| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 189, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015873015873015872, | |
| "grad_norm": 2.3559653063882835, | |
| "learning_rate": 0.0, | |
| "loss": 1.0469, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.031746031746031744, | |
| "grad_norm": 2.29659253609106, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.9929, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.047619047619047616, | |
| "grad_norm": 2.8899785663629123, | |
| "learning_rate": 6.25e-07, | |
| "loss": 1.0189, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06349206349206349, | |
| "grad_norm": 2.2527917765154153, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 0.9098, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.07936507936507936, | |
| "grad_norm": 2.2029116064708907, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.0462, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 2.122312207060731, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.9986, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 2.045608321522422, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.9554, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.12698412698412698, | |
| "grad_norm": 1.8619152723479657, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.9522, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 1.8663709940802706, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.9994, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.15873015873015872, | |
| "grad_norm": 1.756651050051264, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.9373, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1746031746031746, | |
| "grad_norm": 1.6604489437599113, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.8839, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 1.292095856227553, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.9907, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.20634920634920634, | |
| "grad_norm": 1.2693344466908103, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.9322, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 1.6344663181288221, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 1.0934, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 1.1511973531225708, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.9771, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.25396825396825395, | |
| "grad_norm": 1.1464903643465947, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 1.0442, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2698412698412698, | |
| "grad_norm": 1.1549964376534243, | |
| "learning_rate": 5e-06, | |
| "loss": 0.949, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 1.2909225755543452, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 1.0098, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.30158730158730157, | |
| "grad_norm": 1.0413894105842352, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.7468, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.31746031746031744, | |
| "grad_norm": 1.2863204827157997, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.9232, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.901486908411037, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.9885, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3492063492063492, | |
| "grad_norm": 1.0229681930848715, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 1.0508, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.36507936507936506, | |
| "grad_norm": 1.0772270927236638, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.8728, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 0.9393952588410857, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.8349, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3968253968253968, | |
| "grad_norm": 1.0822345499912303, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.0283, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4126984126984127, | |
| "grad_norm": 0.8124841375138875, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.8612, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.7868328056966778, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.7232, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.8283002452965974, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.9752, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4603174603174603, | |
| "grad_norm": 0.8117454313345658, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.9053, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 0.8594519052279771, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.9479, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.49206349206349204, | |
| "grad_norm": 0.9958158956912483, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.9188, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5079365079365079, | |
| "grad_norm": 0.8918575253813723, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.8847, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5238095238095238, | |
| "grad_norm": 0.8125329449215294, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0379, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5396825396825397, | |
| "grad_norm": 0.7108806956906407, | |
| "learning_rate": 9.999691920767945e-06, | |
| "loss": 0.8376, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.641926463787557, | |
| "learning_rate": 9.998767721036901e-06, | |
| "loss": 0.8241, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.7048636529194373, | |
| "learning_rate": 9.997227514697568e-06, | |
| "loss": 0.9693, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5873015873015873, | |
| "grad_norm": 0.6041864409794199, | |
| "learning_rate": 9.99507149155218e-06, | |
| "loss": 0.9839, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6031746031746031, | |
| "grad_norm": 0.6529381186048961, | |
| "learning_rate": 9.992299917291118e-06, | |
| "loss": 0.8479, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6190476190476191, | |
| "grad_norm": 0.7457758141141355, | |
| "learning_rate": 9.98891313346017e-06, | |
| "loss": 0.9095, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6349206349206349, | |
| "grad_norm": 0.6700791615416641, | |
| "learning_rate": 9.984911557418444e-06, | |
| "loss": 0.7685, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6507936507936508, | |
| "grad_norm": 0.6202447937301818, | |
| "learning_rate": 9.980295682286924e-06, | |
| "loss": 0.8387, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.6888680420644837, | |
| "learning_rate": 9.97506607688772e-06, | |
| "loss": 0.9107, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6825396825396826, | |
| "grad_norm": 0.5229452850388104, | |
| "learning_rate": 9.969223385673958e-06, | |
| "loss": 0.8308, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6984126984126984, | |
| "grad_norm": 0.5679326043532053, | |
| "learning_rate": 9.962768328650367e-06, | |
| "loss": 0.7516, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.5234412349262514, | |
| "learning_rate": 9.95570170128455e-06, | |
| "loss": 0.8443, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7301587301587301, | |
| "grad_norm": 0.5148736685750067, | |
| "learning_rate": 9.94802437440896e-06, | |
| "loss": 0.7959, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.746031746031746, | |
| "grad_norm": 0.6223703419413371, | |
| "learning_rate": 9.939737294113585e-06, | |
| "loss": 0.8964, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 0.4712938980573866, | |
| "learning_rate": 9.930841481629358e-06, | |
| "loss": 0.8884, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.6385581101993485, | |
| "learning_rate": 9.92133803320231e-06, | |
| "loss": 0.7817, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7936507936507936, | |
| "grad_norm": 0.47528095545287, | |
| "learning_rate": 9.91122811995848e-06, | |
| "loss": 0.819, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8095238095238095, | |
| "grad_norm": 0.5522186664203698, | |
| "learning_rate": 9.90051298775959e-06, | |
| "loss": 0.8691, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8253968253968254, | |
| "grad_norm": 0.3924890188917555, | |
| "learning_rate": 9.88919395704952e-06, | |
| "loss": 0.8259, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8412698412698413, | |
| "grad_norm": 0.5584015479821739, | |
| "learning_rate": 9.877272422691583e-06, | |
| "loss": 0.9318, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.5472693893320031, | |
| "learning_rate": 9.864749853796642e-06, | |
| "loss": 0.7983, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.873015873015873, | |
| "grad_norm": 0.5011856989250408, | |
| "learning_rate": 9.85162779354206e-06, | |
| "loss": 0.7289, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.48176520075987733, | |
| "learning_rate": 9.837907858981536e-06, | |
| "loss": 0.8795, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.9047619047619048, | |
| "grad_norm": 0.4693619944653085, | |
| "learning_rate": 9.823591740845831e-06, | |
| "loss": 0.8625, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9206349206349206, | |
| "grad_norm": 0.5158078748351012, | |
| "learning_rate": 9.808681203334416e-06, | |
| "loss": 0.7975, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9365079365079365, | |
| "grad_norm": 0.467299048377056, | |
| "learning_rate": 9.793178083898073e-06, | |
| "loss": 0.878, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 0.4360100853426926, | |
| "learning_rate": 9.777084293012448e-06, | |
| "loss": 0.842, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9682539682539683, | |
| "grad_norm": 0.4999196363033725, | |
| "learning_rate": 9.760401813942641e-06, | |
| "loss": 0.7661, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9841269841269841, | |
| "grad_norm": 0.49451715958225617, | |
| "learning_rate": 9.743132702498785e-06, | |
| "loss": 0.8685, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.51449429417728, | |
| "learning_rate": 9.725279086782719e-06, | |
| "loss": 0.7676, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.0158730158730158, | |
| "grad_norm": 0.5392465569053122, | |
| "learning_rate": 9.706843166925733e-06, | |
| "loss": 0.7978, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.0317460317460316, | |
| "grad_norm": 0.49426185655546884, | |
| "learning_rate": 9.687827214817433e-06, | |
| "loss": 0.8264, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.0476190476190477, | |
| "grad_norm": 0.5050909892528982, | |
| "learning_rate": 9.668233573825794e-06, | |
| "loss": 0.8898, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0634920634920635, | |
| "grad_norm": 0.45134127922296613, | |
| "learning_rate": 9.64806465850836e-06, | |
| "loss": 0.7317, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0793650793650793, | |
| "grad_norm": 0.5413266326970981, | |
| "learning_rate": 9.62732295431471e-06, | |
| "loss": 0.7307, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0952380952380953, | |
| "grad_norm": 0.4781316290575908, | |
| "learning_rate": 9.606011017280166e-06, | |
| "loss": 0.8977, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 0.5064140744904799, | |
| "learning_rate": 9.5841314737108e-06, | |
| "loss": 0.7648, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.126984126984127, | |
| "grad_norm": 0.5543523877170532, | |
| "learning_rate": 9.56168701985981e-06, | |
| "loss": 0.7995, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.4891764300467825, | |
| "learning_rate": 9.538680421595236e-06, | |
| "loss": 0.8072, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1587301587301586, | |
| "grad_norm": 0.48203192054287314, | |
| "learning_rate": 9.515114514059127e-06, | |
| "loss": 0.8128, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1746031746031746, | |
| "grad_norm": 0.499915788005329, | |
| "learning_rate": 9.490992201318165e-06, | |
| "loss": 0.7876, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1904761904761905, | |
| "grad_norm": 0.43129451868532453, | |
| "learning_rate": 9.466316456005783e-06, | |
| "loss": 0.7755, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.2063492063492063, | |
| "grad_norm": 0.49436944947590167, | |
| "learning_rate": 9.441090318955843e-06, | |
| "loss": 0.7015, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.2222222222222223, | |
| "grad_norm": 0.5018080177691097, | |
| "learning_rate": 9.415316898827923e-06, | |
| "loss": 0.7346, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.2380952380952381, | |
| "grad_norm": 0.42117192002428844, | |
| "learning_rate": 9.388999371724212e-06, | |
| "loss": 0.8242, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.253968253968254, | |
| "grad_norm": 0.47397540901194374, | |
| "learning_rate": 9.362140980798127e-06, | |
| "loss": 0.8928, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.2698412698412698, | |
| "grad_norm": 0.48823131897505534, | |
| "learning_rate": 9.334745035854646e-06, | |
| "loss": 0.7581, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.5170460810325518, | |
| "learning_rate": 9.306814912942445e-06, | |
| "loss": 0.8361, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.3015873015873016, | |
| "grad_norm": 0.41118521047488926, | |
| "learning_rate": 9.278354053937848e-06, | |
| "loss": 0.7794, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.3174603174603174, | |
| "grad_norm": 0.4827654705693697, | |
| "learning_rate": 9.249365966120692e-06, | |
| "loss": 0.8542, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.45176873751511454, | |
| "learning_rate": 9.219854221742106e-06, | |
| "loss": 0.8101, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3492063492063493, | |
| "grad_norm": 0.44526540495239475, | |
| "learning_rate": 9.189822457584311e-06, | |
| "loss": 0.7419, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3650793650793651, | |
| "grad_norm": 0.41133066066087726, | |
| "learning_rate": 9.159274374512444e-06, | |
| "loss": 0.6576, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.380952380952381, | |
| "grad_norm": 0.4500027229237173, | |
| "learning_rate": 9.128213737018493e-06, | |
| "loss": 0.8058, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3968253968253967, | |
| "grad_norm": 0.40834920107678924, | |
| "learning_rate": 9.096644372757393e-06, | |
| "loss": 0.8849, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.4126984126984126, | |
| "grad_norm": 0.5843795042717066, | |
| "learning_rate": 9.064570172075349e-06, | |
| "loss": 0.7969, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.5139681695756663, | |
| "learning_rate": 9.031995087530403e-06, | |
| "loss": 0.7983, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.4444444444444444, | |
| "grad_norm": 0.47799160571848326, | |
| "learning_rate": 8.99892313340537e-06, | |
| "loss": 0.6612, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4603174603174602, | |
| "grad_norm": 0.48090290795792257, | |
| "learning_rate": 8.96535838521314e-06, | |
| "loss": 0.8026, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.4761904761904763, | |
| "grad_norm": 0.48955363216016506, | |
| "learning_rate": 8.931304979194452e-06, | |
| "loss": 0.8051, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.492063492063492, | |
| "grad_norm": 0.47949685756309185, | |
| "learning_rate": 8.896767111808177e-06, | |
| "loss": 0.7354, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.507936507936508, | |
| "grad_norm": 0.5732670061875946, | |
| "learning_rate": 8.861749039214177e-06, | |
| "loss": 0.9129, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 0.48050508555262206, | |
| "learning_rate": 8.826255076748823e-06, | |
| "loss": 0.8445, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.5396825396825395, | |
| "grad_norm": 0.4329532952395629, | |
| "learning_rate": 8.790289598393186e-06, | |
