AI NEWS.ccAI正在发生 未来正在更新
♡ 我的收藏 0
AI 模型库/WaveCut/Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-DFlash2-NInfer-v3
MODEL PROFILE · 模型档案

Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-DFlash2-NInfer-v3

WaveCut · Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-DFlash2-NInfer-v3

A single-file NInfer **v3** artifact carrying the community uncensored tune HauhauCS/Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-MTP-GGUF as the target model and the official z-lab/Qwen3.8-27B-DFlash2 draft model as the DFlash2 speculative companion (recipe `qwen3_8_27b-v2`, the first NInfer recipe with the DFlash2 payload). The native NextN/MTP head of the tune is preserved as well, so both `--spec mtp` and `--spec dflash2` work. It runs on **NInfer-all**, the `master` branch of iamwavecut/ninfer-all, which serves the RTX 3090, RTX 4090, RTX 5090 and RTX PRO 6000 Blackwell. Its first (v2) release ran on stock NInfer, which builds only for the RTX 5090; this v3 file was verified on NInfer-all. `--spec mtp --draft-tokens 3 --lm-head-draft` selects the tune's own NextN head instead. Vision (`--vision`) is the official tower: the release's BF16 projector reproduces it bit-for-bit through the FP16 export path, so the artifact carries the official BF16 vision words. **Hardware.** Upstream NInfer targets one RTX 5090 (`sm_120a`, 32 GB). NInfer-all serves the same `qwen3.8-27b / groupwise-int` identity with DFlash2 on the RTX 3090, 4090, 5090 and RTX PRO 6000: its Docker image needs an NVIDIA driver of the CUDA 13 branch (580 or newer) and the NVIDIA Container Toolkit, and a build (its README) sets `CMAKE_CUDA_ARCHITECTURES` to `86`, `89` or `120a`. The speed of the official artifact of this identity on each card is in its reference measurements, this tune itself was not measured there. On 24 GB cards the official artifact of the same size fits DFlash2 with five drafts beside 258,048 tokens of `rk4v4` KV on the 3090 and 245,760 on the 4090 (`--kv-dtype rk4v4 --gdn-state-fp16 --spec dflash2 --draft-tokens 5`).

译文排队中，当前展示来源原文。

视觉问答license:apache-2.0
上下文窗口来源未披露
输入 → 输出文本、图像 → 文本
模型权重待核验
档案更新时间2026/11/2 19:15:54
OVERVIEW

模型介绍

A single-file NInfer **v3** artifact carrying the community uncensored tune HauhauCS/Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-MTP-GGUF as the target model and the official z-lab/Qwen3.8-27B-DFlash2 draft model as the DFlash2 speculative companion (recipe `qwen3_8_27b-v2`, the first NInfer recipe with the DFlash2 payload). The native NextN/MTP head of the tune is preserved as well, so both `--spec mtp` and `--spec dflash2` work. It runs on **NInfer-all**, the `master` branch of iamwavecut/ninfer-all, which serves the RTX 3090, RTX 4090, RTX 5090 and RTX PRO 6000 Blackwell. Its first (v2) release ran on stock NInfer, which builds only for the RTX 5090; this v3 file was verified on NInfer-all. `--spec mtp --draft-tokens 3 --lm-head-draft` selects the tune's own NextN head instead. Vision (`--vision`) is the official tower: the release's BF16 projector reproduces it bit-for-bit through the FP16 export path, so the artifact carries the official BF16 vision words. **Hardware.** Upstream NInfer targets one RTX 5090 (`sm_120a`, 32 GB). NInfer-all serves the same `qwen3.8-27b / groupwise-int` identity with DFlash2 on the RTX 3090, 4090, 5090 and RTX PRO 6000: its Docker image needs an NVIDIA driver of the CUDA 13 branch (580 or newer) and the NVIDIA Container Toolkit, and a build (its README) sets `CMAKE_CUDA_ARCHITECTURES` to `86`, `89` or `120a`. The speed of the official artifact of this identity on each card is in its reference measurements, this tune itself was not measured there. On 24 GB cards the official artifact of the same size fits DFlash2 with five drafts beside 258,048 tokens of `rk4v4` KV on the 3090 and 245,760 on the 4090 (`--kv-dtype rk4v4 --gdn-state-fp16 --spec dflash2 --draft-tokens 5`).

主要特点

特点摘要待依据来源补齐，不凭模型名称推断能力。

局限与注意事项

暂无足够来源支持具体缺点结论。请重点验证实际任务表现、时延、数据政策、许可与版本生命周期。

PRICING & COST

来源报价与费用分析

同一模型在不同渠道、地区和配置下可能采用不同价格。下表独立保留各渠道报价，不将渠道价格视为厂商官方统一价格。

来源 / 渠道输入输出计费单位采集时间
暂未取得可核验报价，请查阅来源渠道。

按月估算基础 Token 费用

—

预估不包含缓存优惠、阶梯定价、多模态附加费用、网络检索、税费及推理额外输出。

INDEPENDENT BENCHMARKS

公开评测与运行配置

暂无已采集的可核验公开评分。模型未上榜不代表能力较低。

查看完整已收录榜单 →

常见问题

开放权重是否代表可以免费商用？

不一定。应核对具体版本许可证、商用门槛和使用政策。自部署还会产生算力和维护成本。

这些价格是实时结算价格吗？

不是实时结算价格，而是带采集时间的来源报价快照。请到对应渠道核对当前版本、区域、折扣和计费方式。

资料来源与可追溯记录

huggingface · WaveCut/Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-DFlash2-NInfer-v3 ↗2026/11/1 05:59:37
huggingface · enrich:a4d836d1-9171-400e-8928-649feb1e4623 ↗2026/10/29 20:12:11

许可证：apache-2.0。模型介绍与第三方评测均以原来源的具体版本和方法为准。