{ "best_metric": 1.3379485607147217, "best_model_checkpoint": "runs/deepseek_20240422-210351/checkpoint-15000", "epoch": 0.75, "eval_steps": 5000, "global_step": 30000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "grad_norm": 21.625, "learning_rate": 4.0000000000000003e-07, "loss": 2.2157, "step": 10 }, { "epoch": 0.0, "grad_norm": 41.0, "learning_rate": 8.000000000000001e-07, "loss": 2.1199, "step": 20 }, { "epoch": 0.0, "grad_norm": 30.875, "learning_rate": 1.2000000000000002e-06, "loss": 2.2071, "step": 30 }, { "epoch": 0.0, "grad_norm": 51.0, "learning_rate": 1.6000000000000001e-06, "loss": 2.2955, "step": 40 }, { "epoch": 0.0, "grad_norm": 61.25, "learning_rate": 2.0000000000000003e-06, "loss": 2.1046, "step": 50 }, { "epoch": 0.0, "grad_norm": 32.25, "learning_rate": 2.4000000000000003e-06, "loss": 2.162, "step": 60 }, { "epoch": 0.0, "grad_norm": 28.5, "learning_rate": 2.8000000000000003e-06, "loss": 2.2084, "step": 70 }, { "epoch": 0.0, "grad_norm": 24.0, "learning_rate": 3.2000000000000003e-06, "loss": 2.1635, "step": 80 }, { "epoch": 0.0, "grad_norm": 31.5, "learning_rate": 3.6000000000000003e-06, "loss": 2.0963, "step": 90 }, { "epoch": 0.0, "grad_norm": 30.0, "learning_rate": 4.000000000000001e-06, "loss": 1.9403, "step": 100 }, { "epoch": 0.0, "grad_norm": 44.5, "learning_rate": 4.4e-06, "loss": 1.8556, "step": 110 }, { "epoch": 0.0, "grad_norm": 22.375, "learning_rate": 4.800000000000001e-06, "loss": 1.9425, "step": 120 }, { "epoch": 0.0, "grad_norm": 44.0, "learning_rate": 5.2e-06, "loss": 1.8286, "step": 130 }, { "epoch": 0.0, "grad_norm": 21.75, "learning_rate": 5.600000000000001e-06, "loss": 2.0117, "step": 140 }, { "epoch": 0.0, "grad_norm": 51.75, "learning_rate": 6e-06, "loss": 1.8744, "step": 150 }, { "epoch": 0.0, "grad_norm": 23.0, "learning_rate": 6.4000000000000006e-06, "loss": 1.8316, "step": 160 }, { "epoch": 0.0, "grad_norm": 25.25, "learning_rate": 6.800000000000001e-06, "loss": 1.63, "step": 170 }, { "epoch": 0.0, "grad_norm": 44.5, "learning_rate": 7.2000000000000005e-06, "loss": 1.7395, "step": 180 }, { "epoch": 0.0, "grad_norm": 20.375, "learning_rate": 7.600000000000001e-06, "loss": 1.8581, "step": 190 }, { "epoch": 0.01, "grad_norm": 44.75, "learning_rate": 8.000000000000001e-06, "loss": 1.7136, "step": 200 }, { "epoch": 0.01, "grad_norm": 59.75, "learning_rate": 8.400000000000001e-06, "loss": 1.5627, "step": 210 }, { "epoch": 0.01, "grad_norm": 33.25, "learning_rate": 8.8e-06, "loss": 1.6858, "step": 220 }, { "epoch": 0.01, "grad_norm": 22.0, "learning_rate": 9.200000000000002e-06, "loss": 1.9664, "step": 230 }, { "epoch": 0.01, "grad_norm": 18.625, "learning_rate": 9.600000000000001e-06, "loss": 1.5142, "step": 240 }, { "epoch": 0.01, "grad_norm": 15.5625, "learning_rate": 1e-05, "loss": 1.7143, "step": 250 }, { "epoch": 0.01, "grad_norm": 13.5625, "learning_rate": 1.04e-05, "loss": 1.8581, "step": 260 }, { "epoch": 0.01, "grad_norm": 21.125, "learning_rate": 1.0800000000000002e-05, "loss": 1.625, "step": 270 }, { "epoch": 0.01, "grad_norm": 15.75, "learning_rate": 1.1200000000000001e-05, "loss": 1.6071, "step": 280 }, { "epoch": 0.01, "grad_norm": 14.5, "learning_rate": 1.16e-05, "loss": 1.8191, "step": 290 }, { "epoch": 0.01, "grad_norm": 22.25, "learning_rate": 1.2e-05, "loss": 1.5947, "step": 300 }, { "epoch": 0.01, "grad_norm": 22.625, "learning_rate": 1.2400000000000002e-05, "loss": 1.6599, "step": 310 }, { "epoch": 0.01, "grad_norm": 88.5, "learning_rate": 1.2800000000000001e-05, "loss": 1.6836, "step": 320 }, { "epoch": 0.01, "grad_norm": 12.5625, "learning_rate": 1.3200000000000002e-05, "loss": 1.8363, "step": 330 }, { "epoch": 0.01, "grad_norm": 15.625, "learning_rate": 1.3600000000000002e-05, "loss": 1.723, "step": 340 }, { "epoch": 0.01, "grad_norm": 18.5, "learning_rate": 1.4e-05, "loss": 1.6806, "step": 350 }, { "epoch": 0.01, "grad_norm": 10.875, "learning_rate": 1.4400000000000001e-05, "loss": 1.618, "step": 360 }, { "epoch": 0.01, "grad_norm": 11.5, "learning_rate": 1.48e-05, "loss": 1.6253, "step": 370 }, { "epoch": 0.01, "grad_norm": 23.375, "learning_rate": 1.5200000000000002e-05, "loss": 1.7142, "step": 380 }, { "epoch": 0.01, "grad_norm": 9.9375, "learning_rate": 1.5600000000000003e-05, "loss": 1.6268, "step": 390 }, { "epoch": 0.01, "grad_norm": 15.375, "learning_rate": 1.6000000000000003e-05, "loss": 1.652, "step": 400 }, { "epoch": 0.01, "grad_norm": 29.375, "learning_rate": 1.64e-05, "loss": 1.6204, "step": 410 }, { "epoch": 0.01, "grad_norm": 33.75, "learning_rate": 1.6800000000000002e-05, "loss": 1.5077, "step": 420 }, { "epoch": 0.01, "grad_norm": 49.0, "learning_rate": 1.72e-05, "loss": 1.7802, "step": 430 }, { "epoch": 0.01, "grad_norm": 13.8125, "learning_rate": 1.76e-05, "loss": 1.4695, "step": 440 }, { "epoch": 0.01, "grad_norm": 25.75, "learning_rate": 1.8e-05, "loss": 1.6256, "step": 450 }, { "epoch": 0.01, "grad_norm": 16.25, "learning_rate": 1.8400000000000003e-05, "loss": 1.6964, "step": 460 }, { "epoch": 0.01, "grad_norm": 20.375, "learning_rate": 1.88e-05, "loss": 1.7083, "step": 470 }, { "epoch": 0.01, "grad_norm": 22.625, "learning_rate": 1.9200000000000003e-05, "loss": 1.6634, "step": 480 }, { "epoch": 0.01, "grad_norm": 14.4375, "learning_rate": 1.9600000000000002e-05, "loss": 1.5189, "step": 490 }, { "epoch": 0.01, "grad_norm": 49.25, "learning_rate": 2e-05, "loss": 1.6792, "step": 500 }, { "epoch": 0.01, "grad_norm": 28.125, "learning_rate": 1.999322033898305e-05, "loss": 1.6381, "step": 510 }, { "epoch": 0.01, "grad_norm": 13.4375, "learning_rate": 1.9986440677966104e-05, "loss": 1.5793, "step": 520 }, { "epoch": 0.01, "grad_norm": 20.25, "learning_rate": 1.9979661016949154e-05, "loss": 1.5979, "step": 530 }, { "epoch": 0.01, "grad_norm": 10.875, "learning_rate": 1.9972881355932204e-05, "loss": 1.7109, "step": 540 }, { "epoch": 0.01, "grad_norm": 19.75, "learning_rate": 1.9966101694915257e-05, "loss": 1.5867, "step": 550 }, { "epoch": 0.01, "grad_norm": 25.375, "learning_rate": 1.9959322033898307e-05, "loss": 1.6398, "step": 560 }, { "epoch": 0.01, "grad_norm": 19.5, "learning_rate": 1.995254237288136e-05, "loss": 1.4327, "step": 570 }, { "epoch": 0.01, "grad_norm": 11.0625, "learning_rate": 1.994576271186441e-05, "loss": 1.5478, "step": 580 }, { "epoch": 0.01, "grad_norm": 23.625, "learning_rate": 1.993898305084746e-05, "loss": 1.8115, "step": 590 }, { "epoch": 0.01, "grad_norm": 11.625, "learning_rate": 1.9932203389830512e-05, "loss": 1.6309, "step": 600 }, { "epoch": 0.02, "grad_norm": 10.4375, "learning_rate": 1.992542372881356e-05, "loss": 1.5757, "step": 610 }, { "epoch": 0.02, "grad_norm": 11.375, "learning_rate": 1.991864406779661e-05, "loss": 1.6906, "step": 620 }, { "epoch": 0.02, "grad_norm": 12.625, "learning_rate": 1.991186440677966e-05, "loss": 1.6413, "step": 630 }, { "epoch": 0.02, "grad_norm": 17.125, "learning_rate": 1.990508474576271e-05, "loss": 1.6699, "step": 640 }, { "epoch": 0.02, "grad_norm": 16.875, "learning_rate": 1.9898305084745764e-05, "loss": 1.5938, "step": 650 }, { "epoch": 0.02, "grad_norm": 12.75, "learning_rate": 1.9891525423728814e-05, "loss": 1.8807, "step": 660 }, { "epoch": 0.02, "grad_norm": 13.5, "learning_rate": 1.9884745762711867e-05, "loss": 1.5723, "step": 670 }, { "epoch": 0.02, "grad_norm": 10.5, "learning_rate": 1.9877966101694917e-05, "loss": 1.4536, "step": 680 }, { "epoch": 0.02, "grad_norm": 17.0, "learning_rate": 1.9871186440677966e-05, "loss": 1.6836, "step": 690 }, { "epoch": 0.02, "grad_norm": 18.125, "learning_rate": 1.986440677966102e-05, "loss": 1.5296, "step": 700 }, { "epoch": 0.02, "grad_norm": 15.375, "learning_rate": 1.985762711864407e-05, "loss": 1.6754, "step": 710 }, { "epoch": 0.02, "grad_norm": 17.625, "learning_rate": 1.985084745762712e-05, "loss": 1.5548, "step": 720 }, { "epoch": 0.02, "grad_norm": 8.875, "learning_rate": 1.9844067796610172e-05, "loss": 1.6737, "step": 730 }, { "epoch": 0.02, "grad_norm": 13.0, "learning_rate": 1.9837288135593222e-05, "loss": 1.5516, "step": 740 }, { "epoch": 0.02, "grad_norm": 13.875, "learning_rate": 1.9830508474576275e-05, "loss": 1.6862, "step": 750 }, { "epoch": 0.02, "grad_norm": 8.8125, "learning_rate": 1.9823728813559324e-05, "loss": 1.6523, "step": 760 }, { "epoch": 0.02, "grad_norm": 10.5625, "learning_rate": 1.9816949152542374e-05, "loss": 1.6752, "step": 770 }, { "epoch": 0.02, "grad_norm": 11.0625, "learning_rate": 1.9810169491525427e-05, "loss": 1.565, "step": 780 }, { "epoch": 0.02, "grad_norm": 11.375, "learning_rate": 1.9803389830508477e-05, "loss": 1.447, "step": 790 }, { "epoch": 0.02, "grad_norm": 16.875, "learning_rate": 1.9796610169491527e-05, "loss": 1.618, "step": 800 }, { "epoch": 0.02, "grad_norm": 12.375, "learning_rate": 1.978983050847458e-05, "loss": 1.5672, "step": 810 }, { "epoch": 0.02, "grad_norm": 8.3125, "learning_rate": 1.9783050847457626e-05, "loss": 1.5638, "step": 820 }, { "epoch": 0.02, "grad_norm": 10.0625, "learning_rate": 1.977627118644068e-05, "loss": 1.6477, "step": 830 }, { "epoch": 0.02, "grad_norm": 27.5, "learning_rate": 1.976949152542373e-05, "loss": 1.5532, "step": 840 }, { "epoch": 0.02, "grad_norm": 17.625, "learning_rate": 1.9762711864406782e-05, "loss": 1.6517, "step": 850 }, { "epoch": 0.02, "grad_norm": 9.75, "learning_rate": 1.9755932203389832e-05, "loss": 1.663, "step": 860 }, { "epoch": 0.02, "grad_norm": 8.3125, "learning_rate": 1.974915254237288e-05, "loss": 1.6119, "step": 870 }, { "epoch": 0.02, "grad_norm": 19.625, "learning_rate": 1.9742372881355935e-05, "loss": 1.5863, "step": 880 }, { "epoch": 0.02, "grad_norm": 9.875, "learning_rate": 1.9735593220338984e-05, "loss": 1.6251, "step": 890 }, { "epoch": 0.02, "grad_norm": 13.875, "learning_rate": 1.9728813559322034e-05, "loss": 1.6313, "step": 900 }, { "epoch": 0.02, "grad_norm": 9.125, "learning_rate": 1.9722033898305087e-05, "loss": 1.5873, "step": 910 }, { "epoch": 0.02, "grad_norm": 25.375, "learning_rate": 1.9715254237288137e-05, "loss": 1.6079, "step": 920 }, { "epoch": 0.02, "grad_norm": 24.25, "learning_rate": 1.970847457627119e-05, "loss": 1.4977, "step": 930 }, { "epoch": 0.02, "grad_norm": 9.3125, "learning_rate": 1.970169491525424e-05, "loss": 1.6284, "step": 940 }, { "epoch": 0.02, "grad_norm": 10.375, "learning_rate": 1.969491525423729e-05, "loss": 1.6224, "step": 950 }, { "epoch": 0.02, "grad_norm": 12.6875, "learning_rate": 1.9688135593220342e-05, "loss": 1.5419, "step": 960 }, { "epoch": 0.02, "grad_norm": 15.0, "learning_rate": 1.9681355932203392e-05, "loss": 1.5991, "step": 970 }, { "epoch": 0.02, "grad_norm": 9.0625, "learning_rate": 1.9674576271186442e-05, "loss": 1.583, "step": 980 }, { "epoch": 0.02, "grad_norm": 16.625, "learning_rate": 1.9667796610169495e-05, "loss": 1.5774, "step": 990 }, { "epoch": 0.03, "grad_norm": 14.625, "learning_rate": 1.9661016949152545e-05, "loss": 1.7158, "step": 1000 }, { "epoch": 0.03, "grad_norm": 15.0, "learning_rate": 1.9654237288135594e-05, "loss": 1.6369, "step": 1010 }, { "epoch": 0.03, "grad_norm": 9.0625, "learning_rate": 1.9647457627118644e-05, "loss": 1.6978, "step": 1020 }, { "epoch": 0.03, "grad_norm": 9.3125, "learning_rate": 1.9640677966101697e-05, "loss": 1.6175, "step": 1030 }, { "epoch": 0.03, "grad_norm": 11.1875, "learning_rate": 1.9633898305084747e-05, "loss": 1.6845, "step": 1040 }, { "epoch": 0.03, "grad_norm": 14.1875, "learning_rate": 1.9627118644067796e-05, "loss": 1.5605, "step": 1050 }, { "epoch": 0.03, "grad_norm": 11.1875, "learning_rate": 1.962033898305085e-05, "loss": 1.734, "step": 1060 }, { "epoch": 0.03, "grad_norm": 11.8125, "learning_rate": 1.96135593220339e-05, "loss": 1.4532, "step": 1070 }, { "epoch": 0.03, "grad_norm": 42.0, "learning_rate": 1.960677966101695e-05, "loss": 1.7891, "step": 1080 }, { "epoch": 0.03, "grad_norm": 18.75, "learning_rate": 1.9600000000000002e-05, "loss": 1.6711, "step": 1090 }, { "epoch": 0.03, "grad_norm": 13.4375, "learning_rate": 1.9593220338983052e-05, "loss": 1.7482, "step": 1100 }, { "epoch": 0.03, "grad_norm": 16.25, "learning_rate": 1.95864406779661e-05, "loss": 1.7121, "step": 1110 }, { "epoch": 0.03, "grad_norm": 12.5, "learning_rate": 1.9579661016949155e-05, "loss": 1.5464, "step": 1120 }, { "epoch": 0.03, "grad_norm": 18.375, "learning_rate": 1.9572881355932204e-05, "loss": 1.5346, "step": 1130 }, { "epoch": 0.03, "grad_norm": 13.4375, "learning_rate": 1.9566101694915257e-05, "loss": 1.5374, "step": 1140 }, { "epoch": 0.03, "grad_norm": 12.875, "learning_rate": 1.9559322033898307e-05, "loss": 1.6641, "step": 1150 }, { "epoch": 0.03, "grad_norm": 10.25, "learning_rate": 1.9552542372881357e-05, "loss": 1.5511, "step": 1160 }, { "epoch": 0.03, "grad_norm": 8.75, "learning_rate": 1.954576271186441e-05, "loss": 1.5672, "step": 1170 }, { "epoch": 0.03, "grad_norm": 29.125, "learning_rate": 1.953898305084746e-05, "loss": 1.4583, "step": 1180 }, { "epoch": 0.03, "grad_norm": 16.5, "learning_rate": 1.953220338983051e-05, "loss": 1.6112, "step": 1190 }, { "epoch": 0.03, "grad_norm": 11.0, "learning_rate": 1.9525423728813562e-05, "loss": 1.6229, "step": 1200 }, { "epoch": 0.03, "grad_norm": 6.6875, "learning_rate": 1.9518644067796612e-05, "loss": 1.5782, "step": 1210 }, { "epoch": 0.03, "grad_norm": 12.9375, "learning_rate": 1.9511864406779665e-05, "loss": 1.5294, "step": 1220 }, { "epoch": 0.03, "grad_norm": 6.96875, "learning_rate": 1.950508474576271e-05, "loss": 1.5371, "step": 1230 }, { "epoch": 0.03, "grad_norm": 18.5, "learning_rate": 1.9498305084745765e-05, "loss": 1.7215, "step": 1240 }, { "epoch": 0.03, "grad_norm": 17.75, "learning_rate": 1.9491525423728814e-05, "loss": 1.7009, "step": 1250 }, { "epoch": 0.03, "grad_norm": 28.25, "learning_rate": 1.9484745762711864e-05, "loss": 1.4353, "step": 1260 }, { "epoch": 0.03, "grad_norm": 11.0625, "learning_rate": 1.9477966101694917e-05, "loss": 1.5369, "step": 1270 }, { "epoch": 0.03, "grad_norm": 13.25, "learning_rate": 1.9471186440677967e-05, "loss": 1.3927, "step": 1280 }, { "epoch": 0.03, "grad_norm": 25.375, "learning_rate": 1.9464406779661017e-05, "loss": 1.7541, "step": 1290 }, { "epoch": 0.03, "grad_norm": 15.3125, "learning_rate": 1.945762711864407e-05, "loss": 1.5196, "step": 1300 }, { "epoch": 0.03, "grad_norm": 10.125, "learning_rate": 1.945084745762712e-05, "loss": 1.545, "step": 1310 }, { "epoch": 0.03, "grad_norm": 10.3125, "learning_rate": 1.9444067796610172e-05, "loss": 1.547, "step": 1320 }, { "epoch": 0.03, "grad_norm": 11.4375, "learning_rate": 1.9437288135593222e-05, "loss": 1.4786, "step": 1330 }, { "epoch": 0.03, "grad_norm": 6.5, "learning_rate": 1.9430508474576272e-05, "loss": 1.6925, "step": 1340 }, { "epoch": 0.03, "grad_norm": 10.4375, "learning_rate": 1.9423728813559325e-05, "loss": 1.3977, "step": 1350 }, { "epoch": 0.03, "grad_norm": 11.0625, "learning_rate": 1.9416949152542375e-05, "loss": 1.5348, "step": 1360 }, { "epoch": 0.03, "grad_norm": 10.875, "learning_rate": 1.9410169491525424e-05, "loss": 1.4933, "step": 1370 }, { "epoch": 0.03, "grad_norm": 10.1875, "learning_rate": 1.9403389830508477e-05, "loss": 1.5145, "step": 1380 }, { "epoch": 0.03, "grad_norm": 18.0, "learning_rate": 1.9396610169491527e-05, "loss": 1.3786, "step": 1390 }, { "epoch": 0.04, "grad_norm": 28.25, "learning_rate": 1.938983050847458e-05, "loss": 1.7004, "step": 1400 }, { "epoch": 0.04, "grad_norm": 24.5, "learning_rate": 1.938305084745763e-05, "loss": 1.5797, "step": 1410 }, { "epoch": 0.04, "grad_norm": 15.5, "learning_rate": 1.937627118644068e-05, "loss": 1.7938, "step": 1420 }, { "epoch": 0.04, "grad_norm": 10.875, "learning_rate": 1.9369491525423733e-05, "loss": 1.4801, "step": 1430 }, { "epoch": 0.04, "grad_norm": 20.75, "learning_rate": 1.936271186440678e-05, "loss": 1.5033, "step": 1440 }, { "epoch": 0.04, "grad_norm": 16.125, "learning_rate": 1.9355932203389832e-05, "loss": 1.4213, "step": 1450 }, { "epoch": 0.04, "grad_norm": 20.25, "learning_rate": 1.9349152542372882e-05, "loss": 1.5956, "step": 1460 }, { "epoch": 0.04, "grad_norm": 13.5625, "learning_rate": 1.934237288135593e-05, "loss": 1.7034, "step": 1470 }, { "epoch": 0.04, "grad_norm": 12.5, "learning_rate": 1.9335593220338985e-05, "loss": 1.5511, "step": 1480 }, { "epoch": 0.04, "grad_norm": 10.5, "learning_rate": 1.9328813559322034e-05, "loss": 1.4598, "step": 1490 }, { "epoch": 0.04, "grad_norm": 10.75, "learning_rate": 1.9322033898305087e-05, "loss": 1.5557, "step": 1500 }, { "epoch": 0.04, "grad_norm": 9.6875, "learning_rate": 1.9315254237288137e-05, "loss": 1.4942, "step": 1510 }, { "epoch": 0.04, "grad_norm": 35.5, "learning_rate": 1.9308474576271187e-05, "loss": 1.6278, "step": 1520 }, { "epoch": 0.04, "grad_norm": 7.96875, "learning_rate": 1.930169491525424e-05, "loss": 1.5048, "step": 1530 }, { "epoch": 0.04, "grad_norm": 11.25, "learning_rate": 1.929491525423729e-05, "loss": 1.515, "step": 1540 }, { "epoch": 0.04, "grad_norm": 7.875, "learning_rate": 1.928813559322034e-05, "loss": 1.6476, "step": 1550 }, { "epoch": 0.04, "grad_norm": 14.5625, "learning_rate": 1.9281355932203392e-05, "loss": 1.4581, "step": 1560 }, { "epoch": 0.04, "grad_norm": 8.0625, "learning_rate": 1.9274576271186442e-05, "loss": 1.5717, "step": 1570 }, { "epoch": 0.04, "grad_norm": 11.3125, "learning_rate": 1.9267796610169492e-05, "loss": 1.4752, "step": 1580 }, { "epoch": 0.04, "grad_norm": 12.875, "learning_rate": 1.9261016949152545e-05, "loss": 1.5378, "step": 1590 }, { "epoch": 0.04, "grad_norm": 23.125, "learning_rate": 1.9254237288135595e-05, "loss": 1.608, "step": 1600 }, { "epoch": 0.04, "grad_norm": 21.75, "learning_rate": 1.9247457627118648e-05, "loss": 1.4953, "step": 1610 }, { "epoch": 0.04, "grad_norm": 19.875, "learning_rate": 1.9240677966101698e-05, "loss": 1.4522, "step": 1620 }, { "epoch": 0.04, "grad_norm": 14.9375, "learning_rate": 1.9233898305084747e-05, "loss": 1.6297, "step": 1630 }, { "epoch": 0.04, "grad_norm": 11.6875, "learning_rate": 1.92271186440678e-05, "loss": 1.698, "step": 1640 }, { "epoch": 0.04, "grad_norm": 10.625, "learning_rate": 1.9220338983050847e-05, "loss": 1.5721, "step": 1650 }, { "epoch": 0.04, "grad_norm": 9.0, "learning_rate": 1.92135593220339e-05, "loss": 1.487, "step": 1660 }, { "epoch": 0.04, "grad_norm": 10.625, "learning_rate": 1.920677966101695e-05, "loss": 1.5897, "step": 1670 }, { "epoch": 0.04, "grad_norm": 15.25, "learning_rate": 1.9200000000000003e-05, "loss": 1.6502, "step": 1680 }, { "epoch": 0.04, "grad_norm": 24.625, "learning_rate": 1.9193220338983052e-05, "loss": 1.6367, "step": 1690 }, { "epoch": 0.04, "grad_norm": 8.3125, "learning_rate": 1.9186440677966102e-05, "loss": 1.4932, "step": 1700 }, { "epoch": 0.04, "grad_norm": 14.8125, "learning_rate": 1.9179661016949155e-05, "loss": 1.6045, "step": 1710 }, { "epoch": 0.04, "grad_norm": 12.0, "learning_rate": 1.9172881355932205e-05, "loss": 1.6201, "step": 1720 }, { "epoch": 0.04, "grad_norm": 9.3125, "learning_rate": 1.9166101694915254e-05, "loss": 1.395, "step": 1730 }, { "epoch": 0.04, "grad_norm": 14.0, "learning_rate": 1.9159322033898308e-05, "loss": 1.4311, "step": 1740 }, { "epoch": 0.04, "grad_norm": 10.875, "learning_rate": 1.9152542372881357e-05, "loss": 1.6827, "step": 1750 }, { "epoch": 0.04, "grad_norm": 10.25, "learning_rate": 1.9145762711864407e-05, "loss": 1.6859, "step": 1760 }, { "epoch": 0.04, "grad_norm": 16.125, "learning_rate": 1.913898305084746e-05, "loss": 1.5337, "step": 1770 }, { "epoch": 0.04, "grad_norm": 8.8125, "learning_rate": 1.913220338983051e-05, "loss": 1.7639, "step": 1780 }, { "epoch": 0.04, "grad_norm": 11.1875, "learning_rate": 1.9125423728813563e-05, "loss": 1.508, "step": 1790 }, { "epoch": 0.04, "grad_norm": 9.5625, "learning_rate": 1.9118644067796613e-05, "loss": 1.5, "step": 1800 }, { "epoch": 0.05, "grad_norm": 11.875, "learning_rate": 1.9111864406779662e-05, "loss": 1.6331, "step": 1810 }, { "epoch": 0.05, "grad_norm": 9.875, "learning_rate": 1.9105084745762715e-05, "loss": 1.3836, "step": 1820 }, { "epoch": 0.05, "grad_norm": 11.0625, "learning_rate": 1.9098305084745765e-05, "loss": 1.5407, "step": 1830 }, { "epoch": 0.05, "grad_norm": 12.25, "learning_rate": 1.9091525423728815e-05, "loss": 1.5822, "step": 1840 }, { "epoch": 0.05, "grad_norm": 10.125, "learning_rate": 1.9084745762711868e-05, "loss": 1.5684, "step": 1850 }, { "epoch": 0.05, "grad_norm": 13.875, "learning_rate": 1.9077966101694914e-05, "loss": 1.5036, "step": 1860 }, { "epoch": 0.05, "grad_norm": 15.0625, "learning_rate": 1.9071186440677967e-05, "loss": 1.5305, "step": 1870 }, { "epoch": 0.05, "grad_norm": 33.5, "learning_rate": 1.9064406779661017e-05, "loss": 1.4381, "step": 1880 }, { "epoch": 0.05, "grad_norm": 13.625, "learning_rate": 1.905762711864407e-05, "loss": 1.5901, "step": 1890 }, { "epoch": 0.05, "grad_norm": 16.625, "learning_rate": 1.905084745762712e-05, "loss": 1.4541, "step": 1900 }, { "epoch": 0.05, "grad_norm": 8.625, "learning_rate": 1.904406779661017e-05, "loss": 1.5044, "step": 1910 }, { "epoch": 0.05, "grad_norm": 15.75, "learning_rate": 1.9037288135593223e-05, "loss": 1.5519, "step": 1920 }, { "epoch": 0.05, "grad_norm": 8.0625, "learning_rate": 1.9030508474576272e-05, "loss": 1.5799, "step": 1930 }, { "epoch": 0.05, "grad_norm": 9.625, "learning_rate": 1.9023728813559322e-05, "loss": 1.5128, "step": 1940 }, { "epoch": 0.05, "grad_norm": 19.125, "learning_rate": 1.9016949152542375e-05, "loss": 1.5036, "step": 1950 }, { "epoch": 0.05, "grad_norm": 11.75, "learning_rate": 1.9010169491525425e-05, "loss": 1.4, "step": 1960 }, { "epoch": 0.05, "grad_norm": 8.625, "learning_rate": 1.9003389830508478e-05, "loss": 1.3903, "step": 1970 }, { "epoch": 0.05, "grad_norm": 10.6875, "learning_rate": 1.8996610169491528e-05, "loss": 1.5759, "step": 1980 }, { "epoch": 0.05, "grad_norm": 12.0625, "learning_rate": 1.8989830508474577e-05, "loss": 1.6143, "step": 1990 }, { "epoch": 0.05, "grad_norm": 43.5, "learning_rate": 1.898305084745763e-05, "loss": 1.5139, "step": 2000 }, { "epoch": 0.05, "grad_norm": 24.625, "learning_rate": 1.897627118644068e-05, "loss": 1.6689, "step": 2010 }, { "epoch": 0.05, "grad_norm": 26.5, "learning_rate": 1.896949152542373e-05, "loss": 1.6536, "step": 2020 }, { "epoch": 0.05, "grad_norm": 10.875, "learning_rate": 1.8962711864406783e-05, "loss": 1.5775, "step": 2030 }, { "epoch": 0.05, "grad_norm": 15.5, "learning_rate": 1.8955932203389833e-05, "loss": 1.5513, "step": 2040 }, { "epoch": 0.05, "grad_norm": 13.5, "learning_rate": 1.8949152542372882e-05, "loss": 1.5543, "step": 2050 }, { "epoch": 0.05, "grad_norm": 11.5625, "learning_rate": 1.8942372881355932e-05, "loss": 1.6387, "step": 2060 }, { "epoch": 0.05, "grad_norm": 10.25, "learning_rate": 1.8935593220338985e-05, "loss": 1.5928, "step": 2070 }, { "epoch": 0.05, "grad_norm": 9.5625, "learning_rate": 1.8928813559322035e-05, "loss": 1.5242, "step": 2080 }, { "epoch": 0.05, "grad_norm": 8.5, "learning_rate": 1.8922033898305085e-05, "loss": 1.5208, "step": 2090 }, { "epoch": 0.05, "grad_norm": 10.5625, "learning_rate": 1.8915254237288138e-05, "loss": 1.4693, "step": 2100 }, { "epoch": 0.05, "grad_norm": 18.625, "learning_rate": 1.8908474576271187e-05, "loss": 1.6504, "step": 2110 }, { "epoch": 0.05, "grad_norm": 15.1875, "learning_rate": 1.8901694915254237e-05, "loss": 1.3865, "step": 2120 }, { "epoch": 0.05, "grad_norm": 18.75, "learning_rate": 1.889491525423729e-05, "loss": 1.5483, "step": 2130 }, { "epoch": 0.05, "grad_norm": 10.5, "learning_rate": 1.888813559322034e-05, "loss": 1.5347, "step": 2140 }, { "epoch": 0.05, "grad_norm": 19.875, "learning_rate": 1.8881355932203393e-05, "loss": 1.4194, "step": 2150 }, { "epoch": 0.05, "grad_norm": 9.8125, "learning_rate": 1.8874576271186443e-05, "loss": 1.617, "step": 2160 }, { "epoch": 0.05, "grad_norm": 8.25, "learning_rate": 1.8867796610169492e-05, "loss": 1.4816, "step": 2170 }, { "epoch": 0.05, "grad_norm": 20.625, "learning_rate": 1.8861016949152545e-05, "loss": 1.6254, "step": 2180 }, { "epoch": 0.05, "grad_norm": 14.6875, "learning_rate": 1.8854237288135595e-05, "loss": 1.6584, "step": 2190 }, { "epoch": 0.06, "grad_norm": 6.96875, "learning_rate": 1.8847457627118645e-05, "loss": 1.6248, "step": 2200 }, { "epoch": 0.06, "grad_norm": 15.9375, "learning_rate": 1.8840677966101698e-05, "loss": 1.5322, "step": 2210 }, { "epoch": 0.06, "grad_norm": 9.625, "learning_rate": 1.8833898305084748e-05, "loss": 1.506, "step": 2220 }, { "epoch": 0.06, "grad_norm": 9.75, "learning_rate": 1.8827118644067797e-05, "loss": 1.6489, "step": 2230 }, { "epoch": 0.06, "grad_norm": 15.4375, "learning_rate": 1.882033898305085e-05, "loss": 1.4596, "step": 2240 }, { "epoch": 0.06, "grad_norm": 11.25, "learning_rate": 1.88135593220339e-05, "loss": 1.5384, "step": 2250 }, { "epoch": 0.06, "grad_norm": 9.3125, "learning_rate": 1.8806779661016953e-05, "loss": 1.5334, "step": 2260 }, { "epoch": 0.06, "grad_norm": 8.6875, "learning_rate": 1.88e-05, "loss": 1.5069, "step": 2270 }, { "epoch": 0.06, "grad_norm": 26.0, "learning_rate": 1.8793220338983053e-05, "loss": 1.4391, "step": 2280 }, { "epoch": 0.06, "grad_norm": 16.0, "learning_rate": 1.8786440677966102e-05, "loss": 1.4752, "step": 2290 }, { "epoch": 0.06, "grad_norm": 14.8125, "learning_rate": 1.8779661016949152e-05, "loss": 1.5738, "step": 2300 }, { "epoch": 0.06, "grad_norm": 12.625, "learning_rate": 1.8772881355932205e-05, "loss": 1.3754, "step": 2310 }, { "epoch": 0.06, "grad_norm": 12.875, "learning_rate": 1.8766101694915255e-05, "loss": 1.5944, "step": 2320 }, { "epoch": 0.06, "grad_norm": 15.0625, "learning_rate": 1.8759322033898305e-05, "loss": 1.4887, "step": 2330 }, { "epoch": 0.06, "grad_norm": 9.3125, "learning_rate": 1.8752542372881358e-05, "loss": 1.4891, "step": 2340 }, { "epoch": 0.06, "grad_norm": 24.25, "learning_rate": 1.8745762711864407e-05, "loss": 1.7288, "step": 2350 }, { "epoch": 0.06, "grad_norm": 12.8125, "learning_rate": 1.873898305084746e-05, "loss": 1.4783, "step": 2360 }, { "epoch": 0.06, "grad_norm": 6.5, "learning_rate": 1.873220338983051e-05, "loss": 1.6752, "step": 2370 }, { "epoch": 0.06, "grad_norm": 17.5, "learning_rate": 1.872542372881356e-05, "loss": 1.3905, "step": 2380 }, { "epoch": 0.06, "grad_norm": 13.9375, "learning_rate": 1.8718644067796613e-05, "loss": 1.4255, "step": 2390 }, { "epoch": 0.06, "grad_norm": 11.5, "learning_rate": 1.8711864406779663e-05, "loss": 1.5024, "step": 2400 }, { "epoch": 0.06, "grad_norm": 15.375, "learning_rate": 1.8705084745762712e-05, "loss": 1.4946, "step": 2410 }, { "epoch": 0.06, "grad_norm": 12.3125, "learning_rate": 1.8698305084745765e-05, "loss": 1.1831, "step": 2420 }, { "epoch": 0.06, "grad_norm": 18.625, "learning_rate": 1.8691525423728815e-05, "loss": 1.4678, "step": 2430 }, { "epoch": 0.06, "grad_norm": 18.125, "learning_rate": 1.8684745762711868e-05, "loss": 1.3618, "step": 2440 }, { "epoch": 0.06, "grad_norm": 8.125, "learning_rate": 1.8677966101694918e-05, "loss": 1.5961, "step": 2450 }, { "epoch": 0.06, "grad_norm": 11.9375, "learning_rate": 1.8671186440677968e-05, "loss": 1.5546, "step": 2460 }, { "epoch": 0.06, "grad_norm": 10.75, "learning_rate": 1.866440677966102e-05, "loss": 1.628, "step": 2470 }, { "epoch": 0.06, "grad_norm": 21.75, "learning_rate": 1.8657627118644067e-05, "loss": 1.3956, "step": 2480 }, { "epoch": 0.06, "grad_norm": 13.875, "learning_rate": 1.865084745762712e-05, "loss": 1.4657, "step": 2490 }, { "epoch": 0.06, "grad_norm": 16.5, "learning_rate": 1.864406779661017e-05, "loss": 1.6003, "step": 2500 }, { "epoch": 0.06, "grad_norm": 11.875, "learning_rate": 1.863728813559322e-05, "loss": 1.5296, "step": 2510 }, { "epoch": 0.06, "grad_norm": 8.5, "learning_rate": 1.8630508474576273e-05, "loss": 1.5027, "step": 2520 }, { "epoch": 0.06, "grad_norm": 6.875, "learning_rate": 1.8623728813559322e-05, "loss": 1.416, "step": 2530 }, { "epoch": 0.06, "grad_norm": 16.875, "learning_rate": 1.8616949152542376e-05, "loss": 1.5729, "step": 2540 }, { "epoch": 0.06, "grad_norm": 10.5, "learning_rate": 1.8610169491525425e-05, "loss": 1.6464, "step": 2550 }, { "epoch": 0.06, "grad_norm": 12.375, "learning_rate": 1.8603389830508475e-05, "loss": 1.5534, "step": 2560 }, { "epoch": 0.06, "grad_norm": 9.0, "learning_rate": 1.8596610169491528e-05, "loss": 1.6106, "step": 2570 }, { "epoch": 0.06, "grad_norm": 84.0, "learning_rate": 1.8589830508474578e-05, "loss": 1.5543, "step": 2580 }, { "epoch": 0.06, "grad_norm": 13.6875, "learning_rate": 1.8583050847457627e-05, "loss": 1.5078, "step": 2590 }, { "epoch": 0.07, "grad_norm": 12.125, "learning_rate": 1.857627118644068e-05, "loss": 1.4837, "step": 2600 }, { "epoch": 0.07, "grad_norm": 17.375, "learning_rate": 1.856949152542373e-05, "loss": 1.3, "step": 2610 }, { "epoch": 0.07, "grad_norm": 13.375, "learning_rate": 1.856271186440678e-05, "loss": 1.6079, "step": 2620 }, { "epoch": 0.07, "grad_norm": 38.0, "learning_rate": 1.8555932203389833e-05, "loss": 1.5849, "step": 2630 }, { "epoch": 0.07, "grad_norm": 17.75, "learning_rate": 1.8549152542372883e-05, "loss": 1.4636, "step": 2640 }, { "epoch": 0.07, "grad_norm": 12.5625, "learning_rate": 1.8542372881355936e-05, "loss": 1.5832, "step": 2650 }, { "epoch": 0.07, "grad_norm": 7.6875, "learning_rate": 1.8535593220338986e-05, "loss": 1.4782, "step": 2660 }, { "epoch": 0.07, "grad_norm": 10.0, "learning_rate": 1.8528813559322035e-05, "loss": 1.5614, "step": 2670 }, { "epoch": 0.07, "grad_norm": 9.375, "learning_rate": 1.852203389830509e-05, "loss": 1.4678, "step": 2680 }, { "epoch": 0.07, "grad_norm": 12.125, "learning_rate": 1.8515254237288135e-05, "loss": 1.4207, "step": 2690 }, { "epoch": 0.07, "grad_norm": 11.6875, "learning_rate": 1.8508474576271188e-05, "loss": 1.5857, "step": 2700 }, { "epoch": 0.07, "grad_norm": 11.5, "learning_rate": 1.8501694915254237e-05, "loss": 1.5383, "step": 2710 }, { "epoch": 0.07, "grad_norm": 12.9375, "learning_rate": 1.849491525423729e-05, "loss": 1.4254, "step": 2720 }, { "epoch": 0.07, "grad_norm": 15.6875, "learning_rate": 1.848813559322034e-05, "loss": 1.4762, "step": 2730 }, { "epoch": 0.07, "grad_norm": 18.625, "learning_rate": 1.848135593220339e-05, "loss": 1.4275, "step": 2740 }, { "epoch": 0.07, "grad_norm": 12.5625, "learning_rate": 1.8474576271186443e-05, "loss": 1.2928, "step": 2750 }, { "epoch": 0.07, "grad_norm": 12.6875, "learning_rate": 1.8467796610169493e-05, "loss": 1.4166, "step": 2760 }, { "epoch": 0.07, "grad_norm": 8.1875, "learning_rate": 1.8461016949152542e-05, "loss": 1.4579, "step": 2770 }, { "epoch": 0.07, "grad_norm": 8.6875, "learning_rate": 1.8454237288135596e-05, "loss": 1.4313, "step": 2780 }, { "epoch": 0.07, "grad_norm": 8.8125, "learning_rate": 1.8447457627118645e-05, "loss": 1.5309, "step": 2790 }, { "epoch": 0.07, "grad_norm": 12.5, "learning_rate": 1.8440677966101695e-05, "loss": 1.4502, "step": 2800 }, { "epoch": 0.07, "grad_norm": 8.1875, "learning_rate": 1.8433898305084748e-05, "loss": 1.4403, "step": 2810 }, { "epoch": 0.07, "grad_norm": 8.125, "learning_rate": 1.8427118644067798e-05, "loss": 1.612, "step": 2820 }, { "epoch": 0.07, "grad_norm": 10.5, "learning_rate": 1.842033898305085e-05, "loss": 1.3711, "step": 2830 }, { "epoch": 0.07, "grad_norm": 11.25, "learning_rate": 1.84135593220339e-05, "loss": 1.4374, "step": 2840 }, { "epoch": 0.07, "grad_norm": 11.125, "learning_rate": 1.840677966101695e-05, "loss": 1.4582, "step": 2850 }, { "epoch": 0.07, "grad_norm": 9.8125, "learning_rate": 1.8400000000000003e-05, "loss": 1.5342, "step": 2860 }, { "epoch": 0.07, "grad_norm": 13.1875, "learning_rate": 1.8393220338983053e-05, "loss": 1.3282, "step": 2870 }, { "epoch": 0.07, "grad_norm": 8.5, "learning_rate": 1.8386440677966103e-05, "loss": 1.5913, "step": 2880 }, { "epoch": 0.07, "grad_norm": 11.0, "learning_rate": 1.8379661016949153e-05, "loss": 1.6801, "step": 2890 }, { "epoch": 0.07, "grad_norm": 19.25, "learning_rate": 1.8372881355932202e-05, "loss": 1.4581, "step": 2900 }, { "epoch": 0.07, "grad_norm": 17.5, "learning_rate": 1.8366101694915255e-05, "loss": 1.3822, "step": 2910 }, { "epoch": 0.07, "grad_norm": 28.125, "learning_rate": 1.8359322033898305e-05, "loss": 1.3937, "step": 2920 }, { "epoch": 0.07, "grad_norm": 11.4375, "learning_rate": 1.8352542372881358e-05, "loss": 1.4311, "step": 2930 }, { "epoch": 0.07, "grad_norm": 6.4375, "learning_rate": 1.8345762711864408e-05, "loss": 1.4237, "step": 2940 }, { "epoch": 0.07, "grad_norm": 20.625, "learning_rate": 1.8338983050847458e-05, "loss": 1.7361, "step": 2950 }, { "epoch": 0.07, "grad_norm": 13.875, "learning_rate": 1.833220338983051e-05, "loss": 1.5519, "step": 2960 }, { "epoch": 0.07, "grad_norm": 8.625, "learning_rate": 1.832542372881356e-05, "loss": 1.3663, "step": 2970 }, { "epoch": 0.07, "grad_norm": 16.875, "learning_rate": 1.831864406779661e-05, "loss": 1.4183, "step": 2980 }, { "epoch": 0.07, "grad_norm": 23.375, "learning_rate": 1.8311864406779663e-05, "loss": 1.4189, "step": 2990 }, { "epoch": 0.07, "grad_norm": 16.125, "learning_rate": 1.8305084745762713e-05, "loss": 1.5008, "step": 3000 }, { "epoch": 0.08, "grad_norm": 19.0, "learning_rate": 1.8298305084745766e-05, "loss": 1.2341, "step": 3010 }, { "epoch": 0.08, "grad_norm": 9.0625, "learning_rate": 1.8291525423728816e-05, "loss": 1.3109, "step": 3020 }, { "epoch": 0.08, "grad_norm": 14.125, "learning_rate": 1.8284745762711865e-05, "loss": 1.4832, "step": 3030 }, { "epoch": 0.08, "grad_norm": 25.75, "learning_rate": 1.827796610169492e-05, "loss": 1.4074, "step": 3040 }, { "epoch": 0.08, "grad_norm": 9.5625, "learning_rate": 1.8271186440677968e-05, "loss": 1.324, "step": 3050 }, { "epoch": 0.08, "grad_norm": 19.0, "learning_rate": 1.8264406779661018e-05, "loss": 1.6133, "step": 3060 }, { "epoch": 0.08, "grad_norm": 12.75, "learning_rate": 1.825762711864407e-05, "loss": 1.4959, "step": 3070 }, { "epoch": 0.08, "grad_norm": 17.75, "learning_rate": 1.825084745762712e-05, "loss": 1.5612, "step": 3080 }, { "epoch": 0.08, "grad_norm": 11.125, "learning_rate": 1.824406779661017e-05, "loss": 1.4213, "step": 3090 }, { "epoch": 0.08, "grad_norm": 22.625, "learning_rate": 1.823728813559322e-05, "loss": 1.6306, "step": 3100 }, { "epoch": 0.08, "grad_norm": 13.375, "learning_rate": 1.8230508474576273e-05, "loss": 1.6571, "step": 3110 }, { "epoch": 0.08, "grad_norm": 11.875, "learning_rate": 1.8223728813559323e-05, "loss": 1.5212, "step": 3120 }, { "epoch": 0.08, "grad_norm": 9.0, "learning_rate": 1.8216949152542373e-05, "loss": 1.5581, "step": 3130 }, { "epoch": 0.08, "grad_norm": 7.03125, "learning_rate": 1.8210169491525426e-05, "loss": 1.4206, "step": 3140 }, { "epoch": 0.08, "grad_norm": 11.8125, "learning_rate": 1.8203389830508475e-05, "loss": 1.5661, "step": 3150 }, { "epoch": 0.08, "grad_norm": 12.625, "learning_rate": 1.8196610169491525e-05, "loss": 1.3835, "step": 3160 }, { "epoch": 0.08, "grad_norm": 15.6875, "learning_rate": 1.8189830508474578e-05, "loss": 1.5146, "step": 3170 }, { "epoch": 0.08, "grad_norm": 8.0625, "learning_rate": 1.8183050847457628e-05, "loss": 1.466, "step": 3180 }, { "epoch": 0.08, "grad_norm": 21.625, "learning_rate": 1.817627118644068e-05, "loss": 1.5828, "step": 3190 }, { "epoch": 0.08, "grad_norm": 6.90625, "learning_rate": 1.816949152542373e-05, "loss": 1.4519, "step": 3200 }, { "epoch": 0.08, "grad_norm": 14.4375, "learning_rate": 1.816271186440678e-05, "loss": 1.6093, "step": 3210 }, { "epoch": 0.08, "grad_norm": 20.625, "learning_rate": 1.8155932203389833e-05, "loss": 1.6915, "step": 3220 }, { "epoch": 0.08, "grad_norm": 10.1875, "learning_rate": 1.8149152542372883e-05, "loss": 1.5521, "step": 3230 }, { "epoch": 0.08, "grad_norm": 10.625, "learning_rate": 1.8142372881355933e-05, "loss": 1.4542, "step": 3240 }, { "epoch": 0.08, "grad_norm": 8.6875, "learning_rate": 1.8135593220338986e-05, "loss": 1.494, "step": 3250 }, { "epoch": 0.08, "grad_norm": 9.625, "learning_rate": 1.8128813559322036e-05, "loss": 1.3653, "step": 3260 }, { "epoch": 0.08, "grad_norm": 12.375, "learning_rate": 1.8122033898305085e-05, "loss": 1.3581, "step": 3270 }, { "epoch": 0.08, "grad_norm": 22.625, "learning_rate": 1.811525423728814e-05, "loss": 1.6393, "step": 3280 }, { "epoch": 0.08, "grad_norm": 19.875, "learning_rate": 1.8108474576271188e-05, "loss": 1.4813, "step": 3290 }, { "epoch": 0.08, "grad_norm": 15.3125, "learning_rate": 1.810169491525424e-05, "loss": 1.3628, "step": 3300 }, { "epoch": 0.08, "grad_norm": 10.25, "learning_rate": 1.8094915254237288e-05, "loss": 1.5978, "step": 3310 }, { "epoch": 0.08, "grad_norm": 12.5625, "learning_rate": 1.808813559322034e-05, "loss": 1.5745, "step": 3320 }, { "epoch": 0.08, "grad_norm": 20.0, "learning_rate": 1.808135593220339e-05, "loss": 1.4374, "step": 3330 }, { "epoch": 0.08, "grad_norm": 8.9375, "learning_rate": 1.807457627118644e-05, "loss": 1.4625, "step": 3340 }, { "epoch": 0.08, "grad_norm": 22.625, "learning_rate": 1.8067796610169493e-05, "loss": 1.6427, "step": 3350 }, { "epoch": 0.08, "grad_norm": 12.6875, "learning_rate": 1.8061016949152543e-05, "loss": 1.6387, "step": 3360 }, { "epoch": 0.08, "grad_norm": 13.1875, "learning_rate": 1.8054237288135593e-05, "loss": 1.405, "step": 3370 }, { "epoch": 0.08, "grad_norm": 22.875, "learning_rate": 1.8047457627118646e-05, "loss": 1.3973, "step": 3380 }, { "epoch": 0.08, "grad_norm": 8.625, "learning_rate": 1.8040677966101695e-05, "loss": 1.4561, "step": 3390 }, { "epoch": 0.09, "grad_norm": 12.625, "learning_rate": 1.803389830508475e-05, "loss": 1.4632, "step": 3400 }, { "epoch": 0.09, "grad_norm": 18.125, "learning_rate": 1.8027118644067798e-05, "loss": 1.4032, "step": 3410 }, { "epoch": 0.09, "grad_norm": 12.75, "learning_rate": 1.8020338983050848e-05, "loss": 1.5775, "step": 3420 }, { "epoch": 0.09, "grad_norm": 17.875, "learning_rate": 1.80135593220339e-05, "loss": 1.4885, "step": 3430 }, { "epoch": 0.09, "grad_norm": 17.25, "learning_rate": 1.800677966101695e-05, "loss": 1.414, "step": 3440 }, { "epoch": 0.09, "grad_norm": 10.5, "learning_rate": 1.8e-05, "loss": 1.4406, "step": 3450 }, { "epoch": 0.09, "grad_norm": 15.75, "learning_rate": 1.7993220338983054e-05, "loss": 1.4872, "step": 3460 }, { "epoch": 0.09, "grad_norm": 10.25, "learning_rate": 1.7986440677966103e-05, "loss": 1.4513, "step": 3470 }, { "epoch": 0.09, "grad_norm": 11.875, "learning_rate": 1.7979661016949156e-05, "loss": 1.5059, "step": 3480 }, { "epoch": 0.09, "grad_norm": 14.25, "learning_rate": 1.7972881355932206e-05, "loss": 1.4446, "step": 3490 }, { "epoch": 0.09, "grad_norm": 14.0, "learning_rate": 1.7966101694915256e-05, "loss": 1.4928, "step": 3500 }, { "epoch": 0.09, "grad_norm": 19.5, "learning_rate": 1.795932203389831e-05, "loss": 1.5305, "step": 3510 }, { "epoch": 0.09, "grad_norm": 10.0, "learning_rate": 1.7952542372881355e-05, "loss": 1.4932, "step": 3520 }, { "epoch": 0.09, "grad_norm": 32.0, "learning_rate": 1.7945762711864408e-05, "loss": 1.2962, "step": 3530 }, { "epoch": 0.09, "grad_norm": 12.0, "learning_rate": 1.7938983050847458e-05, "loss": 1.2428, "step": 3540 }, { "epoch": 0.09, "grad_norm": 12.4375, "learning_rate": 1.7932203389830508e-05, "loss": 1.5858, "step": 3550 }, { "epoch": 0.09, "grad_norm": 7.78125, "learning_rate": 1.792542372881356e-05, "loss": 1.4792, "step": 3560 }, { "epoch": 0.09, "grad_norm": 8.3125, "learning_rate": 1.791864406779661e-05, "loss": 1.4305, "step": 3570 }, { "epoch": 0.09, "grad_norm": 19.5, "learning_rate": 1.7911864406779664e-05, "loss": 1.5068, "step": 3580 }, { "epoch": 0.09, "grad_norm": 12.6875, "learning_rate": 1.7905084745762713e-05, "loss": 1.3162, "step": 3590 }, { "epoch": 0.09, "grad_norm": 16.875, "learning_rate": 1.7898305084745763e-05, "loss": 1.5589, "step": 3600 }, { "epoch": 0.09, "grad_norm": 12.0625, "learning_rate": 1.7891525423728816e-05, "loss": 1.3933, "step": 3610 }, { "epoch": 0.09, "grad_norm": 10.75, "learning_rate": 1.7884745762711866e-05, "loss": 1.409, "step": 3620 }, { "epoch": 0.09, "grad_norm": 10.1875, "learning_rate": 1.7877966101694916e-05, "loss": 1.5357, "step": 3630 }, { "epoch": 0.09, "grad_norm": 8.8125, "learning_rate": 1.787118644067797e-05, "loss": 1.4958, "step": 3640 }, { "epoch": 0.09, "grad_norm": 12.25, "learning_rate": 1.7864406779661018e-05, "loss": 1.4976, "step": 3650 }, { "epoch": 0.09, "grad_norm": 22.125, "learning_rate": 1.785762711864407e-05, "loss": 1.472, "step": 3660 }, { "epoch": 0.09, "grad_norm": 8.6875, "learning_rate": 1.785084745762712e-05, "loss": 1.2726, "step": 3670 }, { "epoch": 0.09, "grad_norm": 12.1875, "learning_rate": 1.784406779661017e-05, "loss": 1.3201, "step": 3680 }, { "epoch": 0.09, "grad_norm": 12.4375, "learning_rate": 1.7837288135593224e-05, "loss": 1.4802, "step": 3690 }, { "epoch": 0.09, "grad_norm": 7.0625, "learning_rate": 1.7830508474576274e-05, "loss": 1.4807, "step": 3700 }, { "epoch": 0.09, "grad_norm": 32.5, "learning_rate": 1.7823728813559323e-05, "loss": 1.7589, "step": 3710 }, { "epoch": 0.09, "grad_norm": 14.8125, "learning_rate": 1.7816949152542376e-05, "loss": 1.4768, "step": 3720 }, { "epoch": 0.09, "grad_norm": 16.5, "learning_rate": 1.7810169491525423e-05, "loss": 1.3863, "step": 3730 }, { "epoch": 0.09, "grad_norm": 10.5625, "learning_rate": 1.7803389830508476e-05, "loss": 1.4855, "step": 3740 }, { "epoch": 0.09, "grad_norm": 12.3125, "learning_rate": 1.7796610169491526e-05, "loss": 1.4989, "step": 3750 }, { "epoch": 0.09, "grad_norm": 15.375, "learning_rate": 1.778983050847458e-05, "loss": 1.4212, "step": 3760 }, { "epoch": 0.09, "grad_norm": 13.6875, "learning_rate": 1.778305084745763e-05, "loss": 1.3991, "step": 3770 }, { "epoch": 0.09, "grad_norm": 11.3125, "learning_rate": 1.7776271186440678e-05, "loss": 1.496, "step": 3780 }, { "epoch": 