End of training

README.md

@@ -3,9 +3,24 @@ license: apache-2.0
 base_model: openai/whisper-large-v3
 tags:
 - generated_from_trainer
+datasets:
+- BembaSpeech
+metrics:
+- wer
 model-index:
 - name: whisper-large-v3-bem-fsv
-  results: []
+  results:
+  - task:
+      name: Automatic Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: BembaSpeech bem
+      type: BembaSpeech
+      args: bem
+    metrics:
+    - name: Wer
+      type: wer
+      value: 0.4033761652809272
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -13,7 +28,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # whisper-large-v3-bem-fsv
 
-This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on an unknown dataset.
+This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the BembaSpeech bem dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.4783
+- Wer: 0.4034
 
 ## Model description
 

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 1.994236311239193,
+    "eval_loss": 0.47826600074768066,
+    "eval_runtime": 323.8582,
+    "eval_samples": 499,
+    "eval_samples_per_second": 1.541,
+    "eval_steps_per_second": 0.386,
+    "eval_wer": 0.4033761652809272,
+    "total_flos": 2.821282593767424e+19,
+    "train_loss": 0.8678894690695526,
+    "train_runtime": 2803.3653,
+    "train_samples": 4164,
+    "train_samples_per_second": 2.971,
+    "train_steps_per_second": 0.123
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 1.994236311239193,
+    "eval_loss": 0.47826600074768066,
+    "eval_runtime": 323.8582,
+    "eval_samples": 499,
+    "eval_samples_per_second": 1.541,
+    "eval_steps_per_second": 0.386,
+    "eval_wer": 0.4033761652809272
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 1.994236311239193,
+    "total_flos": 2.821282593767424e+19,
+    "train_loss": 0.8678894690695526,
+    "train_runtime": 2803.3653,
+    "train_samples": 4164,
+    "train_samples_per_second": 2.971,
+    "train_steps_per_second": 0.123
+}

trainer_state.json

@@ -0,0 +1,133 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.994236311239193,
+  "eval_steps": 500,
+  "global_step": 346,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.1440922190201729,
+      "grad_norm": 16.46904754638672,
+      "learning_rate": 7.699999999999999e-07,
+      "loss": 3.4205,
+      "step": 25
+    },
+    {
+      "epoch": 0.2881844380403458,
+      "grad_norm": 74.49513244628906,
+      "learning_rate": 1.6099999999999998e-06,
+      "loss": 2.0618,
+      "step": 50
+    },
+    {
+      "epoch": 0.4322766570605187,
+      "grad_norm": 6.616641044616699,
+      "learning_rate": 2.4849999999999995e-06,
+      "loss": 1.4862,
+      "step": 75
+    },
+    {
+      "epoch": 0.5763688760806917,
+      "grad_norm": 4.94362211227417,
+      "learning_rate": 3.3599999999999996e-06,
+      "loss": 0.8913,
+      "step": 100
+    },
+    {
+      "epoch": 0.7204610951008645,
+      "grad_norm": 4.745422840118408,
+      "learning_rate": 4.234999999999999e-06,
+      "loss": 0.6597,
+      "step": 125
+    },
+    {
+      "epoch": 0.8645533141210374,
+      "grad_norm": 6.370882511138916,
+      "learning_rate": 5.109999999999999e-06,
+      "loss": 0.539,
+      "step": 150
+    },
+    {
+      "epoch": 1.0086455331412103,
+      "grad_norm": 3.6810574531555176,
+      "learning_rate": 5.985e-06,
+      "loss": 0.4565,
+      "step": 175
+    },
+    {
+      "epoch": 1.1527377521613833,
+      "grad_norm": 4.011445999145508,
+      "learning_rate": 6.8599999999999995e-06,
+      "loss": 0.3858,
+      "step": 200
+    },
+    {
+      "epoch": 1.2968299711815563,
+      "grad_norm": 3.8143043518066406,
+      "learning_rate": 7.735e-06,
+      "loss": 0.3774,
+      "step": 225
+    },
+    {
+      "epoch": 1.440922190201729,
+      "grad_norm": 3.4702436923980713,
+      "learning_rate": 8.609999999999999e-06,
+      "loss": 0.3695,
+      "step": 250
+    },
+    {
+      "epoch": 1.585014409221902,
+      "grad_norm": 3.361118793487549,
+      "learning_rate": 9.485e-06,
+      "loss": 0.364,
+      "step": 275
+    },
+    {
+      "epoch": 1.729106628242075,
+      "grad_norm": 2.591237783432007,
+      "learning_rate": 1.0359999999999999e-05,
+      "loss": 0.3584,
+      "step": 300
+    },
+    {
+      "epoch": 1.8731988472622478,
+      "grad_norm": 2.9504644870758057,
+      "learning_rate": 1.1235e-05,
+      "loss": 0.3474,
+      "step": 325
+    },
+    {
+      "epoch": 1.994236311239193,
+      "step": 346,
+      "total_flos": 2.821282593767424e+19,
+      "train_loss": 0.8678894690695526,
+      "train_runtime": 2803.3653,
+      "train_samples_per_second": 2.971,
+      "train_steps_per_second": 0.123
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 346,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.821282593767424e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}