Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,6 +1,7 @@
|
|
1 |
import os
|
2 |
from threading import Thread
|
3 |
from typing import Iterator
|
|
|
4 |
|
5 |
import gradio as gr
|
6 |
import spaces
|
@@ -29,7 +30,6 @@ model = AutoModelForCausalLM.from_pretrained(
|
|
29 |
model.config.sliding_window = 4096
|
30 |
model.eval()
|
31 |
|
32 |
-
|
33 |
@spaces.GPU
|
34 |
def generate(
|
35 |
message: str,
|
@@ -46,11 +46,11 @@ def generate(
|
|
46 |
if input_ids.shape[1] > MAX_INPUT_TOKEN_LENGTH:
|
47 |
input_ids = input_ids[:, -MAX_INPUT_TOKEN_LENGTH:]
|
48 |
gr.Warning(f"Trimmed input from conversation as it was longer than {MAX_INPUT_TOKEN_LENGTH} tokens.")
|
49 |
-
input_ids = input_ids.to(
|
50 |
|
51 |
streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
|
52 |
generate_kwargs = dict(
|
53 |
-
|
54 |
streamer=streamer,
|
55 |
max_new_tokens=max_new_tokens,
|
56 |
do_sample=True,
|
@@ -64,11 +64,15 @@ def generate(
|
|
64 |
t.start()
|
65 |
|
66 |
outputs = []
|
67 |
-
|
68 |
-
|
|
|
|
|
|
|
|
|
|
|
69 |
yield "".join(outputs)
|
70 |
|
71 |
-
|
72 |
demo = gr.ChatInterface(
|
73 |
fn=generate,
|
74 |
type="messages",
|
@@ -122,6 +126,5 @@ demo = gr.ChatInterface(
|
|
122 |
cache_examples=False,
|
123 |
)
|
124 |
|
125 |
-
|
126 |
if __name__ == "__main__":
|
127 |
-
demo.launch()
|
|
|
1 |
import os
|
2 |
from threading import Thread
|
3 |
from typing import Iterator
|
4 |
+
import queue
|
5 |
|
6 |
import gradio as gr
|
7 |
import spaces
|
|
|
30 |
model.config.sliding_window = 4096
|
31 |
model.eval()
|
32 |
|
|
|
33 |
@spaces.GPU
|
34 |
def generate(
|
35 |
message: str,
|
|
|
46 |
if input_ids.shape[1] > MAX_INPUT_TOKEN_LENGTH:
|
47 |
input_ids = input_ids[:, -MAX_INPUT_TOKEN_LENGTH:]
|
48 |
gr.Warning(f"Trimmed input from conversation as it was longer than {MAX_INPUT_TOKEN_LENGTH} tokens.")
|
49 |
+
input_ids = input_ids.to(device)
|
50 |
|
51 |
streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
|
52 |
generate_kwargs = dict(
|
53 |
+
input_ids=input_ids,
|
54 |
streamer=streamer,
|
55 |
max_new_tokens=max_new_tokens,
|
56 |
do_sample=True,
|
|
|
64 |
t.start()
|
65 |
|
66 |
outputs = []
|
67 |
+
try:
|
68 |
+
for text in streamer:
|
69 |
+
outputs.append(text)
|
70 |
+
yield "".join(outputs)
|
71 |
+
except queue.Empty:
|
72 |
+
# キューが空になった場合の処理
|
73 |
+
gr.Warning("生成プロセスがタイムアウトしました。")
|
74 |
yield "".join(outputs)
|
75 |
|
|
|
76 |
demo = gr.ChatInterface(
|
77 |
fn=generate,
|
78 |
type="messages",
|
|
|
126 |
cache_examples=False,
|
127 |
)
|
128 |
|
|
|
129 |
if __name__ == "__main__":
|
130 |
+
demo.launch()
|