LlamaEdge-RAG 0.8.1

Major changes:

• Improve RagPromptBuilder

• Remove the bindings between --main-gpu and --tensor-split CLI options

• Update to llama-core v0.14.1, chat-prompts v0.11.1, and endpoints v0.12.0.

LlamaEdge-RAG 0.8.0

Major changes:

• Support Meta-Llama-3.1-Instruct and internlm2.5-7b-chat tool use

  • New prompt template: llama-3-tool and internlm-2-tool

• Add --main-gpu and --tensor-split CLI options

• Update to llama-core v0.14.0, chat-prompts v0.11.0, and endpoints v0.11.1.

LlamaEdge-RAG 0.7.5

Major changes:

• Update to llama-core v0.13.1, chat-prompts v0.10.2, and endpoints v0.11.0.

LlamaEdge-RAG 0.7.4

Major changes:

• Update to llama-core v0.13.0, chat-prompts v0.10.1, and endpoints v0.10.2.

LlamaEdge-RAG 0.7.3

Major changes:

• Extend /v1/files endpoint

  • List files on server
  • Retrieve a specific file
  • Delete a specific file

• Update to llama-core v0.12.1, chat-prompts v0.10.0, and endpoints v0.10.1.

LlamaEdge-RAG 0.7.2

Major changes:

• Improve rag_query_handler
• Update deps
  • llama-core v0.12.0
  • chat-prompts v0.9.0
  • endpoints v0.10.0

LlamaEdge-RAG 0.7.1

Major changes:

• Update deps
  • endpoints v0.9.1
  • chat-prompts v0.8.2
  • llama-core v0.11.4

LlamaEdge-RAG 0.7.0

Major change:

• (BREAKING) Update the response data of the /v1/info endpoint

LlamaEdge-RAG 0.6.6

Major change:

• Work around DNS failure in some containers by updating the default value of qdrant-url CLI option from http://localhost:6333 to http://127.0.0.1:6333

LlamaEdge-RAG 0.6.5

Major change:

• Log the buffer of rag query request body while triggering bad request errors.