| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.6510416666666666, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0013020833333333333, | |
| "grad_norm": 0.6808765530586243, | |
| "learning_rate": 2.02e-06, | |
| "loss": 1.5302, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0026041666666666665, | |
| "grad_norm": 0.7672109007835388, | |
| "learning_rate": 4.04e-06, | |
| "loss": 1.5499, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00390625, | |
| "grad_norm": 0.7675047516822815, | |
| "learning_rate": 6.06e-06, | |
| "loss": 1.6097, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.005208333333333333, | |
| "grad_norm": 0.6575904488563538, | |
| "learning_rate": 8.08e-06, | |
| "loss": 1.5117, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.006510416666666667, | |
| "grad_norm": 0.7006857991218567, | |
| "learning_rate": 1.0100000000000002e-05, | |
| "loss": 1.618, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0078125, | |
| "grad_norm": 0.8015255928039551, | |
| "learning_rate": 1.212e-05, | |
| "loss": 1.538, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.009114583333333334, | |
| "grad_norm": 0.7574004530906677, | |
| "learning_rate": 1.4140000000000002e-05, | |
| "loss": 1.5707, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.010416666666666666, | |
| "grad_norm": 0.7140095233917236, | |
| "learning_rate": 1.616e-05, | |
| "loss": 1.4702, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01171875, | |
| "grad_norm": 0.770279049873352, | |
| "learning_rate": 1.818e-05, | |
| "loss": 1.5559, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.013020833333333334, | |
| "grad_norm": 0.7535017728805542, | |
| "learning_rate": 2.0200000000000003e-05, | |
| "loss": 1.5051, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.014322916666666666, | |
| "grad_norm": 0.7053647041320801, | |
| "learning_rate": 2.222e-05, | |
| "loss": 1.4195, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.015625, | |
| "grad_norm": 0.8053725361824036, | |
| "learning_rate": 2.424e-05, | |
| "loss": 1.4431, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.016927083333333332, | |
| "grad_norm": 0.7332000136375427, | |
| "learning_rate": 2.6260000000000003e-05, | |
| "loss": 1.3888, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.018229166666666668, | |
| "grad_norm": 0.7415592670440674, | |
| "learning_rate": 2.8280000000000004e-05, | |
| "loss": 1.3524, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.01953125, | |
| "grad_norm": 0.8849363327026367, | |
| "learning_rate": 3.0299999999999998e-05, | |
| "loss": 1.4205, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.020833333333333332, | |
| "grad_norm": 0.8366559147834778, | |
| "learning_rate": 3.232e-05, | |
| "loss": 1.3502, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.022135416666666668, | |
| "grad_norm": 0.8472776412963867, | |
| "learning_rate": 3.434e-05, | |
| "loss": 1.3127, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0234375, | |
| "grad_norm": 0.9324319958686829, | |
| "learning_rate": 3.636e-05, | |
| "loss": 1.3929, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.024739583333333332, | |
| "grad_norm": 0.7191261649131775, | |
| "learning_rate": 3.838e-05, | |
| "loss": 1.2803, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.026041666666666668, | |
| "grad_norm": 0.9117175936698914, | |
| "learning_rate": 4.0400000000000006e-05, | |
| "loss": 1.2158, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02734375, | |
| "grad_norm": 0.7863415479660034, | |
| "learning_rate": 4.242e-05, | |
| "loss": 1.273, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.028645833333333332, | |
| "grad_norm": 0.7084434032440186, | |
| "learning_rate": 4.444e-05, | |
| "loss": 1.2066, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.029947916666666668, | |
| "grad_norm": 0.695151686668396, | |
| "learning_rate": 4.6460000000000006e-05, | |
| "loss": 1.1198, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.03125, | |
| "grad_norm": 0.49544116854667664, | |
| "learning_rate": 4.848e-05, | |
| "loss": 1.1062, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.032552083333333336, | |
| "grad_norm": 0.47292080521583557, | |
| "learning_rate": 5.05e-05, | |
| "loss": 1.0123, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.033854166666666664, | |
| "grad_norm": 0.5713898539543152, | |
| "learning_rate": 5.2520000000000005e-05, | |
| "loss": 1.057, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.03515625, | |
| "grad_norm": 0.6084786653518677, | |
| "learning_rate": 5.454e-05, | |
| "loss": 1.0196, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.036458333333333336, | |
| "grad_norm": 0.6218786239624023, | |
| "learning_rate": 5.656000000000001e-05, | |
| "loss": 1.0767, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.037760416666666664, | |
| "grad_norm": 0.6100573539733887, | |
| "learning_rate": 5.858e-05, | |
| "loss": 1.0825, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0390625, | |
| "grad_norm": 0.5810414552688599, | |
| "learning_rate": 6.0599999999999996e-05, | |
| "loss": 1.0459, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.040364583333333336, | |
| "grad_norm": 0.48789265751838684, | |
| "learning_rate": 6.262000000000001e-05, | |
| "loss": 0.9882, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.041666666666666664, | |
| "grad_norm": 0.46344229578971863, | |
| "learning_rate": 6.464e-05, | |
| "loss": 1.0365, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.04296875, | |
| "grad_norm": 0.5264561772346497, | |
| "learning_rate": 6.666e-05, | |
| "loss": 1.0536, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.044270833333333336, | |
| "grad_norm": 0.5107188820838928, | |
| "learning_rate": 6.868e-05, | |
| "loss": 1.072, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.045572916666666664, | |
| "grad_norm": 0.49732786417007446, | |
| "learning_rate": 7.07e-05, | |
| "loss": 1.0793, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.046875, | |
| "grad_norm": 0.4549235999584198, | |
| "learning_rate": 7.272e-05, | |
| "loss": 0.9733, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.048177083333333336, | |
| "grad_norm": 0.4460265636444092, | |
| "learning_rate": 7.474e-05, | |
| "loss": 1.0077, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.049479166666666664, | |
| "grad_norm": 0.47252264618873596, | |
| "learning_rate": 7.676e-05, | |
| "loss": 1.0423, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.05078125, | |
| "grad_norm": 0.3926396071910858, | |
| "learning_rate": 7.878e-05, | |
| "loss": 0.9915, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.052083333333333336, | |
| "grad_norm": 0.4565116763114929, | |
| "learning_rate": 8.080000000000001e-05, | |
| "loss": 1.0371, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.053385416666666664, | |
| "grad_norm": 0.36098912358283997, | |
| "learning_rate": 8.282e-05, | |
| "loss": 0.9437, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.0546875, | |
| "grad_norm": 0.3953251242637634, | |
| "learning_rate": 8.484e-05, | |
| "loss": 0.9441, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.055989583333333336, | |
| "grad_norm": 0.4209212064743042, | |
| "learning_rate": 8.686e-05, | |
| "loss": 1.0181, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.057291666666666664, | |
| "grad_norm": 0.3656007647514343, | |
| "learning_rate": 8.888e-05, | |
| "loss": 1.0204, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.05859375, | |
| "grad_norm": 0.4179205894470215, | |
| "learning_rate": 9.09e-05, | |
| "loss": 0.9894, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.059895833333333336, | |
| "grad_norm": 0.4047180116176605, | |
| "learning_rate": 9.292000000000001e-05, | |
| "loss": 0.9743, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.061197916666666664, | |
| "grad_norm": 0.36672013998031616, | |
| "learning_rate": 9.494e-05, | |
| "loss": 0.9355, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.0625, | |
| "grad_norm": 0.4091084599494934, | |
| "learning_rate": 9.696e-05, | |
| "loss": 1.0145, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.06380208333333333, | |
| "grad_norm": 0.4033145606517792, | |
| "learning_rate": 9.898e-05, | |
| "loss": 0.8795, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.06510416666666667, | |
| "grad_norm": 0.4093867838382721, | |
| "learning_rate": 0.000101, | |
| "loss": 0.9118, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06640625, | |
| "grad_norm": 0.4064587354660034, | |
| "learning_rate": 0.00010302, | |
| "loss": 0.9846, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.06770833333333333, | |
| "grad_norm": 0.4279826283454895, | |
| "learning_rate": 0.00010504000000000001, | |
| "loss": 0.9459, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.06901041666666667, | |
| "grad_norm": 0.3995688855648041, | |
| "learning_rate": 0.00010706000000000001, | |
| "loss": 0.8842, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0703125, | |
| "grad_norm": 0.37477144598960876, | |
| "learning_rate": 0.00010908, | |
| "loss": 0.9676, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.07161458333333333, | |
| "grad_norm": 0.4394771456718445, | |
| "learning_rate": 0.00011110000000000002, | |
| "loss": 1.0441, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.07291666666666667, | |
| "grad_norm": 0.5129756331443787, | |
| "learning_rate": 0.00011312000000000001, | |
| "loss": 0.9511, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.07421875, | |
| "grad_norm": 0.3840573728084564, | |
| "learning_rate": 0.00011514, | |
| "loss": 0.9381, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.07552083333333333, | |
| "grad_norm": 0.3669443726539612, | |
| "learning_rate": 0.00011716, | |
| "loss": 0.9452, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.07682291666666667, | |
| "grad_norm": 0.4047168791294098, | |
| "learning_rate": 0.00011918, | |
| "loss": 0.9662, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.078125, | |
| "grad_norm": 0.43623411655426025, | |
| "learning_rate": 0.00012119999999999999, | |
| "loss": 0.9399, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07942708333333333, | |
| "grad_norm": 0.45791196823120117, | |
| "learning_rate": 0.00012322, | |
| "loss": 0.9579, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.08072916666666667, | |
| "grad_norm": 0.3631390929222107, | |
| "learning_rate": 0.00012524000000000001, | |
| "loss": 0.9537, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.08203125, | |
| "grad_norm": 0.44578367471694946, | |
| "learning_rate": 0.00012726, | |
| "loss": 0.969, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 0.3980722427368164, | |
| "learning_rate": 0.00012928, | |
| "loss": 0.9296, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.08463541666666667, | |
| "grad_norm": 0.415637731552124, | |
| "learning_rate": 0.00013130000000000002, | |
| "loss": 0.871, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.0859375, | |
| "grad_norm": 0.4139644503593445, | |
| "learning_rate": 0.00013332, | |
| "loss": 0.9521, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.08723958333333333, | |
| "grad_norm": 0.38350746035575867, | |
| "learning_rate": 0.00013534000000000002, | |
| "loss": 0.8741, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.08854166666666667, | |
| "grad_norm": 0.3766394555568695, | |
| "learning_rate": 0.00013736, | |
| "loss": 0.9806, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.08984375, | |
| "grad_norm": 0.3952779769897461, | |
| "learning_rate": 0.00013937999999999998, | |
| "loss": 0.9444, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.09114583333333333, | |
| "grad_norm": 0.3924815058708191, | |
| "learning_rate": 0.0001414, | |
| "loss": 0.9173, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09244791666666667, | |
| "grad_norm": 0.37496528029441833, | |
| "learning_rate": 0.00014342, | |
| "loss": 0.884, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.09375, | |
| "grad_norm": 0.3437505066394806, | |
| "learning_rate": 0.00014544, | |
| "loss": 0.8915, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.09505208333333333, | |
| "grad_norm": 0.4203055202960968, | |
| "learning_rate": 0.00014746, | |
| "loss": 0.8764, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.09635416666666667, | |
| "grad_norm": 0.37117621302604675, | |
| "learning_rate": 0.00014948, | |
| "loss": 0.9337, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.09765625, | |
| "grad_norm": 0.3810153305530548, | |
| "learning_rate": 0.0001515, | |
| "loss": 0.8872, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.09895833333333333, | |
| "grad_norm": 0.3892213702201843, | |
| "learning_rate": 0.00015352, | |
| "loss": 0.8906, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.10026041666666667, | |
| "grad_norm": 0.37222960591316223, | |
| "learning_rate": 0.00015554000000000002, | |
| "loss": 0.8289, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1015625, | |
| "grad_norm": 0.37449032068252563, | |
| "learning_rate": 0.00015756, | |
| "loss": 0.9172, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.10286458333333333, | |
| "grad_norm": 0.4078482687473297, | |
| "learning_rate": 0.00015958000000000001, | |
| "loss": 0.8892, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.10416666666666667, | |
| "grad_norm": 0.3846503794193268, | |
| "learning_rate": 0.00016160000000000002, | |
| "loss": 0.8464, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.10546875, | |
| "grad_norm": 0.39846357703208923, | |
| "learning_rate": 0.00016362, | |
| "loss": 0.9439, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.10677083333333333, | |
| "grad_norm": 0.3736095726490021, | |
| "learning_rate": 0.00016564, | |
| "loss": 0.9479, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.10807291666666667, | |
| "grad_norm": 0.38269343972206116, | |
| "learning_rate": 0.00016766, | |
| "loss": 0.8499, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.109375, | |
| "grad_norm": 0.4197899103164673, | |
| "learning_rate": 0.00016968, | |
| "loss": 0.9199, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.11067708333333333, | |
| "grad_norm": 0.3637363016605377, | |
| "learning_rate": 0.0001717, | |
| "loss": 0.8569, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.11197916666666667, | |
| "grad_norm": 0.36869698762893677, | |
| "learning_rate": 0.00017372, | |
| "loss": 0.9584, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.11328125, | |
| "grad_norm": 0.35092616081237793, | |
| "learning_rate": 0.00017574, | |
| "loss": 0.9374, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.11458333333333333, | |
| "grad_norm": 0.39582520723342896, | |
| "learning_rate": 0.00017776, | |
| "loss": 0.9239, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.11588541666666667, | |
| "grad_norm": 0.3559093773365021, | |
| "learning_rate": 0.00017978000000000002, | |
| "loss": 0.8901, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.1171875, | |
| "grad_norm": 0.3778141140937805, | |
| "learning_rate": 0.0001818, | |
| "loss": 0.8903, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.11848958333333333, | |
| "grad_norm": 0.36067256331443787, | |
| "learning_rate": 0.00018382, | |
| "loss": 0.8905, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.11979166666666667, | |
| "grad_norm": 0.36430442333221436, | |
| "learning_rate": 0.00018584000000000002, | |
| "loss": 0.8941, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.12109375, | |
| "grad_norm": 0.40255868434906006, | |
| "learning_rate": 0.00018786, | |
| "loss": 0.9192, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.12239583333333333, | |
| "grad_norm": 0.38338273763656616, | |
| "learning_rate": 0.00018988, | |
| "loss": 0.8976, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.12369791666666667, | |
| "grad_norm": 0.3768159747123718, | |
| "learning_rate": 0.0001919, | |
| "loss": 0.9337, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.3879520297050476, | |
| "learning_rate": 0.00019392, | |
| "loss": 0.8761, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.12630208333333334, | |
| "grad_norm": 0.3825957477092743, | |
| "learning_rate": 0.00019594, | |
| "loss": 0.9051, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.12760416666666666, | |
| "grad_norm": 0.3993157148361206, | |
| "learning_rate": 0.00019796, | |
| "loss": 0.8613, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.12890625, | |
| "grad_norm": 0.3906169831752777, | |
| "learning_rate": 0.00019998, | |
| "loss": 0.9948, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.13020833333333334, | |
| "grad_norm": 0.38024818897247314, | |
| "learning_rate": 0.000202, | |
| "loss": 0.981, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13151041666666666, | |
| "grad_norm": 0.3382592499256134, | |
| "learning_rate": 0.00020199688492212377, | |
| "loss": 0.8662, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.1328125, | |
| "grad_norm": 0.42753666639328003, | |
| "learning_rate": 0.00020198753988064772, | |
| "loss": 0.858, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.13411458333333334, | |
| "grad_norm": 0.36199378967285156, | |
| "learning_rate": 0.00020197196545201806, | |
| "loss": 0.863, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.13541666666666666, | |
| "grad_norm": 0.30593058466911316, | |
| "learning_rate": 0.0002019501625969389, | |
| "loss": 0.9211, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.13671875, | |
| "grad_norm": 0.40363338589668274, | |
| "learning_rate": 0.00020192213266031304, | |
| "loss": 0.9512, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.13802083333333334, | |
| "grad_norm": 0.34124210476875305, | |
| "learning_rate": 0.00020188787737115897, | |
| "loss": 0.8821, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.13932291666666666, | |
| "grad_norm": 0.324048787355423, | |
| "learning_rate": 0.00020184739884250436, | |
| "loss": 0.9053, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.140625, | |
| "grad_norm": 0.32893842458724976, | |
| "learning_rate": 0.00020180069957125544, | |
| "loss": 0.8581, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.14192708333333334, | |
| "grad_norm": 0.3708108067512512, | |
| "learning_rate": 0.0002017477824380433, | |
| "loss": 0.908, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.14322916666666666, | |
| "grad_norm": 0.3627496361732483, | |
| "learning_rate": 0.00020168865070704594, | |
| "loss": 0.8493, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.14453125, | |
| "grad_norm": 0.38643914461135864, | |
| "learning_rate": 0.00020162330802578706, | |
| "loss": 0.9305, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.14583333333333334, | |
| "grad_norm": 0.3313356041908264, | |
| "learning_rate": 0.00020155175842491107, | |
| "loss": 0.9209, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.14713541666666666, | |
| "grad_norm": 0.36708641052246094, | |
| "learning_rate": 0.0002014740063179344, | |
| "loss": 0.908, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.1484375, | |
| "grad_norm": 0.3595336973667145, | |
| "learning_rate": 0.00020139005650097317, | |
| "loss": 0.8212, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.14973958333333334, | |
| "grad_norm": 0.43490394949913025, | |
| "learning_rate": 0.00020129991415244756, | |
| "loss": 0.8863, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.15104166666666666, | |
| "grad_norm": 0.35051169991493225, | |
| "learning_rate": 0.00020120358483276227, | |
| "loss": 0.9196, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.15234375, | |
| "grad_norm": 0.36855900287628174, | |
| "learning_rate": 0.00020110107448396346, | |
| "loss": 0.8889, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.15364583333333334, | |
| "grad_norm": 0.32522818446159363, | |
| "learning_rate": 0.0002009923894293723, | |
| "loss": 0.9072, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.15494791666666666, | |
| "grad_norm": 0.3217287063598633, | |
| "learning_rate": 0.00020087753637319499, | |
| "loss": 0.8859, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.15625, | |
| "grad_norm": 0.4293065667152405, | |
| "learning_rate": 0.00020075652240010892, | |
| "loss": 0.8551, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.15755208333333334, | |
| "grad_norm": 0.3320338726043701, | |
| "learning_rate": 0.00020062935497482606, | |
| "loss": 0.9491, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.15885416666666666, | |
| "grad_norm": 0.3622731864452362, | |
| "learning_rate": 0.00020049604194163217, | |
| "loss": 0.9246, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.16015625, | |
| "grad_norm": 0.33044755458831787, | |
| "learning_rate": 0.00020035659152390313, | |
| "loss": 0.8979, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.16145833333333334, | |
| "grad_norm": 0.3863113522529602, | |
| "learning_rate": 0.00020021101232359757, | |
| "loss": 0.8701, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.16276041666666666, | |
| "grad_norm": 0.39442121982574463, | |
| "learning_rate": 0.0002000593133207263, | |
| "loss": 0.8516, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1640625, | |
| "grad_norm": 0.3625333309173584, | |
| "learning_rate": 0.00019990150387279835, | |
| "loss": 0.887, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.16536458333333334, | |
| "grad_norm": 0.3347029983997345, | |
| "learning_rate": 0.00019973759371424388, | |
| "loss": 0.9712, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 0.34266117215156555, | |
| "learning_rate": 0.0001995675929558135, | |
| "loss": 0.9359, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.16796875, | |
| "grad_norm": 0.34118279814720154, | |
| "learning_rate": 0.0001993915120839548, | |
| "loss": 0.9324, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.16927083333333334, | |
| "grad_norm": 0.3674456477165222, | |
| "learning_rate": 0.00019920936196016534, | |
| "loss": 0.8567, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.17057291666666666, | |
| "grad_norm": 0.3397194445133209, | |
| "learning_rate": 0.0001990211538203228, | |
| "loss": 0.9137, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.171875, | |
| "grad_norm": 0.32985278964042664, | |
| "learning_rate": 0.00019882689927399174, | |
| "loss": 0.9193, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.17317708333333334, | |
| "grad_norm": 0.3857951760292053, | |
| "learning_rate": 0.00019862661030370764, | |
| "loss": 0.896, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.17447916666666666, | |
| "grad_norm": 0.3355730175971985, | |
| "learning_rate": 0.00019842029926423762, | |
| "loss": 0.9673, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.17578125, | |
| "grad_norm": 0.3015749156475067, | |
| "learning_rate": 0.00019820797888181837, | |
| "loss": 0.9244, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.17708333333333334, | |
| "grad_norm": 0.3197927474975586, | |
| "learning_rate": 0.00019798966225337126, | |
| "loss": 0.9298, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.17838541666666666, | |
| "grad_norm": 0.31796562671661377, | |
| "learning_rate": 0.00019776536284569425, | |
| "loss": 0.9274, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.1796875, | |
| "grad_norm": 0.37680956721305847, | |
| "learning_rate": 0.00019753509449463134, | |
| "loss": 0.8858, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.18098958333333334, | |
| "grad_norm": 0.33456477522850037, | |
| "learning_rate": 0.00019729887140421912, | |
| "loss": 0.8654, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.18229166666666666, | |
| "grad_norm": 0.3346468508243561, | |
| "learning_rate": 0.00019705670814581052, | |
| "loss": 0.7735, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.18359375, | |
| "grad_norm": 0.33176180720329285, | |
| "learning_rate": 0.00019680861965717597, | |
| "loss": 0.938, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.18489583333333334, | |
| "grad_norm": 0.36904606223106384, | |
| "learning_rate": 0.0001965546212415821, | |
| "loss": 0.9044, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.18619791666666666, | |
| "grad_norm": 0.37553444504737854, | |
| "learning_rate": 0.00019629472856684755, | |
| "loss": 0.8976, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.1875, | |
| "grad_norm": 0.30502942204475403, | |
| "learning_rate": 0.00019602895766437678, | |
| "loss": 0.8745, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.18880208333333334, | |
| "grad_norm": 0.39606553316116333, | |
| "learning_rate": 0.00019575732492817092, | |
| "loss": 0.8426, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.19010416666666666, | |
| "grad_norm": 0.3282630741596222, | |
| "learning_rate": 0.00019547984711381662, | |
| "loss": 0.8231, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.19140625, | |
| "grad_norm": 0.3370037376880646, | |
| "learning_rate": 0.0001951965413374525, | |
| "loss": 0.7844, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.19270833333333334, | |
| "grad_norm": 0.30234625935554504, | |
| "learning_rate": 0.00019490742507471338, | |
| "loss": 0.8674, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.19401041666666666, | |
| "grad_norm": 0.34176871180534363, | |
| "learning_rate": 0.0001946125161596522, | |
| "loss": 0.9853, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.1953125, | |
| "grad_norm": 0.322853684425354, | |
| "learning_rate": 0.00019431183278363997, | |
| "loss": 0.9021, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19661458333333334, | |
| "grad_norm": 0.32170554995536804, | |
| "learning_rate": 0.00019400539349424367, | |
| "loss": 0.8872, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.19791666666666666, | |
| "grad_norm": 0.3576914072036743, | |
| "learning_rate": 0.0001936932171940821, | |
| "loss": 0.9812, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.19921875, | |
| "grad_norm": 0.3470019996166229, | |
| "learning_rate": 0.00019337532313966, | |
| "loss": 0.9353, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.20052083333333334, | |
| "grad_norm": 0.2939004600048065, | |
| "learning_rate": 0.00019305173094017996, | |
| "loss": 0.8564, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.20182291666666666, | |
| "grad_norm": 0.353512167930603, | |
| "learning_rate": 0.0001927224605563332, | |
| "loss": 0.8534, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.203125, | |
| "grad_norm": 0.330636203289032, | |
| "learning_rate": 0.00019238753229906797, | |
| "loss": 0.8046, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.20442708333333334, | |
| "grad_norm": 0.2957157492637634, | |
| "learning_rate": 0.00019204696682833682, | |
| "loss": 0.8823, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.20572916666666666, | |
| "grad_norm": 0.3297777473926544, | |
| "learning_rate": 0.00019170078515182216, | |
| "loss": 0.8739, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.20703125, | |
| "grad_norm": 0.37313520908355713, | |
| "learning_rate": 0.00019134900862364054, | |
| "loss": 0.7445, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.20833333333333334, | |
| "grad_norm": 0.3856740891933441, | |
| "learning_rate": 0.00019099165894302515, | |
| "loss": 0.9027, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.20963541666666666, | |
| "grad_norm": 0.3400990664958954, | |
| "learning_rate": 0.00019062875815298763, | |
| "loss": 0.8591, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2109375, | |
| "grad_norm": 0.33788663148880005, | |
| "learning_rate": 0.00019026032863895805, | |
| "loss": 0.8587, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.21223958333333334, | |
| "grad_norm": 0.31548887491226196, | |
| "learning_rate": 0.00018988639312740433, | |
| "loss": 0.8769, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.21354166666666666, | |
| "grad_norm": 0.319621205329895, | |
| "learning_rate": 0.0001895069746844302, | |
| "loss": 0.8355, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.21484375, | |
| "grad_norm": 0.32653823494911194, | |
| "learning_rate": 0.00018912209671435252, | |
| "loss": 0.9207, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.21614583333333334, | |
| "grad_norm": 0.3283182978630066, | |
| "learning_rate": 0.00018873178295825732, | |
| "loss": 0.8737, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.21744791666666666, | |
| "grad_norm": 0.32019880414009094, | |
| "learning_rate": 0.00018833605749253566, | |
| "loss": 0.9572, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.21875, | |
| "grad_norm": 0.34993138909339905, | |
| "learning_rate": 0.00018793494472739831, | |
| "loss": 0.849, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.22005208333333334, | |
| "grad_norm": 0.33055248856544495, | |
| "learning_rate": 0.00018752846940537003, | |
| "loss": 0.8668, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.22135416666666666, | |
| "grad_norm": 0.31973016262054443, | |
| "learning_rate": 0.0001871166565997633, | |
| "loss": 0.8875, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.22265625, | |
| "grad_norm": 0.3294101357460022, | |
| "learning_rate": 0.00018669953171313188, | |
| "loss": 0.9007, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.22395833333333334, | |
| "grad_norm": 0.30089470744132996, | |
| "learning_rate": 0.00018627712047570352, | |
| "loss": 0.9276, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.22526041666666666, | |
| "grad_norm": 0.3125169575214386, | |
| "learning_rate": 0.0001858494489437931, | |
| "loss": 0.8886, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.2265625, | |
| "grad_norm": 0.34174057841300964, | |
| "learning_rate": 0.0001854165434981953, | |
| "loss": 0.8538, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.22786458333333334, | |
| "grad_norm": 0.31229665875434875, | |
| "learning_rate": 0.00018497843084255708, | |
| "loss": 0.8942, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.22916666666666666, | |
| "grad_norm": 0.35159239172935486, | |
| "learning_rate": 0.00018453513800173072, | |
| "loss": 0.8556, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.23046875, | |
| "grad_norm": 0.3651125431060791, | |
| "learning_rate": 0.00018408669232010684, | |
| "loss": 0.8867, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.23177083333333334, | |
| "grad_norm": 0.34238573908805847, | |
| "learning_rate": 0.00018363312145992737, | |
| "loss": 0.8788, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.23307291666666666, | |
| "grad_norm": 0.3748724162578583, | |
| "learning_rate": 0.0001831744533995795, | |
| "loss": 0.8432, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.234375, | |
| "grad_norm": 0.34040236473083496, | |
| "learning_rate": 0.00018271071643186968, | |
| "loss": 0.8855, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.23567708333333334, | |
| "grad_norm": 0.3309285342693329, | |
| "learning_rate": 0.00018224193916227852, | |
| "loss": 0.7903, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.23697916666666666, | |
| "grad_norm": 0.3496496379375458, | |
| "learning_rate": 0.00018176815050719615, | |
| "loss": 0.8447, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.23828125, | |
| "grad_norm": 0.3104117810726166, | |
| "learning_rate": 0.00018128937969213852, | |
| "loss": 0.9041, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.23958333333333334, | |
| "grad_norm": 0.35145995020866394, | |
| "learning_rate": 0.00018080565624994474, | |
| "loss": 0.8347, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.24088541666666666, | |
| "grad_norm": 0.33385100960731506, | |
| "learning_rate": 0.00018031701001895524, | |
| "loss": 0.8578, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.2421875, | |
| "grad_norm": 0.35553544759750366, | |
| "learning_rate": 0.0001798234711411713, | |
| "loss": 0.8204, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.24348958333333334, | |
| "grad_norm": 0.365278422832489, | |
| "learning_rate": 0.00017932507006039567, | |
| "loss": 0.8957, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.24479166666666666, | |
| "grad_norm": 0.3690166771411896, | |
| "learning_rate": 0.0001788218375203547, | |
| "loss": 0.906, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.24609375, | |
| "grad_norm": 0.30934587121009827, | |
| "learning_rate": 0.00017831380456280192, | |
| "loss": 0.8963, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.24739583333333334, | |
| "grad_norm": 0.3289198875427246, | |
| "learning_rate": 0.00017780100252560313, | |
| "loss": 0.9016, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.24869791666666666, | |
| "grad_norm": 0.3520403206348419, | |
| "learning_rate": 0.00017728346304080357, | |
| "loss": 0.9407, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.3425486981868744, | |
| "learning_rate": 0.0001767612180326764, | |
| "loss": 0.8207, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.2513020833333333, | |
| "grad_norm": 0.36947768926620483, | |
| "learning_rate": 0.00017623429971575384, | |
| "loss": 0.9397, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.2526041666666667, | |
| "grad_norm": 0.3345178961753845, | |
| "learning_rate": 0.0001757027405928396, | |
| "loss": 0.9446, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.25390625, | |
| "grad_norm": 0.35662034153938293, | |
| "learning_rate": 0.00017516657345300425, | |
| "loss": 0.8349, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2552083333333333, | |
| "grad_norm": 0.32738953828811646, | |
| "learning_rate": 0.00017462583136956258, | |
| "loss": 0.8485, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.2565104166666667, | |
| "grad_norm": 0.35002079606056213, | |
| "learning_rate": 0.00017408054769803337, | |
| "loss": 0.8226, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.2578125, | |
| "grad_norm": 0.3495853543281555, | |
| "learning_rate": 0.00017353075607408209, | |
| "loss": 0.8536, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2591145833333333, | |
| "grad_norm": 0.35052958130836487, | |
| "learning_rate": 0.00017297649041144575, | |
| "loss": 0.8592, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.2604166666666667, | |
| "grad_norm": 0.328745573759079, | |
| "learning_rate": 0.0001724177848998413, | |
| "loss": 0.8492, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.26171875, | |
| "grad_norm": 0.32306015491485596, | |
| "learning_rate": 0.00017185467400285644, | |
| "loss": 0.9631, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.2630208333333333, | |
| "grad_norm": 0.33527088165283203, | |
| "learning_rate": 0.00017128719245582374, | |
| "loss": 0.923, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.2643229166666667, | |
| "grad_norm": 0.31656643748283386, | |
| "learning_rate": 0.00017071537526367817, | |
| "loss": 0.814, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.265625, | |
| "grad_norm": 0.3177151679992676, | |
| "learning_rate": 0.00017013925769879755, | |
| "loss": 0.8771, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.2669270833333333, | |
| "grad_norm": 0.3306122124195099, | |
| "learning_rate": 0.00016955887529882714, | |
| "loss": 0.8597, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2682291666666667, | |
| "grad_norm": 0.34116023778915405, | |
| "learning_rate": 0.0001689742638644871, | |
| "loss": 0.8552, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.26953125, | |
| "grad_norm": 0.3636139929294586, | |
| "learning_rate": 0.00016838545945736458, | |
| "loss": 0.8387, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.2708333333333333, | |
| "grad_norm": 0.3404369652271271, | |
| "learning_rate": 0.00016779249839768884, | |
| "loss": 0.8186, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.2721354166666667, | |
| "grad_norm": 0.3426530659198761, | |
| "learning_rate": 0.00016719541726209117, | |
| "loss": 0.9083, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.2734375, | |
| "grad_norm": 0.34500670433044434, | |
| "learning_rate": 0.00016659425288134854, | |
| "loss": 0.8645, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2747395833333333, | |
| "grad_norm": 0.36617475748062134, | |
| "learning_rate": 0.00016598904233811168, | |
| "loss": 0.8138, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.2760416666666667, | |
| "grad_norm": 0.349728524684906, | |
| "learning_rate": 0.00016537982296461768, | |
| "loss": 0.8747, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.27734375, | |
| "grad_norm": 0.3206585645675659, | |
| "learning_rate": 0.00016476663234038717, | |
| "loss": 0.8638, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.2786458333333333, | |
| "grad_norm": 0.3095620274543762, | |
| "learning_rate": 0.00016414950828990625, | |
| "loss": 0.8839, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.2799479166666667, | |
| "grad_norm": 0.3462662100791931, | |
| "learning_rate": 0.00016352848888029326, | |
| "loss": 0.8475, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.28125, | |
| "grad_norm": 0.30881351232528687, | |
| "learning_rate": 0.00016290361241895064, | |
| "loss": 0.875, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.2825520833333333, | |
| "grad_norm": 0.3559829294681549, | |
| "learning_rate": 0.00016227491745120196, | |
| "loss": 0.8667, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.2838541666666667, | |
| "grad_norm": 0.3805384039878845, | |
| "learning_rate": 0.0001616424427579143, | |
| "loss": 0.8644, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.28515625, | |
| "grad_norm": 0.288903146982193, | |
| "learning_rate": 0.0001610062273531059, | |
| "loss": 0.9587, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.2864583333333333, | |
| "grad_norm": 0.3124583959579468, | |
| "learning_rate": 0.00016036631048153979, | |
| "loss": 0.9036, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2877604166666667, | |
| "grad_norm": 0.3762437403202057, | |
| "learning_rate": 0.0001597227316163029, | |
| "loss": 0.8723, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.2890625, | |
| "grad_norm": 0.2969296872615814, | |
| "learning_rate": 0.00015907553045637116, | |
| "loss": 0.9165, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.2903645833333333, | |
| "grad_norm": 0.3156786561012268, | |
| "learning_rate": 0.00015842474692416068, | |
| "loss": 0.8797, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.2916666666666667, | |
| "grad_norm": 0.3499281406402588, | |
| "learning_rate": 0.0001577704211630652, | |
| "loss": 0.9325, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.29296875, | |
| "grad_norm": 0.32044610381126404, | |
| "learning_rate": 0.00015711259353497981, | |
| "loss": 0.9274, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2942708333333333, | |
| "grad_norm": 0.3384315073490143, | |
| "learning_rate": 0.0001564513046178113, | |
| "loss": 0.8486, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.2955729166666667, | |
| "grad_norm": 0.35218653082847595, | |
| "learning_rate": 0.000155786595202975, | |
| "loss": 0.815, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.296875, | |
| "grad_norm": 0.308248907327652, | |
| "learning_rate": 0.00015511850629287865, | |
| "loss": 0.8801, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.2981770833333333, | |
| "grad_norm": 0.3766931891441345, | |
| "learning_rate": 0.00015444707909839325, | |
| "loss": 0.7229, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.2994791666666667, | |
| "grad_norm": 0.3386388421058655, | |
| "learning_rate": 0.00015377235503631083, | |
| "loss": 0.8249, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.30078125, | |
| "grad_norm": 0.3271256983280182, | |
| "learning_rate": 0.0001530943757267898, | |
| "loss": 0.8026, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.3020833333333333, | |
| "grad_norm": 0.3631853461265564, | |
| "learning_rate": 0.00015241318299078751, | |
| "loss": 0.8691, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.3033854166666667, | |
| "grad_norm": 0.32474151253700256, | |
| "learning_rate": 0.00015172881884748063, | |
| "loss": 0.8536, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.3046875, | |
| "grad_norm": 0.3289790153503418, | |
| "learning_rate": 0.00015104132551167318, | |
| "loss": 0.9035, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.3059895833333333, | |
| "grad_norm": 0.32092922925949097, | |
| "learning_rate": 0.00015035074539119248, | |
| "loss": 0.8217, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.3072916666666667, | |
| "grad_norm": 0.350707471370697, | |
| "learning_rate": 0.00014965712108427323, | |
| "loss": 0.8041, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.30859375, | |
| "grad_norm": 0.33856043219566345, | |
| "learning_rate": 0.00014896049537693005, | |
| "loss": 0.8546, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.3098958333333333, | |
| "grad_norm": 0.3133806884288788, | |
| "learning_rate": 0.00014826091124031792, | |
| "loss": 0.9246, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.3111979166666667, | |
| "grad_norm": 0.37403765320777893, | |
| "learning_rate": 0.0001475584118280817, | |
| "loss": 0.8921, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.3125, | |
| "grad_norm": 0.3411153256893158, | |
| "learning_rate": 0.00014685304047369423, | |
| "loss": 0.8605, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3138020833333333, | |
| "grad_norm": 0.3260857164859772, | |
| "learning_rate": 0.00014614484068778324, | |
| "loss": 0.8922, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.3151041666666667, | |
| "grad_norm": 0.32127058506011963, | |
| "learning_rate": 0.00014543385615544744, | |
| "loss": 0.8851, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.31640625, | |
| "grad_norm": 0.30613449215888977, | |
| "learning_rate": 0.00014472013073356184, | |
| "loss": 0.8732, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.3177083333333333, | |
| "grad_norm": 0.3390278220176697, | |
| "learning_rate": 0.00014400370844807234, | |
| "loss": 0.8251, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.3190104166666667, | |
| "grad_norm": 0.34041738510131836, | |
| "learning_rate": 0.00014328463349128025, | |
| "loss": 0.8146, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3203125, | |
| "grad_norm": 0.3534432053565979, | |
| "learning_rate": 0.000142562950219116, | |
| "loss": 0.8614, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.3216145833333333, | |
| "grad_norm": 0.3607141673564911, | |
| "learning_rate": 0.00014183870314840325, | |
| "loss": 0.8102, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.3229166666666667, | |
| "grad_norm": 0.3430651128292084, | |
| "learning_rate": 0.00014111193695411285, | |
| "loss": 0.8703, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.32421875, | |
| "grad_norm": 0.35520392656326294, | |
| "learning_rate": 0.00014038269646660703, | |
| "loss": 0.8424, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.3255208333333333, | |
| "grad_norm": 0.35280680656433105, | |
| "learning_rate": 0.00013965102666887408, | |
| "loss": 0.8588, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3268229166666667, | |
| "grad_norm": 0.29341921210289, | |
| "learning_rate": 0.0001389169726937536, | |
| "loss": 0.8856, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.328125, | |
| "grad_norm": 0.344706267118454, | |
| "learning_rate": 0.0001381805798211525, | |
| "loss": 0.8694, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.3294270833333333, | |
| "grad_norm": 0.3089728355407715, | |
| "learning_rate": 0.00013744189347525182, | |
| "loss": 0.8805, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.3307291666666667, | |
| "grad_norm": 0.2982807159423828, | |
| "learning_rate": 0.00013670095922170498, | |
| "loss": 0.9559, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.33203125, | |
| "grad_norm": 0.35118257999420166, | |
| "learning_rate": 0.00013595782276482678, | |
| "loss": 0.8535, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.31711286306381226, | |
| "learning_rate": 0.00013521252994477446, | |
| "loss": 0.9169, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.3346354166666667, | |
| "grad_norm": 0.28975602984428406, | |
| "learning_rate": 0.00013446512673471965, | |
| "loss": 0.9879, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.3359375, | |
| "grad_norm": 0.3587123453617096, | |
| "learning_rate": 0.0001337156592380131, | |
| "loss": 0.8419, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.3372395833333333, | |
| "grad_norm": 0.32705414295196533, | |
| "learning_rate": 0.0001329641736853402, | |
| "loss": 0.8646, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.3385416666666667, | |
| "grad_norm": 0.301553338766098, | |
| "learning_rate": 0.0001322107164318697, | |
| "loss": 0.8199, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.33984375, | |
| "grad_norm": 0.31590837240219116, | |
| "learning_rate": 0.00013145533395439405, | |
| "loss": 0.8675, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.3411458333333333, | |
| "grad_norm": 0.28727132081985474, | |
| "learning_rate": 0.0001306980728484627, | |
| "loss": 0.8971, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.3424479166666667, | |
| "grad_norm": 0.38094404339790344, | |
| "learning_rate": 0.00012993897982550764, | |
| "loss": 0.8053, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.34375, | |
| "grad_norm": 0.3682737648487091, | |
| "learning_rate": 0.00012917810170996218, | |
| "loss": 0.8066, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.3450520833333333, | |
| "grad_norm": 0.34040960669517517, | |
| "learning_rate": 0.0001284154854363725, | |
| "loss": 1.0175, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.3463541666666667, | |
| "grad_norm": 0.3567488193511963, | |
| "learning_rate": 0.00012765117804650267, | |
| "loss": 0.8878, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.34765625, | |
| "grad_norm": 0.3080446720123291, | |
| "learning_rate": 0.00012688522668643268, | |
| "loss": 0.7767, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.3489583333333333, | |
| "grad_norm": 0.29930922389030457, | |
| "learning_rate": 0.00012611767860365038, | |
| "loss": 0.8593, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.3502604166666667, | |
| "grad_norm": 0.3136909008026123, | |
| "learning_rate": 0.00012534858114413692, | |
| "loss": 0.8393, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.3515625, | |
| "grad_norm": 0.3064602017402649, | |
| "learning_rate": 0.00012457798174944645, | |
| "loss": 0.8666, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.3528645833333333, | |
| "grad_norm": 0.3652801215648651, | |
| "learning_rate": 0.0001238059279537795, | |
| "loss": 0.8608, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.3541666666666667, | |
| "grad_norm": 0.3020021319389343, | |
| "learning_rate": 0.00012303246738105082, | |
| "loss": 0.8949, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.35546875, | |
| "grad_norm": 0.3343137502670288, | |
| "learning_rate": 0.00012225764774195186, | |
| "loss": 0.8392, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.3567708333333333, | |
| "grad_norm": 0.33666467666625977, | |
| "learning_rate": 0.00012148151683100776, | |
| "loss": 0.8073, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.3580729166666667, | |
| "grad_norm": 0.3414088189601898, | |
| "learning_rate": 0.00012070412252362897, | |
| "loss": 0.7982, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.359375, | |
| "grad_norm": 0.34131115674972534, | |
| "learning_rate": 0.0001199255127731582, | |
| "loss": 0.8725, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.3606770833333333, | |
| "grad_norm": 0.3253263235092163, | |
| "learning_rate": 0.00011914573560791246, | |
| "loss": 0.8502, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.3619791666666667, | |
| "grad_norm": 0.2974787652492523, | |
| "learning_rate": 0.00011836483912822035, | |
| "loss": 0.8696, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.36328125, | |
| "grad_norm": 0.310968816280365, | |
| "learning_rate": 0.00011758287150345516, | |
| "loss": 0.8618, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.3645833333333333, | |
| "grad_norm": 0.30811068415641785, | |
| "learning_rate": 0.00011679988096906333, | |
| "loss": 0.7827, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3658854166666667, | |
| "grad_norm": 0.31703218817710876, | |
| "learning_rate": 0.00011601591582358924, | |
| "loss": 0.8557, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.3671875, | |
| "grad_norm": 0.3499086797237396, | |
| "learning_rate": 0.00011523102442569585, | |
| "loss": 0.8819, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.3684895833333333, | |
| "grad_norm": 0.3098037838935852, | |
| "learning_rate": 0.00011444525519118179, | |
| "loss": 0.928, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.3697916666666667, | |
| "grad_norm": 0.30361640453338623, | |
| "learning_rate": 0.00011365865658999474, | |
| "loss": 0.8187, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.37109375, | |
| "grad_norm": 0.3158656060695648, | |
| "learning_rate": 0.00011287127714324162, | |
| "loss": 0.8389, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3723958333333333, | |
| "grad_norm": 0.31333765387535095, | |
| "learning_rate": 0.00011208316542019556, | |
| "loss": 0.874, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.3736979166666667, | |
| "grad_norm": 0.3238910734653473, | |
| "learning_rate": 0.00011129437003530006, | |
| "loss": 0.8417, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.31688323616981506, | |
| "learning_rate": 0.00011050493964516997, | |
| "loss": 0.7645, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.3763020833333333, | |
| "grad_norm": 0.32902631163597107, | |
| "learning_rate": 0.00010971492294559029, | |
| "loss": 0.785, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.3776041666666667, | |
| "grad_norm": 0.3547747731208801, | |
| "learning_rate": 0.00010892436866851235, | |
| "loss": 0.8035, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.37890625, | |
| "grad_norm": 0.34230130910873413, | |
| "learning_rate": 0.00010813332557904784, | |
| "loss": 0.8288, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.3802083333333333, | |
| "grad_norm": 0.3104645013809204, | |
| "learning_rate": 0.00010734184247246066, | |
| "loss": 0.8458, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.3815104166666667, | |
| "grad_norm": 0.3715684711933136, | |
| "learning_rate": 0.00010654996817115704, | |
| "loss": 0.9265, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.3828125, | |
| "grad_norm": 0.30713942646980286, | |
| "learning_rate": 0.00010575775152167391, | |
| "loss": 0.904, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.3841145833333333, | |
| "grad_norm": 0.32410570979118347, | |
| "learning_rate": 0.00010496524139166594, | |
| "loss": 0.8621, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3854166666666667, | |
| "grad_norm": 0.3331278860569, | |
| "learning_rate": 0.00010417248666689095, | |
| "loss": 0.8377, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.38671875, | |
| "grad_norm": 0.3093259334564209, | |
| "learning_rate": 0.00010337953624819464, | |
| "loss": 0.8502, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.3880208333333333, | |
| "grad_norm": 0.3295346796512604, | |
| "learning_rate": 0.0001025864390484939, | |
| "loss": 0.8671, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.3893229166666667, | |
| "grad_norm": 0.2909308075904846, | |
| "learning_rate": 0.00010179324398975984, | |
| "loss": 0.8519, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.390625, | |
| "grad_norm": 0.3205868899822235, | |
| "learning_rate": 0.000101, | |
| "loss": 0.8516, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3919270833333333, | |
| "grad_norm": 0.3373914659023285, | |
| "learning_rate": 0.00010020675601024019, | |
| "loss": 0.9442, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.3932291666666667, | |
| "grad_norm": 0.31464096903800964, | |
| "learning_rate": 9.941356095150613e-05, | |
| "loss": 0.8757, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.39453125, | |
| "grad_norm": 0.32743626832962036, | |
| "learning_rate": 9.862046375180539e-05, | |
| "loss": 0.811, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.3958333333333333, | |
| "grad_norm": 0.36112740635871887, | |
| "learning_rate": 9.782751333310905e-05, | |
| "loss": 0.8523, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.3971354166666667, | |
| "grad_norm": 0.3145286738872528, | |
| "learning_rate": 9.70347586083341e-05, | |
| "loss": 0.9191, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.3984375, | |
| "grad_norm": 0.3485451638698578, | |
| "learning_rate": 9.62422484783261e-05, | |
| "loss": 0.8646, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.3997395833333333, | |
| "grad_norm": 0.33839061856269836, | |
| "learning_rate": 9.5450031828843e-05, | |
| "loss": 0.9219, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.4010416666666667, | |
| "grad_norm": 0.29558366537094116, | |
| "learning_rate": 9.465815752753935e-05, | |
| "loss": 0.8581, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.40234375, | |
| "grad_norm": 0.30999499559402466, | |
| "learning_rate": 9.386667442095219e-05, | |
| "loss": 0.9484, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.4036458333333333, | |
| "grad_norm": 0.3167370557785034, | |
| "learning_rate": 9.307563133148767e-05, | |
| "loss": 0.8414, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4049479166666667, | |
| "grad_norm": 0.34423360228538513, | |
| "learning_rate": 9.228507705440976e-05, | |
| "loss": 0.8577, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.40625, | |
| "grad_norm": 0.33800390362739563, | |
| "learning_rate": 9.149506035483005e-05, | |
| "loss": 0.866, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.4075520833333333, | |
| "grad_norm": 0.31321457028388977, | |
| "learning_rate": 9.070562996469997e-05, | |
| "loss": 0.8437, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.4088541666666667, | |
| "grad_norm": 0.3370513916015625, | |
| "learning_rate": 8.991683457980443e-05, | |
| "loss": 0.8091, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.41015625, | |
| "grad_norm": 0.31457364559173584, | |
| "learning_rate": 8.912872285675841e-05, | |
| "loss": 0.7871, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.4114583333333333, | |
| "grad_norm": 0.3068949282169342, | |
| "learning_rate": 8.834134341000527e-05, | |
| "loss": 0.8504, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.4127604166666667, | |
| "grad_norm": 0.33699652552604675, | |
| "learning_rate": 8.755474480881823e-05, | |
| "loss": 0.8919, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.4140625, | |
| "grad_norm": 0.2944529056549072, | |
| "learning_rate": 8.676897557430415e-05, | |
| "loss": 0.8644, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.4153645833333333, | |
| "grad_norm": 0.3244035840034485, | |
| "learning_rate": 8.598408417641078e-05, | |
| "loss": 0.901, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 0.29464152455329895, | |
| "learning_rate": 8.520011903093666e-05, | |
| "loss": 0.8057, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.41796875, | |
| "grad_norm": 0.347151517868042, | |
| "learning_rate": 8.441712849654485e-05, | |
| "loss": 0.8939, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.4192708333333333, | |
| "grad_norm": 0.3080258369445801, | |
| "learning_rate": 8.363516087177962e-05, | |
| "loss": 0.9819, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.4205729166666667, | |
| "grad_norm": 0.3100193440914154, | |
| "learning_rate": 8.285426439208755e-05, | |
| "loss": 0.8772, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.421875, | |
| "grad_norm": 0.34320497512817383, | |
| "learning_rate": 8.20744872268418e-05, | |
| "loss": 0.7966, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.4231770833333333, | |
| "grad_norm": 0.30067065358161926, | |
| "learning_rate": 8.129587747637105e-05, | |
| "loss": 0.8682, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4244791666666667, | |
| "grad_norm": 0.34988993406295776, | |
| "learning_rate": 8.051848316899227e-05, | |
| "loss": 0.8989, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.42578125, | |
| "grad_norm": 0.34072110056877136, | |
| "learning_rate": 7.974235225804814e-05, | |
| "loss": 0.7937, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.4270833333333333, | |
| "grad_norm": 0.33146417140960693, | |
| "learning_rate": 7.896753261894923e-05, | |
| "loss": 0.7794, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.4283854166666667, | |
| "grad_norm": 0.31865736842155457, | |
| "learning_rate": 7.819407204622054e-05, | |
| "loss": 0.8462, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.4296875, | |
| "grad_norm": 0.35836029052734375, | |
| "learning_rate": 7.74220182505536e-05, | |
| "loss": 0.929, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4309895833333333, | |
| "grad_norm": 0.3302249014377594, | |
| "learning_rate": 7.665141885586312e-05, | |
| "loss": 0.8297, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.4322916666666667, | |
| "grad_norm": 0.36180800199508667, | |
| "learning_rate": 7.588232139634968e-05, | |
| "loss": 0.8518, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.43359375, | |
| "grad_norm": 0.30696389079093933, | |
| "learning_rate": 7.511477331356733e-05, | |
| "loss": 0.8821, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.4348958333333333, | |
| "grad_norm": 0.3090788424015045, | |
| "learning_rate": 7.434882195349736e-05, | |
| "loss": 0.8593, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.4361979166666667, | |
| "grad_norm": 0.32818740606307983, | |
| "learning_rate": 7.358451456362751e-05, | |
| "loss": 0.8474, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4375, | |
| "grad_norm": 0.33500784635543823, | |
| "learning_rate": 7.282189829003785e-05, | |
| "loss": 0.7979, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.4388020833333333, | |
| "grad_norm": 0.37808653712272644, | |
| "learning_rate": 7.206102017449237e-05, | |
| "loss": 0.857, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.4401041666666667, | |
| "grad_norm": 0.31529998779296875, | |
| "learning_rate": 7.130192715153731e-05, | |
| "loss": 0.8831, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.44140625, | |
| "grad_norm": 0.2927025556564331, | |
| "learning_rate": 7.054466604560595e-05, | |
| "loss": 0.8116, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.4427083333333333, | |
| "grad_norm": 0.3082600235939026, | |
| "learning_rate": 6.978928356813031e-05, | |
| "loss": 0.796, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.4440104166666667, | |
| "grad_norm": 0.30604058504104614, | |
| "learning_rate": 6.90358263146598e-05, | |
| "loss": 0.9001, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.4453125, | |
| "grad_norm": 0.320302814245224, | |
| "learning_rate": 6.828434076198693e-05, | |
| "loss": 0.8057, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.4466145833333333, | |
| "grad_norm": 0.34440967440605164, | |
| "learning_rate": 6.753487326528033e-05, | |
| "loss": 0.8786, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.4479166666666667, | |
| "grad_norm": 0.313021719455719, | |
| "learning_rate": 6.678747005522557e-05, | |
| "loss": 0.8573, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.44921875, | |
| "grad_norm": 0.3133790194988251, | |
| "learning_rate": 6.60421772351732e-05, | |
| "loss": 0.8614, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.4505208333333333, | |
| "grad_norm": 0.32770583033561707, | |
| "learning_rate": 6.529904077829505e-05, | |
| "loss": 0.829, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.4518229166666667, | |
| "grad_norm": 0.327179491519928, | |
| "learning_rate": 6.455810652474817e-05, | |
| "loss": 0.8383, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.453125, | |
| "grad_norm": 0.30281123518943787, | |
| "learning_rate": 6.381942017884753e-05, | |
| "loss": 0.871, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.4544270833333333, | |
| "grad_norm": 0.3344803750514984, | |
| "learning_rate": 6.30830273062464e-05, | |
| "loss": 0.7489, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.4557291666666667, | |
| "grad_norm": 0.3110140860080719, | |
| "learning_rate": 6.234897333112594e-05, | |
| "loss": 0.8757, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.45703125, | |
| "grad_norm": 0.33310666680336, | |
| "learning_rate": 6.161730353339302e-05, | |
| "loss": 0.9292, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.4583333333333333, | |
| "grad_norm": 0.3537331521511078, | |
| "learning_rate": 6.088806304588717e-05, | |
| "loss": 0.8505, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.4596354166666667, | |
| "grad_norm": 0.3160480856895447, | |
| "learning_rate": 6.0161296851596766e-05, | |
| "loss": 0.867, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.4609375, | |
| "grad_norm": 0.32747846841812134, | |
| "learning_rate": 5.943704978088402e-05, | |
| "loss": 0.8153, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.4622395833333333, | |
| "grad_norm": 0.342339426279068, | |
| "learning_rate": 5.871536650871979e-05, | |
| "loss": 0.8754, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.4635416666666667, | |
| "grad_norm": 0.31615254282951355, | |
| "learning_rate": 5.7996291551927666e-05, | |
| "loss": 0.8338, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.46484375, | |
| "grad_norm": 0.31168562173843384, | |
| "learning_rate": 5.7279869266438234e-05, | |
| "loss": 0.7774, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.4661458333333333, | |
| "grad_norm": 0.31545719504356384, | |
| "learning_rate": 5.656614384455257e-05, | |
| "loss": 0.8077, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.4674479166666667, | |
| "grad_norm": 0.31938084959983826, | |
| "learning_rate": 5.585515931221677e-05, | |
| "loss": 0.8627, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.46875, | |
| "grad_norm": 0.30571112036705017, | |
| "learning_rate": 5.514695952630578e-05, | |
| "loss": 0.8631, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4700520833333333, | |
| "grad_norm": 0.31153225898742676, | |
| "learning_rate": 5.444158817191832e-05, | |
| "loss": 0.8503, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.4713541666666667, | |
| "grad_norm": 0.32617253065109253, | |
| "learning_rate": 5.373908875968211e-05, | |
| "loss": 0.8148, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.47265625, | |
| "grad_norm": 0.3407813608646393, | |
| "learning_rate": 5.3039504623069965e-05, | |
| "loss": 0.8903, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.4739583333333333, | |
| "grad_norm": 0.32965362071990967, | |
| "learning_rate": 5.234287891572674e-05, | |
| "loss": 0.8629, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.4752604166666667, | |
| "grad_norm": 0.3094812035560608, | |
| "learning_rate": 5.164925460880758e-05, | |
| "loss": 0.8478, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.4765625, | |
| "grad_norm": 0.3095204532146454, | |
| "learning_rate": 5.095867448832683e-05, | |
| "loss": 0.9215, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.4778645833333333, | |
| "grad_norm": 0.3161752223968506, | |
| "learning_rate": 5.027118115251938e-05, | |
| "loss": 0.8831, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.4791666666666667, | |
| "grad_norm": 0.2942347824573517, | |
| "learning_rate": 4.95868170092125e-05, | |
| "loss": 0.8385, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.48046875, | |
| "grad_norm": 0.3367157280445099, | |
| "learning_rate": 4.890562427321021e-05, | |
| "loss": 0.8329, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.4817708333333333, | |
| "grad_norm": 0.3312656283378601, | |
| "learning_rate": 4.822764496368917e-05, | |
| "loss": 0.8909, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4830729166666667, | |
| "grad_norm": 0.2954017221927643, | |
| "learning_rate": 4.755292090160676e-05, | |
| "loss": 0.82, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.484375, | |
| "grad_norm": 0.28649380803108215, | |
| "learning_rate": 4.6881493707121315e-05, | |
| "loss": 0.8349, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.4856770833333333, | |
| "grad_norm": 0.3034970462322235, | |
| "learning_rate": 4.621340479702503e-05, | |
| "loss": 0.8533, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.4869791666666667, | |
| "grad_norm": 0.3515808582305908, | |
| "learning_rate": 4.554869538218868e-05, | |
| "loss": 0.8255, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.48828125, | |
| "grad_norm": 0.3340248763561249, | |
| "learning_rate": 4.48874064650202e-05, | |
| "loss": 0.7953, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4895833333333333, | |
| "grad_norm": 0.3169967532157898, | |
| "learning_rate": 4.422957883693483e-05, | |
| "loss": 0.7667, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.4908854166666667, | |
| "grad_norm": 0.33045902848243713, | |