0.09, "grad_norm": 32.75, "learning_rate": 1.776949152542373e-05, "loss": 1.5595, "step": 3790 }, { "epoch": 0.1, "grad_norm": 11.75, "learning_rate": 1.776271186440678e-05, "loss": 1.5164, "step": 3800 }, { "epoch": 0.1, "grad_norm": 11.125, "learning_rate": 1.775593220338983e-05, "loss": 1.5812, "step": 3810 }, { "epoch": 0.1, "grad_norm": 23.125, "learning_rate": 1.7749152542372884e-05, "loss": 1.5681, "step": 3820 }, { "epoch": 0.1, "grad_norm": 11.875, "learning_rate": 1.7742372881355933e-05, "loss": 1.5163, "step": 3830 }, { "epoch": 0.1, "grad_norm": 14.875, "learning_rate": 1.7735593220338983e-05, "loss": 1.427, "step": 3840 }, { "epoch": 0.1, "grad_norm": 26.25, "learning_rate": 1.7728813559322036e-05, "loss": 1.4703, "step": 3850 }, { "epoch": 0.1, "grad_norm": 9.125, "learning_rate": 1.7722033898305086e-05, "loss": 1.3093, "step": 3860 }, { "epoch": 0.1, "grad_norm": 31.5, "learning_rate": 1.771525423728814e-05, "loss": 1.2996, "step": 3870 }, { "epoch": 0.1, "grad_norm": 18.5, "learning_rate": 1.770847457627119e-05, "loss": 1.5404, "step": 3880 }, { "epoch": 0.1, "grad_norm": 16.25, "learning_rate": 1.770169491525424e-05, "loss": 1.4538, "step": 3890 }, { "epoch": 0.1, "grad_norm": 9.6875, "learning_rate": 1.769491525423729e-05, "loss": 1.5769, "step": 3900 }, { "epoch": 0.1, "grad_norm": 19.375, "learning_rate": 1.768813559322034e-05, "loss": 1.4964, "step": 3910 }, { "epoch": 0.1, "grad_norm": 10.6875, "learning_rate": 1.768135593220339e-05, "loss": 1.5531, "step": 3920 }, { "epoch": 0.1, "grad_norm": 8.4375, "learning_rate": 1.767457627118644e-05, "loss": 1.561, "step": 3930 }, { "epoch": 0.1, "grad_norm": 12.5, "learning_rate": 1.7667796610169494e-05, "loss": 1.5124, "step": 3940 }, { "epoch": 0.1, "grad_norm": 14.9375, "learning_rate": 1.7661016949152543e-05, "loss": 1.396, "step": 3950 }, { "epoch": 0.1, "grad_norm": 6.84375, "learning_rate": 1.7654237288135593e-05, "loss": 1.507, "step": 3960 }, { "epoch": 0.1, "grad_norm": 9.4375, "learning_rate": 1.7647457627118646e-05, "loss": 1.5669, "step": 3970 }, { "epoch": 0.1, "grad_norm": 9.5625, "learning_rate": 1.7640677966101696e-05, "loss": 1.3942, "step": 3980 }, { "epoch": 0.1, "grad_norm": 19.5, "learning_rate": 1.7633898305084746e-05, "loss": 1.4645, "step": 3990 }, { "epoch": 0.1, "grad_norm": 18.375, "learning_rate": 1.76271186440678e-05, "loss": 1.3539, "step": 4000 }, { "epoch": 0.1, "grad_norm": 28.0, "learning_rate": 1.762033898305085e-05, "loss": 1.5859, "step": 4010 }, { "epoch": 0.1, "grad_norm": 11.1875, "learning_rate": 1.7613559322033898e-05, "loss": 1.3802, "step": 4020 }, { "epoch": 0.1, "grad_norm": 21.875, "learning_rate": 1.760677966101695e-05, "loss": 1.3353, "step": 4030 }, { "epoch": 0.1, "grad_norm": 20.75, "learning_rate": 1.76e-05, "loss": 1.5717, "step": 4040 }, { "epoch": 0.1, "grad_norm": 11.0, "learning_rate": 1.7593220338983054e-05, "loss": 1.5115, "step": 4050 }, { "epoch": 0.1, "grad_norm": 22.0, "learning_rate": 1.7586440677966104e-05, "loss": 1.4051, "step": 4060 }, { "epoch": 0.1, "grad_norm": 14.9375, "learning_rate": 1.7579661016949153e-05, "loss": 1.6576, "step": 4070 }, { "epoch": 0.1, "grad_norm": 11.25, "learning_rate": 1.7572881355932206e-05, "loss": 1.6336, "step": 4080 }, { "epoch": 0.1, "grad_norm": 9.75, "learning_rate": 1.7566101694915256e-05, "loss": 1.3511, "step": 4090 }, { "epoch": 0.1, "grad_norm": 13.6875, "learning_rate": 1.7559322033898306e-05, "loss": 1.4982, "step": 4100 }, { "epoch": 0.1, "grad_norm": 25.625, "learning_rate": 1.755254237288136e-05, "loss": 1.2387, "step": 4110 }, { "epoch": 0.1, "grad_norm": 13.3125, "learning_rate": 1.754576271186441e-05, "loss": 1.7004, "step": 4120 }, { "epoch": 0.1, "grad_norm": 15.875, "learning_rate": 1.753898305084746e-05, "loss": 1.5061, "step": 4130 }, { "epoch": 0.1, "grad_norm": 15.25, "learning_rate": 1.7532203389830508e-05, "loss": 1.5026, "step": 4140 }, { "epoch": 0.1, "grad_norm": 6.96875, "learning_rate": 1.752542372881356e-05, "loss": 1.5616, "step": 4150 }, { "epoch": 0.1, "grad_norm": 11.25, "learning_rate": 1.751864406779661e-05, "loss": 1.5947, "step": 4160 }, { "epoch": 0.1, "grad_norm": 12.9375, "learning_rate": 1.751186440677966e-05, "loss": 1.5071, "step": 4170 }, { "epoch": 0.1, "grad_norm": 13.6875, "learning_rate": 1.7505084745762714e-05, "loss": 1.3603, "step": 4180 }, { "epoch": 0.1, "grad_norm": 27.0, "learning_rate": 1.7498305084745763e-05, "loss": 1.413, "step": 4190 }, { "epoch": 0.1, "grad_norm": 15.5625, "learning_rate": 1.7491525423728813e-05, "loss": 1.5441, "step": 4200 }, { "epoch": 0.11, "grad_norm": 7.65625, "learning_rate": 1.7484745762711866e-05, "loss": 1.4837, "step": 4210 }, { "epoch": 0.11, "grad_norm": 18.375, "learning_rate": 1.7477966101694916e-05, "loss": 1.4118, "step": 4220 }, { "epoch": 0.11, "grad_norm": 26.5, "learning_rate": 1.747118644067797e-05, "loss": 1.3557, "step": 4230 }, { "epoch": 0.11, "grad_norm": 20.875, "learning_rate": 1.746440677966102e-05, "loss": 1.3283, "step": 4240 }, { "epoch": 0.11, "grad_norm": 13.6875, "learning_rate": 1.745762711864407e-05, "loss": 1.5952, "step": 4250 }, { "epoch": 0.11, "grad_norm": 14.4375, "learning_rate": 1.745084745762712e-05, "loss": 1.4947, "step": 4260 }, { "epoch": 0.11, "grad_norm": 10.6875, "learning_rate": 1.744406779661017e-05, "loss": 1.5308, "step": 4270 }, { "epoch": 0.11, "grad_norm": 11.75, "learning_rate": 1.743728813559322e-05, "loss": 1.5157, "step": 4280 }, { "epoch": 0.11, "grad_norm": 14.5625, "learning_rate": 1.7430508474576274e-05, "loss": 1.5048, "step": 4290 }, { "epoch": 0.11, "grad_norm": 16.75, "learning_rate": 1.7423728813559324e-05, "loss": 1.4284, "step": 4300 }, { "epoch": 0.11, "grad_norm": 9.3125, "learning_rate": 1.7416949152542373e-05, "loss": 1.5088, "step": 4310 }, { "epoch": 0.11, "grad_norm": 15.8125, "learning_rate": 1.7410169491525427e-05, "loss": 1.5715, "step": 4320 }, { "epoch": 0.11, "grad_norm": 12.4375, "learning_rate": 1.7403389830508476e-05, "loss": 1.441, "step": 4330 }, { "epoch": 0.11, "grad_norm": 11.125, "learning_rate": 1.739661016949153e-05, "loss": 1.3917, "step": 4340 }, { "epoch": 0.11, "grad_norm": 22.25, "learning_rate": 1.7389830508474576e-05, "loss": 1.354, "step": 4350 }, { "epoch": 0.11, "grad_norm": 9.1875, "learning_rate": 1.738305084745763e-05, "loss": 1.5789, "step": 4360 }, { "epoch": 0.11, "grad_norm": 8.8125, "learning_rate": 1.737627118644068e-05, "loss": 1.5641, "step": 4370 }, { "epoch": 0.11, "grad_norm": 23.125, "learning_rate": 1.7369491525423728e-05, "loss": 1.4324, "step": 4380 }, { "epoch": 0.11, "grad_norm": 27.25, "learning_rate": 1.736271186440678e-05, "loss": 1.6862, "step": 4390 }, { "epoch": 0.11, "grad_norm": 10.4375, "learning_rate": 1.735593220338983e-05, "loss": 1.5332, "step": 4400 }, { "epoch": 0.11, "grad_norm": 14.5625, "learning_rate": 1.734915254237288e-05, "loss": 1.4554, "step": 4410 }, { "epoch": 0.11, "grad_norm": 22.125, "learning_rate": 1.7342372881355934e-05, "loss": 1.5159, "step": 4420 }, { "epoch": 0.11, "grad_norm": 26.625, "learning_rate": 1.7335593220338983e-05, "loss": 1.4236, "step": 4430 }, { "epoch": 0.11, "grad_norm": 9.625, "learning_rate": 1.7328813559322037e-05, "loss": 1.5137, "step": 4440 }, { "epoch": 0.11, "grad_norm": 16.75, "learning_rate": 1.7322033898305086e-05, "loss": 1.551, "step": 4450 }, { "epoch": 0.11, "grad_norm": 11.1875, "learning_rate": 1.7315254237288136e-05, "loss": 1.5243, "step": 4460 }, { "epoch": 0.11, "grad_norm": 19.375, "learning_rate": 1.730847457627119e-05, "loss": 1.3952, "step": 4470 }, { "epoch": 0.11, "grad_norm": 10.75, "learning_rate": 1.730169491525424e-05, "loss": 1.5068, "step": 4480 }, { "epoch": 0.11, "grad_norm": 12.75, "learning_rate": 1.729491525423729e-05, "loss": 1.5121, "step": 4490 }, { "epoch": 0.11, "grad_norm": 12.1875, "learning_rate": 1.728813559322034e-05, "loss": 1.5624, "step": 4500 }, { "epoch": 0.11, "grad_norm": 5.5625, "learning_rate": 1.728135593220339e-05, "loss": 1.5784, "step": 4510 }, { "epoch": 0.11, "grad_norm": 20.625, "learning_rate": 1.7274576271186444e-05, "loss": 1.5975, "step": 4520 }, { "epoch": 0.11, "grad_norm": 16.5, "learning_rate": 1.7267796610169494e-05, "loss": 1.5058, "step": 4530 }, { "epoch": 0.11, "grad_norm": 23.125, "learning_rate": 1.7261016949152544e-05, "loss": 1.5025, "step": 4540 }, { "epoch": 0.11, "grad_norm": 8.1875, "learning_rate": 1.7254237288135597e-05, "loss": 1.5016, "step": 4550 }, { "epoch": 0.11, "grad_norm": 18.75, "learning_rate": 1.7247457627118643e-05, "loss": 1.6883, "step": 4560 }, { "epoch": 0.11, "grad_norm": 10.4375, "learning_rate": 1.7240677966101696e-05, "loss": 1.3446, "step": 4570 }, { "epoch": 0.11, "grad_norm": 10.6875, "learning_rate": 1.7233898305084746e-05, "loss": 1.4538, "step": 4580 }, { "epoch": 0.11, "grad_norm": 24.5, "learning_rate": 1.7227118644067796e-05, "loss": 1.6639, "step": 4590 }, { "epoch": 0.12, "grad_norm": 9.125, "learning_rate": 1.722033898305085e-05, "loss": 1.4906, "step": 4600 }, { "epoch": 0.12, "grad_norm": 6.53125, "learning_rate": 1.72135593220339e-05, "loss": 1.4948, "step": 4610 }, { "epoch": 0.12, "grad_norm": 11.875, "learning_rate": 1.720677966101695e-05, "loss": 1.4362, "step": 4620 }, { "epoch": 0.12, "grad_norm": 15.4375, "learning_rate": 1.72e-05, "loss": 1.5428, "step": 4630 }, { "epoch": 0.12, "grad_norm": 8.75, "learning_rate": 1.719322033898305e-05, "loss": 1.2071, "step": 4640 }, { "epoch": 0.12, "grad_norm": 21.125, "learning_rate": 1.7186440677966104e-05, "loss": 1.5102, "step": 4650 }, { "epoch": 0.12, "grad_norm": 13.8125, "learning_rate": 1.7179661016949154e-05, "loss": 1.3775, "step": 4660 }, { "epoch": 0.12, "grad_norm": 18.5, "learning_rate": 1.7172881355932204e-05, "loss": 1.4434, "step": 4670 }, { "epoch": 0.12, "grad_norm": 12.0625, "learning_rate": 1.7166101694915257e-05, "loss": 1.5176, "step": 4680 }, { "epoch": 0.12, "grad_norm": 15.875, "learning_rate": 1.7159322033898306e-05, "loss": 1.4333, "step": 4690 }, { "epoch": 0.12, "grad_norm": 18.375, "learning_rate": 1.715254237288136e-05, "loss": 1.4007, "step": 4700 }, { "epoch": 0.12, "grad_norm": 5.8125, "learning_rate": 1.714576271186441e-05, "loss": 1.5271, "step": 4710 }, { "epoch": 0.12, "grad_norm": 18.875, "learning_rate": 1.713898305084746e-05, "loss": 1.5802, "step": 4720 }, { "epoch": 0.12, "grad_norm": 19.625, "learning_rate": 1.7132203389830512e-05, "loss": 1.684, "step": 4730 }, { "epoch": 0.12, "grad_norm": 17.5, "learning_rate": 1.712542372881356e-05, "loss": 1.3561, "step": 4740 }, { "epoch": 0.12, "grad_norm": 9.125, "learning_rate": 1.711864406779661e-05, "loss": 1.4918, "step": 4750 }, { "epoch": 0.12, "grad_norm": 20.0, "learning_rate": 1.711186440677966e-05, "loss": 1.6214, "step": 4760 }, { "epoch": 0.12, "grad_norm": 18.5, "learning_rate": 1.710508474576271e-05, "loss": 1.5829, "step": 4770 }, { "epoch": 0.12, "grad_norm": 18.25, "learning_rate": 1.7098305084745764e-05, "loss": 1.3968, "step": 4780 }, { "epoch": 0.12, "grad_norm": 42.5, "learning_rate": 1.7091525423728814e-05, "loss": 1.4853, "step": 4790 }, { "epoch": 0.12, "grad_norm": 8.4375, "learning_rate": 1.7084745762711867e-05, "loss": 1.4132, "step": 4800 }, { "epoch": 0.12, "grad_norm": 10.4375, "learning_rate": 1.7077966101694916e-05, "loss": 1.3807, "step": 4810 }, { "epoch": 0.12, "grad_norm": 13.1875, "learning_rate": 1.7071186440677966e-05, "loss": 1.426, "step": 4820 }, { "epoch": 0.12, "grad_norm": 12.875, "learning_rate": 1.706440677966102e-05, "loss": 1.5588, "step": 4830 }, { "epoch": 0.12, "grad_norm": 8.875, "learning_rate": 1.705762711864407e-05, "loss": 1.5123, "step": 4840 }, { "epoch": 0.12, "grad_norm": 18.125, "learning_rate": 1.705084745762712e-05, "loss": 1.3166, "step": 4850 }, { "epoch": 0.12, "grad_norm": 12.5625, "learning_rate": 1.704406779661017e-05, "loss": 1.5191, "step": 4860 }, { "epoch": 0.12, "grad_norm": 14.75, "learning_rate": 1.703728813559322e-05, "loss": 1.4739, "step": 4870 }, { "epoch": 0.12, "grad_norm": 22.875, "learning_rate": 1.703050847457627e-05, "loss": 1.5437, "step": 4880 }, { "epoch": 0.12, "grad_norm": 14.375, "learning_rate": 1.7023728813559324e-05, "loss": 1.501, "step": 4890 }, { "epoch": 0.12, "grad_norm": 14.125, "learning_rate": 1.7016949152542374e-05, "loss": 1.6071, "step": 4900 }, { "epoch": 0.12, "grad_norm": 13.5, "learning_rate": 1.7010169491525427e-05, "loss": 1.4153, "step": 4910 }, { "epoch": 0.12, "grad_norm": 15.875, "learning_rate": 1.7003389830508477e-05, "loss": 1.5077, "step": 4920 }, { "epoch": 0.12, "grad_norm": 18.75, "learning_rate": 1.6996610169491526e-05, "loss": 1.4882, "step": 4930 }, { "epoch": 0.12, "grad_norm": 12.5625, "learning_rate": 1.698983050847458e-05, "loss": 1.3678, "step": 4940 }, { "epoch": 0.12, "grad_norm": 8.25, "learning_rate": 1.698305084745763e-05, "loss": 1.4071, "step": 4950 }, { "epoch": 0.12, "grad_norm": 13.125, "learning_rate": 1.697627118644068e-05, "loss": 1.5199, "step": 4960 }, { "epoch": 0.12, "grad_norm": 23.5, "learning_rate": 1.696949152542373e-05, "loss": 1.484, "step": 4970 }, { "epoch": 0.12, "grad_norm": 15.4375, "learning_rate": 1.6962711864406782e-05, "loss": 1.4716, "step": 4980 }, { "epoch": 0.12, "grad_norm": 18.125, "learning_rate": 1.695593220338983e-05, "loss": 1.4641, "step": 4990 }, { "epoch": 0.12, "grad_norm": 10.75, "learning_rate": 1.694915254237288e-05, "loss": 1.3656, "step": 5000 }, { "epoch": 0.12, "eval_loss": 1.4290132522583008, "eval_runtime": 59.1678, "eval_samples_per_second": 16.901, "eval_steps_per_second": 16.901, "step": 5000 }, { "epoch": 0.13, "grad_norm": 19.0, "learning_rate": 1.6942372881355934e-05, "loss": 1.6571, "step": 5010 }, { "epoch": 0.13, "grad_norm": 17.375, "learning_rate": 1.6935593220338984e-05, "loss": 1.4248, "step": 5020 }, { "epoch": 0.13, "grad_norm": 12.6875, "learning_rate": 1.6928813559322034e-05, "loss": 1.5244, "step": 5030 }, { "epoch": 0.13, "grad_norm": 17.25, "learning_rate": 1.6922033898305087e-05, "loss": 1.4614, "step": 5040 }, { "epoch": 0.13, "grad_norm": 13.0, "learning_rate": 1.6915254237288136e-05, "loss": 1.4683, "step": 5050 }, { "epoch": 0.13, "grad_norm": 11.75, "learning_rate": 1.6908474576271186e-05, "loss": 1.3468, "step": 5060 }, { "epoch": 0.13, "grad_norm": 10.375, "learning_rate": 1.690169491525424e-05, "loss": 1.4622, "step": 5070 }, { "epoch": 0.13, "grad_norm": 17.75, "learning_rate": 1.689491525423729e-05, "loss": 1.4391, "step": 5080 }, { "epoch": 0.13, "grad_norm": 16.0, "learning_rate": 1.6888135593220342e-05, "loss": 1.5315, "step": 5090 }, { "epoch": 0.13, "grad_norm": 14.5, "learning_rate": 1.6881355932203392e-05, "loss": 1.462, "step": 5100 }, { "epoch": 0.13, "grad_norm": 14.625, "learning_rate": 1.687457627118644e-05, "loss": 1.6432, "step": 5110 }, { "epoch": 0.13, "grad_norm": 11.0625, "learning_rate": 1.6867796610169495e-05, "loss": 1.4723, "step": 5120 }, { "epoch": 0.13, "grad_norm": 41.25, "learning_rate": 1.6861016949152544e-05, "loss": 1.5995, "step": 5130 }, { "epoch": 0.13, "grad_norm": 15.3125, "learning_rate": 1.6854237288135594e-05, "loss": 1.416, "step": 5140 }, { "epoch": 0.13, "grad_norm": 14.3125, "learning_rate": 1.6847457627118647e-05, "loss": 1.4539, "step": 5150 }, { "epoch": 0.13, "grad_norm": 18.0, "learning_rate": 1.6840677966101697e-05, "loss": 1.5037, "step": 5160 }, { "epoch": 0.13, "grad_norm": 10.0, "learning_rate": 1.683389830508475e-05, "loss": 1.4723, "step": 5170 }, { "epoch": 0.13, "grad_norm": 13.8125, "learning_rate": 1.6827118644067796e-05, "loss": 1.5438, "step": 5180 }, { "epoch": 0.13, "grad_norm": 22.5, "learning_rate": 1.682033898305085e-05, "loss": 1.4839, "step": 5190 }, { "epoch": 0.13, "grad_norm": 15.75, "learning_rate": 1.68135593220339e-05, "loss": 1.4003, "step": 5200 }, { "epoch": 0.13, "grad_norm": 12.1875, "learning_rate": 1.680677966101695e-05, "loss": 1.4559, "step": 5210 }, { "epoch": 0.13, "grad_norm": 26.875, "learning_rate": 1.6800000000000002e-05, "loss": 1.3201, "step": 5220 }, { "epoch": 0.13, "grad_norm": 21.75, "learning_rate": 1.679322033898305e-05, "loss": 1.5647, "step": 5230 }, { "epoch": 0.13, "grad_norm": 10.75, "learning_rate": 1.67864406779661e-05, "loss": 1.4694, "step": 5240 }, { "epoch": 0.13, "grad_norm": 20.0, "learning_rate": 1.6779661016949154e-05, "loss": 1.5724, "step": 5250 }, { "epoch": 0.13, "grad_norm": 13.0, "learning_rate": 1.6772881355932204e-05, "loss": 1.4456, "step": 5260 }, { "epoch": 0.13, "grad_norm": 16.375, "learning_rate": 1.6766101694915257e-05, "loss": 1.3278, "step": 5270 }, { "epoch": 0.13, "grad_norm": 14.3125, "learning_rate": 1.6759322033898307e-05, "loss": 1.4132, "step": 5280 }, { "epoch": 0.13, "grad_norm": 12.0, "learning_rate": 1.6752542372881357e-05, "loss": 1.3619, "step": 5290 }, { "epoch": 0.13, "grad_norm": 21.25, "learning_rate": 1.674576271186441e-05, "loss": 1.3347, "step": 5300 }, { "epoch": 0.13, "grad_norm": 20.75, "learning_rate": 1.673898305084746e-05, "loss": 1.471, "step": 5310 }, { "epoch": 0.13, "grad_norm": 12.75, "learning_rate": 1.673220338983051e-05, "loss": 1.5976, "step": 5320 }, { "epoch": 0.13, "grad_norm": 11.9375, "learning_rate": 1.6725423728813562e-05, "loss": 1.466, "step": 5330 }, { "epoch": 0.13, "grad_norm": 8.0, "learning_rate": 1.6718644067796612e-05, "loss": 1.5438, "step": 5340 }, { "epoch": 0.13, "grad_norm": 9.9375, "learning_rate": 1.671186440677966e-05, "loss": 1.515, "step": 5350 }, { "epoch": 0.13, "grad_norm": 16.75, "learning_rate": 1.6705084745762715e-05, "loss": 1.4414, "step": 5360 }, { "epoch": 0.13, "grad_norm": 9.5625, "learning_rate": 1.6698305084745764e-05, "loss": 1.4415, "step": 5370 }, { "epoch": 0.13, "grad_norm": 11.25, "learning_rate": 1.6691525423728817e-05, "loss": 1.4419, "step": 5380 }, { "epoch": 0.13, "grad_norm": 11.8125, "learning_rate": 1.6684745762711864e-05, "loss": 1.3952, "step": 5390 }, { "epoch": 0.14, "grad_norm": 12.6875, "learning_rate": 1.6677966101694917e-05, "loss": 1.4437, "step": 5400 }, { "epoch": 0.14, "grad_norm": 9.8125, "learning_rate": 1.6671186440677967e-05, "loss": 1.2443, "step": 5410 }, { "epoch": 0.14, "grad_norm": 14.625, "learning_rate": 1.6664406779661016e-05, "loss": 1.5725, "step": 5420 }, { "epoch": 0.14, "grad_norm": 14.0, "learning_rate": 1.665762711864407e-05, "loss": 1.3723, "step": 5430 }, { "epoch": 0.14, "grad_norm": 12.0, "learning_rate": 1.665084745762712e-05, "loss": 1.4005, "step": 5440 }, { "epoch": 0.14, "grad_norm": 13.875, "learning_rate": 1.6644067796610172e-05, "loss": 1.4065, "step": 5450 }, { "epoch": 0.14, "grad_norm": 12.3125, "learning_rate": 1.6637288135593222e-05, "loss": 1.3542, "step": 5460 }, { "epoch": 0.14, "grad_norm": 18.375, "learning_rate": 1.663050847457627e-05, "loss": 1.5096, "step": 5470 }, { "epoch": 0.14, "grad_norm": 15.625, "learning_rate": 1.6623728813559325e-05, "loss": 1.3777, "step": 5480 }, { "epoch": 0.14, "grad_norm": 12.3125, "learning_rate": 1.6616949152542374e-05, "loss": 1.4372, "step": 5490 }, { "epoch": 0.14, "grad_norm": 14.25, "learning_rate": 1.6610169491525424e-05, "loss": 1.5311, "step": 5500 }, { "epoch": 0.14, "grad_norm": 9.125, "learning_rate": 1.6603389830508477e-05, "loss": 1.2314, "step": 5510 }, { "epoch": 0.14, "grad_norm": 11.75, "learning_rate": 1.6596610169491527e-05, "loss": 1.587, "step": 5520 }, { "epoch": 0.14, "grad_norm": 9.4375, "learning_rate": 1.6589830508474577e-05, "loss": 1.4027, "step": 5530 }, { "epoch": 0.14, "grad_norm": 12.125, "learning_rate": 1.658305084745763e-05, "loss": 1.3296, "step": 5540 }, { "epoch": 0.14, "grad_norm": 15.625, "learning_rate": 1.657627118644068e-05, "loss": 1.3901, "step": 5550 }, { "epoch": 0.14, "grad_norm": 20.5, "learning_rate": 1.6569491525423732e-05, "loss": 1.5559, "step": 5560 }, { "epoch": 0.14, "grad_norm": 16.125, "learning_rate": 1.6562711864406782e-05, "loss": 1.455, "step": 5570 }, { "epoch": 0.14, "grad_norm": 17.0, "learning_rate": 1.6555932203389832e-05, "loss": 1.3029, "step": 5580 }, { "epoch": 0.14, "grad_norm": 8.25, "learning_rate": 1.654915254237288e-05, "loss": 1.3962, "step": 5590 }, { "epoch": 0.14, "grad_norm": 8.125, "learning_rate": 1.654237288135593e-05, "loss": 1.3684, "step": 5600 }, { "epoch": 0.14, "grad_norm": 10.25, "learning_rate": 1.6535593220338984e-05, "loss": 1.3081, "step": 5610 }, { "epoch": 0.14, "grad_norm": 12.125, "learning_rate": 1.6528813559322034e-05, "loss": 1.4176, "step": 5620 }, { "epoch": 0.14, "grad_norm": 11.5625, "learning_rate": 1.6522033898305084e-05, "loss": 1.4085, "step": 5630 }, { "epoch": 0.14, "grad_norm": 14.1875, "learning_rate": 1.6515254237288137e-05, "loss": 1.5883, "step": 5640 }, { "epoch": 0.14, "grad_norm": 19.875, "learning_rate": 1.6508474576271187e-05, "loss": 1.4177, "step": 5650 }, { "epoch": 0.14, "grad_norm": 9.1875, "learning_rate": 1.650169491525424e-05, "loss": 1.4454, "step": 5660 }, { "epoch": 0.14, "grad_norm": 10.3125, "learning_rate": 1.649491525423729e-05, "loss": 1.4518, "step": 5670 }, { "epoch": 0.14, "grad_norm": 11.9375, "learning_rate": 1.648813559322034e-05, "loss": 1.4185, "step": 5680 }, { "epoch": 0.14, "grad_norm": 12.25, "learning_rate": 1.6481355932203392e-05, "loss": 1.5352, "step": 5690 }, { "epoch": 0.14, "grad_norm": 24.5, "learning_rate": 1.6474576271186442e-05, "loss": 1.5149, "step": 5700 }, { "epoch": 0.14, "grad_norm": 60.5, "learning_rate": 1.646779661016949e-05, "loss": 1.4459, "step": 5710 }, { "epoch": 0.14, "grad_norm": 15.0625, "learning_rate": 1.6461016949152545e-05, "loss": 1.5419, "step": 5720 }, { "epoch": 0.14, "grad_norm": 17.5, "learning_rate": 1.6454237288135594e-05, "loss": 1.3723, "step": 5730 }, { "epoch": 0.14, "grad_norm": 11.5625, "learning_rate": 1.6447457627118648e-05, "loss": 1.5236, "step": 5740 }, { "epoch": 0.14, "grad_norm": 13.25, "learning_rate": 1.6440677966101697e-05, "loss": 1.554, "step": 5750 }, { "epoch": 0.14, "grad_norm": 12.1875, "learning_rate": 1.6433898305084747e-05, "loss": 1.4347, "step": 5760 }, { "epoch": 0.14, "grad_norm": 10.125, "learning_rate": 1.64271186440678e-05, "loss": 1.56, "step": 5770 }, { "epoch": 0.14, "grad_norm": 14.8125, "learning_rate": 1.642033898305085e-05, "loss": 1.2835, "step": 5780 }, { "epoch": 0.14, "grad_norm": 10.8125, "learning_rate": 1.64135593220339e-05, "loss": 1.3239, "step": 5790 }, { "epoch": 0.14, "grad_norm": 10.875, "learning_rate": 1.640677966101695e-05, "loss": 1.5509, "step": 5800 }, { "epoch": 0.15, "grad_norm": 15.75, "learning_rate": 1.64e-05, "loss": 1.4835, "step": 5810 }, { "epoch": 0.15, "grad_norm": 23.0, "learning_rate": 1.6393220338983052e-05, "loss": 1.5825, "step": 5820 }, { "epoch": 0.15, "grad_norm": 18.875, "learning_rate": 1.63864406779661e-05, "loss": 1.4388, "step": 5830 }, { "epoch": 0.15, "grad_norm": 18.125, "learning_rate": 1.6379661016949155e-05, "loss": 1.4779, "step": 5840 }, { "epoch": 0.15, "grad_norm": 13.8125, "learning_rate": 1.6372881355932204e-05, "loss": 1.5107, "step": 5850 }, { "epoch": 0.15, "grad_norm": 11.0, "learning_rate": 1.6366101694915254e-05, "loss": 1.3552, "step": 5860 }, { "epoch": 0.15, "grad_norm": 21.625, "learning_rate": 1.6359322033898307e-05, "loss": 1.396, "step": 5870 }, { "epoch": 0.15, "grad_norm": 8.0625, "learning_rate": 1.6352542372881357e-05, "loss": 1.3733, "step": 5880 }, { "epoch": 0.15, "grad_norm": 20.375, "learning_rate": 1.6345762711864407e-05, "loss": 1.6521, "step": 5890 }, { "epoch": 0.15, "grad_norm": 31.25, "learning_rate": 1.633898305084746e-05, "loss": 1.6624, "step": 5900 }, { "epoch": 0.15, "grad_norm": 11.125, "learning_rate": 1.633220338983051e-05, "loss": 1.4172, "step": 5910 }, { "epoch": 0.15, "grad_norm": 10.0625, "learning_rate": 1.6325423728813563e-05, "loss": 1.6346, "step": 5920 }, { "epoch": 0.15, "grad_norm": 20.875, "learning_rate": 1.6318644067796612e-05, "loss": 1.4104, "step": 5930 }, { "epoch": 0.15, "grad_norm": 15.375, "learning_rate": 1.6311864406779662e-05, "loss": 1.4004, "step": 5940 }, { "epoch": 0.15, "grad_norm": 12.5625, "learning_rate": 1.6305084745762715e-05, "loss": 1.5183, "step": 5950 }, { "epoch": 0.15, "grad_norm": 13.6875, "learning_rate": 1.6298305084745765e-05, "loss": 1.3157, "step": 5960 }, { "epoch": 0.15, "grad_norm": 14.5, "learning_rate": 1.6291525423728814e-05, "loss": 1.258, "step": 5970 }, { "epoch": 0.15, "grad_norm": 24.25, "learning_rate": 1.6284745762711868e-05, "loss": 1.4648, "step": 5980 }, { "epoch": 0.15, "grad_norm": 12.5625, "learning_rate": 1.6277966101694917e-05, "loss": 1.4231, "step": 5990 }, { "epoch": 0.15, "grad_norm": 13.0625, "learning_rate": 1.6271186440677967e-05, "loss": 1.3304, "step": 6000 }, { "epoch": 0.15, "grad_norm": 25.75, "learning_rate": 1.6264406779661017e-05, "loss": 1.4515, "step": 6010 }, { "epoch": 0.15, "grad_norm": 12.125, "learning_rate": 1.625762711864407e-05, "loss": 1.5502, "step": 6020 }, { "epoch": 0.15, "grad_norm": 10.625, "learning_rate": 1.625084745762712e-05, "loss": 1.4834, "step": 6030 }, { "epoch": 0.15, "grad_norm": 8.4375, "learning_rate": 1.624406779661017e-05, "loss": 1.344, "step": 6040 }, { "epoch": 0.15, "grad_norm": 17.0, "learning_rate": 1.6237288135593222e-05, "loss": 1.3194, "step": 6050 }, { "epoch": 0.15, "grad_norm": 7.25, "learning_rate": 1.6230508474576272e-05, "loss": 1.3901, "step": 6060 }, { "epoch": 0.15, "grad_norm": 21.375, "learning_rate": 1.6223728813559322e-05, "loss": 1.3026, "step": 6070 }, { "epoch": 0.15, "grad_norm": 17.25, "learning_rate": 1.6216949152542375e-05, "loss": 1.5712, "step": 6080 }, { "epoch": 0.15, "grad_norm": 22.875, "learning_rate": 1.6210169491525424e-05, "loss": 1.3923, "step": 6090 }, { "epoch": 0.15, "grad_norm": 21.5, "learning_rate": 1.6203389830508474e-05, "loss": 1.3557, "step": 6100 }, { "epoch": 0.15, "grad_norm": 16.25, "learning_rate": 1.6196610169491527e-05, "loss": 1.5063, "step": 6110 }, { "epoch": 0.15, "grad_norm": 13.3125, "learning_rate": 1.6189830508474577e-05, "loss": 1.5183, "step": 6120 }, { "epoch": 0.15, "grad_norm": 11.875, "learning_rate": 1.618305084745763e-05, "loss": 1.4881, "step": 6130 }, { "epoch": 0.15, "grad_norm": 15.5, "learning_rate": 1.617627118644068e-05, "loss": 1.4537, "step": 6140 }, { "epoch": 0.15, "grad_norm": 12.3125, "learning_rate": 1.616949152542373e-05, "loss": 1.4031, "step": 6150 }, { "epoch": 0.15, "grad_norm": 14.0, "learning_rate": 1.6162711864406783e-05, "loss": 1.4161, "step": 6160 }, { "epoch": 0.15, "grad_norm": 14.25, "learning_rate": 1.6155932203389832e-05, "loss": 1.4112, "step": 6170 }, { "epoch": 0.15, "grad_norm": 13.1875, "learning_rate": 1.6149152542372882e-05, "loss": 1.3729, "step": 6180 }, { "epoch": 0.15, "grad_norm": 8.3125, "learning_rate": 1.6142372881355935e-05, "loss": 1.2703, "step": 6190 }, { "epoch": 0.15, "grad_norm": 14.0, "learning_rate": 1.6135593220338985e-05, "loss": 1.5176, "step": 6200 }, { "epoch": 0.16, "grad_norm": 14.5, "learning_rate": 1.6128813559322038e-05, "loss": 1.451, "step": 6210 }, { "epoch": 0.16, "grad_norm": 18.25, "learning_rate": 1.6122033898305084e-05, "loss": 1.5495, "step": 6220 }, { "epoch": 0.16, "grad_norm": 11.25, "learning_rate": 1.6115254237288137e-05, "loss": 1.4713, "step": 6230 }, { "epoch": 0.16, "grad_norm": 20.375, "learning_rate": 1.6108474576271187e-05, "loss": 1.5256, "step": 6240 }, { "epoch": 0.16, "grad_norm": 24.375, "learning_rate": 1.6101694915254237e-05, "loss": 1.3237, "step": 6250 }, { "epoch": 0.16, "grad_norm": 16.625, "learning_rate": 1.609491525423729e-05, "loss": 1.5105, "step": 6260 }, { "epoch": 0.16, "grad_norm": 31.875, "learning_rate": 1.608813559322034e-05, "loss": 1.4395, "step": 6270 }, { "epoch": 0.16, "grad_norm": 14.6875, "learning_rate": 1.608135593220339e-05, "loss": 1.4263, "step": 6280 }, { "epoch": 0.16, "grad_norm": 13.875, "learning_rate": 1.6074576271186442e-05, "loss": 1.5833, "step": 6290 }, { "epoch": 0.16, "grad_norm": 12.5, "learning_rate": 1.6067796610169492e-05, "loss": 1.3931, "step": 6300 }, { "epoch": 0.16, "grad_norm": 17.25, "learning_rate": 1.6061016949152545e-05, "loss": 1.5431, "step": 6310 }, { "epoch": 0.16, "grad_norm": 22.25, "learning_rate": 1.6054237288135595e-05, "loss": 1.4012, "step": 6320 }, { "epoch": 0.16, "grad_norm": 11.8125, "learning_rate": 1.6047457627118645e-05, "loss": 1.4685, "step": 6330 }, { "epoch": 0.16, "grad_norm": 18.625, "learning_rate": 1.6040677966101698e-05, "loss": 1.6647, "step": 6340 }, { "epoch": 0.16, "grad_norm": 17.125, "learning_rate": 1.6033898305084747e-05, "loss": 1.5383, "step": 6350 }, { "epoch": 0.16, "grad_norm": 8.0625, "learning_rate": 1.6027118644067797e-05, "loss": 1.5919, "step": 6360 }, { "epoch": 0.16, "grad_norm": 10.0625, "learning_rate": 1.602033898305085e-05, "loss": 1.4546, "step": 6370 }, { "epoch": 0.16, "grad_norm": 10.9375, "learning_rate": 1.60135593220339e-05, "loss": 1.5119, "step": 6380 }, { "epoch": 0.16, "grad_norm": 21.375, "learning_rate": 1.600677966101695e-05, "loss": 1.3847, "step": 6390 }, { "epoch": 0.16, "grad_norm": 14.5625, "learning_rate": 1.6000000000000003e-05, "loss": 1.6196, "step": 6400 }, { "epoch": 0.16, "grad_norm": 14.5, "learning_rate": 1.5993220338983052e-05, "loss": 1.3922, "step": 6410 }, { "epoch": 0.16, "grad_norm": 19.125, "learning_rate": 1.5986440677966105e-05, "loss": 1.522, "step": 6420 }, { "epoch": 0.16, "grad_norm": 15.6875, "learning_rate": 1.5979661016949152e-05, "loss": 1.4102, "step": 6430 }, { "epoch": 0.16, "grad_norm": 20.25, "learning_rate": 1.5972881355932205e-05, "loss": 1.4111, "step": 6440 }, { "epoch": 0.16, "grad_norm": 6.75, "learning_rate": 1.5966101694915255e-05, "loss": 1.528, "step": 6450 }, { "epoch": 0.16, "grad_norm": 10.1875, "learning_rate": 1.5959322033898304e-05, "loss": 1.3356, "step": 6460 }, { "epoch": 0.16, "grad_norm": 9.0, "learning_rate": 1.5952542372881357e-05, "loss": 1.4551, "step": 6470 }, { "epoch": 0.16, "grad_norm": 18.5, "learning_rate": 1.5945762711864407e-05, "loss": 1.478, "step": 6480 }, { "epoch": 0.16, "grad_norm": 24.5, "learning_rate": 1.593898305084746e-05, "loss": 1.4338, "step": 6490 }, { "epoch": 0.16, "grad_norm": 19.75, "learning_rate": 1.593220338983051e-05, "loss": 1.3614, "step": 6500 }, { "epoch": 0.16, "grad_norm": 13.5, "learning_rate": 1.592542372881356e-05, "loss": 1.5299, "step": 6510 }, { "epoch": 0.16, "grad_norm": 35.0, "learning_rate": 1.5918644067796613e-05, "loss": 1.6247, "step": 6520 }, { "epoch": 0.16, "grad_norm": 13.9375, "learning_rate": 1.5911864406779662e-05, "loss": 1.5331, "step": 6530 }, { "epoch": 0.16, "grad_norm": 15.5625, "learning_rate": 1.5905084745762712e-05, "loss": 1.2449, "step": 6540 }, { "epoch": 0.16, "grad_norm": 24.625, "learning_rate": 1.5898305084745765e-05, "loss": 1.6226, "step": 6550 }, { "epoch": 0.16, "grad_norm": 25.875, "learning_rate": 1.5891525423728815e-05, "loss": 1.2374, "step": 6560 }, { "epoch": 0.16, "grad_norm": 9.75, "learning_rate": 1.5884745762711865e-05, "loss": 1.4768, "step": 6570 }, { "epoch": 0.16, "grad_norm": 41.5, "learning_rate": 1.5877966101694918e-05, "loss": 1.6625, "step": 6580 }, { "epoch": 0.16, "grad_norm": 11.9375, "learning_rate": 1.5871186440677967e-05, "loss": 1.3106, "step": 6590 }, { "epoch": 0.17, "grad_norm": 17.25, "learning_rate": 1.586440677966102e-05, "loss": 1.5881, "step": 6600 }, { "epoch": 0.17, "grad_norm": 22.75, "learning_rate": 1.585762711864407e-05, "loss": 1.366, "step": 6610 }, { "epoch": 0.17, "grad_norm": 19.875, "learning_rate": 1.585084745762712e-05, "loss": 1.6092, "step": 6620 }, { "epoch": 0.17, "grad_norm": 14.125, "learning_rate": 1.584406779661017e-05, "loss": 1.4089, "step": 6630 }, { "epoch": 0.17, "grad_norm": 16.375, "learning_rate": 1.583728813559322e-05, "loss": 1.5134, "step": 6640 }, { "epoch": 0.17, "grad_norm": 14.4375, "learning_rate": 1.5830508474576272e-05, "loss": 1.5151, "step": 6650 }, { "epoch": 0.17, "grad_norm": 14.4375, "learning_rate": 1.5823728813559322e-05, "loss": 1.4323, "step": 6660 }, { "epoch": 0.17, "grad_norm": 34.25, "learning_rate": 1.5816949152542372e-05, "loss": 1.6351, "step": 6670 }, { "epoch": 0.17, "grad_norm": 13.9375, "learning_rate": 1.5810169491525425e-05, "loss": 1.5497, "step": 6680 }, { "epoch": 0.17, "grad_norm": 11.0, "learning_rate": 1.5803389830508475e-05, "loss": 1.6565, "step": 6690 }, { "epoch": 0.17, "grad_norm": 9.9375, "learning_rate": 1.5796610169491528e-05, "loss": 1.5064, "step": 6700 }, { "epoch": 0.17, "grad_norm": 24.375, "learning_rate": 1.5789830508474577e-05, "loss": 1.3691, "step": 6710 }, { "epoch": 0.17, "grad_norm": 17.75, "learning_rate": 1.5783050847457627e-05, "loss": 1.3609, "step": 6720 }, { "epoch": 0.17, "grad_norm": 9.0625, "learning_rate": 1.577627118644068e-05, "loss": 1.4671, "step": 6730 }, { "epoch": 0.17, "grad_norm": 16.625, "learning_rate": 1.576949152542373e-05, "loss": 1.5389, "step": 6740 }, { "epoch": 0.17, "grad_norm": 10.4375, "learning_rate": 1.576271186440678e-05, "loss": 1.3601, "step": 6750 }, { "epoch": 0.17, "grad_norm": 21.75, "learning_rate": 1.5755932203389833e-05, "loss": 1.4282, "step": 6760 }, { "epoch": 0.17, "grad_norm": 21.75, "learning_rate": 1.5749152542372882e-05, "loss": 1.3492, "step": 6770 }, { "epoch": 0.17, "grad_norm": 11.1875, "learning_rate": 1.5742372881355936e-05, "loss": 1.5076, "step": 6780 }, { "epoch": 0.17, "grad_norm": 13.6875, "learning_rate": 1.5735593220338985e-05, "loss": 1.4347, "step": 6790 }, { "epoch": 0.17, "grad_norm": 13.5, "learning_rate": 1.5728813559322035e-05, "loss": 1.5, "step": 6800 }, { "epoch": 0.17, "grad_norm": 23.125, "learning_rate": 1.5722033898305088e-05, "loss": 1.2902, "step": 6810 }, { "epoch": 0.17, "grad_norm": 16.125, "learning_rate": 1.5715254237288138e-05, "loss": 1.4831, "step": 6820 }, { "epoch": 0.17, "grad_norm": 19.5, "learning_rate": 1.5708474576271187e-05, "loss": 1.501, "step": 6830 }, { "epoch": 0.17, "grad_norm": 20.125, "learning_rate": 1.5701694915254237e-05, "loss": 1.4927, "step": 6840 }, { "epoch": 0.17, "grad_norm": 9.0625, "learning_rate": 1.5694915254237287e-05, "loss": 1.51, "step": 6850 }, { "epoch": 0.17, "grad_norm": 10.6875, "learning_rate": 1.568813559322034e-05, "loss": 1.5486, "step": 6860 }, { "epoch": 0.17, "grad_norm": 12.5625, "learning_rate": 1.568135593220339e-05, "loss": 1.4239, "step": 6870 }, { "epoch": 0.17, "grad_norm": 11.6875, "learning_rate": 1.5674576271186443e-05, "loss": 1.2807, "step": 6880 }, { "epoch": 0.17, "grad_norm": 14.125, "learning_rate": 1.5667796610169492e-05, "loss": 1.426, "step": 6890 }, { "epoch": 0.17, "grad_norm": 28.5, "learning_rate": 1.5661016949152542e-05, "loss": 1.4277, "step": 6900 }, { "epoch": 0.17, "grad_norm": 17.125, "learning_rate": 1.5654237288135595e-05, "loss": 1.499, "step": 6910 }, { "epoch": 0.17, "grad_norm": 26.5, "learning_rate": 1.5647457627118645e-05, "loss": 1.3421, "step": 6920 }, { "epoch": 0.17, "grad_norm": 18.875, "learning_rate": 1.5640677966101695e-05, "loss": 1.5728, "step": 6930 }, { "epoch": 0.17, "grad_norm": 26.5, "learning_rate": 1.5633898305084748e-05, "loss": 1.2488, "step": 6940 }, { "epoch": 0.17, "grad_norm": 12.5625, "learning_rate": 1.5627118644067798e-05, "loss": 1.4965, "step": 6950 }, { "epoch": 0.17, "grad_norm": 15.5, "learning_rate": 1.562033898305085e-05, "loss": 1.3805, "step": 6960 }, { "epoch": 0.17, "grad_norm": 14.125, "learning_rate": 1.56135593220339e-05, "loss": 1.4851, "step": 6970 }, { "epoch": 0.17, "grad_norm": 13.375, "learning_rate": 1.560677966101695e-05, "loss": 1.4168, "step": 6980 }, { "epoch": 0.17, "grad_norm": 8.1875, "learning_rate": 1.5600000000000003e-05, "loss": 1.4241, "step": 6990 }, { "epoch": 0.17, "grad_norm": 12.25, "learning_rate": 1.5593220338983053e-05, "loss": 1.3725, "step": 7000 }, { "epoch": 0.18, "grad_norm": 21.375, "learning_rate": 1.5586440677966103e-05, "loss": 1.2136, "step": 7010 }, { "epoch": 0.18, "grad_norm": 18.25, "learning_rate": 1.5579661016949156e-05, "loss": 1.4192, "step": 7020 }, { "epoch": 0.18, "grad_norm": 23.125, "learning_rate": 1.5572881355932205e-05, "loss": 1.4821, "step": 7030 }, { "epoch": 0.18, "grad_norm": 6.21875, "learning_rate": 1.5566101694915255e-05, "loss": 1.5629, "step": 7040 }, { "epoch": 0.18, "grad_norm": 15.6875, "learning_rate": 1.5559322033898305e-05, "loss": 1.4349, "step": 7050 }, { "epoch": 0.18, "grad_norm": 10.125, "learning_rate": 1.5552542372881358e-05, "loss": 1.3712, "step": 7060 }, { "epoch": 0.18, "grad_norm": 22.75, "learning_rate": 1.5545762711864408e-05, "loss": 1.3625, "step": 7070 }, { "epoch": 0.18, "grad_norm": 12.0, "learning_rate": 1.5538983050847457e-05, "loss": 1.4636, "step": 7080 }, { "epoch": 0.18, "grad_norm": 19.75, "learning_rate": 1.553220338983051e-05, "loss": 1.2732, "step": 7090 }, { "epoch": 0.18, "grad_norm": 8.375, "learning_rate": 1.552542372881356e-05, "loss": 1.3205, "step": 7100 }, { "epoch": 0.18, "grad_norm": 16.75, "learning_rate": 1.551864406779661e-05, "loss": 1.3579, "step": 7110 }, { "epoch": 0.18, "grad_norm": 8.0625, "learning_rate": 1.5511864406779663e-05, "loss": 1.4791, "step": 7120 }, { "epoch": 0.18, "grad_norm": 18.5, "learning_rate": 1.5505084745762713e-05, "loss": 1.4535, "step": 7130 }, { "epoch": 0.18, "grad_norm": 16.375, "learning_rate": 1.5498305084745762e-05, "loss": 1.4981, "step": 7140 }, { "epoch": 0.18, "grad_norm": 22.125, "learning_rate": 1.5491525423728815e-05, "loss": 1.456, "step": 7150 }, { "epoch": 0.18, "grad_norm": 24.375, "learning_rate": 1.5484745762711865e-05, "loss": 1.2791, "step": 7160 }, { "epoch": 0.18, "grad_norm": 19.875, "learning_rate": 1.5477966101694918e-05, "loss": 1.3166, "step": 7170 }, { "epoch": 0.18, "grad_norm": 7.03125, "learning_rate": 1.5471186440677968e-05, "loss": 1.3394, "step": 7180 }, { "epoch": 0.18, "grad_norm": 12.4375, "learning_rate": 1.5464406779661018e-05, "loss": 1.4032, "step": 7190 }, { "epoch": 0.18, "grad_norm": 11.0625, "learning_rate": 1.545762711864407e-05, "loss": 1.5416, "step": 7200 }, { "epoch": 0.18, "grad_norm": 10.8125, "learning_rate": 1.545084745762712e-05, "loss": 1.5153, "step": 7210 }, { "epoch": 0.18, "grad_norm": 8.4375, "learning_rate": 1.544406779661017e-05, "loss": 1.4548, "step": 7220 }, { "epoch": 0.18, "grad_norm": 25.75, "learning_rate": 1.5437288135593223e-05, "loss": 1.4448, "step": 7230 }, { "epoch": 0.18, "grad_norm": 16.875, "learning_rate": 1.5430508474576273e-05, "loss": 1.4032, "step": 7240 }, { "epoch": 0.18, "grad_norm": 14.1875, "learning_rate": 1.5423728813559326e-05, "loss": 1.3053, "step": 7250 }, { "epoch": 0.18, "grad_norm": 23.375, "learning_rate": 1.5416949152542372e-05, "loss": 1.508, "step": 7260 }, { "epoch": 0.18, "grad_norm": 7.5, "learning_rate": 1.5410169491525425e-05, "loss": 1.5071, "step": 7270 }, { "epoch": 0.18, "grad_norm": 17.75, "learning_rate": 1.5403389830508475e-05, "loss": 1.3446, "step": 7280 }, { "epoch": 0.18, "grad_norm": 16.125, "learning_rate": 1.5396610169491525e-05, "loss": 1.3282, "step": 7290 }, { "epoch": 0.18, "grad_norm": 13.5625, "learning_rate": 1.5389830508474578e-05, "loss": 1.4596, "step": 7300 }, { "epoch": 0.18, "grad_norm": 23.0, "learning_rate": 1.5383050847457628e-05, "loss": 1.3568, "step": 7310 }, { "epoch": 0.18, "grad_norm": 30.875, "learning_rate": 1.5376271186440677e-05, "loss": 1.4984, "step": 7320 }, { "epoch": 0.18, "grad_norm": 31.125, "learning_rate": 1.536949152542373e-05, "loss": 1.2972, "step": 7330 }, { "epoch": 0.18, "grad_norm": 19.375, "learning_rate": 1.536271186440678e-05, "loss": 1.2122, "step": 7340 }, { "epoch": 0.18, "grad_norm": 18.875, "learning_rate": 1.5355932203389833e-05, "loss": 1.3889, "step": 7350 }, { "epoch": 0.18, "grad_norm": 15.3125, "learning_rate": 1.5349152542372883e-05, "loss": 1.3616, "step": 7360 }, { "epoch": 0.18, "grad_norm": 10.75, "learning_rate": 1.5342372881355933e-05, "loss": 1.4538, "step": 7370 }, { "epoch": 0.18, "grad_norm": 18.125, "learning_rate": 1.5335593220338986e-05, "loss": 1.4689, "step": 7380 }, { "epoch": 0.18, "grad_norm": 22.125, "learning_rate": 1.5328813559322035e-05, "loss": 1.3906, "step": 7390 }, { "epoch": 0.18, "grad_norm": 12.8125, "learning_rate": 1.5322033898305085e-05, "loss": 1.3831, "step": 7400 }, { "epoch": 0.19, "grad_norm": 20.25, "learning_rate": 1.5315254237288138e-05, "loss": 1.3703, "step": 7410 }, { "epoch": 0.19, "grad_norm": 17.75, "learning_rate": 1.5308474576271188e-05, "loss": 1.4825, "step": 7420 }, { "epoch": 0.19, "grad_norm": 11.0625, "learning_rate": 1.530169491525424e-05, "loss": 1.5844, "step": 7430 }, { "epoch": 0.19, "grad_norm": 23.25, "learning_rate": 1.529491525423729e-05, "loss": 1.5891, "step": 7440 }, { "epoch": 0.19, "grad_norm": 20.0, "learning_rate": 1.528813559322034e-05, "loss": 1.4705, "step": 7450 }, { "epoch": 0.19, "grad_norm": 8.6875, "learning_rate": 1.528135593220339e-05, "loss": 1.4058, "step": 7460 }, { "epoch": 0.19, "grad_norm": 29.125, "learning_rate": 1.527457627118644e-05, "loss": 1.4728, "step": 7470 }, { "epoch": 0.19, "grad_norm": 33.0, "learning_rate": 1.5267796610169493e-05, "loss": 1.4388, "step": 7480 }, { "epoch": 0.19, "grad_norm": 11.4375, "learning_rate": 1.5261016949152543e-05, "loss": 1.4398, "step": 7490 }, { "epoch": 0.19, "grad_norm": 11.4375, "learning_rate": 1.5254237288135594e-05, "loss": 1.3666, "step": 7500 }, { "epoch": 0.19, "grad_norm": 20.875, "learning_rate": 1.5247457627118645e-05, "loss": 1.4701, "step": 7510 }, { "epoch": 0.19, "grad_norm": 32.0, "learning_rate": 1.5240677966101695e-05, "loss": 1.3738, "step": 7520 }, { "epoch": 0.19, "grad_norm": 11.4375, "learning_rate": 1.5233898305084747e-05, "loss": 1.5328, "step": 7530 }, { "epoch": 0.19, "grad_norm": 7.84375, "learning_rate": 1.5227118644067798e-05, "loss": 1.3978, "step": 7540 }, { "epoch": 0.19, "grad_norm": 14.3125, "learning_rate": 1.522033898305085e-05, "loss": 1.2389, "step": 7550 }, { "epoch": 0.19, "grad_norm": 19.125, "learning_rate": 1.5213559322033899e-05, "loss": 1.4894, "step": 7560 }, { "epoch": 0.19, "grad_norm": 8.0625, "learning_rate": 1.520677966101695e-05, "loss": 1.3802, "step": 7570 }, { "epoch": 0.19, "grad_norm": 11.875, "learning_rate": 1.5200000000000002e-05, "loss": 1.5854, "step": 7580 }, { "epoch": 0.19, "grad_norm": 40.5, "learning_rate": 1.5193220338983052e-05, "loss": 1.4176, "step": 7590 }, { "epoch": 0.19, "grad_norm": 13.625, "learning_rate": 1.5186440677966103e-05, "loss": 1.3442, "step": 7600 }, { "epoch": 0.19, "grad_norm": 17.5, "learning_rate": 1.5179661016949154e-05, "loss": 1.5505, "step": 7610 }, { "epoch": 0.19, "grad_norm": 16.25, "learning_rate": 1.5172881355932206e-05, "loss": 1.3115, "step": 7620 }, { "epoch": 0.19, "grad_norm": 16.125, "learning_rate": 1.5166101694915255e-05, "loss": 1.4113, "step": 7630 }, { "epoch": 0.19, "grad_norm": 6.375, "learning_rate": 1.5159322033898307e-05, "loss": 1.3821, "step": 7640 }, { "epoch": 0.19, "grad_norm": 15.125, "learning_rate": 1.5152542372881358e-05, "loss": 1.5575, "step": 7650 }, { "epoch": 0.19, "grad_norm": 14.625, "learning_rate": 1.514576271186441e-05, "loss": 1.3458, "step": 7660 }, { "epoch": 0.19, "grad_norm": 15.875, "learning_rate": 1.5138983050847458e-05, "loss": 1.5992, "step": 7670 }, { "epoch": 0.19, "grad_norm": 18.875, "learning_rate": 1.5132203389830509e-05, "loss": 1.5372, "step": 7680 }, { "epoch": 0.19, "grad_norm": 11.375, "learning_rate": 1.512542372881356e-05, "loss": 1.4866, "step": 7690 }, { "epoch": 0.19, "grad_norm": 12.25, "learning_rate": 1.511864406779661e-05, "loss": 1.2221, "step": 7700 }, { "epoch": 0.19, "grad_norm": 21.75, "learning_rate": 1.5111864406779662e-05, "loss": 1.2278, "step": 7710 }, { "epoch": 0.19, "grad_norm": 26.5, "learning_rate": 1.5105084745762713e-05, "loss": 1.4566, "step": 7720 }, { "epoch": 0.19, "grad_norm": 20.25, "learning_rate": 1.5098305084745763e-05, "loss": 1.4411, "step": 7730 }, { "epoch": 0.19, "grad_norm": 27.125, "learning_rate": 1.5091525423728814e-05, "loss": 1.4632, "step": 7740 }, { "epoch": 0.19, "grad_norm": 13.9375, "learning_rate": 1.5084745762711865e-05, "loss": 1.3873, "step": 7750 }, { "epoch": 0.19, "grad_norm": 15.5625, "learning_rate": 1.5077966101694917e-05, "loss": 1.5183, "step": 7760 }, { "epoch": 0.19, "grad_norm": 10.8125, "learning_rate": 1.5071186440677967e-05, "loss": 1.4366, "step": 7770 }, { "epoch": 0.19, "grad_norm": 18.875, "learning_rate": 1.5064406779661018e-05, "loss": 1.2684, "step": 7780 }, { "epoch": 0.19, "grad_norm": 33.75, "learning_rate": 1.505762711864407e-05, "loss": 1.2841, "step": 7790 }, { "epoch": 0.2, "grad_norm": 21.625, "learning_rate": 1.505084745762712e-05, "loss": 1.4256, "step": 7800 }, { "epoch": 0.2, "grad_norm": 44.0, "learning_rate": 1.504406779661017e-05, "loss": 1.2481, "step": 7810 }, { "epoch": 0.2, "grad_norm": 14.1875, "learning_rate": 1.5037288135593222e-05, "loss": 1.4066, "step": 7820 }, { "epoch": 0.2, "grad_norm": 12.0625, "learning_rate": 1.5030508474576273e-05, "loss": 1.4502, "step": 7830 }, { "epoch": 0.2, "grad_norm": 13.3125, "learning_rate": 1.5023728813559325e-05, "loss": 1.4257, "step": 7840 }, { "epoch": 0.2, "grad_norm": 15.875, "learning_rate": 1.5016949152542374e-05, "loss": 1.2423, "step": 7850 }, { "epoch": 0.2, "grad_norm": 27.125, "learning_rate": 1.5010169491525426e-05, "loss": 1.4682, "step": 7860 }, { "epoch": 0.2, "grad_norm": 12.375, "learning_rate": 1.5003389830508477e-05, "loss": 1.5076, "step": 7870 }, { "epoch": 0.2, "grad_norm": 14.6875, "learning_rate": 1.4996610169491525e-05, "loss": 1.4543, "step": 7880 }, { "epoch": 0.2, "grad_norm": 13.25, "learning_rate": 1.4989830508474577e-05, "loss": 1.3303, "step": 7890 }, { "epoch": 0.2, "grad_norm": 12.875, "learning_rate": 1.4983050847457628e-05, "loss": 1.4475, "step": 7900 }, { "epoch": 0.2, "grad_norm": 17.125, "learning_rate": 1.4976271186440678e-05, "loss": 1.4678, "step": 7910 }, { "epoch": 0.2, "grad_norm": 12.4375, "learning_rate": 1.4969491525423729e-05, "loss": 1.1695, "step": 7920 }, { "epoch": 0.2, "grad_norm": 12.375, "learning_rate": 1.496271186440678e-05, "loss": 1.4393, "step": 7930 }, { "epoch": 0.2, "grad_norm": 10.625, "learning_rate": 1.4955932203389832e-05, "loss": 1.52, "step": 7940 }, { "epoch": 0.2, "grad_norm": 20.375, "learning_rate": 1.4949152542372882e-05, "loss": 1.4157, "step": 7950 }, { "epoch": 0.2, "grad_norm": 38.0, "learning_rate": 1.4942372881355933e-05, "loss": 1.2901, "step": 7960 }, { "epoch": 0.2, "grad_norm": 11.625, "learning_rate": 1.4935593220338984e-05, "loss": 1.4463, "step": 7970 }, { "epoch": 0.2, "grad_norm": 20.875, "learning_rate": 1.4928813559322036e-05, "loss": 1.3633, "step": 7980 }, { "epoch": 0.2, "grad_norm": 17.5, "learning_rate": 1.4922033898305086e-05, "loss": 1.4328, "step": 7990 }, { "epoch": 0.2, "grad_norm": 31.125, "learning_rate": 1.4915254237288137e-05, "loss": 1.5899, "step": 8000 }, { "epoch": 0.2, "grad_norm": 9.125, "learning_rate": 1.4908474576271188e-05, "loss": 1.4177, "step": 8010 }, { "epoch": 0.2, "grad_norm": 13.375, "learning_rate": 1.490169491525424e-05, "loss": 1.3138, "step": 8020 }, { "epoch": 0.2, "grad_norm": 27.5, "learning_rate": 1.489491525423729e-05, "loss": 1.3029, "step": 8030 }, { "epoch": 0.2, "grad_norm": 19.75, "learning_rate": 1.4888135593220341e-05, "loss": 1.4184, "step": 8040 }, { "epoch": 0.2, "grad_norm": 12.5, "learning_rate": 1.4881355932203392e-05, "loss": 1.4537, "step": 8050 }, { "epoch": 0.2, "grad_norm": 21.75, "learning_rate": 1.4874576271186442e-05, "loss": 1.411, "step": 8060 }, { "epoch": 0.2, "grad_norm": 23.0, "learning_rate": 1.4867796610169493e-05, "loss": 1.1435, "step": 8070 }, { "epoch": 0.2, "grad_norm": 15.25, "learning_rate": 1.4861016949152545e-05, "loss": 1.5546, "step": 8080 }, { "epoch": 0.2, "grad_norm": 16.625, "learning_rate": 1.4854237288135593e-05, "loss": 1.2875, "step": 8090 }, { "epoch": 0.2, "grad_norm": 33.5, "learning_rate": 1.4847457627118644e-05, "loss": 1.5006, "step": 8100 }, { "epoch": 0.2, "grad_norm": 8.9375, "learning_rate": 1.4840677966101696e-05, "loss": 1.5724, "step": 8110 }, { "epoch": 0.2, "grad_norm": 16.75, "learning_rate": 1.4833898305084747e-05, "loss": 1.5309, "step": 8120 }, { "epoch": 0.2, "grad_norm": 16.5, "learning_rate": 1.4827118644067797e-05, "loss": 1.4456, "step": 8130 }, { "epoch": 0.2, "grad_norm": 14.25, "learning_rate": 1.4820338983050848e-05, "loss": 1.3758, "step": 8140 }, { "epoch": 0.2, "grad_norm": 12.4375, "learning_rate": 1.48135593220339e-05, "loss": 1.5042, "step": 8150 }, { "epoch": 0.2, "grad_norm": 26.5, "learning_rate": 1.4806779661016951e-05, "loss": 1.4351, "step": 8160 }, { "epoch": 0.2, "grad_norm": 10.5625, "learning_rate": 1.48e-05, "loss": 1.2969, "step": 8170 }, { "epoch": 0.2, "grad_norm": 14.1875, "learning_rate": 1.4793220338983052e-05, "loss": 1.4189, "step": 8180 }, { "epoch": 0.2, "grad_norm": 21.875, "learning_rate": 1.4786440677966103e-05, "loss": 1.3903, "step": 8190 }, { "epoch": 0.2, "grad_norm": 12.0, "learning_rate": 1.4779661016949153e-05, "loss": 1.4782, "step": 8200 }, { "epoch": 0.21, "grad_norm": 12.5625, "learning_rate": 1.4772881355932205e-05, "loss": 1.4388, "step": 8210 }, { "epoch": 0.21, "grad_norm": 11.25, "learning_rate": 1.4766101694915256e-05, "loss": 1.2787, "step": 8220 }, { "epoch": 0.21, "grad_norm": 19.75, "learning_rate": 1.4759322033898307e-05, "loss": 1.3323, "step": 8230 }, { "epoch": 0.21, "grad_norm": 24.5, "learning_rate": 1.4752542372881357e-05, "loss": 1.3868, "step": 8240 }, { "epoch": 0.21, "grad_norm": 10.625, "learning_rate": 1.4745762711864408e-05, "loss": 1.3775, "step": 8250 }, { "epoch": 0.21, "grad_norm": 20.875, "learning_rate": 1.473898305084746e-05, "loss": 1.388, "step": 8260 }, { "epoch": 0.21, "grad_norm": 19.5, "learning_rate": 1.4732203389830511e-05, "loss": 1.3465, "step": 8270 }, { "epoch": 0.21, "grad_norm": 27.0, "learning_rate": 1.4725423728813561e-05, "loss": 1.4667, "step": 8280 }, { "epoch": 0.21, "grad_norm": 12.375, "learning_rate": 1.4718644067796612e-05, "loss": 1.4507, "step": 8290 }, { "epoch": 0.21, "grad_norm": 7.09375, "learning_rate": 1.4711864406779662e-05, "loss": 1.3994, "step": 8300 }, { "epoch": 0.21, "grad_norm": 16.375, "learning_rate": 1.4705084745762712e-05, "loss": 1.382, "step": 8310 }, { "epoch": 0.21, "grad_norm": 26.25, "learning_rate": 1.4698305084745763e-05, "loss": 1.2215, "step": 8320 }, { "epoch": 0.21, "grad_norm": 9.4375, "learning_rate": 1.4691525423728815e-05, "loss": 1.1387, "step": 8330 }, { "epoch": 0.21, "grad_norm": 19.5, "learning_rate": 1.4684745762711864e-05, "loss": 1.3104, "step": 8340 }, { "epoch": 0.21, "grad_norm": 22.875, "learning_rate": 1.4677966101694916e-05, "loss": 1.392, "step": 8350 }, { "epoch": 0.21, "grad_norm": 30.875, "learning_rate": 1.4671186440677967e-05, "loss": 1.4324, "step": 8360 }, { "epoch": 0.21, "grad_norm": 7.5, "learning_rate": 1.4664406779661018e-05, "loss": 1.3885, "step": 8370 }, { "epoch": 0.21, "grad_norm": 24.125, "learning_rate": 1.4657627118644068e-05, "loss": 1.2777, "step": 8380 }, { "epoch": 0.21, "grad_norm": 21.0, "learning_rate": 1.465084745762712e-05, "loss": 1.2714, "step": 8390 }, { "epoch": 0.21, "grad_norm": 10.1875, "learning_rate": 1.4644067796610171e-05, "loss": 1.5134, "step": 8400 }, { "epoch": 0.21, "grad_norm": 23.125, "learning_rate": 1.4637288135593222e-05, "loss": 1.3744, "step": 8410 }, { "epoch": 0.21, "grad_norm": 12.4375, "learning_rate": 1.4630508474576272e-05, "loss": 1.376, "step": 8420 }, { "epoch": 0.21, "grad_norm": 17.375, "learning_rate": 1.4623728813559323e-05, "loss": 1.4146, "step": 8430 }, { "epoch": 0.21, "grad_norm": 14.4375, "learning_rate": 1.4616949152542375e-05, "loss": 1.5153, "step": 8440 }, { "epoch": 0.21, "grad_norm": 22.25, "learning_rate": 1.4610169491525426e-05, "loss": 1.4103, "step": 8450 }, { "epoch": 0.21, "grad_norm": 12.1875, "learning_rate": 1.4603389830508476e-05, "loss": 1.2765, "step": 8460 }, { "epoch": 0.21, "grad_norm": 8.6875, "learning_rate": 1.4596610169491527e-05, "loss": 1.2245, "step": 8470 }, { "epoch": 0.21, "grad_norm": 14.0, "learning_rate": 1.4589830508474579e-05, "loss": 1.4353, "step": 8480 }, { "epoch": 0.21, "grad_norm": 14.5625, "learning_rate": 1.458305084745763e-05, "loss": 1.3907, "step": 8490 }, { "epoch": 0.21, "grad_norm": 18.125, "learning_rate": 1.4576271186440678e-05, "loss": 1.4604, "step": 8500 }, { "epoch": 0.21, "grad_norm": 12.25, "learning_rate": 1.456949152542373e-05, "loss": 1.3923, "step": 8510 }, { "epoch": 0.21, "grad_norm": 7.90625, "learning_rate": 1.456271186440678e-05, "loss": 1.3248, "step": 8520 }, { "epoch": 0.21, "grad_norm": 25.875, "learning_rate": 1.455593220338983e-05, "loss": 1.4283, "step": 8530 }, { "epoch": 0.21, "grad_norm": 12.5625, "learning_rate": 1.4549152542372882e-05, "loss": 1.4287, "step": 8540 }, { "epoch": 0.21, "grad_norm": 19.875, "learning_rate": 1.4542372881355933e-05, "loss": 1.3291, "step": 8550 }, { "epoch": 0.21, "grad_norm": 13.6875, "learning_rate": 1.4535593220338983e-05, "loss": 1.4037, "step": 8560 }, { "epoch": 0.21, "grad_norm": 14.125, "learning_rate": 1.4528813559322035e-05, "loss": 1.3539, "step": 8570 }, { "epoch": 0.21, "grad_norm": 12.25, "learning_rate": 1.4522033898305086e-05, "loss": 1.4476, "step": 8580 }, { "epoch": 0.21, "grad_norm": 13.375, "learning_rate": 1.4515254237288137e-05, "loss": 1.4544, "step": 8590 }, { "epoch": 0.21, "grad_norm": 12.625, "learning_rate": 1.4508474576271187e-05, "loss": 1.5032, "step": 8600 }, { "epoch": 0.22, "grad_norm": 12.8125, "learning_rate": 1.4501694915254239e-05, "loss": 1.3221, "step": 8610 }, { "epoch": 0.22, "grad_norm": 48.75, "learning_rate": 1.449491525423729e-05, "loss": 1.2948, "step": 8620 }, { "epoch": 0.22, "grad_norm": 11.8125, "learning_rate": 1.4488135593220341e-05, "loss": 1.4861, "step": 8630 }, { "epoch": 0.22, "grad_norm": 12.25, "learning_rate": 1.4481355932203391e-05, "loss": 1.5319, "step": 8640 }, { "epoch": 0.22, "grad_norm": 10.8125, "learning_rate": 1.4474576271186442e-05, "loss": 1.4589, "step": 8650 }, { "epoch": 0.22, "grad_norm": 21.625, "learning_rate": 1.4467796610169494e-05, "loss": 1.4423, "step": 8660 }, { "epoch": 0.22, "grad_norm": 30.5, "learning_rate": 1.4461016949152544e-05, "loss": 1.39, "step": 8670 }, { "epoch": 0.22, "grad_norm": 27.25, "learning_rate": 1.4454237288135595e-05, "loss": 1.5098, "step": 8680 }, { "epoch": 0.22, "grad_norm": 10.5625, "learning_rate": 1.4447457627118646e-05, "loss": 1.5591, "step": 8690 }, { "epoch": 0.22, "grad_norm": 10.125, "learning_rate": 1.4440677966101698e-05, "loss": 1.4686, "step": 8700 }, { "epoch": 0.22, "grad_norm": 24.25, "learning_rate": 1.4433898305084746e-05, "loss": 1.4181, "step": 8710 }, { "epoch": 0.22, "grad_norm": 24.375, "learning_rate": 1.4427118644067797e-05, "loss": 1.3366, "step": 8720 }, { "epoch": 0.22, "grad_norm": 20.125, "learning_rate": 1.4420338983050849e-05, "loss": 1.4106, "step": 8730 }, { "epoch": 0.22, "grad_norm": 13.4375, "learning_rate": 1.4413559322033898e-05, "loss": 1.4431, "step": 8740 }, { "epoch": 0.22, "grad_norm": 16.625, "learning_rate": 1.440677966101695e-05, "loss": 1.5069, "step": 8750 }, { "epoch": 0.22, "grad_norm": 17.125, "learning_rate": 1.4400000000000001e-05, "loss": 1.5823, "step": 8760 }, { "epoch": 0.22, "grad_norm": 11.0, "learning_rate": 1.4393220338983052e-05, "loss": 1.5015, "step": 8770 }, { "epoch": 0.22, "grad_norm": 11.25, "learning_rate": 1.4386440677966102e-05, "loss": 1.3885, "step": 8780 }, { "epoch": 0.22, "grad_norm": 69.5, "learning_rate": 1.4379661016949154e-05, "loss": 1.3728, "step": 8790 }, { "epoch": 0.22, "grad_norm": 13.25, "learning_rate": 1.4372881355932205e-05, "loss": 1.5583, "step": 8800 }, { "epoch": 0.22, "grad_norm": 17.5, "learning_rate": 1.4366101694915255e-05, "loss": 1.3834, "step": 8810 }, { "epoch": 0.22, "grad_norm": 13.3125, "learning_rate": 1.4359322033898306e-05, "loss": 1.4981, "step": 8820 }, { "epoch": 0.22, "grad_norm": 28.75, "learning_rate": 1.4352542372881357e-05, "loss": 1.2518, "step": 8830 }, { "epoch": 0.22, "grad_norm": 17.0, "learning_rate": 1.4345762711864409e-05, "loss": 1.3005, "step": 8840 }, { "epoch": 0.22, "grad_norm": 14.4375, "learning_rate": 1.4338983050847459e-05, "loss": 1.4738, "step": 8850 }, { "epoch": 0.22, "grad_norm": 13.3125, "learning_rate": 1.433220338983051e-05, "loss": 1.5208, "step": 8860 }, { "epoch": 0.22, "grad_norm": 32.25, "learning_rate": 1.4325423728813561e-05, "loss": 1.5122, "step": 8870 }, { "epoch": 0.22, "grad_norm": 16.625, "learning_rate": 1.4318644067796613e-05, "loss": 1.3556, "step": 8880 }, { "epoch": 0.22, "grad_norm": 13.125, "learning_rate": 1.4311864406779662e-05, "loss": 1.4243, "step": 8890 }, { "epoch": 0.22, "grad_norm": 14.6875, "learning_rate": 1.4305084745762714e-05, "loss": 1.4722, "step": 8900 }, { "epoch": 0.22, "grad_norm": 17.25, "learning_rate": 1.4298305084745765e-05, "loss": 1.4912, "step": 8910 }, { "epoch": 0.22, "grad_norm": 23.375, "learning_rate": 1.4291525423728813e-05, "loss": 1.2772, "step": 8920 }, { "epoch": 0.22, "grad_norm": 12.125, "learning_rate": 1.4284745762711865e-05, "loss": 1.2326, "step": 8930 }, { "epoch": 0.22, "grad_norm": 8.25, "learning_rate": 1.4277966101694916e-05, "loss": 1.2836, "step": 8940 }, { "epoch": 0.22, "grad_norm": 9.75, "learning_rate": 1.4271186440677966e-05, "loss": 1.6562, "step": 8950 }, { "epoch": 0.22, "grad_norm": 20.625, "learning_rate": 1.4264406779661017e-05, "loss": 1.377, "step": 8960 }, { "epoch": 0.22, "grad_norm": 11.1875, "learning_rate": 1.4257627118644069e-05, "loss": 1.5435, "step": 8970 }, { "epoch": 0.22, "grad_norm": 6.9375, "learning_rate": 1.425084745762712e-05, "loss": 1.4033, "step": 8980 }, { "epoch": 0.22, "grad_norm": 15.25, "learning_rate": 1.424406779661017e-05, "loss": 1.3591, "step": 8990 }, { "epoch": 0.23, "grad_norm": 13.625, "learning_rate": 1.4237288135593221e-05, "loss": 1.3307, "step": 9000 }, { "epoch": 0.23, "grad_norm": 22.75, "learning_rate": 1.4230508474576273e-05, "loss": 1.434, "step": 9010 }, { "epoch": 0.23, "grad_norm": 15.375, "learning_rate": 1.4223728813559324e-05, "loss": 1.3094, "step": 9020 }, { "epoch": 0.23, "grad_norm": 16.375, "learning_rate": 1.4216949152542374e-05, "loss": 1.6157, "step": 9030 }, { "epoch": 0.23, "grad_norm": 14.5, "learning_rate": 1.4210169491525425e-05, "loss": 1.5208, "step": 9040 }, { "epoch": 0.23, "grad_norm": 15.875, "learning_rate": 1.4203389830508476e-05, "loss": 1.3652, "step": 9050 }, { "epoch": 0.23, "grad_norm": 14.3125, "learning_rate": 1.4196610169491528e-05, "loss": 1.4421, "step": 9060 }, { "epoch": 0.23, "grad_norm": 12.625, "learning_rate": 1.4189830508474578e-05, "loss": 1.0994, "step": 9070 }, { "epoch": 0.23, "grad_norm": 28.875, "learning_rate": 1.4183050847457629e-05, "loss": 1.4243, "step": 9080 }, { "epoch": 0.23, "grad_norm": 11.625, "learning_rate": 1.417627118644068e-05, "loss": 1.487, "step": 9090 }, { "epoch": 0.23, "grad_norm": 37.25, "learning_rate": 1.416949152542373e-05, "loss": 1.5298, "step": 9100 }, { "epoch": 0.23, "grad_norm": 22.75, "learning_rate": 1.4162711864406781e-05, "loss": 1.4225, "step": 9110 }, { "epoch": 0.23, "grad_norm": 18.5, "learning_rate": 1.4155932203389833e-05, "loss": 1.4601, "step": 9120 }, { "epoch": 0.23, "grad_norm": 16.375, "learning_rate": 1.414915254237288e-05, "loss": 1.4158, "step": 9130 }, { "epoch": 0.23, "grad_norm": 28.875, "learning_rate": 1.4142372881355932e-05, "loss": 1.3936, "step": 9140 }, { "epoch": 0.23, "grad_norm": 14.4375, "learning_rate": 1.4135593220338984e-05, "loss": 1.3291, "step": 9150 }, { "epoch": 0.23, "grad_norm": 10.625, "learning_rate": 1.4128813559322035e-05, "loss": 1.4435, "step": 9160 }, { "epoch": 0.23, "grad_norm": 16.625, "learning_rate": 1.4122033898305085e-05, "loss": 1.2596, "step": 9170 }, { "epoch": 0.23, "grad_norm": 50.5, "learning_rate": 1.4115254237288136e-05, "loss": 1.3926, "step": 9180 }, { "epoch": 0.23, "grad_norm": 24.875, "learning_rate": 1.4108474576271188e-05, "loss": 1.4198, "step": 9190 }, { "epoch": 0.23, "grad_norm": 16.5, "learning_rate": 1.4101694915254239e-05, "loss": 1.479, "step": 9200 }, { "epoch": 0.23, "grad_norm": 10.3125, "learning_rate": 1.4094915254237289e-05, "loss": 1.2579, "step": 9210 }, { "epoch": 0.23, "grad_norm": 14.6875, "learning_rate": 1.408813559322034e-05, "loss": 1.4707, "step": 9220 }, { "epoch": 0.23, "grad_norm": 17.75, "learning_rate": 1.4081355932203391e-05, "loss": 1.4451, "step": 9230 }, { "epoch": 0.23, "grad_norm": 20.375, "learning_rate": 1.4074576271186441e-05, "loss": 1.4185, "step": 9240 }, { "epoch": 0.23, "grad_norm": 22.25, "learning_rate": 1.4067796610169493e-05, "loss": 1.3394, "step": 9250 }, { "epoch": 0.23, "grad_norm": 13.25, "learning_rate": 1.4061016949152544e-05, "loss": 1.4731, "step": 9260 }, { "epoch": 0.23, "grad_norm": 14.8125, "learning_rate": 1.4054237288135595e-05, "loss": 1.2973, "step": 9270 }, { "epoch": 0.23, "grad_norm": 30.5, "learning_rate": 1.4047457627118645e-05, "loss": 1.5124, "step": 9280 }, { "epoch": 0.23, "grad_norm": 15.125, "learning_rate": 1.4040677966101696e-05, "loss": 1.2753, "step": 9290 }, { "epoch": 0.23, "grad_norm": 17.125, "learning_rate": 1.4033898305084748e-05, "loss": 1.4415, "step": 9300 }, { "epoch": 0.23, "grad_norm": 18.75, "learning_rate": 1.40271186440678e-05, "loss": 1.3429, "step": 9310 }, { "epoch": 0.23, "grad_norm": 32.0, "learning_rate": 1.4020338983050849e-05, "loss": 1.4191, "step": 9320 }, { "epoch": 0.23, "grad_norm": 17.375, "learning_rate": 1.4013559322033899e-05, "loss": 1.356, "step": 9330 }, { "epoch": 0.23, "grad_norm": 26.25, "learning_rate": 1.400677966101695e-05, "loss": 1.4327, "step": 9340 }, { "epoch": 0.23, "grad_norm": 8.4375, "learning_rate": 1.4e-05, "loss": 1.3378, "step": 9350 }, { "epoch": 0.23, "grad_norm": 27.375, "learning_rate": 1.3993220338983051e-05, "loss": 1.4791, "step": 9360 }, { "epoch": 0.23, "grad_norm": 9.9375, "learning_rate": 1.3986440677966103e-05, "loss": 1.3589, "step": 9370 }, { "epoch": 0.23, "grad_norm": 11.5, "learning_rate": 1.3979661016949152e-05, "loss": 1.585, "step": 9380 }, { "epoch": 0.23, "grad_norm": 9.1875, "learning_rate": 1.3972881355932204e-05, "loss": 1.4928, "step": 9390 }, { "epoch": 0.23, "grad_norm": 29.0, "learning_rate": 1.3966101694915255e-05, "loss": 1.3321, "step": 9400 }, { "epoch": 0.24, "grad_norm": 7.46875, "learning_rate": 1.3959322033898306e-05, "loss": 1.3466, "step": 9410 }, { "epoch": 0.24, "grad_norm": 14.1875, "learning_rate": 1.3952542372881356e-05, "loss": 1.4826, "step": 9420 }, { "epoch": 0.24, "grad_norm": 25.375, "learning_rate": 1.3945762711864408e-05, "loss": 1.4323, "step": 9430 }, { "epoch": 0.24, "grad_norm": 23.625, "learning_rate": 1.3938983050847459e-05, "loss": 1.2847, "step": 9440 }, { "epoch": 0.24, "grad_norm": 21.625, "learning_rate": 1.393220338983051e-05, "loss": 1.6091, "step": 9450 }, { "epoch": 0.24, "grad_norm": 25.5, "learning_rate": 1.392542372881356e-05, "loss": 1.4748, "step": 9460 }, { "epoch": 0.24, "grad_norm": 10.4375, "learning_rate": 1.3918644067796612e-05, "loss": 1.3939, "step": 9470 }, { "epoch": 0.24, "grad_norm": 29.625, "learning_rate": 1.3911864406779663e-05, "loss": 1.5993, "step": 9480 }, { "epoch": 0.24, "grad_norm": 28.125, "learning_rate": 1.3905084745762714e-05, "loss": 1.3029, "step": 9490 }, { "epoch": 0.24, "grad_norm": 11.6875, "learning_rate": 1.3898305084745764e-05, "loss": 1.3153, "step": 9500 }, { "epoch": 0.24, "grad_norm": 9.125, "learning_rate": 1.3891525423728815e-05, "loss": 1.597, "step": 9510 }, { "epoch": 0.24, "grad_norm": 14.9375, "learning_rate": 1.3884745762711867e-05, "loss": 1.4219, "step": 9520 }, { "epoch": 0.24, "grad_norm": 21.75, "learning_rate": 1.3877966101694918e-05, "loss": 1.437, "step": 9530 }, { "epoch": 0.24, "grad_norm": 16.125, "learning_rate": 1.3871186440677966e-05, "loss": 1.4779, "step": 9540 }, { "epoch": 0.24, "grad_norm": 18.875, "learning_rate": 1.3864406779661018e-05, "loss": 1.3931, "step": 9550 }, { "epoch": 0.24, "grad_norm": 27.25, "learning_rate": 1.3857627118644067e-05, "loss": 1.3809, "step": 9560 }, { "epoch": 0.24, "grad_norm": 23.0, "learning_rate": 1.3850847457627119e-05, "loss": 1.2787, "step": 9570 }, { "epoch": 0.24, "grad_norm": 11.875, "learning_rate": 1.384406779661017e-05, "loss": 1.313, "step": 9580 }, { "epoch": 0.24, "grad_norm": 14.8125, "learning_rate": 1.3837288135593222e-05, "loss": 1.5016, "step": 9590 }, { "epoch": 0.24, "grad_norm": 9.0, "learning_rate": 1.3830508474576271e-05, "loss": 1.2676, "step": 9600 }, { "epoch": 0.24, "grad_norm": 12.9375, "learning_rate": 1.3823728813559323e-05, "loss": 1.4059, "step": 9610 }, { "epoch": 0.24, "grad_norm": 23.375, "learning_rate": 1.3816949152542374e-05, "loss": 1.4623, "step": 9620 }, { "epoch": 0.24, "grad_norm": 13.9375, "learning_rate": 1.3810169491525425e-05, "loss": 1.2202, "step": 9630 }, { "epoch": 0.24, "grad_norm": 29.875, "learning_rate": 1.3803389830508475e-05, "loss": 1.3342, "step": 9640 }, { "epoch": 0.24, "grad_norm": 21.125, "learning_rate": 1.3796610169491527e-05, "loss": 1.3207, "step": 9650 }, { "epoch": 0.24, "grad_norm": 48.75, "learning_rate": 1.3789830508474578e-05, "loss": 1.2569, "step": 9660 }, { "epoch": 0.24, "grad_norm": 36.5, "learning_rate": 1.378305084745763e-05, "loss": 1.485, "step": 9670 }, { "epoch": 0.24, "grad_norm": 18.75, "learning_rate": 1.3776271186440679e-05, "loss": 1.2762, "step": 9680 }, { "epoch": 0.24, "grad_norm": 12.9375, "learning_rate": 1.376949152542373e-05, "loss": 1.4295, "step": 9690 }, { "epoch": 0.24, "grad_norm": 25.5, "learning_rate": 1.3762711864406782e-05, "loss": 1.4219, "step": 9700 }, { "epoch": 0.24, "grad_norm": 19.75, "learning_rate": 1.3755932203389832e-05, "loss": 1.4329, "step": 9710 }, { "epoch": 0.24, "grad_norm": 36.0, "learning_rate": 1.3749152542372883e-05, "loss": 1.4465, "step": 9720 }, { "epoch": 0.24, "grad_norm": 17.625, "learning_rate": 1.3742372881355934e-05, "loss": 1.4222, "step": 9730 }, { "epoch": 0.24, "grad_norm": 17.875, "learning_rate": 1.3735593220338986e-05, "loss": 1.4862, "step": 9740 }, { "epoch": 0.24, "grad_norm": 19.25, "learning_rate": 1.3728813559322034e-05, "loss": 1.3689, "step": 9750 }, { "epoch": 0.24, "grad_norm": 7.375, "learning_rate": 1.3722033898305085e-05, "loss": 1.4442, "step": 9760 }, { "epoch": 0.24, "grad_norm": 21.375, "learning_rate": 1.3715254237288137e-05, "loss": 1.5084, "step": 9770 }, { "epoch": 0.24, "grad_norm": 23.0, "learning_rate": 1.3708474576271186e-05, "loss": 1.4052, "step": 9780 }, { "epoch": 0.24, "grad_norm": 15.75, "learning_rate": 1.3701694915254238e-05, "loss": 1.2911, "step": 9790 }, { "epoch": 0.24, "grad_norm": 18.5, "learning_rate": 1.3694915254237289e-05, "loss": 1.5132, "step": 9800 }, { "epoch": 0.25, "grad_norm": 13.25, "learning_rate": 1.368813559322034e-05, "loss": 1.4176, "step": 9810 }, { "epoch": 0.25, "grad_norm": 42.5, "learning_rate": 1.368135593220339e-05, "loss": 1.4619, "step": 9820 }, { "epoch": 0.25, "grad_norm": 29.875, "learning_rate": 1.3674576271186442e-05, "loss": 1.3777, "step": 9830 }, { "epoch": 0.25, "grad_norm": 13.8125, "learning_rate": 1.3667796610169493e-05, "loss": 1.2798, "step": 9840 }, { "epoch": 0.25, "grad_norm": 19.875, "learning_rate": 1.3661016949152543e-05, "loss": 1.4079, "step": 9850 }, { "epoch": 0.25, "grad_norm": 24.0, "learning_rate": 1.3654237288135594e-05, "loss": 1.497, "step": 9860 }, { "epoch": 0.25, "grad_norm": 25.25, "learning_rate": 1.3647457627118646e-05, "loss": 1.3753, "step": 9870 }, { "epoch": 0.25, "grad_norm": 22.375, "learning_rate": 1.3640677966101697e-05, "loss": 1.5533, "step": 9880 }, { "epoch": 0.25, "grad_norm": 19.75, "learning_rate": 1.3633898305084747e-05, "loss": 1.2329, "step": 9890 }, { "epoch": 0.25, "grad_norm": 14.0625, "learning_rate": 1.3627118644067798e-05, "loss": 1.3455, "step": 9900 }, { "epoch": 0.25, "grad_norm": 19.25, "learning_rate": 1.362033898305085e-05, "loss": 1.3209, "step": 9910 }, { "epoch": 0.25, "grad_norm": 43.75, "learning_rate": 1.36135593220339e-05, "loss": 1.31, "step": 9920 }, { "epoch": 0.25, "grad_norm": 41.75, "learning_rate": 1.360677966101695e-05, "loss": 1.4145, "step": 9930 }, { "epoch": 0.25, "grad_norm": 17.5, "learning_rate": 1.3600000000000002e-05, "loss": 1.3466, "step": 9940 }, { "epoch": 0.25, "grad_norm": 22.5, "learning_rate": 1.3593220338983053e-05, "loss": 1.3998, "step": 9950 }, { "epoch": 0.25, "grad_norm": 12.1875, "learning_rate": 1.3586440677966101e-05, "loss": 1.4374, "step": 9960 }, { "epoch": 0.25, "grad_norm": 19.375, "learning_rate": 1.3579661016949153e-05, "loss": 1.4396, "step": 9970 }, { "epoch": 0.25, "grad_norm": 9.6875, "learning_rate": 1.3572881355932204e-05, "loss": 1.4851, "step": 9980 }, { "epoch": 0.25, "grad_norm": 15.9375, "learning_rate": 1.3566101694915254e-05, "loss": 1.4044, "step": 9990 }, { "epoch": 0.25, "grad_norm": 16.125, "learning_rate": 1.3559322033898305e-05, "loss": 1.5382, "step": 10000 }, { "epoch": 0.25, "eval_loss": 1.3570239543914795, "eval_runtime": 59.2806, "eval_samples_per_second": 16.869, "eval_steps_per_second": 16.869, "step": 10000 }, { "epoch": 0.25, "grad_norm": 20.75, "learning_rate": 1.3552542372881357e-05, "loss": 1.3801, "step": 10010 }, { "epoch": 0.25, "grad_norm": 12.875, "learning_rate": 1.3545762711864408e-05, "loss": 1.3604, "step": 10020 }, { "epoch": 0.25, "grad_norm": 10.9375, "learning_rate": 1.3538983050847458e-05, "loss": 1.489, "step": 10030 }, { "epoch": 0.25, "grad_norm": 10.6875, "learning_rate": 1.353220338983051e-05, "loss": 1.5012, "step": 10040 }, { "epoch": 0.25, "grad_norm": 21.625, "learning_rate": 1.352542372881356e-05, "loss": 1.4174, "step": 10050 }, { "epoch": 0.25, "grad_norm": 28.875, "learning_rate": 1.3518644067796612e-05, "loss": 1.2765, "step": 10060 }, { "epoch": 0.25, "grad_norm": 23.875, "learning_rate": 1.3511864406779662e-05, "loss": 1.4869, "step": 10070 }, { "epoch": 0.25, "grad_norm": 20.0, "learning_rate": 1.3505084745762713e-05, "loss": 1.2636, "step": 10080 }, { "epoch": 0.25, "grad_norm": 12.4375, "learning_rate": 1.3498305084745764e-05, "loss": 1.4241, "step": 10090 }, { "epoch": 0.25, "grad_norm": 25.5, "learning_rate": 1.3491525423728816e-05, "loss": 1.3722, "step": 10100 }, { "epoch": 0.25, "grad_norm": 21.5, "learning_rate": 1.3484745762711866e-05, "loss": 1.3611, "step": 10110 }, { "epoch": 0.25, "grad_norm": 14.375, "learning_rate": 1.3477966101694917e-05, "loss": 1.4016, "step": 10120 }, { "epoch": 0.25, "grad_norm": 11.3125, "learning_rate": 1.3471186440677968e-05, "loss": 1.3755, "step": 10130 }, { "epoch": 0.25, "grad_norm": 15.125, "learning_rate": 1.346440677966102e-05, "loss": 1.3698, "step": 10140 }, { "epoch": 0.25, "grad_norm": 23.875, "learning_rate": 1.345762711864407e-05, "loss": 1.3596, "step": 10150 }, { "epoch": 0.25, "grad_norm": 11.9375, "learning_rate": 1.3450847457627121e-05, "loss": 1.3765, "step": 10160 }, { "epoch": 0.25, "grad_norm": 33.0, "learning_rate": 1.3444067796610169e-05, "loss": 1.3891, "step": 10170 }, { "epoch": 0.25, "grad_norm": 17.25, "learning_rate": 1.343728813559322e-05, "loss": 1.3707, "step": 10180 }, { "epoch": 0.25, "grad_norm": 14.0625, "learning_rate": 1.3430508474576272e-05, "loss": 1.3348, "step": 10190 }, { "epoch": 0.26, "grad_norm": 30.625, "learning_rate": 1.3423728813559323e-05, "loss": 1.4954, "step": 10200 }, { "epoch": 0.26, "grad_norm": 7.8125, "learning_rate": 1.3416949152542373e-05, "loss": 1.4154, "step": 10210 }, { "epoch": 0.26, "grad_norm": 29.25, "learning_rate": 1.3410169491525424e-05, "loss": 1.5622, "step": 10220 }, { "epoch": 0.26, "grad_norm": 6.34375, "learning_rate": 1.3403389830508476e-05, "loss": 1.4377, "step": 10230 }, { "epoch": 0.26, "grad_norm": 31.875, "learning_rate": 1.3396610169491527e-05, "loss": 1.3888, "step": 10240 }, { "epoch": 0.26, "grad_norm": 41.5, "learning_rate": 1.3389830508474577e-05, "loss": 1.4047, "step": 10250 }, { "epoch": 0.26, "grad_norm": 20.0, "learning_rate": 1.3383050847457628e-05, "loss": 1.3593, "step": 10260 }, { "epoch": 0.26, "grad_norm": 15.6875, "learning_rate": 1.337627118644068e-05, "loss": 1.4452, "step": 10270 }, { "epoch": 0.26, "grad_norm": 8.8125, "learning_rate": 1.3369491525423731e-05, "loss": 1.4375, "step": 10280 }, { "epoch": 0.26, "grad_norm": 55.0, "learning_rate": 1.336271186440678e-05, "loss": 1.2869, "step": 10290 }, { "epoch": 0.26, "grad_norm": 17.0, "learning_rate": 1.3355932203389832e-05, "loss": 1.3051, "step": 10300 }, { "epoch": 0.26, "grad_norm": 6.71875, "learning_rate": 1.3349152542372883e-05, "loss": 1.4207, "step": 10310 }, { "epoch": 0.26, "grad_norm": 27.375, "learning_rate": 1.3342372881355933e-05, "loss": 1.5179, "step": 10320 }, { "epoch": 0.26, "grad_norm": 22.0, "learning_rate": 1.3335593220338985e-05, "loss": 1.3529, "step": 10330 }, { "epoch": 0.26, "grad_norm": 27.125, "learning_rate": 1.3328813559322036e-05, "loss": 1.2933, "step": 10340 }, { "epoch": 0.26, "grad_norm": 15.375, "learning_rate": 1.3322033898305087e-05, "loss": 1.4026, "step": 10350 }, { "epoch": 0.26, "grad_norm": 20.0, "learning_rate": 1.3315254237288137e-05, "loss": 1.455, "step": 10360 }, { "epoch": 0.26, "grad_norm": 25.75, "learning_rate": 1.3308474576271187e-05, "loss": 1.4478, "step": 10370 }, { "epoch": 0.26, "grad_norm": 17.875, "learning_rate": 1.3301694915254238e-05, "loss": 1.4811, "step": 10380 }, { "epoch": 0.26, "grad_norm": 23.0, "learning_rate": 1.3294915254237288e-05, "loss": 1.2626, "step": 10390 }, { "epoch": 0.26, "grad_norm": 24.375, "learning_rate": 1.328813559322034e-05, "loss": 1.4401, "step": 10400 }, { "epoch": 0.26, "grad_norm": 15.5625, "learning_rate": 1.328135593220339e-05, "loss": 1.3587, "step": 10410 }, { "epoch": 0.26, "grad_norm": 43.75, "learning_rate": 1.3274576271186442e-05, "loss": 1.3697, "step": 10420 }, { "epoch": 0.26, "grad_norm": 16.375, "learning_rate": 1.3267796610169492e-05, "loss": 1.5165, "step": 10430 }, { "epoch": 0.26, "grad_norm": 26.25, "learning_rate": 1.3261016949152543e-05, "loss": 1.3134, "step": 10440 }, { "epoch": 0.26, "grad_norm": 29.375, "learning_rate": 1.3254237288135595e-05, "loss": 1.5107, "step": 10450 }, { "epoch": 0.26, "grad_norm": 15.0625, "learning_rate": 1.3247457627118644e-05, "loss": 1.4568, "step": 10460 }, { "epoch": 0.26, "grad_norm": 19.625, "learning_rate": 1.3240677966101696e-05, "loss": 1.3214, "step": 10470 }, { "epoch": 0.26, "grad_norm": 8.8125, "learning_rate": 1.3233898305084747e-05, "loss": 1.1839, "step": 10480 }, { "epoch": 0.26, "grad_norm": 12.0, "learning_rate": 1.3227118644067798e-05, "loss": 1.4022, "step": 10490 }, { "epoch": 0.26, "grad_norm": 14.5, "learning_rate": 1.3220338983050848e-05, "loss": 1.3453, "step": 10500 }, { "epoch": 0.26, "grad_norm": 12.625, "learning_rate": 1.32135593220339e-05, "loss": 1.6298, "step": 10510 }, { "epoch": 0.26, "grad_norm": 13.625, "learning_rate": 1.3206779661016951e-05, "loss": 1.4567, "step": 10520 }, { "epoch": 0.26, "grad_norm": 23.75, "learning_rate": 1.3200000000000002e-05, "loss": 1.5548, "step": 10530 }, { "epoch": 0.26, "grad_norm": 54.5, "learning_rate": 1.3193220338983052e-05, "loss": 1.4111, "step": 10540 }, { "epoch": 0.26, "grad_norm": 13.0625, "learning_rate": 1.3186440677966103e-05, "loss": 1.433, "step": 10550 }, { "epoch": 0.26, "grad_norm": 17.375, "learning_rate": 1.3179661016949155e-05, "loss": 1.5689, "step": 10560 }, { "epoch": 0.26, "grad_norm": 20.25, "learning_rate": 1.3172881355932206e-05, "loss": 1.3946, "step": 10570 }, { "epoch": 0.26, "grad_norm": 15.5, "learning_rate": 1.3166101694915254e-05, "loss": 1.4482, "step": 10580 }, { "epoch": 0.26, "grad_norm": 9.875, "learning_rate": 1.3159322033898306e-05, "loss": 1.4798, "step": 10590 }, { "epoch": 0.27, "grad_norm": 24.875, "learning_rate": 1.3152542372881355e-05, "loss": 1.4339, "step": 10600 }, { "epoch": 0.27, "grad_norm": 27.875, "learning_rate": 1.3145762711864407e-05, "loss": 1.3277, "step": 10610 }, { "epoch": 0.27, "grad_norm": 15.4375, "learning_rate": 1.3138983050847458e-05, "loss": 1.362, "step": 10620 }, { "epoch": 0.27, "grad_norm": 26.875, "learning_rate": 1.313220338983051e-05, "loss": 1.3184, "step": 10630 }, { "epoch": 0.27, "grad_norm": 10.25, "learning_rate": 1.312542372881356e-05, "loss": 1.5107, "step": 10640 }, { "epoch": 0.27, "grad_norm": 15.0625, "learning_rate": 1.311864406779661e-05, "loss": 1.1496, "step": 10650 }, { "epoch": 0.27, "grad_norm": 14.5, "learning_rate": 1.3111864406779662e-05, "loss": 1.4018, "step": 10660 }, { "epoch": 0.27, "grad_norm": 38.25, "learning_rate": 1.3105084745762714e-05, "loss": 1.5046, "step": 10670 }, { "epoch": 0.27, "grad_norm": 16.25, "learning_rate": 1.3098305084745763e-05, "loss": 1.3867, "step": 10680 }, { "epoch": 0.27, "grad_norm": 9.125, "learning_rate": 1.3091525423728815e-05, "loss": 1.3284, "step": 10690 }, { "epoch": 0.27, "grad_norm": 27.75, "learning_rate": 1.3084745762711866e-05, "loss": 1.4511, "step": 10700 }, { "epoch": 0.27, "grad_norm": 11.0625, "learning_rate": 1.3077966101694917e-05, "loss": 1.4621, "step": 10710 }, { "epoch": 0.27, "grad_norm": 17.875, "learning_rate": 1.3071186440677967e-05, "loss": 1.3628, "step": 10720 }, { "epoch": 0.27, "grad_norm": 19.25, "learning_rate": 1.3064406779661019e-05, "loss": 1.3811, "step": 10730 }, { "epoch": 0.27, "grad_norm": 17.125, "learning_rate": 1.305762711864407e-05, "loss": 1.283, "step": 10740 }, { "epoch": 0.27, "grad_norm": 7.78125, "learning_rate": 1.305084745762712e-05, "loss": 1.3398, "step": 10750 }, { "epoch": 0.27, "grad_norm": 43.5, "learning_rate": 1.3044067796610171e-05, "loss": 1.3942, "step": 10760 }, { "epoch": 0.27, "grad_norm": 14.75, "learning_rate": 1.3037288135593222e-05, "loss": 1.407, "step": 10770 }, { "epoch": 0.27, "grad_norm": 19.625, "learning_rate": 1.3030508474576274e-05, "loss": 1.3883, "step": 10780 }, { "epoch": 0.27, "grad_norm": 22.125, "learning_rate": 1.3023728813559322e-05, "loss": 1.4614, "step": 10790 }, { "epoch": 0.27, "grad_norm": 9.75, "learning_rate": 1.3016949152542373e-05, "loss": 1.4048, "step": 10800 }, { "epoch": 0.27, "grad_norm": 16.625, "learning_rate": 1.3010169491525425e-05, "loss": 1.5016, "step": 10810 }, { "epoch": 0.27, "grad_norm": 16.75, "learning_rate": 1.3003389830508474e-05, "loss": 1.562, "step": 10820 }, { "epoch": 0.27, "grad_norm": 39.0, "learning_rate": 1.2996610169491526e-05, "loss": 1.3651, "step": 10830 }, { "epoch": 0.27, "grad_norm": 18.75, "learning_rate": 1.2989830508474577e-05, "loss": 1.4262, "step": 10840 }, { "epoch": 0.27, "grad_norm": 9.75, "learning_rate": 1.2983050847457629e-05, "loss": 1.3998, "step": 10850 }, { "epoch": 0.27, "grad_norm": 19.625, "learning_rate": 1.2976271186440678e-05, "loss": 1.3164, "step": 10860 }, { "epoch": 0.27, "grad_norm": 18.125, "learning_rate": 1.296949152542373e-05, "loss": 1.2888, "step": 10870 }, { "epoch": 0.27, "grad_norm": 8.5625, "learning_rate": 1.2962711864406781e-05, "loss": 1.3567, "step": 10880 }, { "epoch": 0.27, "grad_norm": 18.75, "learning_rate": 1.295593220338983e-05, "loss": 1.5165, "step": 10890 }, { "epoch": 0.27, "grad_norm": 11.125, "learning_rate": 1.2949152542372882e-05, "loss": 1.4366, "step": 10900 }, { "epoch": 0.27, "grad_norm": 14.875, "learning_rate": 1.2942372881355934e-05, "loss": 1.3693, "step": 10910 }, { "epoch": 0.27, "grad_norm": 12.625, "learning_rate": 1.2935593220338985e-05, "loss": 1.4838, "step": 10920 }, { "epoch": 0.27, "grad_norm": 11.375, "learning_rate": 1.2928813559322035e-05, "loss": 1.2853, "step": 10930 }, { "epoch": 0.27, "grad_norm": 17.375, "learning_rate": 1.2922033898305086e-05, "loss": 1.3973, "step": 10940 }, { "epoch": 0.27, "grad_norm": 26.75, "learning_rate": 1.2915254237288137e-05, "loss": 1.3605, "step": 10950 }, { "epoch": 0.27, "grad_norm": 23.5, "learning_rate": 1.2908474576271189e-05, "loss": 1.5397, "step": 10960 }, { "epoch": 0.27, "grad_norm": 34.5, "learning_rate": 1.2901694915254239e-05, "loss": 1.2128, "step": 10970 }, { "epoch": 0.27, "grad_norm": 15.0625, "learning_rate": 1.289491525423729e-05, "loss": 1.4024, "step": 10980 }, { "epoch": 0.27, "grad_norm": 13.75, "learning_rate": 1.2888135593220341e-05, "loss": 1.448, "step": 10990 }, { "epoch": 0.28, "grad_norm": 10.5, "learning_rate": 1.288135593220339e-05, "loss": 1.564, "step": 11000 }, { "epoch": 0.28, "grad_norm": 22.0, "learning_rate": 1.287457627118644e-05, "loss": 1.4604, "step": 11010 }, { "epoch": 0.28, "grad_norm": 33.25, "learning_rate": 1.2867796610169492e-05, "loss": 1.4426, "step": 11020 }, { "epoch": 0.28, "grad_norm": 18.25, "learning_rate": 1.2861016949152542e-05, "loss": 1.4848, "step": 11030 }, { "epoch": 0.28, "grad_norm": 28.75, "learning_rate": 1.2854237288135593e-05, "loss": 1.5427, "step": 11040 }, { "epoch": 0.28, "grad_norm": 28.25, "learning_rate": 1.2847457627118645e-05, "loss": 1.4344, "step": 11050 }, { "epoch": 0.28, "grad_norm": 12.5625, "learning_rate": 1.2840677966101696e-05, "loss": 1.3914, "step": 11060 }, { "epoch": 0.28, "grad_norm": 9.875, "learning_rate": 1.2833898305084746e-05, "loss": 1.3298, "step": 11070 }, { "epoch": 0.28, "grad_norm": 9.75, "learning_rate": 1.2827118644067797e-05, "loss": 1.5821, "step": 11080 }, { "epoch": 0.28, "grad_norm": 7.8125, "learning_rate": 1.2820338983050849e-05, "loss": 1.3951, "step": 11090 }, { "epoch": 0.28, "grad_norm": 9.4375, "learning_rate": 1.28135593220339e-05, "loss": 1.5711, "step": 11100 }, { "epoch": 0.28, "grad_norm": 16.5, "learning_rate": 1.280677966101695e-05, "loss": 1.3264, "step": 11110 }, { "epoch": 0.28, "grad_norm": 33.25, "learning_rate": 1.2800000000000001e-05, "loss": 1.2913, "step": 11120 }, { "epoch": 0.28, "grad_norm": 23.875, "learning_rate": 1.2793220338983053e-05, "loss": 1.4656, "step": 11130 }, { "epoch": 0.28, "grad_norm": 42.75, "learning_rate": 1.2786440677966104e-05, "loss": 1.3079, "step": 11140 }, { "epoch": 0.28, "grad_norm": 21.875, "learning_rate": 1.2779661016949154e-05, "loss": 1.3085, "step": 11150 }, { "epoch": 0.28, "grad_norm": 13.75, "learning_rate": 1.2772881355932205e-05, "loss": 1.2513, "step": 11160 }, { "epoch": 0.28, "grad_norm": 25.25, "learning_rate": 1.2766101694915256e-05, "loss": 1.2082, "step": 11170 }, { "epoch": 0.28, "grad_norm": 13.0625, "learning_rate": 1.2759322033898308e-05, "loss": 1.466, "step": 11180 }, { "epoch": 0.28, "grad_norm": 20.75, "learning_rate": 1.2752542372881358e-05, "loss": 1.2071, "step": 11190 }, { "epoch": 0.28, "grad_norm": 9.1875, "learning_rate": 1.2745762711864407e-05, "loss": 1.3974, "step": 11200 }, { "epoch": 0.28, "grad_norm": 33.0, "learning_rate": 1.2738983050847457e-05, "loss": 1.3897, "step": 11210 }, { "epoch": 0.28, "grad_norm": 25.625, "learning_rate": 1.2732203389830508e-05, "loss": 1.2995, "step": 11220 }, { "epoch": 0.28, "grad_norm": 24.0, "learning_rate": 1.272542372881356e-05, "loss": 1.5626, "step": 11230 }, { "epoch": 0.28, "grad_norm": 33.25, "learning_rate": 1.2718644067796611e-05, "loss": 1.4729, "step": 11240 }, { "epoch": 0.28, "grad_norm": 34.75, "learning_rate": 1.2711864406779661e-05, "loss": 1.3542, "step": 11250 }, { "epoch": 0.28, "grad_norm": 12.875, "learning_rate": 1.2705084745762712e-05, "loss": 1.3378, "step": 11260 }, { "epoch": 0.28, "grad_norm": 46.25, "learning_rate": 1.2698305084745764e-05, "loss": 1.6005, "step": 11270 }, { "epoch": 0.28, "grad_norm": 25.75, "learning_rate": 1.2691525423728815e-05, "loss": 