Spaces:
Sleeping
Sleeping
File size: 997 Bytes
1ee715a a0e2620 1ee715a a0e2620 1ee715a a0e2620 1ee715a a0e2620 1ee715a a0e2620 3329d00 a0e2620 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 | ---
title: BYOD-Llama-3.1-8B
emoji: 🧬
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 5.49.1
python_version: "3.12"
app_file: app.py
pinned: false
license: other
startup_duration_timeout: 1h
short_description: Masked-diffusion demo for BYOD-Llama-3.1-8B
---
# BYOD-Llama-3.1-8B
Interactive full-precision inference for **BYOD-Llama-3.1-8B**, one of the BYOD
(Build Your Own DLM) models. It uses the exact `best` LoRA checkpoint
from the corresponding experiment and loads its original base model in BF16.
No 4-bit quantization is used.
Generation starts from masked answer positions and refines them in parallel.
Fewer denoising steps than generated tokens provide a sub-autoregressive
inference budget; increasing the step count gives the model more refinement
opportunities.
Model: [Ruurd/BYOD-Llama-3.1-8B](https://huggingface.co/Ruurd/BYOD-Llama-3.1-8B)
This is a research demo. Outputs may be inaccurate or inappropriate and
inherit limitations from the original base model.
|