LLM_MODEL_TYPE=huggingface
# LLM_MODEL_TYPE=openai
# LLM_MODEL_TYPE=hftgi
# LLM_MODEL_TYPE=ollama
# LLM_MODEL_TYPE=google
# LLM_MODEL_TYPE=vllm

HUGGINGFACE_AUTH_TOKEN=

HFTGI_SERVER_URL=

OPENAI_API_KEY=

GOOGLE_API_KEY=

# if unset, default to "gpt-3.5-turbo"
OPENAI_MODEL_NAME=

# GEMINI_MODEL_NAME=gemini-1.5-pro-latest

# OLLAMA_MODEL_NAME=orca2:7b
# OLLAMA_MODEL_NAME=mistral:7b
# OLLAMA_MODEL_NAME=gemma:7b
# OLLAMA_MODEL_NAME=llama2:7b
OLLAMA_MODEL_NAME=llama3:8b

OLLAMA_RP=1.15
HF_RP=1.15

LANGCHAIN_DEBUG=false
BATCH_SIZE=1
APPLY_CHAT_TEMPLATE_FOR_RAG=true

# cpu, mps or cuda:0 - if unset, use whatever detected
HF_EMBEDDINGS_DEVICE_TYPE=
HF_PIPELINE_DEVICE_TYPE=

# uncomment one of the below to load corresponding quantized model
# LOAD_QUANTIZED_MODEL=4bit
# LOAD_QUANTIZED_MODEL=8bit

QA_WITH_RAG=true
# QA_WITH_RAG=false

RETRIEVER_TYPE=questions_file
# RETRIEVER_TYPE=vectorstore

QUESTIONS_FILE_PATH="./data/datasets/ms_macro.json"

DISABLE_MODEL_PRELOADING=true
CHAT_HISTORY_ENABLED=false
SHOW_PARAM_SETTINGS=false
SHARE_GRADIO_APP=false

# if unset, default to "hkunlp/instructor-xl"
HF_EMBEDDINGS_MODEL_NAME="hkunlp/instructor-large"

# number of cpu cores - used to set n_threads for GPT4ALL & LlamaCpp models
NUMBER_OF_CPU_CORES=

USING_TORCH_BFLOAT16=true

# HUGGINGFACE_MODEL_NAME_OR_PATH="databricks/dolly-v2-3b"
# HUGGINGFACE_MODEL_NAME_OR_PATH="databricks/dolly-v2-7b"
# HUGGINGFACE_MODEL_NAME_OR_PATH="databricks/dolly-v2-12b"
# HUGGINGFACE_MODEL_NAME_OR_PATH="TheBloke/wizardLM-7B-HF"
# HUGGINGFACE_MODEL_NAME_OR_PATH="TheBloke/vicuna-7B-1.1-HF"
# HUGGINGFACE_MODEL_NAME_OR_PATH="nomic-ai/gpt4all-j"
# HUGGINGFACE_MODEL_NAME_OR_PATH="nomic-ai/gpt4all-falcon"
# HUGGINGFACE_MODEL_NAME_OR_PATH="lmsys/fastchat-t5-3b-v1.0"
# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Llama-2-7b-chat-hf"
# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Llama-2-13b-chat-hf"
# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Llama-2-70b-chat-hf"
# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Meta-Llama-3-8B-Instruct"
# HUGGINGFACE_MODEL_NAME_OR_PATH="meta-llama/Meta-Llama-3-70B-Instruct"
# HUGGINGFACE_MODEL_NAME_OR_PATH="microsoft/Orca-2-7b"
# HUGGINGFACE_MODEL_NAME_OR_PATH="microsoft/Orca-2-13b"
HUGGINGFACE_MODEL_NAME_OR_PATH="google/gemma-1.1-2b-it"
# HUGGINGFACE_MODEL_NAME_OR_PATH="google/gemma-1.1-7b-it"
# HUGGINGFACE_MODEL_NAME_OR_PATH="microsoft/Phi-3-mini-128k-instruct"
# HUGGINGFACE_MODEL_NAME_OR_PATH="mistralai/Mistral-7B-Instruct-v0.2"