embedl/Cosmos-Reason2-2B-W4A16-Edge2-FlashHead
Image-Text-to-Text • 2B • Updated • 2.31k • 8
FlashHead
Efficient Drop-In Replacement for the Classification Head in Language Model Inference. https://github.com/embedl/flash-head
-
-
embedl/Qwen3-1.7B-FlashHead-W4A16
2B • Updated • 150 • 3 -
embedl/gemma-3-270m-it-FlashHead
0.3B • Updated • 216 • 3 -
embedl/Qwen3-0.6B-FlashHead
0.6B • Updated • 99 • 4 -
embedl/Qwen3-1.7B-FlashHead
2B • Updated • 122 • 3 -
embedl/Llama-3.2-1B-Instruct-FlashHead
1B • Updated • 80 • 4 -
embedl/Llama-3.2-3B-Instruct-FlashHead
3B • Updated • 88 • 4 -
Edge Inference Benchmarks
🚀6On-Device benchmarks across devices and models.
-
embedl/Llama-3.2-3B-Instruct-FlashHead-W4A16
4B • Updated • 113 • 4 -
embedl/Llama-3.2-1B-Instruct-FlashHead-W4A16
2B • Updated • 57 • 6 -
embedl/gemma-3-1b-it-FlashHead
1.0B • Updated • 145 • 3 -
embedl/gemma-3-1b-it-FlashHead-W4A16
1B • Updated • 148 • 3 -
embedl/Qwen3.5-2B-FlashHead
Image-Text-to-Text • 2B • Updated • 437 -
embedl/Qwen3.5-4B-FlashHead
Image-Text-to-Text • 5B • Updated • 424 -
embedl/Qwen3.5-9B-FlashHead
Image-Text-to-Text • 10B • Updated • 411 -
embedl/Qwen3.5-0.8B-FlashHead
Image-Text-to-Text • 0.9B • Updated • 390 -
embedl/Qwen3.5-27B-FlashHead
Image-Text-to-Text • 28B • Updated • 255