ModelMiniCPM-V 4.7
MiniCPM-V-4.7-35B-A3B
A multimodal language model in the MiniCPM-V series, built on a Mixture-of-Experts (MoE) architecture with approximately 35B total parameters and approximately 3B parameters activated during inference. The model is designed for efficient on-device deployment and supports image-text understanding tasks.
Timeline (last 90 days)
Oct 7
A new model entry named MiniCPM-V-4.7-35B-A3B has appeared on the HuggingFace platform.
Expired50%
Oct 7
The MiniCPM-V-4.7-35B-A3B model currently has no model card attached.
Expired50%
Oct 7
The "V" in the name MiniCPM-V-4.7-35B-A3B stands for Vision, referring to visual capability.
Unverified50%