| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 28120, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17780938833570412, | |
| "grad_norm": 1.8023470640182495, | |
| "learning_rate": 0.0002988, | |
| "loss": 4.6629, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.17780938833570412, | |
| "eval_loss": 0.3540094494819641, | |
| "eval_runtime": 77.5992, | |
| "eval_samples_per_second": 19.059, | |
| "eval_steps_per_second": 2.384, | |
| "eval_wer": 0.5421303656597775, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.35561877667140823, | |
| "grad_norm": 1.411382794380188, | |
| "learning_rate": 0.0002945908761766835, | |
| "loss": 0.6579, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.35561877667140823, | |
| "eval_loss": 0.25881391763687134, | |
| "eval_runtime": 77.5785, | |
| "eval_samples_per_second": 19.065, | |
| "eval_steps_per_second": 2.385, | |
| "eval_wer": 0.48825295884119413, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5334281650071123, | |
| "grad_norm": 1.561378836631775, | |
| "learning_rate": 0.00028916002896451845, | |
| "loss": 0.591, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.5334281650071123, | |
| "eval_loss": 0.25524866580963135, | |
| "eval_runtime": 77.3789, | |
| "eval_samples_per_second": 19.114, | |
| "eval_steps_per_second": 2.391, | |
| "eval_wer": 0.47200141317788374, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7112375533428165, | |
| "grad_norm": 1.4678888320922852, | |
| "learning_rate": 0.00028372918175235333, | |
| "loss": 0.5467, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.7112375533428165, | |
| "eval_loss": 0.23701371252536774, | |
| "eval_runtime": 77.4845, | |
| "eval_samples_per_second": 19.088, | |
| "eval_steps_per_second": 2.388, | |
| "eval_wer": 0.4541600423953365, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8890469416785206, | |
| "grad_norm": 2.583188056945801, | |
| "learning_rate": 0.00027829833454018826, | |
| "loss": 0.5405, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.8890469416785206, | |
| "eval_loss": 0.23755988478660583, | |
| "eval_runtime": 78.0759, | |
| "eval_samples_per_second": 18.943, | |
| "eval_steps_per_second": 2.369, | |
| "eval_wer": 0.45557322027910263, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.0668563300142249, | |
| "grad_norm": 1.422250509262085, | |
| "learning_rate": 0.00027286748732802314, | |
| "loss": 0.5027, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.0668563300142249, | |
| "eval_loss": 0.22338581085205078, | |
| "eval_runtime": 77.6024, | |
| "eval_samples_per_second": 19.059, | |
| "eval_steps_per_second": 2.384, | |
| "eval_wer": 0.4306659600777248, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.2446657183499288, | |
| "grad_norm": 1.006721019744873, | |
| "learning_rate": 0.00026743664011585807, | |
| "loss": 0.5001, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.2446657183499288, | |
| "eval_loss": 0.21763387322425842, | |
| "eval_runtime": 77.5866, | |
| "eval_samples_per_second": 19.063, | |
| "eval_steps_per_second": 2.384, | |
| "eval_wer": 0.42130365659777425, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.422475106685633, | |
| "grad_norm": 1.3477824926376343, | |
| "learning_rate": 0.00026200579290369295, | |
| "loss": 0.4962, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.422475106685633, | |
| "eval_loss": 0.21994474530220032, | |
| "eval_runtime": 77.7546, | |
| "eval_samples_per_second": 19.021, | |
| "eval_steps_per_second": 2.379, | |
| "eval_wer": 0.4205087440381558, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.600284495021337, | |
| "grad_norm": 1.08402419090271, | |
| "learning_rate": 0.00025657494569152783, | |
| "loss": 0.486, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.600284495021337, | |
| "eval_loss": 0.21454988420009613, | |
| "eval_runtime": 77.6412, | |
| "eval_samples_per_second": 19.049, | |
| "eval_steps_per_second": 2.383, | |
| "eval_wer": 0.41671082847553437, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.7780938833570412, | |
| "grad_norm": 0.5935032963752747, | |
| "learning_rate": 0.00025114409847936276, | |
| "loss": 0.47, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.7780938833570412, | |
| "eval_loss": 0.21590569615364075, | |
| "eval_runtime": 77.5792, | |
| "eval_samples_per_second": 19.064, | |
| "eval_steps_per_second": 2.385, | |
| "eval_wer": 0.4168874757110051, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.9559032716927454, | |
| "grad_norm": 1.4301828145980835, | |
| "learning_rate": 0.00024571325126719764, | |
| "loss": 0.4557, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.9559032716927454, | |
| "eval_loss": 0.2098698765039444, | |
| "eval_runtime": 78.1464, | |
| "eval_samples_per_second": 18.926, | |
| "eval_steps_per_second": 2.367, | |
| "eval_wer": 0.4135311782370606, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.1337126600284497, | |
| "grad_norm": 1.2725244760513306, | |
| "learning_rate": 0.00024028240405503257, | |
| "loss": 0.4514, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.1337126600284497, | |
| "eval_loss": 0.20907503366470337, | |
| "eval_runtime": 78.1315, | |
| "eval_samples_per_second": 18.93, | |
| "eval_steps_per_second": 2.368, | |
| "eval_wer": 0.4099982335276453, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.3115220483641536, | |
| "grad_norm": 1.0271036624908447, | |
| "learning_rate": 0.00023485155684286748, | |
| "loss": 0.4539, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.3115220483641536, | |
| "eval_loss": 0.2038286179304123, | |
| "eval_runtime": 77.8532, | |
| "eval_samples_per_second": 18.997, | |
| "eval_steps_per_second": 2.376, | |
| "eval_wer": 0.40160748984278394, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.4893314366998576, | |
| "grad_norm": 1.0436575412750244, | |
| "learning_rate": 0.0002294315713251267, | |
| "loss": 0.439, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.4893314366998576, | |
| "eval_loss": 0.20407754182815552, | |
| "eval_runtime": 78.101, | |
| "eval_samples_per_second": 18.937, | |
| "eval_steps_per_second": 2.369, | |
| "eval_wer": 0.4024907260201378, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.667140825035562, | |
| "grad_norm": 1.0521398782730103, | |
| "learning_rate": 0.00022400072411296162, | |
| "loss": 0.4378, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.667140825035562, | |
| "eval_loss": 0.20021408796310425, | |
| "eval_runtime": 78.1969, | |
| "eval_samples_per_second": 18.914, | |
| "eval_steps_per_second": 2.366, | |
| "eval_wer": 0.39162692103868574, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.844950213371266, | |
| "grad_norm": 0.716386616230011, | |
| "learning_rate": 0.0002185698769007965, | |
| "loss": 0.4347, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.844950213371266, | |
| "eval_loss": 0.19606797397136688, | |
| "eval_runtime": 78.0563, | |
| "eval_samples_per_second": 18.948, | |
| "eval_steps_per_second": 2.37, | |
| "eval_wer": 0.39109697933227344, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.0227596017069702, | |
| "grad_norm": 0.6210708022117615, | |
| "learning_rate": 0.0002131390296886314, | |
| "loss": 0.4278, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.0227596017069702, | |
| "eval_loss": 0.1994515061378479, | |
| "eval_runtime": 78.0396, | |
| "eval_samples_per_second": 18.952, | |
| "eval_steps_per_second": 2.371, | |
| "eval_wer": 0.3923335099805688, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.200568990042674, | |
| "grad_norm": 0.4629976153373718, | |
| "learning_rate": 0.0002077081824764663, | |
| "loss": 0.4117, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.200568990042674, | |
| "eval_loss": 0.19594013690948486, | |
| "eval_runtime": 77.8452, | |
| "eval_samples_per_second": 18.999, | |
| "eval_steps_per_second": 2.377, | |
| "eval_wer": 0.38915385974209504, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.3783783783783785, | |
| "grad_norm": 0.9285233020782471, | |
| "learning_rate": 0.0002022773352643012, | |
| "loss": 0.4149, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.3783783783783785, | |
| "eval_loss": 0.1925920844078064, | |
| "eval_runtime": 77.8068, | |
| "eval_samples_per_second": 19.009, | |
| "eval_steps_per_second": 2.378, | |
| "eval_wer": 0.3858858858858859, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.5561877667140824, | |
| "grad_norm": 1.0625221729278564, | |
| "learning_rate": 0.00019684648805213612, | |
| "loss": 0.4148, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.5561877667140824, | |
| "eval_loss": 0.1958448737859726, | |
| "eval_runtime": 77.9401, | |
| "eval_samples_per_second": 18.976, | |
| "eval_steps_per_second": 2.374, | |
| "eval_wer": 0.38040982158629216, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.733997155049787, | |
| "grad_norm": 0.7951219081878662, | |
| "learning_rate": 0.00019142650253439536, | |
| "loss": 0.4009, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.733997155049787, | |
| "eval_loss": 0.19297942519187927, | |
| "eval_runtime": 77.4746, | |
| "eval_samples_per_second": 19.09, | |
| "eval_steps_per_second": 2.388, | |
| "eval_wer": 0.37899664370252606, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.9118065433854907, | |
| "grad_norm": 0.689900279045105, | |
| "learning_rate": 0.00018599565532223026, | |
| "loss": 0.4174, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.9118065433854907, | |
| "eval_loss": 0.19552023708820343, | |
| "eval_runtime": 77.7104, | |
| "eval_samples_per_second": 19.032, | |
| "eval_steps_per_second": 2.381, | |
| "eval_wer": 0.3822646175587352, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.089615931721195, | |
| "grad_norm": 0.5333609580993652, | |
| "learning_rate": 0.00018056480811006514, | |
| "loss": 0.4012, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 4.089615931721195, | |
| "eval_loss": 0.19501054286956787, | |
| "eval_runtime": 77.674, | |
| "eval_samples_per_second": 19.041, | |
| "eval_steps_per_second": 2.382, | |
| "eval_wer": 0.3812047341459106, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 4.2674253200568995, | |
| "grad_norm": 1.5428721904754639, | |
| "learning_rate": 0.00017513396089790005, | |
| "loss": 0.3974, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 4.2674253200568995, | |
| "eval_loss": 0.19340351223945618, | |
| "eval_runtime": 78.0603, | |
| "eval_samples_per_second": 18.947, | |
| "eval_steps_per_second": 2.37, | |
| "eval_wer": 0.3773184949655538, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 4.445234708392603, | |
| "grad_norm": 0.44063669443130493, | |
| "learning_rate": 0.00016970311368573495, | |
| "loss": 0.3943, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 4.445234708392603, | |
| "eval_loss": 0.18450064957141876, | |
| "eval_runtime": 78.009, | |
| "eval_samples_per_second": 18.959, | |
| "eval_steps_per_second": 2.372, | |
| "eval_wer": 0.37201907790143085, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 4.623044096728307, | |
| "grad_norm": 0.9376386404037476, | |
| "learning_rate": 0.0001642831281679942, | |
| "loss": 0.4071, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 4.623044096728307, | |
| "eval_loss": 0.1920066624879837, | |
| "eval_runtime": 77.3116, | |
| "eval_samples_per_second": 19.13, | |
| "eval_steps_per_second": 2.393, | |
| "eval_wer": 0.3838544426779721, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 4.800853485064011, | |
| "grad_norm": 0.6557429432868958, | |
| "learning_rate": 0.0001588522809558291, | |
| "loss": 0.3968, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 4.800853485064011, | |
| "eval_loss": 0.1866944283246994, | |
| "eval_runtime": 77.5081, | |
| "eval_samples_per_second": 19.082, | |
| "eval_steps_per_second": 2.387, | |
| "eval_wer": 0.37431549196255076, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 4.978662873399715, | |
| "grad_norm": 0.5748176574707031, | |
| "learning_rate": 0.000153421433743664, | |
| "loss": 0.3795, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.978662873399715, | |
| "eval_loss": 0.18717192113399506, | |
| "eval_runtime": 77.4836, | |
| "eval_samples_per_second": 19.088, | |
| "eval_steps_per_second": 2.388, | |
| "eval_wer": 0.37131248895954777, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 5.15647226173542, | |
| "grad_norm": 0.8843936920166016, | |
| "learning_rate": 0.0001479905865314989, | |
| "loss": 0.3856, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 5.15647226173542, | |
| "eval_loss": 0.18692350387573242, | |
| "eval_runtime": 77.4161, | |
| "eval_samples_per_second": 19.105, | |
| "eval_steps_per_second": 2.39, | |
| "eval_wer": 0.3736972266384031, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 5.334281650071124, | |
| "grad_norm": 1.5647237300872803, | |
| "learning_rate": 0.0001425597393193338, | |
| "loss": 0.3706, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 5.334281650071124, | |
| "eval_loss": 0.19033658504486084, | |
| "eval_runtime": 77.561, | |
| "eval_samples_per_second": 19.069, | |
| "eval_steps_per_second": 2.385, | |
| "eval_wer": 0.37661190602367073, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 5.512091038406828, | |
| "grad_norm": 2.1427793502807617, | |
| "learning_rate": 0.00013713975380159305, | |
| "loss": 0.3784, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 5.512091038406828, | |
| "eval_loss": 0.18607346713542938, | |
| "eval_runtime": 77.668, | |
| "eval_samples_per_second": 19.043, | |
| "eval_steps_per_second": 2.382, | |
| "eval_wer": 0.3683094859565448, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 5.689900426742532, | |
| "grad_norm": 2.697434186935425, | |
| "learning_rate": 0.00013170890658942793, | |
| "loss": 0.3777, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 5.689900426742532, | |
| "eval_loss": 0.1866033524274826, | |
| "eval_runtime": 77.9873, | |
| "eval_samples_per_second": 18.965, | |
| "eval_steps_per_second": 2.372, | |
| "eval_wer": 0.37131248895954777, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 5.867709815078236, | |
| "grad_norm": 1.5033948421478271, | |
| "learning_rate": 0.00012627805937726286, | |
| "loss": 0.3861, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 5.867709815078236, | |
| "eval_loss": 0.18121445178985596, | |
| "eval_runtime": 77.6152, | |
| "eval_samples_per_second": 19.056, | |
| "eval_steps_per_second": 2.384, | |
| "eval_wer": 0.3637166578343049, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 6.0455192034139404, | |
| "grad_norm": 0.5671353936195374, | |
| "learning_rate": 0.00012084721216509774, | |
| "loss": 0.3711, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 6.0455192034139404, | |
| "eval_loss": 0.18417000770568848, | |
| "eval_runtime": 77.4756, | |
| "eval_samples_per_second": 19.09, | |
| "eval_steps_per_second": 2.388, | |
| "eval_wer": 0.3667196608373079, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 6.223328591749644, | |
| "grad_norm": 1.0996285676956177, | |
| "learning_rate": 0.00011542722664735697, | |
| "loss": 0.374, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 6.223328591749644, | |
| "eval_loss": 0.18148785829544067, | |
| "eval_runtime": 77.7259, | |
| "eval_samples_per_second": 19.028, | |
| "eval_steps_per_second": 2.38, | |
| "eval_wer": 0.3617735382441265, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 6.401137980085348, | |
| "grad_norm": 0.6861454844474792, | |
| "learning_rate": 0.00010999637943519187, | |
| "loss": 0.3539, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 6.401137980085348, | |
| "eval_loss": 0.18153779208660126, | |
| "eval_runtime": 77.7594, | |
| "eval_samples_per_second": 19.02, | |
| "eval_steps_per_second": 2.379, | |
| "eval_wer": 0.3646882176293941, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 6.578947368421053, | |
| "grad_norm": 1.423963189125061, | |
| "learning_rate": 0.00010456553222302678, | |
| "loss": 0.3625, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 6.578947368421053, | |
| "eval_loss": 0.17849859595298767, | |
| "eval_runtime": 77.8982, | |
| "eval_samples_per_second": 18.986, | |
| "eval_steps_per_second": 2.375, | |
| "eval_wer": 0.3588588588588589, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 6.756756756756757, | |
| "grad_norm": 0.9670858383178711, | |
| "learning_rate": 9.913468501086169e-05, | |
| "loss": 0.3599, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 6.756756756756757, | |
| "eval_loss": 0.17952215671539307, | |
| "eval_runtime": 77.3124, | |
| "eval_samples_per_second": 19.13, | |
| "eval_steps_per_second": 2.393, | |
| "eval_wer": 0.362126832715068, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 6.934566145092461, | |
| "grad_norm": 0.523705005645752, | |
| "learning_rate": 9.370383779869658e-05, | |
| "loss": 0.3654, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 6.934566145092461, | |
| "eval_loss": 0.1822131723165512, | |
| "eval_runtime": 77.744, | |
| "eval_samples_per_second": 19.024, | |
| "eval_steps_per_second": 2.38, | |
| "eval_wer": 0.36239180356827416, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 7.112375533428165, | |
| "grad_norm": 1.0047301054000854, | |
| "learning_rate": 8.82729905865315e-05, | |
| "loss": 0.3693, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 7.112375533428165, | |
| "eval_loss": 0.17921391129493713, | |
| "eval_runtime": 77.3373, | |
| "eval_samples_per_second": 19.124, | |
| "eval_steps_per_second": 2.392, | |
| "eval_wer": 0.3611552729199788, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 7.290184921763869, | |
| "grad_norm": 0.6277859807014465, | |
| "learning_rate": 8.285300506879071e-05, | |
| "loss": 0.3519, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 7.290184921763869, | |
| "eval_loss": 0.18002206087112427, | |
| "eval_runtime": 77.6246, | |
| "eval_samples_per_second": 19.053, | |
| "eval_steps_per_second": 2.383, | |
| "eval_wer": 0.36751457339692634, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 7.467994310099574, | |
| "grad_norm": 0.8712663650512695, | |
| "learning_rate": 7.743301955104996e-05, | |
| "loss": 0.3553, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 7.467994310099574, | |
| "eval_loss": 0.1808168739080429, | |
| "eval_runtime": 77.4403, | |
| "eval_samples_per_second": 19.099, | |
| "eval_steps_per_second": 2.389, | |
| "eval_wer": 0.36398162868751105, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 7.6458036984352775, | |
| "grad_norm": 1.000291109085083, | |
| "learning_rate": 7.200217233888485e-05, | |
| "loss": 0.3451, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 7.6458036984352775, | |
| "eval_loss": 0.18079166114330292, | |
| "eval_runtime": 77.717, | |
| "eval_samples_per_second": 19.031, | |
| "eval_steps_per_second": 2.38, | |
| "eval_wer": 0.36195018547959723, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 7.823613086770981, | |
| "grad_norm": 0.9150896668434143, | |
| "learning_rate": 6.657132512671976e-05, | |
| "loss": 0.3558, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 7.823613086770981, | |
| "eval_loss": 0.17938227951526642, | |
| "eval_runtime": 77.6629, | |
| "eval_samples_per_second": 19.044, | |
| "eval_steps_per_second": 2.382, | |
| "eval_wer": 0.36097862568450806, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 8.001422475106686, | |
| "grad_norm": 0.4225611686706543, | |
| "learning_rate": 6.114047791455467e-05, | |
| "loss": 0.3595, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 8.001422475106686, | |
| "eval_loss": 0.17718034982681274, | |
| "eval_runtime": 78.0056, | |
| "eval_samples_per_second": 18.96, | |
| "eval_steps_per_second": 2.372, | |
| "eval_wer": 0.3576223282105635, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 8.17923186344239, | |
| "grad_norm": 0.7559336423873901, | |
| "learning_rate": 5.570963070238957e-05, | |
| "loss": 0.3404, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 8.17923186344239, | |
| "eval_loss": 0.17881204187870026, | |
| "eval_runtime": 78.2489, | |
| "eval_samples_per_second": 18.901, | |
| "eval_steps_per_second": 2.364, | |
| "eval_wer": 0.35806394629924043, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 8.357041251778094, | |
| "grad_norm": 1.2298369407653809, | |
| "learning_rate": 5.027878349022447e-05, | |
| "loss": 0.3593, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 8.357041251778094, | |
| "eval_loss": 0.1782107651233673, | |
| "eval_runtime": 77.8969, | |
| "eval_samples_per_second": 18.987, | |
| "eval_steps_per_second": 2.375, | |
| "eval_wer": 0.357975622681505, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 8.534850640113799, | |
| "grad_norm": 0.5965376496315002, | |
| "learning_rate": 4.4847936278059375e-05, | |
| "loss": 0.3471, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 8.534850640113799, | |
| "eval_loss": 0.17967215180397034, | |
| "eval_runtime": 77.691, | |
| "eval_samples_per_second": 19.037, | |
| "eval_steps_per_second": 2.381, | |
| "eval_wer": 0.3606253312135665, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 8.712660028449502, | |
| "grad_norm": 0.3367222249507904, | |
| "learning_rate": 3.941708906589427e-05, | |
| "loss": 0.3497, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 8.712660028449502, | |
| "eval_loss": 0.17775095999240875, | |
| "eval_runtime": 78.061, | |
| "eval_samples_per_second": 18.947, | |
| "eval_steps_per_second": 2.37, | |
| "eval_wer": 0.35877053524112346, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 8.890469416785207, | |
| "grad_norm": 1.0665998458862305, | |
| "learning_rate": 3.399710354815351e-05, | |
| "loss": 0.3398, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 8.890469416785207, | |
| "eval_loss": 0.1774686574935913, | |
| "eval_runtime": 77.6438, | |
| "eval_samples_per_second": 19.049, | |
| "eval_steps_per_second": 2.383, | |
| "eval_wer": 0.3583289171524466, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 9.06827880512091, | |
| "grad_norm": 1.8358111381530762, | |
| "learning_rate": 2.8566256335988413e-05, | |
| "loss": 0.3444, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 9.06827880512091, | |
| "eval_loss": 0.1796201765537262, | |
| "eval_runtime": 77.7985, | |
| "eval_samples_per_second": 19.011, | |
| "eval_steps_per_second": 2.378, | |
| "eval_wer": 0.35859388800565273, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 9.246088193456615, | |
| "grad_norm": 1.32257878780365, | |
| "learning_rate": 2.3135409123823315e-05, | |
| "loss": 0.3366, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 9.246088193456615, | |
| "eval_loss": 0.1784891039133072, | |
| "eval_runtime": 77.5756, | |
| "eval_samples_per_second": 19.065, | |
| "eval_steps_per_second": 2.385, | |
| "eval_wer": 0.35735735735735735, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 9.42389758179232, | |
| "grad_norm": 1.8535629510879517, | |
| "learning_rate": 1.7704561911658217e-05, | |
| "loss": 0.3434, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 9.42389758179232, | |
| "eval_loss": 0.17805208265781403, | |
| "eval_runtime": 77.8845, | |
| "eval_samples_per_second": 18.99, | |
| "eval_steps_per_second": 2.375, | |
| "eval_wer": 0.3592121533298004, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 9.601706970128022, | |
| "grad_norm": 1.9332554340362549, | |
| "learning_rate": 1.228457639391745e-05, | |
| "loss": 0.3426, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 9.601706970128022, | |
| "eval_loss": 0.17857009172439575, | |
| "eval_runtime": 77.3945, | |
| "eval_samples_per_second": 19.11, | |
| "eval_steps_per_second": 2.39, | |
| "eval_wer": 0.35930047694753575, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 9.779516358463727, | |
| "grad_norm": 1.1602191925048828, | |
| "learning_rate": 6.8537291817523524e-06, | |
| "loss": 0.3496, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 9.779516358463727, | |
| "eval_loss": 0.17868547141551971, | |
| "eval_runtime": 77.5791, | |
| "eval_samples_per_second": 19.064, | |
| "eval_steps_per_second": 2.385, | |
| "eval_wer": 0.3590355060943296, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 9.95732574679943, | |
| "grad_norm": 1.2770110368728638, | |
| "learning_rate": 1.4228819695872554e-06, | |
| "loss": 0.334, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 9.95732574679943, | |
| "eval_loss": 0.17876511812210083, | |
| "eval_runtime": 78.0274, | |
| "eval_samples_per_second": 18.955, | |
| "eval_steps_per_second": 2.371, | |
| "eval_wer": 0.35877053524112346, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 28120, | |
| "total_flos": 2.1329860467760157e+19, | |
| "train_loss": 0.48163046043254915, | |
| "train_runtime": 15300.4436, | |
| "train_samples_per_second": 7.349, | |
| "train_steps_per_second": 1.838 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 28120, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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": 2.1329860467760157e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |