| { | |
| "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": 6.009749882617182, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.9599, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.075590133166555, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.9633, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 5.928848212980959, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.9821, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.593549976084117, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.9514, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.3077749395428535, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.919, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.128554779855043, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.8915, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.2549551244048707, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.8957, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.7864247721768365, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.9121, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.626876340537182, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8756, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.336119804569353, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.8306, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.302511877585399, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8618, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.947677684488028, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8202, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.19015704170681, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8068, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.374494684428963, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.7848, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.2221265369254075, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7701, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.7915097033861274, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7884, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.4773794153628015, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7238, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.6727898010484572, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7702, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.347417471554053, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7207, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.9985165825714708, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7455, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.199371918170394, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.698, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.9050849760524423, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7686, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.8155446761133532, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7313, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8869615636758892, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7026, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.6917381417070398, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.6997, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.9312681751337729, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.7256, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.8444977059054274, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.6729, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.6994161709970882, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.6767, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.8189218721493167, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.6532, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.9273489215645048, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.6799, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.8723659555826176, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6549, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.8396142752709144, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6789, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.9910244731772918, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6593, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.9841857415990768, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6282, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.8409263366332019, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6567, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.6944232358370123, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6613, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.6207687624657097, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6353, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.6739553832801757, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.677, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.8548712846131624, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6756, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.781040590075017, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6393, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.6108854716934051, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.649, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.0928541930952285, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.624, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.9754069896935101, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6609, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.7788059042133489, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6369, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.8431879486005146, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6444, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.7393570543371014, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6388, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.664068178592833, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6243, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.7090097715617976, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.646, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.7334225515283923, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6354, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5941521369676573, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6154, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.7480566332974071, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6174, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.8626029001361554, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6413, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.8269493677998199, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6223, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.6719630586169119, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.64, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.9941341382437822, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6559, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 1.040084879208832, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6422, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.6221282961811937, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6206, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.7972756107195269, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.641, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.6948631849596781, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6404, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5347832110631282, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6248, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.6179128606435014, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6797, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.6617753023513985, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.691, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.6902323709542636, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6188, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.6264402553731189, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6305, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.6407398873224012, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6326, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.7388963190390668, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6231, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.6354524471878831, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6192, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.6177861464013269, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6173, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.7602413599079468, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6387, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.9670565036053368, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6317, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.6613980458221297, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6302, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.8499765441722105, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.633, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.7909399269817926, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6211, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.6696404012538277, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6122, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.7197742512419036, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6301, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.6336606474653554, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.617, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.7134578520353108, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6162, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5763634005391612, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.5923, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.6994997879674727, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 1.1665, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.663668959507631, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5072, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.7534911865116837, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5348, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.7409852817349848, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.5495, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.755354660257437, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5105, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.6981598309526842, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5243, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.7217298436114843, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5632, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.5964946779646126, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5162, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5369623301962194, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.4974, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.6361825718264023, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5741, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.5457482544455791, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.4817, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.621183171441248, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5256, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.6024982075096807, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5183, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.5507577341659401, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.487, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.656548606457835, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5313, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.7473918019474984, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5399, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.6568692533390533, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5436, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.7960579552072998, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5248, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.6943723192682147, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5135, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.6221308719567451, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5007, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.6890273513261123, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5137, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.635227229227924, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5268, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.5098709423988134, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5346, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.5469654163201183, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5012, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.6211714764394449, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.551, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.5741170359016289, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.499, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5253415678462584, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5128, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5213708075624927, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5088, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5770539720009547, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.4835, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.46986192322372733, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5223, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.5591087979056086, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5374, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.5457341799955456, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5098, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.49155641813331263, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.4972, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.509871711392781, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4891, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.45951388885897904, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5008, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5997831428684143, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5442, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5445503397303164, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5109, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.598543791846891, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5037, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.5970998480869277, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5293, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5147813197264381, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5056, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.6502003897371386, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.515, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.49697191157086323, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.4718, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.5651383024683498, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.537, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.45555861780976925, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4788, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5025579740869665, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.5291, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.4894104654576455, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5093, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.4715029827955709, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5091, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.49773608559013427, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4821, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.5176442333342187, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.513, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.4450541600165966, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5081, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.5531998410934351, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5009, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5792434502747217, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.4676, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.5155949718486056, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5243, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.5650792115497688, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5203, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.46784958408700267, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.4836, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5023883615945592, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5308, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5169840502527886, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5037, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.5220543079428224, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5042, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.46130753008505304, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5174, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.4946362975171465, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5214, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.42301383080892785, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.4934, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.4947156483676366, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5012, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.5351124997580846, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5108, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.4508020010500608, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5006, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.5101355555529757, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5115, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.45733849145355376, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.4989, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.5268663623120013, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.5223, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.4868480415665556, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5141, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.5184169302330628, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5081, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.45665532377726525, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5028, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.4614507720440248, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5034, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.5201947939417807, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5356, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.43968993535740386, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5063, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.5094836151144623, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5371, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4232984520926242, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.4919, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.41319100596071945, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.4973, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.5184173804752, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5169, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.4053846544528848, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4925, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.6118749143877253, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.9233, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.7863880134100316, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.4102, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.5781448769183162, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3942, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6457912680059338, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3802, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.8044233307271424, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.4037, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5707628061321618, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.365, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.5897535922883721, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3883, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.7040790635461859, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3884, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6496892996382279, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3831, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.5748517953923707, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3848, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.6860941657808949, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3881, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5896236316982028, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3846, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.558423895659883, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3701, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.607307251174889, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3808, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.45685602456634594, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.38, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.6360020885315628, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.4027, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.5517232094209592, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3617, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.609757611090094, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3883, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.6861543816140283, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3705, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.435954376407554, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3853, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.6718318274646132, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3697, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.4133365013901801, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3661, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.6715649235810961, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3989, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.45391596865024075, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3655, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.6122675475288637, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.395, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.5366163476820688, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3917, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.4968522278131071, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3786, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.5198372960433549, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3847, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.47393331565989677, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3699, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.4758722047932862, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3643, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.47378298297311894, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3742, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.5028783304528256, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.4079, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.5221328341146546, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3824, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.5053615878023786, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3787, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.49210967693399754, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3476, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.5153040225098335, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3809, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.46181922052099905, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3625, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.4495509119379618, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.357, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.45238033379586606, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3817, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.4400789237385193, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3979, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.4191074925796654, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.37, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.4348942512380275, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3974, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.4254653293435295, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.365, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.4419916685325939, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3817, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.43521716053501874, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3762, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.4651911810225462, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.4017, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.4247457811172361, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3538, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.46689627396409833, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3695, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.4270269682955677, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.372, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.46140627698535097, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3842, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.4404033974446217, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3852, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.46447125762165486, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3927, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.4353947473725421, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.4015, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.40293328111530025, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.363, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.3933434586259424, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3803, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.3687846514408937, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3511, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.387862326668434, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3865, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.38677767608899283, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3714, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.38306217915378327, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3908, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.38537521651307916, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3738, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.35016506129651176, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3782, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.37776179205144933, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3604, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3480851336945175, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3723, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3927542762041621, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3772, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.40104383745149, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.4019, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.39991102206654766, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3694, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.41351292800884626, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.4004, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.39880252030549396, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3829, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.39717876773789196, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3678, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.3622607714160862, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3916, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.3738000058256077, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3802, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.352545760097278, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3548, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.3496122365127777, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3971, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3802549782120675, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3826, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.38354248700477533, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3783, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.38276306092995077, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3699, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.3686956478126846, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3876, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3658053166082081, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3811, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 1.063563500653395, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.602, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.5954422509126113, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2871, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.42069038422184357, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2692, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.7915886965547525, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2619, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.6173333342878601, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2647, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.44597472973693175, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2783, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.5642714225916374, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.3088, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.4567078679847236, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.263, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.42604464173888756, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2639, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.4186615150338383, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2603, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.4559825663187837, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2756, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.4854553006810388, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.285, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3607717151057323, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2409, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4335325127832642, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2565, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.38433308580293546, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2624, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3581942692662836, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2752, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.3567420371929066, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2434, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.40407831288797236, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2608, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.3618767247683211, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2831, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.35654069790014364, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2608, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.3602802923594643, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2718, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.3850065780770819, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2563, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.35269816218065053, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.283, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3703703137659632, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2697, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.34260528173301186, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2657, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3524105492324971, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2525, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.34550026795831074, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2458, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.3555525701416852, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2768, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3554087005303088, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2512, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.36427732115988454, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2781, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.33374538383332664, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2467, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.3322099299664637, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2608, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.3316483800021058, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2576, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.3261938119084201, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2878, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.32511496976671606, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2596, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.3279648409518121, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2522, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3237715770110167, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2762, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.36678758793596783, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2828, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.31879336073393627, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2618, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.3339324373064974, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2479, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.3579404212093104, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2705, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.3343250245742033, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2876, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.3573374018353801, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2352, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.35847619978455914, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2689, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.31397015113635274, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2703, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.3452950142420524, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2721, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.3143815600342237, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2507, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.3745899399214708, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2784, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.32211919532534733, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2595, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.32378890714273556, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2538, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.34184126025699085, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2684, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3110744741854127, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2436, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.3016065674970487, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.259, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.30667559061129823, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.249, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3138172725864594, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2706, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.29888449512718346, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2417, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3137930704733289, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.252, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.3134464419602159, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2616, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.29933269043537897, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2466, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.28417536646293373, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2413, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.33336971430954304, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2764, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.32239257945779837, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2701, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.30497308580102833, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2573, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.2867735986252616, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2412, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.3035179636629331, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2621, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.3019909009193723, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2635, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.3196752333734528, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2782, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.28531033626949587, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2427, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.2837772225572217, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2554, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.31061272879894686, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2685, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.28351976489695185, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2362, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.31016059137446284, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2635, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.32693836242997165, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2741, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.2946948363462871, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2513, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.29666435301956406, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2673, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.29507089916379353, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2516, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.3046896254703635, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2634, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.27559853609481577, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2275, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.9616212187558745, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.4554, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.5168258996309555, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.2015, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.4334684883530649, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1873, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3513297977747081, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.2096, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.3110505815079282, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.201, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.4149097298753098, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.2118, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.40728113429528984, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.186, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.5390053303510833, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.2082, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.4376458282753438, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.195, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.345653785794032, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.194, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.3268952330989083, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.2014, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.33498825289451467, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.1927, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.3712287130355773, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.1959, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.3383757727897693, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.1722, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.33700778557053596, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.1741, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.3200959367864112, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.2148, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.3078300912219208, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1981, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.29047506198378087, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.19, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.29730185270328885, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.2008, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.32676242196428573, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.2022, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.30869720647420357, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1911, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.3262804658826441, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1992, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.3057293185305725, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1839, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.2905791118816385, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1742, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.298381131402672, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.1877, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.30212327976653813, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.1721, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.30563617190990083, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.2003, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.2895545426580402, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1855, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.29607266070528987, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1891, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.2897601109504778, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1867, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.29804744072255257, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1873, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2879008327875041, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1708, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.28892910746278955, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.2106, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.2798229526176588, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1802, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.28711572125872065, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1834, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.2886435298266583, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1916, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.2748238421947878, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1909, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.2726980790930926, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.1897, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.28451298649389384, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1956, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.2835346989669834, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1887, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.27604177758933646, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1812, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.2801843278672649, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.202, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.28434714803809824, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.1987, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.2826187293923764, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1997, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2570290632898358, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.1652, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2819837415945326, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.2047, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2796441988630729, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1755, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.2806177505897117, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1873, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.27611592271146684, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1924, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.28159101053124325, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1982, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.28238031393198926, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.1876, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.2788997965465259, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1922, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.2626324817172001, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1849, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.29044745652242704, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.2025, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.26833814012738494, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1925, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.2733338144733219, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1865, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.27917101107270037, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1989, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.26725231095309865, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1896, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.27037942909220003, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.1856, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.2770940123031874, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1996, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.26513327384044433, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1906, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.2771131655529229, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1845, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.27240144216386997, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1979, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.25890135882026305, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1925, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.26633764962859896, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1894, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.27335284697384926, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1932, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2775690156422715, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1884, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.2552692583309363, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.1658, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.27539156465302406, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1974, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.2658059958925005, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.1718, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.27504020476412633, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1672, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.2856354724861289, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.2051, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.27196662083534523, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1726, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.2745951524088046, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.178, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.28225332772815215, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.2011, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.26980364140447516, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1941, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.25887991291423934, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1799, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.26335457282274577, | |
| "learning_rate": 0.0, | |
| "loss": 0.1948, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 9.228474907497595e+17, | |
| "train_loss": 0.413890665769577, | |
| "train_runtime": 15456.3662, | |
| "train_samples_per_second": 3.235, | |
| "train_steps_per_second": 0.025 | |
| } | |
| ], | |
| "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": 9.228474907497595e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |