Spaces:
Runtime error
Runtime error
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import torch
|
2 |
+
import gradio as gr
|
3 |
+
from transformers import BioGptTokenizer, BioGptForCausalLM
|
4 |
+
|
5 |
+
model_names = [
|
6 |
+
"BioGPT",
|
7 |
+
"BioGPT-Large",
|
8 |
+
"BioGPT-QA-PubMedQA-BioGPT",
|
9 |
+
"BioGPT-QA-PubMEDQA-BioGPT-Large",
|
10 |
+
"BioGPT-RE-BC5CDR",
|
11 |
+
"BioGPT-RE-DDI",
|
12 |
+
"BioGPT-RE-DTI",
|
13 |
+
"BioGPT-DC-HoC"
|
14 |
+
]
|
15 |
+
|
16 |
+
def load_model(model_name):
|
17 |
+
model_name_map = {
|
18 |
+
"BioGPT":"microsoft/biogpt",
|
19 |
+
"BioGPT-QA-PubMedQA-BioGPT":"microsoft/BioGPT-Large-PubMedQA"
|
20 |
+
}
|
21 |
+
|
22 |
+
tokenizer = BioGptTokenizer.from_pretrained(model_name_map[model_name])
|
23 |
+
model = BioGptForCausalLM.from_pretrained(model_name_map[model_name])
|
24 |
+
return tokenizer, model
|
25 |
+
|
26 |
+
|
27 |
+
def get_beam_output(sentence, selected_model, min_len,max_len, n_beams):
|
28 |
+
tokenizer, model = load_model(selected_model)
|
29 |
+
inputs = tokenizer(sentence, return_tensors="pt")
|
30 |
+
with torch.no_grad():
|
31 |
+
beam_output = model.generate(**inputs,
|
32 |
+
min_length=100,
|
33 |
+
max_length=1024,
|
34 |
+
num_beams=n_beams,
|
35 |
+
early_stopping=True
|
36 |
+
)
|
37 |
+
output=tokenizer.decode(beam_output[0], skip_special_tokens=True)
|
38 |
+
return output
|
39 |
+
|
40 |
+
inputs = [
|
41 |
+
gr.inputs.Textbox(label="prompt", lines=5, default="Bicalutamide"),
|
42 |
+
gr.Dropdown(model_names, value="microsoft/biogpt", label="selected_model"),
|
43 |
+
gr.inputs.Slider(1, 500, 1, default=100, label="min_len"),
|
44 |
+
gr.inputs.Slider(1, 2048, 1, default=1024, label="max_len"),
|
45 |
+
gr.inputs.Slider(1, 10, 1, default=5, label="num_beams")
|
46 |
+
]
|
47 |
+
outputs = gr.outputs.Textbox(label="output")
|
48 |
+
|
49 |
+
iface = gr.Interface(
|
50 |
+
fn=get_beam_output,
|
51 |
+
inputs=inputs,
|
52 |
+
outputs=outputs,
|
53 |
+
examples=[["Bicalutamide"], ["Janus kinase 3 (JAK-3)"], ["Apricitabine"], ["Xylazine"], ["Psoralen"], ["CP-673451"]]
|
54 |
+
)
|
55 |
+
|
56 |
+
iface.launch(debug=True, enable_queue=True)
|