| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.8023255813953485, | |
| "eval_steps": 500, | |
| "global_step": 168, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03488372093023256, | |
| "grad_norm": 43.01280212402344, | |
| "learning_rate": 5.0000000000000004e-08, | |
| "loss": 2.7496, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06976744186046512, | |
| "grad_norm": 44.429630279541016, | |
| "learning_rate": 1.0000000000000001e-07, | |
| "loss": 2.7809, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.10465116279069768, | |
| "grad_norm": 42.754234313964844, | |
| "learning_rate": 1.5000000000000002e-07, | |
| "loss": 2.6968, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.13953488372093023, | |
| "grad_norm": 45.27891159057617, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 2.8076, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1744186046511628, | |
| "grad_norm": 43.920860290527344, | |
| "learning_rate": 2.5000000000000004e-07, | |
| "loss": 2.7255, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.20930232558139536, | |
| "grad_norm": 44.61790466308594, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 2.7178, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2441860465116279, | |
| "grad_norm": 44.28205871582031, | |
| "learning_rate": 3.5000000000000004e-07, | |
| "loss": 2.7193, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.27906976744186046, | |
| "grad_norm": 45.1516227722168, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 2.6846, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.313953488372093, | |
| "grad_norm": 46.8723258972168, | |
| "learning_rate": 4.5000000000000003e-07, | |
| "loss": 2.67, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3488372093023256, | |
| "grad_norm": 48.33848571777344, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 2.6368, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.38372093023255816, | |
| "grad_norm": 46.772193908691406, | |
| "learning_rate": 5.5e-07, | |
| "loss": 2.5562, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4186046511627907, | |
| "grad_norm": 48.12417984008789, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 2.5362, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.45348837209302323, | |
| "grad_norm": 49.0787239074707, | |
| "learning_rate": 6.5e-07, | |
| "loss": 2.3981, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.4883720930232558, | |
| "grad_norm": 52.367183685302734, | |
| "learning_rate": 7.000000000000001e-07, | |
| "loss": 2.3837, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5232558139534884, | |
| "grad_norm": 53.57649230957031, | |
| "learning_rate": 7.5e-07, | |
| "loss": 2.2885, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.5581395348837209, | |
| "grad_norm": 56.58591079711914, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 2.1447, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5930232558139535, | |
| "grad_norm": 55.064735412597656, | |
| "learning_rate": 8.500000000000001e-07, | |
| "loss": 2.0242, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.627906976744186, | |
| "grad_norm": 52.61149597167969, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 1.9263, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.6627906976744186, | |
| "grad_norm": 44.34925079345703, | |
| "learning_rate": 9.500000000000001e-07, | |
| "loss": 1.7466, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6976744186046512, | |
| "grad_norm": 36.627296447753906, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.6084, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7325581395348837, | |
| "grad_norm": 30.89563751220703, | |
| "learning_rate": 1.0500000000000001e-06, | |
| "loss": 1.4701, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.7674418604651163, | |
| "grad_norm": 29.17167091369629, | |
| "learning_rate": 1.1e-06, | |
| "loss": 1.4218, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8023255813953488, | |
| "grad_norm": 28.237022399902344, | |
| "learning_rate": 1.1500000000000002e-06, | |
| "loss": 1.3014, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.8372093023255814, | |
| "grad_norm": 28.778654098510742, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.1857, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.872093023255814, | |
| "grad_norm": 29.368289947509766, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.082, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9069767441860465, | |
| "grad_norm": 28.46448516845703, | |
| "learning_rate": 1.3e-06, | |
| "loss": 0.8716, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.9418604651162791, | |
| "grad_norm": 24.584318161010742, | |
| "learning_rate": 1.3500000000000002e-06, | |
| "loss": 0.6848, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.9767441860465116, | |
| "grad_norm": 21.166847229003906, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 0.5438, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 21.166847229003906, | |
| "learning_rate": 1.45e-06, | |
| "loss": 0.4084, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.0348837209302326, | |
| "grad_norm": 24.098440170288086, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.3545, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.069767441860465, | |
| "grad_norm": 12.6813325881958, | |
| "learning_rate": 1.5500000000000002e-06, | |
| "loss": 0.2673, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.1046511627906976, | |
| "grad_norm": 7.6524128913879395, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.2033, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.1395348837209303, | |
| "grad_norm": 5.23082160949707, | |
| "learning_rate": 1.6500000000000003e-06, | |
| "loss": 0.1887, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.1744186046511629, | |
| "grad_norm": 4.729929447174072, | |
| "learning_rate": 1.7000000000000002e-06, | |
| "loss": 0.1683, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.2093023255813953, | |
| "grad_norm": 5.831579208374023, | |
| "learning_rate": 1.75e-06, | |
| "loss": 0.1643, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.244186046511628, | |
| "grad_norm": 4.030057430267334, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 0.1528, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.2790697674418605, | |
| "grad_norm": 3.560523509979248, | |
| "learning_rate": 1.85e-06, | |
| "loss": 0.1422, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.3139534883720931, | |
| "grad_norm": 3.3749780654907227, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 0.1327, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.3488372093023255, | |
| "grad_norm": 2.7184131145477295, | |
| "learning_rate": 1.9500000000000004e-06, | |
| "loss": 0.1199, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.3837209302325582, | |
| "grad_norm": 2.8681583404541016, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.1107, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.4186046511627908, | |
| "grad_norm": 2.8731987476348877, | |
| "learning_rate": 2.05e-06, | |
| "loss": 0.1063, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.4534883720930232, | |
| "grad_norm": 2.9522054195404053, | |
| "learning_rate": 2.1000000000000002e-06, | |
| "loss": 0.0989, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.4883720930232558, | |
| "grad_norm": 2.689038038253784, | |
| "learning_rate": 2.15e-06, | |
| "loss": 0.0949, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.5232558139534884, | |
| "grad_norm": 2.6328952312469482, | |
| "learning_rate": 2.2e-06, | |
| "loss": 0.1006, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.558139534883721, | |
| "grad_norm": 2.7056033611297607, | |
| "learning_rate": 2.25e-06, | |
| "loss": 0.0929, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.5930232558139537, | |
| "grad_norm": 2.568206310272217, | |
| "learning_rate": 2.3000000000000004e-06, | |
| "loss": 0.0881, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.627906976744186, | |
| "grad_norm": 2.714211940765381, | |
| "learning_rate": 2.35e-06, | |
| "loss": 0.096, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.6627906976744184, | |
| "grad_norm": 2.3650710582733154, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.086, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.697674418604651, | |
| "grad_norm": 2.1763498783111572, | |
| "learning_rate": 2.4500000000000003e-06, | |
| "loss": 0.0858, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.7325581395348837, | |
| "grad_norm": 2.124727487564087, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0843, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.7674418604651163, | |
| "grad_norm": 1.641269564628601, | |
| "learning_rate": 2.55e-06, | |
| "loss": 0.0803, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.802325581395349, | |
| "grad_norm": 1.5644842386245728, | |
| "learning_rate": 2.6e-06, | |
| "loss": 0.0779, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.8372093023255816, | |
| "grad_norm": 1.1404681205749512, | |
| "learning_rate": 2.6500000000000005e-06, | |
| "loss": 0.0765, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.872093023255814, | |
| "grad_norm": 0.8386123776435852, | |
| "learning_rate": 2.7000000000000004e-06, | |
| "loss": 0.0749, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.9069767441860463, | |
| "grad_norm": 1.8363338708877563, | |
| "learning_rate": 2.7500000000000004e-06, | |
| "loss": 0.0737, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.941860465116279, | |
| "grad_norm": 1.2105377912521362, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.0762, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.9767441860465116, | |
| "grad_norm": 1.5502218008041382, | |
