[Feat] Update for paddleocr 3.5#17820
Merged
Bobholamovic merged 59 commits intoPaddlePaddle:mainfrom Apr 21, 2026
Merged
Conversation
|
Thanks for your contribution! |
… into feat/engine
Per review feedback from changdazhou on PR PaddlePaddle#17820 (L26), update the CUDA 12.6 Docker GPU line to require driver >= 550.54.14, matching the pip section already at L61 (both ZH and EN). Signed-off-by: Bvicii <yizhanhuang2002@gmail.com>
Per follow-up review on PR PaddlePaddle#17820: from a completeness standpoint, None belongs in the "Supports ..." enumeration rather than only in the trailing clarification sentence. Move None into the list as the default value and tighten the follow-on sentence accordingly. - EN: "Supports None (the default), paddle, paddle_static, paddle_dynamic, and transformers. When left as None, PaddleOCR preserves the behavior of earlier versions..." - ZH: "支持 None（默认值）、paddle、paddle_static、paddle_dynamic、 transformers。保持为默认值 None 时..." Applied to all three supported-value variants across the module_usage and pipeline_usage pages — same 48 files / 66 rows as the previous clarification commit. Signed-off-by: Bvicii <yizhanhuang2002@gmail.com>
Resolves conflict in docs/version3.x/pipeline_usage/PaddleOCR-VL.md:
- Accept upstream refactor of CLI and Python instantiation parameter
tables from HTML to markdown pipe-table format.
- Preserve the {#流程导览} anchor on the "流程导览" heading (needed
for mkdocs bilingual link check).
- Re-apply the engine-row clarification (None as default + legacy
behavior note) to the two engine rows in the new pipe-tables.
Incoming commits:
- a874bcb Optimize docs
- 85275d4 Update docs
Signed-off-by: Bvicii <yizhanhuang2002@gmail.com>
docs: release-review fixes for 3.5 docs
cuicheng01
reviewed
Apr 20, 2026
Collaborator
cuicheng01
left a comment
cuicheng01
left a comment
There was a problem hiding this comment.
模型介绍部分的速度指标，建议增加一下说明？比如推理后端等
|
|
||
| 上述命令使用飞桨框架作为默认推理引擎，请在运行前确保相关依赖已经安装。 | ||
|
|
||
| 如果使用 `transformers` 作为推理引擎，可参考如下命令： |
Collaborator
There was a problem hiding this comment.
这一句建议改为如果选择 transformers 作为推理引擎，请确保已配置 PyTorch 环境，参考命令如下：xxx
下边增加一句描述：
- 注：默认的飞桨框架作为推理引擎在多数场景下拥有更出色的推理性能，推荐作为首选。（飞桨框架的描述，可以和下边的paddle_static或paddle_dynamic对齐）
| ## 五、FAQ | ||
| 训练后的模型如果想使用 `paddle_dynamic` 或 `transformers` 引擎，请参考后文 [推理引擎](#五推理引擎) 中的 [权重转换](#52-权重转换) 部分将模型由 `pdparams` 格式通过 PaddleX 转换为 `safetensors` 格式。 | ||
|
|
||
| ## 五、推理引擎 {#五推理引擎} |
Member
Author
There was a problem hiding this comment.
这里主要是为了保证mkdocs的link不出问题，mkdocs和github markdown语法不完全兼容，所以牺牲了markdown换mkdocs功能正常
| paddleocr chart_parsing -i "{'image': 'https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/chart_parsing_02.png'}" | ||
| ``` | ||
|
|
||
| 上述命令使用飞桨框架作为默认推理引擎，请在运行前确保相关依赖已经安装。 |
Collaborator
There was a problem hiding this comment.
这里是不和下边的paddle_static或paddle_dynamic对齐，更具体一些，减少歧义。
| ) | ||
| # ocr = PaddleOCR(lang="en") # 通过 lang 参数来使用英文模型 | ||
| # ocr = PaddleOCR(ocr_version="PP-OCRv4") # 通过 ocr_version 参数来使用 PP-OCR 其他版本 | ||
| # ocr = PaddleOCR(ocr_version="PP-OCRv4", engine="transformers") # 通过 ocr_version 参数来使用 PP-OCR 其他版本 |
| print(f"PaddleOCR版本: {paddleocr.__version__}") | ||
|
|
||
| # 验证GPU是否可用 | ||
| # 若使用 Paddle 推理引擎，可继续验证 PaddlePaddle 与 GPU 是否可用 |
Collaborator
There was a problem hiding this comment.
飞桨框架、Paddle 推理引擎、Paddle_static等，现在的叫法比较多，最好统一一下
| 快速体验虽然可以让您感受到 PaddleOCR-VL 的效果，但在推理速度、显存占用等方面不是最佳状态。如果您希望将 PaddleOCR-VL 应用于生产环境，并且对推理性能有更高的要求，请阅读 [3. 使用推理加速框架提升 VLM 推理性能](#3-vlm) 或其他硬件文档中的对应章节。 | ||
| 1. **本地直接推理（快速体验 / 脚本集成）**： | ||
| 适用于在本机通过 PaddleOCR CLI 或 Python API 直接调用 PaddleOCR-VL。 | ||
| 这一类通常对应本地推理引擎方式，如 PaddlePaddle 或 Transformers。 |
Collaborator
There was a problem hiding this comment.
这里又多了一种叫法：PaddlePaddle
| 1. **希望快速体验 PaddleOCR-VL**： | ||
|
|
||
| 如果您希望快速体验 PaddleOCR-VL 的推理效果，请阅读 [1. 环境准备](#1) 和 [2. 快速开始](#2)，或其他硬件文档中的对应章节。 | ||
| 确认上述信息后，再按使用目标选择路径： |
Collaborator
There was a problem hiding this comment.
现在这块的内容太多了，我建议加一个流程图让用户选择，流程开始变的复杂，文字太多也不是很好理解。
Revert PaddleOCR-VL.md to the pre-a874bcb3e8 "Optimize docs" state, then re-apply the PR's three edits and a small set of the maintainer's prose-only changes from that commit. Full revert rationale: a874bcb converted every HTML table in the file to pipe-tables (17 tables) but the conversion had multiple rendering problems — unescaped `|` inside `str|None` type cells breaking column alignment, trailing rows collapsed into prose, a corrupted `prompt_label` row. The commit also stripped every `<details>/<summary>` wrapper (27 in total), so both the parameter tables and the Docker Compose customization / multi-language invocation examples stopped being collapsed, and it broke the YAML frontmatter (`comments: true` → `## comments: true`), the tabbed `=== "..."` blocks, and some HTML-entity escaping inside C++/PHP code samples. Edits preserved on top of the HTML baseline: - This PR's three edits: {#流程导览} anchor, and the engine-row completeness / default-None clarification on both engine rows. - Maintainer's prose edits from a874bcb that were clearly intentional and did not introduce rendering regressions: * Remove the paddleocr-vl-1.5_metrics.png image. * Expand the "镜像中仅预装飞桨框架" note with a recommendation to install other engines via 方法二. * Generalize the "latest-xxx" TIP wording. * Switch the paddlepaddle.org.cn quick-install URL from autolink to explicit Markdown link syntax. Other prose/structure changes from a874bcb (list-indent reflow inside tabs, `<details>` stripping, table conversions, frontmatter, code-sample re-encoding) are rendering regressions rather than intended edits and are rolled back here; maintainer can re-apply any genuinely desired prose changes cleanly on top of this baseline. Signed-off-by: Bvicii <yizhanhuang2002@gmail.com>
docs: restore PaddleOCR-VL.md rendering
… into feat/engine
cuicheng01
previously approved these changes
Apr 20, 2026
cuicheng01
approved these changes
Apr 20, 2026
cuicheng01
approved these changes
Apr 21, 2026
|
btw I think the image header in the README should be updated to say paddle 3.5 instead of 3.0 haha |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.