| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9994472084024323, | |
| "eval_steps": 70, | |
| "global_step": 1130, | |
| "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 | |
| }, | |
| { | |
| "epoch": 0.5006080707573245, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 5.127348091831755e-06, | |
| "loss": 1.4069, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.5014925373134328, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 5.1132008731404225e-06, | |
| "loss": 1.3273, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.5023770038695412, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 5.099052747666289e-06, | |
| "loss": 1.3473, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.5032614704256495, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 5.084903828741312e-06, | |
| "loss": 1.4105, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.5041459369817579, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 5.070754229703811e-06, | |
| "loss": 1.3755, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5050304035378662, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 5.056604063897549e-06, | |
| "loss": 1.4316, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.5059148700939746, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 5.042453444670829e-06, | |
| "loss": 1.3914, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.5067993366500829, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 5.028302485375588e-06, | |
| "loss": 1.4373, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.5076838032061912, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 5.0141512993664864e-06, | |
| "loss": 1.3995, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.5085682697622996, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 5e-06, | |
| "loss": 1.2879, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.5094527363184079, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 4.985848700633515e-06, | |
| "loss": 1.346, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.5103372028745163, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 4.971697514624414e-06, | |
| "loss": 1.3491, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.5112216694306246, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 4.957546555329174e-06, | |
| "loss": 1.3267, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.512106135986733, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.943395936102453e-06, | |
| "loss": 1.4237, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.5129906025428413, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.929245770296191e-06, | |
| "loss": 1.3134, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5138750690989496, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.91509617125869e-06, | |
| "loss": 1.3902, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.514759535655058, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.900947252333712e-06, | |
| "loss": 1.3079, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.5156440022111664, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.8867991268595775e-06, | |
| "loss": 1.3071, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.5165284687672748, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 4.872651908168244e-06, | |
| "loss": 1.2942, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.5174129353233831, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.85850570958441e-06, | |
| "loss": 1.4196, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.5182974018794915, | |
| "grad_norm": 2.25, | |
| "learning_rate": 4.844360644424599e-06, | |
| "loss": 1.3455, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.5191818684355998, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 4.830216825996257e-06, | |
| "loss": 1.504, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.5200663349917082, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 4.816074367596843e-06, | |
| "loss": 1.3046, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.5209508015478165, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.801933382512923e-06, | |
| "loss": 1.3623, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.5218352681039248, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.78779398401926e-06, | |
| "loss": 1.3359, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5227197346600332, | |
| "grad_norm": 2.125, | |
| "learning_rate": 4.7736562853779066e-06, | |
| "loss": 1.5075, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.5236042012161415, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 4.7595203998373e-06, | |
| "loss": 1.4021, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.5244886677722499, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 4.745386440631354e-06, | |
| "loss": 1.2704, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.5253731343283582, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 4.73125452097855e-06, | |
| "loss": 1.3103, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.5262576008844666, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 4.717124754081038e-06, | |
| "loss": 1.3654, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.5271420674405749, | |
| "grad_norm": 2.125, | |
| "learning_rate": 4.702997253123716e-06, | |
| "loss": 1.322, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.5280265339966832, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.688872131273334e-06, | |
| "loss": 1.2978, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.5289110005527916, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 4.674749501677585e-06, | |
| "loss": 1.4011, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.5297954671088999, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 4.660629477464197e-06, | |
| "loss": 1.3308, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.5306799336650083, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 4.646512171740028e-06, | |
| "loss": 1.3854, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5315644002211166, | |
| "grad_norm": 2.25, | |
| "learning_rate": 4.6323976975901595e-06, | |
| "loss": 1.2851, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.532448866777225, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 4.618286168076993e-06, | |
| "loss": 1.4358, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.6041776962393405e-06, | |
| "loss": 1.34, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.5342177998894417, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 4.590072395091521e-06, | |
| "loss": 1.3445, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.53510226644555, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.575970377622456e-06, | |
| "loss": 1.3103, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.5359867330016583, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.56187175679476e-06, | |
| "loss": 1.2384, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.5368711995577667, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 4.547776645543846e-06, | |
| "loss": 1.3134, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.537755666113875, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.533685156777008e-06, | |
| "loss": 1.3305, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.5386401326699835, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.519597403372524e-06, | |
| "loss": 1.4062, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.5395245992260918, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.505513498178752e-06, | |
| "loss": 1.3568, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5404090657822002, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 4.491433554013221e-06, | |
| "loss": 1.2757, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.5412935323383085, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 1.2589, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.5421779988944168, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.463285999877458e-06, | |
| "loss": 1.4028, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.5430624654505252, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.449218615380029e-06, | |
| "loss": 1.3952, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.5439469320066335, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 4.435155642854637e-06, | |
| "loss": 1.4054, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.5448313985627419, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.421097194951137e-06, | |
| "loss": 1.3026, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.5457158651188502, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.407043384283136e-06, | |
| "loss": 1.4435, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.5466003316749586, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.3929943234270935e-06, | |
| "loss": 1.3515, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.5474847982310669, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 4.37895012492143e-06, | |
| "loss": 1.3434, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.5483692647871752, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.364910901265607e-06, | |
| "loss": 1.479, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5492537313432836, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.350876764919238e-06, | |
| "loss": 1.2995, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.5501381978993919, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.336847828301189e-06, | |
| "loss": 1.4547, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.5510226644555003, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 4.322824203788669e-06, | |
| "loss": 1.3666, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.5519071310116086, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 4.308806003716338e-06, | |
| "loss": 1.3513, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.552791597567717, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 4.294793340375405e-06, | |
| "loss": 1.3136, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.5536760641238253, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.280786326012724e-06, | |
| "loss": 1.3245, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.5545605306799337, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.266785072829901e-06, | |
| "loss": 1.3171, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.555444997236042, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.252789692982394e-06, | |
| "loss": 1.2222, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.5563294637921503, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 4.23880029857861e-06, | |
| "loss": 1.547, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.5572139303482587, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.224817001679011e-06, | |
| "loss": 1.2893, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5572139303482587, | |
| "eval_loss": 1.3222172260284424, | |
| "eval_runtime": 232.1074, | |
| "eval_samples_per_second": 69.278, | |
| "eval_steps_per_second": 8.66, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.558098396904367, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 4.210839914295219e-06, | |
| "loss": 1.3732, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.5589828634604754, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 4.196869148389114e-06, | |
| "loss": 1.4756, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.5598673300165837, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 4.1829048158719344e-06, | |
| "loss": 1.3808, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.5607517965726921, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 4.168947028603389e-06, | |
| "loss": 1.2927, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.5616362631288004, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.154995898390756e-06, | |
| "loss": 1.3053, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.5625207296849087, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 4.141051536987986e-06, | |
| "loss": 1.4368, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.5634051962410171, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.127114056094811e-06, | |
| "loss": 1.3628, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.5642896627971254, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 4.113183567355846e-06, | |
| "loss": 1.2668, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.5651741293532339, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.099260182359698e-06, | |
| "loss": 1.3524, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.5660585959093422, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 4.085344012638067e-06, | |
| "loss": 1.4395, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5669430624654506, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 4.071435169664861e-06, | |
| "loss": 1.4211, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.5678275290215589, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.057533764855291e-06, | |
| "loss": 1.4624, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.5687119955776673, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.043639909564993e-06, | |
| "loss": 1.3607, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.5695964621337756, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.029753715089124e-06, | |
| "loss": 1.4639, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.5704809286898839, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 4.015875292661474e-06, | |
| "loss": 1.2738, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.5713653952459923, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.002004753453577e-06, | |
| "loss": 1.266, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.5722498618021006, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 3.988142208573822e-06, | |
| "loss": 1.3372, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.573134328358209, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.9742877690665575e-06, | |
| "loss": 1.4597, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.5740187949143173, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 3.960441545911205e-06, | |
| "loss": 1.3481, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.5749032614704257, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 3.94660365002137e-06, | |
| "loss": 1.454, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.575787728026534, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.932774192243953e-06, | |
| "loss": 1.3835, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.5766721945826423, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 3.918953283358266e-06, | |
| "loss": 1.4907, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.5775566611387507, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.905141034075135e-06, | |
| "loss": 1.3345, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.578441127694859, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 3.891337555036022e-06, | |
| "loss": 1.5018, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.5793255942509674, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 3.877542956812137e-06, | |
| "loss": 1.3936, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.5802100608070757, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.863757349903551e-06, | |
| "loss": 1.3161, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.5810945273631841, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.849980844738308e-06, | |
| "loss": 1.4803, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.5819789939192924, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.836213551671547e-06, | |
| "loss": 1.3776, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.5828634604754007, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 3.822455580984613e-06, | |
| "loss": 1.3356, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.5837479270315091, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 3.808707042884176e-06, | |
| "loss": 1.4152, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5846323935876174, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.794968047501345e-06, | |
| "loss": 1.549, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.5855168601437258, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 3.781238704890793e-06, | |
| "loss": 1.4049, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.5864013266998341, | |
| "grad_norm": 2.25, | |
| "learning_rate": 3.767519125029865e-06, | |
| "loss": 1.4221, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.5872857932559425, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 3.7538094178177054e-06, | |
| "loss": 1.2047, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.5881702598120508, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.7401096930743753e-06, | |
| "loss": 1.3547, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.5890547263681593, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 3.72642006053997e-06, | |
| "loss": 1.4558, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.5899391929242676, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.7127406298737458e-06, | |
| "loss": 1.4111, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.5908236594803759, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.6990715106532353e-06, | |
| "loss": 1.3221, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.5917081260364843, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 3.6854128123733736e-06, | |
| "loss": 1.2724, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.6717646444456196e-06, | |
| "loss": 1.3462, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.593477059148701, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 3.658127116197079e-06, | |
| "loss": 1.3287, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.5943615257048093, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.6445003368696317e-06, | |
| "loss": 1.2705, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.5952459922609177, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 3.6308844156190546e-06, | |
| "loss": 1.4935, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.596130458817026, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.6172794615141448e-06, | |
| "loss": 1.4558, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 2.25, | |
| "learning_rate": 3.60368558353585e-06, | |
| "loss": 1.4027, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.5978993919292427, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 3.590102890576394e-06, | |
| "loss": 1.3063, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.598783858485351, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 3.576531491438403e-06, | |
| "loss": 1.3965, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.5996683250414594, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.5629714948340354e-06, | |
| "loss": 1.2793, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.6005527915975677, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 3.5494230093841144e-06, | |
| "loss": 1.5579, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.6014372581536761, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 3.5358861436172487e-06, | |
| "loss": 1.2379, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6023217247097844, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 3.5223610059689733e-06, | |
| "loss": 1.2644, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.6032061912658928, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 3.5088477047808738e-06, | |
| "loss": 1.3687, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.6040906578220011, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 3.495346348299724e-06, | |
| "loss": 1.3623, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.6049751243781094, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 3.481857044676613e-06, | |
| "loss": 1.4687, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.6058595909342178, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.4683799019660834e-06, | |
| "loss": 1.4703, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.6067440574903261, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 1.3077, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.6076285240464345, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.4414625310130038e-06, | |
| "loss": 1.3626, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.6085129906025428, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 3.4280225183890124e-06, | |
| "loss": 1.4234, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.6093974571586512, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.4145950979129916e-06, | |
| "loss": 1.3617, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.6102819237147595, | |
| "grad_norm": 2.25, | |
| "learning_rate": 3.401180377143774e-06, | |
| "loss": 1.3452, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6111663902708678, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 3.3877784635384668e-06, | |
| "loss": 1.3217, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.6120508568269762, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 3.374389464451583e-06, | |
| "loss": 1.4221, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.6129353233830845, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.361013487134188e-06, | |
| "loss": 1.3176, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.613819789939193, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.3476506387330355e-06, | |
| "loss": 1.2636, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.6147042564953012, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.3343010262897125e-06, | |
| "loss": 1.199, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.6155887230514097, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.320964756739781e-06, | |
| "loss": 1.3831, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.616473189607518, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.3076419369119216e-06, | |
| "loss": 1.3128, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.6173576561636264, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.2943326735270766e-06, | |
| "loss": 1.3597, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.6182421227197347, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 3.2810370731975956e-06, | |
| "loss": 1.4076, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.619126589275843, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.2677552424263836e-06, | |
| "loss": 1.376, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.619126589275843, | |
| "eval_loss": 1.3213250637054443, | |
| "eval_runtime": 232.1662, | |
| "eval_samples_per_second": 69.261, | |
| "eval_steps_per_second": 8.658, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6200110558319514, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.254487287606044e-06, | |
| "loss": 1.3793, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.6208955223880597, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 3.2412333150180298e-06, | |
| "loss": 1.4251, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.6217799889441681, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.2279934308317905e-06, | |
| "loss": 1.4261, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.6226644555002764, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.2147677411039236e-06, | |
| "loss": 1.4114, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.6235489220563848, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 3.2015563517773214e-06, | |
| "loss": 1.2844, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.6244333886124931, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.1883593686803244e-06, | |
| "loss": 1.4281, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.6253178551686014, | |
| "grad_norm": 2.25, | |
| "learning_rate": 3.1751768975258745e-06, | |
| "loss": 1.3457, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.6262023217247098, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 3.162009043910668e-06, | |
| "loss": 1.3013, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.6270867882808181, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 3.1488559133143065e-06, | |
| "loss": 1.347, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.6279712548369265, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 3.1357176110984578e-06, | |
| "loss": 1.2946, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6288557213930348, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.122594242506006e-06, | |
| "loss": 1.3807, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.6297401879491432, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 3.1094859126602132e-06, | |
| "loss": 1.3381, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.6306246545052515, | |
| "grad_norm": 2.25, | |
| "learning_rate": 3.096392726563874e-06, | |
| "loss": 1.3269, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.6315091210613598, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.0833147890984735e-06, | |
| "loss": 1.2699, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.6323935876174682, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.070252205023356e-06, | |
| "loss": 1.3409, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.6332780541735765, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 3.0572050789748732e-06, | |
| "loss": 1.3592, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.6341625207296849, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.044173515465554e-06, | |
| "loss": 1.4928, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.6350469872857932, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 3.031157618883265e-06, | |
| "loss": 1.3309, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.6359314538419016, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 3.0181574934903745e-06, | |
| "loss": 1.4954, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.6368159203980099, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.0051732434229185e-06, | |
| "loss": 1.2679, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6377003869541183, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.992204972689765e-06, | |
| "loss": 1.3471, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.6385848535102266, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 2.9792527851717805e-06, | |
| "loss": 1.4871, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.639469320066335, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.966316784621e-06, | |
| "loss": 1.3921, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.6403537866224434, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.9533970746597935e-06, | |
| "loss": 1.4691, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.6412382531785517, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.9404937587800374e-06, | |
| "loss": 1.4068, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.6421227197346601, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.9276069403422846e-06, | |
| "loss": 1.437, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.6430071862907684, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.9147367225749425e-06, | |
| "loss": 1.4186, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.6438916528468768, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.90188320857343e-06, | |
| "loss": 1.3671, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.6447761194029851, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.889046501299373e-06, | |
| "loss": 1.2498, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.6456605859590934, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.8762267035797607e-06, | |
| "loss": 1.4189, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6465450525152018, | |
| "grad_norm": 2.25, | |
| "learning_rate": 2.8634239181061384e-06, | |
| "loss": 1.3515, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.6474295190713101, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.850638247433768e-06, | |
| "loss": 1.4201, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.6483139856274185, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.837869793980823e-06, | |
| "loss": 1.345, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.6491984521835268, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 2.8251186600275533e-06, | |
| "loss": 1.2402, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.6500829187396352, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 2.8123849477154808e-06, | |
| "loss": 1.3014, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.6509673852957435, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 2.799668759046564e-06, | |
| "loss": 1.3631, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.6518518518518519, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.786970195882398e-06, | |
| "loss": 1.4583, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.6527363184079602, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.7742893599433875e-06, | |
| "loss": 1.4511, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.6536207849640685, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 2.761626352807938e-06, | |
| "loss": 1.4266, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.6545052515201769, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.748981275911633e-06, | |
| "loss": 1.3963, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6553897180762852, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 2.736354230546437e-06, | |
| "loss": 1.5029, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.6562741846323936, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 2.7237453178598645e-06, | |
| "loss": 1.5213, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.6571586511885019, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 2.71115463885419e-06, | |
| "loss": 1.2602, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.6580431177446103, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.698582294385621e-06, | |
| "loss": 1.4041, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.6589275843007186, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.6860283851635067e-06, | |
| "loss": 1.2874, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.6598120508568269, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.6734930117495133e-06, | |
| "loss": 1.4202, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.6606965174129353, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.6609762745568358e-06, | |
| "loss": 1.3251, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.6615809839690436, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 2.6484782738493843e-06, | |
| "loss": 1.4304, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.662465450525152, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.6359991097409767e-06, | |
| "loss": 1.3746, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.6633499170812603, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 2.6235388821945497e-06, | |
| "loss": 1.3906, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6642343836373688, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.61109769102135e-06, | |
| "loss": 1.3656, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.665118850193477, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.598675635880129e-06, | |
| "loss": 1.3807, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.6660033167495855, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.5862728162763595e-06, | |
| "loss": 1.3882, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.6668877833056938, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.573889331561423e-06, | |
| "loss": 1.4212, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.6677722498618021, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.5615252809318287e-06, | |
| "loss": 1.3437, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.6686567164179105, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.5491807634284016e-06, | |
| "loss": 1.4545, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.6695411829740188, | |
| "grad_norm": 2.390625, | |
| "learning_rate": 2.536855877935507e-06, | |
| "loss": 1.3538, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.6704256495301272, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.524550723180249e-06, | |
| "loss": 1.3707, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.6713101160862355, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.5122653977316758e-06, | |
| "loss": 1.4478, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.6721945826423439, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 1.3143, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6730790491984522, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.487754628235805e-06, | |
| "loss": 1.3425, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.6739635157545605, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.4755293805292572e-06, | |
| "loss": 1.4169, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.6748479823106689, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.4633243548093195e-06, | |
| "loss": 1.3797, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.6757324488667772, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.4511396488429727e-06, | |
| "loss": 1.4054, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.6766169154228856, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.4389753602344298e-06, | |
| "loss": 1.3736, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.6775013819789939, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.4268315864243447e-06, | |
| "loss": 1.3543, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.6783858485351023, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.4147084246890478e-06, | |
| "loss": 1.3911, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.6792703150912106, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 2.402605972139759e-06, | |
| "loss": 1.3499, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.6801547816473189, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.3905243257218007e-06, | |
| "loss": 1.4691, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.6810392482034273, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 2.3784635822138424e-06, | |
| "loss": 1.2953, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6810392482034273, | |
| "eval_loss": 1.3208374977111816, | |
| "eval_runtime": 232.2504, | |
| "eval_samples_per_second": 69.236, | |
| "eval_steps_per_second": 8.654, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6819237147595356, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.366423838227102e-06, | |
| "loss": 1.3865, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.682808181315644, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.354405190204592e-06, | |
| "loss": 1.2972, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.6836926478717523, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.342407734420331e-06, | |
| "loss": 1.2642, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.6845771144278607, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.3304315669785853e-06, | |
| "loss": 1.3523, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.685461580983969, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.318476783813088e-06, | |
| "loss": 1.3066, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.6863460475400774, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.306543480686281e-06, | |
| "loss": 1.4334, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.6872305140961857, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.294631753188541e-06, | |
| "loss": 1.3997, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.688114980652294, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.2827416967374115e-06, | |
| "loss": 1.4262, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.6889994472084024, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.270873406576849e-06, | |
| "loss": 1.4025, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.6898839137645107, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.2590269777764516e-06, | |
| "loss": 1.4295, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.6907683803206192, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 2.247202505230695e-06, | |
| "loss": 1.3638, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.6916528468767275, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.2354000836581834e-06, | |
| "loss": 1.3582, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.6925373134328359, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.223619807600878e-06, | |
| "loss": 1.3499, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.6934217799889442, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.2118617714233524e-06, | |
| "loss": 1.3486, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.6943062465450525, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.2001260693120236e-06, | |
| "loss": 1.3466, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.6951907131011609, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.1884127952744095e-06, | |
| "loss": 1.45, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.6960751796572692, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.1767220431383652e-06, | |
| "loss": 1.4239, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.6969596462133776, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.1650539065513415e-06, | |
| "loss": 1.3557, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.6978441127694859, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 2.153408478979628e-06, | |
| "loss": 1.4248, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.6987285793255943, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 2.141785853707607e-06, | |
| "loss": 1.3066, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.6996130458817026, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.130186123837002e-06, | |
| "loss": 1.3607, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.700497512437811, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.118609382286142e-06, | |
| "loss": 1.3326, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.7013819789939193, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.107055721789203e-06, | |
| "loss": 1.3718, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.7022664455500276, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.0955252348954806e-06, | |
| "loss": 1.3615, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.703150912106136, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 2.0840180139686333e-06, | |
| "loss": 1.4377, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.7040353786622443, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.0725341511859576e-06, | |
| "loss": 1.4253, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.7049198452183527, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 1.3843, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.705804311774461, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.0496368678260094e-06, | |
| "loss": 1.4443, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.7066887783305694, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.038223630664835e-06, | |
| "loss": 1.3969, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.7075732448866777, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.0268341184785674e-06, | |
| "loss": 1.3988, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.708457711442786, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.015468422501599e-06, | |
| "loss": 1.389, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.7093421779988944, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.004126633777559e-06, | |
| "loss": 1.3717, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.7102266445550027, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 1.992808843158559e-06, | |
| "loss": 1.5092, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.9815151413044865e-06, | |
| "loss": 1.377, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.7119955776672194, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.9702456186822595e-06, | |
| "loss": 1.2182, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.7128800442233278, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.959000365565119e-06, | |
| "loss": 1.4019, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.7137645107794361, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.947779472031891e-06, | |
| "loss": 1.4615, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.7146489773355446, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 1.936583027966279e-06, | |
| "loss": 1.4442, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.7155334438916529, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.9254111230561282e-06, | |
| "loss": 1.3233, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.7164179104477612, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9142638467927254e-06, | |
| "loss": 1.3184, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.7173023770038696, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.9031412884700612e-06, | |
| "loss": 1.4767, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.7181868435599779, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.8920435371841394e-06, | |
| "loss": 1.3654, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.7190713101160863, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.8809706818322375e-06, | |
| "loss": 1.3787, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.7199557766721946, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.8699228111122146e-06, | |
| "loss": 1.4234, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.720840243228303, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.8589000135217882e-06, | |
| "loss": 1.4363, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.7217247097844113, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.8479023773578354e-06, | |
| "loss": 1.3332, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.7226091763405196, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.8369299907156745e-06, | |
| "loss": 1.5322, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.723493642896628, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.8259829414883728e-06, | |
| "loss": 1.4411, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.7243781094527363, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.815061317366027e-06, | |
| "loss": 1.3387, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.7252625760088447, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.8041652058350768e-06, | |
| "loss": 1.5099, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.726147042564953, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.7932946941775882e-06, | |
| "loss": 1.3271, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.7270315091210614, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.78244986947057e-06, | |
| "loss": 1.3248, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.7279159756771697, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.7716308185852605e-06, | |
| "loss": 1.4185, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.728800442233278, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 1.7608376281864502e-06, | |
| "loss": 1.3921, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.7296849087893864, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.7500703847317663e-06, | |
| "loss": 1.3979, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.7305693753454947, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.739329174470999e-06, | |
| "loss": 1.3455, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.7314538419016031, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.7286140834453958e-06, | |
| "loss": 1.5016, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.7323383084577114, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.7179251974869858e-06, | |
| "loss": 1.4437, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.7332227750138198, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.7072626022178796e-06, | |
| "loss": 1.4511, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.7341072415699281, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.6966263830495939e-06, | |
| "loss": 1.4144, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7349917081260365, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.6860166251823574e-06, | |
| "loss": 1.4895, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.7358761746821448, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.6754334136044398e-06, | |
| "loss": 1.344, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.7367606412382531, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.6648768330914578e-06, | |
| "loss": 1.3382, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.7376451077943615, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 1.3857, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.7385295743504698, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.6438439032954857e-06, | |
| "loss": 1.4671, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.7394140409065783, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.633367722494404e-06, | |
| "loss": 1.4239, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.7402985074626866, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.6229185097207228e-06, | |
| "loss": 1.3192, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.741182974018795, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.6124963486766844e-06, | |
| "loss": 1.3301, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.7420674405749033, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.6021013228478277e-06, | |
| "loss": 1.2805, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.7429519071310116, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.5917335155023368e-06, | |
| "loss": 1.2969, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7429519071310116, | |
| "eval_loss": 1.3206154108047485, | |
| "eval_runtime": 232.2366, | |
| "eval_samples_per_second": 69.24, | |
| "eval_steps_per_second": 8.655, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.74383637368712, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.5813930096903562e-06, | |
| "loss": 1.2511, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.7447208402432283, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 1.5710798882433432e-06, | |
| "loss": 1.3885, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.7456053067993367, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.5607942337733884e-06, | |
| "loss": 1.3645, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.746489773355445, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.5505361286725679e-06, | |
| "loss": 1.3793, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.7473742399115534, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.5403056551122697e-06, | |
| "loss": 1.3091, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.7482587064676617, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.5301028950425494e-06, | |
| "loss": 1.4169, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.7491431730237701, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.5199279301914592e-06, | |
| "loss": 1.2407, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.7500276395798784, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.5097808420644117e-06, | |
| "loss": 1.5407, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.7509121061359867, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.4996617119435036e-06, | |
| "loss": 1.4126, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.7517965726920951, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.4895706208868876e-06, | |
| "loss": 1.3619, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7526810392482034, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.4795076497281052e-06, | |
| "loss": 1.3949, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.7535655058043118, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.4694728790754514e-06, | |
| "loss": 1.2663, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.7544499723604201, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.459466389311321e-06, | |
| "loss": 1.2417, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.7553344389165285, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.4494882605915717e-06, | |
| "loss": 1.257, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.7562189054726368, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.439538572844873e-06, | |
| "loss": 1.3101, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.7571033720287451, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.4296174057720775e-06, | |
| "loss": 1.3139, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.7579878385848535, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.4197248388455693e-06, | |
| "loss": 1.3913, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.7588723051409618, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.4098609513086414e-06, | |
| "loss": 1.3652, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.7597567716970702, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.4000258221748452e-06, | |
| "loss": 1.399, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.7606412382531785, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.390219530227378e-06, | |
| "loss": 1.4919, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7615257048092869, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 1.3804421540184282e-06, | |
| "loss": 1.3179, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.7624101713653952, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.3706937718685654e-06, | |
| "loss": 1.4212, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.7632946379215035, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.3609744618661013e-06, | |
| "loss": 1.4791, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.764179104477612, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.3512843018664728e-06, | |
| "loss": 1.3187, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.7650635710337202, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.3416233694916086e-06, | |
| "loss": 1.2747, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.7659480375898287, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.3319917421293182e-06, | |
| "loss": 1.366, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.766832504145937, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.3223894969326623e-06, | |
| "loss": 1.4011, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.7677169707020454, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.312816710819343e-06, | |
| "loss": 1.4062, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.7686014372581537, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.3032734604710784e-06, | |
| "loss": 1.4579, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.7694859038142621, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.2937598223330006e-06, | |
| "loss": 1.2437, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7703703703703704, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.2842758726130283e-06, | |
| "loss": 1.3704, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.7712548369264787, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.2748216872812747e-06, | |
| "loss": 1.349, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.7721393034825871, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.2653973420694182e-06, | |
| "loss": 1.4582, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.7730237700386954, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.256002912470115e-06, | |
| "loss": 1.3579, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.7739082365948038, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.246638473736378e-06, | |
| "loss": 1.2888, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.7747927031509121, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.2373041008809888e-06, | |
| "loss": 1.2819, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.7756771697070205, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.2279998686758833e-06, | |
| "loss": 1.469, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.7765616362631288, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.2187258516515644e-06, | |
| "loss": 1.4948, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.7774461028192371, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.2094821240964955e-06, | |
| "loss": 1.3585, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.7783305693753455, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.2002687600565138e-06, | |
| "loss": 1.3644, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.7792150359314538, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.1910858333342279e-06, | |
| "loss": 1.4203, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.7800995024875622, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.1819334174884378e-06, | |
| "loss": 1.3598, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.7809839690436705, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.172811585833532e-06, | |
| "loss": 1.2861, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.7818684355997789, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.1637204114389179e-06, | |
| "loss": 1.4354, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.7827529021558872, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.1546599671284158e-06, | |
| "loss": 1.3396, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.7836373687119956, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.1456303254796941e-06, | |
| "loss": 1.3573, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.7845218352681039, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.1366315588236743e-06, | |
| "loss": 1.4086, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.7854063018242122, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.127663739243962e-06, | |
| "loss": 1.3828, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.7862907683803206, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.1187269385762605e-06, | |
| "loss": 1.4209, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.7871752349364289, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.1098212284078037e-06, | |
| "loss": 1.3129, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.7880597014925373, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.1009466800767738e-06, | |
| "loss": 1.4199, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.7889441680486456, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.09210336467174e-06, | |
| "loss": 1.4955, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.789828634604754, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.0832913530310784e-06, | |
| "loss": 1.33, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.7907131011608624, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.0745107157424155e-06, | |
| "loss": 1.4374, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.7915975677169707, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.0657615231420492e-06, | |
| "loss": 1.3894, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.7924820342730791, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.0570438453144044e-06, | |
| "loss": 1.4137, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.7933665008291874, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.0483577520914495e-06, | |
| "loss": 1.3512, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.7942509673852958, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.0397033130521554e-06, | |
| "loss": 1.3495, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.7951354339414041, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.0310805975219256e-06, | |
| "loss": 1.2432, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.7960199004975125, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 1.0224896745720513e-06, | |
| "loss": 1.4603, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7969043670536208, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 1.0139306130191463e-06, | |
| "loss": 1.3466, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.7977888336097292, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.0054034814246093e-06, | |
| "loss": 1.5046, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.7986733001658375, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 9.969083480940617e-07, | |
| "loss": 1.5009, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.7995577667219458, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 9.884452810768115e-07, | |
| "loss": 1.3271, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.8004422332780542, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 9.80014348165298e-07, | |
| "loss": 1.4613, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.8013266998341625, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 9.71615616894559e-07, | |
| "loss": 1.4525, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.8022111663902709, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 9.632491545416788e-07, | |
| "loss": 1.325, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.8030956329463792, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 1.4775, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.8039800995024876, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 9.466133044048865e-07, | |
| "loss": 1.475, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.8048645660585959, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 9.383440498805712e-07, | |
| "loss": 1.4142, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8048645660585959, | |
| "eval_loss": 1.320538878440857, | |
| "eval_runtime": 232.0847, | |
| "eval_samples_per_second": 69.285, | |
| "eval_steps_per_second": 8.661, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8057490326147042, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 9.301073307922432e-07, | |
| "loss": 1.3282, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.8066334991708126, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 9.219032131192124e-07, | |
| "loss": 1.4358, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.8075179657269209, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 9.137317625796338e-07, | |
| "loss": 1.2671, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.8084024322830293, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 9.055930446299916e-07, | |
| "loss": 1.4075, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.8092868988391376, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.974871244645628e-07, | |
| "loss": 1.3825, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.810171365395246, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.894140670149082e-07, | |
| "loss": 1.3359, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.8110558319513543, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 8.813739369493395e-07, | |
| "loss": 1.4309, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.8119402985074626, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 8.733667986724109e-07, | |
| "loss": 1.4381, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.812824765063571, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 8.653927163243986e-07, | |
| "loss": 1.2608, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.8137092316196793, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 8.574517537807897e-07, | |
| "loss": 1.4192, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8145936981757878, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 8.495439746517648e-07, | |
| "loss": 1.2921, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.815478164731896, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 8.416694422816974e-07, | |
| "loss": 1.4909, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.8163626312880045, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.338282197486364e-07, | |
| "loss": 1.3452, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.8172470978441128, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 8.260203698638114e-07, | |
| "loss": 1.3678, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.8181315644002212, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 8.182459551711197e-07, | |
| "loss": 1.2546, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.8190160309563295, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 8.105050379466334e-07, | |
| "loss": 1.4046, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.8199004975124378, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.027976801980924e-07, | |
| "loss": 1.405, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.8207849640685462, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.95123943664417e-07, | |
| "loss": 1.3807, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.8216694306246545, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 7.87483889815207e-07, | |
| "loss": 1.4121, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.8225538971807629, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 7.798775798502484e-07, | |
| "loss": 1.3807, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.8234383637368712, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 7.723050746990291e-07, | |
| "loss": 1.3344, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.8243228302929796, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 7.647664350202461e-07, | |
| "loss": 1.4637, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.8252072968490879, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 7.572617212013189e-07, | |
| "loss": 1.3765, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.8260917634051962, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.497909933579117e-07, | |
| "loss": 1.4303, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.8269762299613046, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.423543113334436e-07, | |
| "loss": 1.3633, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.8278606965174129, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 7.349517346986173e-07, | |
| "loss": 1.3611, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.8287451630735213, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 7.275833227509344e-07, | |
| "loss": 1.3908, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 2.25, | |
| "learning_rate": 7.202491345142288e-07, | |
| "loss": 1.2941, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.830514096185738, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 7.129492287381828e-07, | |
| "loss": 1.3899, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.8313985627418463, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.056836638978698e-07, | |
| "loss": 1.3957, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8322830292979547, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.984524981932756e-07, | |
| "loss": 1.4181, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.833167495854063, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 6.912557895488342e-07, | |
| "loss": 1.3393, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.8340519624101713, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 6.84093595612968e-07, | |
| "loss": 1.3599, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.8349364289662797, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 6.769659737576229e-07, | |
| "loss": 1.333, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.835820895522388, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.698729810778065e-07, | |
| "loss": 1.3796, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.8367053620784964, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 6.628146743911374e-07, | |
| "loss": 1.2729, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.8375898286346047, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 6.55791110237381e-07, | |
| "loss": 1.449, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.8384742951907131, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 6.488023448780068e-07, | |
| "loss": 1.3217, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.8393587617468214, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.418484342957277e-07, | |
| "loss": 1.3816, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.8402432283029297, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.349294341940593e-07, | |
| "loss": 1.3077, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8411276948590382, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 6.280453999968711e-07, | |
| "loss": 1.3948, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.8420121614151465, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 6.211963868479393e-07, | |
| "loss": 1.4487, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.8428966279712549, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.143824496105122e-07, | |
| "loss": 1.3836, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.8437810945273632, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 6.076036428668613e-07, | |
| "loss": 1.4319, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.8446655610834716, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.008600209178539e-07, | |
| "loss": 1.35, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.8455500276395799, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 5.941516377825102e-07, | |
| "loss": 1.2769, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.8464344941956883, | |
| "grad_norm": 2.125, | |
| "learning_rate": 5.874785471975753e-07, | |
| "loss": 1.3282, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.8473189607517966, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 5.808408026170892e-07, | |
| "loss": 1.3316, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.8482034273079049, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 5.742384572119519e-07, | |
| "loss": 1.3037, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.8490878938640133, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 5.676715638695063e-07, | |
| "loss": 1.3868, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8499723604201216, | |
| "grad_norm": 2.125, | |
| "learning_rate": 5.611401751931112e-07, | |
| "loss": 1.3841, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.85085682697623, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 5.546443435017146e-07, | |
| "loss": 1.43, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.8517412935323383, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 5.481841208294447e-07, | |
| "loss": 1.2815, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.8526257600884467, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 5.417595589251828e-07, | |
| "loss": 1.2582, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.853510226644555, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 5.353707092521581e-07, | |
| "loss": 1.3579, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.8543946932006633, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 5.290176229875271e-07, | |
| "loss": 1.2397, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.8552791597567717, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.227003510219714e-07, | |
| "loss": 1.5733, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.85616362631288, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 5.16418943959282e-07, | |
| "loss": 1.4332, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.8570480928689884, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 5.101734521159613e-07, | |
| "loss": 1.505, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.8579325594250967, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 5.039639255208156e-07, | |
| "loss": 1.3825, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.8588170259812051, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 4.977904139145579e-07, | |
| "loss": 1.4532, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.8597014925373134, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.916529667494024e-07, | |
| "loss": 1.4113, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.8605859590934217, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 4.855516331886783e-07, | |
| "loss": 1.3548, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.8614704256495301, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.794864621064266e-07, | |
| "loss": 1.4388, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.8623548922056384, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 4.734575020870169e-07, | |
| "loss": 1.4733, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.8632393587617468, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.6746480142475015e-07, | |
| "loss": 1.4202, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.8641238253178551, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.6150840812348e-07, | |
| "loss": 1.4845, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.8650082918739636, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 4.5558836989621933e-07, | |
| "loss": 1.5008, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.8658927584300719, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 4.497047341647676e-07, | |
| "loss": 1.3423, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.8667772249861803, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 4.43857548059321e-07, | |
| "loss": 1.3769, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.8667772249861803, | |
| "eval_loss": 1.320468544960022, | |
| "eval_runtime": 232.1389, | |
| "eval_samples_per_second": 69.269, | |
| "eval_steps_per_second": 8.659, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.8676616915422886, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 4.380468584181069e-07, | |
| "loss": 1.5007, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.8685461580983969, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 1.3305, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.8694306246545053, | |
| "grad_norm": 2.125, | |
| "learning_rate": 4.265351544191365e-07, | |
| "loss": 1.321, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.8703150912106136, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 4.208342322745845e-07, | |
| "loss": 1.2659, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.871199557766722, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 4.151699910199336e-07, | |
| "loss": 1.2902, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.8720840243228303, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 4.095424760279454e-07, | |
| "loss": 1.377, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.8729684908789387, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.039517323771958e-07, | |
| "loss": 1.3572, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.873852957435047, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 3.983978048517029e-07, | |
| "loss": 1.3266, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.8747374239911553, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.9288073794057634e-07, | |
| "loss": 1.3651, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.8756218905472637, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.87400575837657e-07, | |
| "loss": 1.3594, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.876506357103372, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.819573624411643e-07, | |
| "loss": 1.3874, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.8773908236594804, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.765511413533429e-07, | |
| "loss": 1.3306, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.8782752902155887, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 3.711819558801183e-07, | |
| "loss": 1.5526, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.8791597567716971, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.6584984903074104e-07, | |
| "loss": 1.3536, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.8800442233278054, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 3.6055486351745327e-07, | |
| "loss": 1.4116, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.8809286898839138, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 3.5529704175513604e-07, | |
| "loss": 1.368, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.8818131564400221, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.500764258609779e-07, | |
| "loss": 1.4066, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.8826976229961304, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.4489305765413097e-07, | |
| "loss": 1.4416, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.8835820895522388, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.3974697865538195e-07, | |
| "loss": 1.397, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.8844665561083471, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 3.346382300868134e-07, | |
| "loss": 1.4165, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8853510226644555, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 3.2956685287148014e-07, | |
| "loss": 1.3116, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.8862354892205638, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.24532887633075e-07, | |
| "loss": 1.3223, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.8871199557766722, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 3.195363746956087e-07, | |
| "loss": 1.4175, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.8880044223327805, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 3.1457735408308153e-07, | |
| "loss": 1.34, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 3.096558655191706e-07, | |
| "loss": 1.3023, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.8897733554449972, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.047719484269018e-07, | |
| "loss": 1.5113, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.8906578220011055, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.9992564192834253e-07, | |
| "loss": 1.3554, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.891542288557214, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.9511698484428144e-07, | |
| "loss": 1.2778, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.8924267551133223, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 2.9034601569392417e-07, | |
| "loss": 1.441, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.8933112216694307, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.85612772694579e-07, | |
| "loss": 1.3236, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.894195688225539, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.809172937613552e-07, | |
| "loss": 1.3639, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.8950801547816473, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 2.762596165068548e-07, | |
| "loss": 1.4892, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.8959646213377557, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.7163977824087696e-07, | |
| "loss": 1.428, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.896849087893864, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.6705781597011147e-07, | |
| "loss": 1.4441, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.8977335544499724, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.6251376639785163e-07, | |
| "loss": 1.3683, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.8986180210060807, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 2.5800766592369077e-07, | |
| "loss": 1.385, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.8995024875621891, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.535395506432403e-07, | |
| "loss": 1.4734, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.9003869541182974, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.4910945634783024e-07, | |
| "loss": 1.4088, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.9012714206744058, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 1.3204, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.9021558872305141, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 2.403634723543674e-07, | |
| "loss": 1.2867, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9030403537866224, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.3604765271503073e-07, | |
| "loss": 1.5298, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.9039248203427308, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 2.3176999417760637e-07, | |
| "loss": 1.4862, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.9048092868988391, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.2753053100779632e-07, | |
| "loss": 1.4299, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.9056937534549475, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 2.2332929716533947e-07, | |
| "loss": 1.1436, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.9065782200110558, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.1916632630374579e-07, | |
| "loss": 1.4004, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.9074626865671642, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.1504165177002212e-07, | |
| "loss": 1.3453, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.9083471531232725, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 2.1095530660440954e-07, | |
| "loss": 1.3255, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.9092316196793808, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.069073235401109e-07, | |
| "loss": 1.3531, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.9101160862354892, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.028977350030409e-07, | |
| "loss": 1.2367, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.9110005527915975, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.989265731115525e-07, | |
| "loss": 1.2896, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.9118850193477059, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.9499386967619104e-07, | |
| "loss": 1.4309, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.9127694859038142, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.910996561994316e-07, | |
| "loss": 1.4226, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.9136539524599226, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.872439638754331e-07, | |
| "loss": 1.3422, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.914538419016031, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 1.834268235897807e-07, | |
| "loss": 1.4291, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.9154228855721394, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.7964826591924722e-07, | |
| "loss": 1.3052, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.9163073521282477, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.7590832113154045e-07, | |
| "loss": 1.4362, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.917191818684356, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.7220701918506666e-07, | |
| "loss": 1.3082, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.9180762852404644, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.685443897286859e-07, | |
| "loss": 1.444, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.9189607517965727, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 1.6492046210147884e-07, | |
| "loss": 1.3503, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.9198452183526811, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.6133526533250566e-07, | |
| "loss": 1.4377, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.9207296849087894, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 1.577888281405826e-07, | |
| "loss": 1.3313, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.9216141514648978, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.5428117893404315e-07, | |
| "loss": 1.3644, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.9224986180210061, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.5081234581051484e-07, | |
| "loss": 1.3837, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.9233830845771144, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 1.473823565566923e-07, | |
| "loss": 1.4241, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.9242675511332228, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 1.4399123864811904e-07, | |
| "loss": 1.3139, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.9251520176893311, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.4063901924895985e-07, | |
| "loss": 1.5017, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.9260364842454395, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.3732572521179043e-07, | |
| "loss": 1.3503, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.9269209508015478, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.3405138307737765e-07, | |
| "loss": 1.3755, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.9278054173576562, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.3081601907447007e-07, | |
| "loss": 1.4103, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 0.9286898839137645, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.2761965911958385e-07, | |
| "loss": 1.3431, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.9286898839137645, | |
| "eval_loss": 1.320536732673645, | |
| "eval_runtime": 232.1531, | |
| "eval_samples_per_second": 69.265, | |
| "eval_steps_per_second": 8.658, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.9295743504698729, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.244623288168012e-07, | |
| "loss": 1.3322, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 0.9304588170259812, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.2134405345755773e-07, | |
| "loss": 1.265, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.9313432835820895, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.1826485802044707e-07, | |
| "loss": 1.2818, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.9322277501381979, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.1522476717101605e-07, | |
| "loss": 1.3637, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.9331122166943062, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.1222380526156929e-07, | |
| "loss": 1.3624, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.9339966832504146, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 1.4057, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.9348811498065229, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.0633936410446077e-07, | |
| "loss": 1.3419, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 0.9357656163626313, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.0345593199344972e-07, | |
| "loss": 1.352, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.9366500829187396, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.0061172309534739e-07, | |
| "loss": 1.3915, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.9375345494748479, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 9.780676019336632e-08, | |
| "loss": 1.2872, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.9384190160309563, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 9.504106575634664e-08, | |
| "loss": 1.3331, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 0.9393034825870646, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 9.231466193856897e-08, | |
| "loss": 1.4363, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.940187949143173, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.96275705795846e-08, | |
| "loss": 1.4318, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 0.9410724156992814, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.697981320403337e-08, | |
| "loss": 1.492, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.9419568822553898, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.437141102147883e-08, | |
| "loss": 1.5143, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.9428413488114981, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 8.18023849262306e-08, | |
| "loss": 1.4165, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.9437258153676064, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 7.927275549718228e-08, | |
| "loss": 1.3239, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 0.9446102819237148, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.678254299764431e-08, | |
| "loss": 1.3245, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.9454947484798231, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 7.433176737518255e-08, | |
| "loss": 1.4069, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 0.9463792150359315, | |
| "grad_norm": 2.25, | |
| "learning_rate": 7.192044826145772e-08, | |
| "loss": 1.2734, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.9472636815920398, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 6.954860497207006e-08, | |
| "loss": 1.36, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 6.721625650640163e-08, | |
| "loss": 1.4026, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.9490326147042565, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 6.49234215474659e-08, | |
| "loss": 1.2729, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 0.9499170812603649, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 6.267011846175896e-08, | |
| "loss": 1.3159, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.9508015478164732, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 6.045636529911025e-08, | |
| "loss": 1.2304, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.9516860143725815, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 5.82821797925387e-08, | |
| "loss": 1.4583, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.9525704809286899, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 5.6147579358112324e-08, | |
| "loss": 1.4346, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.9534549474847982, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 5.4052581094806713e-08, | |
| "loss": 1.359, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.9543394140409066, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 5.19972017843684e-08, | |
| "loss": 1.2627, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 0.9552238805970149, | |
| "grad_norm": 2.125, | |
| "learning_rate": 4.998145789118114e-08, | |
| "loss": 1.3761, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.9561083471531233, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.800536556213375e-08, | |
| "loss": 1.4208, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 0.9569928137092316, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.60689406264897e-08, | |
| "loss": 1.3827, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.9578772802653399, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 4.417219859576383e-08, | |
| "loss": 1.3627, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.9587617468214483, | |
| "grad_norm": 2.25, | |
| "learning_rate": 4.231515466359193e-08, | |
| "loss": 1.3031, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.9596462133775566, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 4.0497823705615836e-08, | |
| "loss": 1.4236, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.960530679933665, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 3.8720220279359065e-08, | |
| "loss": 1.4736, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.9614151464897733, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.698235862411359e-08, | |
| "loss": 1.3541, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 0.9622996130458817, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.528425266082325e-08, | |
| "loss": 1.4398, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.96318407960199, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 3.36259159919744e-08, | |
| "loss": 1.3503, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.9640685461580984, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.2007361901485455e-08, | |
| "loss": 1.3672, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.9649530127142067, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 3.042860335460085e-08, | |
| "loss": 1.3574, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 0.965837479270315, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.8889652997787233e-08, | |
| "loss": 1.3291, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.9667219458264235, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.7390523158633552e-08, | |
| "loss": 1.3926, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 0.9676064123825318, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.593122584574892e-08, | |
| "loss": 1.2835, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.9684908789386402, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.4511772748669894e-08, | |
| "loss": 1.2501, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.9693753454947485, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 2.3132175237765564e-08, | |
| "loss": 1.2829, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.9702598120508569, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.179244436414485e-08, | |
| "loss": 1.3862, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 0.9711442786069652, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.0492590859571558e-08, | |
| "loss": 1.3531, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 0.9720287451630735, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.9232625136374472e-08, | |
| "loss": 1.3427, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 0.9729132117191819, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 1.8012557287367394e-08, | |
| "loss": 1.4296, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9737976782752902, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.6832397085765893e-08, | |
| "loss": 1.3767, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 0.9746821448313986, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.5692153985109037e-08, | |
| "loss": 1.2763, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 0.9755666113875069, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.4591837119186104e-08, | |
| "loss": 1.459, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 0.9764510779436153, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.3531455301960539e-08, | |
| "loss": 1.3972, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.9773355444997236, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.2511017027501682e-08, | |
| "loss": 1.303, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.978220011055832, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.1530530469914258e-08, | |
| "loss": 1.4383, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 0.9791044776119403, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.0590003483275657e-08, | |
| "loss": 1.3901, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 0.9799889441680486, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 9.689443601570425e-09, | |
| "loss": 1.385, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 0.980873410724157, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 8.828858038632538e-09, | |
| "loss": 1.3062, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 0.9817578772802653, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 8.008253688084888e-09, | |
| "loss": 1.4997, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9826423438363737, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 7.227637123285447e-09, | |
| "loss": 1.3024, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 0.983526810392482, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 6.487014597275631e-09, | |
| "loss": 1.4312, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 0.9844112769485904, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 5.7863920427286834e-09, | |
| "loss": 1.5325, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 0.9852957435046987, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 5.125775071901373e-09, | |
| "loss": 1.3651, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 0.986180210060807, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 4.505168976592922e-09, | |
| "loss": 1.3628, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.9870646766169154, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 3.924578728096707e-09, | |
| "loss": 1.2619, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.9879491431730237, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.3840089771658424e-09, | |
| "loss": 1.4065, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 0.9888336097291321, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.8834640539737723e-09, | |
| "loss": 1.3817, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 0.9897180762852404, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.422947968078182e-09, | |
| "loss": 1.2757, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 0.9906025428413489, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.002464408392135e-09, | |
| "loss": 1.4751, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9906025428413489, | |
| "eval_loss": 1.3204842805862427, | |
| "eval_runtime": 232.1096, | |
| "eval_samples_per_second": 69.278, | |
| "eval_steps_per_second": 8.66, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9914870093974572, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.622016743150212e-09, | |
| "loss": 1.3439, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 0.9923714759535655, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 1.2816080198868596e-09, | |
| "loss": 1.422, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 0.9932559425096739, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 9.812409654075261e-10, | |
| "loss": 1.5471, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 0.9941404090657822, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 7.209179857675664e-10, | |
| "loss": 1.3187, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 0.9950248756218906, | |
| "grad_norm": 2.125, | |
| "learning_rate": 5.006411662555888e-10, | |
| "loss": 1.3619, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.9959093421779989, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 3.204122713740265e-10, | |
| "loss": 1.4264, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 0.9967938087341073, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.8023274482636965e-10, | |
| "loss": 1.2781, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 0.9976782752902156, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 8.01037095038426e-11, | |
| "loss": 1.5524, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 0.998562741846324, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 2.0025967479853082e-11, | |
| "loss": 1.522, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 0.9994472084024323, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 0.0, | |
| "loss": 1.279, | |
| "step": 1130 | |
| } | |
| ], | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.3240198238109696e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |