Upload 12 files

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8100fede3ebe12028e1a6a943068124b87c44e904129909600f8aa61534da523
+oid sha256:db26089639bed1d2042ae44508cfe824d8a32650e6c7a551b589f9d746235bc1
 size 1264702096

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:64f466b795a2b3f1ba9a0a42ea3004a4fecc0d41800f8d661e39d9eeb5c1c572
+oid sha256:34a41698f332471e4edf3ba686ae22f0a6f4731aa35ba0f155586809001de0d1
 size 2495948982

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad8e0e131aa62acb514576ec4fc6d35c4209b434e53cc06404b8837a81316972
+oid sha256:ea2014619ab560562b67af3ac493f171046c4761b85b6e9309d2e81fe852477e
 size 14244

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b37cc57efe149b952d1d08baf340d5ff65995cd05283b74a9e6abf14b366d6de
+oid sha256:623a857fed43a56c0eb1d901869237571e3979cc10ac0751899d9842563263de
 size 1064

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "<pad>", "do_lower_case": true, "word_delimiter_token": "|", "tokenizer_class": "Wav2Vec2CTCTokenizer"}

trainer_state.json

@@ -1,96 +1,296 @@
 {
-  "best_metric": 0.14490656554698944,
-  "best_model_checkpoint": "./wav2vec2-base-demo/checkpoint-2000",
-  "epoch": 0.986679822397632,
+  "best_metric": 0.11605404317378998,
+  "best_model_checkpoint": "./wav2vec2-base-demo/checkpoint-7000",
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 2000,
+  "global_step": 7296,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.246669955599408,
-      "grad_norm": 3.482757329940796,
+      "epoch": 0.20559210526315788,
+      "grad_norm": 1.0479871034622192,
       "learning_rate": 4.97e-05,
-      "loss": 0.445,
+      "loss": 0.2427,
       "step": 500
     },
     {
-      "epoch": 0.246669955599408,
-      "eval_cer": 0.03472612578852039,
-      "eval_loss": 0.1506977081298828,
-      "eval_mer": 0.2400241837968561,
-      "eval_runtime": 72.835,
-      "eval_samples_per_second": 16.695,
-      "eval_steps_per_second": 2.087,
-      "eval_wer": 0.24393241167434715,
-      "eval_wil": 0.4027411629459634,
-      "eval_wip": 0.5972588370540366,
+      "epoch": 0.20559210526315788,
+      "eval_cer": 0.03330171560541076,
+      "eval_loss": 0.14631743729114532,
+      "eval_mer": 0.2295178941141674,
+      "eval_runtime": 73.5989,
+      "eval_samples_per_second": 16.522,
+      "eval_steps_per_second": 2.065,
+      "eval_wer": 0.23178016726403824,
+      "eval_wil": 0.38783171622631485,
+      "eval_wip": 0.6121682837736852,
       "step": 500
     },
     {
-      "epoch": 0.493339911198816,
-      "grad_norm": 2.1172962188720703,
-      "learning_rate": 4.5547392940333276e-05,
-      "loss": 0.4275,
+      "epoch": 0.41118421052631576,
+      "grad_norm": 7.891260147094727,
+      "learning_rate": 4.634343731606828e-05,
+      "loss": 0.129,
       "step": 1000
     },
     {
-      "epoch": 0.493339911198816,
-      "eval_cer": 0.03416401224158391,
-      "eval_loss": 0.15176725387573242,
-      "eval_mer": 0.23524069028156222,
-      "eval_runtime": 72.5697,
-      "eval_samples_per_second": 16.756,
-      "eval_steps_per_second": 2.095,
-      "eval_wer": 0.23870967741935484,
-      "eval_wil": 0.3973965187752956,
-      "eval_wip": 0.6026034812247044,
+      "epoch": 0.41118421052631576,
+      "eval_cer": 0.034745133619267565,
+      "eval_loss": 0.15320204198360443,
+      "eval_mer": 0.22595870206489677,
+      "eval_runtime": 73.4082,
+      "eval_samples_per_second": 16.565,
+      "eval_steps_per_second": 2.071,
+      "eval_wer": 0.22879330943847073,
+      "eval_wil": 0.38389337364965037,
+      "eval_wip": 0.6161066263503496,
       "step": 1000
     },
     {
-      "epoch": 0.740009866798224,
-      "grad_norm": 3.9609649181365967,
-      "learning_rate": 4.1067908976885865e-05,
-      "loss": 0.4367,
+      "epoch": 0.6167763157894737,
+      "grad_norm": 3.9030227661132812,
+      "learning_rate": 4.266480282519129e-05,
+      "loss": 0.1431,
       "step": 1500
     },
     {
-      "epoch": 0.740009866798224,
-      "eval_cer": 0.032727499843857345,
-      "eval_loss": 0.14839377999305725,
-      "eval_mer": 0.21904472163066627,
-      "eval_runtime": 72.6193,
-      "eval_samples_per_second": 16.745,
-      "eval_steps_per_second": 2.093,
-      "eval_wer": 0.22119815668202766,
-      "eval_wil": 0.3745997578877486,
-      "eval_wip": 0.6254002421122514,
+      "epoch": 0.6167763157894737,
+      "eval_cer": 0.034745133619267565,
+      "eval_loss": 0.15051577985286713,
+      "eval_mer": 0.23915626856803326,
+      "eval_runtime": 73.4176,
+      "eval_samples_per_second": 16.563,
+      "eval_steps_per_second": 2.07,
+      "eval_wer": 0.240442054958184,
+      "eval_wil": 0.397231780167264,
+      "eval_wip": 0.602768219832736,
       "step": 1500
     },
     {
-      "epoch": 0.986679822397632,
-      "grad_norm": 3.834362268447876,
-      "learning_rate": 3.6588425013438454e-05,
-      "loss": 0.4128,
+      "epoch": 0.8223684210526315,
+      "grad_norm": 1.7316572666168213,
+      "learning_rate": 3.8986168334314304e-05,
+      "loss": 0.1393,
       "step": 2000
     },
     {
-      "epoch": 0.986679822397632,
-      "eval_cer": 0.032602585722315905,
-      "eval_loss": 0.14490656554698944,
-      "eval_mer": 0.21031746031746032,
-      "eval_runtime": 72.7757,
-      "eval_samples_per_second": 16.709,
-      "eval_steps_per_second": 2.089,
-      "eval_wer": 0.21167434715821812,
-      "eval_wil": 0.3614631097880948,
-      "eval_wip": 0.6385368902119052,
+      "epoch": 0.8223684210526315,
+      "eval_cer": 0.03425024744308809,
+      "eval_loss": 0.14699852466583252,
+      "eval_mer": 0.2295903330386089,
+      "eval_runtime": 73.3571,
+      "eval_samples_per_second": 16.576,
+      "eval_steps_per_second": 2.072,
+      "eval_wer": 0.23267622461170848,
+      "eval_wil": 0.38674316181876667,
+      "eval_wip": 0.6132568381812333,
       "step": 2000
+    },
+    {
+      "epoch": 1.0279605263157894,
+      "grad_norm": 1.396478533744812,
+      "learning_rate": 3.530753384343732e-05,
+      "loss": 0.1505,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0279605263157894,
+      "eval_cer": 0.033136753546684265,
+      "eval_loss": 0.14985224604606628,
+      "eval_mer": 0.2275943396226415,
+      "eval_runtime": 74.0264,
+      "eval_samples_per_second": 16.427,
+      "eval_steps_per_second": 2.053,
+      "eval_wer": 0.23058542413381122,
+      "eval_wil": 0.38484887915314825,
+      "eval_wip": 0.6151511208468518,
+      "step": 2500
+    },
+    {
+      "epoch": 1.2335526315789473,
+      "grad_norm": 0.4791443347930908,
+      "learning_rate": 3.163625662154209e-05,
+      "loss": 0.1636,
+      "step": 3000
+    },
+    {
+      "epoch": 1.2335526315789473,
+      "eval_cer": 0.03352853843615968,
+      "eval_loss": 0.14658136665821075,
+      "eval_mer": 0.2253104671791839,
+      "eval_runtime": 73.6214,
+      "eval_samples_per_second": 16.517,
+      "eval_steps_per_second": 2.065,
+      "eval_wer": 0.22759856630824374,
+      "eval_wil": 0.38094846443763386,
+      "eval_wip": 0.6190515355623661,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4391447368421053,
+      "grad_norm": 0.7940099239349365,
+      "learning_rate": 2.79576221306651e-05,
+      "loss": 0.179,
+      "step": 3500
+    },
+    {
+      "epoch": 1.4391447368421053,
+      "eval_cer": 0.03336357637743319,
+      "eval_loss": 0.13871586322784424,
+      "eval_mer": 0.22300884955752212,
+      "eval_runtime": 73.6122,
+      "eval_samples_per_second": 16.519,
+      "eval_steps_per_second": 2.065,
+      "eval_wer": 0.22580645161290322,
+      "eval_wil": 0.3782571805567654,
+      "eval_wip": 0.6217428194432346,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6447368421052633,
+      "grad_norm": 1.963213562965393,
+      "learning_rate": 2.4278987639788112e-05,
+      "loss": 0.2181,
+      "step": 4000
+    },
+    {
+      "epoch": 1.6447368421052633,
+      "eval_cer": 0.032641867370504785,
+      "eval_loss": 0.13555069267749786,
+      "eval_mer": 0.2170839469808542,
+      "eval_runtime": 73.2159,
+      "eval_samples_per_second": 16.608,
+      "eval_steps_per_second": 2.076,
+      "eval_wer": 0.22013142174432496,
+      "eval_wil": 0.36858040063576636,
+      "eval_wip": 0.6314195993642336,
+      "step": 4000
+    },
+    {
+      "epoch": 1.850328947368421,
+      "grad_norm": 6.549672603607178,
+      "learning_rate": 2.0600353148911125e-05,
+      "loss": 0.2413,
+      "step": 4500
+    },
+    {
+      "epoch": 1.850328947368421,
+      "eval_cer": 0.030621082151105247,
+      "eval_loss": 0.1285562515258789,
+      "eval_mer": 0.2132114420524919,
+      "eval_runtime": 73.5544,
+      "eval_samples_per_second": 16.532,
+      "eval_steps_per_second": 2.066,
+      "eval_wer": 0.21594982078853048,
+      "eval_wil": 0.3613360451023606,
+      "eval_wip": 0.6386639548976394,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0559210526315788,
+      "grad_norm": 2.745516061782837,
+      "learning_rate": 1.6929075927015892e-05,
+      "loss": 0.2935,
+      "step": 5000
+    },
+    {
+      "epoch": 2.0559210526315788,
+      "eval_cer": 0.029486967997360607,
+      "eval_loss": 0.12071426212787628,
+      "eval_mer": 0.2039551357733176,
+      "eval_runtime": 73.611,
+      "eval_samples_per_second": 16.519,
+      "eval_steps_per_second": 2.065,
+      "eval_wer": 0.20639187574671444,
+      "eval_wil": 0.34776843790653456,
+      "eval_wip": 0.6522315620934654,
+      "step": 5000
+    },
+    {
+      "epoch": 2.2615131578947367,
+      "grad_norm": 1.6406818628311157,
+      "learning_rate": 1.3265155974102414e-05,
+      "loss": 0.3035,
+      "step": 5500
+    },
+    {
+      "epoch": 2.2615131578947367,
+      "eval_cer": 0.02928076542395249,
+      "eval_loss": 0.11994051933288574,
+      "eval_mer": 0.19940828402366864,
+      "eval_runtime": 73.791,
+      "eval_samples_per_second": 16.479,
+      "eval_steps_per_second": 2.06,
+      "eval_wer": 0.2013142174432497,
+      "eval_wil": 0.34143103677153863,
+      "eval_wip": 0.6585689632284614,
+      "step": 5500
+    },
+    {
+      "epoch": 2.4671052631578947,
+      "grad_norm": 1.5923278331756592,
+      "learning_rate": 9.586521483225428e-06,
+      "loss": 0.3185,
+      "step": 6000
+    },
+    {
+      "epoch": 2.4671052631578947,
+      "eval_cer": 0.028909600791817883,
+      "eval_loss": 0.11905083060264587,
+      "eval_mer": 0.1989946777054997,
+      "eval_runtime": 73.7234,
+      "eval_samples_per_second": 16.494,
+      "eval_steps_per_second": 2.062,
+      "eval_wer": 0.20101553166069294,
+      "eval_wil": 0.34076400679117147,
+      "eval_wip": 0.6592359932088285,
+      "step": 6000
+    },
+    {
+      "epoch": 2.6726973684210527,
+      "grad_norm": 1.212615728378296,
+      "learning_rate": 5.907886992348441e-06,
+      "loss": 0.3023,
+      "step": 6500
+    },
+    {
+      "epoch": 2.6726973684210527,
+      "eval_cer": 0.028579676674364896,
+      "eval_loss": 0.11668401211500168,
+      "eval_mer": 0.19840094758661533,
+      "eval_runtime": 73.4554,
+      "eval_samples_per_second": 16.554,
+      "eval_steps_per_second": 2.069,
+      "eval_wer": 0.2001194743130227,
+      "eval_wil": 0.34014943834106115,
+      "eval_wip": 0.6598505616589389,
+      "step": 6500
+    },
+    {
+      "epoch": 2.8782894736842106,
+      "grad_norm": 1.4651743173599243,
+      "learning_rate": 2.229252501471454e-06,
+      "loss": 0.3116,
+      "step": 7000
+    },
+    {
+      "epoch": 2.8782894736842106,
+      "eval_cer": 0.02839409435829759,
+      "eval_loss": 0.11605404317378998,
+      "eval_mer": 0.19668737060041408,
+      "eval_runtime": 73.5998,
+      "eval_samples_per_second": 16.522,
+      "eval_steps_per_second": 2.065,
+      "eval_wer": 0.19862604540023895,
+      "eval_wil": 0.3373527007482864,
+      "eval_wip": 0.6626472992517136,
+      "step": 7000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 6081,
+  "max_steps": 7296,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 500,
@@ -101,13 +301,13 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 3.2957716566380933e+18,
-  "train_batch_size": 12,
+  "total_flos": 9.992785283464022e+18,
+  "train_batch_size": 10,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1456fbb80fdcd6a8b153f11eeae0fef4bc2173125f866d139aac039bc0656f69
+oid sha256:3bf5af7ff8411218351a777026fc80bf0652d760c3902741ad28215c5270edd0
 size 5176