| { | |
| "best_metric": 1.7225215435028076, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-500", | |
| "epoch": 0.09961647656522389, | |
| "eval_steps": 100, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00019923295313044778, | |
| "grad_norm": 1.2004849910736084, | |
| "learning_rate": 5e-06, | |
| "loss": 2.9446, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00019923295313044778, | |
| "eval_loss": 4.056490898132324, | |
| "eval_runtime": 1316.2296, | |
| "eval_samples_per_second": 6.423, | |
| "eval_steps_per_second": 1.606, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00039846590626089557, | |
| "grad_norm": 1.6586755514144897, | |
| "learning_rate": 1e-05, | |
| "loss": 3.5137, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0005976988593913434, | |
| "grad_norm": 1.1276801824569702, | |
| "learning_rate": 1.5e-05, | |
| "loss": 2.5317, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0007969318125217911, | |
| "grad_norm": 1.4570177793502808, | |
| "learning_rate": 2e-05, | |
| "loss": 3.2148, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0009961647656522388, | |
| "grad_norm": 1.695500135421753, | |
| "learning_rate": 2.5e-05, | |
| "loss": 3.2779, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0011953977187826868, | |
| "grad_norm": 1.5533411502838135, | |
| "learning_rate": 3e-05, | |
| "loss": 3.4182, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0013946306719131345, | |
| "grad_norm": 1.5567388534545898, | |
| "learning_rate": 3.5e-05, | |
| "loss": 3.0221, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0015938636250435823, | |
| "grad_norm": 1.6326138973236084, | |
| "learning_rate": 4e-05, | |
| "loss": 3.1334, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00179309657817403, | |
| "grad_norm": 1.7711933851242065, | |
| "learning_rate": 4.5e-05, | |
| "loss": 3.0372, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0019923295313044776, | |
| "grad_norm": 1.850273609161377, | |
| "learning_rate": 5e-05, | |
| "loss": 2.891, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0021915624844349255, | |
| "grad_norm": 1.782905101776123, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.7233, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0023907954375653735, | |
| "grad_norm": 1.5795241594314575, | |
| "learning_rate": 6e-05, | |
| "loss": 2.9004, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.002590028390695821, | |
| "grad_norm": 1.513824224472046, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 2.299, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.002789261343826269, | |
| "grad_norm": 1.6560556888580322, | |
| "learning_rate": 7e-05, | |
| "loss": 2.1206, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0029884942969567166, | |
| "grad_norm": 1.761472463607788, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.3681, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0031877272500871645, | |
| "grad_norm": 2.254106044769287, | |
| "learning_rate": 8e-05, | |
| "loss": 2.3073, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.003386960203217612, | |
| "grad_norm": 1.6817950010299683, | |
| "learning_rate": 8.5e-05, | |
| "loss": 2.1039, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.00358619315634806, | |
| "grad_norm": 1.5102214813232422, | |
| "learning_rate": 9e-05, | |
| "loss": 2.1165, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0037854261094785076, | |
| "grad_norm": 1.2089601755142212, | |
| "learning_rate": 9.5e-05, | |
| "loss": 2.0751, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.003984659062608955, | |
| "grad_norm": 1.210183024406433, | |
| "learning_rate": 0.0001, | |
| "loss": 1.8907, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0041838920157394035, | |
| "grad_norm": 1.2759207487106323, | |
| "learning_rate": 9.999892908320647e-05, | |
| "loss": 1.8462, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.004383124968869851, | |
| "grad_norm": 3.2929604053497314, | |
| "learning_rate": 9.999571637870036e-05, | |
| "loss": 1.864, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.004582357922000299, | |
| "grad_norm": 1.1962004899978638, | |
| "learning_rate": 9.999036202410325e-05, | |
| "loss": 1.7009, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.004781590875130747, | |
| "grad_norm": 1.1904807090759277, | |
| "learning_rate": 9.998286624877786e-05, | |
| "loss": 2.0888, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0049808238282611945, | |
| "grad_norm": 3.4276819229125977, | |
| "learning_rate": 9.997322937381829e-05, | |
| "loss": 1.858, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.005180056781391642, | |
| "grad_norm": 1.3617855310440063, | |
| "learning_rate": 9.996145181203615e-05, | |
| "loss": 2.1621, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.00537928973452209, | |
| "grad_norm": 1.9927242994308472, | |
| "learning_rate": 9.994753406794301e-05, | |
| "loss": 1.8291, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.005578522687652538, | |
| "grad_norm": 1.01543128490448, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 1.9078, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.005777755640782986, | |
| "grad_norm": 1.1790841817855835, | |
| "learning_rate": 9.991328050923581e-05, | |
| "loss": 1.8481, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.005976988593913433, | |
| "grad_norm": 1.0684229135513306, | |
| "learning_rate": 9.989294616193017e-05, | |
| "loss": 1.9119, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.006176221547043881, | |
| "grad_norm": 1.042448878288269, | |
| "learning_rate": 9.98704745668676e-05, | |
| "loss": 1.8315, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.006375454500174329, | |
| "grad_norm": 1.063533902168274, | |
| "learning_rate": 9.98458666866564e-05, | |
| "loss": 1.9271, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.006574687453304777, | |
| "grad_norm": 0.9447396993637085, | |
| "learning_rate": 9.981912357541627e-05, | |
| "loss": 1.9347, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.006773920406435224, | |
| "grad_norm": 1.0030559301376343, | |
| "learning_rate": 9.97902463787331e-05, | |
| "loss": 2.1377, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0069731533595656725, | |
| "grad_norm": 0.9730034470558167, | |
| "learning_rate": 9.975923633360985e-05, | |
| "loss": 1.869, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.00717238631269612, | |
| "grad_norm": 0.9906373023986816, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 2.0613, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.007371619265826568, | |
| "grad_norm": 1.1251929998397827, | |
| "learning_rate": 9.969082310281891e-05, | |
| "loss": 2.3343, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.007570852218957015, | |
| "grad_norm": 1.1643787622451782, | |
| "learning_rate": 9.965342284774632e-05, | |
| "loss": 2.0388, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0077700851720874636, | |
| "grad_norm": 1.09977388381958, | |
| "learning_rate": 9.961389560529836e-05, | |
| "loss": 2.2025, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.00796931812521791, | |
| "grad_norm": 1.0421956777572632, | |
| "learning_rate": 9.957224306869053e-05, | |
| "loss": 1.9667, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.008168551078348359, | |
| "grad_norm": 1.134215235710144, | |
| "learning_rate": 9.952846702217886e-05, | |
| "loss": 2.1981, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.008367784031478807, | |
| "grad_norm": 1.1163185834884644, | |
| "learning_rate": 9.948256934098352e-05, | |
| "loss": 2.1145, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.008567016984609254, | |
| "grad_norm": 1.0088622570037842, | |
| "learning_rate": 9.943455199120837e-05, | |
| "loss": 2.0563, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.008766249937739702, | |
| "grad_norm": 1.0823618173599243, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 1.9748, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.00896548289087015, | |
| "grad_norm": 1.0469329357147217, | |
| "learning_rate": 9.933216660424395e-05, | |
| "loss": 2.1169, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.009164715844000597, | |
| "grad_norm": 1.0795198678970337, | |
| "learning_rate": 9.927780295290389e-05, | |
| "loss": 2.0546, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.009363948797131046, | |
| "grad_norm": 1.2332130670547485, | |
| "learning_rate": 9.922132840449459e-05, | |
| "loss": 2.3422, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.009563181750261494, | |
| "grad_norm": 1.0598793029785156, | |
| "learning_rate": 9.916274537819775e-05, | |
| "loss": 2.0388, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.00976241470339194, | |
| "grad_norm": 1.042548418045044, | |
| "learning_rate": 9.91020563835152e-05, | |
| "loss": 2.0313, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.009961647656522389, | |
| "grad_norm": 1.0272207260131836, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 2.1789, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.010160880609652836, | |
| "grad_norm": 1.0316256284713745, | |
| "learning_rate": 9.897437097795257e-05, | |
| "loss": 1.6957, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.010360113562783284, | |
| "grad_norm": 0.9458410739898682, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.5179, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.010559346515913733, | |
| "grad_norm": 0.9019896388053894, | |
| "learning_rate": 9.883829406604363e-05, | |
| "loss": 1.8115, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.01075857946904418, | |
| "grad_norm": 0.7449951767921448, | |
| "learning_rate": 9.876711602542563e-05, | |
| "loss": 1.4849, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.010957812422174628, | |
| "grad_norm": 0.690968930721283, | |
| "learning_rate": 9.869384896386668e-05, | |
| "loss": 1.4004, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.011157045375305076, | |
| "grad_norm": 0.6796329021453857, | |
| "learning_rate": 9.861849601988383e-05, | |
| "loss": 1.1753, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.011356278328435523, | |
| "grad_norm": 0.6940993070602417, | |
| "learning_rate": 9.854106042134641e-05, | |
| "loss": 1.3674, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.011555511281565971, | |
| "grad_norm": 0.7557864189147949, | |
| "learning_rate": 9.846154548533773e-05, | |
| "loss": 1.4927, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.01175474423469642, | |
| "grad_norm": 0.7977707982063293, | |
| "learning_rate": 9.837995461801299e-05, | |
| "loss": 1.4762, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.011953977187826866, | |
| "grad_norm": 0.889670193195343, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 1.7018, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.012153210140957315, | |
| "grad_norm": 0.8150140047073364, | |
| "learning_rate": 9.821055915851647e-05, | |
| "loss": 1.5165, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.012352443094087761, | |
| "grad_norm": 0.8231312036514282, | |
| "learning_rate": 9.812276182268236e-05, | |
| "loss": 1.8683, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.01255167604721821, | |
| "grad_norm": 0.7673140168190002, | |
| "learning_rate": 9.803290306789676e-05, | |
| "loss": 1.483, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.012750909000348658, | |
| "grad_norm": 0.8216698169708252, | |
| "learning_rate": 9.794098674340965e-05, | |
| "loss": 1.6038, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.012950141953479105, | |
| "grad_norm": 0.7474381327629089, | |
| "learning_rate": 9.784701678661045e-05, | |
| "loss": 1.4957, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.013149374906609553, | |
| "grad_norm": 0.7596239447593689, | |
| "learning_rate": 9.775099722285935e-05, | |
| "loss": 1.7015, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.013348607859740002, | |
| "grad_norm": 0.6952986717224121, | |
| "learning_rate": 9.765293216531486e-05, | |
| "loss": 1.5342, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.013547840812870448, | |
| "grad_norm": 0.9067381024360657, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 1.8279, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.013747073766000897, | |
| "grad_norm": 0.8189777135848999, | |
| "learning_rate": 9.74506824594107e-05, | |
| "loss": 1.5812, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.013946306719131345, | |
| "grad_norm": 0.8380293846130371, | |
| "learning_rate": 9.73465064747553e-05, | |
| "loss": 1.7342, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.014145539672261792, | |
| "grad_norm": 0.8201037645339966, | |
| "learning_rate": 9.724030232334391e-05, | |
| "loss": 1.6843, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.01434477262539224, | |
| "grad_norm": 0.971545398235321, | |
| "learning_rate": 9.713207455460894e-05, | |
| "loss": 1.7146, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.014544005578522687, | |
| "grad_norm": 1.0521457195281982, | |
| "learning_rate": 9.702182780466775e-05, | |
| "loss": 1.8588, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.014743238531653135, | |
| "grad_norm": 0.8414843082427979, | |
| "learning_rate": 9.690956679612421e-05, | |
| "loss": 1.7867, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.014942471484783584, | |
| "grad_norm": 0.9837526082992554, | |
| "learning_rate": 9.67952963378663e-05, | |
| "loss": 1.8164, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.01514170443791403, | |
| "grad_norm": 0.7853424549102783, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 1.5892, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.015340937391044479, | |
| "grad_norm": 0.8817009925842285, | |
| "learning_rate": 9.656074673794018e-05, | |
| "loss": 1.6805, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.015540170344174927, | |
| "grad_norm": 0.8272745609283447, | |
| "learning_rate": 9.644047764359622e-05, | |
| "loss": 1.6335, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.015739403297305374, | |
| "grad_norm": 1.0323444604873657, | |
| "learning_rate": 9.631821919375591e-05, | |
| "loss": 1.9216, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.01593863625043582, | |
| "grad_norm": 0.7793384790420532, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 1.6618, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.01613786920356627, | |
| "grad_norm": 0.8447017073631287, | |
| "learning_rate": 9.606775526115963e-05, | |
| "loss": 1.934, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.016337102156696717, | |
| "grad_norm": 0.7522584199905396, | |
| "learning_rate": 9.593956050744492e-05, | |
| "loss": 1.8898, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.016536335109827164, | |
| "grad_norm": 0.971231997013092, | |
| "learning_rate": 9.580939785585681e-05, | |
| "loss": 1.9143, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.016735568062957614, | |
| "grad_norm": 0.8041723966598511, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.7362, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.01693480101608806, | |
| "grad_norm": 0.8742603659629822, | |
| "learning_rate": 9.554319124605879e-05, | |
| "loss": 2.0349, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.017134033969218507, | |
| "grad_norm": 0.8700699210166931, | |
| "learning_rate": 9.540715869125407e-05, | |
| "loss": 1.7798, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.017333266922348958, | |
| "grad_norm": 1.0314463376998901, | |
| "learning_rate": 9.526918104489777e-05, | |
| "loss": 1.9499, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.017532499875479404, | |
| "grad_norm": 0.9694502949714661, | |
| "learning_rate": 9.512926421749304e-05, | |
| "loss": 2.1316, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.01773173282860985, | |
| "grad_norm": 0.9032630920410156, | |
| "learning_rate": 9.498741420261108e-05, | |
| "loss": 2.1217, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.0179309657817403, | |
| "grad_norm": 0.9292871356010437, | |
| "learning_rate": 9.484363707663442e-05, | |
| "loss": 1.878, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.018130198734870748, | |
| "grad_norm": 0.9102398157119751, | |
| "learning_rate": 9.469793899849661e-05, | |
| "loss": 1.8715, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.018329431688001194, | |
| "grad_norm": 1.0135326385498047, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 2.2684, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.018528664641131645, | |
| "grad_norm": 0.9862831830978394, | |
| "learning_rate": 9.440080503264037e-05, | |
| "loss": 2.1933, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.01872789759426209, | |
| "grad_norm": 0.8928501009941101, | |
| "learning_rate": 9.42493818731521e-05, | |
| "loss": 1.9306, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.018927130547392538, | |
| "grad_norm": 1.0035126209259033, | |
| "learning_rate": 9.409606321741775e-05, | |
| "loss": 2.1338, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.019126363500522988, | |
| "grad_norm": 1.0360318422317505, | |
| "learning_rate": 9.394085563309827e-05, | |
| "loss": 2.0244, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.019325596453653435, | |
| "grad_norm": 1.0765694379806519, | |
| "learning_rate": 9.378376576876999e-05, | |
| "loss": 2.0452, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.01952482940678388, | |
| "grad_norm": 1.0682713985443115, | |
| "learning_rate": 9.362480035363986e-05, | |
| "loss": 1.9638, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01972406235991433, | |
| "grad_norm": 0.9798291325569153, | |
| "learning_rate": 9.34639661972572e-05, | |
| "loss": 1.8321, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.019923295313044778, | |
| "grad_norm": 1.2100462913513184, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 2.1887, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.019923295313044778, | |
| "eval_loss": 1.8207975625991821, | |
| "eval_runtime": 1307.2968, | |
| "eval_samples_per_second": 6.467, | |
| "eval_steps_per_second": 1.617, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.020122528266175225, | |
| "grad_norm": 0.8175336718559265, | |
| "learning_rate": 9.31367192988896e-05, | |
| "loss": 1.3173, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.02032176121930567, | |
| "grad_norm": 0.9728506803512573, | |
| "learning_rate": 9.297032057507264e-05, | |
| "loss": 1.8635, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.02052099417243612, | |
| "grad_norm": 0.7298185229301453, | |
| "learning_rate": 9.280208114573859e-05, | |
| "loss": 1.5464, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.02072022712556657, | |
| "grad_norm": 0.7254154086112976, | |
| "learning_rate": 9.263200821770461e-05, | |
| "loss": 1.3951, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.020919460078697015, | |
| "grad_norm": 0.8585533499717712, | |
| "learning_rate": 9.246010907632895e-05, | |
| "loss": 1.9375, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.021118693031827465, | |
| "grad_norm": 0.672656774520874, | |
| "learning_rate": 9.228639108519868e-05, | |
| "loss": 1.7358, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.021317925984957912, | |
| "grad_norm": 0.6700272560119629, | |
| "learning_rate": 9.211086168581433e-05, | |
| "loss": 1.373, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.02151715893808836, | |
| "grad_norm": 0.6741132736206055, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 1.472, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.02171639189121881, | |
| "grad_norm": 0.7094087600708008, | |
| "learning_rate": 9.175439881593716e-05, | |
| "loss": 1.4418, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.021915624844349255, | |
| "grad_norm": 0.6957412958145142, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 1.3553, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.022114857797479702, | |
| "grad_norm": 0.81883704662323, | |
| "learning_rate": 9.139078154477512e-05, | |
| "loss": 1.4875, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.022314090750610152, | |
| "grad_norm": 0.6015892624855042, | |
| "learning_rate": 9.120630943110077e-05, | |
| "loss": 1.3376, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0225133237037406, | |
| "grad_norm": 0.7218628525733948, | |
| "learning_rate": 9.102007217627568e-05, | |
| "loss": 1.5389, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.022712556656871045, | |
| "grad_norm": 0.6947839260101318, | |
| "learning_rate": 9.083207775808396e-05, | |
| "loss": 1.6291, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.022911789610001496, | |
| "grad_norm": 0.8209601044654846, | |
| "learning_rate": 9.064233422958077e-05, | |
| "loss": 1.5964, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.023111022563131942, | |
| "grad_norm": 0.7330240607261658, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.607, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.02331025551626239, | |
| "grad_norm": 0.6640767455101013, | |
| "learning_rate": 9.025763242814291e-05, | |
| "loss": 1.4805, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.02350948846939284, | |
| "grad_norm": 0.6990946531295776, | |
| "learning_rate": 9.006269063455304e-05, | |
| "loss": 1.3911, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.023708721422523286, | |
| "grad_norm": 0.6661984324455261, | |
| "learning_rate": 8.986603268863536e-05, | |
| "loss": 1.4391, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.023907954375653732, | |
| "grad_norm": 0.7727532386779785, | |
| "learning_rate": 8.966766701456177e-05, | |
| "loss": 1.6384, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.024107187328784183, | |
| "grad_norm": 0.8111420273780823, | |
| "learning_rate": 8.94676021096575e-05, | |
| "loss": 1.856, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.02430642028191463, | |
| "grad_norm": 0.7776418924331665, | |
| "learning_rate": 8.926584654403724e-05, | |
| "loss": 1.5784, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.024505653235045076, | |
| "grad_norm": 0.8065217733383179, | |
| "learning_rate": 8.906240896023794e-05, | |
| "loss": 1.6247, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.024704886188175523, | |
| "grad_norm": 0.7631449103355408, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 1.658, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.024904119141305973, | |
| "grad_norm": 0.8057156801223755, | |
| "learning_rate": 8.865052266813685e-05, | |
| "loss": 1.7705, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.02510335209443642, | |
| "grad_norm": 0.8324928879737854, | |
| "learning_rate": 8.844209160367299e-05, | |
| "loss": 1.7019, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.025302585047566866, | |
| "grad_norm": 0.9295965433120728, | |
| "learning_rate": 8.823201380795001e-05, | |
| "loss": 1.9604, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.025501818000697316, | |
| "grad_norm": 0.80018550157547, | |
| "learning_rate": 8.802029828000156e-05, | |
| "loss": 1.6104, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.025701050953827763, | |
| "grad_norm": 0.8858182430267334, | |
| "learning_rate": 8.780695408901613e-05, | |
| "loss": 1.805, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.02590028390695821, | |
| "grad_norm": 0.6917589902877808, | |
| "learning_rate": 8.759199037394887e-05, | |
| "loss": 1.5116, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02609951686008866, | |
| "grad_norm": 0.8768478631973267, | |
| "learning_rate": 8.737541634312985e-05, | |
| "loss": 1.9037, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.026298749813219106, | |
| "grad_norm": 0.7829615473747253, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 1.7264, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.026497982766349553, | |
| "grad_norm": 0.7925187349319458, | |
| "learning_rate": 8.693747451206232e-05, | |
| "loss": 1.7334, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.026697215719480003, | |
| "grad_norm": 0.821109414100647, | |
| "learning_rate": 8.671612547178428e-05, | |
| "loss": 1.8394, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.02689644867261045, | |
| "grad_norm": 0.8298285007476807, | |
| "learning_rate": 8.649320363489179e-05, | |
| "loss": 1.8602, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.027095681625740897, | |
| "grad_norm": 0.8185926079750061, | |
| "learning_rate": 8.626871855061438e-05, | |
| "loss": 1.8013, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.027294914578871347, | |
| "grad_norm": 0.9280014038085938, | |
| "learning_rate": 8.604267983514594e-05, | |
| "loss": 1.9981, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.027494147532001793, | |
| "grad_norm": 0.9310111403465271, | |
| "learning_rate": 8.581509717123273e-05, | |
| "loss": 2.0068, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.02769338048513224, | |
| "grad_norm": 0.907690703868866, | |
| "learning_rate": 8.558598030775857e-05, | |
| "loss": 1.9491, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.02789261343826269, | |
| "grad_norm": 0.9870088696479797, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 2.1852, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.028091846391393137, | |
| "grad_norm": 0.8817322850227356, | |
| "learning_rate": 8.51231833058426e-05, | |
| "loss": 2.049, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.028291079344523583, | |
| "grad_norm": 1.048509955406189, | |
| "learning_rate": 8.488952299208401e-05, | |
| "loss": 1.9654, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.028490312297654034, | |
| "grad_norm": 0.8694697618484497, | |
| "learning_rate": 8.46543681272818e-05, | |
| "loss": 1.9318, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.02868954525078448, | |
| "grad_norm": 0.8193637728691101, | |
| "learning_rate": 8.44177287846877e-05, | |
| "loss": 1.6541, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.028888778203914927, | |
| "grad_norm": 1.0921413898468018, | |
| "learning_rate": 8.417961510114356e-05, | |
| "loss": 2.1767, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.029088011157045374, | |
| "grad_norm": 1.0805463790893555, | |
| "learning_rate": 8.39400372766471e-05, | |
| "loss": 2.2352, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.029287244110175824, | |
| "grad_norm": 1.1638261079788208, | |
| "learning_rate": 8.36990055739149e-05, | |
| "loss": 2.2381, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.02948647706330627, | |
| "grad_norm": 1.0771574974060059, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 2.2466, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.029685710016436717, | |
| "grad_norm": 0.8697289228439331, | |
| "learning_rate": 8.321262189556409e-05, | |
| "loss": 1.8392, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.029884942969567167, | |
| "grad_norm": 1.0429342985153198, | |
| "learning_rate": 8.296729075500344e-05, | |
| "loss": 2.0501, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.030084175922697614, | |
| "grad_norm": 0.8660073280334473, | |
| "learning_rate": 8.272054740543052e-05, | |
| "loss": 1.5289, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.03028340887582806, | |
| "grad_norm": 0.7414923310279846, | |
| "learning_rate": 8.247240241650918e-05, | |
| "loss": 1.448, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.03048264182895851, | |
| "grad_norm": 0.7929139733314514, | |
| "learning_rate": 8.222286641794488e-05, | |
| "loss": 1.6223, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.030681874782088957, | |
| "grad_norm": 0.9731504321098328, | |
| "learning_rate": 8.197195009902924e-05, | |
| "loss": 1.6532, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.030881107735219404, | |
| "grad_norm": 0.6364417672157288, | |
| "learning_rate": 8.171966420818228e-05, | |
| "loss": 1.2613, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.031080340688349854, | |
| "grad_norm": 0.7185880541801453, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 1.5937, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.0312795736414803, | |
| "grad_norm": 0.6489309072494507, | |
| "learning_rate": 8.121102699725089e-05, | |
| "loss": 1.4197, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.03147880659461075, | |
| "grad_norm": 0.800514817237854, | |
| "learning_rate": 8.095469746549172e-05, | |
| "loss": 1.6741, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.0316780395477412, | |
| "grad_norm": 0.7517793774604797, | |
| "learning_rate": 8.069704193751832e-05, | |
| "loss": 1.554, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.03187727250087164, | |
| "grad_norm": 0.7281696200370789, | |
| "learning_rate": 8.043807145043604e-05, | |
| "loss": 1.5014, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.03207650545400209, | |
| "grad_norm": 0.7210339307785034, | |
| "learning_rate": 8.017779709767858e-05, | |
| "loss": 1.4519, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.03227573840713254, | |
| "grad_norm": 0.8253017663955688, | |
| "learning_rate": 7.991623002853296e-05, | |
| "loss": 1.5572, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.032474971360262984, | |
| "grad_norm": 0.7311911582946777, | |
| "learning_rate": 7.965338144766186e-05, | |
| "loss": 1.6551, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.032674204313393435, | |
| "grad_norm": 0.6998488306999207, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.5228, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.032873437266523885, | |
| "grad_norm": 0.8242675065994263, | |
| "learning_rate": 7.912388484339012e-05, | |
| "loss": 1.6667, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.03307267021965433, | |
| "grad_norm": 0.8666072487831116, | |
| "learning_rate": 7.88572595018617e-05, | |
| "loss": 1.6337, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.03327190317278478, | |
| "grad_norm": 0.7214533686637878, | |
| "learning_rate": 7.858939801138061e-05, | |
| "loss": 1.5181, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.03347113612591523, | |
| "grad_norm": 0.7866557836532593, | |
| "learning_rate": 7.832031184624164e-05, | |
| "loss": 1.693, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.03367036907904567, | |
| "grad_norm": 0.715961217880249, | |
| "learning_rate": 7.80500125332005e-05, | |
| "loss": 1.5668, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.03386960203217612, | |
| "grad_norm": 0.891643762588501, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 1.9778, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.03406883498530657, | |
| "grad_norm": 0.8939557075500488, | |
| "learning_rate": 7.750582082977467e-05, | |
| "loss": 1.8125, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.034268067938437015, | |
| "grad_norm": 0.7379950881004333, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 1.6326, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.034467300891567465, | |
| "grad_norm": 0.7286401987075806, | |
| "learning_rate": 7.695691614555003e-05, | |
| "loss": 1.7061, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.034666533844697915, | |
| "grad_norm": 0.7566738724708557, | |
| "learning_rate": 7.668072579578058e-05, | |
| "loss": 1.7648, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.03486576679782836, | |
| "grad_norm": 0.8034036755561829, | |
| "learning_rate": 7.64033925325184e-05, | |
| "loss": 1.7197, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.03506499975095881, | |
| "grad_norm": 0.8322769403457642, | |
| "learning_rate": 7.612492823579745e-05, | |
| "loss": 1.8101, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.03526423270408926, | |
| "grad_norm": 0.8380805850028992, | |
| "learning_rate": 7.584534483410137e-05, | |
| "loss": 1.7692, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.0354634656572197, | |
| "grad_norm": 0.8168779015541077, | |
| "learning_rate": 7.55646543038526e-05, | |
| "loss": 1.8212, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.03566269861035015, | |
| "grad_norm": 0.8079612851142883, | |
| "learning_rate": 7.528286866889924e-05, | |
| "loss": 1.6029, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.0358619315634806, | |
| "grad_norm": 0.8072683811187744, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.8446, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.036061164516611045, | |
| "grad_norm": 0.7589475512504578, | |
| "learning_rate": 7.471606041430723e-05, | |
| "loss": 1.6799, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.036260397469741495, | |
| "grad_norm": 0.8317643404006958, | |
| "learning_rate": 7.443106207484776e-05, | |
| "loss": 1.8887, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.036459630422871946, | |
| "grad_norm": 1.1349105834960938, | |
| "learning_rate": 7.414501719000187e-05, | |
| "loss": 1.7912, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.03665886337600239, | |
| "grad_norm": 0.7826024889945984, | |
| "learning_rate": 7.385793801298042e-05, | |
| "loss": 1.81, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.03685809632913284, | |
| "grad_norm": 0.904198944568634, | |
| "learning_rate": 7.35698368412999e-05, | |
| "loss": 1.6053, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.03705732928226329, | |
| "grad_norm": 0.8450482487678528, | |
| "learning_rate": 7.328072601625557e-05, | |
| "loss": 1.7809, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.03725656223539373, | |
| "grad_norm": 0.8804753422737122, | |
| "learning_rate": 7.2990617922393e-05, | |
| "loss": 1.9123, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.03745579518852418, | |
| "grad_norm": 0.9151681661605835, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 1.8679, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.03765502814165463, | |
| "grad_norm": 0.9498852491378784, | |
| "learning_rate": 7.240745967946113e-05, | |
| "loss": 1.9956, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.037854261094785076, | |
| "grad_norm": 1.0988324880599976, | |
| "learning_rate": 7.211443451095007e-05, | |
| "loss": 2.0273, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.038053494047915526, | |
| "grad_norm": 0.9793733358383179, | |
| "learning_rate": 7.18204620336671e-05, | |
| "loss": 2.0765, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.038252727001045976, | |
| "grad_norm": 0.9851742386817932, | |
| "learning_rate": 7.152555484041476e-05, | |
| "loss": 2.0685, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.03845195995417642, | |
| "grad_norm": 0.9779745936393738, | |
| "learning_rate": 7.122972556403567e-05, | |
| "loss": 2.0489, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.03865119290730687, | |
| "grad_norm": 0.9139167070388794, | |
| "learning_rate": 7.09329868768714e-05, | |
| "loss": 1.948, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03885042586043732, | |
| "grad_norm": 0.9257181286811829, | |
| "learning_rate": 7.063535149021973e-05, | |
| "loss": 1.9178, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.03904965881356776, | |
| "grad_norm": 0.8992418050765991, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 1.9382, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.03924889176669821, | |
| "grad_norm": 1.0115379095077515, | |
| "learning_rate": 7.003744165515705e-05, | |
| "loss": 1.9663, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03944812471982866, | |
| "grad_norm": 0.9632532596588135, | |
| "learning_rate": 6.973719281921335e-05, | |
| "loss": 2.0153, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.039647357672959106, | |
| "grad_norm": 0.9716604351997375, | |
| "learning_rate": 6.943609850761979e-05, | |
| "loss": 2.0656, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.039846590626089556, | |
| "grad_norm": 1.0795772075653076, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 1.8788, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.039846590626089556, | |
| "eval_loss": 1.7653837203979492, | |
| "eval_runtime": 1307.447, | |
| "eval_samples_per_second": 6.466, | |
| "eval_steps_per_second": 1.617, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.04004582357922, | |
| "grad_norm": 0.7175781726837158, | |
| "learning_rate": 6.883142508466054e-05, | |
| "loss": 1.3337, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.04024505653235045, | |
| "grad_norm": 0.8620020747184753, | |
| "learning_rate": 6.852787187549182e-05, | |
| "loss": 1.3683, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.0404442894854809, | |
| "grad_norm": 0.8307313323020935, | |
| "learning_rate": 6.82235249939575e-05, | |
| "loss": 1.5546, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.04064352243861134, | |
| "grad_norm": 0.752736508846283, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 1.6792, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.04084275539174179, | |
| "grad_norm": 0.8471193909645081, | |
| "learning_rate": 6.761250239606169e-05, | |
| "loss": 1.6798, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.04104198834487224, | |
| "grad_norm": 0.7192975282669067, | |
| "learning_rate": 6.730585285387465e-05, | |
| "loss": 1.5224, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.04124122129800269, | |
| "grad_norm": 0.804473876953125, | |
| "learning_rate": 6.699846198654971e-05, | |
| "loss": 1.8085, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.04144045425113314, | |
| "grad_norm": 0.7053448557853699, | |
| "learning_rate": 6.669034296168855e-05, | |
| "loss": 1.5087, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.04163968720426359, | |
| "grad_norm": 0.6852158308029175, | |
| "learning_rate": 6.638150897808468e-05, | |
| "loss": 1.5744, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.04183892015739403, | |
| "grad_norm": 0.7240145802497864, | |
| "learning_rate": 6.607197326515808e-05, | |
| "loss": 1.4104, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.04203815311052448, | |
| "grad_norm": 0.8016362190246582, | |
| "learning_rate": 6.57617490823885e-05, | |
| "loss": 1.7332, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.04223738606365493, | |
| "grad_norm": 0.813021719455719, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.5941, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.042436619016785374, | |
| "grad_norm": 0.6962946653366089, | |
| "learning_rate": 6.513928849212873e-05, | |
| "loss": 1.5307, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.042635851969915824, | |
| "grad_norm": 0.7284409999847412, | |
| "learning_rate": 6.482707874877854e-05, | |
| "loss": 1.6151, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.042835084923046274, | |
| "grad_norm": 0.824651300907135, | |
| "learning_rate": 6.451423386272312e-05, | |
| "loss": 1.4702, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.04303431787617672, | |
| "grad_norm": 0.6916880011558533, | |
| "learning_rate": 6.420076723519614e-05, | |
| "loss": 1.6181, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.04323355082930717, | |
| "grad_norm": 0.6723554730415344, | |
| "learning_rate": 6.388669229406462e-05, | |
| "loss": 1.4223, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.04343278378243762, | |
| "grad_norm": 0.8126989603042603, | |
| "learning_rate": 6.357202249325371e-05, | |
| "loss": 1.457, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.04363201673556806, | |
| "grad_norm": 0.606365442276001, | |
| "learning_rate": 6.32567713121704e-05, | |
| "loss": 1.3669, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.04383124968869851, | |
| "grad_norm": 1.0885868072509766, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 1.6269, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.04403048264182896, | |
| "grad_norm": 0.6699976921081543, | |
| "learning_rate": 6.26245788507579e-05, | |
| "loss": 1.4144, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.044229715594959404, | |
| "grad_norm": 0.7255838513374329, | |
| "learning_rate": 6.230766465144967e-05, | |
| "loss": 1.528, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.044428948548089854, | |
| "grad_norm": 0.9336103200912476, | |
| "learning_rate": 6.199022323275083e-05, | |
| "loss": 1.4114, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.044628181501220304, | |
| "grad_norm": 1.1342967748641968, | |
| "learning_rate": 6.167226819279528e-05, | |
| "loss": 1.6462, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.04482741445435075, | |
| "grad_norm": 0.8370170593261719, | |
| "learning_rate": 6.135381315171867e-05, | |
| "loss": 1.6144, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.0450266474074812, | |
| "grad_norm": 0.8431183099746704, | |
| "learning_rate": 6.103487175107507e-05, | |
| "loss": 1.787, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.04522588036061165, | |
| "grad_norm": 0.776308536529541, | |
| "learning_rate": 6.071545765325254e-05, | |
| "loss": 1.4541, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.04542511331374209, | |
| "grad_norm": 0.6885831356048584, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 1.2892, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.04562434626687254, | |
| "grad_norm": 0.848490834236145, | |
| "learning_rate": 6.007526611628086e-05, | |
| "loss": 1.7647, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.04582357922000299, | |
| "grad_norm": 0.8826428055763245, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 1.9109, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.046022812173133434, | |
| "grad_norm": 0.8888834118843079, | |
| "learning_rate": 5.9433348234327765e-05, | |
| "loss": 1.7475, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.046222045126263885, | |
| "grad_norm": 1.0613147020339966, | |
| "learning_rate": 5.911177627460739e-05, | |
| "loss": 1.852, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.046421278079394335, | |
| "grad_norm": 0.91031813621521, | |
| "learning_rate": 5.8789813996717736e-05, | |
| "loss": 1.8365, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.04662051103252478, | |
| "grad_norm": 0.8690152764320374, | |
| "learning_rate": 5.8467475192451226e-05, | |
| "loss": 1.7708, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.04681974398565523, | |
| "grad_norm": 0.9083284139633179, | |
| "learning_rate": 5.814477366972945e-05, | |
| "loss": 1.8436, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.04701897693878568, | |
| "grad_norm": 1.0710521936416626, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 1.9518, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.04721820989191612, | |
| "grad_norm": 0.9421839118003845, | |
| "learning_rate": 5.749833777770225e-05, | |
| "loss": 1.9924, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.04741744284504657, | |
| "grad_norm": 0.9856808185577393, | |
| "learning_rate": 5.717463109955896e-05, | |
| "loss": 1.9579, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.04761667579817702, | |
| "grad_norm": 1.14179527759552, | |
| "learning_rate": 5.685061708409841e-05, | |
| "loss": 1.9985, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.047815908751307465, | |
| "grad_norm": 1.4646105766296387, | |
| "learning_rate": 5.6526309611002594e-05, | |
| "loss": 2.0769, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.048015141704437915, | |
| "grad_norm": 0.9936044216156006, | |
| "learning_rate": 5.6201722572524275e-05, | |
| "loss": 1.7261, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.048214374657568365, | |
| "grad_norm": 1.0193665027618408, | |
| "learning_rate": 5.587686987289189e-05, | |
| "loss": 2.02, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.04841360761069881, | |
| "grad_norm": 1.5562198162078857, | |
| "learning_rate": 5.5551765427713884e-05, | |
| "loss": 1.9928, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.04861284056382926, | |
| "grad_norm": 0.9775562882423401, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 2.0862, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.0488120735169597, | |
| "grad_norm": 1.0259995460510254, | |
| "learning_rate": 5.490085701647805e-05, | |
| "loss": 1.9838, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.04901130647009015, | |
| "grad_norm": 1.0052729845046997, | |
| "learning_rate": 5.457508093317013e-05, | |
| "loss": 2.0598, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.0492105394232206, | |
| "grad_norm": 0.9661989212036133, | |
| "learning_rate": 5.4249108868622086e-05, | |
| "loss": 2.0546, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.049409772376351045, | |
| "grad_norm": 1.0998971462249756, | |
| "learning_rate": 5.392295478639225e-05, | |
| "loss": 2.0453, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.049609005329481495, | |
| "grad_norm": 0.9969896078109741, | |
| "learning_rate": 5.359663265783598e-05, | |
| "loss": 2.0773, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.049808238282611945, | |
| "grad_norm": 1.1022703647613525, | |
| "learning_rate": 5.327015646150716e-05, | |
| "loss": 2.1264, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.05000747123574239, | |
| "grad_norm": 0.7417923808097839, | |
| "learning_rate": 5.294354018255945e-05, | |
| "loss": 1.2255, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.05020670418887284, | |
| "grad_norm": 0.8457772135734558, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 1.3468, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.05040593714200329, | |
| "grad_norm": 0.8699066042900085, | |
| "learning_rate": 5.228994334682604e-05, | |
| "loss": 1.683, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.05060517009513373, | |
| "grad_norm": 0.8360949158668518, | |
| "learning_rate": 5.196299078795344e-05, | |
| "loss": 1.667, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.05080440304826418, | |
| "grad_norm": 0.7682618498802185, | |
| "learning_rate": 5.1635954141088813e-05, | |
| "loss": 1.5661, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.05100363600139463, | |
| "grad_norm": 1.1287474632263184, | |
| "learning_rate": 5.1308847415393666e-05, | |
| "loss": 1.7158, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.051202868954525076, | |
| "grad_norm": 0.7990018725395203, | |
| "learning_rate": 5.0981684623031415e-05, | |
| "loss": 1.4798, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.051402101907655526, | |
| "grad_norm": 0.7507796883583069, | |
| "learning_rate": 5.0654479778567223e-05, | |
| "loss": 1.6002, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.051601334860785976, | |
| "grad_norm": 0.7758486866950989, | |
| "learning_rate": 5.0327246898367597e-05, | |
| "loss": 1.6382, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.05180056781391642, | |
| "grad_norm": 0.756610095500946, | |
| "learning_rate": 5e-05, | |
| "loss": 1.4587, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.05199980076704687, | |
| "grad_norm": 0.7428910732269287, | |
| "learning_rate": 4.9672753101632415e-05, | |
| "loss": 1.4796, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.05219903372017732, | |
| "grad_norm": 0.9282284379005432, | |
| "learning_rate": 4.934552022143279e-05, | |
| "loss": 1.7188, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.05239826667330776, | |
| "grad_norm": 0.8230169415473938, | |
| "learning_rate": 4.901831537696859e-05, | |
| "loss": 1.6737, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.05259749962643821, | |
| "grad_norm": 0.8880282640457153, | |
| "learning_rate": 4.869115258460635e-05, | |
| "loss": 1.6383, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.05279673257956866, | |
| "grad_norm": 0.6525868773460388, | |
| "learning_rate": 4.83640458589112e-05, | |
| "loss": 1.2746, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.052995965532699106, | |
| "grad_norm": 0.8035927414894104, | |
| "learning_rate": 4.8037009212046586e-05, | |
| "loss": 1.5259, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.053195198485829556, | |
| "grad_norm": 0.7835052013397217, | |
| "learning_rate": 4.7710056653173976e-05, | |
| "loss": 1.631, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.053394431438960006, | |
| "grad_norm": 0.7649900317192078, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 1.6074, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.05359366439209045, | |
| "grad_norm": 0.8601043820381165, | |
| "learning_rate": 4.7056459817440544e-05, | |
| "loss": 1.6147, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.0537928973452209, | |
| "grad_norm": 0.775799572467804, | |
| "learning_rate": 4.6729843538492847e-05, | |
| "loss": 1.5416, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.05399213029835135, | |
| "grad_norm": 0.8741908073425293, | |
| "learning_rate": 4.640336734216403e-05, | |
| "loss": 1.7454, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.05419136325148179, | |
| "grad_norm": 0.8385461568832397, | |
| "learning_rate": 4.607704521360776e-05, | |
| "loss": 1.6794, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.05439059620461224, | |
| "grad_norm": 0.7103844285011292, | |
| "learning_rate": 4.575089113137792e-05, | |
| "loss": 1.5788, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.05458982915774269, | |
| "grad_norm": 0.7964368462562561, | |
| "learning_rate": 4.542491906682989e-05, | |
| "loss": 1.5432, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.05478906211087314, | |
| "grad_norm": 0.7998329997062683, | |
| "learning_rate": 4.509914298352197e-05, | |
| "loss": 1.8727, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.05498829506400359, | |
| "grad_norm": 0.7981895208358765, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 1.7219, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.05518752801713404, | |
| "grad_norm": 0.677744448184967, | |
| "learning_rate": 4.444823457228612e-05, | |
| "loss": 1.5126, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.05538676097026448, | |
| "grad_norm": 0.8186181783676147, | |
| "learning_rate": 4.412313012710813e-05, | |
| "loss": 1.7727, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.05558599392339493, | |
| "grad_norm": 0.8363473415374756, | |
| "learning_rate": 4.379827742747575e-05, | |
| "loss": 1.4963, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.05578522687652538, | |
| "grad_norm": 1.204262614250183, | |
| "learning_rate": 4.347369038899744e-05, | |
| "loss": 1.6811, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.055984459829655824, | |
| "grad_norm": 0.9086161851882935, | |
| "learning_rate": 4.3149382915901606e-05, | |
| "loss": 1.8958, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.056183692782786274, | |
| "grad_norm": 0.8663166165351868, | |
| "learning_rate": 4.282536890044104e-05, | |
| "loss": 1.8902, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.056382925735916724, | |
| "grad_norm": 0.9640662670135498, | |
| "learning_rate": 4.250166222229774e-05, | |
| "loss": 2.0409, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.05658215868904717, | |
| "grad_norm": 0.9557836651802063, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 1.9425, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.05678139164217762, | |
| "grad_norm": 1.0826152563095093, | |
| "learning_rate": 4.185522633027057e-05, | |
| "loss": 1.6738, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.05698062459530807, | |
| "grad_norm": 0.7835938334465027, | |
| "learning_rate": 4.153252480754877e-05, | |
| "loss": 1.6434, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.05717985754843851, | |
| "grad_norm": 0.9007090330123901, | |
| "learning_rate": 4.1210186003282275e-05, | |
| "loss": 1.7945, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.05737909050156896, | |
| "grad_norm": 0.9333200454711914, | |
| "learning_rate": 4.088822372539263e-05, | |
| "loss": 2.1531, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.057578323454699404, | |
| "grad_norm": 0.9241042137145996, | |
| "learning_rate": 4.0566651765672246e-05, | |
| "loss": 2.0779, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.057777556407829854, | |
| "grad_norm": 0.9367052912712097, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 1.947, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.057976789360960304, | |
| "grad_norm": 0.8407790660858154, | |
| "learning_rate": 3.992473388371915e-05, | |
| "loss": 1.6544, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.05817602231409075, | |
| "grad_norm": 1.0361599922180176, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 2.0688, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.0583752552672212, | |
| "grad_norm": 1.0046849250793457, | |
| "learning_rate": 3.928454234674747e-05, | |
| "loss": 2.1278, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.05857448822035165, | |
| "grad_norm": 0.9455753564834595, | |
| "learning_rate": 3.896512824892495e-05, | |
| "loss": 1.8499, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.05877372117348209, | |
| "grad_norm": 0.9063624739646912, | |
| "learning_rate": 3.864618684828134e-05, | |
| "loss": 1.9574, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.05897295412661254, | |
| "grad_norm": 0.9712712168693542, | |
| "learning_rate": 3.832773180720475e-05, | |
| "loss": 1.9854, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.05917218707974299, | |
| "grad_norm": 0.953110933303833, | |
| "learning_rate": 3.800977676724919e-05, | |
| "loss": 1.8418, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.059371420032873434, | |
| "grad_norm": 0.8904990553855896, | |
| "learning_rate": 3.769233534855035e-05, | |
| "loss": 1.9232, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.059570652986003884, | |
| "grad_norm": 0.9431455135345459, | |
| "learning_rate": 3.73754211492421e-05, | |
| "loss": 1.7605, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.059769885939134335, | |
| "grad_norm": 1.0603832006454468, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 2.0183, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.059769885939134335, | |
| "eval_loss": 1.7440650463104248, | |
| "eval_runtime": 1312.4715, | |
| "eval_samples_per_second": 6.441, | |
| "eval_steps_per_second": 1.611, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.05996911889226478, | |
| "grad_norm": 0.9234136343002319, | |
| "learning_rate": 3.6743228687829595e-05, | |
| "loss": 1.4519, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.06016835184539523, | |
| "grad_norm": 0.7455631494522095, | |
| "learning_rate": 3.642797750674629e-05, | |
| "loss": 1.4676, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.06036758479852568, | |
| "grad_norm": 0.8243167400360107, | |
| "learning_rate": 3.6113307705935396e-05, | |
| "loss": 1.4413, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.06056681775165612, | |
| "grad_norm": 0.7181432843208313, | |
| "learning_rate": 3.579923276480387e-05, | |
| "loss": 1.719, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.06076605070478657, | |
| "grad_norm": 0.7262921333312988, | |
| "learning_rate": 3.5485766137276894e-05, | |
| "loss": 1.3258, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.06096528365791702, | |
| "grad_norm": 0.6151476502418518, | |
| "learning_rate": 3.5172921251221455e-05, | |
| "loss": 1.2912, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.061164516611047465, | |
| "grad_norm": 0.7192188501358032, | |
| "learning_rate": 3.486071150787128e-05, | |
| "loss": 1.5151, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.061363749564177915, | |
| "grad_norm": 0.7189763784408569, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 1.4975, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.061562982517308365, | |
| "grad_norm": 0.9334692358970642, | |
| "learning_rate": 3.423825091761153e-05, | |
| "loss": 1.8043, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.06176221547043881, | |
| "grad_norm": 0.7157220244407654, | |
| "learning_rate": 3.392802673484193e-05, | |
| "loss": 1.5496, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.06196144842356926, | |
| "grad_norm": 0.7380807399749756, | |
| "learning_rate": 3.361849102191533e-05, | |
| "loss": 1.4888, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.06216068137669971, | |
| "grad_norm": 0.7653237581253052, | |
| "learning_rate": 3.330965703831146e-05, | |
| "loss": 1.5173, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.06235991432983015, | |
| "grad_norm": 0.6636477112770081, | |
| "learning_rate": 3.300153801345028e-05, | |
| "loss": 1.3109, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.0625591472829606, | |
| "grad_norm": 0.6692584156990051, | |
| "learning_rate": 3.2694147146125345e-05, | |
| "loss": 1.5803, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.06275838023609105, | |
| "grad_norm": 0.7734482884407043, | |
| "learning_rate": 3.2387497603938326e-05, | |
| "loss": 1.5582, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.0629576131892215, | |
| "grad_norm": 0.7633421421051025, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 1.474, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.06315684614235194, | |
| "grad_norm": 0.8933658003807068, | |
| "learning_rate": 3.177647500604252e-05, | |
| "loss": 1.4464, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.0633560790954824, | |
| "grad_norm": 0.7234926223754883, | |
| "learning_rate": 3.147212812450819e-05, | |
| "loss": 1.5108, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.06355531204861284, | |
| "grad_norm": 0.8385280966758728, | |
| "learning_rate": 3.116857491533947e-05, | |
| "loss": 1.6382, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.06375454500174328, | |
| "grad_norm": 0.7827748656272888, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 1.7277, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.06395377795487374, | |
| "grad_norm": 0.9369984865188599, | |
| "learning_rate": 3.056390149238022e-05, | |
| "loss": 1.7291, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.06415301090800418, | |
| "grad_norm": 0.8043080568313599, | |
| "learning_rate": 3.0262807180786647e-05, | |
| "loss": 1.7082, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.06435224386113463, | |
| "grad_norm": 0.8048945665359497, | |
| "learning_rate": 2.996255834484296e-05, | |
| "loss": 1.7557, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.06455147681426508, | |
| "grad_norm": 0.7266145348548889, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 1.5275, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.06475070976739553, | |
| "grad_norm": 0.9198572039604187, | |
| "learning_rate": 2.936464850978027e-05, | |
| "loss": 1.7147, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.06494994272052597, | |
| "grad_norm": 0.8027574419975281, | |
| "learning_rate": 2.9067013123128613e-05, | |
| "loss": 1.7335, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.06514917567365643, | |
| "grad_norm": 0.7442224025726318, | |
| "learning_rate": 2.8770274435964355e-05, | |
| "loss": 1.572, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.06534840862678687, | |
| "grad_norm": 0.7306829690933228, | |
| "learning_rate": 2.8474445159585235e-05, | |
| "loss": 1.6521, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.06554764157991731, | |
| "grad_norm": 0.7424048185348511, | |
| "learning_rate": 2.8179537966332887e-05, | |
| "loss": 1.6062, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.06574687453304777, | |
| "grad_norm": 0.7699829936027527, | |
| "learning_rate": 2.7885565489049946e-05, | |
| "loss": 1.5977, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.06594610748617821, | |
| "grad_norm": 0.8047099709510803, | |
| "learning_rate": 2.759254032053888e-05, | |
| "loss": 1.655, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.06614534043930866, | |
| "grad_norm": 0.8257668018341064, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 1.5537, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.06634457339243911, | |
| "grad_norm": 0.7759456634521484, | |
| "learning_rate": 2.700938207760701e-05, | |
| "loss": 1.7591, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.06654380634556956, | |
| "grad_norm": 0.893644392490387, | |
| "learning_rate": 2.671927398374443e-05, | |
| "loss": 1.9046, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.0667430392987, | |
| "grad_norm": 0.8226551413536072, | |
| "learning_rate": 2.6430163158700115e-05, | |
| "loss": 1.6834, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.06694227225183046, | |
| "grad_norm": 0.8337216973304749, | |
| "learning_rate": 2.6142061987019577e-05, | |
| "loss": 1.6095, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.0671415052049609, | |
| "grad_norm": 0.981137752532959, | |
| "learning_rate": 2.5854982809998153e-05, | |
| "loss": 1.9304, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.06734073815809134, | |
| "grad_norm": 1.0048673152923584, | |
| "learning_rate": 2.556893792515227e-05, | |
| "loss": 1.9332, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.0675399711112218, | |
| "grad_norm": 0.9600777626037598, | |
| "learning_rate": 2.5283939585692783e-05, | |
| "loss": 2.1824, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.06773920406435224, | |
| "grad_norm": 0.9388259649276733, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 1.9205, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.06793843701748269, | |
| "grad_norm": 1.0500521659851074, | |
| "learning_rate": 2.471713133110078e-05, | |
| "loss": 1.9183, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.06813766997061314, | |
| "grad_norm": 1.1563540697097778, | |
| "learning_rate": 2.4435345696147403e-05, | |
| "loss": 2.0606, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.06833690292374359, | |
| "grad_norm": 0.9865449070930481, | |
| "learning_rate": 2.4154655165898627e-05, | |
| "loss": 2.1936, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.06853613587687403, | |
| "grad_norm": 1.3468120098114014, | |
| "learning_rate": 2.3875071764202563e-05, | |
| "loss": 1.9796, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.06873536883000449, | |
| "grad_norm": 0.9959917664527893, | |
| "learning_rate": 2.3596607467481603e-05, | |
| "loss": 2.0457, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.06893460178313493, | |
| "grad_norm": 0.9387736320495605, | |
| "learning_rate": 2.3319274204219428e-05, | |
| "loss": 1.8885, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.06913383473626537, | |
| "grad_norm": 1.1370142698287964, | |
| "learning_rate": 2.3043083854449988e-05, | |
| "loss": 2.1721, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.06933306768939583, | |
| "grad_norm": 0.9565383791923523, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 2.0364, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.06953230064252627, | |
| "grad_norm": 0.9984736442565918, | |
| "learning_rate": 2.2494179170225333e-05, | |
| "loss": 1.9395, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.06973153359565672, | |
| "grad_norm": 0.9446450471878052, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 1.8676, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.06993076654878717, | |
| "grad_norm": 0.6334500312805176, | |
| "learning_rate": 2.194998746679952e-05, | |
| "loss": 1.3377, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.07012999950191762, | |
| "grad_norm": 0.8363780975341797, | |
| "learning_rate": 2.167968815375837e-05, | |
| "loss": 1.5063, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.07032923245504806, | |
| "grad_norm": 0.7027566432952881, | |
| "learning_rate": 2.1410601988619394e-05, | |
| "loss": 1.5537, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.07052846540817852, | |
| "grad_norm": 0.726865291595459, | |
| "learning_rate": 2.1142740498138324e-05, | |
| "loss": 1.3377, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.07072769836130896, | |
| "grad_norm": 0.7450829148292542, | |
| "learning_rate": 2.08761151566099e-05, | |
| "loss": 1.4613, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.0709269313144394, | |
| "grad_norm": 0.7997395992279053, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 1.5064, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.07112616426756986, | |
| "grad_norm": 0.6779711842536926, | |
| "learning_rate": 2.034661855233815e-05, | |
| "loss": 1.5039, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.0713253972207003, | |
| "grad_norm": 0.6842930912971497, | |
| "learning_rate": 2.008376997146705e-05, | |
| "loss": 1.4626, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.07152463017383075, | |
| "grad_norm": 0.7847104072570801, | |
| "learning_rate": 1.982220290232143e-05, | |
| "loss": 1.6606, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.0717238631269612, | |
| "grad_norm": 0.6870419383049011, | |
| "learning_rate": 1.9561928549563968e-05, | |
| "loss": 1.4029, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.07192309608009165, | |
| "grad_norm": 0.7871290445327759, | |
| "learning_rate": 1.9302958062481673e-05, | |
| "loss": 1.6475, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.07212232903322209, | |
| "grad_norm": 0.7864654064178467, | |
| "learning_rate": 1.9045302534508297e-05, | |
| "loss": 1.7248, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.07232156198635255, | |
| "grad_norm": 0.7532442808151245, | |
| "learning_rate": 1.8788973002749112e-05, | |
| "loss": 1.5125, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.07252079493948299, | |
| "grad_norm": 0.5970544815063477, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 1.4215, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.07272002789261343, | |
| "grad_norm": 0.5979761481285095, | |
| "learning_rate": 1.8280335791817733e-05, | |
| "loss": 1.2157, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.07291926084574389, | |
| "grad_norm": 0.7799795269966125, | |
| "learning_rate": 1.8028049900970767e-05, | |
| "loss": 1.5365, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.07311849379887433, | |
| "grad_norm": 0.8635774254798889, | |
| "learning_rate": 1.777713358205514e-05, | |
| "loss": 1.535, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.07331772675200478, | |
| "grad_norm": 0.7627410292625427, | |
| "learning_rate": 1.7527597583490822e-05, | |
| "loss": 1.5883, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.07351695970513523, | |
| "grad_norm": 0.8532704710960388, | |
| "learning_rate": 1.7279452594569483e-05, | |
| "loss": 1.8226, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.07371619265826568, | |
| "grad_norm": 0.7512400150299072, | |
| "learning_rate": 1.703270924499656e-05, | |
| "loss": 1.673, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.07391542561139612, | |
| "grad_norm": 0.8353219032287598, | |
| "learning_rate": 1.678737810443593e-05, | |
| "loss": 1.6371, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.07411465856452658, | |
| "grad_norm": 0.7360902428627014, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 1.4375, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.07431389151765702, | |
| "grad_norm": 0.8248170018196106, | |
| "learning_rate": 1.6300994426085103e-05, | |
| "loss": 1.6235, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.07451312447078746, | |
| "grad_norm": 0.7493998408317566, | |
| "learning_rate": 1.605996272335291e-05, | |
| "loss": 1.5191, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.07471235742391792, | |
| "grad_norm": 0.760131299495697, | |
| "learning_rate": 1.5820384898856434e-05, | |
| "loss": 1.5488, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.07491159037704836, | |
| "grad_norm": 0.8441518545150757, | |
| "learning_rate": 1.5582271215312294e-05, | |
| "loss": 1.6011, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.07511082333017881, | |
| "grad_norm": 0.8258650898933411, | |
| "learning_rate": 1.5345631872718214e-05, | |
| "loss": 1.6453, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.07531005628330927, | |
| "grad_norm": 0.7505908012390137, | |
| "learning_rate": 1.5110477007916001e-05, | |
| "loss": 1.4798, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.07550928923643971, | |
| "grad_norm": 0.7128810286521912, | |
| "learning_rate": 1.4876816694157419e-05, | |
| "loss": 1.5665, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.07570852218957015, | |
| "grad_norm": 0.8183525800704956, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 1.6711, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.07590775514270061, | |
| "grad_norm": 0.8301215767860413, | |
| "learning_rate": 1.4414019692241437e-05, | |
| "loss": 1.9458, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.07610698809583105, | |
| "grad_norm": 0.8092257380485535, | |
| "learning_rate": 1.4184902828767287e-05, | |
| "loss": 1.7644, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.0763062210489615, | |
| "grad_norm": 0.8209111094474792, | |
| "learning_rate": 1.3957320164854059e-05, | |
| "loss": 1.7298, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.07650545400209195, | |
| "grad_norm": 0.8352941274642944, | |
| "learning_rate": 1.373128144938563e-05, | |
| "loss": 1.7143, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.0767046869552224, | |
| "grad_norm": 0.8569772243499756, | |
| "learning_rate": 1.3506796365108232e-05, | |
| "loss": 1.8154, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.07690391990835284, | |
| "grad_norm": 0.8412910103797913, | |
| "learning_rate": 1.3283874528215733e-05, | |
| "loss": 1.8742, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.0771031528614833, | |
| "grad_norm": 0.907353937625885, | |
| "learning_rate": 1.3062525487937699e-05, | |
| "loss": 2.0139, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.07730238581461374, | |
| "grad_norm": 0.8931063413619995, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 1.9315, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.07750161876774418, | |
| "grad_norm": 1.0445189476013184, | |
| "learning_rate": 1.2624583656870154e-05, | |
| "loss": 2.0511, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.07770085172087464, | |
| "grad_norm": 0.9549689888954163, | |
| "learning_rate": 1.2408009626051137e-05, | |
| "loss": 2.019, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.07790008467400508, | |
| "grad_norm": 0.8554056286811829, | |
| "learning_rate": 1.2193045910983863e-05, | |
| "loss": 1.7846, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.07809931762713553, | |
| "grad_norm": 1.0174126625061035, | |
| "learning_rate": 1.1979701719998453e-05, | |
| "loss": 2.2389, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.07829855058026598, | |
| "grad_norm": 1.0677056312561035, | |
| "learning_rate": 1.1767986192049984e-05, | |
| "loss": 2.2263, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.07849778353339643, | |
| "grad_norm": 0.9621496200561523, | |
| "learning_rate": 1.1557908396327028e-05, | |
| "loss": 1.9629, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.07869701648652687, | |
| "grad_norm": 0.9656440019607544, | |
| "learning_rate": 1.134947733186315e-05, | |
| "loss": 2.1125, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.07889624943965733, | |
| "grad_norm": 0.8916715979576111, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 1.9651, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.07909548239278777, | |
| "grad_norm": 1.1278259754180908, | |
| "learning_rate": 1.0937591039762085e-05, | |
| "loss": 2.1121, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.07929471534591821, | |
| "grad_norm": 1.0817816257476807, | |
| "learning_rate": 1.0734153455962765e-05, | |
| "loss": 1.9473, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.07949394829904867, | |
| "grad_norm": 0.9848653674125671, | |
| "learning_rate": 1.0532397890342505e-05, | |
| "loss": 2.0839, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.07969318125217911, | |
| "grad_norm": 1.0025157928466797, | |
| "learning_rate": 1.0332332985438248e-05, | |
| "loss": 1.9569, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.07969318125217911, | |
| "eval_loss": 1.725017786026001, | |
| "eval_runtime": 1311.2021, | |
| "eval_samples_per_second": 6.448, | |
| "eval_steps_per_second": 1.612, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.07989241420530956, | |
| "grad_norm": 0.6342359185218811, | |
| "learning_rate": 1.013396731136465e-05, | |
| "loss": 1.1818, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.08009164715844, | |
| "grad_norm": 0.6265669465065002, | |
| "learning_rate": 9.937309365446973e-06, | |
| "loss": 1.3554, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.08029088011157046, | |
| "grad_norm": 0.639958381652832, | |
| "learning_rate": 9.742367571857091e-06, | |
| "loss": 1.4432, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.0804901130647009, | |
| "grad_norm": 0.6715696454048157, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.5088, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.08068934601783134, | |
| "grad_norm": 0.7032715082168579, | |
| "learning_rate": 9.357665770419244e-06, | |
| "loss": 1.4106, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.0808885789709618, | |
| "grad_norm": 0.6766837239265442, | |
| "learning_rate": 9.167922241916055e-06, | |
| "loss": 1.4883, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.08108781192409224, | |
| "grad_norm": 0.7200549244880676, | |
| "learning_rate": 8.97992782372432e-06, | |
| "loss": 1.5611, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.08128704487722269, | |
| "grad_norm": 0.7490854859352112, | |
| "learning_rate": 8.793690568899216e-06, | |
| "loss": 1.4866, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.08148627783035314, | |
| "grad_norm": 0.685373067855835, | |
| "learning_rate": 8.609218455224893e-06, | |
| "loss": 1.493, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.08168551078348359, | |
| "grad_norm": 0.7181761264801025, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 1.4245, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.08188474373661403, | |
| "grad_norm": 0.7508552074432373, | |
| "learning_rate": 8.245601184062852e-06, | |
| "loss": 1.4166, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.08208397668974449, | |
| "grad_norm": 0.7618119120597839, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 1.6266, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.08228320964287493, | |
| "grad_norm": 0.8253456354141235, | |
| "learning_rate": 7.889138314185678e-06, | |
| "loss": 1.5646, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.08248244259600537, | |
| "grad_norm": 0.7223126292228699, | |
| "learning_rate": 7.71360891480134e-06, | |
| "loss": 1.579, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.08268167554913583, | |
| "grad_norm": 0.8785665035247803, | |
| "learning_rate": 7.539890923671062e-06, | |
| "loss": 1.8207, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.08288090850226627, | |
| "grad_norm": 0.7915347218513489, | |
| "learning_rate": 7.367991782295391e-06, | |
| "loss": 1.5869, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.08308014145539672, | |
| "grad_norm": 0.6824432611465454, | |
| "learning_rate": 7.197918854261432e-06, | |
| "loss": 1.3985, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.08327937440852717, | |
| "grad_norm": 0.7498596906661987, | |
| "learning_rate": 7.029679424927365e-06, | |
| "loss": 1.4986, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.08347860736165762, | |
| "grad_norm": 0.7597733736038208, | |
| "learning_rate": 6.863280701110408e-06, | |
| "loss": 1.7079, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.08367784031478806, | |
| "grad_norm": 0.7908221483230591, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 1.5831, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.08387707326791852, | |
| "grad_norm": 0.7998091578483582, | |
| "learning_rate": 6.536033802742813e-06, | |
| "loss": 1.7472, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.08407630622104896, | |
| "grad_norm": 0.7912585735321045, | |
| "learning_rate": 6.375199646360142e-06, | |
| "loss": 1.6609, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.0842755391741794, | |
| "grad_norm": 0.8024159073829651, | |
| "learning_rate": 6.216234231230012e-06, | |
| "loss": 1.8421, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.08447477212730986, | |
| "grad_norm": 0.7846222519874573, | |
| "learning_rate": 6.059144366901736e-06, | |
| "loss": 1.5901, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.0846740050804403, | |
| "grad_norm": 0.8087249994277954, | |
| "learning_rate": 5.903936782582253e-06, | |
| "loss": 1.6698, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.08487323803357075, | |
| "grad_norm": 0.8053250908851624, | |
| "learning_rate": 5.750618126847912e-06, | |
| "loss": 1.6399, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.0850724709867012, | |
| "grad_norm": 0.678579568862915, | |
| "learning_rate": 5.599194967359639e-06, | |
| "loss": 1.5368, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.08527170393983165, | |
| "grad_norm": 0.7634561657905579, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 1.6473, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.08547093689296209, | |
| "grad_norm": 0.8198792338371277, | |
| "learning_rate": 5.302061001503394e-06, | |
| "loss": 1.5986, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.08567016984609255, | |
| "grad_norm": 0.9108977913856506, | |
| "learning_rate": 5.156362923365588e-06, | |
| "loss": 1.8712, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.08586940279922299, | |
| "grad_norm": 0.7726624011993408, | |
| "learning_rate": 5.012585797388936e-06, | |
| "loss": 1.5413, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.08606863575235343, | |
| "grad_norm": 0.8134598135948181, | |
| "learning_rate": 4.87073578250698e-06, | |
| "loss": 1.5729, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.08626786870548389, | |
| "grad_norm": 0.8380071520805359, | |
| "learning_rate": 4.730818955102234e-06, | |
| "loss": 1.683, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.08646710165861433, | |
| "grad_norm": 0.9267144799232483, | |
| "learning_rate": 4.592841308745932e-06, | |
| "loss": 1.8211, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.08666633461174478, | |
| "grad_norm": 0.8285892605781555, | |
| "learning_rate": 4.456808753941205e-06, | |
| "loss": 1.6478, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.08686556756487523, | |
| "grad_norm": 0.8663996458053589, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.0562, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.08706480051800568, | |
| "grad_norm": 0.9194437265396118, | |
| "learning_rate": 4.190602144143207e-06, | |
| "loss": 1.7663, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.08726403347113612, | |
| "grad_norm": 0.8766307234764099, | |
| "learning_rate": 4.06043949255509e-06, | |
| "loss": 1.8105, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.08746326642426658, | |
| "grad_norm": 1.0402116775512695, | |
| "learning_rate": 3.932244738840379e-06, | |
| "loss": 1.9364, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.08766249937739702, | |
| "grad_norm": 0.9294980764389038, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 1.9542, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.08786173233052746, | |
| "grad_norm": 0.8864694237709045, | |
| "learning_rate": 3.681780806244095e-06, | |
| "loss": 1.8497, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.08806096528365792, | |
| "grad_norm": 0.9366459250450134, | |
| "learning_rate": 3.5595223564037884e-06, | |
| "loss": 2.0825, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.08826019823678836, | |
| "grad_norm": 0.9816076159477234, | |
| "learning_rate": 3.4392532620598216e-06, | |
| "loss": 1.9228, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.08845943118991881, | |
| "grad_norm": 0.9872433543205261, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 2.0644, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.08865866414304927, | |
| "grad_norm": 0.8447529673576355, | |
| "learning_rate": 3.2047036621337236e-06, | |
| "loss": 1.8916, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.08885789709617971, | |
| "grad_norm": 0.987470269203186, | |
| "learning_rate": 3.0904332038757977e-06, | |
| "loss": 2.1664, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.08905713004931015, | |
| "grad_norm": 0.905831515789032, | |
| "learning_rate": 2.978172195332263e-06, | |
| "loss": 1.7784, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.08925636300244061, | |
| "grad_norm": 1.0130242109298706, | |
| "learning_rate": 2.8679254453910785e-06, | |
| "loss": 1.9239, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.08945559595557105, | |
| "grad_norm": 1.1352851390838623, | |
| "learning_rate": 2.759697676656098e-06, | |
| "loss": 2.1013, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.0896548289087015, | |
| "grad_norm": 1.1989761590957642, | |
| "learning_rate": 2.653493525244721e-06, | |
| "loss": 2.1471, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.08985406186183195, | |
| "grad_norm": 0.6423122882843018, | |
| "learning_rate": 2.549317540589308e-06, | |
| "loss": 1.3106, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.0900532948149624, | |
| "grad_norm": 0.6389409899711609, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.4082, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.09025252776809284, | |
| "grad_norm": 0.6329924464225769, | |
| "learning_rate": 2.3470678346851518e-06, | |
| "loss": 1.4597, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.0904517607212233, | |
| "grad_norm": 0.6127148866653442, | |
| "learning_rate": 2.2490027771406687e-06, | |
| "loss": 1.176, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.09065099367435374, | |
| "grad_norm": 0.6773777604103088, | |
| "learning_rate": 2.152983213389559e-06, | |
| "loss": 1.2917, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.09085022662748418, | |
| "grad_norm": 0.7712229490280151, | |
| "learning_rate": 2.0590132565903476e-06, | |
| "loss": 1.7185, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.09104945958061464, | |
| "grad_norm": 0.6682973504066467, | |
| "learning_rate": 1.9670969321032407e-06, | |
| "loss": 1.5444, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.09124869253374508, | |
| "grad_norm": 0.6766843795776367, | |
| "learning_rate": 1.8772381773176417e-06, | |
| "loss": 1.3107, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.09144792548687553, | |
| "grad_norm": 0.773476243019104, | |
| "learning_rate": 1.7894408414835362e-06, | |
| "loss": 1.583, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.09164715844000598, | |
| "grad_norm": 0.7078460454940796, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 1.5706, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.09184639139313643, | |
| "grad_norm": 0.7150459289550781, | |
| "learning_rate": 1.620045381987012e-06, | |
| "loss": 1.561, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.09204562434626687, | |
| "grad_norm": 0.9921646118164062, | |
| "learning_rate": 1.5384545146622852e-06, | |
| "loss": 1.9899, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.09224485729939733, | |
| "grad_norm": 0.701726496219635, | |
| "learning_rate": 1.4589395786535953e-06, | |
| "loss": 1.2879, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.09244409025252777, | |
| "grad_norm": 0.7816105484962463, | |
| "learning_rate": 1.3815039801161721e-06, | |
| "loss": 1.5822, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.09264332320565821, | |
| "grad_norm": 0.7791746258735657, | |
| "learning_rate": 1.3061510361333185e-06, | |
| "loss": 1.6751, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.09284255615878867, | |
| "grad_norm": 0.6447046995162964, | |
| "learning_rate": 1.232883974574367e-06, | |
| "loss": 1.3991, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.09304178911191911, | |
| "grad_norm": 0.7769410014152527, | |
| "learning_rate": 1.1617059339563807e-06, | |
| "loss": 1.6797, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.09324102206504956, | |
| "grad_norm": 0.8174847364425659, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 1.786, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.09344025501818001, | |
| "grad_norm": 0.7846056222915649, | |
| "learning_rate": 1.0256290220474307e-06, | |
| "loss": 1.6683, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.09363948797131046, | |
| "grad_norm": 0.8444880843162537, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 1.7476, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.0938387209244409, | |
| "grad_norm": 0.7403116822242737, | |
| "learning_rate": 8.979436164848088e-07, | |
| "loss": 1.4758, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.09403795387757136, | |
| "grad_norm": 0.7375995516777039, | |
| "learning_rate": 8.372546218022747e-07, | |
| "loss": 1.6429, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.0942371868307018, | |
| "grad_norm": 0.8594033718109131, | |
| "learning_rate": 7.786715955054203e-07, | |
| "loss": 1.675, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.09443641978383224, | |
| "grad_norm": 0.7836558222770691, | |
| "learning_rate": 7.221970470961125e-07, | |
| "loss": 1.6502, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.0946356527369627, | |
| "grad_norm": 0.7155219316482544, | |
| "learning_rate": 6.678333957560512e-07, | |
| "loss": 1.4955, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.09483488569009314, | |
| "grad_norm": 0.6998591423034668, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 1.4825, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.09503411864322359, | |
| "grad_norm": 0.9176483154296875, | |
| "learning_rate": 5.654480087916303e-07, | |
| "loss": 1.8354, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.09523335159635404, | |
| "grad_norm": 0.8327824473381042, | |
| "learning_rate": 5.174306590164879e-07, | |
| "loss": 1.7847, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.09543258454948449, | |
| "grad_norm": 0.7299410700798035, | |
| "learning_rate": 4.715329778211375e-07, | |
| "loss": 1.5685, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.09563181750261493, | |
| "grad_norm": 0.9764321446418762, | |
| "learning_rate": 4.277569313094809e-07, | |
| "loss": 1.7331, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.09583105045574539, | |
| "grad_norm": 0.779161810874939, | |
| "learning_rate": 3.8610439470164737e-07, | |
| "loss": 1.5425, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.09603028340887583, | |
| "grad_norm": 1.092887043952942, | |
| "learning_rate": 3.465771522536854e-07, | |
| "loss": 1.5982, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.09622951636200627, | |
| "grad_norm": 1.0371736288070679, | |
| "learning_rate": 3.09176897181096e-07, | |
| "loss": 1.7093, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.09642874931513673, | |
| "grad_norm": 0.9150737524032593, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 2.003, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.09662798226826717, | |
| "grad_norm": 0.88661789894104, | |
| "learning_rate": 2.407636663901591e-07, | |
| "loss": 1.8276, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.09682721522139762, | |
| "grad_norm": 0.8697336316108704, | |
| "learning_rate": 2.0975362126691712e-07, | |
| "loss": 2.0095, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.09702644817452807, | |
| "grad_norm": 0.8557798266410828, | |
| "learning_rate": 1.8087642458373134e-07, | |
| "loss": 1.7249, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.09722568112765852, | |
| "grad_norm": 0.8817839622497559, | |
| "learning_rate": 1.5413331334360182e-07, | |
| "loss": 1.8848, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.09742491408078896, | |
| "grad_norm": 0.8968345522880554, | |
| "learning_rate": 1.2952543313240472e-07, | |
| "loss": 1.9211, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.0976241470339194, | |
| "grad_norm": 0.8569303750991821, | |
| "learning_rate": 1.0705383806982606e-07, | |
| "loss": 1.8268, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.09782337998704986, | |
| "grad_norm": 0.878320574760437, | |
| "learning_rate": 8.671949076420882e-08, | |
| "loss": 1.7088, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.0980226129401803, | |
| "grad_norm": 1.0637781620025635, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 2.1718, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.09822184589331075, | |
| "grad_norm": 1.0918681621551514, | |
| "learning_rate": 5.246593205699424e-08, | |
| "loss": 2.0975, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.0984210788464412, | |
| "grad_norm": 0.9981033205986023, | |
| "learning_rate": 3.8548187963854956e-08, | |
| "loss": 1.9647, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.09862031179957165, | |
| "grad_norm": 0.8573983311653137, | |
| "learning_rate": 2.6770626181715773e-08, | |
| "loss": 1.8779, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.09881954475270209, | |
| "grad_norm": 0.9774297475814819, | |
| "learning_rate": 1.7133751222137007e-08, | |
| "loss": 1.9759, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.09901877770583255, | |
| "grad_norm": 0.9767397046089172, | |
| "learning_rate": 9.637975896759077e-09, | |
| "loss": 2.1979, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.09921801065896299, | |
| "grad_norm": 0.9074400067329407, | |
| "learning_rate": 4.2836212996499865e-09, | |
| "loss": 1.7405, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.09941724361209343, | |
| "grad_norm": 0.9676385521888733, | |
| "learning_rate": 1.0709167935385455e-09, | |
| "loss": 2.088, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.09961647656522389, | |
| "grad_norm": 1.0142868757247925, | |
| "learning_rate": 0.0, | |
| "loss": 1.9986, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09961647656522389, | |
| "eval_loss": 1.7225215435028076, | |
| "eval_runtime": 1306.9234, | |
| "eval_samples_per_second": 6.469, | |
| "eval_steps_per_second": 1.618, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.1094434267136e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |