| { | |
| "best_metric": 7.651927947998047, | |
| "best_model_checkpoint": "/storage4/output/shiraishi/tmp_0426_zero2_1/checkpoint-200", | |
| "epoch": 0.030562347188264057, | |
| "eval_steps": 200, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 6.783477783203125, | |
| "learning_rate": 3.053435114503817e-08, | |
| "loss": 11.3891, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.414793014526367, | |
| "learning_rate": 6.106870229007634e-08, | |
| "loss": 11.3744, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.458027362823486, | |
| "learning_rate": 9.160305343511451e-08, | |
| "loss": 11.3815, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 9.68161392211914, | |
| "learning_rate": 1.2213740458015268e-07, | |
| "loss": 11.393, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.597453594207764, | |
| "learning_rate": 1.5267175572519085e-07, | |
| "loss": 11.4053, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.35366439819336, | |
| "learning_rate": 1.8320610687022902e-07, | |
| "loss": 11.3801, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.251340389251709, | |
| "learning_rate": 2.137404580152672e-07, | |
| "loss": 11.3817, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.854252815246582, | |
| "learning_rate": 2.4427480916030536e-07, | |
| "loss": 11.3886, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.2333526611328125, | |
| "learning_rate": 2.7480916030534355e-07, | |
| "loss": 11.3566, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.312106609344482, | |
| "learning_rate": 3.053435114503817e-07, | |
| "loss": 11.3452, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.145501136779785, | |
| "learning_rate": 3.358778625954199e-07, | |
| "loss": 11.4325, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.581705570220947, | |
| "learning_rate": 3.6641221374045804e-07, | |
| "loss": 11.3761, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.878636360168457, | |
| "learning_rate": 3.9694656488549623e-07, | |
| "loss": 11.3896, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 6.950926780700684, | |
| "learning_rate": 4.274809160305344e-07, | |
| "loss": 11.3989, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 10.297870635986328, | |
| "learning_rate": 4.5801526717557257e-07, | |
| "loss": 11.4054, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 11.376952171325684, | |
| "learning_rate": 4.885496183206107e-07, | |
| "loss": 11.3834, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.478097438812256, | |
| "learning_rate": 5.190839694656489e-07, | |
| "loss": 11.3697, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.463815689086914, | |
| "learning_rate": 5.496183206106871e-07, | |
| "loss": 11.4119, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.785317897796631, | |
| "learning_rate": 5.801526717557252e-07, | |
| "loss": 11.3758, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 6.733437538146973, | |
| "learning_rate": 6.106870229007634e-07, | |
| "loss": 11.3886, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.351161003112793, | |
| "learning_rate": 6.412213740458016e-07, | |
| "loss": 11.3935, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.157550811767578, | |
| "learning_rate": 6.717557251908398e-07, | |
| "loss": 11.3778, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.681120872497559, | |
| "learning_rate": 7.022900763358779e-07, | |
| "loss": 11.3726, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.128857135772705, | |
| "learning_rate": 7.328244274809161e-07, | |
| "loss": 11.3474, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 6.945237159729004, | |
| "learning_rate": 7.633587786259543e-07, | |
| "loss": 11.3574, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.129561901092529, | |
| "learning_rate": 7.938931297709925e-07, | |
| "loss": 11.3895, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.440257549285889, | |
| "learning_rate": 8.244274809160306e-07, | |
| "loss": 11.3618, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.143465518951416, | |
| "learning_rate": 8.549618320610687e-07, | |
| "loss": 11.3794, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 10.438725471496582, | |
| "learning_rate": 8.85496183206107e-07, | |
| "loss": 11.3828, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.326735019683838, | |
| "learning_rate": 9.160305343511451e-07, | |
| "loss": 11.3828, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.470747947692871, | |
| "learning_rate": 9.465648854961833e-07, | |
| "loss": 11.3717, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 7.645094871520996, | |
| "learning_rate": 9.770992366412214e-07, | |
| "loss": 11.3616, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.069401741027832, | |
| "learning_rate": 1.0076335877862597e-06, | |
| "loss": 11.3415, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 11.976044654846191, | |
| "learning_rate": 1.0381679389312977e-06, | |
| "loss": 11.3767, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.605710029602051, | |
| "learning_rate": 1.068702290076336e-06, | |
| "loss": 11.3258, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.739644050598145, | |
| "learning_rate": 1.0992366412213742e-06, | |
| "loss": 11.3289, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.319056034088135, | |
| "learning_rate": 1.1297709923664122e-06, | |
| "loss": 11.3485, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.979043006896973, | |
| "learning_rate": 1.1603053435114505e-06, | |
| "loss": 11.3434, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.536917686462402, | |
| "learning_rate": 1.1908396946564885e-06, | |
| "loss": 11.3105, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.834555625915527, | |
| "learning_rate": 1.2213740458015268e-06, | |
| "loss": 11.3234, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.229254245758057, | |
| "learning_rate": 1.251908396946565e-06, | |
| "loss": 11.2945, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.75545072555542, | |
| "learning_rate": 1.2824427480916033e-06, | |
| "loss": 11.3162, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.102319240570068, | |
| "learning_rate": 1.3129770992366413e-06, | |
| "loss": 11.3081, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.633400917053223, | |
| "learning_rate": 1.3435114503816796e-06, | |
| "loss": 11.2669, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.306519031524658, | |
| "learning_rate": 1.3740458015267178e-06, | |
| "loss": 11.2839, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.957504749298096, | |
| "learning_rate": 1.4045801526717559e-06, | |
| "loss": 11.2745, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.333808898925781, | |
| "learning_rate": 1.435114503816794e-06, | |
| "loss": 11.2366, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.642115592956543, | |
| "learning_rate": 1.4656488549618321e-06, | |
| "loss": 11.2738, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.175962448120117, | |
| "learning_rate": 1.4961832061068704e-06, | |
| "loss": 11.2295, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.367372512817383, | |
| "learning_rate": 1.5267175572519086e-06, | |
| "loss": 11.2532, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.235678195953369, | |
| "learning_rate": 1.5572519083969467e-06, | |
| "loss": 11.2107, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.395421981811523, | |
| "learning_rate": 1.587786259541985e-06, | |
| "loss": 11.2046, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.745922088623047, | |
| "learning_rate": 1.618320610687023e-06, | |
| "loss": 11.1656, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.334342002868652, | |
| "learning_rate": 1.6488549618320612e-06, | |
| "loss": 11.1305, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.523844242095947, | |
| "learning_rate": 1.6793893129770995e-06, | |
| "loss": 11.1534, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 16.16385269165039, | |
| "learning_rate": 1.7099236641221375e-06, | |
| "loss": 11.0576, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.178238868713379, | |
| "learning_rate": 1.7404580152671757e-06, | |
| "loss": 11.0708, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.8342742919921875, | |
| "learning_rate": 1.770992366412214e-06, | |
| "loss": 11.1786, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 12.120439529418945, | |
| "learning_rate": 1.801526717557252e-06, | |
| "loss": 11.0226, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.631428241729736, | |
| "learning_rate": 1.8320610687022903e-06, | |
| "loss": 11.0507, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.0986104011535645, | |
| "learning_rate": 1.8625954198473283e-06, | |
| "loss": 11.0972, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.366494655609131, | |
| "learning_rate": 1.8931297709923666e-06, | |
| "loss": 11.0679, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.760643005371094, | |
| "learning_rate": 1.923664122137405e-06, | |
| "loss": 10.981, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.623432159423828, | |
| "learning_rate": 1.954198473282443e-06, | |
| "loss": 10.9184, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.003146171569824, | |
| "learning_rate": 1.984732824427481e-06, | |
| "loss": 10.8968, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.153034210205078, | |
| "learning_rate": 2.0152671755725194e-06, | |
| "loss": 10.859, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.09888219833374, | |
| "learning_rate": 2.0458015267175574e-06, | |
| "loss": 10.8728, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.097289085388184, | |
| "learning_rate": 2.0763358778625954e-06, | |
| "loss": 10.8831, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.85009765625, | |
| "learning_rate": 2.106870229007634e-06, | |
| "loss": 10.8421, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.426596641540527, | |
| "learning_rate": 2.137404580152672e-06, | |
| "loss": 10.821, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.099856376647949, | |
| "learning_rate": 2.16793893129771e-06, | |
| "loss": 10.8648, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.625555038452148, | |
| "learning_rate": 2.1984732824427484e-06, | |
| "loss": 10.6711, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.467529296875, | |
| "learning_rate": 2.2290076335877865e-06, | |
| "loss": 10.7989, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.075023651123047, | |
| "learning_rate": 2.2595419847328245e-06, | |
| "loss": 10.6594, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.846537113189697, | |
| "learning_rate": 2.2900763358778625e-06, | |
| "loss": 10.7858, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 9.528846740722656, | |
| "learning_rate": 2.320610687022901e-06, | |
| "loss": 10.5162, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.104434013366699, | |
| "learning_rate": 2.351145038167939e-06, | |
| "loss": 10.7615, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.450407028198242, | |
| "learning_rate": 2.381679389312977e-06, | |
| "loss": 10.6087, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.1212968826293945, | |
| "learning_rate": 2.4122137404580155e-06, | |
| "loss": 10.577, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.871504783630371, | |
| "learning_rate": 2.4427480916030536e-06, | |
| "loss": 10.5229, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.003294467926025, | |
| "learning_rate": 2.4732824427480916e-06, | |
| "loss": 10.5099, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.826551914215088, | |
| "learning_rate": 2.50381679389313e-06, | |
| "loss": 10.5681, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.307621955871582, | |
| "learning_rate": 2.534351145038168e-06, | |
| "loss": 10.3494, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.38472318649292, | |
| "learning_rate": 2.5648854961832066e-06, | |
| "loss": 10.2822, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.0979461669921875, | |
| "learning_rate": 2.595419847328244e-06, | |
| "loss": 10.4044, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.168255805969238, | |
| "learning_rate": 2.6259541984732826e-06, | |
| "loss": 10.2981, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.889970779418945, | |
| "learning_rate": 2.6564885496183207e-06, | |
| "loss": 10.3901, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.614548206329346, | |
| "learning_rate": 2.687022900763359e-06, | |
| "loss": 10.3475, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.925837993621826, | |
| "learning_rate": 2.717557251908397e-06, | |
| "loss": 10.338, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.6765899658203125, | |
| "learning_rate": 2.7480916030534356e-06, | |
| "loss": 10.264, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.884031295776367, | |
| "learning_rate": 2.7786259541984732e-06, | |
| "loss": 9.8399, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.458746433258057, | |
| "learning_rate": 2.8091603053435117e-06, | |
| "loss": 10.1623, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.800764560699463, | |
| "learning_rate": 2.8396946564885497e-06, | |
| "loss": 9.9704, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.747128486633301, | |
| "learning_rate": 2.870229007633588e-06, | |
| "loss": 9.752, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.7548346519470215, | |
| "learning_rate": 2.900763358778626e-06, | |
| "loss": 9.8412, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.906667709350586, | |
| "learning_rate": 2.9312977099236643e-06, | |
| "loss": 9.8227, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.52130126953125, | |
| "learning_rate": 2.9618320610687023e-06, | |
| "loss": 9.8856, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.0463080406188965, | |
| "learning_rate": 2.9923664122137408e-06, | |
| "loss": 9.9391, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.734062671661377, | |
| "learning_rate": 3.022900763358779e-06, | |
| "loss": 9.7216, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.168327331542969, | |
| "learning_rate": 3.0534351145038173e-06, | |
| "loss": 9.6673, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.75146484375, | |
