| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.49972360420121614, | |
| "eval_steps": 70, | |
| "global_step": 565, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0008844665561083472, | |
| "grad_norm": 51.0, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 2.9884, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0017689331122166944, | |
| "grad_norm": 45.75, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 2.9128, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0026533996683250414, | |
| "grad_norm": 50.5, | |
| "learning_rate": 1.5e-06, | |
| "loss": 2.8706, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0035378662244333887, | |
| "grad_norm": 51.75, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.9548, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.004422332780541736, | |
| "grad_norm": 49.75, | |
| "learning_rate": 2.5e-06, | |
| "loss": 2.9129, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005306799336650083, | |
| "grad_norm": 40.5, | |
| "learning_rate": 3e-06, | |
| "loss": 2.7836, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00619126589275843, | |
| "grad_norm": 35.5, | |
| "learning_rate": 3.5e-06, | |
| "loss": 2.7584, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0070757324488667775, | |
| "grad_norm": 31.375, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.7227, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.007960199004975124, | |
| "grad_norm": 30.625, | |
| "learning_rate": 4.5e-06, | |
| "loss": 2.6632, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.008844665561083471, | |
| "grad_norm": 25.375, | |
| "learning_rate": 5e-06, | |
| "loss": 2.595, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.009729132117191818, | |
| "grad_norm": 22.75, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 2.5175, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.010613598673300166, | |
| "grad_norm": 19.875, | |
| "learning_rate": 6e-06, | |
| "loss": 2.3391, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.011498065229408513, | |
| "grad_norm": 17.375, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 2.1773, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.01238253178551686, | |
| "grad_norm": 14.875, | |
| "learning_rate": 7e-06, | |
| "loss": 2.2454, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.013266998341625208, | |
| "grad_norm": 16.0, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 2.1009, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.014151464897733555, | |
| "grad_norm": 13.75, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.0104, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.015035931453841902, | |
| "grad_norm": 11.75, | |
| "learning_rate": 8.5e-06, | |
| "loss": 2.0199, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.015920398009950248, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 9e-06, | |
| "loss": 1.8737, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.016804864566058595, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 9.5e-06, | |
| "loss": 1.9285, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.017689331122166942, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6982, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01857379767827529, | |
| "grad_norm": 15.25, | |
| "learning_rate": 9.99997997403252e-06, | |
| "loss": 1.6093, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.019458264234383637, | |
| "grad_norm": 9.75, | |
| "learning_rate": 9.999919896290497e-06, | |
| "loss": 1.7027, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.020342730790491984, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 9.999819767255175e-06, | |
| "loss": 1.6295, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.02122719734660033, | |
| "grad_norm": 8.75, | |
| "learning_rate": 9.999679587728626e-06, | |
| "loss": 1.6263, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.02211166390270868, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 9.999499358833745e-06, | |
| "loss": 1.6724, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.022996130458817026, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 9.999279082014233e-06, | |
| "loss": 1.8481, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.023880597014925373, | |
| "grad_norm": 8.25, | |
| "learning_rate": 9.999018759034594e-06, | |
| "loss": 1.6871, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02476506357103372, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 9.998718391980113e-06, | |
| "loss": 1.676, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.025649530127142068, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 9.998377983256851e-06, | |
| "loss": 1.664, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.026533996683250415, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 9.99799753559161e-06, | |
| "loss": 1.7022, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.027418463239358763, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 9.997577052031922e-06, | |
| "loss": 1.6794, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.02830292979546711, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 9.997116535946028e-06, | |
| "loss": 1.5217, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.029187396351575457, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 9.996615991022835e-06, | |
| "loss": 1.6774, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.030071862907683804, | |
| "grad_norm": 4.5, | |
| "learning_rate": 9.996075421271905e-06, | |
| "loss": 1.5377, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.03095632946379215, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 9.99549483102341e-06, | |
| "loss": 1.6521, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.031840796019900496, | |
| "grad_norm": 3.375, | |
| "learning_rate": 9.9948742249281e-06, | |
| "loss": 1.5706, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.03272526257600884, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 9.994213607957273e-06, | |
| "loss": 1.6699, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03360972913211719, | |
| "grad_norm": 3.375, | |
| "learning_rate": 9.993512985402724e-06, | |
| "loss": 1.5173, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.03449419568822554, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 9.992772362876714e-06, | |
| "loss": 1.7547, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.035378662244333885, | |
| "grad_norm": 3.40625, | |
| "learning_rate": 9.991991746311916e-06, | |
| "loss": 1.6122, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03626312880044223, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 9.99117114196137e-06, | |
| "loss": 1.6103, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.03714759535655058, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 9.99031055639843e-06, | |
| "loss": 1.6719, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.03803206191265893, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 9.989409996516726e-06, | |
| "loss": 1.6497, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.038916528468767274, | |
| "grad_norm": 3.515625, | |
| "learning_rate": 9.988469469530086e-06, | |
| "loss": 1.4652, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.03980099502487562, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 9.9874889829725e-06, | |
| "loss": 1.6086, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04068546158098397, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 9.986468544698041e-06, | |
| "loss": 1.4361, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.041569928137092316, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 9.985408162880813e-06, | |
| "loss": 1.5678, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.04245439469320066, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 9.984307846014891e-06, | |
| "loss": 1.6907, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.04333886124930901, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 9.983167602914236e-06, | |
| "loss": 1.6741, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.04422332780541736, | |
| "grad_norm": 3.25, | |
| "learning_rate": 9.981987442712634e-06, | |
| "loss": 1.5514, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.045107794361525705, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 9.980767374863627e-06, | |
| "loss": 1.5447, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04599226091763405, | |
| "grad_norm": 3.265625, | |
| "learning_rate": 9.97950740914043e-06, | |
| "loss": 1.5559, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0468767274737424, | |
| "grad_norm": 3.625, | |
| "learning_rate": 9.978207555635856e-06, | |
| "loss": 1.5351, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.04776119402985075, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 9.976867824762236e-06, | |
| "loss": 1.5018, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.048645660585959094, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 9.97548822725133e-06, | |
| "loss": 1.6645, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04953012714206744, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 9.974068774154252e-06, | |
| "loss": 1.6275, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.05041459369817579, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 9.972609476841368e-06, | |
| "loss": 1.5563, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.051299060254284136, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 9.971110347002213e-06, | |
| "loss": 1.4901, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.05218352681039248, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 9.9695713966454e-06, | |
| "loss": 1.5567, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.05306799336650083, | |
| "grad_norm": 3.3125, | |
| "learning_rate": 9.967992638098517e-06, | |
| "loss": 1.5663, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05395245992260918, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 9.966374084008025e-06, | |
| "loss": 1.5848, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.054836926478717525, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 9.964715747339178e-06, | |
| "loss": 1.5302, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.05572139303482587, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 9.963017641375888e-06, | |
| "loss": 1.5876, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.05660585959093422, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 9.961279779720641e-06, | |
| "loss": 1.5807, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.05749032614704257, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 9.959502176294384e-06, | |
| "loss": 1.6079, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.058374792703150914, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 9.957684845336409e-06, | |
| "loss": 1.6045, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.05925925925925926, | |
| "grad_norm": 4.25, | |
| "learning_rate": 9.955827801404237e-06, | |
| "loss": 1.3866, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.06014372581536761, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 9.95393105937351e-06, | |
| "loss": 1.5315, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.061028192371475956, | |
| "grad_norm": 4.25, | |
| "learning_rate": 9.951994634437866e-06, | |
| "loss": 1.5193, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0619126589275843, | |
| "grad_norm": 4.0, | |
| "learning_rate": 9.950018542108818e-06, | |
| "loss": 1.5626, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0619126589275843, | |
| "eval_loss": 1.5007312297821045, | |
| "eval_runtime": 233.3408, | |
| "eval_samples_per_second": 68.912, | |
| "eval_steps_per_second": 8.614, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06279712548369265, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 9.948002798215632e-06, | |
| "loss": 1.5859, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.06368159203980099, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 9.945947418905193e-06, | |
| "loss": 1.6998, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.06456605859590935, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 9.943852420641888e-06, | |
| "loss": 1.6379, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.06545052515201769, | |
| "grad_norm": 4.625, | |
| "learning_rate": 9.941717820207461e-06, | |
| "loss": 1.6244, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.06633499170812604, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 9.939543634700891e-06, | |
| "loss": 1.5815, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06721945826423438, | |
| "grad_norm": 4.625, | |
| "learning_rate": 9.937329881538242e-06, | |
| "loss": 1.6174, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.06810392482034273, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 9.935076578452535e-06, | |
| "loss": 1.461, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.06898839137645107, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 9.932783743493599e-06, | |
| "loss": 1.6119, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.06987285793255943, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 9.930451395027932e-06, | |
| "loss": 1.592, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.07075732448866777, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 9.928079551738542e-06, | |
| "loss": 1.551, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07164179104477612, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 9.925668232624819e-06, | |
| "loss": 1.455, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.07252625760088446, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 9.923217457002356e-06, | |
| "loss": 1.6723, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.07341072415699282, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 9.92072724450282e-06, | |
| "loss": 1.4825, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.07429519071310116, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 9.91819761507377e-06, | |
| "loss": 1.5309, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.07517965726920951, | |
| "grad_norm": 8.5, | |
| "learning_rate": 9.915628588978522e-06, | |
| "loss": 1.4844, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07606412382531785, | |
| "grad_norm": 9.125, | |
| "learning_rate": 9.913020186795967e-06, | |
| "loss": 1.3599, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.07694859038142621, | |
| "grad_norm": 8.5, | |
| "learning_rate": 9.910372429420416e-06, | |
| "loss": 1.4616, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.07783305693753455, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 9.907685338061433e-06, | |
| "loss": 1.579, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.0787175234936429, | |
| "grad_norm": 9.375, | |
| "learning_rate": 9.904958934243655e-06, | |
| "loss": 1.4275, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.07960199004975124, | |
| "grad_norm": 7.9375, | |
| "learning_rate": 9.902193239806634e-06, | |
| "loss": 1.593, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0804864566058596, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 9.899388276904653e-06, | |
| "loss": 1.4569, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.08137092316196794, | |
| "grad_norm": 8.5, | |
| "learning_rate": 9.89654406800655e-06, | |
| "loss": 1.4831, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.08225538971807629, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 9.893660635895542e-06, | |
| "loss": 1.4014, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.08313985627418463, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 1.5062, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.08402432283029299, | |
| "grad_norm": 5.5, | |
| "learning_rate": 9.887776194738433e-06, | |
| "loss": 1.6041, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.08490878938640133, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 9.884775232828985e-06, | |
| "loss": 1.3051, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.08579325594250968, | |
| "grad_norm": 5.3125, | |
| "learning_rate": 9.881735141979554e-06, | |
| "loss": 1.5133, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.08667772249861802, | |
| "grad_norm": 4.875, | |
| "learning_rate": 9.878655946542443e-06, | |
| "loss": 1.6156, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.08756218905472637, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 9.8755376711832e-06, | |
| "loss": 1.4838, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.08844665561083472, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 9.872380340880416e-06, | |
| "loss": 1.4328, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08933112216694307, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 9.869183980925531e-06, | |
| "loss": 1.4843, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.09021558872305141, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 9.865948616922623e-06, | |
| "loss": 1.263, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.09110005527915975, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 9.862674274788212e-06, | |
| "loss": 1.3736, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.0919845218352681, | |
| "grad_norm": 3.984375, | |
| "learning_rate": 9.85936098075104e-06, | |
| "loss": 1.3568, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.09286898839137644, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 9.856008761351882e-06, | |
| "loss": 1.3707, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0937534549474848, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 9.85261764344331e-06, | |
| "loss": 1.3825, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.09463792150359314, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 9.849187654189486e-06, | |
| "loss": 1.4497, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.0955223880597015, | |
| "grad_norm": 2.875, | |
| "learning_rate": 9.845718821065957e-06, | |
| "loss": 1.5509, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.09640685461580983, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 9.842211171859418e-06, | |
| "loss": 1.3991, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.09729132117191819, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 9.838664734667496e-06, | |
| "loss": 1.4616, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.09817578772802653, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 9.835079537898523e-06, | |
| "loss": 1.3058, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.09906025428413488, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 9.831455610271315e-06, | |
| "loss": 1.3997, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.09994472084024322, | |
| "grad_norm": 2.25, | |
| "learning_rate": 9.827792980814934e-06, | |
| "loss": 1.4942, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.10082918739635158, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 9.82409167886846e-06, | |
| "loss": 1.4756, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.10171365395245992, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 9.820351734080754e-06, | |
| "loss": 1.4339, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.10259812050856827, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 9.81657317641022e-06, | |
| "loss": 1.4626, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.10348258706467661, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 9.81275603612457e-06, | |
| "loss": 1.3905, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.10436705362078497, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 9.808900343800568e-06, | |
| "loss": 1.3814, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1052515201768933, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 9.80500613032381e-06, | |
| "loss": 1.5248, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.10613598673300166, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 9.801073426888447e-06, | |
| "loss": 1.4267, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.10702045328911, | |
| "grad_norm": 3.328125, | |
| "learning_rate": 9.79710226499696e-06, | |
| "loss": 1.297, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.10790491984521836, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 9.79309267645989e-06, | |
| "loss": 1.4822, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.1087893864013267, | |
| "grad_norm": 3.375, | |
| "learning_rate": 9.789044693395593e-06, | |
| "loss": 1.5098, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.10967385295743505, | |
| "grad_norm": 2.375, | |
| "learning_rate": 9.784958348229978e-06, | |
| "loss": 1.5161, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.11055831951354339, | |
| "grad_norm": 2.625, | |
| "learning_rate": 9.780833673696255e-06, | |
| "loss": 1.636, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.11144278606965174, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 9.776670702834662e-06, | |
| "loss": 1.4871, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.11232725262576009, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 9.772469468992206e-06, | |
| "loss": 1.5421, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.11321171918186844, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 9.768230005822394e-06, | |
| "loss": 1.4452, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.11409618573797678, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 9.76395234728497e-06, | |
| "loss": 1.3874, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.11498065229408513, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 9.759636527645633e-06, | |
| "loss": 1.4366, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11586511885019347, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 1.4365, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.11674958540630183, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 9.750890543652169e-06, | |
| "loss": 1.4513, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.11763405196241017, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 9.74646044935676e-06, | |
| "loss": 1.4241, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 9.741992334076309e-06, | |
| "loss": 1.5274, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.11940298507462686, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 9.737486233602149e-06, | |
| "loss": 1.3942, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.12028745163073522, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 9.73294218402989e-06, | |
| "loss": 1.4371, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.12117191818684356, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 9.728360221759125e-06, | |
| "loss": 1.4276, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.12205638474295191, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 9.723740383493146e-06, | |
| "loss": 1.4581, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.12294085129906025, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 9.719082706238645e-06, | |
| "loss": 1.4475, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1238253178551686, | |
| "grad_norm": 2.25, | |
| "learning_rate": 9.714387227305422e-06, | |
| "loss": 1.4863, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1238253178551686, | |
| "eval_loss": 1.3686763048171997, | |
| "eval_runtime": 232.4295, | |
| "eval_samples_per_second": 69.182, | |
| "eval_steps_per_second": 8.648, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12470978441127695, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 9.709653984306077e-06, | |
| "loss": 1.3077, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.1255942509673853, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 9.70488301515572e-06, | |
| "loss": 1.2957, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.12647871752349366, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 9.700074358071658e-06, | |
| "loss": 1.2871, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.12736318407960198, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 9.695228051573098e-06, | |
| "loss": 1.2346, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.12824765063571034, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 9.69034413448083e-06, | |
| "loss": 1.5424, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1291321171918187, | |
| "grad_norm": 3.53125, | |
| "learning_rate": 9.68542264591692e-06, | |
| "loss": 1.4396, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.13001658374792704, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 9.680463625304393e-06, | |
| "loss": 1.3637, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.13090105030403537, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 9.675467112366925e-06, | |
| "loss": 1.5764, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.13178551686014373, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 9.670433147128522e-06, | |
| "loss": 1.3826, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.13266998341625208, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 9.665361769913187e-06, | |
| "loss": 1.4979, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.13355444997236043, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.660253021344618e-06, | |
| "loss": 1.3917, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.13443891652846876, | |
| "grad_norm": 2.625, | |
| "learning_rate": 9.65510694234587e-06, | |
| "loss": 1.5089, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.13532338308457711, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 9.649923574139023e-06, | |
| "loss": 1.3935, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.13620784964068547, | |
| "grad_norm": 2.78125, | |
| "learning_rate": 9.644702958244865e-06, | |
| "loss": 1.3694, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.1370923161967938, | |
| "grad_norm": 2.5, | |
| "learning_rate": 9.639445136482549e-06, | |
| "loss": 1.3446, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.13797678275290215, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 9.634150150969259e-06, | |
| "loss": 1.4295, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.1388612493090105, | |
| "grad_norm": 2.25, | |
| "learning_rate": 9.628818044119884e-06, | |
| "loss": 1.3697, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.13974571586511886, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 9.623448858646658e-06, | |
| "loss": 1.392, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.14063018242122718, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 9.618042637558838e-06, | |
| "loss": 1.4678, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.14151464897733554, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 9.612599424162344e-06, | |
| "loss": 1.3916, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1423991155334439, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 9.607119262059426e-06, | |
| "loss": 1.4652, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.14328358208955225, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 9.601602195148299e-06, | |
| "loss": 1.4551, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.14416804864566057, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 9.596048267622806e-06, | |
| "loss": 1.4511, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.14505251520176893, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 9.590457523972055e-06, | |
| "loss": 1.4071, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.14593698175787728, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 9.584830008980068e-06, | |
| "loss": 1.4284, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.14682144831398564, | |
| "grad_norm": 2.375, | |
| "learning_rate": 9.579165767725416e-06, | |
| "loss": 1.4594, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.14770591487009396, | |
| "grad_norm": 2.375, | |
| "learning_rate": 9.573464845580864e-06, | |
| "loss": 1.4633, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.14859038142620232, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 1.56, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.14947484798231067, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 9.561953141581895e-06, | |
| "loss": 1.3288, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.15035931453841903, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 9.55614245194068e-06, | |
| "loss": 1.3863, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.15124378109452735, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.550295265835235e-06, | |
| "loss": 1.3928, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.1521282476506357, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 9.544411630103782e-06, | |
| "loss": 1.4527, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.15301271420674406, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 9.538491591876522e-06, | |
| "loss": 1.3693, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.15389718076285241, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 9.53253519857525e-06, | |
| "loss": 1.2812, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.15478164731896074, | |
| "grad_norm": 2.125, | |
| "learning_rate": 9.526542497912984e-06, | |
| "loss": 1.4142, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1556661138750691, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 9.520513537893574e-06, | |
| "loss": 1.4069, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.15655058043117745, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 9.514448366811322e-06, | |
| "loss": 1.3569, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.1574350469872858, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 9.508347033250597e-06, | |
| "loss": 1.4351, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.15831951354339413, | |
| "grad_norm": 2.375, | |
| "learning_rate": 9.502209586085444e-06, | |
| "loss": 1.4548, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.15920398009950248, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 9.496036074479184e-06, | |
| "loss": 1.5012, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.16008844665561084, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 9.48982654788404e-06, | |
| "loss": 1.4076, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.1609729132117192, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 9.48358105604072e-06, | |
| "loss": 1.3964, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.16185737976782752, | |
| "grad_norm": 2.5, | |
| "learning_rate": 9.47729964897803e-06, | |
| "loss": 1.4238, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.16274184632393587, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.470982377012475e-06, | |
| "loss": 1.3678, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.16362631288004423, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 9.464629290747844e-06, | |
| "loss": 1.4618, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.16451077943615258, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 9.458240441074818e-06, | |
| "loss": 1.3658, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.1653952459922609, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 9.451815879170556e-06, | |
| "loss": 1.4175, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.16627971254836926, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.445355656498284e-06, | |
| "loss": 1.4043, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.16716417910447762, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 9.438859824806889e-06, | |
| "loss": 1.3923, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.16804864566058597, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 9.432328436130493e-06, | |
| "loss": 1.3122, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1689331122166943, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 9.425761542788049e-06, | |
| "loss": 1.4381, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.16981757877280265, | |
| "grad_norm": 2.375, | |
| "learning_rate": 9.419159197382913e-06, | |
| "loss": 1.3126, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.170702045328911, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 9.412521452802425e-06, | |
| "loss": 1.4128, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.17158651188501936, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 9.40584836221749e-06, | |
| "loss": 1.2311, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.1724709784411277, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.399139979082148e-06, | |
| "loss": 1.402, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.17335544499723604, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 9.39239635713314e-06, | |
| "loss": 1.3014, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.1742399115533444, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 9.38561755038949e-06, | |
| "loss": 1.3299, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.17512437810945275, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.37880361315206e-06, | |
| "loss": 1.4074, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.17600884466556108, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 9.37195460000313e-06, | |
| "loss": 1.306, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.17689331122166943, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 9.365070565805941e-06, | |
| "loss": 1.445, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 9.358151565704275e-06, | |
| "loss": 1.3732, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.17866224433388614, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.351197655121995e-06, | |
| "loss": 1.3544, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.17954671088999447, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 9.34420888976262e-06, | |
| "loss": 1.367, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.18043117744610282, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 9.337185325608864e-06, | |
| "loss": 1.4574, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.18131564400221117, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 1.4689, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.1822001105583195, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 9.323034026242378e-06, | |
| "loss": 1.2427, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.18308457711442785, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 9.315906404387032e-06, | |
| "loss": 1.3404, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.1839690436705362, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 9.308744210451167e-06, | |
| "loss": 1.4636, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.18485351022664456, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 9.301547501806725e-06, | |
| "loss": 1.3681, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.1857379767827529, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.294316336102132e-06, | |
| "loss": 1.3969, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1857379767827529, | |
| "eval_loss": 1.3509321212768555, | |
| "eval_runtime": 232.201, | |
| "eval_samples_per_second": 69.25, | |
| "eval_steps_per_second": 8.656, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.18662244333886124, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 9.287050771261818e-06, | |
| "loss": 1.3528, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.1875069098949696, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 9.279750865485772e-06, | |
| "loss": 1.3776, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.18839137645107795, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 9.272416677249067e-06, | |
| "loss": 1.3311, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.18927584300718628, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 9.265048265301385e-06, | |
| "loss": 1.4167, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.19016030956329463, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 9.257645688666557e-06, | |
| "loss": 1.2651, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.191044776119403, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 9.25020900664209e-06, | |
| "loss": 1.4641, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.19192924267551134, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 9.242738278798682e-06, | |
| "loss": 1.3616, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.19281370923161967, | |
| "grad_norm": 2.25, | |
| "learning_rate": 9.235233564979756e-06, | |
| "loss": 1.4412, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.19369817578772802, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 9.227694925300972e-06, | |
| "loss": 1.372, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.19458264234383638, | |
| "grad_norm": 2.375, | |
| "learning_rate": 9.220122420149753e-06, | |
| "loss": 1.4577, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.19546710889994473, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 9.212516110184794e-06, | |
| "loss": 1.2607, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.19635157545605306, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 9.204876056335584e-06, | |
| "loss": 1.394, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.1972360420121614, | |
| "grad_norm": 2.875, | |
| "learning_rate": 9.197202319801908e-06, | |
| "loss": 1.4744, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.19812050856826977, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 9.18949496205337e-06, | |
| "loss": 1.3798, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.19900497512437812, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 9.181754044828882e-06, | |
| "loss": 1.5303, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.19988944168048645, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 9.17397963013619e-06, | |
| "loss": 1.5118, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.2007739082365948, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 9.166171780251365e-06, | |
| "loss": 1.4777, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.20165837479270315, | |
| "grad_norm": 2.875, | |
| "learning_rate": 9.158330557718304e-06, | |
| "loss": 1.3949, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.2025428413488115, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 9.150456025348236e-06, | |
| "loss": 1.6279, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.20342730790491984, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 9.142548246219212e-06, | |
| "loss": 1.3632, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2043117744610282, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 9.134607283675601e-06, | |
| "loss": 1.3441, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.20519624101713654, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 9.12663320132759e-06, | |
| "loss": 1.402, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.2060807075732449, | |
| "grad_norm": 3.375, | |
| "learning_rate": 9.118626063050661e-06, | |
| "loss": 1.5275, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.20696517412935322, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 9.110585932985095e-06, | |
| "loss": 1.38, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.20784964068546158, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 9.102512875535439e-06, | |
| "loss": 1.3739, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.20873410724156993, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 9.09440695537001e-06, | |
| "loss": 1.495, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.2096185737976783, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 9.086268237420368e-06, | |
| "loss": 1.3662, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.2105030403537866, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 9.07809678688079e-06, | |
| "loss": 1.4109, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.21138750690989497, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 9.069892669207757e-06, | |
| "loss": 1.3512, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.21227197346600332, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 9.06165595011943e-06, | |
| "loss": 1.4568, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.21315644002211168, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 9.053386695595114e-06, | |
| "loss": 1.4943, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.21404090657822, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 1.503, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.21492537313432836, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.036750845458323e-06, | |
| "loss": 1.4121, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2158098396904367, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 9.028384383105444e-06, | |
| "loss": 1.3212, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.21669430624654507, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 9.019985651834703e-06, | |
| "loss": 1.3527, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2175787728026534, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 9.01155471892319e-06, | |
| "loss": 1.246, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.21846323935876175, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 9.00309165190594e-06, | |
| "loss": 1.4277, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.2193477059148701, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 8.994596518575393e-06, | |
| "loss": 1.4682, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.22023217247097845, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.986069386980855e-06, | |
| "loss": 1.3714, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.22111663902708678, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.97751032542795e-06, | |
| "loss": 1.3973, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.22200110558319514, | |
| "grad_norm": 2.5, | |
| "learning_rate": 8.968919402478076e-06, | |
| "loss": 1.4555, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.2228855721393035, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.960296686947846e-06, | |
| "loss": 1.22, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.22377003869541182, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.95164224790855e-06, | |
| "loss": 1.3659, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.22465450525152017, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 8.942956154685596e-06, | |
| "loss": 1.3025, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.22553897180762852, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.93423847685795e-06, | |
| "loss": 1.3139, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.22642343836373688, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.925489284257587e-06, | |
| "loss": 1.4009, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.2273079049198452, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 8.916708646968924e-06, | |
| "loss": 1.4532, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.22819237147595356, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.907896635328262e-06, | |
| "loss": 1.387, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.2290768380320619, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.899053319923228e-06, | |
| "loss": 1.3644, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.22996130458817027, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 8.890178771592198e-06, | |
| "loss": 1.4128, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2308457711442786, | |
| "grad_norm": 2.25, | |
| "learning_rate": 8.881273061423741e-06, | |
| "loss": 1.5197, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.23173023770038695, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.87233626075604e-06, | |
| "loss": 1.3812, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.2326147042564953, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 8.863368441176326e-06, | |
| "loss": 1.2539, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.23349917081260366, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 8.854369674520306e-06, | |
| "loss": 1.4247, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.23438363736871198, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 8.845340032871584e-06, | |
| "loss": 1.3402, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.23526810392482034, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 8.836279588561084e-06, | |
| "loss": 1.3674, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.2361525704809287, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 8.827188414166468e-06, | |
| "loss": 1.4622, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 8.818066582511564e-06, | |
| "loss": 1.323, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.23792150359314537, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 8.808914166665773e-06, | |
| "loss": 1.4084, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.23880597014925373, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 8.799731239943488e-06, | |
| "loss": 1.3691, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.23969043670536208, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 8.790517875903506e-06, | |
| "loss": 1.2452, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.24057490326147044, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.781274148348438e-06, | |
| "loss": 1.3614, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.24145936981757876, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.772000131324117e-06, | |
| "loss": 1.4875, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.24234383637368712, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.762695899119012e-06, | |
| "loss": 1.2157, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.24322830292979547, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 8.753361526263622e-06, | |
| "loss": 1.6414, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.24411276948590382, | |
| "grad_norm": 2.375, | |
| "learning_rate": 8.743997087529886e-06, | |
| "loss": 1.2842, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.24499723604201215, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 8.734602657930582e-06, | |
| "loss": 1.4622, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.2458817025981205, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.725178312718727e-06, | |
| "loss": 1.4923, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.24676616915422886, | |
| "grad_norm": 2.25, | |
| "learning_rate": 8.715724127386971e-06, | |
| "loss": 1.4421, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2476506357103372, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 8.706240177667003e-06, | |
| "loss": 1.3777, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2476506357103372, | |
| "eval_loss": 1.340869665145874, | |
| "eval_runtime": 232.0202, | |
| "eval_samples_per_second": 69.304, | |
| "eval_steps_per_second": 8.663, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.24853510226644554, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 8.696726539528924e-06, | |
| "loss": 1.2984, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.2494195688225539, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 8.687183289180659e-06, | |
| "loss": 1.4034, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.25030403537866225, | |
| "grad_norm": 2.125, | |
| "learning_rate": 8.67761050306734e-06, | |
| "loss": 1.4313, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.2511885019347706, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.668008257870684e-06, | |
| "loss": 1.3227, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.25207296849087896, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 8.658376630508391e-06, | |
| "loss": 1.3931, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.2529574350469873, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.648715698133529e-06, | |
| "loss": 1.4474, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.2538419016030956, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 8.639025538133899e-06, | |
| "loss": 1.204, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.25472636815920396, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 8.629306228131435e-06, | |
| "loss": 1.1931, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.2556108347153123, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 8.619557845981571e-06, | |
| "loss": 1.4676, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.2564953012714207, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 8.609780469772623e-06, | |
| "loss": 1.3624, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.257379767827529, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 8.599974177825155e-06, | |
| "loss": 1.3875, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.2582642343836374, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 8.590139048691361e-06, | |
| "loss": 1.2976, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.25914870093974574, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.580275161154432e-06, | |
| "loss": 1.4271, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.2600331674958541, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 8.570382594227923e-06, | |
| "loss": 1.377, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.2609176340519624, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 8.56046142715513e-06, | |
| "loss": 1.3283, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.26180210060807074, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 8.550511739408428e-06, | |
| "loss": 1.3989, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.2626865671641791, | |
| "grad_norm": 2.125, | |
| "learning_rate": 8.540533610688679e-06, | |
| "loss": 1.3885, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.26357103372028745, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.53052712092455e-06, | |
| "loss": 1.4217, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.2644555002763958, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.520492350271895e-06, | |
| "loss": 1.2601, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.26533996683250416, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.510429379113114e-06, | |
| "loss": 1.4437, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2662244333886125, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 8.500338288056496e-06, | |
| "loss": 1.3637, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.26710889994472087, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 8.490219157935589e-06, | |
| "loss": 1.3865, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.26799336650082917, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 8.480072069808541e-06, | |
| "loss": 1.4416, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.2688778330569375, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 8.469897104957454e-06, | |
| "loss": 1.3866, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.2697622996130459, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 8.459694344887732e-06, | |
| "loss": 1.4329, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.27064676616915423, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 8.449463871327434e-06, | |
| "loss": 1.4178, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.2715312327252626, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 8.439205766226613e-06, | |
| "loss": 1.3813, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.27241569928137094, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.428920111756658e-06, | |
| "loss": 1.5067, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.2733001658374793, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 8.418606990309645e-06, | |
| "loss": 1.4275, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.2741846323935876, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 8.408266484497664e-06, | |
| "loss": 1.3738, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.27506909894969594, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.397898677152173e-06, | |
| "loss": 1.2003, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.2759535655058043, | |
| "grad_norm": 2.25, | |
| "learning_rate": 8.387503651323317e-06, | |
| "loss": 1.3476, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.27683803206191265, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 8.377081490279278e-06, | |
| "loss": 1.4104, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.277722498618021, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 8.366632277505598e-06, | |
| "loss": 1.3786, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.27860696517412936, | |
| "grad_norm": 2.125, | |
| "learning_rate": 8.356156096704516e-06, | |
| "loss": 1.4073, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.2794914317302377, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 1.5986, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.28037589828634607, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 8.335123166908544e-06, | |
| "loss": 1.4255, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.28126036484245437, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.324566586395563e-06, | |
| "loss": 1.4524, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.2821448313985627, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.313983374817643e-06, | |
| "loss": 1.4044, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.2830292979546711, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.303373616950408e-06, | |
| "loss": 1.3609, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.28391376451077943, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 8.29273739778212e-06, | |
| "loss": 1.5053, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.2847982310668878, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 8.282074802513016e-06, | |
| "loss": 1.3513, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.28568269762299614, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.271385916554605e-06, | |
| "loss": 1.4534, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.2865671641791045, | |
| "grad_norm": 2.375, | |
| "learning_rate": 8.260670825529002e-06, | |
| "loss": 1.2827, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.28745163073521285, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.249929615268234e-06, | |
| "loss": 1.3483, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.28833609729132115, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 8.239162371813552e-06, | |
| "loss": 1.3943, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.2892205638474295, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.22836918141474e-06, | |
| "loss": 1.3578, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.29010503040353786, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 8.217550130529433e-06, | |
| "loss": 1.3517, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.2909894969596462, | |
| "grad_norm": 2.375, | |
| "learning_rate": 8.206705305822414e-06, | |
| "loss": 1.422, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.29187396351575456, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 8.195834794164925e-06, | |
| "loss": 1.3983, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.2927584300718629, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 8.184938682633973e-06, | |
| "loss": 1.3837, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.2936428966279713, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 8.17401705851163e-06, | |
| "loss": 1.4879, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.2945273631840796, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 8.163070009284326e-06, | |
| "loss": 1.3282, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.2954118297401879, | |
| "grad_norm": 2.375, | |
| "learning_rate": 8.152097622642167e-06, | |
| "loss": 1.3357, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 8.141099986478212e-06, | |
| "loss": 1.2913, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.29718076285240463, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 8.130077188887786e-06, | |
| "loss": 1.4665, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.298065229408513, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 8.119029318167763e-06, | |
| "loss": 1.2694, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.29894969596462134, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 8.107956462815862e-06, | |
| "loss": 1.3418, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.2998341625207297, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 8.096858711529938e-06, | |
| "loss": 1.4219, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.30071862907683805, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.085736153207277e-06, | |
| "loss": 1.4296, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3016030956329464, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 8.074588876943872e-06, | |
| "loss": 1.3679, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.3024875621890547, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 8.063416972033725e-06, | |
| "loss": 1.2633, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.30337202874516306, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 8.05222052796811e-06, | |
| "loss": 1.4455, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.3042564953012714, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 8.040999634434883e-06, | |
| "loss": 1.35, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.30514096185737977, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.029754381317741e-06, | |
| "loss": 1.3095, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3060254284134881, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.018484858695514e-06, | |
| "loss": 1.4285, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.3069098949695965, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.00719115684144e-06, | |
| "loss": 1.3694, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.30779436152570483, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 7.995873366222442e-06, | |
| "loss": 1.5045, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.3086788280818132, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 7.9845315774984e-06, | |
| "loss": 1.4376, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.3095632946379215, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.973165881521435e-06, | |
| "loss": 1.3382, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3095632946379215, | |
| "eval_loss": 1.3338196277618408, | |
| "eval_runtime": 231.9317, | |
| "eval_samples_per_second": 69.331, | |
| "eval_steps_per_second": 8.666, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.31044776119402984, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 7.961776369335164e-06, | |
| "loss": 1.4081, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.3113322277501382, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.950363132173992e-06, | |
| "loss": 1.4814, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.31221669430624654, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 1.4584, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.3131011608623549, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 7.927465848814045e-06, | |
| "loss": 1.5012, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.31398562741846325, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 7.915981986031367e-06, | |
| "loss": 1.4007, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3148700939745716, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.90447476510452e-06, | |
| "loss": 1.2796, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.3157545605306799, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 7.892944278210796e-06, | |
| "loss": 1.2907, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.31663902708678826, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 7.88139061771386e-06, | |
| "loss": 1.4433, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.3175234936428966, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 7.869813876162997e-06, | |
| "loss": 1.2836, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.31840796019900497, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.858214146292394e-06, | |
| "loss": 1.3365, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3192924267551133, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.846591521020373e-06, | |
| "loss": 1.437, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.3201768933112217, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.834946093448658e-06, | |
| "loss": 1.2714, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.32106135986733003, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.823277956861635e-06, | |
| "loss": 1.4181, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.3219458264234384, | |
| "grad_norm": 2.25, | |
| "learning_rate": 7.811587204725593e-06, | |
| "loss": 1.4788, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.3228302929795467, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 7.799873930687979e-06, | |
| "loss": 1.3304, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.32371475953565504, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 7.78813822857665e-06, | |
| "loss": 1.3729, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.3245992260917634, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 7.776380192399122e-06, | |
| "loss": 1.3222, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.32548369264787175, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.764599916341817e-06, | |
| "loss": 1.3088, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.3263681592039801, | |
| "grad_norm": 2.5, | |
| "learning_rate": 7.752797494769307e-06, | |
| "loss": 1.3756, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.32725262576008846, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 7.74097302222355e-06, | |
| "loss": 1.3111, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3281370923161968, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.729126593423151e-06, | |
| "loss": 1.386, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.32902155887230516, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.71725830326259e-06, | |
| "loss": 1.3747, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.32990602542841346, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 7.70536824681146e-06, | |
| "loss": 1.3323, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.3307904919845218, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.69345651931372e-06, | |
| "loss": 1.4747, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.33167495854063017, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 7.681523216186912e-06, | |
| "loss": 1.3596, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3325594250967385, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 7.669568433021418e-06, | |
| "loss": 1.2835, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.3334438916528469, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.65759226557967e-06, | |
| "loss": 1.3523, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.33432835820895523, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 7.64559480979541e-06, | |
| "loss": 1.327, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.3352128247650636, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 7.633576161772898e-06, | |
| "loss": 1.2579, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.33609729132117194, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 7.621536417786159e-06, | |
| "loss": 1.362, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.33698175787728024, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 7.6094756742782e-06, | |
| "loss": 1.3091, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.3378662244333886, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 7.597394027860243e-06, | |
| "loss": 1.4317, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.33875069098949695, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.585291575310952e-06, | |
| "loss": 1.3915, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.3396351575456053, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 7.5731684135756566e-06, | |
| "loss": 1.3549, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.34051962410171366, | |
| "grad_norm": 2.25, | |
| "learning_rate": 7.5610246397655715e-06, | |
| "loss": 1.3647, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.341404090657822, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.548860351157028e-06, | |
| "loss": 1.3884, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.34228855721393037, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 7.536675645190682e-06, | |
| "loss": 1.3217, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.3431730237700387, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 7.524470619470745e-06, | |
| "loss": 1.2801, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.344057490326147, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 7.512245371764197e-06, | |
| "loss": 1.4198, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.3449419568822554, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.547, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.34582642343836373, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.487734602268326e-06, | |
| "loss": 1.3439, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.3467108899944721, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 7.475449276819753e-06, | |
| "loss": 1.4559, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.34759535655058044, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 7.463144122064494e-06, | |
| "loss": 1.3188, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.3484798231066888, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 7.4508192365716005e-06, | |
| "loss": 1.3856, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.34936428966279715, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 7.438474719068174e-06, | |
| "loss": 1.3753, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.3502487562189055, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 7.426110668438578e-06, | |
| "loss": 1.3117, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.3511332227750138, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 7.413727183723642e-06, | |
| "loss": 1.3912, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.35201768933112215, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.401324364119872e-06, | |
| "loss": 1.3095, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.3529021558872305, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.388902308978652e-06, | |
| "loss": 1.3545, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.35378662244333886, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.37646111780545e-06, | |
| "loss": 1.4058, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3546710889994472, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 7.364000890259024e-06, | |
| "loss": 1.3587, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 7.351521726150619e-06, | |
| "loss": 1.348, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.3564400221116639, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.339023725443166e-06, | |
| "loss": 1.322, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.3573244886677723, | |
| "grad_norm": 2.25, | |
| "learning_rate": 7.326506988250488e-06, | |
| "loss": 1.3455, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.3582089552238806, | |
| "grad_norm": 2.25, | |
| "learning_rate": 7.313971614836496e-06, | |
| "loss": 1.3546, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.35909342177998893, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.3014177056143795e-06, | |
| "loss": 1.4034, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.3599778883360973, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 7.288845361145812e-06, | |
| "loss": 1.2737, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.36086235489220564, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.276254682140137e-06, | |
| "loss": 1.3589, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.361746821448314, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 7.2636457694535655e-06, | |
| "loss": 1.4803, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.36263128800442235, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 7.251018724088367e-06, | |
| "loss": 1.3554, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3635157545605307, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 7.238373647192063e-06, | |
| "loss": 1.338, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.364400221116639, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 7.225710640056614e-06, | |
| "loss": 1.3909, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.36528468767274735, | |
| "grad_norm": 2.25, | |
| "learning_rate": 7.213029804117604e-06, | |
| "loss": 1.33, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.3661691542288557, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.200331240953439e-06, | |
| "loss": 1.4574, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.36705362078496406, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 7.187615052284522e-06, | |
| "loss": 1.4585, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.3679380873410724, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 7.174881339972448e-06, | |
| "loss": 1.4293, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.36882255389718077, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 7.16213020601918e-06, | |
| "loss": 1.3263, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.3697070204532891, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 7.149361752566233e-06, | |
| "loss": 1.3842, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.3705914870093975, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 7.136576081893863e-06, | |
| "loss": 1.3817, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.3714759535655058, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.12377329642024e-06, | |
| "loss": 1.4406, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.3714759535655058, | |
| "eval_loss": 1.3291162252426147, | |
| "eval_runtime": 231.9227, | |
| "eval_samples_per_second": 69.333, | |
| "eval_steps_per_second": 8.667, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.37236042012161413, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.110953498700629e-06, | |
| "loss": 1.4744, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.3732448866777225, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.09811679142657e-06, | |
| "loss": 1.39, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.37412935323383084, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.085263277425058e-06, | |
| "loss": 1.5196, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.3750138197899392, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.072393059657716e-06, | |
| "loss": 1.4646, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.37589828634604755, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.059506241219964e-06, | |
| "loss": 1.3383, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.3767827529021559, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.046602925340209e-06, | |
| "loss": 1.4669, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.37766721945826426, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.033683215379002e-06, | |
| "loss": 1.3031, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.37855168601437256, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 7.020747214828221e-06, | |
| "loss": 1.4132, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.3794361525704809, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 7.007795027310236e-06, | |
| "loss": 1.3288, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.38032061912658927, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.994826756577082e-06, | |
| "loss": 1.3066, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3812050856826976, | |
| "grad_norm": 3.234375, | |
| "learning_rate": 6.981842506509626e-06, | |
| "loss": 1.4464, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.382089552238806, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 6.968842381116736e-06, | |
| "loss": 1.3772, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.38297401879491433, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 6.955826484534447e-06, | |
| "loss": 1.3471, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.3838584853510227, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 6.942794921025127e-06, | |
| "loss": 1.3815, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.38474295190713104, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 6.9297477949766445e-06, | |
| "loss": 1.3692, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.38562741846323934, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 6.916685210901527e-06, | |
| "loss": 1.4499, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.3865118850193477, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 6.903607273436128e-06, | |
| "loss": 1.4358, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.38739635157545604, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.890514087339789e-06, | |
| "loss": 1.4993, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.3882808181315644, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.8774057574939956e-06, | |
| "loss": 1.3827, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.38916528468767275, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.864282388901544e-06, | |
| "loss": 1.3918, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3900497512437811, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 6.851144086685695e-06, | |
| "loss": 1.3348, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.39093421779988946, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 6.837990956089334e-06, | |
| "loss": 1.4058, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.3918186843559978, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 6.824823102474127e-06, | |
| "loss": 1.2347, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.3927031509121061, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 6.811640631319677e-06, | |
| "loss": 1.3793, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.39358761746821447, | |
| "grad_norm": 2.375, | |
| "learning_rate": 6.79844364822268e-06, | |
| "loss": 1.406, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.3944720840243228, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 6.785232258896078e-06, | |
| "loss": 1.3245, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.3953565505804312, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 6.77200656916821e-06, | |
| "loss": 1.4432, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.39624101713653953, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.758766684981972e-06, | |
| "loss": 1.2411, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.3971254836926479, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 6.745512712393958e-06, | |
| "loss": 1.3898, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.39800995024875624, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 6.732244757573619e-06, | |
| "loss": 1.3282, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.3988944168048646, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 6.718962926802407e-06, | |
| "loss": 1.4219, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.3997788833609729, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 6.705667326472926e-06, | |
| "loss": 1.4502, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.40066334991708125, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 6.69235806308808e-06, | |
| "loss": 1.4133, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.4015478164731896, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 6.67903524326022e-06, | |
| "loss": 1.359, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.40243228302929795, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.665698973710289e-06, | |
| "loss": 1.4398, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.4033167495854063, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 6.652349361266967e-06, | |
| "loss": 1.4904, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.40420121614151466, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 6.638986512865814e-06, | |
| "loss": 1.4179, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.405085682697623, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.625610535548418e-06, | |
| "loss": 1.346, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.4059701492537313, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.612221536461534e-06, | |
| "loss": 1.2897, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.40685461580983967, | |
| "grad_norm": 2.25, | |
| "learning_rate": 6.598819622856227e-06, | |
| "loss": 1.3319, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.407739082365948, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.585404902087011e-06, | |
| "loss": 1.3922, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.4086235489220564, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 6.571977481610989e-06, | |
| "loss": 1.3128, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.40950801547816473, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.558537468986998e-06, | |
| "loss": 1.3726, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.4103924820342731, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 1.3356, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.41127694859038144, | |
| "grad_norm": 2.25, | |
| "learning_rate": 6.531620098033919e-06, | |
| "loss": 1.413, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.4121614151464898, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 6.518142955323389e-06, | |
| "loss": 1.4754, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.4130458817025981, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 6.504653651700278e-06, | |
| "loss": 1.3714, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.41393034825870645, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 6.4911522952191275e-06, | |
| "loss": 1.513, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.4148148148148148, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 6.477638994031028e-06, | |
| "loss": 1.3935, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.41569928137092316, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 6.464113856382752e-06, | |
| "loss": 1.3362, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4165837479270315, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 6.450576990615888e-06, | |
| "loss": 1.4558, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.41746821448313987, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 6.437028505165965e-06, | |
| "loss": 1.3723, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.4183526810392482, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 6.423468508561599e-06, | |
| "loss": 1.3712, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.4192371475953566, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 6.409897109423607e-06, | |
| "loss": 1.3105, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.4201216141514649, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 6.396314416464151e-06, | |
| "loss": 1.5579, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.4210060807075732, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 6.382720538485856e-06, | |
| "loss": 1.3702, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.4218905472636816, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 6.3691155843809475e-06, | |
| "loss": 1.467, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.42277501381978994, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 6.355499663130369e-06, | |
| "loss": 1.243, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.4236594803758983, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 6.341872883802923e-06, | |
| "loss": 1.4165, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.42454394693200664, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.328235355554382e-06, | |
| "loss": 1.3261, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.425428413488115, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 6.314587187626628e-06, | |
| "loss": 1.4347, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.42631288004422335, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 6.3009284893467655e-06, | |
| "loss": 1.4363, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.42719734660033165, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.287259370126255e-06, | |
| "loss": 1.3918, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.42808181315644, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 6.273579939460032e-06, | |
| "loss": 1.2938, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.42896627971254836, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 6.259890306925627e-06, | |
| "loss": 1.3493, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.4298507462686567, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 6.246190582182296e-06, | |
| "loss": 1.3912, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.43073521282476507, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 6.232480874970137e-06, | |
| "loss": 1.3116, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.4316196793808734, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 6.218761295109209e-06, | |
| "loss": 1.3492, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.4325041459369818, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.205031952498656e-06, | |
| "loss": 1.2913, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.43338861249309013, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.191292957115825e-06, | |
| "loss": 1.4216, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.43338861249309013, | |
| "eval_loss": 1.3258094787597656, | |
| "eval_runtime": 232.1391, | |
| "eval_samples_per_second": 69.269, | |
| "eval_steps_per_second": 8.659, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.43427307904919843, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 6.177544419015388e-06, | |
| "loss": 1.2843, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.4351575456053068, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 6.163786448328455e-06, | |
| "loss": 1.2391, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.43604201216141514, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 6.150019155261694e-06, | |
| "loss": 1.4237, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.4369264787175235, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 6.136242650096451e-06, | |
| "loss": 1.4529, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.43781094527363185, | |
| "grad_norm": 2.375, | |
| "learning_rate": 6.122457043187863e-06, | |
| "loss": 1.3786, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.4386954118297402, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.108662444963979e-06, | |
| "loss": 1.2776, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.43957987838584855, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 6.094858965924866e-06, | |
| "loss": 1.3655, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.4404643449419569, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.081046716641735e-06, | |
| "loss": 1.398, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.4413488114980652, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 6.0672258077560475e-06, | |
| "loss": 1.3377, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.44223327805417356, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 6.053396349978632e-06, | |
| "loss": 1.4135, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4431177446102819, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 6.039558454088796e-06, | |
| "loss": 1.3852, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.44400221116639027, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 6.025712230933443e-06, | |
| "loss": 1.3643, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.4448866777224986, | |
| "grad_norm": 2.375, | |
| "learning_rate": 6.0118577914261786e-06, | |
| "loss": 1.1846, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.445771144278607, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.997995246546424e-06, | |
| "loss": 1.1764, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.44665561083471533, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.984124707338528e-06, | |
| "loss": 1.3886, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.44754007739082363, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 5.970246284910877e-06, | |
| "loss": 1.4707, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.448424543946932, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.9563600904350074e-06, | |
| "loss": 1.2391, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.44930901050304034, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 5.942466235144708e-06, | |
| "loss": 1.4436, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.4501934770591487, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 5.9285648303351404e-06, | |
| "loss": 1.2727, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.45107794361525705, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 5.914655987361934e-06, | |
| "loss": 1.377, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.4519624101713654, | |
| "grad_norm": 2.125, | |
| "learning_rate": 5.900739817640304e-06, | |
| "loss": 1.39, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.45284687672747376, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 5.886816432644155e-06, | |
| "loss": 1.4615, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.4537313432835821, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.872885943905191e-06, | |
| "loss": 1.4085, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.4546158098396904, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 5.858948463012015e-06, | |
| "loss": 1.3797, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.45550027639579876, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 5.8450041016092465e-06, | |
| "loss": 1.315, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.4563847429519071, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 5.831052971396613e-06, | |
| "loss": 1.423, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.4572692095080155, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.817095184128067e-06, | |
| "loss": 1.4338, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.4581536760641238, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 5.803130851610887e-06, | |
| "loss": 1.397, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.4590381426202322, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 5.78916008570478e-06, | |
| "loss": 1.4068, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.45992260917634054, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.77518299832099e-06, | |
| "loss": 1.363, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4608070757324489, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.761199701421392e-06, | |
| "loss": 1.4428, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.4616915422885572, | |
| "grad_norm": 2.25, | |
| "learning_rate": 5.747210307017609e-06, | |
| "loss": 1.3775, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.46257600884466554, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 5.733214927170101e-06, | |
| "loss": 1.3901, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.4634604754007739, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 5.719213673987277e-06, | |
| "loss": 1.3603, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.46434494195688225, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.705206659624597e-06, | |
| "loss": 1.477, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.4652294085129906, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.6911939962836625e-06, | |
| "loss": 1.3506, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.46611387506909896, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 5.6771757962113325e-06, | |
| "loss": 1.157, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.4669983416252073, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 5.663152171698813e-06, | |
| "loss": 1.4238, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.46788280818131567, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 5.649123235080762e-06, | |
| "loss": 1.2986, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.46876727473742397, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.635089098734394e-06, | |
| "loss": 1.4711, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4696517412935323, | |
| "grad_norm": 2.125, | |
| "learning_rate": 5.621049875078571e-06, | |
| "loss": 1.3763, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.4705362078496407, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 5.6070056765729065e-06, | |
| "loss": 1.1883, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.47142067440574903, | |
| "grad_norm": 2.375, | |
| "learning_rate": 5.592956615716867e-06, | |
| "loss": 1.4552, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.4723051409618574, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 5.578902805048866e-06, | |
| "loss": 1.2559, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.47318960751796574, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 5.564844357145365e-06, | |
| "loss": 1.3076, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 5.550781384619974e-06, | |
| "loss": 1.3541, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.47495854063018245, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 5.536714000122543e-06, | |
| "loss": 1.2893, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.47584300718629075, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 1.4, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.4767274737423991, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 5.50856644598678e-06, | |
| "loss": 1.3689, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.47761194029850745, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 5.49448650182125e-06, | |
| "loss": 1.3987, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4784964068546158, | |
| "grad_norm": 2.25, | |
| "learning_rate": 5.480402596627477e-06, | |
| "loss": 1.3482, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.47938087341072416, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 5.466314843222993e-06, | |
| "loss": 1.3633, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.4802653399668325, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.452223354456154e-06, | |
| "loss": 1.3443, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.48114980652294087, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 5.438128243205241e-06, | |
| "loss": 1.2519, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.4820342730790492, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 5.4240296223775465e-06, | |
| "loss": 1.2557, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.4829187396351575, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 5.409927604908481e-06, | |
| "loss": 1.4529, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.4838032061912659, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 5.395822303760662e-06, | |
| "loss": 1.3618, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.48468767274737423, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.381713831923008e-06, | |
| "loss": 1.4106, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.4855721393034826, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 5.367602302409841e-06, | |
| "loss": 1.2277, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.48645660585959094, | |
| "grad_norm": 2.125, | |
| "learning_rate": 5.353487828259973e-06, | |
| "loss": 1.3436, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4873410724156993, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 5.339370522535805e-06, | |
| "loss": 1.4189, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.48822553897180765, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 5.325250498322417e-06, | |
| "loss": 1.5008, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.489110005527916, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 5.311127868726667e-06, | |
| "loss": 1.4573, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.4899944720840243, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 5.297002746876284e-06, | |
| "loss": 1.525, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.49087893864013266, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 5.282875245918963e-06, | |
| "loss": 1.308, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.491763405196241, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 5.26874547902145e-06, | |
| "loss": 1.3713, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.49264787175234936, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 5.2546135593686484e-06, | |
| "loss": 1.4282, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.4935323383084577, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 5.240479600162704e-06, | |
| "loss": 1.5133, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.4944168048645661, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 5.226343714622096e-06, | |
| "loss": 1.3118, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.4953012714206744, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 5.212206015980742e-06, | |
| "loss": 1.4079, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.4953012714206744, | |
| "eval_loss": 1.3238087892532349, | |
| "eval_runtime": 232.2455, | |
| "eval_samples_per_second": 69.237, | |
| "eval_steps_per_second": 8.655, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.4961857379767827, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 5.198066617487078e-06, | |
| "loss": 1.4577, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.4970702045328911, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 5.1839256324031575e-06, | |
| "loss": 1.3982, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.49795467108899943, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 5.169783174003744e-06, | |
| "loss": 1.3064, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.4988391376451078, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 5.155639355575402e-06, | |
| "loss": 1.4527, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.49972360420121614, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 5.141494290415592e-06, | |
| "loss": 1.3521, | |
| "step": 565 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 565, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.1620099119054848e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |