Training checkpoint

README.md

@@ -1,21 +1,23 @@
 ---
+language:
+- zh
 license: apache-2.0
-base_model: jun-han/whisper-small-zh
+base_model: openai/whisper-small
 tags:
 - generated_from_trainer
 datasets:
-- common_voice_16_1
+- mozilla-foundation/common_voice_16_1
 model-index:
-- name: whisper-small
+- name: Wisper-Small-zh_test
   results: []
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# whisper-small
+# Wisper-Small-zh_test
 
-This model is a fine-tuned version of [jun-han/whisper-small-zh](https://huggingface.co/jun-han/whisper-small-zh) on the common_voice_16_1 dataset.
+This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Common Voice 16.1 dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.3867
 - Cer: 21.2804

trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 33.39944765252322,
-  "best_model_checkpoint": "/kaggle/working/whisper-small/checkpoint-1000",
-  "epoch": 1.3245033112582782,
+  "best_metric": 21.28044187798142,
+  "best_model_checkpoint": "/kaggle/working/whisper-small/checkpoint-2000",
+  "epoch": 2.6490066225165565,
   "eval_steps": 500,
-  "global_step": 1000,
+  "global_step": 2000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -307,19 +307,317 @@
       "step": 1000
     },
     {
-      "epoch": 1.3245033112582782,
-      "step": 1000,
-      "total_flos": 4.65727933108224e+18,
-      "train_loss": 2.47559251499176,
-      "train_runtime": 12073.0146,
-      "train_samples_per_second": 1.325,
-      "train_steps_per_second": 0.083
+      "epoch": 1.3576158940397351,
+      "grad_norm": 15.552218437194824,
+      "learning_rate": 6.540000000000001e-06,
+      "loss": 0.501,
+      "step": 1025
+    },
+    {
+      "epoch": 1.390728476821192,
+      "grad_norm": 18.500070571899414,
+      "learning_rate": 6.373333333333334e-06,
+      "loss": 0.5625,
+      "step": 1050
+    },
+    {
+      "epoch": 1.423841059602649,
+      "grad_norm": 13.384895324707031,
+      "learning_rate": 6.206666666666668e-06,
+      "loss": 0.4785,
+      "step": 1075
+    },
+    {
+      "epoch": 1.4569536423841059,
+      "grad_norm": 11.979981422424316,
+      "learning_rate": 6.040000000000001e-06,
+      "loss": 0.4728,
+      "step": 1100
+    },
+    {
+      "epoch": 1.490066225165563,
+      "grad_norm": 13.125775337219238,
+      "learning_rate": 5.873333333333334e-06,
+      "loss": 0.4484,
+      "step": 1125
+    },
+    {
+      "epoch": 1.5231788079470199,
+      "grad_norm": 10.14282512664795,
+      "learning_rate": 5.713333333333334e-06,
+      "loss": 0.4227,
+      "step": 1150
+    },
+    {
+      "epoch": 1.5562913907284768,
+      "grad_norm": 14.51291561126709,
+      "learning_rate": 5.546666666666667e-06,
+      "loss": 0.4419,
+      "step": 1175
+    },
+    {
+      "epoch": 1.589403973509934,
+      "grad_norm": 13.863635063171387,
+      "learning_rate": 5.380000000000001e-06,
+      "loss": 0.4429,
+      "step": 1200
+    },
+    {
+      "epoch": 1.6225165562913908,
+      "grad_norm": 11.522802352905273,
+      "learning_rate": 5.213333333333334e-06,
+      "loss": 0.3981,
+      "step": 1225
+    },
+    {
+      "epoch": 1.6556291390728477,
+      "grad_norm": 13.733115196228027,
+      "learning_rate": 5.046666666666668e-06,
+      "loss": 0.3974,
+      "step": 1250
+    },
+    {
+      "epoch": 1.6887417218543046,
+      "grad_norm": 11.775789260864258,
+      "learning_rate": 4.880000000000001e-06,
+      "loss": 0.3978,
+      "step": 1275
+    },
+    {
+      "epoch": 1.7218543046357615,
+      "grad_norm": 8.765869140625,
+      "learning_rate": 4.713333333333334e-06,
+      "loss": 0.3812,
+      "step": 1300
+    },
+    {
+      "epoch": 1.7549668874172184,
+      "grad_norm": 10.572060585021973,
+      "learning_rate": 4.546666666666667e-06,
+      "loss": 0.4257,
+      "step": 1325
+    },
+    {
+      "epoch": 1.7880794701986755,
+      "grad_norm": 11.202813148498535,
+      "learning_rate": 4.38e-06,
+      "loss": 0.3635,
+      "step": 1350
+    },
+    {
+      "epoch": 1.8211920529801324,
+      "grad_norm": 10.402491569519043,
+      "learning_rate": 4.213333333333333e-06,
+      "loss": 0.3636,
+      "step": 1375
+    },
+    {
+      "epoch": 1.8543046357615895,
+      "grad_norm": 22.270811080932617,
+      "learning_rate": 4.046666666666667e-06,
+      "loss": 0.3696,
+      "step": 1400
+    },
+    {
+      "epoch": 1.8874172185430464,
+      "grad_norm": 8.238192558288574,
+      "learning_rate": 3.88e-06,
+      "loss": 0.397,
+      "step": 1425
+    },
+    {
+      "epoch": 1.9205298013245033,
+      "grad_norm": 9.551776885986328,
+      "learning_rate": 3.713333333333334e-06,
+      "loss": 0.3747,
+      "step": 1450
+    },
+    {
+      "epoch": 1.9536423841059603,
+      "grad_norm": 13.149867057800293,
+      "learning_rate": 3.5466666666666673e-06,
+      "loss": 0.3771,
+      "step": 1475
+    },
+    {
+      "epoch": 1.9867549668874172,
+      "grad_norm": 13.081770896911621,
+      "learning_rate": 3.3800000000000007e-06,
+      "loss": 0.352,
+      "step": 1500
+    },
+    {
+      "epoch": 1.9867549668874172,
+      "eval_cer": 23.299020838563898,
+      "eval_loss": 0.4214184284210205,
+      "eval_runtime": 2258.7536,
+      "eval_samples_per_second": 2.203,
+      "eval_steps_per_second": 0.275,
+      "step": 1500
+    },
+    {
+      "epoch": 2.019867549668874,
+      "grad_norm": 7.450255393981934,
+      "learning_rate": 3.213333333333334e-06,
+      "loss": 0.2911,
+      "step": 1525
+    },
+    {
+      "epoch": 2.052980132450331,
+      "grad_norm": 8.604903221130371,
+      "learning_rate": 3.0466666666666666e-06,
+      "loss": 0.2069,
+      "step": 1550
+    },
+    {
+      "epoch": 2.0860927152317883,
+      "grad_norm": 5.367754936218262,
+      "learning_rate": 2.88e-06,
+      "loss": 0.2461,
+      "step": 1575
+    },
+    {
+      "epoch": 2.119205298013245,
+      "grad_norm": 11.53250789642334,
+      "learning_rate": 2.7133333333333333e-06,
+      "loss": 0.2071,
+      "step": 1600
+    },
+    {
+      "epoch": 2.152317880794702,
+      "grad_norm": 9.057580947875977,
+      "learning_rate": 2.5466666666666667e-06,
+      "loss": 0.2255,
+      "step": 1625
+    },
+    {
+      "epoch": 2.185430463576159,
+      "grad_norm": 8.253719329833984,
+      "learning_rate": 2.38e-06,
+      "loss": 0.1968,
+      "step": 1650
+    },
+    {
+      "epoch": 2.218543046357616,
+      "grad_norm": 10.867476463317871,
+      "learning_rate": 2.2133333333333335e-06,
+      "loss": 0.2177,
+      "step": 1675
+    },
+    {
+      "epoch": 2.251655629139073,
+      "grad_norm": 10.779939651489258,
+      "learning_rate": 2.046666666666667e-06,
+      "loss": 0.2191,
+      "step": 1700
+    },
+    {
+      "epoch": 2.2847682119205297,
+      "grad_norm": 8.389144897460938,
+      "learning_rate": 1.8800000000000002e-06,
+      "loss": 0.2137,
+      "step": 1725
+    },
+    {
+      "epoch": 2.3178807947019866,
+      "grad_norm": 11.38824462890625,
+      "learning_rate": 1.7133333333333336e-06,
+      "loss": 0.2326,
+      "step": 1750
+    },
+    {
+      "epoch": 2.3509933774834435,
+      "grad_norm": 10.286229133605957,
+      "learning_rate": 1.546666666666667e-06,
+      "loss": 0.2014,
+      "step": 1775
+    },
+    {
+      "epoch": 2.384105960264901,
+      "grad_norm": 9.5010986328125,
+      "learning_rate": 1.3800000000000001e-06,
+      "loss": 0.2121,
+      "step": 1800
+    },
+    {
+      "epoch": 2.4172185430463577,
+      "grad_norm": 10.394664764404297,
+      "learning_rate": 1.2133333333333335e-06,
+      "loss": 0.2166,
+      "step": 1825
+    },
+    {
+      "epoch": 2.4503311258278146,
+      "grad_norm": 13.732166290283203,
+      "learning_rate": 1.0466666666666669e-06,
+      "loss": 0.2243,
+      "step": 1850
+    },
+    {
+      "epoch": 2.4834437086092715,
+      "grad_norm": 7.431657791137695,
+      "learning_rate": 8.8e-07,
+      "loss": 0.2186,
+      "step": 1875
+    },
+    {
+      "epoch": 2.5165562913907285,
+      "grad_norm": 5.561306476593018,
+      "learning_rate": 7.133333333333334e-07,
+      "loss": 0.1992,
+      "step": 1900
+    },
+    {
+      "epoch": 2.5496688741721854,
+      "grad_norm": 8.894646644592285,
+      "learning_rate": 5.466666666666667e-07,
+      "loss": 0.2195,
+      "step": 1925
+    },
+    {
+      "epoch": 2.5827814569536423,
+      "grad_norm": 8.13320541381836,
+      "learning_rate": 3.8e-07,
+      "loss": 0.2235,
+      "step": 1950
+    },
+    {
+      "epoch": 2.6158940397350996,
+      "grad_norm": 8.711597442626953,
+      "learning_rate": 2.1333333333333334e-07,
+      "loss": 0.2033,
+      "step": 1975
+    },
+    {
+      "epoch": 2.6490066225165565,
+      "grad_norm": 9.557293891906738,
+      "learning_rate": 4.6666666666666674e-08,
+      "loss": 0.2243,
+      "step": 2000
+    },
+    {
+      "epoch": 2.6490066225165565,
+      "eval_cer": 21.28044187798142,
+      "eval_loss": 0.38667094707489014,
+      "eval_runtime": 2236.5916,
+      "eval_samples_per_second": 2.225,
+      "eval_steps_per_second": 0.278,
+      "step": 2000
+    },
+    {
+      "epoch": 2.6490066225165565,
+      "step": 2000,
+      "total_flos": 9.31455866216448e+18,
+      "train_loss": 0.15934584045410155,
+      "train_runtime": 12035.0029,
+      "train_samples_per_second": 2.659,
+      "train_steps_per_second": 0.166
     }
   ],
   "logging_steps": 25,
-  "max_steps": 1000,
+  "max_steps": 2000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
+  "num_train_epochs": 3,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -333,7 +631,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.65727933108224e+18,
+  "total_flos": 9.31455866216448e+18,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null