Configure model and feature extractor

config.json

@@ -0,0 +1,85 @@
+{
+    "activation_dropout": 0.0,
+    "apply_spec_augment": true,
+    "architectures": [
+        "Wav2Vec2ForPreTraining"
+    ],
+    "attention_dropout": 0.1,
+    "bos_token_id": 1,
+    "codevector_dim": 256,
+    "contrastive_logits_temperature": 0.1,
+    "conv_bias": false,
+    "conv_dim": [
+        512,
+        512,
+        512,
+        512,
+        512,
+        512,
+        512
+    ],
+    "conv_kernel": [
+        10,
+        3,
+        3,
+        3,
+        3,
+        2,
+        2
+    ],
+    "conv_stride": [
+        5,
+        2,
+        2,
+        2,
+        2,
+        2,
+        2
+    ],
+    "ctc_loss_reduction": "sum",
+    "ctc_zero_infinity": false,
+    "diversity_loss_weight": 0.1,
+    "do_stable_layer_norm": true,
+    "eos_token_id": 2,
+    "feat_extract_activation": "gelu",
+    "feat_extract_norm": "layer",
+    "feat_proj_dropout": 0.1,
+    "feat_quantizer_dropout": 0.0,
+    "final_dropout": 0.0,
+    "freeze_feat_extract_train": true,
+    "gradient_checkpointing": true,
+    "hidden_act": "gelu",
+    "hidden_dropout": 0.1,
+    "hidden_size": 768,
+    "initializer_range": 0.02,
+    "intermediate_size": 3072,
+    "layer_norm_eps": 1e-05,
+    "layerdrop": 0.05,
+    "mask_channel_length": 10,
+    "mask_channel_min_space": 1,
+    "mask_channel_other": 0.0,
+    "mask_channel_prob": 0.0,
+    "mask_channel_selection": "static",
+    "mask_feature_length": 10,
+    "mask_feature_prob": 0.0,
+    "mask_time_length": 10,
+    "mask_time_min_space": 1,
+    "mask_time_other": 0.0,
+    "mask_time_prob": 0.05,
+    "mask_time_selection": "static",
+    "model_type": "wav2vec2",
+    "no_mask_channel_overlap": false,
+    "no_mask_time_overlap": false,
+    "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": 12,
+    "num_negatives": 100,
+    "pad_token_id": 0,
+    "proj_codevector_dim": 256,
+    "transformers_version": "4.9.0.dev0",
+    "vocab_size": 32
+}

config.py

@@ -0,0 +1,21 @@
+"""
+Create the model configuration (`config.json`) and the feature extractor configuration
+(`preprocessor_config.json`).
+"""
+from transformers import Wav2Vec2Config, Wav2Vec2FeatureExtractor
+
+model_dir="./"
+
+config = Wav2Vec2Config.from_pretrained(
+    "facebook/wav2vec2-base", 
+    mask_time_length=10,
+    mask_time_prob=0.05,
+    diversity_loss_weight=0.1,
+    num_negatives=100,
+    do_stable_layer_norm=True,
+    feat_extract_norm="layer",
+)
+config.save_pretrained(model_dir)
+
+config = Wav2Vec2FeatureExtractor.from_pretrained("facebook/wav2vec2-base", return_attention_mask=True)
+config.save_pretrained(model_dir)

preprocessor_config.json

@@ -0,0 +1,9 @@
+{
+    "do_normalize": true,
+    "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+    "feature_size": 1,
+    "padding_side": "right",
+    "padding_value": 0.0,
+    "return_attention_mask": true,
+    "sampling_rate": 16000
+}