End of training

README.md

@@ -3,11 +3,23 @@ license: apache-2.0
 base_model: openai/whisper-tiny
 tags:
 - generated_from_trainer
+datasets:
+- razhan/common_voice_ckb_16
 metrics:
 - wer
 model-index:
 - name: whisper-tiny-ckb
-  results: []
+  results:
+  - task:
+      name: Automatic Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: razhan/common_voice_ckb_16
+      type: razhan/common_voice_ckb_16
+    metrics:
+    - name: Wer
+      type: wer
+      value: 0.47801004237740824
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -15,7 +27,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # whisper-tiny-ckb
 
-This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on an unknown dataset.
+This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the razhan/common_voice_ckb_16 dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.2612
 - Wer: 0.4780

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.35,
+    "eval_loss": 0.26123046875,
+    "eval_runtime": 151.8487,
+    "eval_samples": 4940,
+    "eval_samples_per_second": 32.532,
+    "eval_steps_per_second": 0.066,
+    "eval_wer": 0.47801004237740824,
+    "train_loss": 0.4515265909830729,
+    "train_runtime": 14971.1378,
+    "train_samples": 105929,
+    "train_samples_per_second": 30.779,
+    "train_steps_per_second": 0.04
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.35,
+    "eval_loss": 0.26123046875,
+    "eval_runtime": 151.8487,
+    "eval_samples": 4940,
+    "eval_samples_per_second": 32.532,
+    "eval_steps_per_second": 0.066,
+    "eval_wer": 0.47801004237740824
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.35,
+    "train_loss": 0.4515265909830729,
+    "train_runtime": 14971.1378,
+    "train_samples": 105929,
+    "train_samples_per_second": 30.779,
+    "train_steps_per_second": 0.04
+}

trainer_state.json

@@ -0,0 +1,444 @@
+{
+  "best_metric": 0.47801004237740824,
+  "best_model_checkpoint": "./whisper-tiny-ckb/checkpoint-600",
+  "epoch": 4.3478260869565215,
+  "eval_steps": 100,
+  "global_step": 600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.6164804129562553e-06,
+      "loss": 3.7615,
+      "step": 10
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.980934789368156e-06,
+      "loss": 2.8759,
+      "step": 20
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.99823228863792e-06,
+      "loss": 2.0115,
+      "step": 30
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 6.655623437887147e-06,
+      "loss": 1.4273,
+      "step": 40
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.1422479480192775e-06,
+      "loss": 1.0629,
+      "step": 50
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 7.528775214088733e-06,
+      "loss": 0.864,
+      "step": 60
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 7.849441238868767e-06,
+      "loss": 0.69,
+      "step": 70
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.123456553166724e-06,
+      "loss": 0.5759,
+      "step": 80
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.36268666504982e-06,
+      "loss": 0.5048,
+      "step": 90
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.574976672697987e-06,
+      "loss": 0.4502,
+      "step": 100
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.498779296875,
+      "eval_runtime": 154.3428,
+      "eval_samples_per_second": 32.007,
+      "eval_steps_per_second": 0.065,
+      "eval_wer": 0.7165699227235497,
+      "step": 100
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 8.765784640355151e-06,
+      "loss": 0.4391,
+      "step": 110
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.93906144297322e-06,
+      "loss": 0.4092,
+      "step": 120
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 9.09775921698996e-06,
+      "loss": 0.386,
+      "step": 130
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.244141981517345e-06,
+      "loss": 0.3711,
+      "step": 140
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.379984164319582e-06,
+      "loss": 0.3478,
+      "step": 150
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 9.50670232443118e-06,
+      "loss": 0.3501,
+      "step": 160
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.625445353695127e-06,
+      "loss": 0.3344,
+      "step": 170
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 9.73715793581418e-06,
+      "loss": 0.3167,
+      "step": 180
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 9.842626246843015e-06,
+      "loss": 0.3082,
+      "step": 190
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.942511545878664e-06,
+      "loss": 0.2977,
+      "step": 200
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 0.355712890625,
+      "eval_runtime": 154.4865,
+      "eval_samples_per_second": 31.977,
+      "eval_steps_per_second": 0.065,
+      "eval_wer": 0.5858765713471742,
+      "step": 200
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.925e-06,
+      "loss": 0.3017,
+      "step": 210
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.675000000000001e-06,
+      "loss": 0.2896,
+      "step": 220
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.425e-06,
+      "loss": 0.2844,
+      "step": 230
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 9.175000000000001e-06,
+      "loss": 0.2729,
+      "step": 240
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 8.925e-06,
+      "loss": 0.2675,
+      "step": 250
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 8.675e-06,
+      "loss": 0.2796,
+      "step": 260
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.425000000000001e-06,
+      "loss": 0.2726,
+      "step": 270
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 8.175e-06,
+      "loss": 0.2623,
+      "step": 280
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 7.925000000000001e-06,
+      "loss": 0.2456,
+      "step": 290
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 7.675e-06,
+      "loss": 0.2494,
+      "step": 300
+    },
+    {
+      "epoch": 2.17,
+      "eval_loss": 0.3095703125,
+      "eval_runtime": 165.9821,
+      "eval_samples_per_second": 29.762,
+      "eval_steps_per_second": 0.06,
+      "eval_wer": 0.5314625547523236,
+      "step": 300
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 7.425000000000001e-06,
+      "loss": 0.2531,
+      "step": 310
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 7.175000000000001e-06,
+      "loss": 0.2496,
+      "step": 320
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 6.925000000000001e-06,
+      "loss": 0.2421,
+      "step": 330
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 6.6750000000000005e-06,
+      "loss": 0.2337,
+      "step": 340
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 6.425e-06,
+      "loss": 0.2358,
+      "step": 350
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.175000000000001e-06,
+      "loss": 0.2433,
+      "step": 360
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.925000000000001e-06,
+      "loss": 0.2335,
+      "step": 370
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5.675000000000001e-06,
+      "loss": 0.2301,
+      "step": 380
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5.4250000000000006e-06,
+      "loss": 0.2223,
+      "step": 390
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.1750000000000004e-06,
+      "loss": 0.2224,
+      "step": 400
+    },
+    {
+      "epoch": 2.9,
+      "eval_loss": 0.28173828125,
+      "eval_runtime": 152.093,
+      "eval_samples_per_second": 32.48,
+      "eval_steps_per_second": 0.066,
+      "eval_wer": 0.5008012535166126,
+      "step": 400
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.925e-06,
+      "loss": 0.2293,
+      "step": 410
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 4.675000000000001e-06,
+      "loss": 0.2255,
+      "step": 420
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 4.425e-06,
+      "loss": 0.2131,
+      "step": 430
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 4.175e-06,
+      "loss": 0.2106,
+      "step": 440
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.9250000000000005e-06,
+      "loss": 0.2176,
+      "step": 450
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.6750000000000004e-06,
+      "loss": 0.22,
+      "step": 460
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 3.4250000000000007e-06,
+      "loss": 0.2151,
+      "step": 470
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 3.175e-06,
+      "loss": 0.2082,
+      "step": 480
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.925e-06,
+      "loss": 0.2014,
+      "step": 490
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 2.6750000000000002e-06,
+      "loss": 0.2148,
+      "step": 500
+    },
+    {
+      "epoch": 3.62,
+      "eval_loss": 0.2666015625,
+      "eval_runtime": 152.4154,
+      "eval_samples_per_second": 32.411,
+      "eval_steps_per_second": 0.066,
+      "eval_wer": 0.481927281791959,
+      "step": 500
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 2.425e-06,
+      "loss": 0.2158,
+      "step": 510
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 2.1750000000000004e-06,
+      "loss": 0.2103,
+      "step": 520
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.925e-06,
+      "loss": 0.2091,
+      "step": 530
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.6750000000000003e-06,
+      "loss": 0.1998,
+      "step": 540
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.425e-06,
+      "loss": 0.2068,
+      "step": 550
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1.175e-06,
+      "loss": 0.212,
+      "step": 560
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 9.25e-07,
+      "loss": 0.2034,
+      "step": 570
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 6.750000000000001e-07,
+      "loss": 0.2013,
+      "step": 580
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 4.2500000000000006e-07,
+      "loss": 0.1919,
+      "step": 590
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.7500000000000002e-07,
+      "loss": 0.2096,
+      "step": 600
+    },
+    {
+      "epoch": 4.35,
+      "eval_loss": 0.26123046875,
+      "eval_runtime": 154.28,
+      "eval_samples_per_second": 32.02,
+      "eval_steps_per_second": 0.065,
+      "eval_wer": 0.47801004237740824,
+      "step": 600
+    },
+    {
+      "epoch": 4.35,
+      "step": 600,
+      "total_flos": 1.1344381987478569e+19,
+      "train_loss": 0.4515265909830729,
+      "train_runtime": 14971.1378,
+      "train_samples_per_second": 30.779,
+      "train_steps_per_second": 0.04
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 600,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 100,
+  "total_flos": 1.1344381987478569e+19,
+  "train_batch_size": 192,
+  "trial_name": null,
+  "trial_params": null
+}