Edit model card
  • Train Config
  • base_model: allganize/Llama-3-Alpha-Ko-8B-Instruct
  • model_type: AutoModelForCausalLM tokenizer_type: AutoTokenizer

HOW TO USE

from transformers import AutoTokenizer, AutoModelForCausalLM

model_id = "MRAIRR/minillama3_8b_all"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype="auto",
    device_map="auto",
)

PROMPT_TEMPLATE = """
# μ§€μ‹œ:
당신은 인곡지λŠ₯ μ–΄μ‹œμŠ€ν„΄νŠΈμž…λ‹ˆλ‹€. μ‚¬μš©μžκ°€ λ¬»λŠ” 말에 μΉœμ ˆν•˜κ³  μ •ν™•ν•˜κ²Œ λ‹΅λ³€ν•˜μ„Έμš”.
"""

messages = [
    {"role": "system", "content":PROMPT_TEMPLATE},
    {"role": "user", "content": "μ•ˆλ…•? λ‚΄ 이름은 ν˜„μˆ˜ γ…Žγ…Ž λ§Œλ‚˜μ„œ λ°˜κ°€μ›Œ"},
]

input_ids = tokenizer.apply_chat_template(
    messages,
    add_generation_prompt=True,
    return_tensors="pt"
).to(model.device)

terminators = [
    tokenizer.eos_token_id,
    tokenizer.convert_tokens_to_ids("<|eot_id|>")
]

outputs = model.generate(
    input_ids,
    max_new_tokens=256,
    temperature = 0.3,
    eos_token_id=terminators,
    do_sample=True,
    repetition_penalty=1.05,
)
response = outputs[0][input_ids.shape[-1]:]
response_text = tokenizer.decode(response, skip_special_tokens=True)
completion = '\n'.join(response_text.split("."))
print(completion)
Downloads last month
1,708
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.