1.3225, "step": 11280 }, { "epoch": 0.28, "grad_norm": 34.0, "learning_rate": 1.2684745762711865e-05, "loss": 1.251, "step": 11290 }, { "epoch": 0.28, "grad_norm": 10.625, "learning_rate": 1.2677966101694916e-05, "loss": 1.3167, "step": 11300 }, { "epoch": 0.28, "grad_norm": 15.375, "learning_rate": 1.2671186440677968e-05, "loss": 1.2932, "step": 11310 }, { "epoch": 0.28, "grad_norm": 16.375, "learning_rate": 1.2664406779661019e-05, "loss": 1.3602, "step": 11320 }, { "epoch": 0.28, "grad_norm": 26.5, "learning_rate": 1.2657627118644069e-05, "loss": 1.4041, "step": 11330 }, { "epoch": 0.28, "grad_norm": 61.5, "learning_rate": 1.265084745762712e-05, "loss": 1.4216, "step": 11340 }, { "epoch": 0.28, "grad_norm": 21.0, "learning_rate": 1.2644067796610171e-05, "loss": 1.3234, "step": 11350 }, { "epoch": 0.28, "grad_norm": 10.25, "learning_rate": 1.2637288135593221e-05, "loss": 1.3886, "step": 11360 }, { "epoch": 0.28, "grad_norm": 28.0, "learning_rate": 1.2630508474576273e-05, "loss": 1.4435, "step": 11370 }, { "epoch": 0.28, "grad_norm": 11.375, "learning_rate": 1.2623728813559324e-05, "loss": 1.429, "step": 11380 }, { "epoch": 0.28, "grad_norm": 20.25, "learning_rate": 1.2616949152542375e-05, "loss": 1.3044, "step": 11390 }, { "epoch": 0.28, "grad_norm": 35.25, "learning_rate": 1.2610169491525425e-05, "loss": 1.1607, "step": 11400 }, { "epoch": 0.29, "grad_norm": 18.75, "learning_rate": 1.2603389830508475e-05, "loss": 1.5203, "step": 11410 }, { "epoch": 0.29, "grad_norm": 19.0, "learning_rate": 1.2596610169491526e-05, "loss": 1.3805, "step": 11420 }, { "epoch": 0.29, "grad_norm": 27.625, "learning_rate": 1.2589830508474576e-05, "loss": 1.4146, "step": 11430 }, { "epoch": 0.29, "grad_norm": 7.6875, "learning_rate": 1.2583050847457627e-05, "loss": 1.3669, "step": 11440 }, { "epoch": 0.29, "grad_norm": 35.75, "learning_rate": 1.2576271186440679e-05, "loss": 1.3515, "step": 11450 }, { "epoch": 0.29, "grad_norm": 12.6875, "learning_rate": 1.256949152542373e-05, "loss": 1.2908, "step": 11460 }, { "epoch": 0.29, "grad_norm": 13.625, "learning_rate": 1.256271186440678e-05, "loss": 1.3585, "step": 11470 }, { "epoch": 0.29, "grad_norm": 19.125, "learning_rate": 1.2555932203389831e-05, "loss": 1.3873, "step": 11480 }, { "epoch": 0.29, "grad_norm": 18.5, "learning_rate": 1.2549152542372883e-05, "loss": 1.3207, "step": 11490 }, { "epoch": 0.29, "grad_norm": 21.125, "learning_rate": 1.2542372881355932e-05, "loss": 1.4772, "step": 11500 }, { "epoch": 0.29, "grad_norm": 7.4375, "learning_rate": 1.2535593220338984e-05, "loss": 1.496, "step": 11510 }, { "epoch": 0.29, "grad_norm": 15.6875, "learning_rate": 1.2528813559322035e-05, "loss": 1.2367, "step": 11520 }, { "epoch": 0.29, "grad_norm": 20.375, "learning_rate": 1.2522033898305087e-05, "loss": 1.4323, "step": 11530 }, { "epoch": 0.29, "grad_norm": 15.3125, "learning_rate": 1.2515254237288136e-05, "loss": 1.2777, "step": 11540 }, { "epoch": 0.29, "grad_norm": 17.25, "learning_rate": 1.2508474576271188e-05, "loss": 1.2263, "step": 11550 }, { "epoch": 0.29, "grad_norm": 10.1875, "learning_rate": 1.2501694915254239e-05, "loss": 1.5418, "step": 11560 }, { "epoch": 0.29, "grad_norm": 22.875, "learning_rate": 1.249491525423729e-05, "loss": 1.4742, "step": 11570 }, { "epoch": 0.29, "grad_norm": 37.5, "learning_rate": 1.248813559322034e-05, "loss": 1.38, "step": 11580 }, { "epoch": 0.29, "grad_norm": 7.15625, "learning_rate": 1.2481355932203392e-05, "loss": 1.2842, "step": 11590 }, { "epoch": 0.29, "grad_norm": 34.5, "learning_rate": 1.2474576271186443e-05, "loss": 1.2457, "step": 11600 }, { "epoch": 0.29, "grad_norm": 13.375, "learning_rate": 1.2467796610169494e-05, "loss": 1.3445, "step": 11610 }, { "epoch": 0.29, "grad_norm": 13.875, "learning_rate": 1.2461016949152542e-05, "loss": 1.3287, "step": 11620 }, { "epoch": 0.29, "grad_norm": 16.625, "learning_rate": 1.2454237288135594e-05, "loss": 1.4088, "step": 11630 }, { "epoch": 0.29, "grad_norm": 17.375, "learning_rate": 1.2447457627118643e-05, "loss": 1.2976, "step": 11640 }, { "epoch": 0.29, "grad_norm": 29.375, "learning_rate": 1.2440677966101695e-05, "loss": 1.257, "step": 11650 }, { "epoch": 0.29, "grad_norm": 15.4375, "learning_rate": 1.2433898305084746e-05, "loss": 1.4099, "step": 11660 }, { "epoch": 0.29, "grad_norm": 19.5, "learning_rate": 1.2427118644067798e-05, "loss": 1.3072, "step": 11670 }, { "epoch": 0.29, "grad_norm": 18.375, "learning_rate": 1.2420338983050847e-05, "loss": 1.3438, "step": 11680 }, { "epoch": 0.29, "grad_norm": 12.125, "learning_rate": 1.2413559322033899e-05, "loss": 1.4044, "step": 11690 }, { "epoch": 0.29, "grad_norm": 20.25, "learning_rate": 1.240677966101695e-05, "loss": 1.3873, "step": 11700 }, { "epoch": 0.29, "grad_norm": 20.25, "learning_rate": 1.2400000000000002e-05, "loss": 1.3976, "step": 11710 }, { "epoch": 0.29, "grad_norm": 25.125, "learning_rate": 1.2393220338983051e-05, "loss": 1.5033, "step": 11720 }, { "epoch": 0.29, "grad_norm": 37.0, "learning_rate": 1.2386440677966103e-05, "loss": 1.2516, "step": 11730 }, { "epoch": 0.29, "grad_norm": 18.625, "learning_rate": 1.2379661016949154e-05, "loss": 1.2863, "step": 11740 }, { "epoch": 0.29, "grad_norm": 29.5, "learning_rate": 1.2372881355932205e-05, "loss": 1.4199, "step": 11750 }, { "epoch": 0.29, "grad_norm": 14.8125, "learning_rate": 1.2366101694915255e-05, "loss": 1.5511, "step": 11760 }, { "epoch": 0.29, "grad_norm": 31.125, "learning_rate": 1.2359322033898307e-05, "loss": 1.5902, "step": 11770 }, { "epoch": 0.29, "grad_norm": 9.125, "learning_rate": 1.2352542372881358e-05, "loss": 1.5406, "step": 11780 }, { "epoch": 0.29, "grad_norm": 15.0625, "learning_rate": 1.234576271186441e-05, "loss": 1.4011, "step": 11790 }, { "epoch": 0.29, "grad_norm": 15.0625, "learning_rate": 1.2338983050847459e-05, "loss": 1.2501, "step": 11800 }, { "epoch": 0.3, "grad_norm": 15.9375, "learning_rate": 1.233220338983051e-05, "loss": 1.3155, "step": 11810 }, { "epoch": 0.3, "grad_norm": 7.28125, "learning_rate": 1.2325423728813562e-05, "loss": 1.3896, "step": 11820 }, { "epoch": 0.3, "grad_norm": 29.25, "learning_rate": 1.231864406779661e-05, "loss": 1.6491, "step": 11830 }, { "epoch": 0.3, "grad_norm": 15.875, "learning_rate": 1.2311864406779661e-05, "loss": 1.3143, "step": 11840 }, { "epoch": 0.3, "grad_norm": 26.75, "learning_rate": 1.2305084745762713e-05, "loss": 1.3963, "step": 11850 }, { "epoch": 0.3, "grad_norm": 38.5, "learning_rate": 1.2298305084745762e-05, "loss": 1.3003, "step": 11860 }, { "epoch": 0.3, "grad_norm": 15.0, "learning_rate": 1.2291525423728814e-05, "loss": 1.51, "step": 11870 }, { "epoch": 0.3, "grad_norm": 23.5, "learning_rate": 1.2284745762711865e-05, "loss": 1.2961, "step": 11880 }, { "epoch": 0.3, "grad_norm": 12.5, "learning_rate": 1.2277966101694917e-05, "loss": 1.2989, "step": 11890 }, { "epoch": 0.3, "grad_norm": 32.0, "learning_rate": 1.2271186440677966e-05, "loss": 1.3863, "step": 11900 }, { "epoch": 0.3, "grad_norm": 22.625, "learning_rate": 1.2264406779661018e-05, "loss": 1.6525, "step": 11910 }, { "epoch": 0.3, "grad_norm": 64.0, "learning_rate": 1.2257627118644069e-05, "loss": 1.1806, "step": 11920 }, { "epoch": 0.3, "grad_norm": 12.0625, "learning_rate": 1.225084745762712e-05, "loss": 1.4508, "step": 11930 }, { "epoch": 0.3, "grad_norm": 29.125, "learning_rate": 1.224406779661017e-05, "loss": 1.5203, "step": 11940 }, { "epoch": 0.3, "grad_norm": 8.6875, "learning_rate": 1.2237288135593222e-05, "loss": 1.5215, "step": 11950 }, { "epoch": 0.3, "grad_norm": 7.40625, "learning_rate": 1.2230508474576273e-05, "loss": 1.2804, "step": 11960 }, { "epoch": 0.3, "grad_norm": 52.75, "learning_rate": 1.2223728813559323e-05, "loss": 1.3751, "step": 11970 }, { "epoch": 0.3, "grad_norm": 13.0, "learning_rate": 1.2216949152542374e-05, "loss": 1.2444, "step": 11980 }, { "epoch": 0.3, "grad_norm": 39.5, "learning_rate": 1.2210169491525426e-05, "loss": 1.2927, "step": 11990 }, { "epoch": 0.3, "grad_norm": 12.5625, "learning_rate": 1.2203389830508477e-05, "loss": 1.4383, "step": 12000 }, { "epoch": 0.3, "grad_norm": 41.5, "learning_rate": 1.2196610169491527e-05, "loss": 1.4106, "step": 12010 }, { "epoch": 0.3, "grad_norm": 16.5, "learning_rate": 1.2189830508474578e-05, "loss": 1.3155, "step": 12020 }, { "epoch": 0.3, "grad_norm": 13.9375, "learning_rate": 1.2183050847457628e-05, "loss": 1.3191, "step": 12030 }, { "epoch": 0.3, "grad_norm": 13.9375, "learning_rate": 1.2176271186440677e-05, "loss": 1.393, "step": 12040 }, { "epoch": 0.3, "grad_norm": 46.25, "learning_rate": 1.2169491525423729e-05, "loss": 1.3792, "step": 12050 }, { "epoch": 0.3, "grad_norm": 60.75, "learning_rate": 1.216271186440678e-05, "loss": 1.3197, "step": 12060 }, { "epoch": 0.3, "grad_norm": 27.875, "learning_rate": 1.2155932203389832e-05, "loss": 1.3858, "step": 12070 }, { "epoch": 0.3, "grad_norm": 42.25, "learning_rate": 1.2149152542372881e-05, "loss": 1.4581, "step": 12080 }, { "epoch": 0.3, "grad_norm": 16.75, "learning_rate": 1.2142372881355933e-05, "loss": 1.4428, "step": 12090 }, { "epoch": 0.3, "grad_norm": 40.75, "learning_rate": 1.2135593220338984e-05, "loss": 1.2876, "step": 12100 }, { "epoch": 0.3, "grad_norm": 34.75, "learning_rate": 1.2128813559322034e-05, "loss": 1.295, "step": 12110 }, { "epoch": 0.3, "grad_norm": 31.625, "learning_rate": 1.2122033898305085e-05, "loss": 1.4799, "step": 12120 }, { "epoch": 0.3, "grad_norm": 23.5, "learning_rate": 1.2115254237288137e-05, "loss": 1.3044, "step": 12130 }, { "epoch": 0.3, "grad_norm": 11.3125, "learning_rate": 1.2108474576271188e-05, "loss": 1.3421, "step": 12140 }, { "epoch": 0.3, "grad_norm": 17.875, "learning_rate": 1.2101694915254238e-05, "loss": 1.2581, "step": 12150 }, { "epoch": 0.3, "grad_norm": 32.75, "learning_rate": 1.209491525423729e-05, "loss": 1.4785, "step": 12160 }, { "epoch": 0.3, "grad_norm": 28.5, "learning_rate": 1.208813559322034e-05, "loss": 1.4118, "step": 12170 }, { "epoch": 0.3, "grad_norm": 11.875, "learning_rate": 1.2081355932203392e-05, "loss": 1.2336, "step": 12180 }, { "epoch": 0.3, "grad_norm": 23.75, "learning_rate": 1.2074576271186442e-05, "loss": 1.2431, "step": 12190 }, { "epoch": 0.3, "grad_norm": 8.4375, "learning_rate": 1.2067796610169493e-05, "loss": 1.3331, "step": 12200 }, { "epoch": 0.31, "grad_norm": 23.125, "learning_rate": 1.2061016949152544e-05, "loss": 1.5897, "step": 12210 }, { "epoch": 0.31, "grad_norm": 29.0, "learning_rate": 1.2054237288135596e-05, "loss": 1.4431, "step": 12220 }, { "epoch": 0.31, "grad_norm": 12.75, "learning_rate": 1.2047457627118646e-05, "loss": 1.207, "step": 12230 }, { "epoch": 0.31, "grad_norm": 17.75, "learning_rate": 1.2040677966101695e-05, "loss": 1.318, "step": 12240 }, { "epoch": 0.31, "grad_norm": 14.6875, "learning_rate": 1.2033898305084745e-05, "loss": 1.3912, "step": 12250 }, { "epoch": 0.31, "grad_norm": 26.0, "learning_rate": 1.2027118644067796e-05, "loss": 1.356, "step": 12260 }, { "epoch": 0.31, "grad_norm": 28.875, "learning_rate": 1.2020338983050848e-05, "loss": 1.4116, "step": 12270 }, { "epoch": 0.31, "grad_norm": 8.875, "learning_rate": 1.20135593220339e-05, "loss": 1.3749, "step": 12280 }, { "epoch": 0.31, "grad_norm": 14.1875, "learning_rate": 1.2006779661016949e-05, "loss": 1.4265, "step": 12290 }, { "epoch": 0.31, "grad_norm": 9.9375, "learning_rate": 1.2e-05, "loss": 1.4168, "step": 12300 }, { "epoch": 0.31, "grad_norm": 30.375, "learning_rate": 1.1993220338983052e-05, "loss": 1.3941, "step": 12310 }, { "epoch": 0.31, "grad_norm": 21.375, "learning_rate": 1.1986440677966103e-05, "loss": 1.3206, "step": 12320 }, { "epoch": 0.31, "grad_norm": 30.5, "learning_rate": 1.1979661016949153e-05, "loss": 1.4332, "step": 12330 }, { "epoch": 0.31, "grad_norm": 11.9375, "learning_rate": 1.1972881355932204e-05, "loss": 1.3988, "step": 12340 }, { "epoch": 0.31, "grad_norm": 19.625, "learning_rate": 1.1966101694915256e-05, "loss": 1.5356, "step": 12350 }, { "epoch": 0.31, "grad_norm": 24.375, "learning_rate": 1.1959322033898307e-05, "loss": 1.371, "step": 12360 }, { "epoch": 0.31, "grad_norm": 21.75, "learning_rate": 1.1952542372881357e-05, "loss": 1.6957, "step": 12370 }, { "epoch": 0.31, "grad_norm": 28.375, "learning_rate": 1.1945762711864408e-05, "loss": 1.4933, "step": 12380 }, { "epoch": 0.31, "grad_norm": 16.5, "learning_rate": 1.193898305084746e-05, "loss": 1.3463, "step": 12390 }, { "epoch": 0.31, "grad_norm": 15.4375, "learning_rate": 1.1932203389830511e-05, "loss": 1.431, "step": 12400 }, { "epoch": 0.31, "grad_norm": 21.875, "learning_rate": 1.192542372881356e-05, "loss": 1.3883, "step": 12410 }, { "epoch": 0.31, "grad_norm": 16.375, "learning_rate": 1.1918644067796612e-05, "loss": 1.4184, "step": 12420 }, { "epoch": 0.31, "grad_norm": 18.0, "learning_rate": 1.1911864406779663e-05, "loss": 1.3175, "step": 12430 }, { "epoch": 0.31, "grad_norm": 17.625, "learning_rate": 1.1905084745762713e-05, "loss": 1.3652, "step": 12440 }, { "epoch": 0.31, "grad_norm": 19.0, "learning_rate": 1.1898305084745763e-05, "loss": 1.494, "step": 12450 }, { "epoch": 0.31, "grad_norm": 31.875, "learning_rate": 1.1891525423728814e-05, "loss": 1.4232, "step": 12460 }, { "epoch": 0.31, "grad_norm": 39.5, "learning_rate": 1.1884745762711864e-05, "loss": 1.3255, "step": 12470 }, { "epoch": 0.31, "grad_norm": 53.75, "learning_rate": 1.1877966101694915e-05, "loss": 1.3136, "step": 12480 }, { "epoch": 0.31, "grad_norm": 31.625, "learning_rate": 1.1871186440677967e-05, "loss": 1.3727, "step": 12490 }, { "epoch": 0.31, "grad_norm": 11.25, "learning_rate": 1.1864406779661018e-05, "loss": 1.1973, "step": 12500 }, { "epoch": 0.31, "grad_norm": 19.125, "learning_rate": 1.1857627118644068e-05, "loss": 1.5978, "step": 12510 }, { "epoch": 0.31, "grad_norm": 10.8125, "learning_rate": 1.185084745762712e-05, "loss": 1.3858, "step": 12520 }, { "epoch": 0.31, "grad_norm": 16.875, "learning_rate": 1.184406779661017e-05, "loss": 1.4182, "step": 12530 }, { "epoch": 0.31, "grad_norm": 25.0, "learning_rate": 1.183728813559322e-05, "loss": 1.1735, "step": 12540 }, { "epoch": 0.31, "grad_norm": 30.375, "learning_rate": 1.1830508474576272e-05, "loss": 1.2487, "step": 12550 }, { "epoch": 0.31, "grad_norm": 40.75, "learning_rate": 1.1823728813559323e-05, "loss": 1.3385, "step": 12560 }, { "epoch": 0.31, "grad_norm": 11.25, "learning_rate": 1.1816949152542375e-05, "loss": 1.3932, "step": 12570 }, { "epoch": 0.31, "grad_norm": 12.875, "learning_rate": 1.1810169491525424e-05, "loss": 1.3053, "step": 12580 }, { "epoch": 0.31, "grad_norm": 16.0, "learning_rate": 1.1803389830508476e-05, "loss": 1.1677, "step": 12590 }, { "epoch": 0.32, "grad_norm": 33.0, "learning_rate": 1.1796610169491527e-05, "loss": 1.3547, "step": 12600 }, { "epoch": 0.32, "grad_norm": 13.25, "learning_rate": 1.1789830508474578e-05, "loss": 1.4628, "step": 12610 }, { "epoch": 0.32, "grad_norm": 36.5, "learning_rate": 1.1783050847457628e-05, "loss": 1.3457, "step": 12620 }, { "epoch": 0.32, "grad_norm": 77.0, "learning_rate": 1.177627118644068e-05, "loss": 1.4649, "step": 12630 }, { "epoch": 0.32, "grad_norm": 26.25, "learning_rate": 1.1769491525423731e-05, "loss": 1.6308, "step": 12640 }, { "epoch": 0.32, "grad_norm": 21.125, "learning_rate": 1.1762711864406782e-05, "loss": 1.315, "step": 12650 }, { "epoch": 0.32, "grad_norm": 18.5, "learning_rate": 1.175593220338983e-05, "loss": 1.3999, "step": 12660 }, { "epoch": 0.32, "grad_norm": 12.4375, "learning_rate": 1.1749152542372882e-05, "loss": 1.3569, "step": 12670 }, { "epoch": 0.32, "grad_norm": 50.0, "learning_rate": 1.1742372881355931e-05, "loss": 1.2881, "step": 12680 }, { "epoch": 0.32, "grad_norm": 14.3125, "learning_rate": 1.1735593220338983e-05, "loss": 1.5333, "step": 12690 }, { "epoch": 0.32, "grad_norm": 32.0, "learning_rate": 1.1728813559322034e-05, "loss": 1.4207, "step": 12700 }, { "epoch": 0.32, "grad_norm": 19.5, "learning_rate": 1.1722033898305086e-05, "loss": 1.3351, "step": 12710 }, { "epoch": 0.32, "grad_norm": 66.0, "learning_rate": 1.1715254237288135e-05, "loss": 1.3974, "step": 12720 }, { "epoch": 0.32, "grad_norm": 17.375, "learning_rate": 1.1708474576271187e-05, "loss": 1.3413, "step": 12730 }, { "epoch": 0.32, "grad_norm": 17.5, "learning_rate": 1.1701694915254238e-05, "loss": 1.4573, "step": 12740 }, { "epoch": 0.32, "grad_norm": 17.25, "learning_rate": 1.169491525423729e-05, "loss": 1.4507, "step": 12750 }, { "epoch": 0.32, "grad_norm": 23.5, "learning_rate": 1.168813559322034e-05, "loss": 1.4594, "step": 12760 }, { "epoch": 0.32, "grad_norm": 27.25, "learning_rate": 1.168135593220339e-05, "loss": 1.3051, "step": 12770 }, { "epoch": 0.32, "grad_norm": 14.8125, "learning_rate": 1.1674576271186442e-05, "loss": 1.4516, "step": 12780 }, { "epoch": 0.32, "grad_norm": 12.1875, "learning_rate": 1.1667796610169494e-05, "loss": 1.3318, "step": 12790 }, { "epoch": 0.32, "grad_norm": 12.0625, "learning_rate": 1.1661016949152543e-05, "loss": 1.5477, "step": 12800 }, { "epoch": 0.32, "grad_norm": 40.75, "learning_rate": 1.1654237288135595e-05, "loss": 1.2766, "step": 12810 }, { "epoch": 0.32, "grad_norm": 18.0, "learning_rate": 1.1647457627118646e-05, "loss": 1.2983, "step": 12820 }, { "epoch": 0.32, "grad_norm": 18.25, "learning_rate": 1.1640677966101697e-05, "loss": 1.2756, "step": 12830 }, { "epoch": 0.32, "grad_norm": 20.625, "learning_rate": 1.1633898305084747e-05, "loss": 1.4396, "step": 12840 }, { "epoch": 0.32, "grad_norm": 28.0, "learning_rate": 1.1627118644067799e-05, "loss": 1.3779, "step": 12850 }, { "epoch": 0.32, "grad_norm": 13.1875, "learning_rate": 1.162033898305085e-05, "loss": 1.4271, "step": 12860 }, { "epoch": 0.32, "grad_norm": 21.875, "learning_rate": 1.1613559322033898e-05, "loss": 1.2491, "step": 12870 }, { "epoch": 0.32, "grad_norm": 43.25, "learning_rate": 1.160677966101695e-05, "loss": 1.4746, "step": 12880 }, { "epoch": 0.32, "grad_norm": 51.75, "learning_rate": 1.16e-05, "loss": 1.4252, "step": 12890 }, { "epoch": 0.32, "grad_norm": 19.25, "learning_rate": 1.159322033898305e-05, "loss": 1.2963, "step": 12900 }, { "epoch": 0.32, "grad_norm": 14.4375, "learning_rate": 1.1586440677966102e-05, "loss": 1.3186, "step": 12910 }, { "epoch": 0.32, "grad_norm": 13.0625, "learning_rate": 1.1579661016949153e-05, "loss": 1.316, "step": 12920 }, { "epoch": 0.32, "grad_norm": 18.875, "learning_rate": 1.1572881355932205e-05, "loss": 1.3342, "step": 12930 }, { "epoch": 0.32, "grad_norm": 10.125, "learning_rate": 1.1566101694915254e-05, "loss": 1.1205, "step": 12940 }, { "epoch": 0.32, "grad_norm": 23.875, "learning_rate": 1.1559322033898306e-05, "loss": 1.4295, "step": 12950 }, { "epoch": 0.32, "grad_norm": 29.5, "learning_rate": 1.1552542372881357e-05, "loss": 1.3836, "step": 12960 }, { "epoch": 0.32, "grad_norm": 24.5, "learning_rate": 1.1545762711864409e-05, "loss": 1.2477, "step": 12970 }, { "epoch": 0.32, "grad_norm": 15.4375, "learning_rate": 1.1538983050847458e-05, "loss": 1.3418, "step": 12980 }, { "epoch": 0.32, "grad_norm": 30.375, "learning_rate": 1.153220338983051e-05, "loss": 1.3886, "step": 12990 }, { "epoch": 0.33, "grad_norm": 29.5, "learning_rate": 1.1525423728813561e-05, "loss": 1.4092, "step": 13000 }, { "epoch": 0.33, "grad_norm": 29.125, "learning_rate": 1.151864406779661e-05, "loss": 1.5852, "step": 13010 }, { "epoch": 0.33, "grad_norm": 18.875, "learning_rate": 1.1511864406779662e-05, "loss": 1.3838, "step": 13020 }, { "epoch": 0.33, "grad_norm": 17.0, "learning_rate": 1.1505084745762714e-05, "loss": 1.3919, "step": 13030 }, { "epoch": 0.33, "grad_norm": 14.125, "learning_rate": 1.1498305084745765e-05, "loss": 1.3076, "step": 13040 }, { "epoch": 0.33, "grad_norm": 10.0625, "learning_rate": 1.1491525423728815e-05, "loss": 1.339, "step": 13050 }, { "epoch": 0.33, "grad_norm": 14.25, "learning_rate": 1.1484745762711866e-05, "loss": 1.4315, "step": 13060 }, { "epoch": 0.33, "grad_norm": 8.6875, "learning_rate": 1.1477966101694916e-05, "loss": 1.3238, "step": 13070 }, { "epoch": 0.33, "grad_norm": 26.0, "learning_rate": 1.1471186440677965e-05, "loss": 1.3567, "step": 13080 }, { "epoch": 0.33, "grad_norm": 44.0, "learning_rate": 1.1464406779661017e-05, "loss": 1.2877, "step": 13090 }, { "epoch": 0.33, "grad_norm": 30.375, "learning_rate": 1.1457627118644068e-05, "loss": 1.3856, "step": 13100 }, { "epoch": 0.33, "grad_norm": 20.875, "learning_rate": 1.145084745762712e-05, "loss": 1.3871, "step": 13110 }, { "epoch": 0.33, "grad_norm": 22.625, "learning_rate": 1.144406779661017e-05, "loss": 1.3937, "step": 13120 }, { "epoch": 0.33, "grad_norm": 19.625, "learning_rate": 1.143728813559322e-05, "loss": 1.4812, "step": 13130 }, { "epoch": 0.33, "grad_norm": 68.5, "learning_rate": 1.1430508474576272e-05, "loss": 1.6372, "step": 13140 }, { "epoch": 0.33, "grad_norm": 29.75, "learning_rate": 1.1423728813559322e-05, "loss": 1.3724, "step": 13150 }, { "epoch": 0.33, "grad_norm": 14.25, "learning_rate": 1.1416949152542373e-05, "loss": 1.2524, "step": 13160 }, { "epoch": 0.33, "grad_norm": 12.375, "learning_rate": 1.1410169491525425e-05, "loss": 1.5147, "step": 13170 }, { "epoch": 0.33, "grad_norm": 59.75, "learning_rate": 1.1403389830508476e-05, "loss": 1.4632, "step": 13180 }, { "epoch": 0.33, "grad_norm": 22.125, "learning_rate": 1.1396610169491526e-05, "loss": 1.3722, "step": 13190 }, { "epoch": 0.33, "grad_norm": 20.5, "learning_rate": 1.1389830508474577e-05, "loss": 1.349, "step": 13200 }, { "epoch": 0.33, "grad_norm": 17.125, "learning_rate": 1.1383050847457629e-05, "loss": 1.3975, "step": 13210 }, { "epoch": 0.33, "grad_norm": 32.5, "learning_rate": 1.137627118644068e-05, "loss": 1.4174, "step": 13220 }, { "epoch": 0.33, "grad_norm": 11.3125, "learning_rate": 1.136949152542373e-05, "loss": 1.2684, "step": 13230 }, { "epoch": 0.33, "grad_norm": 22.375, "learning_rate": 1.1362711864406781e-05, "loss": 1.3821, "step": 13240 }, { "epoch": 0.33, "grad_norm": 15.6875, "learning_rate": 1.1355932203389833e-05, "loss": 1.3647, "step": 13250 }, { "epoch": 0.33, "grad_norm": 9.1875, "learning_rate": 1.1349152542372884e-05, "loss": 1.3544, "step": 13260 }, { "epoch": 0.33, "grad_norm": 27.375, "learning_rate": 1.1342372881355934e-05, "loss": 1.2971, "step": 13270 }, { "epoch": 0.33, "grad_norm": 38.75, "learning_rate": 1.1335593220338983e-05, "loss": 1.5632, "step": 13280 }, { "epoch": 0.33, "grad_norm": 36.0, "learning_rate": 1.1328813559322033e-05, "loss": 1.5077, "step": 13290 }, { "epoch": 0.33, "grad_norm": 31.0, "learning_rate": 1.1322033898305084e-05, "loss": 1.1637, "step": 13300 }, { "epoch": 0.33, "grad_norm": 29.25, "learning_rate": 1.1315254237288136e-05, "loss": 1.4482, "step": 13310 }, { "epoch": 0.33, "grad_norm": 14.0, "learning_rate": 1.1308474576271187e-05, "loss": 1.3316, "step": 13320 }, { "epoch": 0.33, "grad_norm": 26.375, "learning_rate": 1.1301694915254237e-05, "loss": 1.3351, "step": 13330 }, { "epoch": 0.33, "grad_norm": 27.0, "learning_rate": 1.1294915254237288e-05, "loss": 1.4783, "step": 13340 }, { "epoch": 0.33, "grad_norm": 28.5, "learning_rate": 1.128813559322034e-05, "loss": 1.4635, "step": 13350 }, { "epoch": 0.33, "grad_norm": 28.625, "learning_rate": 1.1281355932203391e-05, "loss": 1.2941, "step": 13360 }, { "epoch": 0.33, "grad_norm": 21.25, "learning_rate": 1.1274576271186441e-05, "loss": 1.4337, "step": 13370 }, { "epoch": 0.33, "grad_norm": 9.25, "learning_rate": 1.1267796610169492e-05, "loss": 1.3615, "step": 13380 }, { "epoch": 0.33, "grad_norm": 30.5, "learning_rate": 1.1261016949152544e-05, "loss": 1.4474, "step": 13390 }, { "epoch": 0.34, "grad_norm": 29.5, "learning_rate": 1.1254237288135595e-05, "loss": 1.512, "step": 13400 }, { "epoch": 0.34, "grad_norm": 16.375, "learning_rate": 1.1247457627118645e-05, "loss": 1.3367, "step": 13410 }, { "epoch": 0.34, "grad_norm": 17.25, "learning_rate": 1.1240677966101696e-05, "loss": 1.3797, "step": 13420 }, { "epoch": 0.34, "grad_norm": 21.75, "learning_rate": 1.1233898305084748e-05, "loss": 1.3935, "step": 13430 }, { "epoch": 0.34, "grad_norm": 48.0, "learning_rate": 1.1227118644067799e-05, "loss": 1.2878, "step": 13440 }, { "epoch": 0.34, "grad_norm": 12.3125, "learning_rate": 1.1220338983050849e-05, "loss": 1.1581, "step": 13450 }, { "epoch": 0.34, "grad_norm": 17.75, "learning_rate": 1.12135593220339e-05, "loss": 1.4249, "step": 13460 }, { "epoch": 0.34, "grad_norm": 10.0, "learning_rate": 1.1206779661016951e-05, "loss": 1.3671, "step": 13470 }, { "epoch": 0.34, "grad_norm": 9.6875, "learning_rate": 1.1200000000000001e-05, "loss": 1.4198, "step": 13480 }, { "epoch": 0.34, "grad_norm": 25.5, "learning_rate": 1.1193220338983051e-05, "loss": 1.3878, "step": 13490 }, { "epoch": 0.34, "grad_norm": 26.75, "learning_rate": 1.1186440677966102e-05, "loss": 1.158, "step": 13500 }, { "epoch": 0.34, "grad_norm": 31.375, "learning_rate": 1.1179661016949152e-05, "loss": 1.3105, "step": 13510 }, { "epoch": 0.34, "grad_norm": 15.75, "learning_rate": 1.1172881355932203e-05, "loss": 1.5006, "step": 13520 }, { "epoch": 0.34, "grad_norm": 72.0, "learning_rate": 1.1166101694915255e-05, "loss": 1.4191, "step": 13530 }, { "epoch": 0.34, "grad_norm": 20.875, "learning_rate": 1.1159322033898306e-05, "loss": 1.3523, "step": 13540 }, { "epoch": 0.34, "grad_norm": 22.25, "learning_rate": 1.1152542372881356e-05, "loss": 1.3726, "step": 13550 }, { "epoch": 0.34, "grad_norm": 48.25, "learning_rate": 1.1145762711864407e-05, "loss": 1.3706, "step": 13560 }, { "epoch": 0.34, "grad_norm": 39.25, "learning_rate": 1.1138983050847459e-05, "loss": 1.3988, "step": 13570 }, { "epoch": 0.34, "grad_norm": 34.0, "learning_rate": 1.113220338983051e-05, "loss": 1.36, "step": 13580 }, { "epoch": 0.34, "grad_norm": 18.25, "learning_rate": 1.112542372881356e-05, "loss": 1.5547, "step": 13590 }, { "epoch": 0.34, "grad_norm": 27.875, "learning_rate": 1.1118644067796611e-05, "loss": 1.2315, "step": 13600 }, { "epoch": 0.34, "grad_norm": 26.375, "learning_rate": 1.1111864406779663e-05, "loss": 1.1956, "step": 13610 }, { "epoch": 0.34, "grad_norm": 13.375, "learning_rate": 1.1105084745762712e-05, "loss": 1.2934, "step": 13620 }, { "epoch": 0.34, "grad_norm": 26.625, "learning_rate": 1.1098305084745764e-05, "loss": 1.416, "step": 13630 }, { "epoch": 0.34, "grad_norm": 17.875, "learning_rate": 1.1091525423728815e-05, "loss": 1.2588, "step": 13640 }, { "epoch": 0.34, "grad_norm": 10.5625, "learning_rate": 1.1084745762711867e-05, "loss": 1.3671, "step": 13650 }, { "epoch": 0.34, "grad_norm": 14.3125, "learning_rate": 1.1077966101694916e-05, "loss": 1.4697, "step": 13660 }, { "epoch": 0.34, "grad_norm": 18.625, "learning_rate": 1.1071186440677968e-05, "loss": 1.2121, "step": 13670 }, { "epoch": 0.34, "grad_norm": 21.25, "learning_rate": 1.1064406779661019e-05, "loss": 1.2724, "step": 13680 }, { "epoch": 0.34, "grad_norm": 21.75, "learning_rate": 1.105762711864407e-05, "loss": 1.3785, "step": 13690 }, { "epoch": 0.34, "grad_norm": 14.75, "learning_rate": 1.1050847457627118e-05, "loss": 1.5127, "step": 13700 }, { "epoch": 0.34, "grad_norm": 26.375, "learning_rate": 1.104406779661017e-05, "loss": 1.5337, "step": 13710 }, { "epoch": 0.34, "grad_norm": 18.125, "learning_rate": 1.1037288135593221e-05, "loss": 1.4343, "step": 13720 }, { "epoch": 0.34, "grad_norm": 31.0, "learning_rate": 1.1030508474576271e-05, "loss": 1.4872, "step": 13730 }, { "epoch": 0.34, "grad_norm": 24.125, "learning_rate": 1.1023728813559322e-05, "loss": 1.3495, "step": 13740 }, { "epoch": 0.34, "grad_norm": 12.125, "learning_rate": 1.1016949152542374e-05, "loss": 1.2934, "step": 13750 }, { "epoch": 0.34, "grad_norm": 34.25, "learning_rate": 1.1010169491525423e-05, "loss": 1.4117, "step": 13760 }, { "epoch": 0.34, "grad_norm": 19.625, "learning_rate": 1.1003389830508475e-05, "loss": 1.5345, "step": 13770 }, { "epoch": 0.34, "grad_norm": 21.125, "learning_rate": 1.0996610169491526e-05, "loss": 1.4747, "step": 13780 }, { "epoch": 0.34, "grad_norm": 28.5, "learning_rate": 1.0989830508474578e-05, "loss": 1.3429, "step": 13790 }, { "epoch": 0.34, "grad_norm": 15.4375, "learning_rate": 1.0983050847457627e-05, "loss": 1.0869, "step": 13800 }, { "epoch": 0.35, "grad_norm": 35.25, "learning_rate": 1.0976271186440679e-05, "loss": 1.5203, "step": 13810 }, { "epoch": 0.35, "grad_norm": 55.25, "learning_rate": 1.096949152542373e-05, "loss": 1.3117, "step": 13820 }, { "epoch": 0.35, "grad_norm": 14.3125, "learning_rate": 1.0962711864406782e-05, "loss": 1.4044, "step": 13830 }, { "epoch": 0.35, "grad_norm": 15.125, "learning_rate": 1.0955932203389831e-05, "loss": 1.5186, "step": 13840 }, { "epoch": 0.35, "grad_norm": 20.5, "learning_rate": 1.0949152542372883e-05, "loss": 1.251, "step": 13850 }, { "epoch": 0.35, "grad_norm": 42.75, "learning_rate": 1.0942372881355934e-05, "loss": 1.3674, "step": 13860 }, { "epoch": 0.35, "grad_norm": 12.0, "learning_rate": 1.0935593220338985e-05, "loss": 1.4548, "step": 13870 }, { "epoch": 0.35, "grad_norm": 18.5, "learning_rate": 1.0928813559322035e-05, "loss": 1.2881, "step": 13880 }, { "epoch": 0.35, "grad_norm": 47.0, "learning_rate": 1.0922033898305087e-05, "loss": 1.4235, "step": 13890 }, { "epoch": 0.35, "grad_norm": 15.9375, "learning_rate": 1.0915254237288135e-05, "loss": 1.267, "step": 13900 }, { "epoch": 0.35, "grad_norm": 30.75, "learning_rate": 1.0908474576271186e-05, "loss": 1.2051, "step": 13910 }, { "epoch": 0.35, "grad_norm": 21.625, "learning_rate": 1.0901694915254237e-05, "loss": 1.4902, "step": 13920 }, { "epoch": 0.35, "grad_norm": 19.0, "learning_rate": 1.0894915254237289e-05, "loss": 1.0872, "step": 13930 }, { "epoch": 0.35, "grad_norm": 48.25, "learning_rate": 1.0888135593220339e-05, "loss": 1.3268, "step": 13940 }, { "epoch": 0.35, "grad_norm": 20.0, "learning_rate": 1.088135593220339e-05, "loss": 1.3066, "step": 13950 }, { "epoch": 0.35, "grad_norm": 12.1875, "learning_rate": 1.0874576271186441e-05, "loss": 1.3713, "step": 13960 }, { "epoch": 0.35, "grad_norm": 21.375, "learning_rate": 1.0867796610169493e-05, "loss": 1.4468, "step": 13970 }, { "epoch": 0.35, "grad_norm": 17.0, "learning_rate": 1.0861016949152542e-05, "loss": 1.4413, "step": 13980 }, { "epoch": 0.35, "grad_norm": 21.0, "learning_rate": 1.0854237288135594e-05, "loss": 1.3061, "step": 13990 }, { "epoch": 0.35, "grad_norm": 20.875, "learning_rate": 1.0847457627118645e-05, "loss": 1.4503, "step": 14000 }, { "epoch": 0.35, "grad_norm": 15.125, "learning_rate": 1.0840677966101697e-05, "loss": 1.2215, "step": 14010 }, { "epoch": 0.35, "grad_norm": 24.625, "learning_rate": 1.0833898305084746e-05, "loss": 1.4355, "step": 14020 }, { "epoch": 0.35, "grad_norm": 13.8125, "learning_rate": 1.0827118644067798e-05, "loss": 1.4657, "step": 14030 }, { "epoch": 0.35, "grad_norm": 26.25, "learning_rate": 1.0820338983050849e-05, "loss": 1.268, "step": 14040 }, { "epoch": 0.35, "grad_norm": 20.5, "learning_rate": 1.08135593220339e-05, "loss": 1.306, "step": 14050 }, { "epoch": 0.35, "grad_norm": 12.8125, "learning_rate": 1.080677966101695e-05, "loss": 1.4965, "step": 14060 }, { "epoch": 0.35, "grad_norm": 18.5, "learning_rate": 1.0800000000000002e-05, "loss": 1.3629, "step": 14070 }, { "epoch": 0.35, "grad_norm": 32.0, "learning_rate": 1.0793220338983053e-05, "loss": 1.2364, "step": 14080 }, { "epoch": 0.35, "grad_norm": 14.4375, "learning_rate": 1.0786440677966103e-05, "loss": 1.334, "step": 14090 }, { "epoch": 0.35, "grad_norm": 11.875, "learning_rate": 1.0779661016949154e-05, "loss": 1.4108, "step": 14100 }, { "epoch": 0.35, "grad_norm": 35.5, "learning_rate": 1.0772881355932204e-05, "loss": 1.4491, "step": 14110 }, { "epoch": 0.35, "grad_norm": 16.25, "learning_rate": 1.0766101694915254e-05, "loss": 1.4305, "step": 14120 }, { "epoch": 0.35, "grad_norm": 22.75, "learning_rate": 1.0759322033898305e-05, "loss": 1.324, "step": 14130 }, { "epoch": 0.35, "grad_norm": 16.625, "learning_rate": 1.0752542372881356e-05, "loss": 1.1912, "step": 14140 }, { "epoch": 0.35, "grad_norm": 24.875, "learning_rate": 1.0745762711864408e-05, "loss": 1.3567, "step": 14150 }, { "epoch": 0.35, "grad_norm": 15.25, "learning_rate": 1.0738983050847457e-05, "loss": 1.3639, "step": 14160 }, { "epoch": 0.35, "grad_norm": 11.5625, "learning_rate": 1.0732203389830509e-05, "loss": 1.4455, "step": 14170 }, { "epoch": 0.35, "grad_norm": 13.4375, "learning_rate": 1.072542372881356e-05, "loss": 1.4245, "step": 14180 }, { "epoch": 0.35, "grad_norm": 12.9375, "learning_rate": 1.0718644067796612e-05, "loss": 1.303, "step": 14190 }, { "epoch": 0.35, "grad_norm": 22.125, "learning_rate": 1.0711864406779661e-05, "loss": 1.3533, "step": 14200 }, { "epoch": 0.36, "grad_norm": 13.9375, "learning_rate": 1.0705084745762713e-05, "loss": 1.3825, "step": 14210 }, { "epoch": 0.36, "grad_norm": 30.875, "learning_rate": 1.0698305084745764e-05, "loss": 1.3004, "step": 14220 }, { "epoch": 0.36, "grad_norm": 22.625, "learning_rate": 1.0691525423728814e-05, "loss": 1.3967, "step": 14230 }, { "epoch": 0.36, "grad_norm": 13.125, "learning_rate": 1.0684745762711865e-05, "loss": 1.2226, "step": 14240 }, { "epoch": 0.36, "grad_norm": 33.25, "learning_rate": 1.0677966101694917e-05, "loss": 1.3159, "step": 14250 }, { "epoch": 0.36, "grad_norm": 20.0, "learning_rate": 1.0671186440677968e-05, "loss": 1.239, "step": 14260 }, { "epoch": 0.36, "grad_norm": 16.25, "learning_rate": 1.0664406779661018e-05, "loss": 1.3726, "step": 14270 }, { "epoch": 0.36, "grad_norm": 16.625, "learning_rate": 1.065762711864407e-05, "loss": 1.3982, "step": 14280 }, { "epoch": 0.36, "grad_norm": 23.875, "learning_rate": 1.065084745762712e-05, "loss": 1.4308, "step": 14290 }, { "epoch": 0.36, "grad_norm": 15.0625, "learning_rate": 1.0644067796610172e-05, "loss": 1.299, "step": 14300 }, { "epoch": 0.36, "grad_norm": 40.5, "learning_rate": 1.0637288135593222e-05, "loss": 1.3387, "step": 14310 }, { "epoch": 0.36, "grad_norm": 29.375, "learning_rate": 1.0630508474576271e-05, "loss": 1.5603, "step": 14320 }, { "epoch": 0.36, "grad_norm": 28.0, "learning_rate": 1.0623728813559323e-05, "loss": 1.4149, "step": 14330 }, { "epoch": 0.36, "grad_norm": 51.5, "learning_rate": 1.0616949152542373e-05, "loss": 1.4769, "step": 14340 }, { "epoch": 0.36, "grad_norm": 18.125, "learning_rate": 1.0610169491525424e-05, "loss": 1.5026, "step": 14350 }, { "epoch": 0.36, "grad_norm": 30.5, "learning_rate": 1.0603389830508475e-05, "loss": 1.2405, "step": 14360 }, { "epoch": 0.36, "grad_norm": 19.625, "learning_rate": 1.0596610169491525e-05, "loss": 1.3363, "step": 14370 }, { "epoch": 0.36, "grad_norm": 20.75, "learning_rate": 1.0589830508474576e-05, "loss": 1.4854, "step": 14380 }, { "epoch": 0.36, "grad_norm": 17.875, "learning_rate": 1.0583050847457628e-05, "loss": 1.5275, "step": 14390 }, { "epoch": 0.36, "grad_norm": 20.625, "learning_rate": 1.057627118644068e-05, "loss": 1.3683, "step": 14400 }, { "epoch": 0.36, "grad_norm": 27.0, "learning_rate": 1.0569491525423729e-05, "loss": 1.3453, "step": 14410 }, { "epoch": 0.36, "grad_norm": 15.9375, "learning_rate": 1.056271186440678e-05, "loss": 1.4344, "step": 14420 }, { "epoch": 0.36, "grad_norm": 20.5, "learning_rate": 1.0555932203389832e-05, "loss": 1.4687, "step": 14430 }, { "epoch": 0.36, "grad_norm": 43.5, "learning_rate": 1.0549152542372883e-05, "loss": 1.6406, "step": 14440 }, { "epoch": 0.36, "grad_norm": 21.25, "learning_rate": 1.0542372881355933e-05, "loss": 1.4751, "step": 14450 }, { "epoch": 0.36, "grad_norm": 11.5625, "learning_rate": 1.0535593220338984e-05, "loss": 1.3908, "step": 14460 }, { "epoch": 0.36, "grad_norm": 33.25, "learning_rate": 1.0528813559322036e-05, "loss": 1.3654, "step": 14470 }, { "epoch": 0.36, "grad_norm": 12.5625, "learning_rate": 1.0522033898305087e-05, "loss": 1.3358, "step": 14480 }, { "epoch": 0.36, "grad_norm": 12.5, "learning_rate": 1.0515254237288137e-05, "loss": 1.2108, "step": 14490 }, { "epoch": 0.36, "grad_norm": 8.625, "learning_rate": 1.0508474576271188e-05, "loss": 1.483, "step": 14500 }, { "epoch": 0.36, "grad_norm": 25.75, "learning_rate": 1.050169491525424e-05, "loss": 1.4243, "step": 14510 }, { "epoch": 0.36, "grad_norm": 12.75, "learning_rate": 1.049491525423729e-05, "loss": 1.4495, "step": 14520 }, { "epoch": 0.36, "grad_norm": 26.375, "learning_rate": 1.0488135593220339e-05, "loss": 1.3261, "step": 14530 }, { "epoch": 0.36, "grad_norm": 17.0, "learning_rate": 1.048135593220339e-05, "loss": 1.3303, "step": 14540 }, { "epoch": 0.36, "grad_norm": 16.625, "learning_rate": 1.047457627118644e-05, "loss": 1.2757, "step": 14550 }, { "epoch": 0.36, "grad_norm": 41.0, "learning_rate": 1.0467796610169491e-05, "loss": 1.2599, "step": 14560 }, { "epoch": 0.36, "grad_norm": 18.25, "learning_rate": 1.0461016949152543e-05, "loss": 1.3105, "step": 14570 }, { "epoch": 0.36, "grad_norm": 29.375, "learning_rate": 1.0454237288135594e-05, "loss": 1.4942, "step": 14580 }, { "epoch": 0.36, "grad_norm": 15.8125, "learning_rate": 1.0447457627118644e-05, "loss": 1.5753, "step": 14590 }, { "epoch": 0.36, "grad_norm": 27.0, "learning_rate": 1.0440677966101695e-05, "loss": 1.4115, "step": 14600 }, { "epoch": 0.37, "grad_norm": 32.25, "learning_rate": 1.0433898305084747e-05, "loss": 1.3518, "step": 14610 }, { "epoch": 0.37, "grad_norm": 29.625, "learning_rate": 1.0427118644067798e-05, "loss": 1.522, "step": 14620 }, { "epoch": 0.37, "grad_norm": 25.75, "learning_rate": 1.0420338983050848e-05, "loss": 1.341, "step": 14630 }, { "epoch": 0.37, "grad_norm": 35.25, "learning_rate": 1.04135593220339e-05, "loss": 1.458, "step": 14640 }, { "epoch": 0.37, "grad_norm": 9.5625, "learning_rate": 1.040677966101695e-05, "loss": 1.4959, "step": 14650 }, { "epoch": 0.37, "grad_norm": 30.0, "learning_rate": 1.04e-05, "loss": 1.3038, "step": 14660 }, { "epoch": 0.37, "grad_norm": 29.375, "learning_rate": 1.0393220338983052e-05, "loss": 1.1989, "step": 14670 }, { "epoch": 0.37, "grad_norm": 11.3125, "learning_rate": 1.0386440677966103e-05, "loss": 1.4619, "step": 14680 }, { "epoch": 0.37, "grad_norm": 29.75, "learning_rate": 1.0379661016949155e-05, "loss": 1.386, "step": 14690 }, { "epoch": 0.37, "grad_norm": 20.0, "learning_rate": 1.0372881355932204e-05, "loss": 1.339, "step": 14700 }, { "epoch": 0.37, "grad_norm": 25.0, "learning_rate": 1.0366101694915256e-05, "loss": 1.3886, "step": 14710 }, { "epoch": 0.37, "grad_norm": 24.875, "learning_rate": 1.0359322033898307e-05, "loss": 1.364, "step": 14720 }, { "epoch": 0.37, "grad_norm": 26.0, "learning_rate": 1.0352542372881358e-05, "loss": 1.416, "step": 14730 }, { "epoch": 0.37, "grad_norm": 118.0, "learning_rate": 1.0345762711864406e-05, "loss": 1.3624, "step": 14740 }, { "epoch": 0.37, "grad_norm": 14.375, "learning_rate": 1.0338983050847458e-05, "loss": 1.346, "step": 14750 }, { "epoch": 0.37, "grad_norm": 19.5, "learning_rate": 1.033220338983051e-05, "loss": 1.4017, "step": 14760 }, { "epoch": 0.37, "grad_norm": 14.625, "learning_rate": 1.0325423728813559e-05, "loss": 1.4782, "step": 14770 }, { "epoch": 0.37, "grad_norm": 18.125, "learning_rate": 1.031864406779661e-05, "loss": 1.3452, "step": 14780 }, { "epoch": 0.37, "grad_norm": 8.6875, "learning_rate": 1.0311864406779662e-05, "loss": 1.3968, "step": 14790 }, { "epoch": 0.37, "grad_norm": 29.25, "learning_rate": 1.0305084745762712e-05, "loss": 1.3137, "step": 14800 }, { "epoch": 0.37, "grad_norm": 20.75, "learning_rate": 1.0298305084745763e-05, "loss": 1.4651, "step": 14810 }, { "epoch": 0.37, "grad_norm": 12.4375, "learning_rate": 1.0291525423728814e-05, "loss": 1.468, "step": 14820 }, { "epoch": 0.37, "grad_norm": 19.375, "learning_rate": 1.0284745762711866e-05, "loss": 1.5157, "step": 14830 }, { "epoch": 0.37, "grad_norm": 39.25, "learning_rate": 1.0277966101694915e-05, "loss": 1.4106, "step": 14840 }, { "epoch": 0.37, "grad_norm": 24.25, "learning_rate": 1.0271186440677967e-05, "loss": 1.3044, "step": 14850 }, { "epoch": 0.37, "grad_norm": 7.25, "learning_rate": 1.0264406779661018e-05, "loss": 1.3231, "step": 14860 }, { "epoch": 0.37, "grad_norm": 26.0, "learning_rate": 1.025762711864407e-05, "loss": 1.2322, "step": 14870 }, { "epoch": 0.37, "grad_norm": 16.625, "learning_rate": 1.025084745762712e-05, "loss": 1.1767, "step": 14880 }, { "epoch": 0.37, "grad_norm": 11.625, "learning_rate": 1.024406779661017e-05, "loss": 1.3503, "step": 14890 }, { "epoch": 0.37, "grad_norm": 9.6875, "learning_rate": 1.0237288135593222e-05, "loss": 1.3777, "step": 14900 }, { "epoch": 0.37, "grad_norm": 46.75, "learning_rate": 1.0230508474576274e-05, "loss": 1.3334, "step": 14910 }, { "epoch": 0.37, "grad_norm": 23.25, "learning_rate": 1.0223728813559323e-05, "loss": 1.2998, "step": 14920 }, { "epoch": 0.37, "grad_norm": 45.75, "learning_rate": 1.0216949152542375e-05, "loss": 1.3287, "step": 14930 }, { "epoch": 0.37, "grad_norm": 20.0, "learning_rate": 1.0210169491525423e-05, "loss": 1.4629, "step": 14940 }, { "epoch": 0.37, "grad_norm": 9.375, "learning_rate": 1.0203389830508474e-05, "loss": 1.2114, "step": 14950 }, { "epoch": 0.37, "grad_norm": 35.75, "learning_rate": 1.0196610169491525e-05, "loss": 1.4157, "step": 14960 }, { "epoch": 0.37, "grad_norm": 19.375, "learning_rate": 1.0189830508474577e-05, "loss": 1.3757, "step": 14970 }, { "epoch": 0.37, "grad_norm": 16.75, "learning_rate": 1.0183050847457627e-05, "loss": 1.4091, "step": 14980 }, { "epoch": 0.37, "grad_norm": 12.0, "learning_rate": 1.0176271186440678e-05, "loss": 1.4074, "step": 14990 }, { "epoch": 0.38, "grad_norm": 20.25, "learning_rate": 1.016949152542373e-05, "loss": 1.2817, "step": 15000 }, { "epoch": 0.38, "eval_loss": 1.3379485607147217, "eval_runtime": 59.3062, "eval_samples_per_second": 16.862, "eval_steps_per_second": 16.862, "step": 15000 }, { "epoch": 0.38, "grad_norm": 9.0625, "learning_rate": 1.016271186440678e-05, "loss": 1.4647, "step": 15010 }, { "epoch": 0.38, "grad_norm": 61.0, "learning_rate": 1.015593220338983e-05, "loss": 1.2361, "step": 15020 }, { "epoch": 0.38, "grad_norm": 23.625, "learning_rate": 1.0149152542372882e-05, "loss": 1.2726, "step": 15030 }, { "epoch": 0.38, "grad_norm": 19.375, "learning_rate": 1.0142372881355933e-05, "loss": 1.3574, "step": 15040 }, { "epoch": 0.38, "grad_norm": 49.0, "learning_rate": 1.0135593220338985e-05, "loss": 1.3779, "step": 15050 }, { "epoch": 0.38, "grad_norm": 41.5, "learning_rate": 1.0128813559322034e-05, "loss": 1.1373, "step": 15060 }, { "epoch": 0.38, "grad_norm": 15.1875, "learning_rate": 1.0122033898305086e-05, "loss": 1.5297, "step": 15070 }, { "epoch": 0.38, "grad_norm": 27.375, "learning_rate": 1.0115254237288137e-05, "loss": 1.2211, "step": 15080 }, { "epoch": 0.38, "grad_norm": 10.0625, "learning_rate": 1.0108474576271189e-05, "loss": 1.2089, "step": 15090 }, { "epoch": 0.38, "grad_norm": 23.125, "learning_rate": 1.0101694915254238e-05, "loss": 1.3034, "step": 15100 }, { "epoch": 0.38, "grad_norm": 21.375, "learning_rate": 1.009491525423729e-05, "loss": 1.3993, "step": 15110 }, { "epoch": 0.38, "grad_norm": 25.625, "learning_rate": 1.0088135593220341e-05, "loss": 1.3298, "step": 15120 }, { "epoch": 0.38, "grad_norm": 20.5, "learning_rate": 1.008135593220339e-05, "loss": 1.4189, "step": 15130 }, { "epoch": 0.38, "grad_norm": 35.75, "learning_rate": 1.0074576271186442e-05, "loss": 1.415, "step": 15140 }, { "epoch": 0.38, "grad_norm": 17.625, "learning_rate": 1.0067796610169492e-05, "loss": 1.3125, "step": 15150 }, { "epoch": 0.38, "grad_norm": 10.125, "learning_rate": 1.0061016949152542e-05, "loss": 1.2626, "step": 15160 }, { "epoch": 0.38, "grad_norm": 23.0, "learning_rate": 1.0054237288135593e-05, "loss": 1.2828, "step": 15170 }, { "epoch": 0.38, "grad_norm": 14.75, "learning_rate": 1.0047457627118644e-05, "loss": 1.358, "step": 15180 }, { "epoch": 0.38, "grad_norm": 22.75, "learning_rate": 1.0040677966101696e-05, "loss": 1.2964, "step": 15190 }, { "epoch": 0.38, "grad_norm": 14.0, "learning_rate": 1.0033898305084746e-05, "loss": 1.4093, "step": 15200 }, { "epoch": 0.38, "grad_norm": 58.25, "learning_rate": 1.0027118644067797e-05, "loss": 1.3956, "step": 15210 }, { "epoch": 0.38, "grad_norm": 27.25, "learning_rate": 1.0020338983050848e-05, "loss": 1.3254, "step": 15220 }, { "epoch": 0.38, "grad_norm": 12.75, "learning_rate": 1.00135593220339e-05, "loss": 1.4936, "step": 15230 }, { "epoch": 0.38, "grad_norm": 13.25, "learning_rate": 1.000677966101695e-05, "loss": 1.357, "step": 15240 }, { "epoch": 0.38, "grad_norm": 51.25, "learning_rate": 1e-05, "loss": 1.3777, "step": 15250 }, { "epoch": 0.38, "grad_norm": 33.25, "learning_rate": 9.993220338983052e-06, "loss": 1.3847, "step": 15260 }, { "epoch": 0.38, "grad_norm": 32.75, "learning_rate": 9.986440677966102e-06, "loss": 1.4158, "step": 15270 }, { "epoch": 0.38, "grad_norm": 16.75, "learning_rate": 9.979661016949153e-06, "loss": 1.3884, "step": 15280 }, { "epoch": 0.38, "grad_norm": 35.25, "learning_rate": 9.972881355932205e-06, "loss": 1.3907, "step": 15290 }, { "epoch": 0.38, "grad_norm": 60.5, "learning_rate": 9.966101694915256e-06, "loss": 1.5153, "step": 15300 }, { "epoch": 0.38, "grad_norm": 53.5, "learning_rate": 9.959322033898306e-06, "loss": 1.4849, "step": 15310 }, { "epoch": 0.38, "grad_norm": 17.875, "learning_rate": 9.952542372881356e-06, "loss": 1.4594, "step": 15320 }, { "epoch": 0.38, "grad_norm": 10.25, "learning_rate": 9.945762711864407e-06, "loss": 1.4785, "step": 15330 }, { "epoch": 0.38, "grad_norm": 19.75, "learning_rate": 9.938983050847458e-06, "loss": 1.3351, "step": 15340 }, { "epoch": 0.38, "grad_norm": 10.875, "learning_rate": 9.93220338983051e-06, "loss": 1.3255, "step": 15350 }, { "epoch": 0.38, "grad_norm": 31.0, "learning_rate": 9.92542372881356e-06, "loss": 1.2516, "step": 15360 }, { "epoch": 0.38, "grad_norm": 23.75, "learning_rate": 9.918644067796611e-06, "loss": 1.3803, "step": 15370 }, { "epoch": 0.38, "grad_norm": 13.875, "learning_rate": 9.911864406779662e-06, "loss": 1.5092, "step": 15380 }, { "epoch": 0.38, "grad_norm": 14.25, "learning_rate": 9.905084745762714e-06, "loss": 1.2757, "step": 15390 }, { "epoch": 0.39, "grad_norm": 14.0625, "learning_rate": 9.898305084745763e-06, "loss": 1.44, "step": 15400 }, { "epoch": 0.39, "grad_norm": 35.0, "learning_rate": 9.891525423728813e-06, "loss": 1.2554, "step": 15410 }, { "epoch": 0.39, "grad_norm": 13.0625, "learning_rate": 9.884745762711864e-06, "loss": 1.3457, "step": 15420 }, { "epoch": 0.39, "grad_norm": 55.75, "learning_rate": 9.877966101694916e-06, "loss": 1.3047, "step": 15430 }, { "epoch": 0.39, "grad_norm": 22.625, "learning_rate": 9.871186440677967e-06, "loss": 1.3591, "step": 15440 }, { "epoch": 0.39, "grad_norm": 22.5, "learning_rate": 9.864406779661017e-06, "loss": 1.353, "step": 15450 }, { "epoch": 0.39, "grad_norm": 26.75, "learning_rate": 9.857627118644068e-06, "loss": 1.3824, "step": 15460 }, { "epoch": 0.39, "grad_norm": 20.75, "learning_rate": 9.85084745762712e-06, "loss": 1.5898, "step": 15470 }, { "epoch": 0.39, "grad_norm": 9.25, "learning_rate": 9.844067796610171e-06, "loss": 1.4088, "step": 15480 }, { "epoch": 0.39, "grad_norm": 31.875, "learning_rate": 9.837288135593221e-06, "loss": 1.3672, "step": 15490 }, { "epoch": 0.39, "grad_norm": 24.0, "learning_rate": 9.830508474576272e-06, "loss": 1.364, "step": 15500 }, { "epoch": 0.39, "grad_norm": 32.25, "learning_rate": 9.823728813559322e-06, "loss": 1.4248, "step": 15510 }, { "epoch": 0.39, "grad_norm": 26.875, "learning_rate": 9.816949152542373e-06, "loss": 1.2727, "step": 15520 }, { "epoch": 0.39, "grad_norm": 16.25, "learning_rate": 9.810169491525425e-06, "loss": 1.4068, "step": 15530 }, { "epoch": 0.39, "grad_norm": 33.5, "learning_rate": 9.803389830508474e-06, "loss": 1.4151, "step": 15540 }, { "epoch": 0.39, "grad_norm": 13.0625, "learning_rate": 9.796610169491526e-06, "loss": 1.278, "step": 15550 }, { "epoch": 0.39, "grad_norm": 14.6875, "learning_rate": 9.789830508474577e-06, "loss": 1.2652, "step": 15560 }, { "epoch": 0.39, "grad_norm": 13.5, "learning_rate": 9.783050847457629e-06, "loss": 1.397, "step": 15570 }, { "epoch": 0.39, "grad_norm": 12.9375, "learning_rate": 9.776271186440678e-06, "loss": 1.2682, "step": 15580 }, { "epoch": 0.39, "grad_norm": 34.0, "learning_rate": 9.76949152542373e-06, "loss": 1.152, "step": 15590 }, { "epoch": 0.39, "grad_norm": 81.5, "learning_rate": 9.762711864406781e-06, "loss": 1.3513, "step": 15600 }, { "epoch": 0.39, "grad_norm": 18.5, "learning_rate": 9.755932203389833e-06, "loss": 1.377, "step": 15610 }, { "epoch": 0.39, "grad_norm": 20.625, "learning_rate": 9.749152542372882e-06, "loss": 1.3957, "step": 15620 }, { "epoch": 0.39, "grad_norm": 33.0, "learning_rate": 9.742372881355932e-06, "loss": 1.3222, "step": 15630 }, { "epoch": 0.39, "grad_norm": 12.375, "learning_rate": 9.735593220338983e-06, "loss": 1.4281, "step": 15640 }, { "epoch": 0.39, "grad_norm": 30.25, "learning_rate": 9.728813559322035e-06, "loss": 1.4635, "step": 15650 }, { "epoch": 0.39, "grad_norm": 18.0, "learning_rate": 9.722033898305086e-06, "loss": 1.3443, "step": 15660 }, { "epoch": 0.39, "grad_norm": 76.0, "learning_rate": 9.715254237288136e-06, "loss": 1.4472, "step": 15670 }, { "epoch": 0.39, "grad_norm": 49.5, "learning_rate": 9.708474576271187e-06, "loss": 1.1069, "step": 15680 }, { "epoch": 0.39, "grad_norm": 11.4375, "learning_rate": 9.701694915254239e-06, "loss": 1.5717, "step": 15690 }, { "epoch": 0.39, "grad_norm": 49.75, "learning_rate": 9.69491525423729e-06, "loss": 1.4191, "step": 15700 }, { "epoch": 0.39, "grad_norm": 19.0, "learning_rate": 9.68813559322034e-06, "loss": 1.389, "step": 15710 }, { "epoch": 0.39, "grad_norm": 12.6875, "learning_rate": 9.68135593220339e-06, "loss": 1.1881, "step": 15720 }, { "epoch": 0.39, "grad_norm": 28.5, "learning_rate": 9.674576271186441e-06, "loss": 1.2252, "step": 15730 }, { "epoch": 0.39, "grad_norm": 34.25, "learning_rate": 9.667796610169492e-06, "loss": 1.4493, "step": 15740 }, { "epoch": 0.39, "grad_norm": 21.125, "learning_rate": 9.661016949152544e-06, "loss": 1.4645, "step": 15750 }, { "epoch": 0.39, "grad_norm": 13.1875, "learning_rate": 9.654237288135593e-06, "loss": 1.4717, "step": 15760 }, { "epoch": 0.39, "grad_norm": 11.0625, "learning_rate": 9.647457627118645e-06, "loss": 1.2183, "step": 15770 }, { "epoch": 0.39, "grad_norm": 21.75, "learning_rate": 9.640677966101696e-06, "loss": 1.3174, "step": 15780 }, { "epoch": 0.39, "grad_norm": 11.75, "learning_rate": 9.633898305084746e-06, "loss": 1.3315, "step": 15790 }, { "epoch": 0.4, "grad_norm": 9.875, "learning_rate": 9.627118644067797e-06, "loss": 1.2776, "step": 15800 }, { "epoch": 0.4, "grad_norm": 23.75, "learning_rate": 9.620338983050849e-06, "loss": 1.5261, "step": 15810 }, { "epoch": 0.4, "grad_norm": 11.3125, "learning_rate": 9.6135593220339e-06, "loss": 1.4339, "step": 15820 }, { "epoch": 0.4, "grad_norm": 35.25, "learning_rate": 9.60677966101695e-06, "loss": 1.4765, "step": 15830 }, { "epoch": 0.4, "grad_norm": 108.0, "learning_rate": 9.600000000000001e-06, "loss": 1.3046, "step": 15840 }, { "epoch": 0.4, "grad_norm": 43.75, "learning_rate": 9.593220338983051e-06, "loss": 1.2919, "step": 15850 }, { "epoch": 0.4, "grad_norm": 29.0, "learning_rate": 9.586440677966102e-06, "loss": 1.2598, "step": 15860 }, { "epoch": 0.4, "grad_norm": 13.375, "learning_rate": 9.579661016949154e-06, "loss": 1.3753, "step": 15870 }, { "epoch": 0.4, "grad_norm": 26.75, "learning_rate": 9.572881355932203e-06, "loss": 1.1029, "step": 15880 }, { "epoch": 0.4, "grad_norm": 25.375, "learning_rate": 9.566101694915255e-06, "loss": 1.428, "step": 15890 }, { "epoch": 0.4, "grad_norm": 34.0, "learning_rate": 9.559322033898306e-06, "loss": 1.504, "step": 15900 }, { "epoch": 0.4, "grad_norm": 17.5, "learning_rate": 9.552542372881358e-06, "loss": 1.2561, "step": 15910 }, { "epoch": 0.4, "grad_norm": 22.125, "learning_rate": 9.545762711864407e-06, "loss": 1.4642, "step": 15920 }, { "epoch": 0.4, "grad_norm": 23.75, "learning_rate": 9.538983050847457e-06, "loss": 1.4706, "step": 15930 }, { "epoch": 0.4, "grad_norm": 25.625, "learning_rate": 9.532203389830508e-06, "loss": 1.4855, "step": 15940 }, { "epoch": 0.4, "grad_norm": 52.0, "learning_rate": 9.52542372881356e-06, "loss": 1.3423, "step": 15950 }, { "epoch": 0.4, "grad_norm": 16.875, "learning_rate": 9.518644067796611e-06, "loss": 1.4521, "step": 15960 }, { "epoch": 0.4, "grad_norm": 21.375, "learning_rate": 9.511864406779661e-06, "loss": 1.3876, "step": 15970 }, { "epoch": 0.4, "grad_norm": 26.75, "learning_rate": 9.505084745762712e-06, "loss": 1.3349, "step": 15980 }, { "epoch": 0.4, "grad_norm": 25.0, "learning_rate": 9.498305084745764e-06, "loss": 1.321, "step": 15990 }, { "epoch": 0.4, "grad_norm": 30.25, "learning_rate": 9.491525423728815e-06, "loss": 1.4817, "step": 16000 }, { "epoch": 0.4, "grad_norm": 23.5, "learning_rate": 9.484745762711865e-06, "loss": 1.4348, "step": 16010 }, { "epoch": 0.4, "grad_norm": 6.59375, "learning_rate": 9.477966101694916e-06, "loss": 1.3662, "step": 16020 }, { "epoch": 0.4, "grad_norm": 18.75, "learning_rate": 9.471186440677966e-06, "loss": 1.1267, "step": 16030 }, { "epoch": 0.4, "grad_norm": 16.125, "learning_rate": 9.464406779661017e-06, "loss": 1.3971, "step": 16040 }, { "epoch": 0.4, "grad_norm": 16.125, "learning_rate": 9.457627118644069e-06, "loss": 1.2916, "step": 16050 }, { "epoch": 0.4, "grad_norm": 24.25, "learning_rate": 9.450847457627119e-06, "loss": 1.3208, "step": 16060 }, { "epoch": 0.4, "grad_norm": 11.625, "learning_rate": 9.44406779661017e-06, "loss": 1.2998, "step": 16070 }, { "epoch": 0.4, "grad_norm": 22.25, "learning_rate": 9.437288135593221e-06, "loss": 1.2988, "step": 16080 }, { "epoch": 0.4, "grad_norm": 29.375, "learning_rate": 9.430508474576273e-06, "loss": 1.1747, "step": 16090 }, { "epoch": 0.4, "grad_norm": 51.0, "learning_rate": 9.423728813559322e-06, "loss": 1.4036, "step": 16100 }, { "epoch": 0.4, "grad_norm": 20.625, "learning_rate": 9.416949152542374e-06, "loss": 1.4647, "step": 16110 }, { "epoch": 0.4, "grad_norm": 38.25, "learning_rate": 9.410169491525425e-06, "loss": 1.2687, "step": 16120 }, { "epoch": 0.4, "grad_norm": 21.25, "learning_rate": 9.403389830508477e-06, "loss": 1.2204, "step": 16130 }, { "epoch": 0.4, "grad_norm": 18.625, "learning_rate": 9.396610169491526e-06, "loss": 1.3104, "step": 16140 }, { "epoch": 0.4, "grad_norm": 16.375, "learning_rate": 9.389830508474576e-06, "loss": 1.3218, "step": 16150 }, { "epoch": 0.4, "grad_norm": 12.4375, "learning_rate": 9.383050847457627e-06, "loss": 1.5096, "step": 16160 }, { "epoch": 0.4, "grad_norm": 30.5, "learning_rate": 9.376271186440679e-06, "loss": 1.507, "step": 16170 }, { "epoch": 0.4, "grad_norm": 20.375, "learning_rate": 9.36949152542373e-06, "loss": 1.4757, "step": 16180 }, { "epoch": 0.4, "grad_norm": 20.0, "learning_rate": 9.36271186440678e-06, "loss": 1.2358, "step": 16190 }, { "epoch": 0.41, "grad_norm": 9.25, "learning_rate": 9.355932203389831e-06, "loss": 1.3804, "step": 16200 }, { "epoch": 0.41, "grad_norm": 15.875, "learning_rate": 9.349152542372883e-06, "loss": 1.4316, "step": 16210 }, { "epoch": 0.41, "grad_norm": 14.5, "learning_rate": 9.342372881355934e-06, "loss": 1.2718, "step": 16220 }, { "epoch": 0.41, "grad_norm": 19.25, "learning_rate": 9.335593220338984e-06, "loss": 1.2964, "step": 16230 }, { "epoch": 0.41, "grad_norm": 15.8125, "learning_rate": 9.328813559322034e-06, "loss": 1.4263, "step": 16240 }, { "epoch": 0.41, "grad_norm": 43.5, "learning_rate": 9.322033898305085e-06, "loss": 1.4527, "step": 16250 }, { "epoch": 0.41, "grad_norm": 16.125, "learning_rate": 9.315254237288136e-06, "loss": 1.2035, "step": 16260 }, { "epoch": 0.41, "grad_norm": 25.125, "learning_rate": 9.308474576271188e-06, "loss": 1.2739, "step": 16270 }, { "epoch": 0.41, "grad_norm": 10.0, "learning_rate": 9.301694915254237e-06, "loss": 1.2829, "step": 16280 }, { "epoch": 0.41, "grad_norm": 9.9375, "learning_rate": 9.294915254237289e-06, "loss": 1.4364, "step": 16290 }, { "epoch": 0.41, "grad_norm": 6.5625, "learning_rate": 9.28813559322034e-06, "loss": 1.1994, "step": 16300 }, { "epoch": 0.41, "grad_norm": 14.9375, "learning_rate": 9.28135593220339e-06, "loss": 1.4619, "step": 16310 }, { "epoch": 0.41, "grad_norm": 26.75, "learning_rate": 9.274576271186441e-06, "loss": 1.1941, "step": 16320 }, { "epoch": 0.41, "grad_norm": 20.125, "learning_rate": 9.267796610169493e-06, "loss": 1.3401, "step": 16330 }, { "epoch": 0.41, "grad_norm": 14.6875, "learning_rate": 9.261016949152544e-06, "loss": 1.2734, "step": 16340 }, { "epoch": 0.41, "grad_norm": 14.5, "learning_rate": 9.254237288135594e-06, "loss": 1.4338, "step": 16350 }, { "epoch": 0.41, "grad_norm": 22.75, "learning_rate": 9.247457627118645e-06, "loss": 1.2198, "step": 16360 }, { "epoch": 0.41, "grad_norm": 28.75, "learning_rate": 9.240677966101695e-06, "loss": 1.1812, "step": 16370 }, { "epoch": 0.41, "grad_norm": 13.9375, "learning_rate": 9.233898305084746e-06, "loss": 1.479, "step": 16380 }, { "epoch": 0.41, "grad_norm": 30.375, "learning_rate": 9.227118644067798e-06, "loss": 1.4046, "step": 16390 }, { "epoch": 0.41, "grad_norm": 12.125, "learning_rate": 9.220338983050847e-06, "loss": 1.4306, "step": 16400 }, { "epoch": 0.41, "grad_norm": 8.8125, "learning_rate": 9.213559322033899e-06, "loss": 1.4345, "step": 16410 }, { "epoch": 0.41, "grad_norm": 15.875, "learning_rate": 9.20677966101695e-06, "loss": 1.4688, "step": 16420 }, { "epoch": 0.41, "grad_norm": 28.125, "learning_rate": 9.200000000000002e-06, "loss": 1.1588, "step": 16430 }, { "epoch": 0.41, "grad_norm": 13.375, "learning_rate": 9.193220338983051e-06, "loss": 1.3002, "step": 16440 }, { "epoch": 0.41, "grad_norm": 22.75, "learning_rate": 9.186440677966101e-06, "loss": 1.5033, "step": 16450 }, { "epoch": 0.41, "grad_norm": 14.9375, "learning_rate": 9.179661016949153e-06, "loss": 1.5189, "step": 16460 }, { "epoch": 0.41, "grad_norm": 45.0, "learning_rate": 9.172881355932204e-06, "loss": 1.3223, "step": 16470 }, { "epoch": 0.41, "grad_norm": 34.25, "learning_rate": 9.166101694915255e-06, "loss": 1.4807, "step": 16480 }, { "epoch": 0.41, "grad_norm": 10.125, "learning_rate": 9.159322033898305e-06, "loss": 1.2578, "step": 16490 }, { "epoch": 0.41, "grad_norm": 41.5, "learning_rate": 9.152542372881356e-06, "loss": 1.3199, "step": 16500 }, { "epoch": 0.41, "grad_norm": 27.5, "learning_rate": 9.145762711864408e-06, "loss": 1.5607, "step": 16510 }, { "epoch": 0.41, "grad_norm": 15.625, "learning_rate": 9.13898305084746e-06, "loss": 1.3342, "step": 16520 }, { "epoch": 0.41, "grad_norm": 16.125, "learning_rate": 9.132203389830509e-06, "loss": 1.4272, "step": 16530 }, { "epoch": 0.41, "grad_norm": 25.875, "learning_rate": 9.12542372881356e-06, "loss": 1.2709, "step": 16540 }, { "epoch": 0.41, "grad_norm": 39.0, "learning_rate": 9.11864406779661e-06, "loss": 1.3328, "step": 16550 }, { "epoch": 0.41, "grad_norm": 14.3125, "learning_rate": 9.111864406779661e-06, "loss": 1.4916, "step": 16560 }, { "epoch": 0.41, "grad_norm": 17.875, "learning_rate": 9.105084745762713e-06, "loss": 1.3991, "step": 16570 }, { "epoch": 0.41, "grad_norm": 62.75, "learning_rate": 9.098305084745763e-06, "loss": 1.5126, "step": 16580 }, { "epoch": 0.41, "grad_norm": 30.875, "learning_rate": 9.091525423728814e-06, "loss": 1.4149, "step": 16590 }, { "epoch": 0.41, "grad_norm": 15.0625, "learning_rate": 9.084745762711865e-06, "loss": 1.2595, "step": 16600 }, { "epoch": 0.42, "grad_norm": 14.9375, "learning_rate": 9.077966101694917e-06, "loss": 1.3483, "step": 16610 }, { "epoch": 0.42, "grad_norm": 22.125, "learning_rate": 9.071186440677966e-06, "loss": 1.4082, "step": 16620 }, { "epoch": 0.42, "grad_norm": 10.5, "learning_rate": 9.064406779661018e-06, "loss": 1.5419, "step": 16630 }, { "epoch": 0.42, "grad_norm": 30.625, "learning_rate": 9.05762711864407e-06, "loss": 1.3988, "step": 16640 }, { "epoch": 0.42, "grad_norm": 15.75, "learning_rate": 9.05084745762712e-06, "loss": 1.377, "step": 16650 }, { "epoch": 0.42, "grad_norm": 14.8125, "learning_rate": 9.04406779661017e-06, "loss": 1.2976, "step": 16660 }, { "epoch": 0.42, "grad_norm": 27.75, "learning_rate": 9.03728813559322e-06, "loss": 1.3353, "step": 16670 }, { "epoch": 0.42, "grad_norm": 28.0, "learning_rate": 9.030508474576271e-06, "loss": 1.4809, "step": 16680 }, { "epoch": 0.42, "grad_norm": 29.25, "learning_rate": 9.023728813559323e-06, "loss": 1.2022, "step": 16690 }, { "epoch": 0.42, "grad_norm": 15.1875, "learning_rate": 9.016949152542374e-06, "loss": 1.4174, "step": 16700 }, { "epoch": 0.42, "grad_norm": 27.75, "learning_rate": 9.010169491525424e-06, "loss": 1.2896, "step": 16710 }, { "epoch": 0.42, "grad_norm": 25.875, "learning_rate": 9.003389830508475e-06, "loss": 1.3355, "step": 16720 }, { "epoch": 0.42, "grad_norm": 22.625, "learning_rate": 8.996610169491527e-06, "loss": 1.3372, "step": 16730 }, { "epoch": 0.42, "grad_norm": 19.375, "learning_rate": 8.989830508474578e-06, "loss": 1.3901, "step": 16740 }, { "epoch": 0.42, "grad_norm": 12.8125, "learning_rate": 8.983050847457628e-06, "loss": 1.3096, "step": 16750 }, { "epoch": 0.42, "grad_norm": 13.9375, "learning_rate": 8.976271186440678e-06, "loss": 1.5191, "step": 16760 }, { "epoch": 0.42, "grad_norm": 23.125, "learning_rate": 8.969491525423729e-06, "loss": 1.4531, "step": 16770 }, { "epoch": 0.42, "grad_norm": 25.875, "learning_rate": 8.96271186440678e-06, "loss": 1.5546, "step": 16780 }, { "epoch": 0.42, "grad_norm": 26.125, "learning_rate": 8.955932203389832e-06, "loss": 1.1527, "step": 16790 }, { "epoch": 0.42, "grad_norm": 32.75, "learning_rate": 8.949152542372881e-06, "loss": 1.3397, "step": 16800 }, { "epoch": 0.42, "grad_norm": 30.75, "learning_rate": 8.942372881355933e-06, "loss": 1.3663, "step": 16810 }, { "epoch": 0.42, "grad_norm": 13.5625, "learning_rate": 8.935593220338984e-06, "loss": 1.3837, "step": 16820 }, { "epoch": 0.42, "grad_norm": 12.6875, "learning_rate": 8.928813559322036e-06, "loss": 1.4386, "step": 16830 }, { "epoch": 0.42, "grad_norm": 24.75, "learning_rate": 8.922033898305085e-06, "loss": 1.4903, "step": 16840 }, { "epoch": 0.42, "grad_norm": 19.0, "learning_rate": 8.915254237288137e-06, "loss": 1.4548, "step": 16850 }, { "epoch": 0.42, "grad_norm": 26.25, "learning_rate": 8.908474576271188e-06, "loss": 1.251, "step": 16860 }, { "epoch": 0.42, "grad_norm": 20.375, "learning_rate": 8.901694915254238e-06, "loss": 1.4531, "step": 16870 }, { "epoch": 0.42, "grad_norm": 20.375, "learning_rate": 8.89491525423729e-06, "loss": 1.2677, "step": 16880 }, { "epoch": 0.42, "grad_norm": 22.875, "learning_rate": 8.888135593220339e-06, "loss": 1.3289, "step": 16890 }, { "epoch": 0.42, "grad_norm": 26.125, "learning_rate": 8.88135593220339e-06, "loss": 1.5637, "step": 16900 }, { "epoch": 0.42, "grad_norm": 47.0, "learning_rate": 8.874576271186442e-06, "loss": 1.3166, "step": 16910 }, { "epoch": 0.42, "grad_norm": 19.25, "learning_rate": 8.867796610169492e-06, "loss": 1.2822, "step": 16920 }, { "epoch": 0.42, "grad_norm": 34.25, "learning_rate": 8.861016949152543e-06, "loss": 1.3943, "step": 16930 }, { "epoch": 0.42, "grad_norm": 10.125, "learning_rate": 8.854237288135594e-06, "loss": 1.3883, "step": 16940 }, { "epoch": 0.42, "grad_norm": 33.25, "learning_rate": 8.847457627118646e-06, "loss": 1.3113, "step": 16950 }, { "epoch": 0.42, "grad_norm": 27.75, "learning_rate": 8.840677966101695e-06, "loss": 1.3963, "step": 16960 }, { "epoch": 0.42, "grad_norm": 21.5, "learning_rate": 8.833898305084747e-06, "loss": 1.5098, "step": 16970 }, { "epoch": 0.42, "grad_norm": 20.0, "learning_rate": 8.827118644067797e-06, "loss": 1.4349, "step": 16980 }, { "epoch": 0.42, "grad_norm": 22.25, "learning_rate": 8.820338983050848e-06, "loss": 1.4513, "step": 16990 }, { "epoch": 0.42, "grad_norm": 9.5, "learning_rate": 8.8135593220339e-06, "loss": 1.1621, "step": 17000 }, { "epoch": 0.43, "grad_norm": 13.8125, "learning_rate": 8.806779661016949e-06, "loss": 1.2259, "step": 17010 }, { "epoch": 0.43, "grad_norm": 14.1875, "learning_rate": 8.8e-06, "loss": 1.2985, "step": 17020 }, { "epoch": 0.43, "grad_norm": 36.75, "learning_rate": 8.793220338983052e-06, "loss": 1.4038, "step": 17030 }, { "epoch": 0.43, "grad_norm": 11.875, "learning_rate": 8.786440677966103e-06, "loss": 1.1619, "step": 17040 }, { "epoch": 0.43, "grad_norm": 16.25, "learning_rate": 8.779661016949153e-06, "loss": 1.429, "step": 17050 }, { "epoch": 0.43, "grad_norm": 22.75, "learning_rate": 8.772881355932204e-06, "loss": 1.06, "step": 17060 }, { "epoch": 0.43, "grad_norm": 11.5, "learning_rate": 8.766101694915254e-06, "loss": 1.4852, "step": 17070 }, { "epoch": 0.43, "grad_norm": 29.625, "learning_rate": 8.759322033898305e-06, "loss": 1.2247, "step": 17080 }, { "epoch": 0.43, "grad_norm": 12.1875, "learning_rate": 8.752542372881357e-06, "loss": 1.3069, "step": 17090 }, { "epoch": 0.43, "grad_norm": 21.875, "learning_rate": 8.745762711864407e-06, "loss": 1.3806, "step": 17100 }, { "epoch": 0.43, "grad_norm": 11.25, "learning_rate": 8.738983050847458e-06, "loss": 1.2146, "step": 17110 }, { "epoch": 0.43, "grad_norm": 30.0, "learning_rate": 8.73220338983051e-06, "loss": 1.2134, "step": 17120 }, { "epoch": 0.43, "grad_norm": 16.5, "learning_rate": 8.72542372881356e-06, "loss": 1.479, "step": 17130 }, { "epoch": 0.43, "grad_norm": 21.0, "learning_rate": 8.71864406779661e-06, "loss": 1.2287, "step": 17140 }, { "epoch": 0.43, "grad_norm": 29.0, "learning_rate": 8.711864406779662e-06, "loss": 1.3519, "step": 17150 }, { "epoch": 0.43, "grad_norm": 35.0, "learning_rate": 8.705084745762713e-06, "loss": 1.2435, "step": 17160 }, { "epoch": 0.43, "grad_norm": 22.625, "learning_rate": 8.698305084745765e-06, "loss": 1.2609, "step": 17170 }, { "epoch": 0.43, "grad_norm": 47.5, "learning_rate": 8.691525423728814e-06, "loss": 1.2522, "step": 17180 }, { "epoch": 0.43, "grad_norm": 20.875, "learning_rate": 8.684745762711864e-06, "loss": 1.2606, "step": 17190 }, { "epoch": 0.43, "grad_norm": 38.0, "learning_rate": 8.677966101694915e-06, "loss": 1.3564, "step": 17200 }, { "epoch": 0.43, "grad_norm": 13.0625, "learning_rate": 8.671186440677967e-06, "loss": 1.4235, "step": 17210 }, { "epoch": 0.43, "grad_norm": 24.875, "learning_rate": 8.664406779661018e-06, "loss": 1.1742, "step": 17220 }, { "epoch": 0.43, "grad_norm": 52.0, "learning_rate": 8.657627118644068e-06, "loss": 1.3956, "step": 17230 }, { "epoch": 0.43, "grad_norm": 29.875, "learning_rate": 8.65084745762712e-06, "loss": 1.3366, "step": 17240 }, { "epoch": 0.43, "grad_norm": 19.75, "learning_rate": 8.64406779661017e-06, "loss": 1.4179, "step": 17250 }, { "epoch": 0.43, "grad_norm": 33.0, "learning_rate": 8.637288135593222e-06, "loss": 1.4633, "step": 17260 }, { "epoch": 0.43, "grad_norm": 24.875, "learning_rate": 8.630508474576272e-06, "loss": 1.3481, "step": 17270 }, { "epoch": 0.43, "grad_norm": 14.25, "learning_rate": 8.623728813559322e-06, "loss": 1.3081, "step": 17280 }, { "epoch": 0.43, "grad_norm": 21.125, "learning_rate": 8.616949152542373e-06, "loss": 1.5585, "step": 17290 }, { "epoch": 0.43, "grad_norm": 14.1875, "learning_rate": 8.610169491525424e-06, "loss": 1.4156, "step": 17300 }, { "epoch": 0.43, "grad_norm": 20.25, "learning_rate": 8.603389830508476e-06, "loss": 1.255, "step": 17310 }, { "epoch": 0.43, "grad_norm": 34.0, "learning_rate": 8.596610169491526e-06, "loss": 1.3487, "step": 17320 }, { "epoch": 0.43, "grad_norm": 18.875, "learning_rate": 8.589830508474577e-06, "loss": 1.28, "step": 17330 }, { "epoch": 0.43, "grad_norm": 10.0625, "learning_rate": 8.583050847457628e-06, "loss": 1.3084, "step": 17340 }, { "epoch": 0.43, "grad_norm": 44.75, "learning_rate": 8.57627118644068e-06, "loss": 1.2927, "step": 17350 }, { "epoch": 0.43, "grad_norm": 23.75, "learning_rate": 8.56949152542373e-06, "loss": 1.2896, "step": 17360 }, { "epoch": 0.43, "grad_norm": 9.25, "learning_rate": 8.56271186440678e-06, "loss": 1.3518, "step": 17370 }, { "epoch": 0.43, "grad_norm": 15.0625, "learning_rate": 8.55593220338983e-06, "loss": 1.4414, "step": 17380 }, { "epoch": 0.43, "grad_norm": 23.125, "learning_rate": 8.549152542372882e-06, "loss": 1.4226, "step": 17390 }, { "epoch": 0.43, "grad_norm": 25.25, "learning_rate": 8.542372881355933e-06, "loss": 1.5472, "step": 17400 }, { "epoch": 0.44, "grad_norm": 25.375, "learning_rate": 8.535593220338983e-06, "loss": 1.2825, "step": 17410 }, { "epoch": 0.44, "grad_norm": 13.1875, "learning_rate": 8.528813559322034e-06, "loss": 1.3777, "step": 17420 }, { "epoch": 0.44, "grad_norm": 22.75, "learning_rate": 8.522033898305086e-06, "loss": 1.3355, "step": 17430 }, { "epoch": 0.44, "grad_norm": 13.4375, "learning_rate": 8.515254237288136e-06, "loss": 1.4221, "step": 17440 }, { "epoch": 0.44, "grad_norm": 32.5, "learning_rate": 8.508474576271187e-06, "loss": 1.302, "step": 17450 }, { "epoch": 0.44, "grad_norm": 17.125, "learning_rate": 8.501694915254238e-06, "loss": 1.1499, "step": 17460 }, { "epoch": 0.44, "grad_norm": 32.0, "learning_rate": 8.49491525423729e-06, "loss": 1.1878, "step": 17470 }, { "epoch": 0.44, "grad_norm": 46.25, "learning_rate": 8.48813559322034e-06, "loss": 1.5593, "step": 17480 }, { "epoch": 0.44, "grad_norm": 21.0, "learning_rate": 8.481355932203391e-06, "loss": 1.3942, "step": 17490 }, { "epoch": 0.44, "grad_norm": 11.6875, "learning_rate": 8.47457627118644e-06, "loss": 1.4905, "step": 17500 }, { "epoch": 0.44, "grad_norm": 25.875, "learning_rate": 8.467796610169492e-06, "loss": 1.327, "step": 17510 }, { "epoch": 0.44, "grad_norm": 10.125, "learning_rate": 8.461016949152543e-06, "loss": 1.4815, "step": 17520 }, { "epoch": 0.44, "grad_norm": 17.25, "learning_rate": 8.454237288135593e-06, "loss": 1.3061, "step": 17530 }, { "epoch": 0.44, "grad_norm": 13.25, "learning_rate": 8.447457627118644e-06, "loss": 1.2559, "step": 17540 }, { "epoch": 0.44, "grad_norm": 8.125, "learning_rate": 8.440677966101696e-06, "loss": 1.3552, "step": 17550 }, { "epoch": 0.44, "grad_norm": 8.625, "learning_rate": 8.433898305084747e-06, "loss": 1.3366, "step": 17560 }, { "epoch": 0.44, "grad_norm": 25.5, "learning_rate": 8.427118644067797e-06, "loss": 1.3192, "step": 17570 }, { "epoch": 0.44, "grad_norm": 35.75, "learning_rate": 8.420338983050848e-06, "loss": 1.3305, "step": 17580 }, { "epoch": 0.44, "grad_norm": 22.0, "learning_rate": 8.413559322033898e-06, "loss": 1.2914, "step": 17590 }, { "epoch": 0.44, "grad_norm": 51.75, "learning_rate": 8.40677966101695e-06, "loss": 1.3449, "step": 17600 }, { "epoch": 0.44, "grad_norm": 55.0, "learning_rate": 8.400000000000001e-06, "loss": 1.2464, "step": 17610 }, { "epoch": 0.44, "grad_norm": 20.625, "learning_rate": 8.39322033898305e-06, "loss": 1.4602, "step": 17620 }, { "epoch": 0.44, "grad_norm": 41.25, "learning_rate": 8.386440677966102e-06, "loss": 1.3996, "step": 17630 }, { "epoch": 0.44, "grad_norm": 10.875, "learning_rate": 8.379661016949153e-06, "loss": 1.3882, "step": 17640 }, { "epoch": 0.44, "grad_norm": 28.25, "learning_rate": 8.372881355932205e-06, "loss": 1.4983, "step": 17650 }, { "epoch": 0.44, "grad_norm": 28.5, "learning_rate": 8.366101694915255e-06, "loss": 1.4799, "step": 17660 }, { "epoch": 0.44, "grad_norm": 34.25, "learning_rate": 8.359322033898306e-06, "loss": 1.4551, "step": 17670 }, { "epoch": 0.44, "grad_norm": 9.0625, "learning_rate": 8.352542372881357e-06, "loss": 1.3426, "step": 17680 }, { "epoch": 0.44, "grad_norm": 24.0, "learning_rate": 8.345762711864409e-06, "loss": 1.0709, "step": 17690 }, { "epoch": 0.44, "grad_norm": 21.75, "learning_rate": 8.338983050847458e-06, "loss": 1.2716, "step": 17700 }, { "epoch": 0.44, "grad_norm": 22.125, "learning_rate": 8.332203389830508e-06, "loss": 1.258, "step": 17710 }, { "epoch": 0.44, "grad_norm": 22.375, "learning_rate": 8.32542372881356e-06, "loss": 1.4624, "step": 17720 }, { "epoch": 0.44, "grad_norm": 20.25, "learning_rate": 8.318644067796611e-06, "loss": 1.1973, "step": 17730 }, { "epoch": 0.44, "grad_norm": 23.625, "learning_rate": 8.311864406779662e-06, "loss": 1.3575, "step": 17740 }, { "epoch": 0.44, "grad_norm": 21.75, "learning_rate": 8.305084745762712e-06, "loss": 1.4038, "step": 17750 }, { "epoch": 0.44, "grad_norm": 13.875, "learning_rate": 8.298305084745763e-06, "loss": 1.2004, "step": 17760 }, { "epoch": 0.44, "grad_norm": 13.125, "learning_rate": 8.291525423728815e-06, "loss": 1.4105, "step": 17770 }, { "epoch": 0.44, "grad_norm": 23.75, "learning_rate": 8.284745762711866e-06, "loss": 1.4087, "step": 17780 }, { "epoch": 0.44, "grad_norm": 17.5, "learning_rate": 8.277966101694916e-06, "loss": 1.3011, "step": 17790 }, { "epoch": 0.45, "grad_norm": 22.375, "learning_rate": 8.271186440677966e-06, "loss": 1.3587, "step": 17800 }, { "epoch": 0.45, "grad_norm": 21.25, "learning_rate": 8.264406779661017e-06, "loss": 1.3222, "step": 17810 }, { "epoch": 0.45, "grad_norm": 17.625, "learning_rate": 8.257627118644068e-06, "loss": 1.314, "step": 17820 }, { "epoch": 0.45, "grad_norm": 18.875, "learning_rate": 8.25084745762712e-06, "loss": 1.2684, "step": 17830 }, { "epoch": 0.45, "grad_norm": 14.125, "learning_rate": 8.24406779661017e-06, "loss": 1.3755, "step": 17840 }, { "epoch": 0.45, "grad_norm": 36.0, "learning_rate": 8.237288135593221e-06, "loss": 1.4589, "step": 17850 }, { "epoch": 0.45, "grad_norm": 49.0, "learning_rate": 8.230508474576272e-06, "loss": 1.3277, "step": 17860 }, { "epoch": 0.45, "grad_norm": 15.5, "learning_rate": 8.223728813559324e-06, "loss": 1.1487, "step": 17870 }, { "epoch": 0.45, "grad_norm": 55.0, "learning_rate": 8.216949152542373e-06, "loss": 1.3018, "step": 17880 }, { "epoch": 0.45, "grad_norm": 12.75, "learning_rate": 8.210169491525425e-06, "loss": 1.5137, "step": 17890 }, { "epoch": 0.45, "grad_norm": 14.4375, "learning_rate": 8.203389830508475e-06, "loss": 1.412, "step": 17900 }, { "epoch": 0.45, "grad_norm": 14.8125, "learning_rate": 8.196610169491526e-06, "loss": 1.3836, "step": 17910 }, { "epoch": 0.45, "grad_norm": 30.0, "learning_rate": 8.189830508474577e-06, "loss": 1.3612, "step": 17920 }, { "epoch": 0.45, "grad_norm": 56.0, "learning_rate": 8.183050847457627e-06, "loss": 1.3871, "step": 17930 }, { "epoch": 0.45, "grad_norm": 12.6875, "learning_rate": 8.176271186440678e-06, "loss": 1.335, "step": 17940 }, { "epoch": 0.45, "grad_norm": 38.75, "learning_rate": 8.16949152542373e-06, "loss": 1.3483, "step": 17950 }, { "epoch": 0.45, "grad_norm": 17.5, "learning_rate": 8.162711864406781e-06, "loss": 1.3832, "step": 17960 }, { "epoch": 0.45, "grad_norm": 78.5, "learning_rate": 8.155932203389831e-06, "loss": 1.6168, "step": 17970 }, { "epoch": 0.45, "grad_norm": 18.5, "learning_rate": 8.149152542372882e-06, "loss": 1.2728, "step": 17980 }, { "epoch": 0.45, "grad_norm": 19.5, "learning_rate": 8.142372881355934e-06, "loss": 1.3424, "step": 17990 }, { "epoch": 0.45, "grad_norm": 22.5, "learning_rate": 8.135593220338983e-06, "loss": 1.2738, "step": 18000 }, { "epoch": 0.45, "grad_norm": 9.9375, "learning_rate": 8.128813559322035e-06, "loss": 1.2798, "step": 18010 }, { "epoch": 0.45, "grad_norm": 19.625, "learning_rate": 8.122033898305085e-06, "loss": 1.4365, "step": 18020 }, { "epoch": 0.45, "grad_norm": 12.9375, "learning_rate": 8.115254237288136e-06, "loss": 1.5073, "step": 18030 }, { "epoch": 0.45, "grad_norm": 23.0, "learning_rate": 8.108474576271187e-06, "loss": 1.3122, "step": 18040 }, { "epoch": 0.45, "grad_norm": 22.75, "learning_rate": 8.101694915254237e-06, "loss": 1.4867, "step": 18050 }, { "epoch": 0.45, "grad_norm": 33.5, "learning_rate": 8.094915254237289e-06, "loss": 1.3197, "step": 18060 }, { "epoch": 0.45, "grad_norm": 25.25, "learning_rate": 8.08813559322034e-06, "loss": 1.1822, "step": 18070 }, { "epoch": 0.45, "grad_norm": 28.25, "learning_rate": 8.081355932203391e-06, "loss": 1.3108, "step": 18080 }, { "epoch": 0.45, "grad_norm": 49.75, "learning_rate": 8.074576271186441e-06, "loss": 1.4321, "step": 18090 }, { "epoch": 0.45, "grad_norm": 30.625, "learning_rate": 8.067796610169492e-06, "loss": 1.3081, "step": 18100 }, { "epoch": 0.45, "grad_norm": 10.375, "learning_rate": 8.061016949152542e-06, "loss": 1.2814, "step": 18110 }, { "epoch": 0.45, "grad_norm": 26.0, "learning_rate": 8.054237288135594e-06, "loss": 1.2951, "step": 18120 }, { "epoch": 0.45, "grad_norm": 59.25, "learning_rate": 8.047457627118645e-06, "loss": 1.3096, "step": 18130 }, { "epoch": 0.45, "grad_norm": 29.625, "learning_rate": 8.040677966101695e-06, "loss": 1.3775, "step": 18140 }, { "epoch": 0.45, "grad_norm": 36.0, "learning_rate": 8.033898305084746e-06, "loss": 1.2384, "step": 18150 }, { "epoch": 0.45, "grad_norm": 23.875, "learning_rate": 8.027118644067797e-06, "loss": 1.4262, "step": 18160 }, { "epoch": 0.45, "grad_norm": 26.125, "learning_rate": 8.020338983050849e-06, "loss": 1.1021, "step": 18170 }, { "epoch": 0.45, "grad_norm": 22.75, "learning_rate": 8.013559322033899e-06, "loss": 1.3957, "step": 18180 }, { "epoch": 0.45, "grad_norm": 27.875, "learning_rate": 8.00677966101695e-06, "loss": 1.4021, "step": 18190 }, { "epoch": 0.46, "grad_norm": 22.25, "learning_rate": 8.000000000000001e-06, "loss": 1.4371, "step": 18200 }, { "epoch": 0.46, "grad_norm": 30.5, "learning_rate": 7.993220338983053e-06, "loss": 1.3273, "step": 18210 }, { "epoch": 0.46, "grad_norm": 24.375, "learning_rate": 7.986440677966102e-06, "loss": 1.3039, "step": 18220 }, { "epoch": 0.46, "grad_norm": 18.125, "learning_rate": 7.979661016949152e-06, "loss": 1.2717, "step": 18230 }, { "epoch": 0.46, "grad_norm": 62.75, "learning_rate": 7.972881355932204e-06, "loss": 1.4438, "step": 18240 }, { "epoch": 0.46, "grad_norm": 47.0, "learning_rate": 7.966101694915255e-06, "loss": 1.5599, "step": 18250 }, { "epoch": 0.46, "grad_norm": 11.0, "learning_rate": 7.959322033898306e-06, "loss": 1.3485, "step": 18260 }, { "epoch": 0.46, "grad_norm": 16.875, "learning_rate": 7.952542372881356e-06, "loss": 1.428, "step": 18270 }, { "epoch": 0.46, "grad_norm": 35.5, "learning_rate": 7.945762711864407e-06, "loss": 1.3857, "step": 18280 }, { "epoch": 0.46, "grad_norm": 30.25, "learning_rate": 7.938983050847459e-06, "loss": 1.1532, "step": 18290 }, { "epoch": 0.46, "grad_norm": 22.375, "learning_rate": 7.93220338983051e-06, "loss": 1.2357, "step": 18300 }, { "epoch": 0.46, "grad_norm": 29.25, "learning_rate": 7.92542372881356e-06, "loss": 1.337, "step": 18310 }, { "epoch": 0.46, "grad_norm": 13.3125, "learning_rate": 7.91864406779661e-06, "loss": 1.3956, "step": 18320 }, { "epoch": 0.46, "grad_norm": 36.75, "learning_rate": 7.911864406779661e-06, "loss": 1.2387, "step": 18330 }, { "epoch": 0.46, "grad_norm": 51.75, "learning_rate": 7.905084745762712e-06, "loss": 1.305, "step": 18340 }, { "epoch": 0.46, "grad_norm": 28.625, "learning_rate": 7.898305084745764e-06, "loss": 1.2429, "step": 18350 }, { "epoch": 0.46, "grad_norm": 24.375, "learning_rate": 7.891525423728814e-06, "loss": 1.3913, "step": 18360 }, { "epoch": 0.46, "grad_norm": 20.125, "learning_rate": 7.884745762711865e-06, "loss": 1.5295, "step": 18370 }, { "epoch": 0.46, "grad_norm": 25.625, "learning_rate": 7.877966101694916e-06, "loss": 1.3455, "step": 18380 }, { "epoch": 0.46, "grad_norm": 17.375, "learning_rate": 7.871186440677968e-06, "loss": 1.5508, "step": 18390 }, { "epoch": 0.46, "grad_norm": 71.0, "learning_rate": 7.864406779661017e-06, "loss": 1.2648, "step": 18400 }, { "epoch": 0.46, "grad_norm": 28.875, "learning_rate": 7.857627118644069e-06, "loss": 1.2369, "step": 18410 }, { "epoch": 0.46, "grad_norm": 18.25, "learning_rate": 7.850847457627119e-06, "loss": 1.5172, "step": 18420 }, { "epoch": 0.46, "grad_norm": 20.625, "learning_rate": 7.84406779661017e-06, "loss": 1.3711, "step": 18430 }, { "epoch": 0.46, "grad_norm": 12.6875, "learning_rate": 7.837288135593221e-06, "loss": 1.4377, "step": 18440 }, { "epoch": 0.46, "grad_norm": 28.5, "learning_rate": 7.830508474576271e-06, "loss": 1.303, "step": 18450 }, { "epoch": 0.46, "grad_norm": 18.125, "learning_rate": 7.823728813559322e-06, "loss": 1.2747, "step": 18460 }, { "epoch": 0.46, "grad_norm": 29.75, "learning_rate": 7.816949152542374e-06, "loss": 1.3278, "step": 18470 }, { "epoch": 0.46, "grad_norm": 17.25, "learning_rate": 7.810169491525425e-06, "loss": 1.558, "step": 18480 }, { "epoch": 0.46, "grad_norm": 10.625, "learning_rate": 7.803389830508475e-06, "loss": 1.4647, "step": 18490 }, { "epoch": 0.46, "grad_norm": 21.25, "learning_rate": 7.796610169491526e-06, "loss": 1.3819, "step": 18500 }, { "epoch": 0.46, "grad_norm": 13.875, "learning_rate": 7.789830508474578e-06, "loss": 1.3907, "step": 18510 }, { "epoch": 0.46, "grad_norm": 20.375, "learning_rate": 7.783050847457628e-06, "loss": 1.3471, "step": 18520 }, { "epoch": 0.46, "grad_norm": 18.5, "learning_rate": 7.776271186440679e-06, "loss": 1.5287, "step": 18530 }, { "epoch": 0.46, "grad_norm": 21.25, "learning_rate": 7.769491525423729e-06, "loss": 1.4263, "step": 18540 }, { "epoch": 0.46, "grad_norm": 10.125, "learning_rate": 7.76271186440678e-06, "loss": 1.1069, "step": 18550 }, { "epoch": 0.46, "grad_norm": 13.1875, "learning_rate": 7.755932203389831e-06, "loss": 1.2873, "step": 18560 }, { "epoch": 0.46, "grad_norm": 45.5, "learning_rate": 7.749152542372881e-06, "loss": 1.3418, "step": 18570 }, { "epoch": 0.46, "grad_norm": 15.625, "learning_rate": 7.742372881355933e-06, "loss": 1.2194, "step": 18580 }, { "epoch": 0.46, "grad_norm": 18.25, "learning_rate": 7.735593220338984e-06, "loss": 1.3678, "step": 18590 }, { "epoch": 0.47, "grad_norm": 22.375, "learning_rate": 7.728813559322035e-06, "loss": 1.3842, "step": 18600 }, { "epoch": 0.47, "grad_norm": 35.75, "learning_rate": 7.722033898305085e-06, "loss": 1.1206, "step": 18610 }, { "epoch": 0.47, "grad_norm": 30.25, "learning_rate": 7.715254237288136e-06, "loss": 1.3278, "step": 18620 }, { "epoch": 0.47, "grad_norm": 11.75, "learning_rate": 7.708474576271186e-06, "loss": 1.3577, "step": 18630 }, { "epoch": 0.47, "grad_norm": 22.875, "learning_rate": 7.701694915254238e-06, "loss": 1.4379, "step": 18640 }, { "epoch": 0.47, "grad_norm": 15.75, "learning_rate": 7.694915254237289e-06, "loss": 1.4008, "step": 18650 }, { "epoch": 0.47, "grad_norm": 30.25, "learning_rate": 7.688135593220339e-06, "loss": 1.3135, "step": 18660 }, { "epoch": 0.47, "grad_norm": 18.0, "learning_rate": 7.68135593220339e-06, "loss": 1.3952, "step": 18670 }, { "epoch": 0.47, "grad_norm": 16.875, "learning_rate": 7.674576271186441e-06, "loss": 1.3399, "step": 18680 }, { "epoch": 0.47, "grad_norm": 27.25, "learning_rate": 7.667796610169493e-06, "loss": 1.2071, "step": 18690 }, { "epoch": 0.47, "grad_norm": 20.875, "learning_rate": 7.661016949152543e-06, "loss": 1.3875, "step": 18700 }, { "epoch": 0.47, "grad_norm": 70.0, "learning_rate": 7.654237288135594e-06, "loss": 1.2638, "step": 18710 }, { "epoch": 0.47, "grad_norm": 20.0, "learning_rate": 7.647457627118645e-06, "loss": 1.4459, "step": 18720 }, { "epoch": 0.47, "grad_norm": 17.125, "learning_rate": 7.640677966101695e-06, "loss": 1.4673, "step": 18730 }, { "epoch": 0.47, "grad_norm": 22.625, "learning_rate": 7.633898305084746e-06, "loss": 1.464, "step": 18740 }, { "epoch": 0.47, "grad_norm": 11.0, "learning_rate": 7.627118644067797e-06, "loss": 1.3678, "step": 18750 }, { "epoch": 0.47, "grad_norm": 39.5, "learning_rate": 7.6203389830508476e-06, "loss": 1.4107, "step": 18760 }, { "epoch": 0.47, "grad_norm": 18.25, "learning_rate": 7.613559322033899e-06, "loss": 1.2465, "step": 18770 }, { "epoch": 0.47, "grad_norm": 43.75, "learning_rate": 7.6067796610169495e-06, "loss": 1.3577, "step": 18780 }, { "epoch": 0.47, "grad_norm": 24.0, "learning_rate": 7.600000000000001e-06, "loss": 1.3036, "step": 18790 }, { "epoch": 0.47, "grad_norm": 16.625, "learning_rate": 7.5932203389830515e-06, "loss": 1.3213, "step": 18800 }, { "epoch": 0.47, "grad_norm": 44.0, "learning_rate": 7.586440677966103e-06, "loss": 1.378, "step": 18810 }, { "epoch": 0.47, "grad_norm": 11.1875, "learning_rate": 7.5796610169491534e-06, "loss": 1.2166, "step": 18820 }, { "epoch": 0.47, "grad_norm": 16.75, "learning_rate": 7.572881355932205e-06, "loss": 1.3052, "step": 18830 }, { "epoch": 0.47, "grad_norm": 47.75, "learning_rate": 7.5661016949152545e-06, "loss": 1.2388, "step": 18840 }, { "epoch": 0.47, "grad_norm": 17.25, "learning_rate": 7.559322033898305e-06, "loss": 1.5168, "step": 18850 }, { "epoch": 0.47, "grad_norm": 42.75, "learning_rate": 7.5525423728813565e-06, "loss": 1.062, "step": 18860 }, { "epoch": 0.47, "grad_norm": 33.5, "learning_rate": 7.545762711864407e-06, "loss": 1.4678, "step": 18870 }, { "epoch": 0.47, "grad_norm": 35.0, "learning_rate": 7.5389830508474584e-06, "loss": 1.3447, "step": 18880 }, { "epoch": 0.47, "grad_norm": 18.5, "learning_rate": 7.532203389830509e-06, "loss": 1.3211, "step": 18890 }, { "epoch": 0.47, "grad_norm": 48.5, "learning_rate": 7.52542372881356e-06, "loss": 1.4629, "step": 18900 }, { "epoch": 0.47, "grad_norm": 17.625, "learning_rate": 7.518644067796611e-06, "loss": 1.3007, "step": 18910 }, { "epoch": 0.47, "grad_norm": 30.125, "learning_rate": 7.511864406779662e-06, "loss": 1.3632, "step": 18920 }, { "epoch": 0.47, "grad_norm": 40.0, "learning_rate": 7.505084745762713e-06, "loss": 1.3538, "step": 18930 }, { "epoch": 0.47, "grad_norm": 48.25, "learning_rate": 7.498305084745763e-06, "loss": 1.2519, "step": 18940 }, { "epoch": 0.47, "grad_norm": 34.0, "learning_rate": 7.491525423728814e-06, "loss": 1.1823, "step": 18950 }, { "epoch": 0.47, "grad_norm": 43.5, "learning_rate": 7.4847457627118646e-06, "loss": 1.2593, "step": 18960 }, { "epoch": 0.47, "grad_norm": 19.25, "learning_rate": 7.477966101694916e-06, "loss": 1.3109, "step": 18970 }, { "epoch": 0.47, "grad_norm": 43.0, "learning_rate": 7.4711864406779665e-06, "loss": 1.3609, "step": 18980 }, { "epoch": 0.47, "grad_norm": 23.0, "learning_rate": 7.464406779661018e-06, "loss": 1.366, "step": 18990 }, { "epoch": 0.47, "grad_norm": 25.75, "learning_rate": 7.4576271186440685e-06, "loss": 1.3806, "step": 19000 }, { "epoch": 0.48, "grad_norm": 10.75, "learning_rate": 7.45084745762712e-06, "loss": 1.2656, "step": 19010 }, { "epoch": 0.48, "grad_norm": 28.625, "learning_rate": 7.4440677966101704e-06, "loss": 1.3377, "step": 19020 }, { "epoch": 0.48, "grad_norm": 17.875, "learning_rate": 7.437288135593221e-06, "loss": 1.3975, "step": 19030 }, { "epoch": 0.48, "grad_norm": 45.0, "learning_rate": 7.430508474576272e-06, "loss": 1.1749, "step": 19040 }, { "epoch": 0.48, "grad_norm": 13.375, "learning_rate": 7.423728813559322e-06, "loss": 1.4238, "step": 19050 }, { "epoch": 0.48, "grad_norm": 10.9375, "learning_rate": 7.4169491525423735e-06, "loss": 1.366, "step": 19060 }, { "epoch": 0.48, "grad_norm": 40.25, "learning_rate": 7.410169491525424e-06, "loss": 1.4467, "step": 19070 }, { "epoch": 0.48, "grad_norm": 15.4375, "learning_rate": 7.4033898305084754e-06, "loss": 1.4484, "step": 19080 }, { "epoch": 0.48, "grad_norm": 12.8125, "learning_rate": 7.396610169491526e-06, "loss": 1.4081, "step": 19090 }, { "epoch": 0.48, "grad_norm": 30.0, "learning_rate": 7.3898305084745766e-06, "loss": 1.3336, "step": 19100 }, { "epoch": 0.48, "grad_norm": 36.5, "learning_rate": 7.383050847457628e-06, "loss": 1.3015, "step": 19110 }, { "epoch": 0.48, "grad_norm": 13.875, "learning_rate": 7.3762711864406785e-06, "loss": 1.4283, "step": 19120 }, { "epoch": 0.48, "grad_norm": 41.0, "learning_rate": 7.36949152542373e-06, "loss": 1.2612, "step": 19130 }, { "epoch": 0.48, "grad_norm": 18.0, "learning_rate": 7.3627118644067805e-06, "loss": 1.1913, "step": 19140 }, { "epoch": 0.48, "grad_norm": 19.375, "learning_rate": 7.355932203389831e-06, "loss": 1.338, "step": 19150 }, { "epoch": 0.48, "grad_norm": 51.5, "learning_rate": 7.3491525423728816e-06, "loss": 1.3061, "step": 19160 }, { "epoch": 0.48, "grad_norm": 26.75, "learning_rate": 7.342372881355932e-06, "loss": 1.322, "step": 19170 }, { "epoch": 0.48, "grad_norm": 30.625, "learning_rate": 7.3355932203389835e-06, "loss": 1.4052, "step": 19180 }, { "epoch": 0.48, "grad_norm": 27.125, "learning_rate": 7.328813559322034e-06, "loss": 1.3729, "step": 19190 }, { "epoch": 0.48, "grad_norm": 36.0, "learning_rate": 7.3220338983050855e-06, "loss": 1.225, "step": 19200 }, { "epoch": 0.48, "grad_norm": 11.375, "learning_rate": 7.315254237288136e-06, "loss": 1.3407, "step": 19210 }, { "epoch": 0.48, "grad_norm": 40.0, "learning_rate": 7.3084745762711874e-06, "loss": 1.3733, "step": 19220 }, { "epoch": 0.48, "grad_norm": 42.5, "learning_rate": 7.301694915254238e-06, "loss": 1.2025, "step": 19230 }, { "epoch": 0.48, "grad_norm": 54.25, "learning_rate": 7.294915254237289e-06, "loss": 1.337, "step": 19240 }, { "epoch": 0.48, "grad_norm": 22.75, "learning_rate": 7.288135593220339e-06, "loss": 1.4894, "step": 19250 }, { "epoch": 0.48, "grad_norm": 14.375, "learning_rate": 7.28135593220339e-06, "loss": 1.4047, "step": 19260 }, { "epoch": 0.48, "grad_norm": 44.5, "learning_rate": 7.274576271186441e-06, "loss": 1.3092, "step": 19270 }, { "epoch": 0.48, "grad_norm": 29.75, "learning_rate": 7.267796610169492e-06, "loss": 1.326, "step": 19280 }, { "epoch": 0.48, "grad_norm": 35.0, "learning_rate": 7.261016949152543e-06, "loss": 1.4019, "step": 19290 }, { "epoch": 0.48, "grad_norm": 22.5, "learning_rate": 7.2542372881355936e-06, "loss": 1.5459, "step": 19300 }, { "epoch": 0.48, "grad_norm": 16.25, "learning_rate": 7.247457627118645e-06, "loss": 1.119, "step": 19310 }, { "epoch": 0.48, "grad_norm": 22.75, "learning_rate": 7.2406779661016955e-06, "loss": 1.2458, "step": 19320 }, { "epoch": 0.48, "grad_norm": 15.0, "learning_rate": 7.233898305084747e-06, "loss": 1.2145, "step": 19330 }, { "epoch": 0.48, "grad_norm": 13.9375, "learning_rate": 7.2271186440677975e-06, "loss": 1.2367, "step": 19340 }, { "epoch": 0.48, "grad_norm": 22.875, "learning_rate": 7.220338983050849e-06, "loss": 1.409, "step": 19350 }, { "epoch": 0.48, "grad_norm": 25.75, "learning_rate": 7.2135593220338986e-06, "loss": 1.4789, "step": 19360 }, { "epoch": 0.48, "grad_norm": 8.5, "learning_rate": 7.206779661016949e-06, "loss": 1.1454, "step": 19370 }, { "epoch": 0.48, "grad_norm": 40.0, "learning_rate": 7.2000000000000005e-06, "loss": 1.4167, "step": 19380 }, { "epoch": 0.48, "grad_norm": 17.875, "learning_rate": 7.193220338983051e-06, "loss": 1.4434, "step": 19390 }, { "epoch": 0.48, "grad_norm": 13.4375, "learning_rate": 7.1864406779661025e-06, "loss": 1.3475, "step": 19400 }, { "epoch": 0.49, "grad_norm": 19.75, "learning_rate": 7.179661016949153e-06, "loss": 1.4455, "step": 19410 }, { "epoch": 0.49, "grad_norm": 18.625, "learning_rate": 7.1728813559322044e-06, "loss": 1.2853, "step": 19420 }, { "epoch": 0.49, "grad_norm": 39.5, "learning_rate": 7.166101694915255e-06, "loss": 1.5275, "step": 19430 }, { "epoch": 0.49, "grad_norm": 26.375, "learning_rate": 7.159322033898306e-06, "loss": 1.4343, "step": 19440 }, { "epoch": 0.49, "grad_norm": 46.5, "learning_rate": 7.152542372881357e-06, "loss": 1.2769, "step": 19450 }, { "epoch": 0.49, "grad_norm": 17.625, "learning_rate": 7.145762711864407e-06, "loss": 1.4347, "step": 19460 }, { "epoch": 0.49, "grad_norm": 22.625, "learning_rate": 7.138983050847458e-06, "loss": 1.3595, "step": 19470 }, { "epoch": 0.49, "grad_norm": 19.125, "learning_rate": 7.132203389830509e-06, "loss": 1.2369, "step": 19480 }, { "epoch": 0.49, "grad_norm": 14.1875, "learning_rate": 7.12542372881356e-06, "loss": 1.2498, "step": 19490 }, { "epoch": 0.49, "grad_norm": 11.5625, "learning_rate": 7.1186440677966106e-06, "loss": 1.2684, "step": 19500 }, { "epoch": 0.49, "grad_norm": 14.625, "learning_rate": 7.111864406779662e-06, "loss": 1.4037, "step": 19510 }, { "epoch": 0.49, "grad_norm": 26.375, "learning_rate": 7.1050847457627125e-06, "loss": 1.3848, "step": 19520 }, { "epoch": 0.49, "grad_norm": 23.875, "learning_rate": 7.098305084745764e-06, "loss": 1.2486, "step": 19530 }, { "epoch": 0.49, "grad_norm": 19.125, "learning_rate": 7.0915254237288145e-06, "loss": 1.3605, "step": 19540 }, { "epoch": 0.49, "grad_norm": 16.0, "learning_rate": 7.084745762711865e-06, "loss": 1.3836, "step": 19550 }, { "epoch": 0.49, "grad_norm": 15.25, "learning_rate": 7.077966101694916e-06, "loss": 1.2622, "step": 19560 }, { "epoch": 0.49, "grad_norm": 10.375, "learning_rate": 7.071186440677966e-06, "loss": 1.2034, "step": 19570 }, { "epoch": 0.49, "grad_norm": 26.625, "learning_rate": 7.0644067796610175e-06, "loss": 1.2822, "step": 19580 }, { "epoch": 0.49, "grad_norm": 20.0, "learning_rate": 7.057627118644068e-06, "loss": 1.2962, "step": 19590 }, { "epoch": 0.49, "grad_norm": 54.25, "learning_rate": 7.0508474576271195e-06, "loss": 1.305, "step": 19600 }, { "epoch": 0.49, "grad_norm": 13.375, "learning_rate": 7.04406779661017e-06, "loss": 1.2614, "step": 19610 }, { "epoch": 0.49, "grad_norm": 16.75, "learning_rate": 7.037288135593221e-06, "loss": 1.375, "step": 19620 }, { "epoch": 0.49, "grad_norm": 33.5, "learning_rate": 7.030508474576272e-06, "loss": 1.2791, "step": 19630 }, { "epoch": 0.49, "grad_norm": 35.5, "learning_rate": 7.0237288135593225e-06, "loss": 1.2757, "step": 19640 }, { "epoch": 0.49, "grad_norm": 14.3125, "learning_rate": 7.016949152542374e-06, "loss": 1.1156, "step": 19650 }, { "epoch": 0.49, "grad_norm": 13.4375, "learning_rate": 7.0101694915254245e-06, "loss": 1.3131, "step": 19660 }, { "epoch": 0.49, "grad_norm": 32.25, "learning_rate": 7.003389830508475e-06, "loss": 1.2609, "step": 19670 }, { "epoch": 0.49, "grad_norm": 18.875, "learning_rate": 6.996610169491526e-06, "loss": 1.2951, "step": 19680 }, { "epoch": 0.49, "grad_norm": 7.625, "learning_rate": 6.989830508474576e-06, "loss": 1.2051, "step": 19690 }, { "epoch": 0.49, "grad_norm": 28.375, "learning_rate": 6.9830508474576275e-06, "loss": 1.4015, "step": 19700 }, { "epoch": 0.49, "grad_norm": 26.75, "learning_rate": 6.976271186440678e-06, "loss": 1.5129, "step": 19710 }, { "epoch": 0.49, "grad_norm": 15.0, "learning_rate": 6.9694915254237295e-06, "loss": 1.3309, "step": 19720 }, { "epoch": 0.49, "grad_norm": 57.5, "learning_rate": 6.96271186440678e-06, "loss": 1.3555, "step": 19730 }, { "epoch": 0.49, "grad_norm": 63.75, "learning_rate": 6.9559322033898315e-06, "loss": 1.3705, "step": 19740 }, { "epoch": 0.49, "grad_norm": 41.5, "learning_rate": 6.949152542372882e-06, "loss": 1.4068, "step": 19750 }, { "epoch": 0.49, "grad_norm": 27.75, "learning_rate": 6.942372881355933e-06, "loss": 1.3491, "step": 19760 }, { "epoch": 0.49, "grad_norm": 17.375, "learning_rate": 6.935593220338983e-06, "loss": 1.5199, "step": 19770 }, { "epoch": 0.49, "grad_norm": 24.75, "learning_rate": 6.928813559322034e-06, "loss": 1.3964, "step": 19780 }, { "epoch": 0.49, "grad_norm": 49.75, "learning_rate": 6.922033898305085e-06, "loss": 1.0419, "step": 19790 }, { "epoch": 0.49, "grad_norm": 28.25, "learning_rate": 6.915254237288136e-06, "loss": 1.4914, "step": 19800 }, { "epoch": 0.5, "grad_norm": 36.75, "learning_rate": 6.908474576271187e-06, "loss": 1.37, "step": 19810 }, { "epoch": 0.5, "grad_norm": 9.9375, "learning_rate": 6.901694915254238e-06, "loss": 1.2354, "step": 19820 }, { "epoch": 0.5, "grad_norm": 36.75, "learning_rate": 6.894915254237289e-06, "loss": 1.1537, "step": 19830 }, { "epoch": 0.5, "grad_norm": 24.125, "learning_rate": 6.8881355932203395e-06, "loss": 1.1772, "step": 19840 }, { "epoch": 0.5, "grad_norm": 13.4375, "learning_rate": 6.881355932203391e-06, "loss": 1.3237, "step": 19850 }, { "epoch": 0.5, "grad_norm": 17.875, "learning_rate": 6.8745762711864415e-06, "loss": 1.1969, "step": 19860 }, { "epoch": 0.5, "grad_norm": 20.125, "learning_rate": 6.867796610169493e-06, "loss": 1.4368, "step": 19870 }, { "epoch": 0.5, "grad_norm": 35.5, "learning_rate": 6.861016949152543e-06, "loss": 1.281, "step": 19880 }, { "epoch": 0.5, "grad_norm": 10.6875, "learning_rate": 6.854237288135593e-06, "loss": 1.3889, "step": 19890 }, { "epoch": 0.5, "grad_norm": 25.5, "learning_rate": 6.8474576271186445e-06, "loss": 1.3965, "step": 19900 }, { "epoch": 0.5, "grad_norm": 41.75, "learning_rate": 6.840677966101695e-06, "loss": 1.2596, "step": 19910 }, { "epoch": 0.5, "grad_norm": 27.25, "learning_rate": 6.8338983050847465e-06, "loss": 1.3484, "step": 19920 }, { "epoch": 0.5, "grad_norm": 22.5, "learning_rate": 6.827118644067797e-06, "loss": 1.4299, "step": 19930 }, { "epoch": 0.5, "grad_norm": 16.375, "learning_rate": 6.8203389830508485e-06, "loss": 1.2868, "step": 19940 }, { "epoch": 0.5, "grad_norm": 18.25, "learning_rate": 6.813559322033899e-06, "loss": 1.2739, "step": 19950 }, { "epoch": 0.5, "grad_norm": 50.0, "learning_rate": 6.80677966101695e-06, "loss": 1.4644, "step": 19960 }, { "epoch": 0.5, "grad_norm": 38.5, "learning_rate": 6.800000000000001e-06, "loss": 1.4472, "step": 19970 }, { "epoch": 0.5, "grad_norm": 40.25, "learning_rate": 6.793220338983051e-06, "loss": 1.4583, "step": 19980 }, { "epoch": 0.5, "grad_norm": 52.0, "learning_rate": 6.786440677966102e-06, "loss": 1.4355, "step": 19990 }, { "epoch": 0.5, "grad_norm": 41.25, "learning_rate": 6.779661016949153e-06, "loss": 1.451, "step": 20000 }, { "epoch": 0.5, "eval_loss": 1.349025845527649, "eval_runtime": 59.2373, "eval_samples_per_second": 16.881, "eval_steps_per_second": 16.881, "step": 20000 }, { "epoch": 0.5, "grad_norm": 15.5, "learning_rate": 6.772881355932204e-06, "loss": 1.4218, "step": 20010 }, { "epoch": 0.5, "grad_norm": 10.75, "learning_rate": 6.766101694915255e-06, "loss": 1.4617, "step": 20020 }, { "epoch": 0.5, "grad_norm": 9.0, "learning_rate": 6.759322033898306e-06, "loss": 1.3029, "step": 20030 }, { "epoch": 0.5, "grad_norm": 12.3125, "learning_rate": 6.7525423728813565e-06, "loss": 1.2876, "step": 20040 }, { "epoch": 0.5, "grad_norm": 15.3125, "learning_rate": 6.745762711864408e-06, "loss": 1.4408, "step": 20050 }, { "epoch": 0.5, "grad_norm": 35.25, "learning_rate": 6.7389830508474585e-06, "loss": 1.2928, "step": 20060 }, { "epoch": 0.5, "grad_norm": 23.25, "learning_rate": 6.73220338983051e-06, "loss": 1.3227, "step": 20070 }, { "epoch": 0.5, "grad_norm": 24.75, "learning_rate": 6.7254237288135604e-06, "loss": 1.3638, "step": 20080 }, { "epoch": 0.5, "grad_norm": 19.125, "learning_rate": 6.71864406779661e-06, "loss": 1.3714, "step": 20090 }, { "epoch": 0.5, "grad_norm": 12.625, "learning_rate": 6.7118644067796615e-06, "loss": 1.4923, "step": 20100 }, { "epoch": 0.5, "grad_norm": 26.25, "learning_rate": 6.705084745762712e-06, "loss": 1.4869, "step": 20110 }, { "epoch": 0.5, "grad_norm": 17.75, "learning_rate": 6.6983050847457635e-06, "loss": 1.3427, "step": 20120 }, { "epoch": 0.5, "grad_norm": 33.5, "learning_rate": 6.691525423728814e-06, "loss": 1.3611, "step": 20130 }, { "epoch": 0.5, "grad_norm": 14.1875, "learning_rate": 6.6847457627118655e-06, "loss": 1.4874, "step": 20140 }, { "epoch": 0.5, "grad_norm": 28.25, "learning_rate": 6.677966101694916e-06, "loss": 1.368, "step": 20150 }, { "epoch": 0.5, "grad_norm": 51.5, "learning_rate": 6.6711864406779666e-06, "loss": 1.3734, "step": 20160 }, { "epoch": 0.5, "grad_norm": 44.5, "learning_rate": 6.664406779661018e-06, "loss": 1.3609, "step": 20170 }, { "epoch": 0.5, "grad_norm": 41.0, "learning_rate": 6.6576271186440685e-06, "loss": 1.4929, "step": 20180 }, { "epoch": 0.5, "grad_norm": 49.5, "learning_rate": 6.650847457627119e-06, "loss": 1.3619, "step": 20190 }, { "epoch": 0.51, "grad_norm": 24.125, "learning_rate": 6.64406779661017e-06, "loss": 1.3749, "step": 20200 }, { "epoch": 0.51, "grad_norm": 31.75, "learning_rate": 6.637288135593221e-06, "loss": 1.4952, "step": 20210 }, { "epoch": 0.51, "grad_norm": 21.0, "learning_rate": 6.6305084745762716e-06, "loss": 1.3311, "step": 20220 }, { "epoch": 0.51, "grad_norm": 28.5, "learning_rate": 6.623728813559322e-06, "loss": 1.4247, "step": 20230 }, { "epoch": 0.51, "grad_norm": 20.375, "learning_rate": 6.6169491525423735e-06, "loss": 1.4556, "step": 20240 }, { "epoch": 0.51, "grad_norm": 28.25, "learning_rate": 6.610169491525424e-06, "loss": 1.4477, "step": 20250 }, { "epoch": 0.51, "grad_norm": 21.0, "learning_rate": 6.6033898305084755e-06, "loss": 1.3582, "step": 20260 }, { "epoch": 0.51, "grad_norm": 55.75, "learning_rate": 6.596610169491526e-06, "loss": 1.3473, "step": 20270 }, { "epoch": 0.51, "grad_norm": 9.1875, "learning_rate": 6.5898305084745774e-06, "loss": 1.4635, "step": 20280 }, { "epoch": 0.51, "grad_norm": 8.3125, "learning_rate": 6.583050847457627e-06, "loss": 1.249, "step": 20290 }, { "epoch": 0.51, "grad_norm": 65.0, "learning_rate": 6.576271186440678e-06, "loss": 1.3769, "step": 20300 }, { "epoch": 0.51, "grad_norm": 14.125, "learning_rate": 6.569491525423729e-06, "loss": 1.4276, "step": 20310 }, { "epoch": 0.51, "grad_norm": 44.0, "learning_rate": 6.56271186440678e-06, "loss": 1.4521, "step": 20320 }, { "epoch": 0.51, "grad_norm": 43.5, "learning_rate": 6.555932203389831e-06, "loss": 1.3885, "step": 20330 }, { "epoch": 0.51, "grad_norm": 52.25, "learning_rate": 6.549152542372882e-06, "loss": 1.345, "step": 20340 }, { "epoch": 0.51, "grad_norm": 27.625, "learning_rate": 6.542372881355933e-06, "loss": 1.3546, "step": 20350 }, { "epoch": 0.51, "grad_norm": 42.0, "learning_rate": 6.5355932203389836e-06, "loss": 1.3781, "step": 20360 }, { "epoch": 0.51, "grad_norm": 10.625, "learning_rate": 6.528813559322035e-06, "loss": 1.3201, "step": 20370 }, { "epoch": 0.51, "grad_norm": 34.25, "learning_rate": 6.5220338983050855e-06, "loss": 1.3468, "step": 20380 }, { "epoch": 0.51, "grad_norm": 27.125, "learning_rate": 6.515254237288137e-06, "loss": 1.3345, "step": 20390 }, { "epoch": 0.51, "grad_norm": 13.625, "learning_rate": 6.508474576271187e-06, "loss": 1.2565, "step": 20400 }, { "epoch": 0.51, "grad_norm": 17.625, "learning_rate": 6.501694915254237e-06, "loss": 1.3067, "step": 20410 }, { "epoch": 0.51, "grad_norm": 16.0, "learning_rate": 6.4949152542372886e-06, "loss": 1.4644, "step": 20420 }, { "epoch": 0.51, "grad_norm": 29.25, "learning_rate": 6.488135593220339e-06, "loss": 1.3739, "step": 20430 }, { "epoch": 0.51, "grad_norm": 22.875, "learning_rate": 6.4813559322033905e-06, "loss": 1.4237, "step": 20440 }, { "epoch": 0.51, "grad_norm": 14.0, "learning_rate": 6.474576271186441e-06, "loss": 1.4379, "step": 20450 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 6.4677966101694925e-06, "loss": 1.2757, "step": 20460 }, { "epoch": 0.51, "grad_norm": 15.5, "learning_rate": 6.461016949152543e-06, "loss": 1.4572, "step": 20470 }, { "epoch": 0.51, "grad_norm": 15.5625, "learning_rate": 6.4542372881355944e-06, "loss": 1.3073, "step": 20480 }, { "epoch": 0.51, "grad_norm": 19.5, "learning_rate": 6.447457627118645e-06, "loss": 1.4111, "step": 20490 }, { "epoch": 0.51, "grad_norm": 31.5, "learning_rate": 6.440677966101695e-06, "loss": 1.4395, "step": 20500 }, { "epoch": 0.51, "grad_norm": 99.5, "learning_rate": 6.433898305084746e-06, "loss": 1.2802, "step": 20510 }, { "epoch": 0.51, "grad_norm": 39.25, "learning_rate": 6.427118644067797e-06, "loss": 1.3699, "step": 20520 }, { "epoch": 0.51, "grad_norm": 8.0, "learning_rate": 6.420338983050848e-06, "loss": 1.311, "step": 20530 }, { "epoch": 0.51, "grad_norm": 32.25, "learning_rate": 6.413559322033899e-06, "loss": 1.2298, "step": 20540 }, { "epoch": 0.51, "grad_norm": 65.5, "learning_rate": 6.40677966101695e-06, "loss": 1.1912, "step": 20550 }, { "epoch": 0.51, "grad_norm": 22.375, "learning_rate": 6.4000000000000006e-06, "loss": 1.2825, "step": 20560 }, { "epoch": 0.51, "grad_norm": 6.25, "learning_rate": 6.393220338983052e-06, "loss": 1.3518, "step": 20570 }, { "epoch": 0.51, "grad_norm": 38.0, "learning_rate": 6.3864406779661025e-06, "loss": 1.3, "step": 20580 }, { "epoch": 0.51, "grad_norm": 14.125, "learning_rate": 6.379661016949154e-06, "loss": 1.2622, "step": 20590 }, { "epoch": 0.52, "grad_norm": 52.75, "learning_rate": 6.372881355932204e-06, "loss": 1.1673, "step": 20600 }, { "epoch": 0.52, "grad_norm": 38.25, "learning_rate": 6.366101694915254e-06, "loss": 1.3722, "step": 20610 }, { "epoch": 0.52, "grad_norm": 13.8125, "learning_rate": 6.3593220338983056e-06, "loss": 1.2911, "step": 20620 }, { "epoch": 0.52, "grad_norm": 19.875, "learning_rate": 6.352542372881356e-06, "loss": 1.4929, "step": 20630 }, { "epoch": 0.52, "grad_norm": 15.625, "learning_rate": 6.3457627118644075e-06, "loss": 1.2885, "step": 20640 }, { "epoch": 0.52, "grad_norm": 9.375, "learning_rate": 6.338983050847458e-06, "loss": 1.4606, "step": 20650 }, { "epoch": 0.52, "grad_norm": 13.8125, "learning_rate": 6.3322033898305095e-06, "loss": 1.398, "step": 20660 }, { "epoch": 0.52, "grad_norm": 63.5, "learning_rate": 6.32542372881356e-06, "loss": 1.309, "step": 20670 }, { "epoch": 0.52, "grad_norm": 34.75, "learning_rate": 6.318644067796611e-06, "loss": 1.4643, "step": 20680 }, { "epoch": 0.52, "grad_norm": 17.875, "learning_rate": 6.311864406779662e-06, "loss": 1.4006, "step": 20690 }, { "epoch": 0.52, "grad_norm": 36.5, "learning_rate": 6.3050847457627125e-06, "loss": 1.4196, "step": 20700 }, { "epoch": 0.52, "grad_norm": 26.75, "learning_rate": 6.298305084745763e-06, "loss": 1.5026, "step": 20710 }, { "epoch": 0.52, "grad_norm": 52.75, "learning_rate": 6.291525423728814e-06, "loss": 1.3181, "step": 20720 }, { "epoch": 0.52, "grad_norm": 18.0, "learning_rate": 6.284745762711865e-06, "loss": 1.2799, "step": 20730 }, { "epoch": 0.52, "grad_norm": 13.0625, "learning_rate": 6.277966101694916e-06, "loss": 1.5509, "step": 20740 }, { "epoch": 0.52, "grad_norm": 14.625, "learning_rate": 6.271186440677966e-06, "loss": 1.1255, "step": 20750 }, { "epoch": 0.52, "grad_norm": 11.9375, "learning_rate": 6.2644067796610176e-06, "loss": 1.2412, "step": 20760 }, { "epoch": 0.52, "grad_norm": 40.25, "learning_rate": 6.257627118644068e-06, "loss": 1.3701, "step": 20770 }, { "epoch": 0.52, "grad_norm": 25.875, "learning_rate": 6.2508474576271195e-06, "loss": 1.3874, "step": 20780 }, { "epoch": 0.52, "grad_norm": 11.8125, "learning_rate": 6.24406779661017e-06, "loss": 1.3743, "step": 20790 }, { "epoch": 0.52, "grad_norm": 9.25, "learning_rate": 6.2372881355932215e-06, "loss": 1.4481, "step": 20800 }, { "epoch": 0.52, "grad_norm": 26.875, "learning_rate": 6.230508474576271e-06, "loss": 1.381, "step": 20810 }, { "epoch": 0.52, "grad_norm": 14.1875, "learning_rate": 6.223728813559322e-06, "loss": 1.0738, "step": 20820 }, { "epoch": 0.52, "grad_norm": 24.75, "learning_rate": 6.216949152542373e-06, "loss": 1.2937, "step": 20830 }, { "epoch": 0.52, "grad_norm": 28.625, "learning_rate": 6.210169491525424e-06, "loss": 1.3835, "step": 20840 }, { "epoch": 0.52, "grad_norm": 47.0, "learning_rate": 6.203389830508475e-06, "loss": 1.565, "step": 20850 }, { "epoch": 0.52, "grad_norm": 7.6875, "learning_rate": 6.196610169491526e-06, "loss": 1.3831, "step": 20860 }, { "epoch": 0.52, "grad_norm": 7.53125, "learning_rate": 6.189830508474577e-06, "loss": 1.2528, "step": 20870 }, { "epoch": 0.52, "grad_norm": 20.75, "learning_rate": 6.183050847457628e-06, "loss": 1.4723, "step": 20880 }, { "epoch": 0.52, "grad_norm": 14.3125, "learning_rate": 6.176271186440679e-06, "loss": 1.4442, "step": 20890 }, { "epoch": 0.52, "grad_norm": 12.9375, "learning_rate": 6.1694915254237295e-06, "loss": 1.3466, "step": 20900 }, { "epoch": 0.52, "grad_norm": 38.0, "learning_rate": 6.162711864406781e-06, "loss": 1.3282, "step": 20910 }, { "epoch": 0.52, "grad_norm": 24.625, "learning_rate": 6.155932203389831e-06, "loss": 1.3331, "step": 20920 }, { "epoch": 0.52, "grad_norm": 30.25, "learning_rate": 6.149152542372881e-06, "loss": 1.4733, "step": 20930 }, { "epoch": 0.52, "grad_norm": 25.625, "learning_rate": 6.142372881355933e-06, "loss": 1.558, "step": 20940 }, { "epoch": 0.52, "grad_norm": 16.75, "learning_rate": 6.135593220338983e-06, "loss": 1.225, "step": 20950 }, { "epoch": 0.52, "grad_norm": 18.5, "learning_rate": 6.1288135593220346e-06, "loss": 1.3408, "step": 20960 }, { "epoch": 0.52, "grad_norm": 25.125, "learning_rate": 6.122033898305085e-06, "loss": 1.424, "step": 20970 }, { "epoch": 0.52, "grad_norm": 67.0, "learning_rate": 6.1152542372881365e-06, "loss": 1.2106, "step": 20980 }, { "epoch": 0.52, "grad_norm": 17.25, "learning_rate": 6.108474576271187e-06, "loss": 1.3747, "step": 20990 }, { "epoch": 0.53, "grad_norm": 20.625, "learning_rate": 6.1016949152542385e-06, "loss": 1.3211, "step": 21000 }, { "epoch": 0.53, "grad_norm": 40.0, "learning_rate": 6.094915254237289e-06, "loss": 1.3193, "step": 21010 }, { "epoch": 0.53, "grad_norm": 52.0, "learning_rate": 6.088135593220339e-06, "loss": 1.4027, "step": 21020 }, { "epoch": 0.53, "grad_norm": 12.375, "learning_rate": 6.08135593220339e-06, "loss": 1.3687, "step": 21030 }, { "epoch": 0.53, "grad_norm": 23.0, "learning_rate": 6.074576271186441e-06, "loss": 1.4436, "step": 21040 }, { "epoch": 0.53, "grad_norm": 46.75, "learning_rate": 6.067796610169492e-06, "loss": 1.3343, "step": 21050 }, { "epoch": 0.53, "grad_norm": 21.0, "learning_rate": 6.061016949152543e-06, "loss": 1.3544, "step": 21060 }, { "epoch": 0.53, "grad_norm": 16.875, "learning_rate": 6.054237288135594e-06, "loss": 1.3887, "step": 21070 }, { "epoch": 0.53, "grad_norm": 14.9375, "learning_rate": 6.047457627118645e-06, "loss": 1.385, "step": 21080 }, { "epoch": 0.53, "grad_norm": 78.0, "learning_rate": 6.040677966101696e-06, "loss": 1.4487, "step": 21090 }, { "epoch": 0.53, "grad_norm": 16.625, "learning_rate": 6.0338983050847465e-06, "loss": 1.4415, "step": 21100 }, { "epoch": 0.53, "grad_norm": 33.0, "learning_rate": 6.027118644067798e-06, "loss": 1.3929, "step": 21110 }, { "epoch": 0.53, "grad_norm": 17.375, "learning_rate": 6.020338983050848e-06, "loss": 1.2834, "step": 21120 }, { "epoch": 0.53, "grad_norm": 14.8125, "learning_rate": 6.013559322033898e-06, "loss": 1.4886, "step": 21130 }, { "epoch": 0.53, "grad_norm": 22.0, "learning_rate": 6.00677966101695e-06, "loss": 1.401, "step": 21140 }, { "epoch": 0.53, "grad_norm": 26.875, "learning_rate": 6e-06, "loss": 1.2837, "step": 21150 }, { "epoch": 0.53, "grad_norm": 27.75, "learning_rate": 5.9932203389830516e-06, "loss": 1.1739, "step": 21160 }, { "epoch": 0.53, "grad_norm": 27.0, "learning_rate": 5.986440677966102e-06, "loss": 1.4129, "step": 21170 }, { "epoch": 0.53, "grad_norm": 37.0, "learning_rate": 5.9796610169491535e-06, "loss": 1.3065, "step": 21180 }, { "epoch": 0.53, "grad_norm": 15.625, "learning_rate": 5.972881355932204e-06, "loss": 1.4291, "step": 21190 }, { "epoch": 0.53, "grad_norm": 17.25, "learning_rate": 5.9661016949152555e-06, "loss": 1.3072, "step": 21200 }, { "epoch": 0.53, "grad_norm": 11.375, "learning_rate": 5.959322033898306e-06, "loss": 1.2729, "step": 21210 }, { "epoch": 0.53, "grad_norm": 23.75, "learning_rate": 5.9525423728813566e-06, "loss": 1.3186, "step": 21220 }, { "epoch": 0.53, "grad_norm": 14.125, "learning_rate": 5.945762711864407e-06, "loss": 1.3264, "step": 21230 }, { "epoch": 0.53, "grad_norm": 36.5, "learning_rate": 5.938983050847458e-06, "loss": 1.4535, "step": 21240 }, { "epoch": 0.53, "grad_norm": 12.0625, "learning_rate": 5.932203389830509e-06, "loss": 1.4143, "step": 21250 }, { "epoch": 0.53, "grad_norm": 30.75, "learning_rate": 5.92542372881356e-06, "loss": 1.2338, "step": 21260 }, { "epoch": 0.53, "grad_norm": 31.5, "learning_rate": 5.91864406779661e-06, "loss": 1.3512, "step": 21270 }, { "epoch": 0.53, "grad_norm": 16.25, "learning_rate": 5.911864406779662e-06, "loss": 1.4743, "step": 21280 }, { "epoch": 0.53, "grad_norm": 30.625, "learning_rate": 5.905084745762712e-06, "loss": 1.5777, "step": 21290 }, { "epoch": 0.53, "grad_norm": 15.875, "learning_rate": 5.8983050847457635e-06, "loss": 1.4545, "step": 21300 }, { "epoch": 0.53, "grad_norm": 48.25, "learning_rate": 5.891525423728814e-06, "loss": 1.3676, "step": 21310 }, { "epoch": 0.53, "grad_norm": 70.5, "learning_rate": 5.8847457627118655e-06, "loss": 1.2971, "step": 21320 }, { "epoch": 0.53, "grad_norm": 18.375, "learning_rate": 5.877966101694915e-06, "loss": 1.3533, "step": 21330 }, { "epoch": 0.53, "grad_norm": 17.5, "learning_rate": 5.871186440677966e-06, "loss": 1.4603, "step": 21340 }, { "epoch": 0.53, "grad_norm": 14.3125, "learning_rate": 5.864406779661017e-06, "loss": 1.2108, "step": 21350 }, { "epoch": 0.53, "grad_norm": 22.875, "learning_rate": 5.857627118644068e-06, "loss": 1.4043, "step": 21360 }, { "epoch": 0.53, "grad_norm": 29.5, "learning_rate": 5.850847457627119e-06, "loss": 1.4477, "step": 21370 }, { "epoch": 0.53, "grad_norm": 51.25, "learning_rate": 5.84406779661017e-06, "loss": 1.1954, "step": 21380 }, { "epoch": 0.53, "grad_norm": 17.375, "learning_rate": 5.837288135593221e-06, "loss": 1.4261, "step": 21390 }, { "epoch": 0.54, "grad_norm": 16.75, "learning_rate": 5.830508474576272e-06, "loss": 1.2177, "step": 21400 }, { "epoch": 0.54, "grad_norm": 31.375, "learning_rate": 5.823728813559323e-06, "loss": 1.4369, "step": 21410 }, { "epoch": 0.54, "grad_norm": 38.5, "learning_rate": 5.8169491525423736e-06, "loss": 1.524, "step": 21420 }, { "epoch": 0.54, "grad_norm": 24.125, "learning_rate": 5.810169491525425e-06, "loss": 1.287, "step": 21430 }, { "epoch": 0.54, "grad_norm": 27.25, "learning_rate": 5.803389830508475e-06, "loss": 1.331, "step": 21440 }, { "epoch": 0.54, "grad_norm": 26.25, "learning_rate": 5.796610169491525e-06, "loss": 1.2623, "step": 21450 }, { "epoch": 0.54, "grad_norm": 27.0, "learning_rate": 5.789830508474577e-06, "loss": 1.4062, "step": 21460 }, { "epoch": 0.54, "grad_norm": 33.0, "learning_rate": 5.783050847457627e-06, "loss": 1.282, "step": 21470 }, { "epoch": 0.54, "grad_norm": 10.0, "learning_rate": 5.776271186440679e-06, "loss": 1.1568, "step": 21480 }, { "epoch": 0.54, "grad_norm": 32.5, "learning_rate": 5.769491525423729e-06, "loss": 1.3498, "step": 21490 }, { "epoch": 0.54, "grad_norm": 108.5, "learning_rate": 5.7627118644067805e-06, "loss": 1.527, "step": 21500 }, { "epoch": 0.54, "grad_norm": 32.75, "learning_rate": 5.755932203389831e-06, "loss": 1.2589, "step": 21510 }, { "epoch": 0.54, "grad_norm": 17.5, "learning_rate": 5.7491525423728825e-06, "loss": 1.2433, "step": 21520 }, { "epoch": 0.54, "grad_norm": 18.25, "learning_rate": 5.742372881355933e-06, "loss": 1.2189, "step": 21530 }, { "epoch": 0.54, "grad_norm": 23.125, "learning_rate": 5.735593220338983e-06, "loss": 1.1866, "step": 21540 }, { "epoch": 0.54, "grad_norm": 22.625, "learning_rate": 5.728813559322034e-06, "loss": 1.2153, "step": 21550 }, { "epoch": 0.54, "grad_norm": 87.5, "learning_rate": 5.722033898305085e-06, "loss": 1.4075, "step": 21560 }, { "epoch": 0.54, "grad_norm": 21.5, "learning_rate": 5.715254237288136e-06, "loss": 1.3385, "step": 21570 }, { "epoch": 0.54, "grad_norm": 18.125, "learning_rate": 5.708474576271187e-06, "loss": 1.2616, "step": 21580 }, { "epoch": 0.54, "grad_norm": 17.625, "learning_rate": 5.701694915254238e-06, "loss": 1.2999, "step": 21590 }, { "epoch": 0.54, "grad_norm": 27.375, "learning_rate": 5.694915254237289e-06, "loss": 1.3824, "step": 21600 }, { "epoch": 0.54, "grad_norm": 16.875, "learning_rate": 5.68813559322034e-06, "loss": 1.2897, "step": 21610 }, { "epoch": 0.54, "grad_norm": 37.75, "learning_rate": 5.6813559322033906e-06, "loss": 1.2475, "step": 21620 }, { "epoch": 0.54, "grad_norm": 32.0, "learning_rate": 5.674576271186442e-06, "loss": 1.4961, "step": 21630 }, { "epoch": 0.54, "grad_norm": 28.625, "learning_rate": 5.667796610169492e-06, "loss": 1.5183, "step": 21640 }, { "epoch": 0.54, "grad_norm": 34.5, "learning_rate": 5.661016949152542e-06, "loss": 1.5494, "step": 21650 }, { "epoch": 0.54, "grad_norm": 20.25, "learning_rate": 5.654237288135594e-06, "loss": 1.424, "step": 21660 }, { "epoch": 0.54, "grad_norm": 27.75, "learning_rate": 5.647457627118644e-06, "loss": 1.3111, "step": 21670 }, { "epoch": 0.54, "grad_norm": 17.125, "learning_rate": 5.640677966101696e-06, "loss": 1.3682, "step": 21680 }, { "epoch": 0.54, "grad_norm": 24.625, "learning_rate": 5.633898305084746e-06, "loss": 1.2766, "step": 21690 }, { "epoch": 0.54, "grad_norm": 24.875, "learning_rate": 5.6271186440677975e-06, "loss": 1.3045, "step": 21700 }, { "epoch": 0.54, "grad_norm": 12.3125, "learning_rate": 5.620338983050848e-06, "loss": 1.4166, "step": 21710 }, { "epoch": 0.54, "grad_norm": 13.0, "learning_rate": 5.6135593220338995e-06, "loss": 1.4751, "step": 21720 }, { "epoch": 0.54, "grad_norm": 21.125, "learning_rate": 5.60677966101695e-06, "loss": 1.2947, "step": 21730 }, { "epoch": 0.54, "grad_norm": 19.0, "learning_rate": 5.600000000000001e-06, "loss": 1.3368, "step": 21740 }, { "epoch": 0.54, "grad_norm": 20.625, "learning_rate": 5.593220338983051e-06, "loss": 1.3925, "step": 21750 }, { "epoch": 0.54, "grad_norm": 19.25, "learning_rate": 5.586440677966102e-06, "loss": 1.2478, "step": 21760 }, { "epoch": 0.54, "grad_norm": 50.5, "learning_rate": 5.579661016949153e-06, "loss": 1.3246, "step": 21770 }, { "epoch": 0.54, "grad_norm": 66.0, "learning_rate": 5.572881355932204e-06, "loss": 1.2474, "step": 21780 }, { "epoch": 0.54, "grad_norm": 22.75, "learning_rate": 5.566101694915255e-06, "loss": 1.6124, "step": 21790 }, { "epoch": 0.55, "grad_norm": 24.125, "learning_rate": 5.559322033898306e-06, "loss": 1.4545, "step": 21800 }, { "epoch": 0.55, "grad_norm": 15.5, "learning_rate": 5.552542372881356e-06, "loss": 1.1902, "step": 21810 }, { "epoch": 0.55, "grad_norm": 12.625, "learning_rate": 5.5457627118644076e-06, "loss": 1.2157, "step": 21820 }, { "epoch": 0.55, "grad_norm": 12.5, "learning_rate": 5.538983050847458e-06, "loss": 1.3625, "step": 21830 }, { "epoch": 0.55, "grad_norm": 19.0, "learning_rate": 5.5322033898305095e-06, "loss": 1.3041, "step": 21840 }, { "epoch": 0.55, "grad_norm": 47.0, "learning_rate": 5.525423728813559e-06, "loss": 1.1826, "step": 21850 }, { "epoch": 0.55, "grad_norm": 24.875, "learning_rate": 5.518644067796611e-06, "loss": 1.412, "step": 21860 }, { "epoch": 0.55, "grad_norm": 10.0625, "learning_rate": 5.511864406779661e-06, "loss": 1.3009, "step": 21870 }, { "epoch": 0.55, "grad_norm": 17.5, "learning_rate": 5.505084745762712e-06, "loss": 1.3198, "step": 21880 }, { "epoch": 0.55, "grad_norm": 31.5, "learning_rate": 5.498305084745763e-06, "loss": 1.3365, "step": 21890 }, { "epoch": 0.55, "grad_norm": 36.0, "learning_rate": 5.491525423728814e-06, "loss": 1.4381, "step": 21900 }, { "epoch": 0.55, "grad_norm": 21.625, "learning_rate": 5.484745762711865e-06, "loss": 1.3239, "step": 21910 }, { "epoch": 0.55, "grad_norm": 45.0, "learning_rate": 5.477966101694916e-06, "loss": 1.5603, "step": 21920 }, { "epoch": 0.55, "grad_norm": 32.5, "learning_rate": 5.471186440677967e-06, "loss": 1.3348, "step": 21930 }, { "epoch": 0.55, "grad_norm": 20.375, "learning_rate": 5.464406779661018e-06, "loss": 1.3331, "step": 21940 }, { "epoch": 0.55, "grad_norm": 25.0, "learning_rate": 5.457627118644067e-06, "loss": 1.4195, "step": 21950 }, { "epoch": 0.55, "grad_norm": 39.25, "learning_rate": 5.450847457627119e-06, "loss": 1.5658, "step": 21960 }, { "epoch": 0.55, "grad_norm": 46.5, "learning_rate": 5.444067796610169e-06, "loss": 1.3294, "step": 21970 }, { "epoch": 0.55, "grad_norm": 37.75, "learning_rate": 5.437288135593221e-06, "loss": 1.5364, "step": 21980 }, { "epoch": 0.55, "grad_norm": 16.25, "learning_rate": 5.430508474576271e-06, "loss": 1.4479, "step": 21990 }, { "epoch": 0.55, "grad_norm": 45.25, "learning_rate": 5.423728813559323e-06, "loss": 1.2464, "step": 22000 }, { "epoch": 0.55, "grad_norm": 21.375, "learning_rate": 5.416949152542373e-06, "loss": 1.1403, "step": 22010 }, { "epoch": 0.55, "grad_norm": 23.5, "learning_rate": 5.4101694915254246e-06, "loss": 1.3901, "step": 22020 }, { "epoch": 0.55, "grad_norm": 9.1875, "learning_rate": 5.403389830508475e-06, "loss": 1.394, "step": 22030 }, { "epoch": 0.55, "grad_norm": 10.0625, "learning_rate": 5.3966101694915265e-06, "loss": 1.0434, "step": 22040 }, { "epoch": 0.55, "grad_norm": 18.375, "learning_rate": 5.389830508474577e-06, "loss": 1.3842, "step": 22050 }, { "epoch": 0.55, "grad_norm": 1.734375, "learning_rate": 5.383050847457627e-06, "loss": 1.2067, "step": 22060 }, { "epoch": 0.55, "grad_norm": 13.5, "learning_rate": 5.376271186440678e-06, "loss": 1.3957, "step": 22070 }, { "epoch": 0.55, "grad_norm": 63.5, "learning_rate": 5.369491525423729e-06, "loss": 1.2194, "step": 22080 }, { "epoch": 0.55, "grad_norm": 32.0, "learning_rate": 5.36271186440678e-06, "loss": 1.0633, "step": 22090 }, { "epoch": 0.55, "grad_norm": 20.625, "learning_rate": 5.355932203389831e-06, "loss": 1.2579, "step": 22100 }, { "epoch": 0.55, "grad_norm": 8.3125, "learning_rate": 5.349152542372882e-06, "loss": 1.1221, "step": 22110 }, { "epoch": 0.55, "grad_norm": 25.375, "learning_rate": 5.342372881355933e-06, "loss": 1.2434, "step": 22120 }, { "epoch": 0.55, "grad_norm": 36.0, "learning_rate": 5.335593220338984e-06, "loss": 1.5282, "step": 22130 }, { "epoch": 0.55, "grad_norm": 13.4375, "learning_rate": 5.328813559322035e-06, "loss": 1.3162, "step": 22140 }, { "epoch": 0.55, "grad_norm": 33.25, "learning_rate": 5.322033898305086e-06, "loss": 1.2288, "step": 22150 }, { "epoch": 0.55, "grad_norm": 50.0, "learning_rate": 5.315254237288136e-06, "loss": 1.5033, "step": 22160 }, { "epoch": 0.55, "grad_norm": 19.75, "learning_rate": 5.308474576271186e-06, "loss": 1.3759, "step": 22170 }, { "epoch": 0.55, "grad_norm": 56.0, "learning_rate": 5.301694915254238e-06, "loss": 1.342, "step": 22180 }, { "epoch": 0.55, "grad_norm": 23.375, "learning_rate": 5.294915254237288e-06, "loss": 1.3466, "step": 22190 }, { "epoch": 0.56, "grad_norm": 60.0, "learning_rate": 5.28813559322034e-06, "loss": 1.3293, "step": 22200 }, { "epoch": 0.56, "grad_norm": 30.625, "learning_rate": 5.28135593220339e-06, "loss": 1.4459, "step": 22210 }, { "epoch": 0.56, "grad_norm": 53.5, "learning_rate": 5.2745762711864416e-06, "loss": 1.3259, "step": 22220 }, { "epoch": 0.56, "grad_norm": 45.25, "learning_rate": 5.267796610169492e-06, "loss": 1.273, "step": 22230 }, { "epoch": 0.56, "grad_norm": 35.25, "learning_rate": 5.2610169491525435e-06, "loss": 1.2938, "step": 22240 }, { "epoch": 0.56, "grad_norm": 5.125, "learning_rate": 5.254237288135594e-06, "loss": 1.3133, "step": 22250 }, { "epoch": 0.56, "grad_norm": 47.25, "learning_rate": 5.247457627118645e-06, "loss": 1.4179, "step": 22260 }, { "epoch": 0.56, "grad_norm": 14.4375, "learning_rate": 5.240677966101695e-06, "loss": 1.2183, "step": 22270 }, { "epoch": 0.56, "grad_norm": 12.125, "learning_rate": 5.233898305084746e-06, "loss": 1.3146, "step": 22280 }, { "epoch": 0.56, "grad_norm": 16.5, "learning_rate": 5.227118644067797e-06, "loss": 1.3791, "step": 22290 }, { "epoch": 0.56, "grad_norm": 16.75, "learning_rate": 5.220338983050848e-06, "loss": 1.4089, "step": 22300 }, { "epoch": 0.56, "grad_norm": 15.25, "learning_rate": 5.213559322033899e-06, "loss": 1.2884, "step": 22310 }, { "epoch": 0.56, "grad_norm": 15.5625, "learning_rate": 5.20677966101695e-06, "loss": 1.358, "step": 22320 }, { "epoch": 0.56, "grad_norm": 50.5, "learning_rate": 5.2e-06, "loss": 1.5178, "step": 22330 }, { "epoch": 0.56, "grad_norm": 16.875, "learning_rate": 5.193220338983052e-06, "loss": 1.4127, "step": 22340 }, { "epoch": 0.56, "grad_norm": 14.3125, "learning_rate": 5.186440677966102e-06, "loss": 1.2716, "step": 22350 }, { "epoch": 0.56, "grad_norm": 34.25, "learning_rate": 5.1796610169491535e-06, "loss": 1.2392, "step": 22360 }, { "epoch": 0.56, "grad_norm": 68.0, "learning_rate": 5.172881355932203e-06, "loss": 1.2263, "step": 22370 }, { "epoch": 0.56, "grad_norm": 29.875, "learning_rate": 5.166101694915255e-06, "loss": 1.5481, "step": 22380 }, { "epoch": 0.56, "grad_norm": 27.5, "learning_rate": 5.159322033898305e-06, "loss": 1.4132, "step": 22390 }, { "epoch": 0.56, "grad_norm": 9.6875, "learning_rate": 5.152542372881356e-06, "loss": 1.2786, "step": 22400 }, { "epoch": 0.56, "grad_norm": 15.1875, "learning_rate": 5.145762711864407e-06, "loss": 1.3105, "step": 22410 }, { "epoch": 0.56, "grad_norm": 15.3125, "learning_rate": 5.138983050847458e-06, "loss": 1.2716, "step": 22420 }, { "epoch": 0.56, "grad_norm": 11.5, "learning_rate": 5.132203389830509e-06, "loss": 1.381, "step": 22430 }, { "epoch": 0.56, "grad_norm": 28.75, "learning_rate": 5.12542372881356e-06, "loss": 1.2154, "step": 22440 }, { "epoch": 0.56, "grad_norm": 34.75, "learning_rate": 5.118644067796611e-06, "loss": 1.2588, "step": 22450 }, { "epoch": 0.56, "grad_norm": 27.625, "learning_rate": 5.111864406779662e-06, "loss": 1.2042, "step": 22460 }, { "epoch": 0.56, "grad_norm": 21.75, "learning_rate": 5.105084745762711e-06, "loss": 1.3014, "step": 22470 }, { "epoch": 0.56, "grad_norm": 16.25, "learning_rate": 5.098305084745763e-06, "loss": 1.2872, "step": 22480 }, { "epoch": 0.56, "grad_norm": 6.34375, "learning_rate": 5.091525423728813e-06, "loss": 1.3333, "step": 22490 }, { "epoch": 0.56, "grad_norm": 24.375, "learning_rate": 5.084745762711865e-06, "loss": 1.2379, "step": 22500 }, { "epoch": 0.56, "grad_norm": 15.9375, "learning_rate": 5.077966101694915e-06, "loss": 1.3752, "step": 22510 }, { "epoch": 0.56, "grad_norm": 36.0, "learning_rate": 5.071186440677967e-06, "loss": 1.3533, "step": 22520 }, { "epoch": 0.56, "grad_norm": 15.6875, "learning_rate": 5.064406779661017e-06, "loss": 1.2247, "step": 22530 }, { "epoch": 