edcd7c9d6a
Signed-off-by: Mingyuan Qi <mingyuan.qi@intel.com> Co-authored-by: Copilot Autofix powered by AI <62310815+github-advanced-security[bot]@users.noreply.github.com>
47 lines
884 B
YAML
47 lines
884 B
YAML
# Copyright (C) 2024 Intel Corporation
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
|
|
# Model language for LLM
|
|
model_language: "Chinese"
|
|
splitter_name: "RecursiveCharacter"
|
|
k_rerank: 5
|
|
search_method: "similarity"
|
|
score_threshold: 0.5
|
|
bm25_weight: 0
|
|
|
|
# Pipeline
|
|
name: "default"
|
|
|
|
# Node parser
|
|
node_parser: "simple"
|
|
chunk_size: 192
|
|
chunk_overlap: 48
|
|
|
|
# Indexer
|
|
indexer: "faiss_vector"
|
|
|
|
# Retriever
|
|
retriever: "vectorsimilarity"
|
|
k_retrieval: 30
|
|
|
|
# Post Processor
|
|
postprocessor: "reranker"
|
|
|
|
# Generator
|
|
generator: "chatqna"
|
|
prompt_path: "./default_prompt.txt"
|
|
|
|
# Models
|
|
embedding_model_id: "BAAI/bge-small-en-v1.5"
|
|
# Device for embedding model inference
|
|
embedding_device: "AUTO"
|
|
|
|
rerank_model_id: "BAAI/bge-reranker-large"
|
|
# Device for reranking model inference
|
|
rerank_device: "AUTO"
|
|
|
|
llm_model_id: "Qwen/Qwen2-7B-Instruct"
|
|
llm_weights: "INT4"
|
|
# Device for LLM model inference
|
|
llm_device: "AUTO"
|