| "learning_rate": 3.083969465648855e-06, | |
| "loss": 9.635, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.428467750549316, | |
| "learning_rate": 3.1145038167938934e-06, | |
| "loss": 9.4586, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.898741245269775, | |
| "learning_rate": 3.1450381679389314e-06, | |
| "loss": 9.7941, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.938511371612549, | |
| "learning_rate": 3.17557251908397e-06, | |
| "loss": 9.7426, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.729054927825928, | |
| "learning_rate": 3.206106870229008e-06, | |
| "loss": 9.451, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.630996227264404, | |
| "learning_rate": 3.236641221374046e-06, | |
| "loss": 9.5906, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.786062717437744, | |
| "learning_rate": 3.267175572519084e-06, | |
| "loss": 9.5057, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.723715782165527, | |
| "learning_rate": 3.2977099236641224e-06, | |
| "loss": 9.4507, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.253941059112549, | |
| "learning_rate": 3.3282442748091605e-06, | |
| "loss": 9.5653, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.430165767669678, | |
| "learning_rate": 3.358778625954199e-06, | |
| "loss": 9.4307, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.978307247161865, | |
| "learning_rate": 3.3893129770992365e-06, | |
| "loss": 9.3075, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.934025764465332, | |
| "learning_rate": 3.419847328244275e-06, | |
| "loss": 9.1069, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.9066596031188965, | |
| "learning_rate": 3.450381679389313e-06, | |
| "loss": 9.2956, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.814650058746338, | |
| "learning_rate": 3.4809160305343515e-06, | |
| "loss": 9.3122, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.222245216369629, | |
| "learning_rate": 3.5114503816793895e-06, | |
| "loss": 8.8238, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.756102561950684, | |
| "learning_rate": 3.541984732824428e-06, | |
| "loss": 9.2161, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.839966297149658, | |
| "learning_rate": 3.5725190839694656e-06, | |
| "loss": 8.9416, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.648271083831787, | |
| "learning_rate": 3.603053435114504e-06, | |
| "loss": 8.9465, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.638028621673584, | |
| "learning_rate": 3.633587786259542e-06, | |
| "loss": 9.0576, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.655818462371826, | |
| "learning_rate": 3.6641221374045806e-06, | |
| "loss": 9.1852, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.362784385681152, | |
| "learning_rate": 3.694656488549619e-06, | |
| "loss": 8.9361, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.316291332244873, | |
| "learning_rate": 3.7251908396946566e-06, | |
| "loss": 8.8583, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.02163553237915, | |
| "learning_rate": 3.755725190839695e-06, | |
| "loss": 9.2601, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.8954572677612305, | |
| "learning_rate": 3.786259541984733e-06, | |
| "loss": 8.7542, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.2410197257995605, | |
| "learning_rate": 3.816793893129772e-06, | |
| "loss": 8.91, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.047369003295898, | |
| "learning_rate": 3.84732824427481e-06, | |
| "loss": 8.9428, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.15311861038208, | |
| "learning_rate": 3.877862595419848e-06, | |
| "loss": 8.8639, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.029300689697266, | |
| "learning_rate": 3.908396946564886e-06, | |
| "loss": 8.8831, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.91277813911438, | |
| "learning_rate": 3.938931297709925e-06, | |
| "loss": 8.8629, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.7793161869049072, | |
| "learning_rate": 3.969465648854962e-06, | |
| "loss": 8.6063, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.8341503143310547, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 8.8206, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.8343710899353027, | |
| "learning_rate": 4.030534351145039e-06, | |
| "loss": 8.7022, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.699455738067627, | |
| "learning_rate": 4.061068702290077e-06, | |
| "loss": 8.6788, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.8112385272979736, | |
| "learning_rate": 4.091603053435115e-06, | |
| "loss": 8.5557, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.5949337482452393, | |
| "learning_rate": 4.122137404580153e-06, | |
| "loss": 8.625, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.011490345001221, | |
| "learning_rate": 4.152671755725191e-06, | |
| "loss": 8.4473, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.92791748046875, | |
| "learning_rate": 4.18320610687023e-06, | |
| "loss": 8.437, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.749657154083252, | |
| "learning_rate": 4.213740458015268e-06, | |
| "loss": 8.4635, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.6413307189941406, | |
| "learning_rate": 4.244274809160306e-06, | |
| "loss": 8.2699, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.124506950378418, | |
| "learning_rate": 4.274809160305344e-06, | |
| "loss": 8.4271, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.3390684127807617, | |
| "learning_rate": 4.305343511450382e-06, | |
| "loss": 8.488, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.8475050926208496, | |
| "learning_rate": 4.33587786259542e-06, | |
| "loss": 8.5997, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.3542966842651367, | |
| "learning_rate": 4.366412213740459e-06, | |
| "loss": 8.3446, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.394378900527954, | |
| "learning_rate": 4.396946564885497e-06, | |
| "loss": 8.4899, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.418760061264038, | |
| "learning_rate": 4.427480916030535e-06, | |
| "loss": 8.6685, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.194042205810547, | |
| "learning_rate": 4.458015267175573e-06, | |
| "loss": 8.4058, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.3027408123016357, | |
| "learning_rate": 4.488549618320611e-06, | |
| "loss": 8.3844, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.279644012451172, | |
| "learning_rate": 4.519083969465649e-06, | |
| "loss": 8.3619, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.07478928565979, | |
| "learning_rate": 4.549618320610688e-06, | |
| "loss": 8.3307, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.4793453216552734, | |
| "learning_rate": 4.580152671755725e-06, | |
| "loss": 8.1942, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.0340731143951416, | |
| "learning_rate": 4.610687022900764e-06, | |
| "loss": 8.441, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.8235411643981934, | |
| "learning_rate": 4.641221374045802e-06, | |
| "loss": 8.2814, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.1620304584503174, | |
| "learning_rate": 4.67175572519084e-06, | |
| "loss": 8.2211, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.142979621887207, | |
| "learning_rate": 4.702290076335878e-06, | |
| "loss": 8.1689, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.8217127323150635, | |
| "learning_rate": 4.732824427480917e-06, | |
| "loss": 8.1966, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.3655271530151367, | |
| "learning_rate": 4.763358778625954e-06, | |
| "loss": 8.484, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.0678720474243164, | |
| "learning_rate": 4.793893129770993e-06, | |
| "loss": 8.0157, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.0809926986694336, | |
| "learning_rate": 4.824427480916031e-06, | |
| "loss": 8.3355, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.9575114250183105, | |
| "learning_rate": 4.854961832061069e-06, | |
| "loss": 8.0863, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.0213122367858887, | |
| "learning_rate": 4.885496183206107e-06, | |
| "loss": 8.2348, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.777534246444702, | |
| "learning_rate": 4.916030534351145e-06, | |
| "loss": 8.1467, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.1352999210357666, | |
| "learning_rate": 4.946564885496183e-06, | |
| "loss": 7.8425, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.975677013397217, | |
| "learning_rate": 4.977099236641222e-06, | |
| "loss": 8.1678, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.865999937057495, | |
| "learning_rate": 5.00763358778626e-06, | |
| "loss": 7.7025, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.6866650581359863, | |
| "learning_rate": 5.038167938931297e-06, | |
| "loss": 7.9703, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.7522146701812744, | |
| "learning_rate": 5.068702290076336e-06, | |
| "loss": 8.0509, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.8245480060577393, | |
| "learning_rate": 5.099236641221374e-06, | |
| "loss": 8.0765, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.8313450813293457, | |
| "learning_rate": 5.129770992366413e-06, | |
| "loss": 8.1584, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.2930164337158203, | |
| "learning_rate": 5.16030534351145e-06, | |