0.56, "grad_norm": 30.125, "learning_rate": 5.057627118644069e-06, "loss": 1.3658, "step": 22540 }, { "epoch": 0.56, "grad_norm": 16.0, "learning_rate": 5.050847457627119e-06, "loss": 1.4848, "step": 22550 }, { "epoch": 0.56, "grad_norm": 17.625, "learning_rate": 5.0440677966101705e-06, "loss": 1.3547, "step": 22560 }, { "epoch": 0.56, "grad_norm": 30.5, "learning_rate": 5.037288135593221e-06, "loss": 1.3617, "step": 22570 }, { "epoch": 0.56, "grad_norm": 11.8125, "learning_rate": 5.030508474576271e-06, "loss": 1.3552, "step": 22580 }, { "epoch": 0.56, "grad_norm": 29.375, "learning_rate": 5.023728813559322e-06, "loss": 1.4876, "step": 22590 }, { "epoch": 0.56, "grad_norm": 32.75, "learning_rate": 5.016949152542373e-06, "loss": 1.2906, "step": 22600 }, { "epoch": 0.57, "grad_norm": 28.625, "learning_rate": 5.010169491525424e-06, "loss": 1.3796, "step": 22610 }, { "epoch": 0.57, "grad_norm": 47.75, "learning_rate": 5.003389830508475e-06, "loss": 1.4018, "step": 22620 }, { "epoch": 0.57, "grad_norm": 29.625, "learning_rate": 4.996610169491526e-06, "loss": 1.1936, "step": 22630 }, { "epoch": 0.57, "grad_norm": 62.0, "learning_rate": 4.989830508474577e-06, "loss": 1.4104, "step": 22640 }, { "epoch": 0.57, "grad_norm": 16.0, "learning_rate": 4.983050847457628e-06, "loss": 1.2701, "step": 22650 }, { "epoch": 0.57, "grad_norm": 37.5, "learning_rate": 4.976271186440678e-06, "loss": 1.3059, "step": 22660 }, { "epoch": 0.57, "grad_norm": 16.5, "learning_rate": 4.969491525423729e-06, "loss": 1.313, "step": 22670 }, { "epoch": 0.57, "grad_norm": 6.75, "learning_rate": 4.96271186440678e-06, "loss": 1.3277, "step": 22680 }, { "epoch": 0.57, "grad_norm": 13.25, "learning_rate": 4.955932203389831e-06, "loss": 1.1521, "step": 22690 }, { "epoch": 0.57, "grad_norm": 13.4375, "learning_rate": 4.949152542372882e-06, "loss": 1.4065, "step": 22700 }, { "epoch": 0.57, "grad_norm": 25.0, "learning_rate": 4.942372881355932e-06, "loss": 1.3772, "step": 22710 }, { "epoch": 0.57, "grad_norm": 31.125, "learning_rate": 4.935593220338984e-06, "loss": 1.3597, "step": 22720 }, { "epoch": 0.57, "grad_norm": 12.625, "learning_rate": 4.928813559322034e-06, "loss": 1.3534, "step": 22730 }, { "epoch": 0.57, "grad_norm": 15.5, "learning_rate": 4.922033898305086e-06, "loss": 1.4141, "step": 22740 }, { "epoch": 0.57, "grad_norm": 26.375, "learning_rate": 4.915254237288136e-06, "loss": 1.4019, "step": 22750 }, { "epoch": 0.57, "grad_norm": 40.25, "learning_rate": 4.908474576271187e-06, "loss": 1.273, "step": 22760 }, { "epoch": 0.57, "grad_norm": 19.25, "learning_rate": 4.901694915254237e-06, "loss": 1.3954, "step": 22770 }, { "epoch": 0.57, "grad_norm": 26.875, "learning_rate": 4.894915254237289e-06, "loss": 1.4013, "step": 22780 }, { "epoch": 0.57, "grad_norm": 28.0, "learning_rate": 4.888135593220339e-06, "loss": 1.428, "step": 22790 }, { "epoch": 0.57, "grad_norm": 27.0, "learning_rate": 4.881355932203391e-06, "loss": 1.1701, "step": 22800 }, { "epoch": 0.57, "grad_norm": 78.0, "learning_rate": 4.874576271186441e-06, "loss": 1.3312, "step": 22810 }, { "epoch": 0.57, "grad_norm": 26.375, "learning_rate": 4.867796610169492e-06, "loss": 1.2204, "step": 22820 }, { "epoch": 0.57, "grad_norm": 32.5, "learning_rate": 4.861016949152543e-06, "loss": 1.4364, "step": 22830 }, { "epoch": 0.57, "grad_norm": 65.5, "learning_rate": 4.854237288135594e-06, "loss": 1.5421, "step": 22840 }, { "epoch": 0.57, "grad_norm": 48.25, "learning_rate": 4.847457627118645e-06, "loss": 1.4326, "step": 22850 }, { "epoch": 0.57, "grad_norm": 47.25, "learning_rate": 4.840677966101695e-06, "loss": 1.243, "step": 22860 }, { "epoch": 0.57, "grad_norm": 12.1875, "learning_rate": 4.833898305084746e-06, "loss": 1.3397, "step": 22870 }, { "epoch": 0.57, "grad_norm": 20.5, "learning_rate": 4.827118644067797e-06, "loss": 1.2613, "step": 22880 }, { "epoch": 0.57, "grad_norm": 37.5, "learning_rate": 4.820338983050848e-06, "loss": 1.3743, "step": 22890 }, { "epoch": 0.57, "grad_norm": 43.25, "learning_rate": 4.813559322033899e-06, "loss": 1.4017, "step": 22900 }, { "epoch": 0.57, "grad_norm": 11.375, "learning_rate": 4.80677966101695e-06, "loss": 1.268, "step": 22910 }, { "epoch": 0.57, "grad_norm": 59.0, "learning_rate": 4.800000000000001e-06, "loss": 1.347, "step": 22920 }, { "epoch": 0.57, "grad_norm": 42.75, "learning_rate": 4.793220338983051e-06, "loss": 1.3816, "step": 22930 }, { "epoch": 0.57, "grad_norm": 38.5, "learning_rate": 4.786440677966102e-06, "loss": 1.4551, "step": 22940 }, { "epoch": 0.57, "grad_norm": 28.875, "learning_rate": 4.779661016949153e-06, "loss": 1.3871, "step": 22950 }, { "epoch": 0.57, "grad_norm": 14.375, "learning_rate": 4.772881355932204e-06, "loss": 1.2011, "step": 22960 }, { "epoch": 0.57, "grad_norm": 31.875, "learning_rate": 4.766101694915254e-06, "loss": 1.3172, "step": 22970 }, { "epoch": 0.57, "grad_norm": 41.75, "learning_rate": 4.759322033898306e-06, "loss": 1.1602, "step": 22980 }, { "epoch": 0.57, "grad_norm": 27.875, "learning_rate": 4.752542372881356e-06, "loss": 1.1804, "step": 22990 }, { "epoch": 0.57, "grad_norm": 17.0, "learning_rate": 4.745762711864408e-06, "loss": 1.4837, "step": 23000 }, { "epoch": 0.58, "grad_norm": 61.75, "learning_rate": 4.738983050847458e-06, "loss": 1.4385, "step": 23010 }, { "epoch": 0.58, "grad_norm": 44.0, "learning_rate": 4.732203389830509e-06, "loss": 1.3842, "step": 23020 }, { "epoch": 0.58, "grad_norm": 51.25, "learning_rate": 4.725423728813559e-06, "loss": 1.4563, "step": 23030 }, { "epoch": 0.58, "grad_norm": 28.875, "learning_rate": 4.718644067796611e-06, "loss": 1.3894, "step": 23040 }, { "epoch": 0.58, "grad_norm": 144.0, "learning_rate": 4.711864406779661e-06, "loss": 1.266, "step": 23050 }, { "epoch": 0.58, "grad_norm": 9.8125, "learning_rate": 4.705084745762713e-06, "loss": 1.446, "step": 23060 }, { "epoch": 0.58, "grad_norm": 17.125, "learning_rate": 4.698305084745763e-06, "loss": 1.2639, "step": 23070 }, { "epoch": 0.58, "grad_norm": 11.125, "learning_rate": 4.691525423728814e-06, "loss": 1.3835, "step": 23080 }, { "epoch": 0.58, "grad_norm": 30.75, "learning_rate": 4.684745762711865e-06, "loss": 1.3717, "step": 23090 }, { "epoch": 0.58, "grad_norm": 28.125, "learning_rate": 4.677966101694916e-06, "loss": 1.4138, "step": 23100 }, { "epoch": 0.58, "grad_norm": 10.75, "learning_rate": 4.671186440677967e-06, "loss": 1.4599, "step": 23110 }, { "epoch": 0.58, "grad_norm": 36.75, "learning_rate": 4.664406779661017e-06, "loss": 1.2974, "step": 23120 }, { "epoch": 0.58, "grad_norm": 11.9375, "learning_rate": 4.657627118644068e-06, "loss": 1.4729, "step": 23130 }, { "epoch": 0.58, "grad_norm": 19.25, "learning_rate": 4.650847457627119e-06, "loss": 1.4423, "step": 23140 }, { "epoch": 0.58, "grad_norm": 22.375, "learning_rate": 4.64406779661017e-06, "loss": 1.2612, "step": 23150 }, { "epoch": 0.58, "grad_norm": 40.5, "learning_rate": 4.637288135593221e-06, "loss": 1.1974, "step": 23160 }, { "epoch": 0.58, "grad_norm": 28.75, "learning_rate": 4.630508474576272e-06, "loss": 1.3801, "step": 23170 }, { "epoch": 0.58, "grad_norm": 33.75, "learning_rate": 4.623728813559323e-06, "loss": 1.248, "step": 23180 }, { "epoch": 0.58, "grad_norm": 34.0, "learning_rate": 4.616949152542373e-06, "loss": 1.4849, "step": 23190 }, { "epoch": 0.58, "grad_norm": 31.25, "learning_rate": 4.610169491525424e-06, "loss": 1.2063, "step": 23200 }, { "epoch": 0.58, "grad_norm": 33.25, "learning_rate": 4.603389830508475e-06, "loss": 1.449, "step": 23210 }, { "epoch": 0.58, "grad_norm": 24.875, "learning_rate": 4.596610169491526e-06, "loss": 1.327, "step": 23220 }, { "epoch": 0.58, "grad_norm": 14.5, "learning_rate": 4.589830508474576e-06, "loss": 1.2771, "step": 23230 }, { "epoch": 0.58, "grad_norm": 16.75, "learning_rate": 4.583050847457628e-06, "loss": 1.1508, "step": 23240 }, { "epoch": 0.58, "grad_norm": 31.125, "learning_rate": 4.576271186440678e-06, "loss": 1.2125, "step": 23250 }, { "epoch": 0.58, "grad_norm": 17.625, "learning_rate": 4.56949152542373e-06, "loss": 1.3442, "step": 23260 }, { "epoch": 0.58, "grad_norm": 37.25, "learning_rate": 4.56271186440678e-06, "loss": 1.238, "step": 23270 }, { "epoch": 0.58, "grad_norm": 21.0, "learning_rate": 4.555932203389831e-06, "loss": 1.3523, "step": 23280 }, { "epoch": 0.58, "grad_norm": 16.375, "learning_rate": 4.549152542372881e-06, "loss": 1.3336, "step": 23290 }, { "epoch": 0.58, "grad_norm": 28.75, "learning_rate": 4.542372881355933e-06, "loss": 1.1508, "step": 23300 }, { "epoch": 0.58, "grad_norm": 47.25, "learning_rate": 4.535593220338983e-06, "loss": 1.3125, "step": 23310 }, { "epoch": 0.58, "grad_norm": 23.875, "learning_rate": 4.528813559322035e-06, "loss": 1.2634, "step": 23320 }, { "epoch": 0.58, "grad_norm": 23.25, "learning_rate": 4.522033898305085e-06, "loss": 1.2934, "step": 23330 }, { "epoch": 0.58, "grad_norm": 58.5, "learning_rate": 4.515254237288136e-06, "loss": 1.2535, "step": 23340 }, { "epoch": 0.58, "grad_norm": 17.5, "learning_rate": 4.508474576271187e-06, "loss": 1.1762, "step": 23350 }, { "epoch": 0.58, "grad_norm": 25.5, "learning_rate": 4.501694915254238e-06, "loss": 1.396, "step": 23360 }, { "epoch": 0.58, "grad_norm": 63.5, "learning_rate": 4.494915254237289e-06, "loss": 1.3609, "step": 23370 }, { "epoch": 0.58, "grad_norm": 42.5, "learning_rate": 4.488135593220339e-06, "loss": 1.3194, "step": 23380 }, { "epoch": 0.58, "grad_norm": 50.5, "learning_rate": 4.48135593220339e-06, "loss": 1.4122, "step": 23390 }, { "epoch": 0.58, "grad_norm": 50.5, "learning_rate": 4.474576271186441e-06, "loss": 1.4809, "step": 23400 }, { "epoch": 0.59, "grad_norm": 44.25, "learning_rate": 4.467796610169492e-06, "loss": 1.2724, "step": 23410 }, { "epoch": 0.59, "grad_norm": 28.375, "learning_rate": 4.461016949152543e-06, "loss": 1.3694, "step": 23420 }, { "epoch": 0.59, "grad_norm": 10.3125, "learning_rate": 4.454237288135594e-06, "loss": 1.4755, "step": 23430 }, { "epoch": 0.59, "grad_norm": 16.375, "learning_rate": 4.447457627118645e-06, "loss": 1.2465, "step": 23440 }, { "epoch": 0.59, "grad_norm": 32.25, "learning_rate": 4.440677966101695e-06, "loss": 1.302, "step": 23450 }, { "epoch": 0.59, "grad_norm": 44.75, "learning_rate": 4.433898305084746e-06, "loss": 1.1708, "step": 23460 }, { "epoch": 0.59, "grad_norm": 55.5, "learning_rate": 4.427118644067797e-06, "loss": 1.4715, "step": 23470 }, { "epoch": 0.59, "grad_norm": 29.625, "learning_rate": 4.420338983050848e-06, "loss": 1.316, "step": 23480 }, { "epoch": 0.59, "grad_norm": 15.625, "learning_rate": 4.413559322033898e-06, "loss": 1.4471, "step": 23490 }, { "epoch": 0.59, "grad_norm": 27.625, "learning_rate": 4.40677966101695e-06, "loss": 1.1327, "step": 23500 }, { "epoch": 0.59, "grad_norm": 21.25, "learning_rate": 4.4e-06, "loss": 1.1244, "step": 23510 }, { "epoch": 0.59, "grad_norm": 15.8125, "learning_rate": 4.393220338983052e-06, "loss": 1.278, "step": 23520 }, { "epoch": 0.59, "grad_norm": 14.1875, "learning_rate": 4.386440677966102e-06, "loss": 1.3043, "step": 23530 }, { "epoch": 0.59, "grad_norm": 22.5, "learning_rate": 4.379661016949153e-06, "loss": 1.3161, "step": 23540 }, { "epoch": 0.59, "grad_norm": 14.8125, "learning_rate": 4.372881355932203e-06, "loss": 1.5512, "step": 23550 }, { "epoch": 0.59, "grad_norm": 46.25, "learning_rate": 4.366101694915255e-06, "loss": 1.2756, "step": 23560 }, { "epoch": 0.59, "grad_norm": 37.75, "learning_rate": 4.359322033898305e-06, "loss": 1.2049, "step": 23570 }, { "epoch": 0.59, "grad_norm": 14.4375, "learning_rate": 4.352542372881357e-06, "loss": 1.5071, "step": 23580 }, { "epoch": 0.59, "grad_norm": 24.75, "learning_rate": 4.345762711864407e-06, "loss": 1.3423, "step": 23590 }, { "epoch": 0.59, "grad_norm": 24.75, "learning_rate": 4.338983050847458e-06, "loss": 1.1705, "step": 23600 }, { "epoch": 0.59, "grad_norm": 39.25, "learning_rate": 4.332203389830509e-06, "loss": 1.2632, "step": 23610 }, { "epoch": 0.59, "grad_norm": 17.5, "learning_rate": 4.32542372881356e-06, "loss": 1.4266, "step": 23620 }, { "epoch": 0.59, "grad_norm": 51.0, "learning_rate": 4.318644067796611e-06, "loss": 1.4648, "step": 23630 }, { "epoch": 0.59, "grad_norm": 19.25, "learning_rate": 4.311864406779661e-06, "loss": 1.3572, "step": 23640 }, { "epoch": 0.59, "grad_norm": 11.75, "learning_rate": 4.305084745762712e-06, "loss": 1.3391, "step": 23650 }, { "epoch": 0.59, "grad_norm": 24.0, "learning_rate": 4.298305084745763e-06, "loss": 1.3104, "step": 23660 }, { "epoch": 0.59, "grad_norm": 15.25, "learning_rate": 4.291525423728814e-06, "loss": 1.2981, "step": 23670 }, { "epoch": 0.59, "grad_norm": 14.875, "learning_rate": 4.284745762711865e-06, "loss": 1.3867, "step": 23680 }, { "epoch": 0.59, "grad_norm": 12.875, "learning_rate": 4.277966101694915e-06, "loss": 1.2998, "step": 23690 }, { "epoch": 0.59, "grad_norm": 25.75, "learning_rate": 4.271186440677967e-06, "loss": 1.4142, "step": 23700 }, { "epoch": 0.59, "grad_norm": 11.0625, "learning_rate": 4.264406779661017e-06, "loss": 1.4814, "step": 23710 }, { "epoch": 0.59, "grad_norm": 21.375, "learning_rate": 4.257627118644068e-06, "loss": 1.527, "step": 23720 }, { "epoch": 0.59, "grad_norm": 11.375, "learning_rate": 4.250847457627119e-06, "loss": 1.2311, "step": 23730 }, { "epoch": 0.59, "grad_norm": 35.75, "learning_rate": 4.24406779661017e-06, "loss": 1.4337, "step": 23740 }, { "epoch": 0.59, "grad_norm": 12.1875, "learning_rate": 4.23728813559322e-06, "loss": 1.2348, "step": 23750 }, { "epoch": 0.59, "grad_norm": 11.5, "learning_rate": 4.230508474576272e-06, "loss": 1.2744, "step": 23760 }, { "epoch": 0.59, "grad_norm": 24.875, "learning_rate": 4.223728813559322e-06, "loss": 1.1926, "step": 23770 }, { "epoch": 0.59, "grad_norm": 33.5, "learning_rate": 4.216949152542374e-06, "loss": 1.3709, "step": 23780 }, { "epoch": 0.59, "grad_norm": 44.0, "learning_rate": 4.210169491525424e-06, "loss": 1.2787, "step": 23790 }, { "epoch": 0.59, "grad_norm": 60.5, "learning_rate": 4.203389830508475e-06, "loss": 1.5057, "step": 23800 }, { "epoch": 0.6, "grad_norm": 24.875, "learning_rate": 4.196610169491525e-06, "loss": 1.3983, "step": 23810 }, { "epoch": 0.6, "grad_norm": 87.0, "learning_rate": 4.189830508474577e-06, "loss": 1.266, "step": 23820 }, { "epoch": 0.6, "grad_norm": 35.75, "learning_rate": 4.183050847457627e-06, "loss": 1.5536, "step": 23830 }, { "epoch": 0.6, "grad_norm": 13.25, "learning_rate": 4.176271186440679e-06, "loss": 1.4579, "step": 23840 }, { "epoch": 0.6, "grad_norm": 13.0625, "learning_rate": 4.169491525423729e-06, "loss": 1.3378, "step": 23850 }, { "epoch": 0.6, "grad_norm": 11.6875, "learning_rate": 4.16271186440678e-06, "loss": 1.4234, "step": 23860 }, { "epoch": 0.6, "grad_norm": 36.75, "learning_rate": 4.155932203389831e-06, "loss": 1.1912, "step": 23870 }, { "epoch": 0.6, "grad_norm": 26.5, "learning_rate": 4.149152542372882e-06, "loss": 1.3767, "step": 23880 }, { "epoch": 0.6, "grad_norm": 12.375, "learning_rate": 4.142372881355933e-06, "loss": 1.4622, "step": 23890 }, { "epoch": 0.6, "grad_norm": 42.25, "learning_rate": 4.135593220338983e-06, "loss": 1.4049, "step": 23900 }, { "epoch": 0.6, "grad_norm": 61.5, "learning_rate": 4.128813559322034e-06, "loss": 1.4161, "step": 23910 }, { "epoch": 0.6, "grad_norm": 12.5625, "learning_rate": 4.122033898305085e-06, "loss": 1.2148, "step": 23920 }, { "epoch": 0.6, "grad_norm": 43.0, "learning_rate": 4.115254237288136e-06, "loss": 1.4088, "step": 23930 }, { "epoch": 0.6, "grad_norm": 34.25, "learning_rate": 4.108474576271187e-06, "loss": 1.4125, "step": 23940 }, { "epoch": 0.6, "grad_norm": 69.0, "learning_rate": 4.101694915254237e-06, "loss": 1.4205, "step": 23950 }, { "epoch": 0.6, "grad_norm": 29.375, "learning_rate": 4.094915254237289e-06, "loss": 1.1675, "step": 23960 }, { "epoch": 0.6, "grad_norm": 27.125, "learning_rate": 4.088135593220339e-06, "loss": 1.2709, "step": 23970 }, { "epoch": 0.6, "grad_norm": 6.84375, "learning_rate": 4.081355932203391e-06, "loss": 1.236, "step": 23980 }, { "epoch": 0.6, "grad_norm": 40.75, "learning_rate": 4.074576271186441e-06, "loss": 1.2984, "step": 23990 }, { "epoch": 0.6, "grad_norm": 24.5, "learning_rate": 4.067796610169492e-06, "loss": 1.2543, "step": 24000 }, { "epoch": 0.6, "grad_norm": 17.0, "learning_rate": 4.061016949152542e-06, "loss": 1.1977, "step": 24010 }, { "epoch": 0.6, "grad_norm": 9.8125, "learning_rate": 4.054237288135594e-06, "loss": 1.2005, "step": 24020 }, { "epoch": 0.6, "grad_norm": 28.875, "learning_rate": 4.047457627118644e-06, "loss": 1.3207, "step": 24030 }, { "epoch": 0.6, "grad_norm": 19.25, "learning_rate": 4.040677966101696e-06, "loss": 1.2664, "step": 24040 }, { "epoch": 0.6, "grad_norm": 26.0, "learning_rate": 4.033898305084746e-06, "loss": 1.2086, "step": 24050 }, { "epoch": 0.6, "grad_norm": 36.5, "learning_rate": 4.027118644067797e-06, "loss": 1.4535, "step": 24060 }, { "epoch": 0.6, "grad_norm": 22.25, "learning_rate": 4.020338983050847e-06, "loss": 1.411, "step": 24070 }, { "epoch": 0.6, "grad_norm": 7.90625, "learning_rate": 4.013559322033899e-06, "loss": 1.3128, "step": 24080 }, { "epoch": 0.6, "grad_norm": 11.875, "learning_rate": 4.006779661016949e-06, "loss": 1.1817, "step": 24090 }, { "epoch": 0.6, "grad_norm": 11.6875, "learning_rate": 4.000000000000001e-06, "loss": 1.382, "step": 24100 }, { "epoch": 0.6, "grad_norm": 9.4375, "learning_rate": 3.993220338983051e-06, "loss": 1.2053, "step": 24110 }, { "epoch": 0.6, "grad_norm": 32.5, "learning_rate": 3.986440677966102e-06, "loss": 1.4975, "step": 24120 }, { "epoch": 0.6, "grad_norm": 13.9375, "learning_rate": 3.979661016949153e-06, "loss": 1.3662, "step": 24130 }, { "epoch": 0.6, "grad_norm": 13.375, "learning_rate": 3.972881355932204e-06, "loss": 1.1596, "step": 24140 }, { "epoch": 0.6, "grad_norm": 16.75, "learning_rate": 3.966101694915255e-06, "loss": 1.4116, "step": 24150 }, { "epoch": 0.6, "grad_norm": 13.1875, "learning_rate": 3.959322033898305e-06, "loss": 1.4245, "step": 24160 }, { "epoch": 0.6, "grad_norm": 14.125, "learning_rate": 3.952542372881356e-06, "loss": 1.5293, "step": 24170 }, { "epoch": 0.6, "grad_norm": 16.125, "learning_rate": 3.945762711864407e-06, "loss": 1.324, "step": 24180 }, { "epoch": 0.6, "grad_norm": 15.5625, "learning_rate": 3.938983050847458e-06, "loss": 1.3789, "step": 24190 }, { "epoch": 0.6, "grad_norm": 23.125, "learning_rate": 3.932203389830509e-06, "loss": 1.4892, "step": 24200 }, { "epoch": 0.61, "grad_norm": 18.125, "learning_rate": 3.925423728813559e-06, "loss": 1.3624, "step": 24210 }, { "epoch": 0.61, "grad_norm": 32.5, "learning_rate": 3.918644067796611e-06, "loss": 1.3137, "step": 24220 }, { "epoch": 0.61, "grad_norm": 32.0, "learning_rate": 3.911864406779661e-06, "loss": 1.5286, "step": 24230 }, { "epoch": 0.61, "grad_norm": 52.75, "learning_rate": 3.905084745762713e-06, "loss": 1.4272, "step": 24240 }, { "epoch": 0.61, "grad_norm": 54.25, "learning_rate": 3.898305084745763e-06, "loss": 1.3123, "step": 24250 }, { "epoch": 0.61, "grad_norm": 79.5, "learning_rate": 3.891525423728814e-06, "loss": 1.2272, "step": 24260 }, { "epoch": 0.61, "grad_norm": 22.375, "learning_rate": 3.884745762711864e-06, "loss": 1.369, "step": 24270 }, { "epoch": 0.61, "grad_norm": 25.5, "learning_rate": 3.877966101694916e-06, "loss": 1.3113, "step": 24280 }, { "epoch": 0.61, "grad_norm": 30.375, "learning_rate": 3.871186440677966e-06, "loss": 1.3477, "step": 24290 }, { "epoch": 0.61, "grad_norm": 24.125, "learning_rate": 3.864406779661018e-06, "loss": 1.3172, "step": 24300 }, { "epoch": 0.61, "grad_norm": 19.75, "learning_rate": 3.857627118644068e-06, "loss": 1.3028, "step": 24310 }, { "epoch": 0.61, "grad_norm": 30.375, "learning_rate": 3.850847457627119e-06, "loss": 1.4591, "step": 24320 }, { "epoch": 0.61, "grad_norm": 35.25, "learning_rate": 3.844067796610169e-06, "loss": 1.3009, "step": 24330 }, { "epoch": 0.61, "grad_norm": 24.375, "learning_rate": 3.837288135593221e-06, "loss": 1.35, "step": 24340 }, { "epoch": 0.61, "grad_norm": 31.0, "learning_rate": 3.830508474576271e-06, "loss": 1.4239, "step": 24350 }, { "epoch": 0.61, "grad_norm": 50.25, "learning_rate": 3.823728813559323e-06, "loss": 1.2657, "step": 24360 }, { "epoch": 0.61, "grad_norm": 45.75, "learning_rate": 3.816949152542373e-06, "loss": 1.3529, "step": 24370 }, { "epoch": 0.61, "grad_norm": 31.5, "learning_rate": 3.8101694915254238e-06, "loss": 1.3574, "step": 24380 }, { "epoch": 0.61, "grad_norm": 21.125, "learning_rate": 3.8033898305084748e-06, "loss": 1.2754, "step": 24390 }, { "epoch": 0.61, "grad_norm": 11.3125, "learning_rate": 3.7966101694915257e-06, "loss": 1.2281, "step": 24400 }, { "epoch": 0.61, "grad_norm": 22.125, "learning_rate": 3.7898305084745767e-06, "loss": 1.3828, "step": 24410 }, { "epoch": 0.61, "grad_norm": 13.0, "learning_rate": 3.7830508474576273e-06, "loss": 1.4197, "step": 24420 }, { "epoch": 0.61, "grad_norm": 33.0, "learning_rate": 3.7762711864406782e-06, "loss": 1.2412, "step": 24430 }, { "epoch": 0.61, "grad_norm": 22.375, "learning_rate": 3.7694915254237292e-06, "loss": 1.1763, "step": 24440 }, { "epoch": 0.61, "grad_norm": 29.625, "learning_rate": 3.76271186440678e-06, "loss": 1.2184, "step": 24450 }, { "epoch": 0.61, "grad_norm": 15.375, "learning_rate": 3.755932203389831e-06, "loss": 1.381, "step": 24460 }, { "epoch": 0.61, "grad_norm": 12.9375, "learning_rate": 3.7491525423728813e-06, "loss": 1.2347, "step": 24470 }, { "epoch": 0.61, "grad_norm": 18.125, "learning_rate": 3.7423728813559323e-06, "loss": 1.2788, "step": 24480 }, { "epoch": 0.61, "grad_norm": 9.0625, "learning_rate": 3.7355932203389833e-06, "loss": 1.3998, "step": 24490 }, { "epoch": 0.61, "grad_norm": 19.375, "learning_rate": 3.7288135593220342e-06, "loss": 1.489, "step": 24500 }, { "epoch": 0.61, "grad_norm": 50.0, "learning_rate": 3.7220338983050852e-06, "loss": 1.3069, "step": 24510 }, { "epoch": 0.61, "grad_norm": 21.25, "learning_rate": 3.715254237288136e-06, "loss": 1.2861, "step": 24520 }, { "epoch": 0.61, "grad_norm": 18.875, "learning_rate": 3.7084745762711867e-06, "loss": 1.2172, "step": 24530 }, { "epoch": 0.61, "grad_norm": 56.75, "learning_rate": 3.7016949152542377e-06, "loss": 1.2433, "step": 24540 }, { "epoch": 0.61, "grad_norm": 54.25, "learning_rate": 3.6949152542372883e-06, "loss": 1.3888, "step": 24550 }, { "epoch": 0.61, "grad_norm": 25.25, "learning_rate": 3.6881355932203393e-06, "loss": 1.3927, "step": 24560 }, { "epoch": 0.61, "grad_norm": 48.25, "learning_rate": 3.6813559322033902e-06, "loss": 1.3315, "step": 24570 }, { "epoch": 0.61, "grad_norm": 76.0, "learning_rate": 3.6745762711864408e-06, "loss": 1.3999, "step": 24580 }, { "epoch": 0.61, "grad_norm": 50.5, "learning_rate": 3.6677966101694918e-06, "loss": 1.3443, "step": 24590 }, { "epoch": 0.61, "grad_norm": 25.375, "learning_rate": 3.6610169491525427e-06, "loss": 1.4185, "step": 24600 }, { "epoch": 0.62, "grad_norm": 14.3125, "learning_rate": 3.6542372881355937e-06, "loss": 1.2463, "step": 24610 }, { "epoch": 0.62, "grad_norm": 13.25, "learning_rate": 3.6474576271186447e-06, "loss": 1.1997, "step": 24620 }, { "epoch": 0.62, "grad_norm": 16.875, "learning_rate": 3.640677966101695e-06, "loss": 1.3394, "step": 24630 }, { "epoch": 0.62, "grad_norm": 32.25, "learning_rate": 3.633898305084746e-06, "loss": 1.2986, "step": 24640 }, { "epoch": 0.62, "grad_norm": 13.25, "learning_rate": 3.6271186440677968e-06, "loss": 1.4319, "step": 24650 }, { "epoch": 0.62, "grad_norm": 17.75, "learning_rate": 3.6203389830508478e-06, "loss": 1.2953, "step": 24660 }, { "epoch": 0.62, "grad_norm": 7.84375, "learning_rate": 3.6135593220338987e-06, "loss": 1.4395, "step": 24670 }, { "epoch": 0.62, "grad_norm": 12.5, "learning_rate": 3.6067796610169493e-06, "loss": 1.294, "step": 24680 }, { "epoch": 0.62, "grad_norm": 23.75, "learning_rate": 3.6000000000000003e-06, "loss": 1.3854, "step": 24690 }, { "epoch": 0.62, "grad_norm": 24.5, "learning_rate": 3.5932203389830512e-06, "loss": 1.2867, "step": 24700 }, { "epoch": 0.62, "grad_norm": 27.0, "learning_rate": 3.5864406779661022e-06, "loss": 1.2254, "step": 24710 }, { "epoch": 0.62, "grad_norm": 10.125, "learning_rate": 3.579661016949153e-06, "loss": 1.3512, "step": 24720 }, { "epoch": 0.62, "grad_norm": 59.75, "learning_rate": 3.5728813559322033e-06, "loss": 1.3847, "step": 24730 }, { "epoch": 0.62, "grad_norm": 51.0, "learning_rate": 3.5661016949152543e-06, "loss": 1.3099, "step": 24740 }, { "epoch": 0.62, "grad_norm": 29.625, "learning_rate": 3.5593220338983053e-06, "loss": 1.3505, "step": 24750 }, { "epoch": 0.62, "grad_norm": 43.5, "learning_rate": 3.5525423728813563e-06, "loss": 1.055, "step": 24760 }, { "epoch": 0.62, "grad_norm": 14.1875, "learning_rate": 3.5457627118644072e-06, "loss": 1.4283, "step": 24770 }, { "epoch": 0.62, "grad_norm": 31.625, "learning_rate": 3.538983050847458e-06, "loss": 1.5257, "step": 24780 }, { "epoch": 0.62, "grad_norm": 17.5, "learning_rate": 3.5322033898305088e-06, "loss": 1.3837, "step": 24790 }, { "epoch": 0.62, "grad_norm": 26.875, "learning_rate": 3.5254237288135597e-06, "loss": 1.3111, "step": 24800 }, { "epoch": 0.62, "grad_norm": 13.9375, "learning_rate": 3.5186440677966103e-06, "loss": 1.4159, "step": 24810 }, { "epoch": 0.62, "grad_norm": 47.25, "learning_rate": 3.5118644067796613e-06, "loss": 1.4506, "step": 24820 }, { "epoch": 0.62, "grad_norm": 28.375, "learning_rate": 3.5050847457627122e-06, "loss": 1.235, "step": 24830 }, { "epoch": 0.62, "grad_norm": 41.25, "learning_rate": 3.498305084745763e-06, "loss": 1.3643, "step": 24840 }, { "epoch": 0.62, "grad_norm": 15.5625, "learning_rate": 3.4915254237288138e-06, "loss": 1.1906, "step": 24850 }, { "epoch": 0.62, "grad_norm": 29.625, "learning_rate": 3.4847457627118648e-06, "loss": 1.3653, "step": 24860 }, { "epoch": 0.62, "grad_norm": 19.625, "learning_rate": 3.4779661016949157e-06, "loss": 1.2675, "step": 24870 }, { "epoch": 0.62, "grad_norm": 18.375, "learning_rate": 3.4711864406779667e-06, "loss": 1.4436, "step": 24880 }, { "epoch": 0.62, "grad_norm": 20.5, "learning_rate": 3.464406779661017e-06, "loss": 1.3635, "step": 24890 }, { "epoch": 0.62, "grad_norm": 18.125, "learning_rate": 3.457627118644068e-06, "loss": 1.4103, "step": 24900 }, { "epoch": 0.62, "grad_norm": 28.5, "learning_rate": 3.450847457627119e-06, "loss": 1.2253, "step": 24910 }, { "epoch": 0.62, "grad_norm": 41.25, "learning_rate": 3.4440677966101698e-06, "loss": 1.3705, "step": 24920 }, { "epoch": 0.62, "grad_norm": 39.25, "learning_rate": 3.4372881355932207e-06, "loss": 1.4641, "step": 24930 }, { "epoch": 0.62, "grad_norm": 30.875, "learning_rate": 3.4305084745762713e-06, "loss": 1.4391, "step": 24940 }, { "epoch": 0.62, "grad_norm": 33.0, "learning_rate": 3.4237288135593223e-06, "loss": 1.2609, "step": 24950 }, { "epoch": 0.62, "grad_norm": 13.0, "learning_rate": 3.4169491525423733e-06, "loss": 1.4069, "step": 24960 }, { "epoch": 0.62, "grad_norm": 30.125, "learning_rate": 3.4101694915254242e-06, "loss": 1.3547, "step": 24970 }, { "epoch": 0.62, "grad_norm": 33.5, "learning_rate": 3.403389830508475e-06, "loss": 1.0802, "step": 24980 }, { "epoch": 0.62, "grad_norm": 16.625, "learning_rate": 3.3966101694915253e-06, "loss": 1.4374, "step": 24990 }, { "epoch": 0.62, "grad_norm": 18.375, "learning_rate": 3.3898305084745763e-06, "loss": 1.3566, "step": 25000 }, { "epoch": 0.62, "eval_loss": 1.3512009382247925, "eval_runtime": 59.1664, "eval_samples_per_second": 16.901, "eval_steps_per_second": 16.901, "step": 25000 }, { "epoch": 0.63, "grad_norm": 36.25, "learning_rate": 3.3830508474576273e-06, "loss": 1.5974, "step": 25010 }, { "epoch": 0.63, "grad_norm": 21.125, "learning_rate": 3.3762711864406783e-06, "loss": 1.2943, "step": 25020 }, { "epoch": 0.63, "grad_norm": 13.9375, "learning_rate": 3.3694915254237292e-06, "loss": 1.2029, "step": 25030 }, { "epoch": 0.63, "grad_norm": 25.375, "learning_rate": 3.3627118644067802e-06, "loss": 1.4546, "step": 25040 }, { "epoch": 0.63, "grad_norm": 23.5, "learning_rate": 3.3559322033898308e-06, "loss": 1.4137, "step": 25050 }, { "epoch": 0.63, "grad_norm": 17.125, "learning_rate": 3.3491525423728817e-06, "loss": 1.3464, "step": 25060 }, { "epoch": 0.63, "grad_norm": 16.75, "learning_rate": 3.3423728813559327e-06, "loss": 1.3147, "step": 25070 }, { "epoch": 0.63, "grad_norm": 14.5, "learning_rate": 3.3355932203389833e-06, "loss": 1.3724, "step": 25080 }, { "epoch": 0.63, "grad_norm": 33.5, "learning_rate": 3.3288135593220343e-06, "loss": 1.4675, "step": 25090 }, { "epoch": 0.63, "grad_norm": 9.4375, "learning_rate": 3.322033898305085e-06, "loss": 1.337, "step": 25100 }, { "epoch": 0.63, "grad_norm": 19.875, "learning_rate": 3.3152542372881358e-06, "loss": 1.3212, "step": 25110 }, { "epoch": 0.63, "grad_norm": 26.375, "learning_rate": 3.3084745762711868e-06, "loss": 1.3812, "step": 25120 }, { "epoch": 0.63, "grad_norm": 14.4375, "learning_rate": 3.3016949152542377e-06, "loss": 1.2622, "step": 25130 }, { "epoch": 0.63, "grad_norm": 51.25, "learning_rate": 3.2949152542372887e-06, "loss": 1.3327, "step": 25140 }, { "epoch": 0.63, "grad_norm": 7.03125, "learning_rate": 3.288135593220339e-06, "loss": 1.3353, "step": 25150 }, { "epoch": 0.63, "grad_norm": 11.25, "learning_rate": 3.28135593220339e-06, "loss": 1.1754, "step": 25160 }, { "epoch": 0.63, "grad_norm": 12.8125, "learning_rate": 3.274576271186441e-06, "loss": 1.4266, "step": 25170 }, { "epoch": 0.63, "grad_norm": 20.25, "learning_rate": 3.2677966101694918e-06, "loss": 1.2969, "step": 25180 }, { "epoch": 0.63, "grad_norm": 44.0, "learning_rate": 3.2610169491525428e-06, "loss": 1.4229, "step": 25190 }, { "epoch": 0.63, "grad_norm": 28.875, "learning_rate": 3.2542372881355933e-06, "loss": 1.2987, "step": 25200 }, { "epoch": 0.63, "grad_norm": 31.0, "learning_rate": 3.2474576271186443e-06, "loss": 1.1972, "step": 25210 }, { "epoch": 0.63, "grad_norm": 48.75, "learning_rate": 3.2406779661016953e-06, "loss": 1.3853, "step": 25220 }, { "epoch": 0.63, "grad_norm": 38.5, "learning_rate": 3.2338983050847462e-06, "loss": 1.1954, "step": 25230 }, { "epoch": 0.63, "grad_norm": 28.625, "learning_rate": 3.2271186440677972e-06, "loss": 1.4971, "step": 25240 }, { "epoch": 0.63, "grad_norm": 20.0, "learning_rate": 3.2203389830508473e-06, "loss": 1.4897, "step": 25250 }, { "epoch": 0.63, "grad_norm": 25.625, "learning_rate": 3.2135593220338983e-06, "loss": 1.2711, "step": 25260 }, { "epoch": 0.63, "grad_norm": 13.1875, "learning_rate": 3.2067796610169493e-06, "loss": 1.4647, "step": 25270 }, { "epoch": 0.63, "grad_norm": 25.125, "learning_rate": 3.2000000000000003e-06, "loss": 1.3632, "step": 25280 }, { "epoch": 0.63, "grad_norm": 23.125, "learning_rate": 3.1932203389830513e-06, "loss": 1.3991, "step": 25290 }, { "epoch": 0.63, "grad_norm": 31.75, "learning_rate": 3.186440677966102e-06, "loss": 1.4431, "step": 25300 }, { "epoch": 0.63, "grad_norm": 23.625, "learning_rate": 3.1796610169491528e-06, "loss": 1.2923, "step": 25310 }, { "epoch": 0.63, "grad_norm": 20.25, "learning_rate": 3.1728813559322038e-06, "loss": 1.3391, "step": 25320 }, { "epoch": 0.63, "grad_norm": 17.5, "learning_rate": 3.1661016949152547e-06, "loss": 1.3396, "step": 25330 }, { "epoch": 0.63, "grad_norm": 12.6875, "learning_rate": 3.1593220338983053e-06, "loss": 1.451, "step": 25340 }, { "epoch": 0.63, "grad_norm": 52.75, "learning_rate": 3.1525423728813563e-06, "loss": 1.1884, "step": 25350 }, { "epoch": 0.63, "grad_norm": 12.1875, "learning_rate": 3.145762711864407e-06, "loss": 1.4216, "step": 25360 }, { "epoch": 0.63, "grad_norm": 18.5, "learning_rate": 3.138983050847458e-06, "loss": 1.2726, "step": 25370 }, { "epoch": 0.63, "grad_norm": 20.75, "learning_rate": 3.1322033898305088e-06, "loss": 1.1607, "step": 25380 }, { "epoch": 0.63, "grad_norm": 24.375, "learning_rate": 3.1254237288135598e-06, "loss": 1.3368, "step": 25390 }, { "epoch": 0.64, "grad_norm": 22.5, "learning_rate": 3.1186440677966107e-06, "loss": 1.3483, "step": 25400 }, { "epoch": 0.64, "grad_norm": 15.3125, "learning_rate": 3.111864406779661e-06, "loss": 1.376, "step": 25410 }, { "epoch": 0.64, "grad_norm": 30.625, "learning_rate": 3.105084745762712e-06, "loss": 1.5073, "step": 25420 }, { "epoch": 0.64, "grad_norm": 20.375, "learning_rate": 3.098305084745763e-06, "loss": 1.2712, "step": 25430 }, { "epoch": 0.64, "grad_norm": 30.0, "learning_rate": 3.091525423728814e-06, "loss": 1.4744, "step": 25440 }, { "epoch": 0.64, "grad_norm": 45.75, "learning_rate": 3.0847457627118648e-06, "loss": 1.1748, "step": 25450 }, { "epoch": 0.64, "grad_norm": 17.75, "learning_rate": 3.0779661016949153e-06, "loss": 1.3216, "step": 25460 }, { "epoch": 0.64, "grad_norm": 10.25, "learning_rate": 3.0711864406779663e-06, "loss": 1.3373, "step": 25470 }, { "epoch": 0.64, "grad_norm": 22.25, "learning_rate": 3.0644067796610173e-06, "loss": 1.1978, "step": 25480 }, { "epoch": 0.64, "grad_norm": 30.125, "learning_rate": 3.0576271186440683e-06, "loss": 1.2682, "step": 25490 }, { "epoch": 0.64, "grad_norm": 16.0, "learning_rate": 3.0508474576271192e-06, "loss": 1.1775, "step": 25500 }, { "epoch": 0.64, "grad_norm": 23.125, "learning_rate": 3.0440677966101694e-06, "loss": 1.1059, "step": 25510 }, { "epoch": 0.64, "grad_norm": 25.5, "learning_rate": 3.0372881355932203e-06, "loss": 1.2425, "step": 25520 }, { "epoch": 0.64, "grad_norm": 14.3125, "learning_rate": 3.0305084745762713e-06, "loss": 1.4788, "step": 25530 }, { "epoch": 0.64, "grad_norm": 41.25, "learning_rate": 3.0237288135593223e-06, "loss": 1.4762, "step": 25540 }, { "epoch": 0.64, "grad_norm": 19.375, "learning_rate": 3.0169491525423733e-06, "loss": 1.3921, "step": 25550 }, { "epoch": 0.64, "grad_norm": 28.125, "learning_rate": 3.010169491525424e-06, "loss": 1.4866, "step": 25560 }, { "epoch": 0.64, "grad_norm": 76.5, "learning_rate": 3.003389830508475e-06, "loss": 1.2202, "step": 25570 }, { "epoch": 0.64, "grad_norm": 49.5, "learning_rate": 2.9966101694915258e-06, "loss": 1.3687, "step": 25580 }, { "epoch": 0.64, "grad_norm": 25.0, "learning_rate": 2.9898305084745768e-06, "loss": 1.5083, "step": 25590 }, { "epoch": 0.64, "grad_norm": 20.5, "learning_rate": 2.9830508474576277e-06, "loss": 1.3506, "step": 25600 }, { "epoch": 0.64, "grad_norm": 40.5, "learning_rate": 2.9762711864406783e-06, "loss": 1.5132, "step": 25610 }, { "epoch": 0.64, "grad_norm": 40.75, "learning_rate": 2.969491525423729e-06, "loss": 1.3987, "step": 25620 }, { "epoch": 0.64, "grad_norm": 37.25, "learning_rate": 2.96271186440678e-06, "loss": 1.3267, "step": 25630 }, { "epoch": 0.64, "grad_norm": 17.625, "learning_rate": 2.955932203389831e-06, "loss": 1.2356, "step": 25640 }, { "epoch": 0.64, "grad_norm": 23.875, "learning_rate": 2.9491525423728818e-06, "loss": 1.5515, "step": 25650 }, { "epoch": 0.64, "grad_norm": 32.25, "learning_rate": 2.9423728813559327e-06, "loss": 1.3458, "step": 25660 }, { "epoch": 0.64, "grad_norm": 24.25, "learning_rate": 2.935593220338983e-06, "loss": 1.3487, "step": 25670 }, { "epoch": 0.64, "grad_norm": 10.0, "learning_rate": 2.928813559322034e-06, "loss": 1.2515, "step": 25680 }, { "epoch": 0.64, "grad_norm": 11.75, "learning_rate": 2.922033898305085e-06, "loss": 1.4637, "step": 25690 }, { "epoch": 0.64, "grad_norm": 33.0, "learning_rate": 2.915254237288136e-06, "loss": 1.2734, "step": 25700 }, { "epoch": 0.64, "grad_norm": 31.125, "learning_rate": 2.9084745762711868e-06, "loss": 1.3431, "step": 25710 }, { "epoch": 0.64, "grad_norm": 20.375, "learning_rate": 2.9016949152542373e-06, "loss": 1.2379, "step": 25720 }, { "epoch": 0.64, "grad_norm": 27.5, "learning_rate": 2.8949152542372883e-06, "loss": 1.4185, "step": 25730 }, { "epoch": 0.64, "grad_norm": 23.625, "learning_rate": 2.8881355932203393e-06, "loss": 1.3128, "step": 25740 }, { "epoch": 0.64, "grad_norm": 16.875, "learning_rate": 2.8813559322033903e-06, "loss": 1.3461, "step": 25750 }, { "epoch": 0.64, "grad_norm": 29.375, "learning_rate": 2.8745762711864412e-06, "loss": 1.3569, "step": 25760 }, { "epoch": 0.64, "grad_norm": 40.5, "learning_rate": 2.8677966101694914e-06, "loss": 1.1441, "step": 25770 }, { "epoch": 0.64, "grad_norm": 20.75, "learning_rate": 2.8610169491525424e-06, "loss": 1.2368, "step": 25780 }, { "epoch": 0.64, "grad_norm": 23.5, "learning_rate": 2.8542372881355933e-06, "loss": 1.167, "step": 25790 }, { "epoch": 0.65, "grad_norm": 16.625, "learning_rate": 2.8474576271186443e-06, "loss": 1.2205, "step": 25800 }, { "epoch": 0.65, "grad_norm": 7.40625, "learning_rate": 2.8406779661016953e-06, "loss": 1.2521, "step": 25810 }, { "epoch": 0.65, "grad_norm": 15.4375, "learning_rate": 2.833898305084746e-06, "loss": 1.4285, "step": 25820 }, { "epoch": 0.65, "grad_norm": 25.875, "learning_rate": 2.827118644067797e-06, "loss": 1.4341, "step": 25830 }, { "epoch": 0.65, "grad_norm": 41.25, "learning_rate": 2.820338983050848e-06, "loss": 1.3918, "step": 25840 }, { "epoch": 0.65, "grad_norm": 17.0, "learning_rate": 2.8135593220338988e-06, "loss": 1.3055, "step": 25850 }, { "epoch": 0.65, "grad_norm": 22.375, "learning_rate": 2.8067796610169497e-06, "loss": 1.3505, "step": 25860 }, { "epoch": 0.65, "grad_norm": 16.75, "learning_rate": 2.8000000000000003e-06, "loss": 1.3144, "step": 25870 }, { "epoch": 0.65, "grad_norm": 32.25, "learning_rate": 2.793220338983051e-06, "loss": 1.3681, "step": 25880 }, { "epoch": 0.65, "grad_norm": 50.75, "learning_rate": 2.786440677966102e-06, "loss": 1.3335, "step": 25890 }, { "epoch": 0.65, "grad_norm": 23.5, "learning_rate": 2.779661016949153e-06, "loss": 1.493, "step": 25900 }, { "epoch": 0.65, "grad_norm": 32.5, "learning_rate": 2.7728813559322038e-06, "loss": 1.574, "step": 25910 }, { "epoch": 0.65, "grad_norm": 17.375, "learning_rate": 2.7661016949152548e-06, "loss": 1.415, "step": 25920 }, { "epoch": 0.65, "grad_norm": 13.75, "learning_rate": 2.7593220338983053e-06, "loss": 1.1766, "step": 25930 }, { "epoch": 0.65, "grad_norm": 20.875, "learning_rate": 2.752542372881356e-06, "loss": 1.3461, "step": 25940 }, { "epoch": 0.65, "grad_norm": 36.0, "learning_rate": 2.745762711864407e-06, "loss": 1.4238, "step": 25950 }, { "epoch": 0.65, "grad_norm": 24.5, "learning_rate": 2.738983050847458e-06, "loss": 1.3314, "step": 25960 }, { "epoch": 0.65, "grad_norm": 12.0, "learning_rate": 2.732203389830509e-06, "loss": 1.4093, "step": 25970 }, { "epoch": 0.65, "grad_norm": 47.75, "learning_rate": 2.7254237288135593e-06, "loss": 1.1776, "step": 25980 }, { "epoch": 0.65, "grad_norm": 9.5, "learning_rate": 2.7186440677966103e-06, "loss": 1.3146, "step": 25990 }, { "epoch": 0.65, "grad_norm": 26.0, "learning_rate": 2.7118644067796613e-06, "loss": 1.4606, "step": 26000 }, { "epoch": 0.65, "grad_norm": 10.375, "learning_rate": 2.7050847457627123e-06, "loss": 1.2379, "step": 26010 }, { "epoch": 0.65, "grad_norm": 21.0, "learning_rate": 2.6983050847457633e-06, "loss": 1.3822, "step": 26020 }, { "epoch": 0.65, "grad_norm": 14.5625, "learning_rate": 2.6915254237288134e-06, "loss": 1.3943, "step": 26030 }, { "epoch": 0.65, "grad_norm": 25.625, "learning_rate": 2.6847457627118644e-06, "loss": 1.4265, "step": 26040 }, { "epoch": 0.65, "grad_norm": 24.125, "learning_rate": 2.6779661016949153e-06, "loss": 1.3966, "step": 26050 }, { "epoch": 0.65, "grad_norm": 21.875, "learning_rate": 2.6711864406779663e-06, "loss": 1.4399, "step": 26060 }, { "epoch": 0.65, "grad_norm": 38.75, "learning_rate": 2.6644067796610173e-06, "loss": 1.3613, "step": 26070 }, { "epoch": 0.65, "grad_norm": 42.25, "learning_rate": 2.657627118644068e-06, "loss": 1.2613, "step": 26080 }, { "epoch": 0.65, "grad_norm": 43.25, "learning_rate": 2.650847457627119e-06, "loss": 1.2263, "step": 26090 }, { "epoch": 0.65, "grad_norm": 44.25, "learning_rate": 2.64406779661017e-06, "loss": 1.2948, "step": 26100 }, { "epoch": 0.65, "grad_norm": 19.0, "learning_rate": 2.6372881355932208e-06, "loss": 1.3783, "step": 26110 }, { "epoch": 0.65, "grad_norm": 35.5, "learning_rate": 2.6305084745762718e-06, "loss": 1.3806, "step": 26120 }, { "epoch": 0.65, "grad_norm": 20.0, "learning_rate": 2.6237288135593223e-06, "loss": 1.3714, "step": 26130 }, { "epoch": 0.65, "grad_norm": 28.0, "learning_rate": 2.616949152542373e-06, "loss": 1.2989, "step": 26140 }, { "epoch": 0.65, "grad_norm": 28.875, "learning_rate": 2.610169491525424e-06, "loss": 1.3626, "step": 26150 }, { "epoch": 0.65, "grad_norm": 23.125, "learning_rate": 2.603389830508475e-06, "loss": 1.3074, "step": 26160 }, { "epoch": 0.65, "grad_norm": 11.625, "learning_rate": 2.596610169491526e-06, "loss": 1.5129, "step": 26170 }, { "epoch": 0.65, "grad_norm": 11.0625, "learning_rate": 2.5898305084745768e-06, "loss": 1.3508, "step": 26180 }, { "epoch": 0.65, "grad_norm": 7.25, "learning_rate": 2.5830508474576273e-06, "loss": 1.3963, "step": 26190 }, { "epoch": 0.66, "grad_norm": 7.0625, "learning_rate": 2.576271186440678e-06, "loss": 1.4194, "step": 26200 }, { "epoch": 0.66, "grad_norm": 34.0, "learning_rate": 2.569491525423729e-06, "loss": 1.2905, "step": 26210 }, { "epoch": 0.66, "grad_norm": 12.875, "learning_rate": 2.56271186440678e-06, "loss": 1.554, "step": 26220 }, { "epoch": 0.66, "grad_norm": 24.375, "learning_rate": 2.555932203389831e-06, "loss": 1.1871, "step": 26230 }, { "epoch": 0.66, "grad_norm": 62.75, "learning_rate": 2.5491525423728814e-06, "loss": 1.3252, "step": 26240 }, { "epoch": 0.66, "grad_norm": 52.0, "learning_rate": 2.5423728813559323e-06, "loss": 1.3434, "step": 26250 }, { "epoch": 0.66, "grad_norm": 32.0, "learning_rate": 2.5355932203389833e-06, "loss": 1.5209, "step": 26260 }, { "epoch": 0.66, "grad_norm": 34.0, "learning_rate": 2.5288135593220343e-06, "loss": 1.2505, "step": 26270 }, { "epoch": 0.66, "grad_norm": 23.875, "learning_rate": 2.5220338983050853e-06, "loss": 1.3513, "step": 26280 }, { "epoch": 0.66, "grad_norm": 26.625, "learning_rate": 2.5152542372881354e-06, "loss": 1.4316, "step": 26290 }, { "epoch": 0.66, "grad_norm": 29.25, "learning_rate": 2.5084745762711864e-06, "loss": 1.367, "step": 26300 }, { "epoch": 0.66, "grad_norm": 22.75, "learning_rate": 2.5016949152542374e-06, "loss": 1.3901, "step": 26310 }, { "epoch": 0.66, "grad_norm": 26.75, "learning_rate": 2.4949152542372883e-06, "loss": 1.4063, "step": 26320 }, { "epoch": 0.66, "grad_norm": 12.0625, "learning_rate": 2.488135593220339e-06, "loss": 1.4608, "step": 26330 }, { "epoch": 0.66, "grad_norm": 16.375, "learning_rate": 2.48135593220339e-06, "loss": 1.1745, "step": 26340 }, { "epoch": 0.66, "grad_norm": 23.25, "learning_rate": 2.474576271186441e-06, "loss": 1.4735, "step": 26350 }, { "epoch": 0.66, "grad_norm": 38.0, "learning_rate": 2.467796610169492e-06, "loss": 1.1955, "step": 26360 }, { "epoch": 0.66, "grad_norm": 34.0, "learning_rate": 2.461016949152543e-06, "loss": 1.3215, "step": 26370 }, { "epoch": 0.66, "grad_norm": 30.875, "learning_rate": 2.4542372881355933e-06, "loss": 1.3611, "step": 26380 }, { "epoch": 0.66, "grad_norm": 28.625, "learning_rate": 2.4474576271186443e-06, "loss": 1.2287, "step": 26390 }, { "epoch": 0.66, "grad_norm": 26.625, "learning_rate": 2.4406779661016953e-06, "loss": 1.2022, "step": 26400 }, { "epoch": 0.66, "grad_norm": 30.625, "learning_rate": 2.433898305084746e-06, "loss": 1.3603, "step": 26410 }, { "epoch": 0.66, "grad_norm": 13.8125, "learning_rate": 2.427118644067797e-06, "loss": 1.463, "step": 26420 }, { "epoch": 0.66, "grad_norm": 14.9375, "learning_rate": 2.4203389830508474e-06, "loss": 1.3554, "step": 26430 }, { "epoch": 0.66, "grad_norm": 17.125, "learning_rate": 2.4135593220338984e-06, "loss": 1.4788, "step": 26440 }, { "epoch": 0.66, "grad_norm": 40.75, "learning_rate": 2.4067796610169493e-06, "loss": 1.3611, "step": 26450 }, { "epoch": 0.66, "grad_norm": 28.0, "learning_rate": 2.4000000000000003e-06, "loss": 1.3523, "step": 26460 }, { "epoch": 0.66, "grad_norm": 23.375, "learning_rate": 2.393220338983051e-06, "loss": 1.3743, "step": 26470 }, { "epoch": 0.66, "grad_norm": 54.25, "learning_rate": 2.386440677966102e-06, "loss": 1.3022, "step": 26480 }, { "epoch": 0.66, "grad_norm": 13.0, "learning_rate": 2.379661016949153e-06, "loss": 1.3887, "step": 26490 }, { "epoch": 0.66, "grad_norm": 15.9375, "learning_rate": 2.372881355932204e-06, "loss": 1.4052, "step": 26500 }, { "epoch": 0.66, "grad_norm": 29.25, "learning_rate": 2.3661016949152544e-06, "loss": 1.3447, "step": 26510 }, { "epoch": 0.66, "grad_norm": 22.375, "learning_rate": 2.3593220338983053e-06, "loss": 1.3072, "step": 26520 }, { "epoch": 0.66, "grad_norm": 14.125, "learning_rate": 2.3525423728813563e-06, "loss": 1.4062, "step": 26530 }, { "epoch": 0.66, "grad_norm": 41.75, "learning_rate": 2.345762711864407e-06, "loss": 1.3262, "step": 26540 }, { "epoch": 0.66, "grad_norm": 24.25, "learning_rate": 2.338983050847458e-06, "loss": 1.4441, "step": 26550 }, { "epoch": 0.66, "grad_norm": 15.3125, "learning_rate": 2.3322033898305084e-06, "loss": 1.3227, "step": 26560 }, { "epoch": 0.66, "grad_norm": 52.25, "learning_rate": 2.3254237288135594e-06, "loss": 1.3981, "step": 26570 }, { "epoch": 0.66, "grad_norm": 30.125, "learning_rate": 2.3186440677966103e-06, "loss": 1.4248, "step": 26580 }, { "epoch": 0.66, "grad_norm": 70.0, "learning_rate": 2.3118644067796613e-06, "loss": 1.3605, "step": 26590 }, { "epoch": 0.67, "grad_norm": 35.5, "learning_rate": 2.305084745762712e-06, "loss": 1.2909, "step": 26600 }, { "epoch": 0.67, "grad_norm": 13.0, "learning_rate": 2.298305084745763e-06, "loss": 1.3771, "step": 26610 }, { "epoch": 0.67, "grad_norm": 49.25, "learning_rate": 2.291525423728814e-06, "loss": 1.2699, "step": 26620 }, { "epoch": 0.67, "grad_norm": 24.5, "learning_rate": 2.284745762711865e-06, "loss": 1.1961, "step": 26630 }, { "epoch": 0.67, "grad_norm": 38.5, "learning_rate": 2.2779661016949154e-06, "loss": 1.3764, "step": 26640 }, { "epoch": 0.67, "grad_norm": 35.25, "learning_rate": 2.2711864406779663e-06, "loss": 1.2731, "step": 26650 }, { "epoch": 0.67, "grad_norm": 20.875, "learning_rate": 2.2644067796610173e-06, "loss": 1.2816, "step": 26660 }, { "epoch": 0.67, "grad_norm": 46.75, "learning_rate": 2.257627118644068e-06, "loss": 1.2159, "step": 26670 }, { "epoch": 0.67, "grad_norm": 9.25, "learning_rate": 2.250847457627119e-06, "loss": 1.371, "step": 26680 }, { "epoch": 0.67, "grad_norm": 25.625, "learning_rate": 2.2440677966101694e-06, "loss": 1.5302, "step": 26690 }, { "epoch": 0.67, "grad_norm": 15.0, "learning_rate": 2.2372881355932204e-06, "loss": 1.3585, "step": 26700 }, { "epoch": 0.67, "grad_norm": 16.5, "learning_rate": 2.2305084745762714e-06, "loss": 1.201, "step": 26710 }, { "epoch": 0.67, "grad_norm": 27.25, "learning_rate": 2.2237288135593223e-06, "loss": 1.3247, "step": 26720 }, { "epoch": 0.67, "grad_norm": 17.75, "learning_rate": 2.216949152542373e-06, "loss": 1.4727, "step": 26730 }, { "epoch": 0.67, "grad_norm": 22.5, "learning_rate": 2.210169491525424e-06, "loss": 1.3304, "step": 26740 }, { "epoch": 0.67, "grad_norm": 25.125, "learning_rate": 2.203389830508475e-06, "loss": 1.3933, "step": 26750 }, { "epoch": 0.67, "grad_norm": 19.75, "learning_rate": 2.196610169491526e-06, "loss": 1.3612, "step": 26760 }, { "epoch": 0.67, "grad_norm": 30.0, "learning_rate": 2.1898305084745764e-06, "loss": 1.3657, "step": 26770 }, { "epoch": 0.67, "grad_norm": 43.5, "learning_rate": 2.1830508474576273e-06, "loss": 1.4132, "step": 26780 }, { "epoch": 0.67, "grad_norm": 37.5, "learning_rate": 2.1762711864406783e-06, "loss": 1.5674, "step": 26790 }, { "epoch": 0.67, "grad_norm": 24.625, "learning_rate": 2.169491525423729e-06, "loss": 1.2488, "step": 26800 }, { "epoch": 0.67, "grad_norm": 27.625, "learning_rate": 2.16271186440678e-06, "loss": 1.3611, "step": 26810 }, { "epoch": 0.67, "grad_norm": 35.75, "learning_rate": 2.1559322033898304e-06, "loss": 1.4191, "step": 26820 }, { "epoch": 0.67, "grad_norm": 19.625, "learning_rate": 2.1491525423728814e-06, "loss": 1.3845, "step": 26830 }, { "epoch": 0.67, "grad_norm": 20.0, "learning_rate": 2.1423728813559324e-06, "loss": 1.2857, "step": 26840 }, { "epoch": 0.67, "grad_norm": 13.875, "learning_rate": 2.1355932203389833e-06, "loss": 1.4068, "step": 26850 }, { "epoch": 0.67, "grad_norm": 20.625, "learning_rate": 2.128813559322034e-06, "loss": 1.5439, "step": 26860 }, { "epoch": 0.67, "grad_norm": 42.5, "learning_rate": 2.122033898305085e-06, "loss": 1.2449, "step": 26870 }, { "epoch": 0.67, "grad_norm": 32.75, "learning_rate": 2.115254237288136e-06, "loss": 1.2669, "step": 26880 }, { "epoch": 0.67, "grad_norm": 11.625, "learning_rate": 2.108474576271187e-06, "loss": 1.2574, "step": 26890 }, { "epoch": 0.67, "grad_norm": 20.25, "learning_rate": 2.1016949152542374e-06, "loss": 1.4733, "step": 26900 }, { "epoch": 0.67, "grad_norm": 41.75, "learning_rate": 2.0949152542372883e-06, "loss": 1.1929, "step": 26910 }, { "epoch": 0.67, "grad_norm": 14.625, "learning_rate": 2.0881355932203393e-06, "loss": 1.0785, "step": 26920 }, { "epoch": 0.67, "grad_norm": 14.6875, "learning_rate": 2.08135593220339e-06, "loss": 1.6247, "step": 26930 }, { "epoch": 0.67, "grad_norm": 36.5, "learning_rate": 2.074576271186441e-06, "loss": 1.451, "step": 26940 }, { "epoch": 0.67, "grad_norm": 31.5, "learning_rate": 2.0677966101694914e-06, "loss": 1.2365, "step": 26950 }, { "epoch": 0.67, "grad_norm": 16.125, "learning_rate": 2.0610169491525424e-06, "loss": 1.4036, "step": 26960 }, { "epoch": 0.67, "grad_norm": 47.75, "learning_rate": 2.0542372881355934e-06, "loss": 1.402, "step": 26970 }, { "epoch": 0.67, "grad_norm": 31.25, "learning_rate": 2.0474576271186443e-06, "loss": 1.3498, "step": 26980 }, { "epoch": 0.67, "grad_norm": 23.0, "learning_rate": 2.0406779661016953e-06, "loss": 1.2872, "step": 26990 }, { "epoch": 0.68, "grad_norm": 26.625, "learning_rate": 2.033898305084746e-06, "loss": 1.4502, "step": 27000 }, { "epoch": 0.68, "grad_norm": 14.4375, "learning_rate": 2.027118644067797e-06, "loss": 1.3091, "step": 27010 }, { "epoch": 0.68, "grad_norm": 22.875, "learning_rate": 2.020338983050848e-06, "loss": 1.3883, "step": 27020 }, { "epoch": 0.68, "grad_norm": 16.0, "learning_rate": 2.0135593220338984e-06, "loss": 1.3933, "step": 27030 }, { "epoch": 0.68, "grad_norm": 14.3125, "learning_rate": 2.0067796610169494e-06, "loss": 1.4509, "step": 27040 }, { "epoch": 0.68, "grad_norm": 12.5625, "learning_rate": 2.0000000000000003e-06, "loss": 1.2744, "step": 27050 }, { "epoch": 0.68, "grad_norm": 43.5, "learning_rate": 1.993220338983051e-06, "loss": 1.2796, "step": 27060 }, { "epoch": 0.68, "grad_norm": 25.875, "learning_rate": 1.986440677966102e-06, "loss": 1.3602, "step": 27070 }, { "epoch": 0.68, "grad_norm": 20.5, "learning_rate": 1.9796610169491524e-06, "loss": 1.312, "step": 27080 }, { "epoch": 0.68, "grad_norm": 15.75, "learning_rate": 1.9728813559322034e-06, "loss": 1.2828, "step": 27090 }, { "epoch": 0.68, "grad_norm": 43.0, "learning_rate": 1.9661016949152544e-06, "loss": 1.448, "step": 27100 }, { "epoch": 0.68, "grad_norm": 7.28125, "learning_rate": 1.9593220338983053e-06, "loss": 1.3548, "step": 27110 }, { "epoch": 0.68, "grad_norm": 20.5, "learning_rate": 1.9525423728813563e-06, "loss": 1.4944, "step": 27120 }, { "epoch": 0.68, "grad_norm": 59.25, "learning_rate": 1.945762711864407e-06, "loss": 1.4544, "step": 27130 }, { "epoch": 0.68, "grad_norm": 13.25, "learning_rate": 1.938983050847458e-06, "loss": 1.4506, "step": 27140 }, { "epoch": 0.68, "grad_norm": 35.5, "learning_rate": 1.932203389830509e-06, "loss": 1.3822, "step": 27150 }, { "epoch": 0.68, "grad_norm": 23.125, "learning_rate": 1.9254237288135594e-06, "loss": 1.3075, "step": 27160 }, { "epoch": 0.68, "grad_norm": 27.125, "learning_rate": 1.9186440677966104e-06, "loss": 1.2924, "step": 27170 }, { "epoch": 0.68, "grad_norm": 12.5625, "learning_rate": 1.9118644067796613e-06, "loss": 1.3947, "step": 27180 }, { "epoch": 0.68, "grad_norm": 40.75, "learning_rate": 1.9050847457627119e-06, "loss": 1.4555, "step": 27190 }, { "epoch": 0.68, "grad_norm": 8.9375, "learning_rate": 1.8983050847457629e-06, "loss": 1.3187, "step": 27200 }, { "epoch": 0.68, "grad_norm": 33.5, "learning_rate": 1.8915254237288136e-06, "loss": 1.2239, "step": 27210 }, { "epoch": 0.68, "grad_norm": 43.5, "learning_rate": 1.8847457627118646e-06, "loss": 1.4288, "step": 27220 }, { "epoch": 0.68, "grad_norm": 43.5, "learning_rate": 1.8779661016949156e-06, "loss": 1.2993, "step": 27230 }, { "epoch": 0.68, "grad_norm": 38.75, "learning_rate": 1.8711864406779661e-06, "loss": 1.4127, "step": 27240 }, { "epoch": 0.68, "grad_norm": 17.75, "learning_rate": 1.8644067796610171e-06, "loss": 1.3089, "step": 27250 }, { "epoch": 0.68, "grad_norm": 41.25, "learning_rate": 1.857627118644068e-06, "loss": 1.3116, "step": 27260 }, { "epoch": 0.68, "grad_norm": 15.5, "learning_rate": 1.8508474576271189e-06, "loss": 1.4069, "step": 27270 }, { "epoch": 0.68, "grad_norm": 8.375, "learning_rate": 1.8440677966101696e-06, "loss": 1.4496, "step": 27280 }, { "epoch": 0.68, "grad_norm": 33.0, "learning_rate": 1.8372881355932204e-06, "loss": 1.3556, "step": 27290 }, { "epoch": 0.68, "grad_norm": 17.375, "learning_rate": 1.8305084745762714e-06, "loss": 1.4954, "step": 27300 }, { "epoch": 0.68, "grad_norm": 20.25, "learning_rate": 1.8237288135593223e-06, "loss": 1.439, "step": 27310 }, { "epoch": 0.68, "grad_norm": 34.25, "learning_rate": 1.816949152542373e-06, "loss": 1.1777, "step": 27320 }, { "epoch": 0.68, "grad_norm": 27.875, "learning_rate": 1.8101694915254239e-06, "loss": 1.4065, "step": 27330 }, { "epoch": 0.68, "grad_norm": 14.25, "learning_rate": 1.8033898305084746e-06, "loss": 1.2855, "step": 27340 }, { "epoch": 0.68, "grad_norm": 13.625, "learning_rate": 1.7966101694915256e-06, "loss": 1.2503, "step": 27350 }, { "epoch": 0.68, "grad_norm": 16.625, "learning_rate": 1.7898305084745766e-06, "loss": 1.3458, "step": 27360 }, { "epoch": 0.68, "grad_norm": 11.6875, "learning_rate": 1.7830508474576271e-06, "loss": 1.2963, "step": 27370 }, { "epoch": 0.68, "grad_norm": 23.625, "learning_rate": 1.7762711864406781e-06, "loss": 1.4009, "step": 27380 }, { "epoch": 0.68, "grad_norm": 33.75, "learning_rate": 1.769491525423729e-06, "loss": 1.444, "step": 27390 }, { "epoch": 0.69, "grad_norm": 16.25, "learning_rate": 1.7627118644067799e-06, "loss": 1.4343, "step": 27400 }, { "epoch": 0.69, "grad_norm": 16.0, "learning_rate": 1.7559322033898306e-06, "loss": 1.2348, "step": 27410 }, { "epoch": 0.69, "grad_norm": 21.25, "learning_rate": 1.7491525423728814e-06, "loss": 1.4386, "step": 27420 }, { "epoch": 0.69, "grad_norm": 20.625, "learning_rate": 1.7423728813559324e-06, "loss": 1.3253, "step": 27430 }, { "epoch": 0.69, "grad_norm": 19.875, "learning_rate": 1.7355932203389834e-06, "loss": 1.3057, "step": 27440 }, { "epoch": 0.69, "grad_norm": 35.75, "learning_rate": 1.728813559322034e-06, "loss": 1.2917, "step": 27450 }, { "epoch": 0.69, "grad_norm": 29.625, "learning_rate": 1.7220338983050849e-06, "loss": 1.4069, "step": 27460 }, { "epoch": 0.69, "grad_norm": 72.5, "learning_rate": 1.7152542372881356e-06, "loss": 1.274, "step": 27470 }, { "epoch": 0.69, "grad_norm": 13.25, "learning_rate": 1.7084745762711866e-06, "loss": 1.4452, "step": 27480 }, { "epoch": 0.69, "grad_norm": 21.75, "learning_rate": 1.7016949152542376e-06, "loss": 1.3647, "step": 27490 }, { "epoch": 0.69, "grad_norm": 35.0, "learning_rate": 1.6949152542372882e-06, "loss": 1.3659, "step": 27500 }, { "epoch": 0.69, "grad_norm": 20.125, "learning_rate": 1.6881355932203391e-06, "loss": 1.3932, "step": 27510 }, { "epoch": 0.69, "grad_norm": 12.875, "learning_rate": 1.6813559322033901e-06, "loss": 1.2859, "step": 27520 }, { "epoch": 0.69, "grad_norm": 13.875, "learning_rate": 1.6745762711864409e-06, "loss": 1.48, "step": 27530 }, { "epoch": 0.69, "grad_norm": 27.5, "learning_rate": 1.6677966101694916e-06, "loss": 1.4621, "step": 27540 }, { "epoch": 0.69, "grad_norm": 12.25, "learning_rate": 1.6610169491525424e-06, "loss": 1.3061, "step": 27550 }, { "epoch": 0.69, "grad_norm": 26.625, "learning_rate": 1.6542372881355934e-06, "loss": 1.272, "step": 27560 }, { "epoch": 0.69, "grad_norm": 15.8125, "learning_rate": 1.6474576271186444e-06, "loss": 1.3301, "step": 27570 }, { "epoch": 0.69, "grad_norm": 11.625, "learning_rate": 1.640677966101695e-06, "loss": 1.5456, "step": 27580 }, { "epoch": 0.69, "grad_norm": 17.5, "learning_rate": 1.6338983050847459e-06, "loss": 1.3963, "step": 27590 }, { "epoch": 0.69, "grad_norm": 34.5, "learning_rate": 1.6271186440677967e-06, "loss": 1.2749, "step": 27600 }, { "epoch": 0.69, "grad_norm": 21.375, "learning_rate": 1.6203389830508476e-06, "loss": 1.3479, "step": 27610 }, { "epoch": 0.69, "grad_norm": 21.375, "learning_rate": 1.6135593220338986e-06, "loss": 1.3659, "step": 27620 }, { "epoch": 0.69, "grad_norm": 20.5, "learning_rate": 1.6067796610169492e-06, "loss": 1.4773, "step": 27630 }, { "epoch": 0.69, "grad_norm": 27.25, "learning_rate": 1.6000000000000001e-06, "loss": 1.3929, "step": 27640 }, { "epoch": 0.69, "grad_norm": 20.875, "learning_rate": 1.593220338983051e-06, "loss": 1.2243, "step": 27650 }, { "epoch": 0.69, "grad_norm": 24.125, "learning_rate": 1.5864406779661019e-06, "loss": 1.3985, "step": 27660 }, { "epoch": 0.69, "grad_norm": 35.75, "learning_rate": 1.5796610169491526e-06, "loss": 1.3317, "step": 27670 }, { "epoch": 0.69, "grad_norm": 29.125, "learning_rate": 1.5728813559322034e-06, "loss": 1.3862, "step": 27680 }, { "epoch": 0.69, "grad_norm": 29.875, "learning_rate": 1.5661016949152544e-06, "loss": 1.2297, "step": 27690 }, { "epoch": 0.69, "grad_norm": 25.625, "learning_rate": 1.5593220338983054e-06, "loss": 1.2997, "step": 27700 }, { "epoch": 0.69, "grad_norm": 27.5, "learning_rate": 1.552542372881356e-06, "loss": 1.2711, "step": 27710 }, { "epoch": 0.69, "grad_norm": 12.9375, "learning_rate": 1.545762711864407e-06, "loss": 1.3008, "step": 27720 }, { "epoch": 0.69, "grad_norm": 27.875, "learning_rate": 1.5389830508474577e-06, "loss": 1.2319, "step": 27730 }, { "epoch": 0.69, "grad_norm": 53.75, "learning_rate": 1.5322033898305086e-06, "loss": 1.323, "step": 27740 }, { "epoch": 0.69, "grad_norm": 24.875, "learning_rate": 1.5254237288135596e-06, "loss": 1.4262, "step": 27750 }, { "epoch": 0.69, "grad_norm": 23.5, "learning_rate": 1.5186440677966102e-06, "loss": 1.2772, "step": 27760 }, { "epoch": 0.69, "grad_norm": 23.625, "learning_rate": 1.5118644067796611e-06, "loss": 1.3022, "step": 27770 }, { "epoch": 0.69, "grad_norm": 53.25, "learning_rate": 1.505084745762712e-06, "loss": 1.3384, "step": 27780 }, { "epoch": 0.69, "grad_norm": 34.0, "learning_rate": 1.4983050847457629e-06, "loss": 1.335, "step": 27790 }, { "epoch": 0.69, "grad_norm": 20.75, "learning_rate": 1.4915254237288139e-06, "loss": 1.4021, "step": 27800 }, { "epoch": 0.7, "grad_norm": 23.0, "learning_rate": 1.4847457627118644e-06, "loss": 1.417, "step": 27810 }, { "epoch": 0.7, "grad_norm": 39.0, "learning_rate": 1.4779661016949154e-06, "loss": 1.3886, "step": 27820 }, { "epoch": 0.7, "grad_norm": 30.75, "learning_rate": 1.4711864406779664e-06, "loss": 1.3769, "step": 27830 }, { "epoch": 0.7, "grad_norm": 27.75, "learning_rate": 1.464406779661017e-06, "loss": 1.2846, "step": 27840 }, { "epoch": 0.7, "grad_norm": 42.5, "learning_rate": 1.457627118644068e-06, "loss": 1.4831, "step": 27850 }, { "epoch": 0.7, "grad_norm": 24.25, "learning_rate": 1.4508474576271187e-06, "loss": 1.3099, "step": 27860 }, { "epoch": 0.7, "grad_norm": 27.0, "learning_rate": 1.4440677966101696e-06, "loss": 1.3877, "step": 27870 }, { "epoch": 0.7, "grad_norm": 15.75, "learning_rate": 1.4372881355932206e-06, "loss": 1.4706, "step": 27880 }, { "epoch": 0.7, "grad_norm": 17.875, "learning_rate": 1.4305084745762712e-06, "loss": 1.2598, "step": 27890 }, { "epoch": 0.7, "grad_norm": 14.5625, "learning_rate": 1.4237288135593222e-06, "loss": 1.3585, "step": 27900 }, { "epoch": 0.7, "grad_norm": 29.125, "learning_rate": 1.416949152542373e-06, "loss": 1.3198, "step": 27910 }, { "epoch": 0.7, "grad_norm": 43.0, "learning_rate": 1.410169491525424e-06, "loss": 1.3106, "step": 27920 }, { "epoch": 0.7, "grad_norm": 9.25, "learning_rate": 1.4033898305084749e-06, "loss": 1.2443, "step": 27930 }, { "epoch": 0.7, "grad_norm": 9.5, "learning_rate": 1.3966101694915254e-06, "loss": 1.4339, "step": 27940 }, { "epoch": 0.7, "grad_norm": 29.125, "learning_rate": 1.3898305084745764e-06, "loss": 1.3682, "step": 27950 }, { "epoch": 0.7, "grad_norm": 21.875, "learning_rate": 1.3830508474576274e-06, "loss": 1.3385, "step": 27960 }, { "epoch": 0.7, "grad_norm": 33.5, "learning_rate": 1.376271186440678e-06, "loss": 1.2603, "step": 27970 }, { "epoch": 0.7, "grad_norm": 38.25, "learning_rate": 1.369491525423729e-06, "loss": 1.3336, "step": 27980 }, { "epoch": 0.7, "grad_norm": 18.125, "learning_rate": 1.3627118644067797e-06, "loss": 1.291, "step": 27990 }, { "epoch": 0.7, "grad_norm": 72.0, "learning_rate": 1.3559322033898307e-06, "loss": 1.3963, "step": 28000 }, { "epoch": 0.7, "grad_norm": 17.125, "learning_rate": 1.3491525423728816e-06, "loss": 1.2018, "step": 28010 }, { "epoch": 0.7, "grad_norm": 18.375, "learning_rate": 1.3423728813559322e-06, "loss": 1.4179, "step": 28020 }, { "epoch": 0.7, "grad_norm": 16.25, "learning_rate": 1.3355932203389832e-06, "loss": 1.1736, "step": 28030 }, { "epoch": 0.7, "grad_norm": 34.0, "learning_rate": 1.328813559322034e-06, "loss": 1.4772, "step": 28040 }, { "epoch": 0.7, "grad_norm": 39.75, "learning_rate": 1.322033898305085e-06, "loss": 1.4616, "step": 28050 }, { "epoch": 0.7, "grad_norm": 8.8125, "learning_rate": 1.3152542372881359e-06, "loss": 1.3562, "step": 28060 }, { "epoch": 0.7, "grad_norm": 11.25, "learning_rate": 1.3084745762711864e-06, "loss": 1.2376, "step": 28070 }, { "epoch": 0.7, "grad_norm": 25.5, "learning_rate": 1.3016949152542374e-06, "loss": 1.4133, "step": 28080 }, { "epoch": 0.7, "grad_norm": 33.5, "learning_rate": 1.2949152542372884e-06, "loss": 1.303, "step": 28090 }, { "epoch": 0.7, "grad_norm": 26.375, "learning_rate": 1.288135593220339e-06, "loss": 1.2879, "step": 28100 }, { "epoch": 0.7, "grad_norm": 15.625, "learning_rate": 1.28135593220339e-06, "loss": 1.3765, "step": 28110 }, { "epoch": 0.7, "grad_norm": 19.375, "learning_rate": 1.2745762711864407e-06, "loss": 1.4118, "step": 28120 }, { "epoch": 0.7, "grad_norm": 46.25, "learning_rate": 1.2677966101694917e-06, "loss": 1.191, "step": 28130 }, { "epoch": 0.7, "grad_norm": 25.25, "learning_rate": 1.2610169491525426e-06, "loss": 1.5409, "step": 28140 }, { "epoch": 0.7, "grad_norm": 24.5, "learning_rate": 1.2542372881355932e-06, "loss": 1.4233, "step": 28150 }, { "epoch": 0.7, "grad_norm": 22.0, "learning_rate": 1.2474576271186442e-06, "loss": 1.2828, "step": 28160 }, { "epoch": 0.7, "grad_norm": 21.5, "learning_rate": 1.240677966101695e-06, "loss": 1.3378, "step": 28170 }, { "epoch": 0.7, "grad_norm": 18.375, "learning_rate": 1.233898305084746e-06, "loss": 1.3357, "step": 28180 }, { "epoch": 0.7, "grad_norm": 42.5, "learning_rate": 1.2271186440677967e-06, "loss": 1.4315, "step": 28190 }, { "epoch": 0.7, "grad_norm": 19.5, "learning_rate": 1.2203389830508477e-06, "loss": 1.392, "step": 28200 }, { "epoch": 0.71, "grad_norm": 18.0, "learning_rate": 1.2135593220338984e-06, "loss": 1.4079, "step": 28210 }, { "epoch": 0.71, "grad_norm": 16.875, "learning_rate": 1.2067796610169492e-06, "loss": 1.2263, "step": 28220 }, { "epoch": 0.71, "grad_norm": 13.8125, "learning_rate": 1.2000000000000002e-06, "loss": 1.2481, "step": 28230 }, { "epoch": 0.71, "grad_norm": 6.84375, "learning_rate": 1.193220338983051e-06, "loss": 1.3511, "step": 28240 }, { "epoch": 0.71, "grad_norm": 24.875, "learning_rate": 1.186440677966102e-06, "loss": 1.3507, "step": 28250 }, { "epoch": 0.71, "grad_norm": 32.25, "learning_rate": 1.1796610169491527e-06, "loss": 1.2826, "step": 28260 }, { "epoch": 0.71, "grad_norm": 15.4375, "learning_rate": 1.1728813559322034e-06, "loss": 1.4907, "step": 28270 }, { "epoch": 0.71, "grad_norm": 12.4375, "learning_rate": 1.1661016949152542e-06, "loss": 1.5181, "step": 28280 }, { "epoch": 0.71, "grad_norm": 27.125, "learning_rate": 1.1593220338983052e-06, "loss": 1.2625, "step": 28290 }, { "epoch": 0.71, "grad_norm": 14.5, "learning_rate": 1.152542372881356e-06, "loss": 1.199, "step": 28300 }, { "epoch": 0.71, "grad_norm": 23.75, "learning_rate": 1.145762711864407e-06, "loss": 1.3387, "step": 28310 }, { "epoch": 0.71, "grad_norm": 8.875, "learning_rate": 1.1389830508474577e-06, "loss": 1.3493, "step": 28320 }, { "epoch": 0.71, "grad_norm": 19.75, "learning_rate": 1.1322033898305087e-06, "loss": 1.3027, "step": 28330 }, { "epoch": 0.71, "grad_norm": 51.5, "learning_rate": 1.1254237288135594e-06, "loss": 1.3349, "step": 28340 }, { "epoch": 0.71, "grad_norm": 78.5, "learning_rate": 1.1186440677966102e-06, "loss": 1.2642, "step": 28350 }, { "epoch": 0.71, "grad_norm": 37.25, "learning_rate": 1.1118644067796612e-06, "loss": 1.3023, "step": 28360 }, { "epoch": 0.71, "grad_norm": 51.0, "learning_rate": 1.105084745762712e-06, "loss": 1.3719, "step": 28370 }, { "epoch": 0.71, "grad_norm": 16.5, "learning_rate": 1.098305084745763e-06, "loss": 1.2886, "step": 28380 }, { "epoch": 0.71, "grad_norm": 31.25, "learning_rate": 1.0915254237288137e-06, "loss": 1.2252, "step": 28390 }, { "epoch": 0.71, "grad_norm": 26.875, "learning_rate": 1.0847457627118644e-06, "loss": 1.2735, "step": 28400 }, { "epoch": 0.71, "grad_norm": 17.75, "learning_rate": 1.0779661016949152e-06, "loss": 1.4295, "step": 28410 }, { "epoch": 0.71, "grad_norm": 22.25, "learning_rate": 1.0711864406779662e-06, "loss": 1.1851, "step": 28420 }, { "epoch": 0.71, "grad_norm": 10.4375, "learning_rate": 1.064406779661017e-06, "loss": 1.3764, "step": 28430 }, { "epoch": 0.71, "grad_norm": 28.875, "learning_rate": 1.057627118644068e-06, "loss": 1.2924, "step": 28440 }, { "epoch": 0.71, "grad_norm": 23.375, "learning_rate": 1.0508474576271187e-06, "loss": 1.4333, "step": 28450 }, { "epoch": 0.71, "grad_norm": 26.5, "learning_rate": 1.0440677966101697e-06, "loss": 1.3717, "step": 28460 }, { "epoch": 0.71, "grad_norm": 44.75, "learning_rate": 1.0372881355932204e-06, "loss": 1.2711, "step": 28470 }, { "epoch": 0.71, "grad_norm": 16.5, "learning_rate": 1.0305084745762712e-06, "loss": 1.3858, "step": 28480 }, { "epoch": 0.71, "grad_norm": 28.875, "learning_rate": 1.0237288135593222e-06, "loss": 1.4771, "step": 28490 }, { "epoch": 0.71, "grad_norm": 15.375, "learning_rate": 1.016949152542373e-06, "loss": 1.3463, "step": 28500 }, { "epoch": 0.71, "grad_norm": 57.75, "learning_rate": 1.010169491525424e-06, "loss": 1.292, "step": 28510 }, { "epoch": 0.71, "grad_norm": 21.0, "learning_rate": 1.0033898305084747e-06, "loss": 1.2532, "step": 28520 }, { "epoch": 0.71, "grad_norm": 25.0, "learning_rate": 9.966101694915254e-07, "loss": 1.4307, "step": 28530 }, { "epoch": 0.71, "grad_norm": 21.375, "learning_rate": 9.898305084745762e-07, "loss": 1.5372, "step": 28540 }, { "epoch": 0.71, "grad_norm": 24.875, "learning_rate": 9.830508474576272e-07, "loss": 1.5059, "step": 28550 }, { "epoch": 0.71, "grad_norm": 62.5, "learning_rate": 9.762711864406782e-07, "loss": 1.262, "step": 28560 }, { "epoch": 0.71, "grad_norm": 29.125, "learning_rate": 9.69491525423729e-07, "loss": 1.2998, "step": 28570 }, { "epoch": 0.71, "grad_norm": 25.875, "learning_rate": 9.627118644067797e-07, "loss": 1.3971, "step": 28580 }, { "epoch": 0.71, "grad_norm": 33.25, "learning_rate": 9.559322033898307e-07, "loss": 1.4679, "step": 28590 }, { "epoch": 0.71, "grad_norm": 12.125, "learning_rate": 9.491525423728814e-07, "loss": 1.3046, "step": 28600 }, { "epoch": 0.72, "grad_norm": 37.25, "learning_rate": 9.423728813559323e-07, "loss": 1.3883, "step": 28610 }, { "epoch": 0.72, "grad_norm": 62.75, "learning_rate": 9.355932203389831e-07, "loss": 1.1382, "step": 28620 }, { "epoch": 0.72, "grad_norm": 27.375, "learning_rate": 9.28813559322034e-07, "loss": 1.361, "step": 28630 }, { "epoch": 0.72, "grad_norm": 36.75, "learning_rate": 9.220338983050848e-07, "loss": 1.3811, "step": 28640 }, { "epoch": 0.72, "grad_norm": 10.1875, "learning_rate": 9.152542372881357e-07, "loss": 1.3962, "step": 28650 }, { "epoch": 0.72, "grad_norm": 38.75, "learning_rate": 9.084745762711864e-07, "loss": 1.3429, "step": 28660 }, { "epoch": 0.72, "grad_norm": 19.625, "learning_rate": 9.016949152542373e-07, "loss": 1.2986, "step": 28670 }, { "epoch": 0.72, "grad_norm": 33.75, "learning_rate": 8.949152542372883e-07, "loss": 1.3135, "step": 28680 }, { "epoch": 0.72, "grad_norm": 24.75, "learning_rate": 8.881355932203391e-07, "loss": 1.4679, "step": 28690 }, { "epoch": 0.72, "grad_norm": 7.21875, "learning_rate": 8.813559322033899e-07, "loss": 1.4693, "step": 28700 }, { "epoch": 0.72, "grad_norm": 19.125, "learning_rate": 8.745762711864407e-07, "loss": 1.3307, "step": 28710 }, { "epoch": 0.72, "grad_norm": 17.0, "learning_rate": 8.677966101694917e-07, "loss": 1.4457, "step": 28720 }, { "epoch": 0.72, "grad_norm": 5.75, "learning_rate": 8.610169491525424e-07, "loss": 1.282, "step": 28730 }, { "epoch": 0.72, "grad_norm": 18.625, "learning_rate": 8.542372881355933e-07, "loss": 1.3471, "step": 28740 }, { "epoch": 0.72, "grad_norm": 24.375, "learning_rate": 8.474576271186441e-07, "loss": 1.2901, "step": 28750 }, { "epoch": 0.72, "grad_norm": 13.9375, "learning_rate": 8.406779661016951e-07, "loss": 1.2879, "step": 28760 }, { "epoch": 0.72, "grad_norm": 40.5, "learning_rate": 8.338983050847458e-07, "loss": 1.4003, "step": 28770 }, { "epoch": 0.72, "grad_norm": 40.5, "learning_rate": 8.271186440677967e-07, "loss": 1.3961, "step": 28780 }, { "epoch": 0.72, "grad_norm": 36.5, "learning_rate": 8.203389830508475e-07, "loss": 1.3868, "step": 28790 }, { "epoch": 0.72, "grad_norm": 9.875, "learning_rate": 8.135593220338983e-07, "loss": 1.4772, "step": 28800 }, { "epoch": 0.72, "grad_norm": 24.0, "learning_rate": 8.067796610169493e-07, "loss": 1.3712, "step": 28810 }, { "epoch": 0.72, "grad_norm": 12.9375, "learning_rate": 8.000000000000001e-07, "loss": 1.3414, "step": 28820 }, { "epoch": 0.72, "grad_norm": 9.75, "learning_rate": 7.932203389830509e-07, "loss": 1.2895, "step": 28830 }, { "epoch": 0.72, "grad_norm": 41.5, "learning_rate": 7.864406779661017e-07, "loss": 1.438, "step": 28840 }, { "epoch": 0.72, "grad_norm": 14.5, "learning_rate": 7.796610169491527e-07, "loss": 1.289, "step": 28850 }, { "epoch": 0.72, "grad_norm": 20.125, "learning_rate": 7.728813559322034e-07, "loss": 1.4131, "step": 28860 }, { "epoch": 0.72, "grad_norm": 31.25, "learning_rate": 7.661016949152543e-07, "loss": 1.2683, "step": 28870 }, { "epoch": 0.72, "grad_norm": 18.75, "learning_rate": 7.593220338983051e-07, "loss": 1.2794, "step": 28880 }, { "epoch": 0.72, "grad_norm": 31.375, "learning_rate": 7.52542372881356e-07, "loss": 1.36, "step": 28890 }, { "epoch": 0.72, "grad_norm": 18.5, "learning_rate": 7.457627118644069e-07, "loss": 1.25, "step": 28900 }, { "epoch": 0.72, "grad_norm": 33.5, "learning_rate": 7.389830508474577e-07, "loss": 1.2343, "step": 28910 }, { "epoch": 0.72, "grad_norm": 39.75, "learning_rate": 7.322033898305085e-07, "loss": 1.3837, "step": 28920 }, { "epoch": 0.72, "grad_norm": 21.125, "learning_rate": 7.254237288135593e-07, "loss": 1.4617, "step": 28930 }, { "epoch": 0.72, "grad_norm": 31.875, "learning_rate": 7.186440677966103e-07, "loss": 1.3333, "step": 28940 }, { "epoch": 0.72, "grad_norm": 19.875, "learning_rate": 7.118644067796611e-07, "loss": 1.3642, "step": 28950 }, { "epoch": 0.72, "grad_norm": 11.4375, "learning_rate": 7.05084745762712e-07, "loss": 1.3345, "step": 28960 }, { "epoch": 0.72, "grad_norm": 43.25, "learning_rate": 6.983050847457627e-07, "loss": 1.2499, "step": 28970 }, { "epoch": 0.72, "grad_norm": 10.9375, "learning_rate": 6.915254237288137e-07, "loss": 1.3462, "step": 28980 }, { "epoch": 0.72, "grad_norm": 21.25, "learning_rate": 6.847457627118645e-07, "loss": 1.3835, "step": 28990 }, { "epoch": 0.72, "grad_norm": 17.0, "learning_rate": 6.779661016949153e-07, "loss": 1.3421, "step": 29000 }, { "epoch": 0.73, "grad_norm": 29.875, "learning_rate": 6.711864406779661e-07, "loss": 1.3094, "step": 29010 }, { "epoch": 0.73, "grad_norm": 15.8125, "learning_rate": 6.64406779661017e-07, "loss": 1.3585, "step": 29020 }, { "epoch": 0.73, "grad_norm": 13.8125, "learning_rate": 6.576271186440679e-07, "loss": 1.4441, "step": 29030 }, { "epoch": 0.73, "grad_norm": 28.25, "learning_rate": 6.508474576271187e-07, "loss": 1.3683, "step": 29040 }, { "epoch": 0.73, "grad_norm": 22.75, "learning_rate": 6.440677966101695e-07, "loss": 1.2842, "step": 29050 }, { "epoch": 0.73, "grad_norm": 25.875, "learning_rate": 6.372881355932203e-07, "loss": 1.3392, "step": 29060 }, { "epoch": 0.73, "grad_norm": 15.75, "learning_rate": 6.305084745762713e-07, "loss": 1.3556, "step": 29070 }, { "epoch": 0.73, "grad_norm": 16.75, "learning_rate": 6.237288135593221e-07, "loss": 1.221, "step": 29080 }, { "epoch": 0.73, "grad_norm": 18.375, "learning_rate": 6.16949152542373e-07, "loss": 1.4398, "step": 29090 }, { "epoch": 0.73, "grad_norm": 29.625, "learning_rate": 6.101694915254238e-07, "loss": 1.5108, "step": 29100 }, { "epoch": 0.73, "grad_norm": 41.5, "learning_rate": 6.033898305084746e-07, "loss": 1.0557, "step": 29110 }, { "epoch": 0.73, "grad_norm": 21.75, "learning_rate": 5.966101694915255e-07, "loss": 1.2235, "step": 29120 }, { "epoch": 0.73, "grad_norm": 18.875, "learning_rate": 5.898305084745763e-07, "loss": 1.3943, "step": 29130 }, { "epoch": 0.73, "grad_norm": 12.9375, "learning_rate": 5.830508474576271e-07, "loss": 1.5561, "step": 29140 }, { "epoch": 0.73, "grad_norm": 36.5, "learning_rate": 5.76271186440678e-07, "loss": 1.3348, "step": 29150 }, { "epoch": 0.73, "grad_norm": 31.0, "learning_rate": 5.694915254237288e-07, "loss": 1.4021, "step": 29160 }, { "epoch": 0.73, "grad_norm": 23.0, "learning_rate": 5.627118644067797e-07, "loss": 1.2516, "step": 29170 }, { "epoch": 0.73, "grad_norm": 53.5, "learning_rate": 5.559322033898306e-07, "loss": 1.3736, "step": 29180 }, { "epoch": 0.73, "grad_norm": 21.375, "learning_rate": 5.491525423728815e-07, "loss": 1.3918, "step": 29190 }, { "epoch": 0.73, "grad_norm": 59.25, "learning_rate": 5.423728813559322e-07, "loss": 1.2499, "step": 29200 }, { "epoch": 0.73, "grad_norm": 33.5, "learning_rate": 5.355932203389831e-07, "loss": 1.4177, "step": 29210 }, { "epoch": 0.73, "grad_norm": 26.25, "learning_rate": 5.28813559322034e-07, "loss": 1.2788, "step": 29220 }, { "epoch": 0.73, "grad_norm": 27.0, "learning_rate": 5.220338983050848e-07, "loss": 1.4001, "step": 29230 }, { "epoch": 0.73, "grad_norm": 24.75, "learning_rate": 5.152542372881356e-07, "loss": 1.3464, "step": 29240 }, { "epoch": 0.73, "grad_norm": 13.0, "learning_rate": 5.084745762711865e-07, "loss": 1.2711, "step": 29250 }, { "epoch": 0.73, "grad_norm": 8.875, "learning_rate": 5.016949152542373e-07, "loss": 1.3788, "step": 29260 }, { "epoch": 0.73, "grad_norm": 29.25, "learning_rate": 4.949152542372881e-07, "loss": 1.3081, "step": 29270 }, { "epoch": 0.73, "grad_norm": 65.0, "learning_rate": 4.881355932203391e-07, "loss": 1.1399, "step": 29280 }, { "epoch": 0.73, "grad_norm": 65.5, "learning_rate": 4.813559322033898e-07, "loss": 1.1244, "step": 29290 }, { "epoch": 0.73, "grad_norm": 14.8125, "learning_rate": 4.745762711864407e-07, "loss": 1.2739, "step": 29300 }, { "epoch": 0.73, "grad_norm": 18.125, "learning_rate": 4.6779661016949154e-07, "loss": 1.3434, "step": 29310 }, { "epoch": 0.73, "grad_norm": 47.0, "learning_rate": 4.610169491525424e-07, "loss": 1.3668, "step": 29320 }, { "epoch": 0.73, "grad_norm": 16.875, "learning_rate": 4.542372881355932e-07, "loss": 1.3071, "step": 29330 }, { "epoch": 0.73, "grad_norm": 19.75, "learning_rate": 4.4745762711864415e-07, "loss": 1.4213, "step": 29340 }, { "epoch": 0.73, "grad_norm": 11.625, "learning_rate": 4.4067796610169497e-07, "loss": 1.2687, "step": 29350 }, { "epoch": 0.73, "grad_norm": 32.25, "learning_rate": 4.3389830508474584e-07, "loss": 1.2956, "step": 29360 }, { "epoch": 0.73, "grad_norm": 12.3125, "learning_rate": 4.2711864406779666e-07, "loss": 1.3848, "step": 29370 }, { "epoch": 0.73, "grad_norm": 40.5, "learning_rate": 4.2033898305084753e-07, "loss": 1.3565, "step": 29380 }, { "epoch": 0.73, "grad_norm": 52.5, "learning_rate": 4.1355932203389835e-07, "loss": 1.407, "step": 29390 }, { "epoch": 0.73, "grad_norm": 25.25, "learning_rate": 4.0677966101694916e-07, "loss": 1.3811, "step": 29400 }, { "epoch": 0.74, "grad_norm": 11.1875, "learning_rate": 4.0000000000000003e-07, "loss": 1.3421, "step": 29410 }, { "epoch": 0.74, "grad_norm": 14.625, "learning_rate": 3.9322033898305085e-07, "loss": 1.3286, "step": 29420 }, { "epoch": 0.74, "grad_norm": 24.5, "learning_rate": 3.864406779661017e-07, "loss": 1.3828, "step": 29430 }, { "epoch": 0.74, "grad_norm": 24.25, "learning_rate": 3.7966101694915254e-07, "loss": 1.255, "step": 29440 }, { "epoch": 0.74, "grad_norm": 13.8125, "learning_rate": 3.7288135593220347e-07, "loss": 1.2879, "step": 29450 }, { "epoch": 0.74, "grad_norm": 26.375, "learning_rate": 3.6610169491525423e-07, "loss": 1.2621, "step": 29460 }, { "epoch": 0.74, "grad_norm": 10.0, "learning_rate": 3.5932203389830516e-07, "loss": 1.5059, "step": 29470 }, { "epoch": 0.74, "grad_norm": 40.75, "learning_rate": 3.52542372881356e-07, "loss": 1.332, "step": 29480 }, { "epoch": 0.74, "grad_norm": 24.375, "learning_rate": 3.4576271186440684e-07, "loss": 1.3809, "step": 29490 }, { "epoch": 0.74, "grad_norm": 37.0, "learning_rate": 3.3898305084745766e-07, "loss": 1.3389, "step": 29500 }, { "epoch": 0.74, "grad_norm": 30.0, "learning_rate": 3.322033898305085e-07, "loss": 1.2023, "step": 29510 }, { "epoch": 0.74, "grad_norm": 16.625, "learning_rate": 3.2542372881355935e-07, "loss": 1.3672, "step": 29520 }, { "epoch": 0.74, "grad_norm": 28.5, "learning_rate": 3.1864406779661017e-07, "loss": 1.5764, "step": 29530 }, { "epoch": 0.74, "grad_norm": 14.625, "learning_rate": 3.1186440677966104e-07, "loss": 1.5877, "step": 29540 }, { "epoch": 0.74, "grad_norm": 22.875, "learning_rate": 3.050847457627119e-07, "loss": 1.4373, "step": 29550 }, { "epoch": 0.74, "grad_norm": 25.875, "learning_rate": 2.9830508474576273e-07, "loss": 1.3222, "step": 29560 }, { "epoch": 0.74, "grad_norm": 38.5, "learning_rate": 2.9152542372881355e-07, "loss": 1.2743, "step": 29570 }, { "epoch": 0.74, "grad_norm": 46.75, "learning_rate": 2.847457627118644e-07, "loss": 1.3351, "step": 29580 }, { "epoch": 0.74, "grad_norm": 55.75, "learning_rate": 2.779661016949153e-07, "loss": 1.1375, "step": 29590 }, { "epoch": 0.74, "grad_norm": 19.0, "learning_rate": 2.711864406779661e-07, "loss": 1.1009, "step": 29600 }, { "epoch": 0.74, "grad_norm": 9.9375, "learning_rate": 2.64406779661017e-07, "loss": 1.1769, "step": 29610 }, { "epoch": 0.74, "grad_norm": 43.5, "learning_rate": 2.576271186440678e-07, "loss": 1.2881, "step": 29620 }, { "epoch": 0.74, "grad_norm": 23.5, "learning_rate": 2.5084745762711867e-07, "loss": 1.3584, "step": 29630 }, { "epoch": 0.74, "grad_norm": 29.875, "learning_rate": 2.4406779661016954e-07, "loss": 1.4624, "step": 29640 }, { "epoch": 0.74, "grad_norm": 14.25, "learning_rate": 2.3728813559322036e-07, "loss": 1.6026, "step": 29650 }, { "epoch": 0.74, "grad_norm": 71.5, "learning_rate": 2.305084745762712e-07, "loss": 1.1366, "step": 29660 }, { "epoch": 0.74, "grad_norm": 42.0, "learning_rate": 2.2372881355932207e-07, "loss": 1.378, "step": 29670 }, { "epoch": 0.74, "grad_norm": 34.5, "learning_rate": 2.1694915254237292e-07, "loss": 1.2709, "step": 29680 }, { "epoch": 0.74, "grad_norm": 30.25, "learning_rate": 2.1016949152542376e-07, "loss": 1.4227, "step": 29690 }, { "epoch": 0.74, "grad_norm": 23.0, "learning_rate": 2.0338983050847458e-07, "loss": 1.3385, "step": 29700 }, { "epoch": 0.74, "grad_norm": 18.875, "learning_rate": 1.9661016949152543e-07, "loss": 1.0732, "step": 29710 }, { "epoch": 0.74, "grad_norm": 12.5625, "learning_rate": 1.8983050847457627e-07, "loss": 1.4202, "step": 29720 }, { "epoch": 0.74, "grad_norm": 33.75, "learning_rate": 1.8305084745762712e-07, "loss": 1.1954, "step": 29730 }, { "epoch": 0.74, "grad_norm": 31.875, "learning_rate": 1.76271186440678e-07, "loss": 1.2623, "step": 29740 }, { "epoch": 0.74, "grad_norm": 18.375, "learning_rate": 1.6949152542372883e-07, "loss": 1.4002, "step": 29750 }, { "epoch": 0.74, "grad_norm": 37.5, "learning_rate": 1.6271186440677968e-07, "loss": 1.3116, "step": 29760 }, { "epoch": 0.74, "grad_norm": 21.0, "learning_rate": 1.5593220338983052e-07, "loss": 1.4163, "step": 29770 }, { "epoch": 0.74, "grad_norm": 41.75, "learning_rate": 1.4915254237288137e-07, "loss": 1.3354, "step": 29780 }, { "epoch": 0.74, "grad_norm": 48.75, "learning_rate": 1.423728813559322e-07, "loss": 1.36, "step": 29790 }, { "epoch": 0.74, "grad_norm": 15.6875, "learning_rate": 1.3559322033898305e-07, "loss": 1.3334, "step": 29800 }, { "epoch": 0.75, "grad_norm": 60.75, "learning_rate": 1.288135593220339e-07, "loss": 1.3058, "step": 29810 }, { "epoch": 0.75, "grad_norm": 24.625, "learning_rate": 1.2203389830508477e-07, "loss": 1.3963, "step": 29820 }, { "epoch": 0.75, "grad_norm": 15.3125, "learning_rate": 1.152542372881356e-07, "loss": 1.3757, "step": 29830 }, { "epoch": 0.75, "grad_norm": 18.375, "learning_rate": 1.0847457627118646e-07, "loss": 1.4074, "step": 29840 }, { "epoch": 0.75, "grad_norm": 17.0, "learning_rate": 1.0169491525423729e-07, "loss": 1.502, "step": 29850 }, { "epoch": 0.75, "grad_norm": 67.5, "learning_rate": 9.491525423728814e-08, "loss": 1.4699, "step": 29860 }, { "epoch": 0.75, "grad_norm": 15.875, "learning_rate": 8.8135593220339e-08, "loss": 1.3926, "step": 29870 }, { "epoch": 0.75, "grad_norm": 37.75, "learning_rate": 8.135593220338984e-08, "loss": 1.411, "step": 29880 }, { "epoch": 0.75, "grad_norm": 21.375, "learning_rate": 7.457627118644068e-08, "loss": 1.3486, "step": 29890 }, { "epoch": 0.75, "grad_norm": 47.0, "learning_rate": 6.779661016949153e-08, "loss": 1.4179, "step": 29900 }, { "epoch": 0.75, "grad_norm": 17.0, "learning_rate": 6.101694915254239e-08, "loss": 1.2906, "step": 29910 }, { "epoch": 0.75, "grad_norm": 11.75, "learning_rate": 5.423728813559323e-08, "loss": 1.3449, "step": 29920 }, { "epoch": 0.75, "grad_norm": 17.375, "learning_rate": 4.745762711864407e-08, "loss": 1.4177, "step": 29930 }, { "epoch": 0.75, "grad_norm": 18.0, "learning_rate": 4.067796610169492e-08, "loss": 1.4116, "step": 29940 }, { "epoch": 0.75, "grad_norm": 18.0, "learning_rate": 3.3898305084745764e-08, "loss": 1.1822, "step": 29950 }, { "epoch": 0.75, "grad_norm": 16.375, "learning_rate": 2.7118644067796615e-08, "loss": 1.2627, "step": 29960 }, { "epoch": 0.75, "grad_norm": 23.5, "learning_rate": 2.033898305084746e-08, "loss": 1.3393, "step": 29970 }, { "epoch": 0.75, "grad_norm": 8.9375, "learning_rate": 1.3559322033898307e-08, "loss": 1.3589, "step": 29980 }, { "epoch": 0.75, "grad_norm": 36.25, "learning_rate": 6.779661016949154e-09, "loss": 1.3054, "step": 29990 }, { "epoch": 0.75, "grad_norm": 21.25, "learning_rate": 0.0, "loss": 1.3149, "step": 30000 }, { "epoch": 0.75, "eval_loss": 1.3585624694824219, "eval_runtime": 59.2167, "eval_samples_per_second": 16.887, "eval_steps_per_second": 16.887, "step": 30000 } ], "logging_steps": 10, "max_steps": 30000, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 5000, "total_flos": 4.7201094991872e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }