初始化项目，由ModelHub XC社区提供模型

Model: dysata/Wav2Vec2-Ru-Child
Source: Original Platform

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,130 @@
+---
+language:
+- ru
+license: mit
+library_name: transformers
+pipeline_tag: automatic-speech-recognition
+tags:
+- wav2vec2
+- speech
+- russian
+- children
+- ctc
+- forced-alignment
+- pronunciation
+- phonetics
+datasets:
+- dysata/rwords
+model-index:
+- name: Wav2Vec2-Ru-Child
+  results: []
+---
+
+# Wav2Vec2-Ru-Child
+
+Модель автоматического распознавания речи (ASR) для русского языка, дообученная на записях детского чтения.
+
+## Model Details
+
+### Architecture
+
+- **Base model:** wav2vec2-large
+- **Architecture:** `Wav2Vec2ForCTC`
+- **Hidden size:** 1024
+- **Layers:** 24 transformer layers
+- **Attention heads:** 16
+- **Parameters:** ~317M
+- **Vocabulary:** 37 токенов (33 буквы русского алфавита + 4 служебных)
+- **CTC loss:** mean reduction
+
+### Intended Use
+
+Модель предназначена для:
+- Распознавания русской детской речи
+- Forced alignment (выравнивание текста по аудио на уровне букв)
+- Анализа произношения — выявление ошибок в детском чтении
+- Классификации качества произношения отдельных звуков (например, звука "Р")
+
+## How to Use
+
+### Speech Recognition
+
+```python
+import torch
+import librosa
+from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC
+
+processor = Wav2Vec2Processor.from_pretrained("dysata/Wav2Vec2-Ru-Child")
+model = Wav2Vec2ForCTC.from_pretrained("dysata/Wav2Vec2-Ru-Child")
+
+audio, sr = librosa.load("audio.wav", sr=16000)
+
+processed = processor([audio], sampling_rate=16000,
+                       return_tensors="pt", padding="longest")
+
+with torch.no_grad():
+    logits = model(processed.input_values,
+                   attention_mask=processed.attention_mask).logits
+
+predicted_ids = torch.argmax(logits, dim=-1)
+transcription = processor.decode(predicted_ids[0])
+print(transcription)
+```
+
+### Forced Alignment
+
+Модель может использоваться для побуквенного выравнивания эталонного текста по аудио через CTC forced alignment (trellis + backtrack + merge_repeats). Это позволяет определить временные границы каждой буквы в записи.
+
+### Hidden States для классификации
+
+```python
+with torch.no_grad():
+    outputs = model(processed.input_values,
+                    attention_mask=processed.attention_mask,
+                    output_hidden_states=True, return_dict=True)
+    last_hidden_state = outputs.hidden_states[-1]  # [batch, frames, 1024]
+```
+
+Вектора последнего скрытого слоя (1024-мерные) могут быть использованы как признаки для классификации качества произношения отдельных звуков.
+
+## Training
+
+Модель дообучена на записях детского чтения на русском языке. Аудиозаписи преобразованы в формат WAV 16 кГц и вручную оттранскрибированы.
+
+## Technical Specifications
+
+| Parameter | Value |
+|---|---|
+| Sample rate | 16 kHz |
+| Feature extractor | 7-layer CNN |
+| Transformer layers | 24 |
+| Hidden size | 1024 |
+| Vocab size | 37 |
+| Precision | float32 |
+| Format | Safetensors |
+
+## Vocabulary
+
+Алфавит модели: `<pad>`, `<s>`, `</s>`, `<unk>`, `|` (разделитель слов), а-я (33 буквы русского алфавита).
+
+## Limitations
+
+- Модель обучена на детской речи и может показывать худшие результаты на взрослой речи
+- Только русский язык
+- Оптимальное качество на записях в формате WAV 16 кГц
+
+## Citation
+
+```bibtex
+@misc{wav2vec2-ru-child,
+  author = {Павел Рудич},
+  title = {Wav2Vec2-Ru-Child: Russian Children's Speech Recognition Model},
+  year = {2025},
+  publisher = {Hugging Face},
+  url = {https://huggingface.co/dysata/Wav2Vec2-Ru-Child}
+}
+```
+
+## Funding
+
+Фонд содействия инновациям (fasie).

added_tokens.json

@@ -0,0 +1,4 @@
+{
+  "[PAD]": 38,
+  "[UNK]": 37
+}

config.json

@@ -0,0 +1,116 @@
+{
+  "_name_or_path": "./model6m",
+  "activation_dropout": 0.0,
+  "adapter_attn_dim": null,
+  "adapter_kernel_size": 3,
+  "adapter_stride": 2,
+  "add_adapter": false,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "classifier_proj_size": 256,
+  "codevector_dim": 768,
+  "contrastive_logits_temperature": 0.1,
+  "conv_bias": true,
+  "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": "mean",
+  "ctc_zero_infinity": true,
+  "diversity_loss_weight": 0.1,
+  "do_stable_layer_norm": true,
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_dropout": 0.0,
+  "feat_extract_norm": "layer",
+  "feat_proj_dropout": 0.0,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.0,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.0,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.0,
+  "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_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_min_space": 1,
+  "mask_time_other": 0.0,
+  "mask_time_prob": 0.05,
+  "mask_time_selection": "static",
+  "model_type": "wav2vec2",
+  "num_adapter_layers": 3,
+  "num_attention_heads": 16,
+  "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_negatives": 100,
+  "output_hidden_size": 1024,
+  "pad_token_id": 0,
+  "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": "4.49.0",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 37,
+  "xvector_output_dim": 512
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f5b304a55228829a041a5aac128708c195d9c5280e6ab80ab6b7d217a6e6cba
+size 1261959180

preprocessor_config.json

@@ -0,0 +1,10 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "Wav2Vec2Processor",
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

special_tokens_map.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

tokenizer_config.json

@@ -0,0 +1,49 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<pad>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "do_lower_case": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<pad>",
+  "processor_class": "Wav2Vec2Processor",
+  "replace_word_delimiter_char": " ",
+  "target_lang": null,
+  "tokenizer_class": "Wav2Vec2CTCTokenizer",
+  "unk_token": "<unk>",
+  "word_delimiter_token": "|"
+}

vocab.json

@@ -0,0 +1,39 @@
+{
+  "</s>": 2,
+  "<pad>": 0,
+  "<s>": 1,
+  "<unk>": 3,
+  "|": 4,
+  "а": 5,
+  "б": 6,
+  "в": 7,
+  "г": 8,
+  "д": 9,
+  "е": 10,
+  "ж": 11,
+  "з": 12,
+  "и": 13,
+  "й": 14,
+  "к": 15,
+  "л": 16,
+  "м": 17,
+  "н": 18,
+  "о": 19,
+  "п": 20,
+  "р": 21,
+  "с": 22,
+  "т": 23,
+  "у": 24,
+  "ф": 25,
+  "х": 26,
+  "ц": 27,
+  "ч": 28,
+  "ш": 29,
+  "щ": 30,
+  "ъ": 31,
+  "ы": 32,
+  "ь": 33,
+  "э": 34,
+  "ю": 35,
+  "я": 36
+}