AcciSAGE-S8-SAPO-FP8
task-specific merged FP8 model for AcciSAGE: Error-aware VLM Tuning for Spatio-temporal Grounding of Traffic Accident Videos.
- Task branch: S8 — Spatial accident localization
- Training: SFT + SAPO; merged and FP8-quantized
- Base model:
Qwen/Qwen3-VL-8B-Instruct - Expected output: Accident center in normalized 0–1000 coordinates
This repository contains a standalone merged FP8 model and does not require a separate LoRA adapter.
Usage
swift infer \
--model yzk9/AcciSAGE-S8-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.
- Downloads last month
- 4
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for yzk9/AcciSAGE-S8-SAPO-FP8
Base model
Qwen/Qwen3-VL-8B-Instruct