| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 0, | |
| "global_step": 415, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0024096385542168677, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 9.975903614457833e-06, | |
| "loss": 1.8971, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004819277108433735, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 9.951807228915663e-06, | |
| "loss": 1.8183, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.007228915662650603, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 9.927710843373494e-06, | |
| "loss": 1.7341, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00963855421686747, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 9.903614457831326e-06, | |
| "loss": 1.8587, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.012048192771084338, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.879518072289156e-06, | |
| "loss": 1.8129, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014457831325301205, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.855421686746988e-06, | |
| "loss": 1.7437, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.016867469879518072, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.83132530120482e-06, | |
| "loss": 1.8127, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01927710843373494, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 9.807228915662652e-06, | |
| "loss": 1.7577, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.021686746987951807, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.783132530120483e-06, | |
| "loss": 1.7672, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.024096385542168676, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.759036144578315e-06, | |
| "loss": 1.6981, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02650602409638554, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 9.734939759036145e-06, | |
| "loss": 1.7759, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02891566265060241, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.710843373493977e-06, | |
| "loss": 1.6405, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03132530120481928, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.686746987951809e-06, | |
| "loss": 1.7534, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.033734939759036145, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 9.662650602409639e-06, | |
| "loss": 1.6553, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03614457831325301, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.63855421686747e-06, | |
| "loss": 1.7058, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03855421686746988, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.614457831325302e-06, | |
| "loss": 1.7272, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04096385542168675, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.590361445783132e-06, | |
| "loss": 1.6819, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.043373493975903614, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.566265060240964e-06, | |
| "loss": 1.7019, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04578313253012048, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.542168674698796e-06, | |
| "loss": 1.6333, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04819277108433735, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 9.518072289156628e-06, | |
| "loss": 1.5319, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05060240963855422, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.49397590361446e-06, | |
| "loss": 1.6696, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05301204819277108, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 9.46987951807229e-06, | |
| "loss": 1.6739, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05542168674698795, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 9.445783132530121e-06, | |
| "loss": 1.6864, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.05783132530120482, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 9.421686746987953e-06, | |
| "loss": 1.5915, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.060240963855421686, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 9.397590361445785e-06, | |
| "loss": 1.562, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06265060240963856, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 9.373493975903615e-06, | |
| "loss": 1.5647, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06506024096385542, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 9.349397590361446e-06, | |
| "loss": 1.6104, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06746987951807229, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 9.325301204819278e-06, | |
| "loss": 1.5392, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06987951807228916, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 9.301204819277108e-06, | |
| "loss": 1.5591, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07228915662650602, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 9.27710843373494e-06, | |
| "loss": 1.5547, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0746987951807229, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 9.253012048192772e-06, | |
| "loss": 1.513, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.07710843373493977, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 9.228915662650602e-06, | |
| "loss": 1.5069, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.07951807228915662, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 9.204819277108434e-06, | |
| "loss": 1.5193, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0819277108433735, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 9.180722891566265e-06, | |
| "loss": 1.4837, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.08433734939759036, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 9.156626506024097e-06, | |
| "loss": 1.4584, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08674698795180723, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 9.132530120481929e-06, | |
| "loss": 1.4999, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0891566265060241, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 9.10843373493976e-06, | |
| "loss": 1.4903, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.09156626506024096, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 9.08433734939759e-06, | |
| "loss": 1.4731, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.09397590361445783, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 9.060240963855423e-06, | |
| "loss": 1.4711, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0963855421686747, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 9.036144578313254e-06, | |
| "loss": 1.5278, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09879518072289156, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 9.012048192771084e-06, | |
| "loss": 1.4631, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.10120481927710843, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 8.987951807228916e-06, | |
| "loss": 1.4589, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.10361445783132531, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 8.963855421686748e-06, | |
| "loss": 1.3893, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.10602409638554217, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 8.939759036144578e-06, | |
| "loss": 1.4173, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.10843373493975904, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 8.91566265060241e-06, | |
| "loss": 1.4701, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1108433734939759, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 8.891566265060241e-06, | |
| "loss": 1.4885, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.11325301204819277, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 8.867469879518073e-06, | |
| "loss": 1.4325, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.11566265060240964, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 8.843373493975905e-06, | |
| "loss": 1.4365, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.1180722891566265, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 8.819277108433735e-06, | |
| "loss": 1.3758, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.12048192771084337, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 8.795180722891567e-06, | |
| "loss": 1.4265, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12289156626506025, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 8.771084337349399e-06, | |
| "loss": 1.3894, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.12530120481927712, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 8.74698795180723e-06, | |
| "loss": 1.4533, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12771084337349398, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 8.722891566265062e-06, | |
| "loss": 1.3708, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.13012048192771083, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.698795180722892e-06, | |
| "loss": 1.4268, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.13253012048192772, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 8.674698795180724e-06, | |
| "loss": 1.4155, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.13493975903614458, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.650602409638556e-06, | |
| "loss": 1.4563, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.13734939759036144, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 8.626506024096386e-06, | |
| "loss": 1.4198, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.13975903614457832, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 8.602409638554217e-06, | |
| "loss": 1.3974, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.14216867469879518, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 8.57831325301205e-06, | |
| "loss": 1.4379, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.14457831325301204, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 8.55421686746988e-06, | |
| "loss": 1.4232, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14698795180722893, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 8.530120481927711e-06, | |
| "loss": 1.4119, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1493975903614458, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 8.506024096385543e-06, | |
| "loss": 1.4066, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.15180722891566265, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 8.481927710843375e-06, | |
| "loss": 1.3308, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.15421686746987953, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 8.457831325301206e-06, | |
| "loss": 1.323, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.1566265060240964, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 8.433734939759038e-06, | |
| "loss": 1.4078, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15903614457831325, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 8.409638554216868e-06, | |
| "loss": 1.3577, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1614457831325301, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 8.3855421686747e-06, | |
| "loss": 1.3174, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.163855421686747, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 8.361445783132532e-06, | |
| "loss": 1.3815, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.16626506024096385, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 8.337349397590362e-06, | |
| "loss": 1.3563, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1686746987951807, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 8.313253012048194e-06, | |
| "loss": 1.4056, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1710843373493976, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 8.289156626506025e-06, | |
| "loss": 1.3572, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.17349397590361446, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 8.265060240963855e-06, | |
| "loss": 1.3202, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.17590361445783131, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 8.240963855421687e-06, | |
| "loss": 1.3166, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1783132530120482, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 8.216867469879519e-06, | |
| "loss": 1.4073, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.18072289156626506, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 8.19277108433735e-06, | |
| "loss": 1.3445, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.18313253012048192, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 8.16867469879518e-06, | |
| "loss": 1.3258, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.1855421686746988, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 8.144578313253012e-06, | |
| "loss": 1.3285, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.18795180722891566, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 8.120481927710844e-06, | |
| "loss": 1.3979, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.19036144578313252, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 8.096385542168676e-06, | |
| "loss": 1.3263, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1927710843373494, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 8.072289156626508e-06, | |
| "loss": 1.3622, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19518072289156627, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 8.048192771084338e-06, | |
| "loss": 1.3178, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.19759036144578312, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 8.02409638554217e-06, | |
| "loss": 1.3474, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.3102, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.20240963855421687, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 7.975903614457831e-06, | |
| "loss": 1.2936, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.20481927710843373, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 7.951807228915663e-06, | |
| "loss": 1.2996, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.20722891566265061, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 7.927710843373495e-06, | |
| "loss": 1.3398, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.20963855421686747, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 7.903614457831325e-06, | |
| "loss": 1.3283, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.21204819277108433, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 7.879518072289157e-06, | |
| "loss": 1.3627, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.21445783132530122, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 7.855421686746989e-06, | |
| "loss": 1.3591, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.21686746987951808, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 7.83132530120482e-06, | |
| "loss": 1.2912, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.21927710843373494, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 7.807228915662652e-06, | |
| "loss": 1.2841, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2216867469879518, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 7.783132530120484e-06, | |
| "loss": 1.3041, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.22409638554216868, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 7.759036144578314e-06, | |
| "loss": 1.3713, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.22650602409638554, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 7.734939759036146e-06, | |
| "loss": 1.2214, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2289156626506024, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 7.710843373493977e-06, | |
| "loss": 1.329, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.23132530120481928, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 7.686746987951807e-06, | |
| "loss": 1.3779, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.23373493975903614, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 7.66265060240964e-06, | |
| "loss": 1.3602, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.236144578313253, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 7.638554216867471e-06, | |
| "loss": 1.2542, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.2385542168674699, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 7.614457831325302e-06, | |
| "loss": 1.3328, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.24096385542168675, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 7.590361445783133e-06, | |
| "loss": 1.2587, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2433734939759036, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 7.5662650602409645e-06, | |
| "loss": 1.2801, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.2457831325301205, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 7.5421686746987955e-06, | |
| "loss": 1.3247, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.24819277108433735, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 7.518072289156627e-06, | |
| "loss": 1.24, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.25060240963855424, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 7.493975903614459e-06, | |
| "loss": 1.2623, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.25301204819277107, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.469879518072289e-06, | |
| "loss": 1.2901, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.25542168674698795, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 7.445783132530121e-06, | |
| "loss": 1.2738, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.25783132530120484, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 7.4216867469879526e-06, | |
| "loss": 1.3002, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.26024096385542167, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 7.3975903614457835e-06, | |
| "loss": 1.2721, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.26265060240963856, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 7.373493975903615e-06, | |
| "loss": 1.2388, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.26506024096385544, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 7.349397590361447e-06, | |
| "loss": 1.2336, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2674698795180723, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 7.325301204819277e-06, | |
| "loss": 1.2375, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.26987951807228916, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 7.301204819277109e-06, | |
| "loss": 1.3051, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.27228915662650605, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 7.277108433734941e-06, | |
| "loss": 1.263, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2746987951807229, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 7.2530120481927715e-06, | |
| "loss": 1.3006, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.27710843373493976, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 7.228915662650603e-06, | |
| "loss": 1.2649, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.27951807228915665, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 7.204819277108435e-06, | |
| "loss": 1.3185, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.2819277108433735, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 7.180722891566265e-06, | |
| "loss": 1.308, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.28433734939759037, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 7.156626506024097e-06, | |
| "loss": 1.2969, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.28674698795180725, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 7.132530120481929e-06, | |
| "loss": 1.2991, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2891566265060241, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 7.1084337349397595e-06, | |
| "loss": 1.3087, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.29156626506024097, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 7.084337349397591e-06, | |
| "loss": 1.2659, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.29397590361445786, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 7.060240963855422e-06, | |
| "loss": 1.3115, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2963855421686747, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 7.036144578313253e-06, | |
| "loss": 1.263, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2987951807228916, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 7.012048192771085e-06, | |
| "loss": 1.3126, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.30120481927710846, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.987951807228917e-06, | |
| "loss": 1.2533, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3036144578313253, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 6.963855421686747e-06, | |
| "loss": 1.2449, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3060240963855422, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 6.9397590361445784e-06, | |
| "loss": 1.3121, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.30843373493975906, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 6.91566265060241e-06, | |
| "loss": 1.2479, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3108433734939759, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 6.891566265060241e-06, | |
| "loss": 1.237, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3132530120481928, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 6.867469879518073e-06, | |
| "loss": 1.2766, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3156626506024096, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 6.843373493975905e-06, | |
| "loss": 1.2575, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.3180722891566265, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 6.819277108433735e-06, | |
| "loss": 1.3013, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3204819277108434, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 6.7951807228915665e-06, | |
| "loss": 1.2832, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3228915662650602, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 6.771084337349398e-06, | |
| "loss": 1.2646, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3253012048192771, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 6.746987951807229e-06, | |
| "loss": 1.3012, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.327710843373494, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 6.722891566265061e-06, | |
| "loss": 1.2523, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.3301204819277108, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 6.698795180722893e-06, | |
| "loss": 1.2599, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3325301204819277, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 6.674698795180723e-06, | |
| "loss": 1.2305, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3349397590361446, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 6.6506024096385545e-06, | |
| "loss": 1.2838, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3373493975903614, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 6.626506024096386e-06, | |
| "loss": 1.2453, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3397590361445783, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 6.602409638554217e-06, | |
| "loss": 1.2577, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3421686746987952, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 6.578313253012049e-06, | |
| "loss": 1.2172, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.344578313253012, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 6.554216867469881e-06, | |
| "loss": 1.2571, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3469879518072289, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 6.530120481927711e-06, | |
| "loss": 1.232, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3493975903614458, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 6.5060240963855425e-06, | |
| "loss": 1.2423, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.35180722891566263, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 6.481927710843374e-06, | |
| "loss": 1.2497, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3542168674698795, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 6.457831325301205e-06, | |
| "loss": 1.2085, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.3566265060240964, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 6.433734939759036e-06, | |
| "loss": 1.2676, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.35903614457831323, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 6.409638554216868e-06, | |
| "loss": 1.234, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3614457831325301, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 6.385542168674699e-06, | |
| "loss": 1.2128, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.363855421686747, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 6.3614457831325305e-06, | |
| "loss": 1.2732, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.36626506024096384, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 6.337349397590362e-06, | |
| "loss": 1.2756, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3686746987951807, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 6.313253012048192e-06, | |
| "loss": 1.2577, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.3710843373493976, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 6.289156626506024e-06, | |
| "loss": 1.2496, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.37349397590361444, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 6.265060240963856e-06, | |
