| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.1060723371308483, | |
| "eval_steps": 500, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.011060723371308484, | |
| "grad_norm": 275.2231140136719, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 101.1244, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.022121446742616967, | |
| "grad_norm": 1128.2623291015625, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 123.0051, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03318217011392545, | |
| "grad_norm": 77.56747436523438, | |
| "learning_rate": 3e-06, | |
| "loss": 62.0741, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.044242893485233935, | |
| "grad_norm": 181.99227905273438, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 132.6745, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.05530361685654242, | |
| "grad_norm": 2387.94189453125, | |
| "learning_rate": 5e-06, | |
| "loss": 56.9033, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.0663643402278509, | |
| "grad_norm": 223.02427673339844, | |
| "learning_rate": 6e-06, | |
| "loss": 88.674, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.07742506359915939, | |
| "grad_norm": 1268.653076171875, | |
| "learning_rate": 7e-06, | |
| "loss": 29.1095, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.08848578697046787, | |
| "grad_norm": 813.103759765625, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 37.2381, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.09954651034177635, | |
| "grad_norm": 310.0917053222656, | |
| "learning_rate": 9e-06, | |
| "loss": 46.942, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.11060723371308484, | |
| "grad_norm": 384.6047058105469, | |
| "learning_rate": 1e-05, | |
| "loss": 54.5569, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.12166795708439332, | |
| "grad_norm": 653.60693359375, | |
| "learning_rate": 9.88888888888889e-06, | |
| "loss": 27.13, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.1327286804557018, | |
| "grad_norm": 169.83457946777344, | |
| "learning_rate": 9.777777777777779e-06, | |
| "loss": 30.8385, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.14378940382701028, | |
| "grad_norm": 365.39520263671875, | |
| "learning_rate": 9.666666666666667e-06, | |
| "loss": 41.3138, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.15485012719831878, | |
| "grad_norm": 589.5994873046875, | |
| "learning_rate": 9.555555555555556e-06, | |
| "loss": 30.9588, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.16591085056962726, | |
| "grad_norm": 278.96026611328125, | |
| "learning_rate": 9.444444444444445e-06, | |
| "loss": 30.8774, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.17697157394093574, | |
| "grad_norm": 308.6740417480469, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 34.5223, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.18803229731224422, | |
| "grad_norm": 290.6882629394531, | |
| "learning_rate": 9.222222222222224e-06, | |
| "loss": 73.9331, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.1990930206835527, | |
| "grad_norm": 162.80252075195312, | |
| "learning_rate": 9.111111111111112e-06, | |
| "loss": 75.7575, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.21015374405486117, | |
| "grad_norm": 614.7833251953125, | |
| "learning_rate": 9e-06, | |
| "loss": 46.2102, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.22121446742616968, | |
| "grad_norm": 90.11128997802734, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 32.4061, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.23227519079747816, | |
| "grad_norm": 466.54559326171875, | |
| "learning_rate": 8.777777777777778e-06, | |
| "loss": 40.7987, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.24333591416878664, | |
| "grad_norm": 103.32861328125, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 123.4267, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.2543966375400951, | |
| "grad_norm": 87.52120971679688, | |
| "learning_rate": 8.555555555555556e-06, | |
| "loss": 47.7221, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.2654573609114036, | |
| "grad_norm": 222.90623474121094, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 50.3554, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.2765180842827121, | |
| "grad_norm": 223.30189514160156, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 83.7681, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.28757880765402055, | |
| "grad_norm": 428.62408447265625, | |
| "learning_rate": 8.222222222222222e-06, | |
| "loss": 55.3063, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.29863953102532903, | |
| "grad_norm": 531.6271362304688, | |
| "learning_rate": 8.111111111111112e-06, | |
| "loss": 37.9162, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.30970025439663756, | |
| "grad_norm": 311.58929443359375, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 59.5817, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.32076097776794604, | |
| "grad_norm": 375.3035583496094, | |
| "learning_rate": 7.88888888888889e-06, | |
| "loss": 52.9975, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.3318217011392545, | |
| "grad_norm": 516.3969116210938, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 42.9525, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.342882424510563, | |
| "grad_norm": 379.7697448730469, | |
| "learning_rate": 7.666666666666667e-06, | |
| "loss": 81.1183, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.3539431478818715, | |
| "grad_norm": 312.02691650390625, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 70.5851, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.36500387125317996, | |
| "grad_norm": 13544.4599609375, | |
| "learning_rate": 7.444444444444445e-06, | |
| "loss": 89.4958, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.37606459462448844, | |
| "grad_norm": 272.00518798828125, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 84.8861, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.3871253179957969, | |
| "grad_norm": 2974.26416015625, | |
| "learning_rate": 7.222222222222223e-06, | |
| "loss": 38.6106, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.3981860413671054, | |
| "grad_norm": 1803.48095703125, | |
| "learning_rate": 7.111111111111112e-06, | |
| "loss": 40.9991, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.40924676473841387, | |
| "grad_norm": 4818.87890625, | |
| "learning_rate": 7e-06, | |
| "loss": 32.4676, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.42030748810972235, | |
| "grad_norm": 104.36139678955078, | |
| "learning_rate": 6.88888888888889e-06, | |
| "loss": 31.6814, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.4313682114810309, | |
| "grad_norm": 612.83349609375, | |
| "learning_rate": 6.777777777777779e-06, | |
| "loss": 34.3181, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.44242893485233936, | |
| "grad_norm": 288.1559143066406, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 41.178, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.45348965822364784, | |
| "grad_norm": 912.2635498046875, | |
| "learning_rate": 6.555555555555556e-06, | |
| "loss": 45.9302, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.4645503815949563, | |
| "grad_norm": 488.1215515136719, | |
| "learning_rate": 6.444444444444445e-06, | |
| "loss": 51.5342, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.4756111049662648, | |
| "grad_norm": 132.82635498046875, | |
| "learning_rate": 6.333333333333333e-06, | |
| "loss": 27.757, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.4866718283375733, | |
| "grad_norm": 564.8312377929688, | |
| "learning_rate": 6.222222222222223e-06, | |
| "loss": 31.4658, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.49773255170888175, | |
| "grad_norm": 306.7361755371094, | |
| "learning_rate": 6.111111111111112e-06, | |
| "loss": 43.3153, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5087932750801902, | |
| "grad_norm": 127.41687774658203, | |
| "learning_rate": 6e-06, | |
| "loss": 52.1603, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.5198539984514987, | |
| "grad_norm": 5578.05712890625, | |
| "learning_rate": 5.88888888888889e-06, | |
| "loss": 80.404, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.5309147218228072, | |
| "grad_norm": 163.3619384765625, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 51.359, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.5419754451941157, | |
| "grad_norm": 309.4814453125, | |
| "learning_rate": 5.666666666666667e-06, | |
| "loss": 32.5008, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.5530361685654241, | |
| "grad_norm": 1415.1243896484375, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 34.8322, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5640968919367326, | |
| "grad_norm": 229.50904846191406, | |
| "learning_rate": 5.444444444444445e-06, | |
| "loss": 50.1107, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.5751576153080411, | |
| "grad_norm": 123.28858184814453, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 39.4679, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.5862183386793496, | |
| "grad_norm": 48702.5703125, | |
| "learning_rate": 5.2222222222222226e-06, | |
| "loss": 65.5247, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.5972790620506581, | |
| "grad_norm": 1075.4512939453125, | |
| "learning_rate": 5.1111111111111115e-06, | |
| "loss": 41.7882, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.6083397854219667, | |
| "grad_norm": 150.42401123046875, | |
| "learning_rate": 5e-06, | |
| "loss": 67.5473, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6194005087932751, | |
| "grad_norm": 466.3522644042969, | |
| "learning_rate": 4.888888888888889e-06, | |
| "loss": 51.2565, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.6304612321645836, | |
| "grad_norm": 11005.3369140625, | |
| "learning_rate": 4.777777777777778e-06, | |
| "loss": 53.788, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.6415219555358921, | |
| "grad_norm": 1204.18798828125, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 91.1635, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.6525826789072006, | |
| "grad_norm": 224.9308624267578, | |
| "learning_rate": 4.555555555555556e-06, | |
| "loss": 31.2719, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.663643402278509, | |
| "grad_norm": 232.47068786621094, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 28.8651, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6747041256498175, | |
| "grad_norm": 159.796630859375, | |
| "learning_rate": 4.333333333333334e-06, | |
| "loss": 37.8374, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.685764849021126, | |
| "grad_norm": 763.4359130859375, | |
| "learning_rate": 4.222222222222223e-06, | |
| "loss": 46.3411, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.6968255723924345, | |
| "grad_norm": 339.2966003417969, | |
| "learning_rate": 4.111111111111111e-06, | |
| "loss": 48.4202, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.707886295763743, | |
| "grad_norm": 748.3837280273438, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 50.4229, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.7189470191350514, | |
| "grad_norm": 231.0708770751953, | |
| "learning_rate": 3.88888888888889e-06, | |
| "loss": 51.902, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7300077425063599, | |
| "grad_norm": 53.15177536010742, | |
| "learning_rate": 3.777777777777778e-06, | |
| "loss": 53.8197, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.7410684658776684, | |
| "grad_norm": 444.1663818359375, | |
| "learning_rate": 3.6666666666666666e-06, | |
| "loss": 27.7717, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.7521291892489769, | |
| "grad_norm": 1551.6669921875, | |
| "learning_rate": 3.555555555555556e-06, | |
| "loss": 45.1505, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.7631899126202853, | |
| "grad_norm": 339.9322509765625, | |
| "learning_rate": 3.444444444444445e-06, | |
| "loss": 40.7306, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.7742506359915938, | |
| "grad_norm": 152.95518493652344, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 41.1691, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7853113593629023, | |
| "grad_norm": 565.0484619140625, | |
| "learning_rate": 3.2222222222222227e-06, | |
| "loss": 40.6706, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.7963720827342108, | |
| "grad_norm": 254.9434814453125, | |
| "learning_rate": 3.1111111111111116e-06, | |
| "loss": 37.7834, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.8074328061055193, | |
| "grad_norm": 426.6705627441406, | |
| "learning_rate": 3e-06, | |
| "loss": 51.4484, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.8184935294768277, | |
| "grad_norm": 319.9366149902344, | |
| "learning_rate": 2.888888888888889e-06, | |
| "loss": 26.9082, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.8295542528481362, | |
| "grad_norm": 518.8477172851562, | |
| "learning_rate": 2.7777777777777783e-06, | |
| "loss": 31.8693, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8406149762194447, | |
| "grad_norm": 428.4947814941406, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 53.7889, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.8516756995907533, | |
| "grad_norm": 216.45521545410156, | |
| "learning_rate": 2.5555555555555557e-06, | |
| "loss": 35.5407, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.8627364229620618, | |
| "grad_norm": 180.7081298828125, | |
| "learning_rate": 2.4444444444444447e-06, | |
| "loss": 28.9896, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.8737971463333702, | |
| "grad_norm": 341.2665100097656, | |
| "learning_rate": 2.3333333333333336e-06, | |
| "loss": 48.7084, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.8848578697046787, | |
| "grad_norm": 24.655475616455078, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 47.937, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8959185930759872, | |
| "grad_norm": 244.29147338867188, | |
| "learning_rate": 2.1111111111111114e-06, | |
| "loss": 27.4093, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.9069793164472957, | |
| "grad_norm": 308.5469665527344, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 62.8975, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.9180400398186042, | |
| "grad_norm": 137.78431701660156, | |
| "learning_rate": 1.888888888888889e-06, | |
| "loss": 71.2175, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.9291007631899126, | |
| "grad_norm": 346.1696472167969, | |
| "learning_rate": 1.777777777777778e-06, | |
| "loss": 74.5859, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.9401614865612211, | |
| "grad_norm": 605.5797729492188, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 31.0377, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9512222099325296, | |
| "grad_norm": 82.16192626953125, | |
| "learning_rate": 1.5555555555555558e-06, | |
| "loss": 45.4671, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.9622829333038381, | |
| "grad_norm": 1627.314453125, | |
| "learning_rate": 1.4444444444444445e-06, | |
| "loss": 60.7454, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.9733436566751466, | |
| "grad_norm": 141.41053771972656, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 43.7195, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.984404380046455, | |
| "grad_norm": 60.2154541015625, | |
| "learning_rate": 1.2222222222222223e-06, | |
| "loss": 37.6804, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.9954651034177635, | |
| "grad_norm": 164.14305114746094, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 30.995, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 120.65478515625, | |
| "eval_runtime": 20.9653, | |
| "eval_samples_per_second": 95.873, | |
| "eval_steps_per_second": 47.936, | |
| "step": 9041 | |
| }, | |
| { | |
| "epoch": 1.006525826789072, | |
| "grad_norm": 866.4227294921875, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 62.0281, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.0175865501603805, | |
| "grad_norm": 186.7490234375, | |
| "learning_rate": 8.88888888888889e-07, | |
| "loss": 61.999, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.028647273531689, | |
| "grad_norm": 120.41979217529297, | |
| "learning_rate": 7.777777777777779e-07, | |
| "loss": 39.2587, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.0397079969029974, | |
| "grad_norm": 188.38967895507812, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 34.3708, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.050768720274306, | |
| "grad_norm": 429.96307373046875, | |
| "learning_rate": 5.555555555555555e-07, | |
| "loss": 27.8105, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0618294436456144, | |
| "grad_norm": 321.49444580078125, | |
| "learning_rate": 4.444444444444445e-07, | |
| "loss": 27.4738, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.0728901670169229, | |
| "grad_norm": 232.31663513183594, | |
| "learning_rate": 3.3333333333333335e-07, | |
| "loss": 60.7941, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.0839508903882313, | |
| "grad_norm": 315.6164855957031, | |
| "learning_rate": 2.2222222222222224e-07, | |
| "loss": 27.5082, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.0950116137595398, | |
| "grad_norm": 841.2481689453125, | |
| "learning_rate": 1.1111111111111112e-07, | |
| "loss": 37.086, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.1060723371308483, | |
| "grad_norm": 104.0994873046875, | |
| "learning_rate": 0.0, | |
| "loss": 33.8564, | |
| "step": 10000 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |