Model reference · open weights

Chameleon_7B_mGPT

Chameleon_7B_mGPT is an open-weight language model from Alpha-VLLM. Chameleon_7B_mGPT (BF16) weighs 14.1 GB; the smallest configuration that runs it is RTX 4060 Ti 16 GB.

For self-hosting, compare this build's requirements below with the GPU you plan to rent.

• Chameleon_7B_mGPT is a 7.0B parameter any-to-any model developed by Alpha-VLLM.
• It is a converted checkpoint of the Chameleon-7b model intended to ease the initialization of Lumina-mGPT training.
• The model features a modified qk-norm implementation that ensures parameter consistency within attention head groups during fine-tuning.

Summary of the Alpha-VLLM/Chameleon_7B_mGPT model card, 2026-10-01

Hugging Face ↗