Unverified50% confidenceFactExact time
Aura's optional model is approximately 380 MB and is only downloaded when enabled; it is used to additionally redact individuals who do not appear in the user's data.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/12/2026
First Seen
Valid until: 1/10/2027
Sources
Related Entities
Related Claims
UnverifiedEmbedding Gemma 2 has 740 million parameters, consisting of a 270M text encoder, a 170M vision encoder, and a 300M audio encoder, with runtime dynamic memory of approximately 191MB to 567MB; the 270M version requires a minimum of 0.5GB memory to run, and the full multimodal version can run locally with 1GB of memory.59% similarUnverifiedWhisper原始FP32版本的Tiny模型约150MB，经INT8量化后可压缩至约40MB，在树莓派4B上可实现实时转写59% similarUnverifiedDomestic users downloading models via Hugging Face typically only achieve speeds of a few tens of KB to over one hundred KB per second.58% similarUnverifiedGemma 4的26B MOE版本下载大小约为18GB，可通过LM Studio免费应用零代码在本地运行58% similarUnverifiedUsing NVIDIA B300 chips in a 1-megawatt data center can achieve a total throughput of approximately 450,000 tokens/s, which at 340 tokens/s per user amounts to approximately 1,324 concurrent active users.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1004301API
curl https://kongchang.com/api/v1/knowledge/claims/1004301MCP
get_claim(id=1004301)