End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 1.19,
-    "eval_loss": 0.38643914461135864,
-    "eval_runtime": 204.4048,
-    "eval_samples": 1000,
-    "eval_samples_per_second": 4.892,
-    "eval_steps_per_second": 0.078,
-    "eval_wer": 18.367063400175983,
-    "train_loss": 0.4157425765991211,
-    "train_runtime": 7330.7001,
-    "train_samples_per_second": 69.843,
-    "train_steps_per_second": 0.136
+    "epoch": 14.04,
+    "eval_loss": 0.30335742235183716,
+    "eval_runtime": 1122.4749,
+    "eval_samples": 5000,
+    "eval_samples_per_second": 4.454,
+    "eval_steps_per_second": 0.557,
+    "eval_wer": 14.535443909153475,
+    "train_loss": 0.0768378963470459,
+    "train_runtime": 74057.892,
+    "train_samples_per_second": 17.284,
+    "train_steps_per_second": 0.135
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 1.19,
-    "eval_loss": 0.38643914461135864,
-    "eval_runtime": 204.4048,
-    "eval_samples": 1000,
-    "eval_samples_per_second": 4.892,
-    "eval_steps_per_second": 0.078,
-    "eval_wer": 18.367063400175983
+    "epoch": 14.04,
+    "eval_loss": 0.30335742235183716,
+    "eval_runtime": 1122.4749,
+    "eval_samples": 5000,
+    "eval_samples_per_second": 4.454,
+    "eval_steps_per_second": 0.557,
+    "eval_wer": 14.535443909153475
 }

runs/Dec15_19-32-56_129-146-3-150/events.out.tfevents.1671208059.129-146-3-150.88241.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4348091858c4f36346386139f256e83dac46e73911a6c38b599d7cfdf598df5f
+size 358

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 1.19,
-    "train_loss": 0.4157425765991211,
-    "train_runtime": 7330.7001,
-    "train_samples_per_second": 69.843,
-    "train_steps_per_second": 0.136
+    "epoch": 14.04,
+    "train_loss": 0.0768378963470459,
+    "train_runtime": 74057.892,
+    "train_samples_per_second": 17.284,
+    "train_steps_per_second": 0.135
 }

trainer_state.json

@@ -1,175 +1,235 @@
 {
-  "best_metric": 18.367063400175983,
-  "best_model_checkpoint": "./whisper-small-nl/checkpoint-600",
-  "epoch": 1.194,
-  "global_step": 1000,
+  "best_metric": 14.535443909153475,
+  "best_model_checkpoint": "whisper-small-nl/checkpoint-10000",
+  "epoch": 14.0372,
+  "global_step": 10000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.1,
-      "learning_rate": 9.473684210526315e-06,
-      "loss": 0.8378,
-      "step": 100
+      "epoch": 1.24,
+      "learning_rate": 9.980000000000001e-06,
+      "loss": 0.7181,
+      "step": 500
     },
     {
-      "epoch": 0.1,
-      "eval_loss": 0.4932844042778015,
-      "eval_runtime": 153.0661,
-      "eval_samples_per_second": 6.533,
-      "eval_steps_per_second": 0.105,
-      "eval_wer": 23.882739776779513,
-      "step": 100
+      "epoch": 2.49,
+      "learning_rate": 8.891111111111111e-06,
+      "loss": 0.2045,
+      "step": 1000
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 8.421052631578948e-06,
-      "loss": 0.5547,
-      "step": 200
+      "epoch": 2.49,
+      "eval_loss": 0.31939101219177246,
+      "eval_runtime": 852.4844,
+      "eval_samples_per_second": 5.865,
+      "eval_steps_per_second": 0.733,
+      "eval_wer": 16.16284610090529,
+      "step": 1000
     },
     {
-      "epoch": 0.2,
-      "eval_loss": 0.44761189818382263,
-      "eval_runtime": 154.0782,
-      "eval_samples_per_second": 6.49,
-      "eval_steps_per_second": 0.104,
-      "eval_wer": 21.057750196823044,
-      "step": 200
+      "epoch": 3.73,
+      "learning_rate": 7.78e-06,
+      "loss": 0.1182,
+      "step": 1500
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 7.368421052631579e-06,
-      "loss": 0.3905,
-      "step": 300
+      "epoch": 4.97,
+      "learning_rate": 6.668888888888889e-06,
+      "loss": 0.0652,
+      "step": 2000
     },
     {
-      "epoch": 0.3,
-      "eval_loss": 0.4335392117500305,
-      "eval_runtime": 152.7461,
-      "eval_samples_per_second": 6.547,
-      "eval_steps_per_second": 0.105,
-      "eval_wer": 21.168897327837723,
-      "step": 300
+      "epoch": 4.97,
+      "eval_loss": 0.3425401449203491,
+      "eval_runtime": 853.9079,
+      "eval_samples_per_second": 5.855,
+      "eval_steps_per_second": 0.732,
+      "eval_wer": 16.367197840012707,
+      "step": 2000
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 6.31578947368421e-06,
-      "loss": 0.3766,
-      "step": 400
+      "epoch": 6.22,
+      "learning_rate": 5.557777777777778e-06,
+      "loss": 0.0318,
+      "step": 2500
     },
     {
-      "epoch": 0.4,
-      "eval_loss": 0.4266929030418396,
-      "eval_runtime": 150.336,
-      "eval_samples_per_second": 6.652,
-      "eval_steps_per_second": 0.106,
-      "eval_wer": 20.052794887231972,
-      "step": 400
+      "epoch": 7.46,
+      "learning_rate": 4.446666666666667e-06,
+      "loss": 0.0167,
+      "step": 3000
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 5.263157894736842e-06,
-      "loss": 0.4164,
-      "step": 500
+      "epoch": 7.46,
+      "eval_loss": 0.3914715647697449,
+      "eval_runtime": 851.1992,
+      "eval_samples_per_second": 5.874,
+      "eval_steps_per_second": 0.734,
+      "eval_wer": 15.818730478056011,
+      "step": 3000
     },
     {
-      "epoch": 0.5,
-      "eval_loss": 0.4138640761375427,
-      "eval_runtime": 157.7477,
-      "eval_samples_per_second": 6.339,
-      "eval_steps_per_second": 0.101,
-      "eval_wer": 21.43287176399759,
-      "step": 500
+      "epoch": 8.71,
+      "learning_rate": 3.335555555555556e-06,
+      "loss": 0.0096,
+      "step": 3500
     },
     {
-      "epoch": 0.6,
-      "learning_rate": 4.210526315789474e-06,
-      "loss": 0.2939,
-      "step": 600
+      "epoch": 9.95,
+      "learning_rate": 2.2244444444444447e-06,
+      "loss": 0.0064,
+      "step": 4000
     },
     {
-      "epoch": 0.6,
-      "eval_loss": 0.38643914461135864,
-      "eval_runtime": 151.6414,
-      "eval_samples_per_second": 6.595,
-      "eval_steps_per_second": 0.106,
-      "eval_wer": 18.367063400175983,
-      "step": 600
+      "epoch": 9.95,
+      "eval_loss": 0.41895967721939087,
+      "eval_runtime": 850.8744,
+      "eval_samples_per_second": 5.876,
+      "eval_steps_per_second": 0.735,
+      "eval_wer": 15.729789824765735,
+      "step": 4000
     },
     {
-      "epoch": 0.7,
-      "learning_rate": 3.157894736842105e-06,
-      "loss": 0.2632,
-      "step": 700
+      "epoch": 1.01,
+      "learning_rate": 5.790526315789474e-06,
+      "loss": 0.1798,
+      "step": 4500
     },
     {
-      "epoch": 0.7,
-      "eval_loss": 0.3864086866378784,
-      "eval_runtime": 152.5897,
-      "eval_samples_per_second": 6.554,
-      "eval_steps_per_second": 0.105,
-      "eval_wer": 18.431899226601214,
-      "step": 700
+      "epoch": 2.02,
+      "learning_rate": 5.26421052631579e-06,
+      "loss": 0.1966,
+      "step": 5000
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 2.105263157894737e-06,
-      "loss": 0.6066,
-      "step": 800
+      "epoch": 2.02,
+      "eval_loss": 0.32975926995277405,
+      "eval_runtime": 1015.8008,
+      "eval_samples_per_second": 4.922,
+      "eval_steps_per_second": 0.615,
+      "eval_wer": 15.088146540314467,
+      "step": 5000
     },
     {
-      "epoch": 0.8,
-      "eval_loss": 0.3804214298725128,
-      "eval_runtime": 156.1258,
-      "eval_samples_per_second": 6.405,
-      "eval_steps_per_second": 0.102,
-      "eval_wer": 19.27476497012921,
-      "step": 800
+      "epoch": 3.03,
+      "learning_rate": 4.737894736842106e-06,
+      "loss": 0.1605,
+      "step": 5500
     },
     {
-      "epoch": 1.09,
-      "learning_rate": 1.0526315789473685e-06,
-      "loss": 0.2075,
-      "step": 900
+      "epoch": 4.04,
+      "learning_rate": 4.211578947368422e-06,
+      "loss": 0.1912,
+      "step": 6000
     },
     {
-      "epoch": 1.09,
-      "eval_loss": 0.3793680667877197,
-      "eval_runtime": 153.2522,
-      "eval_samples_per_second": 6.525,
-      "eval_steps_per_second": 0.104,
-      "eval_wer": 18.890381142036773,
-      "step": 900
+      "epoch": 4.04,
+      "eval_loss": 0.3265514671802521,
+      "eval_runtime": 1016.6509,
+      "eval_samples_per_second": 4.918,
+      "eval_steps_per_second": 0.615,
+      "eval_wer": 14.876383080099528,
+      "step": 6000
     },
     {
-      "epoch": 1.19,
-      "learning_rate": 0.0,
-      "loss": 0.2102,
-      "step": 1000
+      "epoch": 6.01,
+      "learning_rate": 3.685263157894737e-06,
+      "loss": 0.1355,
+      "step": 6500
     },
     {
-      "epoch": 1.19,
-      "eval_loss": 0.37768101692199707,
-      "eval_runtime": 153.5387,
-      "eval_samples_per_second": 6.513,
-      "eval_steps_per_second": 0.104,
-      "eval_wer": 19.88144306025101,
-      "step": 1000
+      "epoch": 7.02,
+      "learning_rate": 3.1600000000000002e-06,
+      "loss": 0.1008,
+      "step": 7000
+    },
+    {
+      "epoch": 7.02,
+      "eval_loss": 0.3261025547981262,
+      "eval_runtime": 1005.7569,
+      "eval_samples_per_second": 4.971,
+      "eval_steps_per_second": 0.621,
+      "eval_wer": 14.808618772830748,
+      "step": 7000
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 2.633684210526316e-06,
+      "loss": 0.1092,
+      "step": 7500
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 2.1073684210526317e-06,
+      "loss": 0.0899,
+      "step": 8000
+    },
+    {
+      "epoch": 9.04,
+      "eval_loss": 0.31958696246147156,
+      "eval_runtime": 1016.386,
+      "eval_samples_per_second": 4.919,
+      "eval_steps_per_second": 0.615,
+      "eval_wer": 14.648737360368468,
+      "step": 8000
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 1.5810526315789477e-06,
+      "loss": 0.0683,
+      "step": 8500
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 1.0547368421052632e-06,
+      "loss": 0.1126,
+      "step": 9000
+    },
+    {
+      "epoch": 12.02,
+      "eval_loss": 0.3282873332500458,
+      "eval_runtime": 1020.4433,
+      "eval_samples_per_second": 4.9,
+      "eval_steps_per_second": 0.612,
+      "eval_wer": 14.589443591508285,
+      "step": 9000
+    },
+    {
+      "epoch": 13.03,
+      "learning_rate": 5.284210526315789e-07,
+      "loss": 0.0853,
+      "step": 9500
+    },
+    {
+      "epoch": 14.04,
+      "learning_rate": 2.105263157894737e-09,
+      "loss": 0.1071,
+      "step": 10000
+    },
+    {
+      "epoch": 14.04,
+      "eval_loss": 0.30335742235183716,
+      "eval_runtime": 1026.6502,
+      "eval_samples_per_second": 4.87,
+      "eval_steps_per_second": 0.609,
+      "eval_wer": 14.535443909153475,
+      "step": 10000
     },
     {
-      "epoch": 1.19,
-      "step": 1000,
-      "total_flos": 1.84579222044672e+19,
-      "train_loss": 0.4157425765991211,
-      "train_runtime": 7330.7001,
-      "train_samples_per_second": 69.843,
-      "train_steps_per_second": 0.136
+      "epoch": 14.04,
+      "step": 10000,
+      "total_flos": 3.699734092598477e+20,
+      "train_loss": 0.0768378963470459,
+      "train_runtime": 74057.892,
+      "train_samples_per_second": 17.284,
+      "train_steps_per_second": 0.135
     }
   ],
-  "max_steps": 1000,
+  "max_steps": 10000,
   "num_train_epochs": 9223372036854775807,
-  "total_flos": 1.84579222044672e+19,
+  "total_flos": 3.699734092598477e+20,
   "trial_name": null,
   "trial_params": null
 }