LlamaEdge-RAG 0.6.1

Major changes:

• Support embedding prompt template for embedding model

LlamaEdge-RAG 0.6.0

Major changes:

• Improve --batch-size CLI option: set batch size for chat and embedding models, respectively

LlamaEdge-RAG 0.5.3

Major changes:

• Add user header in chat completion responses
• Support PLUGIN_DEBUG wasm environment variable for debugging the low-level ggml plugin

LlamaEdge-RAG 0.5.2

Major changes:

• Llama-api-server
  • Improve error responses
  • Add content-type:application/json header in responses

LlamaEdge-RAG 0.5.1

Major change:

• Update /v1/embeddings endpoint to be compatible with OpenAI /v1/embeddings API

LlamaEdge-RAG 0.5.0

Major change:

• Update deps: llama-core v0.9.0, endpoints v0.8.0, and chat-prompts v0.7.1

LlamaEdge-RAG 0.4.0

Major changes:

• New /v1/retrieve endpoint
• New --rag-policy CLI option

LlamaEdge-RAG 0.3.8

major change: update the chat-prompts dep to 0.6.2

LlamaEdge-RAG 0.3.7

Major change:

• Post-process the generation of phi-3-chat model in non-stream mode.

LlamaEdge-RAG 0.3.6

Major changes:

• Post-process the generation of llama-2-chat and llama-3-chat models in non-stream mode.