{ "best_metric": null, "best_model_checkpoint": null, "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": 90.0, "learning_rate": 2e-08, "loss": 6.8274, "step": 10 }, { "epoch": 0.0, "grad_norm": 89.5, "learning_rate": 4e-08, "loss": 6.8083, "step": 20 }, { "epoch": 0.0, "grad_norm": 89.5, "learning_rate": 6e-08, "loss": 6.8347, "step": 30 }, { "epoch": 0.0, "grad_norm": 92.0, "learning_rate": 8e-08, "loss": 6.8096, "step": 40 }, { "epoch": 0.0, "grad_norm": 93.0, "learning_rate": 1e-07, "loss": 6.797, "step": 50 }, { "epoch": 0.0, "grad_norm": 91.5, "learning_rate": 1.2e-07, "loss": 6.8509, "step": 60 }, { "epoch": 0.0, "grad_norm": 93.0, "learning_rate": 1.4e-07, "loss": 6.8584, "step": 70 }, { "epoch": 0.0, "grad_norm": 89.0, "learning_rate": 1.6e-07, "loss": 6.8086, "step": 80 }, { "epoch": 0.0, "grad_norm": 95.0, "learning_rate": 1.8e-07, "loss": 6.8356, "step": 90 }, { "epoch": 0.0, "grad_norm": 87.5, "learning_rate": 2e-07, "loss": 6.8266, "step": 100 }, { "epoch": 0.0, "grad_norm": 89.0, "learning_rate": 2.1999999999999998e-07, "loss": 6.8343, "step": 110 }, { "epoch": 0.0, "grad_norm": 88.0, "learning_rate": 2.4e-07, "loss": 6.8105, "step": 120 }, { "epoch": 0.0, "grad_norm": 87.5, "learning_rate": 2.6e-07, "loss": 6.8398, "step": 130 }, { "epoch": 0.0, "grad_norm": 87.0, "learning_rate": 2.8e-07, "loss": 6.8298, "step": 140 }, { "epoch": 0.0, "grad_norm": 93.0, "learning_rate": 3e-07, "loss": 6.7857, "step": 150 }, { "epoch": 0.0, "grad_norm": 90.0, "learning_rate": 3.2e-07, "loss": 6.7761, "step": 160 }, { "epoch": 0.0, "grad_norm": 87.5, "learning_rate": 3.4000000000000003e-07, "loss": 6.7807, "step": 170 }, { "epoch": 0.0, "grad_norm": 90.5, "learning_rate": 3.6e-07, "loss": 6.7772, "step": 180 }, { "epoch": 0.0, "grad_norm": 88.5, "learning_rate": 3.7999999999999996e-07, "loss": 6.7754, "step": 190 }, { "epoch": 0.01, "grad_norm": 95.5, "learning_rate": 4e-07, "loss": 6.7319, "step": 200 }, { "epoch": 0.01, "grad_norm": 96.0, "learning_rate": 4.1999999999999995e-07, "loss": 6.6868, "step": 210 }, { "epoch": 0.01, "grad_norm": 93.0, "learning_rate": 4.3999999999999997e-07, "loss": 6.6913, "step": 220 }, { "epoch": 0.01, "grad_norm": 89.5, "learning_rate": 4.6e-07, "loss": 6.6759, "step": 230 }, { "epoch": 0.01, "grad_norm": 89.0, "learning_rate": 4.8e-07, "loss": 6.6891, "step": 240 }, { "epoch": 0.01, "grad_norm": 87.0, "learning_rate": 5e-07, "loss": 6.6378, "step": 250 }, { "epoch": 0.01, "grad_norm": 103.5, "learning_rate": 5.2e-07, "loss": 6.6767, "step": 260 }, { "epoch": 0.01, "grad_norm": 92.5, "learning_rate": 5.4e-07, "loss": 6.5617, "step": 270 }, { "epoch": 0.01, "grad_norm": 88.5, "learning_rate": 5.6e-07, "loss": 6.5812, "step": 280 }, { "epoch": 0.01, "grad_norm": 80.5, "learning_rate": 5.8e-07, "loss": 6.5442, "step": 290 }, { "epoch": 0.01, "grad_norm": 82.0, "learning_rate": 6e-07, "loss": 6.4611, "step": 300 }, { "epoch": 0.01, "grad_norm": 85.0, "learning_rate": 6.2e-07, "loss": 6.4348, "step": 310 }, { "epoch": 0.01, "grad_norm": 91.0, "learning_rate": 6.4e-07, "loss": 6.4133, "step": 320 }, { "epoch": 0.01, "grad_norm": 91.0, "learning_rate": 6.6e-07, "loss": 6.3978, "step": 330 }, { "epoch": 0.01, "grad_norm": 87.0, "learning_rate": 6.800000000000001e-07, "loss": 6.324, "step": 340 }, { "epoch": 0.01, "grad_norm": 86.5, "learning_rate": 7e-07, "loss": 6.2901, "step": 350 }, { "epoch": 0.01, "grad_norm": 83.0, "learning_rate": 7.2e-07, "loss": 6.243, "step": 360 }, { "epoch": 0.01, "grad_norm": 80.0, "learning_rate": 7.4e-07, "loss": 6.1756, "step": 370 }, { "epoch": 0.01, "grad_norm": 84.5, "learning_rate": 7.599999999999999e-07, "loss": 6.1589, "step": 380 }, { "epoch": 0.01, "grad_norm": 78.5, "learning_rate": 7.799999999999999e-07, "loss": 6.1148, "step": 390 }, { "epoch": 0.01, "grad_norm": 81.5, "learning_rate": 8e-07, "loss": 6.06, "step": 400 }, { "epoch": 0.01, "grad_norm": 87.0, "learning_rate": 8.199999999999999e-07, "loss": 6.0328, "step": 410 }, { "epoch": 0.01, "grad_norm": 84.0, "learning_rate": 8.399999999999999e-07, "loss": 5.9817, "step": 420 }, { "epoch": 0.01, "grad_norm": 85.0, "learning_rate": 8.599999999999999e-07, "loss": 5.9352, "step": 430 }, { "epoch": 0.01, "grad_norm": 87.0, "learning_rate": 8.799999999999999e-07, "loss": 5.8731, "step": 440 }, { "epoch": 0.01, "grad_norm": 82.0, "learning_rate": 9e-07, "loss": 5.8381, "step": 450 }, { "epoch": 0.01, "grad_norm": 85.5, "learning_rate": 9.2e-07, "loss": 5.7847, "step": 460 }, { "epoch": 0.01, "grad_norm": 79.5, "learning_rate": 9.399999999999999e-07, "loss": 5.7146, "step": 470 }, { "epoch": 0.01, "grad_norm": 83.0, "learning_rate": 9.6e-07, "loss": 5.6431, "step": 480 }, { "epoch": 0.01, "grad_norm": 83.0, "learning_rate": 9.8e-07, "loss": 5.5353, "step": 490 }, { "epoch": 0.01, "grad_norm": 78.5, "learning_rate": 1e-06, "loss": 5.4188, "step": 500 }, { "epoch": 0.01, "grad_norm": 80.0, "learning_rate": 9.996610169491525e-07, "loss": 5.3305, "step": 510 }, { "epoch": 0.01, "grad_norm": 77.5, "learning_rate": 9.99322033898305e-07, "loss": 5.1901, "step": 520 }, { "epoch": 0.01, "grad_norm": 79.0, "learning_rate": 9.989830508474575e-07, "loss": 5.1047, "step": 530 }, { "epoch": 0.01, "grad_norm": 78.0, "learning_rate": 9.9864406779661e-07, "loss": 4.9914, "step": 540 }, { "epoch": 0.01, "grad_norm": 79.0, "learning_rate": 9.983050847457626e-07, "loss": 4.9194, "step": 550 }, { "epoch": 0.01, "grad_norm": 81.5, "learning_rate": 9.979661016949153e-07, "loss": 4.799, "step": 560 }, { "epoch": 0.01, "grad_norm": 79.5, "learning_rate": 9.976271186440679e-07, "loss": 4.7051, "step": 570 }, { "epoch": 0.01, "grad_norm": 79.0, "learning_rate": 9.972881355932204e-07, "loss": 4.6324, "step": 580 }, { "epoch": 0.01, "grad_norm": 77.5, "learning_rate": 9.96949152542373e-07, "loss": 4.5426, "step": 590 }, { "epoch": 0.01, "grad_norm": 80.0, "learning_rate": 9.966101694915254e-07, "loss": 4.4262, "step": 600 }, { "epoch": 0.02, "grad_norm": 77.0, "learning_rate": 9.96271186440678e-07, "loss": 4.3657, "step": 610 }, { "epoch": 0.02, "grad_norm": 81.5, "learning_rate": 9.959322033898305e-07, "loss": 4.3496, "step": 620 }, { "epoch": 0.02, "grad_norm": 79.0, "learning_rate": 9.95593220338983e-07, "loss": 4.2415, "step": 630 }, { "epoch": 0.02, "grad_norm": 77.5, "learning_rate": 9.952542372881356e-07, "loss": 4.1497, "step": 640 }, { "epoch": 0.02, "grad_norm": 76.5, "learning_rate": 9.94915254237288e-07, "loss": 4.1079, "step": 650 }, { "epoch": 0.02, "grad_norm": 77.5, "learning_rate": 9.945762711864406e-07, "loss": 4.0541, "step": 660 }, { "epoch": 0.02, "grad_norm": 77.5, "learning_rate": 9.942372881355931e-07, "loss": 3.9799, "step": 670 }, { "epoch": 0.02, "grad_norm": 77.0, "learning_rate": 9.938983050847457e-07, "loss": 3.9074, "step": 680 }, { "epoch": 0.02, "grad_norm": 78.0, "learning_rate": 9.935593220338982e-07, "loss": 3.8532, "step": 690 }, { "epoch": 0.02, "grad_norm": 80.0, "learning_rate": 9.932203389830507e-07, "loss": 3.781, "step": 700 }, { "epoch": 0.02, "grad_norm": 76.5, "learning_rate": 9.928813559322032e-07, "loss": 3.7504, "step": 710 }, { "epoch": 0.02, "grad_norm": 77.5, "learning_rate": 9.92542372881356e-07, "loss": 3.7237, "step": 720 }, { "epoch": 0.02, "grad_norm": 76.0, "learning_rate": 9.922033898305085e-07, "loss": 3.6538, "step": 730 }, { "epoch": 0.02, "grad_norm": 79.0, "learning_rate": 9.91864406779661e-07, "loss": 3.5991, "step": 740 }, { "epoch": 0.02, "grad_norm": 76.0, "learning_rate": 9.915254237288136e-07, "loss": 3.5435, "step": 750 }, { "epoch": 0.02, "grad_norm": 76.5, "learning_rate": 9.91186440677966e-07, "loss": 3.4975, "step": 760 }, { "epoch": 0.02, "grad_norm": 76.5, "learning_rate": 9.908474576271186e-07, "loss": 3.4701, "step": 770 }, { "epoch": 0.02, "grad_norm": 79.0, "learning_rate": 9.905084745762712e-07, "loss": 3.3482, "step": 780 }, { "epoch": 0.02, "grad_norm": 75.0, "learning_rate": 9.901694915254237e-07, "loss": 3.3289, "step": 790 }, { "epoch": 0.02, "grad_norm": 75.0, "learning_rate": 9.898305084745762e-07, "loss": 3.2755, "step": 800 }, { "epoch": 0.02, "grad_norm": 77.5, "learning_rate": 9.894915254237287e-07, "loss": 3.2532, "step": 810 }, { "epoch": 0.02, "grad_norm": 74.5, "learning_rate": 9.891525423728813e-07, "loss": 3.2162, "step": 820 }, { "epoch": 0.02, "grad_norm": 75.0, "learning_rate": 9.888135593220338e-07, "loss": 3.1885, "step": 830 }, { "epoch": 0.02, "grad_norm": 75.0, "learning_rate": 9.884745762711863e-07, "loss": 3.1086, "step": 840 }, { "epoch": 0.02, "grad_norm": 76.0, "learning_rate": 9.881355932203388e-07, "loss": 3.0738, "step": 850 }, { "epoch": 0.02, "grad_norm": 74.5, "learning_rate": 9.877966101694914e-07, "loss": 3.0059, "step": 860 }, { "epoch": 0.02, "grad_norm": 73.5, "learning_rate": 9.87457627118644e-07, "loss": 2.9603, "step": 870 }, { "epoch": 0.02, "grad_norm": 75.0, "learning_rate": 9.871186440677966e-07, "loss": 2.9388, "step": 880 }, { "epoch": 0.02, "grad_norm": 74.5, "learning_rate": 9.867796610169492e-07, "loss": 2.9279, "step": 890 }, { "epoch": 0.02, "grad_norm": 75.5, "learning_rate": 9.864406779661017e-07, "loss": 2.8923, "step": 900 }, { "epoch": 0.02, "grad_norm": 74.5, "learning_rate": 9.861016949152542e-07, "loss": 2.8519, "step": 910 }, { "epoch": 0.02, "grad_norm": 74.5, "learning_rate": 9.857627118644068e-07, "loss": 2.8075, "step": 920 }, { "epoch": 0.02, "grad_norm": 75.5, "learning_rate": 9.854237288135593e-07, "loss": 2.783, "step": 930 }, { "epoch": 0.02, "grad_norm": 74.5, "learning_rate": 9.850847457627118e-07, "loss": 2.7459, "step": 940 }, { "epoch": 0.02, "grad_norm": 74.5, "learning_rate": 9.847457627118643e-07, "loss": 2.7201, "step": 950 }, { "epoch": 0.02, "grad_norm": 74.0, "learning_rate": 9.844067796610169e-07, "loss": 2.6724, "step": 960 }, { "epoch": 0.02, "grad_norm": 75.0, "learning_rate": 9.840677966101694e-07, "loss": 2.6573, "step": 970 }, { "epoch": 0.02, "grad_norm": 74.5, "learning_rate": 9.83728813559322e-07, "loss": 2.6356, "step": 980 }, { "epoch": 0.02, "grad_norm": 76.5, "learning_rate": 9.833898305084747e-07, "loss": 2.5692, "step": 990 }, { "epoch": 0.03, "grad_norm": 72.5, "learning_rate": 9.830508474576272e-07, "loss": 2.5829, "step": 1000 }, { "epoch": 0.03, "grad_norm": 73.5, "learning_rate": 9.827118644067797e-07, "loss": 2.5146, "step": 1010 }, { "epoch": 0.03, "grad_norm": 74.0, "learning_rate": 9.82372881355932e-07, "loss": 2.5395, "step": 1020 }, { "epoch": 0.03, "grad_norm": 74.0, "learning_rate": 9.820338983050846e-07, "loss": 2.4929, "step": 1030 }, { "epoch": 0.03, "grad_norm": 72.5, "learning_rate": 9.816949152542373e-07, "loss": 2.4339, "step": 1040 }, { "epoch": 0.03, "grad_norm": 73.0, "learning_rate": 9.813559322033898e-07, "loss": 2.404, "step": 1050 }, { "epoch": 0.03, "grad_norm": 74.5, "learning_rate": 9.810169491525424e-07, "loss": 2.397, "step": 1060 }, { "epoch": 0.03, "grad_norm": 71.5, "learning_rate": 9.806779661016949e-07, "loss": 2.3533, "step": 1070 }, { "epoch": 0.03, "grad_norm": 72.0, "learning_rate": 9.803389830508474e-07, "loss": 2.2966, "step": 1080 }, { "epoch": 0.03, "grad_norm": 71.5, "learning_rate": 9.8e-07, "loss": 2.2989, "step": 1090 }, { "epoch": 0.03, "grad_norm": 74.5, "learning_rate": 9.796610169491525e-07, "loss": 2.2941, "step": 1100 }, { "epoch": 0.03, "grad_norm": 75.0, "learning_rate": 9.79322033898305e-07, "loss": 2.2903, "step": 1110 }, { "epoch": 0.03, "grad_norm": 71.0, "learning_rate": 9.789830508474575e-07, "loss": 2.1909, "step": 1120 }, { "epoch": 0.03, "grad_norm": 72.5, "learning_rate": 9.7864406779661e-07, "loss": 2.2198, "step": 1130 }, { "epoch": 0.03, "grad_norm": 71.0, "learning_rate": 9.783050847457626e-07, "loss": 2.1776, "step": 1140 }, { "epoch": 0.03, "grad_norm": 72.0, "learning_rate": 9.779661016949153e-07, "loss": 2.196, "step": 1150 }, { "epoch": 0.03, "grad_norm": 73.5, "learning_rate": 9.776271186440678e-07, "loss": 2.1479, "step": 1160 }, { "epoch": 0.03, "grad_norm": 71.0, "learning_rate": 9.772881355932204e-07, "loss": 2.0738, "step": 1170 }, { "epoch": 0.03, "grad_norm": 70.0, "learning_rate": 9.76949152542373e-07, "loss": 2.1353, "step": 1180 }, { "epoch": 0.03, "grad_norm": 73.0, "learning_rate": 9.766101694915254e-07, "loss": 2.0684, "step": 1190 }, { "epoch": 0.03, "grad_norm": 72.5, "learning_rate": 9.76271186440678e-07, "loss": 2.0626, "step": 1200 }, { "epoch": 0.03, "grad_norm": 72.5, "learning_rate": 9.759322033898305e-07, "loss": 2.0793, "step": 1210 }, { "epoch": 0.03, "grad_norm": 68.5, "learning_rate": 9.75593220338983e-07, "loss": 2.0539, "step": 1220 }, { "epoch": 0.03, "grad_norm": 70.0, "learning_rate": 9.752542372881355e-07, "loss": 2.0261, "step": 1230 }, { "epoch": 0.03, "grad_norm": 69.5, "learning_rate": 9.74915254237288e-07, "loss": 2.0405, "step": 1240 }, { "epoch": 0.03, "grad_norm": 69.0, "learning_rate": 9.745762711864406e-07, "loss": 2.0291, "step": 1250 }, { "epoch": 0.03, "grad_norm": 69.5, "learning_rate": 9.742372881355931e-07, "loss": 1.9667, "step": 1260 }, { "epoch": 0.03, "grad_norm": 77.0, "learning_rate": 9.738983050847456e-07, "loss": 1.9781, "step": 1270 }, { "epoch": 0.03, "grad_norm": 72.0, "learning_rate": 9.735593220338982e-07, "loss": 2.0106, "step": 1280 }, { "epoch": 0.03, "grad_norm": 71.5, "learning_rate": 9.732203389830507e-07, "loss": 1.9822, "step": 1290 }, { "epoch": 0.03, "grad_norm": 69.0, "learning_rate": 9.728813559322032e-07, "loss": 1.9494, "step": 1300 }, { "epoch": 0.03, "grad_norm": 68.0, "learning_rate": 9.72542372881356e-07, "loss": 1.9647, "step": 1310 }, { "epoch": 0.03, "grad_norm": 70.5, "learning_rate": 9.722033898305085e-07, "loss": 1.9182, "step": 1320 }, { "epoch": 0.03, "grad_norm": 70.0, "learning_rate": 9.71864406779661e-07, "loss": 1.9238, "step": 1330 }, { "epoch": 0.03, "grad_norm": 72.0, "learning_rate": 9.715254237288135e-07, "loss": 1.9369, "step": 1340 }, { "epoch": 0.03, "grad_norm": 69.5, "learning_rate": 9.71186440677966e-07, "loss": 1.9225, "step": 1350 }, { "epoch": 0.03, "grad_norm": 67.5, "learning_rate": 9.708474576271186e-07, "loss": 1.9043, "step": 1360 }, { "epoch": 0.03, "grad_norm": 74.5, "learning_rate": 9.705084745762711e-07, "loss": 1.924, "step": 1370 }, { "epoch": 0.03, "grad_norm": 70.0, "learning_rate": 9.701694915254237e-07, "loss": 1.8824, "step": 1380 }, { "epoch": 0.03, "grad_norm": 68.0, "learning_rate": 9.698305084745762e-07, "loss": 1.8886, "step": 1390 }, { "epoch": 0.04, "grad_norm": 69.5, "learning_rate": 9.694915254237287e-07, "loss": 1.8851, "step": 1400 }, { "epoch": 0.04, "grad_norm": 68.0, "learning_rate": 9.691525423728815e-07, "loss": 1.9012, "step": 1410 }, { "epoch": 0.04, "grad_norm": 70.0, "learning_rate": 9.68813559322034e-07, "loss": 1.8949, "step": 1420 }, { "epoch": 0.04, "grad_norm": 71.5, "learning_rate": 9.684745762711865e-07, "loss": 1.8203, "step": 1430 }, { "epoch": 0.04, "grad_norm": 69.5, "learning_rate": 9.681355932203388e-07, "loss": 1.8582, "step": 1440 }, { "epoch": 0.04, "grad_norm": 68.5, "learning_rate": 9.677966101694914e-07, "loss": 1.8363, "step": 1450 }, { "epoch": 0.04, "grad_norm": 71.0, "learning_rate": 9.674576271186439e-07, "loss": 1.8069, "step": 1460 }, { "epoch": 0.04, "grad_norm": 68.5, "learning_rate": 9.671186440677966e-07, "loss": 1.8681, "step": 1470 }, { "epoch": 0.04, "grad_norm": 70.0, "learning_rate": 9.667796610169491e-07, "loss": 1.8022, "step": 1480 }, { "epoch": 0.04, "grad_norm": 73.5, "learning_rate": 9.664406779661017e-07, "loss": 1.8337, "step": 1490 }, { "epoch": 0.04, "grad_norm": 67.5, "learning_rate": 9.661016949152542e-07, "loss": 1.7858, "step": 1500 }, { "epoch": 0.04, "grad_norm": 75.5, "learning_rate": 9.657627118644067e-07, "loss": 1.8289, "step": 1510 }, { "epoch": 0.04, "grad_norm": 70.0, "learning_rate": 9.654237288135593e-07, "loss": 1.7982, "step": 1520 }, { "epoch": 0.04, "grad_norm": 70.0, "learning_rate": 9.650847457627118e-07, "loss": 1.8271, "step": 1530 }, { "epoch": 0.04, "grad_norm": 67.0, "learning_rate": 9.647457627118643e-07, "loss": 1.7395, "step": 1540 }, { "epoch": 0.04, "grad_norm": 67.0, "learning_rate": 9.644067796610168e-07, "loss": 1.82, "step": 1550 }, { "epoch": 0.04, "grad_norm": 67.5, "learning_rate": 9.640677966101694e-07, "loss": 1.8157, "step": 1560 }, { "epoch": 0.04, "grad_norm": 67.0, "learning_rate": 9.637288135593221e-07, "loss": 1.8164, "step": 1570 }, { "epoch": 0.04, "grad_norm": 72.0, "learning_rate": 9.633898305084746e-07, "loss": 1.7715, "step": 1580 }, { "epoch": 0.04, "grad_norm": 71.0, "learning_rate": 9.630508474576272e-07, "loss": 1.7947, "step": 1590 }, { "epoch": 0.04, "grad_norm": 69.5, "learning_rate": 9.627118644067797e-07, "loss": 1.8159, "step": 1600 }, { "epoch": 0.04, "grad_norm": 69.0, "learning_rate": 9.623728813559322e-07, "loss": 1.7614, "step": 1610 }, { "epoch": 0.04, "grad_norm": 71.5, "learning_rate": 9.620338983050847e-07, "loss": 1.7962, "step": 1620 }, { "epoch": 0.04, "grad_norm": 73.5, "learning_rate": 9.616949152542373e-07, "loss": 1.7961, "step": 1630 }, { "epoch": 0.04, "grad_norm": 74.0, "learning_rate": 9.613559322033898e-07, "loss": 1.7997, "step": 1640 }, { "epoch": 0.04, "grad_norm": 71.0, "learning_rate": 9.610169491525423e-07, "loss": 1.7955, "step": 1650 }, { "epoch": 0.04, "grad_norm": 69.0, "learning_rate": 9.606779661016949e-07, "loss": 1.7388, "step": 1660 }, { "epoch": 0.04, "grad_norm": 66.5, "learning_rate": 9.603389830508474e-07, "loss": 1.7265, "step": 1670 }, { "epoch": 0.04, "grad_norm": 69.5, "learning_rate": 9.6e-07, "loss": 1.8179, "step": 1680 }, { "epoch": 0.04, "grad_norm": 72.0, "learning_rate": 9.596610169491524e-07, "loss": 1.7152, "step": 1690 }, { "epoch": 0.04, "grad_norm": 69.5, "learning_rate": 9.59322033898305e-07, "loss": 1.7607, "step": 1700 }, { "epoch": 0.04, "grad_norm": 66.5, "learning_rate": 9.589830508474575e-07, "loss": 1.7638, "step": 1710 }, { "epoch": 0.04, "grad_norm": 69.5, "learning_rate": 9.5864406779661e-07, "loss": 1.726, "step": 1720 }, { "epoch": 0.04, "grad_norm": 67.5, "learning_rate": 9.583050847457628e-07, "loss": 1.7371, "step": 1730 }, { "epoch": 0.04, "grad_norm": 69.5, "learning_rate": 9.579661016949153e-07, "loss": 1.6961, "step": 1740 }, { "epoch": 0.04, "grad_norm": 71.5, "learning_rate": 9.576271186440678e-07, "loss": 1.7849, "step": 1750 }, { "epoch": 0.04, "grad_norm": 70.0, "learning_rate": 9.572881355932203e-07, "loss": 1.7208, "step": 1760 }, { "epoch": 0.04, "grad_norm": 71.0, "learning_rate": 9.569491525423729e-07, "loss": 1.7974, "step": 1770 }, { "epoch": 0.04, "grad_norm": 70.5, "learning_rate": 9.566101694915254e-07, "loss": 1.7476, "step": 1780 }, { "epoch": 0.04, "grad_norm": 69.0, "learning_rate": 9.56271186440678e-07, "loss": 1.752, "step": 1790 }, { "epoch": 0.04, "grad_norm": 69.0, "learning_rate": 9.559322033898305e-07, "loss": 1.7256, "step": 1800 }, { "epoch": 0.05, "grad_norm": 72.0, "learning_rate": 9.55593220338983e-07, "loss": 1.7196, "step": 1810 }, { "epoch": 0.05, "grad_norm": 67.5, "learning_rate": 9.552542372881355e-07, "loss": 1.6905, "step": 1820 }, { "epoch": 0.05, "grad_norm": 67.5, "learning_rate": 9.54915254237288e-07, "loss": 1.6826, "step": 1830 }, { "epoch": 0.05, "grad_norm": 69.0, "learning_rate": 9.545762711864408e-07, "loss": 1.7369, "step": 1840 }, { "epoch": 0.05, "grad_norm": 73.0, "learning_rate": 9.542372881355933e-07, "loss": 1.6937, "step": 1850 }, { "epoch": 0.05, "grad_norm": 68.5, "learning_rate": 9.538983050847456e-07, "loss": 1.6722, "step": 1860 }, { "epoch": 0.05, "grad_norm": 68.5, "learning_rate": 9.535593220338983e-07, "loss": 1.688, "step": 1870 }, { "epoch": 0.05, "grad_norm": 70.0, "learning_rate": 9.532203389830508e-07, "loss": 1.692, "step": 1880 }, { "epoch": 0.05, "grad_norm": 71.5, "learning_rate": 9.528813559322033e-07, "loss": 1.6849, "step": 1890 }, { "epoch": 0.05, "grad_norm": 70.0, "learning_rate": 9.525423728813558e-07, "loss": 1.6826, "step": 1900 }, { "epoch": 0.05, "grad_norm": 66.5, "learning_rate": 9.522033898305084e-07, "loss": 1.7051, "step": 1910 }, { "epoch": 0.05, "grad_norm": 66.0, "learning_rate": 9.51864406779661e-07, "loss": 1.6736, "step": 1920 }, { "epoch": 0.05, "grad_norm": 68.0, "learning_rate": 9.515254237288135e-07, "loss": 1.6739, "step": 1930 }, { "epoch": 0.05, "grad_norm": 69.0, "learning_rate": 9.511864406779661e-07, "loss": 1.6862, "step": 1940 }, { "epoch": 0.05, "grad_norm": 75.0, "learning_rate": 9.508474576271186e-07, "loss": 1.647, "step": 1950 }, { "epoch": 0.05, "grad_norm": 67.0, "learning_rate": 9.505084745762711e-07, "loss": 1.6635, "step": 1960 }, { "epoch": 0.05, "grad_norm": 66.5, "learning_rate": 9.501694915254237e-07, "loss": 1.6047, "step": 1970 }, { "epoch": 0.05, "grad_norm": 63.25, "learning_rate": 9.498305084745763e-07, "loss": 1.6744, "step": 1980 }, { "epoch": 0.05, "grad_norm": 67.0, "learning_rate": 9.494915254237288e-07, "loss": 1.7308, "step": 1990 }, { "epoch": 0.05, "grad_norm": 68.0, "learning_rate": 9.491525423728813e-07, "loss": 1.6831, "step": 2000 }, { "epoch": 0.05, "grad_norm": 69.0, "learning_rate": 9.488135593220339e-07, "loss": 1.6635, "step": 2010 }, { "epoch": 0.05, "grad_norm": 67.0, "learning_rate": 9.484745762711864e-07, "loss": 1.6199, "step": 2020 }, { "epoch": 0.05, "grad_norm": 67.0, "learning_rate": 9.48135593220339e-07, "loss": 1.673, "step": 2030 }, { "epoch": 0.05, "grad_norm": 70.5, "learning_rate": 9.477966101694915e-07, "loss": 1.6659, "step": 2040 }, { "epoch": 0.05, "grad_norm": 69.5, "learning_rate": 9.474576271186441e-07, "loss": 1.6399, "step": 2050 }, { "epoch": 0.05, "grad_norm": 68.5, "learning_rate": 9.471186440677965e-07, "loss": 1.6694, "step": 2060 }, { "epoch": 0.05, "grad_norm": 70.5, "learning_rate": 9.46779661016949e-07, "loss": 1.65, "step": 2070 }, { "epoch": 0.05, "grad_norm": 69.0, "learning_rate": 9.464406779661017e-07, "loss": 1.6637, "step": 2080 }, { "epoch": 0.05, "grad_norm": 67.0, "learning_rate": 9.461016949152542e-07, "loss": 1.6598, "step": 2090 }, { "epoch": 0.05, "grad_norm": 70.0, "learning_rate": 9.457627118644067e-07, "loss": 1.6303, "step": 2100 }, { "epoch": 0.05, "grad_norm": 66.0, "learning_rate": 9.454237288135592e-07, "loss": 1.6766, "step": 2110 }, { "epoch": 0.05, "grad_norm": 67.0, "learning_rate": 9.450847457627118e-07, "loss": 1.6083, "step": 2120 }, { "epoch": 0.05, "grad_norm": 67.5, "learning_rate": 9.447457627118643e-07, "loss": 1.665, "step": 2130 }, { "epoch": 0.05, "grad_norm": 68.0, "learning_rate": 9.444067796610169e-07, "loss": 1.644, "step": 2140 }, { "epoch": 0.05, "grad_norm": 70.0, "learning_rate": 9.440677966101695e-07, "loss": 1.6322, "step": 2150 }, { "epoch": 0.05, "grad_norm": 66.5, "learning_rate": 9.43728813559322e-07, "loss": 1.6223, "step": 2160 }, { "epoch": 0.05, "grad_norm": 66.5, "learning_rate": 9.433898305084745e-07, "loss": 1.6727, "step": 2170 }, { "epoch": 0.05, "grad_norm": 69.0, "learning_rate": 9.43050847457627e-07, "loss": 1.6328, "step": 2180 }, { "epoch": 0.05, "grad_norm": 68.5, "learning_rate": 9.427118644067797e-07, "loss": 1.6439, "step": 2190 }, { "epoch": 0.06, "grad_norm": 67.5, "learning_rate": 9.423728813559322e-07, "loss": 1.6537, "step": 2200 }, { "epoch": 0.06, "grad_norm": 64.5, "learning_rate": 9.420338983050847e-07, "loss": 1.6508, "step": 2210 }, { "epoch": 0.06, "grad_norm": 66.5, "learning_rate": 9.416949152542373e-07, "loss": 1.5704, "step": 2220 }, { "epoch": 0.06, "grad_norm": 66.5, "learning_rate": 9.413559322033898e-07, "loss": 1.5994, "step": 2230 }, { "epoch": 0.06, "grad_norm": 69.0, "learning_rate": 9.410169491525424e-07, "loss": 1.6123, "step": 2240 }, { "epoch": 0.06, "grad_norm": 71.5, "learning_rate": 9.406779661016949e-07, "loss": 1.6146, "step": 2250 }, { "epoch": 0.06, "grad_norm": 68.5, "learning_rate": 9.403389830508475e-07, "loss": 1.6318, "step": 2260 }, { "epoch": 0.06, "grad_norm": 66.0, "learning_rate": 9.399999999999999e-07, "loss": 1.584, "step": 2270 }, { "epoch": 0.06, "grad_norm": 66.5, "learning_rate": 9.396610169491524e-07, "loss": 1.6058, "step": 2280 }, { "epoch": 0.06, "grad_norm": 67.0, "learning_rate": 9.39322033898305e-07, "loss": 1.6386, "step": 2290 }, { "epoch": 0.06, "grad_norm": 67.5, "learning_rate": 9.389830508474576e-07, "loss": 1.6293, "step": 2300 }, { "epoch": 0.06, "grad_norm": 67.5, "learning_rate": 9.386440677966101e-07, "loss": 1.596, "step": 2310 }, { "epoch": 0.06, "grad_norm": 67.5, "learning_rate": 9.383050847457626e-07, "loss": 1.6159, "step": 2320 }, { "epoch": 0.06, "grad_norm": 66.5, "learning_rate": 9.379661016949152e-07, "loss": 1.5201, "step": 2330 }, { "epoch": 0.06, "grad_norm": 71.0, "learning_rate": 9.376271186440677e-07, "loss": 1.6464, "step": 2340 }, { "epoch": 0.06, "grad_norm": 66.0, "learning_rate": 9.372881355932203e-07, "loss": 1.6035, "step": 2350 }, { "epoch": 0.06, "grad_norm": 69.0, "learning_rate": 9.369491525423729e-07, "loss": 1.6379, "step": 2360 }, { "epoch": 0.06, "grad_norm": 68.0, "learning_rate": 9.366101694915254e-07, "loss": 1.6182, "step": 2370 }, { "epoch": 0.06, "grad_norm": 67.0, "learning_rate": 9.362711864406779e-07, "loss": 1.6201, "step": 2380 }, { "epoch": 0.06, "grad_norm": 71.0, "learning_rate": 9.359322033898304e-07, "loss": 1.569, "step": 2390 }, { "epoch": 0.06, "grad_norm": 67.5, "learning_rate": 9.355932203389831e-07, "loss": 1.5749, "step": 2400 }, { "epoch": 0.06, "grad_norm": 67.0, "learning_rate": 9.352542372881356e-07, "loss": 1.6154, "step": 2410 }, { "epoch": 0.06, "grad_norm": 67.0, "learning_rate": 9.349152542372881e-07, "loss": 1.5639, "step": 2420 }, { "epoch": 0.06, "grad_norm": 65.5, "learning_rate": 9.345762711864407e-07, "loss": 1.5745, "step": 2430 }, { "epoch": 0.06, "grad_norm": 71.5, "learning_rate": 9.342372881355932e-07, "loss": 1.5988, "step": 2440 }, { "epoch": 0.06, "grad_norm": 72.0, "learning_rate": 9.338983050847458e-07, "loss": 1.6111, "step": 2450 }, { "epoch": 0.06, "grad_norm": 68.5, "learning_rate": 9.335593220338983e-07, "loss": 1.6351, "step": 2460 }, { "epoch": 0.06, "grad_norm": 68.5, "learning_rate": 9.332203389830509e-07, "loss": 1.5946, "step": 2470 }, { "epoch": 0.06, "grad_norm": 69.5, "learning_rate": 9.328813559322033e-07, "loss": 1.5306, "step": 2480 }, { "epoch": 0.06, "grad_norm": 65.5, "learning_rate": 9.325423728813558e-07, "loss": 1.5605, "step": 2490 }, { "epoch": 0.06, "grad_norm": 68.5, "learning_rate": 9.322033898305083e-07, "loss": 1.5907, "step": 2500 }, { "epoch": 0.06, "grad_norm": 66.0, "learning_rate": 9.31864406779661e-07, "loss": 1.5823, "step": 2510 }, { "epoch": 0.06, "grad_norm": 65.5, "learning_rate": 9.315254237288135e-07, "loss": 1.565, "step": 2520 }, { "epoch": 0.06, "grad_norm": 69.0, "learning_rate": 9.31186440677966e-07, "loss": 1.5669, "step": 2530 }, { "epoch": 0.06, "grad_norm": 67.5, "learning_rate": 9.308474576271186e-07, "loss": 1.5595, "step": 2540 }, { "epoch": 0.06, "grad_norm": 65.5, "learning_rate": 9.305084745762711e-07, "loss": 1.5367, "step": 2550 }, { "epoch": 0.06, "grad_norm": 67.5, "learning_rate": 9.301694915254237e-07, "loss": 1.6048, "step": 2560 }, { "epoch": 0.06, "grad_norm": 66.5, "learning_rate": 9.298305084745763e-07, "loss": 1.6001, "step": 2570 }, { "epoch": 0.06, "grad_norm": 71.0, "learning_rate": 9.294915254237288e-07, "loss": 1.6077, "step": 2580 }, { "epoch": 0.06, "grad_norm": 66.5, "learning_rate": 9.291525423728813e-07, "loss": 1.5488, "step": 2590 }, { "epoch": 0.07, "grad_norm": 67.0, "learning_rate": 9.288135593220338e-07, "loss": 1.5676, "step": 2600 }, { "epoch": 0.07, "grad_norm": 67.5, "learning_rate": 9.284745762711864e-07, "loss": 1.5455, "step": 2610 }, { "epoch": 0.07, "grad_norm": 65.0, "learning_rate": 9.28135593220339e-07, "loss": 1.5867, "step": 2620 }, { "epoch": 0.07, "grad_norm": 67.0, "learning_rate": 9.277966101694915e-07, "loss": 1.5505, "step": 2630 }, { "epoch": 0.07, "grad_norm": 68.0, "learning_rate": 9.274576271186441e-07, "loss": 1.5646, "step": 2640 }, { "epoch": 0.07, "grad_norm": 65.0, "learning_rate": 9.271186440677966e-07, "loss": 1.5345, "step": 2650 }, { "epoch": 0.07, "grad_norm": 69.5, "learning_rate": 9.267796610169491e-07, "loss": 1.6042, "step": 2660 }, { "epoch": 0.07, "grad_norm": 66.0, "learning_rate": 9.264406779661017e-07, "loss": 1.5496, "step": 2670 }, { "epoch": 0.07, "grad_norm": 67.0, "learning_rate": 9.261016949152543e-07, "loss": 1.565, "step": 2680 }, { "epoch": 0.07, "grad_norm": 67.0, "learning_rate": 9.257627118644067e-07, "loss": 1.5168, "step": 2690 }, { "epoch": 0.07, "grad_norm": 68.0, "learning_rate": 9.254237288135592e-07, "loss": 1.5704, "step": 2700 }, { "epoch": 0.07, "grad_norm": 68.5, "learning_rate": 9.250847457627117e-07, "loss": 1.5708, "step": 2710 }, { "epoch": 0.07, "grad_norm": 71.0, "learning_rate": 9.247457627118644e-07, "loss": 1.5891, "step": 2720 }, { "epoch": 0.07, "grad_norm": 66.0, "learning_rate": 9.244067796610169e-07, "loss": 1.5333, "step": 2730 }, { "epoch": 0.07, "grad_norm": 69.0, "learning_rate": 9.240677966101694e-07, "loss": 1.5307, "step": 2740 }, { "epoch": 0.07, "grad_norm": 66.0, "learning_rate": 9.23728813559322e-07, "loss": 1.5467, "step": 2750 }, { "epoch": 0.07, "grad_norm": 66.0, "learning_rate": 9.233898305084745e-07, "loss": 1.5557, "step": 2760 }, { "epoch": 0.07, "grad_norm": 69.5, "learning_rate": 9.23050847457627e-07, "loss": 1.54, "step": 2770 }, { "epoch": 0.07, "grad_norm": 69.5, "learning_rate": 9.227118644067797e-07, "loss": 1.5305, "step": 2780 }, { "epoch": 0.07, "grad_norm": 68.0, "learning_rate": 9.223728813559322e-07, "loss": 1.4947, "step": 2790 }, { "epoch": 0.07, "grad_norm": 70.5, "learning_rate": 9.220338983050847e-07, "loss": 1.5353, "step": 2800 }, { "epoch": 0.07, "grad_norm": 66.5, "learning_rate": 9.216949152542372e-07, "loss": 1.578, "step": 2810 }, { "epoch": 0.07, "grad_norm": 70.5, "learning_rate": 9.213559322033898e-07, "loss": 1.5556, "step": 2820 }, { "epoch": 0.07, "grad_norm": 66.5, "learning_rate": 9.210169491525424e-07, "loss": 1.4998, "step": 2830 }, { "epoch": 0.07, "grad_norm": 66.5, "learning_rate": 9.206779661016949e-07, "loss": 1.5517, "step": 2840 }, { "epoch": 0.07, "grad_norm": 67.5, "learning_rate": 9.203389830508475e-07, "loss": 1.5942, "step": 2850 }, { "epoch": 0.07, "grad_norm": 67.5, "learning_rate": 9.2e-07, "loss": 1.5581, "step": 2860 }, { "epoch": 0.07, "grad_norm": 66.0, "learning_rate": 9.196610169491525e-07, "loss": 1.5578, "step": 2870 }, { "epoch": 0.07, "grad_norm": 69.5, "learning_rate": 9.193220338983051e-07, "loss": 1.5153, "step": 2880 }, { "epoch": 0.07, "grad_norm": 66.5, "learning_rate": 9.189830508474576e-07, "loss": 1.5391, "step": 2890 }, { "epoch": 0.07, "grad_norm": 69.0, "learning_rate": 9.186440677966101e-07, "loss": 1.5475, "step": 2900 }, { "epoch": 0.07, "grad_norm": 68.5, "learning_rate": 9.183050847457626e-07, "loss": 1.5681, "step": 2910 }, { "epoch": 0.07, "grad_norm": 67.0, "learning_rate": 9.179661016949151e-07, "loss": 1.5216, "step": 2920 }, { "epoch": 0.07, "grad_norm": 67.0, "learning_rate": 9.176271186440677e-07, "loss": 1.5378, "step": 2930 }, { "epoch": 0.07, "grad_norm": 67.5, "learning_rate": 9.172881355932203e-07, "loss": 1.5092, "step": 2940 }, { "epoch": 0.07, "grad_norm": 68.0, "learning_rate": 9.169491525423728e-07, "loss": 1.5565, "step": 2950 }, { "epoch": 0.07, "grad_norm": 68.5, "learning_rate": 9.166101694915254e-07, "loss": 1.5672, "step": 2960 }, { "epoch": 0.07, "grad_norm": 66.0, "learning_rate": 9.162711864406779e-07, "loss": 1.5431, "step": 2970 }, { "epoch": 0.07, "grad_norm": 68.0, "learning_rate": 9.159322033898304e-07, "loss": 1.5306, "step": 2980 }, { "epoch": 0.07, "grad_norm": 69.0, "learning_rate": 9.15593220338983e-07, "loss": 1.5469, "step": 2990 }, { "epoch": 0.07, "grad_norm": 66.5, "learning_rate": 9.152542372881356e-07, "loss": 1.4952, "step": 3000 }, { "epoch": 0.08, "grad_norm": 66.5, "learning_rate": 9.149152542372881e-07, "loss": 1.5161, "step": 3010 }, { "epoch": 0.08, "grad_norm": 65.5, "learning_rate": 9.145762711864406e-07, "loss": 1.5524, "step": 3020 }, { "epoch": 0.08, "grad_norm": 69.0, "learning_rate": 9.142372881355932e-07, "loss": 1.5231, "step": 3030 }, { "epoch": 0.08, "grad_norm": 70.5, "learning_rate": 9.138983050847458e-07, "loss": 1.5075, "step": 3040 }, { "epoch": 0.08, "grad_norm": 66.5, "learning_rate": 9.135593220338983e-07, "loss": 1.5372, "step": 3050 }, { "epoch": 0.08, "grad_norm": 70.5, "learning_rate": 9.132203389830509e-07, "loss": 1.4916, "step": 3060 }, { "epoch": 0.08, "grad_norm": 64.5, "learning_rate": 9.128813559322034e-07, "loss": 1.4986, "step": 3070 }, { "epoch": 0.08, "grad_norm": 71.0, "learning_rate": 9.125423728813559e-07, "loss": 1.5884, "step": 3080 }, { "epoch": 0.08, "grad_norm": 69.5, "learning_rate": 9.122033898305084e-07, "loss": 1.5365, "step": 3090 }, { "epoch": 0.08, "grad_norm": 68.0, "learning_rate": 9.11864406779661e-07, "loss": 1.5629, "step": 3100 }, { "epoch": 0.08, "grad_norm": 69.0, "learning_rate": 9.115254237288135e-07, "loss": 1.5343, "step": 3110 }, { "epoch": 0.08, "grad_norm": 70.0, "learning_rate": 9.11186440677966e-07, "loss": 1.558, "step": 3120 }, { "epoch": 0.08, "grad_norm": 70.0, "learning_rate": 9.108474576271185e-07, "loss": 1.4962, "step": 3130 }, { "epoch": 0.08, "grad_norm": 67.0, "learning_rate": 9.105084745762711e-07, "loss": 1.4927, "step": 3140 }, { "epoch": 0.08, "grad_norm": 70.0, "learning_rate": 9.101694915254237e-07, "loss": 1.4935, "step": 3150 }, { "epoch": 0.08, "grad_norm": 65.5, "learning_rate": 9.098305084745762e-07, "loss": 1.493, "step": 3160 }, { "epoch": 0.08, "grad_norm": 65.0, "learning_rate": 9.094915254237288e-07, "loss": 1.5289, "step": 3170 }, { "epoch": 0.08, "grad_norm": 69.5, "learning_rate": 9.091525423728813e-07, "loss": 1.4893, "step": 3180 }, { "epoch": 0.08, "grad_norm": 69.0, "learning_rate": 9.088135593220338e-07, "loss": 1.4864, "step": 3190 }, { "epoch": 0.08, "grad_norm": 67.5, "learning_rate": 9.084745762711864e-07, "loss": 1.4915, "step": 3200 }, { "epoch": 0.08, "grad_norm": 65.5, "learning_rate": 9.08135593220339e-07, "loss": 1.4848, "step": 3210 }, { "epoch": 0.08, "grad_norm": 69.5, "learning_rate": 9.077966101694915e-07, "loss": 1.5586, "step": 3220 }, { "epoch": 0.08, "grad_norm": 72.0, "learning_rate": 9.07457627118644e-07, "loss": 1.5623, "step": 3230 }, { "epoch": 0.08, "grad_norm": 71.0, "learning_rate": 9.071186440677966e-07, "loss": 1.529, "step": 3240 }, { "epoch": 0.08, "grad_norm": 67.5, "learning_rate": 9.067796610169491e-07, "loss": 1.5422, "step": 3250 }, { "epoch": 0.08, "grad_norm": 71.5, "learning_rate": 9.064406779661017e-07, "loss": 1.5146, "step": 3260 }, { "epoch": 0.08, "grad_norm": 66.0, "learning_rate": 9.061016949152542e-07, "loss": 1.5405, "step": 3270 }, { "epoch": 0.08, "grad_norm": 69.0, "learning_rate": 9.057627118644068e-07, "loss": 1.5228, "step": 3280 }, { "epoch": 0.08, "grad_norm": 63.0, "learning_rate": 9.054237288135593e-07, "loss": 1.4958, "step": 3290 }, { "epoch": 0.08, "grad_norm": 65.5, "learning_rate": 9.050847457627118e-07, "loss": 1.4941, "step": 3300 }, { "epoch": 0.08, "grad_norm": 67.0, "learning_rate": 9.047457627118644e-07, "loss": 1.4789, "step": 3310 }, { "epoch": 0.08, "grad_norm": 70.5, "learning_rate": 9.044067796610169e-07, "loss": 1.5649, "step": 3320 }, { "epoch": 0.08, "grad_norm": 62.5, "learning_rate": 9.040677966101694e-07, "loss": 1.5423, "step": 3330 }, { "epoch": 0.08, "grad_norm": 66.5, "learning_rate": 9.037288135593219e-07, "loss": 1.5643, "step": 3340 }, { "epoch": 0.08, "grad_norm": 67.5, "learning_rate": 9.033898305084745e-07, "loss": 1.5374, "step": 3350 }, { "epoch": 0.08, "grad_norm": 67.0, "learning_rate": 9.030508474576271e-07, "loss": 1.513, "step": 3360 }, { "epoch": 0.08, "grad_norm": 66.5, "learning_rate": 9.027118644067796e-07, "loss": 1.4677, "step": 3370 }, { "epoch": 0.08, "grad_norm": 66.5, "learning_rate": 9.023728813559322e-07, "loss": 1.5205, "step": 3380 }, { "epoch": 0.08, "grad_norm": 65.5, "learning_rate": 9.020338983050847e-07, "loss": 1.4427, "step": 3390 }, { "epoch": 0.09, "grad_norm": 67.5, "learning_rate": 9.016949152542372e-07, "loss": 1.4932, "step": 3400 }, { "epoch": 0.09, "grad_norm": 65.0, "learning_rate": 9.013559322033897e-07, "loss": 1.4896, "step": 3410 }, { "epoch": 0.09, "grad_norm": 69.5, "learning_rate": 9.010169491525424e-07, "loss": 1.4794, "step": 3420 }, { "epoch": 0.09, "grad_norm": 69.5, "learning_rate": 9.006779661016949e-07, "loss": 1.4937, "step": 3430 }, { "epoch": 0.09, "grad_norm": 63.5, "learning_rate": 9.003389830508474e-07, "loss": 1.505, "step": 3440 }, { "epoch": 0.09, "grad_norm": 68.0, "learning_rate": 9e-07, "loss": 1.5064, "step": 3450 }, { "epoch": 0.09, "grad_norm": 66.0, "learning_rate": 8.996610169491525e-07, "loss": 1.5155, "step": 3460 }, { "epoch": 0.09, "grad_norm": 67.0, "learning_rate": 8.993220338983051e-07, "loss": 1.4963, "step": 3470 }, { "epoch": 0.09, "grad_norm": 69.0, "learning_rate": 8.989830508474576e-07, "loss": 1.5501, "step": 3480 }, { "epoch": 0.09, "grad_norm": 68.0, "learning_rate": 8.986440677966102e-07, "loss": 1.4952, "step": 3490 }, { "epoch": 0.09, "grad_norm": 67.5, "learning_rate": 8.983050847457627e-07, "loss": 1.5111, "step": 3500 }, { "epoch": 0.09, "grad_norm": 72.5, "learning_rate": 8.979661016949152e-07, "loss": 1.458, "step": 3510 }, { "epoch": 0.09, "grad_norm": 65.0, "learning_rate": 8.976271186440677e-07, "loss": 1.4459, "step": 3520 }, { "epoch": 0.09, "grad_norm": 66.0, "learning_rate": 8.972881355932203e-07, "loss": 1.4634, "step": 3530 }, { "epoch": 0.09, "grad_norm": 65.0, "learning_rate": 8.969491525423728e-07, "loss": 1.5086, "step": 3540 }, { "epoch": 0.09, "grad_norm": 69.0, "learning_rate": 8.966101694915253e-07, "loss": 1.5226, "step": 3550 }, { "epoch": 0.09, "grad_norm": 69.0, "learning_rate": 8.962711864406779e-07, "loss": 1.5094, "step": 3560 }, { "epoch": 0.09, "grad_norm": 75.5, "learning_rate": 8.959322033898304e-07, "loss": 1.4869, "step": 3570 }, { "epoch": 0.09, "grad_norm": 65.0, "learning_rate": 8.95593220338983e-07, "loss": 1.4863, "step": 3580 }, { "epoch": 0.09, "grad_norm": 66.0, "learning_rate": 8.952542372881356e-07, "loss": 1.4733, "step": 3590 }, { "epoch": 0.09, "grad_norm": 76.5, "learning_rate": 8.949152542372881e-07, "loss": 1.4918, "step": 3600 }, { "epoch": 0.09, "grad_norm": 69.0, "learning_rate": 8.945762711864406e-07, "loss": 1.5012, "step": 3610 }, { "epoch": 0.09, "grad_norm": 68.5, "learning_rate": 8.942372881355931e-07, "loss": 1.4974, "step": 3620 }, { "epoch": 0.09, "grad_norm": 64.0, "learning_rate": 8.938983050847458e-07, "loss": 1.441, "step": 3630 }, { "epoch": 0.09, "grad_norm": 65.5, "learning_rate": 8.935593220338983e-07, "loss": 1.4837, "step": 3640 }, { "epoch": 0.09, "grad_norm": 67.5, "learning_rate": 8.932203389830508e-07, "loss": 1.4722, "step": 3650 }, { "epoch": 0.09, "grad_norm": 69.5, "learning_rate": 8.928813559322034e-07, "loss": 1.4821, "step": 3660 }, { "epoch": 0.09, "grad_norm": 68.0, "learning_rate": 8.925423728813559e-07, "loss": 1.5346, "step": 3670 }, { "epoch": 0.09, "grad_norm": 65.0, "learning_rate": 8.922033898305085e-07, "loss": 1.5196, "step": 3680 }, { "epoch": 0.09, "grad_norm": 66.0, "learning_rate": 8.91864406779661e-07, "loss": 1.4761, "step": 3690 }, { "epoch": 0.09, "grad_norm": 65.5, "learning_rate": 8.915254237288136e-07, "loss": 1.4885, "step": 3700 }, { "epoch": 0.09, "grad_norm": 66.0, "learning_rate": 8.911864406779661e-07, "loss": 1.5114, "step": 3710 }, { "epoch": 0.09, "grad_norm": 68.0, "learning_rate": 8.908474576271186e-07, "loss": 1.4701, "step": 3720 }, { "epoch": 0.09, "grad_norm": 64.5, "learning_rate": 8.90508474576271e-07, "loss": 1.501, "step": 3730 }, { "epoch": 0.09, "grad_norm": 68.0, "learning_rate": 8.901694915254237e-07, "loss": 1.4863, "step": 3740 }, { "epoch": 0.09, "grad_norm": 66.5, "learning_rate": 8.898305084745762e-07, "loss": 1.4616, "step": 3750 }, { "epoch": 0.09, "grad_norm": 67.5, "learning_rate": 8.894915254237287e-07, "loss": 1.4995, "step": 3760 }, { "epoch": 0.09, "grad_norm": 66.0, "learning_rate": 8.891525423728813e-07, "loss": 1.4846, "step": 3770 }, { "epoch": 0.09, "grad_norm": 67.0, "learning_rate": 8.888135593220338e-07, "loss": 1.4895, "step": 3780 }, { "epoch": 0.09, "grad_norm": 67.5, "learning_rate": 8.884745762711864e-07, "loss": 1.5152, "step": 3790 }, { "epoch": 0.1, "grad_norm": 70.0, "learning_rate": 8.88135593220339e-07, "loss": 1.4775, "step": 3800 }, { "epoch": 0.1, "grad_norm": 62.5, "learning_rate": 8.877966101694915e-07, "loss": 1.454, "step": 3810 }, { "epoch": 0.1, "grad_norm": 65.0, "learning_rate": 8.87457627118644e-07, "loss": 1.5004, "step": 3820 }, { "epoch": 0.1, "grad_norm": 66.0, "learning_rate": 8.871186440677965e-07, "loss": 1.4825, "step": 3830 }, { "epoch": 0.1, "grad_norm": 68.0, "learning_rate": 8.867796610169492e-07, "loss": 1.5067, "step": 3840 }, { "epoch": 0.1, "grad_norm": 67.5, "learning_rate": 8.864406779661017e-07, "loss": 1.5393, "step": 3850 }, { "epoch": 0.1, "grad_norm": 68.5, "learning_rate": 8.861016949152542e-07, "loss": 1.467, "step": 3860 }, { "epoch": 0.1, "grad_norm": 68.0, "learning_rate": 8.857627118644068e-07, "loss": 1.477, "step": 3870 }, { "epoch": 0.1, "grad_norm": 67.5, "learning_rate": 8.854237288135593e-07, "loss": 1.4855, "step": 3880 }, { "epoch": 0.1, "grad_norm": 67.0, "learning_rate": 8.850847457627118e-07, "loss": 1.4544, "step": 3890 }, { "epoch": 0.1, "grad_norm": 66.5, "learning_rate": 8.847457627118644e-07, "loss": 1.4787, "step": 3900 }, { "epoch": 0.1, "grad_norm": 73.0, "learning_rate": 8.84406779661017e-07, "loss": 1.4947, "step": 3910 }, { "epoch": 0.1, "grad_norm": 65.5, "learning_rate": 8.840677966101695e-07, "loss": 1.4749, "step": 3920 }, { "epoch": 0.1, "grad_norm": 66.0, "learning_rate": 8.837288135593219e-07, "loss": 1.5096, "step": 3930 }, { "epoch": 0.1, "grad_norm": 68.5, "learning_rate": 8.833898305084745e-07, "loss": 1.5044, "step": 3940 }, { "epoch": 0.1, "grad_norm": 77.0, "learning_rate": 8.830508474576271e-07, "loss": 1.4945, "step": 3950 }, { "epoch": 0.1, "grad_norm": 67.5, "learning_rate": 8.827118644067796e-07, "loss": 1.4693, "step": 3960 }, { "epoch": 0.1, "grad_norm": 68.5, "learning_rate": 8.823728813559321e-07, "loss": 1.4914, "step": 3970 }, { "epoch": 0.1, "grad_norm": 71.0, "learning_rate": 8.820338983050847e-07, "loss": 1.4551, "step": 3980 }, { "epoch": 0.1, "grad_norm": 65.5, "learning_rate": 8.816949152542372e-07, "loss": 1.5076, "step": 3990 }, { "epoch": 0.1, "grad_norm": 66.0, "learning_rate": 8.813559322033897e-07, "loss": 1.48, "step": 4000 }, { "epoch": 0.1, "grad_norm": 65.5, "learning_rate": 8.810169491525424e-07, "loss": 1.4452, "step": 4010 }, { "epoch": 0.1, "grad_norm": 64.5, "learning_rate": 8.806779661016949e-07, "loss": 1.4619, "step": 4020 }, { "epoch": 0.1, "grad_norm": 65.0, "learning_rate": 8.803389830508474e-07, "loss": 1.454, "step": 4030 }, { "epoch": 0.1, "grad_norm": 67.0, "learning_rate": 8.799999999999999e-07, "loss": 1.4648, "step": 4040 }, { "epoch": 0.1, "grad_norm": 64.5, "learning_rate": 8.796610169491525e-07, "loss": 1.5015, "step": 4050 }, { "epoch": 0.1, "grad_norm": 66.5, "learning_rate": 8.793220338983051e-07, "loss": 1.4932, "step": 4060 }, { "epoch": 0.1, "grad_norm": 66.0, "learning_rate": 8.789830508474576e-07, "loss": 1.4945, "step": 4070 }, { "epoch": 0.1, "grad_norm": 67.5, "learning_rate": 8.786440677966102e-07, "loss": 1.5184, "step": 4080 }, { "epoch": 0.1, "grad_norm": 65.0, "learning_rate": 8.783050847457627e-07, "loss": 1.4599, "step": 4090 }, { "epoch": 0.1, "grad_norm": 65.5, "learning_rate": 8.779661016949152e-07, "loss": 1.5357, "step": 4100 }, { "epoch": 0.1, "grad_norm": 69.0, "learning_rate": 8.776271186440678e-07, "loss": 1.4867, "step": 4110 }, { "epoch": 0.1, "grad_norm": 67.5, "learning_rate": 8.772881355932204e-07, "loss": 1.4618, "step": 4120 }, { "epoch": 0.1, "grad_norm": 66.5, "learning_rate": 8.769491525423729e-07, "loss": 1.4736, "step": 4130 }, { "epoch": 0.1, "grad_norm": 66.0, "learning_rate": 8.766101694915253e-07, "loss": 1.4781, "step": 4140 }, { "epoch": 0.1, "grad_norm": 66.0, "learning_rate": 8.762711864406778e-07, "loss": 1.4892, "step": 4150 }, { "epoch": 0.1, "grad_norm": 68.5, "learning_rate": 8.759322033898304e-07, "loss": 1.4566, "step": 4160 }, { "epoch": 0.1, "grad_norm": 66.0, "learning_rate": 8.75593220338983e-07, "loss": 1.4814, "step": 4170 }, { "epoch": 0.1, "grad_norm": 71.0, "learning_rate": 8.752542372881355e-07, "loss": 1.4921, "step": 4180 }, { "epoch": 0.1, "grad_norm": 70.5, "learning_rate": 8.749152542372881e-07, "loss": 1.5361, "step": 4190 }, { "epoch": 0.1, "grad_norm": 65.0, "learning_rate": 8.745762711864406e-07, "loss": 1.4342, "step": 4200 }, { "epoch": 0.11, "grad_norm": 66.0, "learning_rate": 8.742372881355931e-07, "loss": 1.5141, "step": 4210 }, { "epoch": 0.11, "grad_norm": 67.5, "learning_rate": 8.738983050847458e-07, "loss": 1.5203, "step": 4220 }, { "epoch": 0.11, "grad_norm": 69.5, "learning_rate": 8.735593220338983e-07, "loss": 1.4885, "step": 4230 }, { "epoch": 0.11, "grad_norm": 65.5, "learning_rate": 8.732203389830508e-07, "loss": 1.5292, "step": 4240 }, { "epoch": 0.11, "grad_norm": 65.0, "learning_rate": 8.728813559322033e-07, "loss": 1.5156, "step": 4250 }, { "epoch": 0.11, "grad_norm": 64.0, "learning_rate": 8.725423728813559e-07, "loss": 1.4583, "step": 4260 }, { "epoch": 0.11, "grad_norm": 64.0, "learning_rate": 8.722033898305085e-07, "loss": 1.5098, "step": 4270 }, { "epoch": 0.11, "grad_norm": 67.0, "learning_rate": 8.71864406779661e-07, "loss": 1.4739, "step": 4280 }, { "epoch": 0.11, "grad_norm": 68.0, "learning_rate": 8.715254237288136e-07, "loss": 1.484, "step": 4290 }, { "epoch": 0.11, "grad_norm": 67.5, "learning_rate": 8.711864406779661e-07, "loss": 1.5197, "step": 4300 }, { "epoch": 0.11, "grad_norm": 66.5, "learning_rate": 8.708474576271186e-07, "loss": 1.5088, "step": 4310 }, { "epoch": 0.11, "grad_norm": 68.0, "learning_rate": 8.705084745762711e-07, "loss": 1.4749, "step": 4320 }, { "epoch": 0.11, "grad_norm": 67.0, "learning_rate": 8.701694915254238e-07, "loss": 1.4429, "step": 4330 }, { "epoch": 0.11, "grad_norm": 67.0, "learning_rate": 8.698305084745763e-07, "loss": 1.4921, "step": 4340 }, { "epoch": 0.11, "grad_norm": 65.5, "learning_rate": 8.694915254237287e-07, "loss": 1.4362, "step": 4350 }, { "epoch": 0.11, "grad_norm": 69.0, "learning_rate": 8.691525423728812e-07, "loss": 1.4985, "step": 4360 }, { "epoch": 0.11, "grad_norm": 62.75, "learning_rate": 8.688135593220338e-07, "loss": 1.4431, "step": 4370 }, { "epoch": 0.11, "grad_norm": 67.0, "learning_rate": 8.684745762711864e-07, "loss": 1.4694, "step": 4380 }, { "epoch": 0.11, "grad_norm": 67.0, "learning_rate": 8.681355932203389e-07, "loss": 1.5127, "step": 4390 }, { "epoch": 0.11, "grad_norm": 70.0, "learning_rate": 8.677966101694915e-07, "loss": 1.5039, "step": 4400 }, { "epoch": 0.11, "grad_norm": 66.5, "learning_rate": 8.67457627118644e-07, "loss": 1.4797, "step": 4410 }, { "epoch": 0.11, "grad_norm": 70.0, "learning_rate": 8.671186440677965e-07, "loss": 1.4775, "step": 4420 }, { "epoch": 0.11, "grad_norm": 67.0, "learning_rate": 8.667796610169492e-07, "loss": 1.4614, "step": 4430 }, { "epoch": 0.11, "grad_norm": 66.5, "learning_rate": 8.664406779661017e-07, "loss": 1.4886, "step": 4440 }, { "epoch": 0.11, "grad_norm": 65.5, "learning_rate": 8.661016949152542e-07, "loss": 1.4871, "step": 4450 }, { "epoch": 0.11, "grad_norm": 66.5, "learning_rate": 8.657627118644067e-07, "loss": 1.4607, "step": 4460 }, { "epoch": 0.11, "grad_norm": 68.5, "learning_rate": 8.654237288135593e-07, "loss": 1.4581, "step": 4470 }, { "epoch": 0.11, "grad_norm": 64.0, "learning_rate": 8.650847457627118e-07, "loss": 1.4685, "step": 4480 }, { "epoch": 0.11, "grad_norm": 69.0, "learning_rate": 8.647457627118644e-07, "loss": 1.4421, "step": 4490 }, { "epoch": 0.11, "grad_norm": 64.5, "learning_rate": 8.64406779661017e-07, "loss": 1.4878, "step": 4500 }, { "epoch": 0.11, "grad_norm": 65.0, "learning_rate": 8.640677966101695e-07, "loss": 1.4713, "step": 4510 }, { "epoch": 0.11, "grad_norm": 66.0, "learning_rate": 8.63728813559322e-07, "loss": 1.5214, "step": 4520 }, { "epoch": 0.11, "grad_norm": 63.75, "learning_rate": 8.633898305084745e-07, "loss": 1.4646, "step": 4530 }, { "epoch": 0.11, "grad_norm": 67.0, "learning_rate": 8.630508474576272e-07, "loss": 1.4881, "step": 4540 }, { "epoch": 0.11, "grad_norm": 66.0, "learning_rate": 8.627118644067797e-07, "loss": 1.4766, "step": 4550 }, { "epoch": 0.11, "grad_norm": 70.5, "learning_rate": 8.623728813559321e-07, "loss": 1.4862, "step": 4560 }, { "epoch": 0.11, "grad_norm": 68.5, "learning_rate": 8.620338983050846e-07, "loss": 1.3567, "step": 4570 }, { "epoch": 0.11, "grad_norm": 65.0, "learning_rate": 8.616949152542372e-07, "loss": 1.4422, "step": 4580 }, { "epoch": 0.11, "grad_norm": 67.0, "learning_rate": 8.613559322033898e-07, "loss": 1.4581, "step": 4590 }, { "epoch": 0.12, "grad_norm": 65.0, "learning_rate": 8.610169491525423e-07, "loss": 1.4973, "step": 4600 }, { "epoch": 0.12, "grad_norm": 66.0, "learning_rate": 8.606779661016949e-07, "loss": 1.4999, "step": 4610 }, { "epoch": 0.12, "grad_norm": 68.0, "learning_rate": 8.603389830508474e-07, "loss": 1.4901, "step": 4620 }, { "epoch": 0.12, "grad_norm": 66.0, "learning_rate": 8.599999999999999e-07, "loss": 1.4738, "step": 4630 }, { "epoch": 0.12, "grad_norm": 71.0, "learning_rate": 8.596610169491524e-07, "loss": 1.4951, "step": 4640 }, { "epoch": 0.12, "grad_norm": 65.0, "learning_rate": 8.593220338983051e-07, "loss": 1.4923, "step": 4650 }, { "epoch": 0.12, "grad_norm": 67.5, "learning_rate": 8.589830508474576e-07, "loss": 1.4571, "step": 4660 }, { "epoch": 0.12, "grad_norm": 67.0, "learning_rate": 8.586440677966101e-07, "loss": 1.4913, "step": 4670 }, { "epoch": 0.12, "grad_norm": 65.5, "learning_rate": 8.583050847457627e-07, "loss": 1.4655, "step": 4680 }, { "epoch": 0.12, "grad_norm": 68.5, "learning_rate": 8.579661016949152e-07, "loss": 1.485, "step": 4690 }, { "epoch": 0.12, "grad_norm": 65.5, "learning_rate": 8.576271186440678e-07, "loss": 1.4815, "step": 4700 }, { "epoch": 0.12, "grad_norm": 70.0, "learning_rate": 8.572881355932204e-07, "loss": 1.5056, "step": 4710 }, { "epoch": 0.12, "grad_norm": 68.0, "learning_rate": 8.569491525423729e-07, "loss": 1.4773, "step": 4720 }, { "epoch": 0.12, "grad_norm": 69.5, "learning_rate": 8.566101694915254e-07, "loss": 1.5163, "step": 4730 }, { "epoch": 0.12, "grad_norm": 69.0, "learning_rate": 8.562711864406779e-07, "loss": 1.4273, "step": 4740 }, { "epoch": 0.12, "grad_norm": 67.0, "learning_rate": 8.559322033898306e-07, "loss": 1.4475, "step": 4750 }, { "epoch": 0.12, "grad_norm": 71.5, "learning_rate": 8.55593220338983e-07, "loss": 1.4884, "step": 4760 }, { "epoch": 0.12, "grad_norm": 71.5, "learning_rate": 8.552542372881355e-07, "loss": 1.5041, "step": 4770 }, { "epoch": 0.12, "grad_norm": 66.0, "learning_rate": 8.54915254237288e-07, "loss": 1.4597, "step": 4780 }, { "epoch": 0.12, "grad_norm": 68.0, "learning_rate": 8.545762711864406e-07, "loss": 1.4828, "step": 4790 }, { "epoch": 0.12, "grad_norm": 65.5, "learning_rate": 8.542372881355931e-07, "loss": 1.4992, "step": 4800 }, { "epoch": 0.12, "grad_norm": 66.5, "learning_rate": 8.538983050847457e-07, "loss": 1.4546, "step": 4810 }, { "epoch": 0.12, "grad_norm": 65.5, "learning_rate": 8.535593220338983e-07, "loss": 1.4745, "step": 4820 }, { "epoch": 0.12, "grad_norm": 68.5, "learning_rate": 8.532203389830508e-07, "loss": 1.488, "step": 4830 }, { "epoch": 0.12, "grad_norm": 66.5, "learning_rate": 8.528813559322033e-07, "loss": 1.4607, "step": 4840 }, { "epoch": 0.12, "grad_norm": 67.0, "learning_rate": 8.525423728813558e-07, "loss": 1.4884, "step": 4850 }, { "epoch": 0.12, "grad_norm": 70.5, "learning_rate": 8.522033898305085e-07, "loss": 1.4927, "step": 4860 }, { "epoch": 0.12, "grad_norm": 66.0, "learning_rate": 8.51864406779661e-07, "loss": 1.498, "step": 4870 }, { "epoch": 0.12, "grad_norm": 69.5, "learning_rate": 8.515254237288135e-07, "loss": 1.4846, "step": 4880 }, { "epoch": 0.12, "grad_norm": 65.5, "learning_rate": 8.511864406779661e-07, "loss": 1.5013, "step": 4890 }, { "epoch": 0.12, "grad_norm": 68.0, "learning_rate": 8.508474576271186e-07, "loss": 1.4643, "step": 4900 }, { "epoch": 0.12, "grad_norm": 68.0, "learning_rate": 8.505084745762712e-07, "loss": 1.465, "step": 4910 }, { "epoch": 0.12, "grad_norm": 69.0, "learning_rate": 8.501694915254238e-07, "loss": 1.4796, "step": 4920 }, { "epoch": 0.12, "grad_norm": 64.5, "learning_rate": 8.498305084745763e-07, "loss": 1.4816, "step": 4930 }, { "epoch": 0.12, "grad_norm": 64.0, "learning_rate": 8.494915254237288e-07, "loss": 1.48, "step": 4940 }, { "epoch": 0.12, "grad_norm": 66.5, "learning_rate": 8.491525423728813e-07, "loss": 1.4097, "step": 4950 }, { "epoch": 0.12, "grad_norm": 66.5, "learning_rate": 8.488135593220339e-07, "loss": 1.488, "step": 4960 }, { "epoch": 0.12, "grad_norm": 66.5, "learning_rate": 8.484745762711864e-07, "loss": 1.4347, "step": 4970 }, { "epoch": 0.12, "grad_norm": 69.0, "learning_rate": 8.481355932203389e-07, "loss": 1.4901, "step": 4980 }, { "epoch": 0.12, "grad_norm": 63.0, "learning_rate": 8.477966101694914e-07, "loss": 1.471, "step": 4990 }, { "epoch": 0.12, "grad_norm": 65.5, "learning_rate": 8.47457627118644e-07, "loss": 1.423, "step": 5000 }, { "epoch": 0.12, "eval_loss": 1.4716249704360962, "eval_runtime": 68.0492, "eval_samples_per_second": 14.695, "eval_steps_per_second": 14.695, "step": 5000 }, { "epoch": 0.13, "grad_norm": 70.0, "learning_rate": 8.471186440677965e-07, "loss": 1.5264, "step": 5010 }, { "epoch": 0.13, "grad_norm": 70.0, "learning_rate": 8.467796610169491e-07, "loss": 1.4538, "step": 5020 }, { "epoch": 0.13, "grad_norm": 65.5, "learning_rate": 8.464406779661017e-07, "loss": 1.5095, "step": 5030 }, { "epoch": 0.13, "grad_norm": 67.5, "learning_rate": 8.461016949152542e-07, "loss": 1.4824, "step": 5040 }, { "epoch": 0.13, "grad_norm": 66.0, "learning_rate": 8.457627118644067e-07, "loss": 1.4512, "step": 5050 }, { "epoch": 0.13, "grad_norm": 71.5, "learning_rate": 8.454237288135592e-07, "loss": 1.5047, "step": 5060 }, { "epoch": 0.13, "grad_norm": 68.0, "learning_rate": 8.450847457627119e-07, "loss": 1.4905, "step": 5070 }, { "epoch": 0.13, "grad_norm": 65.0, "learning_rate": 8.447457627118644e-07, "loss": 1.438, "step": 5080 }, { "epoch": 0.13, "grad_norm": 62.75, "learning_rate": 8.444067796610169e-07, "loss": 1.488, "step": 5090 }, { "epoch": 0.13, "grad_norm": 70.5, "learning_rate": 8.440677966101695e-07, "loss": 1.4296, "step": 5100 }, { "epoch": 0.13, "grad_norm": 66.0, "learning_rate": 8.43728813559322e-07, "loss": 1.4613, "step": 5110 }, { "epoch": 0.13, "grad_norm": 66.0, "learning_rate": 8.433898305084745e-07, "loss": 1.458, "step": 5120 }, { "epoch": 0.13, "grad_norm": 67.5, "learning_rate": 8.430508474576271e-07, "loss": 1.4397, "step": 5130 }, { "epoch": 0.13, "grad_norm": 64.5, "learning_rate": 8.427118644067797e-07, "loss": 1.4432, "step": 5140 }, { "epoch": 0.13, "grad_norm": 63.75, "learning_rate": 8.423728813559322e-07, "loss": 1.4499, "step": 5150 }, { "epoch": 0.13, "grad_norm": 63.75, "learning_rate": 8.420338983050847e-07, "loss": 1.4305, "step": 5160 }, { "epoch": 0.13, "grad_norm": 68.0, "learning_rate": 8.416949152542373e-07, "loss": 1.4123, "step": 5170 }, { "epoch": 0.13, "grad_norm": 68.5, "learning_rate": 8.413559322033898e-07, "loss": 1.4079, "step": 5180 }, { "epoch": 0.13, "grad_norm": 68.5, "learning_rate": 8.410169491525423e-07, "loss": 1.5019, "step": 5190 }, { "epoch": 0.13, "grad_norm": 63.0, "learning_rate": 8.406779661016948e-07, "loss": 1.4572, "step": 5200 }, { "epoch": 0.13, "grad_norm": 66.5, "learning_rate": 8.403389830508474e-07, "loss": 1.4558, "step": 5210 }, { "epoch": 0.13, "grad_norm": 60.75, "learning_rate": 8.399999999999999e-07, "loss": 1.4013, "step": 5220 }, { "epoch": 0.13, "grad_norm": 69.0, "learning_rate": 8.396610169491524e-07, "loss": 1.4969, "step": 5230 }, { "epoch": 0.13, "grad_norm": 64.5, "learning_rate": 8.393220338983051e-07, "loss": 1.4688, "step": 5240 }, { "epoch": 0.13, "grad_norm": 67.0, "learning_rate": 8.389830508474576e-07, "loss": 1.5432, "step": 5250 }, { "epoch": 0.13, "grad_norm": 66.5, "learning_rate": 8.386440677966101e-07, "loss": 1.467, "step": 5260 }, { "epoch": 0.13, "grad_norm": 65.5, "learning_rate": 8.383050847457626e-07, "loss": 1.5081, "step": 5270 }, { "epoch": 0.13, "grad_norm": 63.25, "learning_rate": 8.379661016949152e-07, "loss": 1.4594, "step": 5280 }, { "epoch": 0.13, "grad_norm": 67.5, "learning_rate": 8.376271186440678e-07, "loss": 1.4835, "step": 5290 }, { "epoch": 0.13, "grad_norm": 67.5, "learning_rate": 8.372881355932203e-07, "loss": 1.4342, "step": 5300 }, { "epoch": 0.13, "grad_norm": 68.5, "learning_rate": 8.369491525423729e-07, "loss": 1.4906, "step": 5310 }, { "epoch": 0.13, "grad_norm": 66.5, "learning_rate": 8.366101694915254e-07, "loss": 1.4758, "step": 5320 }, { "epoch": 0.13, "grad_norm": 66.5, "learning_rate": 8.362711864406779e-07, "loss": 1.4938, "step": 5330 }, { "epoch": 0.13, "grad_norm": 70.5, "learning_rate": 8.359322033898305e-07, "loss": 1.4942, "step": 5340 }, { "epoch": 0.13, "grad_norm": 66.0, "learning_rate": 8.355932203389831e-07, "loss": 1.4614, "step": 5350 }, { "epoch": 0.13, "grad_norm": 66.0, "learning_rate": 8.352542372881356e-07, "loss": 1.4281, "step": 5360 }, { "epoch": 0.13, "grad_norm": 66.5, "learning_rate": 8.349152542372881e-07, "loss": 1.4212, "step": 5370 }, { "epoch": 0.13, "grad_norm": 74.0, "learning_rate": 8.345762711864407e-07, "loss": 1.5122, "step": 5380 }, { "epoch": 0.13, "grad_norm": 67.0, "learning_rate": 8.342372881355931e-07, "loss": 1.4601, "step": 5390 }, { "epoch": 0.14, "grad_norm": 60.25, "learning_rate": 8.338983050847457e-07, "loss": 1.4513, "step": 5400 }, { "epoch": 0.14, "grad_norm": 66.5, "learning_rate": 8.335593220338982e-07, "loss": 1.4691, "step": 5410 }, { "epoch": 0.14, "grad_norm": 68.5, "learning_rate": 8.332203389830508e-07, "loss": 1.4399, "step": 5420 }, { "epoch": 0.14, "grad_norm": 73.0, "learning_rate": 8.328813559322033e-07, "loss": 1.4794, "step": 5430 }, { "epoch": 0.14, "grad_norm": 66.5, "learning_rate": 8.325423728813558e-07, "loss": 1.4511, "step": 5440 }, { "epoch": 0.14, "grad_norm": 64.5, "learning_rate": 8.322033898305085e-07, "loss": 1.4648, "step": 5450 }, { "epoch": 0.14, "grad_norm": 67.5, "learning_rate": 8.31864406779661e-07, "loss": 1.4878, "step": 5460 }, { "epoch": 0.14, "grad_norm": 69.5, "learning_rate": 8.315254237288135e-07, "loss": 1.4823, "step": 5470 }, { "epoch": 0.14, "grad_norm": 69.5, "learning_rate": 8.31186440677966e-07, "loss": 1.4978, "step": 5480 }, { "epoch": 0.14, "grad_norm": 67.5, "learning_rate": 8.308474576271186e-07, "loss": 1.4107, "step": 5490 }, { "epoch": 0.14, "grad_norm": 65.0, "learning_rate": 8.305084745762712e-07, "loss": 1.4983, "step": 5500 }, { "epoch": 0.14, "grad_norm": 68.0, "learning_rate": 8.301694915254237e-07, "loss": 1.4869, "step": 5510 }, { "epoch": 0.14, "grad_norm": 69.5, "learning_rate": 8.298305084745763e-07, "loss": 1.499, "step": 5520 }, { "epoch": 0.14, "grad_norm": 65.0, "learning_rate": 8.294915254237288e-07, "loss": 1.4603, "step": 5530 }, { "epoch": 0.14, "grad_norm": 64.5, "learning_rate": 8.291525423728813e-07, "loss": 1.411, "step": 5540 }, { "epoch": 0.14, "grad_norm": 64.5, "learning_rate": 8.288135593220339e-07, "loss": 1.4198, "step": 5550 }, { "epoch": 0.14, "grad_norm": 62.75, "learning_rate": 8.284745762711865e-07, "loss": 1.4814, "step": 5560 }, { "epoch": 0.14, "grad_norm": 72.5, "learning_rate": 8.28135593220339e-07, "loss": 1.4894, "step": 5570 }, { "epoch": 0.14, "grad_norm": 70.5, "learning_rate": 8.277966101694915e-07, "loss": 1.4908, "step": 5580 }, { "epoch": 0.14, "grad_norm": 66.0, "learning_rate": 8.27457627118644e-07, "loss": 1.4304, "step": 5590 }, { "epoch": 0.14, "grad_norm": 66.0, "learning_rate": 8.271186440677965e-07, "loss": 1.4504, "step": 5600 }, { "epoch": 0.14, "grad_norm": 66.5, "learning_rate": 8.267796610169491e-07, "loss": 1.4952, "step": 5610 }, { "epoch": 0.14, "grad_norm": 65.5, "learning_rate": 8.264406779661016e-07, "loss": 1.453, "step": 5620 }, { "epoch": 0.14, "grad_norm": 66.5, "learning_rate": 8.261016949152542e-07, "loss": 1.4776, "step": 5630 }, { "epoch": 0.14, "grad_norm": 67.5, "learning_rate": 8.257627118644067e-07, "loss": 1.4135, "step": 5640 }, { "epoch": 0.14, "grad_norm": 70.5, "learning_rate": 8.254237288135592e-07, "loss": 1.4329, "step": 5650 }, { "epoch": 0.14, "grad_norm": 67.0, "learning_rate": 8.250847457627119e-07, "loss": 1.3881, "step": 5660 }, { "epoch": 0.14, "grad_norm": 66.0, "learning_rate": 8.247457627118644e-07, "loss": 1.4916, "step": 5670 }, { "epoch": 0.14, "grad_norm": 67.0, "learning_rate": 8.244067796610169e-07, "loss": 1.4505, "step": 5680 }, { "epoch": 0.14, "grad_norm": 63.75, "learning_rate": 8.240677966101694e-07, "loss": 1.5199, "step": 5690 }, { "epoch": 0.14, "grad_norm": 65.5, "learning_rate": 8.23728813559322e-07, "loss": 1.476, "step": 5700 }, { "epoch": 0.14, "grad_norm": 66.0, "learning_rate": 8.233898305084745e-07, "loss": 1.5154, "step": 5710 }, { "epoch": 0.14, "grad_norm": 67.0, "learning_rate": 8.230508474576271e-07, "loss": 1.4812, "step": 5720 }, { "epoch": 0.14, "grad_norm": 67.5, "learning_rate": 8.227118644067797e-07, "loss": 1.4933, "step": 5730 }, { "epoch": 0.14, "grad_norm": 67.0, "learning_rate": 8.223728813559322e-07, "loss": 1.425, "step": 5740 }, { "epoch": 0.14, "grad_norm": 67.5, "learning_rate": 8.220338983050847e-07, "loss": 1.4669, "step": 5750 }, { "epoch": 0.14, "grad_norm": 67.0, "learning_rate": 8.216949152542372e-07, "loss": 1.4557, "step": 5760 }, { "epoch": 0.14, "grad_norm": 66.5, "learning_rate": 8.213559322033899e-07, "loss": 1.4626, "step": 5770 }, { "epoch": 0.14, "grad_norm": 67.5, "learning_rate": 8.210169491525424e-07, "loss": 1.477, "step": 5780 }, { "epoch": 0.14, "grad_norm": 68.0, "learning_rate": 8.206779661016949e-07, "loss": 1.4195, "step": 5790 }, { "epoch": 0.14, "grad_norm": 69.0, "learning_rate": 8.203389830508474e-07, "loss": 1.4738, "step": 5800 }, { "epoch": 0.15, "grad_norm": 68.5, "learning_rate": 8.199999999999999e-07, "loss": 1.5056, "step": 5810 }, { "epoch": 0.15, "grad_norm": 71.0, "learning_rate": 8.196610169491525e-07, "loss": 1.5026, "step": 5820 }, { "epoch": 0.15, "grad_norm": 67.5, "learning_rate": 8.19322033898305e-07, "loss": 1.4325, "step": 5830 }, { "epoch": 0.15, "grad_norm": 71.0, "learning_rate": 8.189830508474576e-07, "loss": 1.4919, "step": 5840 }, { "epoch": 0.15, "grad_norm": 71.5, "learning_rate": 8.186440677966101e-07, "loss": 1.4314, "step": 5850 }, { "epoch": 0.15, "grad_norm": 66.5, "learning_rate": 8.183050847457626e-07, "loss": 1.4759, "step": 5860 }, { "epoch": 0.15, "grad_norm": 64.5, "learning_rate": 8.179661016949152e-07, "loss": 1.4543, "step": 5870 }, { "epoch": 0.15, "grad_norm": 68.0, "learning_rate": 8.176271186440678e-07, "loss": 1.5114, "step": 5880 }, { "epoch": 0.15, "grad_norm": 71.0, "learning_rate": 8.172881355932203e-07, "loss": 1.5202, "step": 5890 }, { "epoch": 0.15, "grad_norm": 65.5, "learning_rate": 8.169491525423728e-07, "loss": 1.4694, "step": 5900 }, { "epoch": 0.15, "grad_norm": 65.5, "learning_rate": 8.166101694915254e-07, "loss": 1.51, "step": 5910 }, { "epoch": 0.15, "grad_norm": 63.5, "learning_rate": 8.162711864406779e-07, "loss": 1.4607, "step": 5920 }, { "epoch": 0.15, "grad_norm": 66.0, "learning_rate": 8.159322033898305e-07, "loss": 1.5022, "step": 5930 }, { "epoch": 0.15, "grad_norm": 67.5, "learning_rate": 8.155932203389831e-07, "loss": 1.4228, "step": 5940 }, { "epoch": 0.15, "grad_norm": 66.5, "learning_rate": 8.152542372881356e-07, "loss": 1.4538, "step": 5950 }, { "epoch": 0.15, "grad_norm": 65.5, "learning_rate": 8.149152542372881e-07, "loss": 1.4295, "step": 5960 }, { "epoch": 0.15, "grad_norm": 69.0, "learning_rate": 8.145762711864406e-07, "loss": 1.4724, "step": 5970 }, { "epoch": 0.15, "grad_norm": 67.5, "learning_rate": 8.142372881355933e-07, "loss": 1.488, "step": 5980 }, { "epoch": 0.15, "grad_norm": 65.0, "learning_rate": 8.138983050847458e-07, "loss": 1.4538, "step": 5990 }, { "epoch": 0.15, "grad_norm": 66.0, "learning_rate": 8.135593220338983e-07, "loss": 1.4961, "step": 6000 }, { "epoch": 0.15, "grad_norm": 69.5, "learning_rate": 8.132203389830507e-07, "loss": 1.4509, "step": 6010 }, { "epoch": 0.15, "grad_norm": 62.25, "learning_rate": 8.128813559322033e-07, "loss": 1.4322, "step": 6020 }, { "epoch": 0.15, "grad_norm": 66.0, "learning_rate": 8.125423728813558e-07, "loss": 1.438, "step": 6030 }, { "epoch": 0.15, "grad_norm": 68.5, "learning_rate": 8.122033898305084e-07, "loss": 1.5195, "step": 6040 }, { "epoch": 0.15, "grad_norm": 69.0, "learning_rate": 8.11864406779661e-07, "loss": 1.3999, "step": 6050 }, { "epoch": 0.15, "grad_norm": 66.0, "learning_rate": 8.115254237288135e-07, "loss": 1.4916, "step": 6060 }, { "epoch": 0.15, "grad_norm": 65.5, "learning_rate": 8.11186440677966e-07, "loss": 1.4528, "step": 6070 }, { "epoch": 0.15, "grad_norm": 66.0, "learning_rate": 8.108474576271185e-07, "loss": 1.5167, "step": 6080 }, { "epoch": 0.15, "grad_norm": 64.0, "learning_rate": 8.105084745762712e-07, "loss": 1.4408, "step": 6090 }, { "epoch": 0.15, "grad_norm": 66.5, "learning_rate": 8.101694915254237e-07, "loss": 1.4306, "step": 6100 }, { "epoch": 0.15, "grad_norm": 64.0, "learning_rate": 8.098305084745762e-07, "loss": 1.4774, "step": 6110 }, { "epoch": 0.15, "grad_norm": 69.0, "learning_rate": 8.094915254237288e-07, "loss": 1.4728, "step": 6120 }, { "epoch": 0.15, "grad_norm": 68.0, "learning_rate": 8.091525423728813e-07, "loss": 1.4349, "step": 6130 }, { "epoch": 0.15, "grad_norm": 68.0, "learning_rate": 8.088135593220339e-07, "loss": 1.4466, "step": 6140 }, { "epoch": 0.15, "grad_norm": 69.5, "learning_rate": 8.084745762711865e-07, "loss": 1.4641, "step": 6150 }, { "epoch": 0.15, "grad_norm": 69.5, "learning_rate": 8.08135593220339e-07, "loss": 1.5069, "step": 6160 }, { "epoch": 0.15, "grad_norm": 67.5, "learning_rate": 8.077966101694915e-07, "loss": 1.5108, "step": 6170 }, { "epoch": 0.15, "grad_norm": 66.0, "learning_rate": 8.07457627118644e-07, "loss": 1.45, "step": 6180 }, { "epoch": 0.15, "grad_norm": 66.5, "learning_rate": 8.071186440677966e-07, "loss": 1.4731, "step": 6190 }, { "epoch": 0.15, "grad_norm": 70.0, "learning_rate": 8.067796610169492e-07, "loss": 1.5156, "step": 6200 }, { "epoch": 0.16, "grad_norm": 66.5, "learning_rate": 8.064406779661017e-07, "loss": 1.4864, "step": 6210 }, { "epoch": 0.16, "grad_norm": 68.5, "learning_rate": 8.061016949152541e-07, "loss": 1.5115, "step": 6220 }, { "epoch": 0.16, "grad_norm": 71.0, "learning_rate": 8.057627118644067e-07, "loss": 1.4871, "step": 6230 }, { "epoch": 0.16, "grad_norm": 67.0, "learning_rate": 8.054237288135592e-07, "loss": 1.4302, "step": 6240 }, { "epoch": 0.16, "grad_norm": 63.5, "learning_rate": 8.050847457627118e-07, "loss": 1.4427, "step": 6250 }, { "epoch": 0.16, "grad_norm": 63.75, "learning_rate": 8.047457627118644e-07, "loss": 1.479, "step": 6260 }, { "epoch": 0.16, "grad_norm": 66.5, "learning_rate": 8.044067796610169e-07, "loss": 1.4658, "step": 6270 }, { "epoch": 0.16, "grad_norm": 67.5, "learning_rate": 8.040677966101694e-07, "loss": 1.3803, "step": 6280 }, { "epoch": 0.16, "grad_norm": 67.5, "learning_rate": 8.03728813559322e-07, "loss": 1.4912, "step": 6290 }, { "epoch": 0.16, "grad_norm": 64.5, "learning_rate": 8.033898305084746e-07, "loss": 1.4588, "step": 6300 }, { "epoch": 0.16, "grad_norm": 65.0, "learning_rate": 8.030508474576271e-07, "loss": 1.4416, "step": 6310 }, { "epoch": 0.16, "grad_norm": 66.0, "learning_rate": 8.027118644067796e-07, "loss": 1.4303, "step": 6320 }, { "epoch": 0.16, "grad_norm": 66.0, "learning_rate": 8.023728813559322e-07, "loss": 1.4604, "step": 6330 }, { "epoch": 0.16, "grad_norm": 66.5, "learning_rate": 8.020338983050847e-07, "loss": 1.4369, "step": 6340 }, { "epoch": 0.16, "grad_norm": 64.5, "learning_rate": 8.016949152542372e-07, "loss": 1.4253, "step": 6350 }, { "epoch": 0.16, "grad_norm": 68.0, "learning_rate": 8.013559322033899e-07, "loss": 1.4252, "step": 6360 }, { "epoch": 0.16, "grad_norm": 64.0, "learning_rate": 8.010169491525424e-07, "loss": 1.4328, "step": 6370 }, { "epoch": 0.16, "grad_norm": 66.5, "learning_rate": 8.006779661016949e-07, "loss": 1.4616, "step": 6380 }, { "epoch": 0.16, "grad_norm": 64.0, "learning_rate": 8.003389830508474e-07, "loss": 1.4362, "step": 6390 }, { "epoch": 0.16, "grad_norm": 66.5, "learning_rate": 8e-07, "loss": 1.4958, "step": 6400 }, { "epoch": 0.16, "grad_norm": 64.5, "learning_rate": 7.996610169491526e-07, "loss": 1.4978, "step": 6410 }, { "epoch": 0.16, "grad_norm": 65.0, "learning_rate": 7.993220338983051e-07, "loss": 1.4799, "step": 6420 }, { "epoch": 0.16, "grad_norm": 65.5, "learning_rate": 7.989830508474575e-07, "loss": 1.502, "step": 6430 }, { "epoch": 0.16, "grad_norm": 64.5, "learning_rate": 7.986440677966101e-07, "loss": 1.4575, "step": 6440 }, { "epoch": 0.16, "grad_norm": 67.5, "learning_rate": 7.983050847457626e-07, "loss": 1.4848, "step": 6450 }, { "epoch": 0.16, "grad_norm": 65.5, "learning_rate": 7.979661016949152e-07, "loss": 1.428, "step": 6460 }, { "epoch": 0.16, "grad_norm": 67.5, "learning_rate": 7.976271186440678e-07, "loss": 1.4416, "step": 6470 }, { "epoch": 0.16, "grad_norm": 67.0, "learning_rate": 7.972881355932203e-07, "loss": 1.5052, "step": 6480 }, { "epoch": 0.16, "grad_norm": 69.5, "learning_rate": 7.969491525423728e-07, "loss": 1.4919, "step": 6490 }, { "epoch": 0.16, "grad_norm": 69.0, "learning_rate": 7.966101694915253e-07, "loss": 1.4309, "step": 6500 }, { "epoch": 0.16, "grad_norm": 67.0, "learning_rate": 7.962711864406779e-07, "loss": 1.4807, "step": 6510 }, { "epoch": 0.16, "grad_norm": 65.0, "learning_rate": 7.959322033898305e-07, "loss": 1.4638, "step": 6520 }, { "epoch": 0.16, "grad_norm": 69.5, "learning_rate": 7.95593220338983e-07, "loss": 1.4854, "step": 6530 }, { "epoch": 0.16, "grad_norm": 65.0, "learning_rate": 7.952542372881356e-07, "loss": 1.3914, "step": 6540 }, { "epoch": 0.16, "grad_norm": 70.5, "learning_rate": 7.949152542372881e-07, "loss": 1.484, "step": 6550 }, { "epoch": 0.16, "grad_norm": 69.5, "learning_rate": 7.945762711864406e-07, "loss": 1.4798, "step": 6560 }, { "epoch": 0.16, "grad_norm": 67.0, "learning_rate": 7.942372881355933e-07, "loss": 1.4472, "step": 6570 }, { "epoch": 0.16, "grad_norm": 64.5, "learning_rate": 7.938983050847458e-07, "loss": 1.4862, "step": 6580 }, { "epoch": 0.16, "grad_norm": 66.5, "learning_rate": 7.935593220338983e-07, "loss": 1.5081, "step": 6590 }, { "epoch": 0.17, "grad_norm": 66.0, "learning_rate": 7.932203389830508e-07, "loss": 1.467, "step": 6600 }, { "epoch": 0.17, "grad_norm": 65.5, "learning_rate": 7.928813559322034e-07, "loss": 1.4519, "step": 6610 }, { "epoch": 0.17, "grad_norm": 66.5, "learning_rate": 7.92542372881356e-07, "loss": 1.5176, "step": 6620 }, { "epoch": 0.17, "grad_norm": 66.0, "learning_rate": 7.922033898305084e-07, "loss": 1.4862, "step": 6630 }, { "epoch": 0.17, "grad_norm": 65.5, "learning_rate": 7.918644067796609e-07, "loss": 1.4696, "step": 6640 }, { "epoch": 0.17, "grad_norm": 67.5, "learning_rate": 7.915254237288135e-07, "loss": 1.3946, "step": 6650 }, { "epoch": 0.17, "grad_norm": 64.5, "learning_rate": 7.91186440677966e-07, "loss": 1.4382, "step": 6660 }, { "epoch": 0.17, "grad_norm": 65.5, "learning_rate": 7.908474576271185e-07, "loss": 1.4893, "step": 6670 }, { "epoch": 0.17, "grad_norm": 68.5, "learning_rate": 7.905084745762712e-07, "loss": 1.4708, "step": 6680 }, { "epoch": 0.17, "grad_norm": 67.5, "learning_rate": 7.901694915254237e-07, "loss": 1.4692, "step": 6690 }, { "epoch": 0.17, "grad_norm": 68.0, "learning_rate": 7.898305084745762e-07, "loss": 1.4824, "step": 6700 }, { "epoch": 0.17, "grad_norm": 67.5, "learning_rate": 7.894915254237287e-07, "loss": 1.4448, "step": 6710 }, { "epoch": 0.17, "grad_norm": 67.0, "learning_rate": 7.891525423728813e-07, "loss": 1.4395, "step": 6720 }, { "epoch": 0.17, "grad_norm": 72.0, "learning_rate": 7.888135593220339e-07, "loss": 1.4705, "step": 6730 }, { "epoch": 0.17, "grad_norm": 65.5, "learning_rate": 7.884745762711864e-07, "loss": 1.4924, "step": 6740 }, { "epoch": 0.17, "grad_norm": 67.0, "learning_rate": 7.88135593220339e-07, "loss": 1.4965, "step": 6750 }, { "epoch": 0.17, "grad_norm": 63.25, "learning_rate": 7.877966101694915e-07, "loss": 1.4817, "step": 6760 }, { "epoch": 0.17, "grad_norm": 65.0, "learning_rate": 7.87457627118644e-07, "loss": 1.5155, "step": 6770 }, { "epoch": 0.17, "grad_norm": 69.5, "learning_rate": 7.871186440677967e-07, "loss": 1.4577, "step": 6780 }, { "epoch": 0.17, "grad_norm": 70.5, "learning_rate": 7.867796610169492e-07, "loss": 1.5126, "step": 6790 }, { "epoch": 0.17, "grad_norm": 68.0, "learning_rate": 7.864406779661017e-07, "loss": 1.4374, "step": 6800 }, { "epoch": 0.17, "grad_norm": 66.5, "learning_rate": 7.861016949152542e-07, "loss": 1.4364, "step": 6810 }, { "epoch": 0.17, "grad_norm": 65.0, "learning_rate": 7.857627118644068e-07, "loss": 1.4468, "step": 6820 }, { "epoch": 0.17, "grad_norm": 66.5, "learning_rate": 7.854237288135593e-07, "loss": 1.4643, "step": 6830 }, { "epoch": 0.17, "grad_norm": 67.0, "learning_rate": 7.850847457627118e-07, "loss": 1.4105, "step": 6840 }, { "epoch": 0.17, "grad_norm": 66.0, "learning_rate": 7.847457627118643e-07, "loss": 1.4437, "step": 6850 }, { "epoch": 0.17, "grad_norm": 67.0, "learning_rate": 7.844067796610169e-07, "loss": 1.4737, "step": 6860 }, { "epoch": 0.17, "grad_norm": 64.5, "learning_rate": 7.840677966101694e-07, "loss": 1.5024, "step": 6870 }, { "epoch": 0.17, "grad_norm": 64.5, "learning_rate": 7.837288135593219e-07, "loss": 1.4301, "step": 6880 }, { "epoch": 0.17, "grad_norm": 63.5, "learning_rate": 7.833898305084746e-07, "loss": 1.4815, "step": 6890 }, { "epoch": 0.17, "grad_norm": 66.0, "learning_rate": 7.830508474576271e-07, "loss": 1.4301, "step": 6900 }, { "epoch": 0.17, "grad_norm": 64.5, "learning_rate": 7.827118644067796e-07, "loss": 1.4483, "step": 6910 }, { "epoch": 0.17, "grad_norm": 65.0, "learning_rate": 7.823728813559321e-07, "loss": 1.4424, "step": 6920 }, { "epoch": 0.17, "grad_norm": 64.0, "learning_rate": 7.820338983050847e-07, "loss": 1.4913, "step": 6930 }, { "epoch": 0.17, "grad_norm": 64.5, "learning_rate": 7.816949152542373e-07, "loss": 1.4186, "step": 6940 }, { "epoch": 0.17, "grad_norm": 67.0, "learning_rate": 7.813559322033898e-07, "loss": 1.4123, "step": 6950 }, { "epoch": 0.17, "grad_norm": 65.5, "learning_rate": 7.810169491525424e-07, "loss": 1.4426, "step": 6960 }, { "epoch": 0.17, "grad_norm": 67.0, "learning_rate": 7.806779661016949e-07, "loss": 1.4486, "step": 6970 }, { "epoch": 0.17, "grad_norm": 68.0, "learning_rate": 7.803389830508474e-07, "loss": 1.4525, "step": 6980 }, { "epoch": 0.17, "grad_norm": 64.5, "learning_rate": 7.799999999999999e-07, "loss": 1.3955, "step": 6990 }, { "epoch": 0.17, "grad_norm": 65.0, "learning_rate": 7.796610169491526e-07, "loss": 1.4489, "step": 7000 }, { "epoch": 0.18, "grad_norm": 67.0, "learning_rate": 7.793220338983051e-07, "loss": 1.434, "step": 7010 }, { "epoch": 0.18, "grad_norm": 66.0, "learning_rate": 7.789830508474576e-07, "loss": 1.48, "step": 7020 }, { "epoch": 0.18, "grad_norm": 67.0, "learning_rate": 7.786440677966102e-07, "loss": 1.4567, "step": 7030 }, { "epoch": 0.18, "grad_norm": 69.0, "learning_rate": 7.783050847457627e-07, "loss": 1.5248, "step": 7040 }, { "epoch": 0.18, "grad_norm": 68.0, "learning_rate": 7.779661016949152e-07, "loss": 1.4615, "step": 7050 }, { "epoch": 0.18, "grad_norm": 62.5, "learning_rate": 7.776271186440677e-07, "loss": 1.4386, "step": 7060 }, { "epoch": 0.18, "grad_norm": 67.5, "learning_rate": 7.772881355932203e-07, "loss": 1.4156, "step": 7070 }, { "epoch": 0.18, "grad_norm": 69.0, "learning_rate": 7.769491525423728e-07, "loss": 1.4598, "step": 7080 }, { "epoch": 0.18, "grad_norm": 63.75, "learning_rate": 7.766101694915253e-07, "loss": 1.3977, "step": 7090 }, { "epoch": 0.18, "grad_norm": 68.0, "learning_rate": 7.762711864406779e-07, "loss": 1.461, "step": 7100 }, { "epoch": 0.18, "grad_norm": 67.0, "learning_rate": 7.759322033898305e-07, "loss": 1.4649, "step": 7110 }, { "epoch": 0.18, "grad_norm": 65.0, "learning_rate": 7.75593220338983e-07, "loss": 1.4818, "step": 7120 }, { "epoch": 0.18, "grad_norm": 63.75, "learning_rate": 7.752542372881355e-07, "loss": 1.4182, "step": 7130 }, { "epoch": 0.18, "grad_norm": 67.0, "learning_rate": 7.749152542372881e-07, "loss": 1.4647, "step": 7140 }, { "epoch": 0.18, "grad_norm": 66.5, "learning_rate": 7.745762711864406e-07, "loss": 1.4688, "step": 7150 }, { "epoch": 0.18, "grad_norm": 68.0, "learning_rate": 7.742372881355932e-07, "loss": 1.4222, "step": 7160 }, { "epoch": 0.18, "grad_norm": 66.0, "learning_rate": 7.738983050847458e-07, "loss": 1.4743, "step": 7170 }, { "epoch": 0.18, "grad_norm": 67.0, "learning_rate": 7.735593220338983e-07, "loss": 1.4717, "step": 7180 }, { "epoch": 0.18, "grad_norm": 63.0, "learning_rate": 7.732203389830508e-07, "loss": 1.4649, "step": 7190 }, { "epoch": 0.18, "grad_norm": 67.0, "learning_rate": 7.728813559322033e-07, "loss": 1.4846, "step": 7200 }, { "epoch": 0.18, "grad_norm": 66.0, "learning_rate": 7.72542372881356e-07, "loss": 1.4857, "step": 7210 }, { "epoch": 0.18, "grad_norm": 67.5, "learning_rate": 7.722033898305085e-07, "loss": 1.4876, "step": 7220 }, { "epoch": 0.18, "grad_norm": 65.5, "learning_rate": 7.71864406779661e-07, "loss": 1.4131, "step": 7230 }, { "epoch": 0.18, "grad_norm": 69.0, "learning_rate": 7.715254237288136e-07, "loss": 1.4857, "step": 7240 }, { "epoch": 0.18, "grad_norm": 68.0, "learning_rate": 7.711864406779661e-07, "loss": 1.4667, "step": 7250 }, { "epoch": 0.18, "grad_norm": 67.5, "learning_rate": 7.708474576271185e-07, "loss": 1.4476, "step": 7260 }, { "epoch": 0.18, "grad_norm": 64.5, "learning_rate": 7.705084745762711e-07, "loss": 1.4407, "step": 7270 }, { "epoch": 0.18, "grad_norm": 66.5, "learning_rate": 7.701694915254237e-07, "loss": 1.4836, "step": 7280 }, { "epoch": 0.18, "grad_norm": 65.5, "learning_rate": 7.698305084745762e-07, "loss": 1.4366, "step": 7290 }, { "epoch": 0.18, "grad_norm": 72.0, "learning_rate": 7.694915254237287e-07, "loss": 1.5036, "step": 7300 }, { "epoch": 0.18, "grad_norm": 66.5, "learning_rate": 7.691525423728813e-07, "loss": 1.4604, "step": 7310 }, { "epoch": 0.18, "grad_norm": 69.0, "learning_rate": 7.688135593220339e-07, "loss": 1.5105, "step": 7320 }, { "epoch": 0.18, "grad_norm": 64.5, "learning_rate": 7.684745762711864e-07, "loss": 1.4536, "step": 7330 }, { "epoch": 0.18, "grad_norm": 63.5, "learning_rate": 7.681355932203389e-07, "loss": 1.4388, "step": 7340 }, { "epoch": 0.18, "grad_norm": 68.5, "learning_rate": 7.677966101694915e-07, "loss": 1.4629, "step": 7350 }, { "epoch": 0.18, "grad_norm": 67.5, "learning_rate": 7.67457627118644e-07, "loss": 1.4909, "step": 7360 }, { "epoch": 0.18, "grad_norm": 65.5, "learning_rate": 7.671186440677966e-07, "loss": 1.462, "step": 7370 }, { "epoch": 0.18, "grad_norm": 64.5, "learning_rate": 7.667796610169492e-07, "loss": 1.4207, "step": 7380 }, { "epoch": 0.18, "grad_norm": 64.5, "learning_rate": 7.664406779661017e-07, "loss": 1.45, "step": 7390 }, { "epoch": 0.18, "grad_norm": 69.5, "learning_rate": 7.661016949152542e-07, "loss": 1.4954, "step": 7400 }, { "epoch": 0.19, "grad_norm": 65.5, "learning_rate": 7.657627118644067e-07, "loss": 1.4119, "step": 7410 }, { "epoch": 0.19, "grad_norm": 67.5, "learning_rate": 7.654237288135594e-07, "loss": 1.4738, "step": 7420 }, { "epoch": 0.19, "grad_norm": 66.0, "learning_rate": 7.650847457627119e-07, "loss": 1.4676, "step": 7430 }, { "epoch": 0.19, "grad_norm": 66.0, "learning_rate": 7.647457627118644e-07, "loss": 1.4551, "step": 7440 }, { "epoch": 0.19, "grad_norm": 70.0, "learning_rate": 7.64406779661017e-07, "loss": 1.4752, "step": 7450 }, { "epoch": 0.19, "grad_norm": 68.5, "learning_rate": 7.640677966101694e-07, "loss": 1.4917, "step": 7460 }, { "epoch": 0.19, "grad_norm": 68.5, "learning_rate": 7.637288135593219e-07, "loss": 1.4664, "step": 7470 }, { "epoch": 0.19, "grad_norm": 65.0, "learning_rate": 7.633898305084745e-07, "loss": 1.4719, "step": 7480 }, { "epoch": 0.19, "grad_norm": 70.0, "learning_rate": 7.630508474576271e-07, "loss": 1.4446, "step": 7490 }, { "epoch": 0.19, "grad_norm": 68.5, "learning_rate": 7.627118644067796e-07, "loss": 1.4462, "step": 7500 }, { "epoch": 0.19, "grad_norm": 73.5, "learning_rate": 7.623728813559321e-07, "loss": 1.5, "step": 7510 }, { "epoch": 0.19, "grad_norm": 67.5, "learning_rate": 7.620338983050847e-07, "loss": 1.5078, "step": 7520 }, { "epoch": 0.19, "grad_norm": 67.5, "learning_rate": 7.616949152542373e-07, "loss": 1.5165, "step": 7530 }, { "epoch": 0.19, "grad_norm": 66.0, "learning_rate": 7.613559322033898e-07, "loss": 1.4296, "step": 7540 }, { "epoch": 0.19, "grad_norm": 66.5, "learning_rate": 7.610169491525423e-07, "loss": 1.4336, "step": 7550 }, { "epoch": 0.19, "grad_norm": 65.0, "learning_rate": 7.606779661016949e-07, "loss": 1.4855, "step": 7560 }, { "epoch": 0.19, "grad_norm": 68.0, "learning_rate": 7.603389830508474e-07, "loss": 1.4852, "step": 7570 }, { "epoch": 0.19, "grad_norm": 69.5, "learning_rate": 7.599999999999999e-07, "loss": 1.5054, "step": 7580 }, { "epoch": 0.19, "grad_norm": 64.5, "learning_rate": 7.596610169491526e-07, "loss": 1.3943, "step": 7590 }, { "epoch": 0.19, "grad_norm": 68.5, "learning_rate": 7.593220338983051e-07, "loss": 1.4646, "step": 7600 }, { "epoch": 0.19, "grad_norm": 69.5, "learning_rate": 7.589830508474576e-07, "loss": 1.4634, "step": 7610 }, { "epoch": 0.19, "grad_norm": 66.5, "learning_rate": 7.586440677966101e-07, "loss": 1.5062, "step": 7620 }, { "epoch": 0.19, "grad_norm": 67.0, "learning_rate": 7.583050847457627e-07, "loss": 1.4898, "step": 7630 }, { "epoch": 0.19, "grad_norm": 71.0, "learning_rate": 7.579661016949153e-07, "loss": 1.44, "step": 7640 }, { "epoch": 0.19, "grad_norm": 68.5, "learning_rate": 7.576271186440678e-07, "loss": 1.4501, "step": 7650 }, { "epoch": 0.19, "grad_norm": 64.5, "learning_rate": 7.572881355932204e-07, "loss": 1.481, "step": 7660 }, { "epoch": 0.19, "grad_norm": 67.0, "learning_rate": 7.569491525423728e-07, "loss": 1.4384, "step": 7670 }, { "epoch": 0.19, "grad_norm": 65.0, "learning_rate": 7.566101694915253e-07, "loss": 1.4761, "step": 7680 }, { "epoch": 0.19, "grad_norm": 67.0, "learning_rate": 7.562711864406779e-07, "loss": 1.4658, "step": 7690 }, { "epoch": 0.19, "grad_norm": 66.0, "learning_rate": 7.559322033898305e-07, "loss": 1.4365, "step": 7700 }, { "epoch": 0.19, "grad_norm": 66.0, "learning_rate": 7.55593220338983e-07, "loss": 1.48, "step": 7710 }, { "epoch": 0.19, "grad_norm": 68.0, "learning_rate": 7.552542372881355e-07, "loss": 1.5133, "step": 7720 }, { "epoch": 0.19, "grad_norm": 70.0, "learning_rate": 7.54915254237288e-07, "loss": 1.4544, "step": 7730 }, { "epoch": 0.19, "grad_norm": 64.5, "learning_rate": 7.545762711864406e-07, "loss": 1.4914, "step": 7740 }, { "epoch": 0.19, "grad_norm": 69.0, "learning_rate": 7.542372881355932e-07, "loss": 1.4257, "step": 7750 }, { "epoch": 0.19, "grad_norm": 66.5, "learning_rate": 7.538983050847457e-07, "loss": 1.5068, "step": 7760 }, { "epoch": 0.19, "grad_norm": 68.5, "learning_rate": 7.535593220338983e-07, "loss": 1.4756, "step": 7770 }, { "epoch": 0.19, "grad_norm": 67.5, "learning_rate": 7.532203389830508e-07, "loss": 1.4537, "step": 7780 }, { "epoch": 0.19, "grad_norm": 63.5, "learning_rate": 7.528813559322033e-07, "loss": 1.4739, "step": 7790 }, { "epoch": 0.2, "grad_norm": 64.0, "learning_rate": 7.52542372881356e-07, "loss": 1.439, "step": 7800 }, { "epoch": 0.2, "grad_norm": 64.5, "learning_rate": 7.522033898305085e-07, "loss": 1.4954, "step": 7810 }, { "epoch": 0.2, "grad_norm": 65.0, "learning_rate": 7.51864406779661e-07, "loss": 1.4765, "step": 7820 }, { "epoch": 0.2, "grad_norm": 65.0, "learning_rate": 7.515254237288135e-07, "loss": 1.4639, "step": 7830 }, { "epoch": 0.2, "grad_norm": 67.5, "learning_rate": 7.511864406779661e-07, "loss": 1.5038, "step": 7840 }, { "epoch": 0.2, "grad_norm": 67.0, "learning_rate": 7.508474576271187e-07, "loss": 1.4688, "step": 7850 }, { "epoch": 0.2, "grad_norm": 69.5, "learning_rate": 7.505084745762712e-07, "loss": 1.4774, "step": 7860 }, { "epoch": 0.2, "grad_norm": 68.0, "learning_rate": 7.501694915254238e-07, "loss": 1.4253, "step": 7870 }, { "epoch": 0.2, "grad_norm": 66.0, "learning_rate": 7.498305084745762e-07, "loss": 1.4937, "step": 7880 }, { "epoch": 0.2, "grad_norm": 67.0, "learning_rate": 7.494915254237287e-07, "loss": 1.4684, "step": 7890 }, { "epoch": 0.2, "grad_norm": 72.5, "learning_rate": 7.491525423728812e-07, "loss": 1.5098, "step": 7900 }, { "epoch": 0.2, "grad_norm": 64.5, "learning_rate": 7.488135593220339e-07, "loss": 1.4807, "step": 7910 }, { "epoch": 0.2, "grad_norm": 67.5, "learning_rate": 7.484745762711864e-07, "loss": 1.4842, "step": 7920 }, { "epoch": 0.2, "grad_norm": 68.0, "learning_rate": 7.481355932203389e-07, "loss": 1.4593, "step": 7930 }, { "epoch": 0.2, "grad_norm": 66.0, "learning_rate": 7.477966101694914e-07, "loss": 1.5091, "step": 7940 }, { "epoch": 0.2, "grad_norm": 66.5, "learning_rate": 7.47457627118644e-07, "loss": 1.5042, "step": 7950 }, { "epoch": 0.2, "grad_norm": 65.0, "learning_rate": 7.471186440677966e-07, "loss": 1.4542, "step": 7960 }, { "epoch": 0.2, "grad_norm": 66.5, "learning_rate": 7.467796610169491e-07, "loss": 1.4756, "step": 7970 }, { "epoch": 0.2, "grad_norm": 62.25, "learning_rate": 7.464406779661017e-07, "loss": 1.4254, "step": 7980 }, { "epoch": 0.2, "grad_norm": 66.0, "learning_rate": 7.461016949152542e-07, "loss": 1.5089, "step": 7990 }, { "epoch": 0.2, "grad_norm": 68.0, "learning_rate": 7.457627118644067e-07, "loss": 1.4748, "step": 8000 }, { "epoch": 0.2, "grad_norm": 62.5, "learning_rate": 7.454237288135594e-07, "loss": 1.4041, "step": 8010 }, { "epoch": 0.2, "grad_norm": 65.5, "learning_rate": 7.450847457627119e-07, "loss": 1.437, "step": 8020 }, { "epoch": 0.2, "grad_norm": 68.0, "learning_rate": 7.447457627118644e-07, "loss": 1.4915, "step": 8030 }, { "epoch": 0.2, "grad_norm": 68.5, "learning_rate": 7.444067796610169e-07, "loss": 1.4956, "step": 8040 }, { "epoch": 0.2, "grad_norm": 67.5, "learning_rate": 7.440677966101695e-07, "loss": 1.5155, "step": 8050 }, { "epoch": 0.2, "grad_norm": 67.5, "learning_rate": 7.43728813559322e-07, "loss": 1.4615, "step": 8060 }, { "epoch": 0.2, "grad_norm": 65.0, "learning_rate": 7.433898305084746e-07, "loss": 1.471, "step": 8070 }, { "epoch": 0.2, "grad_norm": 67.0, "learning_rate": 7.430508474576272e-07, "loss": 1.4759, "step": 8080 }, { "epoch": 0.2, "grad_norm": 65.5, "learning_rate": 7.427118644067796e-07, "loss": 1.482, "step": 8090 }, { "epoch": 0.2, "grad_norm": 65.5, "learning_rate": 7.423728813559321e-07, "loss": 1.4691, "step": 8100 }, { "epoch": 0.2, "grad_norm": 66.5, "learning_rate": 7.420338983050846e-07, "loss": 1.4789, "step": 8110 }, { "epoch": 0.2, "grad_norm": 67.5, "learning_rate": 7.416949152542373e-07, "loss": 1.5057, "step": 8120 }, { "epoch": 0.2, "grad_norm": 63.75, "learning_rate": 7.413559322033898e-07, "loss": 1.4689, "step": 8130 }, { "epoch": 0.2, "grad_norm": 66.5, "learning_rate": 7.410169491525423e-07, "loss": 1.4715, "step": 8140 }, { "epoch": 0.2, "grad_norm": 66.5, "learning_rate": 7.406779661016948e-07, "loss": 1.45, "step": 8150 }, { "epoch": 0.2, "grad_norm": 70.5, "learning_rate": 7.403389830508474e-07, "loss": 1.4498, "step": 8160 }, { "epoch": 0.2, "grad_norm": 65.0, "learning_rate": 7.4e-07, "loss": 1.4393, "step": 8170 }, { "epoch": 0.2, "grad_norm": 68.5, "learning_rate": 7.396610169491525e-07, "loss": 1.5018, "step": 8180 }, { "epoch": 0.2, "grad_norm": 70.0, "learning_rate": 7.393220338983051e-07, "loss": 1.466, "step": 8190 }, { "epoch": 0.2, "grad_norm": 69.5, "learning_rate": 7.389830508474576e-07, "loss": 1.4894, "step": 8200 }, { "epoch": 0.21, "grad_norm": 64.5, "learning_rate": 7.386440677966101e-07, "loss": 1.4417, "step": 8210 }, { "epoch": 0.21, "grad_norm": 66.0, "learning_rate": 7.383050847457626e-07, "loss": 1.4571, "step": 8220 }, { "epoch": 0.21, "grad_norm": 64.0, "learning_rate": 7.379661016949153e-07, "loss": 1.4398, "step": 8230 }, { "epoch": 0.21, "grad_norm": 63.25, "learning_rate": 7.376271186440678e-07, "loss": 1.473, "step": 8240 }, { "epoch": 0.21, "grad_norm": 76.5, "learning_rate": 7.372881355932203e-07, "loss": 1.4444, "step": 8250 }, { "epoch": 0.21, "grad_norm": 68.0, "learning_rate": 7.369491525423729e-07, "loss": 1.4357, "step": 8260 }, { "epoch": 0.21, "grad_norm": 67.0, "learning_rate": 7.366101694915254e-07, "loss": 1.4366, "step": 8270 }, { "epoch": 0.21, "grad_norm": 67.0, "learning_rate": 7.36271186440678e-07, "loss": 1.5005, "step": 8280 }, { "epoch": 0.21, "grad_norm": 66.5, "learning_rate": 7.359322033898306e-07, "loss": 1.4709, "step": 8290 }, { "epoch": 0.21, "grad_norm": 69.0, "learning_rate": 7.35593220338983e-07, "loss": 1.4488, "step": 8300 }, { "epoch": 0.21, "grad_norm": 67.0, "learning_rate": 7.352542372881355e-07, "loss": 1.4636, "step": 8310 }, { "epoch": 0.21, "grad_norm": 64.0, "learning_rate": 7.34915254237288e-07, "loss": 1.456, "step": 8320 }, { "epoch": 0.21, "grad_norm": 70.0, "learning_rate": 7.345762711864406e-07, "loss": 1.4277, "step": 8330 }, { "epoch": 0.21, "grad_norm": 69.0, "learning_rate": 7.342372881355932e-07, "loss": 1.4861, "step": 8340 }, { "epoch": 0.21, "grad_norm": 66.5, "learning_rate": 7.338983050847457e-07, "loss": 1.4675, "step": 8350 }, { "epoch": 0.21, "grad_norm": 67.0, "learning_rate": 7.335593220338982e-07, "loss": 1.488, "step": 8360 }, { "epoch": 0.21, "grad_norm": 67.0, "learning_rate": 7.332203389830508e-07, "loss": 1.4869, "step": 8370 }, { "epoch": 0.21, "grad_norm": 70.5, "learning_rate": 7.328813559322033e-07, "loss": 1.4884, "step": 8380 }, { "epoch": 0.21, "grad_norm": 65.5, "learning_rate": 7.325423728813559e-07, "loss": 1.4837, "step": 8390 }, { "epoch": 0.21, "grad_norm": 66.5, "learning_rate": 7.322033898305085e-07, "loss": 1.4456, "step": 8400 }, { "epoch": 0.21, "grad_norm": 66.0, "learning_rate": 7.31864406779661e-07, "loss": 1.4689, "step": 8410 }, { "epoch": 0.21, "grad_norm": 67.0, "learning_rate": 7.315254237288135e-07, "loss": 1.4981, "step": 8420 }, { "epoch": 0.21, "grad_norm": 65.0, "learning_rate": 7.31186440677966e-07, "loss": 1.4375, "step": 8430 }, { "epoch": 0.21, "grad_norm": 65.5, "learning_rate": 7.308474576271187e-07, "loss": 1.4639, "step": 8440 }, { "epoch": 0.21, "grad_norm": 65.0, "learning_rate": 7.305084745762712e-07, "loss": 1.4467, "step": 8450 }, { "epoch": 0.21, "grad_norm": 67.0, "learning_rate": 7.301694915254237e-07, "loss": 1.4271, "step": 8460 }, { "epoch": 0.21, "grad_norm": 66.5, "learning_rate": 7.298305084745763e-07, "loss": 1.4157, "step": 8470 }, { "epoch": 0.21, "grad_norm": 64.0, "learning_rate": 7.294915254237288e-07, "loss": 1.4157, "step": 8480 }, { "epoch": 0.21, "grad_norm": 67.0, "learning_rate": 7.291525423728814e-07, "loss": 1.4447, "step": 8490 }, { "epoch": 0.21, "grad_norm": 64.5, "learning_rate": 7.288135593220338e-07, "loss": 1.4418, "step": 8500 }, { "epoch": 0.21, "grad_norm": 64.0, "learning_rate": 7.284745762711864e-07, "loss": 1.4462, "step": 8510 }, { "epoch": 0.21, "grad_norm": 69.0, "learning_rate": 7.281355932203389e-07, "loss": 1.5123, "step": 8520 }, { "epoch": 0.21, "grad_norm": 66.5, "learning_rate": 7.277966101694914e-07, "loss": 1.4333, "step": 8530 }, { "epoch": 0.21, "grad_norm": 65.5, "learning_rate": 7.27457627118644e-07, "loss": 1.4777, "step": 8540 }, { "epoch": 0.21, "grad_norm": 64.5, "learning_rate": 7.271186440677966e-07, "loss": 1.4531, "step": 8550 }, { "epoch": 0.21, "grad_norm": 66.5, "learning_rate": 7.267796610169491e-07, "loss": 1.4566, "step": 8560 }, { "epoch": 0.21, "grad_norm": 64.5, "learning_rate": 7.264406779661016e-07, "loss": 1.4694, "step": 8570 }, { "epoch": 0.21, "grad_norm": 68.5, "learning_rate": 7.261016949152542e-07, "loss": 1.5073, "step": 8580 }, { "epoch": 0.21, "grad_norm": 64.5, "learning_rate": 7.257627118644067e-07, "loss": 1.4471, "step": 8590 }, { "epoch": 0.21, "grad_norm": 67.5, "learning_rate": 7.254237288135593e-07, "loss": 1.4726, "step": 8600 }, { "epoch": 0.22, "grad_norm": 67.5, "learning_rate": 7.250847457627119e-07, "loss": 1.4678, "step": 8610 }, { "epoch": 0.22, "grad_norm": 66.5, "learning_rate": 7.247457627118644e-07, "loss": 1.481, "step": 8620 }, { "epoch": 0.22, "grad_norm": 63.25, "learning_rate": 7.244067796610169e-07, "loss": 1.4522, "step": 8630 }, { "epoch": 0.22, "grad_norm": 65.5, "learning_rate": 7.240677966101694e-07, "loss": 1.4943, "step": 8640 }, { "epoch": 0.22, "grad_norm": 67.0, "learning_rate": 7.237288135593221e-07, "loss": 1.4812, "step": 8650 }, { "epoch": 0.22, "grad_norm": 67.5, "learning_rate": 7.233898305084746e-07, "loss": 1.4757, "step": 8660 }, { "epoch": 0.22, "grad_norm": 69.5, "learning_rate": 7.230508474576271e-07, "loss": 1.4889, "step": 8670 }, { "epoch": 0.22, "grad_norm": 64.5, "learning_rate": 7.227118644067797e-07, "loss": 1.4999, "step": 8680 }, { "epoch": 0.22, "grad_norm": 65.5, "learning_rate": 7.223728813559322e-07, "loss": 1.4799, "step": 8690 }, { "epoch": 0.22, "grad_norm": 66.5, "learning_rate": 7.220338983050847e-07, "loss": 1.4672, "step": 8700 }, { "epoch": 0.22, "grad_norm": 65.0, "learning_rate": 7.216949152542372e-07, "loss": 1.4843, "step": 8710 }, { "epoch": 0.22, "grad_norm": 65.5, "learning_rate": 7.213559322033898e-07, "loss": 1.4668, "step": 8720 }, { "epoch": 0.22, "grad_norm": 67.0, "learning_rate": 7.210169491525423e-07, "loss": 1.4746, "step": 8730 }, { "epoch": 0.22, "grad_norm": 65.5, "learning_rate": 7.206779661016948e-07, "loss": 1.4795, "step": 8740 }, { "epoch": 0.22, "grad_norm": 65.0, "learning_rate": 7.203389830508474e-07, "loss": 1.477, "step": 8750 }, { "epoch": 0.22, "grad_norm": 67.0, "learning_rate": 7.2e-07, "loss": 1.4565, "step": 8760 }, { "epoch": 0.22, "grad_norm": 63.5, "learning_rate": 7.196610169491525e-07, "loss": 1.4507, "step": 8770 }, { "epoch": 0.22, "grad_norm": 70.0, "learning_rate": 7.19322033898305e-07, "loss": 1.4556, "step": 8780 }, { "epoch": 0.22, "grad_norm": 69.0, "learning_rate": 7.189830508474576e-07, "loss": 1.4427, "step": 8790 }, { "epoch": 0.22, "grad_norm": 66.5, "learning_rate": 7.186440677966101e-07, "loss": 1.4899, "step": 8800 }, { "epoch": 0.22, "grad_norm": 66.5, "learning_rate": 7.183050847457626e-07, "loss": 1.4796, "step": 8810 }, { "epoch": 0.22, "grad_norm": 67.5, "learning_rate": 7.179661016949153e-07, "loss": 1.45, "step": 8820 }, { "epoch": 0.22, "grad_norm": 63.25, "learning_rate": 7.176271186440678e-07, "loss": 1.4756, "step": 8830 }, { "epoch": 0.22, "grad_norm": 66.0, "learning_rate": 7.172881355932203e-07, "loss": 1.4737, "step": 8840 }, { "epoch": 0.22, "grad_norm": 67.5, "learning_rate": 7.169491525423728e-07, "loss": 1.478, "step": 8850 }, { "epoch": 0.22, "grad_norm": 68.5, "learning_rate": 7.166101694915254e-07, "loss": 1.4516, "step": 8860 }, { "epoch": 0.22, "grad_norm": 64.5, "learning_rate": 7.16271186440678e-07, "loss": 1.5122, "step": 8870 }, { "epoch": 0.22, "grad_norm": 68.5, "learning_rate": 7.159322033898305e-07, "loss": 1.4333, "step": 8880 }, { "epoch": 0.22, "grad_norm": 67.5, "learning_rate": 7.155932203389831e-07, "loss": 1.4814, "step": 8890 }, { "epoch": 0.22, "grad_norm": 65.0, "learning_rate": 7.152542372881356e-07, "loss": 1.4131, "step": 8900 }, { "epoch": 0.22, "grad_norm": 67.5, "learning_rate": 7.149152542372881e-07, "loss": 1.4601, "step": 8910 }, { "epoch": 0.22, "grad_norm": 66.5, "learning_rate": 7.145762711864406e-07, "loss": 1.4513, "step": 8920 }, { "epoch": 0.22, "grad_norm": 67.5, "learning_rate": 7.142372881355932e-07, "loss": 1.4492, "step": 8930 }, { "epoch": 0.22, "grad_norm": 64.5, "learning_rate": 7.138983050847457e-07, "loss": 1.4499, "step": 8940 }, { "epoch": 0.22, "grad_norm": 73.5, "learning_rate": 7.135593220338982e-07, "loss": 1.5062, "step": 8950 }, { "epoch": 0.22, "grad_norm": 65.0, "learning_rate": 7.132203389830508e-07, "loss": 1.4741, "step": 8960 }, { "epoch": 0.22, "grad_norm": 69.0, "learning_rate": 7.128813559322033e-07, "loss": 1.4736, "step": 8970 }, { "epoch": 0.22, "grad_norm": 68.5, "learning_rate": 7.125423728813559e-07, "loss": 1.4298, "step": 8980 }, { "epoch": 0.22, "grad_norm": 69.5, "learning_rate": 7.122033898305084e-07, "loss": 1.5216, "step": 8990 }, { "epoch": 0.23, "grad_norm": 66.5, "learning_rate": 7.11864406779661e-07, "loss": 1.4357, "step": 9000 }, { "epoch": 0.23, "grad_norm": 67.0, "learning_rate": 7.115254237288135e-07, "loss": 1.4184, "step": 9010 }, { "epoch": 0.23, "grad_norm": 64.5, "learning_rate": 7.11186440677966e-07, "loss": 1.4065, "step": 9020 }, { "epoch": 0.23, "grad_norm": 68.0, "learning_rate": 7.108474576271187e-07, "loss": 1.4903, "step": 9030 }, { "epoch": 0.23, "grad_norm": 66.0, "learning_rate": 7.105084745762712e-07, "loss": 1.4501, "step": 9040 }, { "epoch": 0.23, "grad_norm": 64.5, "learning_rate": 7.101694915254237e-07, "loss": 1.4491, "step": 9050 }, { "epoch": 0.23, "grad_norm": 71.0, "learning_rate": 7.098305084745762e-07, "loss": 1.4244, "step": 9060 }, { "epoch": 0.23, "grad_norm": 64.5, "learning_rate": 7.094915254237288e-07, "loss": 1.4146, "step": 9070 }, { "epoch": 0.23, "grad_norm": 66.0, "learning_rate": 7.091525423728814e-07, "loss": 1.43, "step": 9080 }, { "epoch": 0.23, "grad_norm": 64.0, "learning_rate": 7.088135593220339e-07, "loss": 1.4789, "step": 9090 }, { "epoch": 0.23, "grad_norm": 67.5, "learning_rate": 7.084745762711865e-07, "loss": 1.4713, "step": 9100 }, { "epoch": 0.23, "grad_norm": 68.5, "learning_rate": 7.08135593220339e-07, "loss": 1.4713, "step": 9110 }, { "epoch": 0.23, "grad_norm": 67.0, "learning_rate": 7.077966101694915e-07, "loss": 1.4777, "step": 9120 }, { "epoch": 0.23, "grad_norm": 68.5, "learning_rate": 7.074576271186439e-07, "loss": 1.4526, "step": 9130 }, { "epoch": 0.23, "grad_norm": 68.5, "learning_rate": 7.071186440677966e-07, "loss": 1.5079, "step": 9140 }, { "epoch": 0.23, "grad_norm": 69.5, "learning_rate": 7.067796610169491e-07, "loss": 1.5103, "step": 9150 }, { "epoch": 0.23, "grad_norm": 67.5, "learning_rate": 7.064406779661016e-07, "loss": 1.488, "step": 9160 }, { "epoch": 0.23, "grad_norm": 69.5, "learning_rate": 7.061016949152542e-07, "loss": 1.5082, "step": 9170 }, { "epoch": 0.23, "grad_norm": 65.5, "learning_rate": 7.057627118644067e-07, "loss": 1.4437, "step": 9180 }, { "epoch": 0.23, "grad_norm": 67.5, "learning_rate": 7.054237288135593e-07, "loss": 1.4919, "step": 9190 }, { "epoch": 0.23, "grad_norm": 66.5, "learning_rate": 7.050847457627118e-07, "loss": 1.4769, "step": 9200 }, { "epoch": 0.23, "grad_norm": 67.0, "learning_rate": 7.047457627118644e-07, "loss": 1.472, "step": 9210 }, { "epoch": 0.23, "grad_norm": 63.75, "learning_rate": 7.044067796610169e-07, "loss": 1.4408, "step": 9220 }, { "epoch": 0.23, "grad_norm": 65.0, "learning_rate": 7.040677966101694e-07, "loss": 1.48, "step": 9230 }, { "epoch": 0.23, "grad_norm": 67.5, "learning_rate": 7.037288135593221e-07, "loss": 1.4382, "step": 9240 }, { "epoch": 0.23, "grad_norm": 65.0, "learning_rate": 7.033898305084746e-07, "loss": 1.4758, "step": 9250 }, { "epoch": 0.23, "grad_norm": 64.0, "learning_rate": 7.030508474576271e-07, "loss": 1.4819, "step": 9260 }, { "epoch": 0.23, "grad_norm": 71.0, "learning_rate": 7.027118644067796e-07, "loss": 1.5281, "step": 9270 }, { "epoch": 0.23, "grad_norm": 68.5, "learning_rate": 7.023728813559322e-07, "loss": 1.4569, "step": 9280 }, { "epoch": 0.23, "grad_norm": 65.0, "learning_rate": 7.020338983050847e-07, "loss": 1.4283, "step": 9290 }, { "epoch": 0.23, "grad_norm": 68.0, "learning_rate": 7.016949152542373e-07, "loss": 1.4144, "step": 9300 }, { "epoch": 0.23, "grad_norm": 64.5, "learning_rate": 7.013559322033899e-07, "loss": 1.4482, "step": 9310 }, { "epoch": 0.23, "grad_norm": 65.0, "learning_rate": 7.010169491525424e-07, "loss": 1.4637, "step": 9320 }, { "epoch": 0.23, "grad_norm": 66.0, "learning_rate": 7.006779661016948e-07, "loss": 1.449, "step": 9330 }, { "epoch": 0.23, "grad_norm": 66.0, "learning_rate": 7.003389830508473e-07, "loss": 1.464, "step": 9340 }, { "epoch": 0.23, "grad_norm": 67.0, "learning_rate": 7e-07, "loss": 1.4, "step": 9350 }, { "epoch": 0.23, "grad_norm": 69.0, "learning_rate": 6.996610169491525e-07, "loss": 1.4246, "step": 9360 }, { "epoch": 0.23, "grad_norm": 68.0, "learning_rate": 6.99322033898305e-07, "loss": 1.4711, "step": 9370 }, { "epoch": 0.23, "grad_norm": 66.0, "learning_rate": 6.989830508474576e-07, "loss": 1.5058, "step": 9380 }, { "epoch": 0.23, "grad_norm": 66.0, "learning_rate": 6.986440677966101e-07, "loss": 1.5505, "step": 9390 }, { "epoch": 0.23, "grad_norm": 67.5, "learning_rate": 6.983050847457627e-07, "loss": 1.5049, "step": 9400 }, { "epoch": 0.24, "grad_norm": 69.0, "learning_rate": 6.979661016949152e-07, "loss": 1.5142, "step": 9410 }, { "epoch": 0.24, "grad_norm": 65.5, "learning_rate": 6.976271186440678e-07, "loss": 1.4848, "step": 9420 }, { "epoch": 0.24, "grad_norm": 61.25, "learning_rate": 6.972881355932203e-07, "loss": 1.4221, "step": 9430 }, { "epoch": 0.24, "grad_norm": 67.5, "learning_rate": 6.969491525423728e-07, "loss": 1.4509, "step": 9440 }, { "epoch": 0.24, "grad_norm": 66.0, "learning_rate": 6.966101694915254e-07, "loss": 1.4312, "step": 9450 }, { "epoch": 0.24, "grad_norm": 72.0, "learning_rate": 6.96271186440678e-07, "loss": 1.4495, "step": 9460 }, { "epoch": 0.24, "grad_norm": 69.0, "learning_rate": 6.959322033898305e-07, "loss": 1.4853, "step": 9470 }, { "epoch": 0.24, "grad_norm": 68.5, "learning_rate": 6.95593220338983e-07, "loss": 1.4941, "step": 9480 }, { "epoch": 0.24, "grad_norm": 64.5, "learning_rate": 6.952542372881356e-07, "loss": 1.4442, "step": 9490 }, { "epoch": 0.24, "grad_norm": 67.0, "learning_rate": 6.949152542372881e-07, "loss": 1.4345, "step": 9500 }, { "epoch": 0.24, "grad_norm": 68.5, "learning_rate": 6.945762711864407e-07, "loss": 1.4962, "step": 9510 }, { "epoch": 0.24, "grad_norm": 67.5, "learning_rate": 6.942372881355933e-07, "loss": 1.5057, "step": 9520 }, { "epoch": 0.24, "grad_norm": 69.0, "learning_rate": 6.938983050847458e-07, "loss": 1.4247, "step": 9530 }, { "epoch": 0.24, "grad_norm": 68.5, "learning_rate": 6.935593220338982e-07, "loss": 1.4728, "step": 9540 }, { "epoch": 0.24, "grad_norm": 68.0, "learning_rate": 6.932203389830507e-07, "loss": 1.5011, "step": 9550 }, { "epoch": 0.24, "grad_norm": 63.5, "learning_rate": 6.928813559322034e-07, "loss": 1.4597, "step": 9560 }, { "epoch": 0.24, "grad_norm": 69.5, "learning_rate": 6.925423728813559e-07, "loss": 1.4246, "step": 9570 }, { "epoch": 0.24, "grad_norm": 64.5, "learning_rate": 6.922033898305084e-07, "loss": 1.4148, "step": 9580 }, { "epoch": 0.24, "grad_norm": 68.0, "learning_rate": 6.91864406779661e-07, "loss": 1.4841, "step": 9590 }, { "epoch": 0.24, "grad_norm": 70.0, "learning_rate": 6.915254237288135e-07, "loss": 1.4521, "step": 9600 }, { "epoch": 0.24, "grad_norm": 69.0, "learning_rate": 6.91186440677966e-07, "loss": 1.4784, "step": 9610 }, { "epoch": 0.24, "grad_norm": 70.0, "learning_rate": 6.908474576271186e-07, "loss": 1.5124, "step": 9620 }, { "epoch": 0.24, "grad_norm": 67.0, "learning_rate": 6.905084745762712e-07, "loss": 1.4599, "step": 9630 }, { "epoch": 0.24, "grad_norm": 64.5, "learning_rate": 6.901694915254237e-07, "loss": 1.4627, "step": 9640 }, { "epoch": 0.24, "grad_norm": 66.5, "learning_rate": 6.898305084745762e-07, "loss": 1.4437, "step": 9650 }, { "epoch": 0.24, "grad_norm": 63.75, "learning_rate": 6.894915254237287e-07, "loss": 1.4753, "step": 9660 }, { "epoch": 0.24, "grad_norm": 65.0, "learning_rate": 6.891525423728814e-07, "loss": 1.4958, "step": 9670 }, { "epoch": 0.24, "grad_norm": 67.5, "learning_rate": 6.888135593220339e-07, "loss": 1.466, "step": 9680 }, { "epoch": 0.24, "grad_norm": 65.5, "learning_rate": 6.884745762711864e-07, "loss": 1.4446, "step": 9690 }, { "epoch": 0.24, "grad_norm": 63.25, "learning_rate": 6.88135593220339e-07, "loss": 1.3986, "step": 9700 }, { "epoch": 0.24, "grad_norm": 71.0, "learning_rate": 6.877966101694915e-07, "loss": 1.4598, "step": 9710 }, { "epoch": 0.24, "grad_norm": 64.0, "learning_rate": 6.874576271186441e-07, "loss": 1.5172, "step": 9720 }, { "epoch": 0.24, "grad_norm": 69.0, "learning_rate": 6.871186440677967e-07, "loss": 1.4723, "step": 9730 }, { "epoch": 0.24, "grad_norm": 66.0, "learning_rate": 6.867796610169492e-07, "loss": 1.4825, "step": 9740 }, { "epoch": 0.24, "grad_norm": 66.0, "learning_rate": 6.864406779661016e-07, "loss": 1.4601, "step": 9750 }, { "epoch": 0.24, "grad_norm": 65.0, "learning_rate": 6.861016949152541e-07, "loss": 1.4799, "step": 9760 }, { "epoch": 0.24, "grad_norm": 63.0, "learning_rate": 6.857627118644067e-07, "loss": 1.4512, "step": 9770 }, { "epoch": 0.24, "grad_norm": 62.25, "learning_rate": 6.854237288135593e-07, "loss": 1.4744, "step": 9780 }, { "epoch": 0.24, "grad_norm": 66.5, "learning_rate": 6.850847457627118e-07, "loss": 1.4564, "step": 9790 }, { "epoch": 0.24, "grad_norm": 66.5, "learning_rate": 6.847457627118643e-07, "loss": 1.4685, "step": 9800 }, { "epoch": 0.25, "grad_norm": 66.0, "learning_rate": 6.844067796610169e-07, "loss": 1.5341, "step": 9810 }, { "epoch": 0.25, "grad_norm": 67.5, "learning_rate": 6.840677966101694e-07, "loss": 1.437, "step": 9820 }, { "epoch": 0.25, "grad_norm": 69.5, "learning_rate": 6.83728813559322e-07, "loss": 1.4125, "step": 9830 }, { "epoch": 0.25, "grad_norm": 70.5, "learning_rate": 6.833898305084746e-07, "loss": 1.5021, "step": 9840 }, { "epoch": 0.25, "grad_norm": 65.5, "learning_rate": 6.830508474576271e-07, "loss": 1.479, "step": 9850 }, { "epoch": 0.25, "grad_norm": 64.5, "learning_rate": 6.827118644067796e-07, "loss": 1.4588, "step": 9860 }, { "epoch": 0.25, "grad_norm": 69.0, "learning_rate": 6.823728813559321e-07, "loss": 1.4607, "step": 9870 }, { "epoch": 0.25, "grad_norm": 65.5, "learning_rate": 6.820338983050848e-07, "loss": 1.4385, "step": 9880 }, { "epoch": 0.25, "grad_norm": 64.0, "learning_rate": 6.816949152542373e-07, "loss": 1.4334, "step": 9890 }, { "epoch": 0.25, "grad_norm": 67.5, "learning_rate": 6.813559322033898e-07, "loss": 1.5137, "step": 9900 }, { "epoch": 0.25, "grad_norm": 63.0, "learning_rate": 6.810169491525424e-07, "loss": 1.4347, "step": 9910 }, { "epoch": 0.25, "grad_norm": 67.0, "learning_rate": 6.806779661016949e-07, "loss": 1.4587, "step": 9920 }, { "epoch": 0.25, "grad_norm": 66.5, "learning_rate": 6.803389830508474e-07, "loss": 1.4086, "step": 9930 }, { "epoch": 0.25, "grad_norm": 64.5, "learning_rate": 6.800000000000001e-07, "loss": 1.4411, "step": 9940 }, { "epoch": 0.25, "grad_norm": 68.5, "learning_rate": 6.796610169491526e-07, "loss": 1.4457, "step": 9950 }, { "epoch": 0.25, "grad_norm": 67.5, "learning_rate": 6.79322033898305e-07, "loss": 1.476, "step": 9960 }, { "epoch": 0.25, "grad_norm": 65.5, "learning_rate": 6.789830508474575e-07, "loss": 1.436, "step": 9970 }, { "epoch": 0.25, "grad_norm": 68.0, "learning_rate": 6.786440677966101e-07, "loss": 1.4495, "step": 9980 }, { "epoch": 0.25, "grad_norm": 66.5, "learning_rate": 6.783050847457627e-07, "loss": 1.4609, "step": 9990 }, { "epoch": 0.25, "grad_norm": 67.0, "learning_rate": 6.779661016949152e-07, "loss": 1.4835, "step": 10000 }, { "epoch": 0.25, "eval_loss": 1.4654276371002197, "eval_runtime": 67.9048, "eval_samples_per_second": 14.727, "eval_steps_per_second": 14.727, "step": 10000 }, { "epoch": 0.25, "grad_norm": 69.0, "learning_rate": 6.776271186440677e-07, "loss": 1.4988, "step": 10010 }, { "epoch": 0.25, "grad_norm": 68.5, "learning_rate": 6.772881355932203e-07, "loss": 1.4973, "step": 10020 }, { "epoch": 0.25, "grad_norm": 65.0, "learning_rate": 6.769491525423728e-07, "loss": 1.4256, "step": 10030 }, { "epoch": 0.25, "grad_norm": 68.0, "learning_rate": 6.766101694915254e-07, "loss": 1.4714, "step": 10040 }, { "epoch": 0.25, "grad_norm": 66.0, "learning_rate": 6.76271186440678e-07, "loss": 1.4417, "step": 10050 }, { "epoch": 0.25, "grad_norm": 63.5, "learning_rate": 6.759322033898305e-07, "loss": 1.4292, "step": 10060 }, { "epoch": 0.25, "grad_norm": 64.5, "learning_rate": 6.75593220338983e-07, "loss": 1.4803, "step": 10070 }, { "epoch": 0.25, "grad_norm": 69.0, "learning_rate": 6.752542372881355e-07, "loss": 1.4759, "step": 10080 }, { "epoch": 0.25, "grad_norm": 66.5, "learning_rate": 6.749152542372881e-07, "loss": 1.4646, "step": 10090 }, { "epoch": 0.25, "grad_norm": 66.0, "learning_rate": 6.745762711864407e-07, "loss": 1.4673, "step": 10100 }, { "epoch": 0.25, "grad_norm": 65.5, "learning_rate": 6.742372881355932e-07, "loss": 1.4792, "step": 10110 }, { "epoch": 0.25, "grad_norm": 65.0, "learning_rate": 6.738983050847458e-07, "loss": 1.5136, "step": 10120 }, { "epoch": 0.25, "grad_norm": 70.0, "learning_rate": 6.735593220338983e-07, "loss": 1.457, "step": 10130 }, { "epoch": 0.25, "grad_norm": 66.5, "learning_rate": 6.732203389830508e-07, "loss": 1.4271, "step": 10140 }, { "epoch": 0.25, "grad_norm": 68.0, "learning_rate": 6.728813559322035e-07, "loss": 1.4752, "step": 10150 }, { "epoch": 0.25, "grad_norm": 68.5, "learning_rate": 6.72542372881356e-07, "loss": 1.4579, "step": 10160 }, { "epoch": 0.25, "grad_norm": 66.0, "learning_rate": 6.722033898305084e-07, "loss": 1.477, "step": 10170 }, { "epoch": 0.25, "grad_norm": 66.0, "learning_rate": 6.718644067796609e-07, "loss": 1.4302, "step": 10180 }, { "epoch": 0.25, "grad_norm": 66.0, "learning_rate": 6.715254237288135e-07, "loss": 1.4742, "step": 10190 }, { "epoch": 0.26, "grad_norm": 65.0, "learning_rate": 6.71186440677966e-07, "loss": 1.4706, "step": 10200 }, { "epoch": 0.26, "grad_norm": 67.5, "learning_rate": 6.708474576271186e-07, "loss": 1.4452, "step": 10210 }, { "epoch": 0.26, "grad_norm": 64.0, "learning_rate": 6.705084745762711e-07, "loss": 1.4182, "step": 10220 }, { "epoch": 0.26, "grad_norm": 67.5, "learning_rate": 6.701694915254237e-07, "loss": 1.4845, "step": 10230 }, { "epoch": 0.26, "grad_norm": 70.0, "learning_rate": 6.698305084745762e-07, "loss": 1.4913, "step": 10240 }, { "epoch": 0.26, "grad_norm": 67.5, "learning_rate": 6.694915254237287e-07, "loss": 1.4409, "step": 10250 }, { "epoch": 0.26, "grad_norm": 68.0, "learning_rate": 6.691525423728814e-07, "loss": 1.5124, "step": 10260 }, { "epoch": 0.26, "grad_norm": 65.0, "learning_rate": 6.688135593220339e-07, "loss": 1.4285, "step": 10270 }, { "epoch": 0.26, "grad_norm": 69.0, "learning_rate": 6.684745762711864e-07, "loss": 1.4087, "step": 10280 }, { "epoch": 0.26, "grad_norm": 68.0, "learning_rate": 6.681355932203389e-07, "loss": 1.4562, "step": 10290 }, { "epoch": 0.26, "grad_norm": 63.25, "learning_rate": 6.677966101694915e-07, "loss": 1.4424, "step": 10300 }, { "epoch": 0.26, "grad_norm": 69.0, "learning_rate": 6.674576271186441e-07, "loss": 1.5074, "step": 10310 }, { "epoch": 0.26, "grad_norm": 66.0, "learning_rate": 6.671186440677966e-07, "loss": 1.4922, "step": 10320 }, { "epoch": 0.26, "grad_norm": 72.5, "learning_rate": 6.667796610169492e-07, "loss": 1.4407, "step": 10330 }, { "epoch": 0.26, "grad_norm": 67.5, "learning_rate": 6.664406779661017e-07, "loss": 1.4868, "step": 10340 }, { "epoch": 0.26, "grad_norm": 68.0, "learning_rate": 6.661016949152542e-07, "loss": 1.4928, "step": 10350 }, { "epoch": 0.26, "grad_norm": 66.0, "learning_rate": 6.657627118644069e-07, "loss": 1.4382, "step": 10360 }, { "epoch": 0.26, "grad_norm": 64.0, "learning_rate": 6.654237288135593e-07, "loss": 1.434, "step": 10370 }, { "epoch": 0.26, "grad_norm": 67.5, "learning_rate": 6.650847457627118e-07, "loss": 1.4703, "step": 10380 }, { "epoch": 0.26, "grad_norm": 68.5, "learning_rate": 6.647457627118643e-07, "loss": 1.5073, "step": 10390 }, { "epoch": 0.26, "grad_norm": 65.5, "learning_rate": 6.644067796610169e-07, "loss": 1.3907, "step": 10400 }, { "epoch": 0.26, "grad_norm": 68.0, "learning_rate": 6.640677966101694e-07, "loss": 1.4997, "step": 10410 }, { "epoch": 0.26, "grad_norm": 69.5, "learning_rate": 6.63728813559322e-07, "loss": 1.4372, "step": 10420 }, { "epoch": 0.26, "grad_norm": 67.5, "learning_rate": 6.633898305084745e-07, "loss": 1.4549, "step": 10430 }, { "epoch": 0.26, "grad_norm": 67.5, "learning_rate": 6.630508474576271e-07, "loss": 1.4796, "step": 10440 }, { "epoch": 0.26, "grad_norm": 65.5, "learning_rate": 6.627118644067796e-07, "loss": 1.4355, "step": 10450 }, { "epoch": 0.26, "grad_norm": 69.5, "learning_rate": 6.623728813559321e-07, "loss": 1.4747, "step": 10460 }, { "epoch": 0.26, "grad_norm": 69.5, "learning_rate": 6.620338983050848e-07, "loss": 1.4663, "step": 10470 }, { "epoch": 0.26, "grad_norm": 70.5, "learning_rate": 6.616949152542373e-07, "loss": 1.4823, "step": 10480 }, { "epoch": 0.26, "grad_norm": 67.0, "learning_rate": 6.613559322033898e-07, "loss": 1.4355, "step": 10490 }, { "epoch": 0.26, "grad_norm": 65.5, "learning_rate": 6.610169491525423e-07, "loss": 1.4624, "step": 10500 }, { "epoch": 0.26, "grad_norm": 68.0, "learning_rate": 6.606779661016949e-07, "loss": 1.5301, "step": 10510 }, { "epoch": 0.26, "grad_norm": 69.0, "learning_rate": 6.603389830508475e-07, "loss": 1.4759, "step": 10520 }, { "epoch": 0.26, "grad_norm": 68.5, "learning_rate": 6.6e-07, "loss": 1.5257, "step": 10530 }, { "epoch": 0.26, "grad_norm": 66.5, "learning_rate": 6.596610169491526e-07, "loss": 1.4665, "step": 10540 }, { "epoch": 0.26, "grad_norm": 66.5, "learning_rate": 6.593220338983051e-07, "loss": 1.4267, "step": 10550 }, { "epoch": 0.26, "grad_norm": 66.0, "learning_rate": 6.589830508474576e-07, "loss": 1.4403, "step": 10560 }, { "epoch": 0.26, "grad_norm": 70.0, "learning_rate": 6.586440677966101e-07, "loss": 1.4791, "step": 10570 }, { "epoch": 0.26, "grad_norm": 67.5, "learning_rate": 6.583050847457627e-07, "loss": 1.4623, "step": 10580 }, { "epoch": 0.26, "grad_norm": 64.5, "learning_rate": 6.579661016949152e-07, "loss": 1.4464, "step": 10590 }, { "epoch": 0.27, "grad_norm": 69.0, "learning_rate": 6.576271186440677e-07, "loss": 1.4335, "step": 10600 }, { "epoch": 0.27, "grad_norm": 67.0, "learning_rate": 6.572881355932203e-07, "loss": 1.4635, "step": 10610 }, { "epoch": 0.27, "grad_norm": 65.0, "learning_rate": 6.569491525423728e-07, "loss": 1.4806, "step": 10620 }, { "epoch": 0.27, "grad_norm": 68.0, "learning_rate": 6.566101694915254e-07, "loss": 1.4591, "step": 10630 }, { "epoch": 0.27, "grad_norm": 68.5, "learning_rate": 6.562711864406779e-07, "loss": 1.4608, "step": 10640 }, { "epoch": 0.27, "grad_norm": 68.5, "learning_rate": 6.559322033898305e-07, "loss": 1.4687, "step": 10650 }, { "epoch": 0.27, "grad_norm": 66.0, "learning_rate": 6.55593220338983e-07, "loss": 1.4573, "step": 10660 }, { "epoch": 0.27, "grad_norm": 64.0, "learning_rate": 6.552542372881355e-07, "loss": 1.4453, "step": 10670 }, { "epoch": 0.27, "grad_norm": 67.5, "learning_rate": 6.549152542372881e-07, "loss": 1.5089, "step": 10680 }, { "epoch": 0.27, "grad_norm": 68.5, "learning_rate": 6.545762711864407e-07, "loss": 1.4764, "step": 10690 }, { "epoch": 0.27, "grad_norm": 67.5, "learning_rate": 6.542372881355932e-07, "loss": 1.4393, "step": 10700 }, { "epoch": 0.27, "grad_norm": 65.5, "learning_rate": 6.538983050847457e-07, "loss": 1.4707, "step": 10710 }, { "epoch": 0.27, "grad_norm": 67.5, "learning_rate": 6.535593220338983e-07, "loss": 1.5209, "step": 10720 }, { "epoch": 0.27, "grad_norm": 70.0, "learning_rate": 6.532203389830508e-07, "loss": 1.4655, "step": 10730 }, { "epoch": 0.27, "grad_norm": 66.5, "learning_rate": 6.528813559322034e-07, "loss": 1.4547, "step": 10740 }, { "epoch": 0.27, "grad_norm": 68.5, "learning_rate": 6.52542372881356e-07, "loss": 1.4503, "step": 10750 }, { "epoch": 0.27, "grad_norm": 64.5, "learning_rate": 6.522033898305085e-07, "loss": 1.4815, "step": 10760 }, { "epoch": 0.27, "grad_norm": 65.5, "learning_rate": 6.51864406779661e-07, "loss": 1.4859, "step": 10770 }, { "epoch": 0.27, "grad_norm": 66.5, "learning_rate": 6.515254237288135e-07, "loss": 1.513, "step": 10780 }, { "epoch": 0.27, "grad_norm": 67.5, "learning_rate": 6.511864406779661e-07, "loss": 1.4934, "step": 10790 }, { "epoch": 0.27, "grad_norm": 68.0, "learning_rate": 6.508474576271186e-07, "loss": 1.4527, "step": 10800 }, { "epoch": 0.27, "grad_norm": 65.0, "learning_rate": 6.505084745762711e-07, "loss": 1.4333, "step": 10810 }, { "epoch": 0.27, "grad_norm": 74.0, "learning_rate": 6.501694915254237e-07, "loss": 1.467, "step": 10820 }, { "epoch": 0.27, "grad_norm": 73.0, "learning_rate": 6.498305084745762e-07, "loss": 1.4497, "step": 10830 }, { "epoch": 0.27, "grad_norm": 65.5, "learning_rate": 6.494915254237287e-07, "loss": 1.4795, "step": 10840 }, { "epoch": 0.27, "grad_norm": 63.75, "learning_rate": 6.491525423728813e-07, "loss": 1.3947, "step": 10850 }, { "epoch": 0.27, "grad_norm": 62.75, "learning_rate": 6.488135593220339e-07, "loss": 1.3917, "step": 10860 }, { "epoch": 0.27, "grad_norm": 66.5, "learning_rate": 6.484745762711864e-07, "loss": 1.4264, "step": 10870 }, { "epoch": 0.27, "grad_norm": 67.0, "learning_rate": 6.481355932203389e-07, "loss": 1.4699, "step": 10880 }, { "epoch": 0.27, "grad_norm": 68.5, "learning_rate": 6.477966101694915e-07, "loss": 1.5, "step": 10890 }, { "epoch": 0.27, "grad_norm": 70.0, "learning_rate": 6.474576271186441e-07, "loss": 1.5071, "step": 10900 }, { "epoch": 0.27, "grad_norm": 67.5, "learning_rate": 6.471186440677966e-07, "loss": 1.4718, "step": 10910 }, { "epoch": 0.27, "grad_norm": 64.0, "learning_rate": 6.467796610169491e-07, "loss": 1.4342, "step": 10920 }, { "epoch": 0.27, "grad_norm": 69.5, "learning_rate": 6.464406779661017e-07, "loss": 1.4734, "step": 10930 }, { "epoch": 0.27, "grad_norm": 68.5, "learning_rate": 6.461016949152542e-07, "loss": 1.4744, "step": 10940 }, { "epoch": 0.27, "grad_norm": 65.0, "learning_rate": 6.457627118644068e-07, "loss": 1.4463, "step": 10950 }, { "epoch": 0.27, "grad_norm": 65.0, "learning_rate": 6.454237288135594e-07, "loss": 1.4691, "step": 10960 }, { "epoch": 0.27, "grad_norm": 66.5, "learning_rate": 6.450847457627119e-07, "loss": 1.4341, "step": 10970 }, { "epoch": 0.27, "grad_norm": 67.0, "learning_rate": 6.447457627118644e-07, "loss": 1.4616, "step": 10980 }, { "epoch": 0.27, "grad_norm": 64.5, "learning_rate": 6.444067796610169e-07, "loss": 1.4609, "step": 10990 }, { "epoch": 0.28, "grad_norm": 63.75, "learning_rate": 6.440677966101694e-07, "loss": 1.4632, "step": 11000 }, { "epoch": 0.28, "grad_norm": 66.5, "learning_rate": 6.43728813559322e-07, "loss": 1.4392, "step": 11010 }, { "epoch": 0.28, "grad_norm": 64.5, "learning_rate": 6.433898305084745e-07, "loss": 1.456, "step": 11020 }, { "epoch": 0.28, "grad_norm": 68.5, "learning_rate": 6.43050847457627e-07, "loss": 1.4876, "step": 11030 }, { "epoch": 0.28, "grad_norm": 72.5, "learning_rate": 6.427118644067796e-07, "loss": 1.4529, "step": 11040 }, { "epoch": 0.28, "grad_norm": 65.5, "learning_rate": 6.423728813559321e-07, "loss": 1.4748, "step": 11050 }, { "epoch": 0.28, "grad_norm": 64.0, "learning_rate": 6.420338983050847e-07, "loss": 1.4162, "step": 11060 }, { "epoch": 0.28, "grad_norm": 65.0, "learning_rate": 6.416949152542373e-07, "loss": 1.4641, "step": 11070 }, { "epoch": 0.28, "grad_norm": 66.0, "learning_rate": 6.413559322033898e-07, "loss": 1.4856, "step": 11080 }, { "epoch": 0.28, "grad_norm": 67.5, "learning_rate": 6.410169491525423e-07, "loss": 1.4431, "step": 11090 }, { "epoch": 0.28, "grad_norm": 66.0, "learning_rate": 6.406779661016949e-07, "loss": 1.4102, "step": 11100 }, { "epoch": 0.28, "grad_norm": 66.5, "learning_rate": 6.403389830508475e-07, "loss": 1.4456, "step": 11110 }, { "epoch": 0.28, "grad_norm": 63.25, "learning_rate": 6.4e-07, "loss": 1.4227, "step": 11120 }, { "epoch": 0.28, "grad_norm": 66.0, "learning_rate": 6.396610169491525e-07, "loss": 1.4358, "step": 11130 }, { "epoch": 0.28, "grad_norm": 69.0, "learning_rate": 6.393220338983051e-07, "loss": 1.4876, "step": 11140 }, { "epoch": 0.28, "grad_norm": 66.5, "learning_rate": 6.389830508474576e-07, "loss": 1.4502, "step": 11150 }, { "epoch": 0.28, "grad_norm": 67.5, "learning_rate": 6.386440677966101e-07, "loss": 1.4364, "step": 11160 }, { "epoch": 0.28, "grad_norm": 66.5, "learning_rate": 6.383050847457628e-07, "loss": 1.4213, "step": 11170 }, { "epoch": 0.28, "grad_norm": 64.5, "learning_rate": 6.379661016949153e-07, "loss": 1.4684, "step": 11180 }, { "epoch": 0.28, "grad_norm": 68.0, "learning_rate": 6.376271186440678e-07, "loss": 1.4335, "step": 11190 }, { "epoch": 0.28, "grad_norm": 66.5, "learning_rate": 6.372881355932202e-07, "loss": 1.47, "step": 11200 }, { "epoch": 0.28, "grad_norm": 65.5, "learning_rate": 6.369491525423728e-07, "loss": 1.4497, "step": 11210 }, { "epoch": 0.28, "grad_norm": 69.0, "learning_rate": 6.366101694915254e-07, "loss": 1.4503, "step": 11220 }, { "epoch": 0.28, "grad_norm": 63.75, "learning_rate": 6.362711864406779e-07, "loss": 1.4279, "step": 11230 }, { "epoch": 0.28, "grad_norm": 67.5, "learning_rate": 6.359322033898305e-07, "loss": 1.5155, "step": 11240 }, { "epoch": 0.28, "grad_norm": 67.5, "learning_rate": 6.35593220338983e-07, "loss": 1.5122, "step": 11250 }, { "epoch": 0.28, "grad_norm": 70.5, "learning_rate": 6.352542372881355e-07, "loss": 1.4839, "step": 11260 }, { "epoch": 0.28, "grad_norm": 66.5, "learning_rate": 6.349152542372881e-07, "loss": 1.4678, "step": 11270 }, { "epoch": 0.28, "grad_norm": 65.0, "learning_rate": 6.345762711864407e-07, "loss": 1.4855, "step": 11280 }, { "epoch": 0.28, "grad_norm": 68.0, "learning_rate": 6.342372881355932e-07, "loss": 1.4765, "step": 11290 }, { "epoch": 0.28, "grad_norm": 71.0, "learning_rate": 6.338983050847457e-07, "loss": 1.4538, "step": 11300 }, { "epoch": 0.28, "grad_norm": 68.0, "learning_rate": 6.335593220338983e-07, "loss": 1.4804, "step": 11310 }, { "epoch": 0.28, "grad_norm": 66.5, "learning_rate": 6.332203389830508e-07, "loss": 1.4747, "step": 11320 }, { "epoch": 0.28, "grad_norm": 68.0, "learning_rate": 6.328813559322034e-07, "loss": 1.5159, "step": 11330 }, { "epoch": 0.28, "grad_norm": 67.5, "learning_rate": 6.325423728813559e-07, "loss": 1.4643, "step": 11340 }, { "epoch": 0.28, "grad_norm": 68.0, "learning_rate": 6.322033898305085e-07, "loss": 1.3724, "step": 11350 }, { "epoch": 0.28, "grad_norm": 65.0, "learning_rate": 6.31864406779661e-07, "loss": 1.4464, "step": 11360 }, { "epoch": 0.28, "grad_norm": 63.0, "learning_rate": 6.315254237288135e-07, "loss": 1.4179, "step": 11370 }, { "epoch": 0.28, "grad_norm": 64.5, "learning_rate": 6.311864406779662e-07, "loss": 1.4811, "step": 11380 }, { "epoch": 0.28, "grad_norm": 67.5, "learning_rate": 6.308474576271187e-07, "loss": 1.4384, "step": 11390 }, { "epoch": 0.28, "grad_norm": 68.0, "learning_rate": 6.305084745762712e-07, "loss": 1.5039, "step": 11400 }, { "epoch": 0.29, "grad_norm": 70.5, "learning_rate": 6.301694915254236e-07, "loss": 1.4565, "step": 11410 }, { "epoch": 0.29, "grad_norm": 65.5, "learning_rate": 6.298305084745762e-07, "loss": 1.399, "step": 11420 }, { "epoch": 0.29, "grad_norm": 67.5, "learning_rate": 6.294915254237288e-07, "loss": 1.4315, "step": 11430 }, { "epoch": 0.29, "grad_norm": 65.5, "learning_rate": 6.291525423728813e-07, "loss": 1.4686, "step": 11440 }, { "epoch": 0.29, "grad_norm": 63.75, "learning_rate": 6.288135593220339e-07, "loss": 1.4673, "step": 11450 }, { "epoch": 0.29, "grad_norm": 67.5, "learning_rate": 6.284745762711864e-07, "loss": 1.4691, "step": 11460 }, { "epoch": 0.29, "grad_norm": 66.0, "learning_rate": 6.281355932203389e-07, "loss": 1.4743, "step": 11470 }, { "epoch": 0.29, "grad_norm": 68.0, "learning_rate": 6.277966101694914e-07, "loss": 1.4986, "step": 11480 }, { "epoch": 0.29, "grad_norm": 66.0, "learning_rate": 6.274576271186441e-07, "loss": 1.4161, "step": 11490 }, { "epoch": 0.29, "grad_norm": 64.0, "learning_rate": 6.271186440677966e-07, "loss": 1.4226, "step": 11500 }, { "epoch": 0.29, "grad_norm": 64.5, "learning_rate": 6.267796610169491e-07, "loss": 1.4703, "step": 11510 }, { "epoch": 0.29, "grad_norm": 67.5, "learning_rate": 6.264406779661016e-07, "loss": 1.4648, "step": 11520 }, { "epoch": 0.29, "grad_norm": 66.5, "learning_rate": 6.261016949152542e-07, "loss": 1.456, "step": 11530 }, { "epoch": 0.29, "grad_norm": 66.0, "learning_rate": 6.257627118644068e-07, "loss": 1.4574, "step": 11540 }, { "epoch": 0.29, "grad_norm": 68.0, "learning_rate": 6.254237288135593e-07, "loss": 1.481, "step": 11550 }, { "epoch": 0.29, "grad_norm": 66.5, "learning_rate": 6.250847457627119e-07, "loss": 1.4125, "step": 11560 }, { "epoch": 0.29, "grad_norm": 69.0, "learning_rate": 6.247457627118644e-07, "loss": 1.4701, "step": 11570 }, { "epoch": 0.29, "grad_norm": 68.0, "learning_rate": 6.244067796610169e-07, "loss": 1.4939, "step": 11580 }, { "epoch": 0.29, "grad_norm": 65.5, "learning_rate": 6.240677966101696e-07, "loss": 1.5027, "step": 11590 }, { "epoch": 0.29, "grad_norm": 66.5, "learning_rate": 6.237288135593221e-07, "loss": 1.4748, "step": 11600 }, { "epoch": 0.29, "grad_norm": 65.5, "learning_rate": 6.233898305084746e-07, "loss": 1.4681, "step": 11610 }, { "epoch": 0.29, "grad_norm": 68.0, "learning_rate": 6.23050847457627e-07, "loss": 1.4537, "step": 11620 }, { "epoch": 0.29, "grad_norm": 67.0, "learning_rate": 6.227118644067796e-07, "loss": 1.4798, "step": 11630 }, { "epoch": 0.29, "grad_norm": 68.0, "learning_rate": 6.223728813559321e-07, "loss": 1.4865, "step": 11640 }, { "epoch": 0.29, "grad_norm": 65.5, "learning_rate": 6.220338983050847e-07, "loss": 1.4461, "step": 11650 }, { "epoch": 0.29, "grad_norm": 67.0, "learning_rate": 6.216949152542372e-07, "loss": 1.4776, "step": 11660 }, { "epoch": 0.29, "grad_norm": 71.5, "learning_rate": 6.213559322033898e-07, "loss": 1.4525, "step": 11670 }, { "epoch": 0.29, "grad_norm": 70.0, "learning_rate": 6.210169491525423e-07, "loss": 1.5047, "step": 11680 }, { "epoch": 0.29, "grad_norm": 65.5, "learning_rate": 6.206779661016948e-07, "loss": 1.4818, "step": 11690 }, { "epoch": 0.29, "grad_norm": 70.0, "learning_rate": 6.203389830508475e-07, "loss": 1.476, "step": 11700 }, { "epoch": 0.29, "grad_norm": 65.5, "learning_rate": 6.2e-07, "loss": 1.4598, "step": 11710 }, { "epoch": 0.29, "grad_norm": 66.5, "learning_rate": 6.196610169491525e-07, "loss": 1.4466, "step": 11720 }, { "epoch": 0.29, "grad_norm": 67.0, "learning_rate": 6.19322033898305e-07, "loss": 1.4924, "step": 11730 }, { "epoch": 0.29, "grad_norm": 66.0, "learning_rate": 6.189830508474576e-07, "loss": 1.52, "step": 11740 }, { "epoch": 0.29, "grad_norm": 67.0, "learning_rate": 6.186440677966102e-07, "loss": 1.4954, "step": 11750 }, { "epoch": 0.29, "grad_norm": 67.0, "learning_rate": 6.183050847457627e-07, "loss": 1.5107, "step": 11760 }, { "epoch": 0.29, "grad_norm": 72.0, "learning_rate": 6.179661016949153e-07, "loss": 1.4472, "step": 11770 }, { "epoch": 0.29, "grad_norm": 67.0, "learning_rate": 6.176271186440678e-07, "loss": 1.4821, "step": 11780 }, { "epoch": 0.29, "grad_norm": 71.0, "learning_rate": 6.172881355932203e-07, "loss": 1.4566, "step": 11790 }, { "epoch": 0.29, "grad_norm": 68.5, "learning_rate": 6.169491525423728e-07, "loss": 1.47, "step": 11800 }, { "epoch": 0.3, "grad_norm": 63.25, "learning_rate": 6.166101694915255e-07, "loss": 1.4686, "step": 11810 }, { "epoch": 0.3, "grad_norm": 65.0, "learning_rate": 6.16271186440678e-07, "loss": 1.49, "step": 11820 }, { "epoch": 0.3, "grad_norm": 68.0, "learning_rate": 6.159322033898304e-07, "loss": 1.4919, "step": 11830 }, { "epoch": 0.3, "grad_norm": 68.0, "learning_rate": 6.15593220338983e-07, "loss": 1.4096, "step": 11840 }, { "epoch": 0.3, "grad_norm": 67.5, "learning_rate": 6.152542372881355e-07, "loss": 1.462, "step": 11850 }, { "epoch": 0.3, "grad_norm": 64.0, "learning_rate": 6.149152542372881e-07, "loss": 1.4232, "step": 11860 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.145762711864406e-07, "loss": 1.4604, "step": 11870 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.142372881355932e-07, "loss": 1.4387, "step": 11880 }, { "epoch": 0.3, "grad_norm": 67.5, "learning_rate": 6.138983050847457e-07, "loss": 1.4457, "step": 11890 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.135593220338982e-07, "loss": 1.4275, "step": 11900 }, { "epoch": 0.3, "grad_norm": 65.0, "learning_rate": 6.132203389830508e-07, "loss": 1.5076, "step": 11910 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.128813559322034e-07, "loss": 1.416, "step": 11920 }, { "epoch": 0.3, "grad_norm": 67.5, "learning_rate": 6.125423728813559e-07, "loss": 1.4856, "step": 11930 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.122033898305084e-07, "loss": 1.454, "step": 11940 }, { "epoch": 0.3, "grad_norm": 72.0, "learning_rate": 6.11864406779661e-07, "loss": 1.4184, "step": 11950 }, { "epoch": 0.3, "grad_norm": 66.5, "learning_rate": 6.115254237288135e-07, "loss": 1.4385, "step": 11960 }, { "epoch": 0.3, "grad_norm": 70.0, "learning_rate": 6.111864406779661e-07, "loss": 1.4271, "step": 11970 }, { "epoch": 0.3, "grad_norm": 69.5, "learning_rate": 6.108474576271187e-07, "loss": 1.4539, "step": 11980 }, { "epoch": 0.3, "grad_norm": 65.0, "learning_rate": 6.105084745762712e-07, "loss": 1.4072, "step": 11990 }, { "epoch": 0.3, "grad_norm": 71.5, "learning_rate": 6.101694915254237e-07, "loss": 1.5427, "step": 12000 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.098305084745762e-07, "loss": 1.4361, "step": 12010 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.094915254237289e-07, "loss": 1.4644, "step": 12020 }, { "epoch": 0.3, "grad_norm": 69.0, "learning_rate": 6.091525423728813e-07, "loss": 1.4782, "step": 12030 }, { "epoch": 0.3, "grad_norm": 66.5, "learning_rate": 6.088135593220338e-07, "loss": 1.4871, "step": 12040 }, { "epoch": 0.3, "grad_norm": 63.5, "learning_rate": 6.084745762711864e-07, "loss": 1.4617, "step": 12050 }, { "epoch": 0.3, "grad_norm": 66.5, "learning_rate": 6.081355932203389e-07, "loss": 1.4472, "step": 12060 }, { "epoch": 0.3, "grad_norm": 67.0, "learning_rate": 6.077966101694914e-07, "loss": 1.44, "step": 12070 }, { "epoch": 0.3, "grad_norm": 67.5, "learning_rate": 6.07457627118644e-07, "loss": 1.4738, "step": 12080 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.071186440677966e-07, "loss": 1.4651, "step": 12090 }, { "epoch": 0.3, "grad_norm": 64.5, "learning_rate": 6.067796610169491e-07, "loss": 1.4628, "step": 12100 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.064406779661016e-07, "loss": 1.3952, "step": 12110 }, { "epoch": 0.3, "grad_norm": 71.5, "learning_rate": 6.061016949152542e-07, "loss": 1.49, "step": 12120 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.057627118644068e-07, "loss": 1.4441, "step": 12130 }, { "epoch": 0.3, "grad_norm": 66.0, "learning_rate": 6.054237288135593e-07, "loss": 1.4932, "step": 12140 }, { "epoch": 0.3, "grad_norm": 66.5, "learning_rate": 6.050847457627118e-07, "loss": 1.4573, "step": 12150 }, { "epoch": 0.3, "grad_norm": 69.5, "learning_rate": 6.047457627118644e-07, "loss": 1.4364, "step": 12160 }, { "epoch": 0.3, "grad_norm": 68.0, "learning_rate": 6.044067796610169e-07, "loss": 1.4912, "step": 12170 }, { "epoch": 0.3, "grad_norm": 65.5, "learning_rate": 6.040677966101695e-07, "loss": 1.4458, "step": 12180 }, { "epoch": 0.3, "grad_norm": 66.5, "learning_rate": 6.037288135593221e-07, "loss": 1.408, "step": 12190 }, { "epoch": 0.3, "grad_norm": 65.0, "learning_rate": 6.033898305084746e-07, "loss": 1.4915, "step": 12200 }, { "epoch": 0.31, "grad_norm": 67.0, "learning_rate": 6.030508474576271e-07, "loss": 1.4723, "step": 12210 }, { "epoch": 0.31, "grad_norm": 64.0, "learning_rate": 6.027118644067796e-07, "loss": 1.4804, "step": 12220 }, { "epoch": 0.31, "grad_norm": 69.0, "learning_rate": 6.023728813559323e-07, "loss": 1.4667, "step": 12230 }, { "epoch": 0.31, "grad_norm": 72.5, "learning_rate": 6.020338983050847e-07, "loss": 1.5079, "step": 12240 }, { "epoch": 0.31, "grad_norm": 66.5, "learning_rate": 6.016949152542372e-07, "loss": 1.4185, "step": 12250 }, { "epoch": 0.31, "grad_norm": 65.5, "learning_rate": 6.013559322033898e-07, "loss": 1.4453, "step": 12260 }, { "epoch": 0.31, "grad_norm": 66.5, "learning_rate": 6.010169491525423e-07, "loss": 1.4486, "step": 12270 }, { "epoch": 0.31, "grad_norm": 67.0, "learning_rate": 6.006779661016948e-07, "loss": 1.5067, "step": 12280 }, { "epoch": 0.31, "grad_norm": 67.0, "learning_rate": 6.003389830508474e-07, "loss": 1.4852, "step": 12290 }, { "epoch": 0.31, "grad_norm": 67.0, "learning_rate": 6e-07, "loss": 1.4627, "step": 12300 }, { "epoch": 0.31, "grad_norm": 69.5, "learning_rate": 5.996610169491525e-07, "loss": 1.4469, "step": 12310 }, { "epoch": 0.31, "grad_norm": 74.0, "learning_rate": 5.99322033898305e-07, "loss": 1.4401, "step": 12320 }, { "epoch": 0.31, "grad_norm": 70.5, "learning_rate": 5.989830508474576e-07, "loss": 1.4844, "step": 12330 }, { "epoch": 0.31, "grad_norm": 68.5, "learning_rate": 5.986440677966102e-07, "loss": 1.452, "step": 12340 }, { "epoch": 0.31, "grad_norm": 63.5, "learning_rate": 5.983050847457627e-07, "loss": 1.4492, "step": 12350 }, { "epoch": 0.31, "grad_norm": 69.0, "learning_rate": 5.979661016949152e-07, "loss": 1.411, "step": 12360 }, { "epoch": 0.31, "grad_norm": 66.0, "learning_rate": 5.976271186440678e-07, "loss": 1.4364, "step": 12370 }, { "epoch": 0.31, "grad_norm": 71.0, "learning_rate": 5.972881355932203e-07, "loss": 1.4932, "step": 12380 }, { "epoch": 0.31, "grad_norm": 64.5, "learning_rate": 5.969491525423728e-07, "loss": 1.4521, "step": 12390 }, { "epoch": 0.31, "grad_norm": 68.5, "learning_rate": 5.966101694915255e-07, "loss": 1.4953, "step": 12400 }, { "epoch": 0.31, "grad_norm": 64.5, "learning_rate": 5.96271186440678e-07, "loss": 1.471, "step": 12410 }, { "epoch": 0.31, "grad_norm": 68.0, "learning_rate": 5.959322033898305e-07, "loss": 1.4628, "step": 12420 }, { "epoch": 0.31, "grad_norm": 70.0, "learning_rate": 5.95593220338983e-07, "loss": 1.4655, "step": 12430 }, { "epoch": 0.31, "grad_norm": 66.0, "learning_rate": 5.952542372881356e-07, "loss": 1.4996, "step": 12440 }, { "epoch": 0.31, "grad_norm": 65.5, "learning_rate": 5.949152542372881e-07, "loss": 1.4829, "step": 12450 }, { "epoch": 0.31, "grad_norm": 66.5, "learning_rate": 5.945762711864406e-07, "loss": 1.467, "step": 12460 }, { "epoch": 0.31, "grad_norm": 66.5, "learning_rate": 5.942372881355932e-07, "loss": 1.4451, "step": 12470 }, { "epoch": 0.31, "grad_norm": 63.5, "learning_rate": 5.938983050847457e-07, "loss": 1.4607, "step": 12480 }, { "epoch": 0.31, "grad_norm": 66.0, "learning_rate": 5.935593220338982e-07, "loss": 1.4551, "step": 12490 }, { "epoch": 0.31, "grad_norm": 65.0, "learning_rate": 5.932203389830508e-07, "loss": 1.4357, "step": 12500 }, { "epoch": 0.31, "grad_norm": 65.5, "learning_rate": 5.928813559322034e-07, "loss": 1.4846, "step": 12510 }, { "epoch": 0.31, "grad_norm": 68.5, "learning_rate": 5.925423728813559e-07, "loss": 1.4669, "step": 12520 }, { "epoch": 0.31, "grad_norm": 65.5, "learning_rate": 5.922033898305084e-07, "loss": 1.4722, "step": 12530 }, { "epoch": 0.31, "grad_norm": 67.5, "learning_rate": 5.91864406779661e-07, "loss": 1.4802, "step": 12540 }, { "epoch": 0.31, "grad_norm": 66.5, "learning_rate": 5.915254237288135e-07, "loss": 1.4242, "step": 12550 }, { "epoch": 0.31, "grad_norm": 69.5, "learning_rate": 5.911864406779661e-07, "loss": 1.4311, "step": 12560 }, { "epoch": 0.31, "grad_norm": 66.5, "learning_rate": 5.908474576271186e-07, "loss": 1.4455, "step": 12570 }, { "epoch": 0.31, "grad_norm": 64.5, "learning_rate": 5.905084745762712e-07, "loss": 1.4349, "step": 12580 }, { "epoch": 0.31, "grad_norm": 64.5, "learning_rate": 5.901694915254237e-07, "loss": 1.4402, "step": 12590 }, { "epoch": 0.32, "grad_norm": 68.5, "learning_rate": 5.898305084745762e-07, "loss": 1.4354, "step": 12600 }, { "epoch": 0.32, "grad_norm": 68.0, "learning_rate": 5.894915254237289e-07, "loss": 1.5072, "step": 12610 }, { "epoch": 0.32, "grad_norm": 69.5, "learning_rate": 5.891525423728814e-07, "loss": 1.4497, "step": 12620 }, { "epoch": 0.32, "grad_norm": 66.5, "learning_rate": 5.888135593220339e-07, "loss": 1.4666, "step": 12630 }, { "epoch": 0.32, "grad_norm": 67.5, "learning_rate": 5.884745762711864e-07, "loss": 1.4904, "step": 12640 }, { "epoch": 0.32, "grad_norm": 67.0, "learning_rate": 5.88135593220339e-07, "loss": 1.4449, "step": 12650 }, { "epoch": 0.32, "grad_norm": 63.75, "learning_rate": 5.877966101694915e-07, "loss": 1.4449, "step": 12660 }, { "epoch": 0.32, "grad_norm": 64.0, "learning_rate": 5.87457627118644e-07, "loss": 1.465, "step": 12670 }, { "epoch": 0.32, "grad_norm": 66.0, "learning_rate": 5.871186440677966e-07, "loss": 1.4671, "step": 12680 }, { "epoch": 0.32, "grad_norm": 66.0, "learning_rate": 5.867796610169491e-07, "loss": 1.4559, "step": 12690 }, { "epoch": 0.32, "grad_norm": 66.0, "learning_rate": 5.864406779661016e-07, "loss": 1.4556, "step": 12700 }, { "epoch": 0.32, "grad_norm": 68.0, "learning_rate": 5.861016949152541e-07, "loss": 1.4432, "step": 12710 }, { "epoch": 0.32, "grad_norm": 69.5, "learning_rate": 5.857627118644068e-07, "loss": 1.4757, "step": 12720 }, { "epoch": 0.32, "grad_norm": 74.0, "learning_rate": 5.854237288135593e-07, "loss": 1.4858, "step": 12730 }, { "epoch": 0.32, "grad_norm": 71.0, "learning_rate": 5.850847457627118e-07, "loss": 1.4992, "step": 12740 }, { "epoch": 0.32, "grad_norm": 63.75, "learning_rate": 5.847457627118644e-07, "loss": 1.4566, "step": 12750 }, { "epoch": 0.32, "grad_norm": 70.5, "learning_rate": 5.844067796610169e-07, "loss": 1.4192, "step": 12760 }, { "epoch": 0.32, "grad_norm": 67.5, "learning_rate": 5.840677966101695e-07, "loss": 1.4498, "step": 12770 }, { "epoch": 0.32, "grad_norm": 69.5, "learning_rate": 5.83728813559322e-07, "loss": 1.467, "step": 12780 }, { "epoch": 0.32, "grad_norm": 67.0, "learning_rate": 5.833898305084746e-07, "loss": 1.4984, "step": 12790 }, { "epoch": 0.32, "grad_norm": 65.0, "learning_rate": 5.830508474576271e-07, "loss": 1.4734, "step": 12800 }, { "epoch": 0.32, "grad_norm": 69.0, "learning_rate": 5.827118644067796e-07, "loss": 1.4828, "step": 12810 }, { "epoch": 0.32, "grad_norm": 65.5, "learning_rate": 5.823728813559323e-07, "loss": 1.4826, "step": 12820 }, { "epoch": 0.32, "grad_norm": 70.0, "learning_rate": 5.820338983050848e-07, "loss": 1.4689, "step": 12830 }, { "epoch": 0.32, "grad_norm": 67.0, "learning_rate": 5.816949152542373e-07, "loss": 1.5024, "step": 12840 }, { "epoch": 0.32, "grad_norm": 65.0, "learning_rate": 5.813559322033898e-07, "loss": 1.4602, "step": 12850 }, { "epoch": 0.32, "grad_norm": 68.5, "learning_rate": 5.810169491525424e-07, "loss": 1.5411, "step": 12860 }, { "epoch": 0.32, "grad_norm": 71.5, "learning_rate": 5.806779661016948e-07, "loss": 1.4612, "step": 12870 }, { "epoch": 0.32, "grad_norm": 65.5, "learning_rate": 5.803389830508474e-07, "loss": 1.4638, "step": 12880 }, { "epoch": 0.32, "grad_norm": 66.0, "learning_rate": 5.8e-07, "loss": 1.4767, "step": 12890 }, { "epoch": 0.32, "grad_norm": 70.0, "learning_rate": 5.796610169491525e-07, "loss": 1.4964, "step": 12900 }, { "epoch": 0.32, "grad_norm": 69.0, "learning_rate": 5.79322033898305e-07, "loss": 1.4746, "step": 12910 }, { "epoch": 0.32, "grad_norm": 70.5, "learning_rate": 5.789830508474575e-07, "loss": 1.3965, "step": 12920 }, { "epoch": 0.32, "grad_norm": 68.0, "learning_rate": 5.786440677966102e-07, "loss": 1.4572, "step": 12930 }, { "epoch": 0.32, "grad_norm": 62.5, "learning_rate": 5.783050847457627e-07, "loss": 1.4019, "step": 12940 }, { "epoch": 0.32, "grad_norm": 63.75, "learning_rate": 5.779661016949152e-07, "loss": 1.4887, "step": 12950 }, { "epoch": 0.32, "grad_norm": 64.0, "learning_rate": 5.776271186440678e-07, "loss": 1.4603, "step": 12960 }, { "epoch": 0.32, "grad_norm": 62.75, "learning_rate": 5.772881355932203e-07, "loss": 1.4077, "step": 12970 }, { "epoch": 0.32, "grad_norm": 66.5, "learning_rate": 5.769491525423729e-07, "loss": 1.4842, "step": 12980 }, { "epoch": 0.32, "grad_norm": 65.0, "learning_rate": 5.766101694915254e-07, "loss": 1.4644, "step": 12990 }, { "epoch": 0.33, "grad_norm": 66.0, "learning_rate": 5.76271186440678e-07, "loss": 1.5002, "step": 13000 }, { "epoch": 0.33, "grad_norm": 65.5, "learning_rate": 5.759322033898305e-07, "loss": 1.4804, "step": 13010 }, { "epoch": 0.33, "grad_norm": 64.5, "learning_rate": 5.75593220338983e-07, "loss": 1.4043, "step": 13020 }, { "epoch": 0.33, "grad_norm": 67.0, "learning_rate": 5.752542372881356e-07, "loss": 1.4611, "step": 13030 }, { "epoch": 0.33, "grad_norm": 64.0, "learning_rate": 5.749152542372882e-07, "loss": 1.475, "step": 13040 }, { "epoch": 0.33, "grad_norm": 67.0, "learning_rate": 5.745762711864407e-07, "loss": 1.4915, "step": 13050 }, { "epoch": 0.33, "grad_norm": 66.5, "learning_rate": 5.742372881355932e-07, "loss": 1.4847, "step": 13060 }, { "epoch": 0.33, "grad_norm": 69.0, "learning_rate": 5.738983050847457e-07, "loss": 1.4703, "step": 13070 }, { "epoch": 0.33, "grad_norm": 68.0, "learning_rate": 5.735593220338982e-07, "loss": 1.4717, "step": 13080 }, { "epoch": 0.33, "grad_norm": 68.5, "learning_rate": 5.732203389830508e-07, "loss": 1.4517, "step": 13090 }, { "epoch": 0.33, "grad_norm": 64.0, "learning_rate": 5.728813559322034e-07, "loss": 1.4324, "step": 13100 }, { "epoch": 0.33, "grad_norm": 77.0, "learning_rate": 5.725423728813559e-07, "loss": 1.5026, "step": 13110 }, { "epoch": 0.33, "grad_norm": 67.5, "learning_rate": 5.722033898305084e-07, "loss": 1.4569, "step": 13120 }, { "epoch": 0.33, "grad_norm": 68.0, "learning_rate": 5.718644067796609e-07, "loss": 1.4446, "step": 13130 }, { "epoch": 0.33, "grad_norm": 65.5, "learning_rate": 5.715254237288136e-07, "loss": 1.5021, "step": 13140 }, { "epoch": 0.33, "grad_norm": 66.0, "learning_rate": 5.711864406779661e-07, "loss": 1.4452, "step": 13150 }, { "epoch": 0.33, "grad_norm": 64.5, "learning_rate": 5.708474576271186e-07, "loss": 1.4473, "step": 13160 }, { "epoch": 0.33, "grad_norm": 65.0, "learning_rate": 5.705084745762712e-07, "loss": 1.4854, "step": 13170 }, { "epoch": 0.33, "grad_norm": 63.0, "learning_rate": 5.701694915254237e-07, "loss": 1.4364, "step": 13180 }, { "epoch": 0.33, "grad_norm": 66.5, "learning_rate": 5.698305084745762e-07, "loss": 1.4456, "step": 13190 }, { "epoch": 0.33, "grad_norm": 74.0, "learning_rate": 5.694915254237288e-07, "loss": 1.4668, "step": 13200 }, { "epoch": 0.33, "grad_norm": 67.0, "learning_rate": 5.691525423728814e-07, "loss": 1.4956, "step": 13210 }, { "epoch": 0.33, "grad_norm": 73.0, "learning_rate": 5.688135593220339e-07, "loss": 1.459, "step": 13220 }, { "epoch": 0.33, "grad_norm": 70.0, "learning_rate": 5.684745762711864e-07, "loss": 1.4574, "step": 13230 }, { "epoch": 0.33, "grad_norm": 65.0, "learning_rate": 5.68135593220339e-07, "loss": 1.478, "step": 13240 }, { "epoch": 0.33, "grad_norm": 69.0, "learning_rate": 5.677966101694916e-07, "loss": 1.4499, "step": 13250 }, { "epoch": 0.33, "grad_norm": 65.5, "learning_rate": 5.674576271186441e-07, "loss": 1.5174, "step": 13260 }, { "epoch": 0.33, "grad_norm": 64.5, "learning_rate": 5.671186440677966e-07, "loss": 1.4631, "step": 13270 }, { "epoch": 0.33, "grad_norm": 65.0, "learning_rate": 5.667796610169491e-07, "loss": 1.4167, "step": 13280 }, { "epoch": 0.33, "grad_norm": 68.5, "learning_rate": 5.664406779661016e-07, "loss": 1.4892, "step": 13290 }, { "epoch": 0.33, "grad_norm": 66.5, "learning_rate": 5.661016949152541e-07, "loss": 1.4651, "step": 13300 }, { "epoch": 0.33, "grad_norm": 72.0, "learning_rate": 5.657627118644068e-07, "loss": 1.5192, "step": 13310 }, { "epoch": 0.33, "grad_norm": 67.5, "learning_rate": 5.654237288135593e-07, "loss": 1.491, "step": 13320 }, { "epoch": 0.33, "grad_norm": 67.5, "learning_rate": 5.650847457627118e-07, "loss": 1.4923, "step": 13330 }, { "epoch": 0.33, "grad_norm": 63.75, "learning_rate": 5.647457627118643e-07, "loss": 1.5122, "step": 13340 }, { "epoch": 0.33, "grad_norm": 65.0, "learning_rate": 5.644067796610169e-07, "loss": 1.4867, "step": 13350 }, { "epoch": 0.33, "grad_norm": 63.0, "learning_rate": 5.640677966101695e-07, "loss": 1.4223, "step": 13360 }, { "epoch": 0.33, "grad_norm": 67.0, "learning_rate": 5.63728813559322e-07, "loss": 1.4503, "step": 13370 }, { "epoch": 0.33, "grad_norm": 66.5, "learning_rate": 5.633898305084745e-07, "loss": 1.4825, "step": 13380 }, { "epoch": 0.33, "grad_norm": 67.5, "learning_rate": 5.630508474576271e-07, "loss": 1.4869, "step": 13390 }, { "epoch": 0.34, "grad_norm": 70.5, "learning_rate": 5.627118644067796e-07, "loss": 1.476, "step": 13400 }, { "epoch": 0.34, "grad_norm": 68.0, "learning_rate": 5.623728813559322e-07, "loss": 1.4165, "step": 13410 }, { "epoch": 0.34, "grad_norm": 68.5, "learning_rate": 5.620338983050848e-07, "loss": 1.4483, "step": 13420 }, { "epoch": 0.34, "grad_norm": 64.0, "learning_rate": 5.616949152542373e-07, "loss": 1.47, "step": 13430 }, { "epoch": 0.34, "grad_norm": 66.5, "learning_rate": 5.613559322033898e-07, "loss": 1.4823, "step": 13440 }, { "epoch": 0.34, "grad_norm": 66.0, "learning_rate": 5.610169491525423e-07, "loss": 1.4556, "step": 13450 }, { "epoch": 0.34, "grad_norm": 72.0, "learning_rate": 5.60677966101695e-07, "loss": 1.489, "step": 13460 }, { "epoch": 0.34, "grad_norm": 69.0, "learning_rate": 5.603389830508475e-07, "loss": 1.4897, "step": 13470 }, { "epoch": 0.34, "grad_norm": 73.5, "learning_rate": 5.6e-07, "loss": 1.4651, "step": 13480 }, { "epoch": 0.34, "grad_norm": 65.0, "learning_rate": 5.596610169491525e-07, "loss": 1.4848, "step": 13490 }, { "epoch": 0.34, "grad_norm": 72.0, "learning_rate": 5.59322033898305e-07, "loss": 1.526, "step": 13500 }, { "epoch": 0.34, "grad_norm": 67.5, "learning_rate": 5.589830508474575e-07, "loss": 1.4349, "step": 13510 }, { "epoch": 0.34, "grad_norm": 66.0, "learning_rate": 5.586440677966101e-07, "loss": 1.4623, "step": 13520 }, { "epoch": 0.34, "grad_norm": 67.5, "learning_rate": 5.583050847457627e-07, "loss": 1.4684, "step": 13530 }, { "epoch": 0.34, "grad_norm": 70.0, "learning_rate": 5.579661016949152e-07, "loss": 1.4291, "step": 13540 }, { "epoch": 0.34, "grad_norm": 67.0, "learning_rate": 5.576271186440677e-07, "loss": 1.4468, "step": 13550 }, { "epoch": 0.34, "grad_norm": 66.5, "learning_rate": 5.572881355932203e-07, "loss": 1.5005, "step": 13560 }, { "epoch": 0.34, "grad_norm": 65.5, "learning_rate": 5.569491525423729e-07, "loss": 1.4544, "step": 13570 }, { "epoch": 0.34, "grad_norm": 67.0, "learning_rate": 5.566101694915254e-07, "loss": 1.4554, "step": 13580 }, { "epoch": 0.34, "grad_norm": 64.5, "learning_rate": 5.56271186440678e-07, "loss": 1.4661, "step": 13590 }, { "epoch": 0.34, "grad_norm": 65.0, "learning_rate": 5.559322033898305e-07, "loss": 1.4719, "step": 13600 }, { "epoch": 0.34, "grad_norm": 69.0, "learning_rate": 5.55593220338983e-07, "loss": 1.4078, "step": 13610 }, { "epoch": 0.34, "grad_norm": 64.0, "learning_rate": 5.552542372881356e-07, "loss": 1.434, "step": 13620 }, { "epoch": 0.34, "grad_norm": 65.5, "learning_rate": 5.549152542372882e-07, "loss": 1.4672, "step": 13630 }, { "epoch": 0.34, "grad_norm": 67.5, "learning_rate": 5.545762711864407e-07, "loss": 1.4137, "step": 13640 }, { "epoch": 0.34, "grad_norm": 66.5, "learning_rate": 5.542372881355932e-07, "loss": 1.4375, "step": 13650 }, { "epoch": 0.34, "grad_norm": 67.5, "learning_rate": 5.538983050847457e-07, "loss": 1.4989, "step": 13660 }, { "epoch": 0.34, "grad_norm": 68.0, "learning_rate": 5.535593220338983e-07, "loss": 1.5068, "step": 13670 }, { "epoch": 0.34, "grad_norm": 65.5, "learning_rate": 5.532203389830509e-07, "loss": 1.4507, "step": 13680 }, { "epoch": 0.34, "grad_norm": 68.5, "learning_rate": 5.528813559322034e-07, "loss": 1.5146, "step": 13690 }, { "epoch": 0.34, "grad_norm": 68.5, "learning_rate": 5.525423728813559e-07, "loss": 1.4665, "step": 13700 }, { "epoch": 0.34, "grad_norm": 67.0, "learning_rate": 5.522033898305084e-07, "loss": 1.4784, "step": 13710 }, { "epoch": 0.34, "grad_norm": 70.5, "learning_rate": 5.518644067796609e-07, "loss": 1.4776, "step": 13720 }, { "epoch": 0.34, "grad_norm": 66.5, "learning_rate": 5.515254237288135e-07, "loss": 1.4302, "step": 13730 }, { "epoch": 0.34, "grad_norm": 66.0, "learning_rate": 5.511864406779661e-07, "loss": 1.4244, "step": 13740 }, { "epoch": 0.34, "grad_norm": 73.0, "learning_rate": 5.508474576271186e-07, "loss": 1.498, "step": 13750 }, { "epoch": 0.34, "grad_norm": 63.5, "learning_rate": 5.505084745762711e-07, "loss": 1.4754, "step": 13760 }, { "epoch": 0.34, "grad_norm": 65.5, "learning_rate": 5.501694915254237e-07, "loss": 1.4721, "step": 13770 }, { "epoch": 0.34, "grad_norm": 67.5, "learning_rate": 5.498305084745762e-07, "loss": 1.4564, "step": 13780 }, { "epoch": 0.34, "grad_norm": 66.0, "learning_rate": 5.494915254237288e-07, "loss": 1.4523, "step": 13790 }, { "epoch": 0.34, "grad_norm": 66.0, "learning_rate": 5.491525423728813e-07, "loss": 1.4356, "step": 13800 }, { "epoch": 0.35, "grad_norm": 65.0, "learning_rate": 5.488135593220339e-07, "loss": 1.4499, "step": 13810 }, { "epoch": 0.35, "grad_norm": 68.0, "learning_rate": 5.484745762711864e-07, "loss": 1.4299, "step": 13820 }, { "epoch": 0.35, "grad_norm": 67.5, "learning_rate": 5.481355932203389e-07, "loss": 1.5106, "step": 13830 }, { "epoch": 0.35, "grad_norm": 66.5, "learning_rate": 5.477966101694916e-07, "loss": 1.4445, "step": 13840 }, { "epoch": 0.35, "grad_norm": 70.0, "learning_rate": 5.474576271186441e-07, "loss": 1.4221, "step": 13850 }, { "epoch": 0.35, "grad_norm": 66.5, "learning_rate": 5.471186440677966e-07, "loss": 1.453, "step": 13860 }, { "epoch": 0.35, "grad_norm": 67.0, "learning_rate": 5.467796610169491e-07, "loss": 1.4445, "step": 13870 }, { "epoch": 0.35, "grad_norm": 70.5, "learning_rate": 5.464406779661017e-07, "loss": 1.4593, "step": 13880 }, { "epoch": 0.35, "grad_norm": 68.0, "learning_rate": 5.461016949152543e-07, "loss": 1.4751, "step": 13890 }, { "epoch": 0.35, "grad_norm": 67.5, "learning_rate": 5.457627118644067e-07, "loss": 1.4619, "step": 13900 }, { "epoch": 0.35, "grad_norm": 67.5, "learning_rate": 5.454237288135593e-07, "loss": 1.4717, "step": 13910 }, { "epoch": 0.35, "grad_norm": 65.0, "learning_rate": 5.450847457627118e-07, "loss": 1.4945, "step": 13920 }, { "epoch": 0.35, "grad_norm": 66.5, "learning_rate": 5.447457627118643e-07, "loss": 1.4218, "step": 13930 }, { "epoch": 0.35, "grad_norm": 66.5, "learning_rate": 5.444067796610168e-07, "loss": 1.4192, "step": 13940 }, { "epoch": 0.35, "grad_norm": 68.5, "learning_rate": 5.440677966101695e-07, "loss": 1.4425, "step": 13950 }, { "epoch": 0.35, "grad_norm": 67.5, "learning_rate": 5.43728813559322e-07, "loss": 1.4843, "step": 13960 }, { "epoch": 0.35, "grad_norm": 70.0, "learning_rate": 5.433898305084745e-07, "loss": 1.4605, "step": 13970 }, { "epoch": 0.35, "grad_norm": 67.5, "learning_rate": 5.430508474576271e-07, "loss": 1.4867, "step": 13980 }, { "epoch": 0.35, "grad_norm": 66.0, "learning_rate": 5.427118644067796e-07, "loss": 1.4419, "step": 13990 }, { "epoch": 0.35, "grad_norm": 64.0, "learning_rate": 5.423728813559322e-07, "loss": 1.4836, "step": 14000 }, { "epoch": 0.35, "grad_norm": 64.5, "learning_rate": 5.420338983050847e-07, "loss": 1.4472, "step": 14010 }, { "epoch": 0.35, "grad_norm": 66.5, "learning_rate": 5.416949152542373e-07, "loss": 1.3945, "step": 14020 }, { "epoch": 0.35, "grad_norm": 67.5, "learning_rate": 5.413559322033898e-07, "loss": 1.4432, "step": 14030 }, { "epoch": 0.35, "grad_norm": 66.0, "learning_rate": 5.410169491525423e-07, "loss": 1.4193, "step": 14040 }, { "epoch": 0.35, "grad_norm": 69.5, "learning_rate": 5.40677966101695e-07, "loss": 1.4406, "step": 14050 }, { "epoch": 0.35, "grad_norm": 64.5, "learning_rate": 5.403389830508475e-07, "loss": 1.4219, "step": 14060 }, { "epoch": 0.35, "grad_norm": 67.5, "learning_rate": 5.4e-07, "loss": 1.4649, "step": 14070 }, { "epoch": 0.35, "grad_norm": 70.0, "learning_rate": 5.396610169491525e-07, "loss": 1.4533, "step": 14080 }, { "epoch": 0.35, "grad_norm": 69.5, "learning_rate": 5.393220338983051e-07, "loss": 1.421, "step": 14090 }, { "epoch": 0.35, "grad_norm": 66.0, "learning_rate": 5.389830508474577e-07, "loss": 1.4834, "step": 14100 }, { "epoch": 0.35, "grad_norm": 69.0, "learning_rate": 5.386440677966101e-07, "loss": 1.4371, "step": 14110 }, { "epoch": 0.35, "grad_norm": 69.5, "learning_rate": 5.383050847457627e-07, "loss": 1.4858, "step": 14120 }, { "epoch": 0.35, "grad_norm": 70.5, "learning_rate": 5.379661016949152e-07, "loss": 1.4433, "step": 14130 }, { "epoch": 0.35, "grad_norm": 66.5, "learning_rate": 5.376271186440677e-07, "loss": 1.3802, "step": 14140 }, { "epoch": 0.35, "grad_norm": 63.0, "learning_rate": 5.372881355932202e-07, "loss": 1.4092, "step": 14150 }, { "epoch": 0.35, "grad_norm": 65.0, "learning_rate": 5.369491525423729e-07, "loss": 1.486, "step": 14160 }, { "epoch": 0.35, "grad_norm": 66.0, "learning_rate": 5.366101694915254e-07, "loss": 1.4519, "step": 14170 }, { "epoch": 0.35, "grad_norm": 67.0, "learning_rate": 5.362711864406779e-07, "loss": 1.4361, "step": 14180 }, { "epoch": 0.35, "grad_norm": 68.5, "learning_rate": 5.359322033898305e-07, "loss": 1.494, "step": 14190 }, { "epoch": 0.35, "grad_norm": 68.5, "learning_rate": 5.35593220338983e-07, "loss": 1.4649, "step": 14200 }, { "epoch": 0.36, "grad_norm": 62.5, "learning_rate": 5.352542372881356e-07, "loss": 1.4777, "step": 14210 }, { "epoch": 0.36, "grad_norm": 62.25, "learning_rate": 5.349152542372881e-07, "loss": 1.4786, "step": 14220 }, { "epoch": 0.36, "grad_norm": 67.0, "learning_rate": 5.345762711864407e-07, "loss": 1.4566, "step": 14230 }, { "epoch": 0.36, "grad_norm": 69.5, "learning_rate": 5.342372881355932e-07, "loss": 1.4531, "step": 14240 }, { "epoch": 0.36, "grad_norm": 64.0, "learning_rate": 5.338983050847457e-07, "loss": 1.4665, "step": 14250 }, { "epoch": 0.36, "grad_norm": 67.0, "learning_rate": 5.335593220338983e-07, "loss": 1.4315, "step": 14260 }, { "epoch": 0.36, "grad_norm": 65.5, "learning_rate": 5.332203389830509e-07, "loss": 1.4696, "step": 14270 }, { "epoch": 0.36, "grad_norm": 70.0, "learning_rate": 5.328813559322034e-07, "loss": 1.5306, "step": 14280 }, { "epoch": 0.36, "grad_norm": 69.0, "learning_rate": 5.325423728813559e-07, "loss": 1.4386, "step": 14290 }, { "epoch": 0.36, "grad_norm": 67.5, "learning_rate": 5.322033898305085e-07, "loss": 1.5052, "step": 14300 }, { "epoch": 0.36, "grad_norm": 66.0, "learning_rate": 5.31864406779661e-07, "loss": 1.4745, "step": 14310 }, { "epoch": 0.36, "grad_norm": 67.5, "learning_rate": 5.315254237288135e-07, "loss": 1.4627, "step": 14320 }, { "epoch": 0.36, "grad_norm": 65.0, "learning_rate": 5.311864406779661e-07, "loss": 1.4817, "step": 14330 }, { "epoch": 0.36, "grad_norm": 67.0, "learning_rate": 5.308474576271186e-07, "loss": 1.4466, "step": 14340 }, { "epoch": 0.36, "grad_norm": 67.5, "learning_rate": 5.305084745762711e-07, "loss": 1.4522, "step": 14350 }, { "epoch": 0.36, "grad_norm": 66.0, "learning_rate": 5.301694915254236e-07, "loss": 1.4946, "step": 14360 }, { "epoch": 0.36, "grad_norm": 67.0, "learning_rate": 5.298305084745763e-07, "loss": 1.4345, "step": 14370 }, { "epoch": 0.36, "grad_norm": 66.5, "learning_rate": 5.294915254237288e-07, "loss": 1.4467, "step": 14380 }, { "epoch": 0.36, "grad_norm": 68.0, "learning_rate": 5.291525423728813e-07, "loss": 1.4991, "step": 14390 }, { "epoch": 0.36, "grad_norm": 68.5, "learning_rate": 5.288135593220339e-07, "loss": 1.4386, "step": 14400 }, { "epoch": 0.36, "grad_norm": 66.0, "learning_rate": 5.284745762711864e-07, "loss": 1.4437, "step": 14410 }, { "epoch": 0.36, "grad_norm": 68.0, "learning_rate": 5.281355932203389e-07, "loss": 1.4982, "step": 14420 }, { "epoch": 0.36, "grad_norm": 65.0, "learning_rate": 5.277966101694915e-07, "loss": 1.4396, "step": 14430 }, { "epoch": 0.36, "grad_norm": 66.0, "learning_rate": 5.274576271186441e-07, "loss": 1.4493, "step": 14440 }, { "epoch": 0.36, "grad_norm": 65.5, "learning_rate": 5.271186440677966e-07, "loss": 1.4816, "step": 14450 }, { "epoch": 0.36, "grad_norm": 70.0, "learning_rate": 5.267796610169491e-07, "loss": 1.4594, "step": 14460 }, { "epoch": 0.36, "grad_norm": 68.5, "learning_rate": 5.264406779661017e-07, "loss": 1.4462, "step": 14470 }, { "epoch": 0.36, "grad_norm": 66.5, "learning_rate": 5.261016949152543e-07, "loss": 1.4388, "step": 14480 }, { "epoch": 0.36, "grad_norm": 65.0, "learning_rate": 5.257627118644068e-07, "loss": 1.4086, "step": 14490 }, { "epoch": 0.36, "grad_norm": 67.0, "learning_rate": 5.254237288135593e-07, "loss": 1.495, "step": 14500 }, { "epoch": 0.36, "grad_norm": 66.0, "learning_rate": 5.250847457627119e-07, "loss": 1.4207, "step": 14510 }, { "epoch": 0.36, "grad_norm": 68.5, "learning_rate": 5.247457627118644e-07, "loss": 1.4836, "step": 14520 }, { "epoch": 0.36, "grad_norm": 66.5, "learning_rate": 5.244067796610169e-07, "loss": 1.477, "step": 14530 }, { "epoch": 0.36, "grad_norm": 67.0, "learning_rate": 5.240677966101695e-07, "loss": 1.4873, "step": 14540 }, { "epoch": 0.36, "grad_norm": 67.5, "learning_rate": 5.23728813559322e-07, "loss": 1.485, "step": 14550 }, { "epoch": 0.36, "grad_norm": 68.5, "learning_rate": 5.233898305084745e-07, "loss": 1.4944, "step": 14560 }, { "epoch": 0.36, "grad_norm": 68.0, "learning_rate": 5.23050847457627e-07, "loss": 1.4815, "step": 14570 }, { "epoch": 0.36, "grad_norm": 66.0, "learning_rate": 5.227118644067796e-07, "loss": 1.4754, "step": 14580 }, { "epoch": 0.36, "grad_norm": 70.5, "learning_rate": 5.223728813559322e-07, "loss": 1.461, "step": 14590 }, { "epoch": 0.36, "grad_norm": 65.0, "learning_rate": 5.220338983050847e-07, "loss": 1.4834, "step": 14600 }, { "epoch": 0.37, "grad_norm": 66.0, "learning_rate": 5.216949152542373e-07, "loss": 1.4707, "step": 14610 }, { "epoch": 0.37, "grad_norm": 69.0, "learning_rate": 5.213559322033898e-07, "loss": 1.4374, "step": 14620 }, { "epoch": 0.37, "grad_norm": 65.5, "learning_rate": 5.210169491525423e-07, "loss": 1.4832, "step": 14630 }, { "epoch": 0.37, "grad_norm": 65.5, "learning_rate": 5.206779661016949e-07, "loss": 1.4971, "step": 14640 }, { "epoch": 0.37, "grad_norm": 71.0, "learning_rate": 5.203389830508475e-07, "loss": 1.4704, "step": 14650 }, { "epoch": 0.37, "grad_norm": 65.0, "learning_rate": 5.2e-07, "loss": 1.4605, "step": 14660 }, { "epoch": 0.37, "grad_norm": 66.0, "learning_rate": 5.196610169491525e-07, "loss": 1.5033, "step": 14670 }, { "epoch": 0.37, "grad_norm": 68.5, "learning_rate": 5.193220338983051e-07, "loss": 1.4341, "step": 14680 }, { "epoch": 0.37, "grad_norm": 66.5, "learning_rate": 5.189830508474577e-07, "loss": 1.4542, "step": 14690 }, { "epoch": 0.37, "grad_norm": 71.0, "learning_rate": 5.186440677966102e-07, "loss": 1.4944, "step": 14700 }, { "epoch": 0.37, "grad_norm": 63.25, "learning_rate": 5.183050847457627e-07, "loss": 1.4058, "step": 14710 }, { "epoch": 0.37, "grad_norm": 69.5, "learning_rate": 5.179661016949153e-07, "loss": 1.4591, "step": 14720 }, { "epoch": 0.37, "grad_norm": 66.0, "learning_rate": 5.176271186440678e-07, "loss": 1.4877, "step": 14730 }, { "epoch": 0.37, "grad_norm": 65.0, "learning_rate": 5.172881355932202e-07, "loss": 1.4544, "step": 14740 }, { "epoch": 0.37, "grad_norm": 66.5, "learning_rate": 5.169491525423729e-07, "loss": 1.4525, "step": 14750 }, { "epoch": 0.37, "grad_norm": 66.5, "learning_rate": 5.166101694915254e-07, "loss": 1.4829, "step": 14760 }, { "epoch": 0.37, "grad_norm": 65.0, "learning_rate": 5.162711864406779e-07, "loss": 1.4331, "step": 14770 }, { "epoch": 0.37, "grad_norm": 67.0, "learning_rate": 5.159322033898304e-07, "loss": 1.4832, "step": 14780 }, { "epoch": 0.37, "grad_norm": 68.5, "learning_rate": 5.15593220338983e-07, "loss": 1.5107, "step": 14790 }, { "epoch": 0.37, "grad_norm": 65.5, "learning_rate": 5.152542372881356e-07, "loss": 1.4392, "step": 14800 }, { "epoch": 0.37, "grad_norm": 65.0, "learning_rate": 5.149152542372881e-07, "loss": 1.5373, "step": 14810 }, { "epoch": 0.37, "grad_norm": 65.5, "learning_rate": 5.145762711864407e-07, "loss": 1.4587, "step": 14820 }, { "epoch": 0.37, "grad_norm": 68.0, "learning_rate": 5.142372881355932e-07, "loss": 1.465, "step": 14830 }, { "epoch": 0.37, "grad_norm": 67.5, "learning_rate": 5.138983050847457e-07, "loss": 1.5013, "step": 14840 }, { "epoch": 0.37, "grad_norm": 66.0, "learning_rate": 5.135593220338983e-07, "loss": 1.4892, "step": 14850 }, { "epoch": 0.37, "grad_norm": 69.0, "learning_rate": 5.132203389830509e-07, "loss": 1.5096, "step": 14860 }, { "epoch": 0.37, "grad_norm": 67.5, "learning_rate": 5.128813559322034e-07, "loss": 1.4201, "step": 14870 }, { "epoch": 0.37, "grad_norm": 65.0, "learning_rate": 5.125423728813559e-07, "loss": 1.4701, "step": 14880 }, { "epoch": 0.37, "grad_norm": 65.0, "learning_rate": 5.122033898305085e-07, "loss": 1.3931, "step": 14890 }, { "epoch": 0.37, "grad_norm": 65.5, "learning_rate": 5.11864406779661e-07, "loss": 1.4622, "step": 14900 }, { "epoch": 0.37, "grad_norm": 65.0, "learning_rate": 5.115254237288136e-07, "loss": 1.4666, "step": 14910 }, { "epoch": 0.37, "grad_norm": 69.5, "learning_rate": 5.111864406779661e-07, "loss": 1.437, "step": 14920 }, { "epoch": 0.37, "grad_norm": 67.0, "learning_rate": 5.108474576271187e-07, "loss": 1.4685, "step": 14930 }, { "epoch": 0.37, "grad_norm": 75.0, "learning_rate": 5.105084745762711e-07, "loss": 1.456, "step": 14940 }, { "epoch": 0.37, "grad_norm": 67.0, "learning_rate": 5.101694915254236e-07, "loss": 1.4602, "step": 14950 }, { "epoch": 0.37, "grad_norm": 69.0, "learning_rate": 5.098305084745763e-07, "loss": 1.4573, "step": 14960 }, { "epoch": 0.37, "grad_norm": 68.0, "learning_rate": 5.094915254237288e-07, "loss": 1.4332, "step": 14970 }, { "epoch": 0.37, "grad_norm": 66.5, "learning_rate": 5.091525423728813e-07, "loss": 1.4702, "step": 14980 }, { "epoch": 0.37, "grad_norm": 65.5, "learning_rate": 5.088135593220338e-07, "loss": 1.4619, "step": 14990 }, { "epoch": 0.38, "grad_norm": 69.0, "learning_rate": 5.084745762711864e-07, "loss": 1.3842, "step": 15000 }, { "epoch": 0.38, "eval_loss": 1.4593136310577393, "eval_runtime": 68.1257, "eval_samples_per_second": 14.679, "eval_steps_per_second": 14.679, "step": 15000 }, { "epoch": 0.38, "grad_norm": 67.0, "learning_rate": 5.08135593220339e-07, "loss": 1.477, "step": 15010 }, { "epoch": 0.38, "grad_norm": 67.0, "learning_rate": 5.077966101694915e-07, "loss": 1.5053, "step": 15020 }, { "epoch": 0.38, "grad_norm": 66.5, "learning_rate": 5.07457627118644e-07, "loss": 1.458, "step": 15030 }, { "epoch": 0.38, "grad_norm": 67.5, "learning_rate": 5.071186440677966e-07, "loss": 1.4389, "step": 15040 }, { "epoch": 0.38, "grad_norm": 68.0, "learning_rate": 5.067796610169491e-07, "loss": 1.4935, "step": 15050 }, { "epoch": 0.38, "grad_norm": 66.0, "learning_rate": 5.064406779661016e-07, "loss": 1.4916, "step": 15060 }, { "epoch": 0.38, "grad_norm": 67.5, "learning_rate": 5.061016949152543e-07, "loss": 1.4686, "step": 15070 }, { "epoch": 0.38, "grad_norm": 69.5, "learning_rate": 5.057627118644068e-07, "loss": 1.4162, "step": 15080 }, { "epoch": 0.38, "grad_norm": 68.5, "learning_rate": 5.054237288135593e-07, "loss": 1.4581, "step": 15090 }, { "epoch": 0.38, "grad_norm": 66.5, "learning_rate": 5.050847457627119e-07, "loss": 1.4925, "step": 15100 }, { "epoch": 0.38, "grad_norm": 66.0, "learning_rate": 5.047457627118644e-07, "loss": 1.4789, "step": 15110 }, { "epoch": 0.38, "grad_norm": 65.0, "learning_rate": 5.04406779661017e-07, "loss": 1.4523, "step": 15120 }, { "epoch": 0.38, "grad_norm": 66.0, "learning_rate": 5.040677966101695e-07, "loss": 1.4105, "step": 15130 }, { "epoch": 0.38, "grad_norm": 66.5, "learning_rate": 5.037288135593221e-07, "loss": 1.4677, "step": 15140 }, { "epoch": 0.38, "grad_norm": 69.0, "learning_rate": 5.033898305084745e-07, "loss": 1.4524, "step": 15150 }, { "epoch": 0.38, "grad_norm": 65.0, "learning_rate": 5.03050847457627e-07, "loss": 1.4859, "step": 15160 }, { "epoch": 0.38, "grad_norm": 66.0, "learning_rate": 5.027118644067795e-07, "loss": 1.4569, "step": 15170 }, { "epoch": 0.38, "grad_norm": 66.0, "learning_rate": 5.023728813559322e-07, "loss": 1.4065, "step": 15180 }, { "epoch": 0.38, "grad_norm": 68.0, "learning_rate": 5.020338983050847e-07, "loss": 1.4856, "step": 15190 }, { "epoch": 0.38, "grad_norm": 66.0, "learning_rate": 5.016949152542372e-07, "loss": 1.5271, "step": 15200 }, { "epoch": 0.38, "grad_norm": 66.5, "learning_rate": 5.013559322033898e-07, "loss": 1.4704, "step": 15210 }, { "epoch": 0.38, "grad_norm": 66.0, "learning_rate": 5.010169491525423e-07, "loss": 1.4468, "step": 15220 }, { "epoch": 0.38, "grad_norm": 70.0, "learning_rate": 5.006779661016949e-07, "loss": 1.5051, "step": 15230 }, { "epoch": 0.38, "grad_norm": 64.0, "learning_rate": 5.003389830508474e-07, "loss": 1.4967, "step": 15240 }, { "epoch": 0.38, "grad_norm": 68.5, "learning_rate": 5e-07, "loss": 1.4862, "step": 15250 }, { "epoch": 0.38, "grad_norm": 65.5, "learning_rate": 4.996610169491525e-07, "loss": 1.5323, "step": 15260 }, { "epoch": 0.38, "grad_norm": 66.5, "learning_rate": 4.99322033898305e-07, "loss": 1.4604, "step": 15270 }, { "epoch": 0.38, "grad_norm": 70.0, "learning_rate": 4.989830508474577e-07, "loss": 1.4842, "step": 15280 }, { "epoch": 0.38, "grad_norm": 73.5, "learning_rate": 4.986440677966102e-07, "loss": 1.5155, "step": 15290 }, { "epoch": 0.38, "grad_norm": 69.5, "learning_rate": 4.983050847457627e-07, "loss": 1.4856, "step": 15300 }, { "epoch": 0.38, "grad_norm": 65.0, "learning_rate": 4.979661016949152e-07, "loss": 1.475, "step": 15310 }, { "epoch": 0.38, "grad_norm": 62.75, "learning_rate": 4.976271186440678e-07, "loss": 1.4617, "step": 15320 }, { "epoch": 0.38, "grad_norm": 65.0, "learning_rate": 4.972881355932203e-07, "loss": 1.4433, "step": 15330 }, { "epoch": 0.38, "grad_norm": 67.0, "learning_rate": 4.969491525423728e-07, "loss": 1.423, "step": 15340 }, { "epoch": 0.38, "grad_norm": 70.0, "learning_rate": 4.966101694915254e-07, "loss": 1.4587, "step": 15350 }, { "epoch": 0.38, "grad_norm": 66.0, "learning_rate": 4.96271186440678e-07, "loss": 1.4315, "step": 15360 }, { "epoch": 0.38, "grad_norm": 69.0, "learning_rate": 4.959322033898305e-07, "loss": 1.4785, "step": 15370 }, { "epoch": 0.38, "grad_norm": 67.0, "learning_rate": 4.95593220338983e-07, "loss": 1.4666, "step": 15380 }, { "epoch": 0.38, "grad_norm": 68.0, "learning_rate": 4.952542372881356e-07, "loss": 1.4239, "step": 15390 }, { "epoch": 0.39, "grad_norm": 68.0, "learning_rate": 4.949152542372881e-07, "loss": 1.4976, "step": 15400 }, { "epoch": 0.39, "grad_norm": 67.0, "learning_rate": 4.945762711864406e-07, "loss": 1.4087, "step": 15410 }, { "epoch": 0.39, "grad_norm": 67.0, "learning_rate": 4.942372881355932e-07, "loss": 1.5106, "step": 15420 }, { "epoch": 0.39, "grad_norm": 64.5, "learning_rate": 4.938983050847457e-07, "loss": 1.4234, "step": 15430 }, { "epoch": 0.39, "grad_norm": 68.0, "learning_rate": 4.935593220338983e-07, "loss": 1.4921, "step": 15440 }, { "epoch": 0.39, "grad_norm": 66.5, "learning_rate": 4.932203389830508e-07, "loss": 1.4657, "step": 15450 }, { "epoch": 0.39, "grad_norm": 67.5, "learning_rate": 4.928813559322034e-07, "loss": 1.4753, "step": 15460 }, { "epoch": 0.39, "grad_norm": 63.25, "learning_rate": 4.925423728813559e-07, "loss": 1.4472, "step": 15470 }, { "epoch": 0.39, "grad_norm": 70.0, "learning_rate": 4.922033898305084e-07, "loss": 1.4597, "step": 15480 }, { "epoch": 0.39, "grad_norm": 68.0, "learning_rate": 4.91864406779661e-07, "loss": 1.4448, "step": 15490 }, { "epoch": 0.39, "grad_norm": 70.0, "learning_rate": 4.915254237288136e-07, "loss": 1.5133, "step": 15500 }, { "epoch": 0.39, "grad_norm": 66.0, "learning_rate": 4.91186440677966e-07, "loss": 1.525, "step": 15510 }, { "epoch": 0.39, "grad_norm": 68.5, "learning_rate": 4.908474576271186e-07, "loss": 1.494, "step": 15520 }, { "epoch": 0.39, "grad_norm": 67.0, "learning_rate": 4.905084745762712e-07, "loss": 1.4544, "step": 15530 }, { "epoch": 0.39, "grad_norm": 67.0, "learning_rate": 4.901694915254237e-07, "loss": 1.4752, "step": 15540 }, { "epoch": 0.39, "grad_norm": 68.5, "learning_rate": 4.898305084745762e-07, "loss": 1.4487, "step": 15550 }, { "epoch": 0.39, "grad_norm": 64.0, "learning_rate": 4.894915254237288e-07, "loss": 1.449, "step": 15560 }, { "epoch": 0.39, "grad_norm": 67.5, "learning_rate": 4.891525423728813e-07, "loss": 1.4816, "step": 15570 }, { "epoch": 0.39, "grad_norm": 65.0, "learning_rate": 4.888135593220339e-07, "loss": 1.5077, "step": 15580 }, { "epoch": 0.39, "grad_norm": 65.0, "learning_rate": 4.884745762711864e-07, "loss": 1.4373, "step": 15590 }, { "epoch": 0.39, "grad_norm": 65.5, "learning_rate": 4.88135593220339e-07, "loss": 1.4647, "step": 15600 }, { "epoch": 0.39, "grad_norm": 64.5, "learning_rate": 4.877966101694915e-07, "loss": 1.4399, "step": 15610 }, { "epoch": 0.39, "grad_norm": 67.5, "learning_rate": 4.87457627118644e-07, "loss": 1.4177, "step": 15620 }, { "epoch": 0.39, "grad_norm": 68.0, "learning_rate": 4.871186440677966e-07, "loss": 1.4514, "step": 15630 }, { "epoch": 0.39, "grad_norm": 62.5, "learning_rate": 4.867796610169491e-07, "loss": 1.3574, "step": 15640 }, { "epoch": 0.39, "grad_norm": 66.5, "learning_rate": 4.864406779661016e-07, "loss": 1.4834, "step": 15650 }, { "epoch": 0.39, "grad_norm": 65.5, "learning_rate": 4.861016949152542e-07, "loss": 1.4246, "step": 15660 }, { "epoch": 0.39, "grad_norm": 64.0, "learning_rate": 4.857627118644068e-07, "loss": 1.4518, "step": 15670 }, { "epoch": 0.39, "grad_norm": 68.0, "learning_rate": 4.854237288135593e-07, "loss": 1.3902, "step": 15680 }, { "epoch": 0.39, "grad_norm": 67.0, "learning_rate": 4.850847457627118e-07, "loss": 1.4994, "step": 15690 }, { "epoch": 0.39, "grad_norm": 66.5, "learning_rate": 4.847457627118644e-07, "loss": 1.4048, "step": 15700 }, { "epoch": 0.39, "grad_norm": 70.5, "learning_rate": 4.84406779661017e-07, "loss": 1.4836, "step": 15710 }, { "epoch": 0.39, "grad_norm": 67.5, "learning_rate": 4.840677966101694e-07, "loss": 1.4305, "step": 15720 }, { "epoch": 0.39, "grad_norm": 68.0, "learning_rate": 4.837288135593219e-07, "loss": 1.4205, "step": 15730 }, { "epoch": 0.39, "grad_norm": 66.5, "learning_rate": 4.833898305084746e-07, "loss": 1.4252, "step": 15740 }, { "epoch": 0.39, "grad_norm": 68.0, "learning_rate": 4.830508474576271e-07, "loss": 1.4475, "step": 15750 }, { "epoch": 0.39, "grad_norm": 72.0, "learning_rate": 4.827118644067796e-07, "loss": 1.4458, "step": 15760 }, { "epoch": 0.39, "grad_norm": 70.5, "learning_rate": 4.823728813559322e-07, "loss": 1.4735, "step": 15770 }, { "epoch": 0.39, "grad_norm": 71.0, "learning_rate": 4.820338983050847e-07, "loss": 1.4911, "step": 15780 }, { "epoch": 0.39, "grad_norm": 67.5, "learning_rate": 4.816949152542373e-07, "loss": 1.5031, "step": 15790 }, { "epoch": 0.4, "grad_norm": 64.0, "learning_rate": 4.813559322033898e-07, "loss": 1.4487, "step": 15800 }, { "epoch": 0.4, "grad_norm": 66.0, "learning_rate": 4.810169491525424e-07, "loss": 1.4527, "step": 15810 }, { "epoch": 0.4, "grad_norm": 69.0, "learning_rate": 4.806779661016949e-07, "loss": 1.4017, "step": 15820 }, { "epoch": 0.4, "grad_norm": 67.0, "learning_rate": 4.803389830508474e-07, "loss": 1.4435, "step": 15830 }, { "epoch": 0.4, "grad_norm": 68.5, "learning_rate": 4.8e-07, "loss": 1.4501, "step": 15840 }, { "epoch": 0.4, "grad_norm": 66.0, "learning_rate": 4.796610169491525e-07, "loss": 1.4972, "step": 15850 }, { "epoch": 0.4, "grad_norm": 66.0, "learning_rate": 4.79322033898305e-07, "loss": 1.4248, "step": 15860 }, { "epoch": 0.4, "grad_norm": 65.5, "learning_rate": 4.789830508474576e-07, "loss": 1.4726, "step": 15870 }, { "epoch": 0.4, "grad_norm": 66.5, "learning_rate": 4.786440677966102e-07, "loss": 1.4861, "step": 15880 }, { "epoch": 0.4, "grad_norm": 65.5, "learning_rate": 4.783050847457627e-07, "loss": 1.4566, "step": 15890 }, { "epoch": 0.4, "grad_norm": 64.5, "learning_rate": 4.779661016949152e-07, "loss": 1.4693, "step": 15900 }, { "epoch": 0.4, "grad_norm": 69.5, "learning_rate": 4.776271186440678e-07, "loss": 1.447, "step": 15910 }, { "epoch": 0.4, "grad_norm": 66.0, "learning_rate": 4.772881355932204e-07, "loss": 1.4898, "step": 15920 }, { "epoch": 0.4, "grad_norm": 68.0, "learning_rate": 4.769491525423728e-07, "loss": 1.4918, "step": 15930 }, { "epoch": 0.4, "grad_norm": 64.5, "learning_rate": 4.766101694915254e-07, "loss": 1.4825, "step": 15940 }, { "epoch": 0.4, "grad_norm": 63.25, "learning_rate": 4.762711864406779e-07, "loss": 1.4576, "step": 15950 }, { "epoch": 0.4, "grad_norm": 66.0, "learning_rate": 4.759322033898305e-07, "loss": 1.4638, "step": 15960 }, { "epoch": 0.4, "grad_norm": 63.5, "learning_rate": 4.7559322033898303e-07, "loss": 1.4329, "step": 15970 }, { "epoch": 0.4, "grad_norm": 68.0, "learning_rate": 4.7525423728813556e-07, "loss": 1.4607, "step": 15980 }, { "epoch": 0.4, "grad_norm": 65.5, "learning_rate": 4.7491525423728814e-07, "loss": 1.5177, "step": 15990 }, { "epoch": 0.4, "grad_norm": 66.0, "learning_rate": 4.7457627118644066e-07, "loss": 1.4215, "step": 16000 }, { "epoch": 0.4, "grad_norm": 68.5, "learning_rate": 4.742372881355932e-07, "loss": 1.4385, "step": 16010 }, { "epoch": 0.4, "grad_norm": 65.5, "learning_rate": 4.7389830508474577e-07, "loss": 1.4134, "step": 16020 }, { "epoch": 0.4, "grad_norm": 67.0, "learning_rate": 4.7355932203389825e-07, "loss": 1.4498, "step": 16030 }, { "epoch": 0.4, "grad_norm": 71.5, "learning_rate": 4.7322033898305083e-07, "loss": 1.4621, "step": 16040 }, { "epoch": 0.4, "grad_norm": 67.5, "learning_rate": 4.7288135593220336e-07, "loss": 1.4642, "step": 16050 }, { "epoch": 0.4, "grad_norm": 61.25, "learning_rate": 4.725423728813559e-07, "loss": 1.4379, "step": 16060 }, { "epoch": 0.4, "grad_norm": 66.5, "learning_rate": 4.7220338983050846e-07, "loss": 1.3932, "step": 16070 }, { "epoch": 0.4, "grad_norm": 65.5, "learning_rate": 4.71864406779661e-07, "loss": 1.4649, "step": 16080 }, { "epoch": 0.4, "grad_norm": 65.5, "learning_rate": 4.715254237288135e-07, "loss": 1.4559, "step": 16090 }, { "epoch": 0.4, "grad_norm": 67.5, "learning_rate": 4.711864406779661e-07, "loss": 1.4719, "step": 16100 }, { "epoch": 0.4, "grad_norm": 65.5, "learning_rate": 4.7084745762711863e-07, "loss": 1.471, "step": 16110 }, { "epoch": 0.4, "grad_norm": 65.0, "learning_rate": 4.705084745762712e-07, "loss": 1.4591, "step": 16120 }, { "epoch": 0.4, "grad_norm": 66.5, "learning_rate": 4.7016949152542374e-07, "loss": 1.4673, "step": 16130 }, { "epoch": 0.4, "grad_norm": 67.5, "learning_rate": 4.698305084745762e-07, "loss": 1.4716, "step": 16140 }, { "epoch": 0.4, "grad_norm": 66.0, "learning_rate": 4.694915254237288e-07, "loss": 1.4219, "step": 16150 }, { "epoch": 0.4, "grad_norm": 69.0, "learning_rate": 4.691525423728813e-07, "loss": 1.429, "step": 16160 }, { "epoch": 0.4, "grad_norm": 63.0, "learning_rate": 4.6881355932203385e-07, "loss": 1.4157, "step": 16170 }, { "epoch": 0.4, "grad_norm": 67.5, "learning_rate": 4.6847457627118643e-07, "loss": 1.4218, "step": 16180 }, { "epoch": 0.4, "grad_norm": 65.5, "learning_rate": 4.6813559322033896e-07, "loss": 1.4667, "step": 16190 }, { "epoch": 0.41, "grad_norm": 66.5, "learning_rate": 4.6779661016949154e-07, "loss": 1.4454, "step": 16200 }, { "epoch": 0.41, "grad_norm": 64.5, "learning_rate": 4.6745762711864406e-07, "loss": 1.5011, "step": 16210 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.671186440677966e-07, "loss": 1.4311, "step": 16220 }, { "epoch": 0.41, "grad_norm": 65.5, "learning_rate": 4.6677966101694917e-07, "loss": 1.4825, "step": 16230 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.6644067796610165e-07, "loss": 1.4875, "step": 16240 }, { "epoch": 0.41, "grad_norm": 66.5, "learning_rate": 4.661016949152542e-07, "loss": 1.4448, "step": 16250 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.6576271186440675e-07, "loss": 1.4423, "step": 16260 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.654237288135593e-07, "loss": 1.4494, "step": 16270 }, { "epoch": 0.41, "grad_norm": 66.5, "learning_rate": 4.6508474576271186e-07, "loss": 1.4696, "step": 16280 }, { "epoch": 0.41, "grad_norm": 73.0, "learning_rate": 4.647457627118644e-07, "loss": 1.4202, "step": 16290 }, { "epoch": 0.41, "grad_norm": 65.5, "learning_rate": 4.644067796610169e-07, "loss": 1.5048, "step": 16300 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.640677966101695e-07, "loss": 1.4971, "step": 16310 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.6372881355932203e-07, "loss": 1.4851, "step": 16320 }, { "epoch": 0.41, "grad_norm": 63.75, "learning_rate": 4.6338983050847455e-07, "loss": 1.4816, "step": 16330 }, { "epoch": 0.41, "grad_norm": 68.0, "learning_rate": 4.6305084745762713e-07, "loss": 1.4676, "step": 16340 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.627118644067796e-07, "loss": 1.4612, "step": 16350 }, { "epoch": 0.41, "grad_norm": 66.5, "learning_rate": 4.623728813559322e-07, "loss": 1.4785, "step": 16360 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.620338983050847e-07, "loss": 1.4538, "step": 16370 }, { "epoch": 0.41, "grad_norm": 63.5, "learning_rate": 4.6169491525423725e-07, "loss": 1.4805, "step": 16380 }, { "epoch": 0.41, "grad_norm": 72.0, "learning_rate": 4.613559322033898e-07, "loss": 1.4757, "step": 16390 }, { "epoch": 0.41, "grad_norm": 68.0, "learning_rate": 4.6101694915254235e-07, "loss": 1.4923, "step": 16400 }, { "epoch": 0.41, "grad_norm": 66.5, "learning_rate": 4.606779661016949e-07, "loss": 1.4342, "step": 16410 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.6033898305084746e-07, "loss": 1.4597, "step": 16420 }, { "epoch": 0.41, "grad_norm": 66.0, "learning_rate": 4.6e-07, "loss": 1.4384, "step": 16430 }, { "epoch": 0.41, "grad_norm": 63.75, "learning_rate": 4.5966101694915257e-07, "loss": 1.4537, "step": 16440 }, { "epoch": 0.41, "grad_norm": 70.0, "learning_rate": 4.5932203389830505e-07, "loss": 1.465, "step": 16450 }, { "epoch": 0.41, "grad_norm": 66.0, "learning_rate": 4.5898305084745757e-07, "loss": 1.4596, "step": 16460 }, { "epoch": 0.41, "grad_norm": 63.25, "learning_rate": 4.5864406779661015e-07, "loss": 1.4752, "step": 16470 }, { "epoch": 0.41, "grad_norm": 67.5, "learning_rate": 4.583050847457627e-07, "loss": 1.5154, "step": 16480 }, { "epoch": 0.41, "grad_norm": 65.5, "learning_rate": 4.579661016949152e-07, "loss": 1.4267, "step": 16490 }, { "epoch": 0.41, "grad_norm": 65.0, "learning_rate": 4.576271186440678e-07, "loss": 1.4826, "step": 16500 }, { "epoch": 0.41, "grad_norm": 68.0, "learning_rate": 4.572881355932203e-07, "loss": 1.4987, "step": 16510 }, { "epoch": 0.41, "grad_norm": 69.5, "learning_rate": 4.569491525423729e-07, "loss": 1.5094, "step": 16520 }, { "epoch": 0.41, "grad_norm": 65.0, "learning_rate": 4.566101694915254e-07, "loss": 1.5129, "step": 16530 }, { "epoch": 0.41, "grad_norm": 68.0, "learning_rate": 4.5627118644067795e-07, "loss": 1.4766, "step": 16540 }, { "epoch": 0.41, "grad_norm": 66.5, "learning_rate": 4.559322033898305e-07, "loss": 1.469, "step": 16550 }, { "epoch": 0.41, "grad_norm": 67.5, "learning_rate": 4.55593220338983e-07, "loss": 1.4652, "step": 16560 }, { "epoch": 0.41, "grad_norm": 64.0, "learning_rate": 4.5525423728813554e-07, "loss": 1.4856, "step": 16570 }, { "epoch": 0.41, "grad_norm": 63.5, "learning_rate": 4.549152542372881e-07, "loss": 1.4837, "step": 16580 }, { "epoch": 0.41, "grad_norm": 67.0, "learning_rate": 4.5457627118644064e-07, "loss": 1.4729, "step": 16590 }, { "epoch": 0.41, "grad_norm": 65.5, "learning_rate": 4.542372881355932e-07, "loss": 1.4397, "step": 16600 }, { "epoch": 0.42, "grad_norm": 65.0, "learning_rate": 4.5389830508474575e-07, "loss": 1.457, "step": 16610 }, { "epoch": 0.42, "grad_norm": 65.0, "learning_rate": 4.535593220338983e-07, "loss": 1.4324, "step": 16620 }, { "epoch": 0.42, "grad_norm": 67.5, "learning_rate": 4.5322033898305086e-07, "loss": 1.4802, "step": 16630 }, { "epoch": 0.42, "grad_norm": 67.0, "learning_rate": 4.528813559322034e-07, "loss": 1.4368, "step": 16640 }, { "epoch": 0.42, "grad_norm": 66.0, "learning_rate": 4.525423728813559e-07, "loss": 1.4821, "step": 16650 }, { "epoch": 0.42, "grad_norm": 64.5, "learning_rate": 4.5220338983050844e-07, "loss": 1.4604, "step": 16660 }, { "epoch": 0.42, "grad_norm": 67.0, "learning_rate": 4.5186440677966097e-07, "loss": 1.4423, "step": 16670 }, { "epoch": 0.42, "grad_norm": 62.0, "learning_rate": 4.5152542372881355e-07, "loss": 1.4394, "step": 16680 }, { "epoch": 0.42, "grad_norm": 66.0, "learning_rate": 4.511864406779661e-07, "loss": 1.4973, "step": 16690 }, { "epoch": 0.42, "grad_norm": 65.0, "learning_rate": 4.508474576271186e-07, "loss": 1.4479, "step": 16700 }, { "epoch": 0.42, "grad_norm": 70.5, "learning_rate": 4.505084745762712e-07, "loss": 1.4347, "step": 16710 }, { "epoch": 0.42, "grad_norm": 71.0, "learning_rate": 4.501694915254237e-07, "loss": 1.4444, "step": 16720 }, { "epoch": 0.42, "grad_norm": 68.0, "learning_rate": 4.4983050847457624e-07, "loss": 1.4181, "step": 16730 }, { "epoch": 0.42, "grad_norm": 69.5, "learning_rate": 4.494915254237288e-07, "loss": 1.4429, "step": 16740 }, { "epoch": 0.42, "grad_norm": 70.0, "learning_rate": 4.4915254237288135e-07, "loss": 1.4523, "step": 16750 }, { "epoch": 0.42, "grad_norm": 64.5, "learning_rate": 4.4881355932203383e-07, "loss": 1.4547, "step": 16760 }, { "epoch": 0.42, "grad_norm": 64.0, "learning_rate": 4.484745762711864e-07, "loss": 1.4827, "step": 16770 }, { "epoch": 0.42, "grad_norm": 63.75, "learning_rate": 4.4813559322033893e-07, "loss": 1.4142, "step": 16780 }, { "epoch": 0.42, "grad_norm": 68.5, "learning_rate": 4.477966101694915e-07, "loss": 1.4234, "step": 16790 }, { "epoch": 0.42, "grad_norm": 66.0, "learning_rate": 4.4745762711864404e-07, "loss": 1.4646, "step": 16800 }, { "epoch": 0.42, "grad_norm": 64.5, "learning_rate": 4.4711864406779657e-07, "loss": 1.4629, "step": 16810 }, { "epoch": 0.42, "grad_norm": 66.5, "learning_rate": 4.4677966101694915e-07, "loss": 1.4742, "step": 16820 }, { "epoch": 0.42, "grad_norm": 67.5, "learning_rate": 4.464406779661017e-07, "loss": 1.4679, "step": 16830 }, { "epoch": 0.42, "grad_norm": 65.0, "learning_rate": 4.4610169491525426e-07, "loss": 1.4858, "step": 16840 }, { "epoch": 0.42, "grad_norm": 68.5, "learning_rate": 4.457627118644068e-07, "loss": 1.4727, "step": 16850 }, { "epoch": 0.42, "grad_norm": 66.5, "learning_rate": 4.454237288135593e-07, "loss": 1.4977, "step": 16860 }, { "epoch": 0.42, "grad_norm": 70.5, "learning_rate": 4.4508474576271184e-07, "loss": 1.5104, "step": 16870 }, { "epoch": 0.42, "grad_norm": 66.5, "learning_rate": 4.4474576271186437e-07, "loss": 1.4548, "step": 16880 }, { "epoch": 0.42, "grad_norm": 69.5, "learning_rate": 4.444067796610169e-07, "loss": 1.4371, "step": 16890 }, { "epoch": 0.42, "grad_norm": 65.5, "learning_rate": 4.440677966101695e-07, "loss": 1.4948, "step": 16900 }, { "epoch": 0.42, "grad_norm": 68.5, "learning_rate": 4.43728813559322e-07, "loss": 1.4385, "step": 16910 }, { "epoch": 0.42, "grad_norm": 71.0, "learning_rate": 4.433898305084746e-07, "loss": 1.4525, "step": 16920 }, { "epoch": 0.42, "grad_norm": 66.0, "learning_rate": 4.430508474576271e-07, "loss": 1.434, "step": 16930 }, { "epoch": 0.42, "grad_norm": 68.5, "learning_rate": 4.4271186440677964e-07, "loss": 1.4501, "step": 16940 }, { "epoch": 0.42, "grad_norm": 66.5, "learning_rate": 4.423728813559322e-07, "loss": 1.4758, "step": 16950 }, { "epoch": 0.42, "grad_norm": 67.0, "learning_rate": 4.4203389830508475e-07, "loss": 1.4919, "step": 16960 }, { "epoch": 0.42, "grad_norm": 67.0, "learning_rate": 4.416949152542372e-07, "loss": 1.4909, "step": 16970 }, { "epoch": 0.42, "grad_norm": 64.5, "learning_rate": 4.413559322033898e-07, "loss": 1.4687, "step": 16980 }, { "epoch": 0.42, "grad_norm": 70.0, "learning_rate": 4.4101694915254233e-07, "loss": 1.4871, "step": 16990 }, { "epoch": 0.42, "grad_norm": 68.5, "learning_rate": 4.4067796610169486e-07, "loss": 1.4497, "step": 17000 }, { "epoch": 0.43, "grad_norm": 66.5, "learning_rate": 4.4033898305084744e-07, "loss": 1.4564, "step": 17010 }, { "epoch": 0.43, "grad_norm": 67.5, "learning_rate": 4.3999999999999997e-07, "loss": 1.478, "step": 17020 }, { "epoch": 0.43, "grad_norm": 68.0, "learning_rate": 4.3966101694915255e-07, "loss": 1.4083, "step": 17030 }, { "epoch": 0.43, "grad_norm": 65.5, "learning_rate": 4.393220338983051e-07, "loss": 1.4706, "step": 17040 }, { "epoch": 0.43, "grad_norm": 69.0, "learning_rate": 4.389830508474576e-07, "loss": 1.4383, "step": 17050 }, { "epoch": 0.43, "grad_norm": 67.5, "learning_rate": 4.386440677966102e-07, "loss": 1.4298, "step": 17060 }, { "epoch": 0.43, "grad_norm": 67.0, "learning_rate": 4.3830508474576266e-07, "loss": 1.4398, "step": 17070 }, { "epoch": 0.43, "grad_norm": 67.5, "learning_rate": 4.379661016949152e-07, "loss": 1.473, "step": 17080 }, { "epoch": 0.43, "grad_norm": 67.5, "learning_rate": 4.3762711864406777e-07, "loss": 1.4727, "step": 17090 }, { "epoch": 0.43, "grad_norm": 64.5, "learning_rate": 4.372881355932203e-07, "loss": 1.4619, "step": 17100 }, { "epoch": 0.43, "grad_norm": 68.0, "learning_rate": 4.369491525423729e-07, "loss": 1.4368, "step": 17110 }, { "epoch": 0.43, "grad_norm": 66.0, "learning_rate": 4.366101694915254e-07, "loss": 1.4614, "step": 17120 }, { "epoch": 0.43, "grad_norm": 67.5, "learning_rate": 4.3627118644067793e-07, "loss": 1.5397, "step": 17130 }, { "epoch": 0.43, "grad_norm": 68.5, "learning_rate": 4.359322033898305e-07, "loss": 1.4376, "step": 17140 }, { "epoch": 0.43, "grad_norm": 66.0, "learning_rate": 4.3559322033898304e-07, "loss": 1.5009, "step": 17150 }, { "epoch": 0.43, "grad_norm": 64.5, "learning_rate": 4.3525423728813557e-07, "loss": 1.4829, "step": 17160 }, { "epoch": 0.43, "grad_norm": 68.5, "learning_rate": 4.3491525423728815e-07, "loss": 1.4822, "step": 17170 }, { "epoch": 0.43, "grad_norm": 67.5, "learning_rate": 4.345762711864406e-07, "loss": 1.4392, "step": 17180 }, { "epoch": 0.43, "grad_norm": 68.5, "learning_rate": 4.342372881355932e-07, "loss": 1.4512, "step": 17190 }, { "epoch": 0.43, "grad_norm": 65.0, "learning_rate": 4.3389830508474573e-07, "loss": 1.4694, "step": 17200 }, { "epoch": 0.43, "grad_norm": 68.0, "learning_rate": 4.3355932203389826e-07, "loss": 1.4912, "step": 17210 }, { "epoch": 0.43, "grad_norm": 65.5, "learning_rate": 4.3322033898305084e-07, "loss": 1.4765, "step": 17220 }, { "epoch": 0.43, "grad_norm": 65.0, "learning_rate": 4.3288135593220337e-07, "loss": 1.4689, "step": 17230 }, { "epoch": 0.43, "grad_norm": 64.0, "learning_rate": 4.325423728813559e-07, "loss": 1.4418, "step": 17240 }, { "epoch": 0.43, "grad_norm": 62.25, "learning_rate": 4.322033898305085e-07, "loss": 1.4501, "step": 17250 }, { "epoch": 0.43, "grad_norm": 65.5, "learning_rate": 4.31864406779661e-07, "loss": 1.4524, "step": 17260 }, { "epoch": 0.43, "grad_norm": 67.5, "learning_rate": 4.315254237288136e-07, "loss": 1.4129, "step": 17270 }, { "epoch": 0.43, "grad_norm": 65.5, "learning_rate": 4.3118644067796606e-07, "loss": 1.4296, "step": 17280 }, { "epoch": 0.43, "grad_norm": 66.5, "learning_rate": 4.308474576271186e-07, "loss": 1.5083, "step": 17290 }, { "epoch": 0.43, "grad_norm": 66.0, "learning_rate": 4.3050847457627117e-07, "loss": 1.5027, "step": 17300 }, { "epoch": 0.43, "grad_norm": 68.5, "learning_rate": 4.301694915254237e-07, "loss": 1.4698, "step": 17310 }, { "epoch": 0.43, "grad_norm": 67.5, "learning_rate": 4.298305084745762e-07, "loss": 1.4803, "step": 17320 }, { "epoch": 0.43, "grad_norm": 64.5, "learning_rate": 4.294915254237288e-07, "loss": 1.4467, "step": 17330 }, { "epoch": 0.43, "grad_norm": 67.5, "learning_rate": 4.2915254237288133e-07, "loss": 1.4902, "step": 17340 }, { "epoch": 0.43, "grad_norm": 63.75, "learning_rate": 4.288135593220339e-07, "loss": 1.4568, "step": 17350 }, { "epoch": 0.43, "grad_norm": 69.0, "learning_rate": 4.2847457627118644e-07, "loss": 1.471, "step": 17360 }, { "epoch": 0.43, "grad_norm": 68.5, "learning_rate": 4.2813559322033897e-07, "loss": 1.4891, "step": 17370 }, { "epoch": 0.43, "grad_norm": 66.5, "learning_rate": 4.277966101694915e-07, "loss": 1.4515, "step": 17380 }, { "epoch": 0.43, "grad_norm": 67.0, "learning_rate": 4.27457627118644e-07, "loss": 1.4659, "step": 17390 }, { "epoch": 0.43, "grad_norm": 67.0, "learning_rate": 4.2711864406779655e-07, "loss": 1.4687, "step": 17400 }, { "epoch": 0.44, "grad_norm": 65.5, "learning_rate": 4.2677966101694913e-07, "loss": 1.4337, "step": 17410 }, { "epoch": 0.44, "grad_norm": 66.5, "learning_rate": 4.2644067796610166e-07, "loss": 1.4572, "step": 17420 }, { "epoch": 0.44, "grad_norm": 65.5, "learning_rate": 4.2610169491525424e-07, "loss": 1.4694, "step": 17430 }, { "epoch": 0.44, "grad_norm": 66.5, "learning_rate": 4.2576271186440677e-07, "loss": 1.454, "step": 17440 }, { "epoch": 0.44, "grad_norm": 64.5, "learning_rate": 4.254237288135593e-07, "loss": 1.4098, "step": 17450 }, { "epoch": 0.44, "grad_norm": 66.0, "learning_rate": 4.250847457627119e-07, "loss": 1.3994, "step": 17460 }, { "epoch": 0.44, "grad_norm": 70.5, "learning_rate": 4.247457627118644e-07, "loss": 1.445, "step": 17470 }, { "epoch": 0.44, "grad_norm": 63.25, "learning_rate": 4.2440677966101693e-07, "loss": 1.4478, "step": 17480 }, { "epoch": 0.44, "grad_norm": 72.5, "learning_rate": 4.2406779661016946e-07, "loss": 1.454, "step": 17490 }, { "epoch": 0.44, "grad_norm": 68.5, "learning_rate": 4.23728813559322e-07, "loss": 1.4836, "step": 17500 }, { "epoch": 0.44, "grad_norm": 61.75, "learning_rate": 4.2338983050847457e-07, "loss": 1.3868, "step": 17510 }, { "epoch": 0.44, "grad_norm": 69.5, "learning_rate": 4.230508474576271e-07, "loss": 1.4972, "step": 17520 }, { "epoch": 0.44, "grad_norm": 77.5, "learning_rate": 4.227118644067796e-07, "loss": 1.4787, "step": 17530 }, { "epoch": 0.44, "grad_norm": 66.5, "learning_rate": 4.223728813559322e-07, "loss": 1.47, "step": 17540 }, { "epoch": 0.44, "grad_norm": 65.5, "learning_rate": 4.2203389830508473e-07, "loss": 1.4529, "step": 17550 }, { "epoch": 0.44, "grad_norm": 67.5, "learning_rate": 4.2169491525423726e-07, "loss": 1.4615, "step": 17560 }, { "epoch": 0.44, "grad_norm": 64.5, "learning_rate": 4.2135593220338984e-07, "loss": 1.4484, "step": 17570 }, { "epoch": 0.44, "grad_norm": 71.0, "learning_rate": 4.2101694915254237e-07, "loss": 1.4518, "step": 17580 }, { "epoch": 0.44, "grad_norm": 69.0, "learning_rate": 4.206779661016949e-07, "loss": 1.4691, "step": 17590 }, { "epoch": 0.44, "grad_norm": 64.5, "learning_rate": 4.203389830508474e-07, "loss": 1.4306, "step": 17600 }, { "epoch": 0.44, "grad_norm": 67.0, "learning_rate": 4.1999999999999995e-07, "loss": 1.4389, "step": 17610 }, { "epoch": 0.44, "grad_norm": 70.5, "learning_rate": 4.1966101694915253e-07, "loss": 1.4872, "step": 17620 }, { "epoch": 0.44, "grad_norm": 63.5, "learning_rate": 4.1932203389830506e-07, "loss": 1.4055, "step": 17630 }, { "epoch": 0.44, "grad_norm": 66.5, "learning_rate": 4.189830508474576e-07, "loss": 1.4629, "step": 17640 }, { "epoch": 0.44, "grad_norm": 66.0, "learning_rate": 4.1864406779661017e-07, "loss": 1.4998, "step": 17650 }, { "epoch": 0.44, "grad_norm": 73.0, "learning_rate": 4.183050847457627e-07, "loss": 1.4719, "step": 17660 }, { "epoch": 0.44, "grad_norm": 67.5, "learning_rate": 4.179661016949153e-07, "loss": 1.5044, "step": 17670 }, { "epoch": 0.44, "grad_norm": 67.5, "learning_rate": 4.176271186440678e-07, "loss": 1.4518, "step": 17680 }, { "epoch": 0.44, "grad_norm": 67.0, "learning_rate": 4.1728813559322033e-07, "loss": 1.453, "step": 17690 }, { "epoch": 0.44, "grad_norm": 66.0, "learning_rate": 4.1694915254237286e-07, "loss": 1.4725, "step": 17700 }, { "epoch": 0.44, "grad_norm": 63.75, "learning_rate": 4.166101694915254e-07, "loss": 1.4361, "step": 17710 }, { "epoch": 0.44, "grad_norm": 66.0, "learning_rate": 4.162711864406779e-07, "loss": 1.4407, "step": 17720 }, { "epoch": 0.44, "grad_norm": 65.5, "learning_rate": 4.159322033898305e-07, "loss": 1.4366, "step": 17730 }, { "epoch": 0.44, "grad_norm": 68.0, "learning_rate": 4.15593220338983e-07, "loss": 1.452, "step": 17740 }, { "epoch": 0.44, "grad_norm": 68.0, "learning_rate": 4.152542372881356e-07, "loss": 1.4488, "step": 17750 }, { "epoch": 0.44, "grad_norm": 63.5, "learning_rate": 4.1491525423728813e-07, "loss": 1.4167, "step": 17760 }, { "epoch": 0.44, "grad_norm": 67.5, "learning_rate": 4.1457627118644066e-07, "loss": 1.4601, "step": 17770 }, { "epoch": 0.44, "grad_norm": 68.0, "learning_rate": 4.1423728813559324e-07, "loss": 1.4814, "step": 17780 }, { "epoch": 0.44, "grad_norm": 64.0, "learning_rate": 4.1389830508474576e-07, "loss": 1.4926, "step": 17790 }, { "epoch": 0.45, "grad_norm": 65.0, "learning_rate": 4.1355932203389824e-07, "loss": 1.4815, "step": 17800 }, { "epoch": 0.45, "grad_norm": 69.0, "learning_rate": 4.132203389830508e-07, "loss": 1.5314, "step": 17810 }, { "epoch": 0.45, "grad_norm": 66.5, "learning_rate": 4.1288135593220335e-07, "loss": 1.4809, "step": 17820 }, { "epoch": 0.45, "grad_norm": 67.5, "learning_rate": 4.1254237288135593e-07, "loss": 1.4412, "step": 17830 }, { "epoch": 0.45, "grad_norm": 67.5, "learning_rate": 4.1220338983050846e-07, "loss": 1.442, "step": 17840 }, { "epoch": 0.45, "grad_norm": 66.0, "learning_rate": 4.11864406779661e-07, "loss": 1.4583, "step": 17850 }, { "epoch": 0.45, "grad_norm": 66.0, "learning_rate": 4.1152542372881356e-07, "loss": 1.4767, "step": 17860 }, { "epoch": 0.45, "grad_norm": 67.0, "learning_rate": 4.111864406779661e-07, "loss": 1.4757, "step": 17870 }, { "epoch": 0.45, "grad_norm": 66.0, "learning_rate": 4.108474576271186e-07, "loss": 1.4631, "step": 17880 }, { "epoch": 0.45, "grad_norm": 68.0, "learning_rate": 4.105084745762712e-07, "loss": 1.4761, "step": 17890 }, { "epoch": 0.45, "grad_norm": 69.0, "learning_rate": 4.101694915254237e-07, "loss": 1.4763, "step": 17900 }, { "epoch": 0.45, "grad_norm": 68.0, "learning_rate": 4.0983050847457626e-07, "loss": 1.4997, "step": 17910 }, { "epoch": 0.45, "grad_norm": 64.0, "learning_rate": 4.094915254237288e-07, "loss": 1.4672, "step": 17920 }, { "epoch": 0.45, "grad_norm": 67.5, "learning_rate": 4.091525423728813e-07, "loss": 1.4776, "step": 17930 }, { "epoch": 0.45, "grad_norm": 66.5, "learning_rate": 4.088135593220339e-07, "loss": 1.4677, "step": 17940 }, { "epoch": 0.45, "grad_norm": 67.0, "learning_rate": 4.084745762711864e-07, "loss": 1.4748, "step": 17950 }, { "epoch": 0.45, "grad_norm": 67.5, "learning_rate": 4.0813559322033895e-07, "loss": 1.4693, "step": 17960 }, { "epoch": 0.45, "grad_norm": 67.5, "learning_rate": 4.0779661016949153e-07, "loss": 1.4612, "step": 17970 }, { "epoch": 0.45, "grad_norm": 64.5, "learning_rate": 4.0745762711864406e-07, "loss": 1.4718, "step": 17980 }, { "epoch": 0.45, "grad_norm": 65.5, "learning_rate": 4.0711864406779664e-07, "loss": 1.4759, "step": 17990 }, { "epoch": 0.45, "grad_norm": 69.5, "learning_rate": 4.0677966101694916e-07, "loss": 1.4291, "step": 18000 }, { "epoch": 0.45, "grad_norm": 68.5, "learning_rate": 4.0644067796610164e-07, "loss": 1.4899, "step": 18010 }, { "epoch": 0.45, "grad_norm": 66.5, "learning_rate": 4.061016949152542e-07, "loss": 1.5047, "step": 18020 }, { "epoch": 0.45, "grad_norm": 65.5, "learning_rate": 4.0576271186440675e-07, "loss": 1.4517, "step": 18030 }, { "epoch": 0.45, "grad_norm": 65.0, "learning_rate": 4.054237288135593e-07, "loss": 1.4401, "step": 18040 }, { "epoch": 0.45, "grad_norm": 67.5, "learning_rate": 4.0508474576271186e-07, "loss": 1.4631, "step": 18050 }, { "epoch": 0.45, "grad_norm": 66.0, "learning_rate": 4.047457627118644e-07, "loss": 1.4511, "step": 18060 }, { "epoch": 0.45, "grad_norm": 66.5, "learning_rate": 4.0440677966101696e-07, "loss": 1.5248, "step": 18070 }, { "epoch": 0.45, "grad_norm": 66.5, "learning_rate": 4.040677966101695e-07, "loss": 1.4094, "step": 18080 }, { "epoch": 0.45, "grad_norm": 64.0, "learning_rate": 4.03728813559322e-07, "loss": 1.5003, "step": 18090 }, { "epoch": 0.45, "grad_norm": 66.0, "learning_rate": 4.033898305084746e-07, "loss": 1.4907, "step": 18100 }, { "epoch": 0.45, "grad_norm": 68.0, "learning_rate": 4.030508474576271e-07, "loss": 1.4569, "step": 18110 }, { "epoch": 0.45, "grad_norm": 68.0, "learning_rate": 4.027118644067796e-07, "loss": 1.4431, "step": 18120 }, { "epoch": 0.45, "grad_norm": 66.0, "learning_rate": 4.023728813559322e-07, "loss": 1.4821, "step": 18130 }, { "epoch": 0.45, "grad_norm": 64.5, "learning_rate": 4.020338983050847e-07, "loss": 1.4891, "step": 18140 }, { "epoch": 0.45, "grad_norm": 65.0, "learning_rate": 4.016949152542373e-07, "loss": 1.4185, "step": 18150 }, { "epoch": 0.45, "grad_norm": 70.0, "learning_rate": 4.013559322033898e-07, "loss": 1.4669, "step": 18160 }, { "epoch": 0.45, "grad_norm": 68.0, "learning_rate": 4.0101694915254235e-07, "loss": 1.4913, "step": 18170 }, { "epoch": 0.45, "grad_norm": 66.5, "learning_rate": 4.0067796610169493e-07, "loss": 1.5018, "step": 18180 }, { "epoch": 0.45, "grad_norm": 69.0, "learning_rate": 4.0033898305084745e-07, "loss": 1.4631, "step": 18190 }, { "epoch": 0.46, "grad_norm": 65.5, "learning_rate": 4e-07, "loss": 1.4598, "step": 18200 }, { "epoch": 0.46, "grad_norm": 63.75, "learning_rate": 3.9966101694915256e-07, "loss": 1.4678, "step": 18210 }, { "epoch": 0.46, "grad_norm": 69.5, "learning_rate": 3.9932203389830504e-07, "loss": 1.4552, "step": 18220 }, { "epoch": 0.46, "grad_norm": 68.0, "learning_rate": 3.989830508474576e-07, "loss": 1.4747, "step": 18230 }, { "epoch": 0.46, "grad_norm": 66.5, "learning_rate": 3.9864406779661015e-07, "loss": 1.4982, "step": 18240 }, { "epoch": 0.46, "grad_norm": 72.0, "learning_rate": 3.9830508474576267e-07, "loss": 1.4744, "step": 18250 }, { "epoch": 0.46, "grad_norm": 69.0, "learning_rate": 3.9796610169491525e-07, "loss": 1.5054, "step": 18260 }, { "epoch": 0.46, "grad_norm": 62.75, "learning_rate": 3.976271186440678e-07, "loss": 1.4516, "step": 18270 }, { "epoch": 0.46, "grad_norm": 69.0, "learning_rate": 3.972881355932203e-07, "loss": 1.4837, "step": 18280 }, { "epoch": 0.46, "grad_norm": 66.0, "learning_rate": 3.969491525423729e-07, "loss": 1.4304, "step": 18290 }, { "epoch": 0.46, "grad_norm": 69.0, "learning_rate": 3.966101694915254e-07, "loss": 1.4789, "step": 18300 }, { "epoch": 0.46, "grad_norm": 65.5, "learning_rate": 3.96271186440678e-07, "loss": 1.5042, "step": 18310 }, { "epoch": 0.46, "grad_norm": 65.0, "learning_rate": 3.9593220338983047e-07, "loss": 1.4196, "step": 18320 }, { "epoch": 0.46, "grad_norm": 63.75, "learning_rate": 3.95593220338983e-07, "loss": 1.4104, "step": 18330 }, { "epoch": 0.46, "grad_norm": 66.0, "learning_rate": 3.952542372881356e-07, "loss": 1.5165, "step": 18340 }, { "epoch": 0.46, "grad_norm": 65.5, "learning_rate": 3.949152542372881e-07, "loss": 1.487, "step": 18350 }, { "epoch": 0.46, "grad_norm": 67.0, "learning_rate": 3.9457627118644064e-07, "loss": 1.4779, "step": 18360 }, { "epoch": 0.46, "grad_norm": 66.0, "learning_rate": 3.942372881355932e-07, "loss": 1.4267, "step": 18370 }, { "epoch": 0.46, "grad_norm": 66.5, "learning_rate": 3.9389830508474574e-07, "loss": 1.457, "step": 18380 }, { "epoch": 0.46, "grad_norm": 67.0, "learning_rate": 3.935593220338983e-07, "loss": 1.4677, "step": 18390 }, { "epoch": 0.46, "grad_norm": 65.5, "learning_rate": 3.9322033898305085e-07, "loss": 1.4515, "step": 18400 }, { "epoch": 0.46, "grad_norm": 66.5, "learning_rate": 3.928813559322034e-07, "loss": 1.4915, "step": 18410 }, { "epoch": 0.46, "grad_norm": 69.5, "learning_rate": 3.925423728813559e-07, "loss": 1.4567, "step": 18420 }, { "epoch": 0.46, "grad_norm": 67.5, "learning_rate": 3.9220338983050844e-07, "loss": 1.4941, "step": 18430 }, { "epoch": 0.46, "grad_norm": 69.0, "learning_rate": 3.9186440677966096e-07, "loss": 1.4898, "step": 18440 }, { "epoch": 0.46, "grad_norm": 68.0, "learning_rate": 3.9152542372881354e-07, "loss": 1.4565, "step": 18450 }, { "epoch": 0.46, "grad_norm": 66.5, "learning_rate": 3.9118644067796607e-07, "loss": 1.4362, "step": 18460 }, { "epoch": 0.46, "grad_norm": 66.0, "learning_rate": 3.9084745762711865e-07, "loss": 1.4559, "step": 18470 }, { "epoch": 0.46, "grad_norm": 71.0, "learning_rate": 3.905084745762712e-07, "loss": 1.5043, "step": 18480 }, { "epoch": 0.46, "grad_norm": 69.5, "learning_rate": 3.901694915254237e-07, "loss": 1.4583, "step": 18490 }, { "epoch": 0.46, "grad_norm": 70.5, "learning_rate": 3.898305084745763e-07, "loss": 1.4676, "step": 18500 }, { "epoch": 0.46, "grad_norm": 68.0, "learning_rate": 3.894915254237288e-07, "loss": 1.4402, "step": 18510 }, { "epoch": 0.46, "grad_norm": 64.0, "learning_rate": 3.8915254237288134e-07, "loss": 1.4872, "step": 18520 }, { "epoch": 0.46, "grad_norm": 65.5, "learning_rate": 3.8881355932203387e-07, "loss": 1.518, "step": 18530 }, { "epoch": 0.46, "grad_norm": 68.0, "learning_rate": 3.884745762711864e-07, "loss": 1.5365, "step": 18540 }, { "epoch": 0.46, "grad_norm": 67.0, "learning_rate": 3.8813559322033893e-07, "loss": 1.4352, "step": 18550 }, { "epoch": 0.46, "grad_norm": 70.0, "learning_rate": 3.877966101694915e-07, "loss": 1.4745, "step": 18560 }, { "epoch": 0.46, "grad_norm": 67.0, "learning_rate": 3.8745762711864404e-07, "loss": 1.4814, "step": 18570 }, { "epoch": 0.46, "grad_norm": 68.5, "learning_rate": 3.871186440677966e-07, "loss": 1.472, "step": 18580 }, { "epoch": 0.46, "grad_norm": 70.5, "learning_rate": 3.8677966101694914e-07, "loss": 1.4585, "step": 18590 }, { "epoch": 0.47, "grad_norm": 65.5, "learning_rate": 3.8644067796610167e-07, "loss": 1.442, "step": 18600 }, { "epoch": 0.47, "grad_norm": 69.0, "learning_rate": 3.8610169491525425e-07, "loss": 1.4981, "step": 18610 }, { "epoch": 0.47, "grad_norm": 67.0, "learning_rate": 3.857627118644068e-07, "loss": 1.4924, "step": 18620 }, { "epoch": 0.47, "grad_norm": 67.0, "learning_rate": 3.8542372881355925e-07, "loss": 1.4971, "step": 18630 }, { "epoch": 0.47, "grad_norm": 66.5, "learning_rate": 3.8508474576271183e-07, "loss": 1.4411, "step": 18640 }, { "epoch": 0.47, "grad_norm": 67.0, "learning_rate": 3.8474576271186436e-07, "loss": 1.4552, "step": 18650 }, { "epoch": 0.47, "grad_norm": 66.0, "learning_rate": 3.8440677966101694e-07, "loss": 1.5125, "step": 18660 }, { "epoch": 0.47, "grad_norm": 69.0, "learning_rate": 3.8406779661016947e-07, "loss": 1.4749, "step": 18670 }, { "epoch": 0.47, "grad_norm": 68.0, "learning_rate": 3.83728813559322e-07, "loss": 1.4489, "step": 18680 }, { "epoch": 0.47, "grad_norm": 67.0, "learning_rate": 3.833898305084746e-07, "loss": 1.4281, "step": 18690 }, { "epoch": 0.47, "grad_norm": 66.0, "learning_rate": 3.830508474576271e-07, "loss": 1.4845, "step": 18700 }, { "epoch": 0.47, "grad_norm": 68.5, "learning_rate": 3.827118644067797e-07, "loss": 1.4545, "step": 18710 }, { "epoch": 0.47, "grad_norm": 64.5, "learning_rate": 3.823728813559322e-07, "loss": 1.4862, "step": 18720 }, { "epoch": 0.47, "grad_norm": 65.5, "learning_rate": 3.820338983050847e-07, "loss": 1.4577, "step": 18730 }, { "epoch": 0.47, "grad_norm": 66.5, "learning_rate": 3.8169491525423727e-07, "loss": 1.4923, "step": 18740 }, { "epoch": 0.47, "grad_norm": 65.0, "learning_rate": 3.813559322033898e-07, "loss": 1.4489, "step": 18750 }, { "epoch": 0.47, "grad_norm": 68.0, "learning_rate": 3.810169491525423e-07, "loss": 1.4753, "step": 18760 }, { "epoch": 0.47, "grad_norm": 68.5, "learning_rate": 3.806779661016949e-07, "loss": 1.4583, "step": 18770 }, { "epoch": 0.47, "grad_norm": 64.5, "learning_rate": 3.8033898305084743e-07, "loss": 1.468, "step": 18780 }, { "epoch": 0.47, "grad_norm": 64.5, "learning_rate": 3.7999999999999996e-07, "loss": 1.4527, "step": 18790 }, { "epoch": 0.47, "grad_norm": 64.5, "learning_rate": 3.7966101694915254e-07, "loss": 1.4091, "step": 18800 }, { "epoch": 0.47, "grad_norm": 68.5, "learning_rate": 3.7932203389830507e-07, "loss": 1.4212, "step": 18810 }, { "epoch": 0.47, "grad_norm": 66.5, "learning_rate": 3.7898305084745765e-07, "loss": 1.4513, "step": 18820 }, { "epoch": 0.47, "grad_norm": 67.0, "learning_rate": 3.786440677966102e-07, "loss": 1.5093, "step": 18830 }, { "epoch": 0.47, "grad_norm": 64.5, "learning_rate": 3.7830508474576265e-07, "loss": 1.4576, "step": 18840 }, { "epoch": 0.47, "grad_norm": 69.0, "learning_rate": 3.7796610169491523e-07, "loss": 1.4183, "step": 18850 }, { "epoch": 0.47, "grad_norm": 66.0, "learning_rate": 3.7762711864406776e-07, "loss": 1.4477, "step": 18860 }, { "epoch": 0.47, "grad_norm": 66.0, "learning_rate": 3.772881355932203e-07, "loss": 1.46, "step": 18870 }, { "epoch": 0.47, "grad_norm": 67.0, "learning_rate": 3.7694915254237287e-07, "loss": 1.4601, "step": 18880 }, { "epoch": 0.47, "grad_norm": 69.0, "learning_rate": 3.766101694915254e-07, "loss": 1.499, "step": 18890 }, { "epoch": 0.47, "grad_norm": 66.5, "learning_rate": 3.76271186440678e-07, "loss": 1.4569, "step": 18900 }, { "epoch": 0.47, "grad_norm": 72.0, "learning_rate": 3.759322033898305e-07, "loss": 1.475, "step": 18910 }, { "epoch": 0.47, "grad_norm": 65.0, "learning_rate": 3.7559322033898303e-07, "loss": 1.4602, "step": 18920 }, { "epoch": 0.47, "grad_norm": 68.0, "learning_rate": 3.752542372881356e-07, "loss": 1.4378, "step": 18930 }, { "epoch": 0.47, "grad_norm": 66.0, "learning_rate": 3.749152542372881e-07, "loss": 1.5027, "step": 18940 }, { "epoch": 0.47, "grad_norm": 63.75, "learning_rate": 3.745762711864406e-07, "loss": 1.4356, "step": 18950 }, { "epoch": 0.47, "grad_norm": 67.5, "learning_rate": 3.742372881355932e-07, "loss": 1.475, "step": 18960 }, { "epoch": 0.47, "grad_norm": 66.5, "learning_rate": 3.738983050847457e-07, "loss": 1.4588, "step": 18970 }, { "epoch": 0.47, "grad_norm": 65.0, "learning_rate": 3.735593220338983e-07, "loss": 1.4918, "step": 18980 }, { "epoch": 0.47, "grad_norm": 65.0, "learning_rate": 3.7322033898305083e-07, "loss": 1.5011, "step": 18990 }, { "epoch": 0.47, "grad_norm": 67.0, "learning_rate": 3.7288135593220336e-07, "loss": 1.4228, "step": 19000 }, { "epoch": 0.48, "grad_norm": 69.5, "learning_rate": 3.7254237288135594e-07, "loss": 1.4383, "step": 19010 }, { "epoch": 0.48, "grad_norm": 68.0, "learning_rate": 3.7220338983050847e-07, "loss": 1.4581, "step": 19020 }, { "epoch": 0.48, "grad_norm": 66.0, "learning_rate": 3.71864406779661e-07, "loss": 1.5045, "step": 19030 }, { "epoch": 0.48, "grad_norm": 67.0, "learning_rate": 3.715254237288136e-07, "loss": 1.4666, "step": 19040 }, { "epoch": 0.48, "grad_norm": 68.5, "learning_rate": 3.7118644067796605e-07, "loss": 1.4369, "step": 19050 }, { "epoch": 0.48, "grad_norm": 67.5, "learning_rate": 3.7084745762711863e-07, "loss": 1.4567, "step": 19060 }, { "epoch": 0.48, "grad_norm": 64.0, "learning_rate": 3.7050847457627116e-07, "loss": 1.4246, "step": 19070 }, { "epoch": 0.48, "grad_norm": 65.5, "learning_rate": 3.701694915254237e-07, "loss": 1.4939, "step": 19080 }, { "epoch": 0.48, "grad_norm": 66.5, "learning_rate": 3.6983050847457627e-07, "loss": 1.4255, "step": 19090 }, { "epoch": 0.48, "grad_norm": 68.5, "learning_rate": 3.694915254237288e-07, "loss": 1.4471, "step": 19100 }, { "epoch": 0.48, "grad_norm": 67.5, "learning_rate": 3.691525423728813e-07, "loss": 1.4525, "step": 19110 }, { "epoch": 0.48, "grad_norm": 67.5, "learning_rate": 3.688135593220339e-07, "loss": 1.4645, "step": 19120 }, { "epoch": 0.48, "grad_norm": 63.0, "learning_rate": 3.6847457627118643e-07, "loss": 1.396, "step": 19130 }, { "epoch": 0.48, "grad_norm": 67.0, "learning_rate": 3.68135593220339e-07, "loss": 1.4515, "step": 19140 }, { "epoch": 0.48, "grad_norm": 68.0, "learning_rate": 3.677966101694915e-07, "loss": 1.4262, "step": 19150 }, { "epoch": 0.48, "grad_norm": 68.5, "learning_rate": 3.67457627118644e-07, "loss": 1.4628, "step": 19160 }, { "epoch": 0.48, "grad_norm": 64.5, "learning_rate": 3.671186440677966e-07, "loss": 1.4454, "step": 19170 }, { "epoch": 0.48, "grad_norm": 65.5, "learning_rate": 3.667796610169491e-07, "loss": 1.4779, "step": 19180 }, { "epoch": 0.48, "grad_norm": 68.0, "learning_rate": 3.6644067796610165e-07, "loss": 1.4684, "step": 19190 }, { "epoch": 0.48, "grad_norm": 61.25, "learning_rate": 3.6610169491525423e-07, "loss": 1.4026, "step": 19200 }, { "epoch": 0.48, "grad_norm": 71.5, "learning_rate": 3.6576271186440676e-07, "loss": 1.4901, "step": 19210 }, { "epoch": 0.48, "grad_norm": 66.0, "learning_rate": 3.6542372881355934e-07, "loss": 1.4411, "step": 19220 }, { "epoch": 0.48, "grad_norm": 68.0, "learning_rate": 3.6508474576271187e-07, "loss": 1.4245, "step": 19230 }, { "epoch": 0.48, "grad_norm": 68.0, "learning_rate": 3.647457627118644e-07, "loss": 1.4781, "step": 19240 }, { "epoch": 0.48, "grad_norm": 68.0, "learning_rate": 3.644067796610169e-07, "loss": 1.4893, "step": 19250 }, { "epoch": 0.48, "grad_norm": 66.5, "learning_rate": 3.6406779661016945e-07, "loss": 1.4981, "step": 19260 }, { "epoch": 0.48, "grad_norm": 67.0, "learning_rate": 3.63728813559322e-07, "loss": 1.452, "step": 19270 }, { "epoch": 0.48, "grad_norm": 66.5, "learning_rate": 3.6338983050847456e-07, "loss": 1.4425, "step": 19280 }, { "epoch": 0.48, "grad_norm": 64.5, "learning_rate": 3.630508474576271e-07, "loss": 1.455, "step": 19290 }, { "epoch": 0.48, "grad_norm": 68.0, "learning_rate": 3.6271186440677967e-07, "loss": 1.4815, "step": 19300 }, { "epoch": 0.48, "grad_norm": 65.0, "learning_rate": 3.623728813559322e-07, "loss": 1.4801, "step": 19310 }, { "epoch": 0.48, "grad_norm": 66.0, "learning_rate": 3.620338983050847e-07, "loss": 1.4514, "step": 19320 }, { "epoch": 0.48, "grad_norm": 65.5, "learning_rate": 3.616949152542373e-07, "loss": 1.4262, "step": 19330 }, { "epoch": 0.48, "grad_norm": 67.5, "learning_rate": 3.6135593220338983e-07, "loss": 1.4865, "step": 19340 }, { "epoch": 0.48, "grad_norm": 65.5, "learning_rate": 3.6101694915254236e-07, "loss": 1.4128, "step": 19350 }, { "epoch": 0.48, "grad_norm": 64.5, "learning_rate": 3.606779661016949e-07, "loss": 1.4294, "step": 19360 }, { "epoch": 0.48, "grad_norm": 67.0, "learning_rate": 3.603389830508474e-07, "loss": 1.4874, "step": 19370 }, { "epoch": 0.48, "grad_norm": 70.5, "learning_rate": 3.6e-07, "loss": 1.4748, "step": 19380 }, { "epoch": 0.48, "grad_norm": 66.0, "learning_rate": 3.596610169491525e-07, "loss": 1.4492, "step": 19390 }, { "epoch": 0.48, "grad_norm": 67.5, "learning_rate": 3.5932203389830505e-07, "loss": 1.4416, "step": 19400 }, { "epoch": 0.49, "grad_norm": 62.5, "learning_rate": 3.5898305084745763e-07, "loss": 1.4267, "step": 19410 }, { "epoch": 0.49, "grad_norm": 65.5, "learning_rate": 3.5864406779661016e-07, "loss": 1.4393, "step": 19420 }, { "epoch": 0.49, "grad_norm": 64.0, "learning_rate": 3.583050847457627e-07, "loss": 1.4221, "step": 19430 }, { "epoch": 0.49, "grad_norm": 69.5, "learning_rate": 3.5796610169491527e-07, "loss": 1.4676, "step": 19440 }, { "epoch": 0.49, "grad_norm": 66.5, "learning_rate": 3.576271186440678e-07, "loss": 1.4178, "step": 19450 }, { "epoch": 0.49, "grad_norm": 71.0, "learning_rate": 3.572881355932203e-07, "loss": 1.4566, "step": 19460 }, { "epoch": 0.49, "grad_norm": 68.5, "learning_rate": 3.5694915254237285e-07, "loss": 1.4463, "step": 19470 }, { "epoch": 0.49, "grad_norm": 64.5, "learning_rate": 3.566101694915254e-07, "loss": 1.4502, "step": 19480 }, { "epoch": 0.49, "grad_norm": 68.0, "learning_rate": 3.5627118644067796e-07, "loss": 1.4539, "step": 19490 }, { "epoch": 0.49, "grad_norm": 66.5, "learning_rate": 3.559322033898305e-07, "loss": 1.4279, "step": 19500 }, { "epoch": 0.49, "grad_norm": 67.5, "learning_rate": 3.55593220338983e-07, "loss": 1.4777, "step": 19510 }, { "epoch": 0.49, "grad_norm": 62.5, "learning_rate": 3.552542372881356e-07, "loss": 1.4702, "step": 19520 }, { "epoch": 0.49, "grad_norm": 64.5, "learning_rate": 3.549152542372881e-07, "loss": 1.4739, "step": 19530 }, { "epoch": 0.49, "grad_norm": 69.0, "learning_rate": 3.545762711864407e-07, "loss": 1.4685, "step": 19540 }, { "epoch": 0.49, "grad_norm": 67.0, "learning_rate": 3.5423728813559323e-07, "loss": 1.4407, "step": 19550 }, { "epoch": 0.49, "grad_norm": 66.5, "learning_rate": 3.5389830508474576e-07, "loss": 1.4624, "step": 19560 }, { "epoch": 0.49, "grad_norm": 66.5, "learning_rate": 3.535593220338983e-07, "loss": 1.4639, "step": 19570 }, { "epoch": 0.49, "grad_norm": 67.0, "learning_rate": 3.532203389830508e-07, "loss": 1.4081, "step": 19580 }, { "epoch": 0.49, "grad_norm": 66.0, "learning_rate": 3.5288135593220334e-07, "loss": 1.4779, "step": 19590 }, { "epoch": 0.49, "grad_norm": 64.5, "learning_rate": 3.525423728813559e-07, "loss": 1.3883, "step": 19600 }, { "epoch": 0.49, "grad_norm": 65.0, "learning_rate": 3.5220338983050845e-07, "loss": 1.469, "step": 19610 }, { "epoch": 0.49, "grad_norm": 71.0, "learning_rate": 3.5186440677966103e-07, "loss": 1.4359, "step": 19620 }, { "epoch": 0.49, "grad_norm": 66.0, "learning_rate": 3.5152542372881356e-07, "loss": 1.449, "step": 19630 }, { "epoch": 0.49, "grad_norm": 68.0, "learning_rate": 3.511864406779661e-07, "loss": 1.5043, "step": 19640 }, { "epoch": 0.49, "grad_norm": 64.5, "learning_rate": 3.5084745762711866e-07, "loss": 1.4643, "step": 19650 }, { "epoch": 0.49, "grad_norm": 66.5, "learning_rate": 3.505084745762712e-07, "loss": 1.4514, "step": 19660 }, { "epoch": 0.49, "grad_norm": 71.0, "learning_rate": 3.5016949152542367e-07, "loss": 1.4132, "step": 19670 }, { "epoch": 0.49, "grad_norm": 64.5, "learning_rate": 3.4983050847457625e-07, "loss": 1.4481, "step": 19680 }, { "epoch": 0.49, "grad_norm": 67.0, "learning_rate": 3.494915254237288e-07, "loss": 1.4146, "step": 19690 }, { "epoch": 0.49, "grad_norm": 67.0, "learning_rate": 3.4915254237288136e-07, "loss": 1.4615, "step": 19700 }, { "epoch": 0.49, "grad_norm": 69.0, "learning_rate": 3.488135593220339e-07, "loss": 1.4834, "step": 19710 }, { "epoch": 0.49, "grad_norm": 67.5, "learning_rate": 3.484745762711864e-07, "loss": 1.4662, "step": 19720 }, { "epoch": 0.49, "grad_norm": 65.5, "learning_rate": 3.48135593220339e-07, "loss": 1.4567, "step": 19730 }, { "epoch": 0.49, "grad_norm": 65.5, "learning_rate": 3.477966101694915e-07, "loss": 1.4811, "step": 19740 }, { "epoch": 0.49, "grad_norm": 65.0, "learning_rate": 3.4745762711864405e-07, "loss": 1.462, "step": 19750 }, { "epoch": 0.49, "grad_norm": 67.5, "learning_rate": 3.4711864406779663e-07, "loss": 1.5057, "step": 19760 }, { "epoch": 0.49, "grad_norm": 66.0, "learning_rate": 3.467796610169491e-07, "loss": 1.4534, "step": 19770 }, { "epoch": 0.49, "grad_norm": 67.5, "learning_rate": 3.464406779661017e-07, "loss": 1.4229, "step": 19780 }, { "epoch": 0.49, "grad_norm": 68.0, "learning_rate": 3.461016949152542e-07, "loss": 1.4725, "step": 19790 }, { "epoch": 0.49, "grad_norm": 66.5, "learning_rate": 3.4576271186440674e-07, "loss": 1.4779, "step": 19800 }, { "epoch": 0.5, "grad_norm": 67.5, "learning_rate": 3.454237288135593e-07, "loss": 1.4545, "step": 19810 }, { "epoch": 0.5, "grad_norm": 69.5, "learning_rate": 3.4508474576271185e-07, "loss": 1.4344, "step": 19820 }, { "epoch": 0.5, "grad_norm": 65.0, "learning_rate": 3.447457627118644e-07, "loss": 1.4603, "step": 19830 }, { "epoch": 0.5, "grad_norm": 68.5, "learning_rate": 3.4440677966101696e-07, "loss": 1.437, "step": 19840 }, { "epoch": 0.5, "grad_norm": 66.5, "learning_rate": 3.440677966101695e-07, "loss": 1.4262, "step": 19850 }, { "epoch": 0.5, "grad_norm": 65.5, "learning_rate": 3.4372881355932206e-07, "loss": 1.449, "step": 19860 }, { "epoch": 0.5, "grad_norm": 70.5, "learning_rate": 3.433898305084746e-07, "loss": 1.45, "step": 19870 }, { "epoch": 0.5, "grad_norm": 65.5, "learning_rate": 3.4305084745762707e-07, "loss": 1.5075, "step": 19880 }, { "epoch": 0.5, "grad_norm": 67.5, "learning_rate": 3.4271186440677965e-07, "loss": 1.4895, "step": 19890 }, { "epoch": 0.5, "grad_norm": 66.0, "learning_rate": 3.423728813559322e-07, "loss": 1.509, "step": 19900 }, { "epoch": 0.5, "grad_norm": 68.5, "learning_rate": 3.420338983050847e-07, "loss": 1.4437, "step": 19910 }, { "epoch": 0.5, "grad_norm": 65.0, "learning_rate": 3.416949152542373e-07, "loss": 1.4244, "step": 19920 }, { "epoch": 0.5, "grad_norm": 67.5, "learning_rate": 3.413559322033898e-07, "loss": 1.4851, "step": 19930 }, { "epoch": 0.5, "grad_norm": 69.0, "learning_rate": 3.410169491525424e-07, "loss": 1.4419, "step": 19940 }, { "epoch": 0.5, "grad_norm": 64.5, "learning_rate": 3.406779661016949e-07, "loss": 1.4804, "step": 19950 }, { "epoch": 0.5, "grad_norm": 66.0, "learning_rate": 3.4033898305084745e-07, "loss": 1.4518, "step": 19960 }, { "epoch": 0.5, "grad_norm": 64.5, "learning_rate": 3.4000000000000003e-07, "loss": 1.473, "step": 19970 }, { "epoch": 0.5, "grad_norm": 64.5, "learning_rate": 3.396610169491525e-07, "loss": 1.4077, "step": 19980 }, { "epoch": 0.5, "grad_norm": 65.5, "learning_rate": 3.3932203389830503e-07, "loss": 1.4568, "step": 19990 }, { "epoch": 0.5, "grad_norm": 62.75, "learning_rate": 3.389830508474576e-07, "loss": 1.4683, "step": 20000 }, { "epoch": 0.5, "eval_loss": 1.4545115232467651, "eval_runtime": 68.3983, "eval_samples_per_second": 14.62, "eval_steps_per_second": 14.62, "step": 20000 }, { "epoch": 0.5, "grad_norm": 69.5, "learning_rate": 3.3864406779661014e-07, "loss": 1.5056, "step": 20010 }, { "epoch": 0.5, "grad_norm": 66.5, "learning_rate": 3.383050847457627e-07, "loss": 1.5009, "step": 20020 }, { "epoch": 0.5, "grad_norm": 69.5, "learning_rate": 3.3796610169491525e-07, "loss": 1.4073, "step": 20030 }, { "epoch": 0.5, "grad_norm": 63.25, "learning_rate": 3.376271186440678e-07, "loss": 1.4391, "step": 20040 }, { "epoch": 0.5, "grad_norm": 68.0, "learning_rate": 3.3728813559322035e-07, "loss": 1.4502, "step": 20050 }, { "epoch": 0.5, "grad_norm": 64.5, "learning_rate": 3.369491525423729e-07, "loss": 1.4458, "step": 20060 }, { "epoch": 0.5, "grad_norm": 68.0, "learning_rate": 3.366101694915254e-07, "loss": 1.4603, "step": 20070 }, { "epoch": 0.5, "grad_norm": 66.0, "learning_rate": 3.36271186440678e-07, "loss": 1.4461, "step": 20080 }, { "epoch": 0.5, "grad_norm": 67.0, "learning_rate": 3.3593220338983046e-07, "loss": 1.4921, "step": 20090 }, { "epoch": 0.5, "grad_norm": 64.0, "learning_rate": 3.35593220338983e-07, "loss": 1.4483, "step": 20100 }, { "epoch": 0.5, "grad_norm": 69.0, "learning_rate": 3.3525423728813557e-07, "loss": 1.4376, "step": 20110 }, { "epoch": 0.5, "grad_norm": 66.0, "learning_rate": 3.349152542372881e-07, "loss": 1.5029, "step": 20120 }, { "epoch": 0.5, "grad_norm": 67.5, "learning_rate": 3.345762711864407e-07, "loss": 1.4496, "step": 20130 }, { "epoch": 0.5, "grad_norm": 66.0, "learning_rate": 3.342372881355932e-07, "loss": 1.4555, "step": 20140 }, { "epoch": 0.5, "grad_norm": 65.0, "learning_rate": 3.3389830508474574e-07, "loss": 1.4612, "step": 20150 }, { "epoch": 0.5, "grad_norm": 64.5, "learning_rate": 3.335593220338983e-07, "loss": 1.4043, "step": 20160 }, { "epoch": 0.5, "grad_norm": 66.0, "learning_rate": 3.3322033898305085e-07, "loss": 1.4803, "step": 20170 }, { "epoch": 0.5, "grad_norm": 64.0, "learning_rate": 3.328813559322034e-07, "loss": 1.4877, "step": 20180 }, { "epoch": 0.5, "grad_norm": 63.5, "learning_rate": 3.325423728813559e-07, "loss": 1.4582, "step": 20190 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 3.3220338983050843e-07, "loss": 1.4408, "step": 20200 }, { "epoch": 0.51, "grad_norm": 65.0, "learning_rate": 3.31864406779661e-07, "loss": 1.4368, "step": 20210 }, { "epoch": 0.51, "grad_norm": 66.5, "learning_rate": 3.3152542372881354e-07, "loss": 1.4513, "step": 20220 }, { "epoch": 0.51, "grad_norm": 65.0, "learning_rate": 3.3118644067796606e-07, "loss": 1.4667, "step": 20230 }, { "epoch": 0.51, "grad_norm": 70.5, "learning_rate": 3.3084745762711864e-07, "loss": 1.4526, "step": 20240 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 3.3050847457627117e-07, "loss": 1.4661, "step": 20250 }, { "epoch": 0.51, "grad_norm": 67.0, "learning_rate": 3.3016949152542375e-07, "loss": 1.4909, "step": 20260 }, { "epoch": 0.51, "grad_norm": 65.5, "learning_rate": 3.298305084745763e-07, "loss": 1.4234, "step": 20270 }, { "epoch": 0.51, "grad_norm": 69.0, "learning_rate": 3.294915254237288e-07, "loss": 1.482, "step": 20280 }, { "epoch": 0.51, "grad_norm": 71.0, "learning_rate": 3.2915254237288134e-07, "loss": 1.4114, "step": 20290 }, { "epoch": 0.51, "grad_norm": 70.5, "learning_rate": 3.2881355932203386e-07, "loss": 1.4808, "step": 20300 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 3.284745762711864e-07, "loss": 1.4945, "step": 20310 }, { "epoch": 0.51, "grad_norm": 66.5, "learning_rate": 3.2813559322033897e-07, "loss": 1.4487, "step": 20320 }, { "epoch": 0.51, "grad_norm": 65.0, "learning_rate": 3.277966101694915e-07, "loss": 1.4687, "step": 20330 }, { "epoch": 0.51, "grad_norm": 61.0, "learning_rate": 3.2745762711864403e-07, "loss": 1.4407, "step": 20340 }, { "epoch": 0.51, "grad_norm": 65.0, "learning_rate": 3.271186440677966e-07, "loss": 1.4604, "step": 20350 }, { "epoch": 0.51, "grad_norm": 65.0, "learning_rate": 3.2677966101694914e-07, "loss": 1.4424, "step": 20360 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 3.264406779661017e-07, "loss": 1.4465, "step": 20370 }, { "epoch": 0.51, "grad_norm": 66.0, "learning_rate": 3.2610169491525424e-07, "loss": 1.4652, "step": 20380 }, { "epoch": 0.51, "grad_norm": 65.5, "learning_rate": 3.2576271186440677e-07, "loss": 1.436, "step": 20390 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 3.254237288135593e-07, "loss": 1.4117, "step": 20400 }, { "epoch": 0.51, "grad_norm": 65.5, "learning_rate": 3.2508474576271183e-07, "loss": 1.4606, "step": 20410 }, { "epoch": 0.51, "grad_norm": 63.75, "learning_rate": 3.2474576271186435e-07, "loss": 1.4409, "step": 20420 }, { "epoch": 0.51, "grad_norm": 64.5, "learning_rate": 3.2440677966101694e-07, "loss": 1.4309, "step": 20430 }, { "epoch": 0.51, "grad_norm": 66.0, "learning_rate": 3.2406779661016946e-07, "loss": 1.4505, "step": 20440 }, { "epoch": 0.51, "grad_norm": 67.0, "learning_rate": 3.2372881355932204e-07, "loss": 1.4847, "step": 20450 }, { "epoch": 0.51, "grad_norm": 62.5, "learning_rate": 3.2338983050847457e-07, "loss": 1.4916, "step": 20460 }, { "epoch": 0.51, "grad_norm": 66.0, "learning_rate": 3.230508474576271e-07, "loss": 1.4501, "step": 20470 }, { "epoch": 0.51, "grad_norm": 66.5, "learning_rate": 3.227118644067797e-07, "loss": 1.4273, "step": 20480 }, { "epoch": 0.51, "grad_norm": 69.0, "learning_rate": 3.223728813559322e-07, "loss": 1.4616, "step": 20490 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 3.220338983050847e-07, "loss": 1.4818, "step": 20500 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 3.2169491525423726e-07, "loss": 1.468, "step": 20510 }, { "epoch": 0.51, "grad_norm": 66.0, "learning_rate": 3.213559322033898e-07, "loss": 1.5071, "step": 20520 }, { "epoch": 0.51, "grad_norm": 66.5, "learning_rate": 3.2101694915254237e-07, "loss": 1.4516, "step": 20530 }, { "epoch": 0.51, "grad_norm": 67.0, "learning_rate": 3.206779661016949e-07, "loss": 1.4563, "step": 20540 }, { "epoch": 0.51, "grad_norm": 70.5, "learning_rate": 3.203389830508474e-07, "loss": 1.4821, "step": 20550 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 3.2e-07, "loss": 1.4666, "step": 20560 }, { "epoch": 0.51, "grad_norm": 65.5, "learning_rate": 3.1966101694915253e-07, "loss": 1.4145, "step": 20570 }, { "epoch": 0.51, "grad_norm": 62.75, "learning_rate": 3.1932203389830506e-07, "loss": 1.4483, "step": 20580 }, { "epoch": 0.51, "grad_norm": 67.5, "learning_rate": 3.1898305084745764e-07, "loss": 1.4599, "step": 20590 }, { "epoch": 0.52, "grad_norm": 66.0, "learning_rate": 3.186440677966101e-07, "loss": 1.4915, "step": 20600 }, { "epoch": 0.52, "grad_norm": 67.0, "learning_rate": 3.183050847457627e-07, "loss": 1.4934, "step": 20610 }, { "epoch": 0.52, "grad_norm": 70.0, "learning_rate": 3.179661016949152e-07, "loss": 1.4333, "step": 20620 }, { "epoch": 0.52, "grad_norm": 64.5, "learning_rate": 3.1762711864406775e-07, "loss": 1.5158, "step": 20630 }, { "epoch": 0.52, "grad_norm": 68.5, "learning_rate": 3.1728813559322033e-07, "loss": 1.426, "step": 20640 }, { "epoch": 0.52, "grad_norm": 69.5, "learning_rate": 3.1694915254237286e-07, "loss": 1.448, "step": 20650 }, { "epoch": 0.52, "grad_norm": 64.5, "learning_rate": 3.166101694915254e-07, "loss": 1.4542, "step": 20660 }, { "epoch": 0.52, "grad_norm": 65.5, "learning_rate": 3.1627118644067797e-07, "loss": 1.4202, "step": 20670 }, { "epoch": 0.52, "grad_norm": 67.0, "learning_rate": 3.159322033898305e-07, "loss": 1.4718, "step": 20680 }, { "epoch": 0.52, "grad_norm": 71.0, "learning_rate": 3.155932203389831e-07, "loss": 1.5071, "step": 20690 }, { "epoch": 0.52, "grad_norm": 68.0, "learning_rate": 3.152542372881356e-07, "loss": 1.4972, "step": 20700 }, { "epoch": 0.52, "grad_norm": 67.0, "learning_rate": 3.149152542372881e-07, "loss": 1.4734, "step": 20710 }, { "epoch": 0.52, "grad_norm": 67.5, "learning_rate": 3.1457627118644066e-07, "loss": 1.5047, "step": 20720 }, { "epoch": 0.52, "grad_norm": 70.0, "learning_rate": 3.142372881355932e-07, "loss": 1.5067, "step": 20730 }, { "epoch": 0.52, "grad_norm": 67.0, "learning_rate": 3.138983050847457e-07, "loss": 1.4898, "step": 20740 }, { "epoch": 0.52, "grad_norm": 65.5, "learning_rate": 3.135593220338983e-07, "loss": 1.499, "step": 20750 }, { "epoch": 0.52, "grad_norm": 74.0, "learning_rate": 3.132203389830508e-07, "loss": 1.4741, "step": 20760 }, { "epoch": 0.52, "grad_norm": 66.5, "learning_rate": 3.128813559322034e-07, "loss": 1.4807, "step": 20770 }, { "epoch": 0.52, "grad_norm": 68.0, "learning_rate": 3.1254237288135593e-07, "loss": 1.439, "step": 20780 }, { "epoch": 0.52, "grad_norm": 67.0, "learning_rate": 3.1220338983050846e-07, "loss": 1.4547, "step": 20790 }, { "epoch": 0.52, "grad_norm": 66.0, "learning_rate": 3.1186440677966104e-07, "loss": 1.5078, "step": 20800 }, { "epoch": 0.52, "grad_norm": 67.0, "learning_rate": 3.115254237288135e-07, "loss": 1.4719, "step": 20810 }, { "epoch": 0.52, "grad_norm": 65.5, "learning_rate": 3.1118644067796604e-07, "loss": 1.3959, "step": 20820 }, { "epoch": 0.52, "grad_norm": 63.0, "learning_rate": 3.108474576271186e-07, "loss": 1.467, "step": 20830 }, { "epoch": 0.52, "grad_norm": 67.5, "learning_rate": 3.1050847457627115e-07, "loss": 1.4718, "step": 20840 }, { "epoch": 0.52, "grad_norm": 69.0, "learning_rate": 3.1016949152542373e-07, "loss": 1.4672, "step": 20850 }, { "epoch": 0.52, "grad_norm": 66.5, "learning_rate": 3.0983050847457626e-07, "loss": 1.5616, "step": 20860 }, { "epoch": 0.52, "grad_norm": 69.5, "learning_rate": 3.094915254237288e-07, "loss": 1.494, "step": 20870 }, { "epoch": 0.52, "grad_norm": 67.0, "learning_rate": 3.0915254237288137e-07, "loss": 1.5186, "step": 20880 }, { "epoch": 0.52, "grad_norm": 70.5, "learning_rate": 3.088135593220339e-07, "loss": 1.488, "step": 20890 }, { "epoch": 0.52, "grad_norm": 68.0, "learning_rate": 3.084745762711864e-07, "loss": 1.4227, "step": 20900 }, { "epoch": 0.52, "grad_norm": 63.5, "learning_rate": 3.08135593220339e-07, "loss": 1.432, "step": 20910 }, { "epoch": 0.52, "grad_norm": 66.5, "learning_rate": 3.077966101694915e-07, "loss": 1.4613, "step": 20920 }, { "epoch": 0.52, "grad_norm": 68.0, "learning_rate": 3.0745762711864406e-07, "loss": 1.465, "step": 20930 }, { "epoch": 0.52, "grad_norm": 66.0, "learning_rate": 3.071186440677966e-07, "loss": 1.4968, "step": 20940 }, { "epoch": 0.52, "grad_norm": 71.0, "learning_rate": 3.067796610169491e-07, "loss": 1.4866, "step": 20950 }, { "epoch": 0.52, "grad_norm": 65.5, "learning_rate": 3.064406779661017e-07, "loss": 1.5076, "step": 20960 }, { "epoch": 0.52, "grad_norm": 67.5, "learning_rate": 3.061016949152542e-07, "loss": 1.4341, "step": 20970 }, { "epoch": 0.52, "grad_norm": 66.5, "learning_rate": 3.0576271186440675e-07, "loss": 1.4862, "step": 20980 }, { "epoch": 0.52, "grad_norm": 67.0, "learning_rate": 3.0542372881355933e-07, "loss": 1.5154, "step": 20990 }, { "epoch": 0.53, "grad_norm": 68.0, "learning_rate": 3.0508474576271186e-07, "loss": 1.4475, "step": 21000 }, { "epoch": 0.53, "grad_norm": 64.5, "learning_rate": 3.0474576271186444e-07, "loss": 1.489, "step": 21010 }, { "epoch": 0.53, "grad_norm": 73.0, "learning_rate": 3.044067796610169e-07, "loss": 1.4482, "step": 21020 }, { "epoch": 0.53, "grad_norm": 66.5, "learning_rate": 3.0406779661016944e-07, "loss": 1.5017, "step": 21030 }, { "epoch": 0.53, "grad_norm": 65.5, "learning_rate": 3.03728813559322e-07, "loss": 1.4573, "step": 21040 }, { "epoch": 0.53, "grad_norm": 66.5, "learning_rate": 3.0338983050847455e-07, "loss": 1.4509, "step": 21050 }, { "epoch": 0.53, "grad_norm": 71.0, "learning_rate": 3.030508474576271e-07, "loss": 1.4765, "step": 21060 }, { "epoch": 0.53, "grad_norm": 65.0, "learning_rate": 3.0271186440677966e-07, "loss": 1.4637, "step": 21070 }, { "epoch": 0.53, "grad_norm": 71.5, "learning_rate": 3.023728813559322e-07, "loss": 1.4916, "step": 21080 }, { "epoch": 0.53, "grad_norm": 68.0, "learning_rate": 3.0203389830508477e-07, "loss": 1.5076, "step": 21090 }, { "epoch": 0.53, "grad_norm": 68.0, "learning_rate": 3.016949152542373e-07, "loss": 1.4416, "step": 21100 }, { "epoch": 0.53, "grad_norm": 66.0, "learning_rate": 3.013559322033898e-07, "loss": 1.4528, "step": 21110 }, { "epoch": 0.53, "grad_norm": 67.0, "learning_rate": 3.0101694915254235e-07, "loss": 1.4604, "step": 21120 }, { "epoch": 0.53, "grad_norm": 67.5, "learning_rate": 3.006779661016949e-07, "loss": 1.5248, "step": 21130 }, { "epoch": 0.53, "grad_norm": 68.5, "learning_rate": 3.003389830508474e-07, "loss": 1.5132, "step": 21140 }, { "epoch": 0.53, "grad_norm": 65.5, "learning_rate": 3e-07, "loss": 1.4708, "step": 21150 }, { "epoch": 0.53, "grad_norm": 65.0, "learning_rate": 2.996610169491525e-07, "loss": 1.4542, "step": 21160 }, { "epoch": 0.53, "grad_norm": 68.5, "learning_rate": 2.993220338983051e-07, "loss": 1.4819, "step": 21170 }, { "epoch": 0.53, "grad_norm": 66.0, "learning_rate": 2.989830508474576e-07, "loss": 1.4844, "step": 21180 }, { "epoch": 0.53, "grad_norm": 69.0, "learning_rate": 2.9864406779661015e-07, "loss": 1.4743, "step": 21190 }, { "epoch": 0.53, "grad_norm": 65.5, "learning_rate": 2.9830508474576273e-07, "loss": 1.4118, "step": 21200 }, { "epoch": 0.53, "grad_norm": 68.0, "learning_rate": 2.9796610169491526e-07, "loss": 1.4716, "step": 21210 }, { "epoch": 0.53, "grad_norm": 64.5, "learning_rate": 2.976271186440678e-07, "loss": 1.469, "step": 21220 }, { "epoch": 0.53, "grad_norm": 66.5, "learning_rate": 2.972881355932203e-07, "loss": 1.4809, "step": 21230 }, { "epoch": 0.53, "grad_norm": 64.5, "learning_rate": 2.9694915254237284e-07, "loss": 1.4595, "step": 21240 }, { "epoch": 0.53, "grad_norm": 73.0, "learning_rate": 2.966101694915254e-07, "loss": 1.4719, "step": 21250 }, { "epoch": 0.53, "grad_norm": 66.0, "learning_rate": 2.9627118644067795e-07, "loss": 1.4414, "step": 21260 }, { "epoch": 0.53, "grad_norm": 63.25, "learning_rate": 2.959322033898305e-07, "loss": 1.4531, "step": 21270 }, { "epoch": 0.53, "grad_norm": 65.0, "learning_rate": 2.9559322033898306e-07, "loss": 1.4457, "step": 21280 }, { "epoch": 0.53, "grad_norm": 67.5, "learning_rate": 2.952542372881356e-07, "loss": 1.4854, "step": 21290 }, { "epoch": 0.53, "grad_norm": 67.0, "learning_rate": 2.949152542372881e-07, "loss": 1.4871, "step": 21300 }, { "epoch": 0.53, "grad_norm": 68.0, "learning_rate": 2.945762711864407e-07, "loss": 1.4342, "step": 21310 }, { "epoch": 0.53, "grad_norm": 65.0, "learning_rate": 2.942372881355932e-07, "loss": 1.4727, "step": 21320 }, { "epoch": 0.53, "grad_norm": 71.0, "learning_rate": 2.9389830508474575e-07, "loss": 1.4878, "step": 21330 }, { "epoch": 0.53, "grad_norm": 69.0, "learning_rate": 2.935593220338983e-07, "loss": 1.4561, "step": 21340 }, { "epoch": 0.53, "grad_norm": 69.0, "learning_rate": 2.932203389830508e-07, "loss": 1.414, "step": 21350 }, { "epoch": 0.53, "grad_norm": 66.5, "learning_rate": 2.928813559322034e-07, "loss": 1.436, "step": 21360 }, { "epoch": 0.53, "grad_norm": 69.5, "learning_rate": 2.925423728813559e-07, "loss": 1.4707, "step": 21370 }, { "epoch": 0.53, "grad_norm": 66.5, "learning_rate": 2.9220338983050844e-07, "loss": 1.4839, "step": 21380 }, { "epoch": 0.53, "grad_norm": 70.5, "learning_rate": 2.91864406779661e-07, "loss": 1.4895, "step": 21390 }, { "epoch": 0.54, "grad_norm": 67.0, "learning_rate": 2.9152542372881355e-07, "loss": 1.4373, "step": 21400 }, { "epoch": 0.54, "grad_norm": 65.0, "learning_rate": 2.9118644067796613e-07, "loss": 1.3779, "step": 21410 }, { "epoch": 0.54, "grad_norm": 66.5, "learning_rate": 2.9084745762711866e-07, "loss": 1.4942, "step": 21420 }, { "epoch": 0.54, "grad_norm": 67.0, "learning_rate": 2.905084745762712e-07, "loss": 1.4563, "step": 21430 }, { "epoch": 0.54, "grad_norm": 63.25, "learning_rate": 2.901694915254237e-07, "loss": 1.4411, "step": 21440 }, { "epoch": 0.54, "grad_norm": 68.5, "learning_rate": 2.8983050847457624e-07, "loss": 1.4222, "step": 21450 }, { "epoch": 0.54, "grad_norm": 65.5, "learning_rate": 2.8949152542372877e-07, "loss": 1.4217, "step": 21460 }, { "epoch": 0.54, "grad_norm": 68.0, "learning_rate": 2.8915254237288135e-07, "loss": 1.4356, "step": 21470 }, { "epoch": 0.54, "grad_norm": 67.0, "learning_rate": 2.888135593220339e-07, "loss": 1.477, "step": 21480 }, { "epoch": 0.54, "grad_norm": 62.5, "learning_rate": 2.8847457627118646e-07, "loss": 1.4594, "step": 21490 }, { "epoch": 0.54, "grad_norm": 65.5, "learning_rate": 2.88135593220339e-07, "loss": 1.4453, "step": 21500 }, { "epoch": 0.54, "grad_norm": 62.5, "learning_rate": 2.877966101694915e-07, "loss": 1.4637, "step": 21510 }, { "epoch": 0.54, "grad_norm": 69.5, "learning_rate": 2.874576271186441e-07, "loss": 1.4335, "step": 21520 }, { "epoch": 0.54, "grad_norm": 68.0, "learning_rate": 2.871186440677966e-07, "loss": 1.4404, "step": 21530 }, { "epoch": 0.54, "grad_norm": 66.5, "learning_rate": 2.867796610169491e-07, "loss": 1.489, "step": 21540 }, { "epoch": 0.54, "grad_norm": 64.0, "learning_rate": 2.864406779661017e-07, "loss": 1.4628, "step": 21550 }, { "epoch": 0.54, "grad_norm": 65.0, "learning_rate": 2.861016949152542e-07, "loss": 1.4929, "step": 21560 }, { "epoch": 0.54, "grad_norm": 68.5, "learning_rate": 2.857627118644068e-07, "loss": 1.4815, "step": 21570 }, { "epoch": 0.54, "grad_norm": 66.0, "learning_rate": 2.854237288135593e-07, "loss": 1.4563, "step": 21580 }, { "epoch": 0.54, "grad_norm": 67.5, "learning_rate": 2.8508474576271184e-07, "loss": 1.4754, "step": 21590 }, { "epoch": 0.54, "grad_norm": 67.0, "learning_rate": 2.847457627118644e-07, "loss": 1.5266, "step": 21600 }, { "epoch": 0.54, "grad_norm": 65.0, "learning_rate": 2.8440677966101695e-07, "loss": 1.4389, "step": 21610 }, { "epoch": 0.54, "grad_norm": 66.5, "learning_rate": 2.840677966101695e-07, "loss": 1.4256, "step": 21620 }, { "epoch": 0.54, "grad_norm": 66.0, "learning_rate": 2.8372881355932206e-07, "loss": 1.4931, "step": 21630 }, { "epoch": 0.54, "grad_norm": 65.5, "learning_rate": 2.8338983050847453e-07, "loss": 1.5427, "step": 21640 }, { "epoch": 0.54, "grad_norm": 66.5, "learning_rate": 2.8305084745762706e-07, "loss": 1.4232, "step": 21650 }, { "epoch": 0.54, "grad_norm": 66.0, "learning_rate": 2.8271186440677964e-07, "loss": 1.484, "step": 21660 }, { "epoch": 0.54, "grad_norm": 67.0, "learning_rate": 2.8237288135593217e-07, "loss": 1.4328, "step": 21670 }, { "epoch": 0.54, "grad_norm": 66.5, "learning_rate": 2.8203389830508475e-07, "loss": 1.4981, "step": 21680 }, { "epoch": 0.54, "grad_norm": 68.5, "learning_rate": 2.816949152542373e-07, "loss": 1.4774, "step": 21690 }, { "epoch": 0.54, "grad_norm": 66.5, "learning_rate": 2.813559322033898e-07, "loss": 1.4202, "step": 21700 }, { "epoch": 0.54, "grad_norm": 66.5, "learning_rate": 2.810169491525424e-07, "loss": 1.4588, "step": 21710 }, { "epoch": 0.54, "grad_norm": 69.5, "learning_rate": 2.806779661016949e-07, "loss": 1.5449, "step": 21720 }, { "epoch": 0.54, "grad_norm": 70.5, "learning_rate": 2.803389830508475e-07, "loss": 1.5002, "step": 21730 }, { "epoch": 0.54, "grad_norm": 68.0, "learning_rate": 2.8e-07, "loss": 1.463, "step": 21740 }, { "epoch": 0.54, "grad_norm": 66.5, "learning_rate": 2.796610169491525e-07, "loss": 1.467, "step": 21750 }, { "epoch": 0.54, "grad_norm": 69.5, "learning_rate": 2.793220338983051e-07, "loss": 1.4127, "step": 21760 }, { "epoch": 0.54, "grad_norm": 71.5, "learning_rate": 2.789830508474576e-07, "loss": 1.493, "step": 21770 }, { "epoch": 0.54, "grad_norm": 64.0, "learning_rate": 2.7864406779661013e-07, "loss": 1.4866, "step": 21780 }, { "epoch": 0.54, "grad_norm": 69.0, "learning_rate": 2.783050847457627e-07, "loss": 1.5279, "step": 21790 }, { "epoch": 0.55, "grad_norm": 69.0, "learning_rate": 2.7796610169491524e-07, "loss": 1.5252, "step": 21800 }, { "epoch": 0.55, "grad_norm": 67.0, "learning_rate": 2.776271186440678e-07, "loss": 1.4794, "step": 21810 }, { "epoch": 0.55, "grad_norm": 67.5, "learning_rate": 2.7728813559322035e-07, "loss": 1.4078, "step": 21820 }, { "epoch": 0.55, "grad_norm": 69.0, "learning_rate": 2.769491525423729e-07, "loss": 1.4782, "step": 21830 }, { "epoch": 0.55, "grad_norm": 70.5, "learning_rate": 2.7661016949152545e-07, "loss": 1.4411, "step": 21840 }, { "epoch": 0.55, "grad_norm": 63.75, "learning_rate": 2.7627118644067793e-07, "loss": 1.393, "step": 21850 }, { "epoch": 0.55, "grad_norm": 67.0, "learning_rate": 2.7593220338983046e-07, "loss": 1.4784, "step": 21860 }, { "epoch": 0.55, "grad_norm": 68.0, "learning_rate": 2.7559322033898304e-07, "loss": 1.4012, "step": 21870 }, { "epoch": 0.55, "grad_norm": 65.5, "learning_rate": 2.7525423728813557e-07, "loss": 1.477, "step": 21880 }, { "epoch": 0.55, "grad_norm": 66.5, "learning_rate": 2.749152542372881e-07, "loss": 1.5067, "step": 21890 }, { "epoch": 0.55, "grad_norm": 67.5, "learning_rate": 2.745762711864407e-07, "loss": 1.4631, "step": 21900 }, { "epoch": 0.55, "grad_norm": 66.5, "learning_rate": 2.742372881355932e-07, "loss": 1.5257, "step": 21910 }, { "epoch": 0.55, "grad_norm": 67.5, "learning_rate": 2.738983050847458e-07, "loss": 1.4561, "step": 21920 }, { "epoch": 0.55, "grad_norm": 65.5, "learning_rate": 2.735593220338983e-07, "loss": 1.4773, "step": 21930 }, { "epoch": 0.55, "grad_norm": 67.5, "learning_rate": 2.7322033898305084e-07, "loss": 1.4732, "step": 21940 }, { "epoch": 0.55, "grad_norm": 66.0, "learning_rate": 2.7288135593220336e-07, "loss": 1.4743, "step": 21950 }, { "epoch": 0.55, "grad_norm": 67.0, "learning_rate": 2.725423728813559e-07, "loss": 1.4624, "step": 21960 }, { "epoch": 0.55, "grad_norm": 63.5, "learning_rate": 2.722033898305084e-07, "loss": 1.4451, "step": 21970 }, { "epoch": 0.55, "grad_norm": 64.5, "learning_rate": 2.71864406779661e-07, "loss": 1.4502, "step": 21980 }, { "epoch": 0.55, "grad_norm": 68.0, "learning_rate": 2.7152542372881353e-07, "loss": 1.4902, "step": 21990 }, { "epoch": 0.55, "grad_norm": 65.5, "learning_rate": 2.711864406779661e-07, "loss": 1.4215, "step": 22000 }, { "epoch": 0.55, "grad_norm": 66.0, "learning_rate": 2.7084745762711864e-07, "loss": 1.4646, "step": 22010 }, { "epoch": 0.55, "grad_norm": 65.5, "learning_rate": 2.7050847457627116e-07, "loss": 1.4843, "step": 22020 }, { "epoch": 0.55, "grad_norm": 66.5, "learning_rate": 2.7016949152542375e-07, "loss": 1.4881, "step": 22030 }, { "epoch": 0.55, "grad_norm": 68.5, "learning_rate": 2.6983050847457627e-07, "loss": 1.4136, "step": 22040 }, { "epoch": 0.55, "grad_norm": 63.25, "learning_rate": 2.6949152542372885e-07, "loss": 1.4735, "step": 22050 }, { "epoch": 0.55, "grad_norm": 67.5, "learning_rate": 2.6915254237288133e-07, "loss": 1.4365, "step": 22060 }, { "epoch": 0.55, "grad_norm": 68.0, "learning_rate": 2.6881355932203386e-07, "loss": 1.5033, "step": 22070 }, { "epoch": 0.55, "grad_norm": 65.5, "learning_rate": 2.6847457627118644e-07, "loss": 1.4978, "step": 22080 }, { "epoch": 0.55, "grad_norm": 67.0, "learning_rate": 2.6813559322033896e-07, "loss": 1.4561, "step": 22090 }, { "epoch": 0.55, "grad_norm": 65.5, "learning_rate": 2.677966101694915e-07, "loss": 1.5001, "step": 22100 }, { "epoch": 0.55, "grad_norm": 66.5, "learning_rate": 2.6745762711864407e-07, "loss": 1.4332, "step": 22110 }, { "epoch": 0.55, "grad_norm": 67.5, "learning_rate": 2.671186440677966e-07, "loss": 1.4629, "step": 22120 }, { "epoch": 0.55, "grad_norm": 67.5, "learning_rate": 2.6677966101694913e-07, "loss": 1.5131, "step": 22130 }, { "epoch": 0.55, "grad_norm": 66.0, "learning_rate": 2.664406779661017e-07, "loss": 1.4893, "step": 22140 }, { "epoch": 0.55, "grad_norm": 68.0, "learning_rate": 2.6610169491525424e-07, "loss": 1.4378, "step": 22150 }, { "epoch": 0.55, "grad_norm": 69.5, "learning_rate": 2.6576271186440676e-07, "loss": 1.4219, "step": 22160 }, { "epoch": 0.55, "grad_norm": 66.5, "learning_rate": 2.654237288135593e-07, "loss": 1.4722, "step": 22170 }, { "epoch": 0.55, "grad_norm": 68.0, "learning_rate": 2.650847457627118e-07, "loss": 1.509, "step": 22180 }, { "epoch": 0.55, "grad_norm": 65.5, "learning_rate": 2.647457627118644e-07, "loss": 1.4875, "step": 22190 }, { "epoch": 0.56, "grad_norm": 72.5, "learning_rate": 2.6440677966101693e-07, "loss": 1.4397, "step": 22200 }, { "epoch": 0.56, "grad_norm": 67.5, "learning_rate": 2.6406779661016945e-07, "loss": 1.4862, "step": 22210 }, { "epoch": 0.56, "grad_norm": 70.0, "learning_rate": 2.6372881355932204e-07, "loss": 1.4248, "step": 22220 }, { "epoch": 0.56, "grad_norm": 67.5, "learning_rate": 2.6338983050847456e-07, "loss": 1.485, "step": 22230 }, { "epoch": 0.56, "grad_norm": 67.5, "learning_rate": 2.6305084745762714e-07, "loss": 1.4798, "step": 22240 }, { "epoch": 0.56, "grad_norm": 64.0, "learning_rate": 2.6271186440677967e-07, "loss": 1.4664, "step": 22250 }, { "epoch": 0.56, "grad_norm": 65.5, "learning_rate": 2.623728813559322e-07, "loss": 1.4435, "step": 22260 }, { "epoch": 0.56, "grad_norm": 64.0, "learning_rate": 2.6203389830508473e-07, "loss": 1.5132, "step": 22270 }, { "epoch": 0.56, "grad_norm": 67.0, "learning_rate": 2.6169491525423725e-07, "loss": 1.4555, "step": 22280 }, { "epoch": 0.56, "grad_norm": 69.5, "learning_rate": 2.613559322033898e-07, "loss": 1.4874, "step": 22290 }, { "epoch": 0.56, "grad_norm": 67.0, "learning_rate": 2.6101694915254236e-07, "loss": 1.4806, "step": 22300 }, { "epoch": 0.56, "grad_norm": 69.5, "learning_rate": 2.606779661016949e-07, "loss": 1.4733, "step": 22310 }, { "epoch": 0.56, "grad_norm": 65.5, "learning_rate": 2.6033898305084747e-07, "loss": 1.4037, "step": 22320 }, { "epoch": 0.56, "grad_norm": 67.0, "learning_rate": 2.6e-07, "loss": 1.4826, "step": 22330 }, { "epoch": 0.56, "grad_norm": 69.0, "learning_rate": 2.5966101694915253e-07, "loss": 1.4517, "step": 22340 }, { "epoch": 0.56, "grad_norm": 68.5, "learning_rate": 2.593220338983051e-07, "loss": 1.4646, "step": 22350 }, { "epoch": 0.56, "grad_norm": 66.5, "learning_rate": 2.5898305084745763e-07, "loss": 1.463, "step": 22360 }, { "epoch": 0.56, "grad_norm": 65.0, "learning_rate": 2.586440677966101e-07, "loss": 1.3894, "step": 22370 }, { "epoch": 0.56, "grad_norm": 66.0, "learning_rate": 2.583050847457627e-07, "loss": 1.4653, "step": 22380 }, { "epoch": 0.56, "grad_norm": 64.0, "learning_rate": 2.579661016949152e-07, "loss": 1.4551, "step": 22390 }, { "epoch": 0.56, "grad_norm": 70.5, "learning_rate": 2.576271186440678e-07, "loss": 1.4699, "step": 22400 }, { "epoch": 0.56, "grad_norm": 68.0, "learning_rate": 2.572881355932203e-07, "loss": 1.4967, "step": 22410 }, { "epoch": 0.56, "grad_norm": 66.5, "learning_rate": 2.5694915254237285e-07, "loss": 1.4465, "step": 22420 }, { "epoch": 0.56, "grad_norm": 71.0, "learning_rate": 2.5661016949152543e-07, "loss": 1.5049, "step": 22430 }, { "epoch": 0.56, "grad_norm": 68.0, "learning_rate": 2.5627118644067796e-07, "loss": 1.459, "step": 22440 }, { "epoch": 0.56, "grad_norm": 64.5, "learning_rate": 2.559322033898305e-07, "loss": 1.4719, "step": 22450 }, { "epoch": 0.56, "grad_norm": 68.5, "learning_rate": 2.5559322033898307e-07, "loss": 1.4305, "step": 22460 }, { "epoch": 0.56, "grad_norm": 68.0, "learning_rate": 2.5525423728813555e-07, "loss": 1.4761, "step": 22470 }, { "epoch": 0.56, "grad_norm": 66.5, "learning_rate": 2.549152542372881e-07, "loss": 1.4834, "step": 22480 }, { "epoch": 0.56, "grad_norm": 68.0, "learning_rate": 2.5457627118644065e-07, "loss": 1.4149, "step": 22490 }, { "epoch": 0.56, "grad_norm": 66.0, "learning_rate": 2.542372881355932e-07, "loss": 1.4605, "step": 22500 }, { "epoch": 0.56, "grad_norm": 68.5, "learning_rate": 2.5389830508474576e-07, "loss": 1.4435, "step": 22510 }, { "epoch": 0.56, "grad_norm": 68.5, "learning_rate": 2.535593220338983e-07, "loss": 1.5106, "step": 22520 }, { "epoch": 0.56, "grad_norm": 66.5, "learning_rate": 2.532203389830508e-07, "loss": 1.4488, "step": 22530 }, { "epoch": 0.56, "grad_norm": 63.75, "learning_rate": 2.528813559322034e-07, "loss": 1.4492, "step": 22540 }, { "epoch": 0.56, "grad_norm": 66.5, "learning_rate": 2.525423728813559e-07, "loss": 1.4728, "step": 22550 }, { "epoch": 0.56, "grad_norm": 68.0, "learning_rate": 2.522033898305085e-07, "loss": 1.5016, "step": 22560 }, { "epoch": 0.56, "grad_norm": 68.5, "learning_rate": 2.5186440677966103e-07, "loss": 1.4636, "step": 22570 }, { "epoch": 0.56, "grad_norm": 66.5, "learning_rate": 2.515254237288135e-07, "loss": 1.4976, "step": 22580 }, { "epoch": 0.56, "grad_norm": 70.0, "learning_rate": 2.511864406779661e-07, "loss": 1.47, "step": 22590 }, { "epoch": 0.56, "grad_norm": 65.0, "learning_rate": 2.508474576271186e-07, "loss": 1.4302, "step": 22600 }, { "epoch": 0.57, "grad_norm": 71.5, "learning_rate": 2.5050847457627114e-07, "loss": 1.4743, "step": 22610 }, { "epoch": 0.57, "grad_norm": 72.0, "learning_rate": 2.501694915254237e-07, "loss": 1.4998, "step": 22620 }, { "epoch": 0.57, "grad_norm": 66.0, "learning_rate": 2.4983050847457625e-07, "loss": 1.4342, "step": 22630 }, { "epoch": 0.57, "grad_norm": 66.0, "learning_rate": 2.4949152542372883e-07, "loss": 1.4675, "step": 22640 }, { "epoch": 0.57, "grad_norm": 66.5, "learning_rate": 2.4915254237288136e-07, "loss": 1.4961, "step": 22650 }, { "epoch": 0.57, "grad_norm": 67.5, "learning_rate": 2.488135593220339e-07, "loss": 1.4907, "step": 22660 }, { "epoch": 0.57, "grad_norm": 68.0, "learning_rate": 2.484745762711864e-07, "loss": 1.4894, "step": 22670 }, { "epoch": 0.57, "grad_norm": 66.0, "learning_rate": 2.48135593220339e-07, "loss": 1.4661, "step": 22680 }, { "epoch": 0.57, "grad_norm": 65.0, "learning_rate": 2.477966101694915e-07, "loss": 1.4717, "step": 22690 }, { "epoch": 0.57, "grad_norm": 63.75, "learning_rate": 2.4745762711864405e-07, "loss": 1.4701, "step": 22700 }, { "epoch": 0.57, "grad_norm": 66.0, "learning_rate": 2.471186440677966e-07, "loss": 1.4747, "step": 22710 }, { "epoch": 0.57, "grad_norm": 67.0, "learning_rate": 2.4677966101694916e-07, "loss": 1.4711, "step": 22720 }, { "epoch": 0.57, "grad_norm": 66.5, "learning_rate": 2.464406779661017e-07, "loss": 1.4452, "step": 22730 }, { "epoch": 0.57, "grad_norm": 67.5, "learning_rate": 2.461016949152542e-07, "loss": 1.4918, "step": 22740 }, { "epoch": 0.57, "grad_norm": 68.0, "learning_rate": 2.457627118644068e-07, "loss": 1.4777, "step": 22750 }, { "epoch": 0.57, "grad_norm": 65.0, "learning_rate": 2.454237288135593e-07, "loss": 1.4705, "step": 22760 }, { "epoch": 0.57, "grad_norm": 66.5, "learning_rate": 2.4508474576271185e-07, "loss": 1.4481, "step": 22770 }, { "epoch": 0.57, "grad_norm": 67.0, "learning_rate": 2.447457627118644e-07, "loss": 1.4624, "step": 22780 }, { "epoch": 0.57, "grad_norm": 65.5, "learning_rate": 2.4440677966101696e-07, "loss": 1.4995, "step": 22790 }, { "epoch": 0.57, "grad_norm": 67.0, "learning_rate": 2.440677966101695e-07, "loss": 1.4544, "step": 22800 }, { "epoch": 0.57, "grad_norm": 66.0, "learning_rate": 2.43728813559322e-07, "loss": 1.4868, "step": 22810 }, { "epoch": 0.57, "grad_norm": 65.5, "learning_rate": 2.4338983050847454e-07, "loss": 1.4181, "step": 22820 }, { "epoch": 0.57, "grad_norm": 66.0, "learning_rate": 2.430508474576271e-07, "loss": 1.4431, "step": 22830 }, { "epoch": 0.57, "grad_norm": 66.5, "learning_rate": 2.4271186440677965e-07, "loss": 1.4813, "step": 22840 }, { "epoch": 0.57, "grad_norm": 66.0, "learning_rate": 2.423728813559322e-07, "loss": 1.5211, "step": 22850 }, { "epoch": 0.57, "grad_norm": 65.0, "learning_rate": 2.420338983050847e-07, "loss": 1.4331, "step": 22860 }, { "epoch": 0.57, "grad_norm": 66.5, "learning_rate": 2.416949152542373e-07, "loss": 1.5019, "step": 22870 }, { "epoch": 0.57, "grad_norm": 68.0, "learning_rate": 2.413559322033898e-07, "loss": 1.4635, "step": 22880 }, { "epoch": 0.57, "grad_norm": 67.5, "learning_rate": 2.4101694915254234e-07, "loss": 1.4641, "step": 22890 }, { "epoch": 0.57, "grad_norm": 66.0, "learning_rate": 2.406779661016949e-07, "loss": 1.4351, "step": 22900 }, { "epoch": 0.57, "grad_norm": 71.0, "learning_rate": 2.4033898305084745e-07, "loss": 1.4785, "step": 22910 }, { "epoch": 0.57, "grad_norm": 66.5, "learning_rate": 2.4e-07, "loss": 1.4715, "step": 22920 }, { "epoch": 0.57, "grad_norm": 67.5, "learning_rate": 2.396610169491525e-07, "loss": 1.4631, "step": 22930 }, { "epoch": 0.57, "grad_norm": 66.5, "learning_rate": 2.393220338983051e-07, "loss": 1.5024, "step": 22940 }, { "epoch": 0.57, "grad_norm": 68.0, "learning_rate": 2.389830508474576e-07, "loss": 1.4602, "step": 22950 }, { "epoch": 0.57, "grad_norm": 66.5, "learning_rate": 2.386440677966102e-07, "loss": 1.3829, "step": 22960 }, { "epoch": 0.57, "grad_norm": 67.0, "learning_rate": 2.383050847457627e-07, "loss": 1.4193, "step": 22970 }, { "epoch": 0.57, "grad_norm": 70.5, "learning_rate": 2.3796610169491525e-07, "loss": 1.4878, "step": 22980 }, { "epoch": 0.57, "grad_norm": 65.5, "learning_rate": 2.3762711864406778e-07, "loss": 1.4852, "step": 22990 }, { "epoch": 0.57, "grad_norm": 70.0, "learning_rate": 2.3728813559322033e-07, "loss": 1.4789, "step": 23000 }, { "epoch": 0.58, "grad_norm": 71.0, "learning_rate": 2.3694915254237289e-07, "loss": 1.4596, "step": 23010 }, { "epoch": 0.58, "grad_norm": 64.5, "learning_rate": 2.3661016949152541e-07, "loss": 1.4706, "step": 23020 }, { "epoch": 0.58, "grad_norm": 68.5, "learning_rate": 2.3627118644067794e-07, "loss": 1.4422, "step": 23030 }, { "epoch": 0.58, "grad_norm": 65.0, "learning_rate": 2.359322033898305e-07, "loss": 1.441, "step": 23040 }, { "epoch": 0.58, "grad_norm": 67.5, "learning_rate": 2.3559322033898305e-07, "loss": 1.4376, "step": 23050 }, { "epoch": 0.58, "grad_norm": 64.5, "learning_rate": 2.352542372881356e-07, "loss": 1.4766, "step": 23060 }, { "epoch": 0.58, "grad_norm": 66.0, "learning_rate": 2.349152542372881e-07, "loss": 1.482, "step": 23070 }, { "epoch": 0.58, "grad_norm": 65.0, "learning_rate": 2.3457627118644066e-07, "loss": 1.4598, "step": 23080 }, { "epoch": 0.58, "grad_norm": 67.5, "learning_rate": 2.3423728813559321e-07, "loss": 1.5314, "step": 23090 }, { "epoch": 0.58, "grad_norm": 66.5, "learning_rate": 2.3389830508474577e-07, "loss": 1.4375, "step": 23100 }, { "epoch": 0.58, "grad_norm": 68.0, "learning_rate": 2.335593220338983e-07, "loss": 1.4928, "step": 23110 }, { "epoch": 0.58, "grad_norm": 67.5, "learning_rate": 2.3322033898305082e-07, "loss": 1.4823, "step": 23120 }, { "epoch": 0.58, "grad_norm": 70.0, "learning_rate": 2.3288135593220338e-07, "loss": 1.5182, "step": 23130 }, { "epoch": 0.58, "grad_norm": 68.0, "learning_rate": 2.3254237288135593e-07, "loss": 1.5249, "step": 23140 }, { "epoch": 0.58, "grad_norm": 67.0, "learning_rate": 2.3220338983050846e-07, "loss": 1.4414, "step": 23150 }, { "epoch": 0.58, "grad_norm": 67.0, "learning_rate": 2.3186440677966101e-07, "loss": 1.4908, "step": 23160 }, { "epoch": 0.58, "grad_norm": 66.0, "learning_rate": 2.3152542372881357e-07, "loss": 1.4741, "step": 23170 }, { "epoch": 0.58, "grad_norm": 64.5, "learning_rate": 2.311864406779661e-07, "loss": 1.4474, "step": 23180 }, { "epoch": 0.58, "grad_norm": 65.5, "learning_rate": 2.3084745762711862e-07, "loss": 1.498, "step": 23190 }, { "epoch": 0.58, "grad_norm": 66.5, "learning_rate": 2.3050847457627118e-07, "loss": 1.4339, "step": 23200 }, { "epoch": 0.58, "grad_norm": 67.5, "learning_rate": 2.3016949152542373e-07, "loss": 1.5245, "step": 23210 }, { "epoch": 0.58, "grad_norm": 69.0, "learning_rate": 2.2983050847457629e-07, "loss": 1.5048, "step": 23220 }, { "epoch": 0.58, "grad_norm": 65.0, "learning_rate": 2.2949152542372879e-07, "loss": 1.47, "step": 23230 }, { "epoch": 0.58, "grad_norm": 65.0, "learning_rate": 2.2915254237288134e-07, "loss": 1.4383, "step": 23240 }, { "epoch": 0.58, "grad_norm": 69.0, "learning_rate": 2.288135593220339e-07, "loss": 1.5034, "step": 23250 }, { "epoch": 0.58, "grad_norm": 63.0, "learning_rate": 2.2847457627118645e-07, "loss": 1.4556, "step": 23260 }, { "epoch": 0.58, "grad_norm": 66.5, "learning_rate": 2.2813559322033898e-07, "loss": 1.4525, "step": 23270 }, { "epoch": 0.58, "grad_norm": 65.5, "learning_rate": 2.277966101694915e-07, "loss": 1.4466, "step": 23280 }, { "epoch": 0.58, "grad_norm": 70.0, "learning_rate": 2.2745762711864406e-07, "loss": 1.3894, "step": 23290 }, { "epoch": 0.58, "grad_norm": 62.5, "learning_rate": 2.271186440677966e-07, "loss": 1.356, "step": 23300 }, { "epoch": 0.58, "grad_norm": 67.5, "learning_rate": 2.2677966101694914e-07, "loss": 1.5171, "step": 23310 }, { "epoch": 0.58, "grad_norm": 66.5, "learning_rate": 2.264406779661017e-07, "loss": 1.4655, "step": 23320 }, { "epoch": 0.58, "grad_norm": 67.5, "learning_rate": 2.2610169491525422e-07, "loss": 1.4562, "step": 23330 }, { "epoch": 0.58, "grad_norm": 71.0, "learning_rate": 2.2576271186440678e-07, "loss": 1.4748, "step": 23340 }, { "epoch": 0.58, "grad_norm": 65.0, "learning_rate": 2.254237288135593e-07, "loss": 1.5118, "step": 23350 }, { "epoch": 0.58, "grad_norm": 73.0, "learning_rate": 2.2508474576271186e-07, "loss": 1.4944, "step": 23360 }, { "epoch": 0.58, "grad_norm": 68.5, "learning_rate": 2.247457627118644e-07, "loss": 1.4614, "step": 23370 }, { "epoch": 0.58, "grad_norm": 68.0, "learning_rate": 2.2440677966101691e-07, "loss": 1.4759, "step": 23380 }, { "epoch": 0.58, "grad_norm": 73.0, "learning_rate": 2.2406779661016947e-07, "loss": 1.4913, "step": 23390 }, { "epoch": 0.58, "grad_norm": 63.5, "learning_rate": 2.2372881355932202e-07, "loss": 1.4718, "step": 23400 }, { "epoch": 0.59, "grad_norm": 67.5, "learning_rate": 2.2338983050847458e-07, "loss": 1.4678, "step": 23410 }, { "epoch": 0.59, "grad_norm": 71.0, "learning_rate": 2.2305084745762713e-07, "loss": 1.49, "step": 23420 }, { "epoch": 0.59, "grad_norm": 72.0, "learning_rate": 2.2271186440677966e-07, "loss": 1.4582, "step": 23430 }, { "epoch": 0.59, "grad_norm": 65.5, "learning_rate": 2.2237288135593219e-07, "loss": 1.4385, "step": 23440 }, { "epoch": 0.59, "grad_norm": 67.0, "learning_rate": 2.2203389830508474e-07, "loss": 1.5076, "step": 23450 }, { "epoch": 0.59, "grad_norm": 66.5, "learning_rate": 2.216949152542373e-07, "loss": 1.4824, "step": 23460 }, { "epoch": 0.59, "grad_norm": 68.0, "learning_rate": 2.2135593220338982e-07, "loss": 1.4799, "step": 23470 }, { "epoch": 0.59, "grad_norm": 67.0, "learning_rate": 2.2101694915254238e-07, "loss": 1.4648, "step": 23480 }, { "epoch": 0.59, "grad_norm": 67.0, "learning_rate": 2.206779661016949e-07, "loss": 1.4858, "step": 23490 }, { "epoch": 0.59, "grad_norm": 68.0, "learning_rate": 2.2033898305084743e-07, "loss": 1.5103, "step": 23500 }, { "epoch": 0.59, "grad_norm": 67.5, "learning_rate": 2.1999999999999998e-07, "loss": 1.4786, "step": 23510 }, { "epoch": 0.59, "grad_norm": 66.5, "learning_rate": 2.1966101694915254e-07, "loss": 1.3982, "step": 23520 }, { "epoch": 0.59, "grad_norm": 67.0, "learning_rate": 2.193220338983051e-07, "loss": 1.4859, "step": 23530 }, { "epoch": 0.59, "grad_norm": 64.5, "learning_rate": 2.189830508474576e-07, "loss": 1.4449, "step": 23540 }, { "epoch": 0.59, "grad_norm": 66.0, "learning_rate": 2.1864406779661015e-07, "loss": 1.425, "step": 23550 }, { "epoch": 0.59, "grad_norm": 67.0, "learning_rate": 2.183050847457627e-07, "loss": 1.4306, "step": 23560 }, { "epoch": 0.59, "grad_norm": 66.0, "learning_rate": 2.1796610169491526e-07, "loss": 1.3738, "step": 23570 }, { "epoch": 0.59, "grad_norm": 64.0, "learning_rate": 2.1762711864406778e-07, "loss": 1.4323, "step": 23580 }, { "epoch": 0.59, "grad_norm": 69.0, "learning_rate": 2.172881355932203e-07, "loss": 1.4396, "step": 23590 }, { "epoch": 0.59, "grad_norm": 68.0, "learning_rate": 2.1694915254237287e-07, "loss": 1.4858, "step": 23600 }, { "epoch": 0.59, "grad_norm": 66.5, "learning_rate": 2.1661016949152542e-07, "loss": 1.5004, "step": 23610 }, { "epoch": 0.59, "grad_norm": 70.0, "learning_rate": 2.1627118644067795e-07, "loss": 1.5145, "step": 23620 }, { "epoch": 0.59, "grad_norm": 65.0, "learning_rate": 2.159322033898305e-07, "loss": 1.4797, "step": 23630 }, { "epoch": 0.59, "grad_norm": 66.5, "learning_rate": 2.1559322033898303e-07, "loss": 1.4935, "step": 23640 }, { "epoch": 0.59, "grad_norm": 68.5, "learning_rate": 2.1525423728813558e-07, "loss": 1.4311, "step": 23650 }, { "epoch": 0.59, "grad_norm": 66.0, "learning_rate": 2.149152542372881e-07, "loss": 1.5181, "step": 23660 }, { "epoch": 0.59, "grad_norm": 74.0, "learning_rate": 2.1457627118644067e-07, "loss": 1.4418, "step": 23670 }, { "epoch": 0.59, "grad_norm": 66.5, "learning_rate": 2.1423728813559322e-07, "loss": 1.4931, "step": 23680 }, { "epoch": 0.59, "grad_norm": 67.5, "learning_rate": 2.1389830508474575e-07, "loss": 1.4365, "step": 23690 }, { "epoch": 0.59, "grad_norm": 70.0, "learning_rate": 2.1355932203389828e-07, "loss": 1.4576, "step": 23700 }, { "epoch": 0.59, "grad_norm": 66.5, "learning_rate": 2.1322033898305083e-07, "loss": 1.4804, "step": 23710 }, { "epoch": 0.59, "grad_norm": 65.5, "learning_rate": 2.1288135593220338e-07, "loss": 1.4659, "step": 23720 }, { "epoch": 0.59, "grad_norm": 66.0, "learning_rate": 2.1254237288135594e-07, "loss": 1.4558, "step": 23730 }, { "epoch": 0.59, "grad_norm": 63.75, "learning_rate": 2.1220338983050847e-07, "loss": 1.4657, "step": 23740 }, { "epoch": 0.59, "grad_norm": 71.5, "learning_rate": 2.11864406779661e-07, "loss": 1.4803, "step": 23750 }, { "epoch": 0.59, "grad_norm": 66.5, "learning_rate": 2.1152542372881355e-07, "loss": 1.4303, "step": 23760 }, { "epoch": 0.59, "grad_norm": 66.0, "learning_rate": 2.111864406779661e-07, "loss": 1.4866, "step": 23770 }, { "epoch": 0.59, "grad_norm": 63.5, "learning_rate": 2.1084745762711863e-07, "loss": 1.4306, "step": 23780 }, { "epoch": 0.59, "grad_norm": 67.5, "learning_rate": 2.1050847457627118e-07, "loss": 1.4288, "step": 23790 }, { "epoch": 0.59, "grad_norm": 67.0, "learning_rate": 2.101694915254237e-07, "loss": 1.4637, "step": 23800 }, { "epoch": 0.6, "grad_norm": 67.5, "learning_rate": 2.0983050847457626e-07, "loss": 1.515, "step": 23810 }, { "epoch": 0.6, "grad_norm": 70.0, "learning_rate": 2.094915254237288e-07, "loss": 1.4255, "step": 23820 }, { "epoch": 0.6, "grad_norm": 67.5, "learning_rate": 2.0915254237288135e-07, "loss": 1.4935, "step": 23830 }, { "epoch": 0.6, "grad_norm": 68.5, "learning_rate": 2.088135593220339e-07, "loss": 1.5427, "step": 23840 }, { "epoch": 0.6, "grad_norm": 66.5, "learning_rate": 2.0847457627118643e-07, "loss": 1.5009, "step": 23850 }, { "epoch": 0.6, "grad_norm": 69.5, "learning_rate": 2.0813559322033896e-07, "loss": 1.5038, "step": 23860 }, { "epoch": 0.6, "grad_norm": 67.0, "learning_rate": 2.077966101694915e-07, "loss": 1.4524, "step": 23870 }, { "epoch": 0.6, "grad_norm": 69.0, "learning_rate": 2.0745762711864406e-07, "loss": 1.4575, "step": 23880 }, { "epoch": 0.6, "grad_norm": 67.0, "learning_rate": 2.0711864406779662e-07, "loss": 1.4309, "step": 23890 }, { "epoch": 0.6, "grad_norm": 70.0, "learning_rate": 2.0677966101694912e-07, "loss": 1.4585, "step": 23900 }, { "epoch": 0.6, "grad_norm": 64.0, "learning_rate": 2.0644067796610167e-07, "loss": 1.4503, "step": 23910 }, { "epoch": 0.6, "grad_norm": 71.0, "learning_rate": 2.0610169491525423e-07, "loss": 1.4801, "step": 23920 }, { "epoch": 0.6, "grad_norm": 67.5, "learning_rate": 2.0576271186440678e-07, "loss": 1.4773, "step": 23930 }, { "epoch": 0.6, "grad_norm": 66.0, "learning_rate": 2.054237288135593e-07, "loss": 1.4497, "step": 23940 }, { "epoch": 0.6, "grad_norm": 66.0, "learning_rate": 2.0508474576271184e-07, "loss": 1.4953, "step": 23950 }, { "epoch": 0.6, "grad_norm": 65.0, "learning_rate": 2.047457627118644e-07, "loss": 1.5234, "step": 23960 }, { "epoch": 0.6, "grad_norm": 67.0, "learning_rate": 2.0440677966101695e-07, "loss": 1.4482, "step": 23970 }, { "epoch": 0.6, "grad_norm": 68.0, "learning_rate": 2.0406779661016947e-07, "loss": 1.4598, "step": 23980 }, { "epoch": 0.6, "grad_norm": 67.0, "learning_rate": 2.0372881355932203e-07, "loss": 1.4573, "step": 23990 }, { "epoch": 0.6, "grad_norm": 67.0, "learning_rate": 2.0338983050847458e-07, "loss": 1.4849, "step": 24000 }, { "epoch": 0.6, "grad_norm": 68.0, "learning_rate": 2.030508474576271e-07, "loss": 1.4329, "step": 24010 }, { "epoch": 0.6, "grad_norm": 67.0, "learning_rate": 2.0271186440677964e-07, "loss": 1.481, "step": 24020 }, { "epoch": 0.6, "grad_norm": 65.0, "learning_rate": 2.023728813559322e-07, "loss": 1.4037, "step": 24030 }, { "epoch": 0.6, "grad_norm": 68.5, "learning_rate": 2.0203389830508475e-07, "loss": 1.4665, "step": 24040 }, { "epoch": 0.6, "grad_norm": 64.5, "learning_rate": 2.016949152542373e-07, "loss": 1.4291, "step": 24050 }, { "epoch": 0.6, "grad_norm": 68.5, "learning_rate": 2.013559322033898e-07, "loss": 1.4684, "step": 24060 }, { "epoch": 0.6, "grad_norm": 69.5, "learning_rate": 2.0101694915254235e-07, "loss": 1.4836, "step": 24070 }, { "epoch": 0.6, "grad_norm": 65.5, "learning_rate": 2.006779661016949e-07, "loss": 1.4914, "step": 24080 }, { "epoch": 0.6, "grad_norm": 64.0, "learning_rate": 2.0033898305084746e-07, "loss": 1.4268, "step": 24090 }, { "epoch": 0.6, "grad_norm": 66.5, "learning_rate": 2e-07, "loss": 1.5189, "step": 24100 }, { "epoch": 0.6, "grad_norm": 69.5, "learning_rate": 1.9966101694915252e-07, "loss": 1.4294, "step": 24110 }, { "epoch": 0.6, "grad_norm": 66.0, "learning_rate": 1.9932203389830507e-07, "loss": 1.4163, "step": 24120 }, { "epoch": 0.6, "grad_norm": 66.0, "learning_rate": 1.9898305084745763e-07, "loss": 1.4475, "step": 24130 }, { "epoch": 0.6, "grad_norm": 67.0, "learning_rate": 1.9864406779661015e-07, "loss": 1.4687, "step": 24140 }, { "epoch": 0.6, "grad_norm": 65.0, "learning_rate": 1.983050847457627e-07, "loss": 1.4612, "step": 24150 }, { "epoch": 0.6, "grad_norm": 66.5, "learning_rate": 1.9796610169491524e-07, "loss": 1.4849, "step": 24160 }, { "epoch": 0.6, "grad_norm": 65.0, "learning_rate": 1.976271186440678e-07, "loss": 1.461, "step": 24170 }, { "epoch": 0.6, "grad_norm": 66.0, "learning_rate": 1.9728813559322032e-07, "loss": 1.4663, "step": 24180 }, { "epoch": 0.6, "grad_norm": 68.5, "learning_rate": 1.9694915254237287e-07, "loss": 1.4987, "step": 24190 }, { "epoch": 0.6, "grad_norm": 74.0, "learning_rate": 1.9661016949152543e-07, "loss": 1.47, "step": 24200 }, { "epoch": 0.61, "grad_norm": 66.5, "learning_rate": 1.9627118644067795e-07, "loss": 1.4539, "step": 24210 }, { "epoch": 0.61, "grad_norm": 68.5, "learning_rate": 1.9593220338983048e-07, "loss": 1.4931, "step": 24220 }, { "epoch": 0.61, "grad_norm": 70.0, "learning_rate": 1.9559322033898304e-07, "loss": 1.507, "step": 24230 }, { "epoch": 0.61, "grad_norm": 69.0, "learning_rate": 1.952542372881356e-07, "loss": 1.4408, "step": 24240 }, { "epoch": 0.61, "grad_norm": 71.0, "learning_rate": 1.9491525423728814e-07, "loss": 1.4734, "step": 24250 }, { "epoch": 0.61, "grad_norm": 64.0, "learning_rate": 1.9457627118644067e-07, "loss": 1.4702, "step": 24260 }, { "epoch": 0.61, "grad_norm": 68.0, "learning_rate": 1.942372881355932e-07, "loss": 1.4114, "step": 24270 }, { "epoch": 0.61, "grad_norm": 67.0, "learning_rate": 1.9389830508474575e-07, "loss": 1.4777, "step": 24280 }, { "epoch": 0.61, "grad_norm": 65.5, "learning_rate": 1.935593220338983e-07, "loss": 1.4162, "step": 24290 }, { "epoch": 0.61, "grad_norm": 68.0, "learning_rate": 1.9322033898305084e-07, "loss": 1.4881, "step": 24300 }, { "epoch": 0.61, "grad_norm": 65.0, "learning_rate": 1.928813559322034e-07, "loss": 1.4465, "step": 24310 }, { "epoch": 0.61, "grad_norm": 65.5, "learning_rate": 1.9254237288135592e-07, "loss": 1.4648, "step": 24320 }, { "epoch": 0.61, "grad_norm": 64.5, "learning_rate": 1.9220338983050847e-07, "loss": 1.4391, "step": 24330 }, { "epoch": 0.61, "grad_norm": 71.5, "learning_rate": 1.91864406779661e-07, "loss": 1.488, "step": 24340 }, { "epoch": 0.61, "grad_norm": 65.5, "learning_rate": 1.9152542372881355e-07, "loss": 1.4339, "step": 24350 }, { "epoch": 0.61, "grad_norm": 68.5, "learning_rate": 1.911864406779661e-07, "loss": 1.4563, "step": 24360 }, { "epoch": 0.61, "grad_norm": 69.5, "learning_rate": 1.9084745762711864e-07, "loss": 1.4788, "step": 24370 }, { "epoch": 0.61, "grad_norm": 66.0, "learning_rate": 1.9050847457627116e-07, "loss": 1.4948, "step": 24380 }, { "epoch": 0.61, "grad_norm": 66.5, "learning_rate": 1.9016949152542372e-07, "loss": 1.4515, "step": 24390 }, { "epoch": 0.61, "grad_norm": 65.5, "learning_rate": 1.8983050847457627e-07, "loss": 1.4362, "step": 24400 }, { "epoch": 0.61, "grad_norm": 69.0, "learning_rate": 1.8949152542372883e-07, "loss": 1.4519, "step": 24410 }, { "epoch": 0.61, "grad_norm": 67.5, "learning_rate": 1.8915254237288133e-07, "loss": 1.4351, "step": 24420 }, { "epoch": 0.61, "grad_norm": 68.5, "learning_rate": 1.8881355932203388e-07, "loss": 1.4866, "step": 24430 }, { "epoch": 0.61, "grad_norm": 66.0, "learning_rate": 1.8847457627118643e-07, "loss": 1.4731, "step": 24440 }, { "epoch": 0.61, "grad_norm": 66.0, "learning_rate": 1.88135593220339e-07, "loss": 1.4202, "step": 24450 }, { "epoch": 0.61, "grad_norm": 67.0, "learning_rate": 1.8779661016949152e-07, "loss": 1.4834, "step": 24460 }, { "epoch": 0.61, "grad_norm": 69.5, "learning_rate": 1.8745762711864404e-07, "loss": 1.4601, "step": 24470 }, { "epoch": 0.61, "grad_norm": 69.0, "learning_rate": 1.871186440677966e-07, "loss": 1.4693, "step": 24480 }, { "epoch": 0.61, "grad_norm": 65.0, "learning_rate": 1.8677966101694915e-07, "loss": 1.4383, "step": 24490 }, { "epoch": 0.61, "grad_norm": 63.75, "learning_rate": 1.8644067796610168e-07, "loss": 1.4926, "step": 24500 }, { "epoch": 0.61, "grad_norm": 63.0, "learning_rate": 1.8610169491525423e-07, "loss": 1.4355, "step": 24510 }, { "epoch": 0.61, "grad_norm": 64.5, "learning_rate": 1.857627118644068e-07, "loss": 1.474, "step": 24520 }, { "epoch": 0.61, "grad_norm": 67.5, "learning_rate": 1.8542372881355932e-07, "loss": 1.4319, "step": 24530 }, { "epoch": 0.61, "grad_norm": 68.5, "learning_rate": 1.8508474576271184e-07, "loss": 1.469, "step": 24540 }, { "epoch": 0.61, "grad_norm": 66.0, "learning_rate": 1.847457627118644e-07, "loss": 1.4818, "step": 24550 }, { "epoch": 0.61, "grad_norm": 72.5, "learning_rate": 1.8440677966101695e-07, "loss": 1.4986, "step": 24560 }, { "epoch": 0.61, "grad_norm": 64.5, "learning_rate": 1.840677966101695e-07, "loss": 1.4524, "step": 24570 }, { "epoch": 0.61, "grad_norm": 65.0, "learning_rate": 1.83728813559322e-07, "loss": 1.5097, "step": 24580 }, { "epoch": 0.61, "grad_norm": 68.5, "learning_rate": 1.8338983050847456e-07, "loss": 1.4928, "step": 24590 }, { "epoch": 0.61, "grad_norm": 67.5, "learning_rate": 1.8305084745762712e-07, "loss": 1.4628, "step": 24600 }, { "epoch": 0.62, "grad_norm": 62.75, "learning_rate": 1.8271186440677967e-07, "loss": 1.4052, "step": 24610 }, { "epoch": 0.62, "grad_norm": 67.0, "learning_rate": 1.823728813559322e-07, "loss": 1.4308, "step": 24620 }, { "epoch": 0.62, "grad_norm": 70.5, "learning_rate": 1.8203389830508473e-07, "loss": 1.5198, "step": 24630 }, { "epoch": 0.62, "grad_norm": 65.5, "learning_rate": 1.8169491525423728e-07, "loss": 1.5095, "step": 24640 }, { "epoch": 0.62, "grad_norm": 67.0, "learning_rate": 1.8135593220338983e-07, "loss": 1.4858, "step": 24650 }, { "epoch": 0.62, "grad_norm": 69.0, "learning_rate": 1.8101694915254236e-07, "loss": 1.4812, "step": 24660 }, { "epoch": 0.62, "grad_norm": 66.0, "learning_rate": 1.8067796610169492e-07, "loss": 1.4465, "step": 24670 }, { "epoch": 0.62, "grad_norm": 65.5, "learning_rate": 1.8033898305084744e-07, "loss": 1.4647, "step": 24680 }, { "epoch": 0.62, "grad_norm": 66.5, "learning_rate": 1.8e-07, "loss": 1.5012, "step": 24690 }, { "epoch": 0.62, "grad_norm": 68.0, "learning_rate": 1.7966101694915252e-07, "loss": 1.4985, "step": 24700 }, { "epoch": 0.62, "grad_norm": 64.5, "learning_rate": 1.7932203389830508e-07, "loss": 1.4605, "step": 24710 }, { "epoch": 0.62, "grad_norm": 67.0, "learning_rate": 1.7898305084745763e-07, "loss": 1.4714, "step": 24720 }, { "epoch": 0.62, "grad_norm": 68.0, "learning_rate": 1.7864406779661016e-07, "loss": 1.45, "step": 24730 }, { "epoch": 0.62, "grad_norm": 67.0, "learning_rate": 1.783050847457627e-07, "loss": 1.4958, "step": 24740 }, { "epoch": 0.62, "grad_norm": 69.0, "learning_rate": 1.7796610169491524e-07, "loss": 1.4518, "step": 24750 }, { "epoch": 0.62, "grad_norm": 65.5, "learning_rate": 1.776271186440678e-07, "loss": 1.4454, "step": 24760 }, { "epoch": 0.62, "grad_norm": 66.5, "learning_rate": 1.7728813559322035e-07, "loss": 1.4836, "step": 24770 }, { "epoch": 0.62, "grad_norm": 64.5, "learning_rate": 1.7694915254237288e-07, "loss": 1.4561, "step": 24780 }, { "epoch": 0.62, "grad_norm": 66.0, "learning_rate": 1.766101694915254e-07, "loss": 1.4442, "step": 24790 }, { "epoch": 0.62, "grad_norm": 64.5, "learning_rate": 1.7627118644067796e-07, "loss": 1.4462, "step": 24800 }, { "epoch": 0.62, "grad_norm": 66.0, "learning_rate": 1.7593220338983051e-07, "loss": 1.5105, "step": 24810 }, { "epoch": 0.62, "grad_norm": 66.0, "learning_rate": 1.7559322033898304e-07, "loss": 1.4328, "step": 24820 }, { "epoch": 0.62, "grad_norm": 66.0, "learning_rate": 1.752542372881356e-07, "loss": 1.4657, "step": 24830 }, { "epoch": 0.62, "grad_norm": 68.5, "learning_rate": 1.7491525423728812e-07, "loss": 1.4911, "step": 24840 }, { "epoch": 0.62, "grad_norm": 65.0, "learning_rate": 1.7457627118644068e-07, "loss": 1.4936, "step": 24850 }, { "epoch": 0.62, "grad_norm": 66.5, "learning_rate": 1.742372881355932e-07, "loss": 1.4486, "step": 24860 }, { "epoch": 0.62, "grad_norm": 65.0, "learning_rate": 1.7389830508474576e-07, "loss": 1.4143, "step": 24870 }, { "epoch": 0.62, "grad_norm": 67.0, "learning_rate": 1.7355932203389831e-07, "loss": 1.4867, "step": 24880 }, { "epoch": 0.62, "grad_norm": 66.5, "learning_rate": 1.7322033898305084e-07, "loss": 1.4848, "step": 24890 }, { "epoch": 0.62, "grad_norm": 67.0, "learning_rate": 1.7288135593220337e-07, "loss": 1.4092, "step": 24900 }, { "epoch": 0.62, "grad_norm": 66.5, "learning_rate": 1.7254237288135592e-07, "loss": 1.4708, "step": 24910 }, { "epoch": 0.62, "grad_norm": 65.5, "learning_rate": 1.7220338983050848e-07, "loss": 1.4355, "step": 24920 }, { "epoch": 0.62, "grad_norm": 67.0, "learning_rate": 1.7186440677966103e-07, "loss": 1.4986, "step": 24930 }, { "epoch": 0.62, "grad_norm": 68.5, "learning_rate": 1.7152542372881353e-07, "loss": 1.5264, "step": 24940 }, { "epoch": 0.62, "grad_norm": 71.5, "learning_rate": 1.711864406779661e-07, "loss": 1.5117, "step": 24950 }, { "epoch": 0.62, "grad_norm": 68.5, "learning_rate": 1.7084745762711864e-07, "loss": 1.4339, "step": 24960 }, { "epoch": 0.62, "grad_norm": 67.0, "learning_rate": 1.705084745762712e-07, "loss": 1.4962, "step": 24970 }, { "epoch": 0.62, "grad_norm": 66.5, "learning_rate": 1.7016949152542372e-07, "loss": 1.4068, "step": 24980 }, { "epoch": 0.62, "grad_norm": 67.5, "learning_rate": 1.6983050847457625e-07, "loss": 1.4685, "step": 24990 }, { "epoch": 0.62, "grad_norm": 67.0, "learning_rate": 1.694915254237288e-07, "loss": 1.4548, "step": 25000 }, { "epoch": 0.62, "eval_loss": 1.466233730316162, "eval_runtime": 68.2791, "eval_samples_per_second": 14.646, "eval_steps_per_second": 14.646, "step": 25000 }, { "epoch": 0.63, "grad_norm": 70.5, "learning_rate": 1.6915254237288136e-07, "loss": 1.4549, "step": 25010 }, { "epoch": 0.63, "grad_norm": 66.0, "learning_rate": 1.688135593220339e-07, "loss": 1.482, "step": 25020 }, { "epoch": 0.63, "grad_norm": 69.0, "learning_rate": 1.6847457627118644e-07, "loss": 1.4466, "step": 25030 }, { "epoch": 0.63, "grad_norm": 68.5, "learning_rate": 1.68135593220339e-07, "loss": 1.448, "step": 25040 }, { "epoch": 0.63, "grad_norm": 69.5, "learning_rate": 1.677966101694915e-07, "loss": 1.5145, "step": 25050 }, { "epoch": 0.63, "grad_norm": 63.0, "learning_rate": 1.6745762711864405e-07, "loss": 1.4911, "step": 25060 }, { "epoch": 0.63, "grad_norm": 70.5, "learning_rate": 1.671186440677966e-07, "loss": 1.4467, "step": 25070 }, { "epoch": 0.63, "grad_norm": 67.5, "learning_rate": 1.6677966101694916e-07, "loss": 1.4316, "step": 25080 }, { "epoch": 0.63, "grad_norm": 66.0, "learning_rate": 1.664406779661017e-07, "loss": 1.4829, "step": 25090 }, { "epoch": 0.63, "grad_norm": 73.5, "learning_rate": 1.6610169491525421e-07, "loss": 1.4723, "step": 25100 }, { "epoch": 0.63, "grad_norm": 66.0, "learning_rate": 1.6576271186440677e-07, "loss": 1.4702, "step": 25110 }, { "epoch": 0.63, "grad_norm": 65.0, "learning_rate": 1.6542372881355932e-07, "loss": 1.4329, "step": 25120 }, { "epoch": 0.63, "grad_norm": 65.5, "learning_rate": 1.6508474576271188e-07, "loss": 1.481, "step": 25130 }, { "epoch": 0.63, "grad_norm": 65.0, "learning_rate": 1.647457627118644e-07, "loss": 1.4415, "step": 25140 }, { "epoch": 0.63, "grad_norm": 69.0, "learning_rate": 1.6440677966101693e-07, "loss": 1.4572, "step": 25150 }, { "epoch": 0.63, "grad_norm": 67.0, "learning_rate": 1.6406779661016949e-07, "loss": 1.4667, "step": 25160 }, { "epoch": 0.63, "grad_norm": 68.5, "learning_rate": 1.6372881355932201e-07, "loss": 1.4629, "step": 25170 }, { "epoch": 0.63, "grad_norm": 65.5, "learning_rate": 1.6338983050847457e-07, "loss": 1.4532, "step": 25180 }, { "epoch": 0.63, "grad_norm": 67.0, "learning_rate": 1.6305084745762712e-07, "loss": 1.4474, "step": 25190 }, { "epoch": 0.63, "grad_norm": 67.5, "learning_rate": 1.6271186440677965e-07, "loss": 1.4358, "step": 25200 }, { "epoch": 0.63, "grad_norm": 65.5, "learning_rate": 1.6237288135593218e-07, "loss": 1.4745, "step": 25210 }, { "epoch": 0.63, "grad_norm": 65.0, "learning_rate": 1.6203389830508473e-07, "loss": 1.4517, "step": 25220 }, { "epoch": 0.63, "grad_norm": 65.0, "learning_rate": 1.6169491525423729e-07, "loss": 1.4568, "step": 25230 }, { "epoch": 0.63, "grad_norm": 68.5, "learning_rate": 1.6135593220338984e-07, "loss": 1.4866, "step": 25240 }, { "epoch": 0.63, "grad_norm": 65.0, "learning_rate": 1.6101694915254234e-07, "loss": 1.426, "step": 25250 }, { "epoch": 0.63, "grad_norm": 65.5, "learning_rate": 1.606779661016949e-07, "loss": 1.4852, "step": 25260 }, { "epoch": 0.63, "grad_norm": 67.0, "learning_rate": 1.6033898305084745e-07, "loss": 1.4411, "step": 25270 }, { "epoch": 0.63, "grad_norm": 68.5, "learning_rate": 1.6e-07, "loss": 1.4613, "step": 25280 }, { "epoch": 0.63, "grad_norm": 68.0, "learning_rate": 1.5966101694915253e-07, "loss": 1.4734, "step": 25290 }, { "epoch": 0.63, "grad_norm": 67.5, "learning_rate": 1.5932203389830506e-07, "loss": 1.5376, "step": 25300 }, { "epoch": 0.63, "grad_norm": 65.0, "learning_rate": 1.589830508474576e-07, "loss": 1.4272, "step": 25310 }, { "epoch": 0.63, "grad_norm": 67.5, "learning_rate": 1.5864406779661017e-07, "loss": 1.4695, "step": 25320 }, { "epoch": 0.63, "grad_norm": 67.0, "learning_rate": 1.583050847457627e-07, "loss": 1.4553, "step": 25330 }, { "epoch": 0.63, "grad_norm": 65.0, "learning_rate": 1.5796610169491525e-07, "loss": 1.4733, "step": 25340 }, { "epoch": 0.63, "grad_norm": 65.5, "learning_rate": 1.576271186440678e-07, "loss": 1.435, "step": 25350 }, { "epoch": 0.63, "grad_norm": 64.5, "learning_rate": 1.5728813559322033e-07, "loss": 1.4751, "step": 25360 }, { "epoch": 0.63, "grad_norm": 66.5, "learning_rate": 1.5694915254237286e-07, "loss": 1.4824, "step": 25370 }, { "epoch": 0.63, "grad_norm": 65.0, "learning_rate": 1.566101694915254e-07, "loss": 1.4431, "step": 25380 }, { "epoch": 0.63, "grad_norm": 73.5, "learning_rate": 1.5627118644067797e-07, "loss": 1.516, "step": 25390 }, { "epoch": 0.64, "grad_norm": 65.5, "learning_rate": 1.5593220338983052e-07, "loss": 1.4409, "step": 25400 }, { "epoch": 0.64, "grad_norm": 67.0, "learning_rate": 1.5559322033898302e-07, "loss": 1.4274, "step": 25410 }, { "epoch": 0.64, "grad_norm": 66.5, "learning_rate": 1.5525423728813558e-07, "loss": 1.48, "step": 25420 }, { "epoch": 0.64, "grad_norm": 65.0, "learning_rate": 1.5491525423728813e-07, "loss": 1.4892, "step": 25430 }, { "epoch": 0.64, "grad_norm": 66.0, "learning_rate": 1.5457627118644068e-07, "loss": 1.4686, "step": 25440 }, { "epoch": 0.64, "grad_norm": 68.0, "learning_rate": 1.542372881355932e-07, "loss": 1.4634, "step": 25450 }, { "epoch": 0.64, "grad_norm": 68.0, "learning_rate": 1.5389830508474574e-07, "loss": 1.4937, "step": 25460 }, { "epoch": 0.64, "grad_norm": 63.75, "learning_rate": 1.535593220338983e-07, "loss": 1.4743, "step": 25470 }, { "epoch": 0.64, "grad_norm": 67.0, "learning_rate": 1.5322033898305085e-07, "loss": 1.4774, "step": 25480 }, { "epoch": 0.64, "grad_norm": 65.0, "learning_rate": 1.5288135593220338e-07, "loss": 1.4126, "step": 25490 }, { "epoch": 0.64, "grad_norm": 64.5, "learning_rate": 1.5254237288135593e-07, "loss": 1.4223, "step": 25500 }, { "epoch": 0.64, "grad_norm": 66.5, "learning_rate": 1.5220338983050846e-07, "loss": 1.4963, "step": 25510 }, { "epoch": 0.64, "grad_norm": 66.5, "learning_rate": 1.51864406779661e-07, "loss": 1.4755, "step": 25520 }, { "epoch": 0.64, "grad_norm": 66.0, "learning_rate": 1.5152542372881354e-07, "loss": 1.4751, "step": 25530 }, { "epoch": 0.64, "grad_norm": 70.0, "learning_rate": 1.511864406779661e-07, "loss": 1.5208, "step": 25540 }, { "epoch": 0.64, "grad_norm": 64.0, "learning_rate": 1.5084745762711865e-07, "loss": 1.4453, "step": 25550 }, { "epoch": 0.64, "grad_norm": 65.0, "learning_rate": 1.5050847457627118e-07, "loss": 1.4952, "step": 25560 }, { "epoch": 0.64, "grad_norm": 69.5, "learning_rate": 1.501694915254237e-07, "loss": 1.4834, "step": 25570 }, { "epoch": 0.64, "grad_norm": 70.0, "learning_rate": 1.4983050847457626e-07, "loss": 1.4714, "step": 25580 }, { "epoch": 0.64, "grad_norm": 69.5, "learning_rate": 1.494915254237288e-07, "loss": 1.5123, "step": 25590 }, { "epoch": 0.64, "grad_norm": 66.0, "learning_rate": 1.4915254237288137e-07, "loss": 1.4964, "step": 25600 }, { "epoch": 0.64, "grad_norm": 67.5, "learning_rate": 1.488135593220339e-07, "loss": 1.4979, "step": 25610 }, { "epoch": 0.64, "grad_norm": 68.0, "learning_rate": 1.4847457627118642e-07, "loss": 1.4971, "step": 25620 }, { "epoch": 0.64, "grad_norm": 65.5, "learning_rate": 1.4813559322033897e-07, "loss": 1.5237, "step": 25630 }, { "epoch": 0.64, "grad_norm": 65.0, "learning_rate": 1.4779661016949153e-07, "loss": 1.466, "step": 25640 }, { "epoch": 0.64, "grad_norm": 67.0, "learning_rate": 1.4745762711864406e-07, "loss": 1.4765, "step": 25650 }, { "epoch": 0.64, "grad_norm": 62.25, "learning_rate": 1.471186440677966e-07, "loss": 1.417, "step": 25660 }, { "epoch": 0.64, "grad_norm": 69.5, "learning_rate": 1.4677966101694914e-07, "loss": 1.4432, "step": 25670 }, { "epoch": 0.64, "grad_norm": 67.0, "learning_rate": 1.464406779661017e-07, "loss": 1.4499, "step": 25680 }, { "epoch": 0.64, "grad_norm": 73.0, "learning_rate": 1.4610169491525422e-07, "loss": 1.4715, "step": 25690 }, { "epoch": 0.64, "grad_norm": 66.0, "learning_rate": 1.4576271186440677e-07, "loss": 1.4751, "step": 25700 }, { "epoch": 0.64, "grad_norm": 67.0, "learning_rate": 1.4542372881355933e-07, "loss": 1.4418, "step": 25710 }, { "epoch": 0.64, "grad_norm": 65.5, "learning_rate": 1.4508474576271186e-07, "loss": 1.4736, "step": 25720 }, { "epoch": 0.64, "grad_norm": 67.0, "learning_rate": 1.4474576271186438e-07, "loss": 1.4995, "step": 25730 }, { "epoch": 0.64, "grad_norm": 71.5, "learning_rate": 1.4440677966101694e-07, "loss": 1.4869, "step": 25740 }, { "epoch": 0.64, "grad_norm": 68.0, "learning_rate": 1.440677966101695e-07, "loss": 1.5158, "step": 25750 }, { "epoch": 0.64, "grad_norm": 69.0, "learning_rate": 1.4372881355932205e-07, "loss": 1.5294, "step": 25760 }, { "epoch": 0.64, "grad_norm": 63.75, "learning_rate": 1.4338983050847455e-07, "loss": 1.4689, "step": 25770 }, { "epoch": 0.64, "grad_norm": 66.0, "learning_rate": 1.430508474576271e-07, "loss": 1.4569, "step": 25780 }, { "epoch": 0.64, "grad_norm": 66.5, "learning_rate": 1.4271186440677966e-07, "loss": 1.4544, "step": 25790 }, { "epoch": 0.65, "grad_norm": 66.5, "learning_rate": 1.423728813559322e-07, "loss": 1.4838, "step": 25800 }, { "epoch": 0.65, "grad_norm": 66.5, "learning_rate": 1.4203389830508474e-07, "loss": 1.4641, "step": 25810 }, { "epoch": 0.65, "grad_norm": 66.0, "learning_rate": 1.4169491525423727e-07, "loss": 1.4856, "step": 25820 }, { "epoch": 0.65, "grad_norm": 63.75, "learning_rate": 1.4135593220338982e-07, "loss": 1.4885, "step": 25830 }, { "epoch": 0.65, "grad_norm": 66.0, "learning_rate": 1.4101694915254237e-07, "loss": 1.4051, "step": 25840 }, { "epoch": 0.65, "grad_norm": 67.0, "learning_rate": 1.406779661016949e-07, "loss": 1.4431, "step": 25850 }, { "epoch": 0.65, "grad_norm": 65.5, "learning_rate": 1.4033898305084746e-07, "loss": 1.4973, "step": 25860 }, { "epoch": 0.65, "grad_norm": 66.0, "learning_rate": 1.4e-07, "loss": 1.4481, "step": 25870 }, { "epoch": 0.65, "grad_norm": 66.5, "learning_rate": 1.3966101694915254e-07, "loss": 1.473, "step": 25880 }, { "epoch": 0.65, "grad_norm": 65.5, "learning_rate": 1.3932203389830506e-07, "loss": 1.4835, "step": 25890 }, { "epoch": 0.65, "grad_norm": 65.0, "learning_rate": 1.3898305084745762e-07, "loss": 1.4437, "step": 25900 }, { "epoch": 0.65, "grad_norm": 69.5, "learning_rate": 1.3864406779661017e-07, "loss": 1.5014, "step": 25910 }, { "epoch": 0.65, "grad_norm": 64.5, "learning_rate": 1.3830508474576273e-07, "loss": 1.413, "step": 25920 }, { "epoch": 0.65, "grad_norm": 71.5, "learning_rate": 1.3796610169491523e-07, "loss": 1.5194, "step": 25930 }, { "epoch": 0.65, "grad_norm": 66.5, "learning_rate": 1.3762711864406778e-07, "loss": 1.439, "step": 25940 }, { "epoch": 0.65, "grad_norm": 72.0, "learning_rate": 1.3728813559322034e-07, "loss": 1.5017, "step": 25950 }, { "epoch": 0.65, "grad_norm": 65.0, "learning_rate": 1.369491525423729e-07, "loss": 1.4643, "step": 25960 }, { "epoch": 0.65, "grad_norm": 65.0, "learning_rate": 1.3661016949152542e-07, "loss": 1.4582, "step": 25970 }, { "epoch": 0.65, "grad_norm": 66.0, "learning_rate": 1.3627118644067795e-07, "loss": 1.4518, "step": 25980 }, { "epoch": 0.65, "grad_norm": 66.0, "learning_rate": 1.359322033898305e-07, "loss": 1.4446, "step": 25990 }, { "epoch": 0.65, "grad_norm": 67.0, "learning_rate": 1.3559322033898305e-07, "loss": 1.5096, "step": 26000 }, { "epoch": 0.65, "grad_norm": 66.0, "learning_rate": 1.3525423728813558e-07, "loss": 1.385, "step": 26010 }, { "epoch": 0.65, "grad_norm": 66.5, "learning_rate": 1.3491525423728814e-07, "loss": 1.4715, "step": 26020 }, { "epoch": 0.65, "grad_norm": 65.5, "learning_rate": 1.3457627118644066e-07, "loss": 1.4763, "step": 26030 }, { "epoch": 0.65, "grad_norm": 66.0, "learning_rate": 1.3423728813559322e-07, "loss": 1.4966, "step": 26040 }, { "epoch": 0.65, "grad_norm": 69.0, "learning_rate": 1.3389830508474575e-07, "loss": 1.4786, "step": 26050 }, { "epoch": 0.65, "grad_norm": 66.5, "learning_rate": 1.335593220338983e-07, "loss": 1.4363, "step": 26060 }, { "epoch": 0.65, "grad_norm": 65.5, "learning_rate": 1.3322033898305085e-07, "loss": 1.4564, "step": 26070 }, { "epoch": 0.65, "grad_norm": 66.0, "learning_rate": 1.3288135593220338e-07, "loss": 1.4775, "step": 26080 }, { "epoch": 0.65, "grad_norm": 68.5, "learning_rate": 1.325423728813559e-07, "loss": 1.5025, "step": 26090 }, { "epoch": 0.65, "grad_norm": 67.5, "learning_rate": 1.3220338983050846e-07, "loss": 1.4541, "step": 26100 }, { "epoch": 0.65, "grad_norm": 68.5, "learning_rate": 1.3186440677966102e-07, "loss": 1.4984, "step": 26110 }, { "epoch": 0.65, "grad_norm": 63.5, "learning_rate": 1.3152542372881357e-07, "loss": 1.4578, "step": 26120 }, { "epoch": 0.65, "grad_norm": 68.5, "learning_rate": 1.311864406779661e-07, "loss": 1.4594, "step": 26130 }, { "epoch": 0.65, "grad_norm": 68.0, "learning_rate": 1.3084745762711863e-07, "loss": 1.4632, "step": 26140 }, { "epoch": 0.65, "grad_norm": 67.0, "learning_rate": 1.3050847457627118e-07, "loss": 1.4885, "step": 26150 }, { "epoch": 0.65, "grad_norm": 68.0, "learning_rate": 1.3016949152542374e-07, "loss": 1.4458, "step": 26160 }, { "epoch": 0.65, "grad_norm": 66.5, "learning_rate": 1.2983050847457626e-07, "loss": 1.4944, "step": 26170 }, { "epoch": 0.65, "grad_norm": 67.5, "learning_rate": 1.2949152542372882e-07, "loss": 1.4712, "step": 26180 }, { "epoch": 0.65, "grad_norm": 66.0, "learning_rate": 1.2915254237288135e-07, "loss": 1.4466, "step": 26190 }, { "epoch": 0.66, "grad_norm": 65.0, "learning_rate": 1.288135593220339e-07, "loss": 1.474, "step": 26200 }, { "epoch": 0.66, "grad_norm": 64.5, "learning_rate": 1.2847457627118643e-07, "loss": 1.4741, "step": 26210 }, { "epoch": 0.66, "grad_norm": 68.5, "learning_rate": 1.2813559322033898e-07, "loss": 1.4871, "step": 26220 }, { "epoch": 0.66, "grad_norm": 65.5, "learning_rate": 1.2779661016949154e-07, "loss": 1.4677, "step": 26230 }, { "epoch": 0.66, "grad_norm": 68.5, "learning_rate": 1.2745762711864406e-07, "loss": 1.4899, "step": 26240 }, { "epoch": 0.66, "grad_norm": 72.0, "learning_rate": 1.271186440677966e-07, "loss": 1.5358, "step": 26250 }, { "epoch": 0.66, "grad_norm": 65.5, "learning_rate": 1.2677966101694914e-07, "loss": 1.4496, "step": 26260 }, { "epoch": 0.66, "grad_norm": 69.0, "learning_rate": 1.264406779661017e-07, "loss": 1.5128, "step": 26270 }, { "epoch": 0.66, "grad_norm": 63.5, "learning_rate": 1.2610169491525425e-07, "loss": 1.4466, "step": 26280 }, { "epoch": 0.66, "grad_norm": 66.0, "learning_rate": 1.2576271186440675e-07, "loss": 1.4976, "step": 