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
 - generated_from_trainer
 metrics:
 - wer
@@ -15,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # wav2vec2-xlsr-53-ft-btb-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 - DEFAULT dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.5418
 - Wer: 0.3991

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 2.847380410022779,
-    "eval_loss": 0.8192350268363953,
-    "eval_runtime": 179.8741,
+    "epoch": 2.277904328018223,
+    "eval_loss": 0.5418316721916199,
+    "eval_runtime": 182.0828,
     "eval_samples": 7022,
-    "eval_samples_per_second": 39.038,
-    "eval_steps_per_second": 4.881,
-    "eval_wer": 0.47557102804980134,
-    "total_flos": 9.363138171711945e+18,
-    "train_loss": 1.33308916015625,
-    "train_runtime": 4617.7601,
+    "eval_samples_per_second": 38.565,
+    "eval_steps_per_second": 4.822,
+    "eval_wer": 0.3990751936645451,
+    "total_flos": 7.50335828249239e+18,
+    "train_loss": 1.2249342193603516,
+    "train_runtime": 3791.7318,
     "train_samples": 28086,
-    "train_samples_per_second": 17.324,
-    "train_steps_per_second": 0.541
+    "train_samples_per_second": 16.879,
+    "train_steps_per_second": 0.527
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.847380410022779,
-    "eval_loss": 0.8192350268363953,
-    "eval_runtime": 179.8741,
+    "epoch": 2.277904328018223,
+    "eval_loss": 0.5418316721916199,
+    "eval_runtime": 182.0828,
     "eval_samples": 7022,
-    "eval_samples_per_second": 39.038,
-    "eval_steps_per_second": 4.881,
-    "eval_wer": 0.47557102804980134
+    "eval_samples_per_second": 38.565,
+    "eval_steps_per_second": 4.822,
+    "eval_wer": 0.3990751936645451
 }

runs/Aug15_17-24-01_a142c9bbb9ba/events.out.tfevents.1723743599.a142c9bbb9ba.1418.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:086f60e788d852c3de43ad710636191662ca907c700b7cd98f184b5d3d3b0bad
+size 406

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.847380410022779,
-    "total_flos": 9.363138171711945e+18,
-    "train_loss": 1.33308916015625,
-    "train_runtime": 4617.7601,
+    "epoch": 2.277904328018223,
+    "total_flos": 7.50335828249239e+18,
+    "train_loss": 1.2249342193603516,
+    "train_runtime": 3791.7318,
     "train_samples": 28086,
-    "train_samples_per_second": 17.324,
-    "train_steps_per_second": 0.541
+    "train_samples_per_second": 16.879,
+    "train_steps_per_second": 0.527
 }

trainer_state.json

@@ -1,168 +1,143 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.847380410022779,
+  "epoch": 2.277904328018223,
   "eval_steps": 200,
-  "global_step": 2500,
+  "global_step": 2000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.22779043280182232,
-      "eval_loss": 2.9656879901885986,
-      "eval_runtime": 180.6727,
-      "eval_samples_per_second": 38.866,
-      "eval_steps_per_second": 4.86,
+      "eval_loss": 2.7455391883850098,
+      "eval_runtime": 186.8417,
+      "eval_samples_per_second": 37.583,
+      "eval_steps_per_second": 4.699,
       "eval_wer": 1.0,
       "step": 200
     },
     {
       "epoch": 0.45558086560364464,
-      "eval_loss": 1.256484031677246,
-      "eval_runtime": 177.1703,
-      "eval_samples_per_second": 39.634,
-      "eval_steps_per_second": 4.956,
-      "eval_wer": 0.8238084481590242,
+      "eval_loss": 1.1437467336654663,
+      "eval_runtime": 180.9385,
+      "eval_samples_per_second": 38.809,
+      "eval_steps_per_second": 4.852,
+      "eval_wer": 0.7405758779680836,
       "step": 400
     },
     {
       "epoch": 0.5694760820045558,
-      "grad_norm": 2.394413948059082,
-      "learning_rate": 0.00029699999999999996,
-      "loss": 3.3708,
+      "grad_norm": 1.8708776235580444,
+      "learning_rate": 0.0002508333333333333,
+      "loss": 2.6988,
       "step": 500
     },
     {
       "epoch": 0.683371298405467,
-      "eval_loss": 0.8798568248748779,
-      "eval_runtime": 177.6107,
-      "eval_samples_per_second": 39.536,
-      "eval_steps_per_second": 4.943,
-      "eval_wer": 0.6336517891547855,
+      "eval_loss": 0.7285588979721069,
+      "eval_runtime": 182.6087,
+      "eval_samples_per_second": 38.454,
+      "eval_steps_per_second": 4.808,
+      "eval_wer": 0.5377164192986885,
       "step": 600
     },
     {
       "epoch": 0.9111617312072893,
-      "eval_loss": 0.7572081685066223,
-      "eval_runtime": 179.3544,
-      "eval_samples_per_second": 39.152,
-      "eval_steps_per_second": 4.895,
-      "eval_wer": 0.5467651709430101,
+      "eval_loss": 0.6604043245315552,
+      "eval_runtime": 181.72,
+      "eval_samples_per_second": 38.642,
+      "eval_steps_per_second": 4.832,
+      "eval_wer": 0.4896557222392006,
       "step": 800
     },
     {
       "epoch": 1.1389521640091116,
-      "grad_norm": 0.749163806438446,
-      "learning_rate": 0.00022664999999999995,
-      "loss": 0.9128,
+      "grad_norm": 0.762070894241333,
+      "learning_rate": 0.00016849999999999998,
+      "loss": 0.8304,
       "step": 1000
     },
     {
       "epoch": 1.1389521640091116,
-      "eval_loss": 0.6133933663368225,
-      "eval_runtime": 178.9414,
-      "eval_samples_per_second": 39.242,
-      "eval_steps_per_second": 4.907,
-      "eval_wer": 0.4571280511300974,
+      "eval_loss": 0.5619407892227173,
+      "eval_runtime": 181.9456,
+      "eval_samples_per_second": 38.594,
+      "eval_steps_per_second": 4.826,
+      "eval_wer": 0.4338617308228916,
       "step": 1000
     },
     {
       "epoch": 1.366742596810934,
-      "eval_loss": 0.5878995060920715,
-      "eval_runtime": 179.0164,
-      "eval_samples_per_second": 39.225,
-      "eval_steps_per_second": 4.905,
-      "eval_wer": 0.44289719501986474,
+      "eval_loss": 0.5691641569137573,
+      "eval_runtime": 182.4948,
+      "eval_samples_per_second": 38.478,
+      "eval_steps_per_second": 4.811,
+      "eval_wer": 0.4244675055475093,
       "step": 1200
     },
     {
       "epoch": 1.5945330296127562,
-      "eval_loss": 0.6282891631126404,
-      "eval_runtime": 178.8026,
-      "eval_samples_per_second": 39.272,
-      "eval_steps_per_second": 4.91,
-      "eval_wer": 0.46545928061760056,
+      "eval_loss": 0.563896656036377,
+      "eval_runtime": 182.6815,
+      "eval_samples_per_second": 38.438,
+      "eval_steps_per_second": 4.806,
+      "eval_wer": 0.4210260566842504,
       "step": 1400
     },
     {
       "epoch": 1.7084282460136673,
-      "grad_norm": 2.0273807048797607,
-      "learning_rate": 0.00015209999999999998,
-      "loss": 0.7779,
+      "grad_norm": 4.044058322906494,
+      "learning_rate": 8.566666666666666e-05,
+      "loss": 0.71,
       "step": 1500
     },
     {
       "epoch": 1.8223234624145785,
-      "eval_loss": 0.6165733337402344,
-      "eval_runtime": 178.9936,
-      "eval_samples_per_second": 39.23,
-      "eval_steps_per_second": 4.905,
-      "eval_wer": 0.4566364155782033,
+      "eval_loss": 0.5502046346664429,
+      "eval_runtime": 183.2379,
+      "eval_samples_per_second": 38.322,
+      "eval_steps_per_second": 4.792,
+      "eval_wer": 0.4186476036088707,
       "step": 1600
     },
     {
       "epoch": 2.050113895216401,
-      "eval_loss": 0.6130892634391785,
-      "eval_runtime": 180.5144,
-      "eval_samples_per_second": 38.9,
-      "eval_steps_per_second": 4.864,
-      "eval_wer": 0.46482148314487304,
+      "eval_loss": 0.5441319942474365,
+      "eval_runtime": 182.7119,
+      "eval_samples_per_second": 38.432,
+      "eval_steps_per_second": 4.805,
+      "eval_wer": 0.4073266984679573,
       "step": 1800
     },
     {
       "epoch": 2.277904328018223,
-      "grad_norm": 2.559767246246338,
-      "learning_rate": 7.725e-05,
-      "loss": 0.7465,
+      "grad_norm": 9.03835678100586,
+      "learning_rate": 2.4999999999999998e-06,
+      "loss": 0.6605,
       "step": 2000
     },
     {
       "epoch": 2.277904328018223,
-      "eval_loss": 0.6311540603637695,
-      "eval_runtime": 180.4386,
-      "eval_samples_per_second": 38.916,
-      "eval_steps_per_second": 4.866,
-      "eval_wer": 0.4721694415285879,
+      "eval_loss": 0.5418316721916199,
+      "eval_runtime": 182.4712,
+      "eval_samples_per_second": 38.483,
+      "eval_steps_per_second": 4.812,
+      "eval_wer": 0.3990751936645451,
       "step": 2000
     },
     {
-      "epoch": 2.5056947608200457,
-      "eval_loss": 0.7629979252815247,
-      "eval_runtime": 179.3715,
-      "eval_samples_per_second": 39.148,
-      "eval_steps_per_second": 4.895,
-      "eval_wer": 0.4721694415285879,
-      "step": 2200
-    },
-    {
-      "epoch": 2.733485193621868,
-      "eval_loss": 0.8067795038223267,
-      "eval_runtime": 179.4035,
-      "eval_samples_per_second": 39.141,
-      "eval_steps_per_second": 4.894,
-      "eval_wer": 0.48269309982859193,
-      "step": 2400
-    },
-    {
-      "epoch": 2.847380410022779,
-      "grad_norm": 5.402973175048828,
-      "learning_rate": 2.9999999999999997e-06,
-      "loss": 0.8574,
-      "step": 2500
-    },
-    {
-      "epoch": 2.847380410022779,
-      "step": 2500,
-      "total_flos": 9.363138171711945e+18,
-      "train_loss": 1.33308916015625,
-      "train_runtime": 4617.7601,
-      "train_samples_per_second": 17.324,
-      "train_steps_per_second": 0.541
+      "epoch": 2.277904328018223,
+      "step": 2000,
+      "total_flos": 7.50335828249239e+18,
+      "train_loss": 1.2249342193603516,
+      "train_runtime": 3791.7318,
+      "train_samples_per_second": 16.879,
+      "train_steps_per_second": 0.527
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2500,
+  "max_steps": 2000,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,
@@ -178,7 +153,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 9.363138171711945e+18,
+  "total_flos": 7.50335828249239e+18,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null