End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.3996790945529938,
+    "eval_runtime": 71.9707,
+    "eval_samples": 1647,
+    "eval_samples_per_second": 22.884,
+    "eval_steps_per_second": 2.862,
+    "eval_wer": 0.3537943008885711,
+    "train_loss": 1.071420776297193,
+    "train_runtime": 3969.9688,
+    "train_samples": 3478,
+    "train_samples_per_second": 13.141,
+    "train_steps_per_second": 0.412
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.3996790945529938,
+    "eval_runtime": 71.9707,
+    "eval_samples": 1647,
+    "eval_samples_per_second": 22.884,
+    "eval_steps_per_second": 2.862,
+    "eval_wer": 0.3537943008885711
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 15.0,
+    "train_loss": 1.071420776297193,
+    "train_runtime": 3969.9688,
+    "train_samples": 3478,
+    "train_samples_per_second": 13.141,
+    "train_steps_per_second": 0.412
+}

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": 3.606069326400757,
+      "eval_runtime": 74.9143,
+      "eval_samples_per_second": 21.985,
+      "eval_steps_per_second": 2.75,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 3.0202510356903076,
+      "eval_runtime": 73.818,
+      "eval_samples_per_second": 22.312,
+      "eval_steps_per_second": 2.791,
+      "eval_wer": 0.9998978653865795,
+      "step": 200
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 0.9479126930236816,
+      "eval_runtime": 73.9013,
+      "eval_samples_per_second": 22.286,
+      "eval_steps_per_second": 2.788,
+      "eval_wer": 0.791645388622204,
+      "step": 300
+    },
+    {
+      "epoch": 3.67,
+      "eval_loss": 0.6023567914962769,
+      "eval_runtime": 73.7383,
+      "eval_samples_per_second": 22.336,
+      "eval_steps_per_second": 2.794,
+      "eval_wer": 0.6285364109896844,
+      "step": 400
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.0002988,
+      "loss": 3.1561,
+      "step": 500
+    },
+    {
+      "epoch": 4.59,
+      "eval_loss": 0.511204183101654,
+      "eval_runtime": 73.7804,
+      "eval_samples_per_second": 22.323,
+      "eval_steps_per_second": 2.792,
+      "eval_wer": 0.5369216627515064,
+      "step": 500
+    },
+    {
+      "epoch": 5.5,
+      "eval_loss": 0.45808982849121094,
+      "eval_runtime": 73.2344,
+      "eval_samples_per_second": 22.489,
+      "eval_steps_per_second": 2.813,
+      "eval_wer": 0.4900418751915024,
+      "step": 600
+    },
+    {
+      "epoch": 6.42,
+      "eval_loss": 0.43211525678634644,
+      "eval_runtime": 73.883,
+      "eval_samples_per_second": 22.292,
+      "eval_steps_per_second": 2.788,
+      "eval_wer": 0.4632826064753345,
+      "step": 700
+    },
+    {
+      "epoch": 7.34,
+      "eval_loss": 0.4252323508262634,
+      "eval_runtime": 73.1676,
+      "eval_samples_per_second": 22.51,
+      "eval_steps_per_second": 2.815,
+      "eval_wer": 0.43999591461546317,
+      "step": 800
+    },
+    {
+      "epoch": 8.26,
+      "eval_loss": 0.4203605055809021,
+      "eval_runtime": 73.3352,
+      "eval_samples_per_second": 22.459,
+      "eval_steps_per_second": 2.809,
+      "eval_wer": 0.4229394341742416,
+      "step": 900
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 0.00016837004405286342,
+      "loss": 0.2247,
+      "step": 1000
+    },
+    {
+      "epoch": 9.17,
+      "eval_loss": 0.3948039710521698,
+      "eval_runtime": 73.35,
+      "eval_samples_per_second": 22.454,
+      "eval_steps_per_second": 2.808,
+      "eval_wer": 0.3970993769788581,
+      "step": 1000
+    },
+    {
+      "epoch": 10.09,
+      "eval_loss": 0.3996581435203552,
+      "eval_runtime": 73.0894,
+      "eval_samples_per_second": 22.534,
+      "eval_steps_per_second": 2.818,
+      "eval_wer": 0.39628230007149423,
+      "step": 1100
+    },
+    {
+      "epoch": 11.01,
+      "eval_loss": 0.4157370626926422,
+      "eval_runtime": 73.6448,
+      "eval_samples_per_second": 22.364,
+      "eval_steps_per_second": 2.797,
+      "eval_wer": 0.3894392809723215,
+      "step": 1200
+    },
+    {
+      "epoch": 11.93,
+      "eval_loss": 0.41419464349746704,
+      "eval_runtime": 72.7569,
+      "eval_samples_per_second": 22.637,
+      "eval_steps_per_second": 2.831,
+      "eval_wer": 0.38545603104892245,
+      "step": 1300
+    },
+    {
+      "epoch": 12.84,
+      "eval_loss": 0.4107553958892822,
+      "eval_runtime": 72.5948,
+      "eval_samples_per_second": 22.688,
+      "eval_steps_per_second": 2.838,
+      "eval_wer": 0.36380349300377895,
+      "step": 1400
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 3.621145374449339e-05,
+      "loss": 0.1022,
+      "step": 1500
+    },
+    {
+      "epoch": 13.76,
+      "eval_loss": 0.3928930163383484,
+      "eval_runtime": 73.1498,
+      "eval_samples_per_second": 22.515,
+      "eval_steps_per_second": 2.816,
+      "eval_wer": 0.3617608007353692,
+      "step": 1500
+    },
+    {
+      "epoch": 14.68,
+      "eval_loss": 0.4004472494125366,
+      "eval_runtime": 72.6133,
+      "eval_samples_per_second": 22.682,
+      "eval_steps_per_second": 2.837,
+      "eval_wer": 0.3544071085690941,
+      "step": 1600
+    },
+    {
+      "epoch": 15.0,
+      "step": 1635,
+      "total_flos": 6.464559143942451e+18,
+      "train_loss": 1.071420776297193,
+      "train_runtime": 3969.9688,
+      "train_samples_per_second": 13.141,
+      "train_steps_per_second": 0.412
+    }
+  ],
+  "max_steps": 1635,
+  "num_train_epochs": 15,
+  "total_flos": 6.464559143942451e+18,
+  "trial_name": null,
+  "trial_params": null
+}