26290 }, { "epoch": 0.66, "grad_norm": 69.0, "learning_rate": 1.254237288135593e-07, "loss": 1.4673, "step": 26300 }, { "epoch": 0.66, "grad_norm": 63.75, "learning_rate": 1.2508474576271186e-07, "loss": 1.4478, "step": 26310 }, { "epoch": 0.66, "grad_norm": 66.0, "learning_rate": 1.2474576271186442e-07, "loss": 1.4564, "step": 26320 }, { "epoch": 0.66, "grad_norm": 67.5, "learning_rate": 1.2440677966101694e-07, "loss": 1.4362, "step": 26330 }, { "epoch": 0.66, "grad_norm": 68.5, "learning_rate": 1.240677966101695e-07, "loss": 1.4838, "step": 26340 }, { "epoch": 0.66, "grad_norm": 67.0, "learning_rate": 1.2372881355932203e-07, "loss": 1.4604, "step": 26350 }, { "epoch": 0.66, "grad_norm": 66.0, "learning_rate": 1.2338983050847458e-07, "loss": 1.3902, "step": 26360 }, { "epoch": 0.66, "grad_norm": 64.0, "learning_rate": 1.230508474576271e-07, "loss": 1.4454, "step": 26370 }, { "epoch": 0.66, "grad_norm": 70.0, "learning_rate": 1.2271186440677966e-07, "loss": 1.4411, "step": 26380 }, { "epoch": 0.66, "grad_norm": 67.5, "learning_rate": 1.223728813559322e-07, "loss": 1.4241, "step": 26390 }, { "epoch": 0.66, "grad_norm": 67.5, "learning_rate": 1.2203389830508474e-07, "loss": 1.4765, "step": 26400 }, { "epoch": 0.66, "grad_norm": 64.0, "learning_rate": 1.2169491525423727e-07, "loss": 1.4272, "step": 26410 }, { "epoch": 0.66, "grad_norm": 67.0, "learning_rate": 1.2135593220338983e-07, "loss": 1.487, "step": 26420 }, { "epoch": 0.66, "grad_norm": 68.0, "learning_rate": 1.2101694915254235e-07, "loss": 1.5069, "step": 26430 }, { "epoch": 0.66, "grad_norm": 66.5, "learning_rate": 1.206779661016949e-07, "loss": 1.4431, "step": 26440 }, { "epoch": 0.66, "grad_norm": 67.0, "learning_rate": 1.2033898305084746e-07, "loss": 1.4667, "step": 26450 }, { "epoch": 0.66, "grad_norm": 67.5, "learning_rate": 1.2e-07, "loss": 1.4413, "step": 26460 }, { "epoch": 0.66, "grad_norm": 64.0, "learning_rate": 1.1966101694915254e-07, "loss": 1.442, "step": 26470 }, { "epoch": 0.66, "grad_norm": 67.0, "learning_rate": 1.193220338983051e-07, "loss": 1.5008, "step": 26480 }, { "epoch": 0.66, "grad_norm": 67.0, "learning_rate": 1.1898305084745763e-07, "loss": 1.4875, "step": 26490 }, { "epoch": 0.66, "grad_norm": 73.0, "learning_rate": 1.1864406779661017e-07, "loss": 1.5183, "step": 26500 }, { "epoch": 0.66, "grad_norm": 66.5, "learning_rate": 1.1830508474576271e-07, "loss": 1.4863, "step": 26510 }, { "epoch": 0.66, "grad_norm": 65.0, "learning_rate": 1.1796610169491525e-07, "loss": 1.456, "step": 26520 }, { "epoch": 0.66, "grad_norm": 69.5, "learning_rate": 1.176271186440678e-07, "loss": 1.5094, "step": 26530 }, { "epoch": 0.66, "grad_norm": 71.5, "learning_rate": 1.1728813559322033e-07, "loss": 1.4483, "step": 26540 }, { "epoch": 0.66, "grad_norm": 67.5, "learning_rate": 1.1694915254237288e-07, "loss": 1.4301, "step": 26550 }, { "epoch": 0.66, "grad_norm": 71.5, "learning_rate": 1.1661016949152541e-07, "loss": 1.5129, "step": 26560 }, { "epoch": 0.66, "grad_norm": 63.75, "learning_rate": 1.1627118644067797e-07, "loss": 1.3979, "step": 26570 }, { "epoch": 0.66, "grad_norm": 63.0, "learning_rate": 1.1593220338983051e-07, "loss": 1.472, "step": 26580 }, { "epoch": 0.66, "grad_norm": 66.5, "learning_rate": 1.1559322033898305e-07, "loss": 1.4346, "step": 26590 }, { "epoch": 0.67, "grad_norm": 68.0, "learning_rate": 1.1525423728813559e-07, "loss": 1.4705, "step": 26600 }, { "epoch": 0.67, "grad_norm": 68.0, "learning_rate": 1.1491525423728814e-07, "loss": 1.4424, "step": 26610 }, { "epoch": 0.67, "grad_norm": 67.0, "learning_rate": 1.1457627118644067e-07, "loss": 1.4568, "step": 26620 }, { "epoch": 0.67, "grad_norm": 66.0, "learning_rate": 1.1423728813559322e-07, "loss": 1.4427, "step": 26630 }, { "epoch": 0.67, "grad_norm": 63.0, "learning_rate": 1.1389830508474575e-07, "loss": 1.4333, "step": 26640 }, { "epoch": 0.67, "grad_norm": 65.5, "learning_rate": 1.135593220338983e-07, "loss": 1.465, "step": 26650 }, { "epoch": 0.67, "grad_norm": 64.0, "learning_rate": 1.1322033898305085e-07, "loss": 1.4452, "step": 26660 }, { "epoch": 0.67, "grad_norm": 62.75, "learning_rate": 1.1288135593220339e-07, "loss": 1.4542, "step": 26670 }, { "epoch": 0.67, "grad_norm": 65.5, "learning_rate": 1.1254237288135593e-07, "loss": 1.431, "step": 26680 }, { "epoch": 0.67, "grad_norm": 65.5, "learning_rate": 1.1220338983050846e-07, "loss": 1.4976, "step": 26690 }, { "epoch": 0.67, "grad_norm": 66.5, "learning_rate": 1.1186440677966101e-07, "loss": 1.4866, "step": 26700 }, { "epoch": 0.67, "grad_norm": 67.0, "learning_rate": 1.1152542372881356e-07, "loss": 1.451, "step": 26710 }, { "epoch": 0.67, "grad_norm": 67.0, "learning_rate": 1.1118644067796609e-07, "loss": 1.4699, "step": 26720 }, { "epoch": 0.67, "grad_norm": 68.0, "learning_rate": 1.1084745762711865e-07, "loss": 1.4611, "step": 26730 }, { "epoch": 0.67, "grad_norm": 69.0, "learning_rate": 1.1050847457627119e-07, "loss": 1.4227, "step": 26740 }, { "epoch": 0.67, "grad_norm": 67.5, "learning_rate": 1.1016949152542372e-07, "loss": 1.4797, "step": 26750 }, { "epoch": 0.67, "grad_norm": 64.5, "learning_rate": 1.0983050847457627e-07, "loss": 1.4143, "step": 26760 }, { "epoch": 0.67, "grad_norm": 66.5, "learning_rate": 1.094915254237288e-07, "loss": 1.4982, "step": 26770 }, { "epoch": 0.67, "grad_norm": 64.0, "learning_rate": 1.0915254237288135e-07, "loss": 1.3862, "step": 26780 }, { "epoch": 0.67, "grad_norm": 69.0, "learning_rate": 1.0881355932203389e-07, "loss": 1.4549, "step": 26790 }, { "epoch": 0.67, "grad_norm": 66.5, "learning_rate": 1.0847457627118643e-07, "loss": 1.427, "step": 26800 }, { "epoch": 0.67, "grad_norm": 66.0, "learning_rate": 1.0813559322033897e-07, "loss": 1.4717, "step": 26810 }, { "epoch": 0.67, "grad_norm": 64.5, "learning_rate": 1.0779661016949151e-07, "loss": 1.4578, "step": 26820 }, { "epoch": 0.67, "grad_norm": 64.5, "learning_rate": 1.0745762711864406e-07, "loss": 1.4316, "step": 26830 }, { "epoch": 0.67, "grad_norm": 67.5, "learning_rate": 1.0711864406779661e-07, "loss": 1.4959, "step": 26840 }, { "epoch": 0.67, "grad_norm": 69.0, "learning_rate": 1.0677966101694914e-07, "loss": 1.4684, "step": 26850 }, { "epoch": 0.67, "grad_norm": 68.5, "learning_rate": 1.0644067796610169e-07, "loss": 1.4891, "step": 26860 }, { "epoch": 0.67, "grad_norm": 63.25, "learning_rate": 1.0610169491525423e-07, "loss": 1.4411, "step": 26870 }, { "epoch": 0.67, "grad_norm": 67.5, "learning_rate": 1.0576271186440677e-07, "loss": 1.4614, "step": 26880 }, { "epoch": 0.67, "grad_norm": 69.5, "learning_rate": 1.0542372881355931e-07, "loss": 1.4461, "step": 26890 }, { "epoch": 0.67, "grad_norm": 65.0, "learning_rate": 1.0508474576271186e-07, "loss": 1.4477, "step": 26900 }, { "epoch": 0.67, "grad_norm": 68.5, "learning_rate": 1.047457627118644e-07, "loss": 1.4497, "step": 26910 }, { "epoch": 0.67, "grad_norm": 66.0, "learning_rate": 1.0440677966101695e-07, "loss": 1.4276, "step": 26920 }, { "epoch": 0.67, "grad_norm": 71.5, "learning_rate": 1.0406779661016948e-07, "loss": 1.5078, "step": 26930 }, { "epoch": 0.67, "grad_norm": 66.5, "learning_rate": 1.0372881355932203e-07, "loss": 1.4674, "step": 26940 }, { "epoch": 0.67, "grad_norm": 67.0, "learning_rate": 1.0338983050847456e-07, "loss": 1.4352, "step": 26950 }, { "epoch": 0.67, "grad_norm": 70.5, "learning_rate": 1.0305084745762711e-07, "loss": 1.4248, "step": 26960 }, { "epoch": 0.67, "grad_norm": 65.5, "learning_rate": 1.0271186440677965e-07, "loss": 1.4597, "step": 26970 }, { "epoch": 0.67, "grad_norm": 65.5, "learning_rate": 1.023728813559322e-07, "loss": 1.4352, "step": 26980 }, { "epoch": 0.67, "grad_norm": 67.5, "learning_rate": 1.0203389830508474e-07, "loss": 1.4893, "step": 26990 }, { "epoch": 0.68, "grad_norm": 65.0, "learning_rate": 1.0169491525423729e-07, "loss": 1.4631, "step": 27000 }, { "epoch": 0.68, "grad_norm": 65.5, "learning_rate": 1.0135593220338982e-07, "loss": 1.4241, "step": 27010 }, { "epoch": 0.68, "grad_norm": 68.0, "learning_rate": 1.0101694915254237e-07, "loss": 1.4746, "step": 27020 }, { "epoch": 0.68, "grad_norm": 70.5, "learning_rate": 1.006779661016949e-07, "loss": 1.4814, "step": 27030 }, { "epoch": 0.68, "grad_norm": 69.5, "learning_rate": 1.0033898305084745e-07, "loss": 1.4733, "step": 27040 }, { "epoch": 0.68, "grad_norm": 66.0, "learning_rate": 1e-07, "loss": 1.414, "step": 27050 }, { "epoch": 0.68, "grad_norm": 66.5, "learning_rate": 9.966101694915254e-08, "loss": 1.4638, "step": 27060 }, { "epoch": 0.68, "grad_norm": 65.5, "learning_rate": 9.932203389830508e-08, "loss": 1.4267, "step": 27070 }, { "epoch": 0.68, "grad_norm": 67.0, "learning_rate": 9.898305084745762e-08, "loss": 1.4608, "step": 27080 }, { "epoch": 0.68, "grad_norm": 66.0, "learning_rate": 9.864406779661016e-08, "loss": 1.4722, "step": 27090 }, { "epoch": 0.68, "grad_norm": 71.5, "learning_rate": 9.830508474576271e-08, "loss": 1.4615, "step": 27100 }, { "epoch": 0.68, "grad_norm": 66.0, "learning_rate": 9.796610169491524e-08, "loss": 1.4675, "step": 27110 }, { "epoch": 0.68, "grad_norm": 68.5, "learning_rate": 9.76271186440678e-08, "loss": 1.4645, "step": 27120 }, { "epoch": 0.68, "grad_norm": 65.0, "learning_rate": 9.728813559322034e-08, "loss": 1.4702, "step": 27130 }, { "epoch": 0.68, "grad_norm": 69.0, "learning_rate": 9.694915254237288e-08, "loss": 1.4844, "step": 27140 }, { "epoch": 0.68, "grad_norm": 68.5, "learning_rate": 9.661016949152542e-08, "loss": 1.4466, "step": 27150 }, { "epoch": 0.68, "grad_norm": 69.0, "learning_rate": 9.627118644067796e-08, "loss": 1.4445, "step": 27160 }, { "epoch": 0.68, "grad_norm": 66.5, "learning_rate": 9.59322033898305e-08, "loss": 1.4458, "step": 27170 }, { "epoch": 0.68, "grad_norm": 66.5, "learning_rate": 9.559322033898305e-08, "loss": 1.4263, "step": 27180 }, { "epoch": 0.68, "grad_norm": 65.5, "learning_rate": 9.525423728813558e-08, "loss": 1.4717, "step": 27190 }, { "epoch": 0.68, "grad_norm": 69.0, "learning_rate": 9.491525423728814e-08, "loss": 1.4649, "step": 27200 }, { "epoch": 0.68, "grad_norm": 65.5, "learning_rate": 9.457627118644066e-08, "loss": 1.4142, "step": 27210 }, { "epoch": 0.68, "grad_norm": 65.5, "learning_rate": 9.423728813559322e-08, "loss": 1.4443, "step": 27220 }, { "epoch": 0.68, "grad_norm": 65.5, "learning_rate": 9.389830508474576e-08, "loss": 1.4432, "step": 27230 }, { "epoch": 0.68, "grad_norm": 67.0, "learning_rate": 9.35593220338983e-08, "loss": 1.4888, "step": 27240 }, { "epoch": 0.68, "grad_norm": 70.0, "learning_rate": 9.322033898305084e-08, "loss": 1.4675, "step": 27250 }, { "epoch": 0.68, "grad_norm": 68.0, "learning_rate": 9.28813559322034e-08, "loss": 1.4734, "step": 27260 }, { "epoch": 0.68, "grad_norm": 69.5, "learning_rate": 9.254237288135592e-08, "loss": 1.4639, "step": 27270 }, { "epoch": 0.68, "grad_norm": 64.5, "learning_rate": 9.220338983050848e-08, "loss": 1.4594, "step": 27280 }, { "epoch": 0.68, "grad_norm": 65.5, "learning_rate": 9.1864406779661e-08, "loss": 1.4698, "step": 27290 }, { "epoch": 0.68, "grad_norm": 67.0, "learning_rate": 9.152542372881356e-08, "loss": 1.5159, "step": 27300 }, { "epoch": 0.68, "grad_norm": 66.5, "learning_rate": 9.11864406779661e-08, "loss": 1.4563, "step": 27310 }, { "epoch": 0.68, "grad_norm": 69.0, "learning_rate": 9.084745762711864e-08, "loss": 1.4545, "step": 27320 }, { "epoch": 0.68, "grad_norm": 64.5, "learning_rate": 9.050847457627118e-08, "loss": 1.4652, "step": 27330 }, { "epoch": 0.68, "grad_norm": 67.5, "learning_rate": 9.016949152542372e-08, "loss": 1.4417, "step": 27340 }, { "epoch": 0.68, "grad_norm": 69.5, "learning_rate": 8.983050847457626e-08, "loss": 1.456, "step": 27350 }, { "epoch": 0.68, "grad_norm": 67.5, "learning_rate": 8.949152542372882e-08, "loss": 1.4768, "step": 27360 }, { "epoch": 0.68, "grad_norm": 72.5, "learning_rate": 8.915254237288134e-08, "loss": 1.4866, "step": 27370 }, { "epoch": 0.68, "grad_norm": 67.0, "learning_rate": 8.88135593220339e-08, "loss": 1.4376, "step": 27380 }, { "epoch": 0.68, "grad_norm": 65.5, "learning_rate": 8.847457627118644e-08, "loss": 1.492, "step": 27390 }, { "epoch": 0.69, "grad_norm": 68.5, "learning_rate": 8.813559322033898e-08, "loss": 1.4365, "step": 27400 }, { "epoch": 0.69, "grad_norm": 65.5, "learning_rate": 8.779661016949152e-08, "loss": 1.426, "step": 27410 }, { "epoch": 0.69, "grad_norm": 66.5, "learning_rate": 8.745762711864406e-08, "loss": 1.4611, "step": 27420 }, { "epoch": 0.69, "grad_norm": 63.25, "learning_rate": 8.71186440677966e-08, "loss": 1.4911, "step": 27430 }, { "epoch": 0.69, "grad_norm": 67.0, "learning_rate": 8.677966101694916e-08, "loss": 1.4346, "step": 27440 }, { "epoch": 0.69, "grad_norm": 65.5, "learning_rate": 8.644067796610168e-08, "loss": 1.4557, "step": 27450 }, { "epoch": 0.69, "grad_norm": 68.5, "learning_rate": 8.610169491525424e-08, "loss": 1.4695, "step": 27460 }, { "epoch": 0.69, "grad_norm": 64.5, "learning_rate": 8.576271186440677e-08, "loss": 1.4524, "step": 27470 }, { "epoch": 0.69, "grad_norm": 66.5, "learning_rate": 8.542372881355932e-08, "loss": 1.4999, "step": 27480 }, { "epoch": 0.69, "grad_norm": 74.5, "learning_rate": 8.508474576271186e-08, "loss": 1.4584, "step": 27490 }, { "epoch": 0.69, "grad_norm": 66.0, "learning_rate": 8.47457627118644e-08, "loss": 1.4943, "step": 27500 }, { "epoch": 0.69, "grad_norm": 63.25, "learning_rate": 8.440677966101694e-08, "loss": 1.4495, "step": 27510 }, { "epoch": 0.69, "grad_norm": 66.5, "learning_rate": 8.40677966101695e-08, "loss": 1.4801, "step": 27520 }, { "epoch": 0.69, "grad_norm": 66.0, "learning_rate": 8.372881355932203e-08, "loss": 1.4397, "step": 27530 }, { "epoch": 0.69, "grad_norm": 65.5, "learning_rate": 8.338983050847458e-08, "loss": 1.4397, "step": 27540 }, { "epoch": 0.69, "grad_norm": 68.5, "learning_rate": 8.305084745762711e-08, "loss": 1.4723, "step": 27550 }, { "epoch": 0.69, "grad_norm": 65.5, "learning_rate": 8.271186440677966e-08, "loss": 1.4612, "step": 27560 }, { "epoch": 0.69, "grad_norm": 66.5, "learning_rate": 8.23728813559322e-08, "loss": 1.4544, "step": 27570 }, { "epoch": 0.69, "grad_norm": 70.5, "learning_rate": 8.203389830508474e-08, "loss": 1.5102, "step": 27580 }, { "epoch": 0.69, "grad_norm": 68.5, "learning_rate": 8.169491525423728e-08, "loss": 1.4415, "step": 27590 }, { "epoch": 0.69, "grad_norm": 65.0, "learning_rate": 8.135593220338982e-08, "loss": 1.3897, "step": 27600 }, { "epoch": 0.69, "grad_norm": 67.5, "learning_rate": 8.101694915254237e-08, "loss": 1.5114, "step": 27610 }, { "epoch": 0.69, "grad_norm": 68.0, "learning_rate": 8.067796610169492e-08, "loss": 1.5128, "step": 27620 }, { "epoch": 0.69, "grad_norm": 67.0, "learning_rate": 8.033898305084745e-08, "loss": 1.4465, "step": 27630 }, { "epoch": 0.69, "grad_norm": 64.5, "learning_rate": 8e-08, "loss": 1.442, "step": 27640 }, { "epoch": 0.69, "grad_norm": 64.5, "learning_rate": 7.966101694915253e-08, "loss": 1.4609, "step": 27650 }, { "epoch": 0.69, "grad_norm": 65.0, "learning_rate": 7.932203389830508e-08, "loss": 1.4393, "step": 27660 }, { "epoch": 0.69, "grad_norm": 68.5, "learning_rate": 7.898305084745762e-08, "loss": 1.4636, "step": 27670 }, { "epoch": 0.69, "grad_norm": 64.5, "learning_rate": 7.864406779661017e-08, "loss": 1.4555, "step": 27680 }, { "epoch": 0.69, "grad_norm": 64.0, "learning_rate": 7.83050847457627e-08, "loss": 1.443, "step": 27690 }, { "epoch": 0.69, "grad_norm": 67.0, "learning_rate": 7.796610169491526e-08, "loss": 1.4717, "step": 27700 }, { "epoch": 0.69, "grad_norm": 71.5, "learning_rate": 7.762711864406779e-08, "loss": 1.4653, "step": 27710 }, { "epoch": 0.69, "grad_norm": 67.5, "learning_rate": 7.728813559322034e-08, "loss": 1.4794, "step": 27720 }, { "epoch": 0.69, "grad_norm": 69.5, "learning_rate": 7.694915254237287e-08, "loss": 1.4852, "step": 27730 }, { "epoch": 0.69, "grad_norm": 65.0, "learning_rate": 7.661016949152542e-08, "loss": 1.4311, "step": 27740 }, { "epoch": 0.69, "grad_norm": 70.0, "learning_rate": 7.627118644067796e-08, "loss": 1.4993, "step": 27750 }, { "epoch": 0.69, "grad_norm": 66.5, "learning_rate": 7.59322033898305e-08, "loss": 1.4761, "step": 27760 }, { "epoch": 0.69, "grad_norm": 66.0, "learning_rate": 7.559322033898305e-08, "loss": 1.4759, "step": 27770 }, { "epoch": 0.69, "grad_norm": 64.5, "learning_rate": 7.525423728813559e-08, "loss": 1.4645, "step": 27780 }, { "epoch": 0.69, "grad_norm": 66.0, "learning_rate": 7.491525423728813e-08, "loss": 1.4371, "step": 27790 }, { "epoch": 0.69, "grad_norm": 69.0, "learning_rate": 7.457627118644068e-08, "loss": 1.4188, "step": 27800 }, { "epoch": 0.7, "grad_norm": 67.0, "learning_rate": 7.423728813559321e-08, "loss": 1.4593, "step": 27810 }, { "epoch": 0.7, "grad_norm": 67.5, "learning_rate": 7.389830508474576e-08, "loss": 1.4744, "step": 27820 }, { "epoch": 0.7, "grad_norm": 67.5, "learning_rate": 7.35593220338983e-08, "loss": 1.3997, "step": 27830 }, { "epoch": 0.7, "grad_norm": 64.5, "learning_rate": 7.322033898305085e-08, "loss": 1.4356, "step": 27840 }, { "epoch": 0.7, "grad_norm": 65.0, "learning_rate": 7.288135593220339e-08, "loss": 1.4756, "step": 27850 }, { "epoch": 0.7, "grad_norm": 70.5, "learning_rate": 7.254237288135593e-08, "loss": 1.4537, "step": 27860 }, { "epoch": 0.7, "grad_norm": 69.0, "learning_rate": 7.220338983050847e-08, "loss": 1.4577, "step": 27870 }, { "epoch": 0.7, "grad_norm": 68.0, "learning_rate": 7.186440677966102e-08, "loss": 1.4765, "step": 27880 }, { "epoch": 0.7, "grad_norm": 69.5, "learning_rate": 7.152542372881355e-08, "loss": 1.4104, "step": 27890 }, { "epoch": 0.7, "grad_norm": 69.0, "learning_rate": 7.11864406779661e-08, "loss": 1.4665, "step": 27900 }, { "epoch": 0.7, "grad_norm": 67.0, "learning_rate": 7.084745762711863e-08, "loss": 1.4816, "step": 27910 }, { "epoch": 0.7, "grad_norm": 66.0, "learning_rate": 7.050847457627119e-08, "loss": 1.4724, "step": 27920 }, { "epoch": 0.7, "grad_norm": 67.5, "learning_rate": 7.016949152542373e-08, "loss": 1.4205, "step": 27930 }, { "epoch": 0.7, "grad_norm": 72.0, "learning_rate": 6.983050847457627e-08, "loss": 1.4627, "step": 27940 }, { "epoch": 0.7, "grad_norm": 66.5, "learning_rate": 6.949152542372881e-08, "loss": 1.4729, "step": 27950 }, { "epoch": 0.7, "grad_norm": 65.5, "learning_rate": 6.915254237288136e-08, "loss": 1.4396, "step": 27960 }, { "epoch": 0.7, "grad_norm": 64.5, "learning_rate": 6.881355932203389e-08, "loss": 1.4283, "step": 27970 }, { "epoch": 0.7, "grad_norm": 67.5, "learning_rate": 6.847457627118645e-08, "loss": 1.4881, "step": 27980 }, { "epoch": 0.7, "grad_norm": 67.5, "learning_rate": 6.813559322033897e-08, "loss": 1.5324, "step": 27990 }, { "epoch": 0.7, "grad_norm": 65.0, "learning_rate": 6.779661016949153e-08, "loss": 1.4599, "step": 28000 }, { "epoch": 0.7, "grad_norm": 68.5, "learning_rate": 6.745762711864407e-08, "loss": 1.4843, "step": 28010 }, { "epoch": 0.7, "grad_norm": 69.5, "learning_rate": 6.711864406779661e-08, "loss": 1.4745, "step": 28020 }, { "epoch": 0.7, "grad_norm": 67.5, "learning_rate": 6.677966101694915e-08, "loss": 1.5165, "step": 28030 }, { "epoch": 0.7, "grad_norm": 68.0, "learning_rate": 6.644067796610169e-08, "loss": 1.479, "step": 28040 }, { "epoch": 0.7, "grad_norm": 66.5, "learning_rate": 6.610169491525423e-08, "loss": 1.4236, "step": 28050 }, { "epoch": 0.7, "grad_norm": 72.0, "learning_rate": 6.576271186440679e-08, "loss": 1.4542, "step": 28060 }, { "epoch": 0.7, "grad_norm": 68.0, "learning_rate": 6.542372881355931e-08, "loss": 1.466, "step": 28070 }, { "epoch": 0.7, "grad_norm": 69.5, "learning_rate": 6.508474576271187e-08, "loss": 1.5094, "step": 28080 }, { "epoch": 0.7, "grad_norm": 68.0, "learning_rate": 6.474576271186441e-08, "loss": 1.4899, "step": 28090 }, { "epoch": 0.7, "grad_norm": 70.5, "learning_rate": 6.440677966101695e-08, "loss": 1.5057, "step": 28100 }, { "epoch": 0.7, "grad_norm": 67.5, "learning_rate": 6.406779661016949e-08, "loss": 1.5005, "step": 28110 }, { "epoch": 0.7, "grad_norm": 65.5, "learning_rate": 6.372881355932203e-08, "loss": 1.4355, "step": 28120 }, { "epoch": 0.7, "grad_norm": 67.0, "learning_rate": 6.338983050847457e-08, "loss": 1.4489, "step": 28130 }, { "epoch": 0.7, "grad_norm": 69.0, "learning_rate": 6.305084745762713e-08, "loss": 1.4299, "step": 28140 }, { "epoch": 0.7, "grad_norm": 64.5, "learning_rate": 6.271186440677965e-08, "loss": 1.4528, "step": 28150 }, { "epoch": 0.7, "grad_norm": 65.5, "learning_rate": 6.237288135593221e-08, "loss": 1.4766, "step": 28160 }, { "epoch": 0.7, "grad_norm": 64.0, "learning_rate": 6.203389830508475e-08, "loss": 1.4354, "step": 28170 }, { "epoch": 0.7, "grad_norm": 62.25, "learning_rate": 6.169491525423729e-08, "loss": 1.4779, "step": 28180 }, { "epoch": 0.7, "grad_norm": 65.0, "learning_rate": 6.135593220338983e-08, "loss": 1.5016, "step": 28190 }, { "epoch": 0.7, "grad_norm": 64.5, "learning_rate": 6.101694915254237e-08, "loss": 1.4592, "step": 28200 }, { "epoch": 0.71, "grad_norm": 67.0, "learning_rate": 6.067796610169491e-08, "loss": 1.4624, "step": 28210 }, { "epoch": 0.71, "grad_norm": 65.0, "learning_rate": 6.033898305084745e-08, "loss": 1.4414, "step": 28220 }, { "epoch": 0.71, "grad_norm": 70.0, "learning_rate": 6e-08, "loss": 1.4958, "step": 28230 }, { "epoch": 0.71, "grad_norm": 69.5, "learning_rate": 5.966101694915255e-08, "loss": 1.4772, "step": 28240 }, { "epoch": 0.71, "grad_norm": 68.0, "learning_rate": 5.932203389830508e-08, "loss": 1.4262, "step": 28250 }, { "epoch": 0.71, "grad_norm": 67.5, "learning_rate": 5.8983050847457624e-08, "loss": 1.4118, "step": 28260 }, { "epoch": 0.71, "grad_norm": 69.5, "learning_rate": 5.8644067796610165e-08, "loss": 1.5084, "step": 28270 }, { "epoch": 0.71, "grad_norm": 66.5, "learning_rate": 5.8305084745762706e-08, "loss": 1.4377, "step": 28280 }, { "epoch": 0.71, "grad_norm": 72.0, "learning_rate": 5.7966101694915253e-08, "loss": 1.4479, "step": 28290 }, { "epoch": 0.71, "grad_norm": 68.0, "learning_rate": 5.7627118644067794e-08, "loss": 1.4761, "step": 28300 }, { "epoch": 0.71, "grad_norm": 66.5, "learning_rate": 5.7288135593220335e-08, "loss": 1.4407, "step": 28310 }, { "epoch": 0.71, "grad_norm": 69.5, "learning_rate": 5.6949152542372876e-08, "loss": 1.4617, "step": 28320 }, { "epoch": 0.71, "grad_norm": 67.5, "learning_rate": 5.6610169491525424e-08, "loss": 1.4758, "step": 28330 }, { "epoch": 0.71, "grad_norm": 69.5, "learning_rate": 5.6271186440677964e-08, "loss": 1.461, "step": 28340 }, { "epoch": 0.71, "grad_norm": 64.0, "learning_rate": 5.5932203389830505e-08, "loss": 1.4355, "step": 28350 }, { "epoch": 0.71, "grad_norm": 66.5, "learning_rate": 5.5593220338983046e-08, "loss": 1.4778, "step": 28360 }, { "epoch": 0.71, "grad_norm": 67.5, "learning_rate": 5.5254237288135594e-08, "loss": 1.4279, "step": 28370 }, { "epoch": 0.71, "grad_norm": 68.5, "learning_rate": 5.4915254237288135e-08, "loss": 1.4686, "step": 28380 }, { "epoch": 0.71, "grad_norm": 63.5, "learning_rate": 5.4576271186440676e-08, "loss": 1.4389, "step": 28390 }, { "epoch": 0.71, "grad_norm": 66.5, "learning_rate": 5.4237288135593217e-08, "loss": 1.4694, "step": 28400 }, { "epoch": 0.71, "grad_norm": 66.5, "learning_rate": 5.389830508474576e-08, "loss": 1.471, "step": 28410 }, { "epoch": 0.71, "grad_norm": 67.0, "learning_rate": 5.3559322033898305e-08, "loss": 1.479, "step": 28420 }, { "epoch": 0.71, "grad_norm": 64.0, "learning_rate": 5.3220338983050846e-08, "loss": 1.4752, "step": 28430 }, { "epoch": 0.71, "grad_norm": 64.5, "learning_rate": 5.288135593220339e-08, "loss": 1.4571, "step": 28440 }, { "epoch": 0.71, "grad_norm": 69.5, "learning_rate": 5.254237288135593e-08, "loss": 1.4891, "step": 28450 }, { "epoch": 0.71, "grad_norm": 65.0, "learning_rate": 5.2203389830508475e-08, "loss": 1.4618, "step": 28460 }, { "epoch": 0.71, "grad_norm": 70.0, "learning_rate": 5.1864406779661016e-08, "loss": 1.5078, "step": 28470 }, { "epoch": 0.71, "grad_norm": 67.5, "learning_rate": 5.152542372881356e-08, "loss": 1.4645, "step": 28480 }, { "epoch": 0.71, "grad_norm": 65.5, "learning_rate": 5.11864406779661e-08, "loss": 1.4452, "step": 28490 }, { "epoch": 0.71, "grad_norm": 63.0, "learning_rate": 5.0847457627118645e-08, "loss": 1.457, "step": 28500 }, { "epoch": 0.71, "grad_norm": 65.5, "learning_rate": 5.0508474576271186e-08, "loss": 1.5021, "step": 28510 }, { "epoch": 0.71, "grad_norm": 66.0, "learning_rate": 5.016949152542373e-08, "loss": 1.4595, "step": 28520 }, { "epoch": 0.71, "grad_norm": 64.5, "learning_rate": 4.983050847457627e-08, "loss": 1.4542, "step": 28530 }, { "epoch": 0.71, "grad_norm": 66.5, "learning_rate": 4.949152542372881e-08, "loss": 1.4501, "step": 28540 }, { "epoch": 0.71, "grad_norm": 67.0, "learning_rate": 4.9152542372881357e-08, "loss": 1.4917, "step": 28550 }, { "epoch": 0.71, "grad_norm": 68.5, "learning_rate": 4.88135593220339e-08, "loss": 1.4516, "step": 28560 }, { "epoch": 0.71, "grad_norm": 65.0, "learning_rate": 4.847457627118644e-08, "loss": 1.4666, "step": 28570 }, { "epoch": 0.71, "grad_norm": 69.0, "learning_rate": 4.813559322033898e-08, "loss": 1.5123, "step": 28580 }, { "epoch": 0.71, "grad_norm": 69.0, "learning_rate": 4.779661016949153e-08, "loss": 1.4659, "step": 28590 }, { "epoch": 0.71, "grad_norm": 69.0, "learning_rate": 4.745762711864407e-08, "loss": 1.4839, "step": 28600 }, { "epoch": 0.72, "grad_norm": 67.5, "learning_rate": 4.711864406779661e-08, "loss": 1.4728, "step": 28610 }, { "epoch": 0.72, "grad_norm": 67.0, "learning_rate": 4.677966101694915e-08, "loss": 1.4596, "step": 28620 }, { "epoch": 0.72, "grad_norm": 64.0, "learning_rate": 4.64406779661017e-08, "loss": 1.4157, "step": 28630 }, { "epoch": 0.72, "grad_norm": 68.0, "learning_rate": 4.610169491525424e-08, "loss": 1.4506, "step": 28640 }, { "epoch": 0.72, "grad_norm": 63.25, "learning_rate": 4.576271186440678e-08, "loss": 1.4663, "step": 28650 }, { "epoch": 0.72, "grad_norm": 66.0, "learning_rate": 4.542372881355932e-08, "loss": 1.4602, "step": 28660 }, { "epoch": 0.72, "grad_norm": 66.0, "learning_rate": 4.508474576271186e-08, "loss": 1.4337, "step": 28670 }, { "epoch": 0.72, "grad_norm": 66.0, "learning_rate": 4.474576271186441e-08, "loss": 1.4953, "step": 28680 }, { "epoch": 0.72, "grad_norm": 67.5, "learning_rate": 4.440677966101695e-08, "loss": 1.4918, "step": 28690 }, { "epoch": 0.72, "grad_norm": 67.0, "learning_rate": 4.406779661016949e-08, "loss": 1.5065, "step": 28700 }, { "epoch": 0.72, "grad_norm": 69.0, "learning_rate": 4.372881355932203e-08, "loss": 1.4844, "step": 28710 }, { "epoch": 0.72, "grad_norm": 68.5, "learning_rate": 4.338983050847458e-08, "loss": 1.4525, "step": 28720 }, { "epoch": 0.72, "grad_norm": 67.5, "learning_rate": 4.305084745762712e-08, "loss": 1.4415, "step": 28730 }, { "epoch": 0.72, "grad_norm": 66.0, "learning_rate": 4.271186440677966e-08, "loss": 1.4609, "step": 28740 }, { "epoch": 0.72, "grad_norm": 68.0, "learning_rate": 4.23728813559322e-08, "loss": 1.41, "step": 28750 }, { "epoch": 0.72, "grad_norm": 69.0, "learning_rate": 4.203389830508475e-08, "loss": 1.4901, "step": 28760 }, { "epoch": 0.72, "grad_norm": 68.5, "learning_rate": 4.169491525423729e-08, "loss": 1.4615, "step": 28770 }, { "epoch": 0.72, "grad_norm": 64.0, "learning_rate": 4.135593220338983e-08, "loss": 1.4815, "step": 28780 }, { "epoch": 0.72, "grad_norm": 68.5, "learning_rate": 4.101694915254237e-08, "loss": 1.4814, "step": 28790 }, { "epoch": 0.72, "grad_norm": 69.5, "learning_rate": 4.067796610169491e-08, "loss": 1.4773, "step": 28800 }, { "epoch": 0.72, "grad_norm": 67.0, "learning_rate": 4.033898305084746e-08, "loss": 1.4626, "step": 28810 }, { "epoch": 0.72, "grad_norm": 66.5, "learning_rate": 4e-08, "loss": 1.4785, "step": 28820 }, { "epoch": 0.72, "grad_norm": 67.0, "learning_rate": 3.966101694915254e-08, "loss": 1.4811, "step": 28830 }, { "epoch": 0.72, "grad_norm": 65.5, "learning_rate": 3.932203389830508e-08, "loss": 1.4873, "step": 28840 }, { "epoch": 0.72, "grad_norm": 63.75, "learning_rate": 3.898305084745763e-08, "loss": 1.4226, "step": 28850 }, { "epoch": 0.72, "grad_norm": 69.0, "learning_rate": 3.864406779661017e-08, "loss": 1.485, "step": 28860 }, { "epoch": 0.72, "grad_norm": 65.0, "learning_rate": 3.830508474576271e-08, "loss": 1.4259, "step": 28870 }, { "epoch": 0.72, "grad_norm": 68.5, "learning_rate": 3.796610169491525e-08, "loss": 1.4619, "step": 28880 }, { "epoch": 0.72, "grad_norm": 73.0, "learning_rate": 3.7627118644067794e-08, "loss": 1.4714, "step": 28890 }, { "epoch": 0.72, "grad_norm": 65.0, "learning_rate": 3.728813559322034e-08, "loss": 1.4456, "step": 28900 }, { "epoch": 0.72, "grad_norm": 65.0, "learning_rate": 3.694915254237288e-08, "loss": 1.4864, "step": 28910 }, { "epoch": 0.72, "grad_norm": 63.5, "learning_rate": 3.661016949152542e-08, "loss": 1.394, "step": 28920 }, { "epoch": 0.72, "grad_norm": 67.0, "learning_rate": 3.6271186440677964e-08, "loss": 1.4562, "step": 28930 }, { "epoch": 0.72, "grad_norm": 69.0, "learning_rate": 3.593220338983051e-08, "loss": 1.5021, "step": 28940 }, { "epoch": 0.72, "grad_norm": 65.5, "learning_rate": 3.559322033898305e-08, "loss": 1.4753, "step": 28950 }, { "epoch": 0.72, "grad_norm": 64.5, "learning_rate": 3.5254237288135593e-08, "loss": 1.4206, "step": 28960 }, { "epoch": 0.72, "grad_norm": 66.0, "learning_rate": 3.4915254237288134e-08, "loss": 1.4772, "step": 28970 }, { "epoch": 0.72, "grad_norm": 67.5, "learning_rate": 3.457627118644068e-08, "loss": 1.4506, "step": 28980 }, { "epoch": 0.72, "grad_norm": 68.0, "learning_rate": 3.423728813559322e-08, "loss": 1.5169, "step": 28990 }, { "epoch": 0.72, "grad_norm": 64.5, "learning_rate": 3.3898305084745764e-08, "loss": 1.4311, "step": 29000 }, { "epoch": 0.73, "grad_norm": 66.5, "learning_rate": 3.3559322033898305e-08, "loss": 1.5013, "step": 29010 }, { "epoch": 0.73, "grad_norm": 66.5, "learning_rate": 3.3220338983050845e-08, "loss": 1.3978, "step": 29020 }, { "epoch": 0.73, "grad_norm": 69.5, "learning_rate": 3.288135593220339e-08, "loss": 1.5471, "step": 29030 }, { "epoch": 0.73, "grad_norm": 66.0, "learning_rate": 3.2542372881355934e-08, "loss": 1.438, "step": 29040 }, { "epoch": 0.73, "grad_norm": 64.5, "learning_rate": 3.2203389830508475e-08, "loss": 1.4891, "step": 29050 }, { "epoch": 0.73, "grad_norm": 70.0, "learning_rate": 3.1864406779661016e-08, "loss": 1.4335, "step": 29060 }, { "epoch": 0.73, "grad_norm": 66.5, "learning_rate": 3.152542372881356e-08, "loss": 1.4434, "step": 29070 }, { "epoch": 0.73, "grad_norm": 66.5, "learning_rate": 3.1186440677966104e-08, "loss": 1.461, "step": 29080 }, { "epoch": 0.73, "grad_norm": 68.0, "learning_rate": 3.0847457627118645e-08, "loss": 1.5126, "step": 29090 }, { "epoch": 0.73, "grad_norm": 67.0, "learning_rate": 3.0508474576271186e-08, "loss": 1.4445, "step": 29100 }, { "epoch": 0.73, "grad_norm": 64.5, "learning_rate": 3.016949152542373e-08, "loss": 1.4168, "step": 29110 }, { "epoch": 0.73, "grad_norm": 64.5, "learning_rate": 2.9830508474576274e-08, "loss": 1.4507, "step": 29120 }, { "epoch": 0.73, "grad_norm": 65.5, "learning_rate": 2.9491525423728812e-08, "loss": 1.458, "step": 29130 }, { "epoch": 0.73, "grad_norm": 69.5, "learning_rate": 2.9152542372881353e-08, "loss": 1.4649, "step": 29140 }, { "epoch": 0.73, "grad_norm": 66.0, "learning_rate": 2.8813559322033897e-08, "loss": 1.457, "step": 29150 }, { "epoch": 0.73, "grad_norm": 72.5, "learning_rate": 2.8474576271186438e-08, "loss": 1.4955, "step": 29160 }, { "epoch": 0.73, "grad_norm": 67.0, "learning_rate": 2.8135593220338982e-08, "loss": 1.4841, "step": 29170 }, { "epoch": 0.73, "grad_norm": 70.5, "learning_rate": 2.7796610169491523e-08, "loss": 1.4464, "step": 29180 }, { "epoch": 0.73, "grad_norm": 68.0, "learning_rate": 2.7457627118644067e-08, "loss": 1.4687, "step": 29190 }, { "epoch": 0.73, "grad_norm": 65.0, "learning_rate": 2.7118644067796608e-08, "loss": 1.5202, "step": 29200 }, { "epoch": 0.73, "grad_norm": 68.0, "learning_rate": 2.6779661016949152e-08, "loss": 1.449, "step": 29210 }, { "epoch": 0.73, "grad_norm": 70.0, "learning_rate": 2.6440677966101693e-08, "loss": 1.5073, "step": 29220 }, { "epoch": 0.73, "grad_norm": 65.0, "learning_rate": 2.6101694915254238e-08, "loss": 1.4933, "step": 29230 }, { "epoch": 0.73, "grad_norm": 64.0, "learning_rate": 2.576271186440678e-08, "loss": 1.4657, "step": 29240 }, { "epoch": 0.73, "grad_norm": 66.5, "learning_rate": 2.5423728813559323e-08, "loss": 1.4532, "step": 29250 }, { "epoch": 0.73, "grad_norm": 69.5, "learning_rate": 2.5084745762711864e-08, "loss": 1.4642, "step": 29260 }, { "epoch": 0.73, "grad_norm": 68.0, "learning_rate": 2.4745762711864405e-08, "loss": 1.4637, "step": 29270 }, { "epoch": 0.73, "grad_norm": 63.5, "learning_rate": 2.440677966101695e-08, "loss": 1.4633, "step": 29280 }, { "epoch": 0.73, "grad_norm": 63.0, "learning_rate": 2.406779661016949e-08, "loss": 1.44, "step": 29290 }, { "epoch": 0.73, "grad_norm": 67.5, "learning_rate": 2.3728813559322034e-08, "loss": 1.5197, "step": 29300 }, { "epoch": 0.73, "grad_norm": 64.5, "learning_rate": 2.3389830508474575e-08, "loss": 1.464, "step": 29310 }, { "epoch": 0.73, "grad_norm": 73.5, "learning_rate": 2.305084745762712e-08, "loss": 1.5056, "step": 29320 }, { "epoch": 0.73, "grad_norm": 67.5, "learning_rate": 2.271186440677966e-08, "loss": 1.4384, "step": 29330 }, { "epoch": 0.73, "grad_norm": 68.0, "learning_rate": 2.2372881355932204e-08, "loss": 1.47, "step": 29340 }, { "epoch": 0.73, "grad_norm": 71.5, "learning_rate": 2.2033898305084745e-08, "loss": 1.4637, "step": 29350 }, { "epoch": 0.73, "grad_norm": 65.5, "learning_rate": 2.169491525423729e-08, "loss": 1.4459, "step": 29360 }, { "epoch": 0.73, "grad_norm": 66.0, "learning_rate": 2.135593220338983e-08, "loss": 1.4714, "step": 29370 }, { "epoch": 0.73, "grad_norm": 64.0, "learning_rate": 2.1016949152542374e-08, "loss": 1.4877, "step": 29380 }, { "epoch": 0.73, "grad_norm": 66.0, "learning_rate": 2.0677966101694915e-08, "loss": 1.4625, "step": 29390 }, { "epoch": 0.73, "grad_norm": 68.0, "learning_rate": 2.0338983050847456e-08, "loss": 1.4429, "step": 29400 }, { "epoch": 0.74, "grad_norm": 66.5, "learning_rate": 2e-08, "loss": 1.4736, "step": 29410 }, { "epoch": 0.74, "grad_norm": 71.0, "learning_rate": 1.966101694915254e-08, "loss": 1.4478, "step": 29420 }, { "epoch": 0.74, "grad_norm": 68.0, "learning_rate": 1.9322033898305086e-08, "loss": 1.4806, "step": 29430 }, { "epoch": 0.74, "grad_norm": 65.0, "learning_rate": 1.8983050847457626e-08, "loss": 1.458, "step": 29440 }, { "epoch": 0.74, "grad_norm": 64.5, "learning_rate": 1.864406779661017e-08, "loss": 1.4625, "step": 29450 }, { "epoch": 0.74, "grad_norm": 64.5, "learning_rate": 1.830508474576271e-08, "loss": 1.4605, "step": 29460 }, { "epoch": 0.74, "grad_norm": 68.5, "learning_rate": 1.7966101694915256e-08, "loss": 1.465, "step": 29470 }, { "epoch": 0.74, "grad_norm": 66.5, "learning_rate": 1.7627118644067797e-08, "loss": 1.4571, "step": 29480 }, { "epoch": 0.74, "grad_norm": 64.5, "learning_rate": 1.728813559322034e-08, "loss": 1.438, "step": 29490 }, { "epoch": 0.74, "grad_norm": 69.5, "learning_rate": 1.6949152542372882e-08, "loss": 1.5273, "step": 29500 }, { "epoch": 0.74, "grad_norm": 64.5, "learning_rate": 1.6610169491525423e-08, "loss": 1.4298, "step": 29510 }, { "epoch": 0.74, "grad_norm": 65.5, "learning_rate": 1.6271186440677967e-08, "loss": 1.4049, "step": 29520 }, { "epoch": 0.74, "grad_norm": 69.0, "learning_rate": 1.5932203389830508e-08, "loss": 1.4414, "step": 29530 }, { "epoch": 0.74, "grad_norm": 68.0, "learning_rate": 1.5593220338983052e-08, "loss": 1.4952, "step": 29540 }, { "epoch": 0.74, "grad_norm": 65.5, "learning_rate": 1.5254237288135593e-08, "loss": 1.4542, "step": 29550 }, { "epoch": 0.74, "grad_norm": 67.5, "learning_rate": 1.4915254237288137e-08, "loss": 1.4709, "step": 29560 }, { "epoch": 0.74, "grad_norm": 66.5, "learning_rate": 1.4576271186440676e-08, "loss": 1.4756, "step": 29570 }, { "epoch": 0.74, "grad_norm": 63.75, "learning_rate": 1.4237288135593219e-08, "loss": 1.4611, "step": 29580 }, { "epoch": 0.74, "grad_norm": 67.0, "learning_rate": 1.3898305084745762e-08, "loss": 1.4593, "step": 29590 }, { "epoch": 0.74, "grad_norm": 68.0, "learning_rate": 1.3559322033898304e-08, "loss": 1.4649, "step": 29600 }, { "epoch": 0.74, "grad_norm": 66.0, "learning_rate": 1.3220338983050847e-08, "loss": 1.4545, "step": 29610 }, { "epoch": 0.74, "grad_norm": 65.0, "learning_rate": 1.288135593220339e-08, "loss": 1.4639, "step": 29620 }, { "epoch": 0.74, "grad_norm": 65.5, "learning_rate": 1.2542372881355932e-08, "loss": 1.4659, "step": 29630 }, { "epoch": 0.74, "grad_norm": 64.5, "learning_rate": 1.2203389830508474e-08, "loss": 1.4277, "step": 29640 }, { "epoch": 0.74, "grad_norm": 69.0, "learning_rate": 1.1864406779661017e-08, "loss": 1.473, "step": 29650 }, { "epoch": 0.74, "grad_norm": 68.5, "learning_rate": 1.152542372881356e-08, "loss": 1.4905, "step": 29660 }, { "epoch": 0.74, "grad_norm": 67.0, "learning_rate": 1.1186440677966102e-08, "loss": 1.4237, "step": 29670 }, { "epoch": 0.74, "grad_norm": 69.5, "learning_rate": 1.0847457627118645e-08, "loss": 1.4994, "step": 29680 }, { "epoch": 0.74, "grad_norm": 67.0, "learning_rate": 1.0508474576271187e-08, "loss": 1.4373, "step": 29690 }, { "epoch": 0.74, "grad_norm": 65.5, "learning_rate": 1.0169491525423728e-08, "loss": 1.4231, "step": 29700 }, { "epoch": 0.74, "grad_norm": 69.0, "learning_rate": 9.83050847457627e-09, "loss": 1.4697, "step": 29710 }, { "epoch": 0.74, "grad_norm": 66.0, "learning_rate": 9.491525423728813e-09, "loss": 1.5293, "step": 29720 }, { "epoch": 0.74, "grad_norm": 64.5, "learning_rate": 9.152542372881356e-09, "loss": 1.4447, "step": 29730 }, { "epoch": 0.74, "grad_norm": 66.0, "learning_rate": 8.813559322033898e-09, "loss": 1.4482, "step": 29740 }, { "epoch": 0.74, "grad_norm": 66.0, "learning_rate": 8.474576271186441e-09, "loss": 1.4338, "step": 29750 }, { "epoch": 0.74, "grad_norm": 69.5, "learning_rate": 8.135593220338983e-09, "loss": 1.5099, "step": 29760 }, { "epoch": 0.74, "grad_norm": 66.5, "learning_rate": 7.796610169491526e-09, "loss": 1.5048, "step": 29770 }, { "epoch": 0.74, "grad_norm": 63.25, "learning_rate": 7.457627118644069e-09, "loss": 1.4926, "step": 29780 }, { "epoch": 0.74, "grad_norm": 66.0, "learning_rate": 7.1186440677966095e-09, "loss": 1.4132, "step": 29790 }, { "epoch": 0.74, "grad_norm": 69.0, "learning_rate": 6.779661016949152e-09, "loss": 1.4814, "step": 29800 }, { "epoch": 0.75, "grad_norm": 67.0, "learning_rate": 6.440677966101695e-09, "loss": 1.4551, "step": 29810 }, { "epoch": 0.75, "grad_norm": 67.0, "learning_rate": 6.101694915254237e-09, "loss": 1.4826, "step": 29820 }, { "epoch": 0.75, "grad_norm": 71.0, "learning_rate": 5.76271186440678e-09, "loss": 1.458, "step": 29830 }, { "epoch": 0.75, "grad_norm": 66.5, "learning_rate": 5.423728813559322e-09, "loss": 1.5004, "step": 29840 }, { "epoch": 0.75, "grad_norm": 65.5, "learning_rate": 5.084745762711864e-09, "loss": 1.4904, "step": 29850 }, { "epoch": 0.75, "grad_norm": 66.5, "learning_rate": 4.745762711864407e-09, "loss": 1.4144, "step": 29860 }, { "epoch": 0.75, "grad_norm": 69.5, "learning_rate": 4.406779661016949e-09, "loss": 1.4279, "step": 29870 }, { "epoch": 0.75, "grad_norm": 65.0, "learning_rate": 4.067796610169492e-09, "loss": 1.4719, "step": 29880 }, { "epoch": 0.75, "grad_norm": 72.5, "learning_rate": 3.728813559322034e-09, "loss": 1.5085, "step": 29890 }, { "epoch": 0.75, "grad_norm": 70.5, "learning_rate": 3.389830508474576e-09, "loss": 1.4336, "step": 29900 }, { "epoch": 0.75, "grad_norm": 63.25, "learning_rate": 3.0508474576271186e-09, "loss": 1.4601, "step": 29910 }, { "epoch": 0.75, "grad_norm": 65.0, "learning_rate": 2.711864406779661e-09, "loss": 1.4593, "step": 29920 }, { "epoch": 0.75, "grad_norm": 67.0, "learning_rate": 2.3728813559322033e-09, "loss": 1.4755, "step": 29930 }, { "epoch": 0.75, "grad_norm": 60.5, "learning_rate": 2.033898305084746e-09, "loss": 1.4302, "step": 29940 }, { "epoch": 0.75, "grad_norm": 65.5, "learning_rate": 1.694915254237288e-09, "loss": 1.4244, "step": 29950 }, { "epoch": 0.75, "grad_norm": 67.0, "learning_rate": 1.3559322033898306e-09, "loss": 1.4924, "step": 29960 }, { "epoch": 0.75, "grad_norm": 67.5, "learning_rate": 1.016949152542373e-09, "loss": 1.4943, "step": 29970 }, { "epoch": 0.75, "grad_norm": 65.5, "learning_rate": 6.779661016949153e-10, "loss": 1.4661, "step": 29980 }, { "epoch": 0.75, "grad_norm": 67.0, "learning_rate": 3.3898305084745764e-10, "loss": 1.4674, "step": 29990 }, { "epoch": 0.75, "grad_norm": 71.0, "learning_rate": 0.0, "loss": 1.4675, "step": 30000 }, { "epoch": 0.75, "eval_loss": 1.463437795639038, "eval_runtime": 66.0956, "eval_samples_per_second": 15.13, "eval_steps_per_second": 15.13, "step": 30000 } ], "logging_steps": 10, "max_steps": 30000, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 5000, "total_flos": 4.841766125568e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }