phi3_short_story_merged_bfloat16

from transformers import AutoModelForCausalLM, AutoTokenizer

# Load the base model and tokenizer
tokenizer_model = "unsloth/Phi-3-mini-4k-instruct"
lora_model = "oztrkoguz/phi3_short_story_merged_bfloat16"
tokenizer = AutoTokenizer.from_pretrained(tokenizer_model)
model = AutoModelForCausalLM.from_pretrained(lora_model).to("cuda")

alpaca_prompt = """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:
create a short story from this keywords

### Input:
{}

### Response:
{}"""

# Use the merged model for inference
inputs = tokenizer(
[
    alpaca_prompt.format(
        "cat, dog, human",
        "", # output - leave this blank for generation!
    )
], return_tensors = "pt").to("cuda")


with torch.no_grad():
    output = model.generate(
        **inputs,
        max_length=100
    )

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