Add model files

config.json

@@ -71,6 +71,6 @@
   "num_feat_extract_layers": 7,
   "num_hidden_layers": 24,
   "pad_token_id": 30,
-  "transformers_version": "4.4.2",
+  "transformers_version": "4.5.0.dev0",
   "vocab_size": 31
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9ab46384537a8ea6bbfd53bb2a4715eb721a0de52c3746a25aa3d5931eae1a37
+oid sha256:2e816e78ee7fd0e498f93f933dd0b18034841e0ba38042639fceb7d006be4270
 size 1262055847

trainer_state.json

@@ -1,240 +1,30 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 18.95583596214511,
-  "global_step": 8000,
+  "epoch": 1.1845425867507886,
+  "global_step": 500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.18,
-      "learning_rate": 0.00011848341232227489,
-      "loss": 5.8456,
+      "learning_rate": 0.0003470916966876807,
+      "loss": 1.587,
       "step": 500
     },
     {
       "epoch": 1.18,
-      "eval_loss": 2.693875551223755,
-      "eval_runtime": 1054.431,
-      "eval_samples_per_second": 4.905,
-      "eval_wer": 1.0,
+      "eval_loss": 0.2940497398376465,
+      "eval_runtime": 783.4997,
+      "eval_samples_per_second": 6.601,
+      "eval_wer": 0.5440472766257878,
       "step": 500
-    },
-    {
-      "epoch": 2.37,
-      "learning_rate": 0.00023696682464454977,
-      "loss": 0.5706,
-      "step": 1000
-    },
-    {
-      "epoch": 2.37,
-      "eval_loss": 0.19720512628555298,
-      "eval_runtime": 965.5246,
-      "eval_samples_per_second": 5.357,
-      "eval_wer": 0.4136704756426013,
-      "step": 1000
-    },
-    {
-      "epoch": 3.55,
-      "learning_rate": 0.0003554502369668247,
-      "loss": 0.2135,
-      "step": 1500
-    },
-    {
-      "epoch": 3.55,
-      "eval_loss": 0.17894236743450165,
-      "eval_runtime": 829.9805,
-      "eval_samples_per_second": 6.231,
-      "eval_wer": 0.3831812357817117,
-      "step": 1500
-    },
-    {
-      "epoch": 4.74,
-      "learning_rate": 0.00039178515007898897,
-      "loss": 0.1761,
-      "step": 2000
-    },
-    {
-      "epoch": 4.74,
-      "eval_loss": 0.1792241483926773,
-      "eval_runtime": 839.1507,
-      "eval_samples_per_second": 6.163,
-      "eval_wer": 0.3637789922338729,
-      "step": 2000
-    },
-    {
-      "epoch": 5.92,
-      "learning_rate": 0.0003786203264876251,
-      "loss": 0.1403,
-      "step": 2500
-    },
-    {
-      "epoch": 5.92,
-      "eval_loss": 0.16667363047599792,
-      "eval_runtime": 773.0615,
-      "eval_samples_per_second": 6.69,
-      "eval_wer": 0.3549407734748843,
-      "step": 2500
-    },
-    {
-      "epoch": 7.11,
-      "learning_rate": 0.00036545550289626125,
-      "loss": 0.1206,
-      "step": 3000
-    },
-    {
-      "epoch": 7.11,
-      "eval_loss": 0.16922971606254578,
-      "eval_runtime": 754.7522,
-      "eval_samples_per_second": 6.853,
-      "eval_wer": 0.35033862406191985,
-      "step": 3000
-    },
-    {
-      "epoch": 8.29,
-      "learning_rate": 0.00035229067930489733,
-      "loss": 0.1065,
-      "step": 3500
-    },
-    {
-      "epoch": 8.29,
-      "eval_loss": 0.17697516083717346,
-      "eval_runtime": 830.9983,
-      "eval_samples_per_second": 6.224,
-      "eval_wer": 0.34748842925502704,
-      "step": 3500
-    },
-    {
-      "epoch": 9.48,
-      "learning_rate": 0.00033912585571353347,
-      "loss": 0.0949,
-      "step": 4000
-    },
-    {
-      "epoch": 9.48,
-      "eval_loss": 0.17623142898082733,
-      "eval_runtime": 808.622,
-      "eval_samples_per_second": 6.396,
-      "eval_wer": 0.34633789190178593,
-      "step": 4000
-    },
-    {
-      "epoch": 10.66,
-      "learning_rate": 0.0003259610321221696,
-      "loss": 0.0834,
-      "step": 4500
-    },
-    {
-      "epoch": 10.66,
-      "eval_loss": 0.17101719975471497,
-      "eval_runtime": 805.7401,
-      "eval_samples_per_second": 6.419,
-      "eval_wer": 0.33004732892294014,
-      "step": 4500
-    },
-    {
-      "epoch": 11.85,
-      "learning_rate": 0.0003127962085308057,
-      "loss": 0.0781,
-      "step": 5000
-    },
-    {
-      "epoch": 11.85,
-      "eval_loss": 0.1730242371559143,
-      "eval_runtime": 801.4866,
-      "eval_samples_per_second": 6.453,
-      "eval_wer": 0.3337342781685537,
-      "step": 5000
-    },
-    {
-      "epoch": 13.03,
-      "learning_rate": 0.00029963138493944183,
-      "loss": 0.0702,
-      "step": 5500
-    },
-    {
-      "epoch": 13.03,
-      "eval_loss": 0.1908612698316574,
-      "eval_runtime": 817.2678,
-      "eval_samples_per_second": 6.328,
-      "eval_wer": 0.3248437622571451,
-      "step": 5500
-    },
-    {
-      "epoch": 14.22,
-      "learning_rate": 0.00028646656134807797,
-      "loss": 0.065,
-      "step": 6000
-    },
-    {
-      "epoch": 14.22,
-      "eval_loss": 0.19133996963500977,
-      "eval_runtime": 815.1424,
-      "eval_samples_per_second": 6.345,
-      "eval_wer": 0.3376304160238475,
-      "step": 6000
-    },
-    {
-      "epoch": 15.4,
-      "learning_rate": 0.00027330173775671406,
-      "loss": 0.0634,
-      "step": 6500
-    },
-    {
-      "epoch": 15.4,
-      "eval_loss": 0.17903825640678406,
-      "eval_runtime": 788.5559,
-      "eval_samples_per_second": 6.559,
-      "eval_wer": 0.32520984232408545,
-      "step": 6500
-    },
-    {
-      "epoch": 16.59,
-      "learning_rate": 0.0002601369141653502,
-      "loss": 0.0574,
-      "step": 7000
-    },
-    {
-      "epoch": 16.59,
-      "eval_loss": 0.2020258903503418,
-      "eval_runtime": 1572.1771,
-      "eval_samples_per_second": 3.29,
-      "eval_wer": 0.33182543210522186,
-      "step": 7000
-    },
-    {
-      "epoch": 17.77,
-      "learning_rate": 0.00024697209057398633,
-      "loss": 0.0539,
-      "step": 7500
-    },
-    {
-      "epoch": 17.77,
-      "eval_loss": 0.18831460177898407,
-      "eval_runtime": 887.0107,
-      "eval_samples_per_second": 5.831,
-      "eval_wer": 0.3172083780038177,
-      "step": 7500
-    },
-    {
-      "epoch": 18.96,
-      "learning_rate": 0.00023380726698262242,
-      "loss": 0.0518,
-      "step": 8000
-    },
-    {
-      "epoch": 18.96,
-      "eval_loss": 0.19699327647686005,
-      "eval_runtime": 927.4568,
-      "eval_samples_per_second": 5.577,
-      "eval_wer": 0.31524723478806577,
-      "step": 8000
     }
   ],
-  "max_steps": 16880,
-  "num_train_epochs": 40,
-  "total_flos": 3.909573475423295e+19,
+  "max_steps": 2110,
+  "num_train_epochs": 5,
+  "total_flos": 2.4510852111832704e+18,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d3ff2eca68c59215249e3d45e21b893cb00ca2fcf912465cd2d8b55458c36c22
+oid sha256:474caa22f1c7f0c4e89880f0d02b07015187ad5c2b1a354e1f8340ea7bbd6181
 size 2287

vocab.json

@@ -1 +1 @@
-{"c": 0, "\u00f1": 1, "f": 2, "i": 3, "m": 4, "a": 5, "w": 6, "n": 7, "g": 8, "e": 9, "t": 10, "d": 11, "o": 12, "v": 13, "j": 14, "z": 16, "b": 17, "h": 18, "k": 19, "p": 20, "q": 21, "r": 22, "l": 23, "y": 24, "s": 25, "x": 26, "\u00ed": 27, "u": 28, "|": 15, "[UNK]": 29, "[PAD]": 30}
+{"r": 0, "w": 1, "e": 2, "a": 3, "b": 4, "p": 5, "m": 7, "x": 8, "q": 9, "d": 10, "g": 11, "h": 12, "f": 13, "\u00ed": 14, "v": 15, "c": 16, "\u00f1": 17, "i": 18, "t": 19, "y": 20, "n": 21, "z": 22, "o": 23, "l": 24, "s": 25, "k": 26, "j": 27, "u": 28, "|": 6, "[UNK]": 29, "[PAD]": 30}