GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
DJLougen
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
DJLougen
Model Tree
unsloth/Qwen3.6-35B-A3B
Input Modalities
Output Modalities
Supported Functionality
I'm a PhD student in visual neuroscience at the University of Toronto who also happens to spend way too much time fine-tuning, merging, and quantizing open-weight models on rented H100s and a local DGX Spark. All training compute is self-funded — balancing GPU costs against a student budget. If my uploads have been useful to you, consider buying a PhD student a coffee. It goes a long way toward keeping these experiments running.
Vision/video restoration: The repository now includes the Qwen3.6 base visual tower plus image/video processor files, so the source safetensors checkpoint should load as the multimodal conditional-generation architecture. Existing GGUF artifacts made before this restoration remain text-only until rebuilt.
Qwen3_5MoeForConditionalGeneration — Qwen 3.6 MoE with linear + full attention interleaved (Gated Delta Net)python
from transformers import AutoModelForCausalLM, AutoTokenizermodel_id = "DJLougen/Ornstein3.6-35B-A3B"tok = AutoTokenizer.from_pretrained(model_id)model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype="auto", device_map="auto")messages = [{"role": "user", "content": "Explain mixture-of-experts routing in one paragraph."}]inputs = tok.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to(model.device)out = model.generate(inputs, max_new_tokens=512)print(tok.decode(out[0][inputs.shape[-1]:], skip_special_tokens=True))
See the GGUF repo — pick a quant that fits your memory (Q4_K_M is a strong default for 24GB).
Apache 2.0 — inherited from the Qwen 3.6 base release.