π¬ MDLM AR Model (Korean) - Hanbin42
μ΄ λͺ¨λΈμ MDLM (Masked Diffusion Language Model) ꡬ쑰λ₯Ό κΈ°λ°μΌλ‘ ν Autoregressive Korean Language Modelμ
λλ€.Hanbin42/my-mdlm-ar-model
μ skt/kogpt2-base-v2
ν ν¬λμ΄μ μ parkseongjun/psjkodata
νκ΅μ΄ λ°μ΄ν°μ
μΌλ‘ νμ΅λμμ΅λλ€.
π§ Model Details
- Backbone: Autoregressive (AR)
- Diffusion Type: Absorbing State
- Input Length: 1024 tokens
- Vocab Size: 51200 (KoGPT2 κΈ°μ€)
- Training Steps: 50,000
- Sampling Steps: 128 (DDPM-style)
- Precision: bfloat16
- EMA: Enabled (0.9999)
π¦ Files
File | Description |
---|---|
best.ckpt |
PyTorch Lightning λͺ¨λΈ 체ν¬ν¬μΈνΈ |
config.yaml |
νμ΅ μ μ¬μ©ν νμ΄νΌνλΌλ―Έν° μ€μ |
README.md |
λͺ¨λΈ μ€λͺ λ¬Έμ |
π How to Use
import torch
from lightning.pytorch import LightningModule
from diffusion import Diffusion # μ΄ νλ‘μ νΈ κΈ°μ€μΌλ‘ μ μλ¨
model = Diffusion.load_from_checkpoint("best.ckpt", config=..., tokenizer=...)
model.eval()
- Downloads last month
- 5
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
π
Ask for provider support