End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.37601369619369507,
+    "eval_runtime": 107.4226,
+    "eval_samples": 1647,
+    "eval_samples_per_second": 15.332,
+    "eval_steps_per_second": 1.918,
+    "eval_wer": 0.35205801246042284,
+    "train_loss": 1.070046563688039,
+    "train_runtime": 6409.9307,
+    "train_samples": 3478,
+    "train_samples_per_second": 8.139,
+    "train_steps_per_second": 0.255
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.37601369619369507,
+    "eval_runtime": 107.4226,
+    "eval_samples": 1647,
+    "eval_samples_per_second": 15.332,
+    "eval_steps_per_second": 1.918,
+    "eval_wer": 0.35205801246042284
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 15.0,
+    "train_loss": 1.070046563688039,
+    "train_runtime": 6409.9307,
+    "train_samples": 3478,
+    "train_samples_per_second": 8.139,
+    "train_steps_per_second": 0.255
+}

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": 3.599937915802002,
+      "eval_runtime": 108.2949,
+      "eval_samples_per_second": 15.208,
+      "eval_steps_per_second": 1.902,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 2.994201183319092,
+      "eval_runtime": 108.6924,
+      "eval_samples_per_second": 15.153,
+      "eval_steps_per_second": 1.895,
+      "eval_wer": 0.9998978653865795,
+      "step": 200
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 0.9031277298927307,
+      "eval_runtime": 110.4415,
+      "eval_samples_per_second": 14.913,
+      "eval_steps_per_second": 1.865,
+      "eval_wer": 0.7882749463793279,
+      "step": 300
+    },
+    {
+      "epoch": 3.67,
+      "eval_loss": 0.5929930210113525,
+      "eval_runtime": 108.5294,
+      "eval_samples_per_second": 15.176,
+      "eval_steps_per_second": 1.898,
+      "eval_wer": 0.6226126034112961,
+      "step": 400
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.0002988,
+      "loss": 3.1501,
+      "step": 500
+    },
+    {
+      "epoch": 4.59,
+      "eval_loss": 0.4966978430747986,
+      "eval_runtime": 109.0578,
+      "eval_samples_per_second": 15.102,
+      "eval_steps_per_second": 1.889,
+      "eval_wer": 0.523439893780002,
+      "step": 500
+    },
+    {
+      "epoch": 5.5,
+      "eval_loss": 0.4888415038585663,
+      "eval_runtime": 109.9028,
+      "eval_samples_per_second": 14.986,
+      "eval_steps_per_second": 1.874,
+      "eval_wer": 0.5052599325911551,
+      "step": 600
+    },
+    {
+      "epoch": 6.42,
+      "eval_loss": 0.4392887353897095,
+      "eval_runtime": 109.1459,
+      "eval_samples_per_second": 15.09,
+      "eval_steps_per_second": 1.887,
+      "eval_wer": 0.4745174139515882,
+      "step": 700
+    },
+    {
+      "epoch": 7.34,
+      "eval_loss": 0.43620193004608154,
+      "eval_runtime": 109.4439,
+      "eval_samples_per_second": 15.049,
+      "eval_steps_per_second": 1.882,
+      "eval_wer": 0.437034010826269,
+      "step": 800
+    },
+    {
+      "epoch": 8.26,
+      "eval_loss": 0.4384012520313263,
+      "eval_runtime": 110.6137,
+      "eval_samples_per_second": 14.89,
+      "eval_steps_per_second": 1.862,
+      "eval_wer": 0.42242876110713923,
+      "step": 900
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 0.00016837004405286342,
+      "loss": 0.2259,
+      "step": 1000
+    },
+    {
+      "epoch": 9.17,
+      "eval_loss": 0.416930228471756,
+      "eval_runtime": 108.6343,
+      "eval_samples_per_second": 15.161,
+      "eval_steps_per_second": 1.896,
+      "eval_wer": 0.4008783576754162,
+      "step": 1000
+    },
+    {
+      "epoch": 10.09,
+      "eval_loss": 0.39647600054740906,
+      "eval_runtime": 108.7436,
+      "eval_samples_per_second": 15.146,
+      "eval_steps_per_second": 1.894,
+      "eval_wer": 0.3887243386783781,
+      "step": 1100
+    },
+    {
+      "epoch": 11.01,
+      "eval_loss": 0.4072294235229492,
+      "eval_runtime": 109.186,
+      "eval_samples_per_second": 15.084,
+      "eval_steps_per_second": 1.887,
+      "eval_wer": 0.38402614646103567,
+      "step": 1200
+    },
+    {
+      "epoch": 11.93,
+      "eval_loss": 0.3937165439128876,
+      "eval_runtime": 109.4893,
+      "eval_samples_per_second": 15.043,
+      "eval_steps_per_second": 1.881,
+      "eval_wer": 0.3703401082626902,
+      "step": 1300
+    },
+    {
+      "epoch": 12.84,
+      "eval_loss": 0.3900948762893677,
+      "eval_runtime": 108.9312,
+      "eval_samples_per_second": 15.12,
+      "eval_steps_per_second": 1.891,
+      "eval_wer": 0.3655397814319273,
+      "step": 1400
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 3.621145374449339e-05,
+      "loss": 0.1024,
+      "step": 1500
+    },
+    {
+      "epoch": 13.76,
+      "eval_loss": 0.38345181941986084,
+      "eval_runtime": 109.4545,
+      "eval_samples_per_second": 15.047,
+      "eval_steps_per_second": 1.882,
+      "eval_wer": 0.3559391277704014,
+      "step": 1500
+    },
+    {
+      "epoch": 14.68,
+      "eval_loss": 0.3779906928539276,
+      "eval_runtime": 109.5836,
+      "eval_samples_per_second": 15.03,
+      "eval_steps_per_second": 1.88,
+      "eval_wer": 0.3533857624348892,
+      "step": 1600
+    },
+    {
+      "epoch": 15.0,
+      "step": 1635,
+      "total_flos": 6.464559143942451e+18,
+      "train_loss": 1.070046563688039,
+      "train_runtime": 6409.9307,
+      "train_samples_per_second": 8.139,
+      "train_steps_per_second": 0.255
+    }
+  ],
+  "max_steps": 1635,
+  "num_train_epochs": 15,
+  "total_flos": 6.464559143942451e+18,
+  "trial_name": null,
+  "trial_params": null
+}