Comments (4)
As described in the research paper by Lukas Blecher et al., Nougat was trained mainly for English (which makes sense since most papers on Arxiv are in English -- English is still the world/common language), which means that other languages are unlikely to work exceptionally well. However, the paper mentions that Nougat works acceptably well for other Roman languages (Italian, German, French, etc.). Chinese is not a Roman language, and so as he mentions, this often results in repetitions (the missing page not found error).
If you want Nougat to recognize languages that are not Roman, such as Chinese, Japanese, or languages with the Cyrillic alphabet, the model would have to be fine-tuned. I'm working on a project to make preparing the training data for fine-tuning Nougat easier.
from nougat.
@marwinsteiner Hi, may I inquire where I can find the project related to preparing training data for fine-tuning Nougat? I'm highly interested. :)
@XixuHu It is a currently private repo under my name. It is a WIP, does not currently support generation of finetuning datasets for nougat-ocr
. I'm still trying to figure out how I cand o that. However, this is the plan, with a light Streamlit frontend so you can choose some parameters like which language(s) you want, how many pages of training data, etc.
If you want to collaborate lmk \sorry for late reply
from nougat.
@marwinsteiner Hi, may I inquire where I can find the project related to preparing training data for fine-tuning Nougat? I'm highly interested. :)
from nougat.
@marwinsteiner how do things work? im ready to help if you need :)
from nougat.
Related Issues (20)
- Have anybody trained a Chinese version?
- Why citations come out like this?
- Why can't I run nougat-ocr on pdfs? HOT 3
- What does this issue mean?
- TypeError: BARTDecoder.prepare_inputs_for_inference() got an unexpected keyword argument 'cache_position' HOT 11
- Big Bug: Can not detect some kinds of whitespace in arxiv computer science paper.
- UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:3587.) return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined] and TypeError: BARTDecoder.prepare_inputs_for_inference() got an unexpected keyword argument 'cache_position' -> Cannot close object, library is destroyed. This may cause a memory leak! HOT 5
- Questions Regarding the Nougat Model's Pre-training Process
- pydantic error HOT 17
- The command to generate mmd from pdf, the output result is empty
- TypeError: BARTDecoder HOT 3
- Low amount of recognised pages
- pydantic_core._pydantic_core.ValidationError: 1 validation error for InitSchema | Windows, Python 3.11.5 HOT 3
- Training set format problem
- nought项目可以下载直接部署到服务器对外暴露接口提供使用吗
- Why? Cannot close object, library is destroyed. This may cause a memory leak! HOT 3
- How to view .mmd format in better format ? HOT 1
- buy the license for the model
- Maintenance status; community fork HOT 8
- Input should be None [type=none_required, input_value=1.2, input_type=float] For further information visit https://errors.pydantic.dev/2.8/v/none_required HOT 4
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from nougat.