| "loss": 1.2258, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3759036144578313, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 6.240963855421688e-06, | |
| "loss": 1.1938, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3783132530120482, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 6.2168674698795185e-06, | |
| "loss": 1.2642, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.38072289156626504, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 6.19277108433735e-06, | |
| "loss": 1.2423, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.38313253012048193, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 6.168674698795182e-06, | |
| "loss": 1.2773, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.3855421686746988, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 6.144578313253012e-06, | |
| "loss": 1.2747, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.38795180722891565, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 6.120481927710844e-06, | |
| "loss": 1.2425, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.39036144578313253, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 6.096385542168676e-06, | |
| "loss": 1.2592, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3927710843373494, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 6.0722891566265066e-06, | |
| "loss": 1.2554, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.39518072289156625, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 6.048192771084338e-06, | |
| "loss": 1.2929, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.39759036144578314, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 6.02409638554217e-06, | |
| "loss": 1.1996, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 6e-06, | |
| "loss": 1.2782, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.40240963855421685, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 5.975903614457832e-06, | |
| "loss": 1.1855, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.40481927710843374, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 5.951807228915664e-06, | |
| "loss": 1.1886, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.4072289156626506, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 5.927710843373495e-06, | |
| "loss": 1.1398, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.40963855421686746, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 5.9036144578313255e-06, | |
| "loss": 1.1729, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.41204819277108434, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 5.879518072289157e-06, | |
| "loss": 1.2716, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.41445783132530123, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 5.855421686746988e-06, | |
| "loss": 1.2669, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.41686746987951806, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 5.83132530120482e-06, | |
| "loss": 1.2207, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.41927710843373495, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 5.807228915662652e-06, | |
| "loss": 1.2459, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.42168674698795183, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 5.783132530120482e-06, | |
| "loss": 1.2252, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.42409638554216866, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 5.7590361445783135e-06, | |
| "loss": 1.2537, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.42650602409638555, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 5.734939759036145e-06, | |
| "loss": 1.2581, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.42891566265060244, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 5.710843373493976e-06, | |
| "loss": 1.2493, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.43132530120481927, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 5.686746987951808e-06, | |
| "loss": 1.188, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.43373493975903615, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 5.66265060240964e-06, | |
| "loss": 1.2578, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.43614457831325304, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 5.63855421686747e-06, | |
| "loss": 1.2119, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.43855421686746987, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 5.6144578313253015e-06, | |
| "loss": 1.2372, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.44096385542168676, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 5.590361445783133e-06, | |
| "loss": 1.2965, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4433734939759036, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 5.566265060240964e-06, | |
| "loss": 1.1978, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4457831325301205, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 5.542168674698796e-06, | |
| "loss": 1.2581, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.44819277108433736, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 5.518072289156628e-06, | |
| "loss": 1.2148, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4506024096385542, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 5.493975903614458e-06, | |
| "loss": 1.2362, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4530120481927711, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 5.4698795180722896e-06, | |
| "loss": 1.1945, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.45542168674698796, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 5.445783132530121e-06, | |
| "loss": 1.2047, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4578313253012048, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 5.421686746987952e-06, | |
| "loss": 1.1945, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4602409638554217, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 5.397590361445784e-06, | |
| "loss": 1.2674, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.46265060240963857, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 5.373493975903615e-06, | |
| "loss": 1.2381, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.4650602409638554, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 5.349397590361446e-06, | |
| "loss": 1.236, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.4674698795180723, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 5.325301204819278e-06, | |
| "loss": 1.2428, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.46987951807228917, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 5.301204819277109e-06, | |
| "loss": 1.1794, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.472289156626506, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 5.27710843373494e-06, | |
| "loss": 1.1801, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.4746987951807229, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 5.253012048192771e-06, | |
| "loss": 1.2689, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.4771084337349398, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 5.228915662650603e-06, | |
| "loss": 1.2742, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.4795180722891566, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 5.204819277108434e-06, | |
| "loss": 1.1817, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.4819277108433735, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 5.180722891566266e-06, | |
| "loss": 1.2414, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4843373493975904, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 5.156626506024097e-06, | |
| "loss": 1.1819, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.4867469879518072, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 5.132530120481927e-06, | |
| "loss": 1.2304, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.4891566265060241, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 5.108433734939759e-06, | |
| "loss": 1.2907, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.491566265060241, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 5.084337349397591e-06, | |
| "loss": 1.2876, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.4939759036144578, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 5.060240963855422e-06, | |
| "loss": 1.2352, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4963855421686747, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 5.036144578313254e-06, | |
| "loss": 1.2529, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.4987951807228916, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 5.012048192771085e-06, | |
| "loss": 1.2677, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.5012048192771085, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 4.987951807228916e-06, | |
| "loss": 1.227, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.5036144578313253, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 4.963855421686747e-06, | |
| "loss": 1.2165, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.5060240963855421, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 4.939759036144578e-06, | |
| "loss": 1.2693, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5084337349397591, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 4.91566265060241e-06, | |
| "loss": 1.2258, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.5108433734939759, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 4.891566265060242e-06, | |
| "loss": 1.2487, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.5132530120481927, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 4.8674698795180725e-06, | |
| "loss": 1.2432, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.5156626506024097, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 4.843373493975904e-06, | |
| "loss": 1.2707, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.5180722891566265, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 4.819277108433735e-06, | |
| "loss": 1.2248, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5204819277108433, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 4.795180722891566e-06, | |
| "loss": 1.2202, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.5228915662650603, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 4.771084337349398e-06, | |
| "loss": 1.2058, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.5253012048192771, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 4.74698795180723e-06, | |
| "loss": 1.2477, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5277108433734939, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 4.7228915662650606e-06, | |
| "loss": 1.1872, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.5301204819277109, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 4.698795180722892e-06, | |
| "loss": 1.1966, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5325301204819277, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 4.674698795180723e-06, | |
| "loss": 1.188, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5349397590361445, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 4.650602409638554e-06, | |
| "loss": 1.2359, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5373493975903615, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 4.626506024096386e-06, | |
| "loss": 1.2064, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5397590361445783, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 4.602409638554217e-06, | |
| "loss": 1.2431, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5421686746987951, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 4.578313253012049e-06, | |
| "loss": 1.2543, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5445783132530121, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 4.55421686746988e-06, | |
| "loss": 1.2109, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5469879518072289, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 4.530120481927711e-06, | |
| "loss": 1.1845, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5493975903614458, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 4.506024096385542e-06, | |
| "loss": 1.1951, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5518072289156627, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 4.481927710843374e-06, | |
| "loss": 1.1969, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5542168674698795, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 4.457831325301205e-06, | |
| "loss": 1.2563, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5566265060240964, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 4.433734939759037e-06, | |
| "loss": 1.2213, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5590361445783133, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 4.4096385542168675e-06, | |
| "loss": 1.2261, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5614457831325301, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 4.385542168674699e-06, | |
| "loss": 1.21, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.563855421686747, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 4.361445783132531e-06, | |
| "loss": 1.197, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5662650602409639, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 4.337349397590362e-06, | |
| "loss": 1.2003, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5686746987951807, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 4.313253012048193e-06, | |
| "loss": 1.2401, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5710843373493976, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 4.289156626506025e-06, | |
| "loss": 1.2506, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.5734939759036145, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 4.2650602409638555e-06, | |
| "loss": 1.2157, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5759036144578313, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 4.240963855421687e-06, | |
| "loss": 1.2192, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.5783132530120482, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 4.216867469879519e-06, | |
| "loss": 1.1676, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5807228915662651, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 4.19277108433735e-06, | |
| "loss": 1.2432, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.5831325301204819, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 4.168674698795181e-06, | |
| "loss": 1.2353, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.5855421686746988, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 4.144578313253013e-06, | |
| "loss": 1.187, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5879518072289157, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 4.1204819277108436e-06, | |
| "loss": 1.1805, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.5903614457831325, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 4.096385542168675e-06, | |
| "loss": 1.2429, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5927710843373494, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 4.072289156626506e-06, | |
| "loss": 1.0815, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.5951807228915663, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 4.048192771084338e-06, | |
| "loss": 1.2279, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.5975903614457831, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 4.024096385542169e-06, | |
| "loss": 1.1839, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.2672, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.6024096385542169, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 3.975903614457832e-06, | |
| "loss": 1.2033, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6048192771084338, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 3.9518072289156625e-06, | |
| "loss": 1.25, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.6072289156626506, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 3.927710843373494e-06, | |
| "loss": 1.2099, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.6096385542168675, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 3.903614457831326e-06, | |
| "loss": 1.157, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.6120481927710844, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 3.879518072289157e-06, | |
| "loss": 1.1643, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.6144578313253012, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 3.855421686746989e-06, | |
| "loss": 1.2308, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6168674698795181, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 3.83132530120482e-06, | |
| "loss": 1.1898, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.619277108433735, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 3.807228915662651e-06, | |
| "loss": 1.1833, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.6216867469879518, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 3.7831325301204823e-06, | |
| "loss": 1.1651, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.6240963855421687, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.7590361445783136e-06, | |
| "loss": 1.2409, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.6265060240963856, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 3.7349397590361445e-06, | |
| "loss": 1.2054, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6289156626506024, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 3.7108433734939763e-06, | |
| "loss": 1.1575, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.6313253012048192, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 3.6867469879518076e-06, | |
| "loss": 1.2138, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6337349397590362, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 3.6626506024096385e-06, | |
| "loss": 1.1803, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.636144578313253, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 3.6385542168674703e-06, | |
| "loss": 1.1605, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.6385542168674698, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 3.6144578313253016e-06, | |
| "loss": 1.1491, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6409638554216868, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 3.5903614457831325e-06, | |
| "loss": 1.1631, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6433734939759036, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 3.5662650602409643e-06, | |
| "loss": 1.2348, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6457831325301204, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 3.5421686746987956e-06, | |
| "loss": 1.308, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6481927710843374, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 3.5180722891566266e-06, | |
| "loss": 1.19, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6506024096385542, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 3.4939759036144583e-06, | |
| "loss": 1.2012, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.653012048192771, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 3.4698795180722892e-06, | |
| "loss": 1.2212, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.655421686746988, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 3.4457831325301206e-06, | |
| "loss": 1.2351, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.6578313253012048, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 3.4216867469879523e-06, | |
| "loss": 1.2123, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.6602409638554216, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 3.3975903614457832e-06, | |
| "loss": 1.2499, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.6626506024096386, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 3.3734939759036146e-06, | |
| "loss": 1.1171, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6650602409638554, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 3.3493975903614463e-06, | |
| "loss": 1.2002, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.6674698795180722, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.3253012048192772e-06, | |
| "loss": 1.2602, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.6698795180722892, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 3.3012048192771086e-06, | |
| "loss": 1.214, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.672289156626506, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 3.2771084337349403e-06, | |
| "loss": 1.1642, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6746987951807228, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 3.2530120481927713e-06, | |
| "loss": 1.2339, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6771084337349398, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 3.2289156626506026e-06, | |
| "loss": 1.2056, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6795180722891566, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 3.204819277108434e-06, | |
| "loss": 1.223, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.6819277108433734, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 3.1807228915662653e-06, | |
| "loss": 1.1978, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6843373493975904, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 3.156626506024096e-06, | |
| "loss": 1.1659, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.6867469879518072, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 3.132530120481928e-06, | |
| "loss": 1.1631, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.689156626506024, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.1084337349397593e-06, | |
| "loss": 1.2063, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.691566265060241, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 3.084337349397591e-06, | |
| "loss": 1.1784, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.6939759036144578, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 3.060240963855422e-06, | |
| "loss": 1.2651, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.6963855421686747, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 3.0361445783132533e-06, | |
| "loss": 1.1946, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.6987951807228916, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 3.012048192771085e-06, | |
| "loss": 1.19, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7012048192771084, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 2.987951807228916e-06, | |
| "loss": 1.1271, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.7036144578313253, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 2.9638554216867473e-06, | |
| "loss": 1.186, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.7060240963855422, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 2.9397590361445786e-06, | |
| "loss": 1.2006, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.708433734939759, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 2.91566265060241e-06, | |
| "loss": 1.1859, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.7108433734939759, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 2.891566265060241e-06, | |
| "loss": 1.2297, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7132530120481928, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 2.8674698795180726e-06, | |
| "loss": 1.1706, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.7156626506024096, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 2.843373493975904e-06, | |
| "loss": 1.2615, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.7180722891566265, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 2.819277108433735e-06, | |
| "loss": 1.2033, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.7204819277108434, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 2.7951807228915666e-06, | |
| "loss": 1.1958, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.7228915662650602, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 2.771084337349398e-06, | |
| "loss": 1.1919, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7253012048192771, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 2.746987951807229e-06, | |
| "loss": 1.2025, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.727710843373494, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 2.7228915662650607e-06, | |
| "loss": 1.1796, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.7301204819277108, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 2.698795180722892e-06, | |
| "loss": 1.1958, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.7325301204819277, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 2.674698795180723e-06, | |
| "loss": 1.123, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.7349397590361446, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 2.6506024096385547e-06, | |
| "loss": 1.2265, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7373493975903614, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 2.6265060240963856e-06, | |
| "loss": 1.2097, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.7397590361445783, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 2.602409638554217e-06, | |
| "loss": 1.1958, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.7421686746987952, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 2.5783132530120487e-06, | |
| "loss": 1.1628, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.744578313253012, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 2.5542168674698796e-06, | |
| "loss": 1.1568, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7469879518072289, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 2.530120481927711e-06, | |
| "loss": 1.2034, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7493975903614458, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 2.5060240963855427e-06, | |
| "loss": 1.1065, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.7518072289156627, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 2.4819277108433736e-06, | |
| "loss": 1.1767, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.7542168674698795, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 2.457831325301205e-06, | |
| "loss": 1.1741, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7566265060240964, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 2.4337349397590363e-06, | |
| "loss": 1.177, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.7590361445783133, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 2.4096385542168676e-06, | |
| "loss": 1.1835, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7614457831325301, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 2.385542168674699e-06, | |
| "loss": 1.1393, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.763855421686747, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 2.3614457831325303e-06, | |
| "loss": 1.1554, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.7662650602409639, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 2.3373493975903616e-06, | |
| "loss": 1.2462, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.7686746987951807, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 2.313253012048193e-06, | |
| "loss": 1.1781, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.7710843373493976, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 2.2891566265060243e-06, | |
| "loss": 1.1799, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7734939759036145, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 2.2650602409638556e-06, | |
| "loss": 1.1082, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.7759036144578313, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 2.240963855421687e-06, | |
| "loss": 1.259, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7783132530120482, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 2.2168674698795183e-06, | |
| "loss": 1.158, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.7807228915662651, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 2.1927710843373496e-06, | |
| "loss": 1.2057, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7831325301204819, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 2.168674698795181e-06, | |
| "loss": 1.1866, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7855421686746988, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 2.1445783132530123e-06, | |
| "loss": 1.222, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7879518072289157, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 2.1204819277108437e-06, | |
| "loss": 1.2492, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.7903614457831325, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 2.096385542168675e-06, | |
| "loss": 1.2351, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.7927710843373494, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 2.0722891566265063e-06, | |
| "loss": 1.1282, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.7951807228915663, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 2.0481927710843377e-06, | |
| "loss": 1.1673, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7975903614457831, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 2.024096385542169e-06, | |
| "loss": 1.238, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.1519, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.8024096385542169, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 1.9759036144578312e-06, | |
| "loss": 1.1692, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.8048192771084337, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 1.951807228915663e-06, | |
| "loss": 1.1471, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.8072289156626506, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 1.9277108433734943e-06, | |
| "loss": 1.1541, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8096385542168675, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 1.9036144578313255e-06, | |
| "loss": 1.2038, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.8120481927710843, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 1.8795180722891568e-06, | |
| "loss": 1.1794, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.8144578313253013, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 1.8554216867469881e-06, | |
| "loss": 1.2553, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.8168674698795181, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 1.8313253012048193e-06, | |
| "loss": 1.185, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.8192771084337349, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 1.8072289156626508e-06, | |
| "loss": 1.2193, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8216867469879519, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 1.7831325301204822e-06, | |
| "loss": 1.1436, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.8240963855421687, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 1.7590361445783133e-06, | |
| "loss": 1.1608, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.8265060240963855, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 1.7349397590361446e-06, | |
| "loss": 1.2343, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.8289156626506025, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 1.7108433734939762e-06, | |
| "loss": 1.1779, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.8313253012048193, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 1.6867469879518073e-06, | |
| "loss": 1.2414, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.8337349397590361, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 1.6626506024096386e-06, | |
| "loss": 1.2011, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.8361445783132531, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 1.6385542168674702e-06, | |
| "loss": 1.1586, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.8385542168674699, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 1.6144578313253013e-06, | |
| "loss": 1.1782, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.8409638554216867, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 1.5903614457831326e-06, | |
| "loss": 1.1461, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.8433734939759037, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 1.566265060240964e-06, | |
| "loss": 1.2264, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8457831325301205, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 1.5421686746987955e-06, | |
| "loss": 1.1675, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.8481927710843373, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 1.5180722891566266e-06, | |
| "loss": 1.1975, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.8506024096385543, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 1.493975903614458e-06, | |
| "loss": 1.1335, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.8530120481927711, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 1.4698795180722893e-06, | |
| "loss": 1.2112, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.8554216867469879, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 1.4457831325301204e-06, | |
| "loss": 1.2036, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8578313253012049, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 1.421686746987952e-06, | |
| "loss": 1.2405, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.8602409638554217, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.3975903614457833e-06, | |
| "loss": 1.1708, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.8626506024096385, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 1.3734939759036144e-06, | |
| "loss": 1.1913, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.8650602409638555, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 1.349397590361446e-06, | |
| "loss": 1.207, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.8674698795180723, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 1.3253012048192773e-06, | |
| "loss": 1.1908, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8698795180722891, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 1.3012048192771085e-06, | |
| "loss": 1.1619, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.8722891566265061, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 1.2771084337349398e-06, | |
| "loss": 1.2102, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.8746987951807229, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 1.2530120481927713e-06, | |
| "loss": 1.1496, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.8771084337349397, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 1.2289156626506025e-06, | |
| "loss": 1.226, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.8795180722891566, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 1.2048192771084338e-06, | |
| "loss": 1.2046, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8819277108433735, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 1.1807228915662651e-06, | |
| "loss": 1.2166, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.8843373493975903, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 1.1566265060240965e-06, | |
| "loss": 1.1674, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.8867469879518072, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 1.1325301204819278e-06, | |
| "loss": 1.2386, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.8891566265060241, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 1.1084337349397592e-06, | |
| "loss": 1.1736, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.891566265060241, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 1.0843373493975905e-06, | |
| "loss": 1.188, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8939759036144578, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 1.0602409638554218e-06, | |
| "loss": 1.1511, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.8963855421686747, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 1.0361445783132532e-06, | |
| "loss": 1.156, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.8987951807228916, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 1.0120481927710845e-06, | |
| "loss": 1.2058, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.9012048192771084, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 9.879518072289156e-07, | |
| "loss": 1.1479, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.9036144578313253, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 9.638554216867472e-07, | |
| "loss": 1.1658, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.9060240963855422, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 9.397590361445784e-07, | |
| "loss": 1.246, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.908433734939759, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 9.156626506024096e-07, | |
| "loss": 1.2199, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.9108433734939759, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 8.915662650602411e-07, | |
| "loss": 1.157, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.9132530120481928, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 8.674698795180723e-07, | |
| "loss": 1.1644, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.9156626506024096, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 8.433734939759036e-07, | |
| "loss": 1.2226, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9180722891566265, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 8.192771084337351e-07, | |
| "loss": 1.1723, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.9204819277108434, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 7.951807228915663e-07, | |
| "loss": 1.205, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.9228915662650602, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 7.710843373493978e-07, | |
| "loss": 1.1976, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.9253012048192771, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 7.46987951807229e-07, | |
| "loss": 1.1185, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.927710843373494, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 7.228915662650602e-07, | |
| "loss": 1.128, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.9301204819277108, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 6.987951807228917e-07, | |
| "loss": 1.1403, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.9325301204819277, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 6.74698795180723e-07, | |
| "loss": 1.2571, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.9349397590361446, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 6.506024096385542e-07, | |
| "loss": 1.222, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.9373493975903614, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 6.265060240963857e-07, | |
| "loss": 1.2048, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.9397590361445783, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.024096385542169e-07, | |
| "loss": 1.2411, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9421686746987952, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 5.783132530120482e-07, | |
| "loss": 1.1832, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.944578313253012, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 5.542168674698796e-07, | |
| "loss": 1.2639, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.946987951807229, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 5.301204819277109e-07, | |
| "loss": 1.2135, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.9493975903614458, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 5.060240963855422e-07, | |
| "loss": 1.1594, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.9518072289156626, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 4.819277108433736e-07, | |
| "loss": 1.1236, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9542168674698795, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 4.578313253012048e-07, | |
| "loss": 1.1344, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.9566265060240964, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 4.3373493975903615e-07, | |
| "loss": 1.1536, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.9590361445783132, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 4.0963855421686754e-07, | |
| "loss": 1.1962, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.9614457831325302, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 3.855421686746989e-07, | |
| "loss": 1.1711, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.963855421686747, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 3.614457831325301e-07, | |
| "loss": 1.2135, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9662650602409638, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 3.373493975903615e-07, | |
| "loss": 1.1843, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.9686746987951808, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 3.1325301204819284e-07, | |
| "loss": 1.1817, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.9710843373493976, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 2.891566265060241e-07, | |
| "loss": 1.2055, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.9734939759036144, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 2.6506024096385546e-07, | |
| "loss": 1.1925, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.9759036144578314, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 2.409638554216868e-07, | |
| "loss": 1.1721, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9783132530120482, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 2.1686746987951808e-07, | |
| "loss": 1.2074, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.980722891566265, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 1.9277108433734944e-07, | |
| "loss": 1.1375, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.983132530120482, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 1.6867469879518075e-07, | |
| "loss": 1.1968, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.9855421686746988, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 1.4457831325301206e-07, | |
| "loss": 1.2086, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.9879518072289156, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 1.204819277108434e-07, | |
| "loss": 1.1967, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9903614457831326, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 9.638554216867472e-08, | |
| "loss": 1.2235, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.9927710843373494, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 7.228915662650603e-08, | |
| "loss": 1.1924, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.9951807228915662, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 4.819277108433736e-08, | |
| "loss": 1.1543, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.9975903614457832, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 2.409638554216868e-08, | |
| "loss": 1.2282, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.0, | |
| "loss": 1.2109, | |
| "step": 415 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 415, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3144272689357128e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |