Downloads · 30 days
7
16% of all-time downloads
AcciSAGE-C8-SAPO-FP8
yzk9/AcciSAGE-C8-SAPO-FP8
AcciSAGE-C8-SAPO-FP8 is a video-text-to-text model from yzk9. Use it for the video-text-to-text task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.
task-specific merged FP8 model for AcciSAGE: Error-aware VLM Tuning for Spatio-temporal Grounding of Traffic Accident Videos.
All-time downloads
44
Public
Parameters
8.8B
10.6 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors10.6 GB · 100%
Parameter types
How the weights are stored.
F8_E4M36.9B · 79%
Base models
Model card
From the Hugging Face model README
AcciSAGE-C8-SAPO-FP8
task-specific merged FP8 model for AcciSAGE: Error-aware VLM Tuning for Spatio-temporal Grounding of Traffic Accident Videos.
- Task branch: C8 — Accident type classification
- Training: SFT + SAPO; merged and FP8-quantized
- Base model:
Qwen/Qwen3-VL-8B-Instruct - Expected output: One of rear-end, t-bone, sideswipe, head-on, or single
This repository contains a standalone merged FP8 model and does not require a separate LoRA adapter.
Usage
swift infer \
--model yzk9/AcciSAGE-C8-SAPO-FP8 \
--infer_backend vllm
Task-specific prompts and video preprocessing settings are required. This model was developed for research on traffic-accident video understanding.
- Task
- Video-Text-to-Text
- Library
- transformers
- Type
- qwen3_vl
- License
- apache-2.0
- Created
- Jul 18, 2026
- Updated
- Jul 18, 2026