Changes / Feeds / XiaomiMiMo/MiMo-V2.5
XiaomiMiMo/MiMo-V2.5
4 changes to a tracked key since 2026-04-27, newest first. Head as last read: 63651580ca774f8504f676040460aed3e1244ac1.
GPU memory to fine-tune it, card by card: MiMo-V2.5 on /m/.
| Date | Advisory | What changed, and what it does |
|---|---|---|
| 2026-05-08 | x_NXMC-2026-0032 | the chat template text changed but rendered byte-identically on the fixed conversation set (system+user, multi-turn, tool call) (2026-05-08).13b5e3f92ab9 -> 2fd4f899a491 |
| 2026-05-07 | x_NXMC-2026-0031 breaking | EOS changed from [151643,151645,1561672] to [151643,151645,151672]; anything fine-tuned or served against the old value stops at a different token (2026-05-07).4da2748fc0a0 -> 13b5e3f92ab9 |
| 2026-04-28 | x_NXMC-2026-0030 breaking | context window 262144 to 1048576 tokens; rope_theta 5000000 to 10000000; every position embedding changes; the chat template changed and renders differently (system+user, multi-turn, tool-call); prompts rendered with the old one no longer match what the model now expects, so cached prompts and eval sets need re-rendering (2026-04-28).c3f3abb0657e -> 4da2748fc0a0 |
| 2026-04-27 | x_NXMC-2026-0029 breaking | EOS changed from [151643,151645] to [151643,151645,1561672]; anything fine-tuned or served against the old value stops at a different token; the chat template changed and renders differently (system+user, multi-turn, tool-call); prompts rendered with the old one no longer match what the model now expects, so cached prompts and eval sets need re-rendering; generation defaults changed (temperature unset to 1, top_p unset to 0.95); a call that does not pass these explicitly now generates differently (2026-04-27).4f78a4d18f53 -> c3f3abb0657e |
Pin it, and let a bot move the pin
from_pretrained("XiaomiMiMo/MiMo-V2.5", revision="63651580ca774f8504f676040460aed3e1244ac1")With the Renovate preset, a pull request moves this pin when the head moves, and links back here.
[](https://www.neurarch.com/changes-feeds/models/XiaomiMiMo/MiMo-V2.5.html)