| { | |
| "best_metric": 0.7119987566055331, | |
| "best_model_checkpoint": "xls-r-greek-aivaliot/checkpoint-14196", | |
| "epoch": 35.0, | |
| "eval_steps": 500, | |
| "global_step": 19110, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 6.1718, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 3.1837, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 0.6650782444993528, | |
| "eval_loss": 2.3546931743621826, | |
| "eval_runtime": 42.745, | |
| "eval_samples_per_second": 25.547, | |
| "eval_steps_per_second": 3.205, | |
| "eval_wer": 0.972023624494871, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 0.00029838796346050506, | |
| "loss": 2.2974, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 0.0002951638903815153, | |
| "loss": 2.0321, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 0.00029193981730252553, | |
| "loss": 1.9583, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 0.45720084715848924, | |
| "eval_loss": 1.9236657619476318, | |
| "eval_runtime": 44.6726, | |
| "eval_samples_per_second": 24.444, | |
| "eval_steps_per_second": 3.067, | |
| "eval_wer": 0.8734846129934721, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 0.0002887157442235357, | |
| "loss": 1.7559, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 0.00028549167114454594, | |
| "loss": 1.6492, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 0.0002822675980655561, | |
| "loss": 1.6148, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 0.450847158489234, | |
| "eval_loss": 1.817557692527771, | |
| "eval_runtime": 41.1726, | |
| "eval_samples_per_second": 26.522, | |
| "eval_steps_per_second": 3.327, | |
| "eval_wer": 0.8234379857009636, | |
| "step": 1638 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 0.00027904352498656635, | |
| "loss": 1.421, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "learning_rate": 0.0002758194519075766, | |
| "loss": 1.423, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.4232556771384869, | |
| "eval_loss": 1.8342238664627075, | |
| "eval_runtime": 39.2094, | |
| "eval_samples_per_second": 27.85, | |
| "eval_steps_per_second": 3.494, | |
| "eval_wer": 0.820329499533727, | |
| "step": 2184 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 0.00027259537882858676, | |
| "loss": 1.4186, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 0.000269371305749597, | |
| "loss": 1.2185, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 0.00026614723267060717, | |
| "loss": 1.2324, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.4242852100247088, | |
| "eval_loss": 1.7897675037384033, | |
| "eval_runtime": 45.221, | |
| "eval_samples_per_second": 24.148, | |
| "eval_steps_per_second": 3.03, | |
| "eval_wer": 0.8047870686975443, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "learning_rate": 0.00026292315959161735, | |
| "loss": 1.135, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "learning_rate": 0.0002596990865126276, | |
| "loss": 1.0774, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "learning_rate": 0.0002564750134336378, | |
| "loss": 1.0871, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 0.40566537239675254, | |
| "eval_loss": 1.783974051475525, | |
| "eval_runtime": 39.7961, | |
| "eval_samples_per_second": 27.44, | |
| "eval_steps_per_second": 3.443, | |
| "eval_wer": 0.7671743860739819, | |
| "step": 3276 | |
| }, | |
| { | |
| "epoch": 6.23, | |
| "learning_rate": 0.000253250940354648, | |
| "loss": 1.0111, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 6.59, | |
| "learning_rate": 0.0002500268672756582, | |
| "loss": 0.9552, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "learning_rate": 0.00024680279419666846, | |
| "loss": 0.9449, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 0.3969584657018473, | |
| "eval_loss": 1.9680403470993042, | |
| "eval_runtime": 41.0562, | |
| "eval_samples_per_second": 26.598, | |
| "eval_steps_per_second": 3.337, | |
| "eval_wer": 0.7777432390425862, | |
| "step": 3822 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "learning_rate": 0.00024357872111767866, | |
| "loss": 0.8435, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 0.00024035464803868887, | |
| "loss": 0.8168, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.4033415695964231, | |
| "eval_loss": 1.956693172454834, | |
| "eval_runtime": 38.8124, | |
| "eval_samples_per_second": 28.135, | |
| "eval_steps_per_second": 3.53, | |
| "eval_wer": 0.7690394777743239, | |
| "step": 4368 | |
| }, | |
| { | |
| "epoch": 8.06, | |
| "learning_rate": 0.00023713057495969905, | |
| "loss": 0.8366, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 8.42, | |
| "learning_rate": 0.00023390650188070925, | |
| "loss": 0.728, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "learning_rate": 0.00023068242880171949, | |
| "loss": 0.7254, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 0.408224496999647, | |
| "eval_loss": 2.184379816055298, | |
| "eval_runtime": 39.1148, | |
| "eval_samples_per_second": 27.918, | |
| "eval_steps_per_second": 3.503, | |
| "eval_wer": 0.7878458190861051, | |
| "step": 4914 | |
| }, | |
| { | |
| "epoch": 9.16, | |
| "learning_rate": 0.0002274583557227297, | |
| "loss": 0.7016, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "learning_rate": 0.0002242342826437399, | |
| "loss": 0.6563, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 9.89, | |
| "learning_rate": 0.0002210102095647501, | |
| "loss": 0.6584, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 0.3924579362277915, | |
| "eval_loss": 2.313096046447754, | |
| "eval_runtime": 38.9682, | |
| "eval_samples_per_second": 28.023, | |
| "eval_steps_per_second": 3.516, | |
| "eval_wer": 0.7681069319241529, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 10.26, | |
| "learning_rate": 0.0002177861364857603, | |
| "loss": 0.6053, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 10.62, | |
| "learning_rate": 0.00021456206340677054, | |
| "loss": 0.5844, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 10.99, | |
| "learning_rate": 0.00021133799032778075, | |
| "loss": 0.5825, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 0.39419343452170846, | |
| "eval_loss": 2.4024157524108887, | |
| "eval_runtime": 38.7244, | |
| "eval_samples_per_second": 28.199, | |
| "eval_steps_per_second": 3.538, | |
| "eval_wer": 0.7614236866645944, | |
| "step": 6006 | |
| }, | |
| { | |
| "epoch": 11.36, | |
| "learning_rate": 0.00020811391724879095, | |
| "loss": 0.4944, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 11.72, | |
| "learning_rate": 0.00020488984416980116, | |
| "loss": 0.5214, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 0.37542652076714905, | |
| "eval_loss": 2.2336461544036865, | |
| "eval_runtime": 39.0775, | |
| "eval_samples_per_second": 27.944, | |
| "eval_steps_per_second": 3.506, | |
| "eval_wer": 0.7415293751942804, | |
| "step": 6552 | |
| }, | |
| { | |
| "epoch": 12.09, | |
| "learning_rate": 0.0002016657710908114, | |
| "loss": 0.4977, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 12.45, | |
| "learning_rate": 0.0001984416980118216, | |
| "loss": 0.4326, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 12.82, | |
| "learning_rate": 0.0001952176249328318, | |
| "loss": 0.4618, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 0.38816331333098014, | |
| "eval_loss": 2.4235401153564453, | |
| "eval_runtime": 38.6611, | |
| "eval_samples_per_second": 28.245, | |
| "eval_steps_per_second": 3.544, | |
| "eval_wer": 0.7475909232203917, | |
| "step": 7098 | |
| }, | |
| { | |
| "epoch": 13.19, | |
| "learning_rate": 0.000191993551853842, | |
| "loss": 0.4269, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 13.55, | |
| "learning_rate": 0.0001887694787748522, | |
| "loss": 0.3915, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 13.92, | |
| "learning_rate": 0.00018554540569586244, | |
| "loss": 0.4034, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 0.38966348982233207, | |
| "eval_loss": 2.432621479034424, | |
| "eval_runtime": 39.0269, | |
| "eval_samples_per_second": 27.981, | |
| "eval_steps_per_second": 3.51, | |
| "eval_wer": 0.7384208890270438, | |
| "step": 7644 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "learning_rate": 0.00018232133261687265, | |
| "loss": 0.3616, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 14.65, | |
| "learning_rate": 0.00017909725953788285, | |
| "loss": 0.3638, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 0.38351570773032123, | |
| "eval_loss": 2.5917038917541504, | |
| "eval_runtime": 38.9489, | |
| "eval_samples_per_second": 28.037, | |
| "eval_steps_per_second": 3.517, | |
| "eval_wer": 0.7409076779608331, | |
| "step": 8190 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "learning_rate": 0.00017587318645889303, | |
| "loss": 0.3469, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 0.00017264911337990324, | |
| "loss": 0.2989, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 15.75, | |
| "learning_rate": 0.0001694250403009135, | |
| "loss": 0.307, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 0.38233909871749616, | |
| "eval_loss": 2.653606414794922, | |
| "eval_runtime": 38.6877, | |
| "eval_samples_per_second": 28.226, | |
| "eval_steps_per_second": 3.541, | |
| "eval_wer": 0.7483680447622008, | |
| "step": 8736 | |
| }, | |
| { | |
| "epoch": 16.12, | |
| "learning_rate": 0.00016620096722192368, | |
| "loss": 0.3028, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 16.48, | |
| "learning_rate": 0.00016297689414293388, | |
| "loss": 0.2857, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 16.85, | |
| "learning_rate": 0.00015975282106394409, | |
| "loss": 0.2794, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_cer": 0.3855159430521238, | |
| "eval_loss": 2.7668089866638184, | |
| "eval_runtime": 38.9844, | |
| "eval_samples_per_second": 28.011, | |
| "eval_steps_per_second": 3.514, | |
| "eval_wer": 0.7413739508859185, | |
| "step": 9282 | |
| }, | |
| { | |
| "epoch": 17.22, | |
| "learning_rate": 0.00015652874798495432, | |
| "loss": 0.2484, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 17.58, | |
| "learning_rate": 0.00015330467490596452, | |
| "loss": 0.245, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 17.95, | |
| "learning_rate": 0.00015008060182697473, | |
| "loss": 0.2445, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_cer": 0.3990469466996117, | |
| "eval_loss": 2.955599069595337, | |
| "eval_runtime": 38.9844, | |
| "eval_samples_per_second": 28.011, | |
| "eval_steps_per_second": 3.514, | |
| "eval_wer": 0.7597140192726143, | |
| "step": 9828 | |
| }, | |
| { | |
| "epoch": 18.32, | |
| "learning_rate": 0.00014685652874798494, | |
| "loss": 0.208, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 18.68, | |
| "learning_rate": 0.00014363245566899517, | |
| "loss": 0.2209, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_cer": 0.38439816448994, | |
| "eval_loss": 2.9723663330078125, | |
| "eval_runtime": 38.8984, | |
| "eval_samples_per_second": 28.073, | |
| "eval_steps_per_second": 3.522, | |
| "eval_wer": 0.7396642834939384, | |
| "step": 10374 | |
| }, | |
| { | |
| "epoch": 19.05, | |
| "learning_rate": 0.00014040838259000535, | |
| "loss": 0.2107, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 19.41, | |
| "learning_rate": 0.00013718430951101558, | |
| "loss": 0.1935, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 19.78, | |
| "learning_rate": 0.00013396023643202578, | |
| "loss": 0.193, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_cer": 0.3829568184492293, | |
| "eval_loss": 3.105555534362793, | |
| "eval_runtime": 38.7297, | |
| "eval_samples_per_second": 28.195, | |
| "eval_steps_per_second": 3.537, | |
| "eval_wer": 0.7412185265775567, | |
| "step": 10920 | |
| }, | |
| { | |
| "epoch": 20.15, | |
| "learning_rate": 0.000130736163353036, | |
| "loss": 0.1764, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 20.51, | |
| "learning_rate": 0.0001275120902740462, | |
| "loss": 0.1683, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 20.88, | |
| "learning_rate": 0.0001242880171950564, | |
| "loss": 0.1608, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_cer": 0.3850452994469938, | |
| "eval_loss": 3.2178432941436768, | |
| "eval_runtime": 38.643, | |
| "eval_samples_per_second": 28.259, | |
| "eval_steps_per_second": 3.545, | |
| "eval_wer": 0.7332918868511035, | |
| "step": 11466 | |
| }, | |
| { | |
| "epoch": 21.25, | |
| "learning_rate": 0.00012106394411606662, | |
| "loss": 0.1514, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 21.61, | |
| "learning_rate": 0.00011783987103707682, | |
| "loss": 0.1421, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 21.98, | |
| "learning_rate": 0.00011461579795808704, | |
| "loss": 0.1454, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_cer": 0.3789563478056242, | |
| "eval_loss": 3.3512697219848633, | |
| "eval_runtime": 38.841, | |
| "eval_samples_per_second": 28.115, | |
| "eval_steps_per_second": 3.527, | |
| "eval_wer": 0.7322039166925707, | |
| "step": 12012 | |
| }, | |
| { | |
| "epoch": 22.34, | |
| "learning_rate": 0.00011139172487909725, | |
| "loss": 0.1378, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 22.71, | |
| "learning_rate": 0.00010816765180010745, | |
| "loss": 0.1269, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_cer": 0.38342746205435935, | |
| "eval_loss": 3.4092280864715576, | |
| "eval_runtime": 39.0183, | |
| "eval_samples_per_second": 27.987, | |
| "eval_steps_per_second": 3.511, | |
| "eval_wer": 0.7311159465340379, | |
| "step": 12558 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "learning_rate": 0.00010494357872111767, | |
| "loss": 0.1334, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 23.44, | |
| "learning_rate": 0.00010171950564212788, | |
| "loss": 0.1069, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 23.81, | |
| "learning_rate": 9.84954325631381e-05, | |
| "loss": 0.114, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_cer": 0.3863101541357807, | |
| "eval_loss": 3.3670547008514404, | |
| "eval_runtime": 38.7589, | |
| "eval_samples_per_second": 28.174, | |
| "eval_steps_per_second": 3.535, | |
| "eval_wer": 0.7300279763755051, | |
| "step": 13104 | |
| }, | |
| { | |
| "epoch": 24.18, | |
| "learning_rate": 9.527135948414829e-05, | |
| "loss": 0.1064, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 24.54, | |
| "learning_rate": 9.204728640515852e-05, | |
| "loss": 0.1056, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 24.91, | |
| "learning_rate": 8.882321332616871e-05, | |
| "loss": 0.1086, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_cer": 0.38175079421108365, | |
| "eval_loss": 3.6167728900909424, | |
| "eval_runtime": 38.7877, | |
| "eval_samples_per_second": 28.153, | |
| "eval_steps_per_second": 3.532, | |
| "eval_wer": 0.7179048803232826, | |
| "step": 13650 | |
| }, | |
| { | |
| "epoch": 25.27, | |
| "learning_rate": 8.559914024717892e-05, | |
| "loss": 0.0853, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 25.64, | |
| "learning_rate": 8.237506716818914e-05, | |
| "loss": 0.0877, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_cer": 0.3746617249088128, | |
| "eval_loss": 3.5180928707122803, | |
| "eval_runtime": 38.7044, | |
| "eval_samples_per_second": 28.214, | |
| "eval_steps_per_second": 3.54, | |
| "eval_wer": 0.7119987566055331, | |
| "step": 14196 | |
| }, | |
| { | |
| "epoch": 26.01, | |
| "learning_rate": 7.915099408919934e-05, | |
| "loss": 0.0966, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 26.37, | |
| "learning_rate": 7.592692101020956e-05, | |
| "loss": 0.0855, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 26.74, | |
| "learning_rate": 7.270284793121977e-05, | |
| "loss": 0.0856, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_cer": 0.39136957289092833, | |
| "eval_loss": 3.875309467315674, | |
| "eval_runtime": 38.7918, | |
| "eval_samples_per_second": 28.15, | |
| "eval_steps_per_second": 3.532, | |
| "eval_wer": 0.7429281939695368, | |
| "step": 14742 | |
| }, | |
| { | |
| "epoch": 27.11, | |
| "learning_rate": 6.947877485222997e-05, | |
| "loss": 0.081, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 27.47, | |
| "learning_rate": 6.625470177324019e-05, | |
| "loss": 0.0708, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 27.84, | |
| "learning_rate": 6.30306286942504e-05, | |
| "loss": 0.0734, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_cer": 0.37516178373926345, | |
| "eval_loss": 3.8759613037109375, | |
| "eval_runtime": 38.881, | |
| "eval_samples_per_second": 28.086, | |
| "eval_steps_per_second": 3.524, | |
| "eval_wer": 0.7227230338824993, | |
| "step": 15288 | |
| }, | |
| { | |
| "epoch": 28.21, | |
| "learning_rate": 5.980655561526061e-05, | |
| "loss": 0.0734, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "learning_rate": 5.658248253627082e-05, | |
| "loss": 0.0642, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 28.94, | |
| "learning_rate": 5.335840945728102e-05, | |
| "loss": 0.0702, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_cer": 0.37551476644311094, | |
| "eval_loss": 3.9142181873321533, | |
| "eval_runtime": 38.8332, | |
| "eval_samples_per_second": 28.12, | |
| "eval_steps_per_second": 3.528, | |
| "eval_wer": 0.723811004041032, | |
| "step": 15834 | |
| }, | |
| { | |
| "epoch": 29.3, | |
| "learning_rate": 5.0134336378291234e-05, | |
| "loss": 0.0617, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 29.67, | |
| "learning_rate": 4.6910263299301446e-05, | |
| "loss": 0.0661, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_cer": 0.3737498529238734, | |
| "eval_loss": 4.010254859924316, | |
| "eval_runtime": 38.9811, | |
| "eval_samples_per_second": 28.014, | |
| "eval_steps_per_second": 3.515, | |
| "eval_wer": 0.7196145477152627, | |
| "step": 16380 | |
| }, | |
| { | |
| "epoch": 30.04, | |
| "learning_rate": 4.368619022031166e-05, | |
| "loss": 0.0559, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 30.4, | |
| "learning_rate": 4.046211714132187e-05, | |
| "loss": 0.0575, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 30.77, | |
| "learning_rate": 3.7238044062332076e-05, | |
| "loss": 0.0569, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_cer": 0.3768090363572185, | |
| "eval_loss": 4.117391109466553, | |
| "eval_runtime": 39.1729, | |
| "eval_samples_per_second": 27.876, | |
| "eval_steps_per_second": 3.497, | |
| "eval_wer": 0.729250854833696, | |
| "step": 16926 | |
| }, | |
| { | |
| "epoch": 31.14, | |
| "learning_rate": 3.401397098334229e-05, | |
| "loss": 0.0559, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 31.5, | |
| "learning_rate": 3.078989790435249e-05, | |
| "loss": 0.0501, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 31.87, | |
| "learning_rate": 2.756582482536271e-05, | |
| "loss": 0.0494, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_cer": 0.37404400517707964, | |
| "eval_loss": 4.161037921905518, | |
| "eval_runtime": 38.9916, | |
| "eval_samples_per_second": 28.006, | |
| "eval_steps_per_second": 3.514, | |
| "eval_wer": 0.7207025178737955, | |
| "step": 17472 | |
| }, | |
| { | |
| "epoch": 32.23, | |
| "learning_rate": 2.4341751746372914e-05, | |
| "loss": 0.048, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 32.6, | |
| "learning_rate": 2.1117678667383127e-05, | |
| "loss": 0.0509, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 32.97, | |
| "learning_rate": 1.7893605588393335e-05, | |
| "loss": 0.0493, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_cer": 0.37219084598188024, | |
| "eval_loss": 4.1340861320495605, | |
| "eval_runtime": 38.9947, | |
| "eval_samples_per_second": 28.004, | |
| "eval_steps_per_second": 3.513, | |
| "eval_wer": 0.7165060615480261, | |
| "step": 18018 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 1.4669532509403546e-05, | |
| "loss": 0.0415, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 33.7, | |
| "learning_rate": 1.1445459430413755e-05, | |
| "loss": 0.0395, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_cer": 0.372132015531239, | |
| "eval_loss": 4.2251057624816895, | |
| "eval_runtime": 38.8607, | |
| "eval_samples_per_second": 28.1, | |
| "eval_steps_per_second": 3.525, | |
| "eval_wer": 0.7135529996891514, | |
| "step": 18564 | |
| }, | |
| { | |
| "epoch": 34.07, | |
| "learning_rate": 8.221386351423965e-06, | |
| "loss": 0.0472, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 34.43, | |
| "learning_rate": 4.997313272434174e-06, | |
| "loss": 0.037, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 34.8, | |
| "learning_rate": 1.7732401934443846e-06, | |
| "loss": 0.0377, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_cer": 0.3725438286857277, | |
| "eval_loss": 4.262682914733887, | |
| "eval_runtime": 38.6349, | |
| "eval_samples_per_second": 28.265, | |
| "eval_steps_per_second": 3.546, | |
| "eval_wer": 0.715728940006217, | |
| "step": 19110 | |
| } | |
| ], | |
| "logging_steps": 200, | |
| "max_steps": 19110, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 35, | |
| "save_steps": 500, | |
| "total_flos": 2.6904981744760697e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |