| { | |
| "best_metric": 3.1603705883026123, | |
| "best_model_checkpoint": "./models/full-finetuning/LLaMmlein_120M/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": 40.14425277709961, | |
| "learning_rate": 4.96e-05, | |
| "loss": 4.8477, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.016712626389237067, | |
| "grad_norm": 22.107275009155273, | |
| "learning_rate": 4.958203421252212e-05, | |
| "loss": 4.4466, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.016712626389237067, | |
| "eval_loss": 4.3118462562561035, | |
| "eval_runtime": 22.0404, | |
| "eval_samples_per_second": 202.446, | |
| "eval_steps_per_second": 25.317, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.025068939583855605, | |
| "grad_norm": 19.609399795532227, | |
| "learning_rate": 4.916069773320974e-05, | |
| "loss": 4.2986, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.033425252778474135, | |
| "grad_norm": 18.64438247680664, | |
| "learning_rate": 4.873936125389736e-05, | |
| "loss": 4.2295, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.033425252778474135, | |
| "eval_loss": 4.121812343597412, | |
| "eval_runtime": 22.1428, | |
| "eval_samples_per_second": 201.51, | |
| "eval_steps_per_second": 25.2, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.04178156597309267, | |
| "grad_norm": 19.098369598388672, | |
| "learning_rate": 4.8318024774584986e-05, | |
| "loss": 4.1165, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.05013787916771121, | |
| "grad_norm": 17.83785629272461, | |
| "learning_rate": 4.789668829527261e-05, | |
| "loss": 4.029, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.05013787916771121, | |
| "eval_loss": 4.040452480316162, | |
| "eval_runtime": 22.088, | |
| "eval_samples_per_second": 202.01, | |
| "eval_steps_per_second": 25.263, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.05849419236232974, | |
| "grad_norm": 14.394288063049316, | |
| "learning_rate": 4.747535181596023e-05, | |
| "loss": 4.0701, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.06685050555694827, | |
| "grad_norm": 17.005945205688477, | |
| "learning_rate": 4.705401533664785e-05, | |
| "loss": 4.0239, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.06685050555694827, | |
| "eval_loss": 3.948943853378296, | |
| "eval_runtime": 22.0828, | |
| "eval_samples_per_second": 202.058, | |
| "eval_steps_per_second": 25.269, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.0752068187515668, | |
| "grad_norm": 20.297489166259766, | |
| "learning_rate": 4.663267885733547e-05, | |
| "loss": 3.9704, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.08356313194618534, | |
| "grad_norm": 17.280521392822266, | |
| "learning_rate": 4.621134237802309e-05, | |
| "loss": 3.9118, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.08356313194618534, | |
| "eval_loss": 3.891711950302124, | |
| "eval_runtime": 22.1454, | |
| "eval_samples_per_second": 201.487, | |
| "eval_steps_per_second": 25.197, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.09191944514080387, | |
| "grad_norm": 18.847597122192383, | |
| "learning_rate": 4.579000589871071e-05, | |
| "loss": 3.9024, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.10027575833542242, | |
| "grad_norm": 14.654472351074219, | |
| "learning_rate": 4.5368669419398335e-05, | |
| "loss": 3.8479, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.10027575833542242, | |
| "eval_loss": 3.8378305435180664, | |
| "eval_runtime": 22.1051, | |
| "eval_samples_per_second": 201.854, | |
| "eval_steps_per_second": 25.243, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.10863207153004095, | |
| "grad_norm": 18.267274856567383, | |
| "learning_rate": 4.494817561304458e-05, | |
| "loss": 3.8116, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.11698838472465949, | |
| "grad_norm": 13.313653945922852, | |
| "learning_rate": 4.45268391337322e-05, | |
| "loss": 3.8592, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.11698838472465949, | |
| "eval_loss": 3.8072171211242676, | |
| "eval_runtime": 22.2351, | |
| "eval_samples_per_second": 200.674, | |
| "eval_steps_per_second": 25.095, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.12534469791927802, | |
| "grad_norm": 16.960010528564453, | |
| "learning_rate": 4.410550265441982e-05, | |
| "loss": 3.8291, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.13370101111389654, | |
| "grad_norm": 15.129623413085938, | |
| "learning_rate": 4.3684166175107444e-05, | |
| "loss": 3.7697, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.13370101111389654, | |
| "eval_loss": 3.777130126953125, | |
| "eval_runtime": 22.2765, | |
| "eval_samples_per_second": 200.3, | |
| "eval_steps_per_second": 25.049, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.1420573243085151, | |
| "grad_norm": 18.825756072998047, | |
| "learning_rate": 4.326535771467094e-05, | |
| "loss": 3.7714, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.1504136375031336, | |
| "grad_norm": 13.93099308013916, | |
| "learning_rate": 4.284402123535856e-05, | |
| "loss": 3.7719, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.1504136375031336, | |
| "eval_loss": 3.747012138366699, | |
| "eval_runtime": 22.2951, | |
| "eval_samples_per_second": 200.134, | |
| "eval_steps_per_second": 25.028, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.15876995069775215, | |
| "grad_norm": 12.885889053344727, | |
| "learning_rate": 4.2422684756046185e-05, | |
| "loss": 3.7127, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.16712626389237067, | |
| "grad_norm": 15.362942695617676, | |
| "learning_rate": 4.200134827673381e-05, | |
| "loss": 3.7135, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.16712626389237067, | |
| "eval_loss": 3.7190206050872803, | |
| "eval_runtime": 22.2344, | |
| "eval_samples_per_second": 200.68, | |
| "eval_steps_per_second": 25.096, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.17548257708698922, | |
| "grad_norm": 12.432964324951172, | |
| "learning_rate": 4.158001179742142e-05, | |
| "loss": 3.6715, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.18383889028160774, | |
| "grad_norm": 25.97146987915039, | |
| "learning_rate": 4.115867531810904e-05, | |
| "loss": 3.6599, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.18383889028160774, | |
| "eval_loss": 3.6818652153015137, | |
| "eval_runtime": 22.2972, | |
| "eval_samples_per_second": 200.115, | |
| "eval_steps_per_second": 25.026, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.1921952034762263, | |
| "grad_norm": 21.378082275390625, | |
| "learning_rate": 4.073733883879666e-05, | |
| "loss": 3.6754, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.20055151667084484, | |
| "grad_norm": 11.59192943572998, | |
| "learning_rate": 4.031600235948428e-05, | |
| "loss": 3.669, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.20055151667084484, | |
| "eval_loss": 3.661188840866089, | |
| "eval_runtime": 22.2553, | |
| "eval_samples_per_second": 200.492, | |
| "eval_steps_per_second": 25.073, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.20890782986546336, | |
| "grad_norm": 17.75707244873047, | |
| "learning_rate": 3.9894665880171905e-05, | |
| "loss": 3.6794, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.2172641430600819, | |
| "grad_norm": 10.825678825378418, | |
| "learning_rate": 3.947332940085953e-05, | |
| "loss": 3.6113, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.2172641430600819, | |
| "eval_loss": 3.6508119106292725, | |
| "eval_runtime": 22.2982, | |
| "eval_samples_per_second": 200.106, | |
| "eval_steps_per_second": 25.024, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.22562045625470042, | |
| "grad_norm": 10.04261302947998, | |
| "learning_rate": 3.905199292154715e-05, | |
| "loss": 3.5963, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.23397676944931897, | |
| "grad_norm": 13.926618576049805, | |
| "learning_rate": 3.863065644223477e-05, | |
| "loss": 3.5997, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.23397676944931897, | |
| "eval_loss": 3.6223905086517334, | |
| "eval_runtime": 22.2717, | |
| "eval_samples_per_second": 200.344, | |
| "eval_steps_per_second": 25.054, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.2423330826439375, | |
| "grad_norm": 11.00304889678955, | |
| "learning_rate": 3.820931996292239e-05, | |
| "loss": 3.5991, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.25068939583855604, | |
| "grad_norm": 16.099769592285156, | |
| "learning_rate": 3.778798348361001e-05, | |
| "loss": 3.6042, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.25068939583855604, | |
| "eval_loss": 3.5953731536865234, | |
| "eval_runtime": 22.2814, | |
| "eval_samples_per_second": 200.257, | |
| "eval_steps_per_second": 25.043, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.2590457090331746, | |
| "grad_norm": 12.459487915039062, | |
| "learning_rate": 3.736664700429763e-05, | |
| "loss": 3.5871, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.2674020222277931, | |
| "grad_norm": 16.979909896850586, | |
| "learning_rate": 3.6946153197943875e-05, | |
| "loss": 3.5238, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.2674020222277931, | |
| "eval_loss": 3.590113401412964, | |
| "eval_runtime": 22.2293, | |
| "eval_samples_per_second": 200.726, | |
| "eval_steps_per_second": 25.102, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.2757583354224116, | |
| "grad_norm": 23.20758056640625, | |
| "learning_rate": 3.65248167186315e-05, | |
| "loss": 3.5646, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.2841146486170302, | |
| "grad_norm": 18.35931396484375, | |
| "learning_rate": 3.610348023931912e-05, | |
| "loss": 3.5445, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.2841146486170302, | |
| "eval_loss": 3.563676595687866, | |
| "eval_runtime": 22.2571, | |
| "eval_samples_per_second": 200.475, | |
| "eval_steps_per_second": 25.071, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.2924709618116487, | |
| "grad_norm": 17.187950134277344, | |
| "learning_rate": 3.568214376000674e-05, | |
| "loss": 3.494, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.3008272750062672, | |
| "grad_norm": 15.331987380981445, | |
| "learning_rate": 3.5261649953652984e-05, | |
| "loss": 3.4913, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.3008272750062672, | |
| "eval_loss": 3.541306495666504, | |
| "eval_runtime": 22.2598, | |
| "eval_samples_per_second": 200.451, | |
| "eval_steps_per_second": 25.068, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.30918358820088576, | |
| "grad_norm": 16.340852737426758, | |
| "learning_rate": 3.484031347434061e-05, | |
| "loss": 3.4969, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.3175399013955043, | |
| "grad_norm": 12.265207290649414, | |
| "learning_rate": 3.441897699502823e-05, | |
| "loss": 3.4934, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.3175399013955043, | |
| "eval_loss": 3.520357847213745, | |
| "eval_runtime": 22.2447, | |
| "eval_samples_per_second": 200.587, | |
| "eval_steps_per_second": 25.085, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.32589621459012286, | |
| "grad_norm": 15.456232070922852, | |
| "learning_rate": 3.399764051571585e-05, | |
| "loss": 3.5013, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.33425252778474135, | |
| "grad_norm": 15.721699714660645, | |
| "learning_rate": 3.3576304036403474e-05, | |
| "loss": 3.4627, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.33425252778474135, | |
| "eval_loss": 3.5179378986358643, | |
| "eval_runtime": 22.2594, | |
| "eval_samples_per_second": 200.455, | |
| "eval_steps_per_second": 25.068, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.3426088409793599, | |
| "grad_norm": 12.118553161621094, | |
| "learning_rate": 3.3154967557091096e-05, | |
| "loss": 3.5006, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.35096515417397844, | |
| "grad_norm": 8.990864753723145, | |
| "learning_rate": 3.273447375073734e-05, | |
| "loss": 3.4367, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.35096515417397844, | |
| "eval_loss": 3.5118658542633057, | |
| "eval_runtime": 22.2415, | |
| "eval_samples_per_second": 200.616, | |
| "eval_steps_per_second": 25.088, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.359321467368597, | |
| "grad_norm": 9.9972562789917, | |
| "learning_rate": 3.231313727142496e-05, | |
| "loss": 3.4498, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.3676777805632155, | |
| "grad_norm": 10.996673583984375, | |
| "learning_rate": 3.189180079211258e-05, | |
| "loss": 3.4643, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.3676777805632155, | |
| "eval_loss": 3.483738899230957, | |
| "eval_runtime": 22.2582, | |
| "eval_samples_per_second": 200.465, | |
| "eval_steps_per_second": 25.069, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.37603409375783403, | |
| "grad_norm": 14.55636978149414, | |
| "learning_rate": 3.14704643128002e-05, | |
| "loss": 3.5215, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.3843904069524526, | |
| "grad_norm": 13.585105895996094, | |
| "learning_rate": 3.104912783348782e-05, | |
| "loss": 3.419, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.3843904069524526, | |
| "eval_loss": 3.47660231590271, | |
| "eval_runtime": 22.226, | |
| "eval_samples_per_second": 200.756, | |
| "eval_steps_per_second": 25.106, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.3927467201470711, | |
| "grad_norm": 11.853238105773926, | |
| "learning_rate": 3.062779135417544e-05, | |
| "loss": 3.4438, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.4011030333416897, | |
| "grad_norm": 13.06174373626709, | |
| "learning_rate": 3.020729754782169e-05, | |
| "loss": 3.4029, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.4011030333416897, | |
| "eval_loss": 3.4587268829345703, | |
| "eval_runtime": 22.2726, | |
| "eval_samples_per_second": 200.336, | |
| "eval_steps_per_second": 25.053, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.40945934653630817, | |
| "grad_norm": 16.874757766723633, | |
| "learning_rate": 2.9786803741467938e-05, | |
| "loss": 3.3971, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.4178156597309267, | |
| "grad_norm": 11.108474731445312, | |
| "learning_rate": 2.936546726215556e-05, | |
| "loss": 3.3574, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.4178156597309267, | |
| "eval_loss": 3.446179151535034, | |
| "eval_runtime": 22.3522, | |
| "eval_samples_per_second": 199.622, | |
| "eval_steps_per_second": 24.964, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.42617197292554526, | |
| "grad_norm": 12.936110496520996, | |
| "learning_rate": 2.8944130782843183e-05, | |
| "loss": 3.3829, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.4345282861201638, | |
| "grad_norm": 12.90854549407959, | |
| "learning_rate": 2.8522794303530802e-05, | |
| "loss": 3.4156, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.4345282861201638, | |
| "eval_loss": 3.44026517868042, | |
| "eval_runtime": 22.2661, | |
| "eval_samples_per_second": 200.394, | |
| "eval_steps_per_second": 25.061, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.4428845993147823, | |
| "grad_norm": 10.326555252075195, | |
| "learning_rate": 2.8101457824218424e-05, | |
| "loss": 3.3607, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.45124091250940085, | |
| "grad_norm": 12.372066497802734, | |
| "learning_rate": 2.7681806690823293e-05, | |
| "loss": 3.3836, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.45124091250940085, | |
| "eval_loss": 3.4253649711608887, | |
| "eval_runtime": 22.2507, | |
| "eval_samples_per_second": 200.533, | |
| "eval_steps_per_second": 25.078, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.4595972257040194, | |
| "grad_norm": 9.778299331665039, | |
| "learning_rate": 2.7260470211510912e-05, | |
| "loss": 3.3671, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.46795353889863794, | |
| "grad_norm": 20.047178268432617, | |
| "learning_rate": 2.6839133732198535e-05, | |
| "loss": 3.3395, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.46795353889863794, | |
| "eval_loss": 3.41679048538208, | |
| "eval_runtime": 22.2707, | |
| "eval_samples_per_second": 200.353, | |
| "eval_steps_per_second": 25.055, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.47630985209325644, | |
| "grad_norm": 9.312335968017578, | |
| "learning_rate": 2.6417797252886157e-05, | |
| "loss": 3.3616, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.484666165287875, | |
| "grad_norm": 10.994682312011719, | |
| "learning_rate": 2.5996460773573776e-05, | |
| "loss": 3.3719, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.484666165287875, | |
| "eval_loss": 3.4018924236297607, | |
| "eval_runtime": 22.2565, | |
| "eval_samples_per_second": 200.481, | |
| "eval_steps_per_second": 25.071, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.49302247848249353, | |
| "grad_norm": 13.464505195617676, | |
| "learning_rate": 2.5575124294261398e-05, | |
| "loss": 3.3312, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.5013787916771121, | |
| "grad_norm": 12.18619441986084, | |
| "learning_rate": 2.515378781494902e-05, | |
| "loss": 3.386, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.5013787916771121, | |
| "eval_loss": 3.3899354934692383, | |
| "eval_runtime": 22.2658, | |
| "eval_samples_per_second": 200.397, | |
| "eval_steps_per_second": 25.061, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.5097351048717306, | |
| "grad_norm": 14.552848815917969, | |
| "learning_rate": 2.4732451335636643e-05, | |
| "loss": 3.3377, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.5180914180663492, | |
| "grad_norm": 15.032088279724121, | |
| "learning_rate": 2.4311114856324262e-05, | |
| "loss": 3.3131, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.5180914180663492, | |
| "eval_loss": 3.378127336502075, | |
| "eval_runtime": 22.2423, | |
| "eval_samples_per_second": 200.609, | |
| "eval_steps_per_second": 25.087, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.5264477312609677, | |
| "grad_norm": 14.666757583618164, | |
| "learning_rate": 2.388977837701188e-05, | |
| "loss": 3.3457, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.5348040444555862, | |
| "grad_norm": 11.800482749938965, | |
| "learning_rate": 2.3468441897699503e-05, | |
| "loss": 3.3192, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.5348040444555862, | |
| "eval_loss": 3.3670458793640137, | |
| "eval_runtime": 22.256, | |
| "eval_samples_per_second": 200.485, | |
| "eval_steps_per_second": 25.072, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.5431603576502048, | |
| "grad_norm": 10.835103034973145, | |
| "learning_rate": 2.3047105418387125e-05, | |
| "loss": 3.3235, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.5515166708448233, | |
| "grad_norm": 12.06092357635498, | |
| "learning_rate": 2.2625768939074744e-05, | |
| "loss": 3.2969, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.5515166708448233, | |
| "eval_loss": 3.356658935546875, | |
| "eval_runtime": 22.2404, | |
| "eval_samples_per_second": 200.626, | |
| "eval_steps_per_second": 25.089, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.5598729840394417, | |
| "grad_norm": 15.398877143859863, | |
| "learning_rate": 2.2204432459762367e-05, | |
| "loss": 3.3181, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 0.5682292972340603, | |
| "grad_norm": 10.425477027893066, | |
| "learning_rate": 2.178309598044999e-05, | |
| "loss": 3.3202, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.5682292972340603, | |
| "eval_loss": 3.34324312210083, | |
| "eval_runtime": 22.2237, | |
| "eval_samples_per_second": 200.777, | |
| "eval_steps_per_second": 25.108, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.5765856104286788, | |
| "grad_norm": 13.118115425109863, | |
| "learning_rate": 2.136175950113761e-05, | |
| "loss": 3.3028, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 0.5849419236232974, | |
| "grad_norm": 8.235157012939453, | |
| "learning_rate": 2.0941265694783854e-05, | |
| "loss": 3.2403, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.5849419236232974, | |
| "eval_loss": 3.3430681228637695, | |
| "eval_runtime": 22.2974, | |
| "eval_samples_per_second": 200.113, | |
| "eval_steps_per_second": 25.025, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.5932982368179159, | |
| "grad_norm": 15.389208793640137, | |
| "learning_rate": 2.0519929215471476e-05, | |
| "loss": 3.3105, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 0.6016545500125344, | |
| "grad_norm": 12.708732604980469, | |
| "learning_rate": 2.0098592736159098e-05, | |
| "loss": 3.2775, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.6016545500125344, | |
| "eval_loss": 3.3276991844177246, | |
| "eval_runtime": 22.2643, | |
| "eval_samples_per_second": 200.411, | |
| "eval_steps_per_second": 25.063, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.610010863207153, | |
| "grad_norm": 13.642451286315918, | |
| "learning_rate": 1.9677256256846717e-05, | |
| "loss": 3.2902, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 0.6183671764017715, | |
| "grad_norm": 12.606600761413574, | |
| "learning_rate": 1.9256762450492966e-05, | |
| "loss": 3.271, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.6183671764017715, | |
| "eval_loss": 3.3122496604919434, | |
| "eval_runtime": 22.2804, | |
| "eval_samples_per_second": 200.266, | |
| "eval_steps_per_second": 25.044, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.6267234895963901, | |
| "grad_norm": 11.484159469604492, | |
| "learning_rate": 1.8835425971180585e-05, | |
| "loss": 3.2833, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 0.6350798027910086, | |
| "grad_norm": 12.317131996154785, | |
| "learning_rate": 1.8414089491868204e-05, | |
| "loss": 3.2848, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.6350798027910086, | |
| "eval_loss": 3.3035213947296143, | |
| "eval_runtime": 22.2937, | |
| "eval_samples_per_second": 200.147, | |
| "eval_steps_per_second": 25.03, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.6434361159856271, | |
| "grad_norm": 11.45077896118164, | |
| "learning_rate": 1.7992753012555827e-05, | |
| "loss": 3.202, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 0.6517924291802457, | |
| "grad_norm": 12.859657287597656, | |
| "learning_rate": 1.7572259206202076e-05, | |
| "loss": 3.2376, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.6517924291802457, | |
| "eval_loss": 3.2956559658050537, | |
| "eval_runtime": 22.3804, | |
| "eval_samples_per_second": 199.371, | |
| "eval_steps_per_second": 24.933, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.6601487423748642, | |
| "grad_norm": 14.472012519836426, | |
| "learning_rate": 1.7150922726889695e-05, | |
| "loss": 3.1924, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 0.6685050555694827, | |
| "grad_norm": 13.051079750061035, | |
| "learning_rate": 1.673042892053594e-05, | |
| "loss": 3.2598, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.6685050555694827, | |
| "eval_loss": 3.2878499031066895, | |
| "eval_runtime": 22.2464, | |
| "eval_samples_per_second": 200.572, | |
| "eval_steps_per_second": 25.083, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.6768613687641013, | |
| "grad_norm": 15.44560718536377, | |
| "learning_rate": 1.6309092441223563e-05, | |
| "loss": 3.1978, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 0.6852176819587198, | |
| "grad_norm": 16.988996505737305, | |
| "learning_rate": 1.588775596191118e-05, | |
| "loss": 3.2247, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.6852176819587198, | |
| "eval_loss": 3.279550313949585, | |
| "eval_runtime": 22.2386, | |
| "eval_samples_per_second": 200.642, | |
| "eval_steps_per_second": 25.091, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.6935739951533384, | |
| "grad_norm": 8.293917655944824, | |
| "learning_rate": 1.5466419482598804e-05, | |
| "loss": 3.1682, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 0.7019303083479569, | |
| "grad_norm": 10.755880355834961, | |
| "learning_rate": 1.5045925676245051e-05, | |
| "loss": 3.1849, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.7019303083479569, | |
| "eval_loss": 3.2791192531585693, | |
| "eval_runtime": 22.2554, | |
| "eval_samples_per_second": 200.491, | |
| "eval_steps_per_second": 25.073, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.7102866215425754, | |
| "grad_norm": 17.822643280029297, | |
| "learning_rate": 1.462458919693267e-05, | |
| "loss": 3.1714, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 0.718642934737194, | |
| "grad_norm": 18.230485916137695, | |
| "learning_rate": 1.4203252717620291e-05, | |
| "loss": 3.2112, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.718642934737194, | |
| "eval_loss": 3.260193109512329, | |
| "eval_runtime": 22.2518, | |
| "eval_samples_per_second": 200.523, | |
| "eval_steps_per_second": 25.077, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.7269992479318125, | |
| "grad_norm": 13.363430976867676, | |
| "learning_rate": 1.3781916238307913e-05, | |
| "loss": 3.1655, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 0.735355561126431, | |
| "grad_norm": 11.570181846618652, | |
| "learning_rate": 1.3360579758995534e-05, | |
| "loss": 3.174, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.735355561126431, | |
| "eval_loss": 3.2490386962890625, | |
| "eval_runtime": 22.2683, | |
| "eval_samples_per_second": 200.374, | |
| "eval_steps_per_second": 25.058, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.7437118743210496, | |
| "grad_norm": 19.80602264404297, | |
| "learning_rate": 1.2939243279683155e-05, | |
| "loss": 3.1987, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 0.7520681875156681, | |
| "grad_norm": 10.821731567382812, | |
| "learning_rate": 1.2518749473329402e-05, | |
| "loss": 3.1799, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.7520681875156681, | |
| "eval_loss": 3.240847587585449, | |
| "eval_runtime": 22.2794, | |
| "eval_samples_per_second": 200.275, | |
| "eval_steps_per_second": 25.046, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.7604245007102867, | |
| "grad_norm": 16.301612854003906, | |
| "learning_rate": 1.2097412994017023e-05, | |
| "loss": 3.2029, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 0.7687808139049052, | |
| "grad_norm": 14.699359893798828, | |
| "learning_rate": 1.1676076514704643e-05, | |
| "loss": 3.1752, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.7687808139049052, | |
| "eval_loss": 3.233914852142334, | |
| "eval_runtime": 22.269, | |
| "eval_samples_per_second": 200.369, | |
| "eval_steps_per_second": 25.057, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.7771371270995237, | |
| "grad_norm": 15.696563720703125, | |
| "learning_rate": 1.1254740035392266e-05, | |
| "loss": 3.132, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 0.7854934402941423, | |
| "grad_norm": 13.062487602233887, | |
| "learning_rate": 1.0833403556079886e-05, | |
| "loss": 3.131, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.7854934402941423, | |
| "eval_loss": 3.2280752658843994, | |
| "eval_runtime": 22.2955, | |
| "eval_samples_per_second": 200.13, | |
| "eval_steps_per_second": 25.027, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.7938497534887607, | |
| "grad_norm": 18.67305564880371, | |
| "learning_rate": 1.0412909749726132e-05, | |
| "loss": 3.1571, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 0.8022060666833793, | |
| "grad_norm": 10.377827644348145, | |
| "learning_rate": 9.992415943372378e-06, | |
| "loss": 3.181, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.8022060666833793, | |
| "eval_loss": 3.2206084728240967, | |
| "eval_runtime": 22.2943, | |
| "eval_samples_per_second": 200.141, | |
| "eval_steps_per_second": 25.029, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.8105623798779978, | |
| "grad_norm": 12.836233139038086, | |
| "learning_rate": 9.571922137018624e-06, | |
| "loss": 3.139, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 0.8189186930726163, | |
| "grad_norm": 11.736408233642578, | |
| "learning_rate": 9.150585657706244e-06, | |
| "loss": 3.0932, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 0.8189186930726163, | |
| "eval_loss": 3.2135069370269775, | |
| "eval_runtime": 22.2506, | |
| "eval_samples_per_second": 200.534, | |
| "eval_steps_per_second": 25.078, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 0.8272750062672349, | |
| "grad_norm": 16.016298294067383, | |
| "learning_rate": 8.729249178393865e-06, | |
| "loss": 3.1634, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 0.8356313194618534, | |
| "grad_norm": 10.488819122314453, | |
| "learning_rate": 8.307912699081487e-06, | |
| "loss": 3.1376, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.8356313194618534, | |
| "eval_loss": 3.2051162719726562, | |
| "eval_runtime": 22.294, | |
| "eval_samples_per_second": 200.144, | |
| "eval_steps_per_second": 25.029, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.8439876326564719, | |
| "grad_norm": 16.168071746826172, | |
| "learning_rate": 7.886576219769108e-06, | |
| "loss": 3.1121, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 0.8523439458510905, | |
| "grad_norm": 19.903099060058594, | |
| "learning_rate": 7.465239740456729e-06, | |
| "loss": 3.1084, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 0.8523439458510905, | |
| "eval_loss": 3.198310375213623, | |
| "eval_runtime": 22.3049, | |
| "eval_samples_per_second": 200.046, | |
| "eval_steps_per_second": 25.017, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 0.860700259045709, | |
| "grad_norm": 12.082676887512207, | |
| "learning_rate": 7.043903261144351e-06, | |
| "loss": 3.0957, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 0.8690565722403276, | |
| "grad_norm": 11.764552116394043, | |
| "learning_rate": 6.622566781831971e-06, | |
| "loss": 3.099, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 0.8690565722403276, | |
| "eval_loss": 3.193253993988037, | |
| "eval_runtime": 22.2469, | |
| "eval_samples_per_second": 200.567, | |
| "eval_steps_per_second": 25.082, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 0.8774128854349461, | |
| "grad_norm": 12.482972145080566, | |
| "learning_rate": 6.201230302519592e-06, | |
| "loss": 3.0779, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 0.8857691986295646, | |
| "grad_norm": 14.11436939239502, | |
| "learning_rate": 5.7798938232072135e-06, | |
| "loss": 3.1278, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 0.8857691986295646, | |
| "eval_loss": 3.1867904663085938, | |
| "eval_runtime": 22.2664, | |
| "eval_samples_per_second": 200.392, | |
| "eval_steps_per_second": 25.06, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 0.8941255118241832, | |
| "grad_norm": 19.69700813293457, | |
| "learning_rate": 5.358557343894835e-06, | |
| "loss": 3.0968, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 0.9024818250188017, | |
| "grad_norm": 14.537339210510254, | |
| "learning_rate": 4.937220864582456e-06, | |
| "loss": 3.1436, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 0.9024818250188017, | |
| "eval_loss": 3.180774688720703, | |
| "eval_runtime": 22.293, | |
| "eval_samples_per_second": 200.152, | |
| "eval_steps_per_second": 25.03, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 0.9108381382134202, | |
| "grad_norm": 16.117996215820312, | |
| "learning_rate": 4.515884385270077e-06, | |
| "loss": 3.1288, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 0.9191944514080388, | |
| "grad_norm": 12.458276748657227, | |
| "learning_rate": 4.094547905957698e-06, | |
| "loss": 3.0763, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.9191944514080388, | |
| "eval_loss": 3.175370216369629, | |
| "eval_runtime": 22.253, | |
| "eval_samples_per_second": 200.513, | |
| "eval_steps_per_second": 25.075, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.9275507646026573, | |
| "grad_norm": 14.115385055541992, | |
| "learning_rate": 3.6732114266453192e-06, | |
| "loss": 3.0642, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 0.9359070777972759, | |
| "grad_norm": 19.65464210510254, | |
| "learning_rate": 3.2518749473329403e-06, | |
| "loss": 3.1248, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 0.9359070777972759, | |
| "eval_loss": 3.1690962314605713, | |
| "eval_runtime": 22.3258, | |
| "eval_samples_per_second": 199.858, | |
| "eval_steps_per_second": 24.993, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 0.9442633909918944, | |
| "grad_norm": 11.953753471374512, | |
| "learning_rate": 2.831381140979186e-06, | |
| "loss": 3.1361, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 0.9526197041865129, | |
| "grad_norm": 10.821110725402832, | |
| "learning_rate": 2.4108873346254323e-06, | |
| "loss": 3.0418, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 0.9526197041865129, | |
| "eval_loss": 3.164776563644409, | |
| "eval_runtime": 22.2464, | |
| "eval_samples_per_second": 200.572, | |
| "eval_steps_per_second": 25.083, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 0.9609760173811315, | |
| "grad_norm": 11.476717948913574, | |
| "learning_rate": 1.9895508553130533e-06, | |
| "loss": 3.0504, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 0.96933233057575, | |
| "grad_norm": 10.973363876342773, | |
| "learning_rate": 1.5682143760006742e-06, | |
| "loss": 3.0755, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 0.96933233057575, | |
| "eval_loss": 3.1603705883026123, | |
| "eval_runtime": 22.272, | |
| "eval_samples_per_second": 200.341, | |
| "eval_steps_per_second": 25.054, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 0.9776886437703685, | |
| "grad_norm": 10.836787223815918, | |
| "learning_rate": 1.1468778966882954e-06, | |
| "loss": 3.1001, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 0.9860449569649871, | |
| "grad_norm": 13.901703834533691, | |
| "learning_rate": 7.255414173759165e-07, | |
| "loss": 3.0633, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 0.9860449569649871, | |
| "eval_loss": 3.1587648391723633, | |
| "eval_runtime": 22.3994, | |
| "eval_samples_per_second": 199.202, | |
| "eval_steps_per_second": 24.911, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 0.9944012701596056, | |
| "grad_norm": 19.022567749023438, | |
| "learning_rate": 3.0420493806353753e-07, | |
| "loss": 3.0751, | |
| "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": 2.943723113325527e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |