from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model = AutoModelForCausalLM.from_pretrained("Sebastianpro88/Chichu-Video-1.0", torch_dtype=torch.float16, device_map="cpu")
tokenizer = AutoTokenizer.from_pretrained("Sebastianpro88/Chichu-Video-1.0")

messages = [
    {"role": "system", "content": "You are Chichu-Video-1.0, a video script writer."},
    {"role": "user", "content": "Create a 15-second video about a cat in a garden. Include dialogue."}
]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(text, return_tensors="pt")
out = model.generate(**inputs, max_new_tokens=300, do_sample=True, temperature=0.5)
print(tokenizer.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))