| { | |
| "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": 7.441595063490035, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.12, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 7.484638177083275, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.1255, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 7.202111396440288, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0896, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 6.452170654218077, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.0644, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 5.188815843122055, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 1.0445, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 3.247377461004119, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.997, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.7719390197110103, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.9664, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.4823818390219525, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.9717, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.242204249260898, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.9553, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.255605010455584, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.9199, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 3.9237316733011736, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8879, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.693596381156768, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8633, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.4889591499979207, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8242, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.3543634039701615, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.8138, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.075920207094057, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8056, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.7155500914814588, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7874, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.4914607967548483, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7711, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.5009937594874494, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7573, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.2398047186940726, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7462, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.1963702252810566, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7413, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.3150779973188758, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7376, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.159843431165733, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7255, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 1.2063555822294771, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7215, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.1109265403583821, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.703, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.1720897213199182, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7012, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 1.2825211703773385, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7153, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.8308063777209478, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7024, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 1.0281482576483572, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.6948, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.8200341157177883, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.701, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.060566942365839, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.6869, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 1.1242162874486763, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6801, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.7775219332751483, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6814, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 1.2352462802201458, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6894, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.2277510115788453, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6977, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.9228331091805946, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6591, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.7174685276202803, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.679, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.176892260603213, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6894, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 1.0278248870635087, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6669, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 1.3605533899907176, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6717, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.7285485484642533, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6762, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.9941589848148505, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6515, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.3925480817395375, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6684, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 1.3738521223323201, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6763, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.0279109549483008, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6664, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.3481950601071695, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6638, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.9119821810608826, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6697, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.9919432290349403, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6577, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 1.2138896158517192, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6666, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.2080681941689464, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6559, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.1239281784032824, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6605, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.8166856534066423, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6521, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 1.196527954842215, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6509, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.7983369444352492, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6635, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.9218480580978282, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6476, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.7282292792713069, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6559, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.8484906554368669, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6494, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.9186950061345318, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6521, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.7725234236884608, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.65, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.6135046864434999, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6526, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.0567785588566585, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6396, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.7404946727222057, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.64, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.6639952047181514, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6573, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.7816541565608149, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6432, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.8480913553934821, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6436, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.7674933034474372, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6415, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6962553280648376, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6462, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.754547010446202, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6475, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.6914569903553703, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.653, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.7183778947689864, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6359, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.6814811992435078, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.64, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.7651047855634444, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6381, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.7111755768601328, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6334, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.9250555747835286, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6218, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.7731240952783506, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6329, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.8881072363547095, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.628, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.7660357253718983, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6406, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.9274912155298295, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6324, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 1.230886069065375, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6385, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 0.6536507394497568, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.5951, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 1.019856000451166, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5657, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 1.1582963989458797, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5762, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.7378208821850025, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.5495, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.9629443781107777, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5622, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.8215898497844466, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5651, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.8062659683557859, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5595, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.5974630084173416, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5453, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.6590847969848663, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5507, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.6924831332236423, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5676, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.72991228553215, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5577, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.7421947536546772, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5523, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.5816156716395124, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5529, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.8598648942477825, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5513, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.7438234185802748, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5612, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.6483403429863566, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5603, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.7542774522709634, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5619, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.7461751255762334, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5532, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.9624590582142856, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5351, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.5633453029603849, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5636, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.7716354751517465, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5589, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.7282392649542447, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5562, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.6468222150865308, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5604, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.7663780069012006, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5444, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.6728408269448793, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5646, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.7107729370436198, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5445, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.6501208118140137, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5598, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.6825664278161179, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5642, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.6763795683820295, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5584, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.6040370208325733, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5523, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.7594631571840751, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5391, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.7008429919676409, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5479, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5904255842137021, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5527, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.6422304747709738, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5403, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.6162161895097944, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5444, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5696385308300248, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5443, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5745052658890734, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5556, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.4974602909361061, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5471, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.5577532605223752, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5558, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5174391252499035, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5421, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.5512260293603698, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.5448, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5605182595480298, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.5602, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.6193684260808888, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.5527, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.5553229356906002, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5683, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5850872989870655, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5488, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.4813388614506171, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5494, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.6002838304443686, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5372, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5977081380361673, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5488, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.6541333572193411, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.5444, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.5951206528492121, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5668, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.6221414440418981, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.559, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5934743951126129, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5304, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.7838559999591367, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5468, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.5268273239649981, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5459, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.6092008468432318, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.5553, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.497438975374191, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5405, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5740690973359671, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5464, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.5221731863629598, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5404, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.5387878340861486, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5535, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5347687783365402, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5495, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.6045959368465982, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5467, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.4467725422798922, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5553, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.6663767836670907, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5539, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.4552317360738846, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5469, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.6230492875626062, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5452, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.4761975761713638, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5491, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.47978873745405004, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.549, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.4432295081859974, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5347, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.5003305750000421, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5462, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.5509077577746613, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5547, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.48536813097970655, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5621, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.6461268166103642, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5461, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.5471698672388323, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5545, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.5535315618487886, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5473, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.6282730017895743, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5582, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.5067594555727714, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5674, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.7060287576955648, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5556, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.5237861552370068, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5375, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 0.6678202622563446, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.4848, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6669750923521043, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.4739, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.6399431732450359, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.46, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6751011332197655, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.4612, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.5650723452771215, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.4394, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.586041262485064, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.4426, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.6366589171610696, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.4408, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.6201744872862791, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.4456, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.5829184961363081, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.467, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.5952560690768975, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.4319, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.6156397228610098, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.468, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5592355867355706, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.448, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.5627648775048841, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.4185, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5237278883591286, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.444, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.5658592816380061, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.4459, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.5682196983793549, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.4566, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.5102529062277341, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.4574, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.5013276527024729, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.4727, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.5133783163473737, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.4313, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4695179972729291, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.4692, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.4886224691047343, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.4704, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.496802153085282, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.4503, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.5154406753706321, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.466, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.5608451939426503, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.4587, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.45659520332211434, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.4564, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.5227311360549532, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.4461, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.4387756309191944, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.4479, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.5367015463063387, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.4704, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.47577751129327395, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.4644, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.5384805497260282, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.4353, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.5171349602083627, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.4636, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.46372877012752045, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.4144, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.5267695877549445, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.4444, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.5128841134077331, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.4597, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.5101613385632291, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.437, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.4980424808938728, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.4524, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.5019146512115813, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.449, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.4725650152106896, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.4246, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4522344461295345, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.4573, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.4532537198307878, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.4576, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.38250836302947344, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.4478, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.4804346749670196, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.4521, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.42746502313598084, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.4551, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.5397281638613642, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.4445, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.4036003190873001, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.461, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.5676833483469378, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.4538, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.4082650136958347, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.4532, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.5941132393449255, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.4378, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.42177809042160447, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.4313, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.5829494893057481, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.459, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.44368898960586683, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.4723, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.5281206973339344, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.43, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.4434927023159996, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.4435, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.39764211620379664, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.4709, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.43281731792688005, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.4246, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.4076949243522016, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.4426, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.40937943071521105, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.4488, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.39395551988457694, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.4333, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.4094675463233994, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.4565, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.4246909053508187, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.4442, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.4087712133593184, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.4662, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.41914350963201125, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.4334, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3792408388470156, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.4591, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.41023209297203744, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.4434, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.46330559828002704, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.4524, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.35441939230672964, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.4485, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.43564932421426406, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.4744, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.3814117483935704, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.4592, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.4177127225983221, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.4538, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.42605461927652116, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.442, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.5298260332878959, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.4601, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.45173150756411196, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.4221, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.500521718892224, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.4453, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.459707412496701, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.4467, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.41593816766575187, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.4707, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.46245731898241266, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.4331, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.4016276573512303, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.4566, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.46362716194562703, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.4301, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 0.5004097429818956, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.3689, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.45397929253767844, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.3781, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.7279501653096108, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.3607, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.5271763244671553, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.3798, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.6230623123260388, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.3588, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4691487581943571, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.374, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.5267052496426081, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.3706, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.5067182048115253, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.3671, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.421278553921468, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.3621, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.4343364134376989, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.3757, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.4524399849288063, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.3695, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.39338140551630585, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.3601, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.4365587466463919, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.3577, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4234946069608991, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.3479, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.40138255715707816, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.3643, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.40135399562351226, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.3787, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.3790300509466168, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.3707, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.41758458935097315, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.3905, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.35407080309322747, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.369, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.40656317353230165, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.3549, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.420463928822835, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.3326, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.35160892607322725, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.396, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.423290535153289, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.3702, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3323483152890749, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.3733, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.382062003788085, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.3652, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3527822057086127, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.3505, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.31132924993032884, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.3574, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.38023126218354447, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.3845, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3760193075816827, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.3913, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3584397701304726, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.3488, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.3730714526513745, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.3543, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.3579123444897427, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.3525, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.37105707673062116, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.3747, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.3293600206042143, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.397, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3390868639291035, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.353, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.3280116889287868, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.3698, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3577592658215991, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.3429, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.34859372800510313, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.3464, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.31023285833474384, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.3851, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.3364000022645315, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.356, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.3321914029103981, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.3605, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.3258934278472488, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.3916, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.3584353252301879, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.3623, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.31423092055549057, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.3722, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.3689395382305111, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.3616, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.3533964902772209, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.3708, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.3346117111839511, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.3932, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.31607262685467563, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.3931, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.3382646333246466, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.3719, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.3262881762850147, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.3565, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.31639797972938905, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.3621, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3240547796655519, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.3632, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.34275181674679317, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.3849, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.3453184261266086, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.3758, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.34071235830320895, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.3414, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.3255810804275077, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.3723, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3287782006959308, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.378, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.34162576634422614, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.3785, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.32365384074358655, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.354, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.29539672261333944, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.371, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.3024603487823895, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.3501, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.31182858580612854, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.3709, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.35091428277028275, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.3913, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3319035157120333, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.362, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.29656496926230397, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.3658, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.2998731990744445, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.3909, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.3312793264864784, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.3556, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.30593350276281084, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.3557, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.3194680377559201, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.3356, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.28711000905224604, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.3617, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.2836580888304628, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.3621, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.29883718617423993, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.3673, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.31604534065967155, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.3575, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.3029471310685604, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.3903, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.2966706945691347, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.3574, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.2764131067907005, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.3849, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.31223514877690434, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.3744, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.29567041539338607, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.3415, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.3504411331683098, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3231, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.3931568680835711, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.3025, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.28936012811319806, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.3507, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.32407742733998635, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.3099, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.4680933320551794, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.317, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.4813797502731172, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.353, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.3068835831880434, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.293, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.351051620730367, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.2994, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.37657202784876687, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.288, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.340429910116198, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.2945, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.27336555580482963, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.3367, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.2882006523229666, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.3352, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.34620306890696, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.3337, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.32192969905720603, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.2992, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.2884580223975795, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.3486, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.27459186006874053, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.3206, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.2870978017024338, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.3264, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.3061825714793093, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.3021, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.28007327004321064, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.2989, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.3073486540360964, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.2884, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.26515751186571046, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.328, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.26092068470132884, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.3155, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.2748074640925594, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.3059, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.2860285883258981, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.3278, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.27360963003754485, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.3202, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.2625289125272169, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.3045, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.2677590682338024, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.3413, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.2567974519755871, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.3571, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.25401556042640333, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.3319, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.24749745478560123, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.3153, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.26422956481216253, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.3258, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.24372985394844288, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.3031, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2581077508191922, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.3085, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.25038284641790104, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.3295, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.28978185708152177, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.2994, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.24794574801242528, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.3105, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.25624266514753086, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.3173, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.24324806761672718, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.3235, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.25149631251423993, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.3042, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.23759462417106564, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.3204, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.2500928218978635, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.3411, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.24291903474326795, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.3319, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.23988922674456525, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.3234, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.24176744472261644, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.3401, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2476012377698525, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.3088, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.24029504916945307, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.3153, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.23741894121230409, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.3216, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.24800566532952512, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.3093, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.243947415013688, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.3091, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.2525272399066849, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.2905, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.2534656791817323, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.2995, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.2430105888149135, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.2783, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.23569535625632956, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.3508, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.2540291353288265, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.3058, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.2380598216424529, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.3228, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.24963221995398183, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.3045, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.24346615558606977, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.308, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.23401178265699293, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.3214, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.25175625041590033, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.3227, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.2391601883627694, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.3052, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.23348992719137646, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.3402, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.23720217424937548, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.325, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.2445933538130364, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.3366, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.23489019631785116, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.3392, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.2374303126311836, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.2974, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.2365588987551183, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.3141, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.23431121077879877, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.3358, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.23021249846944067, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.3114, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.2455488381419304, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.3149, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.23795742806289083, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.3388, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.2332652517526661, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.3179, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.24280066249605625, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.2996, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.2546053075094607, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.295, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.24792933487577504, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.3035, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2469568104249538, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.2965, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.22797976337429476, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.34, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2429980732264221, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.2919, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2468825131784473, | |
| "learning_rate": 0.0, | |
| "loss": 0.3017, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 1.349858349918716e+18, | |
| "train_loss": 0.48315360164031, | |
| "train_runtime": 57610.018, | |
| "train_samples_per_second": 0.868, | |
| "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": 1.349858349918716e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |