947368421052e-05,
"loss": 0.0001,
"step": 3490
},
{
"epoch": 0.17,
"grad_norm": 0.005950927734375,
"learning_rate": 1.0263157894736843e-05,
"loss": 0.0001,
"step": 3500
},
{
"epoch": 0.18,
"grad_norm": 0.00823974609375,
"learning_rate": 1.0247368421052631e-05,
"loss": 0.0001,
"step": 3510
},
{
"epoch": 0.18,
"grad_norm": 0.0031280517578125,
"learning_rate": 1.0231578947368422e-05,
"loss": 0.0001,
"step": 3520
},
{
"epoch": 0.18,
"grad_norm": 0.003448486328125,
"learning_rate": 1.021578947368421e-05,
"loss": 0.0014,
"step": 3530
},
{
"epoch": 0.18,
"grad_norm": 0.007537841796875,
"learning_rate": 1.02e-05,
"loss": 0.0002,
"step": 3540
},
{
"epoch": 0.18,
"grad_norm": 0.006591796875,
"learning_rate": 1.018421052631579e-05,
"loss": 0.0001,
"step": 3550
},
{
"epoch": 0.18,
"grad_norm": 0.007293701171875,
"learning_rate": 1.0168421052631578e-05,
"loss": 0.0001,
"step": 3560
},
{
"epoch": 0.18,
"grad_norm": 0.0047607421875,
"learning_rate": 1.0152631578947368e-05,
"loss": 0.0003,
"step": 3570
},
{
"epoch": 0.18,
"grad_norm": 0.004150390625,
"learning_rate": 1.0136842105263157e-05,
"loss": 0.0002,
"step": 3580
},
{
"epoch": 0.18,
"grad_norm": 0.0086669921875,
"learning_rate": 1.0121052631578949e-05,
"loss": 0.0001,
"step": 3590
},
{
"epoch": 0.18,
"grad_norm": 0.0016632080078125,
"learning_rate": 1.0105263157894738e-05,
"loss": 0.0001,
"step": 3600
},
{
"epoch": 0.18,
"grad_norm": 0.0028839111328125,
"learning_rate": 1.0089473684210528e-05,
"loss": 0.0001,
"step": 3610
},
{
"epoch": 0.18,
"grad_norm": 0.0048828125,
"learning_rate": 1.0073684210526317e-05,
"loss": 0.0001,
"step": 3620
},
{
"epoch": 0.18,
"grad_norm": 0.00482177734375,
"learning_rate": 1.0057894736842105e-05,
"loss": 0.0001,
"step": 3630
},
{
"epoch": 0.18,
"grad_norm": 0.00213623046875,
"learning_rate": 1.0042105263157896e-05,
"loss": 0.0001,
"step": 3640
},
{
"epoch": 0.18,
"grad_norm": 0.002410888671875,
"learning_rate": 1.0026315789473684e-05,
"loss": 0.0001,
"step": 3650
},
{
"epoch": 0.18,
"grad_norm": 0.0031890869140625,
"learning_rate": 1.0010526315789474e-05,
"loss": 0.0001,
"step": 3660
},
{
"epoch": 0.18,
"grad_norm": 0.004730224609375,
"learning_rate": 9.994736842105263e-06,
"loss": 0.0001,
"step": 3670
},
{
"epoch": 0.18,
"grad_norm": 0.00167083740234375,
"learning_rate": 9.978947368421053e-06,
"loss": 0.0001,
"step": 3680
},
{
"epoch": 0.18,
"grad_norm": 0.00433349609375,
"learning_rate": 9.963157894736842e-06,
"loss": 0.0001,
"step": 3690
},
{
"epoch": 0.18,
"grad_norm": 0.0030517578125,
"learning_rate": 9.94736842105263e-06,
"loss": 0.0001,
"step": 3700
},
{
"epoch": 0.19,
"grad_norm": 0.003662109375,
"learning_rate": 9.931578947368421e-06,
"loss": 0.0001,
"step": 3710
},
{
"epoch": 0.19,
"grad_norm": 0.00518798828125,
"learning_rate": 9.91578947368421e-06,
"loss": 0.0001,
"step": 3720
},
{
"epoch": 0.19,
"grad_norm": 0.0021514892578125,
"learning_rate": 9.9e-06,
"loss": 0.0002,
"step": 3730
},
{
"epoch": 0.19,
"grad_norm": 0.00445556640625,
"learning_rate": 9.884210526315789e-06,
"loss": 0.0001,
"step": 3740
},
{
"epoch": 0.19,
"grad_norm": 0.00179290771484375,
"learning_rate": 9.868421052631579e-06,
"loss": 0.0001,
"step": 3750
},
{
"epoch": 0.19,
"grad_norm": 0.005401611328125,
"learning_rate": 9.85263157894737e-06,
"loss": 0.0001,
"step": 3760
},
{
"epoch": 0.19,
"grad_norm": 0.00164031982421875,
"learning_rate": 9.836842105263158e-06,
"loss": 0.0001,
"step": 3770
},
{
"epoch": 0.19,
"grad_norm": 0.002593994140625,
"learning_rate": 9.821052631578948e-06,
"loss": 0.0001,
"step": 3780
},
{
"epoch": 0.19,
"grad_norm": 0.01043701171875,
"learning_rate": 9.805263157894737e-06,
"loss": 0.0001,
"step": 3790
},
{
"epoch": 0.19,
"grad_norm": 0.00262451171875,
"learning_rate": 9.789473684210527e-06,
"loss": 0.0001,
"step": 3800
},
{
"epoch": 0.19,
"grad_norm": 0.002227783203125,
"learning_rate": 9.773684210526316e-06,
"loss": 0.0001,
"step": 3810
},
{
"epoch": 0.19,
"grad_norm": 0.00299072265625,
"learning_rate": 9.757894736842106e-06,
"loss": 0.0001,
"step": 3820
},
{
"epoch": 0.19,
"grad_norm": 0.00421142578125,
"learning_rate": 9.742105263157895e-06,
"loss": 0.0001,
"step": 3830
},
{
"epoch": 0.19,
"grad_norm": 0.00347900390625,
"learning_rate": 9.726315789473684e-06,
"loss": 0.0001,
"step": 3840
},
{
"epoch": 0.19,
"grad_norm": 0.0035400390625,
"learning_rate": 9.710526315789474e-06,
"loss": 0.0001,
"step": 3850
},
{
"epoch": 0.19,
"grad_norm": 0.00555419921875,
"learning_rate": 9.694736842105263e-06,
"loss": 0.0001,
"step": 3860
},
{
"epoch": 0.19,
"grad_norm": 0.001708984375,
"learning_rate": 9.678947368421053e-06,
"loss": 0.0002,
"step": 3870
},
{
"epoch": 0.19,
"grad_norm": 0.166015625,
"learning_rate": 9.663157894736842e-06,
"loss": 0.0003,
"step": 3880
},
{
"epoch": 0.19,
"grad_norm": 0.00689697265625,
"learning_rate": 9.647368421052632e-06,
"loss": 0.0001,
"step": 3890
},
{
"epoch": 0.2,
"grad_norm": 0.004180908203125,
"learning_rate": 9.63157894736842e-06,
"loss": 0.0001,
"step": 3900
},
{
"epoch": 0.2,
"grad_norm": 0.00885009765625,
"learning_rate": 9.61578947368421e-06,
"loss": 0.0001,
"step": 3910
},
{
"epoch": 0.2,
"grad_norm": 0.001983642578125,
"learning_rate": 9.600000000000001e-06,
"loss": 0.0001,
"step": 3920
},
{
"epoch": 0.2,
"grad_norm": 0.006439208984375,
"learning_rate": 9.58421052631579e-06,
"loss": 0.0001,
"step": 3930
},
{
"epoch": 0.2,
"grad_norm": 0.00439453125,
"learning_rate": 9.56842105263158e-06,
"loss": 0.0001,
"step": 3940
},
{
"epoch": 0.2,
"grad_norm": 0.006805419921875,
"learning_rate": 9.552631578947369e-06,
"loss": 0.0001,
"step": 3950
},
{
"epoch": 0.2,
"grad_norm": 0.0194091796875,
"learning_rate": 9.53684210526316e-06,
"loss": 0.0001,
"step": 3960
},
{
"epoch": 0.2,
"grad_norm": 0.004669189453125,
"learning_rate": 9.521052631578948e-06,
"loss": 0.0001,
"step": 3970
},
{
"epoch": 0.2,
"grad_norm": 0.01385498046875,
"learning_rate": 9.505263157894737e-06,
"loss": 0.0001,
"step": 3980
},
{
"epoch": 0.2,
"grad_norm": 0.00921630859375,
"learning_rate": 9.489473684210527e-06,
"loss": 0.0001,
"step": 3990
},
{
"epoch": 0.2,
"grad_norm": 0.0023040771484375,
"learning_rate": 9.473684210526315e-06,
"loss": 0.0001,
"step": 4000
},
{
"epoch": 0.2,
"eval_loss": 7.299598655663431e-05,
"eval_runtime": 30.2472,
"eval_samples_per_second": 33.061,
"eval_steps_per_second": 33.061,
"step": 4000
},
{
"epoch": 0.2,
"grad_norm": 0.0035400390625,
"learning_rate": 9.457894736842106e-06,
"loss": 0.0001,
"step": 4010
},
{
"epoch": 0.2,
"grad_norm": 0.00494384765625,
"learning_rate": 9.442105263157894e-06,
"loss": 0.0001,
"step": 4020
},
{
"epoch": 0.2,
"grad_norm": 0.002685546875,
"learning_rate": 9.426315789473685e-06,
"loss": 0.0001,
"step": 4030
},
{
"epoch": 0.2,
"grad_norm": 0.00213623046875,
"learning_rate": 9.410526315789473e-06,
"loss": 0.0001,
"step": 4040
},
{
"epoch": 0.2,
"grad_norm": 0.0021820068359375,
"learning_rate": 9.394736842105262e-06,
"loss": 0.0001,
"step": 4050
},
{
"epoch": 0.2,
"grad_norm": 0.00506591796875,
"learning_rate": 9.378947368421052e-06,
"loss": 0.0001,
"step": 4060
},
{
"epoch": 0.2,
"grad_norm": 0.006622314453125,
"learning_rate": 9.363157894736841e-06,
"loss": 0.0001,
"step": 4070
},
{
"epoch": 0.2,
"grad_norm": 0.002410888671875,
"learning_rate": 9.347368421052631e-06,
"loss": 0.0001,
"step": 4080
},
{
"epoch": 0.2,
"grad_norm": 0.0031280517578125,
"learning_rate": 9.331578947368422e-06,
"loss": 0.0001,
"step": 4090
},
{
"epoch": 0.2,
"grad_norm": 0.00335693359375,
"learning_rate": 9.315789473684212e-06,
"loss": 0.0001,
"step": 4100
},
{
"epoch": 0.21,
"grad_norm": 0.0025634765625,
"learning_rate": 9.3e-06,
"loss": 0.0001,
"step": 4110
},
{
"epoch": 0.21,
"grad_norm": 0.002044677734375,
"learning_rate": 9.28421052631579e-06,
"loss": 0.0004,
"step": 4120
},
{
"epoch": 0.21,
"grad_norm": 0.0037078857421875,
"learning_rate": 9.26842105263158e-06,
"loss": 0.0001,
"step": 4130
},
{
"epoch": 0.21,
"grad_norm": 0.0019989013671875,
"learning_rate": 9.252631578947368e-06,
"loss": 0.0001,
"step": 4140
},
{
"epoch": 0.21,
"grad_norm": 0.0034942626953125,
"learning_rate": 9.236842105263159e-06,
"loss": 0.0001,
"step": 4150
},
{
"epoch": 0.21,
"grad_norm": 0.00164794921875,
"learning_rate": 9.221052631578947e-06,
"loss": 0.0,
"step": 4160
},
{
"epoch": 0.21,
"grad_norm": 0.00408935546875,
"learning_rate": 9.205263157894738e-06,
"loss": 0.0001,
"step": 4170
},
{
"epoch": 0.21,
"grad_norm": 0.00811767578125,
"learning_rate": 9.189473684210526e-06,
"loss": 0.0001,
"step": 4180
},
{
"epoch": 0.21,
"grad_norm": 0.0029754638671875,
"learning_rate": 9.173684210526315e-06,
"loss": 0.0001,
"step": 4190
},
{
"epoch": 0.21,
"grad_norm": 0.004730224609375,
"learning_rate": 9.157894736842105e-06,
"loss": 0.0001,
"step": 4200
},
{
"epoch": 0.21,
"grad_norm": 0.005035400390625,
"learning_rate": 9.142105263157894e-06,
"loss": 0.0001,
"step": 4210
},
{
"epoch": 0.21,
"grad_norm": 0.003753662109375,
"learning_rate": 9.126315789473684e-06,
"loss": 0.0001,
"step": 4220
},
{
"epoch": 0.21,
"grad_norm": 0.0036468505859375,
"learning_rate": 9.110526315789473e-06,
"loss": 0.0001,
"step": 4230
},
{
"epoch": 0.21,
"grad_norm": 0.0022430419921875,
"learning_rate": 9.094736842105263e-06,
"loss": 0.0001,
"step": 4240
},
{
"epoch": 0.21,
"grad_norm": 0.002166748046875,
"learning_rate": 9.078947368421054e-06,
"loss": 0.0001,
"step": 4250
},
{
"epoch": 0.21,
"grad_norm": 0.004425048828125,
"learning_rate": 9.063157894736842e-06,
"loss": 0.0001,
"step": 4260
},
{
"epoch": 0.21,
"grad_norm": 0.0045166015625,
"learning_rate": 9.047368421052633e-06,
"loss": 0.0001,
"step": 4270
},
{
"epoch": 0.21,
"grad_norm": 0.00390625,
"learning_rate": 9.031578947368421e-06,
"loss": 0.0001,
"step": 4280
},
{
"epoch": 0.21,
"grad_norm": 0.00167083740234375,
"learning_rate": 9.015789473684212e-06,
"loss": 0.0001,
"step": 4290
},
{
"epoch": 0.21,
"grad_norm": 0.0023345947265625,
"learning_rate": 9e-06,
"loss": 0.0001,
"step": 4300
},
{
"epoch": 0.22,
"grad_norm": 0.0024871826171875,
"learning_rate": 8.98421052631579e-06,
"loss": 0.0001,
"step": 4310
},
{
"epoch": 0.22,
"grad_norm": 0.0028839111328125,
"learning_rate": 8.96842105263158e-06,
"loss": 0.0001,
"step": 4320
},
{
"epoch": 0.22,
"grad_norm": 0.006805419921875,
"learning_rate": 8.952631578947368e-06,
"loss": 0.0001,
"step": 4330
},
{
"epoch": 0.22,
"grad_norm": 0.00445556640625,
"learning_rate": 8.936842105263158e-06,
"loss": 0.0001,
"step": 4340
},
{
"epoch": 0.22,
"grad_norm": 0.003387451171875,
"learning_rate": 8.921052631578947e-06,
"loss": 0.0001,
"step": 4350
},
{
"epoch": 0.22,
"grad_norm": 0.0038909912109375,
"learning_rate": 8.905263157894737e-06,
"loss": 0.0001,
"step": 4360
},
{
"epoch": 0.22,
"grad_norm": 0.006622314453125,
"learning_rate": 8.889473684210526e-06,
"loss": 0.0001,
"step": 4370
},
{
"epoch": 0.22,
"grad_norm": 0.0029449462890625,
"learning_rate": 8.873684210526316e-06,
"loss": 0.0001,
"step": 4380
},
{
"epoch": 0.22,
"grad_norm": 0.004180908203125,
"learning_rate": 8.857894736842105e-06,
"loss": 0.0001,
"step": 4390
},
{
"epoch": 0.22,
"grad_norm": 0.0029754638671875,
"learning_rate": 8.842105263157893e-06,
"loss": 0.0001,
"step": 4400
},
{
"epoch": 0.22,
"grad_norm": 0.0038299560546875,
"learning_rate": 8.826315789473684e-06,
"loss": 0.0001,
"step": 4410
},
{
"epoch": 0.22,
"grad_norm": 0.01068115234375,
"learning_rate": 8.810526315789474e-06,
"loss": 0.0001,
"step": 4420
},
{
"epoch": 0.22,
"grad_norm": 0.002288818359375,
"learning_rate": 8.794736842105264e-06,
"loss": 0.0001,
"step": 4430
},
{
"epoch": 0.22,
"grad_norm": 0.0022430419921875,
"learning_rate": 8.778947368421053e-06,
"loss": 0.0001,
"step": 4440
},
{
"epoch": 0.22,
"grad_norm": 0.00408935546875,
"learning_rate": 8.763157894736843e-06,
"loss": 0.0001,
"step": 4450
},
{
"epoch": 0.22,
"grad_norm": 0.00191497802734375,
"learning_rate": 8.747368421052632e-06,
"loss": 0.0001,
"step": 4460
},
{
"epoch": 0.22,
"grad_norm": 0.00201416015625,
"learning_rate": 8.731578947368422e-06,
"loss": 0.0001,
"step": 4470
},
{
"epoch": 0.22,
"grad_norm": 0.0059814453125,
"learning_rate": 8.715789473684211e-06,
"loss": 0.0001,
"step": 4480
},
{
"epoch": 0.22,
"grad_norm": 0.0040283203125,
"learning_rate": 8.7e-06,
"loss": 0.0001,
"step": 4490
},
{
"epoch": 0.23,
"grad_norm": 0.0016326904296875,
"learning_rate": 8.68421052631579e-06,
"loss": 0.0001,
"step": 4500
},
{
"epoch": 0.23,
"grad_norm": 0.0035400390625,
"learning_rate": 8.668421052631579e-06,
"loss": 0.0001,
"step": 4510
},
{
"epoch": 0.23,
"grad_norm": 0.0028839111328125,
"learning_rate": 8.652631578947369e-06,
"loss": 0.0001,
"step": 4520
},
{
"epoch": 0.23,
"grad_norm": 0.006683349609375,
"learning_rate": 8.636842105263158e-06,
"loss": 0.0001,
"step": 4530
},
{
"epoch": 0.23,
"grad_norm": 0.001983642578125,
"learning_rate": 8.621052631578948e-06,
"loss": 0.0001,
"step": 4540
},
{
"epoch": 0.23,
"grad_norm": 0.001983642578125,
"learning_rate": 8.605263157894737e-06,
"loss": 0.0001,
"step": 4550
},
{
"epoch": 0.23,
"grad_norm": 0.003692626953125,
"learning_rate": 8.589473684210525e-06,
"loss": 0.0001,
"step": 4560
},
{
"epoch": 0.23,
"grad_norm": 0.0022430419921875,
"learning_rate": 8.573684210526316e-06,
"loss": 0.0001,
"step": 4570
},
{
"epoch": 0.23,
"grad_norm": 0.001708984375,
"learning_rate": 8.557894736842104e-06,
"loss": 0.0001,
"step": 4580
},
{
"epoch": 0.23,
"grad_norm": 0.001678466796875,
"learning_rate": 8.542105263157896e-06,
"loss": 0.0001,
"step": 4590
},
{
"epoch": 0.23,
"grad_norm": 0.00506591796875,
"learning_rate": 8.526315789473685e-06,
"loss": 0.0001,
"step": 4600
},
{
"epoch": 0.23,
"grad_norm": 0.0020294189453125,
"learning_rate": 8.510526315789475e-06,
"loss": 0.0001,
"step": 4610
},
{
"epoch": 0.23,
"grad_norm": 0.0025634765625,
"learning_rate": 8.494736842105264e-06,
"loss": 0.0001,
"step": 4620
},
{
"epoch": 0.23,
"grad_norm": 0.004974365234375,
"learning_rate": 8.478947368421053e-06,
"loss": 0.0001,
"step": 4630
},
{
"epoch": 0.23,
"grad_norm": 0.0030059814453125,
"learning_rate": 8.463157894736843e-06,
"loss": 0.0001,
"step": 4640
},
{
"epoch": 0.23,
"grad_norm": 0.00732421875,
"learning_rate": 8.447368421052632e-06,
"loss": 0.0001,
"step": 4650
},
{
"epoch": 0.23,
"grad_norm": 0.0024566650390625,
"learning_rate": 8.431578947368422e-06,
"loss": 0.0001,
"step": 4660
},
{
"epoch": 0.23,
"grad_norm": 0.004119873046875,
"learning_rate": 8.41578947368421e-06,
"loss": 0.0001,
"step": 4670
},
{
"epoch": 0.23,
"grad_norm": 0.002471923828125,
"learning_rate": 8.400000000000001e-06,
"loss": 0.0001,
"step": 4680
},
{
"epoch": 0.23,
"grad_norm": 0.00537109375,
"learning_rate": 8.38421052631579e-06,
"loss": 0.0001,
"step": 4690
},
{
"epoch": 0.23,
"grad_norm": 0.005096435546875,
"learning_rate": 8.368421052631578e-06,
"loss": 0.0001,
"step": 4700
},
{
"epoch": 0.24,
"grad_norm": 0.0033416748046875,
"learning_rate": 8.352631578947369e-06,
"loss": 0.0001,
"step": 4710
},
{
"epoch": 0.24,
"grad_norm": 0.009521484375,
"learning_rate": 8.336842105263157e-06,
"loss": 0.0001,
"step": 4720
},
{
"epoch": 0.24,
"grad_norm": 0.002197265625,
"learning_rate": 8.321052631578947e-06,
"loss": 0.0001,
"step": 4730
},
{
"epoch": 0.24,
"grad_norm": 0.00567626953125,
"learning_rate": 8.305263157894736e-06,
"loss": 0.0001,
"step": 4740
},
{
"epoch": 0.24,
"grad_norm": 0.00176239013671875,
"learning_rate": 8.289473684210526e-06,
"loss": 0.0002,
"step": 4750
},
{
"epoch": 0.24,
"grad_norm": 0.0029144287109375,
"learning_rate": 8.273684210526317e-06,
"loss": 0.0001,
"step": 4760
},
{
"epoch": 0.24,
"grad_norm": 0.0019683837890625,
"learning_rate": 8.257894736842105e-06,
"loss": 0.0001,
"step": 4770
},
{
"epoch": 0.24,
"grad_norm": 0.00421142578125,
"learning_rate": 8.242105263157896e-06,
"loss": 0.0002,
"step": 4780
},
{
"epoch": 0.24,
"grad_norm": 0.06298828125,
"learning_rate": 8.226315789473684e-06,
"loss": 0.0002,
"step": 4790
},
{
"epoch": 0.24,
"grad_norm": 0.00191497802734375,
"learning_rate": 8.210526315789475e-06,
"loss": 0.0001,
"step": 4800
},
{
"epoch": 0.24,
"grad_norm": 0.00958251953125,
"learning_rate": 8.194736842105263e-06,
"loss": 0.0001,
"step": 4810
},
{
"epoch": 0.24,
"grad_norm": 0.00421142578125,
"learning_rate": 8.178947368421054e-06,
"loss": 0.0001,
"step": 4820
},
{
"epoch": 0.24,
"grad_norm": 0.00225830078125,
"learning_rate": 8.163157894736842e-06,
"loss": 0.0001,
"step": 4830
},
{
"epoch": 0.24,
"grad_norm": 0.002288818359375,
"learning_rate": 8.147368421052631e-06,
"loss": 0.0001,
"step": 4840
},
{
"epoch": 0.24,
"grad_norm": 0.00750732421875,
"learning_rate": 8.131578947368421e-06,
"loss": 0.0001,
"step": 4850
},
{
"epoch": 0.24,
"grad_norm": 0.002838134765625,
"learning_rate": 8.11578947368421e-06,
"loss": 0.0001,
"step": 4860
},
{
"epoch": 0.24,
"grad_norm": 0.00323486328125,
"learning_rate": 8.1e-06,
"loss": 0.0001,
"step": 4870
},
{
"epoch": 0.24,
"grad_norm": 0.00180816650390625,
"learning_rate": 8.084210526315789e-06,
"loss": 0.0001,
"step": 4880
},
{
"epoch": 0.24,
"grad_norm": 0.00616455078125,
"learning_rate": 8.06842105263158e-06,
"loss": 0.0001,
"step": 4890
},
{
"epoch": 0.24,
"grad_norm": 0.0021820068359375,
"learning_rate": 8.052631578947368e-06,
"loss": 0.0001,
"step": 4900
},
{
"epoch": 0.25,
"grad_norm": 0.00787353515625,
"learning_rate": 8.036842105263157e-06,
"loss": 0.0001,
"step": 4910
},
{
"epoch": 0.25,
"grad_norm": 0.00167083740234375,
"learning_rate": 8.021052631578949e-06,
"loss": 0.0001,
"step": 4920
},
{
"epoch": 0.25,
"grad_norm": 0.0034027099609375,
"learning_rate": 8.005263157894737e-06,
"loss": 0.0001,
"step": 4930
},
{
"epoch": 0.25,
"grad_norm": 0.00469970703125,
"learning_rate": 7.989473684210528e-06,
"loss": 0.0001,
"step": 4940
},
{
"epoch": 0.25,
"grad_norm": 0.0037078857421875,
"learning_rate": 7.973684210526316e-06,
"loss": 0.0002,
"step": 4950
},
{
"epoch": 0.25,
"grad_norm": 0.005889892578125,
"learning_rate": 7.957894736842107e-06,
"loss": 0.0001,
"step": 4960
},
{
"epoch": 0.25,
"grad_norm": 0.00162506103515625,
"learning_rate": 7.942105263157895e-06,
"loss": 0.0001,
"step": 4970
},
{
"epoch": 0.25,
"grad_norm": 0.00164031982421875,
"learning_rate": 7.926315789473684e-06,
"loss": 0.0001,
"step": 4980
},
{
"epoch": 0.25,
"grad_norm": 0.00982666015625,
"learning_rate": 7.910526315789474e-06,
"loss": 0.0001,
"step": 4990
},
{
"epoch": 0.25,
"grad_norm": 0.0023345947265625,
"learning_rate": 7.894736842105263e-06,
"loss": 0.0002,
"step": 5000
},
{
"epoch": 0.25,
"grad_norm": 0.0086669921875,
"learning_rate": 7.878947368421053e-06,
"loss": 0.0001,
"step": 5010
},
{
"epoch": 0.25,
"grad_norm": 0.002105712890625,
"learning_rate": 7.863157894736842e-06,
"loss": 0.0001,
"step": 5020
},
{
"epoch": 0.25,
"grad_norm": 0.0031890869140625,
"learning_rate": 7.847368421052632e-06,
"loss": 0.0002,
"step": 5030
},
{
"epoch": 0.25,
"grad_norm": 0.006134033203125,
"learning_rate": 7.831578947368421e-06,
"loss": 0.0001,
"step": 5040
},
{
"epoch": 0.25,
"grad_norm": 0.0032806396484375,
"learning_rate": 7.81578947368421e-06,
"loss": 0.0001,
"step": 5050
},
{
"epoch": 0.25,
"grad_norm": 0.0033416748046875,
"learning_rate": 7.8e-06,
"loss": 0.0001,
"step": 5060
},
{
"epoch": 0.25,
"grad_norm": 0.00347900390625,
"learning_rate": 7.784210526315789e-06,
"loss": 0.0001,
"step": 5070
},
{
"epoch": 0.25,
"grad_norm": 0.0030059814453125,
"learning_rate": 7.768421052631579e-06,
"loss": 0.0001,
"step": 5080
},
{
"epoch": 0.25,
"grad_norm": 0.0086669921875,
"learning_rate": 7.75263157894737e-06,
"loss": 0.0001,
"step": 5090
},
{
"epoch": 0.26,
"grad_norm": 0.00160980224609375,
"learning_rate": 7.73684210526316e-06,
"loss": 0.0001,
"step": 5100
},
{
"epoch": 0.26,
"grad_norm": 0.0019683837890625,
"learning_rate": 7.721052631578948e-06,
"loss": 0.0001,
"step": 5110
},
{
"epoch": 0.26,
"grad_norm": 0.00390625,
"learning_rate": 7.705263157894737e-06,
"loss": 0.0002,
"step": 5120
},
{
"epoch": 0.26,
"grad_norm": 0.0023345947265625,
"learning_rate": 7.689473684210527e-06,
"loss": 0.0001,
"step": 5130
},
{
"epoch": 0.26,
"grad_norm": 0.0023193359375,
"learning_rate": 7.673684210526316e-06,
"loss": 0.0001,
"step": 5140
},
{
"epoch": 0.26,
"grad_norm": 0.003204345703125,
"learning_rate": 7.657894736842106e-06,
"loss": 0.0001,
"step": 5150
},
{
"epoch": 0.26,
"grad_norm": 0.00323486328125,
"learning_rate": 7.642105263157895e-06,
"loss": 0.0001,
"step": 5160
},
{
"epoch": 0.26,
"grad_norm": 0.00982666015625,
"learning_rate": 7.626315789473685e-06,
"loss": 0.0001,
"step": 5170
},
{
"epoch": 0.26,
"grad_norm": 0.00191497802734375,
"learning_rate": 7.610526315789474e-06,
"loss": 0.0001,
"step": 5180
},
{
"epoch": 0.26,
"grad_norm": 0.00860595703125,
"learning_rate": 7.594736842105262e-06,
"loss": 0.0001,
"step": 5190
},
{
"epoch": 0.26,
"grad_norm": 0.00396728515625,
"learning_rate": 7.578947368421053e-06,
"loss": 0.0001,
"step": 5200
},
{
"epoch": 0.26,
"grad_norm": 0.0020599365234375,
"learning_rate": 7.563157894736842e-06,
"loss": 0.0001,
"step": 5210
},
{
"epoch": 0.26,
"grad_norm": 0.0030517578125,
"learning_rate": 7.5473684210526326e-06,
"loss": 0.0001,
"step": 5220
},
{
"epoch": 0.26,
"grad_norm": 0.00274658203125,
"learning_rate": 7.531578947368421e-06,
"loss": 0.0001,
"step": 5230
},
{
"epoch": 0.26,
"grad_norm": 0.007568359375,
"learning_rate": 7.5157894736842115e-06,
"loss": 0.0001,
"step": 5240
},
{
"epoch": 0.26,
"grad_norm": 0.01055908203125,
"learning_rate": 7.5e-06,
"loss": 0.0001,
"step": 5250
},
{
"epoch": 0.26,
"grad_norm": 0.0067138671875,
"learning_rate": 7.48421052631579e-06,
"loss": 0.0001,
"step": 5260
},
{
"epoch": 0.26,
"grad_norm": 0.003753662109375,
"learning_rate": 7.468421052631579e-06,
"loss": 0.0001,
"step": 5270
},
{
"epoch": 0.26,
"grad_norm": 0.005523681640625,
"learning_rate": 7.452631578947369e-06,
"loss": 0.0002,
"step": 5280
},
{
"epoch": 0.26,
"grad_norm": 0.01080322265625,
"learning_rate": 7.436842105263157e-06,
"loss": 0.0001,
"step": 5290
},
{
"epoch": 0.27,
"grad_norm": 0.0020904541015625,
"learning_rate": 7.421052631578948e-06,
"loss": 0.0001,
"step": 5300
},
{
"epoch": 0.27,
"grad_norm": 0.003753662109375,
"learning_rate": 7.405263157894737e-06,
"loss": 0.0001,
"step": 5310
},
{
"epoch": 0.27,
"grad_norm": 0.00396728515625,
"learning_rate": 7.389473684210527e-06,
"loss": 0.0001,
"step": 5320
},
{
"epoch": 0.27,
"grad_norm": 0.0037078857421875,
"learning_rate": 7.373684210526316e-06,
"loss": 0.0001,
"step": 5330
},
{
"epoch": 0.27,
"grad_norm": 0.006378173828125,
"learning_rate": 7.357894736842106e-06,
"loss": 0.0001,
"step": 5340
},
{
"epoch": 0.27,
"grad_norm": 0.00665283203125,
"learning_rate": 7.342105263157895e-06,
"loss": 0.0001,
"step": 5350
},
{
"epoch": 0.27,
"grad_norm": 0.00186920166015625,
"learning_rate": 7.326315789473684e-06,
"loss": 0.0001,
"step": 5360
},
{
"epoch": 0.27,
"grad_norm": 0.0029144287109375,
"learning_rate": 7.310526315789473e-06,
"loss": 0.0001,
"step": 5370
},
{
"epoch": 0.27,
"grad_norm": 0.006561279296875,
"learning_rate": 7.2947368421052636e-06,
"loss": 0.0001,
"step": 5380
},
{
"epoch": 0.27,
"grad_norm": 0.01513671875,
"learning_rate": 7.278947368421053e-06,
"loss": 0.0001,
"step": 5390
},
{
"epoch": 0.27,
"grad_norm": 0.0023956298828125,
"learning_rate": 7.2631578947368426e-06,
"loss": 0.0001,
"step": 5400
},
{
"epoch": 0.27,
"grad_norm": 0.002166748046875,
"learning_rate": 7.247368421052632e-06,
"loss": 0.0001,
"step": 5410
},
{
"epoch": 0.27,
"grad_norm": 0.0029296875,
"learning_rate": 7.2315789473684215e-06,
"loss": 0.0001,
"step": 5420
},
{
"epoch": 0.27,
"grad_norm": 0.0029296875,
"learning_rate": 7.21578947368421e-06,
"loss": 0.0001,
"step": 5430
},
{
"epoch": 0.27,
"grad_norm": 0.01165771484375,
"learning_rate": 7.2e-06,
"loss": 0.0001,
"step": 5440
},
{
"epoch": 0.27,
"grad_norm": 0.006011962890625,
"learning_rate": 7.184210526315789e-06,
"loss": 0.0001,
"step": 5450
},
{
"epoch": 0.27,
"grad_norm": 0.00225830078125,
"learning_rate": 7.1684210526315795e-06,
"loss": 0.0,
"step": 5460
},
{
"epoch": 0.27,
"grad_norm": 0.00341796875,
"learning_rate": 7.152631578947369e-06,
"loss": 0.0,
"step": 5470
},
{
"epoch": 0.27,
"grad_norm": 0.0030670166015625,
"learning_rate": 7.1368421052631585e-06,
"loss": 0.0001,
"step": 5480
},
{
"epoch": 0.27,
"grad_norm": 0.00185394287109375,
"learning_rate": 7.121052631578948e-06,
"loss": 0.0001,
"step": 5490
},
{
"epoch": 0.28,
"grad_norm": 0.00194549560546875,
"learning_rate": 7.105263157894737e-06,
"loss": 0.0001,
"step": 5500
},
{
"epoch": 0.28,
"grad_norm": 0.003204345703125,
"learning_rate": 7.089473684210526e-06,
"loss": 0.0001,
"step": 5510
},
{
"epoch": 0.28,
"grad_norm": 0.00164794921875,
"learning_rate": 7.073684210526316e-06,
"loss": 0.0001,
"step": 5520
},
{
"epoch": 0.28,
"grad_norm": 0.00634765625,
"learning_rate": 7.057894736842105e-06,
"loss": 0.0001,
"step": 5530
},
{
"epoch": 0.28,
"grad_norm": 0.0021209716796875,
"learning_rate": 7.042105263157895e-06,
"loss": 0.0001,
"step": 5540
},
{
"epoch": 0.28,
"grad_norm": 0.00592041015625,
"learning_rate": 7.026315789473685e-06,
"loss": 0.0001,
"step": 5550
},
{
"epoch": 0.28,
"grad_norm": 0.001983642578125,
"learning_rate": 7.010526315789474e-06,
"loss": 0.0001,
"step": 5560
},
{
"epoch": 0.28,
"grad_norm": 0.0020599365234375,
"learning_rate": 6.994736842105263e-06,
"loss": 0.0001,
"step": 5570
},
{
"epoch": 0.28,
"grad_norm": 0.002105712890625,
"learning_rate": 6.9789473684210525e-06,
"loss": 0.0001,
"step": 5580
},
{
"epoch": 0.28,
"grad_norm": 0.0078125,
"learning_rate": 6.963157894736842e-06,
"loss": 0.0001,
"step": 5590
},
{
"epoch": 0.28,
"grad_norm": 0.00250244140625,
"learning_rate": 6.9473684210526315e-06,
"loss": 0.0001,
"step": 5600
},
{
"epoch": 0.28,
"grad_norm": 0.004364013671875,
"learning_rate": 6.931578947368421e-06,
"loss": 0.0001,
"step": 5610
},
{
"epoch": 0.28,
"grad_norm": 0.003662109375,
"learning_rate": 6.9157894736842105e-06,
"loss": 0.0001,
"step": 5620
},
{
"epoch": 0.28,
"grad_norm": 0.00170135498046875,
"learning_rate": 6.900000000000001e-06,
"loss": 0.0001,
"step": 5630
},
{
"epoch": 0.28,
"grad_norm": 0.002777099609375,
"learning_rate": 6.8842105263157895e-06,
"loss": 0.0001,
"step": 5640
},
{
"epoch": 0.28,
"grad_norm": 0.0027313232421875,
"learning_rate": 6.868421052631579e-06,
"loss": 0.0001,
"step": 5650
},
{
"epoch": 0.28,
"grad_norm": 0.00469970703125,
"learning_rate": 6.8526315789473685e-06,
"loss": 0.0001,
"step": 5660
},
{
"epoch": 0.28,
"grad_norm": 0.006500244140625,
"learning_rate": 6.836842105263158e-06,
"loss": 0.0001,
"step": 5670
},
{
"epoch": 0.28,
"grad_norm": 0.00299072265625,
"learning_rate": 6.8210526315789475e-06,
"loss": 0.0001,
"step": 5680
},
{
"epoch": 0.28,
"grad_norm": 0.0030364990234375,
"learning_rate": 6.805263157894737e-06,
"loss": 0.0001,
"step": 5690
},
{
"epoch": 0.28,
"grad_norm": 0.0017242431640625,
"learning_rate": 6.7894736842105264e-06,
"loss": 0.0002,
"step": 5700
},
{
"epoch": 0.29,
"grad_norm": 0.00174713134765625,
"learning_rate": 6.773684210526317e-06,
"loss": 0.0001,
"step": 5710
},
{
"epoch": 0.29,
"grad_norm": 0.006744384765625,
"learning_rate": 6.7578947368421054e-06,
"loss": 0.0001,
"step": 5720
},
{
"epoch": 0.29,
"grad_norm": 0.00640869140625,
"learning_rate": 6.742105263157895e-06,
"loss": 0.0001,
"step": 5730
},
{
"epoch": 0.29,
"grad_norm": 0.00274658203125,
"learning_rate": 6.726315789473684e-06,
"loss": 0.0001,
"step": 5740
},
{
"epoch": 0.29,
"grad_norm": 0.0017242431640625,
"learning_rate": 6.710526315789474e-06,
"loss": 0.0001,
"step": 5750
},
{
"epoch": 0.29,
"grad_norm": 0.00836181640625,
"learning_rate": 6.694736842105263e-06,
"loss": 0.0001,
"step": 5760
},
{
"epoch": 0.29,
"grad_norm": 0.0086669921875,
"learning_rate": 6.678947368421053e-06,
"loss": 0.0001,
"step": 5770
},
{
"epoch": 0.29,
"grad_norm": 0.0224609375,
"learning_rate": 6.663157894736842e-06,
"loss": 0.0002,
"step": 5780
},
{
"epoch": 0.29,
"grad_norm": 0.00506591796875,
"learning_rate": 6.647368421052631e-06,
"loss": 0.0001,
"step": 5790
},
{
"epoch": 0.29,
"grad_norm": 0.0036773681640625,
"learning_rate": 6.631578947368421e-06,
"loss": 0.0001,
"step": 5800
},
{
"epoch": 0.29,
"grad_norm": 0.0034942626953125,
"learning_rate": 6.615789473684211e-06,
"loss": 0.0001,
"step": 5810
},
{
"epoch": 0.29,
"grad_norm": 0.00213623046875,
"learning_rate": 6.6e-06,
"loss": 0.0001,
"step": 5820
},
{
"epoch": 0.29,
"grad_norm": 0.003204345703125,
"learning_rate": 6.58421052631579e-06,
"loss": 0.0001,
"step": 5830
},
{
"epoch": 0.29,
"grad_norm": 0.003173828125,
"learning_rate": 6.568421052631579e-06,
"loss": 0.0001,
"step": 5840
},
{
"epoch": 0.29,
"grad_norm": 0.00567626953125,
"learning_rate": 6.552631578947369e-06,
"loss": 0.0001,
"step": 5850
},
{
"epoch": 0.29,
"grad_norm": 0.0017852783203125,
"learning_rate": 6.5368421052631575e-06,
"loss": 0.0001,
"step": 5860
},
{
"epoch": 0.29,
"grad_norm": 0.00286865234375,
"learning_rate": 6.521052631578947e-06,
"loss": 0.0001,
"step": 5870
},
{
"epoch": 0.29,
"grad_norm": 0.00518798828125,
"learning_rate": 6.505263157894737e-06,
"loss": 0.0001,
"step": 5880
},
{
"epoch": 0.29,
"grad_norm": 0.001983642578125,
"learning_rate": 6.489473684210527e-06,
"loss": 0.0001,
"step": 5890
},
{
"epoch": 0.29,
"grad_norm": 0.00445556640625,
"learning_rate": 6.473684210526316e-06,
"loss": 0.0001,
"step": 5900
},
{
"epoch": 0.3,
"grad_norm": 0.003662109375,
"learning_rate": 6.457894736842106e-06,
"loss": 0.0001,
"step": 5910
},
{
"epoch": 0.3,
"grad_norm": 0.004150390625,
"learning_rate": 6.442105263157895e-06,
"loss": 0.0001,
"step": 5920
},
{
"epoch": 0.3,
"grad_norm": 0.001983642578125,
"learning_rate": 6.426315789473684e-06,
"loss": 0.0001,
"step": 5930
},
{
"epoch": 0.3,
"grad_norm": 0.00201416015625,
"learning_rate": 6.410526315789473e-06,
"loss": 0.0,
"step": 5940
},
{
"epoch": 0.3,
"grad_norm": 0.001708984375,
"learning_rate": 6.394736842105263e-06,
"loss": 0.0,
"step": 5950
},
{
"epoch": 0.3,
"grad_norm": 0.0032958984375,
"learning_rate": 6.378947368421053e-06,
"loss": 0.0001,
"step": 5960
},
{
"epoch": 0.3,
"grad_norm": 0.0235595703125,
"learning_rate": 6.363157894736843e-06,
"loss": 0.0001,
"step": 5970
},
{
"epoch": 0.3,
"grad_norm": 0.006439208984375,
"learning_rate": 6.347368421052632e-06,
"loss": 0.0001,
"step": 5980
},
{
"epoch": 0.3,
"grad_norm": 0.0019989013671875,
"learning_rate": 6.331578947368422e-06,
"loss": 0.0001,
"step": 5990
},
{
"epoch": 0.3,
"grad_norm": 0.004425048828125,
"learning_rate": 6.31578947368421e-06,
"loss": 0.0001,
"step": 6000
},
{
"epoch": 0.3,
"eval_loss": 8.526918099960312e-05,
"eval_runtime": 30.1393,
"eval_samples_per_second": 33.179,
"eval_steps_per_second": 33.179,
"step": 6000
},
{
"epoch": 0.3,
"grad_norm": 0.0020599365234375,
"learning_rate": 6.3e-06,
"loss": 0.0001,
"step": 6010
},
{
"epoch": 0.3,
"grad_norm": 0.002166748046875,
"learning_rate": 6.284210526315789e-06,
"loss": 0.0001,
"step": 6020
},
{
"epoch": 0.3,
"grad_norm": 0.00726318359375,
"learning_rate": 6.268421052631579e-06,
"loss": 0.0001,
"step": 6030
},
{
"epoch": 0.3,
"grad_norm": 0.0029296875,
"learning_rate": 6.252631578947368e-06,
"loss": 0.0001,
"step": 6040
},
{
"epoch": 0.3,
"grad_norm": 0.00238037109375,
"learning_rate": 6.236842105263159e-06,
"loss": 0.0001,
"step": 6050
},
{
"epoch": 0.3,
"grad_norm": 0.002471923828125,
"learning_rate": 6.221052631578948e-06,
"loss": 0.0001,
"step": 6060
},
{
"epoch": 0.3,
"grad_norm": 0.006744384765625,
"learning_rate": 6.205263157894737e-06,
"loss": 0.0001,
"step": 6070
},
{
"epoch": 0.3,
"grad_norm": 0.002349853515625,
"learning_rate": 6.189473684210526e-06,
"loss": 0.0001,
"step": 6080
},
{
"epoch": 0.3,
"grad_norm": 0.01123046875,
"learning_rate": 6.173684210526316e-06,
"loss": 0.0001,
"step": 6090
},
{
"epoch": 0.3,
"grad_norm": 0.00193023681640625,
"learning_rate": 6.157894736842105e-06,
"loss": 0.0002,
"step": 6100
},
{
"epoch": 0.31,
"grad_norm": 0.0035247802734375,
"learning_rate": 6.142105263157895e-06,
"loss": 0.0,
"step": 6110
},
{
"epoch": 0.31,
"grad_norm": 0.00213623046875,
"learning_rate": 6.126315789473684e-06,
"loss": 0.0001,
"step": 6120
},
{
"epoch": 0.31,
"grad_norm": 0.00665283203125,
"learning_rate": 6.1105263157894746e-06,
"loss": 0.0001,
"step": 6130
},
{
"epoch": 0.31,
"grad_norm": 0.002166748046875,
"learning_rate": 6.094736842105263e-06,
"loss": 0.0001,
"step": 6140
},
{
"epoch": 0.31,
"grad_norm": 0.0029296875,
"learning_rate": 6.078947368421053e-06,
"loss": 0.0001,
"step": 6150
},
{
"epoch": 0.31,
"grad_norm": 0.0108642578125,
"learning_rate": 6.063157894736842e-06,
"loss": 0.0001,
"step": 6160
},
{
"epoch": 0.31,
"grad_norm": 0.0020294189453125,
"learning_rate": 6.047368421052632e-06,
"loss": 0.0001,
"step": 6170
},
{
"epoch": 0.31,
"grad_norm": 0.00494384765625,
"learning_rate": 6.031578947368421e-06,
"loss": 0.0001,
"step": 6180
},
{
"epoch": 0.31,
"grad_norm": 0.0037078857421875,
"learning_rate": 6.015789473684211e-06,
"loss": 0.0001,
"step": 6190
},
{
"epoch": 0.31,
"grad_norm": 0.004058837890625,
"learning_rate": 6e-06,
"loss": 0.0001,
"step": 6200
},
{
"epoch": 0.31,
"grad_norm": 0.002105712890625,
"learning_rate": 5.98421052631579e-06,
"loss": 0.0001,
"step": 6210
},
{
"epoch": 0.31,
"grad_norm": 0.006072998046875,
"learning_rate": 5.968421052631579e-06,
"loss": 0.0001,
"step": 6220
},
{
"epoch": 0.31,
"grad_norm": 0.0052490234375,
"learning_rate": 5.952631578947369e-06,
"loss": 0.0001,
"step": 6230
},
{
"epoch": 0.31,
"grad_norm": 0.001556396484375,
"learning_rate": 5.936842105263158e-06,
"loss": 0.0001,
"step": 6240
},
{
"epoch": 0.31,
"grad_norm": 0.0029296875,
"learning_rate": 5.921052631578948e-06,
"loss": 0.0001,
"step": 6250
},
{
"epoch": 0.31,
"grad_norm": 0.0028228759765625,
"learning_rate": 5.905263157894737e-06,
"loss": 0.0,
"step": 6260
},
{
"epoch": 0.31,
"grad_norm": 0.0035400390625,
"learning_rate": 5.889473684210527e-06,
"loss": 0.0001,
"step": 6270
},
{
"epoch": 0.31,
"grad_norm": 0.006561279296875,
"learning_rate": 5.873684210526315e-06,
"loss": 0.0001,
"step": 6280
},
{
"epoch": 0.31,
"grad_norm": 0.00592041015625,
"learning_rate": 5.857894736842105e-06,
"loss": 0.0001,
"step": 6290
},
{
"epoch": 0.32,
"grad_norm": 0.003570556640625,
"learning_rate": 5.842105263157895e-06,
"loss": 0.0001,
"step": 6300
},
{
"epoch": 0.32,
"grad_norm": 0.00616455078125,
"learning_rate": 5.8263157894736846e-06,
"loss": 0.0001,
"step": 6310
},
{
"epoch": 0.32,
"grad_norm": 0.006805419921875,
"learning_rate": 5.810526315789474e-06,
"loss": 0.0001,
"step": 6320
},
{
"epoch": 0.32,
"grad_norm": 0.0027618408203125,
"learning_rate": 5.7947368421052635e-06,
"loss": 0.0001,
"step": 6330
},
{
"epoch": 0.32,
"grad_norm": 0.0020904541015625,
"learning_rate": 5.778947368421053e-06,
"loss": 0.0001,
"step": 6340
},
{
"epoch": 0.32,
"grad_norm": 0.004486083984375,
"learning_rate": 5.763157894736842e-06,
"loss": 0.0001,
"step": 6350
},
{
"epoch": 0.32,
"grad_norm": 0.001678466796875,
"learning_rate": 5.747368421052631e-06,
"loss": 0.0001,
"step": 6360
},
{
"epoch": 0.32,
"grad_norm": 0.0038909912109375,
"learning_rate": 5.731578947368421e-06,
"loss": 0.0001,
"step": 6370
},
{
"epoch": 0.32,
"grad_norm": 0.0023956298828125,
"learning_rate": 5.715789473684211e-06,
"loss": 0.0001,
"step": 6380
},
{
"epoch": 0.32,
"grad_norm": 0.002349853515625,
"learning_rate": 5.7000000000000005e-06,
"loss": 0.0001,
"step": 6390
},
{
"epoch": 0.32,
"grad_norm": 0.005340576171875,
"learning_rate": 5.68421052631579e-06,
"loss": 0.0001,
"step": 6400
},
{
"epoch": 0.32,
"grad_norm": 0.0020294189453125,
"learning_rate": 5.6684210526315795e-06,
"loss": 0.0001,
"step": 6410
},
{
"epoch": 0.32,
"grad_norm": 0.004974365234375,
"learning_rate": 5.652631578947368e-06,
"loss": 0.0001,
"step": 6420
},
{
"epoch": 0.32,
"grad_norm": 0.0022735595703125,
"learning_rate": 5.636842105263158e-06,
"loss": 0.0002,
"step": 6430
},
{
"epoch": 0.32,
"grad_norm": 0.002349853515625,
"learning_rate": 5.621052631578947e-06,
"loss": 0.0001,
"step": 6440
},
{
"epoch": 0.32,
"grad_norm": 0.0037994384765625,
"learning_rate": 5.605263157894737e-06,
"loss": 0.0001,
"step": 6450
},
{
"epoch": 0.32,
"grad_norm": 0.004150390625,
"learning_rate": 5.589473684210527e-06,
"loss": 0.0001,
"step": 6460
},
{
"epoch": 0.32,
"grad_norm": 0.005218505859375,
"learning_rate": 5.573684210526316e-06,
"loss": 0.0001,
"step": 6470
},
{
"epoch": 0.32,
"grad_norm": 0.038330078125,
"learning_rate": 5.557894736842106e-06,
"loss": 0.0001,
"step": 6480
},
{
"epoch": 0.32,
"grad_norm": 0.003204345703125,
"learning_rate": 5.542105263157895e-06,
"loss": 0.0001,
"step": 6490
},
{
"epoch": 0.33,
"grad_norm": 0.0081787109375,
"learning_rate": 5.526315789473684e-06,
"loss": 0.0001,
"step": 6500
},
{
"epoch": 0.33,
"grad_norm": 0.00604248046875,
"learning_rate": 5.5105263157894735e-06,
"loss": 0.0001,
"step": 6510
},
{
"epoch": 0.33,
"grad_norm": 0.00186920166015625,
"learning_rate": 5.494736842105263e-06,
"loss": 0.0001,
"step": 6520
},
{
"epoch": 0.33,
"grad_norm": 0.003204345703125,
"learning_rate": 5.4789473684210525e-06,
"loss": 0.0001,
"step": 6530
},
{
"epoch": 0.33,
"grad_norm": 0.00885009765625,
"learning_rate": 5.463157894736842e-06,
"loss": 0.0001,
"step": 6540
},
{
"epoch": 0.33,
"grad_norm": 0.003326416015625,
"learning_rate": 5.447368421052632e-06,
"loss": 0.0001,
"step": 6550
},
{
"epoch": 0.33,
"grad_norm": 0.0034332275390625,
"learning_rate": 5.431578947368422e-06,
"loss": 0.0001,
"step": 6560
},
{
"epoch": 0.33,
"grad_norm": 0.003936767578125,
"learning_rate": 5.4157894736842105e-06,
"loss": 0.0001,
"step": 6570
},
{
"epoch": 0.33,
"grad_norm": 0.007354736328125,
"learning_rate": 5.4e-06,
"loss": 0.0001,
"step": 6580
},
{
"epoch": 0.33,
"grad_norm": 0.00396728515625,
"learning_rate": 5.3842105263157895e-06,
"loss": 0.0001,
"step": 6590
},
{
"epoch": 0.33,
"grad_norm": 0.00165557861328125,
"learning_rate": 5.368421052631579e-06,
"loss": 0.0001,
"step": 6600
},
{
"epoch": 0.33,
"grad_norm": 0.0021514892578125,
"learning_rate": 5.3526315789473684e-06,
"loss": 0.0001,
"step": 6610
},
{
"epoch": 0.33,
"grad_norm": 0.004730224609375,
"learning_rate": 5.336842105263158e-06,
"loss": 0.0001,
"step": 6620
},
{
"epoch": 0.33,
"grad_norm": 0.002197265625,
"learning_rate": 5.321052631578948e-06,
"loss": 0.0001,
"step": 6630
},
{
"epoch": 0.33,
"grad_norm": 0.00183868408203125,
"learning_rate": 5.305263157894737e-06,
"loss": 0.0001,
"step": 6640
},
{
"epoch": 0.33,
"grad_norm": 0.00885009765625,
"learning_rate": 5.289473684210526e-06,
"loss": 0.0001,
"step": 6650
},
{
"epoch": 0.33,
"grad_norm": 0.00188446044921875,
"learning_rate": 5.273684210526316e-06,
"loss": 0.0001,
"step": 6660
},
{
"epoch": 0.33,
"grad_norm": 0.004241943359375,
"learning_rate": 5.257894736842105e-06,
"loss": 0.0001,
"step": 6670
},
{
"epoch": 0.33,
"grad_norm": 0.001708984375,
"learning_rate": 5.242105263157895e-06,
"loss": 0.0,
"step": 6680
},
{
"epoch": 0.33,
"grad_norm": 0.0034332275390625,
"learning_rate": 5.226315789473684e-06,
"loss": 0.0001,
"step": 6690
},
{
"epoch": 0.34,
"grad_norm": 0.0029144287109375,
"learning_rate": 5.210526315789474e-06,
"loss": 0.0001,
"step": 6700
},
{
"epoch": 0.34,
"grad_norm": 0.0081787109375,
"learning_rate": 5.194736842105263e-06,
"loss": 0.0001,
"step": 6710
},
{
"epoch": 0.34,
"grad_norm": 0.0177001953125,
"learning_rate": 5.178947368421053e-06,
"loss": 0.0001,
"step": 6720
},
{
"epoch": 0.34,
"grad_norm": 0.0023956298828125,
"learning_rate": 5.163157894736842e-06,
"loss": 0.0001,
"step": 6730
},
{
"epoch": 0.34,
"grad_norm": 0.00323486328125,
"learning_rate": 5.147368421052632e-06,
"loss": 0.0001,
"step": 6740
},
{
"epoch": 0.34,
"grad_norm": 0.002777099609375,
"learning_rate": 5.131578947368421e-06,
"loss": 0.0001,
"step": 6750
},
{
"epoch": 0.34,
"grad_norm": 0.00396728515625,
"learning_rate": 5.115789473684211e-06,
"loss": 0.0001,
"step": 6760
},
{
"epoch": 0.34,
"grad_norm": 0.018310546875,
"learning_rate": 5.1e-06,
"loss": 0.0001,
"step": 6770
},
{
"epoch": 0.34,
"grad_norm": 0.0021514892578125,
"learning_rate": 5.084210526315789e-06,
"loss": 0.0001,
"step": 6780
},
{
"epoch": 0.34,
"grad_norm": 0.002777099609375,
"learning_rate": 5.0684210526315784e-06,
"loss": 0.0001,
"step": 6790
},
{
"epoch": 0.34,
"grad_norm": 0.005767822265625,
"learning_rate": 5.052631578947369e-06,
"loss": 0.0001,
"step": 6800
},
{
"epoch": 0.34,
"grad_norm": 0.0040283203125,
"learning_rate": 5.036842105263158e-06,
"loss": 0.0001,
"step": 6810
},
{
"epoch": 0.34,
"grad_norm": 0.00396728515625,
"learning_rate": 5.021052631578948e-06,
"loss": 0.0001,
"step": 6820
},
{
"epoch": 0.34,
"grad_norm": 0.0029296875,
"learning_rate": 5.005263157894737e-06,
"loss": 0.0001,
"step": 6830
},
{
"epoch": 0.34,
"grad_norm": 0.01385498046875,
"learning_rate": 4.989473684210527e-06,
"loss": 0.0002,
"step": 6840
},
{
"epoch": 0.34,
"grad_norm": 0.0017242431640625,
"learning_rate": 4.973684210526315e-06,
"loss": 0.0001,
"step": 6850
},
{
"epoch": 0.34,
"grad_norm": 0.00335693359375,
"learning_rate": 4.957894736842105e-06,
"loss": 0.0001,
"step": 6860
},
{
"epoch": 0.34,
"grad_norm": 0.0031280517578125,
"learning_rate": 4.942105263157894e-06,
"loss": 0.0001,
"step": 6870
},
{
"epoch": 0.34,
"grad_norm": 0.0031280517578125,
"learning_rate": 4.926315789473685e-06,
"loss": 0.0001,
"step": 6880
},
{
"epoch": 0.34,
"grad_norm": 0.008544921875,
"learning_rate": 4.910526315789474e-06,
"loss": 0.0001,
"step": 6890
},
{
"epoch": 0.34,
"grad_norm": 0.0098876953125,
"learning_rate": 4.894736842105264e-06,
"loss": 0.0001,
"step": 6900
},
{
"epoch": 0.35,
"grad_norm": 0.01031494140625,
"learning_rate": 4.878947368421053e-06,
"loss": 0.0001,
"step": 6910
},
{
"epoch": 0.35,
"grad_norm": 0.01123046875,
"learning_rate": 4.863157894736842e-06,
"loss": 0.0001,
"step": 6920
},
{
"epoch": 0.35,
"grad_norm": 0.0021514892578125,
"learning_rate": 4.847368421052631e-06,
"loss": 0.0001,
"step": 6930
},
{
"epoch": 0.35,
"grad_norm": 0.004364013671875,
"learning_rate": 4.831578947368421e-06,
"loss": 0.0001,
"step": 6940
},
{
"epoch": 0.35,
"grad_norm": 0.006195068359375,
"learning_rate": 4.81578947368421e-06,
"loss": 0.0001,
"step": 6950
},
{
"epoch": 0.35,
"grad_norm": 0.003509521484375,
"learning_rate": 4.800000000000001e-06,
"loss": 0.0001,
"step": 6960
},
{
"epoch": 0.35,
"grad_norm": 0.0036468505859375,
"learning_rate": 4.78421052631579e-06,
"loss": 0.0001,
"step": 6970
},
{
"epoch": 0.35,
"grad_norm": 0.0016326904296875,
"learning_rate": 4.76842105263158e-06,
"loss": 0.0001,
"step": 6980
},
{
"epoch": 0.35,
"grad_norm": 0.00173187255859375,
"learning_rate": 4.752631578947368e-06,
"loss": 0.0001,
"step": 6990
},
{
"epoch": 0.35,
"grad_norm": 0.00653076171875,
"learning_rate": 4.736842105263158e-06,
"loss": 0.0001,
"step": 7000
},
{
"epoch": 0.35,
"grad_norm": 0.0062255859375,
"learning_rate": 4.721052631578947e-06,
"loss": 0.0001,
"step": 7010
},
{
"epoch": 0.35,
"grad_norm": 0.00738525390625,
"learning_rate": 4.705263157894737e-06,
"loss": 0.0001,
"step": 7020
},
{
"epoch": 0.35,
"grad_norm": 0.00396728515625,
"learning_rate": 4.689473684210526e-06,
"loss": 0.0001,
"step": 7030
},
{
"epoch": 0.35,
"grad_norm": 0.0081787109375,
"learning_rate": 4.673684210526316e-06,
"loss": 0.0001,
"step": 7040
},
{
"epoch": 0.35,
"grad_norm": 0.00165557861328125,
"learning_rate": 4.657894736842106e-06,
"loss": 0.0001,
"step": 7050
},
{
"epoch": 0.35,
"grad_norm": 0.007049560546875,
"learning_rate": 4.642105263157895e-06,
"loss": 0.0001,
"step": 7060
},
{
"epoch": 0.35,
"grad_norm": 0.00384521484375,
"learning_rate": 4.626315789473684e-06,
"loss": 0.0001,
"step": 7070
},
{
"epoch": 0.35,
"grad_norm": 0.004302978515625,
"learning_rate": 4.610526315789474e-06,
"loss": 0.0001,
"step": 7080
},
{
"epoch": 0.35,
"grad_norm": 0.0034027099609375,
"learning_rate": 4.594736842105263e-06,
"loss": 0.0002,
"step": 7090
},
{
"epoch": 0.35,
"grad_norm": 0.00360107421875,
"learning_rate": 4.578947368421053e-06,
"loss": 0.0001,
"step": 7100
},
{
"epoch": 0.36,
"grad_norm": 0.0052490234375,
"learning_rate": 4.563157894736842e-06,
"loss": 0.0001,
"step": 7110
},
{
"epoch": 0.36,
"grad_norm": 0.0019989013671875,
"learning_rate": 4.547368421052632e-06,
"loss": 0.0001,
"step": 7120
},
{
"epoch": 0.36,
"grad_norm": 0.01263427734375,
"learning_rate": 4.531578947368421e-06,
"loss": 0.0001,
"step": 7130
},
{
"epoch": 0.36,
"grad_norm": 0.05908203125,
"learning_rate": 4.515789473684211e-06,
"loss": 0.0001,
"step": 7140
},
{
"epoch": 0.36,
"grad_norm": 0.00653076171875,
"learning_rate": 4.5e-06,
"loss": 0.0001,
"step": 7150
},
{
"epoch": 0.36,
"grad_norm": 0.0020294189453125,
"learning_rate": 4.48421052631579e-06,
"loss": 0.0001,
"step": 7160
},
{
"epoch": 0.36,
"grad_norm": 0.003387451171875,
"learning_rate": 4.468421052631579e-06,
"loss": 0.0001,
"step": 7170
},
{
"epoch": 0.36,
"grad_norm": 0.0052490234375,
"learning_rate": 4.452631578947369e-06,
"loss": 0.0001,
"step": 7180
},
{
"epoch": 0.36,
"grad_norm": 0.0023651123046875,
"learning_rate": 4.436842105263158e-06,
"loss": 0.0001,
"step": 7190
},
{
"epoch": 0.36,
"grad_norm": 0.00872802734375,
"learning_rate": 4.421052631578947e-06,
"loss": 0.0001,
"step": 7200
},
{
"epoch": 0.36,
"grad_norm": 0.002685546875,
"learning_rate": 4.405263157894737e-06,
"loss": 0.0002,
"step": 7210
},
{
"epoch": 0.36,
"grad_norm": 0.0042724609375,
"learning_rate": 4.3894736842105266e-06,
"loss": 0.0002,
"step": 7220
},
{
"epoch": 0.36,
"grad_norm": 0.00164794921875,
"learning_rate": 4.373684210526316e-06,
"loss": 0.0001,
"step": 7230
},
{
"epoch": 0.36,
"grad_norm": 0.00201416015625,
"learning_rate": 4.3578947368421055e-06,
"loss": 0.0001,
"step": 7240
},
{
"epoch": 0.36,
"grad_norm": 0.0017242431640625,
"learning_rate": 4.342105263157895e-06,
"loss": 0.0001,
"step": 7250
},
{
"epoch": 0.36,
"grad_norm": 0.0057373046875,
"learning_rate": 4.3263157894736845e-06,
"loss": 0.0002,
"step": 7260
},
{
"epoch": 0.36,
"grad_norm": 0.002166748046875,
"learning_rate": 4.310526315789474e-06,
"loss": 0.0001,
"step": 7270
},
{
"epoch": 0.36,
"grad_norm": 0.0030975341796875,
"learning_rate": 4.294736842105263e-06,
"loss": 0.0,
"step": 7280
},
{
"epoch": 0.36,
"grad_norm": 0.001922607421875,
"learning_rate": 4.278947368421052e-06,
"loss": 0.0001,
"step": 7290
},
{
"epoch": 0.36,
"grad_norm": 0.004547119140625,
"learning_rate": 4.2631578947368425e-06,
"loss": 0.0001,
"step": 7300
},
{
"epoch": 0.37,
"grad_norm": 0.00372314453125,
"learning_rate": 4.247368421052632e-06,
"loss": 0.0001,
"step": 7310
},
{
"epoch": 0.37,
"grad_norm": 0.0029754638671875,
"learning_rate": 4.2315789473684215e-06,
"loss": 0.0001,
"step": 7320
},
{
"epoch": 0.37,
"grad_norm": 0.00445556640625,
"learning_rate": 4.215789473684211e-06,
"loss": 0.0002,
"step": 7330
},
{
"epoch": 0.37,
"grad_norm": 0.003265380859375,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.0,
"step": 7340
},
{
"epoch": 0.37,
"grad_norm": 0.01123046875,
"learning_rate": 4.184210526315789e-06,
"loss": 0.0001,
"step": 7350
},
{
"epoch": 0.37,
"grad_norm": 0.0074462890625,
"learning_rate": 4.168421052631579e-06,
"loss": 0.0001,
"step": 7360
},
{
"epoch": 0.37,
"grad_norm": 0.00311279296875,
"learning_rate": 4.152631578947368e-06,
"loss": 0.0001,
"step": 7370
},
{
"epoch": 0.37,
"grad_norm": 0.002349853515625,
"learning_rate": 4.136842105263158e-06,
"loss": 0.0001,
"step": 7380
},
{
"epoch": 0.37,
"grad_norm": 0.003387451171875,
"learning_rate": 4.121052631578948e-06,
"loss": 0.0001,
"step": 7390
},
{
"epoch": 0.37,
"grad_norm": 0.00189208984375,
"learning_rate": 4.105263157894737e-06,
"loss": 0.0001,
"step": 7400
},
{
"epoch": 0.37,
"grad_norm": 0.0022125244140625,
"learning_rate": 4.089473684210527e-06,
"loss": 0.0001,
"step": 7410
},
{
"epoch": 0.37,
"grad_norm": 0.035888671875,
"learning_rate": 4.0736842105263155e-06,
"loss": 0.0001,
"step": 7420
},
{
"epoch": 0.37,
"grad_norm": 0.00177001953125,
"learning_rate": 4.057894736842105e-06,
"loss": 0.0001,
"step": 7430
},
{
"epoch": 0.37,
"grad_norm": 0.003692626953125,
"learning_rate": 4.0421052631578945e-06,
"loss": 0.0001,
"step": 7440
},
{
"epoch": 0.37,
"grad_norm": 0.0030975341796875,
"learning_rate": 4.026315789473684e-06,
"loss": 0.0001,
"step": 7450
},
{
"epoch": 0.37,
"grad_norm": 0.0020599365234375,
"learning_rate": 4.010526315789474e-06,
"loss": 0.0001,
"step": 7460
},
{
"epoch": 0.37,
"grad_norm": 0.002685546875,
"learning_rate": 3.994736842105264e-06,
"loss": 0.0001,
"step": 7470
},
{
"epoch": 0.37,
"grad_norm": 0.005279541015625,
"learning_rate": 3.978947368421053e-06,
"loss": 0.0001,
"step": 7480
},
{
"epoch": 0.37,
"grad_norm": 0.0019989013671875,
"learning_rate": 3.963157894736842e-06,
"loss": 0.0001,
"step": 7490
},
{
"epoch": 0.38,
"grad_norm": 0.0087890625,
"learning_rate": 3.9473684210526315e-06,
"loss": 0.0001,
"step": 7500
},
{
"epoch": 0.38,
"grad_norm": 0.006927490234375,
"learning_rate": 3.931578947368421e-06,
"loss": 0.0001,
"step": 7510
},
{
"epoch": 0.38,
"grad_norm": 0.00191497802734375,
"learning_rate": 3.9157894736842104e-06,
"loss": 0.0001,
"step": 7520
},
{
"epoch": 0.38,
"grad_norm": 0.0111083984375,
"learning_rate": 3.9e-06,
"loss": 0.0001,
"step": 7530
},
{
"epoch": 0.38,
"grad_norm": 0.00787353515625,
"learning_rate": 3.884210526315789e-06,
"loss": 0.0001,
"step": 7540
},
{
"epoch": 0.38,
"grad_norm": 0.0034942626953125,
"learning_rate": 3.86842105263158e-06,
"loss": 0.0001,
"step": 7550
},
{
"epoch": 0.38,
"grad_norm": 0.007232666015625,
"learning_rate": 3.852631578947368e-06,
"loss": 0.0001,
"step": 7560
},
{
"epoch": 0.38,
"grad_norm": 0.0021514892578125,
"learning_rate": 3.836842105263158e-06,
"loss": 0.0001,
"step": 7570
},
{
"epoch": 0.38,
"grad_norm": 0.00457763671875,
"learning_rate": 3.821052631578947e-06,
"loss": 0.0001,
"step": 7580
},
{
"epoch": 0.38,
"grad_norm": 0.0028533935546875,
"learning_rate": 3.805263157894737e-06,
"loss": 0.0001,
"step": 7590
},
{
"epoch": 0.38,
"grad_norm": 0.0030670166015625,
"learning_rate": 3.7894736842105264e-06,
"loss": 0.0001,
"step": 7600
},
{
"epoch": 0.38,
"grad_norm": 0.003753662109375,
"learning_rate": 3.7736842105263163e-06,
"loss": 0.0001,
"step": 7610
},
{
"epoch": 0.38,
"grad_norm": 0.005035400390625,
"learning_rate": 3.7578947368421058e-06,
"loss": 0.0001,
"step": 7620
},
{
"epoch": 0.38,
"grad_norm": 0.0084228515625,
"learning_rate": 3.742105263157895e-06,
"loss": 0.0001,
"step": 7630
},
{
"epoch": 0.38,
"grad_norm": 0.0020294189453125,
"learning_rate": 3.7263157894736843e-06,
"loss": 0.0001,
"step": 7640
},
{
"epoch": 0.38,
"grad_norm": 0.0029449462890625,
"learning_rate": 3.710526315789474e-06,
"loss": 0.0001,
"step": 7650
},
{
"epoch": 0.38,
"grad_norm": 0.00250244140625,
"learning_rate": 3.6947368421052633e-06,
"loss": 0.0001,
"step": 7660
},
{
"epoch": 0.38,
"grad_norm": 0.00640869140625,
"learning_rate": 3.678947368421053e-06,
"loss": 0.0001,
"step": 7670
},
{
"epoch": 0.38,
"grad_norm": 0.0034942626953125,
"learning_rate": 3.663157894736842e-06,
"loss": 0.0001,
"step": 7680
},
{
"epoch": 0.38,
"grad_norm": 0.004364013671875,
"learning_rate": 3.6473684210526318e-06,
"loss": 0.0001,
"step": 7690
},
{
"epoch": 0.39,
"grad_norm": 0.0018768310546875,
"learning_rate": 3.6315789473684213e-06,
"loss": 0.0001,
"step": 7700
},
{
"epoch": 0.39,
"grad_norm": 0.00201416015625,
"learning_rate": 3.6157894736842108e-06,
"loss": 0.0001,
"step": 7710
},
{
"epoch": 0.39,
"grad_norm": 0.001617431640625,
"learning_rate": 3.6e-06,
"loss": 0.0001,
"step": 7720
},
{
"epoch": 0.39,
"grad_norm": 0.00439453125,
"learning_rate": 3.5842105263157898e-06,
"loss": 0.0001,
"step": 7730
},
{
"epoch": 0.39,
"grad_norm": 0.0025634765625,
"learning_rate": 3.5684210526315792e-06,
"loss": 0.0001,
"step": 7740
},
{
"epoch": 0.39,
"grad_norm": 0.004730224609375,
"learning_rate": 3.5526315789473683e-06,
"loss": 0.0001,
"step": 7750
},
{
"epoch": 0.39,
"grad_norm": 0.0038909912109375,
"learning_rate": 3.536842105263158e-06,
"loss": 0.0001,
"step": 7760
},
{
"epoch": 0.39,
"grad_norm": 0.00238037109375,
"learning_rate": 3.5210526315789473e-06,
"loss": 0.0001,
"step": 7770
},
{
"epoch": 0.39,
"grad_norm": 0.00872802734375,
"learning_rate": 3.505263157894737e-06,
"loss": 0.0001,
"step": 7780
},
{
"epoch": 0.39,
"grad_norm": 0.0033111572265625,
"learning_rate": 3.4894736842105263e-06,
"loss": 0.0001,
"step": 7790
},
{
"epoch": 0.39,
"grad_norm": 0.0021209716796875,
"learning_rate": 3.4736842105263158e-06,
"loss": 0.0,
"step": 7800
},
{
"epoch": 0.39,
"grad_norm": 0.00787353515625,
"learning_rate": 3.4578947368421053e-06,
"loss": 0.0001,
"step": 7810
},
{
"epoch": 0.39,
"grad_norm": 0.005523681640625,
"learning_rate": 3.4421052631578947e-06,
"loss": 0.0001,
"step": 7820
},
{
"epoch": 0.39,
"grad_norm": 0.00347900390625,
"learning_rate": 3.4263157894736842e-06,
"loss": 0.0001,
"step": 7830
},
{
"epoch": 0.39,
"grad_norm": 0.00335693359375,
"learning_rate": 3.4105263157894737e-06,
"loss": 0.0001,
"step": 7840
},
{
"epoch": 0.39,
"grad_norm": 0.0024871826171875,
"learning_rate": 3.3947368421052632e-06,
"loss": 0.0001,
"step": 7850
},
{
"epoch": 0.39,
"grad_norm": 0.0035400390625,
"learning_rate": 3.3789473684210527e-06,
"loss": 0.0001,
"step": 7860
},
{
"epoch": 0.39,
"grad_norm": 0.006378173828125,
"learning_rate": 3.363157894736842e-06,
"loss": 0.0001,
"step": 7870
},
{
"epoch": 0.39,
"grad_norm": 0.0032501220703125,
"learning_rate": 3.3473684210526317e-06,
"loss": 0.0001,
"step": 7880
},
{
"epoch": 0.39,
"grad_norm": 0.0016326904296875,
"learning_rate": 3.331578947368421e-06,
"loss": 0.0001,
"step": 7890
},
{
"epoch": 0.4,
"grad_norm": 0.0020294189453125,
"learning_rate": 3.3157894736842107e-06,
"loss": 0.0001,
"step": 7900
},
{
"epoch": 0.4,
"grad_norm": 0.001617431640625,
"learning_rate": 3.3e-06,
"loss": 0.0,
"step": 7910
},
{
"epoch": 0.4,
"grad_norm": 0.003021240234375,
"learning_rate": 3.2842105263157897e-06,
"loss": 0.0001,
"step": 7920
},
{
"epoch": 0.4,
"grad_norm": 0.0086669921875,
"learning_rate": 3.2684210526315787e-06,
"loss": 0.0001,
"step": 7930
},
{
"epoch": 0.4,
"grad_norm": 0.00167083740234375,
"learning_rate": 3.2526315789473686e-06,
"loss": 0.0001,
"step": 7940
},
{
"epoch": 0.4,
"grad_norm": 0.00482177734375,
"learning_rate": 3.236842105263158e-06,
"loss": 0.0001,
"step": 7950
},
{
"epoch": 0.4,
"grad_norm": 0.006591796875,
"learning_rate": 3.2210526315789476e-06,
"loss": 0.0001,
"step": 7960
},
{
"epoch": 0.4,
"grad_norm": 0.00347900390625,
"learning_rate": 3.2052631578947367e-06,
"loss": 0.0001,
"step": 7970
},
{
"epoch": 0.4,
"grad_norm": 0.003448486328125,
"learning_rate": 3.1894736842105266e-06,
"loss": 0.0001,
"step": 7980
},
{
"epoch": 0.4,
"grad_norm": 0.00164794921875,
"learning_rate": 3.173684210526316e-06,
"loss": 0.0001,
"step": 7990
},
{
"epoch": 0.4,
"grad_norm": 0.0022430419921875,
"learning_rate": 3.157894736842105e-06,
"loss": 0.0001,
"step": 8000
},
{
"epoch": 0.4,
"eval_loss": 7.415514846798033e-05,
"eval_runtime": 30.241,
"eval_samples_per_second": 33.068,
"eval_steps_per_second": 33.068,
"step": 8000
}
],
"logging_steps": 10,
"max_steps": 10000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2000,
"total_flos": 6.455688167424e+16,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}