| "learning_rate": 2.85e-06, | |
| "loss": 0.07, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.0641449689865112, | |
| "learning_rate": 2.9e-06, | |
| "loss": 0.0679, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.0348837209302326, | |
| "grad_norm": 0.9201306104660034, | |
| "learning_rate": 2.95e-06, | |
| "loss": 0.0677, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.0697674418604652, | |
| "grad_norm": 0.5951386094093323, | |
| "learning_rate": 3e-06, | |
| "loss": 0.0654, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.104651162790698, | |
| "grad_norm": 0.8307608962059021, | |
| "learning_rate": 3.05e-06, | |
| "loss": 0.0706, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.13953488372093, | |
| "grad_norm": 0.7110892534255981, | |
| "learning_rate": 3.1000000000000004e-06, | |
| "loss": 0.0665, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.1744186046511627, | |
| "grad_norm": 0.6766234040260315, | |
| "learning_rate": 3.1500000000000003e-06, | |
| "loss": 0.0618, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.2093023255813953, | |
| "grad_norm": 0.3967410922050476, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.0644, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.244186046511628, | |
| "grad_norm": 0.3713420331478119, | |
| "learning_rate": 3.2500000000000002e-06, | |
| "loss": 0.0607, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.2790697674418605, | |
| "grad_norm": 0.5613359212875366, | |
| "learning_rate": 3.3000000000000006e-06, | |
| "loss": 0.0623, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.313953488372093, | |
| "grad_norm": 0.5458635687828064, | |
| "learning_rate": 3.3500000000000005e-06, | |
| "loss": 0.0563, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.3488372093023258, | |
| "grad_norm": 0.4196176528930664, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.0547, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.383720930232558, | |
| "grad_norm": 0.5757117867469788, | |
| "learning_rate": 3.45e-06, | |
| "loss": 0.0551, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.4186046511627906, | |
| "grad_norm": 0.45777687430381775, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.0575, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.453488372093023, | |
| "grad_norm": 0.5204553008079529, | |
| "learning_rate": 3.5500000000000003e-06, | |
| "loss": 0.057, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.488372093023256, | |
| "grad_norm": 0.6110821962356567, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.0556, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.5232558139534884, | |
| "grad_norm": 0.45246991515159607, | |
| "learning_rate": 3.65e-06, | |
| "loss": 0.0564, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.558139534883721, | |
| "grad_norm": 0.441976398229599, | |
| "learning_rate": 3.7e-06, | |
| "loss": 0.0534, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.5930232558139537, | |
| "grad_norm": 0.46637651324272156, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.0498, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.6279069767441863, | |
| "grad_norm": 0.482038289308548, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 0.0521, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.6627906976744184, | |
| "grad_norm": 0.5914385318756104, | |
| "learning_rate": 3.85e-06, | |
| "loss": 0.0511, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.697674418604651, | |
| "grad_norm": 0.34532907605171204, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 0.0527, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.7325581395348837, | |
| "grad_norm": 0.35315006971359253, | |
| "learning_rate": 3.95e-06, | |
| "loss": 0.0486, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.7674418604651163, | |
| "grad_norm": 0.4521324634552002, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.0485, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.802325581395349, | |
| "grad_norm": 0.49457868933677673, | |
| "learning_rate": 4.05e-06, | |
| "loss": 0.0525, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.8372093023255816, | |
| "grad_norm": 0.7616601586341858, | |
| "learning_rate": 4.1e-06, | |
| "loss": 0.0528, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.8720930232558137, | |
| "grad_norm": 0.4791123569011688, | |
| "learning_rate": 4.15e-06, | |
| "loss": 0.0512, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.9069767441860463, | |
| "grad_norm": 0.38186997175216675, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 0.0442, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.941860465116279, | |
| "grad_norm": 0.32648414373397827, | |
| "learning_rate": 4.25e-06, | |
| "loss": 0.0432, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.9767441860465116, | |
| "grad_norm": 0.410888671875, | |
| "learning_rate": 4.3e-06, | |
| "loss": 0.042, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.4732547104358673, | |
| "learning_rate": 4.350000000000001e-06, | |
| "loss": 0.0483, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.0348837209302326, | |
| "grad_norm": 0.6372231841087341, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.0382, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.0697674418604652, | |
| "grad_norm": 0.393078088760376, | |
| "learning_rate": 4.450000000000001e-06, | |
| "loss": 0.0403, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.104651162790698, | |
| "grad_norm": 0.4366248846054077, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.0367, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.13953488372093, | |
| "grad_norm": 0.301724374294281, | |
| "learning_rate": 4.5500000000000005e-06, | |
| "loss": 0.0351, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.1744186046511627, | |
| "grad_norm": 0.37665656208992004, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 0.03, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.2093023255813953, | |
| "grad_norm": 0.581331729888916, | |
| "learning_rate": 4.65e-06, | |
| "loss": 0.0316, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.244186046511628, | |
| "grad_norm": 0.44136878848075867, | |
| "learning_rate": 4.7e-06, | |
| "loss": 0.0312, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.2790697674418605, | |
| "grad_norm": 0.7624006271362305, | |
| "learning_rate": 4.75e-06, | |
| "loss": 0.0323, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.313953488372093, | |
| "grad_norm": 0.35294386744499207, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.0286, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.3488372093023258, | |
| "grad_norm": 0.6240035891532898, | |
| "learning_rate": 4.85e-06, | |
| "loss": 0.0289, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.383720930232558, | |
| "grad_norm": 0.40580570697784424, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 0.0259, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.4186046511627906, | |
| "grad_norm": 0.2971636652946472, | |
| "learning_rate": 4.95e-06, | |
| "loss": 0.0251, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.453488372093023, | |
| "grad_norm": 0.3758476972579956, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0244, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.488372093023256, | |
| "grad_norm": 0.3845921754837036, | |
| "learning_rate": 4.997332437005932e-06, | |
| "loss": 0.0242, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.5232558139534884, | |
| "grad_norm": 0.48894616961479187, | |
| "learning_rate": 4.989335440737587e-06, | |
| "loss": 0.0252, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.558139534883721, | |
| "grad_norm": 0.36844325065612793, | |
| "learning_rate": 4.976026077188013e-06, | |
| "loss": 0.0197, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.5930232558139537, | |
| "grad_norm": 0.2800253629684448, | |
| "learning_rate": 4.957432749209755e-06, | |
| "loss": 0.0187, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.6279069767441863, | |
| "grad_norm": 0.34042924642562866, | |
| "learning_rate": 4.933595135901733e-06, | |
| "loss": 0.0182, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.6627906976744184, | |
| "grad_norm": 0.33217477798461914, | |
| "learning_rate": 4.904564107932048e-06, | |
| "loss": 0.0166, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.697674418604651, | |
| "grad_norm": 0.32867398858070374, | |
| "learning_rate": 4.870401618977415e-06, | |
| "loss": 0.017, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.7325581395348837, | |
| "grad_norm": 0.24199356138706207, | |
| "learning_rate": 4.83118057351089e-06, | |
| "loss": 0.0138, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.7674418604651163, | |
| "grad_norm": 0.3216392695903778, | |
| "learning_rate": 4.786984671220053e-06, | |
| "loss": 0.0171, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.802325581395349, | |
| "grad_norm": 0.3574189841747284, | |
| "learning_rate": 4.737908228387656e-06, | |
| "loss": 0.0146, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.8372093023255816, | |
| "grad_norm": 0.30931738018989563, | |
| "learning_rate": 4.684055976615924e-06, | |
| "loss": 0.0128, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.8720930232558137, | |
| "grad_norm": 0.2074785828590393, | |
| "learning_rate": 4.625542839324036e-06, | |
| "loss": 0.0089, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.9069767441860463, | |
| "grad_norm": 0.28993067145347595, | |
| "learning_rate": 4.562493686495756e-06, | |
| "loss": 0.0088, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.941860465116279, | |
| "grad_norm": 0.21493716537952423, | |
| "learning_rate": 4.4950430682005995e-06, | |
| "loss": 0.0092, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.9767441860465116, | |
| "grad_norm": 0.4238521158695221, | |
| "learning_rate": 4.423334927457198e-06, | |
| "loss": 0.0096, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.4238521158695221, | |
| "learning_rate": 4.3475222930516484e-06, | |
| "loss": 0.0092, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.034883720930233, | |
| "grad_norm": 0.31130537390708923, | |
| "learning_rate": 4.267766952966369e-06, | |
| "loss": 0.0052, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.069767441860465, | |
| "grad_norm": 0.19592078030109406, | |
| "learning_rate": 4.184239109116393e-06, | |
| "loss": 0.0052, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.104651162790698, | |
| "grad_norm": 0.1717974841594696, | |
| "learning_rate": 4.097117014129903e-06, | |
| "loss": 0.0041, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.1395348837209305, | |
| "grad_norm": 0.15367873013019562, | |
| "learning_rate": 4.006586590948141e-06, | |
| "loss": 0.0061, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.174418604651163, | |
| "grad_norm": 0.19444844126701355, | |
| "learning_rate": 3.91284103605648e-06, | |
| "loss": 0.0072, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.209302325581396, | |
| "grad_norm": 0.22090725600719452, | |
| "learning_rate": 3.81608040719339e-06, | |
| "loss": 0.0041, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.2441860465116275, | |
| "grad_norm": 0.13613103330135345, | |
| "learning_rate": 3.7165111964171407e-06, | |
| "loss": 0.0042, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.27906976744186, | |
| "grad_norm": 0.5199090838432312, | |
| "learning_rate": 3.6143458894413463e-06, | |
| "loss": 0.0056, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.313953488372093, | |
| "grad_norm": 0.2258583903312683, | |
| "learning_rate": 3.5098025121797375e-06, | |
| "loss": 0.004, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.348837209302325, | |
| "grad_norm": 0.2459285408258438, | |
| "learning_rate": 3.403104165467883e-06, | |
| "loss": 0.0045, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 4.383720930232558, | |
| "grad_norm": 0.18142816424369812, | |
| "learning_rate": 3.2944785489547544e-06, | |
| "loss": 0.0043, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 4.4186046511627906, | |
| "grad_norm": 0.14586737751960754, | |
| "learning_rate": 3.184157475180208e-06, | |
| "loss": 0.0027, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 4.453488372093023, | |
| "grad_norm": 0.17918141186237335, | |
| "learning_rate": 3.0723763748753354e-06, | |
| "loss": 0.0048, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 4.488372093023256, | |
| "grad_norm": 0.23981280624866486, | |
| "learning_rate": 2.9593737945414264e-06, | |
| "loss": 0.0048, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.523255813953488, | |
| "grad_norm": 0.20855571329593658, | |
| "learning_rate": 2.845390887379706e-06, | |
| "loss": 0.002, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.558139534883721, | |
| "grad_norm": 0.15569192171096802, | |
| "learning_rate": 2.730670898658255e-06, | |
| "loss": 0.0053, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.593023255813954, | |
| "grad_norm": 0.17529034614562988, | |
| "learning_rate": 2.6154586466143495e-06, | |
| "loss": 0.0037, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.627906976744186, | |
| "grad_norm": 0.2721562683582306, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0047, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.662790697674419, | |
| "grad_norm": 0.15811768174171448, | |
| "learning_rate": 2.3845413533856517e-06, | |
| "loss": 0.0031, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.6976744186046515, | |
| "grad_norm": 0.3956407904624939, | |
| "learning_rate": 2.269329101341745e-06, | |
| "loss": 0.0048, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.732558139534884, | |
| "grad_norm": 1.2278478145599365, | |
| "learning_rate": 2.1546091126202955e-06, | |
| "loss": 0.0036, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.767441860465116, | |
| "grad_norm": 0.16240213811397552, | |
| "learning_rate": 2.040626205458574e-06, | |
| "loss": 0.0032, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.8023255813953485, | |
| "grad_norm": 0.15741980075836182, | |
| "learning_rate": 1.9276236251246655e-06, | |
| "loss": 0.0027, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.837209302325581, | |
| "grad_norm": 0.15352095663547516, | |
| "learning_rate": 1.8158425248197931e-06, | |
| "loss": 0.0021, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.872093023255814, | |
| "grad_norm": 0.16562862694263458, | |
| "learning_rate": 1.7055214510452462e-06, | |
| "loss": 0.0017, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.906976744186046, | |
| "grad_norm": 0.07386556267738342, | |
| "learning_rate": 1.5968958345321178e-06, | |
| "loss": 0.0028, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.941860465116279, | |
| "grad_norm": 0.11063098907470703, | |
| "learning_rate": 1.490197487820263e-06, | |
| "loss": 0.0022, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.976744186046512, | |
| "grad_norm": 0.1346900910139084, | |
| "learning_rate": 1.3856541105586545e-06, | |
| "loss": 0.0021, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.11798722296953201, | |
| "learning_rate": 1.2834888035828597e-06, | |
| "loss": 0.0031, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 5.034883720930233, | |
| "grad_norm": 0.18651635944843292, | |
| "learning_rate": 1.1839195928066101e-06, | |
| "loss": 0.002, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 5.069767441860465, | |
| "grad_norm": 0.10228271782398224, | |
| "learning_rate": 1.0871589639435204e-06, | |
| "loss": 0.0019, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 5.104651162790698, | |
| "grad_norm": 0.12258665263652802, | |
| "learning_rate": 9.934134090518593e-07, | |
| "loss": 0.002, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 5.1395348837209305, | |
| "grad_norm": 0.0909138098359108, | |
| "learning_rate": 9.028829858700974e-07, | |
| "loss": 0.0019, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 5.174418604651163, | |
| "grad_norm": 0.1315290331840515, | |
| "learning_rate": 8.157608908836071e-07, | |
| "loss": 0.0013, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.209302325581396, | |
| "grad_norm": 0.0621444433927536, | |
| "learning_rate": 7.322330470336314e-07, | |
| "loss": 0.0017, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 5.2441860465116275, | |
| "grad_norm": 0.11794009059667587, | |
| "learning_rate": 6.524777069483526e-07, | |
| "loss": 0.0016, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 5.27906976744186, | |
| "grad_norm": 0.07475598901510239, | |
| "learning_rate": 5.766650725428027e-07, | |
| "loss": 0.0024, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 5.313953488372093, | |
| "grad_norm": 0.12244551628828049, | |
| "learning_rate": 5.049569317994013e-07, | |
| "loss": 0.0021, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 5.348837209302325, | |
| "grad_norm": 0.12658093869686127, | |
| "learning_rate": 4.3750631350424456e-07, | |
| "loss": 0.0015, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 5.383720930232558, | |
| "grad_norm": 0.13053877651691437, | |
| "learning_rate": 3.7445716067596506e-07, | |
| "loss": 0.0019, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 5.4186046511627906, | |
| "grad_norm": 0.07347141206264496, | |
| "learning_rate": 3.1594402338407633e-07, | |
| "loss": 0.0014, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 5.453488372093023, | |
| "grad_norm": 0.09592650085687637, | |
| "learning_rate": 2.620917716123444e-07, | |
| "loss": 0.0026, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 5.488372093023256, | |
| "grad_norm": 0.10409682989120483, | |
| "learning_rate": 2.1301532877994747e-07, | |
| "loss": 0.0014, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 5.523255813953488, | |
| "grad_norm": 0.06076663359999657, | |
| "learning_rate": 1.6881942648911077e-07, | |
| "loss": 0.002, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 5.558139534883721, | |
| "grad_norm": 0.0755394697189331, | |
| "learning_rate": 1.2959838102258537e-07, | |
| "loss": 0.0023, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 5.593023255813954, | |
| "grad_norm": 0.05655893310904503, | |
| "learning_rate": 9.54358920679524e-08, | |
| "loss": 0.001, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 5.627906976744186, | |
| "grad_norm": 0.06078566238284111, | |
| "learning_rate": 6.640486409826785e-08, | |
| "loss": 0.0014, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 5.662790697674419, | |
| "grad_norm": 0.09118495136499405, | |
| "learning_rate": 4.256725079024554e-08, | |
| "loss": 0.0017, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.6976744186046515, | |
| "grad_norm": 0.10137756168842316, | |
| "learning_rate": 2.3973922811987295e-08, | |
| "loss": 0.0015, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.732558139534884, | |
| "grad_norm": 0.12966589629650116, | |
| "learning_rate": 1.0664559262413831e-08, | |
| "loss": 0.0034, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.767441860465116, | |
| "grad_norm": 0.1168215349316597, | |
| "learning_rate": 2.6675629940689508e-09, | |
| "loss": 0.0012, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 5.8023255813953485, | |
| "grad_norm": 0.08311488479375839, | |
| "learning_rate": 0.0, | |
| "loss": 0.0017, | |
| "step": 168 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 168, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 28, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4267371372895273e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |