End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.37983575463294983,
+    "eval_runtime": 126.1466,
+    "eval_samples": 1647,
+    "eval_samples_per_second": 13.056,
+    "eval_steps_per_second": 1.633,
+    "eval_wer": 0.34480645490756817,
+    "train_loss": 1.0781283684826772,
+    "train_runtime": 7410.3512,
+    "train_samples": 3478,
+    "train_samples_per_second": 7.04,
+    "train_steps_per_second": 0.221
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 15.0,
+    "eval_loss": 0.37983575463294983,
+    "eval_runtime": 126.1466,
+    "eval_samples": 1647,
+    "eval_samples_per_second": 13.056,
+    "eval_steps_per_second": 1.633,
+    "eval_wer": 0.34480645490756817
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 15.0,
+    "train_loss": 1.0781283684826772,
+    "train_runtime": 7410.3512,
+    "train_samples": 3478,
+    "train_samples_per_second": 7.04,
+    "train_steps_per_second": 0.221
+}

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": 3.5931618213653564,
+      "eval_runtime": 126.9872,
+      "eval_samples_per_second": 12.97,
+      "eval_steps_per_second": 1.622,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 3.0185344219207764,
+      "eval_runtime": 122.4223,
+      "eval_samples_per_second": 13.453,
+      "eval_steps_per_second": 1.683,
+      "eval_wer": 0.9998978653865795,
+      "step": 200
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 0.9357358813285828,
+      "eval_runtime": 124.4606,
+      "eval_samples_per_second": 13.233,
+      "eval_steps_per_second": 1.655,
+      "eval_wer": 0.8007353692166275,
+      "step": 300
+    },
+    {
+      "epoch": 3.67,
+      "eval_loss": 0.5944719910621643,
+      "eval_runtime": 124.5538,
+      "eval_samples_per_second": 13.223,
+      "eval_steps_per_second": 1.654,
+      "eval_wer": 0.63180471861914,
+      "step": 400
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.0002988,
+      "loss": 3.1829,
+      "step": 500
+    },
+    {
+      "epoch": 4.59,
+      "eval_loss": 0.4931354224681854,
+      "eval_runtime": 132.5462,
+      "eval_samples_per_second": 12.426,
+      "eval_steps_per_second": 1.554,
+      "eval_wer": 0.5265039321826167,
+      "step": 500
+    },
+    {
+      "epoch": 5.5,
+      "eval_loss": 0.47567814588546753,
+      "eval_runtime": 131.8607,
+      "eval_samples_per_second": 12.49,
+      "eval_steps_per_second": 1.562,
+      "eval_wer": 0.47839852926156673,
+      "step": 600
+    },
+    {
+      "epoch": 6.42,
+      "eval_loss": 0.4281878173351288,
+      "eval_runtime": 125.8942,
+      "eval_samples_per_second": 13.082,
+      "eval_steps_per_second": 1.636,
+      "eval_wer": 0.45398835665407006,
+      "step": 700
+    },
+    {
+      "epoch": 7.34,
+      "eval_loss": 0.3994791507720947,
+      "eval_runtime": 125.4372,
+      "eval_samples_per_second": 13.13,
+      "eval_steps_per_second": 1.642,
+      "eval_wer": 0.4251863956694924,
+      "step": 800
+    },
+    {
+      "epoch": 8.26,
+      "eval_loss": 0.4046003818511963,
+      "eval_runtime": 121.2245,
+      "eval_samples_per_second": 13.586,
+      "eval_steps_per_second": 1.699,
+      "eval_wer": 0.4148707997140231,
+      "step": 900
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 0.00016837004405286342,
+      "loss": 0.2215,
+      "step": 1000
+    },
+    {
+      "epoch": 9.17,
+      "eval_loss": 0.4047873914241791,
+      "eval_runtime": 124.6186,
+      "eval_samples_per_second": 13.216,
+      "eval_steps_per_second": 1.653,
+      "eval_wer": 0.39505668471044836,
+      "step": 1000
+    },
+    {
+      "epoch": 10.09,
+      "eval_loss": 0.39440903067588806,
+      "eval_runtime": 125.2639,
+      "eval_samples_per_second": 13.148,
+      "eval_steps_per_second": 1.645,
+      "eval_wer": 0.38647737718312736,
+      "step": 1100
+    },
+    {
+      "epoch": 11.01,
+      "eval_loss": 0.38526079058647156,
+      "eval_runtime": 122.7703,
+      "eval_samples_per_second": 13.415,
+      "eval_steps_per_second": 1.678,
+      "eval_wer": 0.3643141660708814,
+      "step": 1200
+    },
+    {
+      "epoch": 11.93,
+      "eval_loss": 0.395017147064209,
+      "eval_runtime": 125.4537,
+      "eval_samples_per_second": 13.128,
+      "eval_steps_per_second": 1.642,
+      "eval_wer": 0.36319068532325605,
+      "step": 1300
+    },
+    {
+      "epoch": 12.84,
+      "eval_loss": 0.3835676610469818,
+      "eval_runtime": 124.6558,
+      "eval_samples_per_second": 13.212,
+      "eval_steps_per_second": 1.653,
+      "eval_wer": 0.350628127872536,
+      "step": 1400
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 3.621145374449339e-05,
+      "loss": 0.1009,
+      "step": 1500
+    },
+    {
+      "epoch": 13.76,
+      "eval_loss": 0.38080519437789917,
+      "eval_runtime": 122.1546,
+      "eval_samples_per_second": 13.483,
+      "eval_steps_per_second": 1.686,
+      "eval_wer": 0.35103666632621794,
+      "step": 1500
+    },
+    {
+      "epoch": 14.68,
+      "eval_loss": 0.38068610429763794,
+      "eval_runtime": 125.0336,
+      "eval_samples_per_second": 13.172,
+      "eval_steps_per_second": 1.648,
+      "eval_wer": 0.3456235318149321,
+      "step": 1600
+    },
+    {
+      "epoch": 15.0,
+      "step": 1635,
+      "total_flos": 6.464559143942451e+18,
+      "train_loss": 1.0781283684826772,
+      "train_runtime": 7410.3512,
+      "train_samples_per_second": 7.04,
+      "train_steps_per_second": 0.221
+    }
+  ],
+  "max_steps": 1635,
+  "num_train_epochs": 15,
+  "total_flos": 6.464559143942451e+18,
+  "trial_name": null,
+  "trial_params": null
+}