Qwen3.5-122B-A10B-NVFP4

This is a quantized version of Qwen/Qwen3.5-122B-A10B using the NVFP4 quantization scheme.

Please use nightly vLLM for support.

Changelog

• 02/03/2026: Added MTP (multi-token prediction) weights from source checkpoint, enabling speculative decoding with vLLM.
• 25/02/2026: Initial upload.

Calibration

• Samples: 512 (256 from each dataset)
• Datasets:
  • HuggingFaceH4/ultrachat_200k (train_sft split)
  • nvidia/Nemotron-Post-Training-Dataset-v2 (chat split)
• Max sequence length: 4096
• All experts calibrated: moe_calibrate_all_experts=True

Creation

This model was created using VLLM's LLM Compressor with Qwen3.5 MoE support added via PR #2383. The PR adds a custom CalibrationQwen3MoeSparseMoeBlock that routes calibration data to all experts during quantization, ensuring every expert receives proper calibration for accurate NVFP4 quantization.