YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/model-cards#model-card-metadata)
used chat vector extraction method
base model = beomi/Llama-3-Open-Ko-8B v2 base model = sosoai/hansoldeco-beomi-llama3-8b-ko-v0.1 (hansoldeco domain own dataset)
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
import torch
tokenizer = AutoTokenizer.from_pretrained("sosoai/hansoldeco-beomi-llama3-8b-ko-v0.1")
model_name = "sosoai/hansoldeco-beomi-llama3-8b-ko-v0.2-chatvector"
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype=torch.bfloat16,
device_map="auto",
)
conversation = [
{'role': 'user', 'content': "μλ
νμΈμ! λλ λꡬμΈμ?"},
{'role': 'assistant', 'content': "μ λ νμλ°μ½ λλ°°, λ²½μ§ κ·Έλ¦¬κ³ λ§κ°μμ¬ νμ μ λ¬Έ μ±λ΄μ
λλ€. 무μμ λμλ릴κΉμ? μ΄μ κ΄λ ¨λ λν λͺ¨λ μ§λ¬Έμ ν΄μ£ΌμΈμ."},
{'role': 'user', 'content': "μλ
νμΈμ! λΉμ μ μ΄λ¦μ μλ €μ£ΌμΈμ. λ²½μ§νμλ μ΄λ»κ² ν΄κ²°ν΄μ?"}
]
prompt = tokenizer.apply_chat_template(conversation, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, use_cache=True, max_length=256)
output_text = tokenizer.decode(outputs[0])
print(output_text)
- Downloads last month
- 0