Spaces:
Runtime error
Runtime error
Commit
•
c78ba93
1
Parent(s):
c111383
adapt to my model
Browse files
app.py
CHANGED
@@ -1,7 +1,68 @@
|
|
1 |
-
import
|
|
|
|
|
2 |
|
3 |
-
|
4 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
|
6 |
-
|
7 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import torch
|
2 |
+
from peft import PeftModel, PeftConfig
|
3 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
4 |
|
5 |
+
peft_model_id = f"jmartin233/bloom-1b7-lora-reading-comprehension"
|
6 |
+
config = PeftConfig.from_pretrained(peft_model_id)
|
7 |
+
model = AutoModelForCausalLM.from_pretrained(
|
8 |
+
config.base_model_name_or_path,
|
9 |
+
return_dict=True,
|
10 |
+
load_in_8bit=True,
|
11 |
+
device_map="auto",
|
12 |
+
)
|
13 |
+
tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
|
14 |
|
15 |
+
# Load the Lora model
|
16 |
+
model = PeftModel.from_pretrained(model, peft_model_id)
|
17 |
+
|
18 |
+
|
19 |
+
def make_inference(product_name, product_description):
|
20 |
+
batch = tokenizer(
|
21 |
+
f"### Product and Description:\n{product_name}: {product_description}\n\n### Ad:",
|
22 |
+
return_tensors="pt",
|
23 |
+
)
|
24 |
+
|
25 |
+
with torch.cuda.amp.autocast():
|
26 |
+
output_tokens = model.generate(**batch, max_new_tokens=50)
|
27 |
+
|
28 |
+
|
29 |
+
|
30 |
+
def make_inference(person, location, grammer, level):
|
31 |
+
|
32 |
+
batch = tokenizer(f"""
|
33 |
+
Below is a set of requirements for a short passage of English. Please write a passage that meets these requirements:
|
34 |
+
|
35 |
+
### Requirements:
|
36 |
+
person: {person}
|
37 |
+
location: {location}.
|
38 |
+
grammar: {grammar}
|
39 |
+
level: {level}
|
40 |
+
|
41 |
+
### Passage:
|
42 |
+
Passage:""",
|
43 |
+
return_tensors='pt')
|
44 |
+
|
45 |
+
with torch.cuda.amp.autocast():
|
46 |
+
output_tokens = model.generate(**batch, max_new_tokens=50)
|
47 |
+
|
48 |
+
return tokenizer.decode(output_tokens[0], skip_special_tokens=True)
|
49 |
+
|
50 |
+
|
51 |
+
|
52 |
+
if __name__ == "__main__":
|
53 |
+
# make a gradio interface
|
54 |
+
import gradio as gr
|
55 |
+
|
56 |
+
gr.Interface(
|
57 |
+
make_inference,
|
58 |
+
[
|
59 |
+
gr.inputs.Textbox(lines=2, label="Someone's name"),
|
60 |
+
gr.inputs.Textbox(lines=2, label="A location they might visit"),
|
61 |
+
gr.inputs.Textbox(lines=2, label="A type of grammar to use"),
|
62 |
+
gr.inputs.Textbox(lines=2, label="The leve of English to use (beginner, intermediate, advanced))"),
|
63 |
+
|
64 |
+
],
|
65 |
+
gr.outputs.Textbox(label="Passage"),
|
66 |
+
title="Reading Comprehension",
|
67 |
+
description="A generative model that generates simple texts for testing reading comprehension.",
|
68 |
+
).launch()
|