| { | |
| "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": 6.505747529855917, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.037, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.560852269618801, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0615, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.297724789047534, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0368, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.873504662538556, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.0204, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.260219296200858, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.9878, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.8226498262289894, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9503, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.4814856283348576, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.9426, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.4472339963511995, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.9631, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.2452868354337046, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.9487, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.073394526067116, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.926, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.5593666985232364, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8998, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.503183770041401, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8677, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.6581259245716637, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8532, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.6247963160720575, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.8214, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.3657729627592703, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8154, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.345374794880054, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.8029, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.1040494988871448, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.784, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.1197287715806619, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7778, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.1386731281098734, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7733, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.2028279856088686, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7695, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.9650970506021546, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7588, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.2234719505580478, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.742, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 1.0259658673857637, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7614, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.10643945152428, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7286, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.1095769010970096, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.737, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.9761591577884013, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7132, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.9290069048161637, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7003, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.8281825117654518, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.7189, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.9108454848461486, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.7383, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.874417940004658, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.707, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 1.1312159124204537, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.704, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 1.0176318141391008, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.7043, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.839050900261088, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6938, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.9910989495874221, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.7013, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.053641939312259, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.7013, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.0830842904943496, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6812, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.2965657706346354, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6924, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.7067387238358549, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6998, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.8299809452023478, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6843, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.0303843817695786, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.683, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 1.336041158895854, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.7029, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.2263901719907044, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.7072, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 1.0529146573986454, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6608, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.586204252591105, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6938, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.0880221516969555, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.67, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 1.7002247914736648, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6744, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 1.4497178148540242, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6817, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.2280463384232854, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6685, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.4957243167941556, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6629, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.3135523978551822, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6803, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 1.6125339909399647, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6683, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 1.5305881479764225, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.682, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 1.1001238047348845, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6949, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 1.5798483939223087, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6673, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.0082766354607982, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6703, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 1.3422788359537228, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.693, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.059295829997347, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.683, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.8788638110720364, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.656, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 1.0686729053217592, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6692, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.1201463636193725, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6862, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.8559069553016561, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.67, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.9594052865179052, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6715, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 1.1211901210353805, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6512, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.7307537653308905, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6396, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.1601355647681502, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6751, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.8086148186887785, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6772, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.6390400829517844, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6463, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.8517679495121439, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6379, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.6404387302507338, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6485, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 1.094204843701778, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6492, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.9895790429658228, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6575, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5129909217790354, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6607, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.6517715838303199, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6561, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.6016007228527741, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6328, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6075697725308501, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6736, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.5801981604462901, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6588, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.5203710553237849, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.632, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.6212761462486293, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6394, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.6278311348239496, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.6234, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.5673895435016599, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5986, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.9310910409868183, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.6006, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 1.5466451511298254, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.607, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.7332847552442167, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.6087, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 1.01742608188952, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.6032, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 1.06572274158088, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.6019, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.7436152339326781, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5743, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.758140015099645, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5779, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.7580395238395826, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.6161, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.6451558758941932, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5793, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.8956814143718566, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5856, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.8324646477305917, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5738, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.8324848310252281, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5907, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.7151899660755547, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5741, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.6955181797117225, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5837, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.6526987338049262, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.587, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.8216393160611759, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5854, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.6978494971788809, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5832, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.6245459269391584, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5752, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.705216437162525, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5873, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.7065243017186413, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5821, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.7005390865524744, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5779, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.7017453290655963, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5736, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.665455906086348, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5627, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.6040822448749334, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5879, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.594803911451487, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5819, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.485398515306605, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5795, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5218022624908898, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5602, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.5054385639391414, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5861, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.5199852072862294, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5827, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.4637707990598707, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5847, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5291293842644805, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5643, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.5043363436493806, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5764, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.47242650603067554, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5755, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5518150964455962, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.568, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5224071175894263, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5897, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.5344847267477552, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5757, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.5219457944856948, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5792, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5259599717929431, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5787, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.5765311889493613, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5843, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5327677035476945, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.574, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.5930645571466301, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.5751, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.6342759940845253, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5813, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.48893214339925417, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5831, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5647678112481018, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5629, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.449133967218433, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5683, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5063749980986962, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5823, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.5384760726674587, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.5768, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.49653353728746846, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5755, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.533339384381669, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5857, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5120960901069407, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5737, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.543183043341828, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5869, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.5169503919418712, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5902, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.5039442241500608, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.5958, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.45321280216649795, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.588, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5556043673374238, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5655, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.5973369067030252, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5672, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.49639523554292275, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5619, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5340197021946086, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5494, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.6491880061183375, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5813, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.546412539608057, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5727, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.4885545484226667, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5766, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.46543308886146867, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5784, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.4769303758780081, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5633, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.4622281712035914, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5903, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.4080778019883451, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.5732, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.4902378466504111, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5958, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.5328465121163308, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5791, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.41319230509186344, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5796, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.5124687104015371, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5869, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.3990570138333518, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5798, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.5701054658943702, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.57, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.49262188689391084, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.588, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4493359276428374, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.6051, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.4376021711271567, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5903, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.45186829891460095, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5681, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.39621802994504, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5686, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.4455687685905931, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.5128, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.4950940160018366, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.4992, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.5962708491988533, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.5034, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.5705221556476063, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.4955, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.5266122887889625, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.4864, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.4828539198647265, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.4974, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.664332411609878, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.4929, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.49684891767453887, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.4886, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6173844319794962, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.5059, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.6185899744976351, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.4931, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.6306079550908135, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.4846, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5029994064199304, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.4995, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.48247444707199016, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.4937, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.44870669524781803, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.4924, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.44339227471416787, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.4973, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.40066357035072003, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.4877, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.39776641795055945, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.4853, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.49162584289757133, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.4988, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.39954986716181123, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.4783, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4526262538802144, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.4917, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.3576299330547625, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.4812, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.42376036267752043, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.4942, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.3358173028900977, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.4973, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.38778341099392927, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.4944, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.46211039098549717, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.4896, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.36513811455550704, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.4968, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.4777844900010711, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.4989, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.33354829623315135, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.4983, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.42129371314935, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.485, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.34749568893092425, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.4963, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.34239103504819346, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.4903, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.38153414780695555, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.4935, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.42448009768341244, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.4953, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.35999734178814113, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.494, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.4177534542860526, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.4971, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.36614733367029884, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.49, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.3629898887228354, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.4852, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.42164087991155624, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.501, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.3656740313988037, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.5068, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.3412529225542247, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.4909, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.36394592915518625, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.5004, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.3977520392517829, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.5027, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.39864770864687243, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.491, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.45988025280273587, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.4932, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.3546135394052879, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.5188, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.4618601705659484, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.4905, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.3585916516119542, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.4887, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.38124066737860346, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.4743, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.3621227727378548, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.4848, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.35992970989896667, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.4851, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.42114634962477715, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.4946, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.34299656901356307, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.4854, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.3613742984045961, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.4919, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.36699563374978, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.4818, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.38569411272389953, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.4832, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.351362373737474, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.4851, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.32965590050667626, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.4968, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.3588053691920234, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.5014, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.32474815666912576, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.486, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3547954675543647, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.477, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.33714252571554143, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.4823, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.3334975346532029, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.5008, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.34222937404032383, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.5082, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.38345392178288507, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.483, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.34990266060655323, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.5169, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3807577512849688, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.4929, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.373071861585422, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.4848, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.37889759627756314, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.4909, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.40100078003762757, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.4946, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.30764547855271984, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.4689, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.3698265310731593, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.4988, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3267836189939952, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.4934, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.3371861051483228, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.5008, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.36401939606068423, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.4884, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.31703438302188913, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.4929, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.3845426890198222, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.4798, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.3091458428078054, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.472, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3772945785914719, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.4759, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.4571467635241678, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.4717, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.4028002846921973, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.4433, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.6820582355550069, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.4285, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.4056991997120835, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.4167, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.5211806687138262, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.424, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.36529791684800383, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.4265, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.42697884443951184, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.4242, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.38598208321475, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.4156, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.40166934667311494, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.4285, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.3453676456399864, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.4332, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.43244023029113954, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.4, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.31845478103294494, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.4118, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.36264818372252705, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.4361, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.37162158467221174, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.4207, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.3621435380200018, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.417, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.39785127385776103, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.4047, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.3320536307525037, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.4174, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.3487901808850689, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.4065, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.3253802566432799, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.4119, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.3542620704801422, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.4154, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.32690488879784907, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.4086, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.3371738833265028, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.4421, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.2973807352876402, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.4132, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3268544820181663, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.4045, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.3122600328277054, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.4023, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3119871226916449, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.4251, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.3046890124826961, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.4099, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.29856493534107525, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.4136, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.34381708669262084, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.4009, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.2859674479950849, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.4257, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.38519333602943867, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.4088, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.27666917731180446, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.4253, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.31045511439888457, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.4137, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.3064610638812842, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.4063, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3141935716400805, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.4287, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.29111101358636754, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.4192, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.2956183665834541, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.3997, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.29005540179618783, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.4152, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.28290805212326664, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.4304, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.30310967758971785, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.4106, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.30287471467342203, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.4175, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.2719696548648095, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.4037, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.2899661973448743, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.4181, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.2588170019747225, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.4093, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.2888704451594072, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.4264, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.2837738692602651, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.4143, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.2513312536013756, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.4283, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.2829288918717478, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.4098, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.2943511931701685, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.421, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.2504237668330089, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.4345, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.2722020719297627, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.4112, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.2613503479910776, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.4255, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.24873238194791902, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.4302, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.27468461483172774, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.4116, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.26279435284454206, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.4123, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.25842123842807174, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.4137, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.2542193875851777, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.4205, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.28371884420881716, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.4064, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.2736582886132325, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.4258, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.25498388539426536, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.4234, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.24724201903664575, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.4162, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.2502405252119946, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.4279, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.2690069016659926, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.3935, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.25959229005991064, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.4107, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.2403156005885191, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.4438, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.27570234448588693, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.4184, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.265119969859078, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.4104, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.27513305634954666, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.408, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.28848517807921864, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.4199, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.2591173579414495, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.4175, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.2770146299256752, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.4311, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.27644111801793986, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.421, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.24589692673140315, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.4067, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.2605714132372804, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.403, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.2569195095458623, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.4306, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.25435460637415475, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.431, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.25434555401757375, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.42, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.25847082059352217, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.4262, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.3078836967914429, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.4081, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.3897456159789277, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.3757, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.27818616231947, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.3797, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3086382203339711, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.3774, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.4222041747862514, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.3639, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.33504576758682486, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.3674, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.2908160967324533, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.3581, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.3170950976592653, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.3835, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.3184290229549622, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.3694, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.31114641978405194, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.3842, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.28491189296455693, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.371, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.27387529528555665, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.3853, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.26309335177219156, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.3786, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.30160287904440136, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.382, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.29927424315024315, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.366, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.272326049814458, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.3998, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.2502193014427093, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.3791, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.2728654969279862, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.3701, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.2714056788321398, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.3723, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.2751739558506907, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.3863, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.24742383567080395, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.3694, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.24259595387693322, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.3627, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.24217750702784588, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.3805, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.25913091700415253, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.3663, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.2586239046376112, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.3563, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.23579942630271755, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.3543, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.23373961445761668, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.3837, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.23591825589341625, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.3829, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.2353869613244477, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.367, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.2357626051562776, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.3859, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.24024177768466073, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.3762, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.21474250757431293, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.3791, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2230994305247215, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.3881, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.21437251622971168, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.4006, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.23472458900465512, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.3714, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.22834685454037576, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.3719, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.2321116838852033, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.3726, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.25799653704355013, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.3611, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.20430724621101382, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.3721, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.22113708085646827, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.3723, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.22012548360875808, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.3732, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.22222847245134936, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.3768, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.21450792013557687, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.3631, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.21395918035222308, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.3745, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2295027175541579, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.3735, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2074700163824343, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.3808, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.20627809231126568, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.3881, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.2042926461719478, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.3829, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.22097127858331148, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.3726, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.2132949161183552, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.3746, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.21350811403401795, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.3648, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.21619081913763516, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.3658, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.22040882740180187, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.3544, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.20686894830760133, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.3632, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.24523317979996895, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.3614, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.2094299975341201, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.3592, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.20902438069966647, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.3762, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.20521596935129183, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.3515, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.2153152039476982, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.3874, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.21742854143936244, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.3981, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.20306555223195863, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.3915, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.21261935023625392, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.3624, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.21265692761867103, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.3762, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.2092693170717507, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.3573, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.21381173241770798, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.3757, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.20722467428358596, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.3699, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.21363168538840632, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.3723, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.21187736559106565, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.3733, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.21096163827566547, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.3638, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.21188626449673448, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.3594, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.20552467769646116, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.3754, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.21318494767306526, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.3672, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.2090068621882043, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.363, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.20763808048762358, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.3819, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.20795948050505664, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.3558, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2054478121680009, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.3678, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.21770520240117117, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.3493, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.20845976699164964, | |
| "learning_rate": 0.0, | |
| "loss": 0.3647, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 1.9327518915700982e+18, | |
| "train_loss": 0.5209115049013725, | |
| "train_runtime": 62363.4021, | |
| "train_samples_per_second": 0.802, | |
| "train_steps_per_second": 0.006 | |
| } | |
| ], | |
| "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": 1.9327518915700982e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |