MODEL PROFILE · 模型档案

vit-casting-defect

Zainulabedin1 · vit-casting-defect

A **tiny Vision Transformer trained from scratch** (no pretraining, no ImageNet) that classifies steel surface-defect images into six defect classes. The point of the project is to show a ViT *can* be built for manufacturing and trained on CPU: 64x64 image -> 64 patches of 8x8 -> 2-layer transformer encoder -> class, ~113k parameters, trained on CPU in minutes. Per-class F1 (TinyViT): crazing 85.71% · patches 91.23% · pitted_surface 75.86% · inclusion 68.85% · scratches 68.97% · rolled-in_scale 66.67%. **Honest reading:** the small CNN wins by ~9 points — convolutional inductive bias beats a from-scratch transformer on a 1,440-image dataset. The ViT is still the point of the project: 113k parameters, no pretraining, CPU-trained in ~14 minutes, and 76% is far above the 16.7% chance level. On larger industrial datasets (or with a conv stem) the gap closes; see "What I'd try next" in the notebook.

译文排队中,当前展示来源原文。

图像分类license:cc-by-sa-4.0
上下文窗口来源未披露
输入 → 输出图像 → 待核验
模型权重待核验
档案更新时间2026/11/2 19:16:43
OVERVIEW

模型介绍

A **tiny Vision Transformer trained from scratch** (no pretraining, no ImageNet) that classifies steel surface-defect images into six defect classes. The point of the project is to show a ViT *can* be built for manufacturing and trained on CPU: 64x64 image -> 64 patches of 8x8 -> 2-layer transformer encoder -> class, ~113k parameters, trained on CPU in minutes. Per-class F1 (TinyViT): crazing 85.71% · patches 91.23% · pitted_surface 75.86% · inclusion 68.85% · scratches 68.97% · rolled-in_scale 66.67%. **Honest reading:** the small CNN wins by ~9 points — convolutional inductive bias beats a from-scratch transformer on a 1,440-image dataset. The ViT is still the point of the project: 113k parameters, no pretraining, CPU-trained in ~14 minutes, and 76% is far above the 16.7% chance level. On larger industrial datasets (or with a conv stem) the gap closes; see "What I'd try next" in the notebook.

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

许可证:cc-by-sa-4.0。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。