## Quickstart
Build and start Machine Learning backend on `http://localhost:9090`
```bash
docker-compose up
```
Check if it works:
```bash
$ curl http://localhost:9090/health
{"status":"UP"}
```
Then connect running backend to Label Studio using Machine Learning settings.
## Writing your own model
1. Place your scripts for model training & inference inside root directory. Follow the [API guidelines](#api-guidelines) described bellow. You can put everything in a single file, or create 2 separate one say `my_training_module.py` and `my_inference_module.py`
2. Write down your python dependencies in `requirements.txt`
3. Open `wsgi.py` and make your configurations under `init_model_server` arguments:
```python
from my_training_module import training_script
from my_inference_module import InferenceModel
init_model_server(
create_model_func=InferenceModel,
train_script=training_script,
...
```
4. Make sure you have docker & docker-compose installed on your system, then run
```bash
docker-compose up --build
```
## API guidelines
#### Inference module
In order to create module for inference, you have to declare the following class:
```python
from htx.base_model import BaseModel
# use BaseModel inheritance provided by pyheartex SDK
class MyModel(BaseModel):
# Describe input types (Label Studio object tags names)
INPUT_TYPES = ('Image',)
# Describe output types (Label Studio control tags names)
INPUT_TYPES = ('Choices',)
def load(self, resources, **kwargs):
"""Here you load the model into the memory. resources is a dict returned by training script"""
self.model_path = resources["model_path"]
self.labels = resources["labels"]
def predict(self, tasks, **kwargs):
"""Here you create list of model results with Label Studio's prediction format, task by task"""
predictions = []
for task in tasks:
# do inference...
predictions.append(task_prediction)
return predictions
```
#### Training module
Training could be made in a separate environment. The only one convention is that data iterator and working directory are specified as input arguments for training function which outputs JSON-serializable resources consumed later by `load()` function in inference module.
```python
def train(input_iterator, working_dir, **kwargs):
"""Here you gather input examples and output labels and train your model"""
resources = {"model_path": "some/model/path", "labels": ["aaa", "bbb", "ccc"]}
return resources
```
基于labelstudio智能标注码源.zip （90个子文件） 
基于labelstudio智能标注码源 
utils.py 25KB checkpoint model_best model_config.json 474B config.json 614B tokenizer_config.json 235B static inference.pdmodel 3.57MB inference.pdiparams.info 18KB inference.pdiparams 447.68MB special_tokens_map.json 112B model_state.pdparams 449.95MB
vocab.txt 182KB .vscode settings.json 924B dev.txt 14KB evaluate.py 4KB my_ml_backend 1659438492 event.json 8KB _wsgi.py 4KB 1659437585 event.json 10KB 1658821186 event.json 9KB 1658821389 event.json 12KB 1658821343 event.json 11KB 1658821204 event.json 9KB 1659437927 event.json 12KB 1658821283 event.json 9KB docker-compose.yml 563B 1658821585 event.json 7KB 1.1658818930 1659439152 train_data.json 51KB train_data_info.json 13B job_result.json 234B 1659440436 train_data.json 51KB train_data_info.json 13B job_result.json 234B 1659437038 train_data.json 30KB train_data_info.json 13B job_result.json 235B 1659440761 train_data.json 51KB train_data_info.json 13B job_result.json 235B 1659508608 train_data.json 51KB train_data_info.json 13B job_result.json 235B 1659438584 train_data.json 51KB train_data_info.json 13B job_result.json 233B 1659440190 train_data.json 51KB train_data_info.json 13B job_result.json 233B 1659439265 train_data.json 51KB train_data_info.json 13B job_result.json 235B 1659438995 train_data.json 51KB train_data_info.json 13B job_result.json 234B 1659438885 train_data.json 51KB train_data_info.json 13B job_result.json 235B 1659438151 train_data.json 45KB train_data_info.json 13B job_result.json 235B 1659437685 train_data.json 45KB train_data_info.json 13B job_result.json 235B 1658821790 train_data.json 30KB train_data_info.json 13B job_result.json 238B 1658821599 event.json 8KB 1659438569 event.json 6KB Dockerfile 267B 1659438555 event.json 8KB 1658821220 event.json 8KB requirements.txt 28B 1659438516 event.json 5KB 1659437558 event.json 16KB __pycache__ my_ml_backend.cpython-38.pyc 4KB README.md 3KB my_ml_backend.py 5KB 1658821566 event.json 11KB 1658821303 event.json 9KB 1659437662 event.json 13KB data dev.txt 15KB test.txt 0B train.txt 18KB main.ipynb 17KB finetune.py 8KB label-studio-data media upload 1 9e4d363f-weibodata_8.json 2KB 3164eae4-corpus.txt 4KB label_studio.sqlite3 1000KB model.py 2KB doccano_ext.jsonl 12KB doccano.py 8KB corpus.txt 4KB my_ml_backend.py 5KB资源评论
2301_768660712023-09-20资源值得借鉴的内容很多，那就浅学一下吧，值得下载！
我不是企鹅2023-07-10简直是宝藏资源，实用价值很高，支持！
weixin_440299002023-07-07发现一个宝藏资源，资源有很高的参考价值，赶紧学起来~- m0_741663112023-06-26资源内容详细全面，与描述一致，对我很有用，有一定的使用价值。
- m0_750400092023-04-10简直是宝藏资源，实用价值很高，支持！
汀、人工智能
- 粉丝: 5w+
- 资源: 277
