AI NEWS.ccAI正在发生 未来正在更新
♡ 我的收藏 0
AI 模型库/philbert440/Qwen3.8-27B-W4A16-AWQ
MODEL PROFILE · 模型档案

Qwen3.8-27B-W4A16-AWQ

philbert440 · Qwen3.8-27B-W4A16-AWQ

INT4 weight-only quantization of Qwen/Qwen3.8-27B — **the whole model, nothing stripped**: vision tower intact, MTP speculative-decoding head intact, thinking mode calibrated with real reasoning traces. 55 GB BF16 → **19.5 GB**, which puts a current-generation 27B VLM (with headroom for 32K+ context) on a pair of 32 GB V100s — and it serves on modern GPUs with stock vLLM, no ceremony. Validated end-to-end on **2×V100-32GB under 1Cat-vLLM 1.2.2** (SM70 TurboMind dequant path): coherence, think-tag discipline, image understanding, math, long-form generation, and MTP speculative decoding in both draft modes. **Methodology:** warm serve (2 discarded warmup generations), fixed-length generations via `ignore_eos` so every run produces exactly the stated token count, official Qwen3.8 sampling per mode (thinking `1.0/0.95/20`, instruct `0.7/0.80/20 + presence 1.5`), varied prompts. Reported as mean ± sd tokens/s. Rig: 2×V100-32GB, 1Cat-vLLM 1.2.2, TP2, `fp8_e5m2` KV, `max_num_seqs 4`, MTP K=2.

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2026/10/30 07:19:58
OVERVIEW

模型介绍

INT4 weight-only quantization of Qwen/Qwen3.8-27B — **the whole model, nothing stripped**: vision tower intact, MTP speculative-decoding head intact, thinking mode calibrated with real reasoning traces. 55 GB BF16 → **19.5 GB**, which puts a current-generation 27B VLM (with headroom for 32K+ context) on a pair of 32 GB V100s — and it serves on modern GPUs with stock vLLM, no ceremony. Validated end-to-end on **2×V100-32GB under 1Cat-vLLM 1.2.2** (SM70 TurboMind dequant path): coherence, think-tag discipline, image understanding, math, long-form generation, and MTP speculative decoding in both draft modes. **Methodology:** warm serve (2 discarded warmup generations), fixed-length generations via `ignore_eos` so every run produces exactly the stated token count, official Qwen3.8 sampling per mode (thinking `1.0/0.95/20`, instruct `0.7/0.80/20 + presence 1.5`), varied prompts. Reported as mean ± sd tokens/s. Rig: 2×V100-32GB, 1Cat-vLLM 1.2.2, TP2, `fp8_e5m2` KV, `max_num_seqs 4`, MTP K=2.

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 · philbert440/Qwen3.8-27B-W4A16-AWQ ↗2026/10/30 07:19:58
huggingface · enrich:d69f81aa-c56b-44fd-9512-0997e1ef47d8 ↗2026/10/29 08:57:59

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。