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: inf
 - Wer: 1.0

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 3.0864197530864197,
+    "epoch": 6.172839506172839,
     "eval_loss": Infinity,
-    "eval_runtime": 194.9808,
+    "eval_runtime": 176.2004,
     "eval_samples": 7022,
-    "eval_samples_per_second": 36.014,
-    "eval_steps_per_second": 4.503,
-    "eval_wer": 0.36645761105267943,
-    "total_flos": 1.6964840215738495e+19,
-    "train_loss": 1.4283542098999022,
-    "train_runtime": 6370.4271,
+    "eval_samples_per_second": 39.852,
+    "eval_steps_per_second": 4.983,
+    "eval_wer": 1.0,
+    "total_flos": 3.3889468875697558e+19,
+    "train_loss": 2.1502342529296876,
+    "train_runtime": 11441.7295,
     "train_samples": 41435,
-    "train_samples_per_second": 20.093,
-    "train_steps_per_second": 0.314
+    "train_samples_per_second": 22.374,
+    "train_steps_per_second": 0.35
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 3.0864197530864197,
+    "epoch": 6.172839506172839,
     "eval_loss": Infinity,
-    "eval_runtime": 194.9808,
+    "eval_runtime": 176.2004,
     "eval_samples": 7022,
-    "eval_samples_per_second": 36.014,
-    "eval_steps_per_second": 4.503,
-    "eval_wer": 0.36645761105267943
+    "eval_samples_per_second": 39.852,
+    "eval_steps_per_second": 4.983,
+    "eval_wer": 1.0
 }

runs/Aug10_18-57-49_940cfb8ec62d/events.out.tfevents.1723324349.940cfb8ec62d.724.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:535d1591f475f1d8fcd45d568326c471d44a7e926b091aa3a4d917a666fdcaef
+size 406

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 3.0864197530864197,
-    "total_flos": 1.6964840215738495e+19,
-    "train_loss": 1.4283542098999022,
-    "train_runtime": 6370.4271,
+    "epoch": 6.172839506172839,
+    "total_flos": 3.3889468875697558e+19,
+    "train_loss": 2.1502342529296876,
+    "train_runtime": 11441.7295,
     "train_samples": 41435,
-    "train_samples_per_second": 20.093,
-    "train_steps_per_second": 0.314
+    "train_samples_per_second": 22.374,
+    "train_steps_per_second": 0.35
 }

trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.0864197530864197,
+  "epoch": 6.172839506172839,
   "eval_steps": 200,
-  "global_step": 2000,
+  "global_step": 4000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -11,135 +11,253 @@
     {
       "epoch": 0.30864197530864196,
       "eval_loss": Infinity,
-      "eval_runtime": 198.0112,
-      "eval_samples_per_second": 35.463,
-      "eval_steps_per_second": 4.434,
+      "eval_runtime": 176.8693,
+      "eval_samples_per_second": 39.702,
+      "eval_steps_per_second": 4.964,
       "eval_wer": 1.0,
       "step": 200
     },
     {
       "epoch": 0.6172839506172839,
       "eval_loss": Infinity,
-      "eval_runtime": 196.4137,
-      "eval_samples_per_second": 35.751,
-      "eval_steps_per_second": 4.47,
-      "eval_wer": 0.9468401438718138,
+      "eval_runtime": 175.9503,
+      "eval_samples_per_second": 39.909,
+      "eval_steps_per_second": 4.99,
+      "eval_wer": 0.8138755794476257,
       "step": 400
     },
     {
       "epoch": 0.7716049382716049,
-      "grad_norm": 2.496718645095825,
+      "grad_norm": 2.980449914932251,
       "learning_rate": 0.0002958,
-      "loss": 3.8579,
+      "loss": 3.6854,
       "step": 500
     },
     {
       "epoch": 0.9259259259259259,
       "eval_loss": Infinity,
-      "eval_runtime": 194.7508,
-      "eval_samples_per_second": 36.056,
-      "eval_steps_per_second": 4.508,
-      "eval_wer": 0.682339345305338,
+      "eval_runtime": 174.7168,
+      "eval_samples_per_second": 40.191,
+      "eval_steps_per_second": 5.025,
+      "eval_wer": 0.6042746029891057,
       "step": 600
     },
     {
       "epoch": 1.2345679012345678,
       "eval_loss": Infinity,
-      "eval_runtime": 195.8415,
-      "eval_samples_per_second": 35.856,
-      "eval_steps_per_second": 4.483,
-      "eval_wer": 0.5246257125420384,
+      "eval_runtime": 176.3896,
+      "eval_samples_per_second": 39.81,
+      "eval_steps_per_second": 4.978,
+      "eval_wer": 0.503447469907678,
       "step": 800
     },
     {
       "epoch": 1.5432098765432098,
-      "grad_norm": 0.836155354976654,
-      "learning_rate": 0.0002016,
-      "loss": 0.7662,
+      "grad_norm": 0.7878520488739014,
+      "learning_rate": 0.0002578285714285714,
+      "loss": 0.7236,
       "step": 1000
     },
     {
       "epoch": 1.5432098765432098,
       "eval_loss": Infinity,
-      "eval_runtime": 194.2625,
-      "eval_samples_per_second": 36.147,
-      "eval_steps_per_second": 4.52,
-      "eval_wer": 0.45603988936932727,
+      "eval_runtime": 175.5785,
+      "eval_samples_per_second": 39.994,
+      "eval_steps_per_second": 5.001,
+      "eval_wer": 0.44518457922688376,
       "step": 1000
     },
     {
       "epoch": 1.8518518518518519,
       "eval_loss": Infinity,
-      "eval_runtime": 195.3379,
-      "eval_samples_per_second": 35.948,
-      "eval_steps_per_second": 4.495,
-      "eval_wer": 0.4314076844169166,
+      "eval_runtime": 175.3263,
+      "eval_samples_per_second": 40.051,
+      "eval_steps_per_second": 5.008,
+      "eval_wer": 0.4584940204900471,
       "step": 1200
     },
     {
       "epoch": 2.1604938271604937,
       "eval_loss": Infinity,
-      "eval_runtime": 194.6062,
-      "eval_samples_per_second": 36.083,
-      "eval_steps_per_second": 4.512,
-      "eval_wer": 0.43706906626154024,
+      "eval_runtime": 174.0761,
+      "eval_samples_per_second": 40.339,
+      "eval_steps_per_second": 5.044,
+      "eval_wer": 0.4735564125537247,
       "step": 1400
     },
     {
       "epoch": 2.314814814814815,
-      "grad_norm": 0.4859907329082489,
-      "learning_rate": 0.00010239999999999998,
-      "loss": 0.5916,
+      "grad_norm": 1.6092010736465454,
+      "learning_rate": 0.0002153142857142857,
+      "loss": 0.6244,
       "step": 1500
     },
     {
       "epoch": 2.4691358024691357,
       "eval_loss": Infinity,
-      "eval_runtime": 194.4662,
-      "eval_samples_per_second": 36.109,
-      "eval_steps_per_second": 4.515,
-      "eval_wer": 0.38610364483918297,
+      "eval_runtime": 174.8676,
+      "eval_samples_per_second": 40.156,
+      "eval_steps_per_second": 5.021,
+      "eval_wer": 0.4445742926518899,
       "step": 1600
     },
     {
       "epoch": 2.7777777777777777,
       "eval_loss": Infinity,
-      "eval_runtime": 194.3135,
-      "eval_samples_per_second": 36.137,
-      "eval_steps_per_second": 4.518,
-      "eval_wer": 0.36491241738407804,
+      "eval_runtime": 174.4183,
+      "eval_samples_per_second": 40.26,
+      "eval_steps_per_second": 5.034,
+      "eval_wer": 0.4992274031656993,
       "step": 1800
     },
     {
       "epoch": 3.0864197530864197,
-      "grad_norm": 0.7149534821510315,
-      "learning_rate": 2.9999999999999997e-06,
-      "loss": 0.4977,
+      "grad_norm": 13.045821189880371,
+      "learning_rate": 0.0001727142857142857,
+      "loss": 0.8045,
       "step": 2000
     },
     {
       "epoch": 3.0864197530864197,
       "eval_loss": Infinity,
-      "eval_runtime": 195.0164,
-      "eval_samples_per_second": 36.007,
-      "eval_steps_per_second": 4.502,
-      "eval_wer": 0.36645761105267943,
+      "eval_runtime": 176.0054,
+      "eval_samples_per_second": 39.897,
+      "eval_steps_per_second": 4.988,
+      "eval_wer": 0.699323490839209,
       "step": 2000
     },
     {
-      "epoch": 3.0864197530864197,
-      "step": 2000,
-      "total_flos": 1.6964840215738495e+19,
-      "train_loss": 1.4283542098999022,
-      "train_runtime": 6370.4271,
-      "train_samples_per_second": 20.093,
-      "train_steps_per_second": 0.314
+      "epoch": 3.3950617283950617,
+      "eval_loss": Infinity,
+      "eval_runtime": 175.9958,
+      "eval_samples_per_second": 39.899,
+      "eval_steps_per_second": 4.989,
+      "eval_wer": 0.9154558321322375,
+      "step": 2200
+    },
+    {
+      "epoch": 3.7037037037037037,
+      "eval_loss": Infinity,
+      "eval_runtime": 175.4175,
+      "eval_samples_per_second": 40.03,
+      "eval_steps_per_second": 5.005,
+      "eval_wer": 1.0,
+      "step": 2400
+    },
+    {
+      "epoch": 3.8580246913580245,
+      "grad_norm": 1.601144552230835,
+      "learning_rate": 0.0001302,
+      "loss": 2.3067,
+      "step": 2500
+    },
+    {
+      "epoch": 4.012345679012346,
+      "eval_loss": Infinity,
+      "eval_runtime": 175.9931,
+      "eval_samples_per_second": 39.899,
+      "eval_steps_per_second": 4.989,
+      "eval_wer": 0.998558684897355,
+      "step": 2600
+    },
+    {
+      "epoch": 4.320987654320987,
+      "eval_loss": Infinity,
+      "eval_runtime": 175.0274,
+      "eval_samples_per_second": 40.119,
+      "eval_steps_per_second": 5.016,
+      "eval_wer": 1.0,
+      "step": 2800
+    },
+    {
+      "epoch": 4.62962962962963,
+      "grad_norm": 0.7471032738685608,
+      "learning_rate": 8.759999999999999e-05,
+      "loss": 3.008,
+      "step": 3000
+    },
+    {
+      "epoch": 4.62962962962963,
+      "eval_loss": Infinity,
+      "eval_runtime": 175.5031,
+      "eval_samples_per_second": 40.011,
+      "eval_steps_per_second": 5.003,
+      "eval_wer": 1.0,
+      "step": 3000
+    },
+    {
+      "epoch": 4.938271604938271,
+      "eval_loss": Infinity,
+      "eval_runtime": 175.2017,
+      "eval_samples_per_second": 40.08,
+      "eval_steps_per_second": 5.011,
+      "eval_wer": 1.0,
+      "step": 3200
+    },
+    {
+      "epoch": 5.246913580246914,
+      "eval_loss": Infinity,
+      "eval_runtime": 176.0971,
+      "eval_samples_per_second": 39.876,
+      "eval_steps_per_second": 4.986,
+      "eval_wer": 1.0,
+      "step": 3400
+    },
+    {
+      "epoch": 5.401234567901234,
+      "grad_norm": 1.3811966180801392,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 3.032,
+      "step": 3500
+    },
+    {
+      "epoch": 5.555555555555555,
+      "eval_loss": Infinity,
+      "eval_runtime": 176.4078,
+      "eval_samples_per_second": 39.805,
+      "eval_steps_per_second": 4.977,
+      "eval_wer": 1.0,
+      "step": 3600
+    },
+    {
+      "epoch": 5.864197530864198,
+      "eval_loss": Infinity,
+      "eval_runtime": 175.4996,
+      "eval_samples_per_second": 40.011,
+      "eval_steps_per_second": 5.003,
+      "eval_wer": 1.0,
+      "step": 3800
+    },
+    {
+      "epoch": 6.172839506172839,
+      "grad_norm": 0.011152578517794609,
+      "learning_rate": 2.314285714285714e-06,
+      "loss": 3.0173,
+      "step": 4000
+    },
+    {
+      "epoch": 6.172839506172839,
+      "eval_loss": Infinity,
+      "eval_runtime": 175.8629,
+      "eval_samples_per_second": 39.929,
+      "eval_steps_per_second": 4.993,
+      "eval_wer": 1.0,
+      "step": 4000
+    },
+    {
+      "epoch": 6.172839506172839,
+      "step": 4000,
+      "total_flos": 3.3889468875697558e+19,
+      "train_loss": 2.1502342529296876,
+      "train_runtime": 11441.7295,
+      "train_samples_per_second": 22.374,
+      "train_steps_per_second": 0.35
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2000,
+  "max_steps": 4000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 7,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -153,7 +271,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.6964840215738495e+19,
+  "total_flos": 3.3889468875697558e+19,
   "train_batch_size": 64,
   "trial_name": null,
   "trial_params": null