GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
yingfanbot
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
yingfanbot
Model Tree
meta-llama/Llama-3.2-3B-Instruct
Input Modalities
Output Modalities
Supported Functionality
from_pretrained loads the weights only — the looped padded architecture needs the LOTUS code
(code repo).
Run in the pinned env (torch 2.7 / transformers 4.46.2). This loads the safetensors straight from this repo and runs the latent loop — no separate checkpoint file needed:
bash
python scripts/eval.py \--model_id yingfanbot/gsm-lotus-llama3b-codi \--datasets gsm8k --n_looped_iters 6 --c_thought 25 --bf16
Yields 70.58% on GSM8k-Aug (verified by loading this repo's safetensors directly).
bibtex
@article{fan2026bridging,title={Bridging the Gap Between Latent and Explicit Reasoning with Looped Transformers},author={Fan, Ying and Svete, Anej and Lee, Kangwook},journal={arXiv preprint arXiv:2606.31779},year={2026}}