试着使用gradio
Browse files- README.md +1 -1
- app.py +2 -53
- app_history.py +54 -0
README.md
CHANGED
@@ -5,7 +5,7 @@ colorFrom: indigo
|
|
5 |
colorTo: purple
|
6 |
sdk: streamlit
|
7 |
sdk_version: 1.31.1
|
8 |
-
app_file:
|
9 |
pinned: false
|
10 |
license: mit
|
11 |
---
|
|
|
5 |
colorTo: purple
|
6 |
sdk: streamlit
|
7 |
sdk_version: 1.31.1
|
8 |
+
app_file: app_history.py
|
9 |
pinned: false
|
10 |
license: mit
|
11 |
---
|
app.py
CHANGED
@@ -1,54 +1,3 @@
|
|
1 |
-
|
2 |
-
import streamlit as st
|
3 |
-
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
4 |
-
import torch
|
5 |
-
import os
|
6 |
-
from transformers import AutoTokenizer, AutoModel
|
7 |
-
import requests
|
8 |
|
9 |
-
|
10 |
-
huggingface_token = os.getenv('HF_TOKEN')
|
11 |
-
# Set up the token to use with the Hugging Face API
|
12 |
-
if huggingface_token is not None:
|
13 |
-
os.environ['HUGGINGFACE_CO_API_TOKEN'] = huggingface_token
|
14 |
-
API_URL = "https://api-inference.huggingface.co/models/Tokymin/Mood_Anxiety_Disorder_Classify_Model"
|
15 |
-
headers = {"Authorization": f"Tokymin {huggingface_token}"}
|
16 |
-
else:
|
17 |
-
print("error, no token")
|
18 |
-
exit(0)
|
19 |
-
|
20 |
-
# def query(payload):
|
21 |
-
# response = requests.post(API_URL, headers=headers, json=payload)
|
22 |
-
# return response.json()
|
23 |
-
# data = query("Can you please let us know more details about your ")
|
24 |
-
path: Path = Path('Tokymin/Mood_Anxiety_Disorder_Classify_Model')
|
25 |
-
tokenizer = AutoTokenizer.from_pretrained(pretrained_model_name_or_path=path, cache_dir='/home/user', token=huggingface_token)
|
26 |
-
|
27 |
-
# tokenizer = AutoTokenizer.from_pretrained('Tokymin/Mood_Anxiety_Disorder_Classify_Model')
|
28 |
-
model = AutoModelForSequenceClassification.from_pretrained("Tokymin/Mood_Anxiety_Disorder_Classify_Model",num_labels=8)
|
29 |
-
model.eval()
|
30 |
-
|
31 |
-
|
32 |
-
def predict(text):
|
33 |
-
inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True, max_length=512)
|
34 |
-
with torch.no_grad():
|
35 |
-
outputs = model(**inputs)
|
36 |
-
logits = outputs.logits
|
37 |
-
probabilities = torch.softmax(logits, dim=1).squeeze()
|
38 |
-
# 假设每个类别(SAS_Class和SDS_Class)都有4个概率值
|
39 |
-
sas_probs = probabilities[:4] # 获取SAS_Class的概率
|
40 |
-
sds_probs = probabilities[4:] # 获取SDS_Class的概率
|
41 |
-
return sas_probs, sds_probs
|
42 |
-
|
43 |
-
|
44 |
-
# 创建Streamlit应用
|
45 |
-
st.title("Multi-label Classification App")
|
46 |
-
|
47 |
-
# 用户输入文本
|
48 |
-
user_input = st.text_area("Enter text here", "Type something...")
|
49 |
-
|
50 |
-
if st.button("Predict"):
|
51 |
-
# 显示预测结果
|
52 |
-
sas_probs, sds_probs = predict(user_input)
|
53 |
-
st.write("SAS_Class probabilities:", sas_probs.numpy())
|
54 |
-
st.write("SDS_Class probabilities:", sds_probs.numpy())
|
|
|
1 |
+
import gradio as gr
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
3 |
+
gr.load("models/Tokymin/Mood_Anxiety_Disorder_Classify_Model").launch()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
app_history.py
ADDED
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from pathlib import Path
|
2 |
+
import streamlit as st
|
3 |
+
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
4 |
+
import torch
|
5 |
+
import os
|
6 |
+
from transformers import AutoTokenizer, AutoModel
|
7 |
+
import requests
|
8 |
+
|
9 |
+
# Assuming you have set the HF_TOKEN environment variable with your Hugging Face token
|
10 |
+
huggingface_token = os.getenv('HF_TOKEN')
|
11 |
+
# Set up the token to use with the Hugging Face API
|
12 |
+
if huggingface_token is not None:
|
13 |
+
os.environ['HUGGINGFACE_CO_API_TOKEN'] = huggingface_token
|
14 |
+
API_URL = "https://api-inference.huggingface.co/models/Tokymin/Mood_Anxiety_Disorder_Classify_Model"
|
15 |
+
headers = {"Authorization": f"Tokymin {huggingface_token}"}
|
16 |
+
else:
|
17 |
+
print("error, no token")
|
18 |
+
exit(0)
|
19 |
+
|
20 |
+
# def query(payload):
|
21 |
+
# response = requests.post(API_URL, headers=headers, json=payload)
|
22 |
+
# return response.json()
|
23 |
+
# data = query("Can you please let us know more details about your ")
|
24 |
+
path: Path = Path('Tokymin/Mood_Anxiety_Disorder_Classify_Model')
|
25 |
+
tokenizer = AutoTokenizer.from_pretrained(pretrained_model_name_or_path=path, cache_dir='/home/user', token=huggingface_token)
|
26 |
+
|
27 |
+
# tokenizer = AutoTokenizer.from_pretrained('Tokymin/Mood_Anxiety_Disorder_Classify_Model')
|
28 |
+
model = AutoModelForSequenceClassification.from_pretrained("Tokymin/Mood_Anxiety_Disorder_Classify_Model",num_labels=8)
|
29 |
+
model.eval()
|
30 |
+
|
31 |
+
|
32 |
+
def predict(text):
|
33 |
+
inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True, max_length=512)
|
34 |
+
with torch.no_grad():
|
35 |
+
outputs = model(**inputs)
|
36 |
+
logits = outputs.logits
|
37 |
+
probabilities = torch.softmax(logits, dim=1).squeeze()
|
38 |
+
# 假设每个类别(SAS_Class和SDS_Class)都有4个概率值
|
39 |
+
sas_probs = probabilities[:4] # 获取SAS_Class的概率
|
40 |
+
sds_probs = probabilities[4:] # 获取SDS_Class的概率
|
41 |
+
return sas_probs, sds_probs
|
42 |
+
|
43 |
+
|
44 |
+
# 创建Streamlit应用
|
45 |
+
st.title("Multi-label Classification App")
|
46 |
+
|
47 |
+
# 用户输入文本
|
48 |
+
user_input = st.text_area("Enter text here", "Type something...")
|
49 |
+
|
50 |
+
if st.button("Predict"):
|
51 |
+
# 显示预测结果
|
52 |
+
sas_probs, sds_probs = predict(user_input)
|
53 |
+
st.write("SAS_Class probabilities:", sas_probs.numpy())
|
54 |
+
st.write("SDS_Class probabilities:", sds_probs.numpy())
|