End of training

README.md

@@ -3,6 +3,7 @@ library_name: transformers
 license: apache-2.0
 base_model: openai/whisper-small
 tags:
+- whisper-event
 - generated_from_trainer
 metrics:
 - wer
@@ -18,8 +19,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2.2260
-- Wer: 102.5878
+- Loss: 1.9993
+- Wer: 99.2606
 
 ## Model description
 

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 5.0761421319796955,
+    "eval_loss": 1.9993243217468262,
+    "eval_runtime": 144.5521,
+    "eval_samples": 500,
+    "eval_samples_per_second": 3.459,
+    "eval_steps_per_second": 1.729,
+    "eval_wer": 99.26062846580407,
+    "total_flos": 2.8844110761984e+18,
+    "train_loss": 0.9328738340377808,
+    "train_runtime": 2999.0534,
+    "train_samples": 2000,
+    "train_samples_per_second": 3.334,
+    "train_steps_per_second": 1.667
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 5.0761421319796955,
+    "eval_loss": 1.9993243217468262,
+    "eval_runtime": 144.5521,
+    "eval_samples": 500,
+    "eval_samples_per_second": 3.459,
+    "eval_steps_per_second": 1.729,
+    "eval_wer": 99.26062846580407
+}

runs/Aug21_16-08-48_DESKTOP-8I5F6JA/events.out.tfevents.1724230985.DESKTOP-8I5F6JA

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e1466826497c153a6cbee81892be4a2994bde530c1850d844fcb9b94625141e
+size 40

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 5.0761421319796955,
+    "total_flos": 2.8844110761984e+18,
+    "train_loss": 0.9328738340377808,
+    "train_runtime": 2999.0534,
+    "train_samples": 2000,
+    "train_samples_per_second": 3.334,
+    "train_steps_per_second": 1.667
+}

trainer_state.json

@@ -0,0 +1,437 @@
+{
+  "best_metric": 99.26062846580407,
+  "best_model_checkpoint": "./output_dir/checkpoint-2000",
+  "epoch": 5.0761421319796955,
+  "eval_steps": 1000,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.10152284263959391,
+      "grad_norm": 45.89750671386719,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 2.9244,
+      "step": 100
+    },
+    {
+      "epoch": 0.20304568527918782,
+      "grad_norm": 45.53427505493164,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 2.3216,
+      "step": 200
+    },
+    {
+      "epoch": 0.30456852791878175,
+      "grad_norm": 48.774688720703125,
+      "learning_rate": 5.9e-06,
+      "loss": 2.3683,
+      "step": 300
+    },
+    {
+      "epoch": 0.40609137055837563,
+      "grad_norm": 57.26857376098633,
+      "learning_rate": 7.9e-06,
+      "loss": 2.2015,
+      "step": 400
+    },
+    {
+      "epoch": 0.5076142131979695,
+      "grad_norm": 50.28277587890625,
+      "learning_rate": 9.9e-06,
+      "loss": 2.1731,
+      "step": 500
+    },
+    {
+      "epoch": 0.6091370558375635,
+      "grad_norm": 55.22574234008789,
+      "learning_rate": 9.78888888888889e-06,
+      "loss": 2.1768,
+      "step": 600
+    },
+    {
+      "epoch": 0.7106598984771574,
+      "grad_norm": 52.53302764892578,
+      "learning_rate": 9.566666666666668e-06,
+      "loss": 2.0982,
+      "step": 700
+    },
+    {
+      "epoch": 0.8121827411167513,
+      "grad_norm": 54.07991409301758,
+      "learning_rate": 9.344444444444446e-06,
+      "loss": 1.9572,
+      "step": 800
+    },
+    {
+      "epoch": 0.9137055837563451,
+      "grad_norm": 34.251190185546875,
+      "learning_rate": 9.124444444444444e-06,
+      "loss": 2.0769,
+      "step": 900
+    },
+    {
+      "epoch": 1.015228426395939,
+      "grad_norm": 47.28948211669922,
+      "learning_rate": 8.902222222222224e-06,
+      "loss": 1.9198,
+      "step": 1000
+    },
+    {
+      "epoch": 1.015228426395939,
+      "eval_loss": 1.9924733638763428,
+      "eval_runtime": 140.0717,
+      "eval_samples_per_second": 3.57,
+      "eval_steps_per_second": 1.785,
+      "eval_wer": 100.73937153419594,
+      "step": 1000
+    },
+    {
+      "epoch": 1.116751269035533,
+      "grad_norm": 34.96185302734375,
+      "learning_rate": 8.68e-06,
+      "loss": 1.321,
+      "step": 1100
+    },
+    {
+      "epoch": 1.218274111675127,
+      "grad_norm": 32.89934539794922,
+      "learning_rate": 8.457777777777778e-06,
+      "loss": 1.3555,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3197969543147208,
+      "grad_norm": 62.16946029663086,
+      "learning_rate": 8.235555555555557e-06,
+      "loss": 1.3613,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4213197969543148,
+      "grad_norm": 43.142662048339844,
+      "learning_rate": 8.013333333333333e-06,
+      "loss": 1.3281,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5228426395939088,
+      "grad_norm": 47.314292907714844,
+      "learning_rate": 7.791111111111111e-06,
+      "loss": 1.367,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6243654822335025,
+      "grad_norm": 52.20304870605469,
+      "learning_rate": 7.56888888888889e-06,
+      "loss": 1.4075,
+      "step": 1600
+    },
+    {
+      "epoch": 1.7258883248730963,
+      "grad_norm": 59.97978210449219,
+      "learning_rate": 7.346666666666668e-06,
+      "loss": 1.2577,
+      "step": 1700
+    },
+    {
+      "epoch": 1.8274111675126905,
+      "grad_norm": 40.61113357543945,
+      "learning_rate": 7.124444444444445e-06,
+      "loss": 1.4001,
+      "step": 1800
+    },
+    {
+      "epoch": 1.9289340101522843,
+      "grad_norm": 40.51613235473633,
+      "learning_rate": 6.902222222222223e-06,
+      "loss": 1.2778,
+      "step": 1900
+    },
+    {
+      "epoch": 2.030456852791878,
+      "grad_norm": 29.502897262573242,
+      "learning_rate": 6.680000000000001e-06,
+      "loss": 1.1588,
+      "step": 2000
+    },
+    {
+      "epoch": 2.030456852791878,
+      "eval_loss": 1.9993243217468262,
+      "eval_runtime": 139.6714,
+      "eval_samples_per_second": 3.58,
+      "eval_steps_per_second": 1.79,
+      "eval_wer": 99.26062846580407,
+      "step": 2000
+    },
+    {
+      "epoch": 2.1319796954314723,
+      "grad_norm": 32.84286880493164,
+      "learning_rate": 6.457777777777778e-06,
+      "loss": 0.6099,
+      "step": 2100
+    },
+    {
+      "epoch": 2.233502538071066,
+      "grad_norm": 30.602407455444336,
+      "learning_rate": 6.235555555555556e-06,
+      "loss": 0.6896,
+      "step": 2200
+    },
+    {
+      "epoch": 2.33502538071066,
+      "grad_norm": 27.283344268798828,
+      "learning_rate": 6.013333333333335e-06,
+      "loss": 0.6841,
+      "step": 2300
+    },
+    {
+      "epoch": 2.436548223350254,
+      "grad_norm": 50.34073257446289,
+      "learning_rate": 5.791111111111112e-06,
+      "loss": 0.7125,
+      "step": 2400
+    },
+    {
+      "epoch": 2.5380710659898478,
+      "grad_norm": 27.890342712402344,
+      "learning_rate": 5.56888888888889e-06,
+      "loss": 0.7412,
+      "step": 2500
+    },
+    {
+      "epoch": 2.6395939086294415,
+      "grad_norm": 43.19711685180664,
+      "learning_rate": 5.346666666666667e-06,
+      "loss": 0.6532,
+      "step": 2600
+    },
+    {
+      "epoch": 2.7411167512690353,
+      "grad_norm": 43.589752197265625,
+      "learning_rate": 5.124444444444445e-06,
+      "loss": 0.7458,
+      "step": 2700
+    },
+    {
+      "epoch": 2.8426395939086295,
+      "grad_norm": 29.730897903442383,
+      "learning_rate": 4.902222222222222e-06,
+      "loss": 0.7502,
+      "step": 2800
+    },
+    {
+      "epoch": 2.9441624365482233,
+      "grad_norm": 34.04936599731445,
+      "learning_rate": 4.680000000000001e-06,
+      "loss": 0.6897,
+      "step": 2900
+    },
+    {
+      "epoch": 3.045685279187817,
+      "grad_norm": 27.477937698364258,
+      "learning_rate": 4.457777777777778e-06,
+      "loss": 0.4916,
+      "step": 3000
+    },
+    {
+      "epoch": 3.045685279187817,
+      "eval_loss": 2.087052822113037,
+      "eval_runtime": 141.6154,
+      "eval_samples_per_second": 3.531,
+      "eval_steps_per_second": 1.765,
+      "eval_wer": 103.51201478743069,
+      "step": 3000
+    },
+    {
+      "epoch": 3.1472081218274113,
+      "grad_norm": 11.8717679977417,
+      "learning_rate": 4.235555555555556e-06,
+      "loss": 0.3133,
+      "step": 3100
+    },
+    {
+      "epoch": 3.248730964467005,
+      "grad_norm": 26.114721298217773,
+      "learning_rate": 4.013333333333334e-06,
+      "loss": 0.3778,
+      "step": 3200
+    },
+    {
+      "epoch": 3.350253807106599,
+      "grad_norm": 25.646799087524414,
+      "learning_rate": 3.7911111111111114e-06,
+      "loss": 0.3168,
+      "step": 3300
+    },
+    {
+      "epoch": 3.451776649746193,
+      "grad_norm": 21.65736198425293,
+      "learning_rate": 3.568888888888889e-06,
+      "loss": 0.3316,
+      "step": 3400
+    },
+    {
+      "epoch": 3.553299492385787,
+      "grad_norm": 28.374004364013672,
+      "learning_rate": 3.346666666666667e-06,
+      "loss": 0.3265,
+      "step": 3500
+    },
+    {
+      "epoch": 3.6548223350253806,
+      "grad_norm": 43.69049072265625,
+      "learning_rate": 3.124444444444445e-06,
+      "loss": 0.3226,
+      "step": 3600
+    },
+    {
+      "epoch": 3.7563451776649748,
+      "grad_norm": 7.046782493591309,
+      "learning_rate": 2.9022222222222223e-06,
+      "loss": 0.2844,
+      "step": 3700
+    },
+    {
+      "epoch": 3.8578680203045685,
+      "grad_norm": 37.4542236328125,
+      "learning_rate": 2.68e-06,
+      "loss": 0.3246,
+      "step": 3800
+    },
+    {
+      "epoch": 3.9593908629441623,
+      "grad_norm": 20.5508975982666,
+      "learning_rate": 2.457777777777778e-06,
+      "loss": 0.2773,
+      "step": 3900
+    },
+    {
+      "epoch": 4.060913705583756,
+      "grad_norm": 15.81009292602539,
+      "learning_rate": 2.235555555555556e-06,
+      "loss": 0.1894,
+      "step": 4000
+    },
+    {
+      "epoch": 4.060913705583756,
+      "eval_loss": 2.1815617084503174,
+      "eval_runtime": 144.067,
+      "eval_samples_per_second": 3.471,
+      "eval_steps_per_second": 1.735,
+      "eval_wer": 102.95748613678373,
+      "step": 4000
+    },
+    {
+      "epoch": 4.16243654822335,
+      "grad_norm": 2.2856040000915527,
+      "learning_rate": 2.0133333333333337e-06,
+      "loss": 0.1466,
+      "step": 4100
+    },
+    {
+      "epoch": 4.2639593908629445,
+      "grad_norm": 5.688824653625488,
+      "learning_rate": 1.7911111111111113e-06,
+      "loss": 0.1439,
+      "step": 4200
+    },
+    {
+      "epoch": 4.365482233502538,
+      "grad_norm": 1.2593058347702026,
+      "learning_rate": 1.568888888888889e-06,
+      "loss": 0.1348,
+      "step": 4300
+    },
+    {
+      "epoch": 4.467005076142132,
+      "grad_norm": 23.989395141601562,
+      "learning_rate": 1.3466666666666668e-06,
+      "loss": 0.11,
+      "step": 4400
+    },
+    {
+      "epoch": 4.568527918781726,
+      "grad_norm": 2.5791878700256348,
+      "learning_rate": 1.1244444444444446e-06,
+      "loss": 0.1594,
+      "step": 4500
+    },
+    {
+      "epoch": 4.67005076142132,
+      "grad_norm": 21.947067260742188,
+      "learning_rate": 9.022222222222222e-07,
+      "loss": 0.1805,
+      "step": 4600
+    },
+    {
+      "epoch": 4.771573604060913,
+      "grad_norm": 33.04467010498047,
+      "learning_rate": 6.800000000000001e-07,
+      "loss": 0.1427,
+      "step": 4700
+    },
+    {
+      "epoch": 4.873096446700508,
+      "grad_norm": 2.3702778816223145,
+      "learning_rate": 4.5777777777777784e-07,
+      "loss": 0.1359,
+      "step": 4800
+    },
+    {
+      "epoch": 4.974619289340102,
+      "grad_norm": 12.491398811340332,
+      "learning_rate": 2.3555555555555556e-07,
+      "loss": 0.1331,
+      "step": 4900
+    },
+    {
+      "epoch": 5.0761421319796955,
+      "grad_norm": 1.168066382408142,
+      "learning_rate": 1.3333333333333334e-08,
+      "loss": 0.072,
+      "step": 5000
+    },
+    {
+      "epoch": 5.0761421319796955,
+      "eval_loss": 2.2259573936462402,
+      "eval_runtime": 144.5172,
+      "eval_samples_per_second": 3.46,
+      "eval_steps_per_second": 1.73,
+      "eval_wer": 102.58780036968578,
+      "step": 5000
+    },
+    {
+      "epoch": 5.0761421319796955,
+      "step": 5000,
+      "total_flos": 2.8844110761984e+18,
+      "train_loss": 0.9328738340377808,
+      "train_runtime": 2999.0534,
+      "train_samples_per_second": 3.334,
+      "train_steps_per_second": 1.667
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 5000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.8844110761984e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}