| { | |
| "best_metric": 18.436363636363637, | |
| "best_model_checkpoint": "galgame-whisper\\checkpoint-24500", | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 26624, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01220703125, | |
| "grad_norm": 11.14741039276123, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 1.7805, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0244140625, | |
| "grad_norm": 6.63244104385376, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 1.3099, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03662109375, | |
| "grad_norm": 4.578396320343018, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9171, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.048828125, | |
| "grad_norm": 5.572494029998779, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6639, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.048828125, | |
| "eval_cer": 25.6, | |
| "eval_loss": 0.593059778213501, | |
| "eval_normalized_cer": 20.427455815865187, | |
| "eval_runtime": 123.1903, | |
| "eval_samples_per_second": 1.039, | |
| "eval_steps_per_second": 0.13, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06103515625, | |
| "grad_norm": 5.427572250366211, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6159, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.0732421875, | |
| "grad_norm": 4.667797565460205, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5685, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.08544921875, | |
| "grad_norm": 4.875271797180176, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5758, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.09765625, | |
| "grad_norm": 4.109724044799805, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5772, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09765625, | |
| "eval_cer": 25.418181818181818, | |
| "eval_loss": 0.567274808883667, | |
| "eval_normalized_cer": 20.34525277435265, | |
| "eval_runtime": 140.399, | |
| "eval_samples_per_second": 0.912, | |
| "eval_steps_per_second": 0.114, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.10986328125, | |
| "grad_norm": 4.04423189163208, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5723, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1220703125, | |
| "grad_norm": 4.566165447235107, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6474, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.13427734375, | |
| "grad_norm": 4.996761798858643, | |
| "learning_rate": 1e-05, | |
| "loss": 0.611, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.146484375, | |
| "grad_norm": 4.392452716827393, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5462, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.146484375, | |
| "eval_cer": 22.21818181818182, | |
| "eval_loss": 0.5370786190032959, | |
| "eval_normalized_cer": 17.673653925195232, | |
| "eval_runtime": 134.3037, | |
| "eval_samples_per_second": 0.953, | |
| "eval_steps_per_second": 0.119, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.15869140625, | |
| "grad_norm": 3.719304084777832, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5986, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.1708984375, | |
| "grad_norm": 11.704819679260254, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6941, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.18310546875, | |
| "grad_norm": 5.188851833343506, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5486, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.1953125, | |
| "grad_norm": 3.8557848930358887, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6085, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1953125, | |
| "eval_cer": 21.854545454545455, | |
| "eval_loss": 0.5271878838539124, | |
| "eval_normalized_cer": 17.221537196876284, | |
| "eval_runtime": 131.3514, | |
| "eval_samples_per_second": 0.974, | |
| "eval_steps_per_second": 0.122, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.20751953125, | |
| "grad_norm": 3.7432358264923096, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6387, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.2197265625, | |
| "grad_norm": 3.6960973739624023, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5481, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.23193359375, | |
| "grad_norm": 5.312159061431885, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5565, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.244140625, | |
| "grad_norm": 5.376567363739014, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5262, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.244140625, | |
| "eval_cer": 24.509090909090908, | |
| "eval_loss": 0.5245115756988525, | |
| "eval_normalized_cer": 19.482120838471022, | |
| "eval_runtime": 136.6132, | |
| "eval_samples_per_second": 0.937, | |
| "eval_steps_per_second": 0.117, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.25634765625, | |
| "grad_norm": 3.251194715499878, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5751, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.2685546875, | |
| "grad_norm": 3.575839042663574, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5391, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.28076171875, | |
| "grad_norm": 4.032149314880371, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5314, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.29296875, | |
| "grad_norm": 4.987039566040039, | |
| "learning_rate": 1e-05, | |
| "loss": 0.576, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.29296875, | |
| "eval_cer": 22.69090909090909, | |
| "eval_loss": 0.5259984135627747, | |
| "eval_normalized_cer": 18.207973695026716, | |
| "eval_runtime": 127.1247, | |
| "eval_samples_per_second": 1.007, | |
| "eval_steps_per_second": 0.126, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.30517578125, | |
| "grad_norm": 4.042311191558838, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5671, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.3173828125, | |
| "grad_norm": 5.254710674285889, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5525, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.32958984375, | |
| "grad_norm": 3.7549564838409424, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5489, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.341796875, | |
| "grad_norm": 4.815126419067383, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5844, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.341796875, | |
| "eval_cer": 26.581818181818186, | |
| "eval_loss": 0.5143883228302002, | |
| "eval_normalized_cer": 21.98931360460337, | |
| "eval_runtime": 125.4245, | |
| "eval_samples_per_second": 1.021, | |
| "eval_steps_per_second": 0.128, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.35400390625, | |
| "grad_norm": 3.240372657775879, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5684, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.3662109375, | |
| "grad_norm": 4.83163595199585, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5157, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.37841796875, | |
| "grad_norm": 3.209136724472046, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5348, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.390625, | |
| "grad_norm": 3.9407966136932373, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5689, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.390625, | |
| "eval_cer": 23.200000000000003, | |
| "eval_loss": 0.5060851573944092, | |
| "eval_normalized_cer": 18.372379778051787, | |
| "eval_runtime": 132.0722, | |
| "eval_samples_per_second": 0.969, | |
| "eval_steps_per_second": 0.121, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.40283203125, | |
| "grad_norm": 5.201825141906738, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5981, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.4150390625, | |
| "grad_norm": 4.420695781707764, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6028, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.42724609375, | |
| "grad_norm": 4.9300856590271, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5807, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.439453125, | |
| "grad_norm": 4.098893165588379, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5731, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.439453125, | |
| "eval_cer": 22.0, | |
| "eval_loss": 0.4978080987930298, | |
| "eval_normalized_cer": 17.7147554459515, | |
| "eval_runtime": 131.4521, | |
| "eval_samples_per_second": 0.974, | |
| "eval_steps_per_second": 0.122, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.45166015625, | |
| "grad_norm": 3.947633981704712, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4836, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.4638671875, | |
| "grad_norm": 3.02842378616333, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5248, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.47607421875, | |
| "grad_norm": 5.237096309661865, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5299, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.48828125, | |
| "grad_norm": 4.260601997375488, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5058, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.48828125, | |
| "eval_cer": 23.599999999999998, | |
| "eval_loss": 0.494682252407074, | |
| "eval_normalized_cer": 18.988902589395806, | |
| "eval_runtime": 132.3816, | |
| "eval_samples_per_second": 0.967, | |
| "eval_steps_per_second": 0.121, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.50048828125, | |
| "grad_norm": 4.105225086212158, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4992, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.5126953125, | |
| "grad_norm": 4.636739730834961, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5517, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.52490234375, | |
| "grad_norm": 4.676263809204102, | |
| "learning_rate": 1e-05, | |
| "loss": 0.511, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.537109375, | |
| "grad_norm": 4.2934465408325195, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5556, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.537109375, | |
| "eval_cer": 23.454545454545457, | |
| "eval_loss": 0.4881322383880615, | |
| "eval_normalized_cer": 19.399917796958487, | |
| "eval_runtime": 148.1276, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.054, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.54931640625, | |
| "grad_norm": 3.7905611991882324, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5309, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.5615234375, | |
| "grad_norm": 3.6019198894500732, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4845, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.57373046875, | |
| "grad_norm": 3.77414870262146, | |
| "learning_rate": 1e-05, | |
| "loss": 0.491, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.5859375, | |
| "grad_norm": 3.13875412940979, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5033, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5859375, | |
| "eval_cer": 23.81818181818182, | |
| "eval_loss": 0.48237502574920654, | |
| "eval_normalized_cer": 19.31771475544595, | |
| "eval_runtime": 167.3993, | |
| "eval_samples_per_second": 0.765, | |
| "eval_steps_per_second": 0.048, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.59814453125, | |
| "grad_norm": 5.719494819641113, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5085, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.6103515625, | |
| "grad_norm": 3.9417619705200195, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4764, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.62255859375, | |
| "grad_norm": 4.024901390075684, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4834, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.634765625, | |
| "grad_norm": 4.798065185546875, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5041, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.634765625, | |
| "eval_cer": 21.672727272727276, | |
| "eval_loss": 0.47094789147377014, | |
| "eval_normalized_cer": 17.632552404438965, | |
| "eval_runtime": 167.9645, | |
| "eval_samples_per_second": 0.762, | |
| "eval_steps_per_second": 0.048, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.64697265625, | |
| "grad_norm": 3.018202781677246, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5181, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.6591796875, | |
| "grad_norm": 4.9327311515808105, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5419, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.67138671875, | |
| "grad_norm": 4.631186485290527, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4995, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.68359375, | |
| "grad_norm": 5.249241828918457, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5324, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.68359375, | |
| "eval_cer": 21.01818181818182, | |
| "eval_loss": 0.4812983274459839, | |
| "eval_normalized_cer": 16.605014385532264, | |
| "eval_runtime": 164.7775, | |
| "eval_samples_per_second": 0.777, | |
| "eval_steps_per_second": 0.049, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.69580078125, | |
| "grad_norm": 3.599472761154175, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5059, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.7080078125, | |
| "grad_norm": 2.960740089416504, | |
| "learning_rate": 1e-05, | |
| "loss": 0.474, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.72021484375, | |
| "grad_norm": 5.463324546813965, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4646, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.732421875, | |
| "grad_norm": 2.3991730213165283, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5374, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.732421875, | |
| "eval_cer": 21.30909090909091, | |
| "eval_loss": 0.4700665771961212, | |
| "eval_normalized_cer": 17.057131113851213, | |
| "eval_runtime": 168.2533, | |
| "eval_samples_per_second": 0.761, | |
| "eval_steps_per_second": 0.048, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.74462890625, | |
| "grad_norm": 3.7515509128570557, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4979, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.7568359375, | |
| "grad_norm": 3.6961734294891357, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5254, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.76904296875, | |
| "grad_norm": 5.233846187591553, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5757, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.78125, | |
| "grad_norm": 3.659038543701172, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4875, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.78125, | |
| "eval_cer": 20.872727272727275, | |
| "eval_loss": 0.4664301574230194, | |
| "eval_normalized_cer": 16.48170982326346, | |
| "eval_runtime": 170.1628, | |
| "eval_samples_per_second": 0.752, | |
| "eval_steps_per_second": 0.047, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.79345703125, | |
| "grad_norm": 3.4504241943359375, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4794, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.8056640625, | |
| "grad_norm": 3.2377774715423584, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5338, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.81787109375, | |
| "grad_norm": 3.6238794326782227, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4747, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.830078125, | |
| "grad_norm": 3.8324549198150635, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5155, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.830078125, | |
| "eval_cer": 22.254545454545454, | |
| "eval_loss": 0.4597916305065155, | |
| "eval_normalized_cer": 18.12577065351418, | |
| "eval_runtime": 166.0402, | |
| "eval_samples_per_second": 0.771, | |
| "eval_steps_per_second": 0.048, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.84228515625, | |
| "grad_norm": 4.575786113739014, | |
| "learning_rate": 1e-05, | |
| "loss": 0.515, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.8544921875, | |
| "grad_norm": 3.636530637741089, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4976, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.86669921875, | |
| "grad_norm": 4.921797752380371, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5211, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.87890625, | |
| "grad_norm": 3.620969533920288, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4824, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.87890625, | |
| "eval_cer": 26.181818181818183, | |
| "eval_loss": 0.46164897084236145, | |
| "eval_normalized_cer": 21.537196876284423, | |
| "eval_runtime": 150.2, | |
| "eval_samples_per_second": 0.852, | |
| "eval_steps_per_second": 0.053, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.89111328125, | |
| "grad_norm": 4.432845592498779, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5138, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.9033203125, | |
| "grad_norm": 3.5764822959899902, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5413, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.91552734375, | |
| "grad_norm": 4.142552852630615, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4824, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.927734375, | |
| "grad_norm": 3.680915594100952, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5295, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.927734375, | |
| "eval_cer": 20.327272727272728, | |
| "eval_loss": 0.45038798451423645, | |
| "eval_normalized_cer": 15.906288532675708, | |
| "eval_runtime": 173.3587, | |
| "eval_samples_per_second": 0.738, | |
| "eval_steps_per_second": 0.046, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.93994140625, | |
| "grad_norm": 2.5720534324645996, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5137, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.9521484375, | |
| "grad_norm": 2.782604455947876, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5086, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.96435546875, | |
| "grad_norm": 4.596216201782227, | |
| "learning_rate": 1e-05, | |
| "loss": 0.525, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.9765625, | |
| "grad_norm": 4.53118371963501, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5004, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9765625, | |
| "eval_cer": 22.145454545454545, | |
| "eval_loss": 0.4644533395767212, | |
| "eval_normalized_cer": 17.26263871763255, | |
| "eval_runtime": 170.2375, | |
| "eval_samples_per_second": 0.752, | |
| "eval_steps_per_second": 0.047, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.98876953125, | |
| "grad_norm": 3.8377697467803955, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5364, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.0009765625, | |
| "grad_norm": 6.334020614624023, | |
| "learning_rate": 1e-05, | |
| "loss": 0.619, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.01318359375, | |
| "grad_norm": 4.159861087799072, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5219, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 1.025390625, | |
| "grad_norm": 3.607752799987793, | |
| "learning_rate": 1e-05, | |
| "loss": 0.529, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.03759765625, | |
| "grad_norm": 5.036830425262451, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5207, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 1.0498046875, | |
| "grad_norm": 3.520869016647339, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5213, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.06201171875, | |
| "grad_norm": 4.612700939178467, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4954, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 1.07421875, | |
| "grad_norm": 4.156740188598633, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5406, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.08642578125, | |
| "grad_norm": 3.5834848880767822, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5097, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 1.0986328125, | |
| "grad_norm": 3.4885971546173096, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4756, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.11083984375, | |
| "grad_norm": 5.202791213989258, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5375, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 1.123046875, | |
| "grad_norm": 2.762514114379883, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4792, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.13525390625, | |
| "grad_norm": 2.9590845108032227, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4671, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 1.1474609375, | |
| "grad_norm": 2.85101318359375, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4856, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.15966796875, | |
| "grad_norm": 3.444801092147827, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5393, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 1.171875, | |
| "grad_norm": 4.466598033905029, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5304, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.18408203125, | |
| "grad_norm": 3.2998430728912354, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5008, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 1.1962890625, | |
| "grad_norm": 3.405848264694214, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4788, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.20849609375, | |
| "grad_norm": 3.5294339656829834, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4954, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.220703125, | |
| "grad_norm": 3.5182113647460938, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5109, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.220703125, | |
| "eval_cer": 22.8, | |
| "eval_loss": 0.449319064617157, | |
| "eval_normalized_cer": 18.495684340320594, | |
| "eval_runtime": 122.2523, | |
| "eval_samples_per_second": 1.047, | |
| "eval_steps_per_second": 0.065, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.23291015625, | |
| "grad_norm": 4.339564800262451, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5178, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 1.2451171875, | |
| "grad_norm": 3.9555203914642334, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4995, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.25732421875, | |
| "grad_norm": 2.8713884353637695, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4892, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 1.26953125, | |
| "grad_norm": 3.2301995754241943, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4916, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.28173828125, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5373, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 1.2939453125, | |
| "grad_norm": 3.3091487884521484, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4517, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.30615234375, | |
| "grad_norm": 5.547422409057617, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5115, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 1.318359375, | |
| "grad_norm": 4.931210517883301, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4886, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.33056640625, | |
| "grad_norm": 4.160281658172607, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4653, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 1.3427734375, | |
| "grad_norm": 3.172577381134033, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5092, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.35498046875, | |
| "grad_norm": 4.536301612854004, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5189, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 1.3671875, | |
| "grad_norm": 4.744750499725342, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4882, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.37939453125, | |
| "grad_norm": 4.030979633331299, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4538, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 1.3916015625, | |
| "grad_norm": 4.93550443649292, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4778, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.40380859375, | |
| "grad_norm": 3.9617207050323486, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4542, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 1.416015625, | |
| "grad_norm": 2.711639642715454, | |
| "learning_rate": 1e-05, | |
| "loss": 0.501, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.42822265625, | |
| "grad_norm": 3.6887452602386475, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4909, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 1.4404296875, | |
| "grad_norm": 4.250792026519775, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4814, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.45263671875, | |
| "grad_norm": 3.804023265838623, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4922, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 1.46484375, | |
| "grad_norm": 4.579716205596924, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4816, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.46484375, | |
| "eval_cer": 21.163636363636364, | |
| "eval_loss": 0.4456084370613098, | |
| "eval_normalized_cer": 16.6872174270448, | |
| "eval_runtime": 143.2684, | |
| "eval_samples_per_second": 0.893, | |
| "eval_steps_per_second": 0.056, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.47705078125, | |
| "grad_norm": 6.303469657897949, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5534, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 1.4892578125, | |
| "grad_norm": 3.4602832794189453, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5466, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 1.50146484375, | |
| "grad_norm": 5.2441205978393555, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4971, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 1.513671875, | |
| "grad_norm": 4.630413055419922, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4965, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.52587890625, | |
| "grad_norm": 4.705837726593018, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4555, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 1.5380859375, | |
| "grad_norm": 5.804379463195801, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5324, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 1.55029296875, | |
| "grad_norm": 4.1124701499938965, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4799, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "grad_norm": 2.9505362510681152, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4615, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.57470703125, | |
| "grad_norm": 4.232094764709473, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5032, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 1.5869140625, | |
| "grad_norm": 3.584272861480713, | |
| "learning_rate": 1e-05, | |
| "loss": 0.429, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 1.59912109375, | |
| "grad_norm": 4.316075801849365, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5184, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 1.611328125, | |
| "grad_norm": 2.490178108215332, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5072, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.62353515625, | |
| "grad_norm": 4.204127311706543, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4874, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 1.6357421875, | |
| "grad_norm": 3.575812339782715, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4741, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 1.64794921875, | |
| "grad_norm": 3.3744544982910156, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4931, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 1.66015625, | |
| "grad_norm": 3.7953217029571533, | |
| "learning_rate": 1e-05, | |
| "loss": 0.434, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.67236328125, | |
| "grad_norm": 3.54185152053833, | |
| "learning_rate": 1e-05, | |
| "loss": 0.449, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 1.6845703125, | |
| "grad_norm": 3.5956945419311523, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4836, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.69677734375, | |
| "grad_norm": 3.323965072631836, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5173, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 1.708984375, | |
| "grad_norm": 3.515545606613159, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4816, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.708984375, | |
| "eval_cer": 22.472727272727273, | |
| "eval_loss": 0.44189023971557617, | |
| "eval_normalized_cer": 18.12577065351418, | |
| "eval_runtime": 125.2883, | |
| "eval_samples_per_second": 1.022, | |
| "eval_steps_per_second": 0.064, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.72119140625, | |
| "grad_norm": 2.902378797531128, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4416, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 1.7333984375, | |
| "grad_norm": 5.24100399017334, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4477, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 1.74560546875, | |
| "grad_norm": 2.840927839279175, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4683, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 1.7578125, | |
| "grad_norm": 3.2699358463287354, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4806, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.77001953125, | |
| "grad_norm": 3.1949098110198975, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4652, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 1.7822265625, | |
| "grad_norm": 2.1296651363372803, | |
| "learning_rate": 1e-05, | |
| "loss": 0.473, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 1.79443359375, | |
| "grad_norm": 2.817379951477051, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5153, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 1.806640625, | |
| "grad_norm": 3.289232015609741, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4662, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.81884765625, | |
| "grad_norm": 5.639113426208496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4773, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 1.8310546875, | |
| "grad_norm": 3.2733285427093506, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4468, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.84326171875, | |
| "grad_norm": 3.131972312927246, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5031, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 1.85546875, | |
| "grad_norm": 3.908076524734497, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5018, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.86767578125, | |
| "grad_norm": 3.57212233543396, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4862, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 1.8798828125, | |
| "grad_norm": 3.8183035850524902, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4953, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 1.89208984375, | |
| "grad_norm": 3.292840003967285, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4726, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 1.904296875, | |
| "grad_norm": 2.816706418991089, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4925, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.91650390625, | |
| "grad_norm": 4.149738311767578, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4868, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 1.9287109375, | |
| "grad_norm": 3.7483160495758057, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4371, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 1.94091796875, | |
| "grad_norm": 3.3911263942718506, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4279, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 1.953125, | |
| "grad_norm": 4.811086654663086, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4482, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.953125, | |
| "eval_cer": 23.054545454545455, | |
| "eval_loss": 0.43437713384628296, | |
| "eval_normalized_cer": 18.66009042334566, | |
| "eval_runtime": 152.761, | |
| "eval_samples_per_second": 0.838, | |
| "eval_steps_per_second": 0.052, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.96533203125, | |
| "grad_norm": 2.9040095806121826, | |
| "learning_rate": 1e-05, | |
| "loss": 0.448, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 1.9775390625, | |
| "grad_norm": 3.5257699489593506, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4533, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 1.98974609375, | |
| "grad_norm": 3.5330731868743896, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4978, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 2.001953125, | |
| "grad_norm": 3.301609992980957, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4408, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 2.01416015625, | |
| "grad_norm": 2.8477797508239746, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4858, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 2.0263671875, | |
| "grad_norm": 3.097527503967285, | |
| "learning_rate": 1e-05, | |
| "loss": 0.455, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 2.03857421875, | |
| "grad_norm": 3.6209845542907715, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4461, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 2.05078125, | |
| "grad_norm": 3.182161331176758, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4472, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.06298828125, | |
| "grad_norm": 4.14016056060791, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4556, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 2.0751953125, | |
| "grad_norm": 3.5136237144470215, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4817, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 2.08740234375, | |
| "grad_norm": 4.494429111480713, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4399, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 2.099609375, | |
| "grad_norm": 4.786192417144775, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4755, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 2.11181640625, | |
| "grad_norm": 4.873692512512207, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4622, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 2.1240234375, | |
| "grad_norm": 3.9967992305755615, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4796, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 2.13623046875, | |
| "grad_norm": 4.429341793060303, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4564, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 2.1484375, | |
| "grad_norm": 3.952096939086914, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4753, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.16064453125, | |
| "grad_norm": 4.080462455749512, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4625, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 2.1728515625, | |
| "grad_norm": 3.9949817657470703, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4935, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 2.18505859375, | |
| "grad_norm": 3.152660846710205, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5341, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 2.197265625, | |
| "grad_norm": 3.485163927078247, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4524, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.197265625, | |
| "eval_cer": 22.98181818181818, | |
| "eval_loss": 0.4256882667541504, | |
| "eval_normalized_cer": 18.29017673653925, | |
| "eval_runtime": 133.4464, | |
| "eval_samples_per_second": 0.959, | |
| "eval_steps_per_second": 0.06, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.20947265625, | |
| "grad_norm": 3.786696195602417, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4827, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 2.2216796875, | |
| "grad_norm": 3.4594826698303223, | |
| "learning_rate": 1e-05, | |
| "loss": 0.505, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 2.23388671875, | |
| "grad_norm": 3.2194244861602783, | |
| "learning_rate": 1e-05, | |
| "loss": 0.465, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 2.24609375, | |
| "grad_norm": 3.4851391315460205, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4484, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 2.25830078125, | |
| "grad_norm": 4.488097667694092, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4357, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 2.2705078125, | |
| "grad_norm": 3.071812868118286, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4411, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 2.28271484375, | |
| "grad_norm": 4.025106430053711, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4436, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 2.294921875, | |
| "grad_norm": 3.1780993938446045, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4645, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 2.30712890625, | |
| "grad_norm": 4.2200446128845215, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5211, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 2.3193359375, | |
| "grad_norm": 3.622480869293213, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4763, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 2.33154296875, | |
| "grad_norm": 3.0763843059539795, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4876, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 2.34375, | |
| "grad_norm": 3.9623701572418213, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4683, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.35595703125, | |
| "grad_norm": 4.627608776092529, | |
| "learning_rate": 1e-05, | |
| "loss": 0.426, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 2.3681640625, | |
| "grad_norm": 4.717302322387695, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4368, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 2.38037109375, | |
| "grad_norm": 4.21370792388916, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5348, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 2.392578125, | |
| "grad_norm": 4.240349292755127, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4766, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 2.40478515625, | |
| "grad_norm": 4.513136386871338, | |
| "learning_rate": 1e-05, | |
| "loss": 0.421, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 2.4169921875, | |
| "grad_norm": 3.5849685668945312, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4599, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 2.42919921875, | |
| "grad_norm": 3.148627519607544, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4812, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 2.44140625, | |
| "grad_norm": 3.3476321697235107, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5048, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.44140625, | |
| "eval_cer": 23.78181818181818, | |
| "eval_loss": 0.42172056436538696, | |
| "eval_normalized_cer": 18.7422934648582, | |
| "eval_runtime": 133.6191, | |
| "eval_samples_per_second": 0.958, | |
| "eval_steps_per_second": 0.06, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.45361328125, | |
| "grad_norm": 3.0917651653289795, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4876, | |
| "step": 5025 | |
| }, | |
| { | |
| "epoch": 2.4658203125, | |
| "grad_norm": 4.9287309646606445, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4631, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 2.47802734375, | |
| "grad_norm": 3.54841947555542, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4752, | |
| "step": 5075 | |
| }, | |
| { | |
| "epoch": 2.490234375, | |
| "grad_norm": 3.062551975250244, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4232, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 2.50244140625, | |
| "grad_norm": 5.846357822418213, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4819, | |
| "step": 5125 | |
| }, | |
| { | |
| "epoch": 2.5146484375, | |
| "grad_norm": 5.39830207824707, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4574, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 2.52685546875, | |
| "grad_norm": 3.695359468460083, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5137, | |
| "step": 5175 | |
| }, | |
| { | |
| "epoch": 2.5390625, | |
| "grad_norm": 3.8433547019958496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4097, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.55126953125, | |
| "grad_norm": 3.887317419052124, | |
| "learning_rate": 1e-05, | |
| "loss": 0.481, | |
| "step": 5225 | |
| }, | |
| { | |
| "epoch": 2.5634765625, | |
| "grad_norm": 3.9261627197265625, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4513, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 2.57568359375, | |
| "grad_norm": 2.480574369430542, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4372, | |
| "step": 5275 | |
| }, | |
| { | |
| "epoch": 2.587890625, | |
| "grad_norm": 4.399667263031006, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4252, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 2.60009765625, | |
| "grad_norm": 3.6141607761383057, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4369, | |
| "step": 5325 | |
| }, | |
| { | |
| "epoch": 2.6123046875, | |
| "grad_norm": 2.9269521236419678, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4987, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 2.62451171875, | |
| "grad_norm": 3.667206048965454, | |
| "learning_rate": 1e-05, | |
| "loss": 0.495, | |
| "step": 5375 | |
| }, | |
| { | |
| "epoch": 2.63671875, | |
| "grad_norm": 5.493015289306641, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4439, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 2.64892578125, | |
| "grad_norm": 3.328899383544922, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4539, | |
| "step": 5425 | |
| }, | |
| { | |
| "epoch": 2.6611328125, | |
| "grad_norm": 2.529545783996582, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4698, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 2.67333984375, | |
| "grad_norm": 3.669677495956421, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3939, | |
| "step": 5475 | |
| }, | |
| { | |
| "epoch": 2.685546875, | |
| "grad_norm": 3.070936441421509, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4721, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.685546875, | |
| "eval_cer": 23.01818181818182, | |
| "eval_loss": 0.4226231276988983, | |
| "eval_normalized_cer": 18.33127825729552, | |
| "eval_runtime": 126.3723, | |
| "eval_samples_per_second": 1.013, | |
| "eval_steps_per_second": 0.063, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.69775390625, | |
| "grad_norm": 4.265174865722656, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5078, | |
| "step": 5525 | |
| }, | |
| { | |
| "epoch": 2.7099609375, | |
| "grad_norm": 5.314130783081055, | |
| "learning_rate": 1e-05, | |
| "loss": 0.452, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 2.72216796875, | |
| "grad_norm": 4.048317909240723, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4419, | |
| "step": 5575 | |
| }, | |
| { | |
| "epoch": 2.734375, | |
| "grad_norm": 3.8373677730560303, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4236, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.74658203125, | |
| "grad_norm": 3.880476713180542, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4224, | |
| "step": 5625 | |
| }, | |
| { | |
| "epoch": 2.7587890625, | |
| "grad_norm": 4.03834867477417, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4762, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 2.77099609375, | |
| "grad_norm": 3.321204900741577, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4823, | |
| "step": 5675 | |
| }, | |
| { | |
| "epoch": 2.783203125, | |
| "grad_norm": 3.013662099838257, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4839, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 2.79541015625, | |
| "grad_norm": 4.0579705238342285, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4754, | |
| "step": 5725 | |
| }, | |
| { | |
| "epoch": 2.8076171875, | |
| "grad_norm": 3.929385185241699, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4123, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 2.81982421875, | |
| "grad_norm": 4.071752071380615, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4989, | |
| "step": 5775 | |
| }, | |
| { | |
| "epoch": 2.83203125, | |
| "grad_norm": 3.5751779079437256, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4507, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 2.84423828125, | |
| "grad_norm": 3.4296460151672363, | |
| "learning_rate": 1e-05, | |
| "loss": 0.522, | |
| "step": 5825 | |
| }, | |
| { | |
| "epoch": 2.8564453125, | |
| "grad_norm": 2.707711696624756, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4347, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 2.86865234375, | |
| "grad_norm": 2.0897769927978516, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4971, | |
| "step": 5875 | |
| }, | |
| { | |
| "epoch": 2.880859375, | |
| "grad_norm": 2.9398937225341797, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4507, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 2.89306640625, | |
| "grad_norm": 3.4962351322174072, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4662, | |
| "step": 5925 | |
| }, | |
| { | |
| "epoch": 2.9052734375, | |
| "grad_norm": 3.520770311355591, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4784, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 2.91748046875, | |
| "grad_norm": 3.4182958602905273, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4952, | |
| "step": 5975 | |
| }, | |
| { | |
| "epoch": 2.9296875, | |
| "grad_norm": 2.9308042526245117, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4321, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.9296875, | |
| "eval_cer": 20.472727272727273, | |
| "eval_loss": 0.4113434851169586, | |
| "eval_normalized_cer": 15.577476366625564, | |
| "eval_runtime": 105.0617, | |
| "eval_samples_per_second": 1.218, | |
| "eval_steps_per_second": 0.076, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.94189453125, | |
| "grad_norm": 2.3675143718719482, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5268, | |
| "step": 6025 | |
| }, | |
| { | |
| "epoch": 2.9541015625, | |
| "grad_norm": 3.0600383281707764, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4166, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 2.96630859375, | |
| "grad_norm": 3.9981579780578613, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4758, | |
| "step": 6075 | |
| }, | |
| { | |
| "epoch": 2.978515625, | |
| "grad_norm": 4.047635555267334, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4376, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 2.99072265625, | |
| "grad_norm": 3.3930447101593018, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4391, | |
| "step": 6125 | |
| }, | |
| { | |
| "epoch": 3.0029296875, | |
| "grad_norm": 2.583280563354492, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3899, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 3.01513671875, | |
| "grad_norm": 5.198780059814453, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4538, | |
| "step": 6175 | |
| }, | |
| { | |
| "epoch": 3.02734375, | |
| "grad_norm": 5.145614147186279, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5109, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 3.03955078125, | |
| "grad_norm": 4.1711320877075195, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4124, | |
| "step": 6225 | |
| }, | |
| { | |
| "epoch": 3.0517578125, | |
| "grad_norm": 3.6686923503875732, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4233, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 3.06396484375, | |
| "grad_norm": 3.542102575302124, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5235, | |
| "step": 6275 | |
| }, | |
| { | |
| "epoch": 3.076171875, | |
| "grad_norm": 2.960315227508545, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4599, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 3.08837890625, | |
| "grad_norm": 3.370656728744507, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4668, | |
| "step": 6325 | |
| }, | |
| { | |
| "epoch": 3.1005859375, | |
| "grad_norm": 4.451176643371582, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4846, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 3.11279296875, | |
| "grad_norm": 3.628671646118164, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4636, | |
| "step": 6375 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 3.664491653442383, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4843, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 3.13720703125, | |
| "grad_norm": 3.1877737045288086, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5361, | |
| "step": 6425 | |
| }, | |
| { | |
| "epoch": 3.1494140625, | |
| "grad_norm": 3.7835752964019775, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4808, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 3.16162109375, | |
| "grad_norm": 3.494187355041504, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4401, | |
| "step": 6475 | |
| }, | |
| { | |
| "epoch": 3.173828125, | |
| "grad_norm": 2.837970495223999, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4454, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.173828125, | |
| "eval_cer": 21.709090909090907, | |
| "eval_loss": 0.40472567081451416, | |
| "eval_normalized_cer": 16.152897657213316, | |
| "eval_runtime": 133.6937, | |
| "eval_samples_per_second": 0.957, | |
| "eval_steps_per_second": 0.06, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.18603515625, | |
| "grad_norm": 3.5152645111083984, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4798, | |
| "step": 6525 | |
| }, | |
| { | |
| "epoch": 3.1982421875, | |
| "grad_norm": 3.201646327972412, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4576, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 3.21044921875, | |
| "grad_norm": 3.704028606414795, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4428, | |
| "step": 6575 | |
| }, | |
| { | |
| "epoch": 3.22265625, | |
| "grad_norm": 4.099301815032959, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4787, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 3.23486328125, | |
| "grad_norm": 3.0543127059936523, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4761, | |
| "step": 6625 | |
| }, | |
| { | |
| "epoch": 3.2470703125, | |
| "grad_norm": 3.4532554149627686, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4559, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 3.25927734375, | |
| "grad_norm": 2.8139491081237793, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4705, | |
| "step": 6675 | |
| }, | |
| { | |
| "epoch": 3.271484375, | |
| "grad_norm": 3.291689157485962, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4738, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 3.28369140625, | |
| "grad_norm": 3.0359091758728027, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3979, | |
| "step": 6725 | |
| }, | |
| { | |
| "epoch": 3.2958984375, | |
| "grad_norm": 4.325143337249756, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4309, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 3.30810546875, | |
| "grad_norm": 3.864635705947876, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4905, | |
| "step": 6775 | |
| }, | |
| { | |
| "epoch": 3.3203125, | |
| "grad_norm": 3.659311532974243, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5003, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 3.33251953125, | |
| "grad_norm": 3.6954805850982666, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4197, | |
| "step": 6825 | |
| }, | |
| { | |
| "epoch": 3.3447265625, | |
| "grad_norm": 2.9357662200927734, | |
| "learning_rate": 1e-05, | |
| "loss": 0.425, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 3.35693359375, | |
| "grad_norm": 2.9969289302825928, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4883, | |
| "step": 6875 | |
| }, | |
| { | |
| "epoch": 3.369140625, | |
| "grad_norm": 3.333348035812378, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4687, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 3.38134765625, | |
| "grad_norm": 4.444482803344727, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4324, | |
| "step": 6925 | |
| }, | |
| { | |
| "epoch": 3.3935546875, | |
| "grad_norm": 2.3350095748901367, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4476, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 3.40576171875, | |
| "grad_norm": 3.4752862453460693, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4601, | |
| "step": 6975 | |
| }, | |
| { | |
| "epoch": 3.41796875, | |
| "grad_norm": 3.1359448432922363, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4894, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.41796875, | |
| "eval_cer": 20.363636363636363, | |
| "eval_loss": 0.41460537910461426, | |
| "eval_normalized_cer": 15.988491574188245, | |
| "eval_runtime": 131.3226, | |
| "eval_samples_per_second": 0.975, | |
| "eval_steps_per_second": 0.061, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.43017578125, | |
| "grad_norm": 3.0270822048187256, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4288, | |
| "step": 7025 | |
| }, | |
| { | |
| "epoch": 3.4423828125, | |
| "grad_norm": 3.487501382827759, | |
| "learning_rate": 1e-05, | |
| "loss": 0.473, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 3.45458984375, | |
| "grad_norm": 3.1691272258758545, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4516, | |
| "step": 7075 | |
| }, | |
| { | |
| "epoch": 3.466796875, | |
| "grad_norm": 3.07665753364563, | |
| "learning_rate": 1e-05, | |
| "loss": 0.429, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 3.47900390625, | |
| "grad_norm": 3.697643756866455, | |
| "learning_rate": 1e-05, | |
| "loss": 0.479, | |
| "step": 7125 | |
| }, | |
| { | |
| "epoch": 3.4912109375, | |
| "grad_norm": 3.787980556488037, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4837, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 3.50341796875, | |
| "grad_norm": 3.792851686477661, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4339, | |
| "step": 7175 | |
| }, | |
| { | |
| "epoch": 3.515625, | |
| "grad_norm": 4.371828079223633, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4364, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.52783203125, | |
| "grad_norm": 2.8231394290924072, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4776, | |
| "step": 7225 | |
| }, | |
| { | |
| "epoch": 3.5400390625, | |
| "grad_norm": 3.1332738399505615, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4158, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 3.55224609375, | |
| "grad_norm": 3.6352245807647705, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4797, | |
| "step": 7275 | |
| }, | |
| { | |
| "epoch": 3.564453125, | |
| "grad_norm": 4.453933238983154, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4355, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 3.57666015625, | |
| "grad_norm": 2.9594037532806396, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4696, | |
| "step": 7325 | |
| }, | |
| { | |
| "epoch": 3.5888671875, | |
| "grad_norm": 5.014936447143555, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4622, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 3.60107421875, | |
| "grad_norm": 3.2860615253448486, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4536, | |
| "step": 7375 | |
| }, | |
| { | |
| "epoch": 3.61328125, | |
| "grad_norm": 5.162105083465576, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4456, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 3.62548828125, | |
| "grad_norm": 3.706555128097534, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4843, | |
| "step": 7425 | |
| }, | |
| { | |
| "epoch": 3.6376953125, | |
| "grad_norm": 2.490443706512451, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4284, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 3.64990234375, | |
| "grad_norm": 3.4704225063323975, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4562, | |
| "step": 7475 | |
| }, | |
| { | |
| "epoch": 3.662109375, | |
| "grad_norm": 3.8659839630126953, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4603, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.662109375, | |
| "eval_cer": 23.78181818181818, | |
| "eval_loss": 0.4173641800880432, | |
| "eval_normalized_cer": 19.35881627620222, | |
| "eval_runtime": 134.5888, | |
| "eval_samples_per_second": 0.951, | |
| "eval_steps_per_second": 0.059, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.67431640625, | |
| "grad_norm": 3.280311346054077, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4501, | |
| "step": 7525 | |
| }, | |
| { | |
| "epoch": 3.6865234375, | |
| "grad_norm": 4.28645133972168, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4437, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 3.69873046875, | |
| "grad_norm": 3.8637635707855225, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4302, | |
| "step": 7575 | |
| }, | |
| { | |
| "epoch": 3.7109375, | |
| "grad_norm": 3.35193133354187, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4298, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 3.72314453125, | |
| "grad_norm": 3.156805992126465, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4294, | |
| "step": 7625 | |
| }, | |
| { | |
| "epoch": 3.7353515625, | |
| "grad_norm": 3.2977190017700195, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4471, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 3.74755859375, | |
| "grad_norm": 3.3117992877960205, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4024, | |
| "step": 7675 | |
| }, | |
| { | |
| "epoch": 3.759765625, | |
| "grad_norm": 4.135869979858398, | |
| "learning_rate": 1e-05, | |
| "loss": 0.431, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 3.77197265625, | |
| "grad_norm": 3.1641712188720703, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4494, | |
| "step": 7725 | |
| }, | |
| { | |
| "epoch": 3.7841796875, | |
| "grad_norm": 3.1952223777770996, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4415, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 3.79638671875, | |
| "grad_norm": 3.2214698791503906, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4409, | |
| "step": 7775 | |
| }, | |
| { | |
| "epoch": 3.80859375, | |
| "grad_norm": 3.14152193069458, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4407, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 3.82080078125, | |
| "grad_norm": 3.8332841396331787, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4438, | |
| "step": 7825 | |
| }, | |
| { | |
| "epoch": 3.8330078125, | |
| "grad_norm": 3.2300031185150146, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4377, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 3.84521484375, | |
| "grad_norm": 3.9044997692108154, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4376, | |
| "step": 7875 | |
| }, | |
| { | |
| "epoch": 3.857421875, | |
| "grad_norm": 2.9638853073120117, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4605, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 3.86962890625, | |
| "grad_norm": 3.6924691200256348, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4632, | |
| "step": 7925 | |
| }, | |
| { | |
| "epoch": 3.8818359375, | |
| "grad_norm": 2.571397542953491, | |
| "learning_rate": 1e-05, | |
| "loss": 0.439, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 3.89404296875, | |
| "grad_norm": 2.900402069091797, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4431, | |
| "step": 7975 | |
| }, | |
| { | |
| "epoch": 3.90625, | |
| "grad_norm": 3.216660737991333, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4479, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.90625, | |
| "eval_cer": 21.527272727272727, | |
| "eval_loss": 0.40407997369766235, | |
| "eval_normalized_cer": 16.974928072338678, | |
| "eval_runtime": 135.4416, | |
| "eval_samples_per_second": 0.945, | |
| "eval_steps_per_second": 0.059, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.91845703125, | |
| "grad_norm": 3.2848970890045166, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4152, | |
| "step": 8025 | |
| }, | |
| { | |
| "epoch": 3.9306640625, | |
| "grad_norm": 2.901078939437866, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4292, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 3.94287109375, | |
| "grad_norm": 3.3636679649353027, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4624, | |
| "step": 8075 | |
| }, | |
| { | |
| "epoch": 3.955078125, | |
| "grad_norm": 4.242199420928955, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4427, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 3.96728515625, | |
| "grad_norm": 3.422555446624756, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4161, | |
| "step": 8125 | |
| }, | |
| { | |
| "epoch": 3.9794921875, | |
| "grad_norm": 2.359015703201294, | |
| "learning_rate": 1e-05, | |
| "loss": 0.422, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 3.99169921875, | |
| "grad_norm": 3.1188254356384277, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4346, | |
| "step": 8175 | |
| }, | |
| { | |
| "epoch": 4.00390625, | |
| "grad_norm": 3.1094157695770264, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4511, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 4.01611328125, | |
| "grad_norm": 3.269327163696289, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4548, | |
| "step": 8225 | |
| }, | |
| { | |
| "epoch": 4.0283203125, | |
| "grad_norm": 2.589052677154541, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4267, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 4.04052734375, | |
| "grad_norm": 3.032090187072754, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4194, | |
| "step": 8275 | |
| }, | |
| { | |
| "epoch": 4.052734375, | |
| "grad_norm": 3.1409432888031006, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4423, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 4.06494140625, | |
| "grad_norm": 4.070183753967285, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4171, | |
| "step": 8325 | |
| }, | |
| { | |
| "epoch": 4.0771484375, | |
| "grad_norm": 4.6292619705200195, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4689, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 4.08935546875, | |
| "grad_norm": 3.5778212547302246, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4341, | |
| "step": 8375 | |
| }, | |
| { | |
| "epoch": 4.1015625, | |
| "grad_norm": 2.5463714599609375, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4193, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 4.11376953125, | |
| "grad_norm": 2.498852491378784, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4656, | |
| "step": 8425 | |
| }, | |
| { | |
| "epoch": 4.1259765625, | |
| "grad_norm": 2.5242180824279785, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4491, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 4.13818359375, | |
| "grad_norm": 4.114852428436279, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4202, | |
| "step": 8475 | |
| }, | |
| { | |
| "epoch": 4.150390625, | |
| "grad_norm": 2.866490125656128, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4523, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.150390625, | |
| "eval_cer": 19.70909090909091, | |
| "eval_loss": 0.40719008445739746, | |
| "eval_normalized_cer": 15.371968762844226, | |
| "eval_runtime": 136.9983, | |
| "eval_samples_per_second": 0.934, | |
| "eval_steps_per_second": 0.058, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.16259765625, | |
| "grad_norm": 4.2479448318481445, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4742, | |
| "step": 8525 | |
| }, | |
| { | |
| "epoch": 4.1748046875, | |
| "grad_norm": 2.736863851547241, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4366, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 4.18701171875, | |
| "grad_norm": 3.3167812824249268, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4509, | |
| "step": 8575 | |
| }, | |
| { | |
| "epoch": 4.19921875, | |
| "grad_norm": 2.8058440685272217, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4179, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 4.21142578125, | |
| "grad_norm": 3.10642409324646, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4799, | |
| "step": 8625 | |
| }, | |
| { | |
| "epoch": 4.2236328125, | |
| "grad_norm": 3.9041507244110107, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4788, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 4.23583984375, | |
| "grad_norm": 4.171154022216797, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4534, | |
| "step": 8675 | |
| }, | |
| { | |
| "epoch": 4.248046875, | |
| "grad_norm": 2.6511294841766357, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4395, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 4.26025390625, | |
| "grad_norm": 3.9899098873138428, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4421, | |
| "step": 8725 | |
| }, | |
| { | |
| "epoch": 4.2724609375, | |
| "grad_norm": 2.973851442337036, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4166, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 4.28466796875, | |
| "grad_norm": 3.789973497390747, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4312, | |
| "step": 8775 | |
| }, | |
| { | |
| "epoch": 4.296875, | |
| "grad_norm": 4.157674789428711, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4571, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 4.30908203125, | |
| "grad_norm": 4.191178321838379, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4222, | |
| "step": 8825 | |
| }, | |
| { | |
| "epoch": 4.3212890625, | |
| "grad_norm": 2.899761915206909, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4174, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 4.33349609375, | |
| "grad_norm": 2.9615023136138916, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4684, | |
| "step": 8875 | |
| }, | |
| { | |
| "epoch": 4.345703125, | |
| "grad_norm": 2.9529151916503906, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4079, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 4.35791015625, | |
| "grad_norm": 3.4950997829437256, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4239, | |
| "step": 8925 | |
| }, | |
| { | |
| "epoch": 4.3701171875, | |
| "grad_norm": 4.151655673980713, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4737, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 4.38232421875, | |
| "grad_norm": 3.2207164764404297, | |
| "learning_rate": 1e-05, | |
| "loss": 0.468, | |
| "step": 8975 | |
| }, | |
| { | |
| "epoch": 4.39453125, | |
| "grad_norm": 4.433598041534424, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4742, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.39453125, | |
| "eval_cer": 20.836363636363636, | |
| "eval_loss": 0.40726813673973083, | |
| "eval_normalized_cer": 16.563912864775997, | |
| "eval_runtime": 148.3773, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.054, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.40673828125, | |
| "grad_norm": 2.8774595260620117, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4355, | |
| "step": 9025 | |
| }, | |
| { | |
| "epoch": 4.4189453125, | |
| "grad_norm": 3.043325185775757, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4267, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 4.43115234375, | |
| "grad_norm": 3.7946414947509766, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4424, | |
| "step": 9075 | |
| }, | |
| { | |
| "epoch": 4.443359375, | |
| "grad_norm": 2.7561936378479004, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4054, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 4.45556640625, | |
| "grad_norm": 3.3554115295410156, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4442, | |
| "step": 9125 | |
| }, | |
| { | |
| "epoch": 4.4677734375, | |
| "grad_norm": 3.1039364337921143, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4289, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 4.47998046875, | |
| "grad_norm": 2.8910741806030273, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4103, | |
| "step": 9175 | |
| }, | |
| { | |
| "epoch": 4.4921875, | |
| "grad_norm": 3.005373001098633, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3944, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 4.50439453125, | |
| "grad_norm": 2.7834503650665283, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4235, | |
| "step": 9225 | |
| }, | |
| { | |
| "epoch": 4.5166015625, | |
| "grad_norm": 3.2540178298950195, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4617, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 4.52880859375, | |
| "grad_norm": 3.1600005626678467, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4445, | |
| "step": 9275 | |
| }, | |
| { | |
| "epoch": 4.541015625, | |
| "grad_norm": 3.318638801574707, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4164, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 4.55322265625, | |
| "grad_norm": 4.483547210693359, | |
| "learning_rate": 1e-05, | |
| "loss": 0.468, | |
| "step": 9325 | |
| }, | |
| { | |
| "epoch": 4.5654296875, | |
| "grad_norm": 3.946882963180542, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4349, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 4.57763671875, | |
| "grad_norm": 2.6997079849243164, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4483, | |
| "step": 9375 | |
| }, | |
| { | |
| "epoch": 4.58984375, | |
| "grad_norm": 5.617648601531982, | |
| "learning_rate": 1e-05, | |
| "loss": 0.477, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 4.60205078125, | |
| "grad_norm": 3.01735520362854, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4198, | |
| "step": 9425 | |
| }, | |
| { | |
| "epoch": 4.6142578125, | |
| "grad_norm": 3.301004648208618, | |
| "learning_rate": 1e-05, | |
| "loss": 0.466, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 4.62646484375, | |
| "grad_norm": 3.2421813011169434, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4366, | |
| "step": 9475 | |
| }, | |
| { | |
| "epoch": 4.638671875, | |
| "grad_norm": 4.813492774963379, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4737, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.638671875, | |
| "eval_cer": 21.78181818181818, | |
| "eval_loss": 0.40351322293281555, | |
| "eval_normalized_cer": 16.892725030826142, | |
| "eval_runtime": 136.3812, | |
| "eval_samples_per_second": 0.939, | |
| "eval_steps_per_second": 0.059, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.65087890625, | |
| "grad_norm": 3.3861124515533447, | |
| "learning_rate": 1e-05, | |
| "loss": 0.447, | |
| "step": 9525 | |
| }, | |
| { | |
| "epoch": 4.6630859375, | |
| "grad_norm": 3.244462013244629, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4422, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 4.67529296875, | |
| "grad_norm": 3.824782133102417, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4431, | |
| "step": 9575 | |
| }, | |
| { | |
| "epoch": 4.6875, | |
| "grad_norm": 3.746281862258911, | |
| "learning_rate": 1e-05, | |
| "loss": 0.48, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 4.69970703125, | |
| "grad_norm": 4.126583099365234, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4643, | |
| "step": 9625 | |
| }, | |
| { | |
| "epoch": 4.7119140625, | |
| "grad_norm": 3.3953585624694824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4826, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 4.72412109375, | |
| "grad_norm": 3.077698230743408, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4138, | |
| "step": 9675 | |
| }, | |
| { | |
| "epoch": 4.736328125, | |
| "grad_norm": 4.628427982330322, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4102, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 4.74853515625, | |
| "grad_norm": 3.1975038051605225, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4374, | |
| "step": 9725 | |
| }, | |
| { | |
| "epoch": 4.7607421875, | |
| "grad_norm": 3.606600522994995, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4095, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 4.77294921875, | |
| "grad_norm": 3.183274507522583, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4759, | |
| "step": 9775 | |
| }, | |
| { | |
| "epoch": 4.78515625, | |
| "grad_norm": 3.52425479888916, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4429, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 4.79736328125, | |
| "grad_norm": 3.2504448890686035, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4271, | |
| "step": 9825 | |
| }, | |
| { | |
| "epoch": 4.8095703125, | |
| "grad_norm": 3.7762293815612793, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4164, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 4.82177734375, | |
| "grad_norm": 2.5560717582702637, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4622, | |
| "step": 9875 | |
| }, | |
| { | |
| "epoch": 4.833984375, | |
| "grad_norm": 2.2996389865875244, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4728, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 4.84619140625, | |
| "grad_norm": 4.10749626159668, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4954, | |
| "step": 9925 | |
| }, | |
| { | |
| "epoch": 4.8583984375, | |
| "grad_norm": 3.2446017265319824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.485, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 4.87060546875, | |
| "grad_norm": 2.7468326091766357, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4026, | |
| "step": 9975 | |
| }, | |
| { | |
| "epoch": 4.8828125, | |
| "grad_norm": 4.006360054016113, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4785, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.8828125, | |
| "eval_cer": 21.490909090909092, | |
| "eval_loss": 0.39705541729927063, | |
| "eval_normalized_cer": 16.48170982326346, | |
| "eval_runtime": 135.3957, | |
| "eval_samples_per_second": 0.945, | |
| "eval_steps_per_second": 0.059, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.89501953125, | |
| "grad_norm": 1.9145240783691406, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4676, | |
| "step": 10025 | |
| }, | |
| { | |
| "epoch": 4.9072265625, | |
| "grad_norm": 3.474172830581665, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4759, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 4.91943359375, | |
| "grad_norm": 4.628629684448242, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4614, | |
| "step": 10075 | |
| }, | |
| { | |
| "epoch": 4.931640625, | |
| "grad_norm": 3.4078354835510254, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4194, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 4.94384765625, | |
| "grad_norm": 2.336055278778076, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4508, | |
| "step": 10125 | |
| }, | |
| { | |
| "epoch": 4.9560546875, | |
| "grad_norm": 3.3985817432403564, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5009, | |
| "step": 10150 | |
| }, | |
| { | |
| "epoch": 4.96826171875, | |
| "grad_norm": 3.0444228649139404, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4485, | |
| "step": 10175 | |
| }, | |
| { | |
| "epoch": 4.98046875, | |
| "grad_norm": 3.737316131591797, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4723, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 4.99267578125, | |
| "grad_norm": 2.8594186305999756, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4411, | |
| "step": 10225 | |
| }, | |
| { | |
| "epoch": 5.0048828125, | |
| "grad_norm": 3.403825044631958, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3681, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 5.01708984375, | |
| "grad_norm": 2.896949291229248, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4386, | |
| "step": 10275 | |
| }, | |
| { | |
| "epoch": 5.029296875, | |
| "grad_norm": 3.3506815433502197, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4113, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 5.04150390625, | |
| "grad_norm": 4.682892322540283, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4773, | |
| "step": 10325 | |
| }, | |
| { | |
| "epoch": 5.0537109375, | |
| "grad_norm": 4.2950439453125, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4289, | |
| "step": 10350 | |
| }, | |
| { | |
| "epoch": 5.06591796875, | |
| "grad_norm": 3.2204432487487793, | |
| "learning_rate": 1e-05, | |
| "loss": 0.423, | |
| "step": 10375 | |
| }, | |
| { | |
| "epoch": 5.078125, | |
| "grad_norm": 4.998103618621826, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4115, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 5.09033203125, | |
| "grad_norm": 2.7820868492126465, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4685, | |
| "step": 10425 | |
| }, | |
| { | |
| "epoch": 5.1025390625, | |
| "grad_norm": 3.236323356628418, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4255, | |
| "step": 10450 | |
| }, | |
| { | |
| "epoch": 5.11474609375, | |
| "grad_norm": 3.809432029724121, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4964, | |
| "step": 10475 | |
| }, | |
| { | |
| "epoch": 5.126953125, | |
| "grad_norm": 5.030189514160156, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4664, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.126953125, | |
| "eval_cer": 22.581818181818182, | |
| "eval_loss": 0.4009712040424347, | |
| "eval_normalized_cer": 18.33127825729552, | |
| "eval_runtime": 135.5914, | |
| "eval_samples_per_second": 0.944, | |
| "eval_steps_per_second": 0.059, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.13916015625, | |
| "grad_norm": 2.855731248855591, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4279, | |
| "step": 10525 | |
| }, | |
| { | |
| "epoch": 5.1513671875, | |
| "grad_norm": 3.264129161834717, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4094, | |
| "step": 10550 | |
| }, | |
| { | |
| "epoch": 5.16357421875, | |
| "grad_norm": 3.299495220184326, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4343, | |
| "step": 10575 | |
| }, | |
| { | |
| "epoch": 5.17578125, | |
| "grad_norm": 3.564100503921509, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4514, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 5.18798828125, | |
| "grad_norm": 2.6773531436920166, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4512, | |
| "step": 10625 | |
| }, | |
| { | |
| "epoch": 5.2001953125, | |
| "grad_norm": 3.721902370452881, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4326, | |
| "step": 10650 | |
| }, | |
| { | |
| "epoch": 5.21240234375, | |
| "grad_norm": 2.919142961502075, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4466, | |
| "step": 10675 | |
| }, | |
| { | |
| "epoch": 5.224609375, | |
| "grad_norm": 3.0286405086517334, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3802, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 5.23681640625, | |
| "grad_norm": 4.255770206451416, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4289, | |
| "step": 10725 | |
| }, | |
| { | |
| "epoch": 5.2490234375, | |
| "grad_norm": 4.174068450927734, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4699, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 5.26123046875, | |
| "grad_norm": 2.8515052795410156, | |
| "learning_rate": 1e-05, | |
| "loss": 0.464, | |
| "step": 10775 | |
| }, | |
| { | |
| "epoch": 5.2734375, | |
| "grad_norm": 2.487253427505493, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4222, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 5.28564453125, | |
| "grad_norm": 2.592994213104248, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3916, | |
| "step": 10825 | |
| }, | |
| { | |
| "epoch": 5.2978515625, | |
| "grad_norm": 3.5928993225097656, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4278, | |
| "step": 10850 | |
| }, | |
| { | |
| "epoch": 5.31005859375, | |
| "grad_norm": 3.690007448196411, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4369, | |
| "step": 10875 | |
| }, | |
| { | |
| "epoch": 5.322265625, | |
| "grad_norm": 3.200507879257202, | |
| "learning_rate": 1e-05, | |
| "loss": 0.391, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 5.33447265625, | |
| "grad_norm": 2.6833174228668213, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4602, | |
| "step": 10925 | |
| }, | |
| { | |
| "epoch": 5.3466796875, | |
| "grad_norm": 4.44938325881958, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4648, | |
| "step": 10950 | |
| }, | |
| { | |
| "epoch": 5.35888671875, | |
| "grad_norm": 3.660412073135376, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4193, | |
| "step": 10975 | |
| }, | |
| { | |
| "epoch": 5.37109375, | |
| "grad_norm": 3.3475520610809326, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4497, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.37109375, | |
| "eval_cer": 23.381818181818183, | |
| "eval_loss": 0.40254130959510803, | |
| "eval_normalized_cer": 18.824496506370735, | |
| "eval_runtime": 134.1905, | |
| "eval_samples_per_second": 0.954, | |
| "eval_steps_per_second": 0.06, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.38330078125, | |
| "grad_norm": 2.6810033321380615, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4332, | |
| "step": 11025 | |
| }, | |
| { | |
| "epoch": 5.3955078125, | |
| "grad_norm": 3.203681707382202, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4225, | |
| "step": 11050 | |
| }, | |
| { | |
| "epoch": 5.40771484375, | |
| "grad_norm": 2.916377305984497, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4133, | |
| "step": 11075 | |
| }, | |
| { | |
| "epoch": 5.419921875, | |
| "grad_norm": 3.2195537090301514, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4842, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 5.43212890625, | |
| "grad_norm": 3.1595821380615234, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4547, | |
| "step": 11125 | |
| }, | |
| { | |
| "epoch": 5.4443359375, | |
| "grad_norm": 3.207057237625122, | |
| "learning_rate": 1e-05, | |
| "loss": 0.394, | |
| "step": 11150 | |
| }, | |
| { | |
| "epoch": 5.45654296875, | |
| "grad_norm": 2.9330296516418457, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3902, | |
| "step": 11175 | |
| }, | |
| { | |
| "epoch": 5.46875, | |
| "grad_norm": 2.4040415287017822, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4565, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 5.48095703125, | |
| "grad_norm": 5.465662479400635, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4368, | |
| "step": 11225 | |
| }, | |
| { | |
| "epoch": 5.4931640625, | |
| "grad_norm": 3.320988178253174, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4346, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 5.50537109375, | |
| "grad_norm": 2.748600959777832, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4632, | |
| "step": 11275 | |
| }, | |
| { | |
| "epoch": 5.517578125, | |
| "grad_norm": 2.1236891746520996, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4194, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 5.52978515625, | |
| "grad_norm": 3.5435776710510254, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4293, | |
| "step": 11325 | |
| }, | |
| { | |
| "epoch": 5.5419921875, | |
| "grad_norm": 3.9727423191070557, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4373, | |
| "step": 11350 | |
| }, | |
| { | |
| "epoch": 5.55419921875, | |
| "grad_norm": 4.122025966644287, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3914, | |
| "step": 11375 | |
| }, | |
| { | |
| "epoch": 5.56640625, | |
| "grad_norm": 3.523620128631592, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4883, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 5.57861328125, | |
| "grad_norm": 3.5819413661956787, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4592, | |
| "step": 11425 | |
| }, | |
| { | |
| "epoch": 5.5908203125, | |
| "grad_norm": 3.022115468978882, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4231, | |
| "step": 11450 | |
| }, | |
| { | |
| "epoch": 5.60302734375, | |
| "grad_norm": 3.9213130474090576, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4391, | |
| "step": 11475 | |
| }, | |
| { | |
| "epoch": 5.615234375, | |
| "grad_norm": 2.55260968208313, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4375, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.615234375, | |
| "eval_cer": 23.745454545454546, | |
| "eval_loss": 0.3965916633605957, | |
| "eval_normalized_cer": 20.098643649815042, | |
| "eval_runtime": 136.6172, | |
| "eval_samples_per_second": 0.937, | |
| "eval_steps_per_second": 0.059, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.62744140625, | |
| "grad_norm": 4.3812479972839355, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4251, | |
| "step": 11525 | |
| }, | |
| { | |
| "epoch": 5.6396484375, | |
| "grad_norm": 3.3777589797973633, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4634, | |
| "step": 11550 | |
| }, | |
| { | |
| "epoch": 5.65185546875, | |
| "grad_norm": 3.418990135192871, | |
| "learning_rate": 1e-05, | |
| "loss": 0.453, | |
| "step": 11575 | |
| }, | |
| { | |
| "epoch": 5.6640625, | |
| "grad_norm": 3.4627134799957275, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4335, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 5.67626953125, | |
| "grad_norm": 3.955687999725342, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4648, | |
| "step": 11625 | |
| }, | |
| { | |
| "epoch": 5.6884765625, | |
| "grad_norm": 3.9896974563598633, | |
| "learning_rate": 1e-05, | |
| "loss": 0.468, | |
| "step": 11650 | |
| }, | |
| { | |
| "epoch": 5.70068359375, | |
| "grad_norm": 4.082157611846924, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4801, | |
| "step": 11675 | |
| }, | |
| { | |
| "epoch": 5.712890625, | |
| "grad_norm": 2.509510040283203, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4414, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 5.72509765625, | |
| "grad_norm": 4.054114818572998, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4341, | |
| "step": 11725 | |
| }, | |
| { | |
| "epoch": 5.7373046875, | |
| "grad_norm": 3.3360326290130615, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4294, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 5.74951171875, | |
| "grad_norm": 3.023287773132324, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4177, | |
| "step": 11775 | |
| }, | |
| { | |
| "epoch": 5.76171875, | |
| "grad_norm": 2.7048487663269043, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4284, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 5.77392578125, | |
| "grad_norm": 3.7769124507904053, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4297, | |
| "step": 11825 | |
| }, | |
| { | |
| "epoch": 5.7861328125, | |
| "grad_norm": 2.583249092102051, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4328, | |
| "step": 11850 | |
| }, | |
| { | |
| "epoch": 5.79833984375, | |
| "grad_norm": 4.198108196258545, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4499, | |
| "step": 11875 | |
| }, | |
| { | |
| "epoch": 5.810546875, | |
| "grad_norm": 2.8993310928344727, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4081, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 5.82275390625, | |
| "grad_norm": 3.5087976455688477, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4716, | |
| "step": 11925 | |
| }, | |
| { | |
| "epoch": 5.8349609375, | |
| "grad_norm": 3.527189254760742, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4229, | |
| "step": 11950 | |
| }, | |
| { | |
| "epoch": 5.84716796875, | |
| "grad_norm": 2.4347596168518066, | |
| "learning_rate": 1e-05, | |
| "loss": 0.448, | |
| "step": 11975 | |
| }, | |
| { | |
| "epoch": 5.859375, | |
| "grad_norm": 3.4706249237060547, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4526, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.859375, | |
| "eval_cer": 24.87272727272727, | |
| "eval_loss": 0.3955676555633545, | |
| "eval_normalized_cer": 20.263049732840116, | |
| "eval_runtime": 133.1253, | |
| "eval_samples_per_second": 0.961, | |
| "eval_steps_per_second": 0.06, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.87158203125, | |
| "grad_norm": 3.7815253734588623, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4, | |
| "step": 12025 | |
| }, | |
| { | |
| "epoch": 5.8837890625, | |
| "grad_norm": 4.76561164855957, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4176, | |
| "step": 12050 | |
| }, | |
| { | |
| "epoch": 5.89599609375, | |
| "grad_norm": 3.329918622970581, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4232, | |
| "step": 12075 | |
| }, | |
| { | |
| "epoch": 5.908203125, | |
| "grad_norm": 4.160305023193359, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4123, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 5.92041015625, | |
| "grad_norm": 3.11075758934021, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4468, | |
| "step": 12125 | |
| }, | |
| { | |
| "epoch": 5.9326171875, | |
| "grad_norm": 3.1409554481506348, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4624, | |
| "step": 12150 | |
| }, | |
| { | |
| "epoch": 5.94482421875, | |
| "grad_norm": 2.9979705810546875, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4369, | |
| "step": 12175 | |
| }, | |
| { | |
| "epoch": 5.95703125, | |
| "grad_norm": 3.538670301437378, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4299, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 5.96923828125, | |
| "grad_norm": 3.377985954284668, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3944, | |
| "step": 12225 | |
| }, | |
| { | |
| "epoch": 5.9814453125, | |
| "grad_norm": 3.6021058559417725, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4314, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 5.99365234375, | |
| "grad_norm": 3.2901623249053955, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4104, | |
| "step": 12275 | |
| }, | |
| { | |
| "epoch": 6.005859375, | |
| "grad_norm": 4.384257793426514, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5273, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 6.01806640625, | |
| "grad_norm": 3.564580202102661, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4292, | |
| "step": 12325 | |
| }, | |
| { | |
| "epoch": 6.0302734375, | |
| "grad_norm": 3.4145617485046387, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4183, | |
| "step": 12350 | |
| }, | |
| { | |
| "epoch": 6.04248046875, | |
| "grad_norm": 3.1783013343811035, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4193, | |
| "step": 12375 | |
| }, | |
| { | |
| "epoch": 6.0546875, | |
| "grad_norm": 3.3418772220611572, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4635, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 6.06689453125, | |
| "grad_norm": 3.057255744934082, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4068, | |
| "step": 12425 | |
| }, | |
| { | |
| "epoch": 6.0791015625, | |
| "grad_norm": 3.5321204662323, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3614, | |
| "step": 12450 | |
| }, | |
| { | |
| "epoch": 6.09130859375, | |
| "grad_norm": 3.536223888397217, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4195, | |
| "step": 12475 | |
| }, | |
| { | |
| "epoch": 6.103515625, | |
| "grad_norm": 2.386425018310547, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4545, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 6.103515625, | |
| "eval_cer": 26.47272727272727, | |
| "eval_loss": 0.38523948192596436, | |
| "eval_normalized_cer": 20.550760378133994, | |
| "eval_runtime": 117.0609, | |
| "eval_samples_per_second": 1.093, | |
| "eval_steps_per_second": 0.068, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 6.11572265625, | |
| "grad_norm": 2.7935492992401123, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4149, | |
| "step": 12525 | |
| }, | |
| { | |
| "epoch": 6.1279296875, | |
| "grad_norm": 3.3430776596069336, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4718, | |
| "step": 12550 | |
| }, | |
| { | |
| "epoch": 6.14013671875, | |
| "grad_norm": 2.8501319885253906, | |
| "learning_rate": 1e-05, | |
| "loss": 0.467, | |
| "step": 12575 | |
| }, | |
| { | |
| "epoch": 6.15234375, | |
| "grad_norm": 3.3480148315429688, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4245, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 6.16455078125, | |
| "grad_norm": 3.101698160171509, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4185, | |
| "step": 12625 | |
| }, | |
| { | |
| "epoch": 6.1767578125, | |
| "grad_norm": 2.7815115451812744, | |
| "learning_rate": 1e-05, | |
| "loss": 0.446, | |
| "step": 12650 | |
| }, | |
| { | |
| "epoch": 6.18896484375, | |
| "grad_norm": 2.8936548233032227, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4136, | |
| "step": 12675 | |
| }, | |
| { | |
| "epoch": 6.201171875, | |
| "grad_norm": 4.183928966522217, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4384, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 6.21337890625, | |
| "grad_norm": 3.2840652465820312, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4782, | |
| "step": 12725 | |
| }, | |
| { | |
| "epoch": 6.2255859375, | |
| "grad_norm": 3.9204938411712646, | |
| "learning_rate": 1e-05, | |
| "loss": 0.432, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 6.23779296875, | |
| "grad_norm": 4.316317081451416, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4644, | |
| "step": 12775 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 3.1822116374969482, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4331, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 6.26220703125, | |
| "grad_norm": 2.9394965171813965, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4263, | |
| "step": 12825 | |
| }, | |
| { | |
| "epoch": 6.2744140625, | |
| "grad_norm": 3.419806718826294, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4219, | |
| "step": 12850 | |
| }, | |
| { | |
| "epoch": 6.28662109375, | |
| "grad_norm": 3.1678640842437744, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4425, | |
| "step": 12875 | |
| }, | |
| { | |
| "epoch": 6.298828125, | |
| "grad_norm": 4.1609697341918945, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4355, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 6.31103515625, | |
| "grad_norm": 3.895420789718628, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4633, | |
| "step": 12925 | |
| }, | |
| { | |
| "epoch": 6.3232421875, | |
| "grad_norm": 3.244763135910034, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4219, | |
| "step": 12950 | |
| }, | |
| { | |
| "epoch": 6.33544921875, | |
| "grad_norm": 4.2628984451293945, | |
| "learning_rate": 1e-05, | |
| "loss": 0.432, | |
| "step": 12975 | |
| }, | |
| { | |
| "epoch": 6.34765625, | |
| "grad_norm": 4.046095848083496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4199, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.34765625, | |
| "eval_cer": 20.909090909090907, | |
| "eval_loss": 0.38916099071502686, | |
| "eval_normalized_cer": 16.31730374023839, | |
| "eval_runtime": 128.9363, | |
| "eval_samples_per_second": 0.993, | |
| "eval_steps_per_second": 0.062, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.35986328125, | |
| "grad_norm": 3.0475289821624756, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4105, | |
| "step": 13025 | |
| }, | |
| { | |
| "epoch": 6.3720703125, | |
| "grad_norm": 3.288393974304199, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4627, | |
| "step": 13050 | |
| }, | |
| { | |
| "epoch": 6.38427734375, | |
| "grad_norm": 4.572826862335205, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4259, | |
| "step": 13075 | |
| }, | |
| { | |
| "epoch": 6.396484375, | |
| "grad_norm": 3.2055158615112305, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4124, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 6.40869140625, | |
| "grad_norm": 3.237438917160034, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4525, | |
| "step": 13125 | |
| }, | |
| { | |
| "epoch": 6.4208984375, | |
| "grad_norm": 3.7672817707061768, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4218, | |
| "step": 13150 | |
| }, | |
| { | |
| "epoch": 6.43310546875, | |
| "grad_norm": 2.8958332538604736, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4501, | |
| "step": 13175 | |
| }, | |
| { | |
| "epoch": 6.4453125, | |
| "grad_norm": 3.3539748191833496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4294, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 6.45751953125, | |
| "grad_norm": 3.745314836502075, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4038, | |
| "step": 13225 | |
| }, | |
| { | |
| "epoch": 6.4697265625, | |
| "grad_norm": 4.028524398803711, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4313, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 6.48193359375, | |
| "grad_norm": 3.058047294616699, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4215, | |
| "step": 13275 | |
| }, | |
| { | |
| "epoch": 6.494140625, | |
| "grad_norm": 2.5536253452301025, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4306, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 6.50634765625, | |
| "grad_norm": 2.894327402114868, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4211, | |
| "step": 13325 | |
| }, | |
| { | |
| "epoch": 6.5185546875, | |
| "grad_norm": 4.001802444458008, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4212, | |
| "step": 13350 | |
| }, | |
| { | |
| "epoch": 6.53076171875, | |
| "grad_norm": 3.1187326908111572, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4371, | |
| "step": 13375 | |
| }, | |
| { | |
| "epoch": 6.54296875, | |
| "grad_norm": 2.8355751037597656, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4356, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 6.55517578125, | |
| "grad_norm": 4.038667678833008, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4009, | |
| "step": 13425 | |
| }, | |
| { | |
| "epoch": 6.5673828125, | |
| "grad_norm": 3.5054690837860107, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4561, | |
| "step": 13450 | |
| }, | |
| { | |
| "epoch": 6.57958984375, | |
| "grad_norm": 3.1248323917388916, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4244, | |
| "step": 13475 | |
| }, | |
| { | |
| "epoch": 6.591796875, | |
| "grad_norm": 3.1242165565490723, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4456, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 6.591796875, | |
| "eval_cer": 22.545454545454547, | |
| "eval_loss": 0.3937221169471741, | |
| "eval_normalized_cer": 17.79695848746404, | |
| "eval_runtime": 131.9352, | |
| "eval_samples_per_second": 0.97, | |
| "eval_steps_per_second": 0.061, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 6.60400390625, | |
| "grad_norm": 2.5615017414093018, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4451, | |
| "step": 13525 | |
| }, | |
| { | |
| "epoch": 6.6162109375, | |
| "grad_norm": 3.3750381469726562, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4233, | |
| "step": 13550 | |
| }, | |
| { | |
| "epoch": 6.62841796875, | |
| "grad_norm": 3.462449073791504, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4426, | |
| "step": 13575 | |
| }, | |
| { | |
| "epoch": 6.640625, | |
| "grad_norm": 2.785858154296875, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4073, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 6.65283203125, | |
| "grad_norm": 2.7007555961608887, | |
| "learning_rate": 1e-05, | |
| "loss": 0.438, | |
| "step": 13625 | |
| }, | |
| { | |
| "epoch": 6.6650390625, | |
| "grad_norm": 2.3689684867858887, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4226, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 6.67724609375, | |
| "grad_norm": 3.3723831176757812, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4279, | |
| "step": 13675 | |
| }, | |
| { | |
| "epoch": 6.689453125, | |
| "grad_norm": 2.9226441383361816, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4242, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 6.70166015625, | |
| "grad_norm": 2.7876479625701904, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4328, | |
| "step": 13725 | |
| }, | |
| { | |
| "epoch": 6.7138671875, | |
| "grad_norm": 2.7610867023468018, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4412, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 6.72607421875, | |
| "grad_norm": 2.8302724361419678, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4526, | |
| "step": 13775 | |
| }, | |
| { | |
| "epoch": 6.73828125, | |
| "grad_norm": 3.1445977687835693, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4544, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 6.75048828125, | |
| "grad_norm": 3.8864002227783203, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3965, | |
| "step": 13825 | |
| }, | |
| { | |
| "epoch": 6.7626953125, | |
| "grad_norm": 3.2896971702575684, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4182, | |
| "step": 13850 | |
| }, | |
| { | |
| "epoch": 6.77490234375, | |
| "grad_norm": 4.539292335510254, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4413, | |
| "step": 13875 | |
| }, | |
| { | |
| "epoch": 6.787109375, | |
| "grad_norm": 3.2103347778320312, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4346, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 6.79931640625, | |
| "grad_norm": 3.929469585418701, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4181, | |
| "step": 13925 | |
| }, | |
| { | |
| "epoch": 6.8115234375, | |
| "grad_norm": 3.3086280822753906, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3981, | |
| "step": 13950 | |
| }, | |
| { | |
| "epoch": 6.82373046875, | |
| "grad_norm": 3.101243734359741, | |
| "learning_rate": 1e-05, | |
| "loss": 0.503, | |
| "step": 13975 | |
| }, | |
| { | |
| "epoch": 6.8359375, | |
| "grad_norm": 3.602478265762329, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4186, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.8359375, | |
| "eval_cer": 22.472727272727273, | |
| "eval_loss": 0.3899889588356018, | |
| "eval_normalized_cer": 18.33127825729552, | |
| "eval_runtime": 144.5925, | |
| "eval_samples_per_second": 0.885, | |
| "eval_steps_per_second": 0.055, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.84814453125, | |
| "grad_norm": 3.0590033531188965, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4268, | |
| "step": 14025 | |
| }, | |
| { | |
| "epoch": 6.8603515625, | |
| "grad_norm": 4.508790969848633, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4063, | |
| "step": 14050 | |
| }, | |
| { | |
| "epoch": 6.87255859375, | |
| "grad_norm": 3.374329090118408, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4482, | |
| "step": 14075 | |
| }, | |
| { | |
| "epoch": 6.884765625, | |
| "grad_norm": 2.569153308868408, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4477, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 6.89697265625, | |
| "grad_norm": 2.5716049671173096, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4282, | |
| "step": 14125 | |
| }, | |
| { | |
| "epoch": 6.9091796875, | |
| "grad_norm": 2.990652561187744, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4409, | |
| "step": 14150 | |
| }, | |
| { | |
| "epoch": 6.92138671875, | |
| "grad_norm": 2.772998094558716, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3622, | |
| "step": 14175 | |
| }, | |
| { | |
| "epoch": 6.93359375, | |
| "grad_norm": 3.0832324028015137, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4017, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 6.94580078125, | |
| "grad_norm": 2.2825822830200195, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4341, | |
| "step": 14225 | |
| }, | |
| { | |
| "epoch": 6.9580078125, | |
| "grad_norm": 2.8773045539855957, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4512, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 6.97021484375, | |
| "grad_norm": 2.9340527057647705, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4004, | |
| "step": 14275 | |
| }, | |
| { | |
| "epoch": 6.982421875, | |
| "grad_norm": 3.3255691528320312, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3775, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 6.99462890625, | |
| "grad_norm": 3.9721550941467285, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4239, | |
| "step": 14325 | |
| }, | |
| { | |
| "epoch": 7.0068359375, | |
| "grad_norm": 2.075686454772949, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3713, | |
| "step": 14350 | |
| }, | |
| { | |
| "epoch": 7.01904296875, | |
| "grad_norm": 3.62715744972229, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4379, | |
| "step": 14375 | |
| }, | |
| { | |
| "epoch": 7.03125, | |
| "grad_norm": 2.931847333908081, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3971, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 7.04345703125, | |
| "grad_norm": 3.470655679702759, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4529, | |
| "step": 14425 | |
| }, | |
| { | |
| "epoch": 7.0556640625, | |
| "grad_norm": 3.3625166416168213, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4197, | |
| "step": 14450 | |
| }, | |
| { | |
| "epoch": 7.06787109375, | |
| "grad_norm": 3.200965404510498, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4553, | |
| "step": 14475 | |
| }, | |
| { | |
| "epoch": 7.080078125, | |
| "grad_norm": 3.83722186088562, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4601, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 7.080078125, | |
| "eval_cer": 20.727272727272727, | |
| "eval_loss": 0.3850056231021881, | |
| "eval_normalized_cer": 16.399506781750926, | |
| "eval_runtime": 135.6446, | |
| "eval_samples_per_second": 0.944, | |
| "eval_steps_per_second": 0.059, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 7.09228515625, | |
| "grad_norm": 4.523788928985596, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4101, | |
| "step": 14525 | |
| }, | |
| { | |
| "epoch": 7.1044921875, | |
| "grad_norm": 3.117316722869873, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3924, | |
| "step": 14550 | |
| }, | |
| { | |
| "epoch": 7.11669921875, | |
| "grad_norm": 2.9018518924713135, | |
| "learning_rate": 1e-05, | |
| "loss": 0.421, | |
| "step": 14575 | |
| }, | |
| { | |
| "epoch": 7.12890625, | |
| "grad_norm": 4.131909370422363, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4576, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 7.14111328125, | |
| "grad_norm": 3.206359386444092, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4068, | |
| "step": 14625 | |
| }, | |
| { | |
| "epoch": 7.1533203125, | |
| "grad_norm": 4.659736156463623, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3906, | |
| "step": 14650 | |
| }, | |
| { | |
| "epoch": 7.16552734375, | |
| "grad_norm": 3.376612901687622, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4197, | |
| "step": 14675 | |
| }, | |
| { | |
| "epoch": 7.177734375, | |
| "grad_norm": 3.268672227859497, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4122, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 7.18994140625, | |
| "grad_norm": 3.0294389724731445, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4154, | |
| "step": 14725 | |
| }, | |
| { | |
| "epoch": 7.2021484375, | |
| "grad_norm": 3.547778367996216, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4375, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 7.21435546875, | |
| "grad_norm": 2.7753546237945557, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4212, | |
| "step": 14775 | |
| }, | |
| { | |
| "epoch": 7.2265625, | |
| "grad_norm": 3.724691152572632, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4074, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 7.23876953125, | |
| "grad_norm": 2.489899158477783, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4094, | |
| "step": 14825 | |
| }, | |
| { | |
| "epoch": 7.2509765625, | |
| "grad_norm": 3.542677879333496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4034, | |
| "step": 14850 | |
| }, | |
| { | |
| "epoch": 7.26318359375, | |
| "grad_norm": 3.15748929977417, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4553, | |
| "step": 14875 | |
| }, | |
| { | |
| "epoch": 7.275390625, | |
| "grad_norm": 3.477710008621216, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4394, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 7.28759765625, | |
| "grad_norm": 4.362137794494629, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4311, | |
| "step": 14925 | |
| }, | |
| { | |
| "epoch": 7.2998046875, | |
| "grad_norm": 2.3837854862213135, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4179, | |
| "step": 14950 | |
| }, | |
| { | |
| "epoch": 7.31201171875, | |
| "grad_norm": 3.1717562675476074, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4504, | |
| "step": 14975 | |
| }, | |
| { | |
| "epoch": 7.32421875, | |
| "grad_norm": 2.988471269607544, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4357, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 7.32421875, | |
| "eval_cer": 23.599999999999998, | |
| "eval_loss": 0.38635486364364624, | |
| "eval_normalized_cer": 20.098643649815042, | |
| "eval_runtime": 132.5727, | |
| "eval_samples_per_second": 0.966, | |
| "eval_steps_per_second": 0.06, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 7.33642578125, | |
| "grad_norm": 3.9990575313568115, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4263, | |
| "step": 15025 | |
| }, | |
| { | |
| "epoch": 7.3486328125, | |
| "grad_norm": 2.6988320350646973, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4311, | |
| "step": 15050 | |
| }, | |
| { | |
| "epoch": 7.36083984375, | |
| "grad_norm": 3.537297487258911, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4552, | |
| "step": 15075 | |
| }, | |
| { | |
| "epoch": 7.373046875, | |
| "grad_norm": 2.706217050552368, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4365, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 7.38525390625, | |
| "grad_norm": 3.2613677978515625, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4512, | |
| "step": 15125 | |
| }, | |
| { | |
| "epoch": 7.3974609375, | |
| "grad_norm": 3.039867877960205, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4485, | |
| "step": 15150 | |
| }, | |
| { | |
| "epoch": 7.40966796875, | |
| "grad_norm": 3.6644127368927, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4218, | |
| "step": 15175 | |
| }, | |
| { | |
| "epoch": 7.421875, | |
| "grad_norm": 3.3461201190948486, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4292, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 7.43408203125, | |
| "grad_norm": 2.8256707191467285, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4205, | |
| "step": 15225 | |
| }, | |
| { | |
| "epoch": 7.4462890625, | |
| "grad_norm": 3.3773632049560547, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4155, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 7.45849609375, | |
| "grad_norm": 2.6571528911590576, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4567, | |
| "step": 15275 | |
| }, | |
| { | |
| "epoch": 7.470703125, | |
| "grad_norm": 3.0483086109161377, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4293, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 7.48291015625, | |
| "grad_norm": 3.6026933193206787, | |
| "learning_rate": 1e-05, | |
| "loss": 0.426, | |
| "step": 15325 | |
| }, | |
| { | |
| "epoch": 7.4951171875, | |
| "grad_norm": 4.094547748565674, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4562, | |
| "step": 15350 | |
| }, | |
| { | |
| "epoch": 7.50732421875, | |
| "grad_norm": 3.6871438026428223, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4438, | |
| "step": 15375 | |
| }, | |
| { | |
| "epoch": 7.51953125, | |
| "grad_norm": 3.5841362476348877, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3874, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 7.53173828125, | |
| "grad_norm": 3.3375258445739746, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4234, | |
| "step": 15425 | |
| }, | |
| { | |
| "epoch": 7.5439453125, | |
| "grad_norm": 4.966141223907471, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4329, | |
| "step": 15450 | |
| }, | |
| { | |
| "epoch": 7.55615234375, | |
| "grad_norm": 2.2988944053649902, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4377, | |
| "step": 15475 | |
| }, | |
| { | |
| "epoch": 7.568359375, | |
| "grad_norm": 3.4808993339538574, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4375, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 7.568359375, | |
| "eval_cer": 23.854545454545452, | |
| "eval_loss": 0.38098400831222534, | |
| "eval_normalized_cer": 19.399917796958487, | |
| "eval_runtime": 137.4032, | |
| "eval_samples_per_second": 0.932, | |
| "eval_steps_per_second": 0.058, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 7.58056640625, | |
| "grad_norm": 4.054482936859131, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3687, | |
| "step": 15525 | |
| }, | |
| { | |
| "epoch": 7.5927734375, | |
| "grad_norm": 2.874267578125, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4033, | |
| "step": 15550 | |
| }, | |
| { | |
| "epoch": 7.60498046875, | |
| "grad_norm": 2.669652223587036, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4262, | |
| "step": 15575 | |
| }, | |
| { | |
| "epoch": 7.6171875, | |
| "grad_norm": 3.2882497310638428, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4163, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 7.62939453125, | |
| "grad_norm": 4.546947479248047, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4668, | |
| "step": 15625 | |
| }, | |
| { | |
| "epoch": 7.6416015625, | |
| "grad_norm": 2.2762033939361572, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4313, | |
| "step": 15650 | |
| }, | |
| { | |
| "epoch": 7.65380859375, | |
| "grad_norm": 3.4660661220550537, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4181, | |
| "step": 15675 | |
| }, | |
| { | |
| "epoch": 7.666015625, | |
| "grad_norm": 3.2215335369110107, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4114, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 7.67822265625, | |
| "grad_norm": 3.1161818504333496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4599, | |
| "step": 15725 | |
| }, | |
| { | |
| "epoch": 7.6904296875, | |
| "grad_norm": 3.0069620609283447, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4498, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 7.70263671875, | |
| "grad_norm": 2.84161114692688, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4252, | |
| "step": 15775 | |
| }, | |
| { | |
| "epoch": 7.71484375, | |
| "grad_norm": 4.6650590896606445, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4297, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 7.72705078125, | |
| "grad_norm": 2.6274569034576416, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3753, | |
| "step": 15825 | |
| }, | |
| { | |
| "epoch": 7.7392578125, | |
| "grad_norm": 3.5292246341705322, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4168, | |
| "step": 15850 | |
| }, | |
| { | |
| "epoch": 7.75146484375, | |
| "grad_norm": 2.8617982864379883, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4809, | |
| "step": 15875 | |
| }, | |
| { | |
| "epoch": 7.763671875, | |
| "grad_norm": 3.669074773788452, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4613, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 7.77587890625, | |
| "grad_norm": 2.994079113006592, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3902, | |
| "step": 15925 | |
| }, | |
| { | |
| "epoch": 7.7880859375, | |
| "grad_norm": 2.6492769718170166, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3756, | |
| "step": 15950 | |
| }, | |
| { | |
| "epoch": 7.80029296875, | |
| "grad_norm": 3.2743895053863525, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4476, | |
| "step": 15975 | |
| }, | |
| { | |
| "epoch": 7.8125, | |
| "grad_norm": 3.279031276702881, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4563, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 7.8125, | |
| "eval_cer": 23.381818181818183, | |
| "eval_loss": 0.38140711188316345, | |
| "eval_normalized_cer": 18.90669954788327, | |
| "eval_runtime": 135.0002, | |
| "eval_samples_per_second": 0.948, | |
| "eval_steps_per_second": 0.059, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 7.82470703125, | |
| "grad_norm": 2.7851359844207764, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4344, | |
| "step": 16025 | |
| }, | |
| { | |
| "epoch": 7.8369140625, | |
| "grad_norm": 3.965728282928467, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4328, | |
| "step": 16050 | |
| }, | |
| { | |
| "epoch": 7.84912109375, | |
| "grad_norm": 4.293801307678223, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4593, | |
| "step": 16075 | |
| }, | |
| { | |
| "epoch": 7.861328125, | |
| "grad_norm": 3.0265841484069824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4039, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 7.87353515625, | |
| "grad_norm": 3.9985835552215576, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3911, | |
| "step": 16125 | |
| }, | |
| { | |
| "epoch": 7.8857421875, | |
| "grad_norm": 4.016017436981201, | |
| "learning_rate": 1e-05, | |
| "loss": 0.458, | |
| "step": 16150 | |
| }, | |
| { | |
| "epoch": 7.89794921875, | |
| "grad_norm": 3.816105842590332, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4125, | |
| "step": 16175 | |
| }, | |
| { | |
| "epoch": 7.91015625, | |
| "grad_norm": 2.5083696842193604, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3769, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 7.92236328125, | |
| "grad_norm": 3.251554012298584, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4705, | |
| "step": 16225 | |
| }, | |
| { | |
| "epoch": 7.9345703125, | |
| "grad_norm": 3.370077610015869, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4565, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 7.94677734375, | |
| "grad_norm": 2.996699810028076, | |
| "learning_rate": 1e-05, | |
| "loss": 0.461, | |
| "step": 16275 | |
| }, | |
| { | |
| "epoch": 7.958984375, | |
| "grad_norm": 4.30794620513916, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4035, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 7.97119140625, | |
| "grad_norm": 3.3062832355499268, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4067, | |
| "step": 16325 | |
| }, | |
| { | |
| "epoch": 7.9833984375, | |
| "grad_norm": 2.4719371795654297, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4868, | |
| "step": 16350 | |
| }, | |
| { | |
| "epoch": 7.99560546875, | |
| "grad_norm": 3.1429994106292725, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4341, | |
| "step": 16375 | |
| }, | |
| { | |
| "epoch": 8.0078125, | |
| "grad_norm": 4.408067226409912, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4874, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 8.02001953125, | |
| "grad_norm": 3.3049044609069824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4783, | |
| "step": 16425 | |
| }, | |
| { | |
| "epoch": 8.0322265625, | |
| "grad_norm": 2.1289191246032715, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3918, | |
| "step": 16450 | |
| }, | |
| { | |
| "epoch": 8.04443359375, | |
| "grad_norm": 3.2784934043884277, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3953, | |
| "step": 16475 | |
| }, | |
| { | |
| "epoch": 8.056640625, | |
| "grad_norm": 2.8000524044036865, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4145, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 8.056640625, | |
| "eval_cer": 19.236363636363638, | |
| "eval_loss": 0.3803868889808655, | |
| "eval_normalized_cer": 15.207562679819153, | |
| "eval_runtime": 142.7418, | |
| "eval_samples_per_second": 0.897, | |
| "eval_steps_per_second": 0.056, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 8.06884765625, | |
| "grad_norm": 4.039844989776611, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4458, | |
| "step": 16525 | |
| }, | |
| { | |
| "epoch": 8.0810546875, | |
| "grad_norm": 3.5788333415985107, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4211, | |
| "step": 16550 | |
| }, | |
| { | |
| "epoch": 8.09326171875, | |
| "grad_norm": 2.7754452228546143, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3947, | |
| "step": 16575 | |
| }, | |
| { | |
| "epoch": 8.10546875, | |
| "grad_norm": 3.6685307025909424, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4371, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 8.11767578125, | |
| "grad_norm": 4.283354759216309, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4196, | |
| "step": 16625 | |
| }, | |
| { | |
| "epoch": 8.1298828125, | |
| "grad_norm": 3.37106990814209, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4286, | |
| "step": 16650 | |
| }, | |
| { | |
| "epoch": 8.14208984375, | |
| "grad_norm": 3.3839213848114014, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4081, | |
| "step": 16675 | |
| }, | |
| { | |
| "epoch": 8.154296875, | |
| "grad_norm": 2.9974148273468018, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4218, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 8.16650390625, | |
| "grad_norm": 2.861589193344116, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3866, | |
| "step": 16725 | |
| }, | |
| { | |
| "epoch": 8.1787109375, | |
| "grad_norm": 3.7804067134857178, | |
| "learning_rate": 1e-05, | |
| "loss": 0.402, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 8.19091796875, | |
| "grad_norm": 4.124424457550049, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4198, | |
| "step": 16775 | |
| }, | |
| { | |
| "epoch": 8.203125, | |
| "grad_norm": 3.785266876220703, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4744, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 8.21533203125, | |
| "grad_norm": 2.816866636276245, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4509, | |
| "step": 16825 | |
| }, | |
| { | |
| "epoch": 8.2275390625, | |
| "grad_norm": 2.5543909072875977, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4379, | |
| "step": 16850 | |
| }, | |
| { | |
| "epoch": 8.23974609375, | |
| "grad_norm": 3.561476707458496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4109, | |
| "step": 16875 | |
| }, | |
| { | |
| "epoch": 8.251953125, | |
| "grad_norm": 2.819086790084839, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4265, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 8.26416015625, | |
| "grad_norm": 3.327193021774292, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3561, | |
| "step": 16925 | |
| }, | |
| { | |
| "epoch": 8.2763671875, | |
| "grad_norm": 2.7188267707824707, | |
| "learning_rate": 1e-05, | |
| "loss": 0.405, | |
| "step": 16950 | |
| }, | |
| { | |
| "epoch": 8.28857421875, | |
| "grad_norm": 4.3515543937683105, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4097, | |
| "step": 16975 | |
| }, | |
| { | |
| "epoch": 8.30078125, | |
| "grad_norm": 3.3475663661956787, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3892, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 8.30078125, | |
| "eval_cer": 22.581818181818182, | |
| "eval_loss": 0.38389354944229126, | |
| "eval_normalized_cer": 18.16687217427045, | |
| "eval_runtime": 140.8271, | |
| "eval_samples_per_second": 0.909, | |
| "eval_steps_per_second": 0.057, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 8.31298828125, | |
| "grad_norm": 3.1067376136779785, | |
| "learning_rate": 1e-05, | |
| "loss": 0.402, | |
| "step": 17025 | |
| }, | |
| { | |
| "epoch": 8.3251953125, | |
| "grad_norm": 4.873087406158447, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4059, | |
| "step": 17050 | |
| }, | |
| { | |
| "epoch": 8.33740234375, | |
| "grad_norm": 3.115588665008545, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4014, | |
| "step": 17075 | |
| }, | |
| { | |
| "epoch": 8.349609375, | |
| "grad_norm": 4.50831413269043, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4337, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 8.36181640625, | |
| "grad_norm": 2.399965763092041, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4034, | |
| "step": 17125 | |
| }, | |
| { | |
| "epoch": 8.3740234375, | |
| "grad_norm": 3.1204044818878174, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4344, | |
| "step": 17150 | |
| }, | |
| { | |
| "epoch": 8.38623046875, | |
| "grad_norm": 3.457709789276123, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4311, | |
| "step": 17175 | |
| }, | |
| { | |
| "epoch": 8.3984375, | |
| "grad_norm": 2.5090980529785156, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4631, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 8.41064453125, | |
| "grad_norm": 2.9333646297454834, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4329, | |
| "step": 17225 | |
| }, | |
| { | |
| "epoch": 8.4228515625, | |
| "grad_norm": 2.8422629833221436, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3823, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 8.43505859375, | |
| "grad_norm": 2.7649292945861816, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4538, | |
| "step": 17275 | |
| }, | |
| { | |
| "epoch": 8.447265625, | |
| "grad_norm": 3.3747799396514893, | |
| "learning_rate": 1e-05, | |
| "loss": 0.416, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 8.45947265625, | |
| "grad_norm": 4.490537643432617, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3744, | |
| "step": 17325 | |
| }, | |
| { | |
| "epoch": 8.4716796875, | |
| "grad_norm": 3.027010202407837, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3874, | |
| "step": 17350 | |
| }, | |
| { | |
| "epoch": 8.48388671875, | |
| "grad_norm": 2.616095542907715, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3695, | |
| "step": 17375 | |
| }, | |
| { | |
| "epoch": 8.49609375, | |
| "grad_norm": 3.389639377593994, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4259, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 8.50830078125, | |
| "grad_norm": 4.086320400238037, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4047, | |
| "step": 17425 | |
| }, | |
| { | |
| "epoch": 8.5205078125, | |
| "grad_norm": 2.4645962715148926, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4357, | |
| "step": 17450 | |
| }, | |
| { | |
| "epoch": 8.53271484375, | |
| "grad_norm": 2.2121071815490723, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4381, | |
| "step": 17475 | |
| }, | |
| { | |
| "epoch": 8.544921875, | |
| "grad_norm": 2.978116512298584, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3699, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 8.544921875, | |
| "eval_cer": 20.0, | |
| "eval_loss": 0.382442444562912, | |
| "eval_normalized_cer": 16.276202219482123, | |
| "eval_runtime": 146.3607, | |
| "eval_samples_per_second": 0.875, | |
| "eval_steps_per_second": 0.055, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 8.55712890625, | |
| "grad_norm": 3.6570897102355957, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4495, | |
| "step": 17525 | |
| }, | |
| { | |
| "epoch": 8.5693359375, | |
| "grad_norm": 3.3602068424224854, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4132, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 8.58154296875, | |
| "grad_norm": 3.4608097076416016, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4641, | |
| "step": 17575 | |
| }, | |
| { | |
| "epoch": 8.59375, | |
| "grad_norm": 3.122091293334961, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4325, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 8.60595703125, | |
| "grad_norm": 4.238564491271973, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4002, | |
| "step": 17625 | |
| }, | |
| { | |
| "epoch": 8.6181640625, | |
| "grad_norm": 3.2410008907318115, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4678, | |
| "step": 17650 | |
| }, | |
| { | |
| "epoch": 8.63037109375, | |
| "grad_norm": 2.8786799907684326, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3947, | |
| "step": 17675 | |
| }, | |
| { | |
| "epoch": 8.642578125, | |
| "grad_norm": 2.8009581565856934, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4154, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 8.65478515625, | |
| "grad_norm": 3.148653507232666, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3856, | |
| "step": 17725 | |
| }, | |
| { | |
| "epoch": 8.6669921875, | |
| "grad_norm": 3.803799867630005, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4285, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 8.67919921875, | |
| "grad_norm": 3.944755792617798, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4598, | |
| "step": 17775 | |
| }, | |
| { | |
| "epoch": 8.69140625, | |
| "grad_norm": 3.1085519790649414, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4379, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 8.70361328125, | |
| "grad_norm": 5.655707836151123, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4121, | |
| "step": 17825 | |
| }, | |
| { | |
| "epoch": 8.7158203125, | |
| "grad_norm": 2.179668664932251, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4212, | |
| "step": 17850 | |
| }, | |
| { | |
| "epoch": 8.72802734375, | |
| "grad_norm": 3.687593460083008, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4643, | |
| "step": 17875 | |
| }, | |
| { | |
| "epoch": 8.740234375, | |
| "grad_norm": 2.9461138248443604, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4126, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 8.75244140625, | |
| "grad_norm": 3.524925470352173, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4407, | |
| "step": 17925 | |
| }, | |
| { | |
| "epoch": 8.7646484375, | |
| "grad_norm": 4.609610080718994, | |
| "learning_rate": 1e-05, | |
| "loss": 0.42, | |
| "step": 17950 | |
| }, | |
| { | |
| "epoch": 8.77685546875, | |
| "grad_norm": 4.5199079513549805, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3921, | |
| "step": 17975 | |
| }, | |
| { | |
| "epoch": 8.7890625, | |
| "grad_norm": 3.358597993850708, | |
| "learning_rate": 1e-05, | |
| "loss": 0.441, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 8.7890625, | |
| "eval_cer": 21.854545454545455, | |
| "eval_loss": 0.38204386830329895, | |
| "eval_normalized_cer": 16.358405260994658, | |
| "eval_runtime": 141.6466, | |
| "eval_samples_per_second": 0.904, | |
| "eval_steps_per_second": 0.056, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 8.80126953125, | |
| "grad_norm": 3.380112648010254, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3832, | |
| "step": 18025 | |
| }, | |
| { | |
| "epoch": 8.8134765625, | |
| "grad_norm": 3.249871015548706, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4449, | |
| "step": 18050 | |
| }, | |
| { | |
| "epoch": 8.82568359375, | |
| "grad_norm": 3.3038084506988525, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4919, | |
| "step": 18075 | |
| }, | |
| { | |
| "epoch": 8.837890625, | |
| "grad_norm": 3.6768815517425537, | |
| "learning_rate": 1e-05, | |
| "loss": 0.42, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 8.85009765625, | |
| "grad_norm": 2.7637171745300293, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4275, | |
| "step": 18125 | |
| }, | |
| { | |
| "epoch": 8.8623046875, | |
| "grad_norm": 2.896497964859009, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3999, | |
| "step": 18150 | |
| }, | |
| { | |
| "epoch": 8.87451171875, | |
| "grad_norm": 2.759514570236206, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4459, | |
| "step": 18175 | |
| }, | |
| { | |
| "epoch": 8.88671875, | |
| "grad_norm": 3.696629762649536, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5002, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 8.89892578125, | |
| "grad_norm": 2.6874115467071533, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4143, | |
| "step": 18225 | |
| }, | |
| { | |
| "epoch": 8.9111328125, | |
| "grad_norm": 3.066502809524536, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3943, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 8.92333984375, | |
| "grad_norm": 3.2830264568328857, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3878, | |
| "step": 18275 | |
| }, | |
| { | |
| "epoch": 8.935546875, | |
| "grad_norm": 3.5021724700927734, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4332, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 8.94775390625, | |
| "grad_norm": 4.337588310241699, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4256, | |
| "step": 18325 | |
| }, | |
| { | |
| "epoch": 8.9599609375, | |
| "grad_norm": 3.243098258972168, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4593, | |
| "step": 18350 | |
| }, | |
| { | |
| "epoch": 8.97216796875, | |
| "grad_norm": 2.650510787963867, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3951, | |
| "step": 18375 | |
| }, | |
| { | |
| "epoch": 8.984375, | |
| "grad_norm": 3.37617564201355, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4143, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 8.99658203125, | |
| "grad_norm": 3.7461116313934326, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4147, | |
| "step": 18425 | |
| }, | |
| { | |
| "epoch": 9.0087890625, | |
| "grad_norm": 4.165828704833984, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4538, | |
| "step": 18450 | |
| }, | |
| { | |
| "epoch": 9.02099609375, | |
| "grad_norm": 3.3768396377563477, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4128, | |
| "step": 18475 | |
| }, | |
| { | |
| "epoch": 9.033203125, | |
| "grad_norm": 3.006622314453125, | |
| "learning_rate": 1e-05, | |
| "loss": 0.427, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 9.033203125, | |
| "eval_cer": 21.054545454545455, | |
| "eval_loss": 0.3800523281097412, | |
| "eval_normalized_cer": 16.85162351006987, | |
| "eval_runtime": 92.9904, | |
| "eval_samples_per_second": 1.376, | |
| "eval_steps_per_second": 0.086, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 9.04541015625, | |
| "grad_norm": 2.577570915222168, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4233, | |
| "step": 18525 | |
| }, | |
| { | |
| "epoch": 9.0576171875, | |
| "grad_norm": 3.2769463062286377, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4209, | |
| "step": 18550 | |
| }, | |
| { | |
| "epoch": 9.06982421875, | |
| "grad_norm": 3.296905040740967, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4576, | |
| "step": 18575 | |
| }, | |
| { | |
| "epoch": 9.08203125, | |
| "grad_norm": 2.849393367767334, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4033, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 9.09423828125, | |
| "grad_norm": 4.293519496917725, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4604, | |
| "step": 18625 | |
| }, | |
| { | |
| "epoch": 9.1064453125, | |
| "grad_norm": 3.5369253158569336, | |
| "learning_rate": 1e-05, | |
| "loss": 0.444, | |
| "step": 18650 | |
| }, | |
| { | |
| "epoch": 9.11865234375, | |
| "grad_norm": 4.512628555297852, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4008, | |
| "step": 18675 | |
| }, | |
| { | |
| "epoch": 9.130859375, | |
| "grad_norm": 3.0092594623565674, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4219, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 9.14306640625, | |
| "grad_norm": 3.3797607421875, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4102, | |
| "step": 18725 | |
| }, | |
| { | |
| "epoch": 9.1552734375, | |
| "grad_norm": 3.1966211795806885, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4152, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 9.16748046875, | |
| "grad_norm": 3.016399383544922, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4277, | |
| "step": 18775 | |
| }, | |
| { | |
| "epoch": 9.1796875, | |
| "grad_norm": 3.313685417175293, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3932, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 9.19189453125, | |
| "grad_norm": 2.8252785205841064, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4041, | |
| "step": 18825 | |
| }, | |
| { | |
| "epoch": 9.2041015625, | |
| "grad_norm": 2.935676097869873, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4107, | |
| "step": 18850 | |
| }, | |
| { | |
| "epoch": 9.21630859375, | |
| "grad_norm": 2.280316114425659, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4701, | |
| "step": 18875 | |
| }, | |
| { | |
| "epoch": 9.228515625, | |
| "grad_norm": 2.718478202819824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4174, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 9.24072265625, | |
| "grad_norm": 4.804378986358643, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3947, | |
| "step": 18925 | |
| }, | |
| { | |
| "epoch": 9.2529296875, | |
| "grad_norm": 4.070915222167969, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4158, | |
| "step": 18950 | |
| }, | |
| { | |
| "epoch": 9.26513671875, | |
| "grad_norm": 3.2516276836395264, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4379, | |
| "step": 18975 | |
| }, | |
| { | |
| "epoch": 9.27734375, | |
| "grad_norm": 3.5097761154174805, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3767, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 9.27734375, | |
| "eval_cer": 24.254545454545454, | |
| "eval_loss": 0.36628666520118713, | |
| "eval_normalized_cer": 18.701191944101932, | |
| "eval_runtime": 93.7861, | |
| "eval_samples_per_second": 1.365, | |
| "eval_steps_per_second": 0.085, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 9.28955078125, | |
| "grad_norm": 4.052979469299316, | |
| "learning_rate": 1e-05, | |
| "loss": 0.446, | |
| "step": 19025 | |
| }, | |
| { | |
| "epoch": 9.3017578125, | |
| "grad_norm": 1.887149691581726, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3933, | |
| "step": 19050 | |
| }, | |
| { | |
| "epoch": 9.31396484375, | |
| "grad_norm": 3.2341785430908203, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4361, | |
| "step": 19075 | |
| }, | |
| { | |
| "epoch": 9.326171875, | |
| "grad_norm": 2.683950185775757, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4107, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 9.33837890625, | |
| "grad_norm": 3.6661105155944824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.396, | |
| "step": 19125 | |
| }, | |
| { | |
| "epoch": 9.3505859375, | |
| "grad_norm": 2.379519462585449, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3875, | |
| "step": 19150 | |
| }, | |
| { | |
| "epoch": 9.36279296875, | |
| "grad_norm": 2.952665090560913, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4471, | |
| "step": 19175 | |
| }, | |
| { | |
| "epoch": 9.375, | |
| "grad_norm": 3.1039767265319824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4052, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 9.38720703125, | |
| "grad_norm": 3.941380023956299, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3908, | |
| "step": 19225 | |
| }, | |
| { | |
| "epoch": 9.3994140625, | |
| "grad_norm": 2.511928081512451, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4215, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 9.41162109375, | |
| "grad_norm": 2.6684021949768066, | |
| "learning_rate": 1e-05, | |
| "loss": 0.409, | |
| "step": 19275 | |
| }, | |
| { | |
| "epoch": 9.423828125, | |
| "grad_norm": 3.9744958877563477, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4778, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 9.43603515625, | |
| "grad_norm": 2.601891279220581, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4059, | |
| "step": 19325 | |
| }, | |
| { | |
| "epoch": 9.4482421875, | |
| "grad_norm": 3.2740561962127686, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4299, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 9.46044921875, | |
| "grad_norm": 4.9015889167785645, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4427, | |
| "step": 19375 | |
| }, | |
| { | |
| "epoch": 9.47265625, | |
| "grad_norm": 2.882230281829834, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3959, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 9.48486328125, | |
| "grad_norm": 4.404541969299316, | |
| "learning_rate": 1e-05, | |
| "loss": 0.414, | |
| "step": 19425 | |
| }, | |
| { | |
| "epoch": 9.4970703125, | |
| "grad_norm": 3.5113275051116943, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3937, | |
| "step": 19450 | |
| }, | |
| { | |
| "epoch": 9.50927734375, | |
| "grad_norm": 2.4990580081939697, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4262, | |
| "step": 19475 | |
| }, | |
| { | |
| "epoch": 9.521484375, | |
| "grad_norm": 3.3028454780578613, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4519, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 9.521484375, | |
| "eval_cer": 20.472727272727273, | |
| "eval_loss": 0.37869974970817566, | |
| "eval_normalized_cer": 15.454171804356761, | |
| "eval_runtime": 97.3092, | |
| "eval_samples_per_second": 1.315, | |
| "eval_steps_per_second": 0.082, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 9.53369140625, | |
| "grad_norm": 3.5164365768432617, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4119, | |
| "step": 19525 | |
| }, | |
| { | |
| "epoch": 9.5458984375, | |
| "grad_norm": 3.049210786819458, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4198, | |
| "step": 19550 | |
| }, | |
| { | |
| "epoch": 9.55810546875, | |
| "grad_norm": 2.659877300262451, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4047, | |
| "step": 19575 | |
| }, | |
| { | |
| "epoch": 9.5703125, | |
| "grad_norm": 2.617429733276367, | |
| "learning_rate": 1e-05, | |
| "loss": 0.446, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 9.58251953125, | |
| "grad_norm": 3.912100076675415, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4169, | |
| "step": 19625 | |
| }, | |
| { | |
| "epoch": 9.5947265625, | |
| "grad_norm": 2.6492602825164795, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4215, | |
| "step": 19650 | |
| }, | |
| { | |
| "epoch": 9.60693359375, | |
| "grad_norm": 3.404269218444824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3782, | |
| "step": 19675 | |
| }, | |
| { | |
| "epoch": 9.619140625, | |
| "grad_norm": 2.4062957763671875, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4342, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 9.63134765625, | |
| "grad_norm": 2.835569143295288, | |
| "learning_rate": 1e-05, | |
| "loss": 0.372, | |
| "step": 19725 | |
| }, | |
| { | |
| "epoch": 9.6435546875, | |
| "grad_norm": 2.3654327392578125, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4095, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 9.65576171875, | |
| "grad_norm": 2.8574941158294678, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4225, | |
| "step": 19775 | |
| }, | |
| { | |
| "epoch": 9.66796875, | |
| "grad_norm": 3.122192859649658, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4372, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 9.68017578125, | |
| "grad_norm": 2.591271162033081, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4523, | |
| "step": 19825 | |
| }, | |
| { | |
| "epoch": 9.6923828125, | |
| "grad_norm": 3.0938827991485596, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4055, | |
| "step": 19850 | |
| }, | |
| { | |
| "epoch": 9.70458984375, | |
| "grad_norm": 3.643186569213867, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4066, | |
| "step": 19875 | |
| }, | |
| { | |
| "epoch": 9.716796875, | |
| "grad_norm": 3.9704155921936035, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4615, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 9.72900390625, | |
| "grad_norm": 2.1930134296417236, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4453, | |
| "step": 19925 | |
| }, | |
| { | |
| "epoch": 9.7412109375, | |
| "grad_norm": 3.3762011528015137, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3818, | |
| "step": 19950 | |
| }, | |
| { | |
| "epoch": 9.75341796875, | |
| "grad_norm": 3.172435998916626, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4118, | |
| "step": 19975 | |
| }, | |
| { | |
| "epoch": 9.765625, | |
| "grad_norm": 3.6243398189544678, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4184, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 9.765625, | |
| "eval_cer": 21.236363636363638, | |
| "eval_loss": 0.37908729910850525, | |
| "eval_normalized_cer": 16.85162351006987, | |
| "eval_runtime": 97.9066, | |
| "eval_samples_per_second": 1.307, | |
| "eval_steps_per_second": 0.082, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 9.77783203125, | |
| "grad_norm": 3.5656485557556152, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4162, | |
| "step": 20025 | |
| }, | |
| { | |
| "epoch": 9.7900390625, | |
| "grad_norm": 3.304313898086548, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3687, | |
| "step": 20050 | |
| }, | |
| { | |
| "epoch": 9.80224609375, | |
| "grad_norm": 2.4108333587646484, | |
| "learning_rate": 1e-05, | |
| "loss": 0.364, | |
| "step": 20075 | |
| }, | |
| { | |
| "epoch": 9.814453125, | |
| "grad_norm": 3.621305465698242, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4071, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 9.82666015625, | |
| "grad_norm": 3.3533174991607666, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4386, | |
| "step": 20125 | |
| }, | |
| { | |
| "epoch": 9.8388671875, | |
| "grad_norm": 2.8671376705169678, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4299, | |
| "step": 20150 | |
| }, | |
| { | |
| "epoch": 9.85107421875, | |
| "grad_norm": 3.2498745918273926, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4011, | |
| "step": 20175 | |
| }, | |
| { | |
| "epoch": 9.86328125, | |
| "grad_norm": 3.473806381225586, | |
| "learning_rate": 1e-05, | |
| "loss": 0.441, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 9.87548828125, | |
| "grad_norm": 2.80523419380188, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4404, | |
| "step": 20225 | |
| }, | |
| { | |
| "epoch": 9.8876953125, | |
| "grad_norm": 2.332698345184326, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4269, | |
| "step": 20250 | |
| }, | |
| { | |
| "epoch": 9.89990234375, | |
| "grad_norm": 2.8714382648468018, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4263, | |
| "step": 20275 | |
| }, | |
| { | |
| "epoch": 9.912109375, | |
| "grad_norm": 3.2583768367767334, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4206, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 9.92431640625, | |
| "grad_norm": 3.4748706817626953, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4348, | |
| "step": 20325 | |
| }, | |
| { | |
| "epoch": 9.9365234375, | |
| "grad_norm": 2.696352243423462, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4237, | |
| "step": 20350 | |
| }, | |
| { | |
| "epoch": 9.94873046875, | |
| "grad_norm": 3.5071768760681152, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4521, | |
| "step": 20375 | |
| }, | |
| { | |
| "epoch": 9.9609375, | |
| "grad_norm": 3.7230043411254883, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4387, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 9.97314453125, | |
| "grad_norm": 4.502689361572266, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4303, | |
| "step": 20425 | |
| }, | |
| { | |
| "epoch": 9.9853515625, | |
| "grad_norm": 2.849353790283203, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3952, | |
| "step": 20450 | |
| }, | |
| { | |
| "epoch": 9.99755859375, | |
| "grad_norm": 3.3098607063293457, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3903, | |
| "step": 20475 | |
| }, | |
| { | |
| "epoch": 10.009765625, | |
| "grad_norm": 2.970693349838257, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3752, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 10.009765625, | |
| "eval_cer": 24.581818181818182, | |
| "eval_loss": 0.3705242872238159, | |
| "eval_normalized_cer": 19.31771475544595, | |
| "eval_runtime": 88.6677, | |
| "eval_samples_per_second": 1.444, | |
| "eval_steps_per_second": 0.09, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 10.02197265625, | |
| "grad_norm": 2.900347948074341, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4193, | |
| "step": 20525 | |
| }, | |
| { | |
| "epoch": 10.0341796875, | |
| "grad_norm": 2.864896297454834, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4096, | |
| "step": 20550 | |
| }, | |
| { | |
| "epoch": 10.04638671875, | |
| "grad_norm": 2.738795518875122, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4447, | |
| "step": 20575 | |
| }, | |
| { | |
| "epoch": 10.05859375, | |
| "grad_norm": 2.6308414936065674, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3794, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 10.07080078125, | |
| "grad_norm": 2.4290077686309814, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4009, | |
| "step": 20625 | |
| }, | |
| { | |
| "epoch": 10.0830078125, | |
| "grad_norm": 2.612316846847534, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3893, | |
| "step": 20650 | |
| }, | |
| { | |
| "epoch": 10.09521484375, | |
| "grad_norm": 3.2632391452789307, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3999, | |
| "step": 20675 | |
| }, | |
| { | |
| "epoch": 10.107421875, | |
| "grad_norm": 2.5368306636810303, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4016, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 10.11962890625, | |
| "grad_norm": 3.5920157432556152, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4465, | |
| "step": 20725 | |
| }, | |
| { | |
| "epoch": 10.1318359375, | |
| "grad_norm": 2.3094773292541504, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4074, | |
| "step": 20750 | |
| }, | |
| { | |
| "epoch": 10.14404296875, | |
| "grad_norm": 2.4848082065582275, | |
| "learning_rate": 1e-05, | |
| "loss": 0.381, | |
| "step": 20775 | |
| }, | |
| { | |
| "epoch": 10.15625, | |
| "grad_norm": 3.432950496673584, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3846, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 10.16845703125, | |
| "grad_norm": 3.441004514694214, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4343, | |
| "step": 20825 | |
| }, | |
| { | |
| "epoch": 10.1806640625, | |
| "grad_norm": 1.9125665426254272, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3822, | |
| "step": 20850 | |
| }, | |
| { | |
| "epoch": 10.19287109375, | |
| "grad_norm": 2.5689966678619385, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4069, | |
| "step": 20875 | |
| }, | |
| { | |
| "epoch": 10.205078125, | |
| "grad_norm": 2.4284005165100098, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4481, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 10.21728515625, | |
| "grad_norm": 2.9357855319976807, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3985, | |
| "step": 20925 | |
| }, | |
| { | |
| "epoch": 10.2294921875, | |
| "grad_norm": 3.1418330669403076, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4334, | |
| "step": 20950 | |
| }, | |
| { | |
| "epoch": 10.24169921875, | |
| "grad_norm": 3.525568723678589, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4659, | |
| "step": 20975 | |
| }, | |
| { | |
| "epoch": 10.25390625, | |
| "grad_norm": 3.033616304397583, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4532, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 10.25390625, | |
| "eval_cer": 22.363636363636363, | |
| "eval_loss": 0.3776665925979614, | |
| "eval_normalized_cer": 18.372379778051787, | |
| "eval_runtime": 88.8963, | |
| "eval_samples_per_second": 1.44, | |
| "eval_steps_per_second": 0.09, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 10.26611328125, | |
| "grad_norm": 3.0008301734924316, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4229, | |
| "step": 21025 | |
| }, | |
| { | |
| "epoch": 10.2783203125, | |
| "grad_norm": 4.1073102951049805, | |
| "learning_rate": 1e-05, | |
| "loss": 0.43, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 10.29052734375, | |
| "grad_norm": 2.6684110164642334, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4007, | |
| "step": 21075 | |
| }, | |
| { | |
| "epoch": 10.302734375, | |
| "grad_norm": 2.416282892227173, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4006, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 10.31494140625, | |
| "grad_norm": 3.2998218536376953, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4207, | |
| "step": 21125 | |
| }, | |
| { | |
| "epoch": 10.3271484375, | |
| "grad_norm": 4.176929473876953, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4559, | |
| "step": 21150 | |
| }, | |
| { | |
| "epoch": 10.33935546875, | |
| "grad_norm": 2.522796154022217, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4052, | |
| "step": 21175 | |
| }, | |
| { | |
| "epoch": 10.3515625, | |
| "grad_norm": 2.9196386337280273, | |
| "learning_rate": 1e-05, | |
| "loss": 0.414, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 10.36376953125, | |
| "grad_norm": 2.877315044403076, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4454, | |
| "step": 21225 | |
| }, | |
| { | |
| "epoch": 10.3759765625, | |
| "grad_norm": 3.0172317028045654, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4294, | |
| "step": 21250 | |
| }, | |
| { | |
| "epoch": 10.38818359375, | |
| "grad_norm": 3.5081894397735596, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4283, | |
| "step": 21275 | |
| }, | |
| { | |
| "epoch": 10.400390625, | |
| "grad_norm": 2.136151075363159, | |
| "learning_rate": 1e-05, | |
| "loss": 0.429, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 10.41259765625, | |
| "grad_norm": 3.0397658348083496, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4019, | |
| "step": 21325 | |
| }, | |
| { | |
| "epoch": 10.4248046875, | |
| "grad_norm": 3.9129014015197754, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4128, | |
| "step": 21350 | |
| }, | |
| { | |
| "epoch": 10.43701171875, | |
| "grad_norm": 3.1991934776306152, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4, | |
| "step": 21375 | |
| }, | |
| { | |
| "epoch": 10.44921875, | |
| "grad_norm": 2.850768804550171, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4492, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 10.46142578125, | |
| "grad_norm": 3.6122255325317383, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3692, | |
| "step": 21425 | |
| }, | |
| { | |
| "epoch": 10.4736328125, | |
| "grad_norm": 3.7408318519592285, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4239, | |
| "step": 21450 | |
| }, | |
| { | |
| "epoch": 10.48583984375, | |
| "grad_norm": 2.4114573001861572, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3595, | |
| "step": 21475 | |
| }, | |
| { | |
| "epoch": 10.498046875, | |
| "grad_norm": 3.060419797897339, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3665, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 10.498046875, | |
| "eval_cer": 20.945454545454545, | |
| "eval_loss": 0.3751263916492462, | |
| "eval_normalized_cer": 17.13933415536375, | |
| "eval_runtime": 91.0253, | |
| "eval_samples_per_second": 1.406, | |
| "eval_steps_per_second": 0.088, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 10.51025390625, | |
| "grad_norm": 2.921161413192749, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4385, | |
| "step": 21525 | |
| }, | |
| { | |
| "epoch": 10.5224609375, | |
| "grad_norm": 3.127946376800537, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4185, | |
| "step": 21550 | |
| }, | |
| { | |
| "epoch": 10.53466796875, | |
| "grad_norm": 2.4071595668792725, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4172, | |
| "step": 21575 | |
| }, | |
| { | |
| "epoch": 10.546875, | |
| "grad_norm": 3.3871352672576904, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3781, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 10.55908203125, | |
| "grad_norm": 2.465686082839966, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4021, | |
| "step": 21625 | |
| }, | |
| { | |
| "epoch": 10.5712890625, | |
| "grad_norm": 3.5105502605438232, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3874, | |
| "step": 21650 | |
| }, | |
| { | |
| "epoch": 10.58349609375, | |
| "grad_norm": 3.2150614261627197, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4203, | |
| "step": 21675 | |
| }, | |
| { | |
| "epoch": 10.595703125, | |
| "grad_norm": 2.755958080291748, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3961, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 10.60791015625, | |
| "grad_norm": 2.6068344116210938, | |
| "learning_rate": 1e-05, | |
| "loss": 0.425, | |
| "step": 21725 | |
| }, | |
| { | |
| "epoch": 10.6201171875, | |
| "grad_norm": 4.860282897949219, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4188, | |
| "step": 21750 | |
| }, | |
| { | |
| "epoch": 10.63232421875, | |
| "grad_norm": 3.154463529586792, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4108, | |
| "step": 21775 | |
| }, | |
| { | |
| "epoch": 10.64453125, | |
| "grad_norm": 3.5040194988250732, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4295, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 10.65673828125, | |
| "grad_norm": 3.5046403408050537, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4306, | |
| "step": 21825 | |
| }, | |
| { | |
| "epoch": 10.6689453125, | |
| "grad_norm": 3.5825250148773193, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3991, | |
| "step": 21850 | |
| }, | |
| { | |
| "epoch": 10.68115234375, | |
| "grad_norm": 3.1898937225341797, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4365, | |
| "step": 21875 | |
| }, | |
| { | |
| "epoch": 10.693359375, | |
| "grad_norm": 2.6460320949554443, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4053, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 10.70556640625, | |
| "grad_norm": 2.7370963096618652, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3961, | |
| "step": 21925 | |
| }, | |
| { | |
| "epoch": 10.7177734375, | |
| "grad_norm": 2.502699851989746, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3943, | |
| "step": 21950 | |
| }, | |
| { | |
| "epoch": 10.72998046875, | |
| "grad_norm": 3.806084632873535, | |
| "learning_rate": 1e-05, | |
| "loss": 0.424, | |
| "step": 21975 | |
| }, | |
| { | |
| "epoch": 10.7421875, | |
| "grad_norm": 2.743624687194824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3897, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 10.7421875, | |
| "eval_cer": 22.763636363636365, | |
| "eval_loss": 0.3785582482814789, | |
| "eval_normalized_cer": 18.53678586107686, | |
| "eval_runtime": 89.903, | |
| "eval_samples_per_second": 1.424, | |
| "eval_steps_per_second": 0.089, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 10.75439453125, | |
| "grad_norm": 3.1976799964904785, | |
| "learning_rate": 1e-05, | |
| "loss": 0.439, | |
| "step": 22025 | |
| }, | |
| { | |
| "epoch": 10.7666015625, | |
| "grad_norm": 2.4848337173461914, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4722, | |
| "step": 22050 | |
| }, | |
| { | |
| "epoch": 10.77880859375, | |
| "grad_norm": 2.554133176803589, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4607, | |
| "step": 22075 | |
| }, | |
| { | |
| "epoch": 10.791015625, | |
| "grad_norm": 3.2724106311798096, | |
| "learning_rate": 1e-05, | |
| "loss": 0.429, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 10.80322265625, | |
| "grad_norm": 2.920471668243408, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3968, | |
| "step": 22125 | |
| }, | |
| { | |
| "epoch": 10.8154296875, | |
| "grad_norm": 3.323936700820923, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4126, | |
| "step": 22150 | |
| }, | |
| { | |
| "epoch": 10.82763671875, | |
| "grad_norm": 3.5896575450897217, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3986, | |
| "step": 22175 | |
| }, | |
| { | |
| "epoch": 10.83984375, | |
| "grad_norm": 3.3386435508728027, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3979, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 10.85205078125, | |
| "grad_norm": 3.4217638969421387, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3928, | |
| "step": 22225 | |
| }, | |
| { | |
| "epoch": 10.8642578125, | |
| "grad_norm": 2.90092134475708, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3813, | |
| "step": 22250 | |
| }, | |
| { | |
| "epoch": 10.87646484375, | |
| "grad_norm": 2.9017395973205566, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4061, | |
| "step": 22275 | |
| }, | |
| { | |
| "epoch": 10.888671875, | |
| "grad_norm": 3.1124579906463623, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3933, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 10.90087890625, | |
| "grad_norm": 2.763261079788208, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3751, | |
| "step": 22325 | |
| }, | |
| { | |
| "epoch": 10.9130859375, | |
| "grad_norm": 3.8521738052368164, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4537, | |
| "step": 22350 | |
| }, | |
| { | |
| "epoch": 10.92529296875, | |
| "grad_norm": 3.3494768142700195, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3479, | |
| "step": 22375 | |
| }, | |
| { | |
| "epoch": 10.9375, | |
| "grad_norm": 3.752168655395508, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4393, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 10.94970703125, | |
| "grad_norm": 2.980929374694824, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4376, | |
| "step": 22425 | |
| }, | |
| { | |
| "epoch": 10.9619140625, | |
| "grad_norm": 2.931426525115967, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4006, | |
| "step": 22450 | |
| }, | |
| { | |
| "epoch": 10.97412109375, | |
| "grad_norm": 2.798682451248169, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3977, | |
| "step": 22475 | |
| }, | |
| { | |
| "epoch": 10.986328125, | |
| "grad_norm": 3.374363422393799, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3846, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 10.986328125, | |
| "eval_cer": 21.12727272727273, | |
| "eval_loss": 0.3707554340362549, | |
| "eval_normalized_cer": 16.93382655158241, | |
| "eval_runtime": 90.713, | |
| "eval_samples_per_second": 1.411, | |
| "eval_steps_per_second": 0.088, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 10.99853515625, | |
| "grad_norm": 2.9657630920410156, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4397, | |
| "step": 22525 | |
| }, | |
| { | |
| "epoch": 11.0107421875, | |
| "grad_norm": 3.373459577560425, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4813, | |
| "step": 22550 | |
| }, | |
| { | |
| "epoch": 11.02294921875, | |
| "grad_norm": 2.293248176574707, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3939, | |
| "step": 22575 | |
| }, | |
| { | |
| "epoch": 11.03515625, | |
| "grad_norm": 3.1755404472351074, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4173, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 11.04736328125, | |
| "grad_norm": 4.223100185394287, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3965, | |
| "step": 22625 | |
| }, | |
| { | |
| "epoch": 11.0595703125, | |
| "grad_norm": 4.098690986633301, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4371, | |
| "step": 22650 | |
| }, | |
| { | |
| "epoch": 11.07177734375, | |
| "grad_norm": 2.937412977218628, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4287, | |
| "step": 22675 | |
| }, | |
| { | |
| "epoch": 11.083984375, | |
| "grad_norm": 2.240269422531128, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4255, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 11.09619140625, | |
| "grad_norm": 3.1162662506103516, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4157, | |
| "step": 22725 | |
| }, | |
| { | |
| "epoch": 11.1083984375, | |
| "grad_norm": 2.875746726989746, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3941, | |
| "step": 22750 | |
| }, | |
| { | |
| "epoch": 11.12060546875, | |
| "grad_norm": 2.4718353748321533, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3975, | |
| "step": 22775 | |
| }, | |
| { | |
| "epoch": 11.1328125, | |
| "grad_norm": 3.1226015090942383, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4033, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 11.14501953125, | |
| "grad_norm": 3.384474515914917, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4701, | |
| "step": 22825 | |
| }, | |
| { | |
| "epoch": 11.1572265625, | |
| "grad_norm": 3.434800148010254, | |
| "learning_rate": 1e-05, | |
| "loss": 0.404, | |
| "step": 22850 | |
| }, | |
| { | |
| "epoch": 11.16943359375, | |
| "grad_norm": 2.6356847286224365, | |
| "learning_rate": 1e-05, | |
| "loss": 0.408, | |
| "step": 22875 | |
| }, | |
| { | |
| "epoch": 11.181640625, | |
| "grad_norm": 2.8493409156799316, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3589, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 11.19384765625, | |
| "grad_norm": 3.3821358680725098, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4357, | |
| "step": 22925 | |
| }, | |
| { | |
| "epoch": 11.2060546875, | |
| "grad_norm": 2.6961047649383545, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3989, | |
| "step": 22950 | |
| }, | |
| { | |
| "epoch": 11.21826171875, | |
| "grad_norm": 2.9927561283111572, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4523, | |
| "step": 22975 | |
| }, | |
| { | |
| "epoch": 11.23046875, | |
| "grad_norm": 3.8990068435668945, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4638, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 11.23046875, | |
| "eval_cer": 20.545454545454543, | |
| "eval_loss": 0.3684229254722595, | |
| "eval_normalized_cer": 16.974928072338678, | |
| "eval_runtime": 89.0848, | |
| "eval_samples_per_second": 1.437, | |
| "eval_steps_per_second": 0.09, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 11.24267578125, | |
| "grad_norm": 3.1418068408966064, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3857, | |
| "step": 23025 | |
| }, | |
| { | |
| "epoch": 11.2548828125, | |
| "grad_norm": 3.578172445297241, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4817, | |
| "step": 23050 | |
| }, | |
| { | |
| "epoch": 11.26708984375, | |
| "grad_norm": 3.6145339012145996, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3732, | |
| "step": 23075 | |
| }, | |
| { | |
| "epoch": 11.279296875, | |
| "grad_norm": 3.0317516326904297, | |
| "learning_rate": 1e-05, | |
| "loss": 0.429, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 11.29150390625, | |
| "grad_norm": 3.6998438835144043, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4374, | |
| "step": 23125 | |
| }, | |
| { | |
| "epoch": 11.3037109375, | |
| "grad_norm": 3.1519877910614014, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4285, | |
| "step": 23150 | |
| }, | |
| { | |
| "epoch": 11.31591796875, | |
| "grad_norm": 2.30660343170166, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4427, | |
| "step": 23175 | |
| }, | |
| { | |
| "epoch": 11.328125, | |
| "grad_norm": 2.7679004669189453, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4266, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 11.34033203125, | |
| "grad_norm": 4.154956340789795, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3949, | |
| "step": 23225 | |
| }, | |
| { | |
| "epoch": 11.3525390625, | |
| "grad_norm": 3.537022113800049, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4484, | |
| "step": 23250 | |
| }, | |
| { | |
| "epoch": 11.36474609375, | |
| "grad_norm": 3.042243242263794, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4401, | |
| "step": 23275 | |
| }, | |
| { | |
| "epoch": 11.376953125, | |
| "grad_norm": 3.82000994682312, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3968, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 11.38916015625, | |
| "grad_norm": 2.052567958831787, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3856, | |
| "step": 23325 | |
| }, | |
| { | |
| "epoch": 11.4013671875, | |
| "grad_norm": 3.497328758239746, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3915, | |
| "step": 23350 | |
| }, | |
| { | |
| "epoch": 11.41357421875, | |
| "grad_norm": 3.023749589920044, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4381, | |
| "step": 23375 | |
| }, | |
| { | |
| "epoch": 11.42578125, | |
| "grad_norm": 3.8763065338134766, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3879, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 11.43798828125, | |
| "grad_norm": 2.346403121948242, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4255, | |
| "step": 23425 | |
| }, | |
| { | |
| "epoch": 11.4501953125, | |
| "grad_norm": 2.7521772384643555, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4073, | |
| "step": 23450 | |
| }, | |
| { | |
| "epoch": 11.46240234375, | |
| "grad_norm": 2.2675764560699463, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3837, | |
| "step": 23475 | |
| }, | |
| { | |
| "epoch": 11.474609375, | |
| "grad_norm": 3.2309587001800537, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4423, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 11.474609375, | |
| "eval_cer": 21.454545454545453, | |
| "eval_loss": 0.3692309558391571, | |
| "eval_normalized_cer": 16.892725030826142, | |
| "eval_runtime": 91.354, | |
| "eval_samples_per_second": 1.401, | |
| "eval_steps_per_second": 0.088, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 11.48681640625, | |
| "grad_norm": 3.551201581954956, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3808, | |
| "step": 23525 | |
| }, | |
| { | |
| "epoch": 11.4990234375, | |
| "grad_norm": 3.828253984451294, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4081, | |
| "step": 23550 | |
| }, | |
| { | |
| "epoch": 11.51123046875, | |
| "grad_norm": 3.3823349475860596, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3875, | |
| "step": 23575 | |
| }, | |
| { | |
| "epoch": 11.5234375, | |
| "grad_norm": 3.332829713821411, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4167, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 11.53564453125, | |
| "grad_norm": 2.943939685821533, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4397, | |
| "step": 23625 | |
| }, | |
| { | |
| "epoch": 11.5478515625, | |
| "grad_norm": 3.0470168590545654, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4395, | |
| "step": 23650 | |
| }, | |
| { | |
| "epoch": 11.56005859375, | |
| "grad_norm": 3.1324830055236816, | |
| "learning_rate": 1e-05, | |
| "loss": 0.391, | |
| "step": 23675 | |
| }, | |
| { | |
| "epoch": 11.572265625, | |
| "grad_norm": 2.959723711013794, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4063, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 11.58447265625, | |
| "grad_norm": 3.0424516201019287, | |
| "learning_rate": 1e-05, | |
| "loss": 0.457, | |
| "step": 23725 | |
| }, | |
| { | |
| "epoch": 11.5966796875, | |
| "grad_norm": 2.7524912357330322, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3916, | |
| "step": 23750 | |
| }, | |
| { | |
| "epoch": 11.60888671875, | |
| "grad_norm": 3.8842151165008545, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4134, | |
| "step": 23775 | |
| }, | |
| { | |
| "epoch": 11.62109375, | |
| "grad_norm": 2.5505778789520264, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4301, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 11.63330078125, | |
| "grad_norm": 2.3430569171905518, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3913, | |
| "step": 23825 | |
| }, | |
| { | |
| "epoch": 11.6455078125, | |
| "grad_norm": 3.5037317276000977, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4213, | |
| "step": 23850 | |
| }, | |
| { | |
| "epoch": 11.65771484375, | |
| "grad_norm": 3.067307710647583, | |
| "learning_rate": 1e-05, | |
| "loss": 0.395, | |
| "step": 23875 | |
| }, | |
| { | |
| "epoch": 11.669921875, | |
| "grad_norm": 3.3112151622772217, | |
| "learning_rate": 1e-05, | |
| "loss": 0.429, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 11.68212890625, | |
| "grad_norm": 3.318018913269043, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4593, | |
| "step": 23925 | |
| }, | |
| { | |
| "epoch": 11.6943359375, | |
| "grad_norm": 2.73763370513916, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3565, | |
| "step": 23950 | |
| }, | |
| { | |
| "epoch": 11.70654296875, | |
| "grad_norm": 4.161790370941162, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4181, | |
| "step": 23975 | |
| }, | |
| { | |
| "epoch": 11.71875, | |
| "grad_norm": 2.161367893218994, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3991, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 11.71875, | |
| "eval_cer": 20.763636363636365, | |
| "eval_loss": 0.37048545479774475, | |
| "eval_normalized_cer": 16.769420468557335, | |
| "eval_runtime": 97.2817, | |
| "eval_samples_per_second": 1.316, | |
| "eval_steps_per_second": 0.082, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 11.73095703125, | |
| "grad_norm": 2.165503978729248, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3863, | |
| "step": 24025 | |
| }, | |
| { | |
| "epoch": 11.7431640625, | |
| "grad_norm": 3.243612289428711, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4415, | |
| "step": 24050 | |
| }, | |
| { | |
| "epoch": 11.75537109375, | |
| "grad_norm": 3.0755250453948975, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4233, | |
| "step": 24075 | |
| }, | |
| { | |
| "epoch": 11.767578125, | |
| "grad_norm": 3.3624966144561768, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3975, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 11.77978515625, | |
| "grad_norm": 3.3429675102233887, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4349, | |
| "step": 24125 | |
| }, | |
| { | |
| "epoch": 11.7919921875, | |
| "grad_norm": 2.686250925064087, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3938, | |
| "step": 24150 | |
| }, | |
| { | |
| "epoch": 11.80419921875, | |
| "grad_norm": 2.747789144515991, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4164, | |
| "step": 24175 | |
| }, | |
| { | |
| "epoch": 11.81640625, | |
| "grad_norm": 3.79379940032959, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4195, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 11.82861328125, | |
| "grad_norm": 3.6123952865600586, | |
| "learning_rate": 1e-05, | |
| "loss": 0.399, | |
| "step": 24225 | |
| }, | |
| { | |
| "epoch": 11.8408203125, | |
| "grad_norm": 3.1555302143096924, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4238, | |
| "step": 24250 | |
| }, | |
| { | |
| "epoch": 11.85302734375, | |
| "grad_norm": 2.9324278831481934, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4096, | |
| "step": 24275 | |
| }, | |
| { | |
| "epoch": 11.865234375, | |
| "grad_norm": 2.907283067703247, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4128, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 11.87744140625, | |
| "grad_norm": 3.6261141300201416, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4398, | |
| "step": 24325 | |
| }, | |
| { | |
| "epoch": 11.8896484375, | |
| "grad_norm": 4.799058437347412, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4489, | |
| "step": 24350 | |
| }, | |
| { | |
| "epoch": 11.90185546875, | |
| "grad_norm": 2.681626558303833, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4308, | |
| "step": 24375 | |
| }, | |
| { | |
| "epoch": 11.9140625, | |
| "grad_norm": 2.723325729370117, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3909, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 11.92626953125, | |
| "grad_norm": 4.047337055206299, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3749, | |
| "step": 24425 | |
| }, | |
| { | |
| "epoch": 11.9384765625, | |
| "grad_norm": 3.933332920074463, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4232, | |
| "step": 24450 | |
| }, | |
| { | |
| "epoch": 11.95068359375, | |
| "grad_norm": 2.8435702323913574, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4071, | |
| "step": 24475 | |
| }, | |
| { | |
| "epoch": 11.962890625, | |
| "grad_norm": 3.180521249771118, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4023, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 11.962890625, | |
| "eval_cer": 18.436363636363637, | |
| "eval_loss": 0.36682504415512085, | |
| "eval_normalized_cer": 14.26222770242499, | |
| "eval_runtime": 89.1047, | |
| "eval_samples_per_second": 1.437, | |
| "eval_steps_per_second": 0.09, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 11.97509765625, | |
| "grad_norm": 2.902740716934204, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4535, | |
| "step": 24525 | |
| }, | |
| { | |
| "epoch": 11.9873046875, | |
| "grad_norm": 2.6517724990844727, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4093, | |
| "step": 24550 | |
| }, | |
| { | |
| "epoch": 11.99951171875, | |
| "grad_norm": 3.2232935428619385, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4295, | |
| "step": 24575 | |
| }, | |
| { | |
| "epoch": 12.01171875, | |
| "grad_norm": 1.8525205850601196, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4109, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 12.02392578125, | |
| "grad_norm": 2.505067825317383, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3659, | |
| "step": 24625 | |
| }, | |
| { | |
| "epoch": 12.0361328125, | |
| "grad_norm": 2.921861410140991, | |
| "learning_rate": 1e-05, | |
| "loss": 0.387, | |
| "step": 24650 | |
| }, | |
| { | |
| "epoch": 12.04833984375, | |
| "grad_norm": 3.4830503463745117, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4036, | |
| "step": 24675 | |
| }, | |
| { | |
| "epoch": 12.060546875, | |
| "grad_norm": 3.166236400604248, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4014, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 12.07275390625, | |
| "grad_norm": 2.9270448684692383, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3983, | |
| "step": 24725 | |
| }, | |
| { | |
| "epoch": 12.0849609375, | |
| "grad_norm": 3.857215166091919, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4343, | |
| "step": 24750 | |
| }, | |
| { | |
| "epoch": 12.09716796875, | |
| "grad_norm": 3.039315700531006, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4095, | |
| "step": 24775 | |
| }, | |
| { | |
| "epoch": 12.109375, | |
| "grad_norm": 4.983880043029785, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4108, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 12.12158203125, | |
| "grad_norm": 3.1066694259643555, | |
| "learning_rate": 1e-05, | |
| "loss": 0.421, | |
| "step": 24825 | |
| }, | |
| { | |
| "epoch": 12.1337890625, | |
| "grad_norm": 3.10778546333313, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4243, | |
| "step": 24850 | |
| }, | |
| { | |
| "epoch": 12.14599609375, | |
| "grad_norm": 2.57086443901062, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4049, | |
| "step": 24875 | |
| }, | |
| { | |
| "epoch": 12.158203125, | |
| "grad_norm": 2.7861294746398926, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3917, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 12.17041015625, | |
| "grad_norm": 3.35518217086792, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4332, | |
| "step": 24925 | |
| }, | |
| { | |
| "epoch": 12.1826171875, | |
| "grad_norm": 2.97426176071167, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4148, | |
| "step": 24950 | |
| }, | |
| { | |
| "epoch": 12.19482421875, | |
| "grad_norm": 2.733003854751587, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3901, | |
| "step": 24975 | |
| }, | |
| { | |
| "epoch": 12.20703125, | |
| "grad_norm": 2.4438769817352295, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4039, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 12.20703125, | |
| "eval_cer": 20.763636363636365, | |
| "eval_loss": 0.3740121126174927, | |
| "eval_normalized_cer": 15.454171804356761, | |
| "eval_runtime": 88.8688, | |
| "eval_samples_per_second": 1.44, | |
| "eval_steps_per_second": 0.09, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 12.21923828125, | |
| "grad_norm": 2.897085666656494, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4155, | |
| "step": 25025 | |
| }, | |
| { | |
| "epoch": 12.2314453125, | |
| "grad_norm": 3.21028470993042, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4249, | |
| "step": 25050 | |
| }, | |
| { | |
| "epoch": 12.24365234375, | |
| "grad_norm": 2.8493075370788574, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4279, | |
| "step": 25075 | |
| }, | |
| { | |
| "epoch": 12.255859375, | |
| "grad_norm": 3.8995752334594727, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4169, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 12.26806640625, | |
| "grad_norm": 3.4920387268066406, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4164, | |
| "step": 25125 | |
| }, | |
| { | |
| "epoch": 12.2802734375, | |
| "grad_norm": 3.276472330093384, | |
| "learning_rate": 1e-05, | |
| "loss": 0.419, | |
| "step": 25150 | |
| }, | |
| { | |
| "epoch": 12.29248046875, | |
| "grad_norm": 2.8615663051605225, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3974, | |
| "step": 25175 | |
| }, | |
| { | |
| "epoch": 12.3046875, | |
| "grad_norm": 4.341454029083252, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4033, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 12.31689453125, | |
| "grad_norm": 4.287600517272949, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4316, | |
| "step": 25225 | |
| }, | |
| { | |
| "epoch": 12.3291015625, | |
| "grad_norm": 2.183636426925659, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4461, | |
| "step": 25250 | |
| }, | |
| { | |
| "epoch": 12.34130859375, | |
| "grad_norm": 2.487642765045166, | |
| "learning_rate": 1e-05, | |
| "loss": 0.413, | |
| "step": 25275 | |
| }, | |
| { | |
| "epoch": 12.353515625, | |
| "grad_norm": 2.7473361492156982, | |
| "learning_rate": 1e-05, | |
| "loss": 0.445, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 12.36572265625, | |
| "grad_norm": 3.353499174118042, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4473, | |
| "step": 25325 | |
| }, | |
| { | |
| "epoch": 12.3779296875, | |
| "grad_norm": 2.5522449016571045, | |
| "learning_rate": 1e-05, | |
| "loss": 0.41, | |
| "step": 25350 | |
| }, | |
| { | |
| "epoch": 12.39013671875, | |
| "grad_norm": 4.818186283111572, | |
| "learning_rate": 1e-05, | |
| "loss": 0.445, | |
| "step": 25375 | |
| }, | |
| { | |
| "epoch": 12.40234375, | |
| "grad_norm": 4.093599796295166, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4741, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 12.41455078125, | |
| "grad_norm": 3.5378177165985107, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4002, | |
| "step": 25425 | |
| }, | |
| { | |
| "epoch": 12.4267578125, | |
| "grad_norm": 2.7892191410064697, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3966, | |
| "step": 25450 | |
| }, | |
| { | |
| "epoch": 12.43896484375, | |
| "grad_norm": 4.0514044761657715, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4497, | |
| "step": 25475 | |
| }, | |
| { | |
| "epoch": 12.451171875, | |
| "grad_norm": 2.708777904510498, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4542, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 12.451171875, | |
| "eval_cer": 21.78181818181818, | |
| "eval_loss": 0.3693616986274719, | |
| "eval_normalized_cer": 17.879161528976574, | |
| "eval_runtime": 91.1491, | |
| "eval_samples_per_second": 1.404, | |
| "eval_steps_per_second": 0.088, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 12.46337890625, | |
| "grad_norm": 2.597757577896118, | |
| "learning_rate": 1e-05, | |
| "loss": 0.398, | |
| "step": 25525 | |
| }, | |
| { | |
| "epoch": 12.4755859375, | |
| "grad_norm": 4.003383159637451, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3819, | |
| "step": 25550 | |
| }, | |
| { | |
| "epoch": 12.48779296875, | |
| "grad_norm": 2.98598575592041, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3604, | |
| "step": 25575 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 2.1123640537261963, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4159, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 12.51220703125, | |
| "grad_norm": 3.579258441925049, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4355, | |
| "step": 25625 | |
| }, | |
| { | |
| "epoch": 12.5244140625, | |
| "grad_norm": 3.140286922454834, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3937, | |
| "step": 25650 | |
| }, | |
| { | |
| "epoch": 12.53662109375, | |
| "grad_norm": 2.7550101280212402, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4045, | |
| "step": 25675 | |
| }, | |
| { | |
| "epoch": 12.548828125, | |
| "grad_norm": 2.977573871612549, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4117, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 12.56103515625, | |
| "grad_norm": 2.6313016414642334, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3912, | |
| "step": 25725 | |
| }, | |
| { | |
| "epoch": 12.5732421875, | |
| "grad_norm": 1.7519272565841675, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4081, | |
| "step": 25750 | |
| }, | |
| { | |
| "epoch": 12.58544921875, | |
| "grad_norm": 2.9577393531799316, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4349, | |
| "step": 25775 | |
| }, | |
| { | |
| "epoch": 12.59765625, | |
| "grad_norm": 3.945467233657837, | |
| "learning_rate": 1e-05, | |
| "loss": 0.403, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 12.60986328125, | |
| "grad_norm": 3.028316020965576, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4188, | |
| "step": 25825 | |
| }, | |
| { | |
| "epoch": 12.6220703125, | |
| "grad_norm": 2.839320421218872, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4018, | |
| "step": 25850 | |
| }, | |
| { | |
| "epoch": 12.63427734375, | |
| "grad_norm": 2.295088291168213, | |
| "learning_rate": 1e-05, | |
| "loss": 0.35, | |
| "step": 25875 | |
| }, | |
| { | |
| "epoch": 12.646484375, | |
| "grad_norm": 2.7483038902282715, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4129, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 12.65869140625, | |
| "grad_norm": 3.1008427143096924, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3926, | |
| "step": 25925 | |
| }, | |
| { | |
| "epoch": 12.6708984375, | |
| "grad_norm": 2.92093825340271, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3813, | |
| "step": 25950 | |
| }, | |
| { | |
| "epoch": 12.68310546875, | |
| "grad_norm": 2.930253267288208, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3903, | |
| "step": 25975 | |
| }, | |
| { | |
| "epoch": 12.6953125, | |
| "grad_norm": 2.860041379928589, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4042, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 12.6953125, | |
| "eval_cer": 22.10909090909091, | |
| "eval_loss": 0.3708876967430115, | |
| "eval_normalized_cer": 17.79695848746404, | |
| "eval_runtime": 90.9962, | |
| "eval_samples_per_second": 1.407, | |
| "eval_steps_per_second": 0.088, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 12.70751953125, | |
| "grad_norm": 5.0800299644470215, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4279, | |
| "step": 26025 | |
| }, | |
| { | |
| "epoch": 12.7197265625, | |
| "grad_norm": 4.715578556060791, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4126, | |
| "step": 26050 | |
| }, | |
| { | |
| "epoch": 12.73193359375, | |
| "grad_norm": 3.818558692932129, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4176, | |
| "step": 26075 | |
| }, | |
| { | |
| "epoch": 12.744140625, | |
| "grad_norm": 3.0944199562072754, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4155, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 12.75634765625, | |
| "grad_norm": 4.057784080505371, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4359, | |
| "step": 26125 | |
| }, | |
| { | |
| "epoch": 12.7685546875, | |
| "grad_norm": 2.12882661819458, | |
| "learning_rate": 1e-05, | |
| "loss": 0.427, | |
| "step": 26150 | |
| }, | |
| { | |
| "epoch": 12.78076171875, | |
| "grad_norm": 2.9268953800201416, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4437, | |
| "step": 26175 | |
| }, | |
| { | |
| "epoch": 12.79296875, | |
| "grad_norm": 3.1372625827789307, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3814, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 12.80517578125, | |
| "grad_norm": 3.171339511871338, | |
| "learning_rate": 1e-05, | |
| "loss": 0.461, | |
| "step": 26225 | |
| }, | |
| { | |
| "epoch": 12.8173828125, | |
| "grad_norm": 2.2062418460845947, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4147, | |
| "step": 26250 | |
| }, | |
| { | |
| "epoch": 12.82958984375, | |
| "grad_norm": 4.490868091583252, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3782, | |
| "step": 26275 | |
| }, | |
| { | |
| "epoch": 12.841796875, | |
| "grad_norm": 4.018364429473877, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4402, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 12.85400390625, | |
| "grad_norm": 3.0183207988739014, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4353, | |
| "step": 26325 | |
| }, | |
| { | |
| "epoch": 12.8662109375, | |
| "grad_norm": 3.209418535232544, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4267, | |
| "step": 26350 | |
| }, | |
| { | |
| "epoch": 12.87841796875, | |
| "grad_norm": 2.75201678276062, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4098, | |
| "step": 26375 | |
| }, | |
| { | |
| "epoch": 12.890625, | |
| "grad_norm": 2.8569531440734863, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4134, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 12.90283203125, | |
| "grad_norm": 3.4323573112487793, | |
| "learning_rate": 1e-05, | |
| "loss": 0.392, | |
| "step": 26425 | |
| }, | |
| { | |
| "epoch": 12.9150390625, | |
| "grad_norm": 2.4016165733337402, | |
| "learning_rate": 1e-05, | |
| "loss": 0.407, | |
| "step": 26450 | |
| }, | |
| { | |
| "epoch": 12.92724609375, | |
| "grad_norm": 2.4917731285095215, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3779, | |
| "step": 26475 | |
| }, | |
| { | |
| "epoch": 12.939453125, | |
| "grad_norm": 2.2973673343658447, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4162, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 12.939453125, | |
| "eval_cer": 20.10909090909091, | |
| "eval_loss": 0.36156314611434937, | |
| "eval_normalized_cer": 16.029593094944513, | |
| "eval_runtime": 88.7052, | |
| "eval_samples_per_second": 1.443, | |
| "eval_steps_per_second": 0.09, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 12.95166015625, | |
| "grad_norm": 2.8456501960754395, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4392, | |
| "step": 26525 | |
| }, | |
| { | |
| "epoch": 12.9638671875, | |
| "grad_norm": 2.975572109222412, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4097, | |
| "step": 26550 | |
| }, | |
| { | |
| "epoch": 12.97607421875, | |
| "grad_norm": 4.956911563873291, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4098, | |
| "step": 26575 | |
| }, | |
| { | |
| "epoch": 12.98828125, | |
| "grad_norm": 3.6981606483459473, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4451, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 26624, | |
| "total_flos": 6.747472049681203e+20, | |
| "train_loss": 0.031781960636950456, | |
| "train_runtime": 2625.017, | |
| "train_samples_per_second": 162.279, | |
| "train_steps_per_second": 10.142 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 26624, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.747472049681203e+20, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |