End of training

README.md

@@ -2,6 +2,8 @@
 license: apache-2.0
 base_model: facebook/wav2vec2-large-xlsr-53
 tags:
+- automatic-speech-recognition
+- DewiBrynJones/banc-trawsgrifiadau-bangor-clean-with-ccv
 - generated_from_trainer
 metrics:
 - wer
@@ -15,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # wav2vec2-xlsr-53-ft-btb-ccv-cy
 
-This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on an unknown dataset.
+This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the DEWIBRYNJONES/BANC-TRAWSGRIFIADAU-BANGOR-CLEAN-WITH-CCV - DEFAULT dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.7821
 - Wer: 0.4576

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 2.3166023166023164,
-    "eval_loss": Infinity,
-    "eval_runtime": 194.093,
+    "epoch": 1.8933417481855475,
+    "eval_loss": 0.7821305990219116,
+    "eval_runtime": 188.4729,
     "eval_samples": 7022,
-    "eval_samples_per_second": 36.179,
-    "eval_steps_per_second": 0.567,
-    "eval_wer": 0.3263604845935102,
-    "total_flos": 1.2607274134194512e+19,
-    "train_loss": 0.8722912038167318,
-    "train_runtime": 9671.7075,
-    "train_samples": 41435,
-    "train_samples_per_second": 9.926,
-    "train_steps_per_second": 0.62
+    "eval_samples_per_second": 37.257,
+    "eval_steps_per_second": 0.584,
+    "eval_wer": 0.45755324944524906,
+    "total_flos": 1.3297699695003722e+19,
+    "train_loss": 0.9828314208984374,
+    "train_runtime": 9865.427,
+    "train_samples": 50697,
+    "train_samples_per_second": 9.731,
+    "train_steps_per_second": 0.608
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.3166023166023164,
-    "eval_loss": Infinity,
-    "eval_runtime": 194.093,
+    "epoch": 1.8933417481855475,
+    "eval_loss": 0.7821305990219116,
+    "eval_runtime": 188.4729,
     "eval_samples": 7022,
-    "eval_samples_per_second": 36.179,
-    "eval_steps_per_second": 0.567,
-    "eval_wer": 0.3263604845935102
+    "eval_samples_per_second": 37.257,
+    "eval_steps_per_second": 0.584,
+    "eval_wer": 0.45755324944524906
 }

runs/Aug28_14-17-49_6d77da852b30/events.out.tfevents.1724862820.6d77da852b30.31.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e97d4393abe26b409f20689992f7f125f326b45b0281ff9f0aecefd61bd14ba
+size 406

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.3166023166023164,
-    "total_flos": 1.2607274134194512e+19,
-    "train_loss": 0.8722912038167318,
-    "train_runtime": 9671.7075,
-    "train_samples": 41435,
-    "train_samples_per_second": 9.926,
-    "train_steps_per_second": 0.62
+    "epoch": 1.8933417481855475,
+    "total_flos": 1.3297699695003722e+19,
+    "train_loss": 0.9828314208984374,
+    "train_runtime": 9865.427,
+    "train_samples": 50697,
+    "train_samples_per_second": 9.731,
+    "train_steps_per_second": 0.608
 }

trainer_state.json

@@ -1,7 +1,7 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.3166023166023164,
+  "epoch": 1.8933417481855475,
   "eval_steps": 200,
   "global_step": 6000,
   "is_hyper_param_search": false,
@@ -9,373 +9,373 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.07722007722007722,
-      "eval_loss": Infinity,
-      "eval_runtime": 223.773,
-      "eval_samples_per_second": 31.38,
-      "eval_steps_per_second": 0.492,
+      "epoch": 0.06311139160618491,
+      "eval_loss": 3.0427980422973633,
+      "eval_runtime": 195.72,
+      "eval_samples_per_second": 35.878,
+      "eval_steps_per_second": 0.562,
       "eval_wer": 1.0,
       "step": 200
     },
     {
-      "epoch": 0.15444015444015444,
-      "eval_loss": Infinity,
-      "eval_runtime": 201.0458,
-      "eval_samples_per_second": 34.927,
-      "eval_steps_per_second": 0.547,
-      "eval_wer": 0.8963291911755158,
+      "epoch": 0.12622278321236982,
+      "eval_loss": 3.1055634021759033,
+      "eval_runtime": 193.3912,
+      "eval_samples_per_second": 36.31,
+      "eval_steps_per_second": 0.569,
+      "eval_wer": 1.0,
       "step": 400
     },
     {
-      "epoch": 0.19305019305019305,
-      "grad_norm": 4.3846588134765625,
-      "learning_rate": 0.00024799999999999996,
-      "loss": 3.9177,
+      "epoch": 0.1577784790154623,
+      "grad_norm": 6.020185947418213,
+      "learning_rate": 0.00024599999999999996,
+      "loss": 4.232,
       "step": 500
     },
     {
-      "epoch": 0.23166023166023167,
-      "eval_loss": Infinity,
-      "eval_runtime": 194.7647,
-      "eval_samples_per_second": 36.054,
-      "eval_steps_per_second": 0.565,
-      "eval_wer": 0.7594821653487074,
+      "epoch": 0.18933417481855475,
+      "eval_loss": 1.2093147039413452,
+      "eval_runtime": 194.3659,
+      "eval_samples_per_second": 36.128,
+      "eval_steps_per_second": 0.566,
+      "eval_wer": 0.8286185040991775,
       "step": 600
     },
     {
-      "epoch": 0.3088803088803089,
-      "eval_loss": Infinity,
-      "eval_runtime": 194.859,
-      "eval_samples_per_second": 36.036,
-      "eval_steps_per_second": 0.565,
-      "eval_wer": 0.7512108345344293,
+      "epoch": 0.25244556642473964,
+      "eval_loss": 1.0886054039001465,
+      "eval_runtime": 194.9365,
+      "eval_samples_per_second": 36.022,
+      "eval_steps_per_second": 0.564,
+      "eval_wer": 0.7324838225328533,
       "step": 800
     },
     {
-      "epoch": 0.3861003861003861,
-      "grad_norm": 3.36423921585083,
-      "learning_rate": 0.00027805555555555553,
-      "loss": 0.9791,
+      "epoch": 0.3155569580309246,
+      "grad_norm": 4.922729969024658,
+      "learning_rate": 0.00027822222222222224,
+      "loss": 1.0379,
       "step": 1000
     },
     {
-      "epoch": 0.3861003861003861,
-      "eval_loss": Infinity,
-      "eval_runtime": 195.6019,
-      "eval_samples_per_second": 35.899,
-      "eval_steps_per_second": 0.562,
-      "eval_wer": 0.598444418474803,
+      "epoch": 0.3155569580309246,
+      "eval_loss": 0.9017586708068848,
+      "eval_runtime": 193.3111,
+      "eval_samples_per_second": 36.325,
+      "eval_steps_per_second": 0.569,
+      "eval_wer": 0.673049070543058,
       "step": 1000
     },
     {
-      "epoch": 0.46332046332046334,
-      "eval_loss": Infinity,
-      "eval_runtime": 197.8836,
-      "eval_samples_per_second": 35.486,
-      "eval_steps_per_second": 0.556,
-      "eval_wer": 0.5867710646254528,
+      "epoch": 0.3786683496371095,
+      "eval_loss": 0.8131064772605896,
+      "eval_runtime": 192.8235,
+      "eval_samples_per_second": 36.417,
+      "eval_steps_per_second": 0.57,
+      "eval_wer": 0.5803292629452956,
       "step": 1200
     },
     {
-      "epoch": 0.5405405405405406,
-      "eval_loss": Infinity,
-      "eval_runtime": 203.8782,
-      "eval_samples_per_second": 34.442,
-      "eval_steps_per_second": 0.54,
-      "eval_wer": 0.5255476348149014,
+      "epoch": 0.4417797412432944,
+      "eval_loss": 0.7567442655563354,
+      "eval_runtime": 192.8955,
+      "eval_samples_per_second": 36.403,
+      "eval_steps_per_second": 0.57,
+      "eval_wer": 0.5543788782736948,
       "step": 1400
     },
     {
-      "epoch": 0.5791505791505791,
-      "grad_norm": 2.6850786209106445,
-      "learning_rate": 0.0002503333333333333,
-      "loss": 0.805,
+      "epoch": 0.47333543704638686,
+      "grad_norm": 25.353076934814453,
+      "learning_rate": 0.00025049999999999996,
+      "loss": 0.8008,
       "step": 1500
     },
     {
-      "epoch": 0.6177606177606177,
-      "eval_loss": Infinity,
-      "eval_runtime": 199.2388,
-      "eval_samples_per_second": 35.244,
-      "eval_steps_per_second": 0.552,
-      "eval_wer": 0.5281575837845559,
+      "epoch": 0.5048911328494793,
+      "eval_loss": 0.704021155834198,
+      "eval_runtime": 193.2182,
+      "eval_samples_per_second": 36.342,
+      "eval_steps_per_second": 0.569,
+      "eval_wer": 0.5137857266240583,
       "step": 1600
     },
     {
-      "epoch": 0.694980694980695,
-      "eval_loss": Infinity,
-      "eval_runtime": 194.9574,
-      "eval_samples_per_second": 36.018,
-      "eval_steps_per_second": 0.564,
-      "eval_wer": 0.4768805266643294,
+      "epoch": 0.5680025244556642,
+      "eval_loss": 0.6949470043182373,
+      "eval_runtime": 192.7198,
+      "eval_samples_per_second": 36.436,
+      "eval_steps_per_second": 0.571,
+      "eval_wer": 0.5236450125566378,
       "step": 1800
     },
     {
-      "epoch": 0.7722007722007722,
-      "grad_norm": 2.9242658615112305,
-      "learning_rate": 0.0002226111111111111,
-      "loss": 0.7184,
+      "epoch": 0.6311139160618492,
+      "grad_norm": NaN,
+      "learning_rate": 0.00022277777777777774,
+      "loss": 0.7212,
       "step": 2000
     },
     {
-      "epoch": 0.7722007722007722,
-      "eval_loss": Infinity,
-      "eval_runtime": 198.941,
-      "eval_samples_per_second": 35.297,
-      "eval_steps_per_second": 0.553,
-      "eval_wer": 0.4743095321569086,
+      "epoch": 0.6311139160618492,
+      "eval_loss": 0.672233521938324,
+      "eval_runtime": 194.6494,
+      "eval_samples_per_second": 36.075,
+      "eval_steps_per_second": 0.565,
+      "eval_wer": 0.4992758341195073,
       "step": 2000
     },
     {
-      "epoch": 0.8494208494208494,
-      "eval_loss": Infinity,
-      "eval_runtime": 207.8762,
-      "eval_samples_per_second": 33.78,
-      "eval_steps_per_second": 0.529,
-      "eval_wer": 0.46802487891654654,
+      "epoch": 0.694225307668034,
+      "eval_loss": 0.6403974890708923,
+      "eval_runtime": 192.2183,
+      "eval_samples_per_second": 36.531,
+      "eval_steps_per_second": 0.572,
+      "eval_wer": 0.4761689631804834,
       "step": 2200
     },
     {
-      "epoch": 0.9266409266409267,
-      "eval_loss": Infinity,
-      "eval_runtime": 215.534,
-      "eval_samples_per_second": 32.58,
-      "eval_steps_per_second": 0.51,
-      "eval_wer": 0.457026735745913,
+      "epoch": 0.757336699274219,
+      "eval_loss": 0.6335896849632263,
+      "eval_runtime": 192.0564,
+      "eval_samples_per_second": 36.562,
+      "eval_steps_per_second": 0.573,
+      "eval_wer": 0.4685552558497987,
       "step": 2400
     },
     {
-      "epoch": 0.9652509652509652,
-      "grad_norm": 3.865280866622925,
-      "learning_rate": 0.00019483333333333332,
-      "loss": 0.6704,
+      "epoch": 0.7888923950773115,
+      "grad_norm": 5.390285968780518,
+      "learning_rate": 0.00019505555555555555,
+      "loss": 0.6639,
       "step": 2500
     },
     {
-      "epoch": 1.0038610038610039,
-      "eval_loss": Infinity,
-      "eval_runtime": 212.4212,
-      "eval_samples_per_second": 33.057,
-      "eval_steps_per_second": 0.518,
-      "eval_wer": 0.4252528793840001,
+      "epoch": 0.8204480908804039,
+      "eval_loss": 0.593280553817749,
+      "eval_runtime": 191.8655,
+      "eval_samples_per_second": 36.599,
+      "eval_steps_per_second": 0.573,
+      "eval_wer": 0.45866939502252224,
       "step": 2600
     },
     {
-      "epoch": 1.0810810810810811,
-      "eval_loss": Infinity,
-      "eval_runtime": 211.2545,
-      "eval_samples_per_second": 33.24,
-      "eval_steps_per_second": 0.521,
-      "eval_wer": 0.4163972316362173,
+      "epoch": 0.8835594824865888,
+      "eval_loss": 0.5996308326721191,
+      "eval_runtime": 192.1279,
+      "eval_samples_per_second": 36.549,
+      "eval_steps_per_second": 0.573,
+      "eval_wer": 0.45444398676570247,
       "step": 2800
     },
     {
-      "epoch": 1.1583011583011582,
-      "grad_norm": 0.6646206974983215,
-      "learning_rate": 0.00016716666666666665,
-      "loss": 0.5664,
+      "epoch": 0.9466708740927737,
+      "grad_norm": 7.231433391571045,
+      "learning_rate": 0.00016733333333333333,
+      "loss": 0.6278,
       "step": 3000
     },
     {
-      "epoch": 1.1583011583011582,
-      "eval_loss": Infinity,
-      "eval_runtime": 202.1583,
-      "eval_samples_per_second": 34.735,
-      "eval_steps_per_second": 0.544,
-      "eval_wer": 0.41592977808941345,
+      "epoch": 0.9466708740927737,
+      "eval_loss": 0.5639352202415466,
+      "eval_runtime": 195.2391,
+      "eval_samples_per_second": 35.966,
+      "eval_steps_per_second": 0.563,
+      "eval_wer": 0.42424161894258494,
       "step": 3000
     },
     {
-      "epoch": 1.2355212355212355,
-      "eval_loss": Infinity,
-      "eval_runtime": 198.8952,
-      "eval_samples_per_second": 35.305,
-      "eval_steps_per_second": 0.553,
-      "eval_wer": 0.3995039798475582,
+      "epoch": 1.0097822656989586,
+      "eval_loss": 0.566460371017456,
+      "eval_runtime": 191.9621,
+      "eval_samples_per_second": 36.58,
+      "eval_steps_per_second": 0.573,
+      "eval_wer": 0.4227534248395541,
       "step": 3200
     },
     {
-      "epoch": 1.3127413127413128,
-      "eval_loss": Infinity,
-      "eval_runtime": 197.172,
-      "eval_samples_per_second": 35.614,
-      "eval_steps_per_second": 0.558,
-      "eval_wer": 0.3940633399555919,
+      "epoch": 1.0728936573051435,
+      "eval_loss": 0.5476272106170654,
+      "eval_runtime": 193.6396,
+      "eval_samples_per_second": 36.263,
+      "eval_steps_per_second": 0.568,
+      "eval_wer": 0.41909937681871934,
       "step": 3400
     },
     {
-      "epoch": 1.3513513513513513,
-      "grad_norm": 1.06748366355896,
-      "learning_rate": 0.00013944444444444442,
-      "loss": 0.5359,
+      "epoch": 1.104449353108236,
+      "grad_norm": 4.519629955291748,
+      "learning_rate": 0.00013955555555555555,
+      "loss": 0.5528,
       "step": 3500
     },
     {
-      "epoch": 1.3899613899613898,
-      "eval_loss": Infinity,
-      "eval_runtime": 194.4658,
-      "eval_samples_per_second": 36.109,
-      "eval_steps_per_second": 0.566,
-      "eval_wer": 0.38185760845571526,
+      "epoch": 1.1360050489113285,
+      "eval_loss": 0.5440065860748291,
+      "eval_runtime": 191.7996,
+      "eval_samples_per_second": 36.611,
+      "eval_steps_per_second": 0.574,
+      "eval_wer": 0.41887349021379505,
       "step": 3600
     },
     {
-      "epoch": 1.4671814671814671,
-      "eval_loss": Infinity,
-      "eval_runtime": 201.7117,
-      "eval_samples_per_second": 34.812,
-      "eval_steps_per_second": 0.545,
-      "eval_wer": 0.3810785192110423,
+      "epoch": 1.1991164405175134,
+      "eval_loss": 0.5297770500183105,
+      "eval_runtime": 191.4083,
+      "eval_samples_per_second": 36.686,
+      "eval_steps_per_second": 0.575,
+      "eval_wer": 0.4070875244156845,
       "step": 3800
     },
     {
-      "epoch": 1.5444015444015444,
-      "grad_norm": 0.8601678013801575,
-      "learning_rate": 0.00011166666666666667,
-      "loss": 0.5172,
+      "epoch": 1.2622278321236984,
+      "grad_norm": 1.4278947114944458,
+      "learning_rate": 0.00011183333333333332,
+      "loss": 0.5103,
       "step": 4000
     },
     {
-      "epoch": 1.5444015444015444,
-      "eval_loss": Infinity,
-      "eval_runtime": 196.5164,
-      "eval_samples_per_second": 35.732,
-      "eval_steps_per_second": 0.56,
-      "eval_wer": 0.36905457520158935,
+      "epoch": 1.2622278321236984,
+      "eval_loss": 0.5384453535079956,
+      "eval_runtime": 195.1408,
+      "eval_samples_per_second": 35.984,
+      "eval_steps_per_second": 0.564,
+      "eval_wer": 0.4025033550804555,
       "step": 4000
     },
     {
-      "epoch": 1.6216216216216215,
-      "eval_loss": Infinity,
-      "eval_runtime": 193.7969,
-      "eval_samples_per_second": 36.234,
-      "eval_steps_per_second": 0.568,
-      "eval_wer": 0.36086115331177854,
+      "epoch": 1.325339223729883,
+      "eval_loss": 0.531086266040802,
+      "eval_runtime": 192.0473,
+      "eval_samples_per_second": 36.564,
+      "eval_steps_per_second": 0.573,
+      "eval_wer": 0.39976614092666657,
       "step": 4200
     },
     {
-      "epoch": 1.698841698841699,
-      "eval_loss": Infinity,
-      "eval_runtime": 196.3411,
-      "eval_samples_per_second": 35.764,
-      "eval_steps_per_second": 0.56,
-      "eval_wer": 0.3599652006804046,
+      "epoch": 1.388450615336068,
+      "eval_loss": 0.5395579934120178,
+      "eval_runtime": 191.3816,
+      "eval_samples_per_second": 36.691,
+      "eval_steps_per_second": 0.575,
+      "eval_wer": 0.4039383993940924,
       "step": 4400
     },
     {
-      "epoch": 1.7374517374517375,
-      "grad_norm": 0.6527121663093567,
-      "learning_rate": 8.394444444444443e-05,
-      "loss": 0.4817,
+      "epoch": 1.4200063111391605,
+      "grad_norm": 2.4821906089782715,
+      "learning_rate": 8.411111111111111e-05,
+      "loss": 0.5194,
       "step": 4500
     },
     {
-      "epoch": 1.776061776061776,
-      "eval_loss": Infinity,
-      "eval_runtime": 194.384,
-      "eval_samples_per_second": 36.124,
-      "eval_steps_per_second": 0.566,
-      "eval_wer": 0.35086284133847534,
+      "epoch": 1.451562006942253,
+      "eval_loss": 0.5501742959022522,
+      "eval_runtime": 190.6511,
+      "eval_samples_per_second": 36.832,
+      "eval_steps_per_second": 0.577,
+      "eval_wer": 0.4048818081558352,
       "step": 4600
     },
     {
-      "epoch": 1.8532818532818531,
-      "eval_loss": Infinity,
-      "eval_runtime": 196.7828,
-      "eval_samples_per_second": 35.684,
-      "eval_steps_per_second": 0.559,
-      "eval_wer": 0.3529663822990924,
+      "epoch": 1.514673398548438,
+      "eval_loss": 0.6632032990455627,
+      "eval_runtime": 193.4271,
+      "eval_samples_per_second": 36.303,
+      "eval_steps_per_second": 0.569,
+      "eval_wer": 0.4365059328452411,
       "step": 4800
     },
     {
-      "epoch": 1.9305019305019306,
-      "grad_norm": 0.7631692886352539,
-      "learning_rate": 5.6166666666666665e-05,
-      "loss": 0.4818,
+      "epoch": 1.577784790154623,
+      "grad_norm": 11.192009925842285,
+      "learning_rate": 5.649999999999999e-05,
+      "loss": 0.6034,
       "step": 5000
     },
     {
-      "epoch": 1.9305019305019306,
-      "eval_loss": Infinity,
-      "eval_runtime": 194.6791,
-      "eval_samples_per_second": 36.07,
-      "eval_steps_per_second": 0.565,
-      "eval_wer": 0.34340955423110386,
+      "epoch": 1.577784790154623,
+      "eval_loss": 0.7074605226516724,
+      "eval_runtime": 191.8555,
+      "eval_samples_per_second": 36.6,
+      "eval_steps_per_second": 0.573,
+      "eval_wer": 0.4368115441342564,
       "step": 5000
     },
     {
-      "epoch": 2.0077220077220077,
-      "eval_loss": Infinity,
-      "eval_runtime": 200.454,
-      "eval_samples_per_second": 35.03,
-      "eval_steps_per_second": 0.549,
-      "eval_wer": 0.336254918000857,
+      "epoch": 1.6408961817608079,
+      "eval_loss": 0.7465850710868835,
+      "eval_runtime": 191.8575,
+      "eval_samples_per_second": 36.6,
+      "eval_steps_per_second": 0.573,
+      "eval_wer": 0.44187406157403103,
       "step": 5200
     },
     {
-      "epoch": 2.0849420849420848,
-      "eval_loss": Infinity,
-      "eval_runtime": 193.7233,
-      "eval_samples_per_second": 36.248,
-      "eval_steps_per_second": 0.568,
-      "eval_wer": 0.33718982509446455,
+      "epoch": 1.7040075733669928,
+      "eval_loss": 0.7624653577804565,
+      "eval_runtime": 189.4967,
+      "eval_samples_per_second": 37.056,
+      "eval_steps_per_second": 0.58,
+      "eval_wer": 0.449819955088428,
       "step": 5400
     },
     {
-      "epoch": 2.1235521235521237,
-      "grad_norm": 1.005771279335022,
-      "learning_rate": 2.844444444444444e-05,
-      "loss": 0.4196,
+      "epoch": 1.7355632691700853,
+      "grad_norm": 17.51552963256836,
+      "learning_rate": 2.8777777777777776e-05,
+      "loss": 0.74,
       "step": 5500
     },
     {
-      "epoch": 2.1621621621621623,
-      "eval_loss": Infinity,
-      "eval_runtime": 198.8274,
-      "eval_samples_per_second": 35.317,
-      "eval_steps_per_second": 0.553,
-      "eval_wer": 0.3320348512588784,
+      "epoch": 1.7671189649731778,
+      "eval_loss": 0.7502115368843079,
+      "eval_runtime": 192.6564,
+      "eval_samples_per_second": 36.448,
+      "eval_steps_per_second": 0.571,
+      "eval_wer": 0.4474813643550937,
       "step": 5600
     },
     {
-      "epoch": 2.2393822393822393,
-      "eval_loss": Infinity,
-      "eval_runtime": 197.3434,
-      "eval_samples_per_second": 35.583,
-      "eval_steps_per_second": 0.557,
-      "eval_wer": 0.3292690844402893,
+      "epoch": 1.8302303565793627,
+      "eval_loss": 0.7739897966384888,
+      "eval_runtime": 189.3036,
+      "eval_samples_per_second": 37.094,
+      "eval_steps_per_second": 0.581,
+      "eval_wer": 0.45959951633691654,
       "step": 5800
     },
     {
-      "epoch": 2.3166023166023164,
-      "grad_norm": 1.702697515487671,
-      "learning_rate": 7.222222222222222e-07,
-      "loss": 0.3743,
+      "epoch": 1.8933417481855475,
+      "grad_norm": 10.696316719055176,
+      "learning_rate": 1e-06,
+      "loss": 0.7844,
       "step": 6000
     },
     {
-      "epoch": 2.3166023166023164,
-      "eval_loss": Infinity,
-      "eval_runtime": 192.6595,
-      "eval_samples_per_second": 36.448,
-      "eval_steps_per_second": 0.571,
-      "eval_wer": 0.3263604845935102,
+      "epoch": 1.8933417481855475,
+      "eval_loss": 0.7821305990219116,
+      "eval_runtime": 189.7595,
+      "eval_samples_per_second": 37.005,
+      "eval_steps_per_second": 0.58,
+      "eval_wer": 0.45755324944524906,
       "step": 6000
     },
     {
-      "epoch": 2.3166023166023164,
+      "epoch": 1.8933417481855475,
       "step": 6000,
-      "total_flos": 1.2607274134194512e+19,
-      "train_loss": 0.8722912038167318,
-      "train_runtime": 9671.7075,
-      "train_samples_per_second": 9.926,
-      "train_steps_per_second": 0.62
+      "total_flos": 1.3297699695003722e+19,
+      "train_loss": 0.9828314208984374,
+      "train_runtime": 9865.427,
+      "train_samples_per_second": 9.731,
+      "train_steps_per_second": 0.608
     }
   ],
   "logging_steps": 500,
   "max_steps": 6000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
+  "num_train_epochs": 2,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -389,7 +389,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.2607274134194512e+19,
+  "total_flos": 1.3297699695003722e+19,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null