| "loss": 0.7212, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5555555555555556, | |
| "grad_norm": 0.522751486773223, | |
| "learning_rate": 8.753857036234055e-06, | |
| "loss": 0.8149, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.4570961856172299, | |
| "learning_rate": 8.716961879917734e-06, | |
| "loss": 0.7365, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.5873015873015874, | |
| "grad_norm": 0.4363179134183329, | |
| "learning_rate": 8.679608676096793e-06, | |
| "loss": 0.8131, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6031746031746033, | |
| "grad_norm": 0.4655541415571893, | |
| "learning_rate": 8.641802027869774e-06, | |
| "loss": 0.7946, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.619047619047619, | |
| "grad_norm": 0.5743139418639736, | |
| "learning_rate": 8.603546594213935e-06, | |
| "loss": 0.8574, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6349206349206349, | |
| "grad_norm": 0.5267570867681096, | |
| "learning_rate": 8.564847089411128e-06, | |
| "loss": 0.8286, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6507936507936507, | |
| "grad_norm": 0.40799736834923667, | |
| "learning_rate": 8.525708282466839e-06, | |
| "loss": 0.8412, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.4236881481332967, | |
| "learning_rate": 8.486134996522502e-06, | |
| "loss": 0.8172, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.6825396825396826, | |
| "grad_norm": 0.5593679767726464, | |
| "learning_rate": 8.446132108261136e-06, | |
| "loss": 0.8058, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.6984126984126984, | |
| "grad_norm": 0.5031166228419733, | |
| "learning_rate": 8.405704547306379e-06, | |
| "loss": 0.8031, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.45322610730579044, | |
| "learning_rate": 8.364857295615006e-06, | |
| "loss": 0.8903, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.7301587301587302, | |
| "grad_norm": 0.5335556769284883, | |
| "learning_rate": 8.323595386862985e-06, | |
| "loss": 0.7925, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.746031746031746, | |
| "grad_norm": 0.4699718024263939, | |
| "learning_rate": 8.281923905825188e-06, | |
| "loss": 0.7664, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.7619047619047619, | |
| "grad_norm": 0.47207237316096745, | |
| "learning_rate": 8.23984798774876e-06, | |
| "loss": 0.7347, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.4532509556288616, | |
| "learning_rate": 8.197372817720314e-06, | |
| "loss": 0.7369, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.7936507936507935, | |
| "grad_norm": 0.5443221798521994, | |
| "learning_rate": 8.154503630026955e-06, | |
| "loss": 0.7261, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.8095238095238095, | |
| "grad_norm": 0.4456098920838456, | |
| "learning_rate": 8.111245707511253e-06, | |
| "loss": 0.7194, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8253968253968254, | |
| "grad_norm": 0.4159654938486175, | |
| "learning_rate": 8.067604380920228e-06, | |
| "loss": 0.7945, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8412698412698414, | |
| "grad_norm": 0.4706342532274064, | |
| "learning_rate": 8.023585028248435e-06, | |
| "loss": 0.8487, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.5701232470412769, | |
| "learning_rate": 7.979193074075216e-06, | |
| "loss": 0.8887, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.873015873015873, | |
| "grad_norm": 0.530430629054239, | |
| "learning_rate": 7.934433988896233e-06, | |
| "loss": 0.6534, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.48414840419963984, | |
| "learning_rate": 7.889313288449323e-06, | |
| "loss": 0.8214, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 0.4200926363513126, | |
| "learning_rate": 7.843836533034784e-06, | |
| "loss": 0.7614, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.9206349206349205, | |
| "grad_norm": 0.4941849127950555, | |
| "learning_rate": 7.798009326830167e-06, | |
| "loss": 0.7996, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.9365079365079365, | |
| "grad_norm": 0.41647477043231534, | |
| "learning_rate": 7.751837317199673e-06, | |
| "loss": 0.867, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.9523809523809523, | |
| "grad_norm": 0.4462896414872465, | |
| "learning_rate": 7.705326193998207e-06, | |
| "loss": 0.7547, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.9682539682539684, | |
| "grad_norm": 0.46366747032871125, | |
| "learning_rate": 7.658481688870218e-06, | |
| "loss": 0.7582, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.9841269841269842, | |
| "grad_norm": 0.4714130206121814, | |
| "learning_rate": 7.611309574543373e-06, | |
| "loss": 0.7606, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.46690869317456135, | |
| "learning_rate": 7.563815664117173e-06, | |
| "loss": 0.9121, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.015873015873016, | |
| "grad_norm": 0.8060769356732992, | |
| "learning_rate": 7.5160058103465985e-06, | |
| "loss": 0.7122, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.0317460317460316, | |
| "grad_norm": 0.5953210710991091, | |
| "learning_rate": 7.467885904920864e-06, | |
| "loss": 0.7567, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.0476190476190474, | |
| "grad_norm": 0.7560871980312371, | |
| "learning_rate": 7.419461877737373e-06, | |
| "loss": 0.8318, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.0634920634920633, | |
| "grad_norm": 2.1011598702400667, | |
| "learning_rate": 7.370739696170971e-06, | |
| "loss": 0.7428, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.0793650793650795, | |
| "grad_norm": 0.940827314326734, | |
| "learning_rate": 7.321725364338566e-06, | |
| "loss": 0.6161, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.0952380952380953, | |
| "grad_norm": 0.5381465457966281, | |
| "learning_rate": 7.272424922359246e-06, | |
| "loss": 0.6432, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.111111111111111, | |
| "grad_norm": 0.7212158226191104, | |
| "learning_rate": 7.222844445609931e-06, | |
| "loss": 0.7817, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.126984126984127, | |
| "grad_norm": 0.6031927565028607, | |
| "learning_rate": 7.172990043976703e-06, | |
| "loss": 0.7291, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.5554913039306149, | |
| "learning_rate": 7.122867861101868e-06, | |
| "loss": 0.7928, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.1587301587301586, | |
| "grad_norm": 0.4440614086169425, | |
| "learning_rate": 7.072484073626872e-06, | |
| "loss": 0.6864, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.1746031746031744, | |
| "grad_norm": 2.875286251781212, | |
| "learning_rate": 7.021844890431136e-06, | |
| "loss": 0.7627, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.1904761904761907, | |
| "grad_norm": 0.7248206082063566, | |
| "learning_rate": 6.970956551866925e-06, | |
| "loss": 0.728, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.2063492063492065, | |
| "grad_norm": 0.5344769870855947, | |
| "learning_rate": 6.9198253289903515e-06, | |
| "loss": 0.6621, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 0.5334409779130068, | |
| "learning_rate": 6.868457522788561e-06, | |
| "loss": 0.7351, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.238095238095238, | |
| "grad_norm": 0.4791675678917909, | |
| "learning_rate": 6.816859463403271e-06, | |
| "loss": 0.6568, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.253968253968254, | |
| "grad_norm": 0.5667962259074942, | |
| "learning_rate": 6.765037509350685e-06, | |
| "loss": 0.758, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.2698412698412698, | |
| "grad_norm": 0.523154654898243, | |
| "learning_rate": 6.7129980467379265e-06, | |
| "loss": 0.6657, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 0.45239550513741295, | |
| "learning_rate": 6.660747488476066e-06, | |
| "loss": 0.6615, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.3015873015873014, | |
| "grad_norm": 0.6580127713752147, | |
| "learning_rate": 6.608292273489851e-06, | |
| "loss": 0.6112, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.317460317460317, | |
| "grad_norm": 0.6033248382665617, | |
| "learning_rate": 6.555638865924221e-06, | |
| "loss": 0.7033, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.5578461067365529, | |
| "learning_rate": 6.502793754347721e-06, | |
| "loss": 0.7578, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.3492063492063493, | |
| "grad_norm": 0.5451762654132818, | |
| "learning_rate": 6.449763450952912e-06, | |
| "loss": 0.6863, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.365079365079365, | |
| "grad_norm": 0.4667833185680937, | |
| "learning_rate": 6.396554490753848e-06, | |
| "loss": 0.6825, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.380952380952381, | |
| "grad_norm": 0.4209933154088852, | |
| "learning_rate": 6.343173430780769e-06, | |
| "loss": 0.836, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.3968253968253967, | |
| "grad_norm": 0.46876037251704294, | |
| "learning_rate": 6.289626849272062e-06, | |
| "loss": 0.7981, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.4126984126984126, | |
| "grad_norm": 0.47367833829704725, | |
| "learning_rate": 6.2359213448636104e-06, | |
| "loss": 0.751, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.4285714285714284, | |
| "grad_norm": 0.43459439089398605, | |
| "learning_rate": 6.182063535775634e-06, | |
| "loss": 0.7654, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.4444444444444446, | |
| "grad_norm": 0.39767398947957067, | |
| "learning_rate": 6.1280600589971225e-06, | |
| "loss": 0.7896, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.4603174603174605, | |
| "grad_norm": 0.4231324131775063, | |
| "learning_rate": 6.073917569467934e-06, | |
| "loss": 0.8051, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.4761904761904763, | |
| "grad_norm": 0.3983830637612639, | |
| "learning_rate": 6.0196427392587085e-06, | |
| "loss": 0.7038, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.492063492063492, | |
| "grad_norm": 0.4585701856768339, | |
| "learning_rate": 5.96524225674865e-06, | |
| "loss": 0.7422, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.507936507936508, | |
| "grad_norm": 0.4299692751487169, | |
| "learning_rate": 5.9107228258013085e-06, | |
| "loss": 0.7053, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.5238095238095237, | |
| "grad_norm": 0.42827944956580943, | |
| "learning_rate": 5.856091164938451e-06, | |
