| { | |
| "best_metric": 3.3120391368865967, | |
| "best_model_checkpoint": "./models/full-finetuning/german-gpt2/checkpoint-58000", | |
| "epoch": 1.0, | |
| "eval_steps": 1000, | |
| "global_step": 59835, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008356313194618534, | |
| "grad_norm": 7.7337164878845215, | |
| "learning_rate": 4.97e-05, | |
| "loss": 5.7213, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.016712626389237067, | |
| "grad_norm": 6.002913951873779, | |
| "learning_rate": 4.95811915395635e-05, | |
| "loss": 4.8528, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.016712626389237067, | |
| "eval_loss": 4.570379734039307, | |
| "eval_runtime": 33.8087, | |
| "eval_samples_per_second": 131.978, | |
| "eval_steps_per_second": 16.505, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.025068939583855605, | |
| "grad_norm": 5.462852478027344, | |
| "learning_rate": 4.915985506025112e-05, | |
| "loss": 4.6135, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.033425252778474135, | |
| "grad_norm": 6.940282821655273, | |
| "learning_rate": 4.873851858093874e-05, | |
| "loss": 4.4938, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.033425252778474135, | |
| "eval_loss": 4.2595720291137695, | |
| "eval_runtime": 36.1247, | |
| "eval_samples_per_second": 123.516, | |
| "eval_steps_per_second": 15.446, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.04178156597309267, | |
| "grad_norm": 6.612778663635254, | |
| "learning_rate": 4.8317182101626365e-05, | |
| "loss": 4.3683, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.05013787916771121, | |
| "grad_norm": 5.179454803466797, | |
| "learning_rate": 4.789584562231398e-05, | |
| "loss": 4.2089, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.05013787916771121, | |
| "eval_loss": 4.104772567749023, | |
| "eval_runtime": 33.8319, | |
| "eval_samples_per_second": 131.887, | |
| "eval_steps_per_second": 16.493, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.05849419236232974, | |
| "grad_norm": 4.007822036743164, | |
| "learning_rate": 4.74745091430016e-05, | |
| "loss": 4.2532, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.06685050555694827, | |
| "grad_norm": 5.069295883178711, | |
| "learning_rate": 4.7053172663689226e-05, | |
| "loss": 4.2046, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.06685050555694827, | |
| "eval_loss": 4.0018157958984375, | |
| "eval_runtime": 33.841, | |
| "eval_samples_per_second": 131.852, | |
| "eval_steps_per_second": 16.489, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.0752068187515668, | |
| "grad_norm": 8.424234390258789, | |
| "learning_rate": 4.663183618437685e-05, | |
| "loss": 4.1513, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.08356313194618534, | |
| "grad_norm": 5.6522016525268555, | |
| "learning_rate": 4.621049970506447e-05, | |
| "loss": 4.0912, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.08356313194618534, | |
| "eval_loss": 3.927964687347412, | |
| "eval_runtime": 33.8173, | |
| "eval_samples_per_second": 131.944, | |
| "eval_steps_per_second": 16.5, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.09191944514080387, | |
| "grad_norm": 6.062402248382568, | |
| "learning_rate": 4.5789163225752086e-05, | |
| "loss": 4.0602, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.10027575833542242, | |
| "grad_norm": 4.522783279418945, | |
| "learning_rate": 4.536782674643971e-05, | |
| "loss": 4.0243, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.10027575833542242, | |
| "eval_loss": 3.8684792518615723, | |
| "eval_runtime": 33.8061, | |
| "eval_samples_per_second": 131.988, | |
| "eval_steps_per_second": 16.506, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.10863207153004095, | |
| "grad_norm": 5.732029914855957, | |
| "learning_rate": 4.494649026712733e-05, | |
| "loss": 3.9523, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.11698838472465949, | |
| "grad_norm": 5.034173011779785, | |
| "learning_rate": 4.452599646077358e-05, | |
| "loss": 3.9991, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.11698838472465949, | |
| "eval_loss": 3.818037748336792, | |
| "eval_runtime": 33.7911, | |
| "eval_samples_per_second": 132.047, | |
| "eval_steps_per_second": 16.513, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.12534469791927802, | |
| "grad_norm": 6.326226711273193, | |
| "learning_rate": 4.41046599814612e-05, | |
| "loss": 3.9681, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.13370101111389654, | |
| "grad_norm": 5.3602495193481445, | |
| "learning_rate": 4.368332350214882e-05, | |
| "loss": 3.9101, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.13370101111389654, | |
| "eval_loss": 3.7742578983306885, | |
| "eval_runtime": 33.7987, | |
| "eval_samples_per_second": 132.017, | |
| "eval_steps_per_second": 16.51, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.1420573243085151, | |
| "grad_norm": 6.7508111000061035, | |
| "learning_rate": 4.326198702283644e-05, | |
| "loss": 3.9032, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.1504136375031336, | |
| "grad_norm": 5.866630554199219, | |
| "learning_rate": 4.2840650543524055e-05, | |
| "loss": 3.8962, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.1504136375031336, | |
| "eval_loss": 3.7523410320281982, | |
| "eval_runtime": 33.8046, | |
| "eval_samples_per_second": 131.994, | |
| "eval_steps_per_second": 16.507, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.15876995069775215, | |
| "grad_norm": 4.693000316619873, | |
| "learning_rate": 4.241931406421168e-05, | |
| "loss": 3.853, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.16712626389237067, | |
| "grad_norm": 5.653314113616943, | |
| "learning_rate": 4.199882025785793e-05, | |
| "loss": 3.8244, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.16712626389237067, | |
| "eval_loss": 3.711071014404297, | |
| "eval_runtime": 33.8163, | |
| "eval_samples_per_second": 131.948, | |
| "eval_steps_per_second": 16.501, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.17548257708698922, | |
| "grad_norm": 4.951623439788818, | |
| "learning_rate": 4.157748377854555e-05, | |
| "loss": 3.8097, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.18383889028160774, | |
| "grad_norm": 9.94375228881836, | |
| "learning_rate": 4.115614729923317e-05, | |
| "loss": 3.7687, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.18383889028160774, | |
| "eval_loss": 3.6785600185394287, | |
| "eval_runtime": 33.809, | |
| "eval_samples_per_second": 131.977, | |
| "eval_steps_per_second": 16.504, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.1921952034762263, | |
| "grad_norm": 7.155759334564209, | |
| "learning_rate": 4.0734810819920794e-05, | |
| "loss": 3.8008, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.20055151667084484, | |
| "grad_norm": 3.9872701168060303, | |
| "learning_rate": 4.031347434060841e-05, | |
| "loss": 3.7853, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.20055151667084484, | |
| "eval_loss": 3.655622959136963, | |
| "eval_runtime": 33.807, | |
| "eval_samples_per_second": 131.985, | |
| "eval_steps_per_second": 16.505, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.20890782986546336, | |
| "grad_norm": 7.3907647132873535, | |
| "learning_rate": 3.989298053425466e-05, | |
| "loss": 3.8082, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.2172641430600819, | |
| "grad_norm": 4.089804172515869, | |
| "learning_rate": 3.947164405494228e-05, | |
| "loss": 3.7216, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.2172641430600819, | |
| "eval_loss": 3.641737937927246, | |
| "eval_runtime": 33.7976, | |
| "eval_samples_per_second": 132.021, | |
| "eval_steps_per_second": 16.51, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.22562045625470042, | |
| "grad_norm": 3.7814671993255615, | |
| "learning_rate": 3.90503075756299e-05, | |
| "loss": 3.7132, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.23397676944931897, | |
| "grad_norm": 5.00490665435791, | |
| "learning_rate": 3.8628971096317526e-05, | |
| "loss": 3.7508, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.23397676944931897, | |
| "eval_loss": 3.6097123622894287, | |
| "eval_runtime": 33.8097, | |
| "eval_samples_per_second": 131.974, | |
| "eval_steps_per_second": 16.504, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.2423330826439375, | |
| "grad_norm": 4.1177659034729, | |
| "learning_rate": 3.820763461700515e-05, | |
| "loss": 3.7317, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.25068939583855604, | |
| "grad_norm": 5.807891845703125, | |
| "learning_rate": 3.7786298137692763e-05, | |
| "loss": 3.7186, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.25068939583855604, | |
| "eval_loss": 3.5940768718719482, | |
| "eval_runtime": 33.8398, | |
| "eval_samples_per_second": 131.856, | |
| "eval_steps_per_second": 16.489, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.2590457090331746, | |
| "grad_norm": 5.580984592437744, | |
| "learning_rate": 3.736496165838038e-05, | |
| "loss": 3.7059, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.2674020222277931, | |
| "grad_norm": 5.553014278411865, | |
| "learning_rate": 3.6943625179068e-05, | |
| "loss": 3.6317, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.2674020222277931, | |
| "eval_loss": 3.5815696716308594, | |
| "eval_runtime": 33.7945, | |
| "eval_samples_per_second": 132.033, | |
| "eval_steps_per_second": 16.512, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.2757583354224116, | |
| "grad_norm": 7.818964958190918, | |
| "learning_rate": 3.652313137271425e-05, | |
| "loss": 3.6837, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.2841146486170302, | |
| "grad_norm": 7.232082843780518, | |
| "learning_rate": 3.610179489340187e-05, | |
| "loss": 3.6965, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.2841146486170302, | |
| "eval_loss": 3.559220552444458, | |
| "eval_runtime": 33.8032, | |
| "eval_samples_per_second": 131.999, | |
| "eval_steps_per_second": 16.507, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.2924709618116487, | |
| "grad_norm": 5.058260917663574, | |
| "learning_rate": 3.5680458414089495e-05, | |
| "loss": 3.6218, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.3008272750062672, | |
| "grad_norm": 6.645322322845459, | |
| "learning_rate": 3.525912193477712e-05, | |
| "loss": 3.6203, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.3008272750062672, | |
| "eval_loss": 3.5445735454559326, | |
| "eval_runtime": 33.826, | |
| "eval_samples_per_second": 131.911, | |
| "eval_steps_per_second": 16.496, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.30918358820088576, | |
| "grad_norm": 6.7204084396362305, | |
| "learning_rate": 3.483862812842336e-05, | |
| "loss": 3.6135, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.3175399013955043, | |
| "grad_norm": 5.846601486206055, | |
| "learning_rate": 3.441729164911098e-05, | |
| "loss": 3.6139, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.3175399013955043, | |
| "eval_loss": 3.529482364654541, | |
| "eval_runtime": 33.8084, | |
| "eval_samples_per_second": 131.979, | |
| "eval_steps_per_second": 16.505, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.32589621459012286, | |
| "grad_norm": 4.412099838256836, | |
| "learning_rate": 3.3995955169798604e-05, | |
| "loss": 3.6202, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.33425252778474135, | |
| "grad_norm": 6.884310722351074, | |
| "learning_rate": 3.357461869048623e-05, | |
| "loss": 3.5862, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.33425252778474135, | |
| "eval_loss": 3.518963098526001, | |
| "eval_runtime": 33.8217, | |
| "eval_samples_per_second": 131.927, | |
| "eval_steps_per_second": 16.498, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.3426088409793599, | |
| "grad_norm": 4.762236595153809, | |
| "learning_rate": 3.315328221117385e-05, | |
| "loss": 3.613, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.35096515417397844, | |
| "grad_norm": 3.5604190826416016, | |
| "learning_rate": 3.273278840482009e-05, | |
| "loss": 3.5767, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.35096515417397844, | |
| "eval_loss": 3.5067789554595947, | |
| "eval_runtime": 33.8036, | |
| "eval_samples_per_second": 131.998, | |
| "eval_steps_per_second": 16.507, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.359321467368597, | |
| "grad_norm": 3.5096914768218994, | |
| "learning_rate": 3.2311451925507714e-05, | |
| "loss": 3.5692, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.3676777805632155, | |
| "grad_norm": 4.54230260848999, | |
| "learning_rate": 3.1890115446195336e-05, | |
| "loss": 3.5914, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.3676777805632155, | |
| "eval_loss": 3.491032361984253, | |
| "eval_runtime": 33.8108, | |
| "eval_samples_per_second": 131.97, | |
| "eval_steps_per_second": 16.504, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.37603409375783403, | |
| "grad_norm": 5.874061584472656, | |
| "learning_rate": 3.146877896688296e-05, | |
| "loss": 3.6494, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.3843904069524526, | |
| "grad_norm": 5.265365123748779, | |
| "learning_rate": 3.10482851605292e-05, | |
| "loss": 3.5796, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.3843904069524526, | |
| "eval_loss": 3.4834258556365967, | |
| "eval_runtime": 33.7961, | |
| "eval_samples_per_second": 132.027, | |
| "eval_steps_per_second": 16.511, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.3927467201470711, | |
| "grad_norm": 4.7764129638671875, | |
| "learning_rate": 3.062694868121682e-05, | |
| "loss": 3.5603, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.4011030333416897, | |
| "grad_norm": 4.89961576461792, | |
| "learning_rate": 3.0205612201904442e-05, | |
| "loss": 3.5427, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.4011030333416897, | |
| "eval_loss": 3.476203203201294, | |
| "eval_runtime": 33.8616, | |
| "eval_samples_per_second": 131.772, | |
| "eval_steps_per_second": 16.479, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.40945934653630817, | |
| "grad_norm": 6.659942150115967, | |
| "learning_rate": 2.9784275722592064e-05, | |
| "loss": 3.5559, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.4178156597309267, | |
| "grad_norm": 4.591799259185791, | |
| "learning_rate": 2.936378191623831e-05, | |
| "loss": 3.5087, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.4178156597309267, | |
| "eval_loss": 3.465503454208374, | |
| "eval_runtime": 33.8191, | |
| "eval_samples_per_second": 131.937, | |
| "eval_steps_per_second": 16.5, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.42617197292554526, | |
| "grad_norm": 4.84548282623291, | |
| "learning_rate": 2.894244543692593e-05, | |
| "loss": 3.5474, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.4345282861201638, | |
| "grad_norm": 5.839838027954102, | |
| "learning_rate": 2.852110895761355e-05, | |
| "loss": 3.5723, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.4345282861201638, | |
| "eval_loss": 3.458843946456909, | |
| "eval_runtime": 33.8846, | |
| "eval_samples_per_second": 131.682, | |
| "eval_steps_per_second": 16.468, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.4428845993147823, | |
| "grad_norm": 3.7624571323394775, | |
| "learning_rate": 2.8099772478301174e-05, | |
| "loss": 3.4898, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.45124091250940085, | |
| "grad_norm": 4.763157844543457, | |
| "learning_rate": 2.7678435998988793e-05, | |
| "loss": 3.5493, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.45124091250940085, | |
| "eval_loss": 3.4483096599578857, | |
| "eval_runtime": 33.7859, | |
| "eval_samples_per_second": 132.067, | |
| "eval_steps_per_second": 16.516, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.4595972257040194, | |
| "grad_norm": 4.820297718048096, | |
| "learning_rate": 2.725794219263504e-05, | |
| "loss": 3.5162, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.46795353889863794, | |
| "grad_norm": 8.342415809631348, | |
| "learning_rate": 2.683660571332266e-05, | |
| "loss": 3.4815, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.46795353889863794, | |
| "eval_loss": 3.4425435066223145, | |
| "eval_runtime": 33.8093, | |
| "eval_samples_per_second": 131.976, | |
| "eval_steps_per_second": 16.504, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.47630985209325644, | |
| "grad_norm": 3.9816768169403076, | |
| "learning_rate": 2.641526923401028e-05, | |
| "loss": 3.5168, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.484666165287875, | |
| "grad_norm": 4.662781715393066, | |
| "learning_rate": 2.5993932754697902e-05, | |
| "loss": 3.5442, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.484666165287875, | |
| "eval_loss": 3.4301340579986572, | |
| "eval_runtime": 33.7983, | |
| "eval_samples_per_second": 132.018, | |
| "eval_steps_per_second": 16.51, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.49302247848249353, | |
| "grad_norm": 5.855069160461426, | |
| "learning_rate": 2.5573438948344148e-05, | |
| "loss": 3.5048, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.5013787916771121, | |
| "grad_norm": 3.8957674503326416, | |
| "learning_rate": 2.515210246903177e-05, | |
| "loss": 3.5452, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.5013787916771121, | |
| "eval_loss": 3.4219019412994385, | |
| "eval_runtime": 33.8062, | |
| "eval_samples_per_second": 131.988, | |
| "eval_steps_per_second": 16.506, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.5097351048717306, | |
| "grad_norm": 5.273784160614014, | |
| "learning_rate": 2.473076598971939e-05, | |
| "loss": 3.4951, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.5180914180663492, | |
| "grad_norm": 6.109909534454346, | |
| "learning_rate": 2.430942951040701e-05, | |
| "loss": 3.4879, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.5180914180663492, | |
| "eval_loss": 3.419405698776245, | |
| "eval_runtime": 33.7875, | |
| "eval_samples_per_second": 132.061, | |
| "eval_steps_per_second": 16.515, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.5264477312609677, | |
| "grad_norm": 7.194368839263916, | |
| "learning_rate": 2.3888935704053257e-05, | |
| "loss": 3.5259, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.5348040444555862, | |
| "grad_norm": 4.303890228271484, | |
| "learning_rate": 2.3468441897699503e-05, | |
| "loss": 3.4956, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.5348040444555862, | |
| "eval_loss": 3.4062435626983643, | |
| "eval_runtime": 33.8183, | |
| "eval_samples_per_second": 131.94, | |
| "eval_steps_per_second": 16.5, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.5431603576502048, | |
| "grad_norm": 4.180033206939697, | |
| "learning_rate": 2.3047105418387125e-05, | |
| "loss": 3.5137, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.5515166708448233, | |
| "grad_norm": 4.65141487121582, | |
| "learning_rate": 2.2625768939074744e-05, | |
| "loss": 3.4635, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.5515166708448233, | |
| "eval_loss": 3.3979876041412354, | |
| "eval_runtime": 33.8238, | |
| "eval_samples_per_second": 131.919, | |
| "eval_steps_per_second": 16.497, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.5598729840394417, | |
| "grad_norm": 7.3677215576171875, | |
| "learning_rate": 2.2204432459762367e-05, | |
| "loss": 3.5071, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 0.5682292972340603, | |
| "grad_norm": 4.072124481201172, | |
| "learning_rate": 2.178309598044999e-05, | |
| "loss": 3.4836, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.5682292972340603, | |
| "eval_loss": 3.3961925506591797, | |
| "eval_runtime": 33.8074, | |
| "eval_samples_per_second": 131.983, | |
| "eval_steps_per_second": 16.505, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.5765856104286788, | |
| "grad_norm": 4.479409217834473, | |
| "learning_rate": 2.136175950113761e-05, | |
| "loss": 3.4763, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 0.5849419236232974, | |
| "grad_norm": 3.3214428424835205, | |
| "learning_rate": 2.094042302182523e-05, | |
| "loss": 3.4378, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.5849419236232974, | |
| "eval_loss": 3.389758825302124, | |
| "eval_runtime": 33.7986, | |
| "eval_samples_per_second": 132.017, | |
| "eval_steps_per_second": 16.51, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.5932982368179159, | |
| "grad_norm": 6.104400157928467, | |
| "learning_rate": 2.0519086542512852e-05, | |
| "loss": 3.4959, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 0.6016545500125344, | |
| "grad_norm": 5.734145641326904, | |
| "learning_rate": 2.0097750063200475e-05, | |
| "loss": 3.4673, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.6016545500125344, | |
| "eval_loss": 3.3848955631256104, | |
| "eval_runtime": 33.8025, | |
| "eval_samples_per_second": 132.002, | |
| "eval_steps_per_second": 16.508, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.610010863207153, | |
| "grad_norm": 4.3256683349609375, | |
| "learning_rate": 1.9677256256846717e-05, | |
| "loss": 3.4903, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 0.6183671764017715, | |
| "grad_norm": 5.769290447235107, | |
| "learning_rate": 1.925591977753434e-05, | |
| "loss": 3.4722, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.6183671764017715, | |
| "eval_loss": 3.3791496753692627, | |
| "eval_runtime": 33.8066, | |
| "eval_samples_per_second": 131.986, | |
| "eval_steps_per_second": 16.506, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.6267234895963901, | |
| "grad_norm": 4.285965442657471, | |
| "learning_rate": 1.8834583298221962e-05, | |
| "loss": 3.4922, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 0.6350798027910086, | |
| "grad_norm": 4.97435998916626, | |
| "learning_rate": 1.8413246818909584e-05, | |
| "loss": 3.4814, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.6350798027910086, | |
| "eval_loss": 3.3762271404266357, | |
| "eval_runtime": 33.8323, | |
| "eval_samples_per_second": 131.886, | |
| "eval_steps_per_second": 16.493, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.6434361159856271, | |
| "grad_norm": 4.531788349151611, | |
| "learning_rate": 1.7992753012555827e-05, | |
| "loss": 3.4144, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 0.6517924291802457, | |
| "grad_norm": 5.197137355804443, | |
| "learning_rate": 1.757141653324345e-05, | |
| "loss": 3.4358, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.6517924291802457, | |
| "eval_loss": 3.3685717582702637, | |
| "eval_runtime": 33.7899, | |
| "eval_samples_per_second": 132.051, | |
| "eval_steps_per_second": 16.514, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.6601487423748642, | |
| "grad_norm": 5.673033714294434, | |
| "learning_rate": 1.715008005393107e-05, | |
| "loss": 3.3936, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 0.6685050555694827, | |
| "grad_norm": 4.699774742126465, | |
| "learning_rate": 1.672874357461869e-05, | |
| "loss": 3.4711, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.6685050555694827, | |
| "eval_loss": 3.3646111488342285, | |
| "eval_runtime": 33.8073, | |
| "eval_samples_per_second": 131.983, | |
| "eval_steps_per_second": 16.505, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.6768613687641013, | |
| "grad_norm": 4.5016889572143555, | |
| "learning_rate": 1.630824976826494e-05, | |
| "loss": 3.4073, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 0.6852176819587198, | |
| "grad_norm": 9.633309364318848, | |
| "learning_rate": 1.5886913288952558e-05, | |
| "loss": 3.4437, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.6852176819587198, | |
| "eval_loss": 3.3601598739624023, | |
| "eval_runtime": 33.7786, | |
| "eval_samples_per_second": 132.096, | |
| "eval_steps_per_second": 16.519, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.6935739951533384, | |
| "grad_norm": 3.356966733932495, | |
| "learning_rate": 1.5465576809640177e-05, | |
| "loss": 3.3706, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 0.7019303083479569, | |
| "grad_norm": 4.3639678955078125, | |
| "learning_rate": 1.50442403303278e-05, | |
| "loss": 3.4171, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.7019303083479569, | |
| "eval_loss": 3.358569622039795, | |
| "eval_runtime": 33.8063, | |
| "eval_samples_per_second": 131.987, | |
| "eval_steps_per_second": 16.506, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.7102866215425754, | |
| "grad_norm": 6.138998985290527, | |
| "learning_rate": 1.4623746523974047e-05, | |
| "loss": 3.3822, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 0.718642934737194, | |
| "grad_norm": 5.975950717926025, | |
| "learning_rate": 1.4202410044661668e-05, | |
| "loss": 3.442, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.718642934737194, | |
| "eval_loss": 3.3521432876586914, | |
| "eval_runtime": 33.7743, | |
| "eval_samples_per_second": 132.112, | |
| "eval_steps_per_second": 16.521, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.7269992479318125, | |
| "grad_norm": 5.316736698150635, | |
| "learning_rate": 1.378107356534929e-05, | |
| "loss": 3.3837, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 0.735355561126431, | |
| "grad_norm": 4.64693021774292, | |
| "learning_rate": 1.335973708603691e-05, | |
| "loss": 3.3846, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.735355561126431, | |
| "eval_loss": 3.3473236560821533, | |
| "eval_runtime": 33.8058, | |
| "eval_samples_per_second": 131.989, | |
| "eval_steps_per_second": 16.506, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.7437118743210496, | |
| "grad_norm": 7.507499694824219, | |
| "learning_rate": 1.293840060672453e-05, | |
| "loss": 3.4229, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 0.7520681875156681, | |
| "grad_norm": 4.695137023925781, | |
| "learning_rate": 1.2517906800370777e-05, | |
| "loss": 3.4064, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.7520681875156681, | |
| "eval_loss": 3.3478496074676514, | |
| "eval_runtime": 33.7907, | |
| "eval_samples_per_second": 132.048, | |
| "eval_steps_per_second": 16.513, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.7604245007102867, | |
| "grad_norm": 6.782580375671387, | |
| "learning_rate": 1.2096570321058398e-05, | |
| "loss": 3.4665, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 0.7687808139049052, | |
| "grad_norm": 7.160044193267822, | |
| "learning_rate": 1.167523384174602e-05, | |
| "loss": 3.4181, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.7687808139049052, | |
| "eval_loss": 3.339061737060547, | |
| "eval_runtime": 33.8357, | |
| "eval_samples_per_second": 131.872, | |
| "eval_steps_per_second": 16.491, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.7771371270995237, | |
| "grad_norm": 6.559309482574463, | |
| "learning_rate": 1.1253897362433639e-05, | |
| "loss": 3.3691, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 0.7854934402941423, | |
| "grad_norm": 4.2476043701171875, | |
| "learning_rate": 1.0832560883121261e-05, | |
| "loss": 3.3825, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.7854934402941423, | |
| "eval_loss": 3.3343887329101562, | |
| "eval_runtime": 33.7813, | |
| "eval_samples_per_second": 132.085, | |
| "eval_steps_per_second": 16.518, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.7938497534887607, | |
| "grad_norm": 8.163984298706055, | |
| "learning_rate": 1.0411224403808882e-05, | |
| "loss": 3.4073, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 0.8022060666833793, | |
| "grad_norm": 4.705440998077393, | |
| "learning_rate": 9.989887924496504e-06, | |
| "loss": 3.4277, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.8022060666833793, | |
| "eval_loss": 3.334386110305786, | |
| "eval_runtime": 33.8549, | |
| "eval_samples_per_second": 131.798, | |
| "eval_steps_per_second": 16.482, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.8105623798779978, | |
| "grad_norm": 5.080317497253418, | |
| "learning_rate": 9.56939411814275e-06, | |
| "loss": 3.4014, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 0.8189186930726163, | |
| "grad_norm": 4.523305892944336, | |
| "learning_rate": 9.14805763883037e-06, | |
| "loss": 3.331, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 0.8189186930726163, | |
| "eval_loss": 3.33245587348938, | |
| "eval_runtime": 33.9406, | |
| "eval_samples_per_second": 131.465, | |
| "eval_steps_per_second": 16.44, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 0.8272750062672349, | |
| "grad_norm": 5.878600597381592, | |
| "learning_rate": 8.726721159517993e-06, | |
| "loss": 3.4097, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 0.8356313194618534, | |
| "grad_norm": 4.23352575302124, | |
| "learning_rate": 8.305384680205612e-06, | |
| "loss": 3.4073, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.8356313194618534, | |
| "eval_loss": 3.328063488006592, | |
| "eval_runtime": 33.9523, | |
| "eval_samples_per_second": 131.42, | |
| "eval_steps_per_second": 16.435, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.8439876326564719, | |
| "grad_norm": 6.142092227935791, | |
| "learning_rate": 7.884890873851858e-06, | |
| "loss": 3.3904, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 0.8523439458510905, | |
| "grad_norm": 7.910597324371338, | |
| "learning_rate": 7.463554394539479e-06, | |
| "loss": 3.3741, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 0.8523439458510905, | |
| "eval_loss": 3.327683448791504, | |
| "eval_runtime": 33.9417, | |
| "eval_samples_per_second": 131.461, | |
| "eval_steps_per_second": 16.44, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 0.860700259045709, | |
| "grad_norm": 4.868673324584961, | |
| "learning_rate": 7.0422179152271005e-06, | |
| "loss": 3.3615, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 0.8690565722403276, | |
| "grad_norm": 4.47282075881958, | |
| "learning_rate": 6.620881435914722e-06, | |
| "loss": 3.3467, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 0.8690565722403276, | |
| "eval_loss": 3.3206417560577393, | |
| "eval_runtime": 33.9544, | |
| "eval_samples_per_second": 131.412, | |
| "eval_steps_per_second": 16.434, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 0.8774128854349461, | |
| "grad_norm": 5.413076877593994, | |
| "learning_rate": 6.200387629560968e-06, | |
| "loss": 3.3221, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 0.8857691986295646, | |
| "grad_norm": 5.672983169555664, | |
| "learning_rate": 5.779051150248588e-06, | |
| "loss": 3.418, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 0.8857691986295646, | |
| "eval_loss": 3.320605516433716, | |
| "eval_runtime": 33.884, | |
| "eval_samples_per_second": 131.684, | |
| "eval_steps_per_second": 16.468, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 0.8941255118241832, | |
| "grad_norm": 8.368507385253906, | |
| "learning_rate": 5.35771467093621e-06, | |
| "loss": 3.3747, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 0.9024818250188017, | |
| "grad_norm": 6.918625831604004, | |
| "learning_rate": 4.936378191623831e-06, | |
| "loss": 3.416, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 0.9024818250188017, | |
| "eval_loss": 3.317139148712158, | |
| "eval_runtime": 33.9331, | |
| "eval_samples_per_second": 131.494, | |
| "eval_steps_per_second": 16.444, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 0.9108381382134202, | |
| "grad_norm": 6.566315174102783, | |
| "learning_rate": 4.515884385270077e-06, | |
| "loss": 3.3985, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 0.9191944514080388, | |
| "grad_norm": 4.790822982788086, | |
| "learning_rate": 4.094547905957698e-06, | |
| "loss": 3.3607, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.9191944514080388, | |
| "eval_loss": 3.3170626163482666, | |
| "eval_runtime": 33.9426, | |
| "eval_samples_per_second": 131.457, | |
| "eval_steps_per_second": 16.439, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.9275507646026573, | |
| "grad_norm": 5.501828193664551, | |
| "learning_rate": 3.6732114266453192e-06, | |
| "loss": 3.3586, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 0.9359070777972759, | |
| "grad_norm": 8.011107444763184, | |
| "learning_rate": 3.2518749473329403e-06, | |
| "loss": 3.4076, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 0.9359070777972759, | |
| "eval_loss": 3.3138890266418457, | |
| "eval_runtime": 33.9309, | |
| "eval_samples_per_second": 131.503, | |
| "eval_steps_per_second": 16.445, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 0.9442633909918944, | |
| "grad_norm": 5.630584239959717, | |
| "learning_rate": 2.8305384680205613e-06, | |
| "loss": 3.4309, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 0.9526197041865129, | |
| "grad_norm": 3.990445137023926, | |
| "learning_rate": 2.4092019887081824e-06, | |
| "loss": 3.3257, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 0.9526197041865129, | |
| "eval_loss": 3.3135273456573486, | |
| "eval_runtime": 33.8948, | |
| "eval_samples_per_second": 131.643, | |
| "eval_steps_per_second": 16.463, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 0.9609760173811315, | |
| "grad_norm": 4.2760396003723145, | |
| "learning_rate": 1.9878655093958034e-06, | |
| "loss": 3.3362, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 0.96933233057575, | |
| "grad_norm": 4.5603532791137695, | |
| "learning_rate": 1.5665290300834245e-06, | |
| "loss": 3.3657, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 0.96933233057575, | |
| "eval_loss": 3.3120391368865967, | |
| "eval_runtime": 33.957, | |
| "eval_samples_per_second": 131.401, | |
| "eval_steps_per_second": 16.433, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 0.9776886437703685, | |
| "grad_norm": 3.6152896881103516, | |
| "learning_rate": 1.1460352237296705e-06, | |
| "loss": 3.3847, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 0.9860449569649871, | |
| "grad_norm": 5.7242751121521, | |
| "learning_rate": 7.246987444172916e-07, | |
| "loss": 3.3509, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 0.9860449569649871, | |
| "eval_loss": 3.3113863468170166, | |
| "eval_runtime": 33.9749, | |
| "eval_samples_per_second": 131.332, | |
| "eval_steps_per_second": 16.424, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 0.9944012701596056, | |
| "grad_norm": 8.690220832824707, | |
| "learning_rate": 3.033622651049128e-07, | |
| "loss": 3.3511, | |
| "step": 59500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59835, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 2000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.50149494587392e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |