Spaces:
Runtime error
Runtime error
Commit
·
bfa01d0
1
Parent(s):
57a7c34
Add application file
Browse files
app.py
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import torch
|
| 3 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 4 |
+
|
| 5 |
+
device = "cuda" if torch.cuda.is_available() else "cpu"
|
| 6 |
+
model_path = "ibm-granite/granite-3b-code-base"
|
| 7 |
+
tokenizer = AutoTokenizer.from_pretrained(model_path)
|
| 8 |
+
model = AutoModelForCausalLM.from_pretrained(model_path, device_map=device)
|
| 9 |
+
model.eval()
|
| 10 |
+
|
| 11 |
+
def generate_code(input_text):
|
| 12 |
+
input_tokens = tokenizer(input_text, return_tensors="pt")
|
| 13 |
+
for i in input_tokens:
|
| 14 |
+
input_tokens[i] = input_tokens[i].to(device)
|
| 15 |
+
output = model.generate(**input_tokens, max_new_tokens=200)
|
| 16 |
+
output_text = tokenizer.batch_decode(output, skip_special_tokens=True)[0]
|
| 17 |
+
return output_text
|
| 18 |
+
|
| 19 |
+
# Gradio Interface
|
| 20 |
+
iface = gr.Interface(
|
| 21 |
+
fn=generate_code,
|
| 22 |
+
inputs=gr.inputs.Textbox(lines=2, placeholder="Enter code snippet here..."),
|
| 23 |
+
outputs="text"
|
| 24 |
+
)
|
| 25 |
+
|
| 26 |
+
iface.launch()
|