Tencent releases WeMM-Video-2B AWQ W4A16 checkpoint
Tencent published an Apache 2.0-licensed repository for WeMM-Video-2B-AWQ, an AWQ W4A16 Safetensors checkpoint with 4.86 GiB of weight files. The model card says the 3B-parameter model accepts text, images, audio, and video and generates text responses in Chinese or English, while the quantization uses group-128 INT4 for feed-forward weights and BF16 for activations and attention weights. Tencent reports successful loading and generation checks on one NVIDIA L20 using Transformers 5.2.0 and a patched vLLM 0.18.0.
The release provides a downloadable quantized checkpoint and concrete loading instructions for deploying Tencent’s open-weight audiovisual language model, though its capability claims remain vendor-supplied.
tencent/WeMM-Video-2B-AWQ Video-Text-to-Text • 3B • Updated about 4 hours ago • 3 • 1
license: apache-2.0 language:
- en
- zh pipeline_tag: video-text-to-text library_name: transformers base_model: tencent/WeMM-Video-2B base_model_relation: quantized tags:
- wemm
- multimodal
- video
- audio
- quantized
- awq
WeMM-Video-2B · AWQ W4A16
Overview
This repository contains the AWQ W4A16 Safetensors variant of WeMM-Video-2B. WeMM-Video is a Chinese–English bilingual family for joint audio-visual understanding and reasoning. It accepts text, images, audio, and video and generates text responses.
#Tencent#WeMM-Video-2B#WeMM-Video-2B-AWQ#Transformers#vLLM#NVIDIA L20