Unverified50% confidenceFactExact time
Each RL training step of MiMo V2.6 processes approximately 2B tokens, using 1,568 prompts with 16 rollouts per prompt.
1
Sources
50%
Confidence
Long-term
Relevance
9/18/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedLLaMA 1 was trained on 1.4T tokens of data, and LLaMA 2 was trained on 2T tokens of data.73% similarUnverifiedMeta's Llama 2 (70-billion-parameter version) used 1.8 million GPU hours for training.70% similarUnverifiedTraining uses batch size 256, 400,000 steps, EMA 0.9999, linear learning rate decay over the final quarter, and torch.compile achieves a 2.4× speedup over eager mode.62% similarUnverifiedMeta在OPT-175B训练日志中披露，175天的训练过程中经历了超过100次硬件相关的中断事件，平均每1-2天出现一次需要人工干预的故障59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/930925API
curl https://kongchang.com/api/v1/knowledge/claims/930925MCP
get_claim(id=930925)