| "loss": 7.9879, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.8687310218811035, | |
| "learning_rate": 5.190839694656488e-06, | |
| "loss": 8.0552, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.7692668437957764, | |
| "learning_rate": 5.221374045801527e-06, | |
| "loss": 8.1351, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.6235339641571045, | |
| "learning_rate": 5.251908396946565e-06, | |
| "loss": 7.8954, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.965816020965576, | |
| "learning_rate": 5.282442748091603e-06, | |
| "loss": 7.8535, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.6493630409240723, | |
| "learning_rate": 5.312977099236641e-06, | |
| "loss": 8.0987, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.6753995418548584, | |
| "learning_rate": 5.34351145038168e-06, | |
| "loss": 7.8885, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.7723007202148438, | |
| "learning_rate": 5.374045801526718e-06, | |
| "loss": 7.9147, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.466517686843872, | |
| "learning_rate": 5.4045801526717555e-06, | |
| "loss": 7.8116, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.0574333667755127, | |
| "learning_rate": 5.435114503816794e-06, | |
| "loss": 7.9566, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.5315887928009033, | |
| "learning_rate": 5.465648854961832e-06, | |
| "loss": 7.5873, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.6536879539489746, | |
| "learning_rate": 5.496183206106871e-06, | |
| "loss": 7.8048, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.183786153793335, | |
| "learning_rate": 5.526717557251909e-06, | |
| "loss": 7.7694, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.6889312267303467, | |
| "learning_rate": 5.5572519083969465e-06, | |
| "loss": 7.9259, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.6340017318725586, | |
| "learning_rate": 5.587786259541985e-06, | |
| "loss": 7.9812, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.981607437133789, | |
| "learning_rate": 5.618320610687023e-06, | |
| "loss": 7.6588, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.7326838970184326, | |
| "learning_rate": 5.648854961832062e-06, | |
| "loss": 7.3726, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.5863561630249023, | |
| "learning_rate": 5.6793893129770995e-06, | |
| "loss": 7.5351, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.467935562133789, | |
| "learning_rate": 5.7099236641221375e-06, | |
| "loss": 7.4953, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.7849318981170654, | |
| "learning_rate": 5.740458015267176e-06, | |
| "loss": 8.1349, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.6400296688079834, | |
| "learning_rate": 5.7709923664122144e-06, | |
| "loss": 7.6883, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.6569929122924805, | |
| "learning_rate": 5.801526717557252e-06, | |
| "loss": 7.6134, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.578665018081665, | |
| "learning_rate": 5.8320610687022905e-06, | |
| "loss": 7.6199, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.7117719650268555, | |
| "learning_rate": 5.8625954198473286e-06, | |
| "loss": 7.1073, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.5729572772979736, | |
| "learning_rate": 5.8931297709923674e-06, | |
| "loss": 7.6978, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.4764816761016846, | |
| "learning_rate": 5.923664122137405e-06, | |
| "loss": 7.803, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.416189432144165, | |
| "learning_rate": 5.9541984732824435e-06, | |
| "loss": 7.6627, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.5054240226745605, | |
| "learning_rate": 5.9847328244274816e-06, | |
| "loss": 7.4827, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.4899439811706543, | |
| "learning_rate": 6.0152671755725204e-06, | |
| "loss": 7.7537, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.766172409057617, | |
| "learning_rate": 6.045801526717558e-06, | |
| "loss": 7.6807, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.860541343688965, | |
| "learning_rate": 6.076335877862596e-06, | |
| "loss": 7.5989, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.411839723587036, | |
| "learning_rate": 6.1068702290076346e-06, | |
| "loss": 7.5964, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "eval_loss": 7.651927947998047, | |
| "eval_runtime": 48.4699, | |
| "eval_samples_per_second": 227.894, | |
| "eval_steps_per_second": 14.256, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 6544, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 200, | |
| "total_flos": 1.369950149148672e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |