Google· @Google · X·· 5 小时前AI 评分
Gemini 4 Argon 将 Gemma 4 E2B 推理提速 2.85 倍
Google 宣布 Gemini 4 Argon 正在加速其 Gemma 开源模型，使其速度提升近三倍。几天内 Argon 优化 Gemma 4 E2B，在 A100 上通过 vLLM 实现 2.85x 推理提速，吞吐从约 37,900 tokens per second 提升到超过 108,000 tokens per second，保持 100% token parity 且质量无下降。Google 还展示了同时生成 64 个实时 Python 模拟的提速效果。
Gemini 4 Argon is also speeding up our @GoogleGemma open models — making them nearly three times as fast.
In just a few days, Argon optimized Gemma 4 E2B to achieve a 2.85x inference speedup on an A100 via vLLM, jumping from ~37,900 to over 108,000 tokens per second with 100% token parity and zero quality drop.
Here’s that speedup in action, generating 64 live Python simulations at once ↓
来源：Google · x.com