Upload model

README.md

@@ -0,0 +1,199 @@
+---
+library_name: transformers
+tags: []
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]

config.json

@@ -0,0 +1,16 @@
+{
+  "_name_or_path": "HKUSTAudio/xcodec2",
+  "architectures": [
+    "XCodec2Model"
+  ],
+  "auto_map": {
+    "AutoModel": "modeling_xcodec2.XCodec2Model"
+  },
+  "codec_decoder_hidden_size": 1024,
+  "codec_encoder_hidden_size": 1024,
+  "model_type": "xcodec",
+  "semantic_hidden_size": 1024,
+  "torch_dtype": "float32",
+  "transformers_version": "4.44.2",
+  "use_vocos": true
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b6b739f220c9e64e525c104afe77002067fe90ba94f4b4110aa6314195b7fcd
+size 3301612648

modeling_xcodec2.py

@@ -0,0 +1,165 @@
+import torch
+import torch.nn as nn
+from transformers import PreTrainedModel
+from configuration_bigcodec import BigCodecConfig
+
+# 请确保这些模块路径是正确的
+from vq.codec_encoder import CodecEncoder_Transformer
+from vq.codec_decoder_vocos import CodecDecoderVocos
+from vq.module import SemanticEncoder
+from transformers import AutoFeatureExtractor, Wav2Vec2BertModel
+import torch.nn.functional as F
+class XCodec2Model(PreTrainedModel):
+    config_class = BigCodecConfig
+
+    def __init__(self, config: BigCodecConfig):
+        super().__init__(config)
+
+        # 1) 语义模型
+        self.semantic_model = Wav2Vec2BertModel.from_pretrained(
+            "facebook/w2v-bert-2.0",
+            output_hidden_states=True
+        )
+        self.semantic_model.eval()
+
+        self.SemanticEncoder_module = SemanticEncoder(
+             config.semantic_hidden_size,
+             config.semantic_hidden_size,
+             config.semantic_hidden_size
+        )
+
+        # 2) Codec Encoder
+        self.CodecEnc = CodecEncoder_Transformer()
+
+        # 3) Codec Decoder
+        self.generator = CodecDecoderVocos()  
+
+        # 4) 两个全连接层
+        self.fc_prior = nn.Linear(2048, 2048)
+        self.fc_post_a = nn.Linear(2048, 1024)
+        feature_extractor = AutoFeatureExtractor.from_pretrained("facebook/w2v-bert-2.0")
+        self.feature_extractor = feature_extractor
+        self.avg_pooler = nn.AvgPool1d(2, stride=2)
+
+    def forward(self, input_waveform, input_features=None, sample_rate=16000):
+        """
+        这里的 forward 不一定要叫 forward，也可以拆成别的方法；
+        但是如果想兼容 pipeline，需要在 forward 里给出核心逻辑。
+        
+        参数：
+          input_waveform: [batch_size, waveform_length]
+          sample_rate: 默认 16000
+        返回：
+          重构后的语音音频 (Tensor)
+        """
+        # 1) 特征提取
+        # 如果需要 padding，可以在这里做
+        wav = input_waveform 
+        with torch.no_grad():
+          if input_features is None:
+              pad_for_wav = (320 - (wav.shape[1] % 320))
+              wav = torch.nn.functional.pad(wav, (0, pad_for_wav))
+              padded = F.pad(wav, (160, 160, 0, 0))
+              input_features = self.feature_extractor(
+                  padded.cpu().numpy(),
+                  sampling_rate=sample_rate, 
+                  return_tensors="pt"
+              ).input_features.to(self.device)
+
+        # 2) 语义层
+        semantic_output = self.semantic_model(input_features)
+        semantic_hidden_16 = semantic_output.hidden_states[16]  # 取第16层
+        semantic_hidden_16 = semantic_hidden_16.transpose(1, 2)  # [batch, hidden_dim, frames]
+        semantic_encoded = self.SemanticEncoder_module(semantic_hidden_16)
+
+        # 3) codec encoder
+        wav = wav.to(self.device)  # shape: [batch, 1, time]
+        vq_emb = self.CodecEnc(wav.unsqueeze(1))  # [batch, time//down, 1024] 只是示例
+        vq_emb = vq_emb.transpose(1, 2)  # -> [batch, 1024, frames]
+
+        concat_emb = torch.cat([semantic_encoded, vq_emb], dim=1)  # [batch, 1024 + 1024, frames]
+
+        # 5) fc_prior
+        concat_emb = self.fc_prior(concat_emb.transpose(1, 2)).transpose(1, 2)
+        concat_emb = self.avg_pooler(concat_emb)
+
+        # 6) decoder 的量化部分
+        _, vq_code, _ = self.generator(concat_emb, vq=True)
+        vq_post_emb = self.generator.quantizer.get_output_from_indices(vq_code.transpose(1, 2))
+        vq_post_emb = vq_post_emb.transpose(1, 2)
+
+        # 7) fc_post_a
+        vq_post_emb = self.fc_post_a(vq_post_emb.transpose(1, 2)).transpose(1, 2)
+
+        # 8) 最后解码成波形
+        recon_audio = self.generator(vq_post_emb.transpose(1, 2), vq=False)[0]
+        # recon_audio: [batch, time]
+        return recon_audio
+
+    def encode_code(self, input_waveform, sample_rate=16000):
+        """
+        将输入的音频编码为代码表示。
+
+        参数：
+          input_waveform: [batch_size, waveform_length]
+          sample_rate: 默认 16000
+        返回：
+          编码后的代码 (Tensor)
+        """
+        with torch.no_grad():
+ 
+    
+            wav = input_waveform 
+            pad_for_wav = (320 - (wav.shape[1] % 320))
+    
+            wav = torch.nn.functional.pad(wav, (0, pad_for_wav))
+
+            input_features = self.feature_extractor(
+                F.pad(wav[0,:].cpu(), (160, 160)), 
+                sampling_rate=sample_rate, 
+                return_tensors="pt"
+            ).input_features.to(self.device)  # [batch, frames, feat_dim]
+
+            # 2) 语义层
+            semantic_output = self.semantic_model(input_features)
+            semantic_hidden_16 = semantic_output.hidden_states[16]  # 取第16层
+            semantic_hidden_16 = semantic_hidden_16.transpose(1, 2)  # [batch, hidden_dim, frames]
+            semantic_encoded = self.SemanticEncoder_module(semantic_hidden_16)
+
+            # 3) codec encoder
+            wav = wav.to(self.device)  # shape: [batch, 1, time]
+            vq_emb = self.CodecEnc(wav.unsqueeze(1))  # [batch, time//down, 1024] 只是示例
+            vq_emb = vq_emb.transpose(1, 2)  # -> [batch, 1024, frames]
+
+            # 4) 拼接
+            concat_emb = torch.cat([semantic_encoded, vq_emb], dim=1)  # [batch, 2048, frames]
+
+            # 5) fc_prior
+            concat_emb = self.fc_prior(concat_emb.transpose(1, 2)).transpose(1, 2)
+
+            # 6) decoder 的量化部分，获取code
+            concat_emb = self.avg_pooler(concat_emb)
+            _, vq_code, _ = self.generator(concat_emb, vq=True)
+            # vq_code: [batch, frames]
+            return vq_code
+
+    def decode_code(self, vq_code):
+        """
+        将编码后的代码解码回音频。
+
+        参数：
+          vq_code: 编码后的代码 (Tensor) [batch, frames]
+        返回：
+          解码后的音频 (Tensor) [batch, waveform_length]
+        """
+        with torch.no_grad():
+            # 获取量化后的嵌入
+            vq_post_emb = self.generator.quantizer.get_output_from_indices(vq_code.transpose(1, 2))
+            vq_post_emb = vq_post_emb.transpose(1, 2)  # [batch, 1024, frames]
+
+            # 7) fc_post_a
+            vq_post_emb = self.fc_post_a(vq_post_emb.transpose(1, 2)).transpose(1, 2)  # [batch, 1024, frames]
+
+            # 8) 最后解码成波形
+            recon_audio = self.generator(vq_post_emb.transpose(1, 2), vq=False)[0]  # [batch, time]
+            return recon_audio