| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.9616838487972509, |
| "eval_steps": 500, |
| "global_step": 11194, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.000859106529209622, |
| "grad_norm": 10.069700241088867, |
| "learning_rate": 1.1061797752808989e-06, |
| "loss": 2.3607, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.001718213058419244, |
| "grad_norm": 2.7327144145965576, |
| "learning_rate": 2.2241573033707866e-06, |
| "loss": 2.1048, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.002577319587628866, |
| "grad_norm": 3.3271842002868652, |
| "learning_rate": 3.342134831460675e-06, |
| "loss": 1.7854, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.003436426116838488, |
| "grad_norm": 1.4632568359375, |
| "learning_rate": 4.4601123595505624e-06, |
| "loss": 1.4144, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.00429553264604811, |
| "grad_norm": 0.96810382604599, |
| "learning_rate": 5.5780898876404495e-06, |
| "loss": 1.2037, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.005154639175257732, |
| "grad_norm": 0.7168840169906616, |
| "learning_rate": 6.6960674157303374e-06, |
| "loss": 1.0357, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.006013745704467354, |
| "grad_norm": 0.6445292830467224, |
| "learning_rate": 7.814044943820226e-06, |
| "loss": 0.9969, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.006872852233676976, |
| "grad_norm": 0.6907113194465637, |
| "learning_rate": 8.932022471910112e-06, |
| "loss": 1.0041, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.007731958762886598, |
| "grad_norm": 0.6093459725379944, |
| "learning_rate": 1.005e-05, |
| "loss": 0.9553, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.00859106529209622, |
| "grad_norm": 0.6630317568778992, |
| "learning_rate": 1.1167977528089888e-05, |
| "loss": 0.9573, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.009450171821305841, |
| "grad_norm": 0.6552422642707825, |
| "learning_rate": 1.2285955056179776e-05, |
| "loss": 0.9811, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.010309278350515464, |
| "grad_norm": 0.673576831817627, |
| "learning_rate": 1.3403932584269664e-05, |
| "loss": 0.9248, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.011168384879725086, |
| "grad_norm": 0.6397550106048584, |
| "learning_rate": 1.4521910112359552e-05, |
| "loss": 0.9155, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.012027491408934709, |
| "grad_norm": 0.6486415863037109, |
| "learning_rate": 1.563988764044944e-05, |
| "loss": 0.9121, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.01288659793814433, |
| "grad_norm": 0.5658463835716248, |
| "learning_rate": 1.6757865168539328e-05, |
| "loss": 0.9371, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.013745704467353952, |
| "grad_norm": 0.5926650762557983, |
| "learning_rate": 1.7875842696629214e-05, |
| "loss": 0.8875, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.014604810996563574, |
| "grad_norm": 0.576126217842102, |
| "learning_rate": 1.89938202247191e-05, |
| "loss": 0.9006, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.015463917525773196, |
| "grad_norm": 0.5573123097419739, |
| "learning_rate": 2e-05, |
| "loss": 0.9152, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.01632302405498282, |
| "grad_norm": 0.5773940086364746, |
| "learning_rate": 2e-05, |
| "loss": 0.8955, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.01718213058419244, |
| "grad_norm": 0.5635377764701843, |
| "learning_rate": 2e-05, |
| "loss": 0.8897, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.01804123711340206, |
| "grad_norm": 0.5612330436706543, |
| "learning_rate": 2e-05, |
| "loss": 0.8959, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.018900343642611683, |
| "grad_norm": 0.6228774785995483, |
| "learning_rate": 2e-05, |
| "loss": 0.8906, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.019759450171821305, |
| "grad_norm": 0.5534242391586304, |
| "learning_rate": 2e-05, |
| "loss": 0.8666, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.020618556701030927, |
| "grad_norm": 0.5617905855178833, |
| "learning_rate": 2e-05, |
| "loss": 0.8639, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.02147766323024055, |
| "grad_norm": 0.5597807765007019, |
| "learning_rate": 2e-05, |
| "loss": 0.8846, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.022336769759450172, |
| "grad_norm": 0.5115612149238586, |
| "learning_rate": 2e-05, |
| "loss": 0.8671, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.023195876288659795, |
| "grad_norm": 0.5317944288253784, |
| "learning_rate": 2e-05, |
| "loss": 0.8915, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.024054982817869417, |
| "grad_norm": 0.5044589638710022, |
| "learning_rate": 2e-05, |
| "loss": 0.8656, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.02491408934707904, |
| "grad_norm": 0.5581314563751221, |
| "learning_rate": 2e-05, |
| "loss": 0.8764, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.02577319587628866, |
| "grad_norm": 0.5628443956375122, |
| "learning_rate": 2e-05, |
| "loss": 0.8941, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.02663230240549828, |
| "grad_norm": 0.560197114944458, |
| "learning_rate": 2e-05, |
| "loss": 0.8491, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.027491408934707903, |
| "grad_norm": 0.524337649345398, |
| "learning_rate": 2e-05, |
| "loss": 0.8857, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.028350515463917526, |
| "grad_norm": 0.683029055595398, |
| "learning_rate": 2e-05, |
| "loss": 0.8597, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.029209621993127148, |
| "grad_norm": 0.5435221195220947, |
| "learning_rate": 2e-05, |
| "loss": 0.8483, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.03006872852233677, |
| "grad_norm": 0.5640478134155273, |
| "learning_rate": 2e-05, |
| "loss": 0.8865, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.030927835051546393, |
| "grad_norm": 0.4870714545249939, |
| "learning_rate": 2e-05, |
| "loss": 0.8616, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.03178694158075601, |
| "grad_norm": 0.4941020607948303, |
| "learning_rate": 2e-05, |
| "loss": 0.8745, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.03264604810996564, |
| "grad_norm": 0.4704708755016327, |
| "learning_rate": 2e-05, |
| "loss": 0.8337, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.03350515463917526, |
| "grad_norm": 0.5146850347518921, |
| "learning_rate": 2e-05, |
| "loss": 0.8392, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.03436426116838488, |
| "grad_norm": 0.5625430345535278, |
| "learning_rate": 2e-05, |
| "loss": 0.8679, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.0352233676975945, |
| "grad_norm": 0.5005638003349304, |
| "learning_rate": 2e-05, |
| "loss": 0.8686, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.03608247422680412, |
| "grad_norm": 0.5048443078994751, |
| "learning_rate": 2e-05, |
| "loss": 0.8587, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.036941580756013746, |
| "grad_norm": 0.5067149996757507, |
| "learning_rate": 2e-05, |
| "loss": 0.8821, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.037800687285223365, |
| "grad_norm": 0.5074286460876465, |
| "learning_rate": 2e-05, |
| "loss": 0.8589, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.03865979381443299, |
| "grad_norm": 0.4815230965614319, |
| "learning_rate": 2e-05, |
| "loss": 0.8735, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.03951890034364261, |
| "grad_norm": 0.4682285785675049, |
| "learning_rate": 2e-05, |
| "loss": 0.8988, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.040378006872852236, |
| "grad_norm": 0.4762243628501892, |
| "learning_rate": 2e-05, |
| "loss": 0.8572, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.041237113402061855, |
| "grad_norm": 0.48740312457084656, |
| "learning_rate": 2e-05, |
| "loss": 0.819, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.04209621993127148, |
| "grad_norm": 0.4676920473575592, |
| "learning_rate": 2e-05, |
| "loss": 0.8169, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.0429553264604811, |
| "grad_norm": 0.4818402826786041, |
| "learning_rate": 2e-05, |
| "loss": 0.8188, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.04381443298969072, |
| "grad_norm": 0.536251425743103, |
| "learning_rate": 2e-05, |
| "loss": 0.8857, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.044673539518900345, |
| "grad_norm": 0.46843159198760986, |
| "learning_rate": 2e-05, |
| "loss": 0.8162, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.04553264604810996, |
| "grad_norm": 0.47590792179107666, |
| "learning_rate": 2e-05, |
| "loss": 0.8749, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.04639175257731959, |
| "grad_norm": 0.48866006731987, |
| "learning_rate": 2e-05, |
| "loss": 0.8761, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.04725085910652921, |
| "grad_norm": 0.529727041721344, |
| "learning_rate": 2e-05, |
| "loss": 0.8591, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.048109965635738834, |
| "grad_norm": 0.46046891808509827, |
| "learning_rate": 2e-05, |
| "loss": 0.84, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.04896907216494845, |
| "grad_norm": 0.503052830696106, |
| "learning_rate": 2e-05, |
| "loss": 0.841, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.04982817869415808, |
| "grad_norm": 0.47345227003097534, |
| "learning_rate": 2e-05, |
| "loss": 0.8206, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.0506872852233677, |
| "grad_norm": 0.4749337136745453, |
| "learning_rate": 2e-05, |
| "loss": 0.8191, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.05154639175257732, |
| "grad_norm": 0.45893123745918274, |
| "learning_rate": 2e-05, |
| "loss": 0.8512, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.05240549828178694, |
| "grad_norm": 0.4918938875198364, |
| "learning_rate": 2e-05, |
| "loss": 0.8324, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.05326460481099656, |
| "grad_norm": 0.5130265951156616, |
| "learning_rate": 2e-05, |
| "loss": 0.8261, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.05412371134020619, |
| "grad_norm": 0.48524969816207886, |
| "learning_rate": 2e-05, |
| "loss": 0.8402, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.054982817869415807, |
| "grad_norm": 0.4443276822566986, |
| "learning_rate": 2e-05, |
| "loss": 0.778, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.05584192439862543, |
| "grad_norm": 0.48050761222839355, |
| "learning_rate": 2e-05, |
| "loss": 0.8328, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.05670103092783505, |
| "grad_norm": 0.468749463558197, |
| "learning_rate": 2e-05, |
| "loss": 0.8126, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.05756013745704467, |
| "grad_norm": 0.47499385476112366, |
| "learning_rate": 2e-05, |
| "loss": 0.8567, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.058419243986254296, |
| "grad_norm": 0.47868019342422485, |
| "learning_rate": 2e-05, |
| "loss": 0.8247, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.059278350515463915, |
| "grad_norm": 0.4504624009132385, |
| "learning_rate": 2e-05, |
| "loss": 0.8265, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.06013745704467354, |
| "grad_norm": 0.481806218624115, |
| "learning_rate": 2e-05, |
| "loss": 0.8208, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.06099656357388316, |
| "grad_norm": 2.282531261444092, |
| "learning_rate": 2e-05, |
| "loss": 0.8281, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.061855670103092786, |
| "grad_norm": 0.48688116669654846, |
| "learning_rate": 2e-05, |
| "loss": 0.8329, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.0627147766323024, |
| "grad_norm": 0.5048409104347229, |
| "learning_rate": 2e-05, |
| "loss": 0.8492, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.06357388316151202, |
| "grad_norm": 0.493664413690567, |
| "learning_rate": 2e-05, |
| "loss": 0.8449, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.06443298969072164, |
| "grad_norm": 0.4665951728820801, |
| "learning_rate": 2e-05, |
| "loss": 0.8156, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.06529209621993128, |
| "grad_norm": 0.4632743000984192, |
| "learning_rate": 2e-05, |
| "loss": 0.8592, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.0661512027491409, |
| "grad_norm": 0.5417582392692566, |
| "learning_rate": 2e-05, |
| "loss": 0.8185, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.06701030927835051, |
| "grad_norm": 0.5666975378990173, |
| "learning_rate": 2e-05, |
| "loss": 0.8282, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.06786941580756013, |
| "grad_norm": 0.4715534746646881, |
| "learning_rate": 2e-05, |
| "loss": 0.8251, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.06872852233676977, |
| "grad_norm": 0.4496767818927765, |
| "learning_rate": 2e-05, |
| "loss": 0.8049, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.06958762886597938, |
| "grad_norm": 0.47709542512893677, |
| "learning_rate": 2e-05, |
| "loss": 0.84, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.070446735395189, |
| "grad_norm": 0.49228349328041077, |
| "learning_rate": 2e-05, |
| "loss": 0.8107, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.07130584192439862, |
| "grad_norm": 0.4816240072250366, |
| "learning_rate": 2e-05, |
| "loss": 0.8094, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.07216494845360824, |
| "grad_norm": 1.0014556646347046, |
| "learning_rate": 2e-05, |
| "loss": 0.8432, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.07302405498281787, |
| "grad_norm": 0.4686058461666107, |
| "learning_rate": 2e-05, |
| "loss": 0.8189, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.07388316151202749, |
| "grad_norm": 0.4596993923187256, |
| "learning_rate": 2e-05, |
| "loss": 0.8065, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.07474226804123711, |
| "grad_norm": 0.48179084062576294, |
| "learning_rate": 2e-05, |
| "loss": 0.829, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.07560137457044673, |
| "grad_norm": 0.5536651611328125, |
| "learning_rate": 2e-05, |
| "loss": 0.8229, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.07646048109965636, |
| "grad_norm": 0.49012085795402527, |
| "learning_rate": 2e-05, |
| "loss": 0.8365, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.07731958762886598, |
| "grad_norm": 0.46614527702331543, |
| "learning_rate": 2e-05, |
| "loss": 0.8059, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.0781786941580756, |
| "grad_norm": 0.49526447057724, |
| "learning_rate": 2e-05, |
| "loss": 0.823, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.07903780068728522, |
| "grad_norm": 0.4854036271572113, |
| "learning_rate": 2e-05, |
| "loss": 0.831, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.07989690721649484, |
| "grad_norm": 0.46326735615730286, |
| "learning_rate": 2e-05, |
| "loss": 0.8298, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.08075601374570447, |
| "grad_norm": 0.48059317469596863, |
| "learning_rate": 2e-05, |
| "loss": 0.8074, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.08161512027491409, |
| "grad_norm": 0.4574795365333557, |
| "learning_rate": 2e-05, |
| "loss": 0.8097, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.08247422680412371, |
| "grad_norm": 0.44573479890823364, |
| "learning_rate": 2e-05, |
| "loss": 0.8077, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.08333333333333333, |
| "grad_norm": 0.4629650115966797, |
| "learning_rate": 2e-05, |
| "loss": 0.798, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.08419243986254296, |
| "grad_norm": 0.47595855593681335, |
| "learning_rate": 2e-05, |
| "loss": 0.8105, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.08505154639175258, |
| "grad_norm": 0.4622196555137634, |
| "learning_rate": 2e-05, |
| "loss": 0.8058, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.0859106529209622, |
| "grad_norm": 0.4406432807445526, |
| "learning_rate": 2e-05, |
| "loss": 0.849, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.08676975945017182, |
| "grad_norm": 0.47430476546287537, |
| "learning_rate": 2e-05, |
| "loss": 0.8334, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.08762886597938144, |
| "grad_norm": 0.4646911323070526, |
| "learning_rate": 2e-05, |
| "loss": 0.8302, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.08848797250859107, |
| "grad_norm": 0.503621518611908, |
| "learning_rate": 2e-05, |
| "loss": 0.8303, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.08934707903780069, |
| "grad_norm": 0.42737218737602234, |
| "learning_rate": 2e-05, |
| "loss": 0.7868, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.09020618556701031, |
| "grad_norm": 0.4449130892753601, |
| "learning_rate": 2e-05, |
| "loss": 0.8059, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.09106529209621993, |
| "grad_norm": 0.4552324414253235, |
| "learning_rate": 2e-05, |
| "loss": 0.8151, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.09192439862542956, |
| "grad_norm": 0.4285755455493927, |
| "learning_rate": 2e-05, |
| "loss": 0.7987, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.09278350515463918, |
| "grad_norm": 0.44612982869148254, |
| "learning_rate": 2e-05, |
| "loss": 0.8091, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.0936426116838488, |
| "grad_norm": 0.41484394669532776, |
| "learning_rate": 2e-05, |
| "loss": 0.802, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.09450171821305842, |
| "grad_norm": 0.43857651948928833, |
| "learning_rate": 2e-05, |
| "loss": 0.8051, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.09536082474226804, |
| "grad_norm": 0.4621300995349884, |
| "learning_rate": 2e-05, |
| "loss": 0.8114, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.09621993127147767, |
| "grad_norm": 0.46743395924568176, |
| "learning_rate": 2e-05, |
| "loss": 0.791, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.09707903780068729, |
| "grad_norm": 0.4864065945148468, |
| "learning_rate": 2e-05, |
| "loss": 0.8253, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.0979381443298969, |
| "grad_norm": 0.4746995270252228, |
| "learning_rate": 2e-05, |
| "loss": 0.8027, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.09879725085910653, |
| "grad_norm": 0.43570324778556824, |
| "learning_rate": 2e-05, |
| "loss": 0.7926, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.09965635738831616, |
| "grad_norm": 0.6272392868995667, |
| "learning_rate": 2e-05, |
| "loss": 0.7873, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.10051546391752578, |
| "grad_norm": 0.4445926547050476, |
| "learning_rate": 2e-05, |
| "loss": 0.801, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.1013745704467354, |
| "grad_norm": 0.44602736830711365, |
| "learning_rate": 2e-05, |
| "loss": 0.8186, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.10223367697594501, |
| "grad_norm": 0.49426504969596863, |
| "learning_rate": 2e-05, |
| "loss": 0.8112, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.10309278350515463, |
| "grad_norm": 0.4448431730270386, |
| "learning_rate": 2e-05, |
| "loss": 0.8012, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.10395189003436427, |
| "grad_norm": 0.49004390835762024, |
| "learning_rate": 2e-05, |
| "loss": 0.8141, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.10481099656357389, |
| "grad_norm": 0.48352155089378357, |
| "learning_rate": 2e-05, |
| "loss": 0.8125, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.1056701030927835, |
| "grad_norm": 0.43718385696411133, |
| "learning_rate": 2e-05, |
| "loss": 0.808, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.10652920962199312, |
| "grad_norm": 0.46456244587898254, |
| "learning_rate": 2e-05, |
| "loss": 0.8023, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.10738831615120274, |
| "grad_norm": 0.4470098316669464, |
| "learning_rate": 2e-05, |
| "loss": 0.7836, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.10824742268041238, |
| "grad_norm": 0.44037750363349915, |
| "learning_rate": 2e-05, |
| "loss": 0.8215, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.109106529209622, |
| "grad_norm": 0.4431336224079132, |
| "learning_rate": 2e-05, |
| "loss": 0.8164, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.10996563573883161, |
| "grad_norm": 0.4316923916339874, |
| "learning_rate": 2e-05, |
| "loss": 0.824, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.11082474226804123, |
| "grad_norm": 0.4511774778366089, |
| "learning_rate": 2e-05, |
| "loss": 0.8211, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.11168384879725086, |
| "grad_norm": 0.41615837812423706, |
| "learning_rate": 2e-05, |
| "loss": 0.7879, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.11254295532646048, |
| "grad_norm": 0.42174649238586426, |
| "learning_rate": 2e-05, |
| "loss": 0.7917, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.1134020618556701, |
| "grad_norm": 0.5034571886062622, |
| "learning_rate": 2e-05, |
| "loss": 0.8056, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.11426116838487972, |
| "grad_norm": 0.4601093530654907, |
| "learning_rate": 2e-05, |
| "loss": 0.819, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.11512027491408934, |
| "grad_norm": 0.4226849377155304, |
| "learning_rate": 2e-05, |
| "loss": 0.7694, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.11597938144329897, |
| "grad_norm": 0.4490329921245575, |
| "learning_rate": 2e-05, |
| "loss": 0.8005, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.11683848797250859, |
| "grad_norm": 0.43054598569869995, |
| "learning_rate": 2e-05, |
| "loss": 0.7899, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.11769759450171821, |
| "grad_norm": 0.4534497559070587, |
| "learning_rate": 2e-05, |
| "loss": 0.7901, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.11855670103092783, |
| "grad_norm": 0.4433537423610687, |
| "learning_rate": 2e-05, |
| "loss": 0.8133, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.11941580756013746, |
| "grad_norm": 0.4343333840370178, |
| "learning_rate": 2e-05, |
| "loss": 0.7982, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.12027491408934708, |
| "grad_norm": 0.42904624342918396, |
| "learning_rate": 2e-05, |
| "loss": 0.799, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.1211340206185567, |
| "grad_norm": 0.4655359983444214, |
| "learning_rate": 2e-05, |
| "loss": 0.8211, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.12199312714776632, |
| "grad_norm": 0.4095928966999054, |
| "learning_rate": 2e-05, |
| "loss": 0.812, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.12285223367697594, |
| "grad_norm": 0.41238921880722046, |
| "learning_rate": 2e-05, |
| "loss": 0.7756, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.12371134020618557, |
| "grad_norm": 0.4852188229560852, |
| "learning_rate": 2e-05, |
| "loss": 0.8303, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.12457044673539519, |
| "grad_norm": 0.44924116134643555, |
| "learning_rate": 2e-05, |
| "loss": 0.8325, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.1254295532646048, |
| "grad_norm": 0.4346265494823456, |
| "learning_rate": 2e-05, |
| "loss": 0.7992, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.12628865979381443, |
| "grad_norm": 0.4454062581062317, |
| "learning_rate": 2e-05, |
| "loss": 0.797, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.12714776632302405, |
| "grad_norm": 0.4472755193710327, |
| "learning_rate": 2e-05, |
| "loss": 0.8058, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.12800687285223367, |
| "grad_norm": 0.43495413661003113, |
| "learning_rate": 2e-05, |
| "loss": 0.799, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.12886597938144329, |
| "grad_norm": 0.42517977952957153, |
| "learning_rate": 2e-05, |
| "loss": 0.8023, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.12972508591065293, |
| "grad_norm": 0.5048529505729675, |
| "learning_rate": 2e-05, |
| "loss": 0.7935, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.13058419243986255, |
| "grad_norm": 0.41410166025161743, |
| "learning_rate": 2e-05, |
| "loss": 0.7798, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.13144329896907217, |
| "grad_norm": 0.4294983744621277, |
| "learning_rate": 2e-05, |
| "loss": 0.7878, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.1323024054982818, |
| "grad_norm": 0.42941445112228394, |
| "learning_rate": 2e-05, |
| "loss": 0.7884, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.1331615120274914, |
| "grad_norm": 0.4035552740097046, |
| "learning_rate": 2e-05, |
| "loss": 0.792, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.13402061855670103, |
| "grad_norm": 0.4652338922023773, |
| "learning_rate": 2e-05, |
| "loss": 0.777, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.13487972508591065, |
| "grad_norm": 0.4364311993122101, |
| "learning_rate": 2e-05, |
| "loss": 0.7723, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.13573883161512026, |
| "grad_norm": 0.4207146465778351, |
| "learning_rate": 2e-05, |
| "loss": 0.7612, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.13659793814432988, |
| "grad_norm": 0.4272591769695282, |
| "learning_rate": 2e-05, |
| "loss": 0.8061, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.13745704467353953, |
| "grad_norm": 0.4509749710559845, |
| "learning_rate": 2e-05, |
| "loss": 0.7926, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.13831615120274915, |
| "grad_norm": 0.4478810429573059, |
| "learning_rate": 2e-05, |
| "loss": 0.7896, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.13917525773195877, |
| "grad_norm": 0.4207334518432617, |
| "learning_rate": 2e-05, |
| "loss": 0.7678, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.1400343642611684, |
| "grad_norm": 0.4360485374927521, |
| "learning_rate": 2e-05, |
| "loss": 0.7909, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.140893470790378, |
| "grad_norm": 0.4345305562019348, |
| "learning_rate": 2e-05, |
| "loss": 0.8139, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.14175257731958762, |
| "grad_norm": 0.43382421135902405, |
| "learning_rate": 2e-05, |
| "loss": 0.797, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.14261168384879724, |
| "grad_norm": 0.42108669877052307, |
| "learning_rate": 2e-05, |
| "loss": 0.804, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.14347079037800686, |
| "grad_norm": 0.45418545603752136, |
| "learning_rate": 2e-05, |
| "loss": 0.7852, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.14432989690721648, |
| "grad_norm": 0.4910484552383423, |
| "learning_rate": 2e-05, |
| "loss": 0.7669, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.14518900343642613, |
| "grad_norm": 0.4287989139556885, |
| "learning_rate": 2e-05, |
| "loss": 0.8165, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.14604810996563575, |
| "grad_norm": 0.5834421515464783, |
| "learning_rate": 2e-05, |
| "loss": 0.7962, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.14690721649484537, |
| "grad_norm": 0.5063323378562927, |
| "learning_rate": 2e-05, |
| "loss": 0.7734, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.14776632302405499, |
| "grad_norm": 0.4238131046295166, |
| "learning_rate": 2e-05, |
| "loss": 0.8028, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.1486254295532646, |
| "grad_norm": 0.38663214445114136, |
| "learning_rate": 2e-05, |
| "loss": 0.7988, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.14948453608247422, |
| "grad_norm": 0.44796040654182434, |
| "learning_rate": 2e-05, |
| "loss": 0.7949, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.15034364261168384, |
| "grad_norm": 0.41491448879241943, |
| "learning_rate": 2e-05, |
| "loss": 0.7628, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.15120274914089346, |
| "grad_norm": 0.4768643379211426, |
| "learning_rate": 2e-05, |
| "loss": 0.7741, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.15206185567010308, |
| "grad_norm": 0.42717429995536804, |
| "learning_rate": 2e-05, |
| "loss": 0.7955, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.15292096219931273, |
| "grad_norm": 0.43356406688690186, |
| "learning_rate": 2e-05, |
| "loss": 0.7826, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.15378006872852235, |
| "grad_norm": 0.45616310834884644, |
| "learning_rate": 2e-05, |
| "loss": 0.7821, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.15463917525773196, |
| "grad_norm": 0.4642743468284607, |
| "learning_rate": 2e-05, |
| "loss": 0.7832, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.15549828178694158, |
| "grad_norm": 0.42061248421669006, |
| "learning_rate": 2e-05, |
| "loss": 0.7829, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.1563573883161512, |
| "grad_norm": 0.473254919052124, |
| "learning_rate": 2e-05, |
| "loss": 0.7968, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.15721649484536082, |
| "grad_norm": 0.4449467957019806, |
| "learning_rate": 2e-05, |
| "loss": 0.7904, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.15807560137457044, |
| "grad_norm": 0.4272187054157257, |
| "learning_rate": 2e-05, |
| "loss": 0.7939, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.15893470790378006, |
| "grad_norm": 0.4838572144508362, |
| "learning_rate": 2e-05, |
| "loss": 0.808, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.15979381443298968, |
| "grad_norm": 0.44314128160476685, |
| "learning_rate": 2e-05, |
| "loss": 0.7635, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.16065292096219932, |
| "grad_norm": 0.44182509183883667, |
| "learning_rate": 2e-05, |
| "loss": 0.7908, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.16151202749140894, |
| "grad_norm": 0.4147513508796692, |
| "learning_rate": 2e-05, |
| "loss": 0.7883, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.16237113402061856, |
| "grad_norm": 0.44913968443870544, |
| "learning_rate": 2e-05, |
| "loss": 0.83, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.16323024054982818, |
| "grad_norm": 0.43962720036506653, |
| "learning_rate": 2e-05, |
| "loss": 0.7889, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.1640893470790378, |
| "grad_norm": 0.41255542635917664, |
| "learning_rate": 2e-05, |
| "loss": 0.7715, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.16494845360824742, |
| "grad_norm": 0.4223017692565918, |
| "learning_rate": 2e-05, |
| "loss": 0.7746, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.16580756013745704, |
| "grad_norm": 0.4158386290073395, |
| "learning_rate": 2e-05, |
| "loss": 0.7717, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.16666666666666666, |
| "grad_norm": 0.42182308435440063, |
| "learning_rate": 2e-05, |
| "loss": 0.7707, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.16752577319587628, |
| "grad_norm": 0.43559759855270386, |
| "learning_rate": 2e-05, |
| "loss": 0.7756, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.16838487972508592, |
| "grad_norm": 0.4135313332080841, |
| "learning_rate": 2e-05, |
| "loss": 0.7793, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.16924398625429554, |
| "grad_norm": 0.4393354058265686, |
| "learning_rate": 2e-05, |
| "loss": 0.7774, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.17010309278350516, |
| "grad_norm": 0.40443214774131775, |
| "learning_rate": 2e-05, |
| "loss": 0.7891, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.17096219931271478, |
| "grad_norm": 0.44918447732925415, |
| "learning_rate": 2e-05, |
| "loss": 0.7923, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.1718213058419244, |
| "grad_norm": 0.4612440764904022, |
| "learning_rate": 2e-05, |
| "loss": 0.7896, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.17268041237113402, |
| "grad_norm": 0.43252137303352356, |
| "learning_rate": 2e-05, |
| "loss": 0.7628, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.17353951890034364, |
| "grad_norm": 0.4581339657306671, |
| "learning_rate": 2e-05, |
| "loss": 0.8109, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.17439862542955326, |
| "grad_norm": 0.4455418884754181, |
| "learning_rate": 2e-05, |
| "loss": 0.7978, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.17525773195876287, |
| "grad_norm": 0.4451400637626648, |
| "learning_rate": 2e-05, |
| "loss": 0.7832, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.17611683848797252, |
| "grad_norm": 0.4293883144855499, |
| "learning_rate": 2e-05, |
| "loss": 0.8197, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.17697594501718214, |
| "grad_norm": 0.41408681869506836, |
| "learning_rate": 2e-05, |
| "loss": 0.7809, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.17783505154639176, |
| "grad_norm": 0.4040951132774353, |
| "learning_rate": 2e-05, |
| "loss": 0.794, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.17869415807560138, |
| "grad_norm": 0.8779594898223877, |
| "learning_rate": 2e-05, |
| "loss": 0.7654, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.179553264604811, |
| "grad_norm": 0.45652055740356445, |
| "learning_rate": 2e-05, |
| "loss": 0.7794, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.18041237113402062, |
| "grad_norm": 0.42162561416625977, |
| "learning_rate": 2e-05, |
| "loss": 0.7856, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.18127147766323023, |
| "grad_norm": 0.4086996614933014, |
| "learning_rate": 2e-05, |
| "loss": 0.7851, |
| "step": 2110 |
| }, |
| { |
| "epoch": 0.18213058419243985, |
| "grad_norm": 0.4172867238521576, |
| "learning_rate": 2e-05, |
| "loss": 0.8041, |
| "step": 2120 |
| }, |
| { |
| "epoch": 0.18298969072164947, |
| "grad_norm": 0.4956143796443939, |
| "learning_rate": 2e-05, |
| "loss": 0.7823, |
| "step": 2130 |
| }, |
| { |
| "epoch": 0.18384879725085912, |
| "grad_norm": 0.41764020919799805, |
| "learning_rate": 2e-05, |
| "loss": 0.7866, |
| "step": 2140 |
| }, |
| { |
| "epoch": 0.18470790378006874, |
| "grad_norm": 0.40405750274658203, |
| "learning_rate": 2e-05, |
| "loss": 0.8238, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.18556701030927836, |
| "grad_norm": 0.4923059642314911, |
| "learning_rate": 2e-05, |
| "loss": 0.83, |
| "step": 2160 |
| }, |
| { |
| "epoch": 0.18642611683848798, |
| "grad_norm": 0.44558045268058777, |
| "learning_rate": 2e-05, |
| "loss": 0.7791, |
| "step": 2170 |
| }, |
| { |
| "epoch": 0.1872852233676976, |
| "grad_norm": 0.395918071269989, |
| "learning_rate": 2e-05, |
| "loss": 0.7709, |
| "step": 2180 |
| }, |
| { |
| "epoch": 0.18814432989690721, |
| "grad_norm": 0.40383675694465637, |
| "learning_rate": 2e-05, |
| "loss": 0.7824, |
| "step": 2190 |
| }, |
| { |
| "epoch": 0.18900343642611683, |
| "grad_norm": 0.42041248083114624, |
| "learning_rate": 2e-05, |
| "loss": 0.7589, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.18986254295532645, |
| "grad_norm": 0.406876802444458, |
| "learning_rate": 2e-05, |
| "loss": 0.7947, |
| "step": 2210 |
| }, |
| { |
| "epoch": 0.19072164948453607, |
| "grad_norm": 0.4463612139225006, |
| "learning_rate": 2e-05, |
| "loss": 0.7608, |
| "step": 2220 |
| }, |
| { |
| "epoch": 0.19158075601374572, |
| "grad_norm": 0.42114925384521484, |
| "learning_rate": 2e-05, |
| "loss": 0.8017, |
| "step": 2230 |
| }, |
| { |
| "epoch": 0.19243986254295534, |
| "grad_norm": 0.5391575694084167, |
| "learning_rate": 2e-05, |
| "loss": 0.8105, |
| "step": 2240 |
| }, |
| { |
| "epoch": 0.19329896907216496, |
| "grad_norm": 0.4293856620788574, |
| "learning_rate": 2e-05, |
| "loss": 0.803, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.19415807560137457, |
| "grad_norm": 0.4132872521877289, |
| "learning_rate": 2e-05, |
| "loss": 0.773, |
| "step": 2260 |
| }, |
| { |
| "epoch": 0.1950171821305842, |
| "grad_norm": 0.42714643478393555, |
| "learning_rate": 2e-05, |
| "loss": 0.7969, |
| "step": 2270 |
| }, |
| { |
| "epoch": 0.1958762886597938, |
| "grad_norm": 0.416900634765625, |
| "learning_rate": 2e-05, |
| "loss": 0.7813, |
| "step": 2280 |
| }, |
| { |
| "epoch": 0.19673539518900343, |
| "grad_norm": 0.40540578961372375, |
| "learning_rate": 2e-05, |
| "loss": 0.7846, |
| "step": 2290 |
| }, |
| { |
| "epoch": 0.19759450171821305, |
| "grad_norm": 0.3905494511127472, |
| "learning_rate": 2e-05, |
| "loss": 0.7636, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.19845360824742267, |
| "grad_norm": 0.4042344391345978, |
| "learning_rate": 2e-05, |
| "loss": 0.7696, |
| "step": 2310 |
| }, |
| { |
| "epoch": 0.19931271477663232, |
| "grad_norm": 0.41851627826690674, |
| "learning_rate": 2e-05, |
| "loss": 0.7524, |
| "step": 2320 |
| }, |
| { |
| "epoch": 0.20017182130584193, |
| "grad_norm": 0.45878472924232483, |
| "learning_rate": 2e-05, |
| "loss": 0.7791, |
| "step": 2330 |
| }, |
| { |
| "epoch": 0.20103092783505155, |
| "grad_norm": 0.3952105641365051, |
| "learning_rate": 2e-05, |
| "loss": 0.7652, |
| "step": 2340 |
| }, |
| { |
| "epoch": 0.20189003436426117, |
| "grad_norm": 0.4032493829727173, |
| "learning_rate": 2e-05, |
| "loss": 0.7915, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.2027491408934708, |
| "grad_norm": 0.4258342981338501, |
| "learning_rate": 2e-05, |
| "loss": 0.7769, |
| "step": 2360 |
| }, |
| { |
| "epoch": 0.2036082474226804, |
| "grad_norm": 0.43611422181129456, |
| "learning_rate": 2e-05, |
| "loss": 0.8027, |
| "step": 2370 |
| }, |
| { |
| "epoch": 0.20446735395189003, |
| "grad_norm": 0.41574737429618835, |
| "learning_rate": 2e-05, |
| "loss": 0.7815, |
| "step": 2380 |
| }, |
| { |
| "epoch": 0.20532646048109965, |
| "grad_norm": 0.4091911017894745, |
| "learning_rate": 2e-05, |
| "loss": 0.7698, |
| "step": 2390 |
| }, |
| { |
| "epoch": 0.20618556701030927, |
| "grad_norm": 0.4178768992424011, |
| "learning_rate": 2e-05, |
| "loss": 0.7955, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.20704467353951891, |
| "grad_norm": 0.4205310046672821, |
| "learning_rate": 2e-05, |
| "loss": 0.754, |
| "step": 2410 |
| }, |
| { |
| "epoch": 0.20790378006872853, |
| "grad_norm": 0.3947570323944092, |
| "learning_rate": 2e-05, |
| "loss": 0.7749, |
| "step": 2420 |
| }, |
| { |
| "epoch": 0.20876288659793815, |
| "grad_norm": 0.425428181886673, |
| "learning_rate": 2e-05, |
| "loss": 0.7831, |
| "step": 2430 |
| }, |
| { |
| "epoch": 0.20962199312714777, |
| "grad_norm": 0.4144151210784912, |
| "learning_rate": 2e-05, |
| "loss": 0.7924, |
| "step": 2440 |
| }, |
| { |
| "epoch": 0.2104810996563574, |
| "grad_norm": 0.40834271907806396, |
| "learning_rate": 2e-05, |
| "loss": 0.7712, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.211340206185567, |
| "grad_norm": 0.570460855960846, |
| "learning_rate": 2e-05, |
| "loss": 0.7918, |
| "step": 2460 |
| }, |
| { |
| "epoch": 0.21219931271477663, |
| "grad_norm": 0.41472381353378296, |
| "learning_rate": 2e-05, |
| "loss": 0.7746, |
| "step": 2470 |
| }, |
| { |
| "epoch": 0.21305841924398625, |
| "grad_norm": 0.4356991648674011, |
| "learning_rate": 2e-05, |
| "loss": 0.793, |
| "step": 2480 |
| }, |
| { |
| "epoch": 0.21391752577319587, |
| "grad_norm": 0.425836443901062, |
| "learning_rate": 2e-05, |
| "loss": 0.7826, |
| "step": 2490 |
| }, |
| { |
| "epoch": 0.21477663230240548, |
| "grad_norm": 0.9032119512557983, |
| "learning_rate": 2e-05, |
| "loss": 0.7696, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.21563573883161513, |
| "grad_norm": 0.4030701518058777, |
| "learning_rate": 2e-05, |
| "loss": 0.7539, |
| "step": 2510 |
| }, |
| { |
| "epoch": 0.21649484536082475, |
| "grad_norm": 0.4129731357097626, |
| "learning_rate": 2e-05, |
| "loss": 0.7652, |
| "step": 2520 |
| }, |
| { |
| "epoch": 0.21735395189003437, |
| "grad_norm": 0.4277688264846802, |
| "learning_rate": 2e-05, |
| "loss": 0.7767, |
| "step": 2530 |
| }, |
| { |
| "epoch": 0.218213058419244, |
| "grad_norm": 0.40256595611572266, |
| "learning_rate": 2e-05, |
| "loss": 0.7498, |
| "step": 2540 |
| }, |
| { |
| "epoch": 0.2190721649484536, |
| "grad_norm": 0.39388591051101685, |
| "learning_rate": 2e-05, |
| "loss": 0.7449, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.21993127147766323, |
| "grad_norm": 0.41771769523620605, |
| "learning_rate": 2e-05, |
| "loss": 0.7658, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.22079037800687284, |
| "grad_norm": 0.43381044268608093, |
| "learning_rate": 2e-05, |
| "loss": 0.7379, |
| "step": 2570 |
| }, |
| { |
| "epoch": 0.22164948453608246, |
| "grad_norm": 0.4080195426940918, |
| "learning_rate": 2e-05, |
| "loss": 0.7623, |
| "step": 2580 |
| }, |
| { |
| "epoch": 0.22250859106529208, |
| "grad_norm": 0.4249197840690613, |
| "learning_rate": 2e-05, |
| "loss": 0.7439, |
| "step": 2590 |
| }, |
| { |
| "epoch": 0.22336769759450173, |
| "grad_norm": 0.4250815510749817, |
| "learning_rate": 2e-05, |
| "loss": 0.7824, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.22422680412371135, |
| "grad_norm": 0.4173683226108551, |
| "learning_rate": 2e-05, |
| "loss": 0.7402, |
| "step": 2610 |
| }, |
| { |
| "epoch": 0.22508591065292097, |
| "grad_norm": 0.4079398810863495, |
| "learning_rate": 2e-05, |
| "loss": 0.7959, |
| "step": 2620 |
| }, |
| { |
| "epoch": 0.2259450171821306, |
| "grad_norm": 0.4313492178916931, |
| "learning_rate": 2e-05, |
| "loss": 0.7741, |
| "step": 2630 |
| }, |
| { |
| "epoch": 0.2268041237113402, |
| "grad_norm": 0.3909526467323303, |
| "learning_rate": 2e-05, |
| "loss": 0.7483, |
| "step": 2640 |
| }, |
| { |
| "epoch": 0.22766323024054982, |
| "grad_norm": 0.4261670410633087, |
| "learning_rate": 2e-05, |
| "loss": 0.7814, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.22852233676975944, |
| "grad_norm": 0.4046727418899536, |
| "learning_rate": 2e-05, |
| "loss": 0.7525, |
| "step": 2660 |
| }, |
| { |
| "epoch": 0.22938144329896906, |
| "grad_norm": 0.4096175730228424, |
| "learning_rate": 2e-05, |
| "loss": 0.7568, |
| "step": 2670 |
| }, |
| { |
| "epoch": 0.23024054982817868, |
| "grad_norm": 0.39225271344184875, |
| "learning_rate": 2e-05, |
| "loss": 0.7691, |
| "step": 2680 |
| }, |
| { |
| "epoch": 0.23109965635738833, |
| "grad_norm": 0.45596063137054443, |
| "learning_rate": 2e-05, |
| "loss": 0.7696, |
| "step": 2690 |
| }, |
| { |
| "epoch": 0.23195876288659795, |
| "grad_norm": 0.3877202272415161, |
| "learning_rate": 2e-05, |
| "loss": 0.7551, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.23281786941580757, |
| "grad_norm": 0.4236013889312744, |
| "learning_rate": 2e-05, |
| "loss": 0.7838, |
| "step": 2710 |
| }, |
| { |
| "epoch": 0.23367697594501718, |
| "grad_norm": 0.3999956548213959, |
| "learning_rate": 2e-05, |
| "loss": 0.7712, |
| "step": 2720 |
| }, |
| { |
| "epoch": 0.2345360824742268, |
| "grad_norm": 0.47755834460258484, |
| "learning_rate": 2e-05, |
| "loss": 0.7712, |
| "step": 2730 |
| }, |
| { |
| "epoch": 0.23539518900343642, |
| "grad_norm": 0.4101397693157196, |
| "learning_rate": 2e-05, |
| "loss": 0.7715, |
| "step": 2740 |
| }, |
| { |
| "epoch": 0.23625429553264604, |
| "grad_norm": 0.42583993077278137, |
| "learning_rate": 2e-05, |
| "loss": 0.7851, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.23711340206185566, |
| "grad_norm": 0.4020129442214966, |
| "learning_rate": 2e-05, |
| "loss": 0.7394, |
| "step": 2760 |
| }, |
| { |
| "epoch": 0.23797250859106528, |
| "grad_norm": 0.4051017761230469, |
| "learning_rate": 2e-05, |
| "loss": 0.7705, |
| "step": 2770 |
| }, |
| { |
| "epoch": 0.23883161512027493, |
| "grad_norm": 0.4843144416809082, |
| "learning_rate": 2e-05, |
| "loss": 0.7836, |
| "step": 2780 |
| }, |
| { |
| "epoch": 0.23969072164948454, |
| "grad_norm": 0.4086487591266632, |
| "learning_rate": 2e-05, |
| "loss": 0.7942, |
| "step": 2790 |
| }, |
| { |
| "epoch": 0.24054982817869416, |
| "grad_norm": 0.49289804697036743, |
| "learning_rate": 2e-05, |
| "loss": 0.8043, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.24140893470790378, |
| "grad_norm": 0.4173983037471771, |
| "learning_rate": 2e-05, |
| "loss": 0.7655, |
| "step": 2810 |
| }, |
| { |
| "epoch": 0.2422680412371134, |
| "grad_norm": 0.42129677534103394, |
| "learning_rate": 2e-05, |
| "loss": 0.7828, |
| "step": 2820 |
| }, |
| { |
| "epoch": 0.24312714776632302, |
| "grad_norm": 0.3931340277194977, |
| "learning_rate": 2e-05, |
| "loss": 0.7744, |
| "step": 2830 |
| }, |
| { |
| "epoch": 0.24398625429553264, |
| "grad_norm": 0.4406070113182068, |
| "learning_rate": 2e-05, |
| "loss": 0.7553, |
| "step": 2840 |
| }, |
| { |
| "epoch": 0.24484536082474226, |
| "grad_norm": 0.42197829484939575, |
| "learning_rate": 2e-05, |
| "loss": 0.7825, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.24570446735395188, |
| "grad_norm": 0.4124753773212433, |
| "learning_rate": 2e-05, |
| "loss": 0.7562, |
| "step": 2860 |
| }, |
| { |
| "epoch": 0.24656357388316152, |
| "grad_norm": 0.4178576171398163, |
| "learning_rate": 2e-05, |
| "loss": 0.7618, |
| "step": 2870 |
| }, |
| { |
| "epoch": 0.24742268041237114, |
| "grad_norm": 0.4552730321884155, |
| "learning_rate": 2e-05, |
| "loss": 0.7935, |
| "step": 2880 |
| }, |
| { |
| "epoch": 0.24828178694158076, |
| "grad_norm": 0.42891925573349, |
| "learning_rate": 2e-05, |
| "loss": 0.7706, |
| "step": 2890 |
| }, |
| { |
| "epoch": 0.24914089347079038, |
| "grad_norm": 0.38573965430259705, |
| "learning_rate": 2e-05, |
| "loss": 0.7537, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.25, |
| "grad_norm": 0.4111515283584595, |
| "learning_rate": 2e-05, |
| "loss": 0.8234, |
| "step": 2910 |
| }, |
| { |
| "epoch": 0.2508591065292096, |
| "grad_norm": 0.4234291911125183, |
| "learning_rate": 2e-05, |
| "loss": 0.7947, |
| "step": 2920 |
| }, |
| { |
| "epoch": 0.25171821305841924, |
| "grad_norm": 0.4000537693500519, |
| "learning_rate": 2e-05, |
| "loss": 0.7626, |
| "step": 2930 |
| }, |
| { |
| "epoch": 0.25257731958762886, |
| "grad_norm": 0.4151047170162201, |
| "learning_rate": 2e-05, |
| "loss": 0.773, |
| "step": 2940 |
| }, |
| { |
| "epoch": 0.2534364261168385, |
| "grad_norm": 0.3972582221031189, |
| "learning_rate": 2e-05, |
| "loss": 0.7824, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.2542955326460481, |
| "grad_norm": 0.42536866664886475, |
| "learning_rate": 2e-05, |
| "loss": 0.7875, |
| "step": 2960 |
| }, |
| { |
| "epoch": 0.2551546391752577, |
| "grad_norm": 0.42411258816719055, |
| "learning_rate": 2e-05, |
| "loss": 0.7407, |
| "step": 2970 |
| }, |
| { |
| "epoch": 0.25601374570446733, |
| "grad_norm": 0.4318678677082062, |
| "learning_rate": 2e-05, |
| "loss": 0.7712, |
| "step": 2980 |
| }, |
| { |
| "epoch": 0.25687285223367695, |
| "grad_norm": 0.41649696230888367, |
| "learning_rate": 2e-05, |
| "loss": 0.7668, |
| "step": 2990 |
| }, |
| { |
| "epoch": 0.25773195876288657, |
| "grad_norm": 0.42106330394744873, |
| "learning_rate": 2e-05, |
| "loss": 0.7717, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.25859106529209624, |
| "grad_norm": 0.4055321216583252, |
| "learning_rate": 2e-05, |
| "loss": 0.7986, |
| "step": 3010 |
| }, |
| { |
| "epoch": 0.25945017182130586, |
| "grad_norm": 0.41959694027900696, |
| "learning_rate": 2e-05, |
| "loss": 0.7908, |
| "step": 3020 |
| }, |
| { |
| "epoch": 0.2603092783505155, |
| "grad_norm": 0.4296233654022217, |
| "learning_rate": 2e-05, |
| "loss": 0.7846, |
| "step": 3030 |
| }, |
| { |
| "epoch": 0.2611683848797251, |
| "grad_norm": 0.41830459237098694, |
| "learning_rate": 2e-05, |
| "loss": 0.8027, |
| "step": 3040 |
| }, |
| { |
| "epoch": 0.2620274914089347, |
| "grad_norm": 0.40822726488113403, |
| "learning_rate": 2e-05, |
| "loss": 0.7805, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.26288659793814434, |
| "grad_norm": 0.4108002483844757, |
| "learning_rate": 2e-05, |
| "loss": 0.7665, |
| "step": 3060 |
| }, |
| { |
| "epoch": 0.26374570446735396, |
| "grad_norm": 0.389371782541275, |
| "learning_rate": 2e-05, |
| "loss": 0.7397, |
| "step": 3070 |
| }, |
| { |
| "epoch": 0.2646048109965636, |
| "grad_norm": 0.42747220396995544, |
| "learning_rate": 2e-05, |
| "loss": 0.7874, |
| "step": 3080 |
| }, |
| { |
| "epoch": 0.2654639175257732, |
| "grad_norm": 0.4048026502132416, |
| "learning_rate": 2e-05, |
| "loss": 0.7751, |
| "step": 3090 |
| }, |
| { |
| "epoch": 0.2663230240549828, |
| "grad_norm": 0.5471147298812866, |
| "learning_rate": 2e-05, |
| "loss": 0.7873, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.26718213058419243, |
| "grad_norm": 0.38855239748954773, |
| "learning_rate": 2e-05, |
| "loss": 0.7339, |
| "step": 3110 |
| }, |
| { |
| "epoch": 0.26804123711340205, |
| "grad_norm": 0.4075400233268738, |
| "learning_rate": 2e-05, |
| "loss": 0.7489, |
| "step": 3120 |
| }, |
| { |
| "epoch": 0.26890034364261167, |
| "grad_norm": 0.4032638370990753, |
| "learning_rate": 2e-05, |
| "loss": 0.7778, |
| "step": 3130 |
| }, |
| { |
| "epoch": 0.2697594501718213, |
| "grad_norm": 0.41255396604537964, |
| "learning_rate": 2e-05, |
| "loss": 0.7492, |
| "step": 3140 |
| }, |
| { |
| "epoch": 0.2706185567010309, |
| "grad_norm": 0.3877742290496826, |
| "learning_rate": 2e-05, |
| "loss": 0.7523, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.27147766323024053, |
| "grad_norm": 0.3896322548389435, |
| "learning_rate": 2e-05, |
| "loss": 0.7774, |
| "step": 3160 |
| }, |
| { |
| "epoch": 0.27233676975945015, |
| "grad_norm": 0.40311911702156067, |
| "learning_rate": 2e-05, |
| "loss": 0.7681, |
| "step": 3170 |
| }, |
| { |
| "epoch": 0.27319587628865977, |
| "grad_norm": 0.3949439227581024, |
| "learning_rate": 2e-05, |
| "loss": 0.7683, |
| "step": 3180 |
| }, |
| { |
| "epoch": 0.27405498281786944, |
| "grad_norm": 0.4075079560279846, |
| "learning_rate": 2e-05, |
| "loss": 0.741, |
| "step": 3190 |
| }, |
| { |
| "epoch": 0.27491408934707906, |
| "grad_norm": 0.4325898289680481, |
| "learning_rate": 2e-05, |
| "loss": 0.7927, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.2757731958762887, |
| "grad_norm": 0.41164514422416687, |
| "learning_rate": 2e-05, |
| "loss": 0.7573, |
| "step": 3210 |
| }, |
| { |
| "epoch": 0.2766323024054983, |
| "grad_norm": 0.41085177659988403, |
| "learning_rate": 2e-05, |
| "loss": 0.7886, |
| "step": 3220 |
| }, |
| { |
| "epoch": 0.2774914089347079, |
| "grad_norm": 0.4105769097805023, |
| "learning_rate": 2e-05, |
| "loss": 0.7609, |
| "step": 3230 |
| }, |
| { |
| "epoch": 0.27835051546391754, |
| "grad_norm": 0.41889050602912903, |
| "learning_rate": 2e-05, |
| "loss": 0.7649, |
| "step": 3240 |
| }, |
| { |
| "epoch": 0.27920962199312716, |
| "grad_norm": 0.3920256197452545, |
| "learning_rate": 2e-05, |
| "loss": 0.747, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.2800687285223368, |
| "grad_norm": 0.4325742721557617, |
| "learning_rate": 2e-05, |
| "loss": 0.7725, |
| "step": 3260 |
| }, |
| { |
| "epoch": 0.2809278350515464, |
| "grad_norm": 0.42666274309158325, |
| "learning_rate": 2e-05, |
| "loss": 0.7847, |
| "step": 3270 |
| }, |
| { |
| "epoch": 0.281786941580756, |
| "grad_norm": 0.39878711104393005, |
| "learning_rate": 2e-05, |
| "loss": 0.7614, |
| "step": 3280 |
| }, |
| { |
| "epoch": 0.28264604810996563, |
| "grad_norm": 0.45834362506866455, |
| "learning_rate": 2e-05, |
| "loss": 0.7683, |
| "step": 3290 |
| }, |
| { |
| "epoch": 0.28350515463917525, |
| "grad_norm": 0.39607352018356323, |
| "learning_rate": 2e-05, |
| "loss": 0.7514, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.28436426116838487, |
| "grad_norm": 0.3845934271812439, |
| "learning_rate": 2e-05, |
| "loss": 0.7696, |
| "step": 3310 |
| }, |
| { |
| "epoch": 0.2852233676975945, |
| "grad_norm": 0.4206698536872864, |
| "learning_rate": 2e-05, |
| "loss": 0.768, |
| "step": 3320 |
| }, |
| { |
| "epoch": 0.2860824742268041, |
| "grad_norm": 0.3820017874240875, |
| "learning_rate": 2e-05, |
| "loss": 0.7598, |
| "step": 3330 |
| }, |
| { |
| "epoch": 0.2869415807560137, |
| "grad_norm": 0.3978671729564667, |
| "learning_rate": 2e-05, |
| "loss": 0.7423, |
| "step": 3340 |
| }, |
| { |
| "epoch": 0.28780068728522334, |
| "grad_norm": 0.4384324550628662, |
| "learning_rate": 2e-05, |
| "loss": 0.7646, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.28865979381443296, |
| "grad_norm": 0.4027288556098938, |
| "learning_rate": 2e-05, |
| "loss": 0.7585, |
| "step": 3360 |
| }, |
| { |
| "epoch": 0.28951890034364264, |
| "grad_norm": 0.4325219690799713, |
| "learning_rate": 2e-05, |
| "loss": 0.7806, |
| "step": 3370 |
| }, |
| { |
| "epoch": 0.29037800687285226, |
| "grad_norm": 0.4221478998661041, |
| "learning_rate": 2e-05, |
| "loss": 0.7597, |
| "step": 3380 |
| }, |
| { |
| "epoch": 0.2912371134020619, |
| "grad_norm": 0.44164422154426575, |
| "learning_rate": 2e-05, |
| "loss": 0.7693, |
| "step": 3390 |
| }, |
| { |
| "epoch": 0.2920962199312715, |
| "grad_norm": 0.38554859161376953, |
| "learning_rate": 2e-05, |
| "loss": 0.763, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.2929553264604811, |
| "grad_norm": 0.41126173734664917, |
| "learning_rate": 2e-05, |
| "loss": 0.75, |
| "step": 3410 |
| }, |
| { |
| "epoch": 0.29381443298969073, |
| "grad_norm": 0.42341408133506775, |
| "learning_rate": 2e-05, |
| "loss": 0.7742, |
| "step": 3420 |
| }, |
| { |
| "epoch": 0.29467353951890035, |
| "grad_norm": 0.45448067784309387, |
| "learning_rate": 2e-05, |
| "loss": 0.7734, |
| "step": 3430 |
| }, |
| { |
| "epoch": 0.29553264604810997, |
| "grad_norm": 0.41353639960289, |
| "learning_rate": 2e-05, |
| "loss": 0.7904, |
| "step": 3440 |
| }, |
| { |
| "epoch": 0.2963917525773196, |
| "grad_norm": 0.4858936667442322, |
| "learning_rate": 2e-05, |
| "loss": 0.7834, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.2972508591065292, |
| "grad_norm": 0.4161330461502075, |
| "learning_rate": 2e-05, |
| "loss": 0.7774, |
| "step": 3460 |
| }, |
| { |
| "epoch": 0.2981099656357388, |
| "grad_norm": 0.4159027338027954, |
| "learning_rate": 2e-05, |
| "loss": 0.7833, |
| "step": 3470 |
| }, |
| { |
| "epoch": 0.29896907216494845, |
| "grad_norm": 0.4200475513935089, |
| "learning_rate": 2e-05, |
| "loss": 0.7544, |
| "step": 3480 |
| }, |
| { |
| "epoch": 0.29982817869415807, |
| "grad_norm": 0.3868395984172821, |
| "learning_rate": 2e-05, |
| "loss": 0.7608, |
| "step": 3490 |
| }, |
| { |
| "epoch": 0.3006872852233677, |
| "grad_norm": 0.44515305757522583, |
| "learning_rate": 2e-05, |
| "loss": 0.8271, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.3015463917525773, |
| "grad_norm": 0.41017138957977295, |
| "learning_rate": 2e-05, |
| "loss": 0.7438, |
| "step": 3510 |
| }, |
| { |
| "epoch": 0.3024054982817869, |
| "grad_norm": 0.3779100179672241, |
| "learning_rate": 2e-05, |
| "loss": 0.7563, |
| "step": 3520 |
| }, |
| { |
| "epoch": 0.30326460481099654, |
| "grad_norm": 0.40496891736984253, |
| "learning_rate": 2e-05, |
| "loss": 0.781, |
| "step": 3530 |
| }, |
| { |
| "epoch": 0.30412371134020616, |
| "grad_norm": 0.382946640253067, |
| "learning_rate": 2e-05, |
| "loss": 0.7728, |
| "step": 3540 |
| }, |
| { |
| "epoch": 0.30498281786941583, |
| "grad_norm": 0.4126337468624115, |
| "learning_rate": 2e-05, |
| "loss": 0.7727, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.30584192439862545, |
| "grad_norm": 0.38314563035964966, |
| "learning_rate": 2e-05, |
| "loss": 0.7638, |
| "step": 3560 |
| }, |
| { |
| "epoch": 0.30670103092783507, |
| "grad_norm": 0.3995760679244995, |
| "learning_rate": 2e-05, |
| "loss": 0.795, |
| "step": 3570 |
| }, |
| { |
| "epoch": 0.3075601374570447, |
| "grad_norm": 0.40734052658081055, |
| "learning_rate": 2e-05, |
| "loss": 0.7333, |
| "step": 3580 |
| }, |
| { |
| "epoch": 0.3084192439862543, |
| "grad_norm": 0.41933971643447876, |
| "learning_rate": 2e-05, |
| "loss": 0.7641, |
| "step": 3590 |
| }, |
| { |
| "epoch": 0.30927835051546393, |
| "grad_norm": 0.4051254093647003, |
| "learning_rate": 2e-05, |
| "loss": 0.7494, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.31013745704467355, |
| "grad_norm": 0.40811339020729065, |
| "learning_rate": 2e-05, |
| "loss": 0.772, |
| "step": 3610 |
| }, |
| { |
| "epoch": 0.31099656357388317, |
| "grad_norm": 0.3880476653575897, |
| "learning_rate": 2e-05, |
| "loss": 0.7377, |
| "step": 3620 |
| }, |
| { |
| "epoch": 0.3118556701030928, |
| "grad_norm": 0.4227207601070404, |
| "learning_rate": 2e-05, |
| "loss": 0.7676, |
| "step": 3630 |
| }, |
| { |
| "epoch": 0.3127147766323024, |
| "grad_norm": 0.36887118220329285, |
| "learning_rate": 2e-05, |
| "loss": 0.7683, |
| "step": 3640 |
| }, |
| { |
| "epoch": 0.313573883161512, |
| "grad_norm": 0.3982653319835663, |
| "learning_rate": 2e-05, |
| "loss": 0.7394, |
| "step": 3650 |
| }, |
| { |
| "epoch": 0.31443298969072164, |
| "grad_norm": 0.4001864492893219, |
| "learning_rate": 2e-05, |
| "loss": 0.7578, |
| "step": 3660 |
| }, |
| { |
| "epoch": 0.31529209621993126, |
| "grad_norm": 0.39527979493141174, |
| "learning_rate": 2e-05, |
| "loss": 0.7644, |
| "step": 3670 |
| }, |
| { |
| "epoch": 0.3161512027491409, |
| "grad_norm": 0.4183419942855835, |
| "learning_rate": 2e-05, |
| "loss": 0.7808, |
| "step": 3680 |
| }, |
| { |
| "epoch": 0.3170103092783505, |
| "grad_norm": 0.41624853014945984, |
| "learning_rate": 2e-05, |
| "loss": 0.7962, |
| "step": 3690 |
| }, |
| { |
| "epoch": 0.3178694158075601, |
| "grad_norm": 0.3790307939052582, |
| "learning_rate": 2e-05, |
| "loss": 0.7506, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.31872852233676974, |
| "grad_norm": 0.3806716799736023, |
| "learning_rate": 2e-05, |
| "loss": 0.7644, |
| "step": 3710 |
| }, |
| { |
| "epoch": 0.31958762886597936, |
| "grad_norm": 0.39428409934043884, |
| "learning_rate": 2e-05, |
| "loss": 0.7638, |
| "step": 3720 |
| }, |
| { |
| "epoch": 0.32044673539518903, |
| "grad_norm": 0.43839022517204285, |
| "learning_rate": 2e-05, |
| "loss": 0.7521, |
| "step": 3730 |
| }, |
| { |
| "epoch": 0.32130584192439865, |
| "grad_norm": 0.40026065707206726, |
| "learning_rate": 2e-05, |
| "loss": 0.7707, |
| "step": 3740 |
| }, |
| { |
| "epoch": 0.32216494845360827, |
| "grad_norm": 0.4190294146537781, |
| "learning_rate": 2e-05, |
| "loss": 0.7841, |
| "step": 3750 |
| }, |
| { |
| "epoch": 0.3230240549828179, |
| "grad_norm": 0.38157919049263, |
| "learning_rate": 2e-05, |
| "loss": 0.7627, |
| "step": 3760 |
| }, |
| { |
| "epoch": 0.3238831615120275, |
| "grad_norm": 0.408637672662735, |
| "learning_rate": 2e-05, |
| "loss": 0.7676, |
| "step": 3770 |
| }, |
| { |
| "epoch": 0.3247422680412371, |
| "grad_norm": 0.4058610200881958, |
| "learning_rate": 2e-05, |
| "loss": 0.7519, |
| "step": 3780 |
| }, |
| { |
| "epoch": 0.32560137457044674, |
| "grad_norm": 0.47251054644584656, |
| "learning_rate": 2e-05, |
| "loss": 0.7542, |
| "step": 3790 |
| }, |
| { |
| "epoch": 0.32646048109965636, |
| "grad_norm": 0.4172389805316925, |
| "learning_rate": 2e-05, |
| "loss": 0.7818, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.327319587628866, |
| "grad_norm": 0.4114549160003662, |
| "learning_rate": 2e-05, |
| "loss": 0.7731, |
| "step": 3810 |
| }, |
| { |
| "epoch": 0.3281786941580756, |
| "grad_norm": 0.4147323966026306, |
| "learning_rate": 2e-05, |
| "loss": 0.8056, |
| "step": 3820 |
| }, |
| { |
| "epoch": 0.3290378006872852, |
| "grad_norm": 0.43751680850982666, |
| "learning_rate": 2e-05, |
| "loss": 0.7631, |
| "step": 3830 |
| }, |
| { |
| "epoch": 0.32989690721649484, |
| "grad_norm": 0.3997228741645813, |
| "learning_rate": 2e-05, |
| "loss": 0.7781, |
| "step": 3840 |
| }, |
| { |
| "epoch": 0.33075601374570446, |
| "grad_norm": 0.39175930619239807, |
| "learning_rate": 2e-05, |
| "loss": 0.7651, |
| "step": 3850 |
| }, |
| { |
| "epoch": 0.3316151202749141, |
| "grad_norm": 0.38528063893318176, |
| "learning_rate": 2e-05, |
| "loss": 0.7554, |
| "step": 3860 |
| }, |
| { |
| "epoch": 0.3324742268041237, |
| "grad_norm": 0.38999444246292114, |
| "learning_rate": 2e-05, |
| "loss": 0.7723, |
| "step": 3870 |
| }, |
| { |
| "epoch": 0.3333333333333333, |
| "grad_norm": 0.3951996862888336, |
| "learning_rate": 2e-05, |
| "loss": 0.7475, |
| "step": 3880 |
| }, |
| { |
| "epoch": 0.33419243986254293, |
| "grad_norm": 0.4000983238220215, |
| "learning_rate": 2e-05, |
| "loss": 0.7865, |
| "step": 3890 |
| }, |
| { |
| "epoch": 0.33505154639175255, |
| "grad_norm": 0.47464287281036377, |
| "learning_rate": 2e-05, |
| "loss": 0.7517, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.33591065292096217, |
| "grad_norm": 0.4179074764251709, |
| "learning_rate": 2e-05, |
| "loss": 0.7352, |
| "step": 3910 |
| }, |
| { |
| "epoch": 0.33676975945017185, |
| "grad_norm": 0.3948587477207184, |
| "learning_rate": 2e-05, |
| "loss": 0.7687, |
| "step": 3920 |
| }, |
| { |
| "epoch": 0.33762886597938147, |
| "grad_norm": 0.4247511923313141, |
| "learning_rate": 2e-05, |
| "loss": 0.7564, |
| "step": 3930 |
| }, |
| { |
| "epoch": 0.3384879725085911, |
| "grad_norm": 0.39262357354164124, |
| "learning_rate": 2e-05, |
| "loss": 0.7518, |
| "step": 3940 |
| }, |
| { |
| "epoch": 0.3393470790378007, |
| "grad_norm": 0.40266096591949463, |
| "learning_rate": 2e-05, |
| "loss": 0.7515, |
| "step": 3950 |
| }, |
| { |
| "epoch": 0.3402061855670103, |
| "grad_norm": 0.4001063108444214, |
| "learning_rate": 2e-05, |
| "loss": 0.7494, |
| "step": 3960 |
| }, |
| { |
| "epoch": 0.34106529209621994, |
| "grad_norm": 0.3866184651851654, |
| "learning_rate": 2e-05, |
| "loss": 0.7547, |
| "step": 3970 |
| }, |
| { |
| "epoch": 0.34192439862542956, |
| "grad_norm": 0.38115713000297546, |
| "learning_rate": 2e-05, |
| "loss": 0.7501, |
| "step": 3980 |
| }, |
| { |
| "epoch": 0.3427835051546392, |
| "grad_norm": 0.4157053232192993, |
| "learning_rate": 2e-05, |
| "loss": 0.759, |
| "step": 3990 |
| }, |
| { |
| "epoch": 0.3436426116838488, |
| "grad_norm": 0.40975072979927063, |
| "learning_rate": 2e-05, |
| "loss": 0.7584, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.3445017182130584, |
| "grad_norm": 0.40007585287094116, |
| "learning_rate": 2e-05, |
| "loss": 0.7475, |
| "step": 4010 |
| }, |
| { |
| "epoch": 0.34536082474226804, |
| "grad_norm": 0.43180954456329346, |
| "learning_rate": 2e-05, |
| "loss": 0.7479, |
| "step": 4020 |
| }, |
| { |
| "epoch": 0.34621993127147765, |
| "grad_norm": 0.42279163002967834, |
| "learning_rate": 2e-05, |
| "loss": 0.7318, |
| "step": 4030 |
| }, |
| { |
| "epoch": 0.3470790378006873, |
| "grad_norm": 0.3827601969242096, |
| "learning_rate": 2e-05, |
| "loss": 0.7591, |
| "step": 4040 |
| }, |
| { |
| "epoch": 0.3479381443298969, |
| "grad_norm": 0.4175439476966858, |
| "learning_rate": 2e-05, |
| "loss": 0.7446, |
| "step": 4050 |
| }, |
| { |
| "epoch": 0.3487972508591065, |
| "grad_norm": 0.40465012192726135, |
| "learning_rate": 2e-05, |
| "loss": 0.7611, |
| "step": 4060 |
| }, |
| { |
| "epoch": 0.34965635738831613, |
| "grad_norm": 0.4096953868865967, |
| "learning_rate": 2e-05, |
| "loss": 0.7598, |
| "step": 4070 |
| }, |
| { |
| "epoch": 0.35051546391752575, |
| "grad_norm": 0.4131430983543396, |
| "learning_rate": 2e-05, |
| "loss": 0.7869, |
| "step": 4080 |
| }, |
| { |
| "epoch": 0.35137457044673537, |
| "grad_norm": 0.4112975001335144, |
| "learning_rate": 2e-05, |
| "loss": 0.7799, |
| "step": 4090 |
| }, |
| { |
| "epoch": 0.35223367697594504, |
| "grad_norm": 0.4253504276275635, |
| "learning_rate": 2e-05, |
| "loss": 0.7339, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.35309278350515466, |
| "grad_norm": 0.4268980920314789, |
| "learning_rate": 2e-05, |
| "loss": 0.7537, |
| "step": 4110 |
| }, |
| { |
| "epoch": 0.3539518900343643, |
| "grad_norm": 0.38124239444732666, |
| "learning_rate": 2e-05, |
| "loss": 0.7729, |
| "step": 4120 |
| }, |
| { |
| "epoch": 0.3548109965635739, |
| "grad_norm": 0.37831252813339233, |
| "learning_rate": 2e-05, |
| "loss": 0.7397, |
| "step": 4130 |
| }, |
| { |
| "epoch": 0.3556701030927835, |
| "grad_norm": 0.41880711913108826, |
| "learning_rate": 2e-05, |
| "loss": 0.7757, |
| "step": 4140 |
| }, |
| { |
| "epoch": 0.35652920962199314, |
| "grad_norm": 0.43672290444374084, |
| "learning_rate": 2e-05, |
| "loss": 0.7425, |
| "step": 4150 |
| }, |
| { |
| "epoch": 0.35738831615120276, |
| "grad_norm": 0.4058953523635864, |
| "learning_rate": 2e-05, |
| "loss": 0.7309, |
| "step": 4160 |
| }, |
| { |
| "epoch": 0.3582474226804124, |
| "grad_norm": 0.39708343148231506, |
| "learning_rate": 2e-05, |
| "loss": 0.755, |
| "step": 4170 |
| }, |
| { |
| "epoch": 0.359106529209622, |
| "grad_norm": 0.40715470910072327, |
| "learning_rate": 2e-05, |
| "loss": 0.7618, |
| "step": 4180 |
| }, |
| { |
| "epoch": 0.3599656357388316, |
| "grad_norm": 0.37325555086135864, |
| "learning_rate": 2e-05, |
| "loss": 0.7381, |
| "step": 4190 |
| }, |
| { |
| "epoch": 0.36082474226804123, |
| "grad_norm": 0.49437960982322693, |
| "learning_rate": 2e-05, |
| "loss": 0.7475, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.36168384879725085, |
| "grad_norm": 0.40334492921829224, |
| "learning_rate": 2e-05, |
| "loss": 0.7775, |
| "step": 4210 |
| }, |
| { |
| "epoch": 0.36254295532646047, |
| "grad_norm": 0.42243167757987976, |
| "learning_rate": 2e-05, |
| "loss": 0.7548, |
| "step": 4220 |
| }, |
| { |
| "epoch": 0.3634020618556701, |
| "grad_norm": 0.4097069501876831, |
| "learning_rate": 2e-05, |
| "loss": 0.7514, |
| "step": 4230 |
| }, |
| { |
| "epoch": 0.3642611683848797, |
| "grad_norm": 0.40296828746795654, |
| "learning_rate": 2e-05, |
| "loss": 0.7689, |
| "step": 4240 |
| }, |
| { |
| "epoch": 0.3651202749140893, |
| "grad_norm": 0.4230409562587738, |
| "learning_rate": 2e-05, |
| "loss": 0.7513, |
| "step": 4250 |
| }, |
| { |
| "epoch": 0.36597938144329895, |
| "grad_norm": 0.4135167896747589, |
| "learning_rate": 2e-05, |
| "loss": 0.7525, |
| "step": 4260 |
| }, |
| { |
| "epoch": 0.36683848797250856, |
| "grad_norm": 0.38695311546325684, |
| "learning_rate": 2e-05, |
| "loss": 0.7546, |
| "step": 4270 |
| }, |
| { |
| "epoch": 0.36769759450171824, |
| "grad_norm": 0.4157143533229828, |
| "learning_rate": 2e-05, |
| "loss": 0.7656, |
| "step": 4280 |
| }, |
| { |
| "epoch": 0.36855670103092786, |
| "grad_norm": 0.4034111499786377, |
| "learning_rate": 2e-05, |
| "loss": 0.7428, |
| "step": 4290 |
| }, |
| { |
| "epoch": 0.3694158075601375, |
| "grad_norm": 0.42762601375579834, |
| "learning_rate": 2e-05, |
| "loss": 0.7734, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.3702749140893471, |
| "grad_norm": 0.38261547684669495, |
| "learning_rate": 2e-05, |
| "loss": 0.7536, |
| "step": 4310 |
| }, |
| { |
| "epoch": 0.3711340206185567, |
| "grad_norm": 0.559786319732666, |
| "learning_rate": 2e-05, |
| "loss": 0.7356, |
| "step": 4320 |
| }, |
| { |
| "epoch": 0.37199312714776633, |
| "grad_norm": 0.4105216860771179, |
| "learning_rate": 2e-05, |
| "loss": 0.7488, |
| "step": 4330 |
| }, |
| { |
| "epoch": 0.37285223367697595, |
| "grad_norm": 0.4059801995754242, |
| "learning_rate": 2e-05, |
| "loss": 0.7527, |
| "step": 4340 |
| }, |
| { |
| "epoch": 0.37371134020618557, |
| "grad_norm": 0.3986341953277588, |
| "learning_rate": 2e-05, |
| "loss": 0.7524, |
| "step": 4350 |
| }, |
| { |
| "epoch": 0.3745704467353952, |
| "grad_norm": 0.38521715998649597, |
| "learning_rate": 2e-05, |
| "loss": 0.7466, |
| "step": 4360 |
| }, |
| { |
| "epoch": 0.3754295532646048, |
| "grad_norm": 0.40247276425361633, |
| "learning_rate": 2e-05, |
| "loss": 0.7576, |
| "step": 4370 |
| }, |
| { |
| "epoch": 0.37628865979381443, |
| "grad_norm": 0.4033831059932709, |
| "learning_rate": 2e-05, |
| "loss": 0.753, |
| "step": 4380 |
| }, |
| { |
| "epoch": 0.37714776632302405, |
| "grad_norm": 0.3811618387699127, |
| "learning_rate": 2e-05, |
| "loss": 0.742, |
| "step": 4390 |
| }, |
| { |
| "epoch": 0.37800687285223367, |
| "grad_norm": 0.38687312602996826, |
| "learning_rate": 2e-05, |
| "loss": 0.7357, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.3788659793814433, |
| "grad_norm": 0.39860835671424866, |
| "learning_rate": 2e-05, |
| "loss": 0.771, |
| "step": 4410 |
| }, |
| { |
| "epoch": 0.3797250859106529, |
| "grad_norm": 0.3995939791202545, |
| "learning_rate": 2e-05, |
| "loss": 0.7678, |
| "step": 4420 |
| }, |
| { |
| "epoch": 0.3805841924398625, |
| "grad_norm": 0.390959769487381, |
| "learning_rate": 2e-05, |
| "loss": 0.7644, |
| "step": 4430 |
| }, |
| { |
| "epoch": 0.38144329896907214, |
| "grad_norm": 0.39426323771476746, |
| "learning_rate": 2e-05, |
| "loss": 0.7513, |
| "step": 4440 |
| }, |
| { |
| "epoch": 0.38230240549828176, |
| "grad_norm": 0.3981815278530121, |
| "learning_rate": 2e-05, |
| "loss": 0.7519, |
| "step": 4450 |
| }, |
| { |
| "epoch": 0.38316151202749144, |
| "grad_norm": 0.3980715870857239, |
| "learning_rate": 2e-05, |
| "loss": 0.7576, |
| "step": 4460 |
| }, |
| { |
| "epoch": 0.38402061855670105, |
| "grad_norm": 0.3965037167072296, |
| "learning_rate": 2e-05, |
| "loss": 0.7469, |
| "step": 4470 |
| }, |
| { |
| "epoch": 0.3848797250859107, |
| "grad_norm": 0.377102792263031, |
| "learning_rate": 2e-05, |
| "loss": 0.7681, |
| "step": 4480 |
| }, |
| { |
| "epoch": 0.3857388316151203, |
| "grad_norm": 0.40167808532714844, |
| "learning_rate": 2e-05, |
| "loss": 0.7422, |
| "step": 4490 |
| }, |
| { |
| "epoch": 0.3865979381443299, |
| "grad_norm": 0.3795527517795563, |
| "learning_rate": 2e-05, |
| "loss": 0.7476, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.38745704467353953, |
| "grad_norm": 0.39987045526504517, |
| "learning_rate": 2e-05, |
| "loss": 0.748, |
| "step": 4510 |
| }, |
| { |
| "epoch": 0.38831615120274915, |
| "grad_norm": 0.41090428829193115, |
| "learning_rate": 2e-05, |
| "loss": 0.7552, |
| "step": 4520 |
| }, |
| { |
| "epoch": 0.38917525773195877, |
| "grad_norm": 0.39889150857925415, |
| "learning_rate": 2e-05, |
| "loss": 0.7397, |
| "step": 4530 |
| }, |
| { |
| "epoch": 0.3900343642611684, |
| "grad_norm": 0.394302636384964, |
| "learning_rate": 2e-05, |
| "loss": 0.7651, |
| "step": 4540 |
| }, |
| { |
| "epoch": 0.390893470790378, |
| "grad_norm": 0.39259299635887146, |
| "learning_rate": 2e-05, |
| "loss": 0.7599, |
| "step": 4550 |
| }, |
| { |
| "epoch": 0.3917525773195876, |
| "grad_norm": 0.4073551297187805, |
| "learning_rate": 2e-05, |
| "loss": 0.758, |
| "step": 4560 |
| }, |
| { |
| "epoch": 0.39261168384879724, |
| "grad_norm": 0.4444946050643921, |
| "learning_rate": 2e-05, |
| "loss": 0.7486, |
| "step": 4570 |
| }, |
| { |
| "epoch": 0.39347079037800686, |
| "grad_norm": 0.40218037366867065, |
| "learning_rate": 2e-05, |
| "loss": 0.7365, |
| "step": 4580 |
| }, |
| { |
| "epoch": 0.3943298969072165, |
| "grad_norm": 0.4302692115306854, |
| "learning_rate": 2e-05, |
| "loss": 0.7506, |
| "step": 4590 |
| }, |
| { |
| "epoch": 0.3951890034364261, |
| "grad_norm": 0.41661903262138367, |
| "learning_rate": 2e-05, |
| "loss": 0.754, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.3960481099656357, |
| "grad_norm": 0.40314728021621704, |
| "learning_rate": 2e-05, |
| "loss": 0.7241, |
| "step": 4610 |
| }, |
| { |
| "epoch": 0.39690721649484534, |
| "grad_norm": 0.388095498085022, |
| "learning_rate": 2e-05, |
| "loss": 0.7596, |
| "step": 4620 |
| }, |
| { |
| "epoch": 0.39776632302405496, |
| "grad_norm": 0.3835248053073883, |
| "learning_rate": 2e-05, |
| "loss": 0.7593, |
| "step": 4630 |
| }, |
| { |
| "epoch": 0.39862542955326463, |
| "grad_norm": 0.38472631573677063, |
| "learning_rate": 2e-05, |
| "loss": 0.7353, |
| "step": 4640 |
| }, |
| { |
| "epoch": 0.39948453608247425, |
| "grad_norm": 0.3875240087509155, |
| "learning_rate": 2e-05, |
| "loss": 0.786, |
| "step": 4650 |
| }, |
| { |
| "epoch": 0.40034364261168387, |
| "grad_norm": 0.44080010056495667, |
| "learning_rate": 2e-05, |
| "loss": 0.7499, |
| "step": 4660 |
| }, |
| { |
| "epoch": 0.4012027491408935, |
| "grad_norm": 0.41625672578811646, |
| "learning_rate": 2e-05, |
| "loss": 0.7563, |
| "step": 4670 |
| }, |
| { |
| "epoch": 0.4020618556701031, |
| "grad_norm": 0.4100339412689209, |
| "learning_rate": 2e-05, |
| "loss": 0.7286, |
| "step": 4680 |
| }, |
| { |
| "epoch": 0.4029209621993127, |
| "grad_norm": 0.40571466088294983, |
| "learning_rate": 2e-05, |
| "loss": 0.7447, |
| "step": 4690 |
| }, |
| { |
| "epoch": 0.40378006872852235, |
| "grad_norm": 0.4143446087837219, |
| "learning_rate": 2e-05, |
| "loss": 0.7568, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.40463917525773196, |
| "grad_norm": 0.44597920775413513, |
| "learning_rate": 2e-05, |
| "loss": 0.7653, |
| "step": 4710 |
| }, |
| { |
| "epoch": 0.4054982817869416, |
| "grad_norm": 0.3949943780899048, |
| "learning_rate": 2e-05, |
| "loss": 0.73, |
| "step": 4720 |
| }, |
| { |
| "epoch": 0.4063573883161512, |
| "grad_norm": 0.38090425729751587, |
| "learning_rate": 2e-05, |
| "loss": 0.7606, |
| "step": 4730 |
| }, |
| { |
| "epoch": 0.4072164948453608, |
| "grad_norm": 0.4613902270793915, |
| "learning_rate": 2e-05, |
| "loss": 0.7535, |
| "step": 4740 |
| }, |
| { |
| "epoch": 0.40807560137457044, |
| "grad_norm": 0.4230399429798126, |
| "learning_rate": 2e-05, |
| "loss": 0.7563, |
| "step": 4750 |
| }, |
| { |
| "epoch": 0.40893470790378006, |
| "grad_norm": 0.38796481490135193, |
| "learning_rate": 2e-05, |
| "loss": 0.7077, |
| "step": 4760 |
| }, |
| { |
| "epoch": 0.4097938144329897, |
| "grad_norm": 0.4022347629070282, |
| "learning_rate": 2e-05, |
| "loss": 0.7531, |
| "step": 4770 |
| }, |
| { |
| "epoch": 0.4106529209621993, |
| "grad_norm": 0.4280019700527191, |
| "learning_rate": 2e-05, |
| "loss": 0.7438, |
| "step": 4780 |
| }, |
| { |
| "epoch": 0.4115120274914089, |
| "grad_norm": 0.39707037806510925, |
| "learning_rate": 2e-05, |
| "loss": 0.7514, |
| "step": 4790 |
| }, |
| { |
| "epoch": 0.41237113402061853, |
| "grad_norm": 0.39433586597442627, |
| "learning_rate": 2e-05, |
| "loss": 0.7213, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.41323024054982815, |
| "grad_norm": 0.42996808886528015, |
| "learning_rate": 2e-05, |
| "loss": 0.7466, |
| "step": 4810 |
| }, |
| { |
| "epoch": 0.41408934707903783, |
| "grad_norm": 0.40540334582328796, |
| "learning_rate": 2e-05, |
| "loss": 0.7525, |
| "step": 4820 |
| }, |
| { |
| "epoch": 0.41494845360824745, |
| "grad_norm": 0.4186977744102478, |
| "learning_rate": 2e-05, |
| "loss": 0.7855, |
| "step": 4830 |
| }, |
| { |
| "epoch": 0.41580756013745707, |
| "grad_norm": 0.4322900176048279, |
| "learning_rate": 2e-05, |
| "loss": 0.7456, |
| "step": 4840 |
| }, |
| { |
| "epoch": 0.4166666666666667, |
| "grad_norm": 0.572149395942688, |
| "learning_rate": 2e-05, |
| "loss": 0.7579, |
| "step": 4850 |
| }, |
| { |
| "epoch": 0.4175257731958763, |
| "grad_norm": 0.37685662508010864, |
| "learning_rate": 2e-05, |
| "loss": 0.7379, |
| "step": 4860 |
| }, |
| { |
| "epoch": 0.4183848797250859, |
| "grad_norm": 0.46816933155059814, |
| "learning_rate": 2e-05, |
| "loss": 0.7356, |
| "step": 4870 |
| }, |
| { |
| "epoch": 0.41924398625429554, |
| "grad_norm": 0.4115220308303833, |
| "learning_rate": 2e-05, |
| "loss": 0.7605, |
| "step": 4880 |
| }, |
| { |
| "epoch": 0.42010309278350516, |
| "grad_norm": 0.40049082040786743, |
| "learning_rate": 2e-05, |
| "loss": 0.7376, |
| "step": 4890 |
| }, |
| { |
| "epoch": 0.4209621993127148, |
| "grad_norm": 0.4084942936897278, |
| "learning_rate": 2e-05, |
| "loss": 0.787, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.4218213058419244, |
| "grad_norm": 0.40400949120521545, |
| "learning_rate": 2e-05, |
| "loss": 0.7686, |
| "step": 4910 |
| }, |
| { |
| "epoch": 0.422680412371134, |
| "grad_norm": 0.3788411021232605, |
| "learning_rate": 2e-05, |
| "loss": 0.7396, |
| "step": 4920 |
| }, |
| { |
| "epoch": 0.42353951890034364, |
| "grad_norm": 0.4002997577190399, |
| "learning_rate": 2e-05, |
| "loss": 0.7565, |
| "step": 4930 |
| }, |
| { |
| "epoch": 0.42439862542955326, |
| "grad_norm": 0.42359012365341187, |
| "learning_rate": 2e-05, |
| "loss": 0.7486, |
| "step": 4940 |
| }, |
| { |
| "epoch": 0.4252577319587629, |
| "grad_norm": 0.4005122184753418, |
| "learning_rate": 2e-05, |
| "loss": 0.7424, |
| "step": 4950 |
| }, |
| { |
| "epoch": 0.4261168384879725, |
| "grad_norm": 0.40040847659111023, |
| "learning_rate": 2e-05, |
| "loss": 0.7431, |
| "step": 4960 |
| }, |
| { |
| "epoch": 0.4269759450171821, |
| "grad_norm": 0.3970852196216583, |
| "learning_rate": 2e-05, |
| "loss": 0.7524, |
| "step": 4970 |
| }, |
| { |
| "epoch": 0.42783505154639173, |
| "grad_norm": 0.4006688892841339, |
| "learning_rate": 2e-05, |
| "loss": 0.7419, |
| "step": 4980 |
| }, |
| { |
| "epoch": 0.42869415807560135, |
| "grad_norm": 0.3915325999259949, |
| "learning_rate": 2e-05, |
| "loss": 0.7512, |
| "step": 4990 |
| }, |
| { |
| "epoch": 0.42955326460481097, |
| "grad_norm": 0.43504029512405396, |
| "learning_rate": 2e-05, |
| "loss": 0.7499, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.43041237113402064, |
| "grad_norm": 0.4006860852241516, |
| "learning_rate": 2e-05, |
| "loss": 0.7722, |
| "step": 5010 |
| }, |
| { |
| "epoch": 0.43127147766323026, |
| "grad_norm": 0.42165443301200867, |
| "learning_rate": 2e-05, |
| "loss": 0.7694, |
| "step": 5020 |
| }, |
| { |
| "epoch": 0.4321305841924399, |
| "grad_norm": 0.39642852544784546, |
| "learning_rate": 2e-05, |
| "loss": 0.7544, |
| "step": 5030 |
| }, |
| { |
| "epoch": 0.4329896907216495, |
| "grad_norm": 0.3671112656593323, |
| "learning_rate": 2e-05, |
| "loss": 0.7534, |
| "step": 5040 |
| }, |
| { |
| "epoch": 0.4338487972508591, |
| "grad_norm": 0.40421435236930847, |
| "learning_rate": 2e-05, |
| "loss": 0.7248, |
| "step": 5050 |
| }, |
| { |
| "epoch": 0.43470790378006874, |
| "grad_norm": 0.3971313536167145, |
| "learning_rate": 2e-05, |
| "loss": 0.795, |
| "step": 5060 |
| }, |
| { |
| "epoch": 0.43556701030927836, |
| "grad_norm": 0.407399445772171, |
| "learning_rate": 2e-05, |
| "loss": 0.7299, |
| "step": 5070 |
| }, |
| { |
| "epoch": 0.436426116838488, |
| "grad_norm": 0.39452338218688965, |
| "learning_rate": 2e-05, |
| "loss": 0.7574, |
| "step": 5080 |
| }, |
| { |
| "epoch": 0.4372852233676976, |
| "grad_norm": 0.4038243889808655, |
| "learning_rate": 2e-05, |
| "loss": 0.7799, |
| "step": 5090 |
| }, |
| { |
| "epoch": 0.4381443298969072, |
| "grad_norm": 0.3823649287223816, |
| "learning_rate": 2e-05, |
| "loss": 0.7309, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.43900343642611683, |
| "grad_norm": 0.38972368836402893, |
| "learning_rate": 2e-05, |
| "loss": 0.7321, |
| "step": 5110 |
| }, |
| { |
| "epoch": 0.43986254295532645, |
| "grad_norm": 0.4127161502838135, |
| "learning_rate": 2e-05, |
| "loss": 0.7457, |
| "step": 5120 |
| }, |
| { |
| "epoch": 0.44072164948453607, |
| "grad_norm": 0.41461309790611267, |
| "learning_rate": 2e-05, |
| "loss": 0.7511, |
| "step": 5130 |
| }, |
| { |
| "epoch": 0.4415807560137457, |
| "grad_norm": 0.37044236063957214, |
| "learning_rate": 2e-05, |
| "loss": 0.73, |
| "step": 5140 |
| }, |
| { |
| "epoch": 0.4424398625429553, |
| "grad_norm": 0.4120967984199524, |
| "learning_rate": 2e-05, |
| "loss": 0.7581, |
| "step": 5150 |
| }, |
| { |
| "epoch": 0.44329896907216493, |
| "grad_norm": 0.40208619832992554, |
| "learning_rate": 2e-05, |
| "loss": 0.758, |
| "step": 5160 |
| }, |
| { |
| "epoch": 0.44415807560137455, |
| "grad_norm": 0.3970443904399872, |
| "learning_rate": 2e-05, |
| "loss": 0.7444, |
| "step": 5170 |
| }, |
| { |
| "epoch": 0.44501718213058417, |
| "grad_norm": 0.3957621157169342, |
| "learning_rate": 2e-05, |
| "loss": 0.7481, |
| "step": 5180 |
| }, |
| { |
| "epoch": 0.44587628865979384, |
| "grad_norm": 0.4279804527759552, |
| "learning_rate": 2e-05, |
| "loss": 0.7477, |
| "step": 5190 |
| }, |
| { |
| "epoch": 0.44673539518900346, |
| "grad_norm": 0.39507153630256653, |
| "learning_rate": 2e-05, |
| "loss": 0.7339, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.4475945017182131, |
| "grad_norm": 0.384280264377594, |
| "learning_rate": 2e-05, |
| "loss": 0.7243, |
| "step": 5210 |
| }, |
| { |
| "epoch": 0.4484536082474227, |
| "grad_norm": 0.38282859325408936, |
| "learning_rate": 2e-05, |
| "loss": 0.7664, |
| "step": 5220 |
| }, |
| { |
| "epoch": 0.4493127147766323, |
| "grad_norm": 0.4108511209487915, |
| "learning_rate": 2e-05, |
| "loss": 0.7354, |
| "step": 5230 |
| }, |
| { |
| "epoch": 0.45017182130584193, |
| "grad_norm": 0.38676372170448303, |
| "learning_rate": 2e-05, |
| "loss": 0.77, |
| "step": 5240 |
| }, |
| { |
| "epoch": 0.45103092783505155, |
| "grad_norm": 0.38403913378715515, |
| "learning_rate": 2e-05, |
| "loss": 0.757, |
| "step": 5250 |
| }, |
| { |
| "epoch": 0.4518900343642612, |
| "grad_norm": 0.41679251194000244, |
| "learning_rate": 2e-05, |
| "loss": 0.7527, |
| "step": 5260 |
| }, |
| { |
| "epoch": 0.4527491408934708, |
| "grad_norm": 0.38778433203697205, |
| "learning_rate": 2e-05, |
| "loss": 0.7092, |
| "step": 5270 |
| }, |
| { |
| "epoch": 0.4536082474226804, |
| "grad_norm": 0.3975287973880768, |
| "learning_rate": 2e-05, |
| "loss": 0.7357, |
| "step": 5280 |
| }, |
| { |
| "epoch": 0.45446735395189003, |
| "grad_norm": 0.3857387900352478, |
| "learning_rate": 2e-05, |
| "loss": 0.7403, |
| "step": 5290 |
| }, |
| { |
| "epoch": 0.45532646048109965, |
| "grad_norm": 0.37999626994132996, |
| "learning_rate": 2e-05, |
| "loss": 0.7525, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.45618556701030927, |
| "grad_norm": 0.37595221400260925, |
| "learning_rate": 2e-05, |
| "loss": 0.749, |
| "step": 5310 |
| }, |
| { |
| "epoch": 0.4570446735395189, |
| "grad_norm": 0.3842063248157501, |
| "learning_rate": 2e-05, |
| "loss": 0.7491, |
| "step": 5320 |
| }, |
| { |
| "epoch": 0.4579037800687285, |
| "grad_norm": 1.773705005645752, |
| "learning_rate": 2e-05, |
| "loss": 0.7663, |
| "step": 5330 |
| }, |
| { |
| "epoch": 0.4587628865979381, |
| "grad_norm": 0.3880942463874817, |
| "learning_rate": 2e-05, |
| "loss": 0.7464, |
| "step": 5340 |
| }, |
| { |
| "epoch": 0.45962199312714774, |
| "grad_norm": 0.40811359882354736, |
| "learning_rate": 2e-05, |
| "loss": 0.7292, |
| "step": 5350 |
| }, |
| { |
| "epoch": 0.46048109965635736, |
| "grad_norm": 0.36452025175094604, |
| "learning_rate": 2e-05, |
| "loss": 0.7588, |
| "step": 5360 |
| }, |
| { |
| "epoch": 0.46134020618556704, |
| "grad_norm": 0.3871554732322693, |
| "learning_rate": 2e-05, |
| "loss": 0.7423, |
| "step": 5370 |
| }, |
| { |
| "epoch": 0.46219931271477666, |
| "grad_norm": 0.4043223559856415, |
| "learning_rate": 2e-05, |
| "loss": 0.7387, |
| "step": 5380 |
| }, |
| { |
| "epoch": 0.4630584192439863, |
| "grad_norm": 0.3838101625442505, |
| "learning_rate": 2e-05, |
| "loss": 0.7361, |
| "step": 5390 |
| }, |
| { |
| "epoch": 0.4639175257731959, |
| "grad_norm": 0.4027150869369507, |
| "learning_rate": 2e-05, |
| "loss": 0.7435, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.4647766323024055, |
| "grad_norm": 0.40335702896118164, |
| "learning_rate": 2e-05, |
| "loss": 0.7566, |
| "step": 5410 |
| }, |
| { |
| "epoch": 0.46563573883161513, |
| "grad_norm": 0.3775824308395386, |
| "learning_rate": 2e-05, |
| "loss": 0.7376, |
| "step": 5420 |
| }, |
| { |
| "epoch": 0.46649484536082475, |
| "grad_norm": 0.378854900598526, |
| "learning_rate": 2e-05, |
| "loss": 0.741, |
| "step": 5430 |
| }, |
| { |
| "epoch": 0.46735395189003437, |
| "grad_norm": 0.382784366607666, |
| "learning_rate": 2e-05, |
| "loss": 0.7546, |
| "step": 5440 |
| }, |
| { |
| "epoch": 0.468213058419244, |
| "grad_norm": 0.41972804069519043, |
| "learning_rate": 2e-05, |
| "loss": 0.7331, |
| "step": 5450 |
| }, |
| { |
| "epoch": 0.4690721649484536, |
| "grad_norm": 0.4076957106590271, |
| "learning_rate": 2e-05, |
| "loss": 0.753, |
| "step": 5460 |
| }, |
| { |
| "epoch": 0.4699312714776632, |
| "grad_norm": 0.38812845945358276, |
| "learning_rate": 2e-05, |
| "loss": 0.7328, |
| "step": 5470 |
| }, |
| { |
| "epoch": 0.47079037800687284, |
| "grad_norm": 0.3751866817474365, |
| "learning_rate": 2e-05, |
| "loss": 0.7625, |
| "step": 5480 |
| }, |
| { |
| "epoch": 0.47164948453608246, |
| "grad_norm": 0.39034783840179443, |
| "learning_rate": 2e-05, |
| "loss": 0.7416, |
| "step": 5490 |
| }, |
| { |
| "epoch": 0.4725085910652921, |
| "grad_norm": 0.372744619846344, |
| "learning_rate": 2e-05, |
| "loss": 0.7314, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.4733676975945017, |
| "grad_norm": 0.43096768856048584, |
| "learning_rate": 2e-05, |
| "loss": 0.7506, |
| "step": 5510 |
| }, |
| { |
| "epoch": 0.4742268041237113, |
| "grad_norm": 0.3995991349220276, |
| "learning_rate": 2e-05, |
| "loss": 0.7764, |
| "step": 5520 |
| }, |
| { |
| "epoch": 0.47508591065292094, |
| "grad_norm": 0.4010429084300995, |
| "learning_rate": 2e-05, |
| "loss": 0.7609, |
| "step": 5530 |
| }, |
| { |
| "epoch": 0.47594501718213056, |
| "grad_norm": 0.4783953130245209, |
| "learning_rate": 2e-05, |
| "loss": 0.7457, |
| "step": 5540 |
| }, |
| { |
| "epoch": 0.47680412371134023, |
| "grad_norm": 0.391328364610672, |
| "learning_rate": 2e-05, |
| "loss": 0.7708, |
| "step": 5550 |
| }, |
| { |
| "epoch": 0.47766323024054985, |
| "grad_norm": 0.4106771945953369, |
| "learning_rate": 2e-05, |
| "loss": 0.7385, |
| "step": 5560 |
| }, |
| { |
| "epoch": 0.47852233676975947, |
| "grad_norm": 0.3909521698951721, |
| "learning_rate": 2e-05, |
| "loss": 0.7279, |
| "step": 5570 |
| }, |
| { |
| "epoch": 0.4793814432989691, |
| "grad_norm": 0.3801710903644562, |
| "learning_rate": 2e-05, |
| "loss": 0.7066, |
| "step": 5580 |
| }, |
| { |
| "epoch": 0.4802405498281787, |
| "grad_norm": 0.4061938524246216, |
| "learning_rate": 2e-05, |
| "loss": 0.7403, |
| "step": 5590 |
| }, |
| { |
| "epoch": 0.48109965635738833, |
| "grad_norm": 0.40357154607772827, |
| "learning_rate": 2e-05, |
| "loss": 0.7556, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.48195876288659795, |
| "grad_norm": 0.35953155159950256, |
| "learning_rate": 2e-05, |
| "loss": 0.7394, |
| "step": 5610 |
| }, |
| { |
| "epoch": 0.48281786941580757, |
| "grad_norm": 0.38126692175865173, |
| "learning_rate": 2e-05, |
| "loss": 0.7351, |
| "step": 5620 |
| }, |
| { |
| "epoch": 0.4836769759450172, |
| "grad_norm": 0.3793662488460541, |
| "learning_rate": 2e-05, |
| "loss": 0.7592, |
| "step": 5630 |
| }, |
| { |
| "epoch": 0.4845360824742268, |
| "grad_norm": 0.3970396816730499, |
| "learning_rate": 2e-05, |
| "loss": 0.761, |
| "step": 5640 |
| }, |
| { |
| "epoch": 0.4853951890034364, |
| "grad_norm": 0.39472565054893494, |
| "learning_rate": 2e-05, |
| "loss": 0.733, |
| "step": 5650 |
| }, |
| { |
| "epoch": 0.48625429553264604, |
| "grad_norm": 0.3791460394859314, |
| "learning_rate": 2e-05, |
| "loss": 0.7356, |
| "step": 5660 |
| }, |
| { |
| "epoch": 0.48711340206185566, |
| "grad_norm": 0.40617644786834717, |
| "learning_rate": 2e-05, |
| "loss": 0.7358, |
| "step": 5670 |
| }, |
| { |
| "epoch": 0.4879725085910653, |
| "grad_norm": 0.4667048156261444, |
| "learning_rate": 2e-05, |
| "loss": 0.7082, |
| "step": 5680 |
| }, |
| { |
| "epoch": 0.4888316151202749, |
| "grad_norm": 0.3840990960597992, |
| "learning_rate": 2e-05, |
| "loss": 0.7581, |
| "step": 5690 |
| }, |
| { |
| "epoch": 0.4896907216494845, |
| "grad_norm": 0.40064701437950134, |
| "learning_rate": 2e-05, |
| "loss": 0.7427, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.49054982817869414, |
| "grad_norm": 0.4124237298965454, |
| "learning_rate": 2e-05, |
| "loss": 0.7175, |
| "step": 5710 |
| }, |
| { |
| "epoch": 0.49140893470790376, |
| "grad_norm": 0.3909745514392853, |
| "learning_rate": 2e-05, |
| "loss": 0.7517, |
| "step": 5720 |
| }, |
| { |
| "epoch": 0.49226804123711343, |
| "grad_norm": 0.3688066303730011, |
| "learning_rate": 2e-05, |
| "loss": 0.7226, |
| "step": 5730 |
| }, |
| { |
| "epoch": 0.49312714776632305, |
| "grad_norm": 0.4286164343357086, |
| "learning_rate": 2e-05, |
| "loss": 0.7287, |
| "step": 5740 |
| }, |
| { |
| "epoch": 0.49398625429553267, |
| "grad_norm": 0.37897518277168274, |
| "learning_rate": 2e-05, |
| "loss": 0.7346, |
| "step": 5750 |
| }, |
| { |
| "epoch": 0.4948453608247423, |
| "grad_norm": 0.4024289548397064, |
| "learning_rate": 2e-05, |
| "loss": 0.7531, |
| "step": 5760 |
| }, |
| { |
| "epoch": 0.4957044673539519, |
| "grad_norm": 0.39345040917396545, |
| "learning_rate": 2e-05, |
| "loss": 0.7242, |
| "step": 5770 |
| }, |
| { |
| "epoch": 0.4965635738831615, |
| "grad_norm": 0.4561581611633301, |
| "learning_rate": 2e-05, |
| "loss": 0.7411, |
| "step": 5780 |
| }, |
| { |
| "epoch": 0.49742268041237114, |
| "grad_norm": 0.3749309182167053, |
| "learning_rate": 2e-05, |
| "loss": 0.7469, |
| "step": 5790 |
| }, |
| { |
| "epoch": 0.49828178694158076, |
| "grad_norm": 0.4012472629547119, |
| "learning_rate": 2e-05, |
| "loss": 0.7099, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.4991408934707904, |
| "grad_norm": 0.3969298005104065, |
| "learning_rate": 2e-05, |
| "loss": 0.7455, |
| "step": 5810 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 0.4022328555583954, |
| "learning_rate": 2e-05, |
| "loss": 0.7283, |
| "step": 5820 |
| }, |
| { |
| "epoch": 0.5008591065292096, |
| "grad_norm": 0.36776313185691833, |
| "learning_rate": 2e-05, |
| "loss": 0.7448, |
| "step": 5830 |
| }, |
| { |
| "epoch": 0.5017182130584192, |
| "grad_norm": 0.4025474190711975, |
| "learning_rate": 2e-05, |
| "loss": 0.7296, |
| "step": 5840 |
| }, |
| { |
| "epoch": 0.5025773195876289, |
| "grad_norm": 0.4022384285926819, |
| "learning_rate": 2e-05, |
| "loss": 0.7522, |
| "step": 5850 |
| }, |
| { |
| "epoch": 0.5034364261168385, |
| "grad_norm": 0.41585469245910645, |
| "learning_rate": 2e-05, |
| "loss": 0.7423, |
| "step": 5860 |
| }, |
| { |
| "epoch": 0.5042955326460481, |
| "grad_norm": 0.3819616734981537, |
| "learning_rate": 2e-05, |
| "loss": 0.7691, |
| "step": 5870 |
| }, |
| { |
| "epoch": 0.5051546391752577, |
| "grad_norm": 0.3753138780593872, |
| "learning_rate": 2e-05, |
| "loss": 0.7678, |
| "step": 5880 |
| }, |
| { |
| "epoch": 0.5060137457044673, |
| "grad_norm": 0.39819473028182983, |
| "learning_rate": 2e-05, |
| "loss": 0.7341, |
| "step": 5890 |
| }, |
| { |
| "epoch": 0.506872852233677, |
| "grad_norm": 0.3874221444129944, |
| "learning_rate": 2e-05, |
| "loss": 0.7498, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.5077319587628866, |
| "grad_norm": 0.41702935099601746, |
| "learning_rate": 2e-05, |
| "loss": 0.7442, |
| "step": 5910 |
| }, |
| { |
| "epoch": 0.5085910652920962, |
| "grad_norm": 0.4022543430328369, |
| "learning_rate": 2e-05, |
| "loss": 0.7363, |
| "step": 5920 |
| }, |
| { |
| "epoch": 0.5094501718213058, |
| "grad_norm": 0.401775598526001, |
| "learning_rate": 2e-05, |
| "loss": 0.7527, |
| "step": 5930 |
| }, |
| { |
| "epoch": 0.5103092783505154, |
| "grad_norm": 0.4923630952835083, |
| "learning_rate": 2e-05, |
| "loss": 0.7521, |
| "step": 5940 |
| }, |
| { |
| "epoch": 0.511168384879725, |
| "grad_norm": 0.3755284249782562, |
| "learning_rate": 2e-05, |
| "loss": 0.7314, |
| "step": 5950 |
| }, |
| { |
| "epoch": 0.5120274914089347, |
| "grad_norm": 0.41290706396102905, |
| "learning_rate": 2e-05, |
| "loss": 0.7362, |
| "step": 5960 |
| }, |
| { |
| "epoch": 0.5128865979381443, |
| "grad_norm": 0.38221514225006104, |
| "learning_rate": 2e-05, |
| "loss": 0.739, |
| "step": 5970 |
| }, |
| { |
| "epoch": 0.5137457044673539, |
| "grad_norm": 0.41157266497612, |
| "learning_rate": 2e-05, |
| "loss": 0.7339, |
| "step": 5980 |
| }, |
| { |
| "epoch": 0.5146048109965635, |
| "grad_norm": 0.3943363130092621, |
| "learning_rate": 2e-05, |
| "loss": 0.7628, |
| "step": 5990 |
| }, |
| { |
| "epoch": 0.5154639175257731, |
| "grad_norm": 0.4046788215637207, |
| "learning_rate": 2e-05, |
| "loss": 0.7279, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.5163230240549829, |
| "grad_norm": 0.3751663863658905, |
| "learning_rate": 2e-05, |
| "loss": 0.7474, |
| "step": 6010 |
| }, |
| { |
| "epoch": 0.5171821305841925, |
| "grad_norm": 0.37326687574386597, |
| "learning_rate": 2e-05, |
| "loss": 0.7735, |
| "step": 6020 |
| }, |
| { |
| "epoch": 0.5180412371134021, |
| "grad_norm": 0.4058487117290497, |
| "learning_rate": 2e-05, |
| "loss": 0.7402, |
| "step": 6030 |
| }, |
| { |
| "epoch": 0.5189003436426117, |
| "grad_norm": 0.38418149948120117, |
| "learning_rate": 2e-05, |
| "loss": 0.7447, |
| "step": 6040 |
| }, |
| { |
| "epoch": 0.5197594501718213, |
| "grad_norm": 0.39697104692459106, |
| "learning_rate": 2e-05, |
| "loss": 0.7961, |
| "step": 6050 |
| }, |
| { |
| "epoch": 0.520618556701031, |
| "grad_norm": 0.4206184446811676, |
| "learning_rate": 2e-05, |
| "loss": 0.7417, |
| "step": 6060 |
| }, |
| { |
| "epoch": 0.5214776632302406, |
| "grad_norm": 0.4147883951663971, |
| "learning_rate": 2e-05, |
| "loss": 0.7192, |
| "step": 6070 |
| }, |
| { |
| "epoch": 0.5223367697594502, |
| "grad_norm": 0.39864054322242737, |
| "learning_rate": 2e-05, |
| "loss": 0.7595, |
| "step": 6080 |
| }, |
| { |
| "epoch": 0.5231958762886598, |
| "grad_norm": 0.40809568762779236, |
| "learning_rate": 2e-05, |
| "loss": 0.7566, |
| "step": 6090 |
| }, |
| { |
| "epoch": 0.5240549828178694, |
| "grad_norm": 0.3927410840988159, |
| "learning_rate": 2e-05, |
| "loss": 0.7625, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.5249140893470791, |
| "grad_norm": 0.41640281677246094, |
| "learning_rate": 2e-05, |
| "loss": 0.7513, |
| "step": 6110 |
| }, |
| { |
| "epoch": 0.5257731958762887, |
| "grad_norm": 0.36995160579681396, |
| "learning_rate": 2e-05, |
| "loss": 0.7569, |
| "step": 6120 |
| }, |
| { |
| "epoch": 0.5266323024054983, |
| "grad_norm": 0.41234222054481506, |
| "learning_rate": 2e-05, |
| "loss": 0.7565, |
| "step": 6130 |
| }, |
| { |
| "epoch": 0.5274914089347079, |
| "grad_norm": 0.38027313351631165, |
| "learning_rate": 2e-05, |
| "loss": 0.7611, |
| "step": 6140 |
| }, |
| { |
| "epoch": 0.5283505154639175, |
| "grad_norm": 0.44180914759635925, |
| "learning_rate": 2e-05, |
| "loss": 0.7353, |
| "step": 6150 |
| }, |
| { |
| "epoch": 0.5292096219931272, |
| "grad_norm": 0.4217710793018341, |
| "learning_rate": 2e-05, |
| "loss": 0.7364, |
| "step": 6160 |
| }, |
| { |
| "epoch": 0.5300687285223368, |
| "grad_norm": 0.37701427936553955, |
| "learning_rate": 2e-05, |
| "loss": 0.7369, |
| "step": 6170 |
| }, |
| { |
| "epoch": 0.5309278350515464, |
| "grad_norm": 0.4394442141056061, |
| "learning_rate": 2e-05, |
| "loss": 0.7023, |
| "step": 6180 |
| }, |
| { |
| "epoch": 0.531786941580756, |
| "grad_norm": 0.3926869034767151, |
| "learning_rate": 2e-05, |
| "loss": 0.7612, |
| "step": 6190 |
| }, |
| { |
| "epoch": 0.5326460481099656, |
| "grad_norm": 0.3746460974216461, |
| "learning_rate": 2e-05, |
| "loss": 0.7555, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.5335051546391752, |
| "grad_norm": 0.40384641289711, |
| "learning_rate": 2e-05, |
| "loss": 0.7436, |
| "step": 6210 |
| }, |
| { |
| "epoch": 0.5343642611683849, |
| "grad_norm": 0.37056082487106323, |
| "learning_rate": 2e-05, |
| "loss": 0.7482, |
| "step": 6220 |
| }, |
| { |
| "epoch": 0.5352233676975945, |
| "grad_norm": 0.4089955687522888, |
| "learning_rate": 2e-05, |
| "loss": 0.7451, |
| "step": 6230 |
| }, |
| { |
| "epoch": 0.5360824742268041, |
| "grad_norm": 0.3872886598110199, |
| "learning_rate": 2e-05, |
| "loss": 0.7254, |
| "step": 6240 |
| }, |
| { |
| "epoch": 0.5369415807560137, |
| "grad_norm": 0.3914802074432373, |
| "learning_rate": 2e-05, |
| "loss": 0.7292, |
| "step": 6250 |
| }, |
| { |
| "epoch": 0.5378006872852233, |
| "grad_norm": 0.769835889339447, |
| "learning_rate": 2e-05, |
| "loss": 0.7186, |
| "step": 6260 |
| }, |
| { |
| "epoch": 0.538659793814433, |
| "grad_norm": 0.39080721139907837, |
| "learning_rate": 2e-05, |
| "loss": 0.7381, |
| "step": 6270 |
| }, |
| { |
| "epoch": 0.5395189003436426, |
| "grad_norm": 0.3812302052974701, |
| "learning_rate": 2e-05, |
| "loss": 0.739, |
| "step": 6280 |
| }, |
| { |
| "epoch": 0.5403780068728522, |
| "grad_norm": 0.38450372219085693, |
| "learning_rate": 2e-05, |
| "loss": 0.7174, |
| "step": 6290 |
| }, |
| { |
| "epoch": 0.5412371134020618, |
| "grad_norm": 0.41678953170776367, |
| "learning_rate": 2e-05, |
| "loss": 0.7299, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.5420962199312714, |
| "grad_norm": 0.4016650915145874, |
| "learning_rate": 2e-05, |
| "loss": 0.7422, |
| "step": 6310 |
| }, |
| { |
| "epoch": 0.5429553264604811, |
| "grad_norm": 0.5218998789787292, |
| "learning_rate": 2e-05, |
| "loss": 0.7403, |
| "step": 6320 |
| }, |
| { |
| "epoch": 0.5438144329896907, |
| "grad_norm": 0.7454551458358765, |
| "learning_rate": 2e-05, |
| "loss": 0.7382, |
| "step": 6330 |
| }, |
| { |
| "epoch": 0.5446735395189003, |
| "grad_norm": 0.3671092391014099, |
| "learning_rate": 2e-05, |
| "loss": 0.7227, |
| "step": 6340 |
| }, |
| { |
| "epoch": 0.5455326460481099, |
| "grad_norm": 0.39646652340888977, |
| "learning_rate": 2e-05, |
| "loss": 0.7398, |
| "step": 6350 |
| }, |
| { |
| "epoch": 0.5463917525773195, |
| "grad_norm": 0.3983212411403656, |
| "learning_rate": 2e-05, |
| "loss": 0.7235, |
| "step": 6360 |
| }, |
| { |
| "epoch": 0.5472508591065293, |
| "grad_norm": 0.3650527000427246, |
| "learning_rate": 2e-05, |
| "loss": 0.7124, |
| "step": 6370 |
| }, |
| { |
| "epoch": 0.5481099656357389, |
| "grad_norm": 0.4242032766342163, |
| "learning_rate": 2e-05, |
| "loss": 0.7428, |
| "step": 6380 |
| }, |
| { |
| "epoch": 0.5489690721649485, |
| "grad_norm": 0.3994031250476837, |
| "learning_rate": 2e-05, |
| "loss": 0.7378, |
| "step": 6390 |
| }, |
| { |
| "epoch": 0.5498281786941581, |
| "grad_norm": 0.42586684226989746, |
| "learning_rate": 2e-05, |
| "loss": 0.7231, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.5506872852233677, |
| "grad_norm": 0.3687209486961365, |
| "learning_rate": 2e-05, |
| "loss": 0.7459, |
| "step": 6410 |
| }, |
| { |
| "epoch": 0.5515463917525774, |
| "grad_norm": 0.40230828523635864, |
| "learning_rate": 2e-05, |
| "loss": 0.7391, |
| "step": 6420 |
| }, |
| { |
| "epoch": 0.552405498281787, |
| "grad_norm": 0.3987503945827484, |
| "learning_rate": 2e-05, |
| "loss": 0.72, |
| "step": 6430 |
| }, |
| { |
| "epoch": 0.5532646048109966, |
| "grad_norm": 0.40480706095695496, |
| "learning_rate": 2e-05, |
| "loss": 0.7244, |
| "step": 6440 |
| }, |
| { |
| "epoch": 0.5541237113402062, |
| "grad_norm": 0.3924468457698822, |
| "learning_rate": 2e-05, |
| "loss": 0.7617, |
| "step": 6450 |
| }, |
| { |
| "epoch": 0.5549828178694158, |
| "grad_norm": 0.35725754499435425, |
| "learning_rate": 2e-05, |
| "loss": 0.7363, |
| "step": 6460 |
| }, |
| { |
| "epoch": 0.5558419243986255, |
| "grad_norm": 0.4191160500049591, |
| "learning_rate": 2e-05, |
| "loss": 0.7657, |
| "step": 6470 |
| }, |
| { |
| "epoch": 0.5567010309278351, |
| "grad_norm": 0.38524308800697327, |
| "learning_rate": 2e-05, |
| "loss": 0.7133, |
| "step": 6480 |
| }, |
| { |
| "epoch": 0.5575601374570447, |
| "grad_norm": 0.3856598436832428, |
| "learning_rate": 2e-05, |
| "loss": 0.7546, |
| "step": 6490 |
| }, |
| { |
| "epoch": 0.5584192439862543, |
| "grad_norm": 0.41385617852211, |
| "learning_rate": 2e-05, |
| "loss": 0.7519, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.5592783505154639, |
| "grad_norm": 0.3790697455406189, |
| "learning_rate": 2e-05, |
| "loss": 0.7295, |
| "step": 6510 |
| }, |
| { |
| "epoch": 0.5601374570446735, |
| "grad_norm": 0.41983598470687866, |
| "learning_rate": 2e-05, |
| "loss": 0.7348, |
| "step": 6520 |
| }, |
| { |
| "epoch": 0.5609965635738832, |
| "grad_norm": 0.36119964718818665, |
| "learning_rate": 2e-05, |
| "loss": 0.7308, |
| "step": 6530 |
| }, |
| { |
| "epoch": 0.5618556701030928, |
| "grad_norm": 0.38680288195610046, |
| "learning_rate": 2e-05, |
| "loss": 0.7339, |
| "step": 6540 |
| }, |
| { |
| "epoch": 0.5627147766323024, |
| "grad_norm": 0.4145539700984955, |
| "learning_rate": 2e-05, |
| "loss": 0.7091, |
| "step": 6550 |
| }, |
| { |
| "epoch": 0.563573883161512, |
| "grad_norm": 0.3693397045135498, |
| "learning_rate": 2e-05, |
| "loss": 0.7098, |
| "step": 6560 |
| }, |
| { |
| "epoch": 0.5644329896907216, |
| "grad_norm": 0.3949684500694275, |
| "learning_rate": 2e-05, |
| "loss": 0.7416, |
| "step": 6570 |
| }, |
| { |
| "epoch": 0.5652920962199313, |
| "grad_norm": 0.396666020154953, |
| "learning_rate": 2e-05, |
| "loss": 0.7716, |
| "step": 6580 |
| }, |
| { |
| "epoch": 0.5661512027491409, |
| "grad_norm": 0.4408302307128906, |
| "learning_rate": 2e-05, |
| "loss": 0.7492, |
| "step": 6590 |
| }, |
| { |
| "epoch": 0.5670103092783505, |
| "grad_norm": 0.3889305889606476, |
| "learning_rate": 2e-05, |
| "loss": 0.775, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.5678694158075601, |
| "grad_norm": 0.400215208530426, |
| "learning_rate": 2e-05, |
| "loss": 0.7451, |
| "step": 6610 |
| }, |
| { |
| "epoch": 0.5687285223367697, |
| "grad_norm": 0.41504040360450745, |
| "learning_rate": 2e-05, |
| "loss": 0.7547, |
| "step": 6620 |
| }, |
| { |
| "epoch": 0.5695876288659794, |
| "grad_norm": 0.4264198839664459, |
| "learning_rate": 2e-05, |
| "loss": 0.7512, |
| "step": 6630 |
| }, |
| { |
| "epoch": 0.570446735395189, |
| "grad_norm": 0.3995652496814728, |
| "learning_rate": 2e-05, |
| "loss": 0.6935, |
| "step": 6640 |
| }, |
| { |
| "epoch": 0.5713058419243986, |
| "grad_norm": 0.42268258333206177, |
| "learning_rate": 2e-05, |
| "loss": 0.7336, |
| "step": 6650 |
| }, |
| { |
| "epoch": 0.5721649484536082, |
| "grad_norm": 0.38238558173179626, |
| "learning_rate": 2e-05, |
| "loss": 0.7337, |
| "step": 6660 |
| }, |
| { |
| "epoch": 0.5730240549828178, |
| "grad_norm": 0.3691173195838928, |
| "learning_rate": 2e-05, |
| "loss": 0.7196, |
| "step": 6670 |
| }, |
| { |
| "epoch": 0.5738831615120275, |
| "grad_norm": 0.3902357518672943, |
| "learning_rate": 2e-05, |
| "loss": 0.7063, |
| "step": 6680 |
| }, |
| { |
| "epoch": 0.5747422680412371, |
| "grad_norm": 0.3873693346977234, |
| "learning_rate": 2e-05, |
| "loss": 0.7489, |
| "step": 6690 |
| }, |
| { |
| "epoch": 0.5756013745704467, |
| "grad_norm": 0.39792609214782715, |
| "learning_rate": 2e-05, |
| "loss": 0.7353, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.5764604810996563, |
| "grad_norm": 0.40667009353637695, |
| "learning_rate": 2e-05, |
| "loss": 0.7751, |
| "step": 6710 |
| }, |
| { |
| "epoch": 0.5773195876288659, |
| "grad_norm": 0.40490615367889404, |
| "learning_rate": 2e-05, |
| "loss": 0.7612, |
| "step": 6720 |
| }, |
| { |
| "epoch": 0.5781786941580757, |
| "grad_norm": 0.4047369360923767, |
| "learning_rate": 2e-05, |
| "loss": 0.7129, |
| "step": 6730 |
| }, |
| { |
| "epoch": 0.5790378006872853, |
| "grad_norm": 0.40595370531082153, |
| "learning_rate": 2e-05, |
| "loss": 0.7379, |
| "step": 6740 |
| }, |
| { |
| "epoch": 0.5798969072164949, |
| "grad_norm": 0.4153591990470886, |
| "learning_rate": 2e-05, |
| "loss": 0.7214, |
| "step": 6750 |
| }, |
| { |
| "epoch": 0.5807560137457045, |
| "grad_norm": 0.40520018339157104, |
| "learning_rate": 2e-05, |
| "loss": 0.7485, |
| "step": 6760 |
| }, |
| { |
| "epoch": 0.5816151202749141, |
| "grad_norm": 0.3709152638912201, |
| "learning_rate": 2e-05, |
| "loss": 0.7327, |
| "step": 6770 |
| }, |
| { |
| "epoch": 0.5824742268041238, |
| "grad_norm": 0.40015506744384766, |
| "learning_rate": 2e-05, |
| "loss": 0.7182, |
| "step": 6780 |
| }, |
| { |
| "epoch": 0.5833333333333334, |
| "grad_norm": 0.4174875319004059, |
| "learning_rate": 2e-05, |
| "loss": 0.7237, |
| "step": 6790 |
| }, |
| { |
| "epoch": 0.584192439862543, |
| "grad_norm": 0.38705530762672424, |
| "learning_rate": 2e-05, |
| "loss": 0.716, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.5850515463917526, |
| "grad_norm": 0.4316568076610565, |
| "learning_rate": 2e-05, |
| "loss": 0.7257, |
| "step": 6810 |
| }, |
| { |
| "epoch": 0.5859106529209622, |
| "grad_norm": 0.38438260555267334, |
| "learning_rate": 2e-05, |
| "loss": 0.7118, |
| "step": 6820 |
| }, |
| { |
| "epoch": 0.5867697594501718, |
| "grad_norm": 0.4032870829105377, |
| "learning_rate": 2e-05, |
| "loss": 0.7407, |
| "step": 6830 |
| }, |
| { |
| "epoch": 0.5876288659793815, |
| "grad_norm": 0.38428691029548645, |
| "learning_rate": 2e-05, |
| "loss": 0.7312, |
| "step": 6840 |
| }, |
| { |
| "epoch": 0.5884879725085911, |
| "grad_norm": 0.4077114760875702, |
| "learning_rate": 2e-05, |
| "loss": 0.7478, |
| "step": 6850 |
| }, |
| { |
| "epoch": 0.5893470790378007, |
| "grad_norm": 0.4109855890274048, |
| "learning_rate": 2e-05, |
| "loss": 0.7447, |
| "step": 6860 |
| }, |
| { |
| "epoch": 0.5902061855670103, |
| "grad_norm": 0.41067856550216675, |
| "learning_rate": 2e-05, |
| "loss": 0.7443, |
| "step": 6870 |
| }, |
| { |
| "epoch": 0.5910652920962199, |
| "grad_norm": 0.4081782400608063, |
| "learning_rate": 2e-05, |
| "loss": 0.7123, |
| "step": 6880 |
| }, |
| { |
| "epoch": 0.5919243986254296, |
| "grad_norm": 0.4068760573863983, |
| "learning_rate": 2e-05, |
| "loss": 0.7316, |
| "step": 6890 |
| }, |
| { |
| "epoch": 0.5927835051546392, |
| "grad_norm": 0.4102189242839813, |
| "learning_rate": 2e-05, |
| "loss": 0.718, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.5936426116838488, |
| "grad_norm": 0.3934837877750397, |
| "learning_rate": 2e-05, |
| "loss": 0.7291, |
| "step": 6910 |
| }, |
| { |
| "epoch": 0.5945017182130584, |
| "grad_norm": 0.4376651644706726, |
| "learning_rate": 2e-05, |
| "loss": 0.7354, |
| "step": 6920 |
| }, |
| { |
| "epoch": 0.595360824742268, |
| "grad_norm": 0.39079388976097107, |
| "learning_rate": 2e-05, |
| "loss": 0.7029, |
| "step": 6930 |
| }, |
| { |
| "epoch": 0.5962199312714777, |
| "grad_norm": 0.3903258442878723, |
| "learning_rate": 2e-05, |
| "loss": 0.7269, |
| "step": 6940 |
| }, |
| { |
| "epoch": 0.5970790378006873, |
| "grad_norm": 0.39610376954078674, |
| "learning_rate": 2e-05, |
| "loss": 0.7565, |
| "step": 6950 |
| }, |
| { |
| "epoch": 0.5979381443298969, |
| "grad_norm": 0.39760294556617737, |
| "learning_rate": 2e-05, |
| "loss": 0.7412, |
| "step": 6960 |
| }, |
| { |
| "epoch": 0.5987972508591065, |
| "grad_norm": 0.3585233688354492, |
| "learning_rate": 2e-05, |
| "loss": 0.7295, |
| "step": 6970 |
| }, |
| { |
| "epoch": 0.5996563573883161, |
| "grad_norm": 0.36380884051322937, |
| "learning_rate": 2e-05, |
| "loss": 0.7186, |
| "step": 6980 |
| }, |
| { |
| "epoch": 0.6005154639175257, |
| "grad_norm": 0.39174193143844604, |
| "learning_rate": 2e-05, |
| "loss": 0.7357, |
| "step": 6990 |
| }, |
| { |
| "epoch": 0.6013745704467354, |
| "grad_norm": 0.39799588918685913, |
| "learning_rate": 2e-05, |
| "loss": 0.7371, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.602233676975945, |
| "grad_norm": 0.36917436122894287, |
| "learning_rate": 2e-05, |
| "loss": 0.746, |
| "step": 7010 |
| }, |
| { |
| "epoch": 0.6030927835051546, |
| "grad_norm": 0.42522454261779785, |
| "learning_rate": 2e-05, |
| "loss": 0.7262, |
| "step": 7020 |
| }, |
| { |
| "epoch": 0.6039518900343642, |
| "grad_norm": 0.40242502093315125, |
| "learning_rate": 2e-05, |
| "loss": 0.7664, |
| "step": 7030 |
| }, |
| { |
| "epoch": 0.6048109965635738, |
| "grad_norm": 0.4184524118900299, |
| "learning_rate": 2e-05, |
| "loss": 0.7394, |
| "step": 7040 |
| }, |
| { |
| "epoch": 0.6056701030927835, |
| "grad_norm": 0.3913715183734894, |
| "learning_rate": 2e-05, |
| "loss": 0.7463, |
| "step": 7050 |
| }, |
| { |
| "epoch": 0.6065292096219931, |
| "grad_norm": 0.3786400258541107, |
| "learning_rate": 2e-05, |
| "loss": 0.7385, |
| "step": 7060 |
| }, |
| { |
| "epoch": 0.6073883161512027, |
| "grad_norm": 0.38487502932548523, |
| "learning_rate": 2e-05, |
| "loss": 0.7764, |
| "step": 7070 |
| }, |
| { |
| "epoch": 0.6082474226804123, |
| "grad_norm": 0.36665862798690796, |
| "learning_rate": 2e-05, |
| "loss": 0.7374, |
| "step": 7080 |
| }, |
| { |
| "epoch": 0.6091065292096219, |
| "grad_norm": 0.37223145365715027, |
| "learning_rate": 2e-05, |
| "loss": 0.7158, |
| "step": 7090 |
| }, |
| { |
| "epoch": 0.6099656357388317, |
| "grad_norm": 0.4014417827129364, |
| "learning_rate": 2e-05, |
| "loss": 0.7488, |
| "step": 7100 |
| }, |
| { |
| "epoch": 0.6108247422680413, |
| "grad_norm": 0.3741561472415924, |
| "learning_rate": 2e-05, |
| "loss": 0.7219, |
| "step": 7110 |
| }, |
| { |
| "epoch": 0.6116838487972509, |
| "grad_norm": 0.3873651623725891, |
| "learning_rate": 2e-05, |
| "loss": 0.7224, |
| "step": 7120 |
| }, |
| { |
| "epoch": 0.6125429553264605, |
| "grad_norm": 0.4114565849304199, |
| "learning_rate": 2e-05, |
| "loss": 0.7128, |
| "step": 7130 |
| }, |
| { |
| "epoch": 0.6134020618556701, |
| "grad_norm": 0.37538430094718933, |
| "learning_rate": 2e-05, |
| "loss": 0.7374, |
| "step": 7140 |
| }, |
| { |
| "epoch": 0.6142611683848798, |
| "grad_norm": 0.40238508582115173, |
| "learning_rate": 2e-05, |
| "loss": 0.7271, |
| "step": 7150 |
| }, |
| { |
| "epoch": 0.6151202749140894, |
| "grad_norm": 0.40835902094841003, |
| "learning_rate": 2e-05, |
| "loss": 0.7446, |
| "step": 7160 |
| }, |
| { |
| "epoch": 0.615979381443299, |
| "grad_norm": 0.38308578729629517, |
| "learning_rate": 2e-05, |
| "loss": 0.7321, |
| "step": 7170 |
| }, |
| { |
| "epoch": 0.6168384879725086, |
| "grad_norm": 0.3902871906757355, |
| "learning_rate": 2e-05, |
| "loss": 0.693, |
| "step": 7180 |
| }, |
| { |
| "epoch": 0.6176975945017182, |
| "grad_norm": 0.389384925365448, |
| "learning_rate": 2e-05, |
| "loss": 0.716, |
| "step": 7190 |
| }, |
| { |
| "epoch": 0.6185567010309279, |
| "grad_norm": 0.3872620463371277, |
| "learning_rate": 2e-05, |
| "loss": 0.7214, |
| "step": 7200 |
| }, |
| { |
| "epoch": 0.6194158075601375, |
| "grad_norm": 0.419077068567276, |
| "learning_rate": 2e-05, |
| "loss": 0.7421, |
| "step": 7210 |
| }, |
| { |
| "epoch": 0.6202749140893471, |
| "grad_norm": 0.36774230003356934, |
| "learning_rate": 2e-05, |
| "loss": 0.7319, |
| "step": 7220 |
| }, |
| { |
| "epoch": 0.6211340206185567, |
| "grad_norm": 0.3957841098308563, |
| "learning_rate": 2e-05, |
| "loss": 0.7191, |
| "step": 7230 |
| }, |
| { |
| "epoch": 0.6219931271477663, |
| "grad_norm": 0.39523938298225403, |
| "learning_rate": 2e-05, |
| "loss": 0.6992, |
| "step": 7240 |
| }, |
| { |
| "epoch": 0.622852233676976, |
| "grad_norm": 0.38072100281715393, |
| "learning_rate": 2e-05, |
| "loss": 0.746, |
| "step": 7250 |
| }, |
| { |
| "epoch": 0.6237113402061856, |
| "grad_norm": 0.3953589200973511, |
| "learning_rate": 2e-05, |
| "loss": 0.7453, |
| "step": 7260 |
| }, |
| { |
| "epoch": 0.6245704467353952, |
| "grad_norm": 0.39391908049583435, |
| "learning_rate": 2e-05, |
| "loss": 0.7346, |
| "step": 7270 |
| }, |
| { |
| "epoch": 0.6254295532646048, |
| "grad_norm": 0.38824668526649475, |
| "learning_rate": 2e-05, |
| "loss": 0.7088, |
| "step": 7280 |
| }, |
| { |
| "epoch": 0.6262886597938144, |
| "grad_norm": 0.3857765793800354, |
| "learning_rate": 2e-05, |
| "loss": 0.7424, |
| "step": 7290 |
| }, |
| { |
| "epoch": 0.627147766323024, |
| "grad_norm": 0.3899436295032501, |
| "learning_rate": 2e-05, |
| "loss": 0.7152, |
| "step": 7300 |
| }, |
| { |
| "epoch": 0.6280068728522337, |
| "grad_norm": 0.3713277578353882, |
| "learning_rate": 2e-05, |
| "loss": 0.708, |
| "step": 7310 |
| }, |
| { |
| "epoch": 0.6288659793814433, |
| "grad_norm": 0.39456284046173096, |
| "learning_rate": 2e-05, |
| "loss": 0.7407, |
| "step": 7320 |
| }, |
| { |
| "epoch": 0.6297250859106529, |
| "grad_norm": 0.3972882628440857, |
| "learning_rate": 2e-05, |
| "loss": 0.719, |
| "step": 7330 |
| }, |
| { |
| "epoch": 0.6305841924398625, |
| "grad_norm": 0.34951624274253845, |
| "learning_rate": 2e-05, |
| "loss": 0.7327, |
| "step": 7340 |
| }, |
| { |
| "epoch": 0.6314432989690721, |
| "grad_norm": 0.4226081073284149, |
| "learning_rate": 2e-05, |
| "loss": 0.7198, |
| "step": 7350 |
| }, |
| { |
| "epoch": 0.6323024054982818, |
| "grad_norm": 0.40421581268310547, |
| "learning_rate": 2e-05, |
| "loss": 0.7528, |
| "step": 7360 |
| }, |
| { |
| "epoch": 0.6331615120274914, |
| "grad_norm": 0.4154883325099945, |
| "learning_rate": 2e-05, |
| "loss": 0.7258, |
| "step": 7370 |
| }, |
| { |
| "epoch": 0.634020618556701, |
| "grad_norm": 0.38089707493782043, |
| "learning_rate": 2e-05, |
| "loss": 0.744, |
| "step": 7380 |
| }, |
| { |
| "epoch": 0.6348797250859106, |
| "grad_norm": 0.39381060004234314, |
| "learning_rate": 2e-05, |
| "loss": 0.7281, |
| "step": 7390 |
| }, |
| { |
| "epoch": 0.6357388316151202, |
| "grad_norm": 0.39000505208969116, |
| "learning_rate": 2e-05, |
| "loss": 0.7148, |
| "step": 7400 |
| }, |
| { |
| "epoch": 0.6365979381443299, |
| "grad_norm": 0.3838815689086914, |
| "learning_rate": 2e-05, |
| "loss": 0.7122, |
| "step": 7410 |
| }, |
| { |
| "epoch": 0.6374570446735395, |
| "grad_norm": 0.393691748380661, |
| "learning_rate": 2e-05, |
| "loss": 0.7387, |
| "step": 7420 |
| }, |
| { |
| "epoch": 0.6383161512027491, |
| "grad_norm": 0.3641030192375183, |
| "learning_rate": 2e-05, |
| "loss": 0.739, |
| "step": 7430 |
| }, |
| { |
| "epoch": 0.6391752577319587, |
| "grad_norm": 0.36873501539230347, |
| "learning_rate": 2e-05, |
| "loss": 0.6965, |
| "step": 7440 |
| }, |
| { |
| "epoch": 0.6400343642611683, |
| "grad_norm": 0.40716084837913513, |
| "learning_rate": 2e-05, |
| "loss": 0.7215, |
| "step": 7450 |
| }, |
| { |
| "epoch": 0.6408934707903781, |
| "grad_norm": 0.41122761368751526, |
| "learning_rate": 2e-05, |
| "loss": 0.739, |
| "step": 7460 |
| }, |
| { |
| "epoch": 0.6417525773195877, |
| "grad_norm": 0.3731881082057953, |
| "learning_rate": 2e-05, |
| "loss": 0.7169, |
| "step": 7470 |
| }, |
| { |
| "epoch": 0.6426116838487973, |
| "grad_norm": 0.39967629313468933, |
| "learning_rate": 2e-05, |
| "loss": 0.7311, |
| "step": 7480 |
| }, |
| { |
| "epoch": 0.6434707903780069, |
| "grad_norm": 0.3921056091785431, |
| "learning_rate": 2e-05, |
| "loss": 0.7284, |
| "step": 7490 |
| }, |
| { |
| "epoch": 0.6443298969072165, |
| "grad_norm": 0.37582868337631226, |
| "learning_rate": 2e-05, |
| "loss": 0.7454, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.6451890034364262, |
| "grad_norm": 0.3824794888496399, |
| "learning_rate": 2e-05, |
| "loss": 0.7568, |
| "step": 7510 |
| }, |
| { |
| "epoch": 0.6460481099656358, |
| "grad_norm": 0.38574734330177307, |
| "learning_rate": 2e-05, |
| "loss": 0.714, |
| "step": 7520 |
| }, |
| { |
| "epoch": 0.6469072164948454, |
| "grad_norm": 0.39589419960975647, |
| "learning_rate": 2e-05, |
| "loss": 0.7365, |
| "step": 7530 |
| }, |
| { |
| "epoch": 0.647766323024055, |
| "grad_norm": 0.3802694082260132, |
| "learning_rate": 2e-05, |
| "loss": 0.7393, |
| "step": 7540 |
| }, |
| { |
| "epoch": 0.6486254295532646, |
| "grad_norm": 0.4521954357624054, |
| "learning_rate": 2e-05, |
| "loss": 0.7383, |
| "step": 7550 |
| }, |
| { |
| "epoch": 0.6494845360824743, |
| "grad_norm": 0.4696226119995117, |
| "learning_rate": 2e-05, |
| "loss": 0.7271, |
| "step": 7560 |
| }, |
| { |
| "epoch": 0.6503436426116839, |
| "grad_norm": 0.40010038018226624, |
| "learning_rate": 2e-05, |
| "loss": 0.7485, |
| "step": 7570 |
| }, |
| { |
| "epoch": 0.6512027491408935, |
| "grad_norm": 0.36703062057495117, |
| "learning_rate": 2e-05, |
| "loss": 0.7162, |
| "step": 7580 |
| }, |
| { |
| "epoch": 0.6520618556701031, |
| "grad_norm": 0.4171064794063568, |
| "learning_rate": 2e-05, |
| "loss": 0.7308, |
| "step": 7590 |
| }, |
| { |
| "epoch": 0.6529209621993127, |
| "grad_norm": 0.3983151912689209, |
| "learning_rate": 2e-05, |
| "loss": 0.7158, |
| "step": 7600 |
| }, |
| { |
| "epoch": 0.6537800687285223, |
| "grad_norm": 0.3905238211154938, |
| "learning_rate": 2e-05, |
| "loss": 0.7351, |
| "step": 7610 |
| }, |
| { |
| "epoch": 0.654639175257732, |
| "grad_norm": 0.4532459080219269, |
| "learning_rate": 2e-05, |
| "loss": 0.7399, |
| "step": 7620 |
| }, |
| { |
| "epoch": 0.6554982817869416, |
| "grad_norm": 0.41000667214393616, |
| "learning_rate": 2e-05, |
| "loss": 0.7469, |
| "step": 7630 |
| }, |
| { |
| "epoch": 0.6563573883161512, |
| "grad_norm": 0.3942127525806427, |
| "learning_rate": 2e-05, |
| "loss": 0.7256, |
| "step": 7640 |
| }, |
| { |
| "epoch": 0.6572164948453608, |
| "grad_norm": 0.3943314254283905, |
| "learning_rate": 2e-05, |
| "loss": 0.6928, |
| "step": 7650 |
| }, |
| { |
| "epoch": 0.6580756013745704, |
| "grad_norm": 0.3827362656593323, |
| "learning_rate": 2e-05, |
| "loss": 0.7131, |
| "step": 7660 |
| }, |
| { |
| "epoch": 0.6589347079037801, |
| "grad_norm": 0.3889574110507965, |
| "learning_rate": 2e-05, |
| "loss": 0.7049, |
| "step": 7670 |
| }, |
| { |
| "epoch": 0.6597938144329897, |
| "grad_norm": 0.40035438537597656, |
| "learning_rate": 2e-05, |
| "loss": 0.7348, |
| "step": 7680 |
| }, |
| { |
| "epoch": 0.6606529209621993, |
| "grad_norm": 0.42468225955963135, |
| "learning_rate": 2e-05, |
| "loss": 0.6938, |
| "step": 7690 |
| }, |
| { |
| "epoch": 0.6615120274914089, |
| "grad_norm": 0.3995441794395447, |
| "learning_rate": 2e-05, |
| "loss": 0.7018, |
| "step": 7700 |
| }, |
| { |
| "epoch": 0.6623711340206185, |
| "grad_norm": 0.42671045660972595, |
| "learning_rate": 2e-05, |
| "loss": 0.7343, |
| "step": 7710 |
| }, |
| { |
| "epoch": 0.6632302405498282, |
| "grad_norm": 0.4042049050331116, |
| "learning_rate": 2e-05, |
| "loss": 0.7527, |
| "step": 7720 |
| }, |
| { |
| "epoch": 0.6640893470790378, |
| "grad_norm": 0.38925719261169434, |
| "learning_rate": 2e-05, |
| "loss": 0.7456, |
| "step": 7730 |
| }, |
| { |
| "epoch": 0.6649484536082474, |
| "grad_norm": 0.3824930191040039, |
| "learning_rate": 2e-05, |
| "loss": 0.7141, |
| "step": 7740 |
| }, |
| { |
| "epoch": 0.665807560137457, |
| "grad_norm": 0.43186309933662415, |
| "learning_rate": 2e-05, |
| "loss": 0.7384, |
| "step": 7750 |
| }, |
| { |
| "epoch": 0.6666666666666666, |
| "grad_norm": 0.37725701928138733, |
| "learning_rate": 2e-05, |
| "loss": 0.7045, |
| "step": 7760 |
| }, |
| { |
| "epoch": 0.6675257731958762, |
| "grad_norm": 0.3895176649093628, |
| "learning_rate": 2e-05, |
| "loss": 0.7306, |
| "step": 7770 |
| }, |
| { |
| "epoch": 0.6683848797250859, |
| "grad_norm": 0.3959781527519226, |
| "learning_rate": 2e-05, |
| "loss": 0.7203, |
| "step": 7780 |
| }, |
| { |
| "epoch": 0.6692439862542955, |
| "grad_norm": 2.8573715686798096, |
| "learning_rate": 2e-05, |
| "loss": 0.737, |
| "step": 7790 |
| }, |
| { |
| "epoch": 0.6701030927835051, |
| "grad_norm": 0.3638608753681183, |
| "learning_rate": 2e-05, |
| "loss": 0.7167, |
| "step": 7800 |
| }, |
| { |
| "epoch": 0.6709621993127147, |
| "grad_norm": 0.39853817224502563, |
| "learning_rate": 2e-05, |
| "loss": 0.7242, |
| "step": 7810 |
| }, |
| { |
| "epoch": 0.6718213058419243, |
| "grad_norm": 0.3915829658508301, |
| "learning_rate": 2e-05, |
| "loss": 0.7367, |
| "step": 7820 |
| }, |
| { |
| "epoch": 0.6726804123711341, |
| "grad_norm": 0.3991372883319855, |
| "learning_rate": 2e-05, |
| "loss": 0.7176, |
| "step": 7830 |
| }, |
| { |
| "epoch": 0.6735395189003437, |
| "grad_norm": 0.40146610140800476, |
| "learning_rate": 2e-05, |
| "loss": 0.6978, |
| "step": 7840 |
| }, |
| { |
| "epoch": 0.6743986254295533, |
| "grad_norm": 0.43911999464035034, |
| "learning_rate": 2e-05, |
| "loss": 0.7248, |
| "step": 7850 |
| }, |
| { |
| "epoch": 0.6752577319587629, |
| "grad_norm": 0.46387749910354614, |
| "learning_rate": 2e-05, |
| "loss": 0.7448, |
| "step": 7860 |
| }, |
| { |
| "epoch": 0.6761168384879725, |
| "grad_norm": 0.4136525094509125, |
| "learning_rate": 2e-05, |
| "loss": 0.7383, |
| "step": 7870 |
| }, |
| { |
| "epoch": 0.6769759450171822, |
| "grad_norm": 0.4044949412345886, |
| "learning_rate": 2e-05, |
| "loss": 0.7384, |
| "step": 7880 |
| }, |
| { |
| "epoch": 0.6778350515463918, |
| "grad_norm": 0.37605395913124084, |
| "learning_rate": 2e-05, |
| "loss": 0.7263, |
| "step": 7890 |
| }, |
| { |
| "epoch": 0.6786941580756014, |
| "grad_norm": 0.40915948152542114, |
| "learning_rate": 2e-05, |
| "loss": 0.7254, |
| "step": 7900 |
| }, |
| { |
| "epoch": 0.679553264604811, |
| "grad_norm": 0.42178675532341003, |
| "learning_rate": 2e-05, |
| "loss": 0.7171, |
| "step": 7910 |
| }, |
| { |
| "epoch": 0.6804123711340206, |
| "grad_norm": 0.38003844022750854, |
| "learning_rate": 2e-05, |
| "loss": 0.7069, |
| "step": 7920 |
| }, |
| { |
| "epoch": 0.6812714776632303, |
| "grad_norm": 0.38149625062942505, |
| "learning_rate": 2e-05, |
| "loss": 0.7389, |
| "step": 7930 |
| }, |
| { |
| "epoch": 0.6821305841924399, |
| "grad_norm": 0.45195889472961426, |
| "learning_rate": 2e-05, |
| "loss": 0.7304, |
| "step": 7940 |
| }, |
| { |
| "epoch": 0.6829896907216495, |
| "grad_norm": 0.41516923904418945, |
| "learning_rate": 2e-05, |
| "loss": 0.7353, |
| "step": 7950 |
| }, |
| { |
| "epoch": 0.6838487972508591, |
| "grad_norm": 0.38141489028930664, |
| "learning_rate": 2e-05, |
| "loss": 0.7125, |
| "step": 7960 |
| }, |
| { |
| "epoch": 0.6847079037800687, |
| "grad_norm": 0.40352198481559753, |
| "learning_rate": 2e-05, |
| "loss": 0.7253, |
| "step": 7970 |
| }, |
| { |
| "epoch": 0.6855670103092784, |
| "grad_norm": 0.4237516522407532, |
| "learning_rate": 2e-05, |
| "loss": 0.7018, |
| "step": 7980 |
| }, |
| { |
| "epoch": 0.686426116838488, |
| "grad_norm": 0.3906328082084656, |
| "learning_rate": 2e-05, |
| "loss": 0.7228, |
| "step": 7990 |
| }, |
| { |
| "epoch": 0.6872852233676976, |
| "grad_norm": 0.40206363797187805, |
| "learning_rate": 2e-05, |
| "loss": 0.732, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.6881443298969072, |
| "grad_norm": 0.409654825925827, |
| "learning_rate": 2e-05, |
| "loss": 0.7226, |
| "step": 8010 |
| }, |
| { |
| "epoch": 0.6890034364261168, |
| "grad_norm": 0.42076167464256287, |
| "learning_rate": 2e-05, |
| "loss": 0.725, |
| "step": 8020 |
| }, |
| { |
| "epoch": 0.6898625429553265, |
| "grad_norm": 0.40492281317710876, |
| "learning_rate": 2e-05, |
| "loss": 0.7175, |
| "step": 8030 |
| }, |
| { |
| "epoch": 0.6907216494845361, |
| "grad_norm": 0.3942924439907074, |
| "learning_rate": 2e-05, |
| "loss": 0.7406, |
| "step": 8040 |
| }, |
| { |
| "epoch": 0.6915807560137457, |
| "grad_norm": 0.38230153918266296, |
| "learning_rate": 2e-05, |
| "loss": 0.7298, |
| "step": 8050 |
| }, |
| { |
| "epoch": 0.6924398625429553, |
| "grad_norm": 0.4117579162120819, |
| "learning_rate": 2e-05, |
| "loss": 0.7205, |
| "step": 8060 |
| }, |
| { |
| "epoch": 0.6932989690721649, |
| "grad_norm": 0.4027566909790039, |
| "learning_rate": 2e-05, |
| "loss": 0.7117, |
| "step": 8070 |
| }, |
| { |
| "epoch": 0.6941580756013745, |
| "grad_norm": 0.38895997405052185, |
| "learning_rate": 2e-05, |
| "loss": 0.7467, |
| "step": 8080 |
| }, |
| { |
| "epoch": 0.6950171821305842, |
| "grad_norm": 0.39844027161598206, |
| "learning_rate": 2e-05, |
| "loss": 0.7305, |
| "step": 8090 |
| }, |
| { |
| "epoch": 0.6958762886597938, |
| "grad_norm": 0.4430878162384033, |
| "learning_rate": 2e-05, |
| "loss": 0.7009, |
| "step": 8100 |
| }, |
| { |
| "epoch": 0.6967353951890034, |
| "grad_norm": 0.3984880745410919, |
| "learning_rate": 2e-05, |
| "loss": 0.7151, |
| "step": 8110 |
| }, |
| { |
| "epoch": 0.697594501718213, |
| "grad_norm": 0.37220561504364014, |
| "learning_rate": 2e-05, |
| "loss": 0.7017, |
| "step": 8120 |
| }, |
| { |
| "epoch": 0.6984536082474226, |
| "grad_norm": 0.4082733690738678, |
| "learning_rate": 2e-05, |
| "loss": 0.7261, |
| "step": 8130 |
| }, |
| { |
| "epoch": 0.6993127147766323, |
| "grad_norm": 0.37330180406570435, |
| "learning_rate": 2e-05, |
| "loss": 0.7233, |
| "step": 8140 |
| }, |
| { |
| "epoch": 0.7001718213058419, |
| "grad_norm": 0.43090373277664185, |
| "learning_rate": 2e-05, |
| "loss": 0.73, |
| "step": 8150 |
| }, |
| { |
| "epoch": 0.7010309278350515, |
| "grad_norm": 0.44349825382232666, |
| "learning_rate": 2e-05, |
| "loss": 0.7091, |
| "step": 8160 |
| }, |
| { |
| "epoch": 0.7018900343642611, |
| "grad_norm": 0.3924238979816437, |
| "learning_rate": 2e-05, |
| "loss": 0.7055, |
| "step": 8170 |
| }, |
| { |
| "epoch": 0.7027491408934707, |
| "grad_norm": 0.374772846698761, |
| "learning_rate": 2e-05, |
| "loss": 0.733, |
| "step": 8180 |
| }, |
| { |
| "epoch": 0.7036082474226805, |
| "grad_norm": 0.3925633132457733, |
| "learning_rate": 2e-05, |
| "loss": 0.7394, |
| "step": 8190 |
| }, |
| { |
| "epoch": 0.7044673539518901, |
| "grad_norm": 0.38702136278152466, |
| "learning_rate": 2e-05, |
| "loss": 0.7257, |
| "step": 8200 |
| }, |
| { |
| "epoch": 0.7053264604810997, |
| "grad_norm": 0.38145485520362854, |
| "learning_rate": 2e-05, |
| "loss": 0.7485, |
| "step": 8210 |
| }, |
| { |
| "epoch": 0.7061855670103093, |
| "grad_norm": 0.4215205907821655, |
| "learning_rate": 2e-05, |
| "loss": 0.7288, |
| "step": 8220 |
| }, |
| { |
| "epoch": 0.7070446735395189, |
| "grad_norm": 0.3771664798259735, |
| "learning_rate": 2e-05, |
| "loss": 0.7019, |
| "step": 8230 |
| }, |
| { |
| "epoch": 0.7079037800687286, |
| "grad_norm": 0.39383402466773987, |
| "learning_rate": 2e-05, |
| "loss": 0.6993, |
| "step": 8240 |
| }, |
| { |
| "epoch": 0.7087628865979382, |
| "grad_norm": 1.2111461162567139, |
| "learning_rate": 2e-05, |
| "loss": 0.7413, |
| "step": 8250 |
| }, |
| { |
| "epoch": 0.7096219931271478, |
| "grad_norm": 0.37696823477745056, |
| "learning_rate": 2e-05, |
| "loss": 0.7234, |
| "step": 8260 |
| }, |
| { |
| "epoch": 0.7104810996563574, |
| "grad_norm": 0.3929274082183838, |
| "learning_rate": 2e-05, |
| "loss": 0.7099, |
| "step": 8270 |
| }, |
| { |
| "epoch": 0.711340206185567, |
| "grad_norm": 0.3827717900276184, |
| "learning_rate": 2e-05, |
| "loss": 0.7434, |
| "step": 8280 |
| }, |
| { |
| "epoch": 0.7121993127147767, |
| "grad_norm": 0.3648788332939148, |
| "learning_rate": 2e-05, |
| "loss": 0.717, |
| "step": 8290 |
| }, |
| { |
| "epoch": 0.7130584192439863, |
| "grad_norm": 0.3969038426876068, |
| "learning_rate": 2e-05, |
| "loss": 0.7019, |
| "step": 8300 |
| }, |
| { |
| "epoch": 0.7139175257731959, |
| "grad_norm": 0.36151525378227234, |
| "learning_rate": 2e-05, |
| "loss": 0.7023, |
| "step": 8310 |
| }, |
| { |
| "epoch": 0.7147766323024055, |
| "grad_norm": 0.39144566655158997, |
| "learning_rate": 2e-05, |
| "loss": 0.7093, |
| "step": 8320 |
| }, |
| { |
| "epoch": 0.7156357388316151, |
| "grad_norm": 0.36375683546066284, |
| "learning_rate": 2e-05, |
| "loss": 0.7363, |
| "step": 8330 |
| }, |
| { |
| "epoch": 0.7164948453608248, |
| "grad_norm": 0.3640702962875366, |
| "learning_rate": 2e-05, |
| "loss": 0.7019, |
| "step": 8340 |
| }, |
| { |
| "epoch": 0.7173539518900344, |
| "grad_norm": 0.4449024200439453, |
| "learning_rate": 2e-05, |
| "loss": 0.7183, |
| "step": 8350 |
| }, |
| { |
| "epoch": 0.718213058419244, |
| "grad_norm": 0.3944387435913086, |
| "learning_rate": 2e-05, |
| "loss": 0.7281, |
| "step": 8360 |
| }, |
| { |
| "epoch": 0.7190721649484536, |
| "grad_norm": 0.35167938470840454, |
| "learning_rate": 2e-05, |
| "loss": 0.7234, |
| "step": 8370 |
| }, |
| { |
| "epoch": 0.7199312714776632, |
| "grad_norm": 0.4033135175704956, |
| "learning_rate": 2e-05, |
| "loss": 0.706, |
| "step": 8380 |
| }, |
| { |
| "epoch": 0.7207903780068728, |
| "grad_norm": 0.40369006991386414, |
| "learning_rate": 2e-05, |
| "loss": 0.7261, |
| "step": 8390 |
| }, |
| { |
| "epoch": 0.7216494845360825, |
| "grad_norm": 0.39932918548583984, |
| "learning_rate": 2e-05, |
| "loss": 0.7074, |
| "step": 8400 |
| }, |
| { |
| "epoch": 0.7225085910652921, |
| "grad_norm": 0.37931233644485474, |
| "learning_rate": 2e-05, |
| "loss": 0.711, |
| "step": 8410 |
| }, |
| { |
| "epoch": 0.7233676975945017, |
| "grad_norm": 0.37852662801742554, |
| "learning_rate": 2e-05, |
| "loss": 0.6851, |
| "step": 8420 |
| }, |
| { |
| "epoch": 0.7242268041237113, |
| "grad_norm": 0.36685460805892944, |
| "learning_rate": 2e-05, |
| "loss": 0.6586, |
| "step": 8430 |
| }, |
| { |
| "epoch": 0.7250859106529209, |
| "grad_norm": 0.38326916098594666, |
| "learning_rate": 2e-05, |
| "loss": 0.7155, |
| "step": 8440 |
| }, |
| { |
| "epoch": 0.7259450171821306, |
| "grad_norm": 0.3672126829624176, |
| "learning_rate": 2e-05, |
| "loss": 0.719, |
| "step": 8450 |
| }, |
| { |
| "epoch": 0.7268041237113402, |
| "grad_norm": 0.3831152319908142, |
| "learning_rate": 2e-05, |
| "loss": 0.6962, |
| "step": 8460 |
| }, |
| { |
| "epoch": 0.7276632302405498, |
| "grad_norm": 0.3660293221473694, |
| "learning_rate": 2e-05, |
| "loss": 0.7319, |
| "step": 8470 |
| }, |
| { |
| "epoch": 0.7285223367697594, |
| "grad_norm": 0.38653993606567383, |
| "learning_rate": 2e-05, |
| "loss": 0.7496, |
| "step": 8480 |
| }, |
| { |
| "epoch": 0.729381443298969, |
| "grad_norm": 0.40697425603866577, |
| "learning_rate": 2e-05, |
| "loss": 0.7159, |
| "step": 8490 |
| }, |
| { |
| "epoch": 0.7302405498281787, |
| "grad_norm": 0.3905690312385559, |
| "learning_rate": 2e-05, |
| "loss": 0.7484, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.7310996563573883, |
| "grad_norm": 0.37530186772346497, |
| "learning_rate": 2e-05, |
| "loss": 0.7316, |
| "step": 8510 |
| }, |
| { |
| "epoch": 0.7319587628865979, |
| "grad_norm": 0.3827313780784607, |
| "learning_rate": 2e-05, |
| "loss": 0.7178, |
| "step": 8520 |
| }, |
| { |
| "epoch": 0.7328178694158075, |
| "grad_norm": 0.38314542174339294, |
| "learning_rate": 2e-05, |
| "loss": 0.7184, |
| "step": 8530 |
| }, |
| { |
| "epoch": 0.7336769759450171, |
| "grad_norm": 0.3918944001197815, |
| "learning_rate": 2e-05, |
| "loss": 0.7281, |
| "step": 8540 |
| }, |
| { |
| "epoch": 0.7345360824742269, |
| "grad_norm": 0.3943602442741394, |
| "learning_rate": 2e-05, |
| "loss": 0.7254, |
| "step": 8550 |
| }, |
| { |
| "epoch": 0.7353951890034365, |
| "grad_norm": 0.37051719427108765, |
| "learning_rate": 2e-05, |
| "loss": 0.7201, |
| "step": 8560 |
| }, |
| { |
| "epoch": 0.7362542955326461, |
| "grad_norm": 0.4317139685153961, |
| "learning_rate": 2e-05, |
| "loss": 0.736, |
| "step": 8570 |
| }, |
| { |
| "epoch": 0.7371134020618557, |
| "grad_norm": 0.35919347405433655, |
| "learning_rate": 2e-05, |
| "loss": 0.7125, |
| "step": 8580 |
| }, |
| { |
| "epoch": 0.7379725085910653, |
| "grad_norm": 0.4450185000896454, |
| "learning_rate": 2e-05, |
| "loss": 0.741, |
| "step": 8590 |
| }, |
| { |
| "epoch": 0.738831615120275, |
| "grad_norm": 0.40356162190437317, |
| "learning_rate": 2e-05, |
| "loss": 0.7259, |
| "step": 8600 |
| }, |
| { |
| "epoch": 0.7396907216494846, |
| "grad_norm": 0.3860047161579132, |
| "learning_rate": 2e-05, |
| "loss": 0.6955, |
| "step": 8610 |
| }, |
| { |
| "epoch": 0.7405498281786942, |
| "grad_norm": 0.38151559233665466, |
| "learning_rate": 2e-05, |
| "loss": 0.7158, |
| "step": 8620 |
| }, |
| { |
| "epoch": 0.7414089347079038, |
| "grad_norm": 0.3991175591945648, |
| "learning_rate": 2e-05, |
| "loss": 0.7196, |
| "step": 8630 |
| }, |
| { |
| "epoch": 0.7422680412371134, |
| "grad_norm": 0.38009336590766907, |
| "learning_rate": 2e-05, |
| "loss": 0.7382, |
| "step": 8640 |
| }, |
| { |
| "epoch": 0.743127147766323, |
| "grad_norm": 0.4248391091823578, |
| "learning_rate": 2e-05, |
| "loss": 0.733, |
| "step": 8650 |
| }, |
| { |
| "epoch": 0.7439862542955327, |
| "grad_norm": 0.41284582018852234, |
| "learning_rate": 2e-05, |
| "loss": 0.739, |
| "step": 8660 |
| }, |
| { |
| "epoch": 0.7448453608247423, |
| "grad_norm": 0.4079428017139435, |
| "learning_rate": 2e-05, |
| "loss": 0.7089, |
| "step": 8670 |
| }, |
| { |
| "epoch": 0.7457044673539519, |
| "grad_norm": 0.41107913851737976, |
| "learning_rate": 2e-05, |
| "loss": 0.7119, |
| "step": 8680 |
| }, |
| { |
| "epoch": 0.7465635738831615, |
| "grad_norm": 0.40985432267189026, |
| "learning_rate": 2e-05, |
| "loss": 0.6935, |
| "step": 8690 |
| }, |
| { |
| "epoch": 0.7474226804123711, |
| "grad_norm": 0.3884676396846771, |
| "learning_rate": 2e-05, |
| "loss": 0.7146, |
| "step": 8700 |
| }, |
| { |
| "epoch": 0.7482817869415808, |
| "grad_norm": 0.38182690739631653, |
| "learning_rate": 2e-05, |
| "loss": 0.7072, |
| "step": 8710 |
| }, |
| { |
| "epoch": 0.7491408934707904, |
| "grad_norm": 0.3910977840423584, |
| "learning_rate": 2e-05, |
| "loss": 0.7372, |
| "step": 8720 |
| }, |
| { |
| "epoch": 0.75, |
| "grad_norm": 0.3714883327484131, |
| "learning_rate": 2e-05, |
| "loss": 0.7314, |
| "step": 8730 |
| }, |
| { |
| "epoch": 0.7508591065292096, |
| "grad_norm": 0.4015692174434662, |
| "learning_rate": 2e-05, |
| "loss": 0.7207, |
| "step": 8740 |
| }, |
| { |
| "epoch": 0.7517182130584192, |
| "grad_norm": 0.38211649656295776, |
| "learning_rate": 2e-05, |
| "loss": 0.7381, |
| "step": 8750 |
| }, |
| { |
| "epoch": 0.7525773195876289, |
| "grad_norm": 0.4525204598903656, |
| "learning_rate": 2e-05, |
| "loss": 0.7068, |
| "step": 8760 |
| }, |
| { |
| "epoch": 0.7534364261168385, |
| "grad_norm": 0.4487990736961365, |
| "learning_rate": 2e-05, |
| "loss": 0.727, |
| "step": 8770 |
| }, |
| { |
| "epoch": 0.7542955326460481, |
| "grad_norm": 0.4486429989337921, |
| "learning_rate": 2e-05, |
| "loss": 0.7268, |
| "step": 8780 |
| }, |
| { |
| "epoch": 0.7551546391752577, |
| "grad_norm": 0.39998337626457214, |
| "learning_rate": 2e-05, |
| "loss": 0.7285, |
| "step": 8790 |
| }, |
| { |
| "epoch": 0.7560137457044673, |
| "grad_norm": 0.46015822887420654, |
| "learning_rate": 2e-05, |
| "loss": 0.7215, |
| "step": 8800 |
| }, |
| { |
| "epoch": 0.756872852233677, |
| "grad_norm": 0.3716229796409607, |
| "learning_rate": 2e-05, |
| "loss": 0.7145, |
| "step": 8810 |
| }, |
| { |
| "epoch": 0.7577319587628866, |
| "grad_norm": 0.3896833658218384, |
| "learning_rate": 2e-05, |
| "loss": 0.6891, |
| "step": 8820 |
| }, |
| { |
| "epoch": 0.7585910652920962, |
| "grad_norm": 0.3618304431438446, |
| "learning_rate": 2e-05, |
| "loss": 0.7112, |
| "step": 8830 |
| }, |
| { |
| "epoch": 0.7594501718213058, |
| "grad_norm": 0.37147557735443115, |
| "learning_rate": 2e-05, |
| "loss": 0.7195, |
| "step": 8840 |
| }, |
| { |
| "epoch": 0.7603092783505154, |
| "grad_norm": 0.3710358142852783, |
| "learning_rate": 2e-05, |
| "loss": 0.7385, |
| "step": 8850 |
| }, |
| { |
| "epoch": 0.761168384879725, |
| "grad_norm": 0.39484795928001404, |
| "learning_rate": 2e-05, |
| "loss": 0.6865, |
| "step": 8860 |
| }, |
| { |
| "epoch": 0.7620274914089347, |
| "grad_norm": 0.41713300347328186, |
| "learning_rate": 2e-05, |
| "loss": 0.7173, |
| "step": 8870 |
| }, |
| { |
| "epoch": 0.7628865979381443, |
| "grad_norm": 0.3603912591934204, |
| "learning_rate": 2e-05, |
| "loss": 0.6855, |
| "step": 8880 |
| }, |
| { |
| "epoch": 0.7637457044673539, |
| "grad_norm": 0.39539051055908203, |
| "learning_rate": 2e-05, |
| "loss": 0.7111, |
| "step": 8890 |
| }, |
| { |
| "epoch": 0.7646048109965635, |
| "grad_norm": 0.39834073185920715, |
| "learning_rate": 2e-05, |
| "loss": 0.7166, |
| "step": 8900 |
| }, |
| { |
| "epoch": 0.7654639175257731, |
| "grad_norm": 0.4484548568725586, |
| "learning_rate": 2e-05, |
| "loss": 0.7227, |
| "step": 8910 |
| }, |
| { |
| "epoch": 0.7663230240549829, |
| "grad_norm": 0.3844361901283264, |
| "learning_rate": 2e-05, |
| "loss": 0.7448, |
| "step": 8920 |
| }, |
| { |
| "epoch": 0.7671821305841925, |
| "grad_norm": 0.35846951603889465, |
| "learning_rate": 2e-05, |
| "loss": 0.7378, |
| "step": 8930 |
| }, |
| { |
| "epoch": 0.7680412371134021, |
| "grad_norm": 0.38699042797088623, |
| "learning_rate": 2e-05, |
| "loss": 0.71, |
| "step": 8940 |
| }, |
| { |
| "epoch": 0.7689003436426117, |
| "grad_norm": 0.39588233828544617, |
| "learning_rate": 2e-05, |
| "loss": 0.7223, |
| "step": 8950 |
| }, |
| { |
| "epoch": 0.7697594501718213, |
| "grad_norm": 0.37110257148742676, |
| "learning_rate": 2e-05, |
| "loss": 0.71, |
| "step": 8960 |
| }, |
| { |
| "epoch": 0.770618556701031, |
| "grad_norm": 0.3636080026626587, |
| "learning_rate": 2e-05, |
| "loss": 0.7031, |
| "step": 8970 |
| }, |
| { |
| "epoch": 0.7714776632302406, |
| "grad_norm": 0.3489440083503723, |
| "learning_rate": 2e-05, |
| "loss": 0.6925, |
| "step": 8980 |
| }, |
| { |
| "epoch": 0.7723367697594502, |
| "grad_norm": 0.39057454466819763, |
| "learning_rate": 2e-05, |
| "loss": 0.7307, |
| "step": 8990 |
| }, |
| { |
| "epoch": 0.7731958762886598, |
| "grad_norm": 0.4395849108695984, |
| "learning_rate": 2e-05, |
| "loss": 0.7167, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.7740549828178694, |
| "grad_norm": 0.436786949634552, |
| "learning_rate": 2e-05, |
| "loss": 0.709, |
| "step": 9010 |
| }, |
| { |
| "epoch": 0.7749140893470791, |
| "grad_norm": 0.3535119891166687, |
| "learning_rate": 2e-05, |
| "loss": 0.7156, |
| "step": 9020 |
| }, |
| { |
| "epoch": 0.7757731958762887, |
| "grad_norm": 0.42912858724594116, |
| "learning_rate": 2e-05, |
| "loss": 0.7145, |
| "step": 9030 |
| }, |
| { |
| "epoch": 0.7766323024054983, |
| "grad_norm": 0.36896806955337524, |
| "learning_rate": 2e-05, |
| "loss": 0.704, |
| "step": 9040 |
| }, |
| { |
| "epoch": 0.7774914089347079, |
| "grad_norm": 0.392666220664978, |
| "learning_rate": 2e-05, |
| "loss": 0.728, |
| "step": 9050 |
| }, |
| { |
| "epoch": 0.7783505154639175, |
| "grad_norm": 0.41010749340057373, |
| "learning_rate": 2e-05, |
| "loss": 0.7304, |
| "step": 9060 |
| }, |
| { |
| "epoch": 0.7792096219931272, |
| "grad_norm": 0.4841192066669464, |
| "learning_rate": 2e-05, |
| "loss": 0.7318, |
| "step": 9070 |
| }, |
| { |
| "epoch": 0.7800687285223368, |
| "grad_norm": 0.4069053530693054, |
| "learning_rate": 2e-05, |
| "loss": 0.7333, |
| "step": 9080 |
| }, |
| { |
| "epoch": 0.7809278350515464, |
| "grad_norm": 0.42515045404434204, |
| "learning_rate": 2e-05, |
| "loss": 0.7283, |
| "step": 9090 |
| }, |
| { |
| "epoch": 0.781786941580756, |
| "grad_norm": 0.3845844268798828, |
| "learning_rate": 2e-05, |
| "loss": 0.7316, |
| "step": 9100 |
| }, |
| { |
| "epoch": 0.7826460481099656, |
| "grad_norm": 0.3778882324695587, |
| "learning_rate": 2e-05, |
| "loss": 0.7099, |
| "step": 9110 |
| }, |
| { |
| "epoch": 0.7835051546391752, |
| "grad_norm": 0.3805033266544342, |
| "learning_rate": 2e-05, |
| "loss": 0.6934, |
| "step": 9120 |
| }, |
| { |
| "epoch": 0.7843642611683849, |
| "grad_norm": 0.3901061713695526, |
| "learning_rate": 2e-05, |
| "loss": 0.7081, |
| "step": 9130 |
| }, |
| { |
| "epoch": 0.7852233676975945, |
| "grad_norm": 0.394225150346756, |
| "learning_rate": 2e-05, |
| "loss": 0.6887, |
| "step": 9140 |
| }, |
| { |
| "epoch": 0.7860824742268041, |
| "grad_norm": 0.3621334731578827, |
| "learning_rate": 2e-05, |
| "loss": 0.704, |
| "step": 9150 |
| }, |
| { |
| "epoch": 0.7869415807560137, |
| "grad_norm": 0.4231373965740204, |
| "learning_rate": 2e-05, |
| "loss": 0.6894, |
| "step": 9160 |
| }, |
| { |
| "epoch": 0.7878006872852233, |
| "grad_norm": 0.36363667249679565, |
| "learning_rate": 2e-05, |
| "loss": 0.695, |
| "step": 9170 |
| }, |
| { |
| "epoch": 0.788659793814433, |
| "grad_norm": 0.3719753921031952, |
| "learning_rate": 2e-05, |
| "loss": 0.7068, |
| "step": 9180 |
| }, |
| { |
| "epoch": 0.7895189003436426, |
| "grad_norm": 0.4035317599773407, |
| "learning_rate": 2e-05, |
| "loss": 0.7057, |
| "step": 9190 |
| }, |
| { |
| "epoch": 0.7903780068728522, |
| "grad_norm": 0.37357720732688904, |
| "learning_rate": 2e-05, |
| "loss": 0.7441, |
| "step": 9200 |
| }, |
| { |
| "epoch": 0.7912371134020618, |
| "grad_norm": 0.4038161039352417, |
| "learning_rate": 2e-05, |
| "loss": 0.7279, |
| "step": 9210 |
| }, |
| { |
| "epoch": 0.7920962199312714, |
| "grad_norm": 0.36732155084609985, |
| "learning_rate": 2e-05, |
| "loss": 0.7109, |
| "step": 9220 |
| }, |
| { |
| "epoch": 0.7929553264604811, |
| "grad_norm": 0.3961270749568939, |
| "learning_rate": 2e-05, |
| "loss": 0.7384, |
| "step": 9230 |
| }, |
| { |
| "epoch": 0.7938144329896907, |
| "grad_norm": 0.3747657537460327, |
| "learning_rate": 2e-05, |
| "loss": 0.7148, |
| "step": 9240 |
| }, |
| { |
| "epoch": 0.7946735395189003, |
| "grad_norm": 0.37883415818214417, |
| "learning_rate": 2e-05, |
| "loss": 0.7103, |
| "step": 9250 |
| }, |
| { |
| "epoch": 0.7955326460481099, |
| "grad_norm": 0.35879454016685486, |
| "learning_rate": 2e-05, |
| "loss": 0.7081, |
| "step": 9260 |
| }, |
| { |
| "epoch": 0.7963917525773195, |
| "grad_norm": 0.40805768966674805, |
| "learning_rate": 2e-05, |
| "loss": 0.7228, |
| "step": 9270 |
| }, |
| { |
| "epoch": 0.7972508591065293, |
| "grad_norm": 0.3966427743434906, |
| "learning_rate": 2e-05, |
| "loss": 0.7334, |
| "step": 9280 |
| }, |
| { |
| "epoch": 0.7981099656357389, |
| "grad_norm": 0.41182687878608704, |
| "learning_rate": 2e-05, |
| "loss": 0.7092, |
| "step": 9290 |
| }, |
| { |
| "epoch": 0.7989690721649485, |
| "grad_norm": 0.3790788948535919, |
| "learning_rate": 2e-05, |
| "loss": 0.7281, |
| "step": 9300 |
| }, |
| { |
| "epoch": 0.7998281786941581, |
| "grad_norm": 0.3796280026435852, |
| "learning_rate": 2e-05, |
| "loss": 0.7015, |
| "step": 9310 |
| }, |
| { |
| "epoch": 0.8006872852233677, |
| "grad_norm": 0.4023306369781494, |
| "learning_rate": 2e-05, |
| "loss": 0.7153, |
| "step": 9320 |
| }, |
| { |
| "epoch": 0.8015463917525774, |
| "grad_norm": 0.387482225894928, |
| "learning_rate": 2e-05, |
| "loss": 0.7117, |
| "step": 9330 |
| }, |
| { |
| "epoch": 0.802405498281787, |
| "grad_norm": 0.4018867313861847, |
| "learning_rate": 2e-05, |
| "loss": 0.7131, |
| "step": 9340 |
| }, |
| { |
| "epoch": 0.8032646048109966, |
| "grad_norm": 0.3944343328475952, |
| "learning_rate": 2e-05, |
| "loss": 0.7147, |
| "step": 9350 |
| }, |
| { |
| "epoch": 0.8041237113402062, |
| "grad_norm": 0.3635457754135132, |
| "learning_rate": 2e-05, |
| "loss": 0.7185, |
| "step": 9360 |
| }, |
| { |
| "epoch": 0.8049828178694158, |
| "grad_norm": 0.4009116590023041, |
| "learning_rate": 2e-05, |
| "loss": 0.7405, |
| "step": 9370 |
| }, |
| { |
| "epoch": 0.8058419243986255, |
| "grad_norm": 0.3992771506309509, |
| "learning_rate": 2e-05, |
| "loss": 0.7217, |
| "step": 9380 |
| }, |
| { |
| "epoch": 0.8067010309278351, |
| "grad_norm": 0.3812587559223175, |
| "learning_rate": 2e-05, |
| "loss": 0.6896, |
| "step": 9390 |
| }, |
| { |
| "epoch": 0.8075601374570447, |
| "grad_norm": 0.38376256823539734, |
| "learning_rate": 2e-05, |
| "loss": 0.6934, |
| "step": 9400 |
| }, |
| { |
| "epoch": 0.8084192439862543, |
| "grad_norm": 0.4016886055469513, |
| "learning_rate": 2e-05, |
| "loss": 0.7249, |
| "step": 9410 |
| }, |
| { |
| "epoch": 0.8092783505154639, |
| "grad_norm": 0.37972691655158997, |
| "learning_rate": 2e-05, |
| "loss": 0.6867, |
| "step": 9420 |
| }, |
| { |
| "epoch": 0.8101374570446735, |
| "grad_norm": 0.39387190341949463, |
| "learning_rate": 2e-05, |
| "loss": 0.6812, |
| "step": 9430 |
| }, |
| { |
| "epoch": 0.8109965635738832, |
| "grad_norm": 0.39428645372390747, |
| "learning_rate": 2e-05, |
| "loss": 0.7488, |
| "step": 9440 |
| }, |
| { |
| "epoch": 0.8118556701030928, |
| "grad_norm": 0.3848554193973541, |
| "learning_rate": 2e-05, |
| "loss": 0.7232, |
| "step": 9450 |
| }, |
| { |
| "epoch": 0.8127147766323024, |
| "grad_norm": 0.4475501477718353, |
| "learning_rate": 2e-05, |
| "loss": 0.7134, |
| "step": 9460 |
| }, |
| { |
| "epoch": 0.813573883161512, |
| "grad_norm": 0.38900819420814514, |
| "learning_rate": 2e-05, |
| "loss": 0.7096, |
| "step": 9470 |
| }, |
| { |
| "epoch": 0.8144329896907216, |
| "grad_norm": 0.38350027799606323, |
| "learning_rate": 2e-05, |
| "loss": 0.7467, |
| "step": 9480 |
| }, |
| { |
| "epoch": 0.8152920962199313, |
| "grad_norm": 0.3872745931148529, |
| "learning_rate": 2e-05, |
| "loss": 0.7333, |
| "step": 9490 |
| }, |
| { |
| "epoch": 0.8161512027491409, |
| "grad_norm": 0.3947836458683014, |
| "learning_rate": 2e-05, |
| "loss": 0.7293, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.8170103092783505, |
| "grad_norm": 0.39788034558296204, |
| "learning_rate": 2e-05, |
| "loss": 0.7248, |
| "step": 9510 |
| }, |
| { |
| "epoch": 0.8178694158075601, |
| "grad_norm": 0.40340569615364075, |
| "learning_rate": 2e-05, |
| "loss": 0.7241, |
| "step": 9520 |
| }, |
| { |
| "epoch": 0.8187285223367697, |
| "grad_norm": 0.4285871684551239, |
| "learning_rate": 2e-05, |
| "loss": 0.7099, |
| "step": 9530 |
| }, |
| { |
| "epoch": 0.8195876288659794, |
| "grad_norm": 0.38404223322868347, |
| "learning_rate": 2e-05, |
| "loss": 0.7333, |
| "step": 9540 |
| }, |
| { |
| "epoch": 0.820446735395189, |
| "grad_norm": 0.4171121120452881, |
| "learning_rate": 2e-05, |
| "loss": 0.7372, |
| "step": 9550 |
| }, |
| { |
| "epoch": 0.8213058419243986, |
| "grad_norm": 0.5208197236061096, |
| "learning_rate": 2e-05, |
| "loss": 0.7208, |
| "step": 9560 |
| }, |
| { |
| "epoch": 0.8221649484536082, |
| "grad_norm": 0.39956343173980713, |
| "learning_rate": 2e-05, |
| "loss": 0.7018, |
| "step": 9570 |
| }, |
| { |
| "epoch": 0.8230240549828178, |
| "grad_norm": 0.3871324956417084, |
| "learning_rate": 2e-05, |
| "loss": 0.6897, |
| "step": 9580 |
| }, |
| { |
| "epoch": 0.8238831615120275, |
| "grad_norm": 0.37801074981689453, |
| "learning_rate": 2e-05, |
| "loss": 0.7414, |
| "step": 9590 |
| }, |
| { |
| "epoch": 0.8247422680412371, |
| "grad_norm": 0.3848244547843933, |
| "learning_rate": 2e-05, |
| "loss": 0.7034, |
| "step": 9600 |
| }, |
| { |
| "epoch": 0.8256013745704467, |
| "grad_norm": 0.37562504410743713, |
| "learning_rate": 2e-05, |
| "loss": 0.7515, |
| "step": 9610 |
| }, |
| { |
| "epoch": 0.8264604810996563, |
| "grad_norm": 0.3993630111217499, |
| "learning_rate": 2e-05, |
| "loss": 0.7224, |
| "step": 9620 |
| }, |
| { |
| "epoch": 0.8273195876288659, |
| "grad_norm": 0.3967456519603729, |
| "learning_rate": 2e-05, |
| "loss": 0.7055, |
| "step": 9630 |
| }, |
| { |
| "epoch": 0.8281786941580757, |
| "grad_norm": 0.5025458335876465, |
| "learning_rate": 2e-05, |
| "loss": 0.71, |
| "step": 9640 |
| }, |
| { |
| "epoch": 0.8290378006872853, |
| "grad_norm": 0.36649489402770996, |
| "learning_rate": 2e-05, |
| "loss": 0.7195, |
| "step": 9650 |
| }, |
| { |
| "epoch": 0.8298969072164949, |
| "grad_norm": 0.36913812160491943, |
| "learning_rate": 2e-05, |
| "loss": 0.7306, |
| "step": 9660 |
| }, |
| { |
| "epoch": 0.8307560137457045, |
| "grad_norm": 0.38220715522766113, |
| "learning_rate": 2e-05, |
| "loss": 0.7075, |
| "step": 9670 |
| }, |
| { |
| "epoch": 0.8316151202749141, |
| "grad_norm": 0.4619491398334503, |
| "learning_rate": 2e-05, |
| "loss": 0.7276, |
| "step": 9680 |
| }, |
| { |
| "epoch": 0.8324742268041238, |
| "grad_norm": 0.3976234793663025, |
| "learning_rate": 2e-05, |
| "loss": 0.7261, |
| "step": 9690 |
| }, |
| { |
| "epoch": 0.8333333333333334, |
| "grad_norm": 0.39764654636383057, |
| "learning_rate": 2e-05, |
| "loss": 0.7375, |
| "step": 9700 |
| }, |
| { |
| "epoch": 0.834192439862543, |
| "grad_norm": 0.3769300580024719, |
| "learning_rate": 2e-05, |
| "loss": 0.6918, |
| "step": 9710 |
| }, |
| { |
| "epoch": 0.8350515463917526, |
| "grad_norm": 0.37111008167266846, |
| "learning_rate": 2e-05, |
| "loss": 0.7261, |
| "step": 9720 |
| }, |
| { |
| "epoch": 0.8359106529209622, |
| "grad_norm": 0.3931925594806671, |
| "learning_rate": 2e-05, |
| "loss": 0.7418, |
| "step": 9730 |
| }, |
| { |
| "epoch": 0.8367697594501718, |
| "grad_norm": 0.4176950454711914, |
| "learning_rate": 2e-05, |
| "loss": 0.7426, |
| "step": 9740 |
| }, |
| { |
| "epoch": 0.8376288659793815, |
| "grad_norm": 0.4077303111553192, |
| "learning_rate": 2e-05, |
| "loss": 0.7433, |
| "step": 9750 |
| }, |
| { |
| "epoch": 0.8384879725085911, |
| "grad_norm": 0.41396117210388184, |
| "learning_rate": 2e-05, |
| "loss": 0.7096, |
| "step": 9760 |
| }, |
| { |
| "epoch": 0.8393470790378007, |
| "grad_norm": 0.37262779474258423, |
| "learning_rate": 2e-05, |
| "loss": 0.7233, |
| "step": 9770 |
| }, |
| { |
| "epoch": 0.8402061855670103, |
| "grad_norm": 0.3671231269836426, |
| "learning_rate": 2e-05, |
| "loss": 0.7135, |
| "step": 9780 |
| }, |
| { |
| "epoch": 0.8410652920962199, |
| "grad_norm": 0.37881162762641907, |
| "learning_rate": 2e-05, |
| "loss": 0.7088, |
| "step": 9790 |
| }, |
| { |
| "epoch": 0.8419243986254296, |
| "grad_norm": 0.36753877997398376, |
| "learning_rate": 2e-05, |
| "loss": 0.6823, |
| "step": 9800 |
| }, |
| { |
| "epoch": 0.8427835051546392, |
| "grad_norm": 0.39431846141815186, |
| "learning_rate": 2e-05, |
| "loss": 0.7077, |
| "step": 9810 |
| }, |
| { |
| "epoch": 0.8436426116838488, |
| "grad_norm": 0.3776985704898834, |
| "learning_rate": 2e-05, |
| "loss": 0.7215, |
| "step": 9820 |
| }, |
| { |
| "epoch": 0.8445017182130584, |
| "grad_norm": 0.4258396327495575, |
| "learning_rate": 2e-05, |
| "loss": 0.7011, |
| "step": 9830 |
| }, |
| { |
| "epoch": 0.845360824742268, |
| "grad_norm": 0.40351206064224243, |
| "learning_rate": 2e-05, |
| "loss": 0.7294, |
| "step": 9840 |
| }, |
| { |
| "epoch": 0.8462199312714777, |
| "grad_norm": 0.42546284198760986, |
| "learning_rate": 2e-05, |
| "loss": 0.7272, |
| "step": 9850 |
| }, |
| { |
| "epoch": 0.8470790378006873, |
| "grad_norm": 0.4175368845462799, |
| "learning_rate": 2e-05, |
| "loss": 0.7514, |
| "step": 9860 |
| }, |
| { |
| "epoch": 0.8479381443298969, |
| "grad_norm": 0.3797371983528137, |
| "learning_rate": 2e-05, |
| "loss": 0.7181, |
| "step": 9870 |
| }, |
| { |
| "epoch": 0.8487972508591065, |
| "grad_norm": 0.4009544253349304, |
| "learning_rate": 2e-05, |
| "loss": 0.7033, |
| "step": 9880 |
| }, |
| { |
| "epoch": 0.8496563573883161, |
| "grad_norm": 0.40033429861068726, |
| "learning_rate": 2e-05, |
| "loss": 0.7173, |
| "step": 9890 |
| }, |
| { |
| "epoch": 0.8505154639175257, |
| "grad_norm": 0.38404399156570435, |
| "learning_rate": 2e-05, |
| "loss": 0.7306, |
| "step": 9900 |
| }, |
| { |
| "epoch": 0.8513745704467354, |
| "grad_norm": 0.3615536093711853, |
| "learning_rate": 2e-05, |
| "loss": 0.6965, |
| "step": 9910 |
| }, |
| { |
| "epoch": 0.852233676975945, |
| "grad_norm": 0.38004836440086365, |
| "learning_rate": 2e-05, |
| "loss": 0.6933, |
| "step": 9920 |
| }, |
| { |
| "epoch": 0.8530927835051546, |
| "grad_norm": 0.3871254622936249, |
| "learning_rate": 2e-05, |
| "loss": 0.7047, |
| "step": 9930 |
| }, |
| { |
| "epoch": 0.8539518900343642, |
| "grad_norm": 0.3907207250595093, |
| "learning_rate": 2e-05, |
| "loss": 0.702, |
| "step": 9940 |
| }, |
| { |
| "epoch": 0.8548109965635738, |
| "grad_norm": 0.3619770407676697, |
| "learning_rate": 2e-05, |
| "loss": 0.7256, |
| "step": 9950 |
| }, |
| { |
| "epoch": 0.8556701030927835, |
| "grad_norm": 0.3915918469429016, |
| "learning_rate": 2e-05, |
| "loss": 0.7273, |
| "step": 9960 |
| }, |
| { |
| "epoch": 0.8565292096219931, |
| "grad_norm": 0.3925028145313263, |
| "learning_rate": 2e-05, |
| "loss": 0.7379, |
| "step": 9970 |
| }, |
| { |
| "epoch": 0.8573883161512027, |
| "grad_norm": 0.36883440613746643, |
| "learning_rate": 2e-05, |
| "loss": 0.6967, |
| "step": 9980 |
| }, |
| { |
| "epoch": 0.8582474226804123, |
| "grad_norm": 0.3517337739467621, |
| "learning_rate": 2e-05, |
| "loss": 0.72, |
| "step": 9990 |
| }, |
| { |
| "epoch": 0.8591065292096219, |
| "grad_norm": 0.406662255525589, |
| "learning_rate": 2e-05, |
| "loss": 0.7128, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.8599656357388317, |
| "grad_norm": 0.3930582106113434, |
| "learning_rate": 2e-05, |
| "loss": 0.7125, |
| "step": 10010 |
| }, |
| { |
| "epoch": 0.8608247422680413, |
| "grad_norm": 0.38763442635536194, |
| "learning_rate": 2e-05, |
| "loss": 0.7222, |
| "step": 10020 |
| }, |
| { |
| "epoch": 0.8616838487972509, |
| "grad_norm": 0.3780314028263092, |
| "learning_rate": 2e-05, |
| "loss": 0.694, |
| "step": 10030 |
| }, |
| { |
| "epoch": 0.8625429553264605, |
| "grad_norm": 0.38106173276901245, |
| "learning_rate": 2e-05, |
| "loss": 0.716, |
| "step": 10040 |
| }, |
| { |
| "epoch": 0.8634020618556701, |
| "grad_norm": 0.41772711277008057, |
| "learning_rate": 2e-05, |
| "loss": 0.7225, |
| "step": 10050 |
| }, |
| { |
| "epoch": 0.8642611683848798, |
| "grad_norm": 0.37731948494911194, |
| "learning_rate": 2e-05, |
| "loss": 0.7353, |
| "step": 10060 |
| }, |
| { |
| "epoch": 0.8651202749140894, |
| "grad_norm": 0.38751694560050964, |
| "learning_rate": 2e-05, |
| "loss": 0.718, |
| "step": 10070 |
| }, |
| { |
| "epoch": 0.865979381443299, |
| "grad_norm": 0.3983141779899597, |
| "learning_rate": 2e-05, |
| "loss": 0.7114, |
| "step": 10080 |
| }, |
| { |
| "epoch": 0.8668384879725086, |
| "grad_norm": 0.4114827513694763, |
| "learning_rate": 2e-05, |
| "loss": 0.7254, |
| "step": 10090 |
| }, |
| { |
| "epoch": 0.8676975945017182, |
| "grad_norm": 0.35398706793785095, |
| "learning_rate": 2e-05, |
| "loss": 0.6981, |
| "step": 10100 |
| }, |
| { |
| "epoch": 0.8685567010309279, |
| "grad_norm": 0.37684759497642517, |
| "learning_rate": 2e-05, |
| "loss": 0.6977, |
| "step": 10110 |
| }, |
| { |
| "epoch": 0.8694158075601375, |
| "grad_norm": 0.3989711403846741, |
| "learning_rate": 2e-05, |
| "loss": 0.7502, |
| "step": 10120 |
| }, |
| { |
| "epoch": 0.8702749140893471, |
| "grad_norm": 0.36875253915786743, |
| "learning_rate": 2e-05, |
| "loss": 0.7239, |
| "step": 10130 |
| }, |
| { |
| "epoch": 0.8711340206185567, |
| "grad_norm": 0.3745901882648468, |
| "learning_rate": 2e-05, |
| "loss": 0.6891, |
| "step": 10140 |
| }, |
| { |
| "epoch": 0.8719931271477663, |
| "grad_norm": 0.381804883480072, |
| "learning_rate": 2e-05, |
| "loss": 0.6977, |
| "step": 10150 |
| }, |
| { |
| "epoch": 0.872852233676976, |
| "grad_norm": 0.36928004026412964, |
| "learning_rate": 2e-05, |
| "loss": 0.7202, |
| "step": 10160 |
| }, |
| { |
| "epoch": 0.8737113402061856, |
| "grad_norm": 0.4012286961078644, |
| "learning_rate": 2e-05, |
| "loss": 0.6958, |
| "step": 10170 |
| }, |
| { |
| "epoch": 0.8745704467353952, |
| "grad_norm": 0.35307636857032776, |
| "learning_rate": 1.999957358086795e-05, |
| "loss": 0.7045, |
| "step": 10180 |
| }, |
| { |
| "epoch": 0.8754295532646048, |
| "grad_norm": 0.4081621468067169, |
| "learning_rate": 1.9991993813020483e-05, |
| "loss": 0.7194, |
| "step": 10190 |
| }, |
| { |
| "epoch": 0.8762886597938144, |
| "grad_norm": 0.3913891911506653, |
| "learning_rate": 1.9974946373028628e-05, |
| "loss": 0.7096, |
| "step": 10200 |
| }, |
| { |
| "epoch": 0.877147766323024, |
| "grad_norm": 0.3861335515975952, |
| "learning_rate": 1.994844749501764e-05, |
| "loss": 0.6947, |
| "step": 10210 |
| }, |
| { |
| "epoch": 0.8780068728522337, |
| "grad_norm": 0.41753047704696655, |
| "learning_rate": 1.9912522413632988e-05, |
| "loss": 0.7226, |
| "step": 10220 |
| }, |
| { |
| "epoch": 0.8788659793814433, |
| "grad_norm": 0.3903125524520874, |
| "learning_rate": 1.9867205340009656e-05, |
| "loss": 0.7161, |
| "step": 10230 |
| }, |
| { |
| "epoch": 0.8797250859106529, |
| "grad_norm": 0.3661431074142456, |
| "learning_rate": 1.9812539429193164e-05, |
| "loss": 0.7171, |
| "step": 10240 |
| }, |
| { |
| "epoch": 0.8805841924398625, |
| "grad_norm": 0.37936529517173767, |
| "learning_rate": 1.9748576739043423e-05, |
| "loss": 0.6966, |
| "step": 10250 |
| }, |
| { |
| "epoch": 0.8814432989690721, |
| "grad_norm": 0.3925243616104126, |
| "learning_rate": 1.967537818066047e-05, |
| "loss": 0.7328, |
| "step": 10260 |
| }, |
| { |
| "epoch": 0.8823024054982818, |
| "grad_norm": 0.38311198353767395, |
| "learning_rate": 1.959301346037941e-05, |
| "loss": 0.705, |
| "step": 10270 |
| }, |
| { |
| "epoch": 0.8831615120274914, |
| "grad_norm": 0.399787575006485, |
| "learning_rate": 1.9501561013389646e-05, |
| "loss": 0.7302, |
| "step": 10280 |
| }, |
| { |
| "epoch": 0.884020618556701, |
| "grad_norm": 0.34494566917419434, |
| "learning_rate": 1.9401107929041795e-05, |
| "loss": 0.6865, |
| "step": 10290 |
| }, |
| { |
| "epoch": 0.8848797250859106, |
| "grad_norm": 0.38984084129333496, |
| "learning_rate": 1.9291749867913225e-05, |
| "loss": 0.712, |
| "step": 10300 |
| }, |
| { |
| "epoch": 0.8857388316151202, |
| "grad_norm": 0.39119017124176025, |
| "learning_rate": 1.91735909707114e-05, |
| "loss": 0.7328, |
| "step": 10310 |
| }, |
| { |
| "epoch": 0.8865979381443299, |
| "grad_norm": 0.36477139592170715, |
| "learning_rate": 1.9046743759101545e-05, |
| "loss": 0.714, |
| "step": 10320 |
| }, |
| { |
| "epoch": 0.8874570446735395, |
| "grad_norm": 0.4393681585788727, |
| "learning_rate": 1.8911329028553324e-05, |
| "loss": 0.7358, |
| "step": 10330 |
| }, |
| { |
| "epoch": 0.8883161512027491, |
| "grad_norm": 0.37272876501083374, |
| "learning_rate": 1.876747573330835e-05, |
| "loss": 0.7252, |
| "step": 10340 |
| }, |
| { |
| "epoch": 0.8891752577319587, |
| "grad_norm": 0.4037761688232422, |
| "learning_rate": 1.8615320863578244e-05, |
| "loss": 0.7157, |
| "step": 10350 |
| }, |
| { |
| "epoch": 0.8900343642611683, |
| "grad_norm": 0.4025321304798126, |
| "learning_rate": 1.8455009315090036e-05, |
| "loss": 0.7148, |
| "step": 10360 |
| }, |
| { |
| "epoch": 0.8908934707903781, |
| "grad_norm": 0.3650362491607666, |
| "learning_rate": 1.8286693751103294e-05, |
| "loss": 0.7141, |
| "step": 10370 |
| }, |
| { |
| "epoch": 0.8917525773195877, |
| "grad_norm": 0.38585397601127625, |
| "learning_rate": 1.811053445703024e-05, |
| "loss": 0.6893, |
| "step": 10380 |
| }, |
| { |
| "epoch": 0.8926116838487973, |
| "grad_norm": 0.3852981626987457, |
| "learning_rate": 1.7926699187797375e-05, |
| "loss": 0.7053, |
| "step": 10390 |
| }, |
| { |
| "epoch": 0.8934707903780069, |
| "grad_norm": 0.38907700777053833, |
| "learning_rate": 1.7735363008093995e-05, |
| "loss": 0.7046, |
| "step": 10400 |
| }, |
| { |
| "epoch": 0.8943298969072165, |
| "grad_norm": 0.37354832887649536, |
| "learning_rate": 1.7536708125659577e-05, |
| "loss": 0.715, |
| "step": 10410 |
| }, |
| { |
| "epoch": 0.8951890034364262, |
| "grad_norm": 0.39108115434646606, |
| "learning_rate": 1.733092371776905e-05, |
| "loss": 0.6932, |
| "step": 10420 |
| }, |
| { |
| "epoch": 0.8960481099656358, |
| "grad_norm": 0.38209742307662964, |
| "learning_rate": 1.7118205751080915e-05, |
| "loss": 0.6993, |
| "step": 10430 |
| }, |
| { |
| "epoch": 0.8969072164948454, |
| "grad_norm": 0.37446147203445435, |
| "learning_rate": 1.6898756795019935e-05, |
| "loss": 0.7078, |
| "step": 10440 |
| }, |
| { |
| "epoch": 0.897766323024055, |
| "grad_norm": 0.3764054775238037, |
| "learning_rate": 1.667278582887211e-05, |
| "loss": 0.7259, |
| "step": 10450 |
| }, |
| { |
| "epoch": 0.8986254295532646, |
| "grad_norm": 0.3835497200489044, |
| "learning_rate": 1.6440508042775545e-05, |
| "loss": 0.7292, |
| "step": 10460 |
| }, |
| { |
| "epoch": 0.8994845360824743, |
| "grad_norm": 0.3666471242904663, |
| "learning_rate": 1.6202144632796796e-05, |
| "loss": 0.6947, |
| "step": 10470 |
| }, |
| { |
| "epoch": 0.9003436426116839, |
| "grad_norm": 0.39887842535972595, |
| "learning_rate": 1.595792259028783e-05, |
| "loss": 0.7027, |
| "step": 10480 |
| }, |
| { |
| "epoch": 0.9012027491408935, |
| "grad_norm": 0.3692478537559509, |
| "learning_rate": 1.570807448572419e-05, |
| "loss": 0.7338, |
| "step": 10490 |
| }, |
| { |
| "epoch": 0.9020618556701031, |
| "grad_norm": 0.3501831889152527, |
| "learning_rate": 1.5452838247230197e-05, |
| "loss": 0.7056, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.9029209621993127, |
| "grad_norm": 0.43365010619163513, |
| "learning_rate": 1.5192456934002106e-05, |
| "loss": 0.697, |
| "step": 10510 |
| }, |
| { |
| "epoch": 0.9037800687285223, |
| "grad_norm": 0.3910599648952484, |
| "learning_rate": 1.492717850484502e-05, |
| "loss": 0.704, |
| "step": 10520 |
| }, |
| { |
| "epoch": 0.904639175257732, |
| "grad_norm": 0.36650311946868896, |
| "learning_rate": 1.4657255582043904e-05, |
| "loss": 0.6939, |
| "step": 10530 |
| }, |
| { |
| "epoch": 0.9054982817869416, |
| "grad_norm": 0.3840698003768921, |
| "learning_rate": 1.4382945210793663e-05, |
| "loss": 0.6995, |
| "step": 10540 |
| }, |
| { |
| "epoch": 0.9063573883161512, |
| "grad_norm": 0.37365975975990295, |
| "learning_rate": 1.4104508614417267e-05, |
| "loss": 0.7111, |
| "step": 10550 |
| }, |
| { |
| "epoch": 0.9072164948453608, |
| "grad_norm": 0.38615456223487854, |
| "learning_rate": 1.382221094560512e-05, |
| "loss": 0.6918, |
| "step": 10560 |
| }, |
| { |
| "epoch": 0.9080756013745704, |
| "grad_norm": 0.4073936343193054, |
| "learning_rate": 1.3536321033912506e-05, |
| "loss": 0.6825, |
| "step": 10570 |
| }, |
| { |
| "epoch": 0.9089347079037801, |
| "grad_norm": 0.39948225021362305, |
| "learning_rate": 1.3247111129755609e-05, |
| "loss": 0.6935, |
| "step": 10580 |
| }, |
| { |
| "epoch": 0.9097938144329897, |
| "grad_norm": 0.34603285789489746, |
| "learning_rate": 1.2954856645149868e-05, |
| "loss": 0.6897, |
| "step": 10590 |
| }, |
| { |
| "epoch": 0.9106529209621993, |
| "grad_norm": 0.41271108388900757, |
| "learning_rate": 1.2659835891437553e-05, |
| "loss": 0.7212, |
| "step": 10600 |
| }, |
| { |
| "epoch": 0.9115120274914089, |
| "grad_norm": 0.3676590323448181, |
| "learning_rate": 1.236232981425436e-05, |
| "loss": 0.6665, |
| "step": 10610 |
| }, |
| { |
| "epoch": 0.9123711340206185, |
| "grad_norm": 0.3909376263618469, |
| "learning_rate": 1.2062621725987402e-05, |
| "loss": 0.6965, |
| "step": 10620 |
| }, |
| { |
| "epoch": 0.9132302405498282, |
| "grad_norm": 0.37691184878349304, |
| "learning_rate": 1.1760997035979318e-05, |
| "loss": 0.7135, |
| "step": 10630 |
| }, |
| { |
| "epoch": 0.9140893470790378, |
| "grad_norm": 0.3623291552066803, |
| "learning_rate": 1.1457742978735514e-05, |
| "loss": 0.6911, |
| "step": 10640 |
| }, |
| { |
| "epoch": 0.9149484536082474, |
| "grad_norm": 0.37281203269958496, |
| "learning_rate": 1.1153148340393298e-05, |
| "loss": 0.7076, |
| "step": 10650 |
| }, |
| { |
| "epoch": 0.915807560137457, |
| "grad_norm": 0.39836084842681885, |
| "learning_rate": 1.0847503183713424e-05, |
| "loss": 0.6954, |
| "step": 10660 |
| }, |
| { |
| "epoch": 0.9166666666666666, |
| "grad_norm": 0.366367906332016, |
| "learning_rate": 1.0541098571855896e-05, |
| "loss": 0.6964, |
| "step": 10670 |
| }, |
| { |
| "epoch": 0.9175257731958762, |
| "grad_norm": 0.41464763879776, |
| "learning_rate": 1.0234226291203123e-05, |
| "loss": 0.6999, |
| "step": 10680 |
| }, |
| { |
| "epoch": 0.9183848797250859, |
| "grad_norm": 0.3979295492172241, |
| "learning_rate": 9.927178573494333e-06, |
| "loss": 0.6972, |
| "step": 10690 |
| }, |
| { |
| "epoch": 0.9192439862542955, |
| "grad_norm": 0.3682306706905365, |
| "learning_rate": 9.620247817535902e-06, |
| "loss": 0.7077, |
| "step": 10700 |
| }, |
| { |
| "epoch": 0.9201030927835051, |
| "grad_norm": 0.37287646532058716, |
| "learning_rate": 9.313726310752581e-06, |
| "loss": 0.6911, |
| "step": 10710 |
| }, |
| { |
| "epoch": 0.9209621993127147, |
| "grad_norm": 0.3430989980697632, |
| "learning_rate": 9.00790595084477e-06, |
| "loss": 0.6726, |
| "step": 10720 |
| }, |
| { |
| "epoch": 0.9218213058419243, |
| "grad_norm": 0.3784588873386383, |
| "learning_rate": 8.703077967816944e-06, |
| "loss": 0.6851, |
| "step": 10730 |
| }, |
| { |
| "epoch": 0.9226804123711341, |
| "grad_norm": 0.3933582305908203, |
| "learning_rate": 8.39953264664192e-06, |
| "loss": 0.6879, |
| "step": 10740 |
| }, |
| { |
| "epoch": 0.9235395189003437, |
| "grad_norm": 0.4040727913379669, |
| "learning_rate": 8.097559050825058e-06, |
| "loss": 0.7242, |
| "step": 10750 |
| }, |
| { |
| "epoch": 0.9243986254295533, |
| "grad_norm": 0.39874258637428284, |
| "learning_rate": 7.79744474713166e-06, |
| "loss": 0.7017, |
| "step": 10760 |
| }, |
| { |
| "epoch": 0.9252577319587629, |
| "grad_norm": 0.4554367959499359, |
| "learning_rate": 7.499475531739696e-06, |
| "loss": 0.7183, |
| "step": 10770 |
| }, |
| { |
| "epoch": 0.9261168384879725, |
| "grad_norm": 0.38489654660224915, |
| "learning_rate": 7.203935158078662e-06, |
| "loss": 0.6875, |
| "step": 10780 |
| }, |
| { |
| "epoch": 0.9269759450171822, |
| "grad_norm": 0.3875371813774109, |
| "learning_rate": 6.911105066613709e-06, |
| "loss": 0.6965, |
| "step": 10790 |
| }, |
| { |
| "epoch": 0.9278350515463918, |
| "grad_norm": 0.3381649851799011, |
| "learning_rate": 6.621264116832394e-06, |
| "loss": 0.7113, |
| "step": 10800 |
| }, |
| { |
| "epoch": 0.9286941580756014, |
| "grad_norm": 0.36320725083351135, |
| "learning_rate": 6.334688321689274e-06, |
| "loss": 0.7228, |
| "step": 10810 |
| }, |
| { |
| "epoch": 0.929553264604811, |
| "grad_norm": 0.38542211055755615, |
| "learning_rate": 6.051650584761237e-06, |
| "loss": 0.7278, |
| "step": 10820 |
| }, |
| { |
| "epoch": 0.9304123711340206, |
| "grad_norm": 0.3692699372768402, |
| "learning_rate": 5.772420440363841e-06, |
| "loss": 0.7095, |
| "step": 10830 |
| }, |
| { |
| "epoch": 0.9312714776632303, |
| "grad_norm": 0.35440167784690857, |
| "learning_rate": 5.497263796876207e-06, |
| "loss": 0.6949, |
| "step": 10840 |
| }, |
| { |
| "epoch": 0.9321305841924399, |
| "grad_norm": 0.42387256026268005, |
| "learning_rate": 5.226442683518813e-06, |
| "loss": 0.7229, |
| "step": 10850 |
| }, |
| { |
| "epoch": 0.9329896907216495, |
| "grad_norm": 0.3378446102142334, |
| "learning_rate": 4.960215000825432e-06, |
| "loss": 0.654, |
| "step": 10860 |
| }, |
| { |
| "epoch": 0.9338487972508591, |
| "grad_norm": 0.36746713519096375, |
| "learning_rate": 4.698834275046711e-06, |
| "loss": 0.6913, |
| "step": 10870 |
| }, |
| { |
| "epoch": 0.9347079037800687, |
| "grad_norm": 0.3544543981552124, |
| "learning_rate": 4.442549416719423e-06, |
| "loss": 0.6993, |
| "step": 10880 |
| }, |
| { |
| "epoch": 0.9355670103092784, |
| "grad_norm": 0.39454489946365356, |
| "learning_rate": 4.1916044836311565e-06, |
| "loss": 0.6866, |
| "step": 10890 |
| }, |
| { |
| "epoch": 0.936426116838488, |
| "grad_norm": 0.35111212730407715, |
| "learning_rate": 3.94623844840626e-06, |
| "loss": 0.6997, |
| "step": 10900 |
| }, |
| { |
| "epoch": 0.9372852233676976, |
| "grad_norm": 0.365174263715744, |
| "learning_rate": 3.7066849709343467e-06, |
| "loss": 0.7107, |
| "step": 10910 |
| }, |
| { |
| "epoch": 0.9381443298969072, |
| "grad_norm": 0.3847159743309021, |
| "learning_rate": 3.473172175858059e-06, |
| "loss": 0.6745, |
| "step": 10920 |
| }, |
| { |
| "epoch": 0.9390034364261168, |
| "grad_norm": 0.3644939959049225, |
| "learning_rate": 3.2459224353320046e-06, |
| "loss": 0.7077, |
| "step": 10930 |
| }, |
| { |
| "epoch": 0.9398625429553265, |
| "grad_norm": 0.42002615332603455, |
| "learning_rate": 3.025152157259735e-06, |
| "loss": 0.6616, |
| "step": 10940 |
| }, |
| { |
| "epoch": 0.9407216494845361, |
| "grad_norm": 0.38049033284187317, |
| "learning_rate": 2.811071579210425e-06, |
| "loss": 0.7036, |
| "step": 10950 |
| }, |
| { |
| "epoch": 0.9415807560137457, |
| "grad_norm": 0.3963865637779236, |
| "learning_rate": 2.603884568211513e-06, |
| "loss": 0.6817, |
| "step": 10960 |
| }, |
| { |
| "epoch": 0.9424398625429553, |
| "grad_norm": 0.42913562059402466, |
| "learning_rate": 2.4037884266079434e-06, |
| "loss": 0.7018, |
| "step": 10970 |
| }, |
| { |
| "epoch": 0.9432989690721649, |
| "grad_norm": 0.36684727668762207, |
| "learning_rate": 2.2109737041728955e-06, |
| "loss": 0.6731, |
| "step": 10980 |
| }, |
| { |
| "epoch": 0.9441580756013745, |
| "grad_norm": 0.38636747002601624, |
| "learning_rate": 2.0256240166489454e-06, |
| "loss": 0.689, |
| "step": 10990 |
| }, |
| { |
| "epoch": 0.9450171821305842, |
| "grad_norm": 0.37258586287498474, |
| "learning_rate": 1.847915870892413e-06, |
| "loss": 0.7065, |
| "step": 11000 |
| }, |
| { |
| "epoch": 0.9458762886597938, |
| "grad_norm": 0.36746594309806824, |
| "learning_rate": 1.6780184967874715e-06, |
| "loss": 0.6836, |
| "step": 11010 |
| }, |
| { |
| "epoch": 0.9467353951890034, |
| "grad_norm": 0.3938833773136139, |
| "learning_rate": 1.5160936860900294e-06, |
| "loss": 0.6989, |
| "step": 11020 |
| }, |
| { |
| "epoch": 0.947594501718213, |
| "grad_norm": 0.3353445529937744, |
| "learning_rate": 1.3622956383548762e-06, |
| "loss": 0.6621, |
| "step": 11030 |
| }, |
| { |
| "epoch": 0.9484536082474226, |
| "grad_norm": 0.37192410230636597, |
| "learning_rate": 1.216770814092834e-06, |
| "loss": 0.7006, |
| "step": 11040 |
| }, |
| { |
| "epoch": 0.9493127147766323, |
| "grad_norm": 0.3724251389503479, |
| "learning_rate": 1.0796577952977013e-06, |
| "loss": 0.6869, |
| "step": 11050 |
| }, |
| { |
| "epoch": 0.9501718213058419, |
| "grad_norm": 0.3747676610946655, |
| "learning_rate": 9.510871534758705e-07, |
| "loss": 0.7193, |
| "step": 11060 |
| }, |
| { |
| "epoch": 0.9510309278350515, |
| "grad_norm": 0.3892313241958618, |
| "learning_rate": 8.311813253042334e-07, |
| "loss": 0.6928, |
| "step": 11070 |
| }, |
| { |
| "epoch": 0.9518900343642611, |
| "grad_norm": 0.37105125188827515, |
| "learning_rate": 7.200544960348308e-07, |
| "loss": 0.6711, |
| "step": 11080 |
| }, |
| { |
| "epoch": 0.9527491408934707, |
| "grad_norm": 0.3668333888053894, |
| "learning_rate": 6.178124907572545e-07, |
| "loss": 0.6976, |
| "step": 11090 |
| }, |
| { |
| "epoch": 0.9536082474226805, |
| "grad_norm": 0.3635807931423187, |
| "learning_rate": 5.245526736223528e-07, |
| "loss": 0.6961, |
| "step": 11100 |
| }, |
| { |
| "epoch": 0.9544673539518901, |
| "grad_norm": 0.40923115611076355, |
| "learning_rate": 4.403638551232205e-07, |
| "loss": 0.6926, |
| "step": 11110 |
| }, |
| { |
| "epoch": 0.9553264604810997, |
| "grad_norm": 0.38920608162879944, |
| "learning_rate": 3.6532620752175616e-07, |
| "loss": 0.6951, |
| "step": 11120 |
| }, |
| { |
| "epoch": 0.9561855670103093, |
| "grad_norm": 0.3715786039829254, |
| "learning_rate": 2.9951118850133247e-07, |
| "loss": 0.67, |
| "step": 11130 |
| }, |
| { |
| "epoch": 0.9570446735395189, |
| "grad_norm": 0.37026530504226685, |
| "learning_rate": 2.429814731182774e-07, |
| "loss": 0.6999, |
| "step": 11140 |
| }, |
| { |
| "epoch": 0.9579037800687286, |
| "grad_norm": 0.38067036867141724, |
| "learning_rate": 1.957908941169813e-07, |
| "loss": 0.6833, |
| "step": 11150 |
| }, |
| { |
| "epoch": 0.9587628865979382, |
| "grad_norm": 0.3830522298812866, |
| "learning_rate": 1.5798439066545254e-07, |
| "loss": 0.6796, |
| "step": 11160 |
| }, |
| { |
| "epoch": 0.9596219931271478, |
| "grad_norm": 0.36072632670402527, |
| "learning_rate": 1.2959796556014471e-07, |
| "loss": 0.6999, |
| "step": 11170 |
| }, |
| { |
| "epoch": 0.9604810996563574, |
| "grad_norm": 0.34916412830352783, |
| "learning_rate": 1.1065865094081884e-07, |
| "loss": 0.6982, |
| "step": 11180 |
| }, |
| { |
| "epoch": 0.961340206185567, |
| "grad_norm": 0.3502258360385895, |
| "learning_rate": 1.011844825480734e-07, |
| "loss": 0.6721, |
| "step": 11190 |
| }, |
| { |
| "epoch": 0.9616838487972509, |
| "step": 11194, |
| "total_flos": 4.547514943786923e+19, |
| "train_loss": 0.003869807505066616, |
| "train_runtime": 609.7819, |
| "train_samples_per_second": 220.289, |
| "train_steps_per_second": 18.357 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 11194, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 159, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 4.547514943786923e+19, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
