| { | |
| "best_metric": 0.8778727054595947, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-500", | |
| "epoch": 0.32663726931242854, | |
| "eval_steps": 100, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006532745386248571, | |
| "grad_norm": 11.547571182250977, | |
| "learning_rate": 5e-06, | |
| "loss": 4.6466, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0006532745386248571, | |
| "eval_loss": 1.2989917993545532, | |
| "eval_runtime": 190.9993, | |
| "eval_samples_per_second": 13.497, | |
| "eval_steps_per_second": 3.377, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0013065490772497142, | |
| "grad_norm": 11.975031852722168, | |
| "learning_rate": 1e-05, | |
| "loss": 5.0526, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0019598236158745713, | |
| "grad_norm": 10.511598587036133, | |
| "learning_rate": 1.5e-05, | |
| "loss": 4.8616, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0026130981544994283, | |
| "grad_norm": 8.188464164733887, | |
| "learning_rate": 2e-05, | |
| "loss": 4.739, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0032663726931242854, | |
| "grad_norm": 4.608264446258545, | |
| "learning_rate": 2.5e-05, | |
| "loss": 4.2898, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0039196472317491425, | |
| "grad_norm": 4.9322381019592285, | |
| "learning_rate": 3e-05, | |
| "loss": 4.4307, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.004572921770374, | |
| "grad_norm": 5.750156402587891, | |
| "learning_rate": 3.5e-05, | |
| "loss": 4.3233, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.005226196308998857, | |
| "grad_norm": 5.375987529754639, | |
| "learning_rate": 4e-05, | |
| "loss": 4.235, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.005879470847623714, | |
| "grad_norm": 4.336426734924316, | |
| "learning_rate": 4.5e-05, | |
| "loss": 4.2463, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.006532745386248571, | |
| "grad_norm": 3.931546211242676, | |
| "learning_rate": 5e-05, | |
| "loss": 4.0456, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.007186019924873428, | |
| "grad_norm": 4.456380844116211, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 4.2932, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.007839294463498285, | |
| "grad_norm": 4.070809364318848, | |
| "learning_rate": 6e-05, | |
| "loss": 3.968, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.008492569002123142, | |
| "grad_norm": 3.567103147506714, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 4.1319, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.009145843540748, | |
| "grad_norm": 3.41229248046875, | |
| "learning_rate": 7e-05, | |
| "loss": 3.9014, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.009799118079372856, | |
| "grad_norm": 3.476752281188965, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 3.9473, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.010452392617997713, | |
| "grad_norm": 3.421405076980591, | |
| "learning_rate": 8e-05, | |
| "loss": 3.6684, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.01110566715662257, | |
| "grad_norm": 3.7692196369171143, | |
| "learning_rate": 8.5e-05, | |
| "loss": 3.6834, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.011758941695247428, | |
| "grad_norm": 3.8790102005004883, | |
| "learning_rate": 9e-05, | |
| "loss": 4.1705, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.012412216233872285, | |
| "grad_norm": 4.199179172515869, | |
| "learning_rate": 9.5e-05, | |
| "loss": 4.0168, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.013065490772497142, | |
| "grad_norm": 4.3811516761779785, | |
| "learning_rate": 0.0001, | |
| "loss": 4.2636, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.013718765311121999, | |
| "grad_norm": 4.086483478546143, | |
| "learning_rate": 9.999892908320647e-05, | |
| "loss": 3.386, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.014372039849746856, | |
| "grad_norm": 4.643093585968018, | |
| "learning_rate": 9.999571637870036e-05, | |
| "loss": 4.1357, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.015025314388371713, | |
| "grad_norm": 4.607544422149658, | |
| "learning_rate": 9.999036202410325e-05, | |
| "loss": 3.4956, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01567858892699657, | |
| "grad_norm": 4.911647319793701, | |
| "learning_rate": 9.998286624877786e-05, | |
| "loss": 3.5906, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.01633186346562143, | |
| "grad_norm": 5.430598735809326, | |
| "learning_rate": 9.997322937381829e-05, | |
| "loss": 4.3089, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.016985138004246284, | |
| "grad_norm": 5.196270942687988, | |
| "learning_rate": 9.996145181203615e-05, | |
| "loss": 3.7026, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.017638412542871143, | |
| "grad_norm": 6.21434211730957, | |
| "learning_rate": 9.994753406794301e-05, | |
| "loss": 3.8458, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.018291687081496, | |
| "grad_norm": 6.019957542419434, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 3.638, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.018944961620120857, | |
| "grad_norm": 7.344250679016113, | |
| "learning_rate": 9.991328050923581e-05, | |
| "loss": 3.5139, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.019598236158745713, | |
| "grad_norm": 8.33799934387207, | |
| "learning_rate": 9.989294616193017e-05, | |
| "loss": 3.8276, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02025151069737057, | |
| "grad_norm": 7.272817611694336, | |
| "learning_rate": 9.98704745668676e-05, | |
| "loss": 3.7295, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.020904785235995427, | |
| "grad_norm": 8.105911254882812, | |
| "learning_rate": 9.98458666866564e-05, | |
| "loss": 3.6633, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.021558059774620286, | |
| "grad_norm": 10.303070068359375, | |
| "learning_rate": 9.981912357541627e-05, | |
| "loss": 4.163, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.02221133431324514, | |
| "grad_norm": 11.288853645324707, | |
| "learning_rate": 9.97902463787331e-05, | |
| "loss": 3.8152, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.02286460885187, | |
| "grad_norm": 11.985376358032227, | |
| "learning_rate": 9.975923633360985e-05, | |
| "loss": 3.189, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.023517883390494855, | |
| "grad_norm": 11.919270515441895, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 3.3431, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.024171157929119714, | |
| "grad_norm": 11.31579303741455, | |
| "learning_rate": 9.969082310281891e-05, | |
| "loss": 3.6613, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.02482443246774457, | |
| "grad_norm": 13.086581230163574, | |
| "learning_rate": 9.965342284774632e-05, | |
| "loss": 3.2378, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.025477707006369428, | |
| "grad_norm": 13.448277473449707, | |
| "learning_rate": 9.961389560529836e-05, | |
| "loss": 3.7662, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.026130981544994283, | |
| "grad_norm": 15.159260749816895, | |
| "learning_rate": 9.957224306869053e-05, | |
| "loss": 3.0795, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.026784256083619142, | |
| "grad_norm": 16.633817672729492, | |
| "learning_rate": 9.952846702217886e-05, | |
| "loss": 3.2916, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.027437530622243998, | |
| "grad_norm": 17.962066650390625, | |
| "learning_rate": 9.948256934098352e-05, | |
| "loss": 3.0719, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.028090805160868856, | |
| "grad_norm": 19.26710319519043, | |
| "learning_rate": 9.943455199120837e-05, | |
| "loss": 3.4875, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.02874407969949371, | |
| "grad_norm": 23.785097122192383, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 3.5956, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.02939735423811857, | |
| "grad_norm": 25.271921157836914, | |
| "learning_rate": 9.933216660424395e-05, | |
| "loss": 3.0942, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.030050628776743426, | |
| "grad_norm": 26.448965072631836, | |
| "learning_rate": 9.927780295290389e-05, | |
| "loss": 3.391, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.030703903315368285, | |
| "grad_norm": 21.440593719482422, | |
| "learning_rate": 9.922132840449459e-05, | |
| "loss": 2.6666, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03135717785399314, | |
| "grad_norm": 25.863950729370117, | |
| "learning_rate": 9.916274537819775e-05, | |
| "loss": 2.983, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.032010452392617995, | |
| "grad_norm": 31.86448097229004, | |
| "learning_rate": 9.91020563835152e-05, | |
| "loss": 3.6128, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.03266372693124286, | |
| "grad_norm": 47.512001037597656, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 5.6647, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03331700146986771, | |
| "grad_norm": 7.3868536949157715, | |
| "learning_rate": 9.897437097795257e-05, | |
| "loss": 4.1604, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.03397027600849257, | |
| "grad_norm": 4.953103065490723, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 4.5837, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.034623550547117424, | |
| "grad_norm": 3.1169042587280273, | |
| "learning_rate": 9.883829406604363e-05, | |
| "loss": 4.6018, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.035276825085742286, | |
| "grad_norm": 2.86468505859375, | |
| "learning_rate": 9.876711602542563e-05, | |
| "loss": 4.3774, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.03593009962436714, | |
| "grad_norm": 2.981433868408203, | |
| "learning_rate": 9.869384896386668e-05, | |
| "loss": 4.2887, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.036583374162992, | |
| "grad_norm": 2.8601815700531006, | |
| "learning_rate": 9.861849601988383e-05, | |
| "loss": 3.9852, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.03723664870161685, | |
| "grad_norm": 2.5838541984558105, | |
| "learning_rate": 9.854106042134641e-05, | |
| "loss": 3.9778, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.037889923240241714, | |
| "grad_norm": 2.4947381019592285, | |
| "learning_rate": 9.846154548533773e-05, | |
| "loss": 4.0099, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.03854319777886657, | |
| "grad_norm": 2.7639665603637695, | |
| "learning_rate": 9.837995461801299e-05, | |
| "loss": 4.0092, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.039196472317491425, | |
| "grad_norm": 2.836315631866455, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 3.7736, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03984974685611628, | |
| "grad_norm": 2.828981637954712, | |
| "learning_rate": 9.821055915851647e-05, | |
| "loss": 3.8882, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.04050302139474114, | |
| "grad_norm": 3.3011579513549805, | |
| "learning_rate": 9.812276182268236e-05, | |
| "loss": 4.1102, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.041156295933366, | |
| "grad_norm": 2.943286180496216, | |
| "learning_rate": 9.803290306789676e-05, | |
| "loss": 4.0913, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.04180957047199085, | |
| "grad_norm": 2.9910857677459717, | |
| "learning_rate": 9.794098674340965e-05, | |
| "loss": 3.8973, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.04246284501061571, | |
| "grad_norm": 3.0220143795013428, | |
| "learning_rate": 9.784701678661045e-05, | |
| "loss": 3.8295, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04311611954924057, | |
| "grad_norm": 3.3190722465515137, | |
| "learning_rate": 9.775099722285935e-05, | |
| "loss": 4.1248, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.043769394087865426, | |
| "grad_norm": 3.1170575618743896, | |
| "learning_rate": 9.765293216531486e-05, | |
| "loss": 3.876, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.04442266862649028, | |
| "grad_norm": 3.3601784706115723, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 4.2187, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.04507594316511514, | |
| "grad_norm": 3.230534553527832, | |
| "learning_rate": 9.74506824594107e-05, | |
| "loss": 4.0142, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.04572921770374, | |
| "grad_norm": 3.33896541595459, | |
| "learning_rate": 9.73465064747553e-05, | |
| "loss": 3.869, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.046382492242364855, | |
| "grad_norm": 3.4622280597686768, | |
| "learning_rate": 9.724030232334391e-05, | |
| "loss": 3.9758, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.04703576678098971, | |
| "grad_norm": 3.7361583709716797, | |
| "learning_rate": 9.713207455460894e-05, | |
| "loss": 3.7177, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.047689041319614565, | |
| "grad_norm": 4.231381416320801, | |
| "learning_rate": 9.702182780466775e-05, | |
| "loss": 4.1549, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.04834231585823943, | |
| "grad_norm": 4.004554271697998, | |
| "learning_rate": 9.690956679612421e-05, | |
| "loss": 3.6364, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.04899559039686428, | |
| "grad_norm": 3.9489033222198486, | |
| "learning_rate": 9.67952963378663e-05, | |
| "loss": 3.6562, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.04964886493548914, | |
| "grad_norm": 4.792303562164307, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 4.2874, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.050302139474113994, | |
| "grad_norm": 5.080928802490234, | |
| "learning_rate": 9.656074673794018e-05, | |
| "loss": 3.4885, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.050955414012738856, | |
| "grad_norm": 5.956547737121582, | |
| "learning_rate": 9.644047764359622e-05, | |
| "loss": 3.8888, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.05160868855136371, | |
| "grad_norm": 5.492791652679443, | |
| "learning_rate": 9.631821919375591e-05, | |
| "loss": 3.5125, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.05226196308998857, | |
| "grad_norm": 5.798184394836426, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 3.6023, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05291523762861342, | |
| "grad_norm": 6.541443824768066, | |
| "learning_rate": 9.606775526115963e-05, | |
| "loss": 3.9795, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.053568512167238284, | |
| "grad_norm": 6.9007248878479, | |
| "learning_rate": 9.593956050744492e-05, | |
| "loss": 3.9131, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.05422178670586314, | |
| "grad_norm": 7.156078815460205, | |
| "learning_rate": 9.580939785585681e-05, | |
| "loss": 3.5494, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.054875061244487995, | |
| "grad_norm": 7.655080318450928, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 3.526, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.05552833578311285, | |
| "grad_norm": 8.81391429901123, | |
| "learning_rate": 9.554319124605879e-05, | |
| "loss": 3.9594, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.05618161032173771, | |
| "grad_norm": 9.645224571228027, | |
| "learning_rate": 9.540715869125407e-05, | |
| "loss": 3.8675, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.05683488486036257, | |
| "grad_norm": 9.505653381347656, | |
| "learning_rate": 9.526918104489777e-05, | |
| "loss": 3.3039, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.05748815939898742, | |
| "grad_norm": 11.144811630249023, | |
| "learning_rate": 9.512926421749304e-05, | |
| "loss": 3.289, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.05814143393761228, | |
| "grad_norm": 11.904159545898438, | |
| "learning_rate": 9.498741420261108e-05, | |
| "loss": 3.5135, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.05879470847623714, | |
| "grad_norm": 11.668728828430176, | |
| "learning_rate": 9.484363707663442e-05, | |
| "loss": 3.1836, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.059447983014861996, | |
| "grad_norm": 13.778961181640625, | |
| "learning_rate": 9.469793899849661e-05, | |
| "loss": 3.6804, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.06010125755348685, | |
| "grad_norm": 13.867650985717773, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 3.4685, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.06075453209211171, | |
| "grad_norm": 17.512781143188477, | |
| "learning_rate": 9.440080503264037e-05, | |
| "loss": 3.3502, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.06140780663073657, | |
| "grad_norm": 24.185264587402344, | |
| "learning_rate": 9.42493818731521e-05, | |
| "loss": 2.7727, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.062061081169361425, | |
| "grad_norm": 19.322153091430664, | |
| "learning_rate": 9.409606321741775e-05, | |
| "loss": 3.4637, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06271435570798628, | |
| "grad_norm": 20.148555755615234, | |
| "learning_rate": 9.394085563309827e-05, | |
| "loss": 2.7134, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.06336763024661114, | |
| "grad_norm": 19.66134262084961, | |
| "learning_rate": 9.378376576876999e-05, | |
| "loss": 2.1593, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.06402090478523599, | |
| "grad_norm": 21.868074417114258, | |
| "learning_rate": 9.362480035363986e-05, | |
| "loss": 3.1943, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.06467417932386085, | |
| "grad_norm": 24.685863494873047, | |
| "learning_rate": 9.34639661972572e-05, | |
| "loss": 2.8522, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.06532745386248572, | |
| "grad_norm": 32.843997955322266, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 3.7582, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06532745386248572, | |
| "eval_loss": 1.087826132774353, | |
| "eval_runtime": 193.6149, | |
| "eval_samples_per_second": 13.315, | |
| "eval_steps_per_second": 3.331, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06598072840111056, | |
| "grad_norm": 6.478489875793457, | |
| "learning_rate": 9.31367192988896e-05, | |
| "loss": 4.3501, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.06663400293973543, | |
| "grad_norm": 4.0702080726623535, | |
| "learning_rate": 9.297032057507264e-05, | |
| "loss": 4.2697, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.06728727747836027, | |
| "grad_norm": 2.496466875076294, | |
| "learning_rate": 9.280208114573859e-05, | |
| "loss": 3.8445, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.06794055201698514, | |
| "grad_norm": 2.428135633468628, | |
| "learning_rate": 9.263200821770461e-05, | |
| "loss": 4.0675, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.06859382655561, | |
| "grad_norm": 2.5381500720977783, | |
| "learning_rate": 9.246010907632895e-05, | |
| "loss": 4.223, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.06924710109423485, | |
| "grad_norm": 2.4143152236938477, | |
| "learning_rate": 9.228639108519868e-05, | |
| "loss": 3.9627, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.06990037563285971, | |
| "grad_norm": 2.6589813232421875, | |
| "learning_rate": 9.211086168581433e-05, | |
| "loss": 4.0724, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.07055365017148457, | |
| "grad_norm": 2.906999349594116, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 4.0853, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.07120692471010942, | |
| "grad_norm": 2.604168653488159, | |
| "learning_rate": 9.175439881593716e-05, | |
| "loss": 3.9059, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.07186019924873428, | |
| "grad_norm": 2.6407856941223145, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 4.1648, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07251347378735913, | |
| "grad_norm": 2.5845110416412354, | |
| "learning_rate": 9.139078154477512e-05, | |
| "loss": 3.9705, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.073166748325984, | |
| "grad_norm": 2.725156784057617, | |
| "learning_rate": 9.120630943110077e-05, | |
| "loss": 3.8773, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.07382002286460886, | |
| "grad_norm": 2.776826858520508, | |
| "learning_rate": 9.102007217627568e-05, | |
| "loss": 4.1886, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.0744732974032337, | |
| "grad_norm": 2.7510571479797363, | |
| "learning_rate": 9.083207775808396e-05, | |
| "loss": 3.8013, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.07512657194185857, | |
| "grad_norm": 2.9689526557922363, | |
| "learning_rate": 9.064233422958077e-05, | |
| "loss": 3.8871, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.07577984648048343, | |
| "grad_norm": 2.9751687049865723, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 4.0687, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.07643312101910828, | |
| "grad_norm": 3.208277702331543, | |
| "learning_rate": 9.025763242814291e-05, | |
| "loss": 4.0952, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.07708639555773314, | |
| "grad_norm": 2.978297233581543, | |
| "learning_rate": 9.006269063455304e-05, | |
| "loss": 3.7542, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.07773967009635799, | |
| "grad_norm": 3.079564332962036, | |
| "learning_rate": 8.986603268863536e-05, | |
| "loss": 3.7339, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.07839294463498285, | |
| "grad_norm": 3.3465781211853027, | |
| "learning_rate": 8.966766701456177e-05, | |
| "loss": 4.1308, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07904621917360771, | |
| "grad_norm": 3.38362717628479, | |
| "learning_rate": 8.94676021096575e-05, | |
| "loss": 3.6047, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.07969949371223256, | |
| "grad_norm": 3.4256107807159424, | |
| "learning_rate": 8.926584654403724e-05, | |
| "loss": 4.0278, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.08035276825085742, | |
| "grad_norm": 3.728395700454712, | |
| "learning_rate": 8.906240896023794e-05, | |
| "loss": 3.5618, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.08100604278948229, | |
| "grad_norm": 4.323002338409424, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 3.7879, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.08165931732810713, | |
| "grad_norm": 4.338598728179932, | |
| "learning_rate": 8.865052266813685e-05, | |
| "loss": 3.4469, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.082312591866732, | |
| "grad_norm": 4.841395854949951, | |
| "learning_rate": 8.844209160367299e-05, | |
| "loss": 3.6624, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.08296586640535684, | |
| "grad_norm": 5.293141841888428, | |
| "learning_rate": 8.823201380795001e-05, | |
| "loss": 3.6083, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.0836191409439817, | |
| "grad_norm": 6.072646141052246, | |
| "learning_rate": 8.802029828000156e-05, | |
| "loss": 3.7721, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.08427241548260657, | |
| "grad_norm": 5.880528450012207, | |
| "learning_rate": 8.780695408901613e-05, | |
| "loss": 3.6489, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.08492569002123142, | |
| "grad_norm": 6.2913055419921875, | |
| "learning_rate": 8.759199037394887e-05, | |
| "loss": 4.0706, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08557896455985628, | |
| "grad_norm": 6.430176258087158, | |
| "learning_rate": 8.737541634312985e-05, | |
| "loss": 3.6765, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.08623223909848114, | |
| "grad_norm": 6.8034987449646, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 3.7984, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.08688551363710599, | |
| "grad_norm": 8.133237838745117, | |
| "learning_rate": 8.693747451206232e-05, | |
| "loss": 3.7513, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.08753878817573085, | |
| "grad_norm": 9.387150764465332, | |
| "learning_rate": 8.671612547178428e-05, | |
| "loss": 3.943, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0881920627143557, | |
| "grad_norm": 9.971433639526367, | |
| "learning_rate": 8.649320363489179e-05, | |
| "loss": 3.5509, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.08884533725298056, | |
| "grad_norm": 8.350509643554688, | |
| "learning_rate": 8.626871855061438e-05, | |
| "loss": 2.8905, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.08949861179160543, | |
| "grad_norm": 10.003087043762207, | |
| "learning_rate": 8.604267983514594e-05, | |
| "loss": 3.369, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.09015188633023027, | |
| "grad_norm": 11.017991065979004, | |
| "learning_rate": 8.581509717123273e-05, | |
| "loss": 3.826, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.09080516086885514, | |
| "grad_norm": 11.706796646118164, | |
| "learning_rate": 8.558598030775857e-05, | |
| "loss": 2.9855, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.09145843540748, | |
| "grad_norm": 12.903501510620117, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 2.6285, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09211170994610485, | |
| "grad_norm": 12.504664421081543, | |
| "learning_rate": 8.51231833058426e-05, | |
| "loss": 3.0194, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.09276498448472971, | |
| "grad_norm": 14.484477043151855, | |
| "learning_rate": 8.488952299208401e-05, | |
| "loss": 3.1958, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.09341825902335456, | |
| "grad_norm": 17.707530975341797, | |
| "learning_rate": 8.46543681272818e-05, | |
| "loss": 3.957, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.09407153356197942, | |
| "grad_norm": 20.62574005126953, | |
| "learning_rate": 8.44177287846877e-05, | |
| "loss": 3.2235, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.09472480810060428, | |
| "grad_norm": 18.29495620727539, | |
| "learning_rate": 8.417961510114356e-05, | |
| "loss": 2.3518, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.09537808263922913, | |
| "grad_norm": 28.382152557373047, | |
| "learning_rate": 8.39400372766471e-05, | |
| "loss": 4.1565, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.096031357177854, | |
| "grad_norm": 18.329111099243164, | |
| "learning_rate": 8.36990055739149e-05, | |
| "loss": 2.1261, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.09668463171647886, | |
| "grad_norm": 18.869550704956055, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 2.6734, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.0973379062551037, | |
| "grad_norm": 23.500282287597656, | |
| "learning_rate": 8.321262189556409e-05, | |
| "loss": 3.4861, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.09799118079372857, | |
| "grad_norm": 38.71291732788086, | |
| "learning_rate": 8.296729075500344e-05, | |
| "loss": 5.1638, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09864445533235343, | |
| "grad_norm": 3.8299853801727295, | |
| "learning_rate": 8.272054740543052e-05, | |
| "loss": 4.0414, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.09929772987097828, | |
| "grad_norm": 3.310225248336792, | |
| "learning_rate": 8.247240241650918e-05, | |
| "loss": 4.6365, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.09995100440960314, | |
| "grad_norm": 2.5045037269592285, | |
| "learning_rate": 8.222286641794488e-05, | |
| "loss": 4.4531, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.10060427894822799, | |
| "grad_norm": 2.3877153396606445, | |
| "learning_rate": 8.197195009902924e-05, | |
| "loss": 4.7702, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.10125755348685285, | |
| "grad_norm": 2.4667954444885254, | |
| "learning_rate": 8.171966420818228e-05, | |
| "loss": 4.1018, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.10191082802547771, | |
| "grad_norm": 2.424144983291626, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 4.1768, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.10256410256410256, | |
| "grad_norm": 2.476624011993408, | |
| "learning_rate": 8.121102699725089e-05, | |
| "loss": 3.963, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.10321737710272742, | |
| "grad_norm": 2.590161085128784, | |
| "learning_rate": 8.095469746549172e-05, | |
| "loss": 3.9657, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.10387065164135229, | |
| "grad_norm": 2.8294732570648193, | |
| "learning_rate": 8.069704193751832e-05, | |
| "loss": 4.19, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.10452392617997713, | |
| "grad_norm": 2.526905059814453, | |
| "learning_rate": 8.043807145043604e-05, | |
| "loss": 3.7846, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.105177200718602, | |
| "grad_norm": 2.666771411895752, | |
| "learning_rate": 8.017779709767858e-05, | |
| "loss": 4.2755, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.10583047525722684, | |
| "grad_norm": 2.6683263778686523, | |
| "learning_rate": 7.991623002853296e-05, | |
| "loss": 3.8973, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.1064837497958517, | |
| "grad_norm": 2.753737688064575, | |
| "learning_rate": 7.965338144766186e-05, | |
| "loss": 4.0508, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.10713702433447657, | |
| "grad_norm": 2.6762168407440186, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 3.8199, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.10779029887310142, | |
| "grad_norm": 2.7907073497772217, | |
| "learning_rate": 7.912388484339012e-05, | |
| "loss": 4.0572, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.10844357341172628, | |
| "grad_norm": 2.9300825595855713, | |
| "learning_rate": 7.88572595018617e-05, | |
| "loss": 3.85, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.10909684795035114, | |
| "grad_norm": 2.946000337600708, | |
| "learning_rate": 7.858939801138061e-05, | |
| "loss": 4.0084, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.10975012248897599, | |
| "grad_norm": 3.150700569152832, | |
| "learning_rate": 7.832031184624164e-05, | |
| "loss": 4.0213, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.11040339702760085, | |
| "grad_norm": 3.130871057510376, | |
| "learning_rate": 7.80500125332005e-05, | |
| "loss": 3.7754, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.1110566715662257, | |
| "grad_norm": 3.2514853477478027, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 3.8823, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.11170994610485056, | |
| "grad_norm": 3.6097805500030518, | |
| "learning_rate": 7.750582082977467e-05, | |
| "loss": 3.8464, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.11236322064347543, | |
| "grad_norm": 3.4300525188446045, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 3.587, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.11301649518210027, | |
| "grad_norm": 3.962552070617676, | |
| "learning_rate": 7.695691614555003e-05, | |
| "loss": 3.9351, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.11366976972072514, | |
| "grad_norm": 3.7221381664276123, | |
| "learning_rate": 7.668072579578058e-05, | |
| "loss": 3.7333, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.11432304425935, | |
| "grad_norm": 3.8638551235198975, | |
| "learning_rate": 7.64033925325184e-05, | |
| "loss": 3.2305, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.11497631879797485, | |
| "grad_norm": 4.228523254394531, | |
| "learning_rate": 7.612492823579745e-05, | |
| "loss": 3.7991, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.11562959333659971, | |
| "grad_norm": 4.375904083251953, | |
| "learning_rate": 7.584534483410137e-05, | |
| "loss": 3.3348, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.11628286787522456, | |
| "grad_norm": 5.0705156326293945, | |
| "learning_rate": 7.55646543038526e-05, | |
| "loss": 4.2036, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.11693614241384942, | |
| "grad_norm": 5.519981384277344, | |
| "learning_rate": 7.528286866889924e-05, | |
| "loss": 3.987, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.11758941695247428, | |
| "grad_norm": 5.943816184997559, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 3.4623, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.11824269149109913, | |
| "grad_norm": 6.2165751457214355, | |
| "learning_rate": 7.471606041430723e-05, | |
| "loss": 3.4462, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.11889596602972399, | |
| "grad_norm": 6.183679580688477, | |
| "learning_rate": 7.443106207484776e-05, | |
| "loss": 3.6863, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.11954924056834886, | |
| "grad_norm": 7.212845802307129, | |
| "learning_rate": 7.414501719000187e-05, | |
| "loss": 3.4877, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.1202025151069737, | |
| "grad_norm": 7.299278736114502, | |
| "learning_rate": 7.385793801298042e-05, | |
| "loss": 3.1686, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.12085578964559857, | |
| "grad_norm": 7.535823345184326, | |
| "learning_rate": 7.35698368412999e-05, | |
| "loss": 3.7468, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.12150906418422341, | |
| "grad_norm": 9.03598690032959, | |
| "learning_rate": 7.328072601625557e-05, | |
| "loss": 3.3324, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.12216233872284828, | |
| "grad_norm": 8.97368335723877, | |
| "learning_rate": 7.2990617922393e-05, | |
| "loss": 3.0458, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.12281561326147314, | |
| "grad_norm": 11.850152015686035, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 3.6011, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.12346888780009799, | |
| "grad_norm": 12.661114692687988, | |
| "learning_rate": 7.240745967946113e-05, | |
| "loss": 3.4378, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.12412216233872285, | |
| "grad_norm": 15.797039985656738, | |
| "learning_rate": 7.211443451095007e-05, | |
| "loss": 3.9093, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.12477543687734771, | |
| "grad_norm": 13.158418655395508, | |
| "learning_rate": 7.18204620336671e-05, | |
| "loss": 3.2219, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.12542871141597256, | |
| "grad_norm": 14.715095520019531, | |
| "learning_rate": 7.152555484041476e-05, | |
| "loss": 2.8733, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.1260819859545974, | |
| "grad_norm": 15.411552429199219, | |
| "learning_rate": 7.122972556403567e-05, | |
| "loss": 3.4753, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.12673526049322228, | |
| "grad_norm": 15.468315124511719, | |
| "learning_rate": 7.09329868768714e-05, | |
| "loss": 3.6554, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.12738853503184713, | |
| "grad_norm": 17.27188491821289, | |
| "learning_rate": 7.063535149021973e-05, | |
| "loss": 3.4804, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.12804180957047198, | |
| "grad_norm": 21.4166316986084, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 2.967, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.12869508410909686, | |
| "grad_norm": 18.70254135131836, | |
| "learning_rate": 7.003744165515705e-05, | |
| "loss": 2.5688, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.1293483586477217, | |
| "grad_norm": 22.17182159423828, | |
| "learning_rate": 6.973719281921335e-05, | |
| "loss": 3.8092, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.13000163318634655, | |
| "grad_norm": 21.631132125854492, | |
| "learning_rate": 6.943609850761979e-05, | |
| "loss": 3.4749, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.13065490772497143, | |
| "grad_norm": 27.764705657958984, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 3.2158, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13065490772497143, | |
| "eval_loss": 1.0087602138519287, | |
| "eval_runtime": 193.6858, | |
| "eval_samples_per_second": 13.31, | |
| "eval_steps_per_second": 3.33, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13130818226359628, | |
| "grad_norm": 2.8136210441589355, | |
| "learning_rate": 6.883142508466054e-05, | |
| "loss": 3.6461, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.13196145680222113, | |
| "grad_norm": 3.2069029808044434, | |
| "learning_rate": 6.852787187549182e-05, | |
| "loss": 4.9158, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.132614731340846, | |
| "grad_norm": 2.4724817276000977, | |
| "learning_rate": 6.82235249939575e-05, | |
| "loss": 3.7633, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.13326800587947085, | |
| "grad_norm": 2.367910146713257, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 4.6979, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.1339212804180957, | |
| "grad_norm": 2.252790689468384, | |
| "learning_rate": 6.761250239606169e-05, | |
| "loss": 4.5078, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.13457455495672055, | |
| "grad_norm": 2.2511966228485107, | |
| "learning_rate": 6.730585285387465e-05, | |
| "loss": 4.0116, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.13522782949534543, | |
| "grad_norm": 2.3549845218658447, | |
| "learning_rate": 6.699846198654971e-05, | |
| "loss": 4.0373, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.13588110403397027, | |
| "grad_norm": 2.3142459392547607, | |
| "learning_rate": 6.669034296168855e-05, | |
| "loss": 3.7566, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.13653437857259512, | |
| "grad_norm": 2.4672322273254395, | |
| "learning_rate": 6.638150897808468e-05, | |
| "loss": 3.8954, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.13718765311122, | |
| "grad_norm": 2.590864896774292, | |
| "learning_rate": 6.607197326515808e-05, | |
| "loss": 4.3348, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.13784092764984485, | |
| "grad_norm": 2.5048024654388428, | |
| "learning_rate": 6.57617490823885e-05, | |
| "loss": 4.0438, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.1384942021884697, | |
| "grad_norm": 2.490389108657837, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 3.7406, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.13914747672709457, | |
| "grad_norm": 2.4702444076538086, | |
| "learning_rate": 6.513928849212873e-05, | |
| "loss": 3.8706, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.13980075126571942, | |
| "grad_norm": 2.634340763092041, | |
| "learning_rate": 6.482707874877854e-05, | |
| "loss": 4.0907, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.14045402580434427, | |
| "grad_norm": 2.5504937171936035, | |
| "learning_rate": 6.451423386272312e-05, | |
| "loss": 3.6381, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.14110730034296914, | |
| "grad_norm": 2.8212695121765137, | |
| "learning_rate": 6.420076723519614e-05, | |
| "loss": 4.3291, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.141760574881594, | |
| "grad_norm": 3.0288095474243164, | |
| "learning_rate": 6.388669229406462e-05, | |
| "loss": 4.1332, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.14241384942021884, | |
| "grad_norm": 2.837364435195923, | |
| "learning_rate": 6.357202249325371e-05, | |
| "loss": 3.5461, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.14306712395884372, | |
| "grad_norm": 3.0349831581115723, | |
| "learning_rate": 6.32567713121704e-05, | |
| "loss": 4.0614, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.14372039849746857, | |
| "grad_norm": 3.3535542488098145, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 3.998, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.14437367303609341, | |
| "grad_norm": 3.5414798259735107, | |
| "learning_rate": 6.26245788507579e-05, | |
| "loss": 3.9925, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.14502694757471826, | |
| "grad_norm": 3.5899298191070557, | |
| "learning_rate": 6.230766465144967e-05, | |
| "loss": 4.4236, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.14568022211334314, | |
| "grad_norm": 4.002371788024902, | |
| "learning_rate": 6.199022323275083e-05, | |
| "loss": 4.1189, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.146333496651968, | |
| "grad_norm": 4.0465006828308105, | |
| "learning_rate": 6.167226819279528e-05, | |
| "loss": 3.6562, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.14698677119059284, | |
| "grad_norm": 4.780096530914307, | |
| "learning_rate": 6.135381315171867e-05, | |
| "loss": 4.0265, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1476400457292177, | |
| "grad_norm": 4.929449081420898, | |
| "learning_rate": 6.103487175107507e-05, | |
| "loss": 3.3766, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.14829332026784256, | |
| "grad_norm": 4.737488746643066, | |
| "learning_rate": 6.071545765325254e-05, | |
| "loss": 3.2223, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.1489465948064674, | |
| "grad_norm": 5.096072673797607, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 4.0753, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.14959986934509228, | |
| "grad_norm": 5.692816734313965, | |
| "learning_rate": 6.007526611628086e-05, | |
| "loss": 3.2582, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.15025314388371713, | |
| "grad_norm": 6.038702011108398, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 4.0465, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.15090641842234198, | |
| "grad_norm": 5.955114841461182, | |
| "learning_rate": 5.9433348234327765e-05, | |
| "loss": 3.5113, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.15155969296096686, | |
| "grad_norm": 7.852488994598389, | |
| "learning_rate": 5.911177627460739e-05, | |
| "loss": 3.946, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.1522129674995917, | |
| "grad_norm": 7.21500825881958, | |
| "learning_rate": 5.8789813996717736e-05, | |
| "loss": 3.2654, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.15286624203821655, | |
| "grad_norm": 7.874154090881348, | |
| "learning_rate": 5.8467475192451226e-05, | |
| "loss": 3.5822, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.15351951657684143, | |
| "grad_norm": 8.220457077026367, | |
| "learning_rate": 5.814477366972945e-05, | |
| "loss": 3.5959, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.15417279111546628, | |
| "grad_norm": 9.068872451782227, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 3.5749, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.15482606565409113, | |
| "grad_norm": 10.15458869934082, | |
| "learning_rate": 5.749833777770225e-05, | |
| "loss": 3.7681, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.15547934019271598, | |
| "grad_norm": 10.914835929870605, | |
| "learning_rate": 5.717463109955896e-05, | |
| "loss": 3.6604, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.15613261473134085, | |
| "grad_norm": 11.057694435119629, | |
| "learning_rate": 5.685061708409841e-05, | |
| "loss": 3.1422, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.1567858892699657, | |
| "grad_norm": 12.083439826965332, | |
| "learning_rate": 5.6526309611002594e-05, | |
| "loss": 3.7388, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.15743916380859055, | |
| "grad_norm": 11.058955192565918, | |
| "learning_rate": 5.6201722572524275e-05, | |
| "loss": 3.0529, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.15809243834721542, | |
| "grad_norm": 12.308931350708008, | |
| "learning_rate": 5.587686987289189e-05, | |
| "loss": 2.9113, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.15874571288584027, | |
| "grad_norm": 12.517452239990234, | |
| "learning_rate": 5.5551765427713884e-05, | |
| "loss": 2.6273, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.15939898742446512, | |
| "grad_norm": 13.3101167678833, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 2.5797, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.16005226196309, | |
| "grad_norm": 14.140732765197754, | |
| "learning_rate": 5.490085701647805e-05, | |
| "loss": 2.689, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.16070553650171485, | |
| "grad_norm": 17.422927856445312, | |
| "learning_rate": 5.457508093317013e-05, | |
| "loss": 2.6193, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.1613588110403397, | |
| "grad_norm": 19.202436447143555, | |
| "learning_rate": 5.4249108868622086e-05, | |
| "loss": 3.1502, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.16201208557896457, | |
| "grad_norm": 20.904285430908203, | |
| "learning_rate": 5.392295478639225e-05, | |
| "loss": 3.32, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.16266536011758942, | |
| "grad_norm": 22.786781311035156, | |
| "learning_rate": 5.359663265783598e-05, | |
| "loss": 3.4771, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.16331863465621427, | |
| "grad_norm": 31.27584457397461, | |
| "learning_rate": 5.327015646150716e-05, | |
| "loss": 4.7491, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.16397190919483914, | |
| "grad_norm": 2.7455053329467773, | |
| "learning_rate": 5.294354018255945e-05, | |
| "loss": 4.2161, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.164625183733464, | |
| "grad_norm": 2.7494430541992188, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 4.3482, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.16527845827208884, | |
| "grad_norm": 2.4116597175598145, | |
| "learning_rate": 5.228994334682604e-05, | |
| "loss": 4.2538, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.1659317328107137, | |
| "grad_norm": 2.3236160278320312, | |
| "learning_rate": 5.196299078795344e-05, | |
| "loss": 3.9966, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.16658500734933857, | |
| "grad_norm": 2.252965211868286, | |
| "learning_rate": 5.1635954141088813e-05, | |
| "loss": 3.9031, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.1672382818879634, | |
| "grad_norm": 2.2256863117218018, | |
| "learning_rate": 5.1308847415393666e-05, | |
| "loss": 3.8695, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.16789155642658826, | |
| "grad_norm": 2.294617176055908, | |
| "learning_rate": 5.0981684623031415e-05, | |
| "loss": 4.0556, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.16854483096521314, | |
| "grad_norm": 2.3014414310455322, | |
| "learning_rate": 5.0654479778567223e-05, | |
| "loss": 3.731, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.169198105503838, | |
| "grad_norm": 2.2849111557006836, | |
| "learning_rate": 5.0327246898367597e-05, | |
| "loss": 3.9152, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.16985138004246284, | |
| "grad_norm": 2.435673475265503, | |
| "learning_rate": 5e-05, | |
| "loss": 4.1635, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.1705046545810877, | |
| "grad_norm": 2.5315775871276855, | |
| "learning_rate": 4.9672753101632415e-05, | |
| "loss": 4.0719, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.17115792911971256, | |
| "grad_norm": 2.5360610485076904, | |
| "learning_rate": 4.934552022143279e-05, | |
| "loss": 4.1543, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.1718112036583374, | |
| "grad_norm": 2.621554374694824, | |
| "learning_rate": 4.901831537696859e-05, | |
| "loss": 4.1552, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.17246447819696228, | |
| "grad_norm": 2.6578054428100586, | |
| "learning_rate": 4.869115258460635e-05, | |
| "loss": 3.8845, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.17311775273558713, | |
| "grad_norm": 2.6418116092681885, | |
| "learning_rate": 4.83640458589112e-05, | |
| "loss": 3.9069, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.17377102727421198, | |
| "grad_norm": 2.6990976333618164, | |
| "learning_rate": 4.8037009212046586e-05, | |
| "loss": 4.0657, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.17442430181283686, | |
| "grad_norm": 2.7651925086975098, | |
| "learning_rate": 4.7710056653173976e-05, | |
| "loss": 4.0418, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.1750775763514617, | |
| "grad_norm": 2.7356722354888916, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 4.0282, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.17573085089008655, | |
| "grad_norm": 2.977963924407959, | |
| "learning_rate": 4.7056459817440544e-05, | |
| "loss": 3.8089, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.1763841254287114, | |
| "grad_norm": 3.2176568508148193, | |
| "learning_rate": 4.6729843538492847e-05, | |
| "loss": 4.392, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.17703739996733628, | |
| "grad_norm": 3.1798744201660156, | |
| "learning_rate": 4.640336734216403e-05, | |
| "loss": 4.0641, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.17769067450596113, | |
| "grad_norm": 3.06921648979187, | |
| "learning_rate": 4.607704521360776e-05, | |
| "loss": 3.7511, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.17834394904458598, | |
| "grad_norm": 3.3475852012634277, | |
| "learning_rate": 4.575089113137792e-05, | |
| "loss": 3.8854, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.17899722358321085, | |
| "grad_norm": 3.8393049240112305, | |
| "learning_rate": 4.542491906682989e-05, | |
| "loss": 4.1801, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.1796504981218357, | |
| "grad_norm": 3.7368972301483154, | |
| "learning_rate": 4.509914298352197e-05, | |
| "loss": 3.4228, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.18030377266046055, | |
| "grad_norm": 3.9486243724823, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 3.4336, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.18095704719908542, | |
| "grad_norm": 4.318992614746094, | |
| "learning_rate": 4.444823457228612e-05, | |
| "loss": 3.1958, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.18161032173771027, | |
| "grad_norm": 4.600458145141602, | |
| "learning_rate": 4.412313012710813e-05, | |
| "loss": 3.3415, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.18226359627633512, | |
| "grad_norm": 5.333650588989258, | |
| "learning_rate": 4.379827742747575e-05, | |
| "loss": 3.744, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.18291687081496, | |
| "grad_norm": 5.62832498550415, | |
| "learning_rate": 4.347369038899744e-05, | |
| "loss": 3.9508, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.18357014535358485, | |
| "grad_norm": 5.859029293060303, | |
| "learning_rate": 4.3149382915901606e-05, | |
| "loss": 3.8297, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.1842234198922097, | |
| "grad_norm": 6.399336338043213, | |
| "learning_rate": 4.282536890044104e-05, | |
| "loss": 3.8776, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.18487669443083457, | |
| "grad_norm": 6.635499000549316, | |
| "learning_rate": 4.250166222229774e-05, | |
| "loss": 4.042, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.18552996896945942, | |
| "grad_norm": 6.871761322021484, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 3.5912, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.18618324350808427, | |
| "grad_norm": 7.2283935546875, | |
| "learning_rate": 4.185522633027057e-05, | |
| "loss": 3.3263, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.18683651804670912, | |
| "grad_norm": 7.931128978729248, | |
| "learning_rate": 4.153252480754877e-05, | |
| "loss": 3.6491, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.187489792585334, | |
| "grad_norm": 8.871136665344238, | |
| "learning_rate": 4.1210186003282275e-05, | |
| "loss": 3.3801, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.18814306712395884, | |
| "grad_norm": 9.464795112609863, | |
| "learning_rate": 4.088822372539263e-05, | |
| "loss": 3.3301, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.1887963416625837, | |
| "grad_norm": 11.650923728942871, | |
| "learning_rate": 4.0566651765672246e-05, | |
| "loss": 3.6579, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.18944961620120856, | |
| "grad_norm": 12.516644477844238, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 3.1808, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.1901028907398334, | |
| "grad_norm": 12.424504280090332, | |
| "learning_rate": 3.992473388371915e-05, | |
| "loss": 3.5156, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.19075616527845826, | |
| "grad_norm": 17.3675479888916, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 4.341, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.19140943981708314, | |
| "grad_norm": 13.690961837768555, | |
| "learning_rate": 3.928454234674747e-05, | |
| "loss": 2.6183, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.192062714355708, | |
| "grad_norm": 16.243967056274414, | |
| "learning_rate": 3.896512824892495e-05, | |
| "loss": 3.6009, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.19271598889433283, | |
| "grad_norm": 14.707989692687988, | |
| "learning_rate": 3.864618684828134e-05, | |
| "loss": 2.5393, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.1933692634329577, | |
| "grad_norm": 14.387151718139648, | |
| "learning_rate": 3.832773180720475e-05, | |
| "loss": 2.3307, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.19402253797158256, | |
| "grad_norm": 20.52593994140625, | |
| "learning_rate": 3.800977676724919e-05, | |
| "loss": 2.721, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.1946758125102074, | |
| "grad_norm": 17.45858383178711, | |
| "learning_rate": 3.769233534855035e-05, | |
| "loss": 3.0486, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.19532908704883228, | |
| "grad_norm": 19.98239517211914, | |
| "learning_rate": 3.73754211492421e-05, | |
| "loss": 2.4033, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.19598236158745713, | |
| "grad_norm": 26.76961898803711, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 3.0636, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.19598236158745713, | |
| "eval_loss": 0.916141927242279, | |
| "eval_runtime": 193.7367, | |
| "eval_samples_per_second": 13.307, | |
| "eval_steps_per_second": 3.329, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.19663563612608198, | |
| "grad_norm": 2.2509307861328125, | |
| "learning_rate": 3.6743228687829595e-05, | |
| "loss": 4.1521, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.19728891066470686, | |
| "grad_norm": 2.2536087036132812, | |
| "learning_rate": 3.642797750674629e-05, | |
| "loss": 4.0593, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.1979421852033317, | |
| "grad_norm": 2.2652878761291504, | |
| "learning_rate": 3.6113307705935396e-05, | |
| "loss": 4.1726, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.19859545974195655, | |
| "grad_norm": 2.3236594200134277, | |
| "learning_rate": 3.579923276480387e-05, | |
| "loss": 4.1662, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.1992487342805814, | |
| "grad_norm": 2.248436689376831, | |
| "learning_rate": 3.5485766137276894e-05, | |
| "loss": 3.9428, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.19990200881920628, | |
| "grad_norm": 2.3227086067199707, | |
| "learning_rate": 3.5172921251221455e-05, | |
| "loss": 4.159, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.20055528335783113, | |
| "grad_norm": 2.3091423511505127, | |
| "learning_rate": 3.486071150787128e-05, | |
| "loss": 3.8457, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.20120855789645598, | |
| "grad_norm": 2.2768592834472656, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 3.6557, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.20186183243508085, | |
| "grad_norm": 2.4113123416900635, | |
| "learning_rate": 3.423825091761153e-05, | |
| "loss": 3.9928, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.2025151069737057, | |
| "grad_norm": 2.337956666946411, | |
| "learning_rate": 3.392802673484193e-05, | |
| "loss": 3.8495, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.20316838151233055, | |
| "grad_norm": 2.3353707790374756, | |
| "learning_rate": 3.361849102191533e-05, | |
| "loss": 3.7771, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.20382165605095542, | |
| "grad_norm": 2.4087674617767334, | |
| "learning_rate": 3.330965703831146e-05, | |
| "loss": 4.1044, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.20447493058958027, | |
| "grad_norm": 2.4701943397521973, | |
| "learning_rate": 3.300153801345028e-05, | |
| "loss": 3.5573, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.20512820512820512, | |
| "grad_norm": 2.612384557723999, | |
| "learning_rate": 3.2694147146125345e-05, | |
| "loss": 3.8407, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.20578147966683, | |
| "grad_norm": 2.717681884765625, | |
| "learning_rate": 3.2387497603938326e-05, | |
| "loss": 4.164, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.20643475420545485, | |
| "grad_norm": 2.7894811630249023, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 3.6986, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.2070880287440797, | |
| "grad_norm": 2.887406826019287, | |
| "learning_rate": 3.177647500604252e-05, | |
| "loss": 4.1417, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.20774130328270457, | |
| "grad_norm": 2.995980739593506, | |
| "learning_rate": 3.147212812450819e-05, | |
| "loss": 4.0842, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.20839457782132942, | |
| "grad_norm": 3.2103445529937744, | |
| "learning_rate": 3.116857491533947e-05, | |
| "loss": 4.2465, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.20904785235995427, | |
| "grad_norm": 3.404813766479492, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 3.4803, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.20970112689857912, | |
| "grad_norm": 3.4752864837646484, | |
| "learning_rate": 3.056390149238022e-05, | |
| "loss": 4.2533, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.210354401437204, | |
| "grad_norm": 3.54484486579895, | |
| "learning_rate": 3.0262807180786647e-05, | |
| "loss": 3.9656, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.21100767597582884, | |
| "grad_norm": 3.658374547958374, | |
| "learning_rate": 2.996255834484296e-05, | |
| "loss": 3.5864, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.2116609505144537, | |
| "grad_norm": 3.7134335041046143, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 3.587, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.21231422505307856, | |
| "grad_norm": 3.8654329776763916, | |
| "learning_rate": 2.936464850978027e-05, | |
| "loss": 3.8962, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.2129674995917034, | |
| "grad_norm": 4.069035530090332, | |
| "learning_rate": 2.9067013123128613e-05, | |
| "loss": 3.6819, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.21362077413032826, | |
| "grad_norm": 4.072150230407715, | |
| "learning_rate": 2.8770274435964355e-05, | |
| "loss": 3.3375, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.21427404866895314, | |
| "grad_norm": 4.656280040740967, | |
| "learning_rate": 2.8474445159585235e-05, | |
| "loss": 3.5886, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.21492732320757799, | |
| "grad_norm": 4.830178737640381, | |
| "learning_rate": 2.8179537966332887e-05, | |
| "loss": 3.2974, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.21558059774620283, | |
| "grad_norm": 5.456755638122559, | |
| "learning_rate": 2.7885565489049946e-05, | |
| "loss": 3.8774, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.2162338722848277, | |
| "grad_norm": 5.759149551391602, | |
| "learning_rate": 2.759254032053888e-05, | |
| "loss": 3.6094, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.21688714682345256, | |
| "grad_norm": 5.614565372467041, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 3.608, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.2175404213620774, | |
| "grad_norm": 6.076308250427246, | |
| "learning_rate": 2.700938207760701e-05, | |
| "loss": 3.5142, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.21819369590070228, | |
| "grad_norm": 6.575270175933838, | |
| "learning_rate": 2.671927398374443e-05, | |
| "loss": 3.3849, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.21884697043932713, | |
| "grad_norm": 7.774899482727051, | |
| "learning_rate": 2.6430163158700115e-05, | |
| "loss": 3.4158, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.21950024497795198, | |
| "grad_norm": 7.88145637512207, | |
| "learning_rate": 2.6142061987019577e-05, | |
| "loss": 2.9199, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.22015351951657683, | |
| "grad_norm": 10.132671356201172, | |
| "learning_rate": 2.5854982809998153e-05, | |
| "loss": 4.0171, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.2208067940552017, | |
| "grad_norm": 11.227150917053223, | |
| "learning_rate": 2.556893792515227e-05, | |
| "loss": 2.9164, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.22146006859382655, | |
| "grad_norm": 10.218441009521484, | |
| "learning_rate": 2.5283939585692783e-05, | |
| "loss": 2.9718, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.2221133431324514, | |
| "grad_norm": 11.796672821044922, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 3.0505, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.22276661767107628, | |
| "grad_norm": 13.644339561462402, | |
| "learning_rate": 2.471713133110078e-05, | |
| "loss": 3.9324, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.22341989220970113, | |
| "grad_norm": 13.900707244873047, | |
| "learning_rate": 2.4435345696147403e-05, | |
| "loss": 3.3911, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.22407316674832597, | |
| "grad_norm": 18.582195281982422, | |
| "learning_rate": 2.4154655165898627e-05, | |
| "loss": 2.6938, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.22472644128695085, | |
| "grad_norm": 14.891613006591797, | |
| "learning_rate": 2.3875071764202563e-05, | |
| "loss": 2.8819, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.2253797158255757, | |
| "grad_norm": 17.06895637512207, | |
| "learning_rate": 2.3596607467481603e-05, | |
| "loss": 3.0545, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.22603299036420055, | |
| "grad_norm": 18.432598114013672, | |
| "learning_rate": 2.3319274204219428e-05, | |
| "loss": 3.5887, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.22668626490282542, | |
| "grad_norm": 22.221372604370117, | |
| "learning_rate": 2.3043083854449988e-05, | |
| "loss": 2.5782, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.22733953944145027, | |
| "grad_norm": 23.88602638244629, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 3.1699, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.22799281398007512, | |
| "grad_norm": 17.298885345458984, | |
| "learning_rate": 2.2494179170225333e-05, | |
| "loss": 2.5779, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.2286460885187, | |
| "grad_norm": 34.774845123291016, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 4.6921, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.22929936305732485, | |
| "grad_norm": 2.0609993934631348, | |
| "learning_rate": 2.194998746679952e-05, | |
| "loss": 4.05, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.2299526375959497, | |
| "grad_norm": 2.18334698677063, | |
| "learning_rate": 2.167968815375837e-05, | |
| "loss": 4.2886, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.23060591213457454, | |
| "grad_norm": 2.1787545680999756, | |
| "learning_rate": 2.1410601988619394e-05, | |
| "loss": 4.3155, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.23125918667319942, | |
| "grad_norm": 2.1137516498565674, | |
| "learning_rate": 2.1142740498138324e-05, | |
| "loss": 3.7972, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.23191246121182427, | |
| "grad_norm": 2.1815781593322754, | |
| "learning_rate": 2.08761151566099e-05, | |
| "loss": 3.9406, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.23256573575044912, | |
| "grad_norm": 2.2473208904266357, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 3.94, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.233219010289074, | |
| "grad_norm": 2.286938428878784, | |
| "learning_rate": 2.034661855233815e-05, | |
| "loss": 3.8518, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.23387228482769884, | |
| "grad_norm": 2.282809257507324, | |
| "learning_rate": 2.008376997146705e-05, | |
| "loss": 4.0518, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.2345255593663237, | |
| "grad_norm": 2.15885066986084, | |
| "learning_rate": 1.982220290232143e-05, | |
| "loss": 3.7982, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.23517883390494856, | |
| "grad_norm": 2.232218027114868, | |
| "learning_rate": 1.9561928549563968e-05, | |
| "loss": 3.9533, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2358321084435734, | |
| "grad_norm": 2.28951358795166, | |
| "learning_rate": 1.9302958062481673e-05, | |
| "loss": 3.907, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.23648538298219826, | |
| "grad_norm": 2.3329083919525146, | |
| "learning_rate": 1.9045302534508297e-05, | |
| "loss": 4.0736, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.23713865752082314, | |
| "grad_norm": 2.4136269092559814, | |
| "learning_rate": 1.8788973002749112e-05, | |
| "loss": 3.8261, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.23779193205944799, | |
| "grad_norm": 2.6540565490722656, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 4.0891, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.23844520659807283, | |
| "grad_norm": 2.4953386783599854, | |
| "learning_rate": 1.8280335791817733e-05, | |
| "loss": 3.9206, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.2390984811366977, | |
| "grad_norm": 2.7900853157043457, | |
| "learning_rate": 1.8028049900970767e-05, | |
| "loss": 3.7279, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.23975175567532256, | |
| "grad_norm": 2.7315573692321777, | |
| "learning_rate": 1.777713358205514e-05, | |
| "loss": 4.0381, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.2404050302139474, | |
| "grad_norm": 2.8006503582000732, | |
| "learning_rate": 1.7527597583490822e-05, | |
| "loss": 3.7893, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.24105830475257226, | |
| "grad_norm": 3.0075833797454834, | |
| "learning_rate": 1.7279452594569483e-05, | |
| "loss": 3.9685, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.24171157929119713, | |
| "grad_norm": 3.305966854095459, | |
| "learning_rate": 1.703270924499656e-05, | |
| "loss": 4.3688, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.24236485382982198, | |
| "grad_norm": 3.364837646484375, | |
| "learning_rate": 1.678737810443593e-05, | |
| "loss": 4.0271, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.24301812836844683, | |
| "grad_norm": 3.410656213760376, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 3.913, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.2436714029070717, | |
| "grad_norm": 3.7187421321868896, | |
| "learning_rate": 1.6300994426085103e-05, | |
| "loss": 3.6959, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.24432467744569655, | |
| "grad_norm": 3.892642021179199, | |
| "learning_rate": 1.605996272335291e-05, | |
| "loss": 3.8044, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.2449779519843214, | |
| "grad_norm": 3.7945556640625, | |
| "learning_rate": 1.5820384898856434e-05, | |
| "loss": 3.2933, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.24563122652294628, | |
| "grad_norm": 4.004775047302246, | |
| "learning_rate": 1.5582271215312294e-05, | |
| "loss": 3.3998, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.24628450106157113, | |
| "grad_norm": 4.55605936050415, | |
| "learning_rate": 1.5345631872718214e-05, | |
| "loss": 4.0181, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.24693777560019597, | |
| "grad_norm": 4.659095287322998, | |
| "learning_rate": 1.5110477007916001e-05, | |
| "loss": 3.6973, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.24759105013882085, | |
| "grad_norm": 4.90739107131958, | |
| "learning_rate": 1.4876816694157419e-05, | |
| "loss": 3.4013, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.2482443246774457, | |
| "grad_norm": 5.735424995422363, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 3.6558, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.24889759921607055, | |
| "grad_norm": 5.81011962890625, | |
| "learning_rate": 1.4414019692241437e-05, | |
| "loss": 3.5117, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.24955087375469542, | |
| "grad_norm": 5.775092601776123, | |
| "learning_rate": 1.4184902828767287e-05, | |
| "loss": 3.5385, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.25020414829332027, | |
| "grad_norm": 6.017894268035889, | |
| "learning_rate": 1.3957320164854059e-05, | |
| "loss": 3.2936, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.2508574228319451, | |
| "grad_norm": 6.8194708824157715, | |
| "learning_rate": 1.373128144938563e-05, | |
| "loss": 3.1461, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.25151069737056997, | |
| "grad_norm": 7.288778781890869, | |
| "learning_rate": 1.3506796365108232e-05, | |
| "loss": 3.711, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.2521639719091948, | |
| "grad_norm": 7.597186088562012, | |
| "learning_rate": 1.3283874528215733e-05, | |
| "loss": 3.68, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.2528172464478197, | |
| "grad_norm": 9.038799285888672, | |
| "learning_rate": 1.3062525487937699e-05, | |
| "loss": 3.1288, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.25347052098644457, | |
| "grad_norm": 9.388714790344238, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 3.068, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.2541237955250694, | |
| "grad_norm": 11.19589614868164, | |
| "learning_rate": 1.2624583656870154e-05, | |
| "loss": 3.2682, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.25477707006369427, | |
| "grad_norm": 10.183329582214355, | |
| "learning_rate": 1.2408009626051137e-05, | |
| "loss": 2.604, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.2554303446023191, | |
| "grad_norm": 12.288995742797852, | |
| "learning_rate": 1.2193045910983863e-05, | |
| "loss": 3.0615, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.25608361914094396, | |
| "grad_norm": 10.170268058776855, | |
| "learning_rate": 1.1979701719998453e-05, | |
| "loss": 3.0368, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.2567368936795688, | |
| "grad_norm": 12.902504920959473, | |
| "learning_rate": 1.1767986192049984e-05, | |
| "loss": 2.6595, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.2573901682181937, | |
| "grad_norm": 14.612078666687012, | |
| "learning_rate": 1.1557908396327028e-05, | |
| "loss": 3.1923, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.25804344275681856, | |
| "grad_norm": 14.616671562194824, | |
| "learning_rate": 1.134947733186315e-05, | |
| "loss": 3.3654, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.2586967172954434, | |
| "grad_norm": 15.300880432128906, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 2.3551, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.25934999183406826, | |
| "grad_norm": 16.621246337890625, | |
| "learning_rate": 1.0937591039762085e-05, | |
| "loss": 2.6044, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.2600032663726931, | |
| "grad_norm": 20.40192985534668, | |
| "learning_rate": 1.0734153455962765e-05, | |
| "loss": 2.9109, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.26065654091131796, | |
| "grad_norm": 21.176897048950195, | |
| "learning_rate": 1.0532397890342505e-05, | |
| "loss": 2.6236, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.26130981544994286, | |
| "grad_norm": 36.7708740234375, | |
| "learning_rate": 1.0332332985438248e-05, | |
| "loss": 5.0028, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.26130981544994286, | |
| "eval_loss": 0.8865355849266052, | |
| "eval_runtime": 193.5082, | |
| "eval_samples_per_second": 13.322, | |
| "eval_steps_per_second": 3.333, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.2619630899885677, | |
| "grad_norm": 2.0284175872802734, | |
| "learning_rate": 1.013396731136465e-05, | |
| "loss": 3.2076, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.26261636452719256, | |
| "grad_norm": 2.200660467147827, | |
| "learning_rate": 9.937309365446973e-06, | |
| "loss": 4.3066, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.2632696390658174, | |
| "grad_norm": 2.2217345237731934, | |
| "learning_rate": 9.742367571857091e-06, | |
| "loss": 4.3663, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.26392291360444226, | |
| "grad_norm": 2.1234660148620605, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 4.0273, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.2645761881430671, | |
| "grad_norm": 2.140134572982788, | |
| "learning_rate": 9.357665770419244e-06, | |
| "loss": 4.0097, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.265229462681692, | |
| "grad_norm": 2.241824150085449, | |
| "learning_rate": 9.167922241916055e-06, | |
| "loss": 3.8709, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.26588273722031686, | |
| "grad_norm": 2.207245111465454, | |
| "learning_rate": 8.97992782372432e-06, | |
| "loss": 3.7465, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.2665360117589417, | |
| "grad_norm": 2.2445242404937744, | |
| "learning_rate": 8.793690568899216e-06, | |
| "loss": 3.9702, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.26718928629756655, | |
| "grad_norm": 2.209073781967163, | |
| "learning_rate": 8.609218455224893e-06, | |
| "loss": 3.6522, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.2678425608361914, | |
| "grad_norm": 2.3187787532806396, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 3.9904, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.26849583537481625, | |
| "grad_norm": 2.3060576915740967, | |
| "learning_rate": 8.245601184062852e-06, | |
| "loss": 3.645, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.2691491099134411, | |
| "grad_norm": 2.4985568523406982, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 3.94, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.269802384452066, | |
| "grad_norm": 2.4821956157684326, | |
| "learning_rate": 7.889138314185678e-06, | |
| "loss": 3.8423, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.27045565899069085, | |
| "grad_norm": 2.536126136779785, | |
| "learning_rate": 7.71360891480134e-06, | |
| "loss": 3.8344, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.2711089335293157, | |
| "grad_norm": 2.5374646186828613, | |
| "learning_rate": 7.539890923671062e-06, | |
| "loss": 3.9225, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.27176220806794055, | |
| "grad_norm": 2.6912612915039062, | |
| "learning_rate": 7.367991782295391e-06, | |
| "loss": 3.8863, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.2724154826065654, | |
| "grad_norm": 2.7828660011291504, | |
| "learning_rate": 7.197918854261432e-06, | |
| "loss": 3.8709, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.27306875714519024, | |
| "grad_norm": 2.831202507019043, | |
| "learning_rate": 7.029679424927365e-06, | |
| "loss": 3.7255, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.27372203168381515, | |
| "grad_norm": 3.060506582260132, | |
| "learning_rate": 6.863280701110408e-06, | |
| "loss": 4.0642, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.27437530622244, | |
| "grad_norm": 3.0359199047088623, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 3.7836, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.27502858076106484, | |
| "grad_norm": 3.1587436199188232, | |
| "learning_rate": 6.536033802742813e-06, | |
| "loss": 4.0368, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.2756818552996897, | |
| "grad_norm": 3.3620858192443848, | |
| "learning_rate": 6.375199646360142e-06, | |
| "loss": 3.7547, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.27633512983831454, | |
| "grad_norm": 3.528315782546997, | |
| "learning_rate": 6.216234231230012e-06, | |
| "loss": 3.7245, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.2769884043769394, | |
| "grad_norm": 3.4212453365325928, | |
| "learning_rate": 6.059144366901736e-06, | |
| "loss": 3.3729, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.27764167891556424, | |
| "grad_norm": 3.6946189403533936, | |
| "learning_rate": 5.903936782582253e-06, | |
| "loss": 3.5213, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.27829495345418914, | |
| "grad_norm": 3.8112924098968506, | |
| "learning_rate": 5.750618126847912e-06, | |
| "loss": 3.5261, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.278948227992814, | |
| "grad_norm": 4.669442176818848, | |
| "learning_rate": 5.599194967359639e-06, | |
| "loss": 3.9578, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.27960150253143884, | |
| "grad_norm": 4.836587905883789, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 4.1958, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.2802547770700637, | |
| "grad_norm": 4.8962507247924805, | |
| "learning_rate": 5.302061001503394e-06, | |
| "loss": 3.6376, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.28090805160868854, | |
| "grad_norm": 5.25094747543335, | |
| "learning_rate": 5.156362923365588e-06, | |
| "loss": 3.6825, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.2815613261473134, | |
| "grad_norm": 5.470578193664551, | |
| "learning_rate": 5.012585797388936e-06, | |
| "loss": 3.3548, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.2822146006859383, | |
| "grad_norm": 6.3519415855407715, | |
| "learning_rate": 4.87073578250698e-06, | |
| "loss": 3.4006, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.28286787522456314, | |
| "grad_norm": 6.419662952423096, | |
| "learning_rate": 4.730818955102234e-06, | |
| "loss": 3.493, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.283521149763188, | |
| "grad_norm": 6.477165699005127, | |
| "learning_rate": 4.592841308745932e-06, | |
| "loss": 3.3161, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.28417442430181283, | |
| "grad_norm": 6.963860034942627, | |
| "learning_rate": 4.456808753941205e-06, | |
| "loss": 3.4243, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.2848276988404377, | |
| "grad_norm": 7.6668548583984375, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.5756, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.28548097337906253, | |
| "grad_norm": 8.136970520019531, | |
| "learning_rate": 4.190602144143207e-06, | |
| "loss": 3.3489, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.28613424791768743, | |
| "grad_norm": 10.233713150024414, | |
| "learning_rate": 4.06043949255509e-06, | |
| "loss": 3.6143, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.2867875224563123, | |
| "grad_norm": 10.85151481628418, | |
| "learning_rate": 3.932244738840379e-06, | |
| "loss": 3.1783, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.28744079699493713, | |
| "grad_norm": 9.790265083312988, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 2.58, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.288094071533562, | |
| "grad_norm": 10.893715858459473, | |
| "learning_rate": 3.681780806244095e-06, | |
| "loss": 3.0611, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.28874734607218683, | |
| "grad_norm": 12.682134628295898, | |
| "learning_rate": 3.5595223564037884e-06, | |
| "loss": 2.8948, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.2894006206108117, | |
| "grad_norm": 13.385756492614746, | |
| "learning_rate": 3.4392532620598216e-06, | |
| "loss": 2.6341, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.2900538951494365, | |
| "grad_norm": 16.10133934020996, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 3.5305, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.29070716968806143, | |
| "grad_norm": 12.668194770812988, | |
| "learning_rate": 3.2047036621337236e-06, | |
| "loss": 2.3386, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.2913604442266863, | |
| "grad_norm": 14.395100593566895, | |
| "learning_rate": 3.0904332038757977e-06, | |
| "loss": 2.0309, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.2920137187653111, | |
| "grad_norm": 17.13104820251465, | |
| "learning_rate": 2.978172195332263e-06, | |
| "loss": 2.4272, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.292666993303936, | |
| "grad_norm": 17.783287048339844, | |
| "learning_rate": 2.8679254453910785e-06, | |
| "loss": 2.7686, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.2933202678425608, | |
| "grad_norm": 22.626319885253906, | |
| "learning_rate": 2.759697676656098e-06, | |
| "loss": 2.9195, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.29397354238118567, | |
| "grad_norm": 35.02473831176758, | |
| "learning_rate": 2.653493525244721e-06, | |
| "loss": 4.8236, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.2946268169198106, | |
| "grad_norm": 2.0250165462493896, | |
| "learning_rate": 2.549317540589308e-06, | |
| "loss": 4.0218, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.2952800914584354, | |
| "grad_norm": 1.948515772819519, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 3.9039, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.29593336599706027, | |
| "grad_norm": 2.03169322013855, | |
| "learning_rate": 2.3470678346851518e-06, | |
| "loss": 4.3663, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.2965866405356851, | |
| "grad_norm": 2.1053688526153564, | |
| "learning_rate": 2.2490027771406687e-06, | |
| "loss": 4.0127, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.29723991507430997, | |
| "grad_norm": 2.2202868461608887, | |
| "learning_rate": 2.152983213389559e-06, | |
| "loss": 4.0635, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.2978931896129348, | |
| "grad_norm": 2.141794443130493, | |
| "learning_rate": 2.0590132565903476e-06, | |
| "loss": 3.9512, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.2985464641515597, | |
| "grad_norm": 2.1424460411071777, | |
| "learning_rate": 1.9670969321032407e-06, | |
| "loss": 3.928, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.29919973869018457, | |
| "grad_norm": 2.2259063720703125, | |
| "learning_rate": 1.8772381773176417e-06, | |
| "loss": 3.9537, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.2998530132288094, | |
| "grad_norm": 2.1677327156066895, | |
| "learning_rate": 1.7894408414835362e-06, | |
| "loss": 3.7739, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.30050628776743427, | |
| "grad_norm": 2.260230541229248, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 4.0705, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.3011595623060591, | |
| "grad_norm": 2.394634246826172, | |
| "learning_rate": 1.620045381987012e-06, | |
| "loss": 4.1479, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.30181283684468396, | |
| "grad_norm": 2.3297996520996094, | |
| "learning_rate": 1.5384545146622852e-06, | |
| "loss": 3.6953, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.3024661113833088, | |
| "grad_norm": 2.3964977264404297, | |
| "learning_rate": 1.4589395786535953e-06, | |
| "loss": 3.9611, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.3031193859219337, | |
| "grad_norm": 2.533637046813965, | |
| "learning_rate": 1.3815039801161721e-06, | |
| "loss": 3.9468, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.30377266046055856, | |
| "grad_norm": 2.529632329940796, | |
| "learning_rate": 1.3061510361333185e-06, | |
| "loss": 3.6358, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.3044259349991834, | |
| "grad_norm": 2.569408655166626, | |
| "learning_rate": 1.232883974574367e-06, | |
| "loss": 3.5356, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.30507920953780826, | |
| "grad_norm": 2.727712869644165, | |
| "learning_rate": 1.1617059339563807e-06, | |
| "loss": 3.7934, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.3057324840764331, | |
| "grad_norm": 2.8594043254852295, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 3.933, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.30638575861505796, | |
| "grad_norm": 2.939509868621826, | |
| "learning_rate": 1.0256290220474307e-06, | |
| "loss": 3.5852, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.30703903315368286, | |
| "grad_norm": 3.0464086532592773, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 3.7145, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 3.3523106575012207, | |
| "learning_rate": 8.979436164848088e-07, | |
| "loss": 3.8206, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.30834558223093256, | |
| "grad_norm": 3.190808057785034, | |
| "learning_rate": 8.372546218022747e-07, | |
| "loss": 3.8822, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.3089988567695574, | |
| "grad_norm": 3.408557891845703, | |
| "learning_rate": 7.786715955054203e-07, | |
| "loss": 3.5316, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.30965213130818225, | |
| "grad_norm": 3.75345778465271, | |
| "learning_rate": 7.221970470961125e-07, | |
| "loss": 3.8423, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.3103054058468071, | |
| "grad_norm": 4.055991172790527, | |
| "learning_rate": 6.678333957560512e-07, | |
| "loss": 3.6669, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.31095868038543195, | |
| "grad_norm": 4.102751731872559, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 3.4249, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.31161195492405686, | |
| "grad_norm": 4.2560014724731445, | |
| "learning_rate": 5.654480087916303e-07, | |
| "loss": 3.4731, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.3122652294626817, | |
| "grad_norm": 4.73115873336792, | |
| "learning_rate": 5.174306590164879e-07, | |
| "loss": 3.3286, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.31291850400130655, | |
| "grad_norm": 5.362375259399414, | |
| "learning_rate": 4.715329778211375e-07, | |
| "loss": 3.2443, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.3135717785399314, | |
| "grad_norm": 5.289713382720947, | |
| "learning_rate": 4.277569313094809e-07, | |
| "loss": 3.2602, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.31422505307855625, | |
| "grad_norm": 5.72009801864624, | |
| "learning_rate": 3.8610439470164737e-07, | |
| "loss": 3.1795, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.3148783276171811, | |
| "grad_norm": 6.249419212341309, | |
| "learning_rate": 3.465771522536854e-07, | |
| "loss": 3.2568, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.315531602155806, | |
| "grad_norm": 6.340628147125244, | |
| "learning_rate": 3.09176897181096e-07, | |
| "loss": 3.7829, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.31618487669443085, | |
| "grad_norm": 6.803210735321045, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 3.2678, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.3168381512330557, | |
| "grad_norm": 7.402984619140625, | |
| "learning_rate": 2.407636663901591e-07, | |
| "loss": 3.2089, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.31749142577168055, | |
| "grad_norm": 8.722670555114746, | |
| "learning_rate": 2.0975362126691712e-07, | |
| "loss": 3.4273, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.3181447003103054, | |
| "grad_norm": 8.667076110839844, | |
| "learning_rate": 1.8087642458373134e-07, | |
| "loss": 2.9022, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.31879797484893024, | |
| "grad_norm": 9.627767562866211, | |
| "learning_rate": 1.5413331334360182e-07, | |
| "loss": 3.0261, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.31945124938755515, | |
| "grad_norm": 10.799980163574219, | |
| "learning_rate": 1.2952543313240472e-07, | |
| "loss": 2.731, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.32010452392618, | |
| "grad_norm": 12.3025484085083, | |
| "learning_rate": 1.0705383806982606e-07, | |
| "loss": 3.1692, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.32075779846480484, | |
| "grad_norm": 11.816473007202148, | |
| "learning_rate": 8.671949076420882e-08, | |
| "loss": 2.1674, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.3214110730034297, | |
| "grad_norm": 13.557419776916504, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 2.846, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.32206434754205454, | |
| "grad_norm": 11.210492134094238, | |
| "learning_rate": 5.246593205699424e-08, | |
| "loss": 2.5836, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.3227176220806794, | |
| "grad_norm": 13.040658950805664, | |
| "learning_rate": 3.8548187963854956e-08, | |
| "loss": 3.2673, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.32337089661930424, | |
| "grad_norm": 15.785333633422852, | |
| "learning_rate": 2.6770626181715773e-08, | |
| "loss": 2.518, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.32402417115792914, | |
| "grad_norm": 15.67574405670166, | |
| "learning_rate": 1.7133751222137007e-08, | |
| "loss": 2.387, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.324677445696554, | |
| "grad_norm": 15.306534767150879, | |
| "learning_rate": 9.637975896759077e-09, | |
| "loss": 2.3683, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.32533072023517884, | |
| "grad_norm": 20.011672973632812, | |
| "learning_rate": 4.2836212996499865e-09, | |
| "loss": 3.7281, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.3259839947738037, | |
| "grad_norm": 21.42987823486328, | |
| "learning_rate": 1.0709167935385455e-09, | |
| "loss": 2.4191, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.32663726931242854, | |
| "grad_norm": 32.15013885498047, | |
| "learning_rate": 0.0, | |
| "loss": 3.3215, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.32663726931242854, | |
| "eval_loss": 0.8778727054595947, | |
| "eval_runtime": 193.3536, | |
| "eval_samples_per_second": 13.333, | |
| "eval_steps_per_second": 3.336, | |
| "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": 7.243979410448056e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |