End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.38048940896987915,
+    "eval_runtime": 68.4332,
+    "eval_samples": 1647,
+    "eval_samples_per_second": 24.067,
+    "eval_steps_per_second": 3.01,
+    "eval_wer": 0.33888264732917983,
+    "train_loss": 1.0770871532801825,
+    "train_runtime": 3832.0305,
+    "train_samples": 3478,
+    "train_samples_per_second": 13.614,
+    "train_steps_per_second": 0.427
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.38048940896987915,
+    "eval_runtime": 68.4332,
+    "eval_samples": 1647,
+    "eval_samples_per_second": 24.067,
+    "eval_steps_per_second": 3.01,
+    "eval_wer": 0.33888264732917983
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 15.0,
+    "train_loss": 1.0770871532801825,
+    "train_runtime": 3832.0305,
+    "train_samples": 3478,
+    "train_samples_per_second": 13.614,
+    "train_steps_per_second": 0.427
+}

trainer_state.json

@@ -0,0 +1,187 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 15.0,
+  "global_step": 1635,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.92,
+      "eval_loss": 3.6031575202941895,
+      "eval_runtime": 71.2961,
+      "eval_samples_per_second": 23.101,
+      "eval_steps_per_second": 2.889,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 3.015834331512451,
+      "eval_runtime": 69.71,
+      "eval_samples_per_second": 23.626,
+      "eval_steps_per_second": 2.955,
+      "eval_wer": 0.9998978653865795,
+      "step": 200
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 0.9691545367240906,
+      "eval_runtime": 69.4092,
+      "eval_samples_per_second": 23.729,
+      "eval_steps_per_second": 2.968,
+      "eval_wer": 0.8028801960984577,
+      "step": 300
+    },
+    {
+      "epoch": 3.67,
+      "eval_loss": 0.5820251703262329,
+      "eval_runtime": 69.8188,
+      "eval_samples_per_second": 23.59,
+      "eval_steps_per_second": 2.95,
+      "eval_wer": 0.6160759881523848,
+      "step": 400
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.0002988,
+      "loss": 3.1812,
+      "step": 500
+    },
+    {
+      "epoch": 4.59,
+      "eval_loss": 0.4890775978565216,
+      "eval_runtime": 69.3063,
+      "eval_samples_per_second": 23.764,
+      "eval_steps_per_second": 2.972,
+      "eval_wer": 0.5095495863548156,
+      "step": 500
+    },
+    {
+      "epoch": 5.5,
+      "eval_loss": 0.47190332412719727,
+      "eval_runtime": 69.6259,
+      "eval_samples_per_second": 23.655,
+      "eval_steps_per_second": 2.959,
+      "eval_wer": 0.4853436829741599,
+      "step": 600
+    },
+    {
+      "epoch": 6.42,
+      "eval_loss": 0.4359598457813263,
+      "eval_runtime": 69.3092,
+      "eval_samples_per_second": 23.763,
+      "eval_steps_per_second": 2.972,
+      "eval_wer": 0.4538862220406496,
+      "step": 700
+    },
+    {
+      "epoch": 7.34,
+      "eval_loss": 0.4098358154296875,
+      "eval_runtime": 69.3406,
+      "eval_samples_per_second": 23.752,
+      "eval_steps_per_second": 2.971,
+      "eval_wer": 0.42825043407210706,
+      "step": 800
+    },
+    {
+      "epoch": 8.26,
+      "eval_loss": 0.4020206034183502,
+      "eval_runtime": 70.6229,
+      "eval_samples_per_second": 23.321,
+      "eval_steps_per_second": 2.917,
+      "eval_wer": 0.3993463384741089,
+      "step": 900
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 0.00016837004405286342,
+      "loss": 0.2212,
+      "step": 1000
+    },
+    {
+      "epoch": 9.17,
+      "eval_loss": 0.4000774919986725,
+      "eval_runtime": 68.5561,
+      "eval_samples_per_second": 24.024,
+      "eval_steps_per_second": 3.005,
+      "eval_wer": 0.3805535696047391,
+      "step": 1000
+    },
+    {
+      "epoch": 10.09,
+      "eval_loss": 0.3999648094177246,
+      "eval_runtime": 68.8568,
+      "eval_samples_per_second": 23.919,
+      "eval_steps_per_second": 2.992,
+      "eval_wer": 0.3872944540904913,
+      "step": 1100
+    },
+    {
+      "epoch": 11.01,
+      "eval_loss": 0.40702080726623535,
+      "eval_runtime": 68.6579,
+      "eval_samples_per_second": 23.989,
+      "eval_steps_per_second": 3.0,
+      "eval_wer": 0.3751404350934532,
+      "step": 1200
+    },
+    {
+      "epoch": 11.93,
+      "eval_loss": 0.3874248266220093,
+      "eval_runtime": 69.3466,
+      "eval_samples_per_second": 23.75,
+      "eval_steps_per_second": 2.971,
+      "eval_wer": 0.3551220508630375,
+      "step": 1300
+    },
+    {
+      "epoch": 12.84,
+      "eval_loss": 0.39125174283981323,
+      "eval_runtime": 70.1419,
+      "eval_samples_per_second": 23.481,
+      "eval_steps_per_second": 2.937,
+      "eval_wer": 0.35614339699724235,
+      "step": 1400
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 3.621145374449339e-05,
+      "loss": 0.0998,
+      "step": 1500
+    },
+    {
+      "epoch": 13.76,
+      "eval_loss": 0.3882182836532593,
+      "eval_runtime": 70.0328,
+      "eval_samples_per_second": 23.518,
+      "eval_steps_per_second": 2.941,
+      "eval_wer": 0.34919824328464916,
+      "step": 1500
+    },
+    {
+      "epoch": 14.68,
+      "eval_loss": 0.3794955015182495,
+      "eval_runtime": 69.1922,
+      "eval_samples_per_second": 23.803,
+      "eval_steps_per_second": 2.977,
+      "eval_wer": 0.3403125319170667,
+      "step": 1600
+    },
+    {
+      "epoch": 15.0,
+      "step": 1635,
+      "total_flos": 6.464559143942451e+18,
+      "train_loss": 1.0770871532801825,
+      "train_runtime": 3832.0305,
+      "train_samples_per_second": 13.614,
+      "train_steps_per_second": 0.427
+    }
+  ],
+  "max_steps": 1635,
+  "num_train_epochs": 15,
+  "total_flos": 6.464559143942451e+18,
+  "trial_name": null,
+  "trial_params": null
+}