| "learning_rate": 4.357525307583933e-05, | |
| "loss": 0.8221, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.4921875, | |
| "grad_norm": 0.27641358971595764, | |
| "learning_rate": 4.29244695436289e-05, | |
| "loss": 0.8908, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.4934895833333333, | |
| "grad_norm": 0.34106072783470154, | |
| "learning_rate": 4.227726838369711e-05, | |
| "loss": 0.8107, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.4947916666666667, | |
| "grad_norm": 0.3046337962150574, | |
| "learning_rate": 4.1633689518460225e-05, | |
| "loss": 0.8278, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.49609375, | |
| "grad_norm": 0.3120030462741852, | |
| "learning_rate": 4.0993772646894116e-05, | |
| "loss": 0.8378, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.4973958333333333, | |
| "grad_norm": 0.3395197093486786, | |
| "learning_rate": 4.035755724208573e-05, | |
| "loss": 0.8597, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.4986979166666667, | |
| "grad_norm": 0.33718255162239075, | |
| "learning_rate": 3.972508254879805e-05, | |
| "loss": 0.7818, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.3265829086303711, | |
| "learning_rate": 3.90963875810494e-05, | |
| "loss": 0.8203, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.5013020833333334, | |
| "grad_norm": 0.3368302583694458, | |
| "learning_rate": 3.847151111970676e-05, | |
| "loss": 0.8319, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5026041666666666, | |
| "grad_norm": 0.39881670475006104, | |
| "learning_rate": 3.785049171009381e-05, | |
| "loss": 0.7538, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.50390625, | |
| "grad_norm": 0.31735971570014954, | |
| "learning_rate": 3.723336765961285e-05, | |
| "loss": 0.8627, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.5052083333333334, | |
| "grad_norm": 0.2991076707839966, | |
| "learning_rate": 3.662017703538234e-05, | |
| "loss": 0.8189, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.5065104166666666, | |
| "grad_norm": 0.3316771686077118, | |
| "learning_rate": 3.601095766188833e-05, | |
| "loss": 0.8321, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.5078125, | |
| "grad_norm": 0.32285547256469727, | |
| "learning_rate": 3.540574711865146e-05, | |
| "loss": 0.8444, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5091145833333334, | |
| "grad_norm": 0.30543044209480286, | |
| "learning_rate": 3.4804582737908825e-05, | |
| "loss": 0.8559, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.5104166666666666, | |
| "grad_norm": 0.36356985569000244, | |
| "learning_rate": 3.420750160231118e-05, | |
| "loss": 0.7583, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.51171875, | |
| "grad_norm": 0.3065100908279419, | |
| "learning_rate": 3.361454054263541e-05, | |
| "loss": 0.8257, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.5130208333333334, | |
| "grad_norm": 0.3157348334789276, | |
| "learning_rate": 3.302573613551292e-05, | |
| "loss": 0.8502, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.5143229166666666, | |
| "grad_norm": 0.337979257106781, | |
| "learning_rate": 3.244112470117288e-05, | |
| "loss": 0.8172, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.515625, | |
| "grad_norm": 0.34926459193229675, | |
| "learning_rate": 3.186074230120244e-05, | |
| "loss": 0.8366, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.5169270833333334, | |
| "grad_norm": 0.2944015860557556, | |
| "learning_rate": 3.1284624736321846e-05, | |
| "loss": 0.8571, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.5182291666666666, | |
| "grad_norm": 0.27319666743278503, | |
| "learning_rate": 3.071280754417626e-05, | |
| "loss": 0.8694, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.51953125, | |
| "grad_norm": 0.3642820417881012, | |
| "learning_rate": 3.0145325997143577e-05, | |
| "loss": 0.8504, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.5208333333333334, | |
| "grad_norm": 0.33083510398864746, | |
| "learning_rate": 2.9582215100158706e-05, | |
| "loss": 0.8655, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5221354166666666, | |
| "grad_norm": 0.32897526025772095, | |
| "learning_rate": 2.902350958855426e-05, | |
| "loss": 0.9169, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.5234375, | |
| "grad_norm": 0.3111579418182373, | |
| "learning_rate": 2.846924392591794e-05, | |
| "loss": 0.869, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.5247395833333334, | |
| "grad_norm": 0.31903326511383057, | |
| "learning_rate": 2.791945230196663e-05, | |
| "loss": 0.7606, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.5260416666666666, | |
| "grad_norm": 0.32733333110809326, | |
| "learning_rate": 2.7374168630437456e-05, | |
| "loss": 0.815, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.52734375, | |
| "grad_norm": 0.3421482443809509, | |
| "learning_rate": 2.6833426546995782e-05, | |
| "loss": 0.7627, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.5286458333333334, | |
| "grad_norm": 0.3265641927719116, | |
| "learning_rate": 2.629725940716041e-05, | |
| "loss": 0.8013, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.5299479166666666, | |
| "grad_norm": 0.3222253918647766, | |
| "learning_rate": 2.57657002842462e-05, | |
| "loss": 0.83, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.53125, | |
| "grad_norm": 0.36147239804267883, | |
| "learning_rate": 2.523878196732358e-05, | |
| "loss": 0.8342, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.5325520833333334, | |
| "grad_norm": 0.30930349230766296, | |
| "learning_rate": 2.4716536959196462e-05, | |
| "loss": 0.9737, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.5338541666666666, | |
| "grad_norm": 0.33043771982192993, | |
| "learning_rate": 2.4198997474396877e-05, | |
| "loss": 0.7793, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.53515625, | |
| "grad_norm": 0.3458073139190674, | |
| "learning_rate": 2.3686195437198112e-05, | |
| "loss": 0.8822, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.5364583333333334, | |
| "grad_norm": 0.31358662247657776, | |
| "learning_rate": 2.31781624796453e-05, | |
| "loss": 0.8938, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.5377604166666666, | |
| "grad_norm": 0.32773175835609436, | |
| "learning_rate": 2.2674929939604332e-05, | |
| "loss": 0.8978, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.5390625, | |
| "grad_norm": 0.3206295073032379, | |
| "learning_rate": 2.217652885882869e-05, | |
| "loss": 0.8898, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.5403645833333334, | |
| "grad_norm": 0.31542304158210754, | |
| "learning_rate": 2.1682989981044783e-05, | |
| "loss": 0.8635, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.5416666666666666, | |
| "grad_norm": 0.3245406150817871, | |
| "learning_rate": 2.119434375005527e-05, | |
| "loss": 0.8411, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.54296875, | |
| "grad_norm": 0.3141809105873108, | |
| "learning_rate": 2.071062030786149e-05, | |
| "loss": 0.7775, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.5442708333333334, | |
| "grad_norm": 0.2986677885055542, | |
| "learning_rate": 2.0231849492803852e-05, | |
| "loss": 0.8618, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.5455729166666666, | |
| "grad_norm": 0.3049768805503845, | |
| "learning_rate": 1.9758060837721467e-05, | |
| "loss": 0.8662, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.546875, | |
| "grad_norm": 0.2944938838481903, | |
| "learning_rate": 1.928928356813032e-05, | |
| "loss": 0.8116, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5481770833333334, | |
| "grad_norm": 0.31740906834602356, | |
| "learning_rate": 1.882554660042052e-05, | |
| "loss": 0.8354, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.5494791666666666, | |
| "grad_norm": 0.3051731288433075, | |
| "learning_rate": 1.8366878540072614e-05, | |
| "loss": 0.8606, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.55078125, | |
| "grad_norm": 0.3050106465816498, | |
| "learning_rate": 1.7913307679893173e-05, | |
| "loss": 0.8115, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.5520833333333334, | |
| "grad_norm": 0.2921428084373474, | |
| "learning_rate": 1.7464861998269243e-05, | |
| "loss": 0.8507, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.5533854166666666, | |
| "grad_norm": 0.3205814063549042, | |
| "learning_rate": 1.702156915744292e-05, | |
| "loss": 0.8631, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.5546875, | |
| "grad_norm": 0.33637183904647827, | |
| "learning_rate": 1.6583456501804725e-05, | |
| "loss": 0.8691, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.5559895833333334, | |
| "grad_norm": 0.31378889083862305, | |
| "learning_rate": 1.6150551056206867e-05, | |
| "loss": 0.8181, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.5572916666666666, | |
| "grad_norm": 0.33184900879859924, | |
| "learning_rate": 1.57228795242965e-05, | |
| "loss": 0.7768, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.55859375, | |
| "grad_norm": 0.3419734537601471, | |
| "learning_rate": 1.5300468286868137e-05, | |
| "loss": 0.8905, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.5598958333333334, | |
| "grad_norm": 0.33187946677207947, | |
| "learning_rate": 1.488334340023669e-05, | |
| "loss": 0.8324, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5611979166666666, | |
| "grad_norm": 0.30014726519584656, | |
| "learning_rate": 1.4471530594629996e-05, | |
| "loss": 0.8708, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.5625, | |
| "grad_norm": 0.27040472626686096, | |
| "learning_rate": 1.4065055272601703e-05, | |
| "loss": 0.8852, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.5638020833333334, | |
| "grad_norm": 0.31325167417526245, | |
| "learning_rate": 1.3663942507464348e-05, | |
| "loss": 0.7781, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.5651041666666666, | |
| "grad_norm": 0.3078592121601105, | |
| "learning_rate": 1.3268217041742701e-05, | |
| "loss": 0.84, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.56640625, | |
| "grad_norm": 0.32804566621780396, | |
| "learning_rate": 1.2877903285647486e-05, | |
| "loss": 0.8676, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.5677083333333334, | |
| "grad_norm": 0.3207748234272003, | |
| "learning_rate": 1.2493025315569801e-05, | |
| "loss": 0.858, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.5690104166666666, | |
| "grad_norm": 0.3700745105743408, | |
| "learning_rate": 1.2113606872595673e-05, | |
| "loss": 0.9087, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.5703125, | |
| "grad_norm": 0.33619681000709534, | |
| "learning_rate": 1.173967136104196e-05, | |
| "loss": 0.8805, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.5716145833333334, | |
| "grad_norm": 0.3331134617328644, | |
| "learning_rate": 1.1371241847012401e-05, | |
| "loss": 0.8811, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.5729166666666666, | |
| "grad_norm": 0.3446657061576843, | |
| "learning_rate": 1.1008341056974854e-05, | |
| "loss": 0.8863, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.57421875, | |
| "grad_norm": 0.298917680978775, | |
| "learning_rate": 1.0650991376359473e-05, | |
| "loss": 0.7943, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.5755208333333334, | |
| "grad_norm": 0.36596038937568665, | |
| "learning_rate": 1.029921484817783e-05, | |
| "loss": 0.7989, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.5768229166666666, | |
| "grad_norm": 0.3092229664325714, | |
| "learning_rate": 9.953033171663175e-06, | |
| "loss": 0.9328, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.578125, | |
| "grad_norm": 0.2970651090145111, | |
| "learning_rate": 9.612467700932045e-06, | |
| "loss": 0.7675, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.5794270833333334, | |
| "grad_norm": 0.3474785089492798, | |
| "learning_rate": 9.277539443666783e-06, | |
| "loss": 0.846, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.5807291666666666, | |
| "grad_norm": 0.28902187943458557, | |
| "learning_rate": 8.948269059820025e-06, | |
| "loss": 0.8712, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.58203125, | |
| "grad_norm": 0.3022470772266388, | |
| "learning_rate": 8.624676860340025e-06, | |
| "loss": 0.8763, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.5833333333333334, | |
| "grad_norm": 0.2670323848724365, | |
| "learning_rate": 8.306782805917904e-06, | |
| "loss": 0.886, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.5846354166666666, | |
| "grad_norm": 0.2715941369533539, | |
| "learning_rate": 7.994606505756355e-06, | |
| "loss": 0.8477, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.5859375, | |
| "grad_norm": 0.32153239846229553, | |
| "learning_rate": 7.68816721636004e-06, | |
| "loss": 0.8436, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5872395833333334, | |
| "grad_norm": 0.33201107382774353, | |
| "learning_rate": 7.3874838403478e-06, | |
| "loss": 0.887, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.5885416666666666, | |
| "grad_norm": 0.3307580053806305, | |
| "learning_rate": 7.092574925286614e-06, | |
| "loss": 0.8239, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.58984375, | |
| "grad_norm": 0.31865382194519043, | |
| "learning_rate": 6.803458662547507e-06, | |
| "loss": 0.8073, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.5911458333333334, | |
| "grad_norm": 0.30783191323280334, | |
| "learning_rate": 6.520152886183406e-06, | |
| "loss": 0.8604, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.5924479166666666, | |
| "grad_norm": 0.32232406735420227, | |
| "learning_rate": 6.242675071829111e-06, | |
| "loss": 0.8612, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.59375, | |
| "grad_norm": 0.35587257146835327, | |
| "learning_rate": 5.971042335623229e-06, | |
| "loss": 0.8706, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.5950520833333334, | |
| "grad_norm": 0.2914402484893799, | |
| "learning_rate": 5.705271433152458e-06, | |
| "loss": 0.8478, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.5963541666666666, | |
| "grad_norm": 0.36631619930267334, | |
| "learning_rate": 5.445378758417925e-06, | |
| "loss": 0.7445, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.59765625, | |
| "grad_norm": 0.31569549441337585, | |
| "learning_rate": 5.191380342824035e-06, | |
| "loss": 0.8365, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.5989583333333334, | |
| "grad_norm": 0.31928232312202454, | |
| "learning_rate": 4.943291854189493e-06, | |
| "loss": 0.8951, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6002604166666666, | |
| "grad_norm": 0.31012532114982605, | |
| "learning_rate": 4.701128595780878e-06, | |
| "loss": 0.9263, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.6015625, | |
| "grad_norm": 0.30832427740097046, | |
| "learning_rate": 4.464905505368658e-06, | |
| "loss": 0.8536, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.6028645833333334, | |
| "grad_norm": 0.32320111989974976, | |
| "learning_rate": 4.23463715430577e-06, | |
| "loss": 0.8638, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.6041666666666666, | |
| "grad_norm": 0.3501695990562439, | |
| "learning_rate": 4.010337746628751e-06, | |
| "loss": 0.7462, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.60546875, | |
| "grad_norm": 0.3446730673313141, | |
| "learning_rate": 3.792021118181636e-06, | |
| "loss": 0.8484, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.6067708333333334, | |
| "grad_norm": 0.3519136309623718, | |
| "learning_rate": 3.5797007357623945e-06, | |
| "loss": 0.8201, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.6080729166666666, | |
| "grad_norm": 0.30372154712677, | |
| "learning_rate": 3.3733896962923658e-06, | |
| "loss": 0.8871, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.609375, | |
| "grad_norm": 0.3335200250148773, | |
| "learning_rate": 3.1731007260082616e-06, | |
| "loss": 0.8273, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.6106770833333334, | |
| "grad_norm": 0.3028860092163086, | |
| "learning_rate": 2.9788461796772114e-06, | |
| "loss": 0.8267, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.6119791666666666, | |
| "grad_norm": 0.33701807260513306, | |
| "learning_rate": 2.790638039834668e-06, | |
| "loss": 0.8493, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.61328125, | |
| "grad_norm": 0.3008849620819092, | |
| "learning_rate": 2.6084879160452166e-06, | |
| "loss": 0.7696, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.6145833333333334, | |
| "grad_norm": 0.37394464015960693, | |
| "learning_rate": 2.432407044186509e-06, | |
| "loss": 0.939, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.6158854166666666, | |
| "grad_norm": 0.3297758102416992, | |
| "learning_rate": 2.26240628575615e-06, | |
| "loss": 0.8087, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.6171875, | |
| "grad_norm": 0.28165456652641296, | |
| "learning_rate": 2.098496127201648e-06, | |
| "loss": 0.8894, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.6184895833333334, | |
| "grad_norm": 0.315613716840744, | |
| "learning_rate": 1.9406866792737267e-06, | |
| "loss": 0.8807, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.6197916666666666, | |
| "grad_norm": 0.3206324279308319, | |
| "learning_rate": 1.7889876764024505e-06, | |
| "loss": 0.8192, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.62109375, | |
| "grad_norm": 0.3201400339603424, | |
| "learning_rate": 1.6434084760968697e-06, | |
| "loss": 0.8754, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.6223958333333334, | |
| "grad_norm": 0.3509829640388489, | |
| "learning_rate": 1.5039580583678393e-06, | |
| "loss": 0.86, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.6236979166666666, | |
| "grad_norm": 0.2788860499858856, | |
| "learning_rate": 1.3706450251739613e-06, | |
| "loss": 0.8008, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.3174276649951935, | |
| "learning_rate": 1.2434775998910964e-06, | |
| "loss": 0.8038, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6263020833333334, | |
| "grad_norm": 0.3377770483493805, | |
| "learning_rate": 1.1224636268050439e-06, | |
| "loss": 0.8609, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.6276041666666666, | |
| "grad_norm": 0.29502376914024353, | |
| "learning_rate": 1.0076105706276888e-06, | |
| "loss": 0.8764, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.62890625, | |
| "grad_norm": 0.30970874428749084, | |
| "learning_rate": 8.989255160365527e-07, | |
| "loss": 0.8442, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.6302083333333334, | |
| "grad_norm": 0.3286643624305725, | |
| "learning_rate": 7.964151672377458e-07, | |
| "loss": 0.7697, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.6315104166666666, | |
| "grad_norm": 0.32877275347709656, | |
| "learning_rate": 7.000858475524444e-07, | |
| "loss": 0.8386, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.6328125, | |
| "grad_norm": 0.31582364439964294, | |
| "learning_rate": 6.099434990268609e-07, | |
| "loss": 0.799, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.6341145833333334, | |
| "grad_norm": 0.32822880148887634, | |
| "learning_rate": 5.259936820656257e-07, | |
| "loss": 0.7853, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.6354166666666666, | |
| "grad_norm": 0.31782621145248413, | |
| "learning_rate": 4.482415750889204e-07, | |
| "loss": 0.8704, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.63671875, | |
| "grad_norm": 0.30536094307899475, | |
| "learning_rate": 3.766919742129331e-07, | |
| "loss": 0.8136, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.6380208333333334, | |
| "grad_norm": 0.31908681988716125, | |
| "learning_rate": 3.1134929295407564e-07, | |
| "loss": 0.8366, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6393229166666666, | |
| "grad_norm": 0.32564041018486023, | |
| "learning_rate": 2.5221756195672563e-07, | |
| "loss": 0.7645, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.640625, | |
| "grad_norm": 0.30761629343032837, | |
| "learning_rate": 1.9930042874457254e-07, | |
| "loss": 0.8334, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.6419270833333334, | |
| "grad_norm": 0.31079787015914917, | |
| "learning_rate": 1.5260115749566882e-07, | |
| "loss": 0.8626, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.6432291666666666, | |
| "grad_norm": 0.34176650643348694, | |
| "learning_rate": 1.1212262884103974e-07, | |
| "loss": 0.69, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.64453125, | |
| "grad_norm": 0.31278514862060547, | |
| "learning_rate": 7.7867339686987e-08, | |
| "loss": 0.8859, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.6458333333333334, | |
| "grad_norm": 0.30534476041793823, | |
| "learning_rate": 4.98374030611084e-08, | |
| "loss": 0.8918, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.6471354166666666, | |
| "grad_norm": 0.3080224096775055, | |
| "learning_rate": 2.8034547981943713e-08, | |
| "loss": 0.8508, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.6484375, | |
| "grad_norm": 0.3025684654712677, | |
| "learning_rate": 1.246011935228064e-08, | |
| "loss": 0.8545, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.6497395833333334, | |
| "grad_norm": 0.3287234604358673, | |
| "learning_rate": 3.115077876243988e-09, | |
| "loss": 0.8297, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.6510416666666666, | |
| "grad_norm": 0.34251466393470764, | |
| "learning_rate": 0.0, | |
| "loss": 0.8901, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 3.827040194259845e+17, | |
| "train_batch_size": 6, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |