[BUG]: Multiple download of Xenova/all-MiniLM-L6-v2 on initial embed #1262
Description
How are you running AnythingLLM?
AnythingLLM desktop app
What happened?
Embedding the first document will download the the embedding model Xenova/all-MiniLM-L6-v2 for each chunk instead of only once. See Log:
[NativeEmbedder] Embedded Chunk 245 of 248
[NativeEmbedder] The native embedding model has never been run and will be downloaded right now. Subsequent runs will be faster. (~23MB)
[NativeEmbedder] Downloading Xenova/all-MiniLM-L6-v2 from https://huggingface.co/
[NativeEmbedder - Downloading model] tokenizer_config.json 100%
[NativeEmbedder - Downloading model] config.json 100%
[NativeEmbedder - Downloading model] tokenizer.json 100%
[NativeEmbedder - Downloading model] onnx/model_quantized.onnx 100%
[NativeEmbedder] Embedded Chunk 246 of 248
[NativeEmbedder] The native embedding model has never been run and will be downloaded right now. Subsequent runs will be faster. (~23MB)
[NativeEmbedder] Downloading Xenova/all-MiniLM-L6-v2 from https://huggingface.co/
[NativeEmbedder - Downloading model] tokenizer_config.json 100%
[NativeEmbedder - Downloading model] config.json 100%
[NativeEmbedder - Downloading model] tokenizer.json 100%
[NativeEmbedder - Downloading model] onnx/model_quantized.onnx 100%
[NativeEmbedder] Embedded Chunk 247 of 248
[NativeEmbedder] The native embedding model has never been run and will be downloaded right now. Subsequent runs will be faster. (~23MB)
[NativeEmbedder] Downloading Xenova/all-MiniLM-L6-v2 from https://huggingface.co/
[NativeEmbedder - Downloading model] tokenizer_config.json 100%
[NativeEmbedder - Downloading model] config.json 100%
[NativeEmbedder - Downloading model] tokenizer.json 100%
Subsequent embeddings use the existing modell.
Are there known steps to reproduce?
- Install Desktop App for Windows from the official website.
- Start embedding of a big file.
- View the logs.