-
support websockets (inspired by portswigger's use of them in their LLM tutorial material, https://portswigger.net/web-security/all-labs#web-llm-attacks)
-
Motivation: Evaluate the effectiveness of the proposed framework
- prompt:
- system instruction + real-time state info / changes + few-shot
- multi-modal:
- cam + lidar + pos + historica…
-
Problem : AzureOpenAi model should use FinishReason.LENGTH in case of token length limit reached.
If we do a request with too many tokens, then Azure Open AI answer with a http status 400 and a jso…
fb33 updated
19 hours ago
-
### Bug Description
I don't know if this is a bug or I use it wrong. I have a local self-signed certificate, so I have to modify the `http_client` in `llm`. Later if I use `QueryFusionRetriever` with…
-
New data will be harvested periodically which will somehow be reflected in the NLQ component. E.g. it might need to updating of embeddings in a vector store.
New versions of LLMs are made available…
-
_(Ed note, original issue title was: **Prevention of Prompt Injection in Applications Using Large Language Models (LLM)**)_
The popularity of Large Language Models (LLM) like GPT variants from Open…
-
**Title:** Extend Chirps to Scan LLM APIs for Security Issues
**Description:**
Chirps currently provides functionality to scan next-generation AI systems and checks for issues with the vectorDB.…
-
I meet this issue while using ollama on MTL iGPU
![image](https://github.com/intel-analytics/ipex-llm/assets/92354341/b9cc7b61-3b61-4615-b1f2-40a85ac22aee)
my IPEX-LLM version as below
![image](htt…
-
### The Feature
For security, system engineer need to configure APIM setting on endpoint of Azure OpenAI Service.
Some times, Engineer need to assign custom http header and certificate of client aut…
-
### Describe the bug
We encountered a bug when serving Xinference (since 0.8.5 & through docker) through an apache vhost with reverse proxy configuration. Issue is the dynamic resolution of paths (th…