christofid
commited on
Commit
·
89e8857
1
Parent(s):
dc87e4f
Update app.py
Browse files
app.py
CHANGED
@@ -4,25 +4,35 @@ import gradio as gr
|
|
4 |
import pandas as pd
|
5 |
from gt4sd.algorithms.generation.hugging_face import (
|
6 |
HuggingFaceSeq2SeqGenerator,
|
7 |
-
HuggingFaceGenerationAlgorithm
|
8 |
)
|
9 |
from transformers import AutoTokenizer
|
10 |
|
11 |
logger = logging.getLogger(__name__)
|
12 |
logger.addHandler(logging.NullHandler())
|
13 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
def run_inference(
|
15 |
model_name_or_path: str,
|
16 |
-
|
17 |
prompt: str,
|
18 |
num_beams: int,
|
19 |
):
|
|
|
20 |
|
21 |
config = HuggingFaceSeq2SeqGenerator(
|
22 |
algorithm_version=model_name_or_path,
|
23 |
-
prefix=
|
24 |
prompt=prompt,
|
25 |
-
num_beams=num_beams
|
26 |
)
|
27 |
|
28 |
model = HuggingFaceGenerationAlgorithm(config)
|
@@ -30,22 +40,23 @@ def run_inference(
|
|
30 |
|
31 |
text = list(model.sample(1))[0]
|
32 |
|
33 |
-
text = text.replace(
|
34 |
text = text.split(tokenizer.eos_token)[0]
|
35 |
text = text.replace(tokenizer.pad_token, "")
|
36 |
text = text.strip()
|
37 |
|
38 |
-
|
39 |
return text
|
40 |
|
41 |
|
42 |
if __name__ == "__main__":
|
43 |
|
44 |
-
|
45 |
-
|
46 |
-
|
|
|
|
|
|
|
47 |
|
48 |
-
# Load metadata
|
49 |
metadata_root = pathlib.Path(__file__).parent.joinpath("model_cards")
|
50 |
|
51 |
examples = pd.read_csv(metadata_root.joinpath("examples.csv"), header=None).fillna(
|
@@ -67,8 +78,16 @@ if __name__ == "__main__":
|
|
67 |
label="Language model",
|
68 |
value="text-chem-t5-base-augm",
|
69 |
),
|
70 |
-
gr.
|
71 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
72 |
),
|
73 |
gr.Textbox(
|
74 |
label="Text prompt",
|
|
|
4 |
import pandas as pd
|
5 |
from gt4sd.algorithms.generation.hugging_face import (
|
6 |
HuggingFaceSeq2SeqGenerator,
|
7 |
+
HuggingFaceGenerationAlgorithm,
|
8 |
)
|
9 |
from transformers import AutoTokenizer
|
10 |
|
11 |
logger = logging.getLogger(__name__)
|
12 |
logger.addHandler(logging.NullHandler())
|
13 |
|
14 |
+
task2prefix = {
|
15 |
+
"forward": "Predict the product of the following reaction: ",
|
16 |
+
"retrosynthesis": "Predict the reaction that produces the following product: ",
|
17 |
+
"paragraph to actions": "Which actions are described in the following paragraph: ",
|
18 |
+
"molecular captioning": "Caption the following SMILES: ",
|
19 |
+
"text-conditional de novo generation": "Write in SMILES the described molecule: ",
|
20 |
+
}
|
21 |
+
|
22 |
+
|
23 |
def run_inference(
|
24 |
model_name_or_path: str,
|
25 |
+
task: str,
|
26 |
prompt: str,
|
27 |
num_beams: int,
|
28 |
):
|
29 |
+
instruction = task2prefix[task]
|
30 |
|
31 |
config = HuggingFaceSeq2SeqGenerator(
|
32 |
algorithm_version=model_name_or_path,
|
33 |
+
prefix=instruction,
|
34 |
prompt=prompt,
|
35 |
+
num_beams=num_beams,
|
36 |
)
|
37 |
|
38 |
model = HuggingFaceGenerationAlgorithm(config)
|
|
|
40 |
|
41 |
text = list(model.sample(1))[0]
|
42 |
|
43 |
+
text = text.replace(instruction + prompt, "")
|
44 |
text = text.split(tokenizer.eos_token)[0]
|
45 |
text = text.replace(tokenizer.pad_token, "")
|
46 |
text = text.strip()
|
47 |
|
|
|
48 |
return text
|
49 |
|
50 |
|
51 |
if __name__ == "__main__":
|
52 |
|
53 |
+
models = [
|
54 |
+
"text-chem-t5-small-standard",
|
55 |
+
"text-chem-t5-small-augm",
|
56 |
+
"text-chem-t5-base-standard",
|
57 |
+
"text-chem-t5-base-augm",
|
58 |
+
]
|
59 |
|
|
|
60 |
metadata_root = pathlib.Path(__file__).parent.joinpath("model_cards")
|
61 |
|
62 |
examples = pd.read_csv(metadata_root.joinpath("examples.csv"), header=None).fillna(
|
|
|
78 |
label="Language model",
|
79 |
value="text-chem-t5-base-augm",
|
80 |
),
|
81 |
+
gr.Radio(
|
82 |
+
choices=[
|
83 |
+
"forward",
|
84 |
+
"retrosynthesis",
|
85 |
+
"paragraph to actions",
|
86 |
+
"molecular captioning",
|
87 |
+
"text-conditional de novo generation",
|
88 |
+
],
|
89 |
+
label="Task",
|
90 |
+
value="paragraph to actions",
|
91 |
),
|
92 |
gr.Textbox(
|
93 |
label="Text prompt",
|