Xiaomi quietly uploads MiMo-V2.5-DFlash weights to Hugging Face
Xiaomi has uploaded the official DFlash weights for MiMo-V2.5-DFlash to Hugging Face. The model has 300B+ parameters and runs at 8-10 tokens/second on 2x24GB cards with offloading; DFlash could double that speed. The MTP head was shared but does not work yet.
Entities
Related
Tencent Hy Team releases Hy3, a 295B-parameter MoE model
Tencent's Hy Team released Hy3, a 295-billion-parameter Mixture-of-Experts (MoE) language model with 21B active parameters and a 3.8B MTP layer. The model outperforms similar-sized models and rivals open-source models with 2-5x parameters. It is available on Hugging Face in full (598GB) and FP8 quantized (300GB) versions.
Developer adds Nemotron Puzzle 75B support to mlx-lm, benchmarks 4-bit vs 5-bit quantization on M2 Max
A developer contributed native `nemotron_h_puzzle` support to mlx-lm (PR #1535) and ran benchmarks on a 64GB M2 Max. Comparing 4-bit vs 5-bit expert quantization (both with 6-bit dense layers, BF16 output head, group size 64), the 4-bit variant used 42.03 GiB checkpoint (49.68 GB peak memory) and achieved 14.27 tok/s, scoring 24/30 on local task checks, while 5-bit used 49.88 GiB (58.12 GB peak) at 10.53 tok/s scoring 21/30.
Local image-to-3D model for Apple Silicon and iPhone
A port of Hunyuan3D-Paint and Hunyuan3D-Shape to Swift MLX and Python MLX, enabling image-to-3D generation on Apple Silicon Macs and iPhones with low memory usage. It solves the problem of running 3D generation locally without cloud dependencies or high-end GPUs.
User tests GLM 5.2 on consumer hardware, finds performance comparable to Claude and GPT
A user tested GLM 5.2 on a standard computer and was impressed by its capabilities and security, finding them similar to Claude or GPT. They began converting the model to int4 and exploring MTP usage to avoid out-of-memory errors.
User runs 100B+ MoE LLMs on low-end laptop using NVMe swap and Q3 quantization
A Reddit user with a low-spec laptop (i7-8750H, 20GB RAM, GTX 1050 4GB) reports successfully running 100B+ parameter MoE models by offloading parameters to a Samsung NVMe SSD via mmap, using Q3 quantization and quantized KV cache (Q4_0). They note that dense models are unusable on their hardware, but MoE models work with experts offloaded to CPU.