| { | |
| "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": 5.863900157318113, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.9596, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.109932746524333, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0162, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.20387498758172, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0178, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.790412950499749, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.9835, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.158416716109019, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.9483, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.0993792148946016, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9022, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.058882280839032, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.8501, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.522457391421698, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.8768, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.921125491878129, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8887, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.719086125625584, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.8776, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.00819662188803, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8272, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.050397009508778, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8057, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.547425655198485, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8367, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.239649674283963, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.7608, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.4398965199344405, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7656, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.9379897743248704, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7609, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.4123344749511757, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.732, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.560699485232345, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7513, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.2910905898117273, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7624, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.9767602523000253, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7191, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.113128043736566, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7181, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.762279274050193, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.6926, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.8970542754675157, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7319, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8154373738338958, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.6766, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.9043268429451219, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7083, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.7790288918469499, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.686, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7493033550636032, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.6769, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.7814414824186039, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.6911, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.8858153230819887, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.6811, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.7747815330420422, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.6782, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.6964900372216322, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.686, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.7667266533907808, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6669, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.8501524649816993, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6685, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.7828592371443549, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6574, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.8724394670805821, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6749, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.7512321558447025, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6632, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.7750132744507578, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6512, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.7967123908817402, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6663, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.6947515112348691, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6699, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.8838121858158148, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6103, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.937953074232824, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6294, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.6663100984929431, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.63, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.9686394127674997, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6555, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 1.0948792609440945, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6747, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.7646908276106144, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6803, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 1.0848540792128565, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6544, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 1.0115954713396924, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6655, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.9120073804372092, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6521, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 1.050855454401179, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6357, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.5475520022510691, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6443, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.8273962306006929, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6465, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.7881464919038017, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.5951, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.9537798288964097, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6357, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.579153228532081, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6657, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.074195784702514, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.62, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.7439084879563846, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6552, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.9296398158867684, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6265, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 1.0194408640683725, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6316, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.7355167158335016, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.634, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.906333179886472, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6566, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.9984252494246183, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6418, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.6288544699214411, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6489, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.9808612656456591, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6625, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.7864506596926686, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6287, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.7673819391613449, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.615, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 1.0093619808484628, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6237, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.9977549136570084, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6343, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.6027710823264074, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.5971, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.6159304324960394, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6243, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.6213381460421435, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6372, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.8756713112352614, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6413, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.7675343635238856, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6426, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.8581644262943975, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6203, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.9530066853833877, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6412, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6732284794508566, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6329, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.6765379432047343, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6135, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.6269936249438118, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.5911, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.5241257450024129, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6083, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.6719128755921155, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 1.2361, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.8044098947452024, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5482, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 1.0946885767510257, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5149, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.8598307990653626, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4693, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.8552065496627894, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5323, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.8239284111177897, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5492, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6347944568201517, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5168, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.653364996874192, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5329, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.6989776738975938, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5412, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.7181508475505547, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5296, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.6911928102780305, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.534, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6436708383669849, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5218, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.6396133336267467, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5101, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.6423502214891712, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5607, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6214000215124265, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.4887, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.5503754472052366, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5348, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.5973312350989958, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5112, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.485838378982901, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5239, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.5077143548051274, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5178, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.5689308324907205, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.4831, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.6449782413165691, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5062, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.6021971176344099, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5333, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.7193952024706654, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5002, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.49547113809381366, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5048, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.6618235319736041, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5119, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.5644535979761948, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5088, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5525405461144406, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5037, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.8144469278978772, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5186, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.44891431256783426, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5176, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.6221611200366272, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.4977, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.43947676722734164, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4976, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.5218160924853039, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5085, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5034372949967536, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5002, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.45895268200491246, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4976, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.46321651746461323, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.4968, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.47870495552411885, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5287, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.474762382449506, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.4974, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.47651416342315034, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.4622, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.5410630445106255, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5288, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5637184608262586, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5132, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.5727458424587676, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.4653, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5802567977235668, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.527, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.5707103092411799, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.522, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.5896233440772041, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.516, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.47476490550748923, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.4763, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.4995541864941269, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.4972, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5157397952449607, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5135, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5265684040798319, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5075, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.6509664756017955, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4865, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.5230304058607859, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5053, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.564502117499931, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.4876, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5313724748821392, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5109, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.49048494571733037, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5027, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.483002582063503, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.4732, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.5536908988980134, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.4814, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5135758035148594, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5222, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.6693583165294315, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5069, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.6703936126866795, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5171, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.4402756674176734, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.5164, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.4834750840488784, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.4934, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.5001628764630827, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5172, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.5156475824631777, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5196, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.46421714247369034, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.4828, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.4435566342919553, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5066, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.4472454124285204, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.4885, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.45149483986705774, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.4752, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.43737137429775913, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.5025, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.4144061152549108, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.4949, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.45685221044206653, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5056, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.4384227669456062, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.4953, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.4583195214456875, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.4972, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.5036176164018994, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5273, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.3900927170850886, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.4691, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.5187118022065168, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5203, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4231705325668886, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5122, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.44469869668183354, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5395, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.41913855903509245, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5042, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.44923975878690897, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.479, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.265886045380679, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.9111, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.7788939887932237, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3983, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.5814266423710819, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3809, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6541946579068361, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3784, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 1.2254090579151116, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3789, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.6517013681171183, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3872, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.7885832265283199, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3535, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.5877804328511399, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3413, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6404825701419146, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3907, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.6911953292508461, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3687, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.5199471791956721, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3874, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.6434304308376638, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3429, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 5.039222272531735, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.4248, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 1.5682575966736636, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.4148, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 2.2116325260594603, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3839, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 4.965840848460932, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.422, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 8.972880867629566, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.4676, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 6.120391510692015, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.4401, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 1.7351836374094791, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.4232, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 3.5854369520818015, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3999, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 1.9878205895223442, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3703, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 33.41847469258458, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.4428, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.7737787985631442, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3908, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.5908933797743031, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3526, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.8625462111400396, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3676, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.7499809401337406, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.391, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.49289242477309925, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3627, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.6436201609705232, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3823, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.7416369441658253, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.379, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.49608853671818276, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3519, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.5592841113161919, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3657, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.5517950962622802, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.345, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.5506982072115814, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3873, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.5523277228718634, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3686, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.4836562777887041, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3363, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.4856236540285901, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3713, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.43993841071604617, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3343, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.5734289420004812, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3895, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 1.0507912669903445, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3458, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.6631940633125842, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3981, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.46515671795501223, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3588, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.4918452599317938, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.389, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.4554316499011092, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3461, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.455662913331897, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3636, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.3935422562046005, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3426, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.5002040423927975, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3837, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.43073063190022365, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3753, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.41187216224269874, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3587, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.43523550336085925, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3793, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.40840561472711745, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.358, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.4271901296090871, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3506, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.42938531914891453, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3725, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.385551093697211, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3448, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.42732610929649756, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.354, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.4363697439406535, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3588, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.4455666934289362, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.385, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.48232062459395736, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3644, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.43697749519928014, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3639, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.42215823476760755, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3564, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.4533057037652281, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3585, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.412936501647393, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.386, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.44281905988539394, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3527, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3804050542534375, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3483, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.41179401319943626, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3554, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.43315272736795446, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3702, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.398555958378086, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.351, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.42222659633240006, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3752, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.4358304314931661, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3791, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.39597183289865395, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3442, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.4041807692290575, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3729, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.4300731136350461, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3545, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3879260964768763, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3596, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.4055432164000725, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3365, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3958328462192638, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3664, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.38826305855875354, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3612, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.445023011920815, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3524, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.40354636978208697, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.401, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3723224522633442, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3454, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 1.0760263945169621, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.583, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.57256564444408, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2506, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.5147206452202644, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.247, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.7716271271290052, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2848, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.6412161087083598, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2767, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.46626647972419805, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2324, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.5448589396475085, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2555, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.585804456518818, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2895, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.4300412895380309, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2508, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.5168234192574198, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2685, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.4893794924489026, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2709, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.4968317843194202, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2517, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3928163060384461, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2569, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4237663552092556, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2543, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.5067424096750924, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2762, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.36213911630724266, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.235, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.45366427130401843, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2531, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.43311027908123373, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2678, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.37955430937803575, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2341, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.4332081369900881, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2428, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.40689868552476915, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2615, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.37956577708548206, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.255, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.38931315034089564, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2396, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3867181588805245, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2774, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.36661214027127537, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2476, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3917198545486744, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.266, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.38107425210510537, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2668, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.3639150748646448, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2305, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3895921125067074, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2739, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3604470198674436, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2472, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.3573866946305205, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2442, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.3751341303884494, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2434, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.3533824095734159, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2335, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.38925546709181347, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2563, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3667682256849455, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.237, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.3501415228169262, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2521, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.34288244851982835, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2521, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.3343160035531691, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2486, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.35266964091059416, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2634, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.3310664566174733, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2499, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.3347015557547574, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2596, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.32304663423209445, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2273, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.34703664582821353, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2573, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.34287290395368464, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2408, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.34811520345371805, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2707, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.36299928886945515, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2669, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.3168750185679657, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2354, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.3350612454059626, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2355, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.3251488389428146, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2664, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.3181776200149048, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2501, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3283512698070235, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2543, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3330946669696271, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.261, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.3190143994069841, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2405, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.3762307158093425, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2732, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3241138302819481, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2399, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.3168032723774227, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2517, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3266848288775348, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2321, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.3385676680973875, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2338, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.3460231212436604, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2871, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.32517381734622913, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.263, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.30847719176822114, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2309, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.3369417730320049, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.264, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.31869551674524277, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2379, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.31609783042835005, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2465, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.32882321257536323, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2542, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.3068370521320545, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2461, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.32758411147114447, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2409, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.3221173256989185, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2515, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.32264423655414426, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2466, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.32539030618048703, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2592, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.32643194959120964, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2538, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.32112558543154684, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2324, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.32459137125078197, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2573, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.32316349415665657, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2761, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.3350750218103054, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2614, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.32730727205744853, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2495, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.333725463894041, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2475, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.3294470948013299, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2643, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 1.2839108579019607, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.4669, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.5107764857543692, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.1854, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.5032474964702015, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1981, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3739097179041935, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.1925, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.30220129362139286, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.1865, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3292300542252776, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.1717, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.46968083528040183, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.198, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.498790951481086, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1765, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.521705304577139, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1988, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.3679435482726114, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.1705, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.32621795497574985, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.1817, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.35216362102409743, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.2028, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.35599057654806565, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.187, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.37254584373028504, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.1656, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.37540403302100134, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.1909, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.31221887227619116, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.1664, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.31542463544830635, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1779, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.3211841194629076, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1848, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.27762081759466123, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.1492, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.3385807046048009, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1934, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.30736132267170674, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1672, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.3419430497892458, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1709, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.32901820205339777, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1784, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.32284606187029924, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1883, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.2979899558075267, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.19, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.29358109019278905, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.1805, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.27168467699556326, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1561, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.29855700253581063, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.187, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.3102102832565782, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.18, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.32589748777310773, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1962, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.29069882328179, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1859, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.27290994923559936, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1708, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2803971413210815, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.1631, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.2877557641748574, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.175, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.28713797935091373, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1765, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.2961951071739156, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1886, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.31406853686902136, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1824, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.30517288114113955, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.1828, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.2832178998141666, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1722, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.2815109264602268, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1666, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.29166166476356714, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.179, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.2997842315599802, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.2035, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.30570020154695926, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.1784, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.2887441465885464, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1758, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.28667025350190384, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.174, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2759770219797655, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1868, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2759288320823624, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1781, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.2861440779101426, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.2014, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.27680717610008887, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1841, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.27828400360484457, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1771, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.2922440545224223, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.1739, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.2888552049188564, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1891, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.28350616794758077, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1692, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.288102760229299, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.172, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.27176895224481923, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1522, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.28088791328196594, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1645, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.30491036385171727, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1962, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.29595033711883767, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1991, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.2844453957047236, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.1592, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.2730226751995552, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1608, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.30054267330755824, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1766, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.27825341435155493, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1676, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.27728341090713626, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1708, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.27689669432403957, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1707, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.27564867108595503, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1752, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.28092938065824846, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.178, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2985374876648191, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1859, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.270848132846425, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.1667, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.29173327551585365, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1828, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.2751833092982596, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.1761, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.2731068396465843, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1705, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.2832971823575287, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.181, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.27417127684146547, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1646, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.2839052962915241, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.1792, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2736035642134928, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1696, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2849611699605065, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1696, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.285288808839962, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1822, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2833048397064369, | |
| "learning_rate": 0.0, | |
| "loss": 0.172, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 8.005281734805422e+17, | |
| "train_loss": 0.40785385312942357, | |
| "train_runtime": 14479.4896, | |
| "train_samples_per_second": 3.453, | |
| "train_steps_per_second": 0.027 | |
| } | |
| ], | |
| "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": 8.005281734805422e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |