Skip to content

Clarify TensorRT download guidance #56

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
wimi321 merged 1 commit into from
Aug 28, 2026
Merged

Clarify TensorRT download guidance #56

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
23 changes: 22 additions & 1 deletion scripts/check_website.mjs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -109,7 +109,8 @@ for (const keyword of [
'safeGithubReleaseUrl',
'TensorRT 可选版',
'tensorrt-optional',
'RTX 20 / GTX 16 可选',
'RTX 30 / 40 / 50 优先使用 CUDA，通常更快',
'TensorRT 仅为 RTX 20 / GTX 16 提速',
'/\\.7z\\.(001|002)$/',
'usable.length !== requiredAssets',
'CPU 通用版',
Expand All @@ -130,6 +131,26 @@ for (const keyword of ['ROCm 實驗版', 'Experimental ROCm', 'ROCm 実験版',
}
if (downloadChooser.includes('tensorrt-advanced')) fail('download chooser must use the stable TensorRT catalog category')
if (downloadChooser.includes('RTX 30 系及以下可选')) fail('TensorRT recommendation must not target RTX 30')
for (const keyword of [
'TensorRT 僅供 RTX 20 / GTX 16 加速',
'TensorRT is for speeding up RTX 20 / GTX 16 only',
'TensorRT は RTX 20 / GTX 16 の高速化用のみ',
'TensorRT는 RTX 20 / GTX 16 가속용만',
'TensorRT ใช้เร่ง RTX 20 / GTX 16 เท่านั้น',
'TensorRT chỉ để tăng tốc RTX 20 / GTX 16',
]) {
if (!downloadChooser.includes(keyword)) fail(`download chooser must localize TensorRT guidance: ${keyword}`)
}
for (const keyword of [
'通常比 CUDA 更快',
'often faster than CUDA',
'CUDA より高速な場合あり',
'CUDA보다 빠른 경우가 많음',
'มักเร็วกว่า CUDA',
'thường nhanh hơn CUDA',
]) {
if (downloadChooser.includes(keyword)) fail(`download chooser must not imply TensorRT is generally faster: ${keyword}`)
}
for (const keyword of ['www.goagent.top', 'goagent.top', 'Response.redirect', 'env.ASSETS.fetch']) {
if (!edgeWorker.includes(keyword)) fail(`edge worker must contain: ${keyword}`)
}
Expand Down
14 changes: 7 additions & 7 deletions website/src/components/DownloadChooser.astro
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,7 +36,7 @@ const copyByLang = {
rows: {
nvidia: ['NVIDIA CUDA', 'RTX 20 / 30 / 40 / 50 统一版'],
amd: ['AMD RX 9000', 'ROCm 实验版 · RX 9060 / 9070 系列'],
tensorrt: ['TensorRT 可选版', 'RTX 20 / GTX 16 可选 · 通常比 CUDA 更快 · GTX 10 不支持 · 下载两个分卷'],
tensorrt: ['TensorRT 可选版', 'RTX 30 / 40 / 50 优先使用 CUDA，通常更快 · TensorRT 仅为 RTX 20 / GTX 16 提速 · GTX 10 不支持 · 下载两个分卷'],
opencl: ['OpenCL 兼容版', 'AMD / Intel 显卡'],
cpu: ['CPU 通用版', '没有独立显卡也能用'],
noEngine: ['无引擎版', '已有自己的 KataGo'],
Expand All @@ -50,47 +50,47 @@ const copyByLang = {
unavailable: '暫不可用', volume1: '分卷 1', volume2: '分卷 2', macTip: '下載後拖到「應用程式」即可安裝', updateTitle: '已經使用免安裝版？',
updateText: '只更新主程式', updateAction: '下載小更新', help: '安裝說明', history: '歷史版本', error: '版本清單暫時無法載入，請重試。',
retry: '重新載入', openDirect: '開啟 GitHub 備用下載', amdMore: 'RX 6000 / RX 7000 / Ryzen AI Max 實驗版在 GitHub；不確定型號請使用 OpenCL。', amdMoreAction: '查看其他 AMD 實驗版',
rows: { nvidia: ['NVIDIA CUDA', 'RTX 20 / 30 / 40 / 50 統一版'], amd: ['AMD RX 9000', 'ROCm 實驗版 · RX 9060 / 9070 系列'], tensorrt: ['TensorRT 選配版', 'RTX 20 / GTX 16 可選 · 通常比 CUDA 更快 · GTX 10 不支援 · 下載兩個分卷'], opencl: ['OpenCL 相容版', 'AMD / Intel 顯示卡'], cpu: ['CPU 通用版', '沒有獨立顯示卡也能用'], noEngine: ['無引擎版', '已有自己的 KataGo'], apple: ['Apple 晶片', 'M1 / M2 / M3 / M4 / M5'], intel: ['Intel 晶片', '舊款 Intel Mac'] }
rows: { nvidia: ['NVIDIA CUDA', 'RTX 20 / 30 / 40 / 50 統一版'], amd: ['AMD RX 9000', 'ROCm 實驗版 · RX 9060 / 9070 系列'], tensorrt: ['TensorRT 選配版', 'RTX 30 / 40 / 50 優先使用 CUDA，通常更快 · TensorRT 僅供 RTX 20 / GTX 16 加速 · GTX 10 不支援 · 下載兩個分卷'], opencl: ['OpenCL 相容版', 'AMD / Intel 顯示卡'], cpu: ['CPU 通用版', '沒有獨立顯示卡也能用'], noEngine: ['無引擎版', '已有自己的 KataGo'], apple: ['Apple 晶片', 'M1 / M2 / M3 / M4 / M5'], intel: ['Intel 晶片', '舊款 Intel Mac'] }
},
en: {
title: 'Choose your version', lead: 'Pick your computer, then download the matching build', loading: 'Loading the latest stable release…', release: 'Stable release',
windowsLead: 'Use CUDA for NVIDIA RTX 20 / 30 / 40 / 50; RX 9000 can try ROCm, while other graphics should use OpenCL', macLead: 'Choose the build for your Mac chip', download: 'Download', recommended: 'Recommended',
unavailable: 'Unavailable', volume1: 'Part 1', volume2: 'Part 2', macTip: 'Drag the app to Applications after downloading', updateTitle: 'Already use the portable edition?',
updateText: 'Update the app only', updateAction: 'Small update', help: 'Install help', history: 'Older releases', error: 'The version list could not be loaded. Please try again.',
retry: 'Reload', openDirect: 'Open GitHub backup', amdMore: 'Find RX 6000 / RX 7000 / Ryzen AI Max experimental builds on GitHub. If unsure, use OpenCL.', amdMoreAction: 'Other AMD experimental builds',
rows: { nvidia: ['NVIDIA CUDA', 'One build for RTX 20 / 30 / 40 / 50'], amd: ['AMD RX 9000', 'Experimental ROCm · RX 9060 / 9070 series'], tensorrt: ['Optional TensorRT', 'For RTX 20 / GTX 16 · often faster than CUDA · GTX 10 unsupported · download both parts'], opencl: ['OpenCL compatible', 'AMD / Intel graphics'], cpu: ['CPU universal', 'Works without dedicated graphics'], noEngine: ['No-engine build', 'For your own KataGo'], apple: ['Apple silicon', 'M1 / M2 / M3 / M4 / M5'], intel: ['Intel chip', 'Older Intel Macs'] }
rows: { nvidia: ['NVIDIA CUDA', 'One build for RTX 20 / 30 / 40 / 50'], amd: ['AMD RX 9000', 'Experimental ROCm · RX 9060 / 9070 series'], tensorrt: ['Optional TensorRT', 'Use CUDA on RTX 30 / 40 / 50; it is usually faster · TensorRT is for speeding up RTX 20 / GTX 16 only · GTX 10 unsupported · download both parts'], opencl: ['OpenCL compatible', 'AMD / Intel graphics'], cpu: ['CPU universal', 'Works without dedicated graphics'], noEngine: ['No-engine build', 'For your own KataGo'], apple: ['Apple silicon', 'M1 / M2 / M3 / M4 / M5'], intel: ['Intel chip', 'Older Intel Macs'] }
},
ja: {
title: 'バージョンを選択', lead: 'パソコンを選び、環境に合う版をダウンロード', loading: '最新の安定版を取得中…', release: '安定版',
windowsLead: 'NVIDIA RTX 20 / 30 / 40 / 50 は CUDA、AMD RX 9000 は ROCm を試用可能、その他は OpenCL を使用', macLead: 'Mac のチップに合う版を選択', download: 'ダウンロード', recommended: 'おすすめ',
unavailable: '利用不可', volume1: '分割 1', volume2: '分割 2', macTip: 'ダウンロード後「アプリケーション」へドラッグ', updateTitle: 'ポータブル版を利用中ですか？',
updateText: '本体だけ更新', updateAction: '小型更新', help: 'インストール案内', history: '過去の版', error: 'バージョン一覧を読み込めませんでした。再試行してください。',
retry: '再読み込み', openDirect: 'GitHub の予備ダウンロードを開く', amdMore: 'RX 6000 / RX 7000 / Ryzen AI Max 実験版は GitHub から取得できます。型番が不明なら OpenCL を使用してください。', amdMoreAction: 'その他の AMD 実験版',
rows: { nvidia: ['NVIDIA CUDA', 'RTX 20 / 30 / 40 / 50 共通版'], amd: ['AMD RX 9000', 'ROCm 実験版 · RX 9060 / 9070 シリーズ'], tensorrt: ['TensorRT オプション版', 'RTX 20 / GTX 16 向け · CUDA より高速な場合あり · GTX 10 非対応 · 両方の分割を取得'], opencl: ['OpenCL 互換版', 'AMD / Intel GPU'], cpu: ['CPU 汎用版', '専用 GPU なしでも利用可能'], noEngine: ['エンジンなし', '自分の KataGo を使用'], apple: ['Apple シリコン', 'M1 / M2 / M3 / M4 / M5'], intel: ['Intel チップ', '旧型 Intel Mac'] }
rows: { nvidia: ['NVIDIA CUDA', 'RTX 20 / 30 / 40 / 50 共通版'], amd: ['AMD RX 9000', 'ROCm 実験版 · RX 9060 / 9070 シリーズ'], tensorrt: ['TensorRT オプション版', 'RTX 30 / 40 / 50 は通常より高速な CUDA を優先 · TensorRT は RTX 20 / GTX 16 の高速化用のみ · GTX 10 非対応 · 両方の分割を取得'], opencl: ['OpenCL 互換版', 'AMD / Intel GPU'], cpu: ['CPU 汎用版', '専用 GPU なしでも利用可能'], noEngine: ['エンジンなし', '自分の KataGo を使用'], apple: ['Apple シリコン', 'M1 / M2 / M3 / M4 / M5'], intel: ['Intel チップ', '旧型 Intel Mac'] }
},
ko: {
title: '버전 선택', lead: '컴퓨터를 고른 뒤 하드웨어에 맞는 버전을 받으세요', loading: '최신 안정판을 불러오는 중…', release: '안정판',
windowsLead: 'NVIDIA RTX 20 / 30 / 40 / 50은 CUDA, AMD RX 9000은 ROCm을 시험할 수 있으며 나머지는 OpenCL을 사용하세요', macLead: 'Mac 칩에 맞는 버전을 선택하세요', download: '다운로드', recommended: '추천',
unavailable: '사용 불가', volume1: '분할 1', volume2: '분할 2', macTip: '다운로드 후 응용 프로그램으로 드래그하세요', updateTitle: '포터블 버전을 이미 사용 중인가요?',
updateText: '프로그램만 업데이트', updateAction: '작은 업데이트', help: '설치 도움말', history: '이전 버전', error: '버전 목록을 불러오지 못했습니다. 다시 시도해 주세요.',
retry: '다시 불러오기', openDirect: 'GitHub 예비 다운로드 열기', amdMore: 'RX 6000 / RX 7000 / Ryzen AI Max 실험판은 GitHub에서 받을 수 있습니다. 모델을 모르면 OpenCL을 사용하세요.', amdMoreAction: '다른 AMD 실험판 보기',
rows: { nvidia: ['NVIDIA CUDA', 'RTX 20 / 30 / 40 / 50 통합판'], amd: ['AMD RX 9000', 'ROCm 실험판 · RX 9060 / 9070 시리즈'], tensorrt: ['TensorRT 선택형', 'RTX 20 / GTX 16용 · CUDA보다 빠른 경우가 많음 · GTX 10 미지원 · 두 파일 모두 다운로드'], opencl: ['OpenCL 호환', 'AMD / Intel 그래픽'], cpu: ['CPU 범용', '외장 그래픽 없이 사용'], noEngine: ['엔진 없음', '자체 KataGo 사용'], apple: ['Apple 실리콘', 'M1 / M2 / M3 / M4 / M5'], intel: ['Intel 칩', '구형 Intel Mac'] }
rows: { nvidia: ['NVIDIA CUDA', 'RTX 20 / 30 / 40 / 50 통합판'], amd: ['AMD RX 9000', 'ROCm 실험판 · RX 9060 / 9070 시리즈'], tensorrt: ['TensorRT 선택형', 'RTX 30 / 40 / 50은 보통 더 빠른 CUDA 우선 · TensorRT는 RTX 20 / GTX 16 가속용만 · GTX 10 미지원 · 두 파일 모두 다운로드'], opencl: ['OpenCL 호환', 'AMD / Intel 그래픽'], cpu: ['CPU 범용', '외장 그래픽 없이 사용'], noEngine: ['엔진 없음', '자체 KataGo 사용'], apple: ['Apple 실리콘', 'M1 / M2 / M3 / M4 / M5'], intel: ['Intel 칩', '구형 Intel Mac'] }
},
th: {
title: 'เลือกเวอร์ชันของคุณ', lead: 'เลือกคอมพิวเตอร์ แล้วดาวน์โหลดรุ่นที่ตรงกับฮาร์ดแวร์', loading: 'กำลังโหลดรุ่นเสถียรล่าสุด…', release: 'รุ่นเสถียร',
windowsLead: 'ใช้ CUDA กับ NVIDIA RTX 20 / 30 / 40 / 50, ทดลอง ROCm ได้กับ AMD RX 9000 และใช้ OpenCL กับรุ่นอื่น', macLead: 'เลือกรุ่นให้ตรงกับชิป Mac', download: 'ดาวน์โหลด', recommended: 'แนะนำ',
unavailable: 'ยังใช้ไม่ได้', volume1: 'ส่วน 1', volume2: 'ส่วน 2', macTip: 'ดาวน์โหลดแล้วลากไปยัง Applications', updateTitle: 'ใช้รุ่นพกพาอยู่แล้ว?',
updateText: 'อัปเดตเฉพาะโปรแกรม', updateAction: 'อัปเดตขนาดเล็ก', help: 'วิธีติดตั้ง', history: 'รุ่นก่อนหน้า', error: 'โหลดรายการเวอร์ชันไม่สำเร็จ โปรดลองอีกครั้ง',
retry: 'โหลดใหม่', openDirect: 'เปิดดาวน์โหลดสำรองบน GitHub', amdMore: 'รุ่นทดลองสำหรับ RX 6000 / RX 7000 / Ryzen AI Max อยู่บน GitHub หากไม่แน่ใจรุ่น ให้ใช้ OpenCL', amdMoreAction: 'รุ่นทดลอง AMD อื่น',
rows: { nvidia: ['NVIDIA CUDA', 'รุ่นเดียวสำหรับ RTX 20 / 30 / 40 / 50'], amd: ['AMD RX 9000', 'ROCm รุ่นทดลอง · RX 9060 / 9070'], tensorrt: ['TensorRT ทางเลือก', 'สำหรับ RTX 20 / GTX 16 · มักเร็วกว่า CUDA · ไม่รองรับ GTX 10 · ดาวน์โหลดทั้งสองส่วน'], opencl: ['รุ่น OpenCL', 'การ์ดจอ AMD / Intel'], cpu: ['รุ่น CPU', 'ใช้ได้โดยไม่มีการ์ดจอแยก'], noEngine: ['ไม่มีเอนจิน', 'สำหรับ KataGo ของคุณเอง'], apple: ['ชิป Apple', 'M1 / M2 / M3 / M4 / M5'], intel: ['ชิป Intel', 'Mac Intel รุ่นเก่า'] }
rows: { nvidia: ['NVIDIA CUDA', 'รุ่นเดียวสำหรับ RTX 20 / 30 / 40 / 50'], amd: ['AMD RX 9000', 'ROCm รุ่นทดลอง · RX 9060 / 9070'], tensorrt: ['TensorRT ทางเลือก', 'RTX 30 / 40 / 50 ควรใช้ CUDA ซึ่งมักเร็วกว่า · TensorRT ใช้เร่ง RTX 20 / GTX 16 เท่านั้น · ไม่รองรับ GTX 10 · ดาวน์โหลดทั้งสองส่วน'], opencl: ['รุ่น OpenCL', 'การ์ดจอ AMD / Intel'], cpu: ['รุ่น CPU', 'ใช้ได้โดยไม่มีการ์ดจอแยก'], noEngine: ['ไม่มีเอนจิน', 'สำหรับ KataGo ของคุณเอง'], apple: ['ชิป Apple', 'M1 / M2 / M3 / M4 / M5'], intel: ['ชิป Intel', 'Mac Intel รุ่นเก่า'] }
},
vi: {
title: 'Chọn phiên bản', lead: 'Chọn máy tính rồi tải bản phù hợp với phần cứng', loading: 'Đang tải bản ổn định mới nhất…', release: 'Bản ổn định',
windowsLead: 'Dùng CUDA cho NVIDIA RTX 20 / 30 / 40 / 50; RX 9000 có thể thử ROCm, các GPU khác dùng OpenCL', macLead: 'Chọn bản phù hợp với chip Mac', download: 'Tải xuống', recommended: 'Khuyên dùng',
unavailable: 'Chưa có', volume1: 'Phần 1', volume2: 'Phần 2', macTip: 'Sau khi tải, kéo ứng dụng vào Applications', updateTitle: 'Đã dùng bản portable?',
updateText: 'Chỉ cập nhật chương trình', updateAction: 'Bản cập nhật nhỏ', help: 'Hướng dẫn cài', history: 'Phiên bản cũ', error: 'Không tải được danh sách phiên bản. Hãy thử lại.',
retry: 'Tải lại', openDirect: 'Mở bản tải dự phòng trên GitHub', amdMore: 'Các bản thử nghiệm RX 6000 / RX 7000 / Ryzen AI Max có trên GitHub. Nếu không chắc, hãy dùng OpenCL.', amdMoreAction: 'Các bản AMD thử nghiệm khác',
rows: { nvidia: ['NVIDIA CUDA', 'Một bản cho RTX 20 / 30 / 40 / 50'], amd: ['AMD RX 9000', 'ROCm thử nghiệm · dòng RX 9060 / 9070'], tensorrt: ['TensorRT tùy chọn', 'Dành cho RTX 20 / GTX 16 · thường nhanh hơn CUDA · không hỗ trợ GTX 10 · tải cả hai phần'], opencl: ['Tương thích OpenCL', 'Đồ họa AMD / Intel'], cpu: ['CPU phổ thông', 'Dùng được không cần GPU rời'], noEngine: ['Không kèm engine', 'Dùng KataGo riêng'], apple: ['Chip Apple', 'M1 / M2 / M3 / M4 / M5'], intel: ['Chip Intel', 'Mac Intel đời cũ'] }
rows: { nvidia: ['NVIDIA CUDA', 'Một bản cho RTX 20 / 30 / 40 / 50'], amd: ['AMD RX 9000', 'ROCm thử nghiệm · dòng RX 9060 / 9070'], tensorrt: ['TensorRT tùy chọn', 'RTX 30 / 40 / 50 ưu tiên CUDA vì thường nhanh hơn · TensorRT chỉ để tăng tốc RTX 20 / GTX 16 · không hỗ trợ GTX 10 · tải cả hai phần'], opencl: ['Tương thích OpenCL', 'Đồ họa AMD / Intel'], cpu: ['CPU phổ thông', 'Dùng được không cần GPU rời'], noEngine: ['Không kèm engine', 'Dùng KataGo riêng'], apple: ['Chip Apple', 'M1 / M2 / M3 / M4 / M5'], intel: ['Chip Intel', 'Mac Intel đời cũ'] }
}
} as const

Expand Down
Loading