customerSupport.py

from transformers import AutoModelForCausalLM,GenerationConfig,AutoTokenizer,GPTQConfig
from peft import AutoPeftModelForCausalLM
from peft import PeftModel, PeftConfig
import torch

def input_data_preprocessing_1(example):


    processed_example = "<|system|>\n You are a support chatbot who helps with user queries chatbot who always responds in the style of a professional.\n<|user|>\n" + example["instruction"] + "\n<|assistant|>\n"

    return processed_example

def customerConverstaion(prompt):
    def input_data_preprocessing(example):

        processed_example = "<|system|>\n You are a support chatbot who helps with user queries chatbot who always responds in the style of a professional.\n<|user|>\n" + example["instruction"] + "\n<|assistant|>\n"

        return processed_example
    input_string = input_data_preprocessing(
        {
            "instruction": "i have a question about cancelling order {{Order Number}}",
        }
    )
    tokenizer = AutoTokenizer.from_pretrained("zephyrFT/checkpoint-100")
    model = AutoPeftModelForCausalLM.from_pretrained("zephyrFT/checkpoint-100",
        low_cpu_mem_usage=True,
        return_dict=True,
        torch_dtype=torch.float16,
        device_map="cuda")

    inputs = tokenizer(input_string, return_tensors="pt").to("cuda")

    generation_config = GenerationConfig(
        do_sample=True,
        top_k=1,
        temperature=0.1,
        max_new_tokens=256,
        pad_token_id=tokenizer.eos_token_id
    )

    outputs = model.generate(**inputs, generation_config=generation_config)
    print(tokenizer.decode(outputs[0], skip_special_tokens=True))

def customerConverstaion_1(prompt):
    # Check GPU availability
    print("Available GPU devices:", torch.cuda.device_count())
    print("Name of the first available GPU:", torch.cuda.get_device_name(0))
    config = PeftConfig.from_pretrained("DSU-FDP/customer-support")
    base_model = AutoModelForCausalLM.from_pretrained("TheBloke/zephyr-7B-beta-GPTQ", device_map='cuda')
    model = PeftModel.from_pretrained(base_model, "DSU-FDP/customer-support")
    from transformers import AutoTokenizer,GPTQConfig
    tokenizer=AutoTokenizer.from_pretrained(base_model, trust_remote_code=True)
    tokenizer.padding_side = 'right'
    tokenizer.pad_token = tokenizer.eos_token
    tokenizer.add_eos_token = True
    tokenizer.add_bos_token, tokenizer.add_eos_token
    tokenizer = AutoTokenizer.from_pretrained("DSU-FDP/customer-support")
    input_string = input_data_preprocessing(
    {
        "instruction": "i have a question about cancelling order {{Order Number}}",
    }
    )
    inputs = tokenizer(input_string, return_tensors="pt").to("cuda")
    generation_config = GenerationConfig(
        do_sample=True,
        top_k=1,
        temperature=0.1,
        max_new_tokens=256,
        pad_token_id=tokenizer.eos_token_id
    )
    outputs = model.generate(**inputs, generation_config=generation_config)
    return outputs