End of training

all_results.json

@@ -1,14 +1,14 @@
 {
-    "epoch": 1.0,
-    "eval_loss": 28.728288650512695,
-    "eval_runtime": 5.2804,
+    "epoch": 120.0,
+    "eval_loss": 0.657963216304779,
+    "eval_runtime": 5.0644,
     "eval_samples": 112,
-    "eval_samples_per_second": 21.21,
-    "eval_steps_per_second": 2.651,
-    "eval_wer": 1.0,
-    "train_loss": 20.803141874425553,
-    "train_runtime": 35.5064,
+    "eval_samples_per_second": 22.115,
+    "eval_steps_per_second": 1.382,
+    "eval_wer": 0.9712230215827338,
+    "train_loss": 3.015083034365785,
+    "train_runtime": 4204.6915,
     "train_samples": 537,
-    "train_samples_per_second": 15.124,
-    "train_steps_per_second": 0.479
+    "train_samples_per_second": 15.326,
+    "train_steps_per_second": 0.485
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 1.0,
-    "eval_loss": 28.728288650512695,
-    "eval_runtime": 5.2804,
+    "epoch": 120.0,
+    "eval_loss": 0.657963216304779,
+    "eval_runtime": 5.0644,
     "eval_samples": 112,
-    "eval_samples_per_second": 21.21,
-    "eval_steps_per_second": 2.651,
-    "eval_wer": 1.0
+    "eval_samples_per_second": 22.115,
+    "eval_steps_per_second": 1.382,
+    "eval_wer": 0.9712230215827338
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7902a22a0853f221655d97939bcbefcd5b7495587d90f5969510eff1344ed4ea
+oid sha256:d1b9cf50527d581cdd062cab8c516d9699ec415eb041a4a557e6b2dfd2350fb5
 size 1262186097

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.0,
-    "train_loss": 20.803141874425553,
-    "train_runtime": 35.5064,
+    "epoch": 120.0,
+    "train_loss": 3.015083034365785,
+    "train_runtime": 4204.6915,
     "train_samples": 537,
-    "train_samples_per_second": 15.124,
-    "train_steps_per_second": 0.479
+    "train_samples_per_second": 15.326,
+    "train_steps_per_second": 0.485
 }

trainer_state.json

@@ -1,25 +1,181 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.0,
-  "global_step": 17,
+  "epoch": 120.0,
+  "global_step": 2040,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.0,
-      "step": 17,
-      "total_flos": 8.750541089193984e+16,
-      "train_loss": 20.803141874425553,
-      "train_runtime": 35.5064,
-      "train_samples_per_second": 15.124,
-      "train_steps_per_second": 0.479
+      "epoch": 5.88,
+      "learning_rate": 1.455e-05,
+      "loss": 16.0345,
+      "step": 100
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 2.955e-05,
+      "loss": 6.2463,
+      "step": 200
+    },
+    {
+      "epoch": 17.65,
+      "learning_rate": 4.454999999999999e-05,
+      "loss": 4.2277,
+      "step": 300
+    },
+    {
+      "epoch": 23.53,
+      "learning_rate": 5.955e-05,
+      "loss": 3.4915,
+      "step": 400
+    },
+    {
+      "epoch": 29.41,
+      "learning_rate": 7.455e-05,
+      "loss": 3.3231,
+      "step": 500
+    },
+    {
+      "epoch": 29.41,
+      "eval_loss": 3.352241277694702,
+      "eval_runtime": 5.0475,
+      "eval_samples_per_second": 22.189,
+      "eval_steps_per_second": 1.387,
+      "eval_wer": 0.998972250770812,
+      "step": 500
+    },
+    {
+      "epoch": 35.29,
+      "learning_rate": 7.027597402597401e-05,
+      "loss": 3.2264,
+      "step": 600
+    },
+    {
+      "epoch": 41.18,
+      "learning_rate": 6.540584415584416e-05,
+      "loss": 3.1652,
+      "step": 700
+    },
+    {
+      "epoch": 47.06,
+      "learning_rate": 6.0535714285714285e-05,
+      "loss": 3.019,
+      "step": 800
+    },
+    {
+      "epoch": 52.94,
+      "learning_rate": 5.566558441558441e-05,
+      "loss": 2.6429,
+      "step": 900
+    },
+    {
+      "epoch": 58.82,
+      "learning_rate": 5.0795454545454536e-05,
+      "loss": 2.1146,
+      "step": 1000
+    },
+    {
+      "epoch": 58.82,
+      "eval_loss": 1.3367875814437866,
+      "eval_runtime": 5.0633,
+      "eval_samples_per_second": 22.12,
+      "eval_steps_per_second": 1.382,
+      "eval_wer": 0.9383350462487153,
+      "step": 1000
+    },
+    {
+      "epoch": 64.71,
+      "learning_rate": 4.592532467532467e-05,
+      "loss": 1.7769,
+      "step": 1100
+    },
+    {
+      "epoch": 70.59,
+      "learning_rate": 4.10551948051948e-05,
+      "loss": 1.5937,
+      "step": 1200
+    },
+    {
+      "epoch": 76.47,
+      "learning_rate": 3.6185064935064934e-05,
+      "loss": 1.4694,
+      "step": 1300
+    },
+    {
+      "epoch": 82.35,
+      "learning_rate": 3.131493506493506e-05,
+      "loss": 1.3781,
+      "step": 1400
+    },
+    {
+      "epoch": 88.24,
+      "learning_rate": 2.6444805194805193e-05,
+      "loss": 1.3134,
+      "step": 1500
+    },
+    {
+      "epoch": 88.24,
+      "eval_loss": 0.7372016310691833,
+      "eval_runtime": 5.0913,
+      "eval_samples_per_second": 21.998,
+      "eval_steps_per_second": 1.375,
+      "eval_wer": 0.9578622816032888,
+      "step": 1500
+    },
+    {
+      "epoch": 94.12,
+      "learning_rate": 2.162337662337662e-05,
+      "loss": 1.271,
+      "step": 1600
+    },
+    {
+      "epoch": 100.0,
+      "learning_rate": 1.6753246753246752e-05,
+      "loss": 1.2273,
+      "step": 1700
+    },
+    {
+      "epoch": 105.88,
+      "learning_rate": 1.1883116883116881e-05,
+      "loss": 1.2121,
+      "step": 1800
+    },
+    {
+      "epoch": 111.76,
+      "learning_rate": 7.012987012987012e-06,
+      "loss": 1.169,
+      "step": 1900
+    },
+    {
+      "epoch": 117.65,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 1.1506,
+      "step": 2000
+    },
+    {
+      "epoch": 117.65,
+      "eval_loss": 0.6582115292549133,
+      "eval_runtime": 5.0953,
+      "eval_samples_per_second": 21.981,
+      "eval_steps_per_second": 1.374,
+      "eval_wer": 0.9681397738951696,
+      "step": 2000
+    },
+    {
+      "epoch": 120.0,
+      "step": 2040,
+      "total_flos": 1.105779936504619e+19,
+      "train_loss": 3.015083034365785,
+      "train_runtime": 4204.6915,
+      "train_samples_per_second": 15.326,
+      "train_steps_per_second": 0.485
     }
   ],
-  "max_steps": 17,
-  "num_train_epochs": 1,
-  "total_flos": 8.750541089193984e+16,
+  "max_steps": 2040,
+  "num_train_epochs": 120,
+  "total_flos": 1.105779936504619e+19,
   "trial_name": null,
   "trial_params": null
 }