from transformers import AutoModelForCausalLM, AutoTokenizer

import torch

model_path = "sahil2801/persuasion_model_glaive"

tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForCausalLM.from_pretrained(model_path).half().cuda()
prompt = '''Claim: <> \n\n Thought: <>'''

inputs = tokenizer(prompt,return_tensors="pt").to(model.device)
outputs = model.generate(**inputs,do_sample=True,eos_token_id=tokenizer.eos_token_id,temperature=0.9,max_new_tokens=1500)

print(tokenizer.decode(outputs[0],skip_special_tokens=True))