Support for Audio Files (#1416)

embedchain/chunkers/audio.py

@@ -0,0 +1,22 @@
+from typing import Optional
+
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+
+from embedchain.chunkers.base_chunker import BaseChunker
+from embedchain.config.add_config import ChunkerConfig
+from embedchain.helpers.json_serializable import register_deserializable
+
+
+@register_deserializable
+class AudioChunker(BaseChunker):
+    """Chunker for audio."""
+
+    def __init__(self, config: Optional[ChunkerConfig] = None):
+        if config is None:
+            config = ChunkerConfig(chunk_size=1000, chunk_overlap=0, length_function=len)
+        text_splitter = RecursiveCharacterTextSplitter(
+            chunk_size=config.chunk_size,
+            chunk_overlap=config.chunk_overlap,
+            length_function=config.length_function,
+        )
+        super().__init__(text_splitter)