End of training

README.md

@@ -2,38 +2,39 @@
 license: apache-2.0
 base_model: openai/whisper-small
 tags:
+- whisper-event
 - generated_from_trainer
 datasets:
-- fleurs
+- google/fleurs
 metrics:
 - wer
 model-index:
-- name: whisper-small-shona
+- name: Whisper small shona
   results:
   - task:
       name: Automatic Speech Recognition
       type: automatic-speech-recognition
     dataset:
-      name: fleurs
-      type: fleurs
+      name: google/fleurs sn_zw
+      type: google/fleurs
       config: sn_zw
       split: test
       args: sn_zw
     metrics:
     - name: Wer
       type: wer
-      value: 50.04219409282701
+      value: 49.90958408679928
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# whisper-small-shona
+# Whisper small shona
 
-This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the fleurs dataset.
+This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the google/fleurs sn_zw dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.1298
-- Wer: 50.0422
+- Loss: 1.1220
+- Wer: 49.9096
 
 ## Model description
 

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 40.82,
-    "eval_loss": 0.965289831161499,
-    "eval_runtime": 148.1672,
-    "eval_samples_per_second": 6.243,
-    "eval_steps_per_second": 0.391,
-    "eval_wer": 45.9,
-    "train_loss": 0.18361923832818866,
-    "train_runtime": 4653.6924,
-    "train_samples_per_second": 6.876,
-    "train_steps_per_second": 0.43
+    "epoch": 121.21,
+    "eval_loss": 1.1220260858535767,
+    "eval_runtime": 56.2101,
+    "eval_samples_per_second": 16.456,
+    "eval_steps_per_second": 0.356,
+    "eval_wer": 49.90958408679928,
+    "train_loss": 0.143078886593692,
+    "train_runtime": 4289.2081,
+    "train_samples_per_second": 22.382,
+    "train_steps_per_second": 0.466
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 40.82,
-    "eval_loss": 0.965289831161499,
-    "eval_runtime": 148.1672,
-    "eval_samples_per_second": 6.243,
-    "eval_steps_per_second": 0.391,
-    "eval_wer": 45.9
+    "epoch": 121.21,
+    "eval_loss": 1.1220260858535767,
+    "eval_runtime": 56.2101,
+    "eval_samples_per_second": 16.456,
+    "eval_steps_per_second": 0.356,
+    "eval_wer": 49.90958408679928
 }

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 40.82,
-    "train_loss": 0.18361923832818866,
-    "train_runtime": 4653.6924,
-    "train_samples_per_second": 6.876,
-    "train_steps_per_second": 0.43
+    "epoch": 121.21,
+    "train_loss": 0.143078886593692,
+    "train_runtime": 4289.2081,
+    "train_samples_per_second": 22.382,
+    "train_steps_per_second": 0.466
 }

trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 45.9,
+  "best_metric": 49.90958408679928,
   "best_model_checkpoint": "./whisper-small-shona/checkpoint-1600",
-  "epoch": 40.816326530612244,
+  "epoch": 121.21212121212122,
   "eval_steps": 400,
   "global_step": 2000,
   "is_hyper_param_search": false,
@@ -9,546 +9,546 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.51,
+      "epoch": 1.52,
       "learning_rate": 4.800000000000001e-07,
-      "loss": 2.8886,
+      "loss": 2.8865,
       "step": 25
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 9.800000000000001e-07,
-      "loss": 2.4034,
+      "epoch": 3.03,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 2.3463,
       "step": 50
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 1.48e-06,
-      "loss": 1.8889,
+      "epoch": 4.55,
+      "learning_rate": 1.46e-06,
+      "loss": 1.7622,
       "step": 75
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 1.98e-06,
-      "loss": 1.4574,
+      "epoch": 6.06,
+      "learning_rate": 1.9600000000000003e-06,
+      "loss": 1.2611,
       "step": 100
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 2.4800000000000004e-06,
-      "loss": 1.1526,
+      "epoch": 7.58,
+      "learning_rate": 2.46e-06,
+      "loss": 0.9353,
       "step": 125
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 2.9800000000000003e-06,
-      "loss": 0.9975,
+      "epoch": 9.09,
+      "learning_rate": 2.96e-06,
+      "loss": 0.7091,
       "step": 150
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 3.48e-06,
-      "loss": 0.797,
+      "epoch": 10.61,
+      "learning_rate": 3.46e-06,
+      "loss": 0.5145,
       "step": 175
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 3.980000000000001e-06,
-      "loss": 0.6751,
+      "epoch": 12.12,
+      "learning_rate": 3.96e-06,
+      "loss": 0.3821,
       "step": 200
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 4.48e-06,
-      "loss": 0.5246,
+      "epoch": 13.64,
+      "learning_rate": 4.4600000000000005e-06,
+      "loss": 0.2477,
       "step": 225
     },
     {
-      "epoch": 5.1,
-      "learning_rate": 4.980000000000001e-06,
-      "loss": 0.4666,
+      "epoch": 15.15,
+      "learning_rate": 4.960000000000001e-06,
+      "loss": 0.1587,
       "step": 250
     },
     {
-      "epoch": 5.61,
-      "learning_rate": 5.480000000000001e-06,
-      "loss": 0.3392,
+      "epoch": 16.67,
+      "learning_rate": 5.460000000000001e-06,
+      "loss": 0.0896,
       "step": 275
     },
     {
-      "epoch": 6.12,
-      "learning_rate": 5.98e-06,
-      "loss": 0.2899,
+      "epoch": 18.18,
+      "learning_rate": 5.9600000000000005e-06,
+      "loss": 0.0466,
       "step": 300
     },
     {
-      "epoch": 6.63,
-      "learning_rate": 6.480000000000001e-06,
-      "loss": 0.1993,
+      "epoch": 19.7,
+      "learning_rate": 6.460000000000001e-06,
+      "loss": 0.0224,
       "step": 325
     },
     {
-      "epoch": 7.14,
-      "learning_rate": 6.98e-06,
-      "loss": 0.1636,
+      "epoch": 21.21,
+      "learning_rate": 6.96e-06,
+      "loss": 0.0132,
       "step": 350
     },
     {
-      "epoch": 7.65,
-      "learning_rate": 7.48e-06,
-      "loss": 0.1039,
+      "epoch": 22.73,
+      "learning_rate": 7.4600000000000006e-06,
+      "loss": 0.0084,
       "step": 375
     },
     {
-      "epoch": 8.16,
-      "learning_rate": 7.980000000000002e-06,
-      "loss": 0.09,
+      "epoch": 24.24,
+      "learning_rate": 7.960000000000002e-06,
+      "loss": 0.0064,
       "step": 400
     },
     {
-      "epoch": 8.16,
-      "eval_loss": 0.857236921787262,
-      "eval_runtime": 149.0293,
-      "eval_samples_per_second": 6.207,
-      "eval_steps_per_second": 0.389,
-      "eval_wer": 50.6875,
+      "epoch": 24.24,
+      "eval_loss": 0.9630287289619446,
+      "eval_runtime": 56.9771,
+      "eval_samples_per_second": 16.235,
+      "eval_steps_per_second": 0.351,
+      "eval_wer": 50.72332730560579,
       "step": 400
     },
     {
-      "epoch": 8.67,
-      "learning_rate": 8.48e-06,
-      "loss": 0.0536,
+      "epoch": 25.76,
+      "learning_rate": 8.46e-06,
+      "loss": 0.0049,
       "step": 425
     },
     {
-      "epoch": 9.18,
-      "learning_rate": 8.98e-06,
-      "loss": 0.0403,
+      "epoch": 27.27,
+      "learning_rate": 8.96e-06,
+      "loss": 0.0041,
       "step": 450
     },
     {
-      "epoch": 9.69,
-      "learning_rate": 9.48e-06,
-      "loss": 0.0278,
+      "epoch": 28.79,
+      "learning_rate": 9.460000000000001e-06,
+      "loss": 0.0034,
       "step": 475
     },
     {
-      "epoch": 10.2,
-      "learning_rate": 9.980000000000001e-06,
-      "loss": 0.0204,
+      "epoch": 30.3,
+      "learning_rate": 9.960000000000001e-06,
+      "loss": 0.0029,
       "step": 500
     },
     {
-      "epoch": 10.71,
-      "learning_rate": 9.84e-06,
-      "loss": 0.0131,
+      "epoch": 31.82,
+      "learning_rate": 9.846666666666668e-06,
+      "loss": 0.0025,
       "step": 525
     },
     {
-      "epoch": 11.22,
-      "learning_rate": 9.673333333333334e-06,
-      "loss": 0.0117,
+      "epoch": 33.33,
+      "learning_rate": 9.68e-06,
+      "loss": 0.0022,
       "step": 550
     },
     {
-      "epoch": 11.73,
-      "learning_rate": 9.506666666666667e-06,
-      "loss": 0.0083,
+      "epoch": 34.85,
+      "learning_rate": 9.513333333333334e-06,
+      "loss": 0.0019,
       "step": 575
     },
     {
-      "epoch": 12.24,
-      "learning_rate": 9.340000000000002e-06,
-      "loss": 0.0069,
+      "epoch": 36.36,
+      "learning_rate": 9.346666666666666e-06,
+      "loss": 0.0017,
       "step": 600
     },
     {
-      "epoch": 12.76,
-      "learning_rate": 9.173333333333334e-06,
-      "loss": 0.0072,
+      "epoch": 37.88,
+      "learning_rate": 9.180000000000002e-06,
+      "loss": 0.0016,
       "step": 625
     },
     {
-      "epoch": 13.27,
-      "learning_rate": 9.006666666666666e-06,
-      "loss": 0.006,
+      "epoch": 39.39,
+      "learning_rate": 9.013333333333334e-06,
+      "loss": 0.0014,
       "step": 650
     },
     {
-      "epoch": 13.78,
-      "learning_rate": 8.84e-06,
-      "loss": 0.0047,
+      "epoch": 40.91,
+      "learning_rate": 8.846666666666668e-06,
+      "loss": 0.0013,
       "step": 675
     },
     {
-      "epoch": 14.29,
-      "learning_rate": 8.673333333333334e-06,
-      "loss": 0.0034,
+      "epoch": 42.42,
+      "learning_rate": 8.68e-06,
+      "loss": 0.0012,
       "step": 700
     },
     {
-      "epoch": 14.8,
-      "learning_rate": 8.506666666666668e-06,
-      "loss": 0.0029,
+      "epoch": 43.94,
+      "learning_rate": 8.513333333333335e-06,
+      "loss": 0.0012,
       "step": 725
     },
     {
-      "epoch": 15.31,
-      "learning_rate": 8.34e-06,
-      "loss": 0.003,
+      "epoch": 45.45,
+      "learning_rate": 8.346666666666668e-06,
+      "loss": 0.0011,
       "step": 750
     },
     {
-      "epoch": 15.82,
-      "learning_rate": 8.173333333333334e-06,
-      "loss": 0.0021,
+      "epoch": 46.97,
+      "learning_rate": 8.18e-06,
+      "loss": 0.001,
       "step": 775
     },
     {
-      "epoch": 16.33,
-      "learning_rate": 8.006666666666667e-06,
-      "loss": 0.0019,
+      "epoch": 48.48,
+      "learning_rate": 8.013333333333333e-06,
+      "loss": 0.001,
       "step": 800
     },
     {
-      "epoch": 16.33,
-      "eval_loss": 0.9069271087646484,
-      "eval_runtime": 147.3435,
-      "eval_samples_per_second": 6.278,
-      "eval_steps_per_second": 0.394,
-      "eval_wer": 46.44375,
+      "epoch": 48.48,
+      "eval_loss": 1.0617018938064575,
+      "eval_runtime": 56.831,
+      "eval_samples_per_second": 16.276,
+      "eval_steps_per_second": 0.352,
+      "eval_wer": 49.93972272453286,
       "step": 800
     },
     {
-      "epoch": 16.84,
-      "learning_rate": 7.840000000000001e-06,
-      "loss": 0.0019,
+      "epoch": 50.0,
+      "learning_rate": 7.846666666666667e-06,
+      "loss": 0.0009,
       "step": 825
     },
     {
-      "epoch": 17.35,
-      "learning_rate": 7.673333333333333e-06,
-      "loss": 0.0019,
+      "epoch": 51.52,
+      "learning_rate": 7.680000000000001e-06,
+      "loss": 0.0009,
       "step": 850
     },
     {
-      "epoch": 17.86,
-      "learning_rate": 7.506666666666668e-06,
-      "loss": 0.0016,
+      "epoch": 53.03,
+      "learning_rate": 7.513333333333334e-06,
+      "loss": 0.0008,
       "step": 875
     },
     {
-      "epoch": 18.37,
-      "learning_rate": 7.340000000000001e-06,
-      "loss": 0.0018,
+      "epoch": 54.55,
+      "learning_rate": 7.346666666666668e-06,
+      "loss": 0.0008,
       "step": 900
     },
     {
-      "epoch": 18.88,
-      "learning_rate": 7.173333333333335e-06,
-      "loss": 0.0014,
+      "epoch": 56.06,
+      "learning_rate": 7.180000000000001e-06,
+      "loss": 0.0007,
       "step": 925
     },
     {
-      "epoch": 19.39,
-      "learning_rate": 7.006666666666667e-06,
-      "loss": 0.0012,
+      "epoch": 57.58,
+      "learning_rate": 7.0133333333333345e-06,
+      "loss": 0.0007,
       "step": 950
     },
     {
-      "epoch": 19.9,
-      "learning_rate": 6.8400000000000014e-06,
-      "loss": 0.0013,
+      "epoch": 59.09,
+      "learning_rate": 6.846666666666667e-06,
+      "loss": 0.0007,
       "step": 975
     },
     {
-      "epoch": 20.41,
-      "learning_rate": 6.6733333333333335e-06,
-      "loss": 0.0012,
+      "epoch": 60.61,
+      "learning_rate": 6.680000000000001e-06,
+      "loss": 0.0007,
       "step": 1000
     },
     {
-      "epoch": 20.92,
-      "learning_rate": 6.5066666666666665e-06,
-      "loss": 0.0011,
+      "epoch": 62.12,
+      "learning_rate": 6.513333333333333e-06,
+      "loss": 0.0006,
       "step": 1025
     },
     {
-      "epoch": 21.43,
-      "learning_rate": 6.34e-06,
-      "loss": 0.001,
+      "epoch": 63.64,
+      "learning_rate": 6.346666666666668e-06,
+      "loss": 0.0006,
       "step": 1050
     },
     {
-      "epoch": 21.94,
-      "learning_rate": 6.173333333333333e-06,
-      "loss": 0.001,
+      "epoch": 65.15,
+      "learning_rate": 6.18e-06,
+      "loss": 0.0006,
       "step": 1075
     },
     {
-      "epoch": 22.45,
-      "learning_rate": 6.006666666666667e-06,
-      "loss": 0.0009,
+      "epoch": 66.67,
+      "learning_rate": 6.013333333333335e-06,
+      "loss": 0.0006,
       "step": 1100
     },
     {
-      "epoch": 22.96,
-      "learning_rate": 5.84e-06,
-      "loss": 0.0009,
+      "epoch": 68.18,
+      "learning_rate": 5.846666666666667e-06,
+      "loss": 0.0005,
       "step": 1125
     },
     {
-      "epoch": 23.47,
-      "learning_rate": 5.673333333333334e-06,
-      "loss": 0.0009,
+      "epoch": 69.7,
+      "learning_rate": 5.68e-06,
+      "loss": 0.0005,
       "step": 1150
     },
     {
-      "epoch": 23.98,
-      "learning_rate": 5.506666666666667e-06,
-      "loss": 0.0008,
+      "epoch": 71.21,
+      "learning_rate": 5.513333333333334e-06,
+      "loss": 0.0005,
       "step": 1175
     },
     {
-      "epoch": 24.49,
-      "learning_rate": 5.3400000000000005e-06,
-      "loss": 0.0008,
+      "epoch": 72.73,
+      "learning_rate": 5.346666666666667e-06,
+      "loss": 0.0005,
       "step": 1200
     },
     {
-      "epoch": 24.49,
-      "eval_loss": 0.9473171830177307,
-      "eval_runtime": 148.8306,
-      "eval_samples_per_second": 6.215,
-      "eval_steps_per_second": 0.39,
-      "eval_wer": 46.04375,
+      "epoch": 72.73,
+      "eval_loss": 1.101595401763916,
+      "eval_runtime": 56.6617,
+      "eval_samples_per_second": 16.325,
+      "eval_steps_per_second": 0.353,
+      "eval_wer": 49.93972272453286,
       "step": 1200
     },
     {
-      "epoch": 25.0,
-      "learning_rate": 5.1733333333333335e-06,
-      "loss": 0.0008,
+      "epoch": 74.24,
+      "learning_rate": 5.18e-06,
+      "loss": 0.0005,
       "step": 1225
     },
     {
-      "epoch": 25.51,
-      "learning_rate": 5.006666666666667e-06,
-      "loss": 0.0008,
+      "epoch": 75.76,
+      "learning_rate": 5.013333333333333e-06,
+      "loss": 0.0005,
       "step": 1250
     },
     {
-      "epoch": 26.02,
-      "learning_rate": 4.84e-06,
-      "loss": 0.0007,
+      "epoch": 77.27,
+      "learning_rate": 4.846666666666667e-06,
+      "loss": 0.0005,
       "step": 1275
     },
     {
-      "epoch": 26.53,
-      "learning_rate": 4.673333333333333e-06,
-      "loss": 0.0007,
+      "epoch": 78.79,
+      "learning_rate": 4.680000000000001e-06,
+      "loss": 0.0005,
       "step": 1300
     },
     {
-      "epoch": 27.04,
-      "learning_rate": 4.506666666666667e-06,
-      "loss": 0.0007,
+      "epoch": 80.3,
+      "learning_rate": 4.513333333333333e-06,
+      "loss": 0.0004,
       "step": 1325
     },
     {
-      "epoch": 27.55,
-      "learning_rate": 4.34e-06,
-      "loss": 0.0007,
+      "epoch": 81.82,
+      "learning_rate": 4.346666666666667e-06,
+      "loss": 0.0004,
       "step": 1350
     },
     {
-      "epoch": 28.06,
-      "learning_rate": 4.173333333333334e-06,
-      "loss": 0.0007,
+      "epoch": 83.33,
+      "learning_rate": 4.18e-06,
+      "loss": 0.0004,
       "step": 1375
     },
     {
-      "epoch": 28.57,
-      "learning_rate": 4.006666666666667e-06,
-      "loss": 0.0006,
+      "epoch": 84.85,
+      "learning_rate": 4.013333333333334e-06,
+      "loss": 0.0004,
       "step": 1400
     },
     {
-      "epoch": 29.08,
-      "learning_rate": 3.8400000000000005e-06,
-      "loss": 0.0007,
+      "epoch": 86.36,
+      "learning_rate": 3.8466666666666665e-06,
+      "loss": 0.0004,
       "step": 1425
     },
     {
-      "epoch": 29.59,
-      "learning_rate": 3.673333333333334e-06,
-      "loss": 0.0006,
+      "epoch": 87.88,
+      "learning_rate": 3.6800000000000003e-06,
+      "loss": 0.0004,
       "step": 1450
     },
     {
-      "epoch": 30.1,
-      "learning_rate": 3.5066666666666673e-06,
-      "loss": 0.0006,
+      "epoch": 89.39,
+      "learning_rate": 3.5133333333333337e-06,
+      "loss": 0.0004,
       "step": 1475
     },
     {
-      "epoch": 30.61,
-      "learning_rate": 3.3400000000000006e-06,
-      "loss": 0.0006,
+      "epoch": 90.91,
+      "learning_rate": 3.346666666666667e-06,
+      "loss": 0.0004,
       "step": 1500
     },
     {
-      "epoch": 31.12,
-      "learning_rate": 3.173333333333334e-06,
-      "loss": 0.0006,
+      "epoch": 92.42,
+      "learning_rate": 3.1800000000000005e-06,
+      "loss": 0.0004,
       "step": 1525
     },
     {
-      "epoch": 31.63,
-      "learning_rate": 3.0066666666666674e-06,
-      "loss": 0.0006,
+      "epoch": 93.94,
+      "learning_rate": 3.013333333333334e-06,
+      "loss": 0.0004,
       "step": 1550
     },
     {
-      "epoch": 32.14,
-      "learning_rate": 2.84e-06,
-      "loss": 0.0006,
+      "epoch": 95.45,
+      "learning_rate": 2.8466666666666672e-06,
+      "loss": 0.0004,
       "step": 1575
     },
     {
-      "epoch": 32.65,
-      "learning_rate": 2.6733333333333333e-06,
-      "loss": 0.0006,
+      "epoch": 96.97,
+      "learning_rate": 2.68e-06,
+      "loss": 0.0004,
       "step": 1600
     },
     {
-      "epoch": 32.65,
-      "eval_loss": 0.965289831161499,
-      "eval_runtime": 147.3092,
-      "eval_samples_per_second": 6.279,
-      "eval_steps_per_second": 0.394,
-      "eval_wer": 45.9,
+      "epoch": 96.97,
+      "eval_loss": 1.1220260858535767,
+      "eval_runtime": 56.5264,
+      "eval_samples_per_second": 16.364,
+      "eval_steps_per_second": 0.354,
+      "eval_wer": 49.90958408679928,
       "step": 1600
     },
     {
-      "epoch": 33.16,
-      "learning_rate": 2.5066666666666667e-06,
-      "loss": 0.0006,
+      "epoch": 98.48,
+      "learning_rate": 2.5133333333333336e-06,
+      "loss": 0.0004,
       "step": 1625
     },
     {
-      "epoch": 33.67,
-      "learning_rate": 2.3400000000000005e-06,
-      "loss": 0.0005,
+      "epoch": 100.0,
+      "learning_rate": 2.346666666666667e-06,
+      "loss": 0.0004,
       "step": 1650
     },
     {
-      "epoch": 34.18,
-      "learning_rate": 2.1733333333333334e-06,
-      "loss": 0.0006,
+      "epoch": 101.52,
+      "learning_rate": 2.1800000000000003e-06,
+      "loss": 0.0004,
       "step": 1675
     },
     {
-      "epoch": 34.69,
-      "learning_rate": 2.006666666666667e-06,
-      "loss": 0.0005,
+      "epoch": 103.03,
+      "learning_rate": 2.0133333333333337e-06,
+      "loss": 0.0004,
       "step": 1700
     },
     {
-      "epoch": 35.2,
-      "learning_rate": 1.8400000000000002e-06,
-      "loss": 0.0006,
+      "epoch": 104.55,
+      "learning_rate": 1.8466666666666668e-06,
+      "loss": 0.0004,
       "step": 1725
     },
     {
-      "epoch": 35.71,
-      "learning_rate": 1.6733333333333335e-06,
-      "loss": 0.0005,
+      "epoch": 106.06,
+      "learning_rate": 1.6800000000000002e-06,
+      "loss": 0.0003,
       "step": 1750
     },
     {
-      "epoch": 36.22,
-      "learning_rate": 1.506666666666667e-06,
-      "loss": 0.0005,
+      "epoch": 107.58,
+      "learning_rate": 1.5133333333333334e-06,
+      "loss": 0.0003,
       "step": 1775
     },
     {
-      "epoch": 36.73,
-      "learning_rate": 1.34e-06,
-      "loss": 0.0005,
+      "epoch": 109.09,
+      "learning_rate": 1.3466666666666668e-06,
+      "loss": 0.0003,
       "step": 1800
     },
     {
-      "epoch": 37.24,
-      "learning_rate": 1.1733333333333335e-06,
-      "loss": 0.0005,
+      "epoch": 110.61,
+      "learning_rate": 1.1800000000000001e-06,
+      "loss": 0.0003,
       "step": 1825
     },
     {
-      "epoch": 37.76,
-      "learning_rate": 1.0066666666666668e-06,
-      "loss": 0.0005,
+      "epoch": 112.12,
+      "learning_rate": 1.0133333333333333e-06,
+      "loss": 0.0003,
       "step": 1850
     },
     {
-      "epoch": 38.27,
-      "learning_rate": 8.400000000000001e-07,
-      "loss": 0.0005,
+      "epoch": 113.64,
+      "learning_rate": 8.466666666666668e-07,
+      "loss": 0.0003,
       "step": 1875
     },
     {
-      "epoch": 38.78,
-      "learning_rate": 6.733333333333334e-07,
-      "loss": 0.0005,
+      "epoch": 115.15,
+      "learning_rate": 6.800000000000001e-07,
+      "loss": 0.0003,
       "step": 1900
     },
     {
-      "epoch": 39.29,
-      "learning_rate": 5.066666666666667e-07,
-      "loss": 0.0005,
+      "epoch": 116.67,
+      "learning_rate": 5.133333333333334e-07,
+      "loss": 0.0003,
       "step": 1925
     },
     {
-      "epoch": 39.8,
-      "learning_rate": 3.4000000000000003e-07,
-      "loss": 0.0005,
+      "epoch": 118.18,
+      "learning_rate": 3.466666666666667e-07,
+      "loss": 0.0003,
       "step": 1950
     },
     {
-      "epoch": 40.31,
-      "learning_rate": 1.7333333333333335e-07,
-      "loss": 0.0005,
+      "epoch": 119.7,
+      "learning_rate": 1.8e-07,
+      "loss": 0.0003,
       "step": 1975
     },
     {
-      "epoch": 40.82,
-      "learning_rate": 6.666666666666667e-09,
-      "loss": 0.0005,
+      "epoch": 121.21,
+      "learning_rate": 1.3333333333333334e-08,
+      "loss": 0.0003,
       "step": 2000
     },
     {
-      "epoch": 40.82,
-      "eval_loss": 0.9719672799110413,
-      "eval_runtime": 147.2827,
-      "eval_samples_per_second": 6.28,
-      "eval_steps_per_second": 0.394,
-      "eval_wer": 45.94375,
+      "epoch": 121.21,
+      "eval_loss": 1.1297551393508911,
+      "eval_runtime": 56.6433,
+      "eval_samples_per_second": 16.33,
+      "eval_steps_per_second": 0.353,
+      "eval_wer": 50.04219409282701,
       "step": 2000
     },
     {
-      "epoch": 40.82,
+      "epoch": 121.21,
       "step": 2000,
-      "total_flos": 9.188559230976655e+18,
-      "train_loss": 0.18361923832818866,
-      "train_runtime": 4653.6924,
-      "train_samples_per_second": 6.876,
-      "train_steps_per_second": 0.43
+      "total_flos": 2.770419932175794e+19,
+      "train_loss": 0.143078886593692,
+      "train_runtime": 4289.2081,
+      "train_samples_per_second": 22.382,
+      "train_steps_per_second": 0.466
     }
   ],
   "logging_steps": 25,
   "max_steps": 2000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 41,
+  "num_train_epochs": 125,
   "save_steps": 400,
-  "total_flos": 9.188559230976655e+18,
+  "total_flos": 2.770419932175794e+19,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null