Training in progress, step 50

config.json

@@ -1,6 +1,6 @@
 {
-  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
-  "activation_dropout": 0.0,
+  "_name_or_path": "nguyenvulebinh/wav2vec2-base-vietnamese-250h",
+  "activation_dropout": 0.1,
   "adapter_kernel_size": 3,
   "adapter_stride": 2,
   "add_adapter": false,
@@ -8,12 +8,12 @@
   "architectures": [
     "Wav2Vec2ForCTC"
   ],
-  "attention_dropout": 0.0,
+  "attention_dropout": 0.05,
   "bos_token_id": 1,
   "classifier_proj_size": 256,
-  "codevector_dim": 768,
+  "codevector_dim": 256,
   "contrastive_logits_temperature": 0.1,
-  "conv_bias": true,
+  "conv_bias": false,
   "conv_dim": [
     512,
     512,
@@ -44,20 +44,20 @@
   "ctc_loss_reduction": "mean",
   "ctc_zero_infinity": false,
   "diversity_loss_weight": 0.1,
-  "do_stable_layer_norm": true,
+  "do_stable_layer_norm": false,
   "eos_token_id": 2,
   "feat_extract_activation": "gelu",
   "feat_extract_dropout": 0.0,
-  "feat_extract_norm": "layer",
+  "feat_extract_norm": "group",
   "feat_proj_dropout": 0.0,
   "feat_quantizer_dropout": 0.0,
-  "final_dropout": 0.0,
-  "gradient_checkpointing": false,
+  "final_dropout": 0.1,
   "hidden_act": "gelu",
-  "hidden_dropout": 0.0,
-  "hidden_size": 1024,
+  "hidden_dropout": 0.05,
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
   "initializer_range": 0.02,
-  "intermediate_size": 4096,
+  "intermediate_size": 3072,
   "layer_norm_eps": 1e-05,
   "layerdrop": 0.0,
   "mask_feature_length": 10,
@@ -68,17 +68,17 @@
   "mask_time_prob": 0.05,
   "model_type": "wav2vec2",
   "num_adapter_layers": 3,
-  "num_attention_heads": 16,
+  "num_attention_heads": 12,
   "num_codevector_groups": 2,
   "num_codevectors_per_group": 320,
   "num_conv_pos_embedding_groups": 16,
   "num_conv_pos_embeddings": 128,
   "num_feat_extract_layers": 7,
-  "num_hidden_layers": 24,
+  "num_hidden_layers": 12,
   "num_negatives": 100,
-  "output_hidden_size": 1024,
-  "pad_token_id": 90,
-  "proj_codevector_dim": 768,
+  "output_hidden_size": 768,
+  "pad_token_id": 109,
+  "proj_codevector_dim": 256,
   "tdnn_dilation": [
     1,
     2,
@@ -101,8 +101,8 @@
     1
   ],
   "torch_dtype": "float32",
-  "transformers_version": "4.20.1",
+  "transformers_version": "4.16.0",
   "use_weighted_layer_sum": false,
-  "vocab_size": 93,
+  "vocab_size": 110,
   "xvector_output_dim": 512
 }

preprocessor_config.json

@@ -4,6 +4,6 @@
   "feature_size": 1,
   "padding_side": "right",
   "padding_value": 0.0,
-  "return_attention_mask": true,
-  "sampling_rate": 44100
+  "return_attention_mask": false,
+  "sampling_rate": 16000
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f5cc9c7b326bc4c29630b031081d1248cd58605d8dffd9098b6251b71669b464
-size 1262279985
+oid sha256:84a60a4bb6c7dcbd71c5d30990d487ad4efe4a530143b160563adefa2d1ce37a
+size 377911777

runs/Jan09_11-03-43_DESKTOP-DRI80NH/1673237554.060922/events.out.tfevents.1673237554.DESKTOP-DRI80NH.2432.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9bfdf7a4418b609ba0541b4518cb762323fa0bf9f7468b7aee52a51ce17402a
+size 4744

runs/Jan09_11-03-43_DESKTOP-DRI80NH/events.out.tfevents.1673237554.DESKTOP-DRI80NH.2432.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89478e59381a1e3bb3565821b18217e81412e94a4d8f190db02e19e7fa7a5a40
+size 5171

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f606955b8bd0f6cfd0be09dd1d9f46d3f94009b663db196ff5cc1a5c6cd5cc88
-size 3311
+oid sha256:00d7eb5cd5542599a6ac675a21095bb69ccfc8e669cf84da6164225e5902fdab
+size 3003