| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.901098901098901, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 5716.10205078125, | |
| "learning_rate": 3.0000000000000004e-08, | |
| "loss": 1383.1007, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 12343.9111328125, | |
| "learning_rate": 6.75e-08, | |
| "loss": 1349.0135, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.380952380952381, | |
| "grad_norm": 10210.48046875, | |
| "learning_rate": 1.05e-07, | |
| "loss": 1340.3351, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 7633.39599609375, | |
| "learning_rate": 1.425e-07, | |
| "loss": 1369.3844, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5494505494505495, | |
| "grad_norm": 12217.7294921875, | |
| "learning_rate": 1.8e-07, | |
| "loss": 1362.663, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.6593406593406593, | |
| "grad_norm": 31348.806640625, | |
| "learning_rate": 2.175e-07, | |
| "loss": 1361.9012, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 7385.05322265625, | |
| "learning_rate": 2.5500000000000005e-07, | |
| "loss": 1365.5517, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.8791208791208791, | |
| "grad_norm": 15594.4677734375, | |
| "learning_rate": 2.925e-07, | |
| "loss": 1341.6402, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.989010989010989, | |
| "grad_norm": 6404.94775390625, | |
| "learning_rate": 3.2999999999999996e-07, | |
| "loss": 1343.9832, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.10989010989011, | |
| "grad_norm": 6927.478515625, | |
| "learning_rate": 3.6750000000000003e-07, | |
| "loss": 1611.4768, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.2197802197802199, | |
| "grad_norm": 12092.9287109375, | |
| "learning_rate": 4.05e-07, | |
| "loss": 1330.508, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.3296703296703296, | |
| "grad_norm": 5830.10693359375, | |
| "learning_rate": 4.425e-07, | |
| "loss": 1333.5215, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.4395604395604396, | |
| "grad_norm": 5059.5302734375, | |
| "learning_rate": 4.800000000000001e-07, | |
| "loss": 1305.5257, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.5494505494505495, | |
| "grad_norm": 14086.837890625, | |
| "learning_rate": 5.175e-07, | |
| "loss": 1293.7212, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.6593406593406592, | |
| "grad_norm": 4606.49462890625, | |
| "learning_rate": 5.55e-07, | |
| "loss": 1293.7292, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 7120.25244140625, | |
| "learning_rate": 5.925e-07, | |
| "loss": 1283.8888, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.879120879120879, | |
| "grad_norm": 4402.51513671875, | |
| "learning_rate": 6.3e-07, | |
| "loss": 1270.6494, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.989010989010989, | |
| "grad_norm": 4826.724609375, | |
| "learning_rate": 6.675e-07, | |
| "loss": 1263.9209, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.087912087912088, | |
| "grad_norm": 4356.83056640625, | |
| "learning_rate": 7.05e-07, | |
| "loss": 1255.111, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.197802197802198, | |
| "grad_norm": 6895.61962890625, | |
| "learning_rate": 7.425000000000001e-07, | |
| "loss": 1224.4364, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 4793.59375, | |
| "learning_rate": 7.799999999999999e-07, | |
| "loss": 1211.3252, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.4175824175824174, | |
| "grad_norm": 6701.6357421875, | |
| "learning_rate": 8.175e-07, | |
| "loss": 1183.1627, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.5274725274725274, | |
| "grad_norm": 4253.81005859375, | |
| "learning_rate": 8.550000000000001e-07, | |
| "loss": 1179.29, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.6373626373626373, | |
| "grad_norm": 4965.86181640625, | |
| "learning_rate": 8.925e-07, | |
| "loss": 1168.3469, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.7472527472527473, | |
| "grad_norm": 3376.816650390625, | |
| "learning_rate": 9.3e-07, | |
| "loss": 1139.6433, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 3226.073486328125, | |
| "learning_rate": 9.675e-07, | |
| "loss": 1133.0777, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.967032967032967, | |
| "grad_norm": 3560.005126953125, | |
| "learning_rate": 1.0050000000000001e-06, | |
| "loss": 1113.3471, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 3.065934065934066, | |
| "grad_norm": 2917.787109375, | |
| "learning_rate": 1.0425000000000002e-06, | |
| "loss": 1094.2038, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.1758241758241756, | |
| "grad_norm": 3291.990478515625, | |
| "learning_rate": 1.08e-06, | |
| "loss": 1069.3087, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.2857142857142856, | |
| "grad_norm": 3082.956298828125, | |
| "learning_rate": 1.1174999999999999e-06, | |
| "loss": 1043.3319, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.3956043956043955, | |
| "grad_norm": 2947.577392578125, | |
| "learning_rate": 1.155e-06, | |
| "loss": 1040.2479, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.5054945054945055, | |
| "grad_norm": 2917.072021484375, | |
| "learning_rate": 1.1925e-06, | |
| "loss": 1014.5039, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.6153846153846154, | |
| "grad_norm": 2572.1064453125, | |
| "learning_rate": 1.23e-06, | |
| "loss": 980.6475, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.7252747252747254, | |
| "grad_norm": 3038.7041015625, | |
| "learning_rate": 1.2675000000000001e-06, | |
| "loss": 969.8687, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.8351648351648353, | |
| "grad_norm": 2459.19482421875, | |
| "learning_rate": 1.305e-06, | |
| "loss": 946.8183, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.9450549450549453, | |
| "grad_norm": 2485.954345703125, | |
| "learning_rate": 1.3425e-06, | |
| "loss": 923.6708, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.043956043956044, | |
| "grad_norm": 2574.14599609375, | |
| "learning_rate": 1.38e-06, | |
| "loss": 906.8975, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 4.153846153846154, | |
| "grad_norm": 2498.670166015625, | |
| "learning_rate": 1.4175e-06, | |
| "loss": 879.6545, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.263736263736264, | |
| "grad_norm": 2621.924072265625, | |
| "learning_rate": 1.455e-06, | |
| "loss": 859.6516, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 4.373626373626374, | |
| "grad_norm": 2298.13671875, | |
| "learning_rate": 1.4925000000000001e-06, | |
| "loss": 835.168, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.483516483516484, | |
| "grad_norm": 2228.95458984375, | |
| "learning_rate": 1.53e-06, | |
| "loss": 817.8176, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 4.593406593406593, | |
| "grad_norm": 2075.57568359375, | |
| "learning_rate": 1.5675e-06, | |
| "loss": 795.5578, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.7032967032967035, | |
| "grad_norm": 2308.193603515625, | |
| "learning_rate": 1.605e-06, | |
| "loss": 769.8123, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 4.813186813186813, | |
| "grad_norm": 2495.91259765625, | |
| "learning_rate": 1.6425e-06, | |
| "loss": 745.4971, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 2490.6796875, | |
| "learning_rate": 1.68e-06, | |
| "loss": 724.5927, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 5.021978021978022, | |
| "grad_norm": 1986.7320556640625, | |
| "learning_rate": 1.7175e-06, | |
| "loss": 711.5865, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 5.131868131868132, | |
| "grad_norm": 1901.4664306640625, | |
| "learning_rate": 1.7550000000000001e-06, | |
| "loss": 692.5993, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 5.241758241758242, | |
| "grad_norm": 2567.854248046875, | |
| "learning_rate": 1.7925e-06, | |
| "loss": 665.2945, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.351648351648351, | |
| "grad_norm": 1668.3482666015625, | |
| "learning_rate": 1.83e-06, | |
| "loss": 648.977, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 5.461538461538462, | |
| "grad_norm": 1845.254150390625, | |
| "learning_rate": 1.8675000000000001e-06, | |
| "loss": 627.8005, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.571428571428571, | |
| "grad_norm": 1811.66845703125, | |
| "learning_rate": 1.905e-06, | |
| "loss": 606.1213, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 5.681318681318682, | |
| "grad_norm": 1720.802734375, | |
| "learning_rate": 1.9425e-06, | |
| "loss": 587.5198, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.791208791208791, | |
| "grad_norm": 1535.228759765625, | |
| "learning_rate": 1.98e-06, | |
| "loss": 570.1765, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 5.9010989010989015, | |
| "grad_norm": 1655.658447265625, | |
| "learning_rate": 2.0175e-06, | |
| "loss": 551.0042, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3979.0732421875, | |
| "learning_rate": 2.0550000000000002e-06, | |
| "loss": 535.4148, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 6.1098901098901095, | |
| "grad_norm": 1727.4771728515625, | |
| "learning_rate": 2.0925000000000003e-06, | |
| "loss": 519.8219, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 6.21978021978022, | |
| "grad_norm": 1350.4666748046875, | |
| "learning_rate": 2.13e-06, | |
| "loss": 510.5004, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 6.329670329670329, | |
| "grad_norm": 2242.578857421875, | |
| "learning_rate": 2.1675e-06, | |
| "loss": 497.497, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 6.43956043956044, | |
| "grad_norm": 1353.7908935546875, | |
| "learning_rate": 2.205e-06, | |
| "loss": 476.5748, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 6.549450549450549, | |
| "grad_norm": 1527.2796630859375, | |
| "learning_rate": 2.2425e-06, | |
| "loss": 468.7679, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.65934065934066, | |
| "grad_norm": 1145.3853759765625, | |
| "learning_rate": 2.28e-06, | |
| "loss": 454.7695, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 6.769230769230769, | |
| "grad_norm": 1233.435302734375, | |
| "learning_rate": 2.3175e-06, | |
| "loss": 438.3474, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 6.8791208791208796, | |
| "grad_norm": 1669.4859619140625, | |
| "learning_rate": 2.355e-06, | |
| "loss": 424.2438, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 6.989010989010989, | |
| "grad_norm": 1667.7239990234375, | |
| "learning_rate": 2.3925e-06, | |
| "loss": 412.884, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 7.087912087912088, | |
| "grad_norm": 1087.6927490234375, | |
| "learning_rate": 2.43e-06, | |
| "loss": 399.3542, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 7.197802197802198, | |
| "grad_norm": 1415.59765625, | |
| "learning_rate": 2.4675e-06, | |
| "loss": 387.3436, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 7.3076923076923075, | |
| "grad_norm": 1952.5645751953125, | |
| "learning_rate": 2.505e-06, | |
| "loss": 377.6118, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 7.417582417582418, | |
| "grad_norm": 1404.5712890625, | |
| "learning_rate": 2.5425000000000002e-06, | |
| "loss": 371.8316, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 7.527472527472527, | |
| "grad_norm": 1185.5135498046875, | |
| "learning_rate": 2.58e-06, | |
| "loss": 363.4571, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 7.637362637362637, | |
| "grad_norm": 4727.9638671875, | |
| "learning_rate": 2.6175e-06, | |
| "loss": 354.3981, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 7.747252747252747, | |
| "grad_norm": 937.0252075195312, | |
| "learning_rate": 2.655e-06, | |
| "loss": 344.7615, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "grad_norm": 1180.9298095703125, | |
| "learning_rate": 2.6925e-06, | |
| "loss": 340.5888, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 7.967032967032967, | |
| "grad_norm": 1405.458984375, | |
| "learning_rate": 2.73e-06, | |
| "loss": 326.6432, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 8.065934065934066, | |
| "grad_norm": 936.8318481445312, | |
| "learning_rate": 2.7675e-06, | |
| "loss": 317.4954, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 8.175824175824175, | |
| "grad_norm": 888.236328125, | |
| "learning_rate": 2.8050000000000002e-06, | |
| "loss": 309.4422, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 8.285714285714286, | |
| "grad_norm": 970.7135620117188, | |
| "learning_rate": 2.8425e-06, | |
| "loss": 301.2726, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 8.395604395604396, | |
| "grad_norm": 1607.8035888671875, | |
| "learning_rate": 2.88e-06, | |
| "loss": 295.7361, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 8.505494505494505, | |
| "grad_norm": 814.48486328125, | |
| "learning_rate": 2.9175e-06, | |
| "loss": 290.0745, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 8.615384615384615, | |
| "grad_norm": 803.1909790039062, | |
| "learning_rate": 2.955e-06, | |
| "loss": 279.9774, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 8.725274725274724, | |
| "grad_norm": 1034.652099609375, | |
| "learning_rate": 2.9925e-06, | |
| "loss": 277.7123, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 8.835164835164836, | |
| "grad_norm": 993.9649658203125, | |
| "learning_rate": 3.0300000000000002e-06, | |
| "loss": 267.6938, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 8.945054945054945, | |
| "grad_norm": 1010.8787841796875, | |
| "learning_rate": 3.0675e-06, | |
| "loss": 266.7134, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 9.043956043956044, | |
| "grad_norm": 762.2285766601562, | |
| "learning_rate": 3.105e-06, | |
| "loss": 264.4567, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 9.153846153846153, | |
| "grad_norm": 5807.85791015625, | |
| "learning_rate": 3.1425e-06, | |
| "loss": 254.0208, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 9.263736263736265, | |
| "grad_norm": 717.2619018554688, | |
| "learning_rate": 3.18e-06, | |
| "loss": 249.0201, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 9.373626373626374, | |
| "grad_norm": 725.6996459960938, | |
| "learning_rate": 3.2175e-06, | |
| "loss": 243.3289, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 9.483516483516484, | |
| "grad_norm": 761.1790161132812, | |
| "learning_rate": 3.255e-06, | |
| "loss": 238.8994, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 9.593406593406593, | |
| "grad_norm": 792.3602905273438, | |
| "learning_rate": 3.2925000000000002e-06, | |
| "loss": 231.8852, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 9.703296703296703, | |
| "grad_norm": 744.7413330078125, | |
| "learning_rate": 3.3300000000000003e-06, | |
| "loss": 229.136, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 9.813186813186814, | |
| "grad_norm": 673.6207885742188, | |
| "learning_rate": 3.3675000000000004e-06, | |
| "loss": 221.0368, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 9.923076923076923, | |
| "grad_norm": 966.46630859375, | |
| "learning_rate": 3.405e-06, | |
| "loss": 217.8422, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 10.021978021978022, | |
| "grad_norm": 740.294921875, | |
| "learning_rate": 3.4425e-06, | |
| "loss": 228.7085, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 10.131868131868131, | |
| "grad_norm": 629.9981689453125, | |
| "learning_rate": 3.48e-06, | |
| "loss": 211.0439, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 10.241758241758241, | |
| "grad_norm": 809.6885375976562, | |
| "learning_rate": 3.5174999999999998e-06, | |
| "loss": 204.4649, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 10.351648351648352, | |
| "grad_norm": 1631.996337890625, | |
| "learning_rate": 3.555e-06, | |
| "loss": 204.0642, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 10.461538461538462, | |
| "grad_norm": 958.5594482421875, | |
| "learning_rate": 3.5925e-06, | |
| "loss": 199.6854, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 10.571428571428571, | |
| "grad_norm": 588.5241088867188, | |
| "learning_rate": 3.63e-06, | |
| "loss": 196.1547, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 10.68131868131868, | |
| "grad_norm": 1439.742919921875, | |
| "learning_rate": 3.6675e-06, | |
| "loss": 194.8075, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 10.791208791208792, | |
| "grad_norm": 547.2682495117188, | |
| "learning_rate": 3.705e-06, | |
| "loss": 190.5422, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 10.901098901098901, | |
| "grad_norm": 550.932373046875, | |
| "learning_rate": 3.7425e-06, | |
| "loss": 188.0797, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 2300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.326589641008087e+19, | |
| "train_batch_size": 24, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |