hf_gemma3_4b_9-checkpoint-57000

์ด ๋ชจ๋ธ์€ Gemma3 4B ๋ชจ๋ธ์„ ํŒŒ์ธํŠœ๋‹ํ•œ ์ฒดํฌํฌ์ธํŠธ์ž…๋‹ˆ๋‹ค.

๋ชจ๋ธ ์ •๋ณด

  • ๋ฒ ์ด์Šค ๋ชจ๋ธ: hf_gemma3_4b_9
  • ์ฒดํฌํฌ์ธํŠธ: checkpoint-57000
  • ๋ชจ๋ธ ํฌ๊ธฐ: 4B parameters
  • ํƒ€์ž…: Causal Language Model
  • ๋ผ์ด์„ ์Šค: Apache 2.0

์‚ฌ์šฉ ๋ฐฉ๋ฒ•

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_name = "NTIS/hf_gemma3_4b_9-checkpoint-57000"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.float16,
    device_map="auto"
)

# ํ…์ŠคํŠธ ์ƒ์„ฑ
text = "์•ˆ๋…•ํ•˜์„ธ์š”"
inputs = tokenizer(text, return_tensors="pt")
outputs = model.generate(
    **inputs, 
    max_length=100, 
    do_sample=True, 
    temperature=0.7,
    pad_token_id=tokenizer.eos_token_id
)
result = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(result)

ํ›ˆ๋ จ ์ •๋ณด

  • ์ด ์ฒดํฌํฌ์ธํŠธ๋Š” ํŠน์ • ์Šคํ…์—์„œ ์ €์žฅ๋œ ๋ชจ๋ธ์ž…๋‹ˆ๋‹ค
  • ์„ฑ๋Šฅ์€ ์ฒดํฌํฌ์ธํŠธ๋งˆ๋‹ค ๋‹ค๋ฅผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค

์ฃผ์˜์‚ฌํ•ญ

  • ์ด ๋ชจ๋ธ์€ ์—ฐ๊ตฌ/์‹คํ—˜ ๋ชฉ์ ์œผ๋กœ ์ œ๊ณต๋ฉ๋‹ˆ๋‹ค
  • ์ƒ์—…์  ์‚ฌ์šฉ ์ „์— ๋ผ์ด์„ ์Šค๋ฅผ ํ™•์ธํ•˜์„ธ์š”
  • GPU ๋ฉ”๋ชจ๋ฆฌ๊ฐ€ ์ถฉ๋ถ„ํ•œ์ง€ ํ™•์ธํ•˜์„ธ์š” (์ตœ์†Œ 8GB ๊ถŒ์žฅ)
Downloads last month
4
Safetensors
Model size
3.88B params
Tensor type
F32
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support