End of training

README.md

@@ -4,11 +4,23 @@ license: apache-2.0
 base_model: openai/whisper-small
 tags:
 - generated_from_trainer
+datasets:
+- abdouaziiz/full_wolof_normalized
 metrics:
 - wer
 model-index:
 - name: MODELING_23
-  results: []
+  results:
+  - task:
+      name: Automatic Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: abdouaziiz/full_wolof_normalized
+      type: abdouaziiz/full_wolof_normalized
+    metrics:
+    - name: Wer
+      type: wer
+      value: 0.26790202840942307
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -16,10 +28,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # MODELING_23
 
-This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on an unknown dataset.
+This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the abdouaziiz/full_wolof_normalized dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.5140
-- Wer: 0.2730
+- Loss: 0.4923
+- Wer: 0.2679
 
 ## Model description
 

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 4.356112169888375,
+    "eval_loss": 0.49226129055023193,
+    "eval_runtime": 608.3552,
+    "eval_samples": 1544,
+    "eval_samples_per_second": 2.538,
+    "eval_steps_per_second": 0.159,
+    "eval_wer": 0.26790202840942307,
+    "total_flos": 7.38663190659072e+19,
+    "train_loss": 0.6509292511940002,
+    "train_runtime": 136739.1848,
+    "train_samples": 58758,
+    "train_samples_per_second": 11.233,
+    "train_steps_per_second": 0.351
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.356112169888375,
+    "eval_loss": 0.49226129055023193,
+    "eval_runtime": 608.3552,
+    "eval_samples": 1544,
+    "eval_samples_per_second": 2.538,
+    "eval_steps_per_second": 0.159,
+    "eval_wer": 0.26790202840942307
+}

runs/Aug03_21-13-12_bambaratts/events.out.tfevents.1754393008.bambaratts.10974.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c44f79eccf5e0426b80acf59ba2c7cfa694aa1a3b5511bc5fe9f07b5a788169d
+size 406

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.356112169888375,
+    "total_flos": 7.38663190659072e+19,
+    "train_loss": 0.6509292511940002,
+    "train_runtime": 136739.1848,
+    "train_samples": 58758,
+    "train_samples_per_second": 11.233,
+    "train_steps_per_second": 0.351
+}

trainer_state.json

@@ -0,0 +1,307 @@
+{
+  "best_metric": 0.26790202840942307,
+  "best_model_checkpoint": "MODELING_23/checkpoint-6500",
+  "epoch": 4.356112169888375,
+  "eval_steps": 500,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.27225701061802343,
+      "grad_norm": 11.537489891052246,
+      "learning_rate": 2.972033368091762e-05,
+      "loss": 1.9983,
+      "step": 500
+    },
+    {
+      "epoch": 0.27225701061802343,
+      "eval_loss": 0.711998701095581,
+      "eval_runtime": 794.9584,
+      "eval_samples_per_second": 1.942,
+      "eval_steps_per_second": 0.122,
+      "eval_wer": 0.6396212076927573,
+      "step": 500
+    },
+    {
+      "epoch": 0.5445140212360469,
+      "grad_norm": 10.056964874267578,
+      "learning_rate": 2.9407507820646507e-05,
+      "loss": 1.2974,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5445140212360469,
+      "eval_loss": 0.6104683876037598,
+      "eval_runtime": 757.7831,
+      "eval_samples_per_second": 2.038,
+      "eval_steps_per_second": 0.128,
+      "eval_wer": 0.578301280177705,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8167710318540703,
+      "grad_norm": 9.043569564819336,
+      "learning_rate": 2.909468196037539e-05,
+      "loss": 1.1407,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8167710318540703,
+      "eval_loss": 0.5547229051589966,
+      "eval_runtime": 785.9187,
+      "eval_samples_per_second": 1.965,
+      "eval_steps_per_second": 0.123,
+      "eval_wer": 0.572397264277781,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0890280424720937,
+      "grad_norm": 8.64743423461914,
+      "learning_rate": 2.8781856100104274e-05,
+      "loss": 0.9514,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0890280424720937,
+      "eval_loss": 0.5316600799560547,
+      "eval_runtime": 748.1024,
+      "eval_samples_per_second": 2.064,
+      "eval_steps_per_second": 0.13,
+      "eval_wer": 0.46518968843163616,
+      "step": 2000
+    },
+    {
+      "epoch": 1.361285053090117,
+      "grad_norm": 8.403170585632324,
+      "learning_rate": 2.846903023983316e-05,
+      "loss": 0.7178,
+      "step": 2500
+    },
+    {
+      "epoch": 1.361285053090117,
+      "eval_loss": 0.5095750689506531,
+      "eval_runtime": 752.8094,
+      "eval_samples_per_second": 2.051,
+      "eval_steps_per_second": 0.129,
+      "eval_wer": 0.4634944759455194,
+      "step": 2500
+    },
+    {
+      "epoch": 1.6335420637081404,
+      "grad_norm": 8.097471237182617,
+      "learning_rate": 2.8156204379562044e-05,
+      "loss": 0.7264,
+      "step": 3000
+    },
+    {
+      "epoch": 1.6335420637081404,
+      "eval_loss": 0.49745145440101624,
+      "eval_runtime": 664.6083,
+      "eval_samples_per_second": 2.323,
+      "eval_steps_per_second": 0.146,
+      "eval_wer": 0.3532472087449582,
+      "step": 3000
+    },
+    {
+      "epoch": 1.905799074326164,
+      "grad_norm": 7.794954776763916,
+      "learning_rate": 2.784337851929093e-05,
+      "loss": 0.7091,
+      "step": 3500
+    },
+    {
+      "epoch": 1.905799074326164,
+      "eval_loss": 0.4781028628349304,
+      "eval_runtime": 631.8344,
+      "eval_samples_per_second": 2.444,
+      "eval_steps_per_second": 0.154,
+      "eval_wer": 0.31595253405038876,
+      "step": 3500
+    },
+    {
+      "epoch": 2.1780560849441875,
+      "grad_norm": 6.23008918762207,
+      "learning_rate": 2.7530552659019814e-05,
+      "loss": 0.5207,
+      "step": 4000
+    },
+    {
+      "epoch": 2.1780560849441875,
+      "eval_loss": 0.47789445519447327,
+      "eval_runtime": 658.2478,
+      "eval_samples_per_second": 2.346,
+      "eval_steps_per_second": 0.147,
+      "eval_wer": 0.3010463552931548,
+      "step": 4000
+    },
+    {
+      "epoch": 2.4503130955622106,
+      "grad_norm": 6.356353282928467,
+      "learning_rate": 2.7217726798748698e-05,
+      "loss": 0.4213,
+      "step": 4500
+    },
+    {
+      "epoch": 2.4503130955622106,
+      "eval_loss": 0.4791910946369171,
+      "eval_runtime": 671.3319,
+      "eval_samples_per_second": 2.3,
+      "eval_steps_per_second": 0.144,
+      "eval_wer": 0.41065061086105104,
+      "step": 4500
+    },
+    {
+      "epoch": 2.722570106180234,
+      "grad_norm": 7.0236334800720215,
+      "learning_rate": 2.690490093847758e-05,
+      "loss": 0.442,
+      "step": 5000
+    },
+    {
+      "epoch": 2.722570106180234,
+      "eval_loss": 0.47505632042884827,
+      "eval_runtime": 644.8379,
+      "eval_samples_per_second": 2.394,
+      "eval_steps_per_second": 0.15,
+      "eval_wer": 0.3007540772783071,
+      "step": 5000
+    },
+    {
+      "epoch": 2.9948271167982576,
+      "grad_norm": 8.722136497497559,
+      "learning_rate": 2.6592700729927007e-05,
+      "loss": 0.4337,
+      "step": 5500
+    },
+    {
+      "epoch": 2.9948271167982576,
+      "eval_loss": 0.4598409831523895,
+      "eval_runtime": 635.9124,
+      "eval_samples_per_second": 2.428,
+      "eval_steps_per_second": 0.153,
+      "eval_wer": 0.3061904483544748,
+      "step": 5500
+    },
+    {
+      "epoch": 3.2670841274162807,
+      "grad_norm": 6.036080360412598,
+      "learning_rate": 2.627987486965589e-05,
+      "loss": 0.2341,
+      "step": 6000
+    },
+    {
+      "epoch": 3.2670841274162807,
+      "eval_loss": 0.49184465408325195,
+      "eval_runtime": 640.365,
+      "eval_samples_per_second": 2.411,
+      "eval_steps_per_second": 0.151,
+      "eval_wer": 0.3112760858128252,
+      "step": 6000
+    },
+    {
+      "epoch": 3.5393411380343043,
+      "grad_norm": 5.870471000671387,
+      "learning_rate": 2.5967049009384774e-05,
+      "loss": 0.2406,
+      "step": 6500
+    },
+    {
+      "epoch": 3.5393411380343043,
+      "eval_loss": 0.49226129055023193,
+      "eval_runtime": 613.8807,
+      "eval_samples_per_second": 2.515,
+      "eval_steps_per_second": 0.158,
+      "eval_wer": 0.26790202840942307,
+      "step": 6500
+    },
+    {
+      "epoch": 3.811598148652328,
+      "grad_norm": 6.475042343139648,
+      "learning_rate": 2.565422314911366e-05,
+      "loss": 0.2479,
+      "step": 7000
+    },
+    {
+      "epoch": 3.811598148652328,
+      "eval_loss": 0.4866102933883667,
+      "eval_runtime": 623.7701,
+      "eval_samples_per_second": 2.475,
+      "eval_steps_per_second": 0.156,
+      "eval_wer": 0.2886537674636114,
+      "step": 7000
+    },
+    {
+      "epoch": 4.083855159270351,
+      "grad_norm": 4.934613227844238,
+      "learning_rate": 2.5341397288842544e-05,
+      "loss": 0.2078,
+      "step": 7500
+    },
+    {
+      "epoch": 4.083855159270351,
+      "eval_loss": 0.501932680606842,
+      "eval_runtime": 633.874,
+      "eval_samples_per_second": 2.436,
+      "eval_steps_per_second": 0.153,
+      "eval_wer": 0.29134272520021043,
+      "step": 7500
+    },
+    {
+      "epoch": 4.356112169888375,
+      "grad_norm": 4.6739068031311035,
+      "learning_rate": 2.502857142857143e-05,
+      "loss": 0.1258,
+      "step": 8000
+    },
+    {
+      "epoch": 4.356112169888375,
+      "eval_loss": 0.5139926075935364,
+      "eval_runtime": 635.4969,
+      "eval_samples_per_second": 2.43,
+      "eval_steps_per_second": 0.153,
+      "eval_wer": 0.2729876658677734,
+      "step": 8000
+    },
+    {
+      "epoch": 4.356112169888375,
+      "step": 8000,
+      "total_flos": 7.38663190659072e+19,
+      "train_loss": 0.6509292511940002,
+      "train_runtime": 136739.1848,
+      "train_samples_per_second": 11.233,
+      "train_steps_per_second": 0.351
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 48000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 27,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 3
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.38663190659072e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}