InternVL

This release introduces two major new models, DiffusionGemma and DeepSeek-V3.2, alongside significant enhancements to the custom kernel API for better fusion capabilities. Numerous bug fixes address issues across various models, CI stability, and dependency compatibility.

This preview release introduces support for the InternVL 2.5 and 3 family of multimodal models, featuring a native multimodal pre-training paradigm and state-of-the-art performance on visual-linguistic tasks.

This release introduces dynamic high-resolution image preprocessing for the InternVL model and updates GGUF metadata handling, alongside various internal cleanups and bug fixes.

This release focuses on V1 engine stability, including a critical memory fix and crash prevention for duplicate request IDs. It introduces FP8 KV Cache support, a new fastsafetensors loader, and expands hardware support for TPU and ROCm.