DeepSeek-Coder-V2-Lite-Instruct 16B running locally on an RTX 2050 laptop

Hello DeepSeek team,

We are developing Bright AI, a local inference technology focused on making large AI models practical on affordable hardware.

We have successfully demonstrated DeepSeek-Coder-V2-Lite-Instruct 16B running locally on an NVIDIA RTX 2050 laptop.

Running a model of this size on hardware in this class is normally considered impractical, but Bright AI made it possible to run the model locally and interact with it.

Recorded demonstration: https://huggingface.co/spaces/waziry/bright-ai-runs-deepseek-16b-on-rtx2050

We are sharing this result because DeepSeek-Coder-V2-Lite-Instruct was the model used in our demonstration, and we would appreciate feedback from the DeepSeek team.

We believe this technology could help make powerful local AI more practical for affordable computers, robots, drones, autonomous machines, and other edge devices.

We are also open to discussing possible technical collaboration or integration with the DeepSeek team.

Thank you,

Bright AI