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: 0.4290
 - Wer: 0.3378

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 2.719665271966527,
-    "eval_loss": 0.41978269815444946,
-    "eval_runtime": 153.1864,
+    "epoch": 2.013162988772745,
+    "eval_loss": 0.4289664328098297,
+    "eval_runtime": 173.7431,
     "eval_samples": 5656,
-    "eval_samples_per_second": 36.922,
-    "eval_steps_per_second": 4.615,
-    "eval_wer": 0.32485435958338016,
-    "total_flos": 1.0779565758702488e+19,
-    "train_loss": 1.1314451643136831,
-    "train_runtime": 7270.6329,
-    "train_samples": 30577,
-    "train_samples_per_second": 11.443,
-    "train_steps_per_second": 0.358
+    "eval_samples_per_second": 32.554,
+    "eval_steps_per_second": 4.069,
+    "eval_wer": 0.33783762096580056,
+    "total_flos": 1.1633941226063049e+19,
+    "train_loss": 1.1463891924344576,
+    "train_runtime": 8433.2953,
+    "train_samples": 41326,
+    "train_samples_per_second": 9.866,
+    "train_steps_per_second": 0.308
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.719665271966527,
-    "eval_loss": 0.41978269815444946,
-    "eval_runtime": 153.1864,
+    "epoch": 2.013162988772745,
+    "eval_loss": 0.4289664328098297,
+    "eval_runtime": 173.7431,
     "eval_samples": 5656,
-    "eval_samples_per_second": 36.922,
-    "eval_steps_per_second": 4.615,
-    "eval_wer": 0.32485435958338016
+    "eval_samples_per_second": 32.554,
+    "eval_steps_per_second": 4.069,
+    "eval_wer": 0.33783762096580056
 }

runs/May23_17-01-20_7571a1a6d5fc/events.out.tfevents.1716489680.7571a1a6d5fc.493.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1229cd6a8b5b49fcd581f45f5616a8c92817922575735c615033946c8d44ce39
+size 406

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.719665271966527,
-    "total_flos": 1.0779565758702488e+19,
-    "train_loss": 1.1314451643136831,
-    "train_runtime": 7270.6329,
-    "train_samples": 30577,
-    "train_samples_per_second": 11.443,
-    "train_steps_per_second": 0.358
+    "epoch": 2.013162988772745,
+    "total_flos": 1.1633941226063049e+19,
+    "train_loss": 1.1463891924344576,
+    "train_runtime": 8433.2953,
+    "train_samples": 41326,
+    "train_samples_per_second": 9.866,
+    "train_steps_per_second": 0.308
 }

trainer_state.json

@@ -1,7 +1,7 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.719665271966527,
+  "epoch": 2.013162988772745,
   "eval_steps": 100,
   "global_step": 2600,
   "is_hyper_param_search": false,
@@ -9,282 +9,282 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.10460251046025104,
-      "eval_loss": 3.505904197692871,
-      "eval_runtime": 155.3669,
-      "eval_samples_per_second": 36.404,
-      "eval_steps_per_second": 4.551,
+      "epoch": 0.07742934572202866,
+      "eval_loss": 3.5325253009796143,
+      "eval_runtime": 173.1277,
+      "eval_samples_per_second": 32.67,
+      "eval_steps_per_second": 4.084,
       "eval_wer": 1.0,
       "step": 100
     },
     {
-      "epoch": 0.20920502092050208,
-      "eval_loss": 3.2686526775360107,
-      "eval_runtime": 151.6355,
-      "eval_samples_per_second": 37.3,
-      "eval_steps_per_second": 4.662,
+      "epoch": 0.1548586914440573,
+      "eval_loss": 2.965233325958252,
+      "eval_runtime": 171.2442,
+      "eval_samples_per_second": 33.029,
+      "eval_steps_per_second": 4.129,
       "eval_wer": 1.0,
       "step": 200
     },
     {
-      "epoch": 0.3138075313807531,
-      "eval_loss": 2.7369139194488525,
-      "eval_runtime": 151.6605,
-      "eval_samples_per_second": 37.294,
-      "eval_steps_per_second": 4.662,
+      "epoch": 0.23228803716608595,
+      "eval_loss": 2.8520941734313965,
+      "eval_runtime": 171.3893,
+      "eval_samples_per_second": 33.001,
+      "eval_steps_per_second": 4.125,
       "eval_wer": 1.0,
       "step": 300
     },
     {
-      "epoch": 0.41841004184100417,
-      "eval_loss": 1.303921103477478,
-      "eval_runtime": 151.8516,
-      "eval_samples_per_second": 37.247,
-      "eval_steps_per_second": 4.656,
-      "eval_wer": 0.8544558745646836,
+      "epoch": 0.3097173828881146,
+      "eval_loss": 1.247292160987854,
+      "eval_runtime": 171.2844,
+      "eval_samples_per_second": 33.021,
+      "eval_steps_per_second": 4.128,
+      "eval_wer": 0.8264832854552165,
       "step": 400
     },
     {
-      "epoch": 0.5230125523012552,
-      "grad_norm": 2.7574479579925537,
-      "learning_rate": 0.0002958,
-      "loss": 3.5426,
+      "epoch": 0.38714672861014326,
+      "grad_norm": 1.9864516258239746,
+      "learning_rate": 0.00029699999999999996,
+      "loss": 3.7403,
       "step": 500
     },
     {
-      "epoch": 0.5230125523012552,
-      "eval_loss": 1.051817536354065,
-      "eval_runtime": 152.6837,
-      "eval_samples_per_second": 37.044,
-      "eval_steps_per_second": 4.63,
-      "eval_wer": 0.7738120075107124,
+      "epoch": 0.38714672861014326,
+      "eval_loss": 0.9730328321456909,
+      "eval_runtime": 172.0915,
+      "eval_samples_per_second": 32.866,
+      "eval_steps_per_second": 4.108,
+      "eval_wer": 0.7234356694644605,
       "step": 500
     },
     {
-      "epoch": 0.6276150627615062,
-      "eval_loss": 0.839580774307251,
-      "eval_runtime": 153.5079,
-      "eval_samples_per_second": 36.845,
-      "eval_steps_per_second": 4.606,
-      "eval_wer": 0.6148994559548073,
+      "epoch": 0.4645760743321719,
+      "eval_loss": 0.8328432440757751,
+      "eval_runtime": 172.9073,
+      "eval_samples_per_second": 32.711,
+      "eval_steps_per_second": 4.089,
+      "eval_wer": 0.6177560944295549,
       "step": 600
     },
     {
-      "epoch": 0.7322175732217573,
-      "eval_loss": 0.7548021674156189,
-      "eval_runtime": 152.4899,
-      "eval_samples_per_second": 37.091,
-      "eval_steps_per_second": 4.636,
-      "eval_wer": 0.5686636388438638,
+      "epoch": 0.5420054200542005,
+      "eval_loss": 0.7426055073738098,
+      "eval_runtime": 173.8661,
+      "eval_samples_per_second": 32.531,
+      "eval_steps_per_second": 4.066,
+      "eval_wer": 0.5505127505576864,
       "step": 700
     },
     {
-      "epoch": 0.8368200836820083,
-      "eval_loss": 0.6868751645088196,
-      "eval_runtime": 153.2403,
-      "eval_samples_per_second": 36.909,
-      "eval_steps_per_second": 4.614,
-      "eval_wer": 0.5185119802282101,
+      "epoch": 0.6194347657762292,
+      "eval_loss": 0.7127000093460083,
+      "eval_runtime": 172.8776,
+      "eval_samples_per_second": 32.717,
+      "eval_steps_per_second": 4.09,
+      "eval_wer": 0.5540113302627144,
       "step": 800
     },
     {
-      "epoch": 0.9414225941422594,
-      "eval_loss": 0.6501098871231079,
-      "eval_runtime": 152.6877,
-      "eval_samples_per_second": 37.043,
-      "eval_steps_per_second": 4.63,
-      "eval_wer": 0.4895283336810515,
+      "epoch": 0.6968641114982579,
+      "eval_loss": 0.6691900491714478,
+      "eval_runtime": 171.858,
+      "eval_samples_per_second": 32.911,
+      "eval_steps_per_second": 4.114,
+      "eval_wer": 0.5079680955208551,
       "step": 900
     },
     {
-      "epoch": 1.0460251046025104,
-      "grad_norm": 1.2487365007400513,
-      "learning_rate": 0.00022957142857142856,
-      "loss": 0.7785,
+      "epoch": 0.7742934572202865,
+      "grad_norm": 3.1098098754882812,
+      "learning_rate": 0.00022928571428571426,
+      "loss": 0.7271,
       "step": 1000
     },
     {
-      "epoch": 1.0460251046025104,
-      "eval_loss": 0.5817079544067383,
-      "eval_runtime": 153.4094,
-      "eval_samples_per_second": 36.869,
-      "eval_steps_per_second": 4.609,
-      "eval_wer": 0.4500649965495659,
+      "epoch": 0.7742934572202865,
+      "eval_loss": 0.6375711560249329,
+      "eval_runtime": 171.5356,
+      "eval_samples_per_second": 32.973,
+      "eval_steps_per_second": 4.122,
+      "eval_wer": 0.5256214793535652,
       "step": 1000
     },
     {
-      "epoch": 1.1506276150627615,
-      "eval_loss": 0.5646732449531555,
-      "eval_runtime": 152.9118,
-      "eval_samples_per_second": 36.989,
-      "eval_steps_per_second": 4.624,
-      "eval_wer": 0.4288006932965287,
+      "epoch": 0.8517228029423152,
+      "eval_loss": 0.6119316816329956,
+      "eval_runtime": 171.7234,
+      "eval_samples_per_second": 32.937,
+      "eval_steps_per_second": 4.117,
+      "eval_wer": 0.47057501885702363,
       "step": 1100
     },
     {
-      "epoch": 1.2552301255230125,
-      "eval_loss": 0.5424243807792664,
-      "eval_runtime": 152.6245,
-      "eval_samples_per_second": 37.058,
-      "eval_steps_per_second": 4.632,
-      "eval_wer": 0.42786987851262215,
+      "epoch": 0.9291521486643438,
+      "eval_loss": 0.5987285375595093,
+      "eval_runtime": 171.8326,
+      "eval_samples_per_second": 32.916,
+      "eval_steps_per_second": 4.114,
+      "eval_wer": 0.4651024698688835,
       "step": 1200
     },
     {
-      "epoch": 1.3598326359832635,
-      "eval_loss": 0.5264406800270081,
-      "eval_runtime": 153.2158,
-      "eval_samples_per_second": 36.915,
-      "eval_steps_per_second": 4.614,
-      "eval_wer": 0.4050809648376691,
+      "epoch": 1.0065814943863725,
+      "eval_loss": 0.56138676404953,
+      "eval_runtime": 173.244,
+      "eval_samples_per_second": 32.648,
+      "eval_steps_per_second": 4.081,
+      "eval_wer": 0.4267304328288745,
       "step": 1300
     },
     {
-      "epoch": 1.4644351464435146,
-      "eval_loss": 0.5098912715911865,
-      "eval_runtime": 153.9853,
-      "eval_samples_per_second": 36.731,
-      "eval_steps_per_second": 4.591,
-      "eval_wer": 0.3977467862817159,
+      "epoch": 1.084010840108401,
+      "eval_loss": 0.5463124513626099,
+      "eval_runtime": 171.8011,
+      "eval_samples_per_second": 32.922,
+      "eval_steps_per_second": 4.115,
+      "eval_wer": 0.4228948339779493,
       "step": 1400
     },
     {
-      "epoch": 1.5690376569037658,
-      "grad_norm": 0.9644107222557068,
-      "learning_rate": 0.0001582857142857143,
-      "loss": 0.5795,
+      "epoch": 1.1614401858304297,
+      "grad_norm": 0.41259104013442993,
+      "learning_rate": 0.00015799999999999996,
+      "loss": 0.5511,
       "step": 1500
     },
     {
-      "epoch": 1.5690376569037658,
-      "eval_loss": 0.5057937502861023,
-      "eval_runtime": 153.7283,
-      "eval_samples_per_second": 36.792,
-      "eval_steps_per_second": 4.599,
-      "eval_wer": 0.3952913610758935,
+      "epoch": 1.1614401858304297,
+      "eval_loss": 0.5231888890266418,
+      "eval_runtime": 173.7205,
+      "eval_samples_per_second": 32.558,
+      "eval_steps_per_second": 4.07,
+      "eval_wer": 0.40788945772014573,
       "step": 1500
     },
     {
-      "epoch": 1.6736401673640167,
-      "eval_loss": 0.48043006658554077,
-      "eval_runtime": 153.7864,
-      "eval_samples_per_second": 36.778,
-      "eval_steps_per_second": 4.597,
-      "eval_wer": 0.37892185970374415,
+      "epoch": 1.2388695315524583,
+      "eval_loss": 0.518454909324646,
+      "eval_runtime": 172.0395,
+      "eval_samples_per_second": 32.876,
+      "eval_steps_per_second": 4.11,
+      "eval_wer": 0.4029465102469869,
       "step": 1600
     },
     {
-      "epoch": 1.778242677824268,
-      "eval_loss": 0.46723026037216187,
-      "eval_runtime": 153.9064,
-      "eval_samples_per_second": 36.75,
-      "eval_steps_per_second": 4.594,
-      "eval_wer": 0.3698062942337629,
+      "epoch": 1.316298877274487,
+      "eval_loss": 0.5089535713195801,
+      "eval_runtime": 171.996,
+      "eval_samples_per_second": 32.884,
+      "eval_steps_per_second": 4.111,
+      "eval_wer": 0.4042303927075476,
       "step": 1700
     },
     {
-      "epoch": 1.8828451882845187,
-      "eval_loss": 0.46052873134613037,
-      "eval_runtime": 153.9609,
-      "eval_samples_per_second": 36.737,
-      "eval_steps_per_second": 4.592,
-      "eval_wer": 0.3712185649403797,
+      "epoch": 1.3937282229965158,
+      "eval_loss": 0.47846707701683044,
+      "eval_runtime": 173.133,
+      "eval_samples_per_second": 32.669,
+      "eval_steps_per_second": 4.084,
+      "eval_wer": 0.38505239845292166,
       "step": 1800
     },
     {
-      "epoch": 1.98744769874477,
-      "eval_loss": 0.4490562677383423,
-      "eval_runtime": 153.6441,
-      "eval_samples_per_second": 36.812,
-      "eval_steps_per_second": 4.602,
-      "eval_wer": 0.35561939304456675,
+      "epoch": 1.4711575687185443,
+      "eval_loss": 0.47747060656547546,
+      "eval_runtime": 172.0577,
+      "eval_samples_per_second": 32.873,
+      "eval_steps_per_second": 4.109,
+      "eval_wer": 0.3802699362873329,
       "step": 1900
     },
     {
-      "epoch": 2.092050209205021,
-      "grad_norm": 1.1409815549850464,
-      "learning_rate": 8.714285714285714e-05,
-      "loss": 0.5057,
+      "epoch": 1.5485869144405728,
+      "grad_norm": 0.49539849162101746,
+      "learning_rate": 8.685714285714285e-05,
+      "loss": 0.4529,
       "step": 2000
     },
     {
-      "epoch": 2.092050209205021,
-      "eval_loss": 0.44942042231559753,
-      "eval_runtime": 154.6428,
-      "eval_samples_per_second": 36.575,
-      "eval_steps_per_second": 4.572,
-      "eval_wer": 0.34534833336008086,
+      "epoch": 1.5485869144405728,
+      "eval_loss": 0.46770602464675903,
+      "eval_runtime": 172.1877,
+      "eval_samples_per_second": 32.848,
+      "eval_steps_per_second": 4.106,
+      "eval_wer": 0.37218147678580027,
       "step": 2000
     },
     {
-      "epoch": 2.196652719665272,
-      "eval_loss": 0.4431760907173157,
-      "eval_runtime": 154.0218,
-      "eval_samples_per_second": 36.722,
-      "eval_steps_per_second": 4.59,
-      "eval_wer": 0.3396832020028566,
+      "epoch": 1.6260162601626016,
+      "eval_loss": 0.4573723077774048,
+      "eval_runtime": 173.1529,
+      "eval_samples_per_second": 32.665,
+      "eval_steps_per_second": 4.083,
+      "eval_wer": 0.3543997047070341,
       "step": 2100
     },
     {
-      "epoch": 2.301255230125523,
-      "eval_loss": 0.43776625394821167,
-      "eval_runtime": 154.8082,
-      "eval_samples_per_second": 36.536,
-      "eval_steps_per_second": 4.567,
-      "eval_wer": 0.335061225144838,
+      "epoch": 1.7034456058846303,
+      "eval_loss": 0.4472625255584717,
+      "eval_runtime": 173.3523,
+      "eval_samples_per_second": 32.627,
+      "eval_steps_per_second": 4.078,
+      "eval_wer": 0.3561810916210621,
       "step": 2200
     },
     {
-      "epoch": 2.405857740585774,
-      "eval_loss": 0.42906874418258667,
-      "eval_runtime": 154.6416,
-      "eval_samples_per_second": 36.575,
-      "eval_steps_per_second": 4.572,
-      "eval_wer": 0.33104909245558567,
+      "epoch": 1.7808749516066589,
+      "eval_loss": 0.4436591863632202,
+      "eval_runtime": 173.0139,
+      "eval_samples_per_second": 32.691,
+      "eval_steps_per_second": 4.086,
+      "eval_wer": 0.34703342908956686,
       "step": 2300
     },
     {
-      "epoch": 2.510460251046025,
-      "eval_loss": 0.4279017746448517,
-      "eval_runtime": 154.5726,
-      "eval_samples_per_second": 36.591,
-      "eval_steps_per_second": 4.574,
-      "eval_wer": 0.3294281908491278,
+      "epoch": 1.8583042973286876,
+      "eval_loss": 0.43528568744659424,
+      "eval_runtime": 173.1967,
+      "eval_samples_per_second": 32.657,
+      "eval_steps_per_second": 4.082,
+      "eval_wer": 0.3450113142141837,
       "step": 2400
     },
     {
-      "epoch": 2.6150627615062763,
-      "grad_norm": 0.6568267345428467,
-      "learning_rate": 1.6e-05,
-      "loss": 0.3986,
+      "epoch": 1.9357336430507162,
+      "grad_norm": 0.6141678094863892,
+      "learning_rate": 1.557142857142857e-05,
+      "loss": 0.4149,
       "step": 2500
     },
     {
-      "epoch": 2.6150627615062763,
-      "eval_loss": 0.4233950078487396,
-      "eval_runtime": 154.8004,
-      "eval_samples_per_second": 36.537,
-      "eval_steps_per_second": 4.567,
-      "eval_wer": 0.3259456596748568,
+      "epoch": 1.9357336430507162,
+      "eval_loss": 0.4299843907356262,
+      "eval_runtime": 172.6246,
+      "eval_samples_per_second": 32.765,
+      "eval_steps_per_second": 4.096,
+      "eval_wer": 0.34008441527178185,
       "step": 2500
     },
     {
-      "epoch": 2.719665271966527,
-      "eval_loss": 0.41978269815444946,
-      "eval_runtime": 153.7416,
-      "eval_samples_per_second": 36.789,
-      "eval_steps_per_second": 4.599,
-      "eval_wer": 0.32485435958338016,
+      "epoch": 2.013162988772745,
+      "eval_loss": 0.4289664328098297,
+      "eval_runtime": 173.2622,
+      "eval_samples_per_second": 32.644,
+      "eval_steps_per_second": 4.081,
+      "eval_wer": 0.33783762096580056,
       "step": 2600
     },
     {
-      "epoch": 2.719665271966527,
+      "epoch": 2.013162988772745,
       "step": 2600,
-      "total_flos": 1.0779565758702488e+19,
-      "train_loss": 1.1314451643136831,
-      "train_runtime": 7270.6329,
-      "train_samples_per_second": 11.443,
-      "train_steps_per_second": 0.358
+      "total_flos": 1.1633941226063049e+19,
+      "train_loss": 1.1463891924344576,
+      "train_runtime": 8433.2953,
+      "train_samples_per_second": 9.866,
+      "train_steps_per_second": 0.308
     }
   ],
   "logging_steps": 500,
@@ -292,7 +292,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 400,
-  "total_flos": 1.0779565758702488e+19,
+  "total_flos": 1.1633941226063049e+19,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null