| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.992, | |
| "eval_steps": 500, | |
| "global_step": 390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 5.839018847816655, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.9618, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.044975028236581, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.9669, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.219047217466672, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.021, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.4838569963975985, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.9571, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.047643058048521, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.8816, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.2364546199415556, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.8964, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.06808539264327, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.8921, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.395847794140312, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.8709, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 5.136863261262288, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.9305, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.504381262936195, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.813, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.9344562843621285, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8366, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.14056368073828, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8205, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.2854369325531834, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.7825, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.55230766379796, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.7745, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.1163531503614808, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7378, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.7545006007454378, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7281, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.4874110741930364, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7435, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.6189725301930469, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.743, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.4046995785004956, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7486, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.9358645050027536, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7467, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.0721376654895611, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.6933, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.1516256662003364, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7067, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.9223029393111457, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7046, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.1474983271538444, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7125, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8955100949974366, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.6728, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 1.2251687166354326, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7381, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7816429553280688, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.6553, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.9763360804698599, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.6798, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.8354114194674345, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.6958, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.6887105015637863, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.6497, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.8779524514630032, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6844, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.7537997504917305, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6672, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.8385727211139176, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6744, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.7602171774049509, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6666, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.7678255126965418, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6642, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.647081056305973, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6638, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.8379083769655634, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6732, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.9149239688006771, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6624, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.9772104074784475, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6962, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.1116947880572285, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6245, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.7227423304768713, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6225, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.9725040218297323, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6604, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.955643555394975, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6133, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.7567970803940156, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6309, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.7996454927598492, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6659, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.7366005891361118, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6569, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.7885369085723203, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6313, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.6751447378852333, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6326, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.8571310551925032, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6201, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.7550362899021348, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6329, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.8280593896023649, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6422, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.8811902923700886, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6354, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 1.1095721931863598, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6667, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.6445072194387576, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6364, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.9734320449566246, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6218, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.6247852399111969, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6401, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.6158328565284048, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6168, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.6280970629663458, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6423, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.876342261245329, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6644, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.6782228770929252, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6425, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.6213554461131593, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6114, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.8970158114530625, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6523, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.6928394207466625, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6168, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.6361855672357039, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6177, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.7475080798399086, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6338, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.7201665058372186, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6004, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.6665201343517307, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6324, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.7733661737414401, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6246, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.6768671192755346, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6219, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.7097685593783307, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.623, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.6536798448840015, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6269, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.6614424957485971, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6351, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.6977238316575177, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6239, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.7934882507910082, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6573, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6575515131086959, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.5857, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.7723302740017406, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6344, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.6941667379950092, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.5906, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.8021010313212673, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6364, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.626971688308395, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 1.068, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.9308988989110186, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5233, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.9488953391338323, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5753, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.7366798820619407, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4947, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.820021588310417, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5649, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.6449418020815564, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.4996, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.9216006193216165, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5333, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.6725609871315669, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.4864, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.9531128581935571, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5237, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.6641456185314702, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5167, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.8823519514010172, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.4995, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5548207275448888, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5106, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.8288761814178127, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5217, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5061675211682982, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5033, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.7190426537289591, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5028, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.6128890949411272, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.4779, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.62309405530766, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5451, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.6334625248101861, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5274, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.5653360526439104, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.539, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.6224327796221331, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.4945, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.7081659483273709, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5272, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5845719830086954, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5066, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.6882406051818343, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5123, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6051334323383424, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5259, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.5879222709126353, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.473, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.5592006776867556, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5262, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5242844070496, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.4905, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5941812532382632, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5479, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5387721208985127, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.4948, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.5854840592046221, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.4785, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.522507074880471, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4754, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.5625209088620615, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.4869, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.43969509643876364, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.4972, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.5619860822540647, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4735, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.4421063815613501, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5055, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5369412032192468, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.4923, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.48910911632729354, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.512, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.5136909704915422, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.4981, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.48631677940560963, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4915, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.4840473244318828, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.4736, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.5064523309799703, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5403, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.589977688148516, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.4694, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.45092131277939423, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4979, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.5694759979218573, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5191, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5065392756850746, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.4805, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5466527945164157, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5199, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.7191169252596367, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5183, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.519742080744315, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5178, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.738483275841055, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4837, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.5192697289694931, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5115, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.7555019608159453, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5102, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.525986636066217, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5035, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.6237338996072936, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.4708, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.5347925953394986, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5117, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.47804164439498226, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.5006, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5097745208171317, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.4956, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.472002785073484, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5138, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.40000804145024127, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.452, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.452939045182372, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5132, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.4492202701835303, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.51, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.5217146312190998, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5055, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.45917544834665075, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5128, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.48608619608415743, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.4807, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.4153980294330535, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.4887, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.4635350891032439, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5051, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.5452295259370951, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5138, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.45699342832042916, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.472, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.5082490486526375, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.4954, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.6384690552060961, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.4987, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.4371303247604508, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.4848, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.41511391912882867, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.4665, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4249372081891107, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.4994, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.44138936645964416, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.4927, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.43910408608963586, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5066, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.43088729689416577, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.4687, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.4278640166834245, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5163, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.4742909576609465, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.471, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.4587718963346401, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5087, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.9966628173120351, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.8075, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6863182224101623, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.4051, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 2.0900892617168663, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3823, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.7960735194376195, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3444, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.655937861814063, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3576, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5911864199285094, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3585, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.6532713894850588, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3538, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.6592758774119294, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3908, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6444342767269957, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.4022, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.6495217170344411, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3443, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.5590722491328468, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3723, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5648244135782439, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3498, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.4820229598242042, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3497, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5561337750606294, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3677, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.5455054065895131, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3433, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.5916420617982933, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3971, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.6508258578592079, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3756, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.47618701005958114, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3403, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.5593921807214691, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3595, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4198057083178645, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3314, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.5110303368655525, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3606, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.5011779517833854, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3799, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.5246653684584236, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3464, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.4855556982779856, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.366, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.5305125098314, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3433, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.4970558826142017, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.36, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.499133817248476, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3337, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.4862098543776659, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3614, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.49489877938395055, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.364, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.47745055778943635, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3369, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.4518599993209865, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3397, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.49190428712388085, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3609, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.42695016108008466, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3403, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.48399579137589493, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3559, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.43535579894833265, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3699, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.42236836330089367, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3389, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.4254052985368455, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3748, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.4124026511474054, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3649, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4358743463877152, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3911, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.3787171361989274, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3594, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.402330181025512, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3528, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.3742935305188853, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.328, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.41761460975640535, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.393, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.3555221908620572, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3705, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.4632160156775803, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3333, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.3854892191886437, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3494, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.3637067298468189, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3405, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.40494042381970147, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3888, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.3755956967947316, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3255, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.4304305110638947, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3865, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.4062101646966682, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3513, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.4000767842561131, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3609, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.3945734513213183, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3219, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.42058336216796155, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3744, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.4132723550394805, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3658, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.39649934238958967, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3405, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.4208572250100482, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.368, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.3978701788145883, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3465, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.4001369867888692, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3532, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.4034806906167456, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.353, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.4307848928295819, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3569, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.3999064020664724, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3568, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.4090043678053143, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.365, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.42532758658903885, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3427, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.39419334493438113, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3772, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.4016749191362796, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3534, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.40847223940384836, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.369, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.4001323176748629, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3387, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.4248079947067711, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3771, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.3945282895021062, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3534, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.3497014179956054, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3563, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3986345065122307, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.376, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.370859934109157, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.339, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3747895589693441, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3758, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.3685623113569053, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.35, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.36104623960562526, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3553, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.40586050975133925, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3772, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.41959521563610697, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3486, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 1.1794707358197836, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.6203, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.6207295094733497, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2424, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.5155664642605005, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2496, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.841858506357091, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2474, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.6482412245876622, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2404, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4828258709239939, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.24, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.5795204158855689, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2274, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.5853029607020683, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.26, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.4828643470497328, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2544, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.4329500194495083, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2206, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.5165455521545157, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2417, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.5275383950137368, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2449, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.39340503477267014, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2199, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4317866921063053, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2351, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.4443169086208162, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2556, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.4005458806639983, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2365, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.4181384968270446, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.245, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.37836401384344864, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2344, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.3713195178763742, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2465, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.36043895430414663, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2338, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.34938091694810997, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.228, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.3625745323291279, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2532, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.32852523370479425, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2086, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3579171159966551, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2375, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.33573195229266234, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2216, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3678272441344075, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2438, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.33596560147428767, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2449, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.35108960586375987, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2298, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3505624560113376, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2163, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3504767757115922, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2189, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.34430891718788276, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2372, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.33778027498109425, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2133, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.33636617129539925, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2301, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.3504754683249517, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2486, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.34795897812025656, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2141, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.36580061553563076, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2304, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3701212255172201, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2431, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.3405446548694439, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2231, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.3553166437308296, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.219, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.37282631650361, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2347, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.32702491987456384, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2257, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.34275469394870306, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2367, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.35933691031419707, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2415, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.3571317614246153, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2333, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.34359614641158914, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2235, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.3521601969192374, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2573, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.3283557605845052, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2111, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.3608233567186079, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2451, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.33885700019167236, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2345, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.3558837243949352, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2463, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3417347616378375, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2451, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3265691533149346, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2292, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.35543632739891745, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2499, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.32721794505380586, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2284, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3473173577929293, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2477, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.3267429658875211, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2256, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3308565691014178, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2525, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.32775713637571846, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2286, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.3230110993671641, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2281, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.34319449892245124, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2574, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.3052306427404134, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2053, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.3389886313628059, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2561, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.3096027527849823, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2167, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3320520175361268, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2372, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.3406340780226723, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2545, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.2988000824396719, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2171, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.30931376951446365, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2231, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.33622255017983976, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2371, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.32360161834108225, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2348, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.30292782196099155, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2048, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.3090810070307293, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2562, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.29309856186061733, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.225, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.31956453050611927, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2279, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.3112087941940966, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2219, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.3153817831789216, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2354, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.31372150880773386, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2422, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.32510096573711394, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2316, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.29906942590577396, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2275, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 1.2139953461603725, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3807, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.5237410389498884, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.183, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.4427029532721314, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1775, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.34230370359902346, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.1509, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.29601757434559484, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.1573, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3814257012353634, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.1619, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.5041931343111281, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1693, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.5823550587186476, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1767, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.4625808142431024, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1588, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.37116127475914895, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.1782, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.3294610298238965, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.1661, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.300193661172271, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.1424, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.4038054890315199, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.1734, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.3513663152569111, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.1608, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.39787351845147634, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.158, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.33655397686316324, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.1596, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.318527802100109, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1704, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.2974865136489592, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1641, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.3016788088406415, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.1561, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.32849297882511297, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1561, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.33069183157158233, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1765, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.3540783313652625, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1731, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.3086978594001584, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1558, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.29331458561932094, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1557, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.3085758487601604, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.1572, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.2906868443213144, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.1557, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.2970918857148492, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1514, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.3173460879970613, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1704, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.2961358934681771, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1469, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.3130008308479644, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1632, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.30139540096728357, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1694, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.30497921291683516, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1515, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2962093457026487, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.1549, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.29145197933072053, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1691, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.2805215505614425, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.16, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.2716802696786645, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1572, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.29116806810849954, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1677, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.30403120086929486, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.1656, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.2888776980210721, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1575, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.28436489114232044, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1512, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.2945919286625033, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1719, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.2915528020141999, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.1525, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.29895984099909056, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.163, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.30792150165168203, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1621, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2909817524180524, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.1517, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2950182202862721, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1579, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2786025132207038, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1589, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.28027422258603657, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1516, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.2751820247938203, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1532, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.2734147517610915, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1528, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.28075210904691406, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.1584, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.2887601716240021, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1637, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.277925460841827, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.154, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.2844367789603032, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1517, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.28931672628419347, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1613, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.2860034658051635, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1563, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.3089159326788425, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1612, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.29457815167214546, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.168, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.27208637514557227, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.1472, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.27590131884272984, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1586, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.28602629295970033, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1578, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.28428689397441836, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1497, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.2836450607351416, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1584, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.2631863358532374, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1423, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.2746300466010014, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1533, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.2916563865381194, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1534, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2801924376684902, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1624, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.3066114462534377, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.1641, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.28936538978957893, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1585, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.29197481999400926, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.1606, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.27877119220344143, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1486, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.29040742011215104, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.172, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.27366147923175643, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1655, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.2928012431572296, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.178, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.28735601580788767, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1551, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2946376966049083, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1647, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2925440423275787, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1696, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2681299045611345, | |
| "learning_rate": 0.0, | |
| "loss": 0.1485, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 7.862516792374067e+17, | |
| "train_loss": 0.3952271275795423, | |
| "train_runtime": 13928.0051, | |
| "train_samples_per_second": 3.59, | |
| "train_steps_per_second": 0.028 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 390, | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.862516792374067e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |