goavinash5/Gradio_LLAMA_Testing

1MODEL_PATH = ""2# if MODEL_PATH is "", default llama.cpp/gptq models 3# will be downloaded to: ./models4 5# Example ggml path:6# MODEL_PATH = "./models/llama-2-7b-chat.ggmlv3.q4_0.bin"7# MODEL_PATH = "./models/Llama-2-7b-Chat-GPTQ"8 9# options: llama.cpp, gptq, transformers10BACKEND_TYPE = "llama.cpp"11 12# only for transformers bitsandbytes 8 bit13LOAD_IN_8BIT = False14 15MAX_MAX_NEW_TOKENS = 204816DEFAULT_MAX_NEW_TOKENS = 102417MAX_INPUT_TOKEN_LENGTH = 400018 19DEFAULT_SYSTEM_PROMPT = "20You are a movie recommender chatbot. You give movie recommendations to users based on their profile. Your job now is to fully understand the user profile based on the given context and give them recommendations based on their input. Here are some rules for you to follow while generating a response:211: Give an explanation for why each of the recommendations is a good fit for the user222: Give a maximum of 5 recommendations, unless specified otherwise by the user233: Give a predicted rating for the movie on a scale of 1 to 5: this is a rating the user would give to the movie if they watched it244: Mention how popular the movie is. Choose from among High, Medium, Low: High being most popular, Low being least255: Avoid recommending movies already rated by the user26 27''' User Context '''28"29