Unverified60% confidenceFactTime unknown
MiniMind-V is an open-source project capable of training a visual multimodal model with 65M (65 million) parameters from scratch within 2 hours.
2
Sources
60%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedTiny-MoE模型在80亿（8B）token上从零训练64% similarUnverifiedB站UP主「程序员小程」发布了一期时长17分钟、制作耗时48小时的AI编程实战教程，从零开发闪卡应用。63% similarUnverifiedCLM's 8-billion-parameter backbone Qwen3-8B is frozen and serves only as a reader; what is actually trained are only the small projection heads on each side with approximately 20 million parameters each, and a complete pretraining run takes about one hour on a single RTX.63% similarUnverifiedThe method operates at inference time, requires no model retraining, can run in real time, and has a 95th-percentile decision latency below 61 milliseconds, which falls within the 80-millisecond single-frame budget.63% similarUnverifiedThinker-WM's training corpus totals approximately 20K hours of first-person video, 10K hours of real robot data, and 1K hours of simulation data, of which about 4K hours of real robot data has aligned depth supervision.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/29981API
curl https://kongchang.com/api/v1/knowledge/claims/29981MCP
get_claim(id=29981)