Added model 38

config.json

@@ -1,6 +1,9 @@
 {
   "_name_or_path": "facebook/wav2vec2-xls-r-300m",
   "activation_dropout": 0.0,
+  "adapter_kernel_size": 3,
+  "adapter_stride": 2,
+  "add_adapter": false,
   "apply_spec_augment": true,
   "architectures": [
     "Wav2Vec2ForCTC"
@@ -58,10 +61,13 @@
   "layer_norm_eps": 1e-05,
   "layerdrop": 0.0,
   "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
   "mask_feature_prob": 0.0,
   "mask_time_length": 10,
+  "mask_time_min_masks": 2,
   "mask_time_prob": 0.05,
   "model_type": "wav2vec2",
+  "num_adapter_layers": 3,
   "num_attention_heads": 16,
   "num_codevector_groups": 2,
   "num_codevectors_per_group": 320,
@@ -70,10 +76,33 @@
   "num_feat_extract_layers": 7,
   "num_hidden_layers": 24,
   "num_negatives": 100,
-  "pad_token_id": 36,
+  "output_hidden_size": 1024,
+  "pad_token_id": 1,
   "proj_codevector_dim": 768,
+  "tdnn_dilation": [
+    1,
+    2,
+    3,
+    1,
+    1
+  ],
+  "tdnn_dim": [
+    512,
+    512,
+    512,
+    512,
+    1500
+  ],
+  "tdnn_kernel": [
+    5,
+    3,
+    3,
+    1,
+    1
+  ],
   "torch_dtype": "float32",
-  "transformers_version": "4.11.3",
+  "transformers_version": "4.17.0",
   "use_weighted_layer_sum": false,
-  "vocab_size": 39
+  "vocab_size": 50,
+  "xvector_output_dim": 512
 }

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:94403fe8c994757aeaec7d1aa29f6eb4939b69e447b973bdfa66315ca28289fc
-size 2490378769
+oid sha256:dbbd74245a32dfee65153384591d3e157fbaadbfa9ed6ba3e84bde67ecc0d26b
+size 2490469009

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d2adb27a665c0f58e24abcea96f3e4924947216cad0eaf5d8890f264304aa5dd
-size 1262083569
+oid sha256:a8832802a6f0497cac9b5409e1fbc323395d51b9df11522d9b57921ec8537ef6
+size 1262131313

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:05b4485ec356bcab49053a2a07a44306616f88219c11bd6cd8c9e1148af4e7d5
+oid sha256:e3887bd9846cc5801fe950d093508213f86dcae29c59bc632aecab4edfdc9648
 size 14503

scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:83a01e5ec7c6e9e42baf4502b3592fe5b3a70801cc1b41dc6175e27562b388c4
+oid sha256:8a6f2f23c0d6be14904a75536e0cd71a3982589a865f06899958ca265e06c661
 size 559

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:acc172b4559bbfe5e4502bb2f0d0f3157e9c54827074c2fd6be9036224745685
+oid sha256:1b50675021671aa65e03d08322cba2953ab219a5178e71b41f5c655d59cdae73
 size 623

trainer_state.json

@@ -1,151 +1,312 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 7.72454448017149,
-  "global_step": 3600,
+  "epoch": 7.909466234149218,
+  "global_step": 6700,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.86,
-      "learning_rate": 0.0002394,
-      "loss": 3.2993,
-      "step": 400
+      "epoch": 0.4,
+      "eval_cer": 0.2550782058484374,
+      "eval_loss": 0.9434235095977783,
+      "eval_runtime": 65.8936,
+      "eval_samples_per_second": 7.588,
+      "eval_steps_per_second": 0.956,
+      "eval_wer": 0.8986276613768566,
+      "step": 335
     },
     {
-      "epoch": 0.86,
-      "eval_loss": Infinity,
-      "eval_runtime": 828.7984,
-      "eval_samples_per_second": 9.006,
-      "eval_steps_per_second": 1.126,
-      "eval_wer": 0.36457931084030065,
+      "epoch": 0.47,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 3.5921,
       "step": 400
     },
     {
-      "epoch": 1.72,
-      "learning_rate": 0.0002723048327137546,
-      "loss": 0.3837,
-      "step": 800
+      "epoch": 0.79,
+      "eval_cer": 0.05617811418940895,
+      "eval_loss": 0.2126482129096985,
+      "eval_runtime": 65.2511,
+      "eval_samples_per_second": 7.663,
+      "eval_steps_per_second": 0.966,
+      "eval_wer": 0.20076805783023663,
+      "step": 670
     },
     {
-      "epoch": 1.72,
-      "eval_loss": Infinity,
-      "eval_runtime": 820.745,
-      "eval_samples_per_second": 9.094,
-      "eval_steps_per_second": 1.137,
-      "eval_wer": 0.25638285452569753,
+      "epoch": 0.94,
+      "learning_rate": 0.00028565965583173995,
+      "loss": 0.3347,
       "step": 800
     },
     {
-      "epoch": 2.57,
-      "learning_rate": 0.00023522304832713752,
-      "loss": 0.2576,
-      "step": 1200
+      "epoch": 1.19,
+      "eval_cer": 0.052689157623961445,
+      "eval_loss": 0.19026830792427063,
+      "eval_runtime": 66.2853,
+      "eval_samples_per_second": 7.543,
+      "eval_steps_per_second": 0.95,
+      "eval_wer": 0.18907776585531147,
+      "step": 1005
     },
     {
-      "epoch": 2.57,
-      "eval_loss": Infinity,
-      "eval_runtime": 816.7249,
-      "eval_samples_per_second": 9.139,
-      "eval_steps_per_second": 1.142,
-      "eval_wer": 0.17959103032697818,
+      "epoch": 1.42,
+      "learning_rate": 0.0002665391969407266,
+      "loss": 0.1948,
       "step": 1200
     },
     {
-      "epoch": 3.43,
-      "learning_rate": 0.0001981412639405204,
-      "loss": 0.1963,
-      "step": 1600
+      "epoch": 1.58,
+      "eval_cer": 0.04130569764347595,
+      "eval_loss": 0.1462544947862625,
+      "eval_runtime": 65.903,
+      "eval_samples_per_second": 7.587,
+      "eval_steps_per_second": 0.956,
+      "eval_wer": 0.13915400688993054,
+      "step": 1340
     },
     {
-      "epoch": 3.43,
-      "eval_loss": Infinity,
-      "eval_runtime": 819.1322,
-      "eval_samples_per_second": 9.112,
-      "eval_steps_per_second": 1.139,
-      "eval_wer": 0.1695059854696326,
+      "epoch": 1.89,
+      "learning_rate": 0.00024741873804971315,
+      "loss": 0.1737,
       "step": 1600
     },
     {
-      "epoch": 4.29,
-      "learning_rate": 0.00016096654275092934,
-      "loss": 0.158,
-      "step": 2000
+      "epoch": 1.98,
+      "eval_cer": 0.040093433074125544,
+      "eval_loss": 0.1409013271331787,
+      "eval_runtime": 66.3053,
+      "eval_samples_per_second": 7.541,
+      "eval_steps_per_second": 0.95,
+      "eval_wer": 0.13599141582425028,
+      "step": 1675
     },
     {
-      "epoch": 4.29,
-      "eval_loss": Infinity,
-      "eval_runtime": 815.3058,
-      "eval_samples_per_second": 9.155,
-      "eval_steps_per_second": 1.144,
-      "eval_wer": 0.16685675284904763,
+      "epoch": 2.36,
+      "learning_rate": 0.00022829827915869978,
+      "loss": 0.1466,
       "step": 2000
     },
     {
-      "epoch": 5.15,
-      "learning_rate": 0.00012388475836431226,
-      "loss": 0.125,
-      "step": 2400
+      "epoch": 2.37,
+      "eval_cer": 0.03887131269526822,
+      "eval_loss": 0.1429334431886673,
+      "eval_runtime": 65.9993,
+      "eval_samples_per_second": 7.576,
+      "eval_steps_per_second": 0.955,
+      "eval_wer": 0.13401479640820016,
+      "step": 2010
     },
     {
-      "epoch": 5.15,
-      "eval_loss": Infinity,
-      "eval_runtime": 815.712,
-      "eval_samples_per_second": 9.15,
-      "eval_steps_per_second": 1.144,
-      "eval_wer": 0.15157202769570643,
+      "epoch": 2.77,
+      "eval_cer": 0.03858549421956772,
+      "eval_loss": 0.1421019285917282,
+      "eval_runtime": 66.1016,
+      "eval_samples_per_second": 7.564,
+      "eval_steps_per_second": 0.953,
+      "eval_wer": 0.12678601682950247,
+      "step": 2345
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.0002091778202676864,
+      "loss": 0.1378,
       "step": 2400
     },
     {
-      "epoch": 6.01,
-      "learning_rate": 8.680297397769517e-05,
-      "loss": 0.101,
-      "step": 2800
+      "epoch": 3.16,
+      "eval_cer": 0.03703813212698225,
+      "eval_loss": 0.1354704052209854,
+      "eval_runtime": 65.9839,
+      "eval_samples_per_second": 7.578,
+      "eval_steps_per_second": 0.955,
+      "eval_wer": 0.12305867735923647,
+      "step": 2680
     },
     {
-      "epoch": 6.01,
-      "eval_loss": Infinity,
-      "eval_runtime": 815.5191,
-      "eval_samples_per_second": 9.152,
-      "eval_steps_per_second": 1.144,
-      "eval_wer": 0.1440015087155263,
+      "epoch": 3.31,
+      "learning_rate": 0.00019005736137667304,
+      "loss": 0.1217,
       "step": 2800
     },
     {
-      "epoch": 6.87,
-      "learning_rate": 4.972118959107806e-05,
-      "loss": 0.0793,
-      "step": 3200
+      "epoch": 3.56,
+      "eval_cer": 0.034948700511516516,
+      "eval_loss": 0.13004331290721893,
+      "eval_runtime": 66.7534,
+      "eval_samples_per_second": 7.49,
+      "eval_steps_per_second": 0.944,
+      "eval_wer": 0.11176370926752131,
+      "step": 3015
     },
     {
-      "epoch": 6.87,
-      "eval_loss": Infinity,
-      "eval_runtime": 816.2609,
-      "eval_samples_per_second": 9.144,
-      "eval_steps_per_second": 1.143,
-      "eval_wer": 0.139223909548912,
+      "epoch": 3.78,
+      "learning_rate": 0.00017093690248565967,
+      "loss": 0.1121,
       "step": 3200
     },
     {
-      "epoch": 7.72,
-      "learning_rate": 1.2639405204460965e-05,
-      "loss": 0.0642,
-      "step": 3600
+      "epoch": 3.95,
+      "eval_cer": 0.03565831879601431,
+      "eval_loss": 0.1295933872461319,
+      "eval_runtime": 67.056,
+      "eval_samples_per_second": 7.456,
+      "eval_steps_per_second": 0.94,
+      "eval_wer": 0.11729824363246173,
+      "step": 3350
     },
     {
-      "epoch": 7.72,
-      "eval_loss": Infinity,
-      "eval_runtime": 815.052,
-      "eval_samples_per_second": 9.158,
-      "eval_steps_per_second": 1.145,
-      "eval_wer": 0.1354611011827252,
+      "epoch": 4.25,
+      "learning_rate": 0.00015181644359464624,
+      "loss": 0.1038,
       "step": 3600
+    },
+    {
+      "epoch": 4.35,
+      "eval_cer": 0.034948700511516516,
+      "eval_loss": 0.13280533254146576,
+      "eval_runtime": 66.5598,
+      "eval_samples_per_second": 7.512,
+      "eval_steps_per_second": 0.947,
+      "eval_wer": 0.11080363697972553,
+      "step": 3685
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.00013269598470363287,
+      "loss": 0.0941,
+      "step": 4000
+    },
+    {
+      "epoch": 4.75,
+      "eval_cer": 0.035214807368203184,
+      "eval_loss": 0.12650151550769806,
+      "eval_runtime": 66.4767,
+      "eval_samples_per_second": 7.521,
+      "eval_steps_per_second": 0.948,
+      "eval_wer": 0.1124978821934828,
+      "step": 4020
+    },
+    {
+      "epoch": 5.14,
+      "eval_cer": 0.034012398608359695,
+      "eval_loss": 0.1327013224363327,
+      "eval_runtime": 66.3907,
+      "eval_samples_per_second": 7.531,
+      "eval_steps_per_second": 0.949,
+      "eval_wer": 0.10718924719037669,
+      "step": 4355
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 0.0001135755258126195,
+      "loss": 0.0862,
+      "step": 4400
+    },
+    {
+      "epoch": 5.54,
+      "eval_cer": 0.0352246631777101,
+      "eval_loss": 0.1414576768875122,
+      "eval_runtime": 66.6897,
+      "eval_samples_per_second": 7.497,
+      "eval_steps_per_second": 0.945,
+      "eval_wer": 0.11294968091715141,
+      "step": 4690
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 9.44550669216061e-05,
+      "loss": 0.0748,
+      "step": 4800
+    },
+    {
+      "epoch": 5.93,
+      "eval_cer": 0.03466288203581601,
+      "eval_loss": 0.13079801201820374,
+      "eval_runtime": 66.6475,
+      "eval_samples_per_second": 7.502,
+      "eval_steps_per_second": 0.945,
+      "eval_wer": 0.1101824137346812,
+      "step": 5025
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 7.533460803059272e-05,
+      "loss": 0.0727,
+      "step": 5200
+    },
+    {
+      "epoch": 6.33,
+      "eval_cer": 0.03377585918019377,
+      "eval_loss": 0.1386057287454605,
+      "eval_runtime": 66.5005,
+      "eval_samples_per_second": 7.519,
+      "eval_steps_per_second": 0.947,
+      "eval_wer": 0.10470435421019936,
+      "step": 5360
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 5.621414913957934e-05,
+      "loss": 0.0622,
+      "step": 5600
+    },
+    {
+      "epoch": 6.72,
+      "eval_cer": 0.0333717709904103,
+      "eval_loss": 0.14488892257213593,
+      "eval_runtime": 66.3875,
+      "eval_samples_per_second": 7.532,
+      "eval_steps_per_second": 0.949,
+      "eval_wer": 0.10509967809340938,
+      "step": 5695
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 3.7093690248565965e-05,
+      "loss": 0.0582,
+      "step": 6000
+    },
+    {
+      "epoch": 7.12,
+      "eval_cer": 0.03320422222879276,
+      "eval_loss": 0.14533209800720215,
+      "eval_runtime": 66.3474,
+      "eval_samples_per_second": 7.536,
+      "eval_steps_per_second": 0.95,
+      "eval_wer": 0.10493025357203366,
+      "step": 6030
+    },
+    {
+      "epoch": 7.51,
+      "eval_cer": 0.03320422222879276,
+      "eval_loss": 0.14391696453094482,
+      "eval_runtime": 66.4699,
+      "eval_samples_per_second": 7.522,
+      "eval_steps_per_second": 0.948,
+      "eval_wer": 0.10515615293386796,
+      "step": 6365
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 1.8021032504780113e-05,
+      "loss": 0.0493,
+      "step": 6400
+    },
+    {
+      "epoch": 7.91,
+      "eval_cer": 0.03264244108689867,
+      "eval_loss": 0.14469173550605774,
+      "eval_runtime": 65.8617,
+      "eval_samples_per_second": 7.592,
+      "eval_steps_per_second": 0.957,
+      "eval_wer": 0.10267125995369063,
+      "step": 6700
     }
   ],
-  "max_steps": 3728,
+  "max_steps": 6776,
   "num_train_epochs": 8,
-  "total_flos": 4.913815144448316e+19,
+  "total_flos": 2.1224029808181802e+20,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e83dad711cbc55efc0a8db3d5cb03c651f7fc58cffbacaf147fa17c9a500477c
-size 2799
+oid sha256:903a958bf893998150c6891a9f369853f085dd31bfb492b34bf63bb942ca1cc7
+size 2991

vocab.json

@@ -1 +1,3 @@
-{"?": 1, "a": 2, "b": 3, "c": 4, "d": 5, "e": 6, "f": 7, "g": 8, "h": 9, "i": 10, "j": 11, "k": 12, "l": 13, "m": 14, "n": 15, "o": 16, "p": 17, "q": 18, "r": 19, "s": 20, "t": 21, "u": 22, "v": 23, "w": 24, "x": 25, "y": 26, "z": 27, "\u00e4": 28, "\u00fc": 29, "\u0107": 30, "\u010d": 31, "\u0111": 32, "\u0161": 33, "\u017e": 34, "|": 0, "[UNK]": 35, "[PAD]": 36}
+{" ": 0, "[PAD]": 1, "[UNK]": 2, "a": 3, "b": 4, "c": 5, "d": 6, "e": 7, "f": 8, "g": 9, "h": 10, "i": 11, "j": 12, "k": 13, "l": 14, "m": 15, "n": 16, "o": 17, "p": 18, "q": 19, "r": 20, "s": 21, "t": 22, "u": 23, "v": 24, "w": 25, "x": 26, "y": 27, "z": 28, "\u00e4": 29, "\u00fc": 30, "\u0107": 31, "\u010d": 32, "\u0111": 33, "\u0161": 34, "\u017e": 35, "\u04e7": 36, 
+"1":37, "2": 38,"3": 39,"4": 40,"5": 41,"6": 42,"7": 43,"8": 44,"9": 45,"0": 46,
+".": 47 }