End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 1.0,
-    "eval_loss": 0.6197173595428467,
-    "eval_runtime": 18.6322,
+    "epoch": 0.33,
+    "eval_loss": 0.5382302403450012,
+    "eval_runtime": 19.0163,
     "eval_samples": 64,
-    "eval_samples_per_second": 3.435,
-    "eval_steps_per_second": 0.107,
-    "eval_wer": 60.43956043956044,
-    "train_loss": 0.6153274965286255,
-    "train_runtime": 782.3972,
-    "train_samples_per_second": 4.09,
-    "train_steps_per_second": 0.128
+    "eval_samples_per_second": 3.366,
+    "eval_steps_per_second": 0.105,
+    "eval_wer": 55.12820512820513,
+    "train_loss": 0.13119232177734375,
+    "train_runtime": 451.4438,
+    "train_samples_per_second": 10.633,
+    "train_steps_per_second": 0.332
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 1.0,
-    "eval_loss": 0.6197173595428467,
-    "eval_runtime": 18.6322,
+    "epoch": 0.33,
+    "eval_loss": 0.5382302403450012,
+    "eval_runtime": 19.0163,
     "eval_samples": 64,
-    "eval_samples_per_second": 3.435,
-    "eval_steps_per_second": 0.107,
-    "eval_wer": 60.43956043956044
+    "eval_samples_per_second": 3.366,
+    "eval_steps_per_second": 0.105,
+    "eval_wer": 55.12820512820513
 }

runs/Dec12_20-18-09_13ea4557542f/events.out.tfevents.1670877055.13ea4557542f.19541.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c9f608849db55cf81235690c55857e2f3e65e9efeb53941d7cd4efce64bc23b
+size 358

train.log

@@ -58,3 +58,9 @@
 {'loss': 0.2946, 'learning_rate': 2.0689655172413796e-06, 'epoch': 0.33}
 {'eval_loss': 0.53872150182724, 'eval_wer': 55.67765567765568, 'eval_runtime': 20.4177, 'eval_samples_per_second': 3.135, 'eval_steps_per_second': 0.098, 'epoch': 0.33}
 {'train_runtime': 451.4438, 'train_samples_per_second': 10.633, 'train_steps_per_second': 0.332, 'train_loss': 0.13119232177734375, 'epoch': 0.33}
+***** train metrics *****
+  epoch                    =       0.33
+  train_loss               =     0.1312
+  train_runtime            = 0:07:31.44
+  train_samples_per_second =     10.633
+  train_steps_per_second   =      0.332

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 1.0,
-    "train_loss": 0.6153274965286255,
-    "train_runtime": 782.3972,
-    "train_samples_per_second": 4.09,
-    "train_steps_per_second": 0.128
+    "epoch": 0.33,
+    "train_loss": 0.13119232177734375,
+    "train_runtime": 451.4438,
+    "train_samples_per_second": 10.633,
+    "train_steps_per_second": 0.332
 }

trainer_state.json

@@ -1,8 +1,8 @@
 {
-  "best_metric": 60.43956043956044,
-  "best_model_checkpoint": "./checkpoint-70",
-  "epoch": 1.0,
-  "global_step": 100,
+  "best_metric": 55.12820512820513,
+  "best_model_checkpoint": "./checkpoint-140",
+  "epoch": 0.3333333333333333,
+  "global_step": 150,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -158,18 +158,93 @@
       "step": 100
     },
     {
-      "epoch": 1.0,
-      "step": 100,
-      "total_flos": 7.8780432384e+16,
-      "train_loss": 0.6153274965286255,
-      "train_runtime": 782.3972,
-      "train_samples_per_second": 4.09,
-      "train_steps_per_second": 0.128
+      "epoch": 0.07,
+      "learning_rate": 2.96551724137931e-05,
+      "loss": 0.4611,
+      "step": 110
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.5625213384628296,
+      "eval_runtime": 19.7626,
+      "eval_samples_per_second": 3.238,
+      "eval_steps_per_second": 0.101,
+      "eval_wer": 57.692307692307686,
+      "step": 110
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.2758620689655175e-05,
+      "loss": 0.4451,
+      "step": 120
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.563637375831604,
+      "eval_runtime": 19.0995,
+      "eval_samples_per_second": 3.351,
+      "eval_steps_per_second": 0.105,
+      "eval_wer": 56.59340659340659,
+      "step": 120
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.586206896551724e-05,
+      "loss": 0.3615,
+      "step": 130
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.5489554405212402,
+      "eval_runtime": 18.5694,
+      "eval_samples_per_second": 3.447,
+      "eval_steps_per_second": 0.108,
+      "eval_wer": 61.172161172161175,
+      "step": 130
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.96551724137931e-06,
+      "loss": 0.4055,
+      "step": 140
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.5382302403450012,
+      "eval_runtime": 22.4274,
+      "eval_samples_per_second": 2.854,
+      "eval_steps_per_second": 0.089,
+      "eval_wer": 55.12820512820513,
+      "step": 140
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 0.2946,
+      "step": 150
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.53872150182724,
+      "eval_runtime": 20.4177,
+      "eval_samples_per_second": 3.135,
+      "eval_steps_per_second": 0.098,
+      "eval_wer": 55.67765567765568,
+      "step": 150
+    },
+    {
+      "epoch": 0.33,
+      "step": 150,
+      "total_flos": 1.18170648576e+17,
+      "train_loss": 0.13119232177734375,
+      "train_runtime": 451.4438,
+      "train_samples_per_second": 10.633,
+      "train_steps_per_second": 0.332
     }
   ],
-  "max_steps": 100,
+  "max_steps": 150,
   "num_train_epochs": 9223372036854775807,
-  "total_flos": 7.8780432384e+16,
+  "total_flos": 1.18170648576e+17,
   "trial_name": null,
   "trial_params": null
 }