Update files
This commit is contained in:
三洋三洋
19
README.md
19
README.md
@@ -1,6 +1,6 @@
|
||||
<div align="center">
|
||||
<h1><img src="./assets/fire.svg" width=30, height=30>
|
||||
𝐓𝐞𝐱𝐓𝐞𝐥𝐥𝐞𝐫 <img src="./assets/fire.svg" width=30, height=30> </h1>
|
||||
𝚃𝚎𝚡𝚃𝚎𝚕𝚕𝚎𝚛 <img src="./assets/fire.svg" width=30, height=30> </h1>
|
||||
|
||||
<p align="center">
|
||||
English | <a href="./assests/README_zh.md">中文版本</a>
|
||||
@@ -23,7 +23,7 @@ python=3.10
|
||||
|
||||
pytorch
|
||||
|
||||
> Note: CUDA version >= 12.0 have been fully tested.
|
||||
> Note: Only CUDA version >= 12.0 have been fully tested, so we recommend using CUDA version>=12.0
|
||||
|
||||
## Getting Started
|
||||
|
||||
@@ -33,7 +33,7 @@ pytorch
|
||||
git clone https://github.com/OleehyO/TexTeller
|
||||
```
|
||||
|
||||
2. After [pytorch installation](https://pytorch.org/get-started), install the required packages:
|
||||
2. After [pytorch installation](https://pytorch.org/get-started/locally/#start-locally), install the required packages:
|
||||
|
||||
```bash
|
||||
pip install -r requirements.txt
|
||||
@@ -47,7 +47,7 @@ pytorch
|
||||
#+e.g. python inference.py -img "./img.jpg" -cuda
|
||||
```
|
||||
|
||||
> chekpoints will be downloaded in your first run.
|
||||
> Checkpoints will be downloaded in your first run.
|
||||
|
||||
## Web Demo
|
||||
|
||||
@@ -57,9 +57,9 @@ You can also run the web demo by navigating to the `TexTeller/src` directory and
|
||||
./start_web.sh
|
||||
```
|
||||
|
||||
Then go to `http://localhost:8000` in your browser to run TexTeller in the web.
|
||||
Then go to `http://localhost:8501` in your browser to run TexTeller in the web.
|
||||
|
||||
> You can change the default settings in start_web.sh, such as inference with GPU(e.g. `USE_CUDA=True`) or increase the number of beams(e.g. `NUM_BEAM=3`) for higher accuracy.
|
||||
> You can change the default settings in `start_web.sh`, such as inference with GPU(e.g. `USE_CUDA=True`) or increase the number of beams(e.g. `NUM_BEAM=3`) for higher accuracy.
|
||||
|
||||
## API
|
||||
|
||||
@@ -82,13 +82,13 @@ You can pass the following arguments to the `serve.py` script to get custom infe
|
||||
| `--ncpu_per_replica` | Number of CPU cores to use per replica, *default is 1*. |
|
||||
| `--ngpu_per_replica` | Number of GPUs to use per replica, *default is 1*. You can set this to 0~1 to run multiple replicas on a single GPU(if --num_replicas 2, --ngpu_per_replica 0.7, then 2 gpus are required) |
|
||||
|
||||
> Client demo can be found in `TexTeller/client/demo.py`.
|
||||
> Client demo can be found in `TexTeller/client/demo.py`, you can refer to `demo.py` to send requests to the server.
|
||||
|
||||
## Training
|
||||
|
||||
### Dataset
|
||||
|
||||
We provide a dataset example in `TexTeller/src/models/ocr_model/train/dataset`, and you can place your own images in the `images` directory and annotate the corresponding formula for each image in `formulas.jsonl`
|
||||
We provide a dataset example in `TexTeller/src/models/ocr_model/train/dataset`, and you can place your own images in the `images` directory and annotate the corresponding formula for each image in `formulas.jsonl`.
|
||||
|
||||
After the dataset is ready, you should **change the `DIR_URL` variable** in `.../dataset/loader.py` to the path of your dataset.
|
||||
|
||||
@@ -96,7 +96,7 @@ After the dataset is ready, you should **change the `DIR_URL` variable** in `...
|
||||
|
||||
If you are using a different dataset, you may need to retrain the tokenizer to match your specific vocabulary. After setting up the dataset, you can do this by:
|
||||
|
||||
1. Change the line `new_tokenizer.save_pretrained('./your_dir_name')` in `TexTeller/src/models/ocr_model/tokenizer/train.py` to your desired output directory name.`
|
||||
1. Change the line `new_tokenizer.save_pretrained('./your_dir_name')` in `TexTeller/src/models/ocr_model/tokenizer/train.py` to your desired output directory name.
|
||||
> To use a different vocabulary size, you should modify the `VOCAB_SIZE` parameter in the `TexTeller/src/models/globals.py`.
|
||||
|
||||
2. Running the following command **under `TexTeller/src` directory**:
|
||||
@@ -114,6 +114,7 @@ python -m models.ocr_model.train.train
|
||||
```
|
||||
|
||||
You can set your own tokenizer and checkpoint path(or fine-tune the default model checkpoint if you don't use your own tokenizer while keeping the same model architecture) in `TexTeller/src/models/ocr_model/train/train.py`.
|
||||
> Please refer to `train.py` for more details.
|
||||
|
||||
Model architecture and training hyperparameters can be adjusted in `TexTeller/src/globals.py` and `TexTeller/src/models/ocr_model/train/train_args.py`.
|
||||
|
||||
|
||||
162
assets/README_zh.md
Normal file
162
assets/README_zh.md
Normal file
@@ -0,0 +1,162 @@
|
||||
<div align="center">
|
||||
<h1><img src="./fire.svg" width=30, height=30>
|
||||
𝚃𝚎𝚡𝚃𝚎𝚕𝚕𝚎𝚛 <img src="./fire.svg" width=30, height=30> </h1>
|
||||
|
||||
<p align="center">
|
||||
<a href="../README.md">English</a> | 中文版本
|
||||
</p>
|
||||
|
||||
<p align="center">
|
||||
<img src="./web_demo.gif" alt="TexTeller_demo" width=800>
|
||||
</p>
|
||||
|
||||
</div>
|
||||
|
||||
TexTeller是一个基于ViT的端到端公式识别模型，可以把图片转换为对应的latex公式
|
||||
|
||||
TexTeller用了550K的图片-公式对进行训练(数据集可以在[这里](https://huggingface.co/datasets/OleehyO/latex-formulas)获取)，相比于[LaTeX-OCR](https://github.com/lukas-blecher/LaTeX-OCR)(使用了一个100K的数据集)，TexTeller具有更强的泛化能力以及更高的精确度，可以覆盖你大部分的使用场景。
|
||||
|
||||
> 我们马上就会发布一个使用5.5M数据集进行训练的TexTeller checkpoint
|
||||
|
||||
## 前置条件
|
||||
|
||||
python=3.10
|
||||
|
||||
pytorch
|
||||
|
||||
> 注意: 只有CUDA版本>= 12.0被完全测试过，所以最好使用>= 12.0的CUDA版本
|
||||
|
||||
|
||||
## Getting Started
|
||||
|
||||
1. 克隆本仓库:
|
||||
|
||||
```bash
|
||||
git clone https://github.com/OleehyO/TexTeller
|
||||
```
|
||||
|
||||
2. [安装pytorch](https://pytorch.org/get-started/locally/#start-locally)后，再安装本项目的依赖包:
|
||||
|
||||
```bash
|
||||
pip install -r requirements.txt
|
||||
```
|
||||
|
||||
3. 进入`TexTeller/src`目录，在终端运行以下命令进行推理:
|
||||
|
||||
```bash
|
||||
python inference.py -img "/path/to/image.{jpg,png}"
|
||||
# use -cuda option to enable GPU inference
|
||||
#+e.g. python inference.py -img "./img.jpg" -cuda
|
||||
```
|
||||
|
||||
> 第一次运行时会在hugging face上下载所需要的checkpoints
|
||||
|
||||
## FAQ：无法连接到Hugging Face
|
||||
|
||||
默认情况下，会在Hugging Face中下载模型权重，**如果你的远端服务器无法连接到Hugging Face**，你可以通过以下命令进行加载：
|
||||
|
||||
1. 安装huggingface hub包
|
||||
|
||||
```bash
|
||||
pip install -U "huggingface_hub[cli]"
|
||||
```
|
||||
|
||||
2. 在能连接Hugging Face的机器上下载模型权重:
|
||||
|
||||
```bash
|
||||
huggingface-cli download OleehyO/TexTeller --include "*.json" "*.bin" "*.txt" --repo-type model --local-dir "your/dir/path"
|
||||
```
|
||||
|
||||
3. 把包含权重的目录上传远端服务器，然后把`TexTeller/src/models/ocr_model/model/TexTeller.py`中的`REPO_NAME = 'OleehyO/TexTeller'`修改为`REPO_NAME = 'your/dir/path'`
|
||||
|
||||
如果你还想在训练模型时开启evaluate，你需要提前下载metric脚本并上传远端服务器：
|
||||
|
||||
1. 在能连接Hugging Face的机器上下载metric脚本
|
||||
|
||||
```bash
|
||||
huggingface-cli download evaluate-metric/google_bleu --repo-type space --local-dir "your/dir/path"
|
||||
```
|
||||
|
||||
2. 把这个目录上传远端服务器，并在`TexTeller/src/models/ocr_model/utils/metrics.py`中把`evaluate.load('google_bleu')`改为`evaluate.load('your/dir/path')`
|
||||
|
||||
## Web Demo
|
||||
|
||||
要想启动web demo，你需要先进入 `TexTeller/src` 目录，然后运行以下命令
|
||||
|
||||
```bash
|
||||
./start_web.sh
|
||||
```
|
||||
|
||||
然后在浏览器里输入`http://localhost:8501`就可以看到web demo
|
||||
|
||||
> 你可以改变`start_web.sh`的默认配置， 例如使用GPU进行推理(e.g. `USE_CUDA=True`) 或者增加beams的数量(e.g. `NUM_BEAM=3`)来获得更高的精确度
|
||||
|
||||
## API
|
||||
|
||||
我们使用[ray serve](https://github.com/ray-project/ray)来对外提供一个TexTeller的API接口，通过使用这个接口，你可以把TexTeller整合到自己的项目里。要想启动server，你需要先进入`TexTeller/src`目录然后运行以下命令:
|
||||
|
||||
```bash
|
||||
python serve.py # default settings
|
||||
```
|
||||
|
||||
你可以给`serve.py`传递以下参数来改变server的推理设置(e.g. `python serve.py --use_gpu` 来启动GPU推理):
|
||||
|
||||
| Argument | Description |
|
||||
| --- | --- |
|
||||
| `-ckpt` | Path to the checkpoint file to load, default is TexTeller pretrained model. |
|
||||
| `-tknz` | Path to the tokenizer, default is TexTeller tokenizer. |
|
||||
| `-port` | Port number to run the server on, *default is 8000*. |
|
||||
| `--use_gpu` | Whether to use GPU for inference. |
|
||||
| `--num_beams` | Number of beams to use for beam search decoding, *default is 1*. |
|
||||
| `--num_replicas` | Number of replicas to run the server on, *default is 1*. You can use this to get higher throughput. |
|
||||
| `--ncpu_per_replica` | Number of CPU cores to use per replica, *default is 1*. |
|
||||
| `--ngpu_per_replica` | Number of GPUs to use per replica, *default is 1*. You can set this to 0~1 to run multiple replicas on a single GPU(if --num_replicas 2, --ngpu_per_replica 0.7, then 2 gpus are required) |
|
||||
|
||||
> 一个客户端demo可以在`TexTeller/client/demo.py`找到，你可以参考`demo.py`来给server发送请求
|
||||
|
||||
## Training
|
||||
|
||||
### Dataset
|
||||
|
||||
我们在`TexTeller/src/models/ocr_model/train/dataset`目录中提供了一个数据集的例子，你可以把自己的图片放在`images`目录然后在`formulas.jsonl`中为每张图片标注对应的公式。
|
||||
|
||||
准备好数据集后，你需要在`.../dataset/loader.py`中把 **`DIR_URL`变量改成你自己数据集的路径**
|
||||
|
||||
### Retrain the tokenizer
|
||||
|
||||
如果你使用了不一样的数据集，你可能需要重新训练tokenizer来得到一个不一样的字典。配置好数据集后，可以通过以下命令来训练自己的tokenizer：
|
||||
|
||||
1. 在`TexTeller/src/models/ocr_model/tokenizer/train.py`中，修改`new_tokenizer.save_pretrained('./your_dir_name')`为你自定义的输出目录
|
||||
> 如果要用一个不一样大小的字典(默认1W个token)，你需要在 `TexTeller/src/models/globals.py`中修改`VOCAB_SIZE`变量
|
||||
|
||||
2. **在 `TexTeller/src` 目录下**运行以下命令:
|
||||
|
||||
```bash
|
||||
python -m models.ocr_model.tokenizer.train
|
||||
```
|
||||
|
||||
### Train the model
|
||||
|
||||
要想训练模型, 你需要在`TexTeller/src`目录下运行以下命令：
|
||||
|
||||
```bash
|
||||
python -m models.ocr_model.train.train
|
||||
```
|
||||
|
||||
你可以在`TexTeller/src/models/ocr_model/train/train.py`中设置自己的tokenizer和checkpoint路径（请参考`train.py`）。如果你使用了与TexTeller一样的架构和相同的字典，你还可以用自己的数据集来微调TexTeller的默认权重。
|
||||
|
||||
在`TexTeller/src/globals.py`和`TexTeller/src/models/ocr_model/train/train_args.py`中，你可以改变模型的架构以及训练的超参数。
|
||||
|
||||
> 我们的训练脚本使用了[Hugging Face Transformers](https://github.com/huggingface/transformers)库, 所以你可以参考他们提供的[文档](https://huggingface.co/docs/transformers/v4.32.1/main_classes/trainer#transformers.TrainingArguments)来获取更多训练参数的细节以及配置。
|
||||
|
||||
## To-Do
|
||||
|
||||
- [ ] 使用更大的数据集来训练模型(5.5M样本，即将发布)
|
||||
|
||||
- [ ] 推理加速
|
||||
|
||||
- [ ] ...
|
||||
|
||||
## Acknowledgements
|
||||
|
||||
Thanks to [LaTeX-OCR](https://github.com/lukas-blecher/LaTeX-OCR) which has brought me a lot of inspiration, and [im2latex-100K](https://zenodo.org/records/56198#.V2px0jXT6eA) which enriches our dataset.
|
||||
@@ -30,14 +30,14 @@ class TexTeller(VisionEncoderDecoderModel):
|
||||
|
||||
@classmethod
|
||||
def from_pretrained(cls, model_path: str = None):
|
||||
if model_path is None or model_path == cls.REPO_NAME:
|
||||
if model_path is None or model_path == 'default':
|
||||
return VisionEncoderDecoderModel.from_pretrained(cls.REPO_NAME)
|
||||
model_path = Path(model_path).resolve()
|
||||
return VisionEncoderDecoderModel.from_pretrained(str(model_path))
|
||||
|
||||
@classmethod
|
||||
def get_tokenizer(cls, tokenizer_path: str = None) -> RobertaTokenizerFast:
|
||||
if tokenizer_path is None or tokenizer_path == cls.REPO_NAME:
|
||||
if tokenizer_path is None or tokenizer_path == 'default':
|
||||
return RobertaTokenizerFast.from_pretrained(cls.REPO_NAME)
|
||||
tokenizer_path = Path(tokenizer_path).resolve()
|
||||
return RobertaTokenizerFast.from_pretrained(str(tokenizer_path))
|
||||
|
||||
@@ -1,9 +1,9 @@
|
||||
#!/usr/bin/env bash
|
||||
set -exu
|
||||
|
||||
export CHECKPOINT_DIR="OleehyO/TexTeller"
|
||||
export TOKENIZER_DIR="OleehyO/TexTeller"
|
||||
export USE_CUDA=True # True or False (case-sensitive)
|
||||
export NUM_BEAM=5
|
||||
export CHECKPOINT_DIR="default"
|
||||
export TOKENIZER_DIR="default"
|
||||
export USE_CUDA=False # True or False (case-sensitive)
|
||||
export NUM_BEAM=1
|
||||
|
||||
streamlit run web.py
|
||||
|
||||
Reference in New Issue
Block a user