End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 100.0,
+    "eval_loss": 1.0855300426483154,
+    "eval_runtime": 13.0411,
+    "eval_samples": 360,
+    "eval_samples_per_second": 27.605,
+    "eval_steps_per_second": 3.451,
+    "eval_wer": 0.475499803998432,
+    "train_loss": 0.7658213693124276,
+    "train_runtime": 5843.2467,
+    "train_samples": 856,
+    "train_samples_per_second": 14.649,
+    "train_steps_per_second": 0.924
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 100.0,
+    "eval_loss": 1.0855300426483154,
+    "eval_runtime": 13.0411,
+    "eval_samples": 360,
+    "eval_samples_per_second": 27.605,
+    "eval_steps_per_second": 3.451,
+    "eval_wer": 0.475499803998432
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 100.0,
+    "train_loss": 0.7658213693124276,
+    "train_runtime": 5843.2467,
+    "train_samples": 856,
+    "train_samples_per_second": 14.649,
+    "train_steps_per_second": 0.924
+}

trainer_state.json

@@ -0,0 +1,277 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 100.0,
+  "global_step": 5400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 3.7,
+      "learning_rate": 4.924999999999999e-05,
+      "loss": 11.9966,
+      "step": 200
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 9.924999999999999e-05,
+      "loss": 3.4607,
+      "step": 400
+    },
+    {
+      "epoch": 9.26,
+      "eval_loss": 2.7746095657348633,
+      "eval_runtime": 13.5943,
+      "eval_samples_per_second": 26.482,
+      "eval_steps_per_second": 3.31,
+      "eval_wer": 1.0415523324186593,
+      "step": 500
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 0.00014925,
+      "loss": 2.4077,
+      "step": 600
+    },
+    {
+      "epoch": 14.81,
+      "learning_rate": 0.00019925,
+      "loss": 0.5977,
+      "step": 800
+    },
+    {
+      "epoch": 18.52,
+      "learning_rate": 0.00024924999999999996,
+      "loss": 0.3442,
+      "step": 1000
+    },
+    {
+      "epoch": 18.52,
+      "eval_loss": 0.9114237427711487,
+      "eval_runtime": 13.1778,
+      "eval_samples_per_second": 27.319,
+      "eval_steps_per_second": 3.415,
+      "eval_wer": 0.591140729125833,
+      "step": 1000
+    },
+    {
+      "epoch": 22.22,
+      "learning_rate": 0.00029925,
+      "loss": 0.28,
+      "step": 1200
+    },
+    {
+      "epoch": 25.93,
+      "learning_rate": 0.0002859285714285714,
+      "loss": 0.2213,
+      "step": 1400
+    },
+    {
+      "epoch": 27.78,
+      "eval_loss": 0.9686752557754517,
+      "eval_runtime": 12.9893,
+      "eval_samples_per_second": 27.715,
+      "eval_steps_per_second": 3.464,
+      "eval_wer": 0.5750686005488044,
+      "step": 1500
+    },
+    {
+      "epoch": 29.63,
+      "learning_rate": 0.0002716428571428571,
+      "loss": 0.1746,
+      "step": 1600
+    },
+    {
+      "epoch": 33.33,
+      "learning_rate": 0.00025735714285714283,
+      "loss": 0.1472,
+      "step": 1800
+    },
+    {
+      "epoch": 37.04,
+      "learning_rate": 0.00024307142857142854,
+      "loss": 0.1242,
+      "step": 2000
+    },
+    {
+      "epoch": 37.04,
+      "eval_loss": 1.0203704833984375,
+      "eval_runtime": 13.2024,
+      "eval_samples_per_second": 27.268,
+      "eval_steps_per_second": 3.408,
+      "eval_wer": 0.5460603684829478,
+      "step": 2000
+    },
+    {
+      "epoch": 40.74,
+      "learning_rate": 0.00022878571428571427,
+      "loss": 0.1135,
+      "step": 2200
+    },
+    {
+      "epoch": 44.44,
+      "learning_rate": 0.00021449999999999998,
+      "loss": 0.0998,
+      "step": 2400
+    },
+    {
+      "epoch": 46.3,
+      "eval_loss": 1.0250211954116821,
+      "eval_runtime": 13.0242,
+      "eval_samples_per_second": 27.641,
+      "eval_steps_per_second": 3.455,
+      "eval_wer": 0.5233241865934928,
+      "step": 2500
+    },
+    {
+      "epoch": 48.15,
+      "learning_rate": 0.00020021428571428572,
+      "loss": 0.0893,
+      "step": 2600
+    },
+    {
+      "epoch": 51.85,
+      "learning_rate": 0.0001859285714285714,
+      "loss": 0.0759,
+      "step": 2800
+    },
+    {
+      "epoch": 55.56,
+      "learning_rate": 0.00017164285714285713,
+      "loss": 0.0727,
+      "step": 3000
+    },
+    {
+      "epoch": 55.56,
+      "eval_loss": 1.1072067022323608,
+      "eval_runtime": 13.8552,
+      "eval_samples_per_second": 25.983,
+      "eval_steps_per_second": 3.248,
+      "eval_wer": 0.5382203057624461,
+      "step": 3000
+    },
+    {
+      "epoch": 59.26,
+      "learning_rate": 0.00015735714285714284,
+      "loss": 0.0694,
+      "step": 3200
+    },
+    {
+      "epoch": 62.96,
+      "learning_rate": 0.00014307142857142855,
+      "loss": 0.0605,
+      "step": 3400
+    },
+    {
+      "epoch": 64.81,
+      "eval_loss": 1.0588016510009766,
+      "eval_runtime": 13.0366,
+      "eval_samples_per_second": 27.614,
+      "eval_steps_per_second": 3.452,
+      "eval_wer": 0.5072520580164641,
+      "step": 3500
+    },
+    {
+      "epoch": 66.67,
+      "learning_rate": 0.00012878571428571426,
+      "loss": 0.0558,
+      "step": 3600
+    },
+    {
+      "epoch": 70.37,
+      "learning_rate": 0.00011449999999999999,
+      "loss": 0.0481,
+      "step": 3800
+    },
+    {
+      "epoch": 74.07,
+      "learning_rate": 0.0001002142857142857,
+      "loss": 0.0458,
+      "step": 4000
+    },
+    {
+      "epoch": 74.07,
+      "eval_loss": 1.0818374156951904,
+      "eval_runtime": 12.9679,
+      "eval_samples_per_second": 27.761,
+      "eval_steps_per_second": 3.47,
+      "eval_wer": 0.5068600548804391,
+      "step": 4000
+    },
+    {
+      "epoch": 77.78,
+      "learning_rate": 8.592857142857142e-05,
+      "loss": 0.0382,
+      "step": 4200
+    },
+    {
+      "epoch": 81.48,
+      "learning_rate": 7.164285714285714e-05,
+      "loss": 0.0338,
+      "step": 4400
+    },
+    {
+      "epoch": 83.33,
+      "eval_loss": 1.094814419746399,
+      "eval_runtime": 13.1264,
+      "eval_samples_per_second": 27.426,
+      "eval_steps_per_second": 3.428,
+      "eval_wer": 0.5107800862406899,
+      "step": 4500
+    },
+    {
+      "epoch": 85.19,
+      "learning_rate": 5.735714285714285e-05,
+      "loss": 0.0308,
+      "step": 4600
+    },
+    {
+      "epoch": 88.89,
+      "learning_rate": 4.307142857142857e-05,
+      "loss": 0.0263,
+      "step": 4800
+    },
+    {
+      "epoch": 92.59,
+      "learning_rate": 2.8785714285714286e-05,
+      "loss": 0.0223,
+      "step": 5000
+    },
+    {
+      "epoch": 92.59,
+      "eval_loss": 1.098594307899475,
+      "eval_runtime": 13.0424,
+      "eval_samples_per_second": 27.602,
+      "eval_steps_per_second": 3.45,
+      "eval_wer": 0.47745981967855744,
+      "step": 5000
+    },
+    {
+      "epoch": 96.3,
+      "learning_rate": 1.4499999999999998e-05,
+      "loss": 0.0205,
+      "step": 5200
+    },
+    {
+      "epoch": 100.0,
+      "learning_rate": 2.1428571428571426e-07,
+      "loss": 0.0206,
+      "step": 5400
+    },
+    {
+      "epoch": 100.0,
+      "step": 5400,
+      "total_flos": 1.3559072525710848e+19,
+      "train_loss": 0.7658213693124276,
+      "train_runtime": 5843.2467,
+      "train_samples_per_second": 14.649,
+      "train_steps_per_second": 0.924
+    }
+  ],
+  "max_steps": 5400,
+  "num_train_epochs": 100,
+  "total_flos": 1.3559072525710848e+19,
+  "trial_name": null,
+  "trial_params": null
+}