add crud ragas evaluation.

[lkk12014402]

Description

add ragas evaluation for crud Chinese dataset.

[lkk12014402]

setup llm-as-judge on 4 cards

docker run -p 8010:80 --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e PT_HPU_ENABLE_LAZY_COLLECTIVES=true -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host ghcr.io/huggingface/tgi-gaudi:2.0.5 --model-id Qwen/Qwen2-72B-Instruct --max-input-tokens 4096 --max-total-tokens 8192 --sharded true --num-shard 4

evaluation

python eval_crud.py --dataset_path ../data/split_merged.json --docs_path ../data/80000_docs  --contain_original_data --llm_endpoint "http://localhost:8010"  --ragas_metrics