🛠️ Maple – Run Gemma 4 26B on a 6GB Laptop GPU at 45 TPS (Sub-2-BPW)
Summary
Maple allows running Gemma 4 26B on a 6GB laptop GPU at 45 tokens per second using sub-2-BPW quantization.
Why it’s interesting
It achieves high performance and runs a large model on very constrained consumer laptop hardware.
Source metrics: Points 3 · Comments 0
HN discussion · Project
Source: #HackerNews / Show HN
Summary
Maple allows running Gemma 4 26B on a 6GB laptop GPU at 45 tokens per second using sub-2-BPW quantization.
Why it’s interesting
It achieves high performance and runs a large model on very constrained consumer laptop hardware.
Source metrics: Points 3 · Comments 0
HN discussion · Project
Source: #HackerNews / Show HN