| { | |
| "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": 6.512522716818102, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.0696, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.564160181602456, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0564, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.593596024999264, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0596, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.985081448984162, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.0466, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.610526434764138, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 1.0029, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.9674251867637347, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9675, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.6183167426455105, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.9643, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.139986932762751, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.954, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.138386824423963, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.9534, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 3.998663659019224, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.8995, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.601600820871555, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8841, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.4878243820396895, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8629, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.6961514903991608, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8292, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.7117706666141828, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.8133, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.4244399364484353, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8074, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.1830469833483066, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7743, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.0960043478039359, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7738, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 0.925926809490321, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.758, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 0.9063053282979348, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7584, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.9596389662207955, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7314, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.9399749494125175, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7422, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.814900129926128, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7038, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 1.006404352335733, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7135, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8831928954701753, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7071, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8702140393430762, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.6963, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.7905059479264891, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7009, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.8222320920404764, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7051, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.881523147841335, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.702, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.683235818479339, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.6728, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.0916899983703134, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.6754, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.9465110514408511, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6673, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.9155354190132595, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.7061, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.7509262620244194, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6809, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.0495172593977244, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6794, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.1286511421752445, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.681, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.2511380187362742, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6802, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.9748855385002707, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6828, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.8156621108069366, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6633, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.9991235641637006, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6751, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.2749953667768767, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6686, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.9412778755551462, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6687, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.0327337726864538, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.671, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 1.3555098275902229, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.677, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.0142701185284615, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6767, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.3579183101745855, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6708, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.973784803214183, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6578, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 1.3153252930699182, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6557, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.8522883488029417, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6478, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.4392745059653276, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6475, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.947603378372481, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6675, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 1.104626066310503, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6453, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.8481175442606228, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6509, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 1.0218965112916476, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6526, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.9434170277879759, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6421, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.8406699901917485, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6424, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.9026206554782883, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6316, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.033689588644658, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6603, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.791725840028453, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.629, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.9294092788143803, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6519, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1389.7577349506907, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.8286, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 287.84776514709336, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.7328, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 2.60146678427347, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6583, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 1.8859944095922938, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6417, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 1.5315662460399988, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6525, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.184843434264107, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6433, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 1.4893895377369322, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6654, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.8748298486341183, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6284, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 1.3699663923970367, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6406, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.9289499371200274, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6276, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 1.3844222693101031, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6503, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 1.0330742305379728, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6254, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 1.3050581038734472, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6415, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.9757426728868531, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6362, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 1.3538131680621317, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6166, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.1431225945777896, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6133, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 1.2079492751237144, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.631, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 1.1441761825751906, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6193, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.9263058024622803, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6161, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.1770323739591833, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.5719, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.9821440975441464, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.569, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.8823439553609027, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5697, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.9112762210935866, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.5594, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.8518848011570124, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5574, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.8512911490008898, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.556, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6560260951817384, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5708, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.8464326366429381, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5669, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5781958051581789, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.548, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.7538860608432648, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.567, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.7236013322656013, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.56, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6456344742128692, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5782, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.5810578790280398, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5545, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.6284996703688887, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5553, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.539844217783587, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5714, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.5806615507593749, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5488, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.4660908011302106, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.573, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.5423768176914702, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5652, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.5298302045696609, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5682, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.4864740532557761, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.554, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.4938140719050684, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5696, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.566671536862956, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5732, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.6612312551578431, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5531, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.5778627367111475, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5408, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.5757871986128803, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5665, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.5585849057932734, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5567, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.6542173505624117, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5556, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.4569668136989876, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5421, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.6529450991504214, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.55, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.6678394214970926, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5573, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.6734626909404589, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5655, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.8400225821815201, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5633, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5303419096191705, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5478, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.6614344684252336, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5455, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.5472102418789047, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5568, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5180083498261117, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5422, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5697847255506858, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5712, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.6040964974688277, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5487, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4742521971792901, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.566, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.6819923342225018, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5404, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.4722306571762546, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5568, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.45741854977275803, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.5596, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.5192906664033172, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.5526, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.4467233339545157, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5627, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5277534067816066, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5657, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.4448021374192618, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5679, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.48955231560697154, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5461, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5491127185002994, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5252, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.539004560507589, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.559, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.4789128117584534, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5297, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.4551171420622061, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5452, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.4062320230041345, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5443, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.5011568209714911, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5596, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.6149446146414052, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5551, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.6047083533465212, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.547, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.46632700703043184, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5464, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5601214392507189, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5508, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.6059686022001739, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5472, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.4227869398475831, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.562, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.4665187980858091, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5407, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.6066621927137116, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5419, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.45539013593733635, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5594, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.4598890146802132, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5456, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.46840313280462353, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5423, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.47793633660719126, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5611, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.4109635129590553, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5596, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.4757780793309859, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.548, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.4667850045505799, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5599, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.40800983146319364, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5455, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.495724755830386, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5506, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.4387238934256952, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5658, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4777624357651288, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5597, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.4795073239219758, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5462, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.425612185278673, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5501, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.41777498513296457, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5544, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.3978988350859362, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5522, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.49169460352223887, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5368, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.5201109396428173, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5557, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.5223620357395014, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.4905, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.42881429340953764, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.4701, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.6464022269396524, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.4682, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6475226688167106, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.4674, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.612117266008549, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.4709, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5011223881087815, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.4664, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.6467725169112344, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.492, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5939346659836927, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.472, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.5027071095650045, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.4959, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.6048110893909238, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.4611, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.48691381474432466, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.4677, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5630771489986927, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.448, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.5150493831886828, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.4596, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.4765000326678474, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.4661, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.4905566091477221, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.4663, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.42542651510717766, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.4563, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.4012248742478175, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.4439, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.5409433225896136, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.4557, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4015271073114425, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.4518, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4721981299304766, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.455, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.43733176558443665, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.4504, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.4130308223543829, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.4716, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.4444052788560941, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.4798, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.4122329176970373, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.4577, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.38139886405102097, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.4691, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.42493686571960576, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.4716, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.3950827300126743, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.4672, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.44805829499136113, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.474, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.4369143120370755, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.4934, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.44502476357356874, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.4488, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.4914622861856519, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.4719, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.49385858341638383, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.4701, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.559542952455097, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.4688, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.4974545523229099, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.4664, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.5021566761179318, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.4636, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.43199282737226535, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.4485, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.4100491369475057, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.4579, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.4555048740503307, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.4641, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4295774465611716, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.4454, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.4785651412027942, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.4388, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.42227580702493306, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.4739, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.44233593861762177, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.4664, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.4150214176585268, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.4892, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.4988147678655741, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.4642, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.4674183051033628, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.4787, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.4529420246577216, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.4903, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.36156313514672506, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.4672, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.43331599376613517, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.4769, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.34250275398477475, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.4793, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.4153769866085747, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.4749, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3640796467211484, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.4641, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.4605091312839679, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.4588, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.3841760161085162, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.4569, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.4297762401237527, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.4826, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.3989050544628582, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.4773, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.4148902267262383, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.4631, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.37315236923456857, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.4557, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.47822952527855817, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.4505, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.38122602440566117, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.4637, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.44550611997452944, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.4711, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.4927928131481893, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.4796, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.3965958103156703, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.4559, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.34579615803746744, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.465, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3993263463900034, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.4679, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.4029764066771043, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.4523, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3419221200613061, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.4592, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.3595384240299151, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.458, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.3620413876501833, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.4756, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3270882856212691, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.457, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.3446593100809947, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.4533, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.40571435960375596, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.4418, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3635415788213531, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.4523, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.412269187518228, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.4694, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.4445468290789392, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.4578, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.3394327173424959, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.4361, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.3947782619670345, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.47, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.35576225634334796, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.4509, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3991955857455497, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.451, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.4852272293028313, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.4085, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.44316609201957585, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.4021, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.6804901436503166, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.4229, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.46064626578777224, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.419, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.5156640732343791, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.4002, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.5124400542001356, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.381, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.4545025150296528, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.3719, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.4764983476665866, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.3951, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.45902267574729466, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.4014, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.42287998012354977, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.3915, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.40844017593827425, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.3705, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.45596793756186865, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.3649, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.38766625246386077, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.3708, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4769480173172975, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.3881, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.4092396043545385, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.3545, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.448234453310902, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.3927, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.44359469507401983, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.4055, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.4088719915070779, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.3856, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.3674312784017301, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.3791, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.3938521997358805, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.3925, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.3623453990180961, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.3924, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.3600044504780135, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.4029, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.3520606689385732, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.3656, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3327793130489215, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.3894, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.39965934167160516, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.3945, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3161346715660535, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.3747, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.3313302846658303, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.3812, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.32079435244947657, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.3837, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.29381848255351123, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.3955, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.32575149113643387, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.3803, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.29157539527768916, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.4054, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.2988204292744495, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.3828, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.29731376217801975, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.387, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.27018508655208406, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.3829, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.286383510499009, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.3813, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.30871337744485106, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.3849, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.2923862683828382, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.3922, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.31472617862809404, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.3712, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.2822166054269409, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.3909, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.28500555437532055, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.3998, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.30256608738210994, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.3951, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.27877028351239436, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.404, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.29405001535468067, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.4283, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.2846756333337361, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.396, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.25839177653428463, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.4037, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.3170799093435875, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.4064, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.29846928577818926, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.3801, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.28359666691655633, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.4006, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.2853900373223349, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.3978, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.30183351823204946, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.3809, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.26513412424846494, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.3837, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.28491958067646267, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.3924, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.2847449046866254, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.406, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.29049536394008524, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.3777, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.26668171489061726, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.3921, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.2824867701723897, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.4009, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.2837737298294316, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.3792, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.2908414313566709, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.384, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.2878412441106774, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.3564, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.30353853584727125, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.3678, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.28504135457069835, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.3811, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.2791963655781665, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.395, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.28620457724872916, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.3818, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.26750728763761783, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.3944, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.27577656010569185, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.3967, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.2831731684757878, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.4102, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.27543329146697637, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.3753, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.28971938443452644, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.3692, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.2852417011416498, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.3899, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.5772779677453664, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.4008, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.2964062916907281, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.365, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.27420611921751054, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.3847, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.2735906025995398, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.3824, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.29961039877972634, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.3903, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.27504216393149944, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.3994, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.2574123710825498, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.3868, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.27208360795264636, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.368, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.28666725143337507, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.3788, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.3756949491808068, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3429, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.3675268483300072, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.3471, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.28792739863501965, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.3582, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.32875207421348396, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.3564, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.3782587360480824, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.3321, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3872900145340423, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.3325, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.3223359591690514, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.3276, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.33574973758010285, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.3245, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.32491634246657564, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.3351, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.3299263674417641, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.3528, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.2878382459323737, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.351, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.3031903041525479, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.309, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.3184646973612724, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.3427, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.3016381771830608, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.3617, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.3030051610528341, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.3619, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.29765952101137166, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.3471, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.2954796322341825, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.3311, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.2607902733239937, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.3817, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.259352438077992, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.3361, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.27425914878006435, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.3233, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.24715987907218878, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.347, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.25636824736040525, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.3553, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.24870904351333006, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.3484, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.23835917870398024, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.3388, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.22920888616083185, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.3554, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.25631081382757875, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.3124, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.2583359373347175, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.3389, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.2396511474610029, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.3627, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.24716904810583543, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.32, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.2326197051401394, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.3464, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.2421341744892754, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.3678, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.24025637141781805, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.3263, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.24097437823375367, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.3458, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.2263409591200407, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.3465, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.22904641123840114, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.3572, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.24350592466909948, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.348, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.23435074168615438, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.3435, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.23783811548314193, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.3479, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.2861107243391282, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.3378, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.23253504506937483, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.3552, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.2246628160130799, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.3679, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.21692743863919423, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.3625, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.22424168871915062, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.3352, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.21387043655972612, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.3598, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.22191490435125394, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.3536, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.23050300947239208, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.3488, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2361544030495828, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.3383, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.2250713610904457, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.3362, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.22925307838835346, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.3325, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.23543085842037606, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.3331, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.23295996119795018, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.3366, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.23330266394800053, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.3307, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.22636221127553688, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.354, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.21858831782914198, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.337, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.22700372130796417, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.3379, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.2234230664222341, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.3454, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.23696571797831542, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.3378, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.22904682230998671, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.3434, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.22497057374402313, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.3522, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.2258287682376544, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.34, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.21724452205355713, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.3521, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.2426189978091419, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.3176, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.2293805302238564, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.3345, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.22779320333802724, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.3617, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.22437051418026907, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.3259, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.23954190278447945, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.3434, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2424599078756838, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.332, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.22993119820564792, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.3262, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.22123537274119692, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.3182, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.21660597193602085, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.3352, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.22493613651849081, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.3525, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.22343841499976613, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.3381, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.23852936531830982, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.3303, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.22543115265579133, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.3179, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.21691606376951345, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.3524, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.22676220521180954, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.3182, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.22371650819974173, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.3207, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.22001504712277892, | |
| "learning_rate": 0.0, | |
| "loss": 0.3288, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 1.6438208865301955e+18, | |
| "train_loss": 0.49524986499395124, | |
| "train_runtime": 58207.9475, | |
| "train_samples_per_second": 0.859, | |
| "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": 1.6438208865301955e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |