Upload 4 files
Browse files- app.py +1 -1
- requirements.txt +3 -1
app.py
CHANGED
@@ -17,7 +17,7 @@ GROUNDING_URLS = []
|
|
17 |
ACCESS_CODE = os.environ.get("SPACE_ACCESS_CODE", "")
|
18 |
ENABLE_DYNAMIC_URLS = True
|
19 |
ENABLE_VECTOR_RAG = True
|
20 |
-
ENABLE_WEB_SEARCH =
|
21 |
RAG_DATA = {"index_base64": "SXhGSYABAAABAAAAAAAAAAAAEAAAAAAAAAAQAAAAAAABAAAAAIABAAAAAAAAq+L8vO1iA71Ey6u7avEnPT/5tTxuLQs7hrgTPJKVCT18lxu9c9UbvZcz+rvtsBs987PLPUS2ET0OLH+73wgnPSK7kD2GeYC8eGOLPFzvBLzgT4m7MnJUPfAGgz33vSu9WGeRvBjTPD0UljK8hdCyvadptLsNVDO9ac2rPU/HPj2u+eo8k7wnPtDTjr0rCeq75DxvvHK0PL1mRkW9pwzevNMce71jGBu9KAuqvSdelz3RUZE9ZIRPPU7Bw73q+qS80jk+PT/soD06Rym+46qgvI/G5ju3uSs9+ooxvWbipjz5tos7xC2ZPCfAmzsa/6m8ID8IO3li8L0dUaE9jbO8vEdoKjyLjCG9PXKKvalusDwX26o8CH7Rvc6lcr1truM9IroGvfIZYj0+Sne9z62xPWCjbzyLoJq9ad5DPVVH+bwcUZ69o/HavKFA3jxtLcy7h3FZPfOwkbsbXts9ig2svRcWS7zXWcc888SUPFf/Ub3oIbM9NIrNO6OZUz3DDGE8m9ikPQ9/Cr3cJLk9Tj0wPFOCkD1Azow9v4poPDFJd71F2o+9QWM2PNM86Twxrsq8a5ORPYmX2L3psZC9502bPMyBG71r4aq9iai4u+3MZL0Xk+C7Y5LqvFePIDsgpIE8CTC7uSoonT3z54S8dluQvOsfpbzIs/48CKDrvL7oEQpmYpe8y77bvOjiRb2nQIE9qGLUu4W0qbz2MYW8iWivPV1qnr2ou468oSryvWF5rz1F5j+847vGPfmy/DxYAN080dbXvIahyTzXzNy82L32vOIHhD2i3hS95GNqPdzPmj3ZeyQ9zvw3vCvVeD3+0q+9SCM3POSd+bw9tse90uCCvRTrvLz0ay89UUHcPHDQPj28LAk8G9kQvT9pXL2zCIg8OPeVvQq6Er1Lk9c8DZorvREmS715Z/M8qkYQveSYdLzWZEU94UM4vVQHsTymWj89vG7tvCzcIT1RP7Q9BQ5mu+Jgrzx8x1y8nE0qPbsmRjxe0jW8wbEyvBC5mz0uih67vLkJvWrHIr0PutA6u1ZGPYqRyT2rN6Q9znmIPeI7Ez0/PxE9BPElvXFa5rytpfU7VexEva176L0rlkG81v0mPfKMIbybQLa9V2IEPUsazr17Thk9agSDvZn+gjwnUh6+N61iOyiZRr2e7c27albBPSeeGL2lUB68d+NYPcJg4IlqYLO8mHA/vXH9mLx3DUs9gfpgPK0Hi7wMU1u83oekPBRzcL1D0jC9ww7VvNwxiL3iR3o9nxAKvRJMp7yNWJw9Ve3cvJqxbL1FBGG9ifKqPeNSir38M6M8KJR4vdreOz0ZIiW8aCggvH8NQLqSATK9soGBvKo20LzKkEi8UhFBvf8NALv8L6M8FDRFPA898bxCOo09UomcvVixyzsAbEE9jRK6PBtlmT21tA69TgYjvQfESb3/Mow7XMjZvZgzVT1CbGU9SzDQvF7eUD1roz+9rPYhPF3AHL1jkuI8zkRnvXqyoL1JBY08/RaBPJAxBz3M6ne8DqkUvKIvnT2hRhg817rJPI/tuL2OGz+9PZbvPFYrob3MARm93w1IPQkCi7wRzRk9vBV8PaM7WT0MBIs8OBZVPcNUBL3pLdo8JGhEvTGSj7y6Z608dR6zPaKYkD2lErg7BlQbPqimq7ykz2C8U9nsvMQsVz12ChG9f5QIPDRRCLweOMA95fQ4PTBNOLPNAMm9wOguvAO8K71gWoi87ysTPRBC+Dv4UbA8t17rPRpRoL1mh8e8rh0NPf1FRbwo8eq8yKiMPLw4qz1qBlA8X40RPeuoGrzSyhY9k6fEO38dnj02gFI9xlEVPPO85z2/nGc8U2CzPHoLlrzeUtM949eDPTQGfr2HRPO89GoHPXtjtrz3dqa91VKlPaq5Kj3z7F08QHYIvGwJzr2rJBw8dX7zPLgyhDuZh+Y83CUUvRQfAz3BO308N/cNvD2wrb0W4ck6rccnPOlxpLybUpG90PS1Oy98yT0q83w8hom3u7K76bxdcJg8NPmZPTRWIb2QPeM7C/ehO9Gq3Dzef4I8", "chunks": {"8570c8c5": {"text": "Vector Database Test Document This is a test document for evaluating the vector database functionality. Section 1: Introduction to Vector Databases Vector databases store and query high-dimensional vector representations of data. They enable semantic search by finding vectors similar to a query vector in an embedding space. Section 2: Use Cases Common applications include: - Document retrieval and question answering - Similarity search for products or content - Recommendation systems - Semantic search in chatbots Section 3: Technical Implementation Vector databases typically use embedding models to convert text into dense vectors, then use algorithms like cosine similarity or approximate nearest neighbor search to find relevant results. Section 4: Benefits - Semantic understanding beyond keyword matching - Scalable retrieval for large document collections - Integration with modern AI systems and large language models - Support for multi-modal data (text, images, audio) This document should generate multiple chunks when processed by the system.", "metadata": {"file_path": "/private/var/folders/0m/_clrz0_d1tzf_fns8rxyy1jr0000gn/T/gradio/c4c745f9c7f069f694a492715df7f50d07f18cee76e93e198029acd8a6c38532/doc.txt", "file_name": "doc.txt", "chunk_index": 0, "start_word": 0, "word_count": 151}, "chunk_id": "8570c8c5"}}, "chunk_ids": ["8570c8c5"], "dimension": 384, "model_name": "all-MiniLM-L6-v2"}
|
22 |
|
23 |
# Get API key from environment - customizable variable name with validation
|
|
|
17 |
ACCESS_CODE = os.environ.get("SPACE_ACCESS_CODE", "")
|
18 |
ENABLE_DYNAMIC_URLS = True
|
19 |
ENABLE_VECTOR_RAG = True
|
20 |
+
ENABLE_WEB_SEARCH = True
|
21 |
RAG_DATA = {"index_base64": "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", "chunks": {"8570c8c5": {"text": "Vector Database Test Document This is a test document for evaluating the vector database functionality. Section 1: Introduction to Vector Databases Vector databases store and query high-dimensional vector representations of data. They enable semantic search by finding vectors similar to a query vector in an embedding space. Section 2: Use Cases Common applications include: - Document retrieval and question answering - Similarity search for products or content - Recommendation systems - Semantic search in chatbots Section 3: Technical Implementation Vector databases typically use embedding models to convert text into dense vectors, then use algorithms like cosine similarity or approximate nearest neighbor search to find relevant results. Section 4: Benefits - Semantic understanding beyond keyword matching - Scalable retrieval for large document collections - Integration with modern AI systems and large language models - Support for multi-modal data (text, images, audio) This document should generate multiple chunks when processed by the system.", "metadata": {"file_path": "/private/var/folders/0m/_clrz0_d1tzf_fns8rxyy1jr0000gn/T/gradio/c4c745f9c7f069f694a492715df7f50d07f18cee76e93e198029acd8a6c38532/doc.txt", "file_name": "doc.txt", "chunk_index": 0, "start_word": 0, "word_count": 151}, "chunk_id": "8570c8c5"}}, "chunk_ids": ["8570c8c5"], "dimension": 384, "model_name": "all-MiniLM-L6-v2"}
|
22 |
|
23 |
# Get API key from environment - customizable variable name with validation
|
requirements.txt
CHANGED
@@ -2,4 +2,6 @@ gradio>=5.35.0
|
|
2 |
requests>=2.32.3
|
3 |
beautifulsoup4>=4.12.3
|
4 |
faiss-cpu==1.7.4
|
5 |
-
numpy==1.24.3
|
|
|
|
|
|
2 |
requests>=2.32.3
|
3 |
beautifulsoup4>=4.12.3
|
4 |
faiss-cpu==1.7.4
|
5 |
+
numpy==1.24.3
|
6 |
+
crawl4ai>=0.2.0
|
7 |
+
aiohttp>=3.8.0
|