Modelpublic

piffie/Llama-3.2-3B-Instruct-Genie-Snapdragon-X2-Elite-v81-16k

sourceHugging Facellama3.2updated 3mo agoView on Hugging Face

1likes9downloads

Model Card

Llama-3.2-3B-Instruct-Genie-Snapdragon-X2-Elite-v81-16k

meta-llama/Llama-3.2-3B-Instruct compiled to a Qualcomm genie bundle for the Snapdragon X2 Elite (Hexagon V81, soc_model 88). Compiled via `qai-hub-models` on Qualcomm AI Hub, targeting Snapdragon X2 Elite CRD (stock weights, just quantized + compiled for V81).

Run via: Genie (qai-appbuilder / AnythingLLMQnnEngine, OpenAI API on :3800). QNN bundles are version-coupled — keep your Hexagon NPU driver current. License: llama3.2.