| "loss": 0.6523, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.5396825396825395, | |
| "grad_norm": 0.4098750307712162, | |
| "learning_rate": 5.801354006512127e-06, | |
| "loss": 0.6895, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.5555555555555554, | |
| "grad_norm": 0.45624383692077836, | |
| "learning_rate": 5.746518095875033e-06, | |
| "loss": 0.6973, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 0.40961695420487504, | |
| "learning_rate": 5.6915901905492586e-06, | |
| "loss": 0.6285, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.5873015873015874, | |
| "grad_norm": 0.5946218628280344, | |
| "learning_rate": 5.6365770593935665e-06, | |
| "loss": 0.5907, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.6031746031746033, | |
| "grad_norm": 0.5401440035651196, | |
| "learning_rate": 5.581485481769231e-06, | |
| "loss": 0.7181, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.619047619047619, | |
| "grad_norm": 0.4378876946579892, | |
| "learning_rate": 5.526322246704628e-06, | |
| "loss": 0.7978, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.634920634920635, | |
| "grad_norm": 0.40853074862176036, | |
| "learning_rate": 5.471094152058592e-06, | |
| "loss": 0.681, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.6507936507936507, | |
| "grad_norm": 0.46389651051528763, | |
| "learning_rate": 5.415808003682717e-06, | |
| "loss": 0.7308, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.42992484133372394, | |
| "learning_rate": 5.360470614582661e-06, | |
| "loss": 0.7136, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.682539682539683, | |
| "grad_norm": 0.557870344379466, | |
| "learning_rate": 5.305088804078559e-06, | |
| "loss": 0.7333, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.6984126984126986, | |
| "grad_norm": 0.45453618737081114, | |
| "learning_rate": 5.249669396964665e-06, | |
| "loss": 0.6349, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "grad_norm": 0.4511080452383348, | |
| "learning_rate": 5.1942192226683385e-06, | |
| "loss": 0.776, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.7301587301587302, | |
| "grad_norm": 0.4900595891663431, | |
| "learning_rate": 5.138745114408427e-06, | |
| "loss": 0.5998, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.746031746031746, | |
| "grad_norm": 0.46419645481002475, | |
| "learning_rate": 5.083253908353193e-06, | |
| "loss": 0.6676, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.761904761904762, | |
| "grad_norm": 0.4905121964865482, | |
| "learning_rate": 5.0277524427778986e-06, | |
| "loss": 0.7831, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 0.4525848803424086, | |
| "learning_rate": 4.972247557222102e-06, | |
| "loss": 0.7164, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.7936507936507935, | |
| "grad_norm": 0.5368330661361714, | |
| "learning_rate": 4.916746091646808e-06, | |
| "loss": 0.6805, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.8095238095238093, | |
| "grad_norm": 0.4224136348005534, | |
| "learning_rate": 4.8612548855915755e-06, | |
| "loss": 0.724, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.825396825396825, | |
| "grad_norm": 0.4869146817578471, | |
| "learning_rate": 4.805780777331662e-06, | |
| "loss": 0.7446, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.8412698412698414, | |
| "grad_norm": 0.4511553539717499, | |
| "learning_rate": 4.750330603035336e-06, | |
| "loss": 0.7124, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.45286461188582156, | |
| "learning_rate": 4.694911195921443e-06, | |
| "loss": 0.7252, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.873015873015873, | |
| "grad_norm": 0.4293856541441545, | |
| "learning_rate": 4.6395293854173395e-06, | |
| "loss": 0.6053, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.888888888888889, | |
| "grad_norm": 0.5836445711298119, | |
| "learning_rate": 4.584191996317285e-06, | |
| "loss": 0.6828, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.9047619047619047, | |
| "grad_norm": 0.4710990644177235, | |
| "learning_rate": 4.528905847941411e-06, | |
| "loss": 0.8414, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.9206349206349205, | |
| "grad_norm": 0.4770727006845428, | |
| "learning_rate": 4.473677753295375e-06, | |
| "loss": 0.6592, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.9365079365079367, | |
| "grad_norm": 0.4087196329651188, | |
| "learning_rate": 4.418514518230769e-06, | |
| "loss": 0.7122, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.9523809523809526, | |
| "grad_norm": 0.48530953017287554, | |
| "learning_rate": 4.363422940606435e-06, | |
| "loss": 0.7454, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.9682539682539684, | |
| "grad_norm": 0.4900655757859956, | |
| "learning_rate": 4.308409809450742e-06, | |
| "loss": 0.7621, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.984126984126984, | |
| "grad_norm": 0.49992358594135816, | |
| "learning_rate": 4.253481904124968e-06, | |
| "loss": 0.7331, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.4385555753140365, | |
| "learning_rate": 4.198645993487872e-06, | |
| "loss": 0.604, | |
| "step": 189 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 315, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 47138450767872.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |