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: 0.7008
 - Wer: 0.4816

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 2.277904328018223,
-    "eval_loss": 0.5418316721916199,
-    "eval_runtime": 182.0828,
+    "epoch": 2.733485193621868,
+    "eval_loss": 0.7008175253868103,
+    "eval_runtime": 193.3669,
     "eval_samples": 7022,
-    "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,
+    "eval_samples_per_second": 36.314,
+    "eval_steps_per_second": 0.569,
+    "eval_wer": 0.4816301040407127,
+    "total_flos": 9.105570625839254e+18,
+    "train_loss": 1.4870800399780273,
+    "train_runtime": 3391.8021,
     "train_samples": 28086,
-    "train_samples_per_second": 16.879,
-    "train_steps_per_second": 0.527
+    "train_samples_per_second": 22.643,
+    "train_steps_per_second": 0.354
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.277904328018223,
-    "eval_loss": 0.5418316721916199,
-    "eval_runtime": 182.0828,
+    "epoch": 2.733485193621868,
+    "eval_loss": 0.7008175253868103,
+    "eval_runtime": 193.3669,
     "eval_samples": 7022,
-    "eval_samples_per_second": 38.565,
-    "eval_steps_per_second": 4.822,
-    "eval_wer": 0.3990751936645451
+    "eval_samples_per_second": 36.314,
+    "eval_steps_per_second": 0.569,
+    "eval_wer": 0.4816301040407127
 }

runs/Aug16_06-47-14_a142c9bbb9ba/events.out.tfevents.1723791319.a142c9bbb9ba.1780.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04808303f26cf31bb952d42ba4a7af9f66a5828691c9d0b6e61048787861118a
+size 406

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.277904328018223,
-    "total_flos": 7.50335828249239e+18,
-    "train_loss": 1.2249342193603516,
-    "train_runtime": 3791.7318,
+    "epoch": 2.733485193621868,
+    "total_flos": 9.105570625839254e+18,
+    "train_loss": 1.4870800399780273,
+    "train_runtime": 3391.8021,
     "train_samples": 28086,
-    "train_samples_per_second": 16.879,
-    "train_steps_per_second": 0.527
+    "train_samples_per_second": 22.643,
+    "train_steps_per_second": 0.354
 }

trainer_state.json

@@ -1,143 +1,93 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.277904328018223,
+  "epoch": 2.733485193621868,
   "eval_steps": 200,
-  "global_step": 2000,
+  "global_step": 1200,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.22779043280182232,
-      "eval_loss": 2.7455391883850098,
-      "eval_runtime": 186.8417,
-      "eval_samples_per_second": 37.583,
-      "eval_steps_per_second": 4.699,
-      "eval_wer": 1.0,
+      "epoch": 0.45558086560364464,
+      "eval_loss": 2.580942392349243,
+      "eval_runtime": 191.346,
+      "eval_samples_per_second": 36.698,
+      "eval_steps_per_second": 0.575,
+      "eval_wer": 0.9998405506318181,
       "step": 200
     },
     {
-      "epoch": 0.45558086560364464,
-      "eval_loss": 1.1437467336654663,
-      "eval_runtime": 180.9385,
-      "eval_samples_per_second": 38.809,
-      "eval_steps_per_second": 4.852,
-      "eval_wer": 0.7405758779680836,
+      "epoch": 0.9111617312072893,
+      "eval_loss": 0.8878806233406067,
+      "eval_runtime": 194.7272,
+      "eval_samples_per_second": 36.061,
+      "eval_steps_per_second": 0.565,
+      "eval_wer": 0.6325356435775124,
       "step": 400
     },
     {
-      "epoch": 0.5694760820045558,
-      "grad_norm": 1.8708776235580444,
-      "learning_rate": 0.0002508333333333333,
-      "loss": 2.6988,
+      "epoch": 1.1389521640091116,
+      "grad_norm": 0.6171550750732422,
+      "learning_rate": 0.0001969444444444444,
+      "loss": 2.4327,
       "step": 500
     },
     {
-      "epoch": 0.683371298405467,
-      "eval_loss": 0.7285588979721069,
-      "eval_runtime": 182.6087,
-      "eval_samples_per_second": 38.454,
-      "eval_steps_per_second": 4.808,
-      "eval_wer": 0.5377164192986885,
+      "epoch": 1.366742596810934,
+      "eval_loss": 0.6608928442001343,
+      "eval_runtime": 195.9363,
+      "eval_samples_per_second": 35.838,
+      "eval_steps_per_second": 0.561,
+      "eval_wer": 0.4896955845812461,
       "step": 600
     },
     {
-      "epoch": 0.9111617312072893,
-      "eval_loss": 0.6604043245315552,
-      "eval_runtime": 181.72,
-      "eval_samples_per_second": 38.642,
-      "eval_steps_per_second": 4.832,
-      "eval_wer": 0.4896557222392006,
+      "epoch": 1.8223234624145785,
+      "eval_loss": 0.668916642665863,
+      "eval_runtime": 191.4141,
+      "eval_samples_per_second": 36.685,
+      "eval_steps_per_second": 0.575,
+      "eval_wer": 0.5018270240104173,
       "step": 800
     },
     {
-      "epoch": 1.1389521640091116,
-      "grad_norm": 0.762070894241333,
-      "learning_rate": 0.00016849999999999998,
-      "loss": 0.8304,
+      "epoch": 2.277904328018223,
+      "grad_norm": 5.059937000274658,
+      "learning_rate": 5.972222222222221e-05,
+      "loss": 0.8178,
       "step": 1000
     },
     {
-      "epoch": 1.1389521640091116,
-      "eval_loss": 0.5619407892227173,
-      "eval_runtime": 181.9456,
-      "eval_samples_per_second": 38.594,
-      "eval_steps_per_second": 4.826,
-      "eval_wer": 0.4338617308228916,
+      "epoch": 2.277904328018223,
+      "eval_loss": 0.6796883344650269,
+      "eval_runtime": 189.9024,
+      "eval_samples_per_second": 36.977,
+      "eval_steps_per_second": 0.579,
+      "eval_wer": 0.4852044273774565,
       "step": 1000
     },
     {
-      "epoch": 1.366742596810934,
-      "eval_loss": 0.5691641569137573,
-      "eval_runtime": 182.4948,
-      "eval_samples_per_second": 38.478,
-      "eval_steps_per_second": 4.811,
-      "eval_wer": 0.4244675055475093,
+      "epoch": 2.733485193621868,
+      "eval_loss": 0.7008175253868103,
+      "eval_runtime": 192.012,
+      "eval_samples_per_second": 36.571,
+      "eval_steps_per_second": 0.573,
+      "eval_wer": 0.4816301040407127,
       "step": 1200
     },
     {
-      "epoch": 1.5945330296127562,
-      "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": 4.044058322906494,
-      "learning_rate": 8.566666666666666e-05,
-      "loss": 0.71,
-      "step": 1500
-    },
-    {
-      "epoch": 1.8223234624145785,
-      "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": 0.5441319942474365,
-      "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": 9.03835678100586,
-      "learning_rate": 2.4999999999999998e-06,
-      "loss": 0.6605,
-      "step": 2000
-    },
-    {
-      "epoch": 2.277904328018223,
-      "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": 2.277904328018223,
-      "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
+      "epoch": 2.733485193621868,
+      "step": 1200,
+      "total_flos": 9.105570625839254e+18,
+      "train_loss": 1.4870800399780273,
+      "train_runtime": 3391.8021,
+      "train_samples_per_second": 22.643,
+      "train_steps_per_second": 0.354
     }
   ],
   "logging_steps": 500,
-  "max_steps": 2000,
+  "max_steps": 1200,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,
@@ -153,8 +103,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 7.50335828249239e+18,
-  "train_batch_size": 32,
+  "total_flos": 9.105570625839254e+18,
+  "train_batch_size": 64,
   "trial_name": null,
   "trial_params": null
 }