AI Model2026
Provides a GGUF-quantized build of GLM-5.3-Flash for local text-generation and inference. Key features: 320B total / 18B active parameters, hybrid sparse+linear attention, native multimodal pretraining and Unsloth Dynamic quantization. Best for developers running GGUF local inference workflows.
AI Model2026
Provides per-tensor non-uniform GGUF quantizations of Qwen3.8-27B using GSQ and RCO, delivering high accuracy at 2.5–3.5 bits and including a BF16 vision projector for multimodal use. Optimized to run unmodified in llama.cpp, Ollama, and LM Studio.
AI Dataset2026
Provides an unattended text-to-video-and-audio streaming toolkit built around FastH3 (a 4-step distillation of MiniMax-H3): generation/retime/HTTP push scripts, a 221-scene prompt library, checkpoint conversion and ComfyUI workflows to run a continuous local stream.
AI Model2026
A TURBO multi-stage fine-tune of Qwen3.8‑27B that shortens internal “thinking” token blocks and raises ARC benchmarks (8‑bit ARC‑C ≈735, ARC‑E ≈882). It ships GGUF quants (regular and MTP, Neo‑Imatrix), vision support and 256k context for local multimodal inference on consumer GPUs.
AI Model2026
NVFP4-quantized checkpoint of Qwen3.8-Flash-Next for GPU-optimized multimodal autoregressive inference — routed MoE experts in W4A4 NVFP4 while attention/ancillary layers remain BF16; ~2.7× smaller than the BF16 source and supports very long contexts.
AI Video2026
Adds a plug-and-play linear-attention branch and LoRA adapters to MiniMax-H3 to run text-to-video generation faster than real-time (near-lossless quality tradeoffs). Includes an optimized FP8 inference stack and a community license with regional restrictions.
Uses lightweight discrete-diffusion adapters to draft blocks of tokens in parallel while keeping the original autoregressive model as the final arbiter, enabling lossless speedups (up to 3×) and compatibility with existing AR weights via a Diffusion Distillation phase.
Shows that fully quantizing all 496 linear layers—including the recurrent Gated DeltaNet—of a hybrid 27B LLM to 4-bit NVFP4 W4A4 preserves benchmark accuracy while reducing model size to 17.5 GiB and improving prefill speed; includes a calibrated NVFP4 checkpoint.
Compresses KV cache for long-chain reasoning by keeping prompt tokens and evicting remaining entries uniformly at random per attention head; across four models and six reasoning tasks it matches the strongest prior evictor while delivering 32–43% higher vLLM throughput. Relies on prompt protection and redundancy across heads/text to retain reasoning traces; suitable when static memory budgets and higher serving throughput are priorities.
AI Model2026
Fine-tuned variant of Qwen3.8-27B optimized to reduce reasoning cost and wall-clock latency for long-running agent workloads. Delivers ~12.8% faster decoding and higher MTP draft acceptance while cutting runaway reasoning, at a small MMLU-Pro accuracy trade-off (−1.45 pp).
Evaluates layer dropout (stochastic depth) in LLM pretraining, finds optimal layer distributions and time schedules that save ~20–25% training FLOPs at comparable or lower validation loss, and enables zero-shot layer skipping and self-speculative decoding for up to ~1.4–1.5× inference speedup.
AI Model2026
A 2B-class causal LLM packaged as a GGUF for local inference; offers 131072-token native long context, XML-style tool-calling support, and is tuned with SFT + RL + OPD using the UltraData family for stronger code, math and agentic abilities.