GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
jarrelscy
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
jarrelscy
Model Tree
XiaomiMiMo/MiMo-V2.6-Pro-RL
Input Modalities
Output Modalities
Supported Functionality
Expert 7 had an FP16 SwiGLU overflow, including on one training token with its original fitted weights. Halving its up-projection scales and doubling its down-projection scales preserves the mathematical function while reducing the intermediate magnitude; activation quantization/rounding can differ. All 104129 training rows routed to that expert passed a finite-range check (peak product 32892.5). The rebalanced gate/up was held fixed during PV; all other eligible projections remained trainable. Layer 69 used codebook/scales learning rates 0.0048/0.0032 and completed 69 updates over 18006461 training tokens.
Relative MoE validation RMS error: 0.787677 → 0.579308. Audit RMS error: 0.870055 → 0.766571. Export replay matched validation exactly. See evaluation/layer69_recovery.json and pv_reports/. These relative errors normalize by MoE output, not the full residual block.
On 16368 scored tokens from documents excluded from PV training, validation and checkpoint selection:
See evaluation/full_model.json for the report and exact test selection. This is calibration-arithmetic emulation, not a serving-runtime parity test. Quality acceptance, SM120 serving, and 1M-context memory fit remain unqualified. The checkpoint is experimental; production_ready=false.
All backbone, MTP, vision, audio, audio-tokenizer and DFlash weights are retained. Calibration and this quality evaluation are text-only. upload_verification.json identifies the snapshot whose full weight hashes were verified.