GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
mbx2
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
mbx2
Model Tree
baidu/Unlimited-OCR
Input Modalities
Output Modalities
Supported Functionality
Early checkpoint of the Frisket OCR student fine-tuned from SmolVLM-256M on nvidia/OCR-Synthetic-Multilingual-v1 (Batch 01).
| Steps | 1000 |
| Train loss | 18.21 → 0.10 |
| Eval gate | not run yet |
| Hardware | RTX 3090 |
Not production-ready — mid-run snapshot for sharing / experiments.
python
from transformers import AutoProcessor, AutoModelForVision2Seqrepo = "mbx2/frisket-ocr-student-batch01-1k"processor = AutoProcessor.from_pretrained(repo)model = AutoModelForVision2Seq.from_pretrained(repo, torch_dtype="auto", device_map="auto")