Spaces:
Sleeping
Sleeping
fawadrashid
commited on
Commit
•
e7db0a6
1
Parent(s):
2389ef4
Upload 3 files
Browse files- Dockerfile +27 -0
- app2.py +43 -0
- requirements.txt +7 -0
Dockerfile
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
FROM python:3.9-slim
|
2 |
+
|
3 |
+
|
4 |
+
RUN useradd -m -u 1000 user
|
5 |
+
USER user
|
6 |
+
ENV HOME=/home/user \
|
7 |
+
PATH=/home/user/.local/bin:$PATH
|
8 |
+
WORKDIR $HOME/app
|
9 |
+
|
10 |
+
COPY --chown=user . $HOME/app
|
11 |
+
COPY ./requirements.txt ~/app/requirements.txt
|
12 |
+
|
13 |
+
USER root
|
14 |
+
RUN rm /var/lib/apt/lists/* -vf
|
15 |
+
RUN apt-get clean
|
16 |
+
RUN apt-get update
|
17 |
+
RUN apt-get upgrade
|
18 |
+
RUN apt-get install -y wget zip unzip uvicorn espeak-ng
|
19 |
+
USER user
|
20 |
+
COPY . .
|
21 |
+
USER root
|
22 |
+
RUN chmod 777 ~/app/*
|
23 |
+
USER user
|
24 |
+
|
25 |
+
RUN pip3 install -r requirements.txt
|
26 |
+
|
27 |
+
CMD ["python", "app.py"]
|
app2.py
ADDED
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
from transformers import pipeline
|
3 |
+
from transformers import BlipForQuestionAnswering
|
4 |
+
from transformers.utils import logging
|
5 |
+
logging.set_verbosity_error()
|
6 |
+
|
7 |
+
from transformers import AutoProcessor
|
8 |
+
|
9 |
+
|
10 |
+
|
11 |
+
od_pipe = pipeline("object-detection", "facebook/detr-resnet-50")
|
12 |
+
tts_pipe = pipeline("text-to-speech",
|
13 |
+
model="kakao-enterprise/vits-ljs")
|
14 |
+
|
15 |
+
model = BlipForQuestionAnswering.from_pretrained(
|
16 |
+
"Salesforce/blip-vqa-base")
|
17 |
+
|
18 |
+
processor = AutoProcessor.from_pretrained(
|
19 |
+
"Salesforce/blip-vqa-base")
|
20 |
+
|
21 |
+
def get_pipeline_prediction(pil_image, question):
|
22 |
+
|
23 |
+
inputs = processor(pil_image, question, return_tensors="pt")
|
24 |
+
|
25 |
+
out = model.generate(**inputs)
|
26 |
+
|
27 |
+
text = processor.decode(out[0], skip_special_tokens=True)
|
28 |
+
|
29 |
+
narrated_text = tts_pipe(text)
|
30 |
+
|
31 |
+
|
32 |
+
return (narrated_text["sampling_rate"], narrated_text["audio"][0] )
|
33 |
+
|
34 |
+
|
35 |
+
demo = gr.Interface(
|
36 |
+
fn=get_pipeline_prediction,
|
37 |
+
inputs=[gr.Image(label="Input image",
|
38 |
+
type="pil"), gr.Textbox(label="Ask your question")],
|
39 |
+
outputs=gr.Audio(label="Narration", type="numpy", autoplay=True)
|
40 |
+
|
41 |
+
)
|
42 |
+
|
43 |
+
demo.launch(server_name="0.0.0.0", server_port=7860)
|
requirements.txt
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
opencv-python-headless<4.3
|
2 |
+
gradio
|
3 |
+
transformers
|
4 |
+
phonemizer
|
5 |
+
py-espeak-ng
|
6 |
+
inflect
|
7 |
+
timm
|