Edit model card

Usage for VLLM

from vllm import LLM, SamplingParams
from transformers import AutoTokenizer, pipeline

BASE_MODEL = "sh2orc/gemma-1.1-korean-7b-it"

llm = LLM(model=BASE_MODEL)

tokenizer = AutoTokenizer.from_pretrained(BASE_MODEL)
tokenizer.pad_token = tokenizer.eos_token
tokenizer.padding_side = 'right'

instruction = '독도에 λŒ€ν•΄μ„œ μ•Œλ €μ€˜'

messages = [
    {
      "role": "user",
      "content": instruction
    }, 
]


prompt_message = tokenizer.apply_chat_template(
        messages, 
        tokenize=False, 
        add_generation_prompt=True,
)

eos_token_id = [tokenizer.eos_token_id, tokenizer.convert_tokens_to_ids("<end_of_turn>")]

outputs = llm.generate(prompt_message, SamplingParams(stop_token_ids=eos_token_id, temperature=0.2, top_p=0.8,max_tokens=4096))

for output in outputs:
    propt = output.prompt
    generated_text = output.outputs[0].text
    print(generated_text)

Result

λ…λ„λŠ” 동해에 μœ„μΉ˜ν•œ λŒ€ν•œλ―Όκ΅­μ˜ νŠΉλ³„μžμΉ˜λ„λ‘œ, 면적은 5.8km이며, 행정ꡬ역은 λ…λ„κ΅°μž…λ‹ˆλ‹€. λ…λ„μ˜ 면적은 λŒ€λž΅ 5.8km이며, μΈκ΅¬λŠ” μ•½ 10λͺ…μž…λ‹ˆλ‹€. λ…λ„λŠ” ν•œκ΅­κ³Ό 일본의 ꡭ경을 ν˜•μ„±ν•˜κ³  있으며, λ…λ„μ˜ 주민듀은 일본의 영ν–₯을 많이 λ°›μ•˜μŠ΅λ‹ˆλ‹€. λ…λ„μ˜ κ²½μ œλŠ” κ΄€κ΄‘κ³Ό μ†Œκ·œλͺ¨μ˜ 어업이 μ£Όλ₯Ό 이루며, λ…λ„λŠ” 1949λ…„λΆ€ν„° 1954λ…„κΉŒμ§€ 일본에 μ˜ν•΄ μ λ Ήλ˜μ—ˆμŠ΅λ‹ˆλ‹€. λ…λ„λŠ” 1954λ…„ 인천 κ°•ν™” μ‘°μ•½μœΌλ‘œ 인해 μš°λ¦¬λ‚˜λΌμ˜ μ˜ν† κ°€ λ˜μ—ˆμŠ΅λ‹ˆλ‹€.
Downloads last month
4
Safetensors
Model size
8.54B params
Tensor type
BF16
Β·
Inference API
Unable to determine this model’s pipeline type. Check the docs .