| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.992, | |
| "eval_steps": 500, | |
| "global_step": 390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 6.924089727970164, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.8489, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.964284693740661, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.8627, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.770522283631703, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.8428, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 6.10578792519572, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.8148, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.504800132843968, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.7728, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.4552438913834176, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.7209, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.3593814254900285, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.7424, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.20678075883569, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.7193, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.643689249493023, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.718, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.495832780796782, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.6851, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.258684851705308, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.6954, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.618851447385731, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.6538, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.9866066883523399, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.6111, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.416221691802126, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.6067, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.7717136924151333, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.5835, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.3862652744191104, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.5935, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.4366478962071656, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.5886, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.2333621770881595, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.5662, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.2039425695249761, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.5617, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.396256400385671, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.5573, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.9023151726405692, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.5406, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.228768712505066, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.547, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.7874065727456556, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.5303, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.9856507920048746, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.5139, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8755724131409769, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.5402, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.8443423195852365, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.5139, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.6992207278663957, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.5065, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.7792416851637417, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.4981, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.6685659514594317, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.4984, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.7536032327642854, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.5155, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.5901247999278916, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.4833, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.6224211315516938, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.4996, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 1.0828320160962968, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.5382, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.7146946512091298, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.5036, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.5816508414335341, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.484, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.6417828097494429, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.5009, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.5683625899810409, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.4973, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.8830206865703908, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.4803, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.5964279929768225, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4903, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.824158607932336, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.476, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.8027729194385715, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.4864, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.1528198063003026, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.4785, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.8095830237066084, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.4798, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.5758644988197384, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.471, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.9909066521510522, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.4875, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 1.0319843874475036, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.4765, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.7521512004563524, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.4723, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.83562245244271, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.4797, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.7674443375001021, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.4735, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.8451963033326747, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.4731, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.7388142879702948, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.4618, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.7649332038229532, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.4564, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.8073033768973511, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.4757, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.5747904683665801, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.4555, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.6917535223430752, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.4774, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.6129951236893196, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.4503, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.5665183402916288, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.4667, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.6843907271010262, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.4569, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.6519756020861179, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.462, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5479517333222286, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.4519, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.5823991047258783, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.4625, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.6602332798789388, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.4634, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.5475906087679264, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.4489, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.5823416385033334, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.4514, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.5624125306225127, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.4627, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.8560787875056831, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.4563, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.6092396657271495, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.4627, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.737041007933656, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.4461, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.44988642866546985, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.4555, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.6021468365025637, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.4595, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.5389224248851311, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.4442, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5303558952301805, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.464, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.5550013812196073, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.462, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.6107978122521343, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.4625, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5287154649205293, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.4592, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.4535441767826748, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.4344, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.5927300672166484, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.4611, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.4949260632389431, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.4454, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.5443213415912672, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.4064, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.4547939146251776, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.3862, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.5390905813461109, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.3908, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.6198513669630196, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.402, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.7882879838756096, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.4103, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.8548359106382617, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.3885, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.7683190763456025, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.399, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.5987239062680368, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.3988, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.7102852001875787, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.4008, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.7589598556123809, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.3916, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.8274483181796952, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.3999, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.7697953781715063, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.3898, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.7938424190005213, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.3883, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.6852446128502925, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.3712, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.8985568921177798, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.4066, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.6535491602848797, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.3869, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.830224233305105, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.3966, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.750358040822685, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.4065, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.7221137567351962, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.3949, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.6902805376735299, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.3917, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.5152077309747354, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.3826, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5768976295641617, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.4053, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.5716819436837993, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.3931, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.4437754879936848, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.3796, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.44049971038126756, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.3898, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 1.1465689567295263, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.4201, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.3970176822743068, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.3921, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5162627953967718, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.389, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5108319103841839, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.3993, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.4929358372139535, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.3837, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.4988786700005865, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.3877, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.4665051243746867, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.3881, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.4648452243040121, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.3777, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.5106417805404915, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.411, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.5315415013290186, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.377, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.3899327030449156, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.3725, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.499753460079324, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.3979, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.43545753262941944, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.3943, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.3989850580726036, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4089, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.44498899242768225, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.394, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.39641729805823556, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.395, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.38915375676226016, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.3957, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.45671013124301657, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.3781, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.4946823048687929, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.3792, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.40279976391046635, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.3778, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.6230433772491878, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.3909, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.4281040257379043, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.4004, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5483240178095553, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.3718, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.45989705505442324, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.3815, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.4690183254039643, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.3919, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.565786543960403, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.3978, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.4044439717899073, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.3795, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.5370629041380162, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.3908, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.477685053807378, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.3883, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.5079551589657898, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.3772, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.4385022837215939, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.4045, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.49902467513848564, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.4034, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.5361666769134471, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.4083, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.47168995551840603, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.3988, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5565410251734092, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.4019, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.4933835881404012, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.3931, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.3935188632443315, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.3795, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.46639822783093865, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.3909, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.35610431486522537, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.3955, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.4706725691116115, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.3873, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.3537001850780939, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.3764, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.47281970960660386, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.375, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.3447011289754408, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.3949, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.4459445615865695, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.3744, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.37309498374878663, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.3836, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.5025908442288188, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.401, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.37285830096919786, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.3778, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.3822097047241446, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.385, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.3752192007885031, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.3959, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4029802335262307, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.3894, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.3846080614609637, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.3952, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.40001914676736294, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.3923, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.4437553457418262, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4012, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.4836132694040789, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.344, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.5160238023362884, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.325, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.5717107628630719, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3276, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.5376110199029949, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3073, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.529905024931671, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3326, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5169923230947776, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3372, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.46295879213320845, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3018, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.44898047571276284, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.2998, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.45699318944284595, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3179, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.38351884873278164, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3081, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.38669804263496715, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.298, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.47479778608920203, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3166, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.3920623307310556, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3183, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.4339772638817777, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.331, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.42448910123624906, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3263, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.4318496162877406, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3182, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.4217973473489407, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3344, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 1.3517161534523745, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3663, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.5039646371874676, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3153, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.39719036095713, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3433, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.44686565346972534, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3019, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 1.1886369815608528, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3241, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.3943997471445818, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3191, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.3999372294891782, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3189, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.40282714820153354, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3196, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.4117958721838313, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3217, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.368496788368057, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3064, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.40433849128917765, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3018, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.3983284598785467, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3112, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.3795370062006202, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3122, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.45485720394867896, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3117, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.38815289137853903, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3038, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.35256926857029, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3077, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.3787558326362315, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3086, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.33597879547330406, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3116, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.3219458489277532, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3035, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.34733374132753825, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3186, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.3178485966531094, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3245, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.35768271711486166, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3049, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.38556839027844914, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.2946, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.3449573917927118, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3245, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.39994897445537847, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3155, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.33435599441301683, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3144, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.40798013974512565, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3224, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.31808427243369025, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3261, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.35770102239123175, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3125, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.34360936954070725, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3042, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.3360102937138561, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3161, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.332088568897091, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3098, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.3088111149162953, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3173, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3037031346787973, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3374, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.33115606043604173, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3266, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.2872465819385532, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3106, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.3145047228943851, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3193, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.2900318473787524, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.308, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.32550180291364367, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3177, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.3123764134316215, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.318, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.2974651922203232, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.2971, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.30375052066204833, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3102, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3296467003653807, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3303, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.32638243443393045, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3216, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.2947016351280737, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3122, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.2964983026821736, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3074, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3427130971539848, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3259, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.2958149965730151, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3016, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3184148157091547, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3241, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 5.115967333409872, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.4779, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.43084777913299394, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3129, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.32417990021362625, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3119, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.3409371750621961, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3137, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.32867888840364384, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3161, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3795674630354955, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3077, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.29443436117454813, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.301, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3969191844391673, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3195, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.34589851364128626, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3298, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.3234256086497945, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.323, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.3411388776792002, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3234, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.29812843637748526, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3103, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.35325762430106616, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.2867, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.9331037316499748, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2801, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.5066470941747184, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2492, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.32803580955653183, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.241, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.4239035973253473, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2791, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 1.7461139039064202, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.3235, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.37846004760606305, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2618, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.37476786819485736, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2758, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.3539870344186331, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.25, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.3349062855670017, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2633, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.3581744465091351, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2618, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.3659092871832494, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2387, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3093393137538967, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2725, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.37794811409107837, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2624, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.29755595161484266, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2597, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.2743829093703299, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2524, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.30007222208041745, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.239, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.2788060212847718, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.244, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.25301328008717006, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2507, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.27526586033724454, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2452, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.31141034554861563, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2648, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.2687991764254246, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.256, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.2809406949223676, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2373, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.26424339081384546, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2624, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.2823345076881774, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2597, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.2571613654645568, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2748, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.2608021189082966, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2517, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.2754077133878195, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2547, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.25422924995662277, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2399, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.2727979120819607, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2593, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.2857007545340251, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2591, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.2569843968790537, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2468, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.2845692911062232, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2638, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.27557399799961535, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.256, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.2682588740629733, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2658, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 1.2418369798435898, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.3092, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.28227835450734984, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2722, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.24993707489162226, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2483, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.27232923923892516, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.264, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.2733671919417325, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2577, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.265102251794001, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2481, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.2634614041723308, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2536, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.2676848021432201, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2545, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.2459112615799778, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2659, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.2545993843083449, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2301, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.26244155509779815, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2533, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.26384759131622887, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.251, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.24689257592302388, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2766, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.26038640454429046, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.256, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.2828560325415062, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2294, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.2405896798661964, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2614, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.2566954380200886, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2386, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.2676403461911548, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2464, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.2766917947490327, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2646, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.23904522845150963, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2401, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.2702937165802826, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2599, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.2506081234252952, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2346, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.24657152733456048, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2465, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.2582494462321876, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2651, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.2521135500940868, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2502, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.24610286108812737, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2395, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.24264785923804955, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.252, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.2480947120260181, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2601, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.2326963417486253, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2687, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.2530177564968595, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2451, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.23734679182741789, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2558, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.24769522929035867, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2538, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.22469706320127494, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2595, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.23667506509131866, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2476, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.23396437321507882, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2472, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.23896719131956434, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2689, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.2427886183720272, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2573, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.24769569984407128, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2422, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.237223231536741, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2594, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.22577301041468506, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2571, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.22744771883906087, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2566, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.2287941257253064, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2451, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.2514832830345657, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2589, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.2618506411344112, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.2478, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.28205963612542284, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.2321, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.24997596557434323, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.2214, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.24764485683722248, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.2151, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.35210213745811003, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.2247, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.374683399802985, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.2277, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.24013578133500085, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.2167, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.31334642007628155, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.2166, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.3327429451087759, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.2107, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.3008925177057534, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.2084, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.23918084817006158, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.2119, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 2.3140484720270376, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.3212, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.3330239312193593, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.2363, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.33556368877300624, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.2096, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.22921402517239972, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.188, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.22307151632823738, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.2302, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.2120605226154532, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.2175, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.24027719022835933, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.2153, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.23170937885232878, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.2093, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.23107385400522465, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.2133, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.21289720409488924, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.2172, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.21284599999600531, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.2198, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.21390804344735034, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.2197, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.20363866115999157, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1961, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.20990499023484815, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.2098, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.21275548448391685, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.2042, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.21033271449474353, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.2106, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.20386279189997755, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.2282, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.20954653147093588, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.218, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.1980469255665047, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.2287, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.19435616038401504, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.2251, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.20625350326402428, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.2162, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.1959479171810325, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.2224, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.19778809581424875, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.2184, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.18951619395610253, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1964, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.1977415602493209, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.2135, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.23778085162054988, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.2265, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.19316340525816594, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.235, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.1951953097170093, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.2274, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.1976999971326611, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.2011, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.20086468998550366, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.2303, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.19186628239872114, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.2376, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.20635371471475802, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.193, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.19838294863532474, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.2199, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.19411614251979062, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.2104, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.848009863286474, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.2327, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.19851507732541096, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.2152, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.19900651722678742, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.2333, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.1985819699173841, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.2204, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.19185142874802685, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.239, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.2017668261258135, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.2311, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.5348084991536339, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.256, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.19769796573260423, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.2274, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.1923082081335489, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.2195, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.1964543175915375, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.2064, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.18450486819743458, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.2427, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.19716276752541123, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.2267, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.1951640323031509, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.2233, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.1966365085381747, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.2134, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.18507754320426112, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.2162, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.1867091787517674, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.2057, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.19700511607753401, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.2233, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.20049628905674952, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.237, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.20391286522042154, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.2077, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.18473031072152943, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.2136, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.18146951681840795, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.2387, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.189825330215684, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.225, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.18758370499306726, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.2301, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.18974686634472085, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.2351, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.1826149267402475, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.2145, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.19195489717583294, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.2136, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.1970012233137704, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.219, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.18378711246844784, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.218, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.1945911774110049, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.2368, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.18601143930389175, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.2154, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.1901355668052435, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.2344, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.1994643353308444, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.2147, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.19378498871024788, | |
| "learning_rate": 0.0, | |
| "loss": 0.2196, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 1.4532057772410798e+18, | |
| "train_loss": 0.34427934441811, | |
| "train_runtime": 52740.7681, | |
| "train_samples_per_second": 0.948, | |
| "train_steps_per_second": 0.007 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 390, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4532057772410798e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |