End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 0.25,
-    "eval_loss": 0.5364237427711487,
-    "eval_runtime": 16.6084,
+    "epoch": 0.33,
+    "eval_loss": 0.5074095726013184,
+    "eval_runtime": 16.8249,
     "eval_samples": 64,
-    "eval_samples_per_second": 3.853,
-    "eval_steps_per_second": 0.12,
-    "eval_wer": 54.57875457875458,
-    "train_loss": 0.0719480574131012,
-    "train_runtime": 406.2172,
-    "train_samples_per_second": 15.755,
-    "train_steps_per_second": 0.492
+    "eval_samples_per_second": 3.804,
+    "eval_steps_per_second": 0.119,
+    "eval_wer": 52.197802197802204,
+    "train_loss": 0.10702953418095906,
+    "train_runtime": 833.6007,
+    "train_samples_per_second": 11.516,
+    "train_steps_per_second": 0.36
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 0.25,
-    "eval_loss": 0.5364237427711487,
-    "eval_runtime": 16.6084,
+    "epoch": 0.33,
+    "eval_loss": 0.5074095726013184,
+    "eval_runtime": 16.8249,
     "eval_samples": 64,
-    "eval_samples_per_second": 3.853,
-    "eval_steps_per_second": 0.12,
-    "eval_wer": 54.57875457875458
+    "eval_samples_per_second": 3.804,
+    "eval_steps_per_second": 0.119,
+    "eval_wer": 52.197802197802204
 }

runs/Dec13_12-14-07_d7f040c448a8/events.out.tfevents.1670934730.d7f040c448a8.15037.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:621b7a2306709878e0351fd168351ff2ec7c15bf93824a12f4627acf91ccc932
+size 358

train.log

@@ -118,3 +118,9 @@
 {'loss': 0.4407, 'learning_rate': 5.033333333333333e-06, 'epoch': 0.33}
 {'eval_loss': 0.5046072602272034, 'eval_wer': 53.11355311355312, 'eval_runtime': 17.9261, 'eval_samples_per_second': 3.57, 'eval_steps_per_second': 0.112, 'epoch': 0.33}
 {'train_runtime': 833.6007, 'train_samples_per_second': 11.516, 'train_steps_per_second': 0.36, 'train_loss': 0.10702953418095906, 'epoch': 0.33}
+***** train metrics *****
+  epoch                    =       0.33
+  train_loss               =      0.107
+  train_runtime            = 0:13:53.60
+  train_samples_per_second =     11.516
+  train_steps_per_second   =       0.36

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 0.25,
-    "train_loss": 0.0719480574131012,
-    "train_runtime": 406.2172,
-    "train_samples_per_second": 15.755,
-    "train_steps_per_second": 0.492
+    "epoch": 0.33,
+    "train_loss": 0.10702953418095906,
+    "train_runtime": 833.6007,
+    "train_samples_per_second": 11.516,
+    "train_steps_per_second": 0.36
 }

trainer_state.json

@@ -1,8 +1,8 @@
 {
-  "best_metric": 54.57875457875458,
-  "best_model_checkpoint": "./checkpoint-160",
-  "epoch": 0.25,
-  "global_step": 200,
+  "best_metric": 52.197802197802204,
+  "best_model_checkpoint": "./checkpoint-280",
+  "epoch": 0.3333333333333333,
+  "global_step": 300,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -317,18 +317,168 @@
       "step": 200
     },
     {
-      "epoch": 0.25,
-      "step": 200,
-      "total_flos": 1.57560864768e+17,
-      "train_loss": 0.0719480574131012,
-      "train_runtime": 406.2172,
-      "train_samples_per_second": 15.755,
-      "train_steps_per_second": 0.492
+      "epoch": 0.03,
+      "learning_rate": 8.033333333333335e-06,
+      "loss": 0.1922,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 0.5239479541778564,
+      "eval_runtime": 16.5572,
+      "eval_samples_per_second": 3.865,
+      "eval_steps_per_second": 0.121,
+      "eval_wer": 55.494505494505496,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 7.7e-06,
+      "loss": 0.2229,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.5312361717224121,
+      "eval_runtime": 17.8694,
+      "eval_samples_per_second": 3.582,
+      "eval_steps_per_second": 0.112,
+      "eval_wer": 54.395604395604394,
+      "step": 220
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 7.3666666666666676e-06,
+      "loss": 0.1976,
+      "step": 230
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.5302589535713196,
+      "eval_runtime": 17.0912,
+      "eval_samples_per_second": 3.745,
+      "eval_steps_per_second": 0.117,
+      "eval_wer": 54.02930402930403,
+      "step": 230
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 7.033333333333334e-06,
+      "loss": 0.2823,
+      "step": 240
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.5269189476966858,
+      "eval_runtime": 17.9989,
+      "eval_samples_per_second": 3.556,
+      "eval_steps_per_second": 0.111,
+      "eval_wer": 54.02930402930403,
+      "step": 240
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 0.2265,
+      "step": 250
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 0.5312862992286682,
+      "eval_runtime": 18.4593,
+      "eval_samples_per_second": 3.467,
+      "eval_steps_per_second": 0.108,
+      "eval_wer": 55.67765567765568,
+      "step": 250
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.366666666666668e-06,
+      "loss": 0.3728,
+      "step": 260
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.5128015279769897,
+      "eval_runtime": 18.2249,
+      "eval_samples_per_second": 3.512,
+      "eval_steps_per_second": 0.11,
+      "eval_wer": 53.47985347985348,
+      "step": 260
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 6.033333333333335e-06,
+      "loss": 0.3738,
+      "step": 270
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.5025143623352051,
+      "eval_runtime": 17.2543,
+      "eval_samples_per_second": 3.709,
+      "eval_steps_per_second": 0.116,
+      "eval_wer": 52.74725274725275,
+      "step": 270
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.7e-06,
+      "loss": 0.488,
+      "step": 280
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.5074095726013184,
+      "eval_runtime": 18.4062,
+      "eval_samples_per_second": 3.477,
+      "eval_steps_per_second": 0.109,
+      "eval_wer": 52.197802197802204,
+      "step": 280
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.366666666666666e-06,
+      "loss": 0.4142,
+      "step": 290
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.5057792663574219,
+      "eval_runtime": 16.5894,
+      "eval_samples_per_second": 3.858,
+      "eval_steps_per_second": 0.121,
+      "eval_wer": 52.56410256410257,
+      "step": 290
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.033333333333333e-06,
+      "loss": 0.4407,
+      "step": 300
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.5046072602272034,
+      "eval_runtime": 17.9261,
+      "eval_samples_per_second": 3.57,
+      "eval_steps_per_second": 0.112,
+      "eval_wer": 53.11355311355312,
+      "step": 300
+    },
+    {
+      "epoch": 0.33,
+      "step": 300,
+      "total_flos": 2.36341297152e+17,
+      "train_loss": 0.10702953418095906,
+      "train_runtime": 833.6007,
+      "train_samples_per_second": 11.516,
+      "train_steps_per_second": 0.36
     }
   ],
-  "max_steps": 200,
+  "max_steps": 300,
   "num_train_epochs": 9223372036854775807,
-  "total_flos": 1.57560864768e+17,
+  "total_flos": 2.36341297152e+17,
   "trial_name": null,
   "trial_params": null
 }