Spaces:
Paused
Paused
File size: 4,133 Bytes
7c99e50 1ca62bc f52050f 7c99e50 f52050f ea69d06 7c99e50 f52050f 7c99e50 f52050f 2fecabe 7c99e50 4f59075 f52050f ec4bd6c f52050f ec4bd6c f52050f bb83a45 f52050f bb83a45 f52050f 7c99e50 f52050f 19d814b 9510f36 bb83a45 9510f36 bb83a45 f52050f 4222109 7c99e50 964f66f 7c99e50 3c2e156 7c99e50 4222109 c659e77 25d42ca 19d814b 7c99e50 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 |
import re
import gradio as gr
import numpy as np
from transformers import Qwen2VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info
import torch
from ast import literal_eval
from PIL import Image
# default: Load the model on the available device(s)
model = Qwen2VLForConditionalGeneration.from_pretrained(
"Qwen/Qwen2-VL-7B-Instruct", torch_dtype="auto", device_map="auto"
)
# default processer
processor = AutoProcessor.from_pretrained("Qwen/Qwen2-VL-7B-Instruct")
other_benifits = '''Extract the following information in the given format:
{'other_benefits_and_information': {
'401k eru: {'This Period':'', 'Year-to-Date':''}},
'quota summary':
{
'sick:': '',
'vacation:': '',
}
'payment method': 'eg. Direct payment',
'Amount': 'eg. 12.99'
}
'''
tax_deductions = '''Extract the following information in the given format:
{
'tax_deductions': {
'federal:': {
'withholding tax:': {'Amount':'', 'Year-To_Date':""},
'ee social security tax:': {'Amount':'', 'Year-To_Date':""},
'ee medicare tax:': {'Amount':'', 'Year-To_Date':""}},
'california:': {
'withholding tax:': {'Amount':'', 'Year-To_Date':""},
'ee disability tax:': {'Amount':'', 'Year-To_Date':""}}},
}
'''
def demo(image_name, prompt):
print("Inside Demo")
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": image_name,
},
{"type": "text", "text": prompt},
],
}
]
print(f"Formulated prompt template {messages}")
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Inference: Generation of the output
generated_ids = model.generate(**inputs, max_new_tokens=1500)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
try:
# almost_json = output_text[0].replace('```\n', '').replace('\n```', '')
almost_json = output_text[0].split('```\n')[-1].split('\n```')[0]
json = literal_eval(almost_json)
except:
try:
# almost_json = output_text[0].replace('```json\n', '').replace('\n```', '')
almost_json = output_text[0].split('```json\n')[-1].split('\n```')[0]
json = literal_eval(almost_json)
except:
json = output_text[0]
return json
def process_document(image):
print(f"Received Image --->>>>>> {image}")
if isinstance(image, np.ndarray):
print("Image is in Numpy array")
image = Image.fromarray(image)
print(type(image))
print("Proceeding with the demo")
one = demo(image, other_benifits)
two = demo(image, tax_deductions)
json_op = {
"tax_deductions": one,
"other_benifits": two
}
return json_op
# article = "<p style='text-align: center'><a href='https://www.xelpmoc.in/' target='_blank'>Made by Xelpmoc</a></p>"
demo = gr.Interface(
fn=process_document,
inputs=gr.Image(type="pil"),
outputs="json",
title="PaySlip_Demo_Model",
# article=article,
# enable_queue=True,
examples=["Slip_1.jpg", "Slip_2.jpg"],
cache_examples=False)
demo.launch() |