private-gpt
private-gpt copied to clipboard
zylon-ai
Interact with your documents using the power of GPT, 100% privately, no data leaks
This allows using OpenAI compatible embedding models such as [text-embeddings-inference](https://github.com/huggingface/text-embeddings-inference). Main changes: - added embedding_api_base, embedding_api_key, embedding_model to openai settings
### Discussed in https://github.com/zylon-ai/private-gpt/discussions/922 Originally posted by **lsimcoates** August 1, 2023 Hi, looking at whether it is feasible to supply private GPT with a text file which contains a list...
contrary to other files types like pdf or pptx, images are not ingested and no error popup. Any idea? tks
``` (privategpt) PS D:\AGI\privategpt> $env:PGPT_PROFILES="ollama"; make run poetry run python -m private_gpt 21:29:52.426 [INFO ] private_gpt.settings.settings_loader - Starting application with profiles=['default', 'ollama'] 21:29:53.085 [INFO ] numexpr.utils - Note: NumExpr detected...
This commit fixes the powershell set up in the documentation in order to ease the setup process.
I am getting very bad responses from the LLM when chatting with the file I uploaded. I am using Qdrant as a vector DB and I checked the contents in...
Uploading a small PDF appears to succeed (no errors reported) but UI doesn't reflect the uploaded file and it can't be queried. Running openAI settings, here are the ingestion results:...
The "make wipe" command does not reset the qdrant database - therefor all previous ingested documents are still present in the model Suggestions for a short term fix ? BR...
I'm running PGPT in docker and following the documentation here to create an "injestion" folder: https://docs.privategpt.dev/manual/document-management/ingestion **After** I run the container I able to get the following command to work...
