# Code adapted from https://github.com/tloen/alpaca-lora
!pip install git+https://github.com/huggingface/transformers.git
!pip install sentencepiece
!pip install peft
!pip install accelerate
!pip install bitsandbytes
import torch
from peft import PeftModel
import transformers

from transformers import LlamaTokenizer, LlamaForCausalLM, GenerationConfig

tokenizer = LlamaTokenizer.from_pretrained("decapoda-research/llama-7b-hf")
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
print(device)
model = LlamaForCausalLM.from_pretrained(
    "decapoda-research/llama-7b-hf",
    load_in_8bit=True,
    torch_dtype=torch.float16,
    device_map="auto",
    offload_folder = "."
)

model = PeftModel.from_pretrained(
    model, 
    "tloen/alpaca-lora-7b",
    torch_dtype=torch.float16,
    device_map="auto",
    load_in_8bit = True,
    offload_folder="."
)

def generate_prompt(instruction, input=None):
    if input:
        return f"""Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.

### Instruction:
{instruction}

### Input:
{input}

### Response:"""
    else:
        return f"""Below is an instruction that describes a task. Write a response that appropriately completes the request.

### Instruction:
{instruction}

### Response:"""


model.eval()


def evaluate(
        instruction,
        input=None,
        temperature=0.1,
        top_p=0.75,
        top_k=40,
        num_beams=4,
        **kwargs,
):
    prompt = generate_prompt(instruction, input)
    inputs = tokenizer(prompt, return_tensors="pt")
    input_ids = inputs["input_ids"].to(device)
    generation_config = GenerationConfig(
        temperature=temperature,
        top_p=top_p,
        top_k=top_k,
        num_beams=num_beams,
        **kwargs,
    )
    with torch.no_grad():
        generation_output = model.generate(
            input_ids=input_ids,
            generation_config=generation_config,
            return_dict_in_generate=True,
            output_scores=True,
            max_new_tokens=2048,
        )
    s = generation_output.sequences[0]
    output = tokenizer.decode(s)
    return output.split("### Response:")[1].strip()
input = """RESEARCH REPORT NO. 1037027Final Clinical Study Report -NV25118: A Randomized, Multicentre, Single Blinded, Parallel Study of the Safety of 100 mg and 200 mg Oseltamivir Administered Intravenously for the Treatment of Influenza in Patients Aged > 13 Years. Report No. 1037027. June 3, 2013 Study Sponsor(s):Dr F Hoffman La-Roche LtdStudy Dates:First patient screened: Jan 15, 2010 Last patient visit: Sept 14, 2012Trial Phase:II/IIIIndication:InfluenzaName of Principal Investigator Thomas MathewSponsor's Signatory: Personnel Responsible for Clinical and Statistical Analyses: Clinical analysis:John ThomasSafety analysis:William RichardStatistical analysis:Charles DanielPharmacokinetic analysis:Robert DavidGCP Compliance: This study was conducted in accordance with the principles of GCP"""
instruction = "Extract Names, Diseases, Dates from below text:"
# input = "RESEARCH REPORT NO. 1037027Final Clinical Study Report -NV25118: A Randomized, Multicentre, Single Blinded, Parallel Study of the Safety of 100 mg and 200 mg Oseltamivir Administered Intravenously for the Treatment of Influenza in Patients Aged > 13 Years. Report No. 1037027. June 3, 2013 Study Sponsor(s):Dr F Hoffman La-Roche LtdStudy Dates:First patient screened: Jan 15, 2010 Last patient visit: Sept 14, 2012"
print("Response:", evaluate(instruction,input))