GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
nchapman
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
nchapman
Model Tree
ornith-ai/Ornith-1.5-35B-A3B
Input Modalities
Output Modalities
Supported Functionality
calib-cascade-v1 — 999×4096-token windows at
dataset-native proportions from
nvidia/Nemotron-Cascade-2-SFT-Data
The exact rows are hash-pinned in
warp-3090-quant-lab-frozen-data
— that pin is the citable calibration generation.See SHA256SUMS in this repo. model.safetensors shard checksums:
text
undefined182e77dd83bd8e9ca818b240b82e28f243762cd5dda32e6eef327df7b1cd107e chat_template.jinja fda8dd193e8dfe66232f01d7719ed679da765952cf8e7b9bbe45af6557a19de1 config.json 9eb15d950e7d08cf1b6ae62e40cde27abf60574f62c86b64a2d56aeeba2a6183 generation_config.json 1a7a211fcd89a7d48d07cf238ce824e94ef1b5884e8d85184c3a467492a46ac0 model-00001-of-00002.safetensors efcafa741d4b8f6a8bf5ebfe2824e3cc20f3680912392d9413fe9486c7bb3a7f model-00002-of-00002.safetensors 47abfc9963df5b1f18c1afc891bd8a294a9d40013daac03fbfed1cc54c67d3c6 model.safetensors.index.json 57e5159f86cc9e9d902f5f376532ccee1cdc96be3aef822601dc9009fc613a1f model_mtp.safetensors bfbc24af59a3e73a9cd0653b8d4ae758dfaec4e3e6c15dfb1c6c8ee8d5683c85 processor_config.json 4e536ddcd36d982387be242e8789c54c05a10d900fe056b84394155dd389c2bf recipe.yaml 67d8113912fabc4af848be59c550922fb4945e3f9ad0853948d5def219fc4e46 tokenizer.json 7ff0211d20c398343cab1300beeb9e038ece910498ff279ecda9e3e5fd41026b tokenizer_config.json
text
undefinedProvenance: quantized on a single B200 via the
warp-3090-quant-lab
champion recipe (unchanged from the 4B release); run log
quant-35b-ornith.20261017T114909Z.13535 (2026-10-17).
python
from vllm import LLMllm = LLM(model="nchapman/Ornith-1.5-35B-A3B-W4A16-GPTQ")