ArcMa commited on
Commit ·
68b7b14
1
Parent(s): fc58215
first
Browse files- Dockerfile +11 -0
- requirements.txt +7 -0
- run.py +70 -0
Dockerfile
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
FROM python:3.9
|
| 2 |
+
|
| 3 |
+
WORKDIR /code
|
| 4 |
+
|
| 5 |
+
COPY ./requirements.txt /code/requirements.txt
|
| 6 |
+
|
| 7 |
+
RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
|
| 8 |
+
|
| 9 |
+
COPY . .
|
| 10 |
+
|
| 11 |
+
CMD ["python", "run.py"]
|
requirements.txt
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
requests
|
| 2 |
+
transformers
|
| 3 |
+
torch
|
| 4 |
+
Pillow
|
| 5 |
+
flask
|
| 6 |
+
werkzeug
|
| 7 |
+
|
run.py
ADDED
|
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import requests
|
| 2 |
+
import base64
|
| 3 |
+
import io
|
| 4 |
+
from transformers import TrOCRProcessor, VisionEncoderDecoderModel
|
| 5 |
+
from PIL import Image
|
| 6 |
+
import torch
|
| 7 |
+
import time
|
| 8 |
+
|
| 9 |
+
processor = TrOCRProcessor.from_pretrained("arcma/decap")
|
| 10 |
+
model = VisionEncoderDecoderModel.from_pretrained("arcma/decap")
|
| 11 |
+
model.eval()
|
| 12 |
+
|
| 13 |
+
def check(x):
|
| 14 |
+
if len(x) < 6:
|
| 15 |
+
return False
|
| 16 |
+
if not set(x).issubset('1234567890abcdefghijklmnopqrstuvwxyz'):
|
| 17 |
+
return False
|
| 18 |
+
return True
|
| 19 |
+
|
| 20 |
+
def process_image(image):
|
| 21 |
+
pixel_values = processor(image, return_tensors="pt").pixel_values
|
| 22 |
+
with torch.no_grad():
|
| 23 |
+
generated_ids = model.generate(pixel_values, num_beams=4, num_return_sequences=4)
|
| 24 |
+
generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)
|
| 25 |
+
generated_text = [x for x in generated_text if check(x)]
|
| 26 |
+
return generated_text[0]
|
| 27 |
+
|
| 28 |
+
def process_html(html):
|
| 29 |
+
|
| 30 |
+
orig_im = Image.open(
|
| 31 |
+
io.BytesIO(
|
| 32 |
+
base64.b64decode(
|
| 33 |
+
html
|
| 34 |
+
.partition('''" style="background:white url('data:image/jpg;base64,''')[2]
|
| 35 |
+
.partition("') no-repeat")[0]
|
| 36 |
+
)
|
| 37 |
+
)
|
| 38 |
+
)
|
| 39 |
+
|
| 40 |
+
return process_image(orig_im)
|
| 41 |
+
|
| 42 |
+
|
| 43 |
+
|
| 44 |
+
from werkzeug.wrappers import Request, Response
|
| 45 |
+
from flask import Flask, request
|
| 46 |
+
from flask import jsonify
|
| 47 |
+
|
| 48 |
+
app = Flask(__name__)
|
| 49 |
+
|
| 50 |
+
@app.route("/", methods=['POST', 'OPTIONS'])
|
| 51 |
+
def hello():
|
| 52 |
+
try:
|
| 53 |
+
return jsonify({
|
| 54 |
+
'x': process_html(request.json['data'])
|
| 55 |
+
})
|
| 56 |
+
except:
|
| 57 |
+
print('fail')
|
| 58 |
+
return "Hello World!"
|
| 59 |
+
|
| 60 |
+
@app.after_request
|
| 61 |
+
def after_request(response):
|
| 62 |
+
response.headers.add("Access-Control-Allow-Origin", "*")
|
| 63 |
+
response.headers.add("Access-Control-Allow-Credentials", "true")
|
| 64 |
+
response.headers.add("Access-Control-Allow-Methods", "GET,HEAD,OPTIONS,POST,PUT")
|
| 65 |
+
response.headers.add("Access-Control-Allow-Headers", "Access-Control-Allow-Headers, Origin,Accept, X-Requested-With, Content-Type, Access-Control-Request-Method, Access-Control-Request-Headers")
|
| 66 |
+
return response
|
| 67 |
+
|
| 68 |
+
if __name__ == '__main__':
|
| 69 |
+
from werkzeug.serving import run_simple
|
| 70 |
+
run_simple('0.0.0.0', 7860, app)
|