Spaces:
Running
on
Zero
Running
on
Zero
Update app.py
Browse files
app.py
CHANGED
@@ -1,31 +1,17 @@
|
|
1 |
import os
|
2 |
-
from collections.abc import Iterator
|
3 |
-
from threading import Thread
|
4 |
import gradio as gr
|
5 |
-
import spaces
|
6 |
import torch
|
7 |
-
import
|
8 |
import asyncio
|
|
|
|
|
|
|
9 |
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
|
10 |
|
11 |
DESCRIPTION = """
|
12 |
-
# QwQ Tiny
|
13 |
"""
|
14 |
|
15 |
-
css ='''
|
16 |
-
h1 {
|
17 |
-
text-align: center;
|
18 |
-
display: block;
|
19 |
-
}
|
20 |
-
|
21 |
-
#duplicate-button {
|
22 |
-
margin: auto;
|
23 |
-
color: #fff;
|
24 |
-
background: #1565c0;
|
25 |
-
border-radius: 100vh;
|
26 |
-
}
|
27 |
-
'''
|
28 |
-
|
29 |
MAX_MAX_NEW_TOKENS = 2048
|
30 |
DEFAULT_MAX_NEW_TOKENS = 1024
|
31 |
MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
|
@@ -41,16 +27,14 @@ model = AutoModelForCausalLM.from_pretrained(
|
|
41 |
)
|
42 |
model.eval()
|
43 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
44 |
|
45 |
-
async def text_to_speech(text: str, output_file="output.mp3"):
|
46 |
-
"""Convert text to speech using Edge TTS and save as MP3"""
|
47 |
-
voice = "en-US-JennyNeural" # Change this to your preferred voice
|
48 |
-
communicate = edge_tts.Communicate(text, voice)
|
49 |
-
await communicate.save(output_file)
|
50 |
-
return output_file
|
51 |
-
|
52 |
-
|
53 |
-
@spaces.GPU
|
54 |
def generate(
|
55 |
message: str,
|
56 |
chat_history: list[dict],
|
@@ -59,47 +43,55 @@ def generate(
|
|
59 |
top_p: float = 0.9,
|
60 |
top_k: int = 50,
|
61 |
repetition_penalty: float = 1.2,
|
62 |
-
):
|
63 |
-
|
64 |
-
is_tts = message.strip().
|
65 |
-
|
66 |
|
67 |
-
|
|
|
|
|
|
|
|
|
68 |
|
|
|
69 |
input_ids = tokenizer.apply_chat_template(conversation, add_generation_prompt=True, return_tensors="pt")
|
|
|
70 |
if input_ids.shape[1] > MAX_INPUT_TOKEN_LENGTH:
|
71 |
input_ids = input_ids[:, -MAX_INPUT_TOKEN_LENGTH:]
|
72 |
gr.Warning(f"Trimmed input from conversation as it was longer than {MAX_INPUT_TOKEN_LENGTH} tokens.")
|
|
|
73 |
input_ids = input_ids.to(model.device)
|
74 |
|
75 |
streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
|
76 |
-
generate_kwargs =
|
77 |
-
|
78 |
-
streamer
|
79 |
-
max_new_tokens
|
80 |
-
do_sample
|
81 |
-
top_p
|
82 |
-
top_k
|
83 |
-
temperature
|
84 |
-
num_beams
|
85 |
-
repetition_penalty
|
86 |
-
|
87 |
t = Thread(target=model.generate, kwargs=generate_kwargs)
|
88 |
t.start()
|
89 |
|
90 |
outputs = []
|
91 |
for text in streamer:
|
92 |
outputs.append(text)
|
93 |
-
yield "".join(outputs)
|
94 |
|
95 |
-
|
96 |
|
|
|
97 |
if is_tts:
|
98 |
-
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
|
|
|
103 |
|
104 |
demo = gr.ChatInterface(
|
105 |
fn=generate,
|
@@ -113,15 +105,13 @@ demo = gr.ChatInterface(
|
|
113 |
stop_btn=None,
|
114 |
examples=[
|
115 |
["A train travels 60 kilometers per hour. If it travels for 5 hours, how far will it travel in total?"],
|
116 |
-
["
|
117 |
-
["
|
118 |
-
["Rewrite the following sentence in passive voice: 'The dog chased the cat.'"],
|
119 |
-
["@tts What is the capital of France?"],
|
120 |
],
|
121 |
cache_examples=False,
|
122 |
type="messages",
|
123 |
description=DESCRIPTION,
|
124 |
-
css=css,
|
125 |
fill_height=True,
|
126 |
)
|
127 |
|
|
|
1 |
import os
|
|
|
|
|
2 |
import gradio as gr
|
|
|
3 |
import torch
|
4 |
+
import tempfile
|
5 |
import asyncio
|
6 |
+
import edge_tts
|
7 |
+
from threading import Thread
|
8 |
+
from collections.abc import Iterator
|
9 |
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
|
10 |
|
11 |
DESCRIPTION = """
|
12 |
+
# QwQ Tiny with Edge TTS
|
13 |
"""
|
14 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
15 |
MAX_MAX_NEW_TOKENS = 2048
|
16 |
DEFAULT_MAX_NEW_TOKENS = 1024
|
17 |
MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
|
|
|
27 |
)
|
28 |
model.eval()
|
29 |
|
30 |
+
async def text_to_speech(text: str) -> str:
|
31 |
+
"""Converts text to speech using Edge TTS and returns the generated audio file path."""
|
32 |
+
communicate = edge_tts.Communicate(text)
|
33 |
+
with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
|
34 |
+
tmp_path = tmp_file.name
|
35 |
+
await communicate.save(tmp_path)
|
36 |
+
return tmp_path
|
37 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
38 |
def generate(
|
39 |
message: str,
|
40 |
chat_history: list[dict],
|
|
|
43 |
top_p: float = 0.9,
|
44 |
top_k: int = 50,
|
45 |
repetition_penalty: float = 1.2,
|
46 |
+
) -> Iterator[str] | str:
|
47 |
+
|
48 |
+
is_tts = message.strip().startswith("@tts")
|
49 |
+
is_text_only = message.strip().startswith("@text")
|
50 |
|
51 |
+
# Remove special tags
|
52 |
+
if is_tts:
|
53 |
+
message = message.replace("@tts", "").strip()
|
54 |
+
elif is_text_only:
|
55 |
+
message = message.replace("@text", "").strip()
|
56 |
|
57 |
+
conversation = [*chat_history, {"role": "user", "content": message}]
|
58 |
input_ids = tokenizer.apply_chat_template(conversation, add_generation_prompt=True, return_tensors="pt")
|
59 |
+
|
60 |
if input_ids.shape[1] > MAX_INPUT_TOKEN_LENGTH:
|
61 |
input_ids = input_ids[:, -MAX_INPUT_TOKEN_LENGTH:]
|
62 |
gr.Warning(f"Trimmed input from conversation as it was longer than {MAX_INPUT_TOKEN_LENGTH} tokens.")
|
63 |
+
|
64 |
input_ids = input_ids.to(model.device)
|
65 |
|
66 |
streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
|
67 |
+
generate_kwargs = {
|
68 |
+
"input_ids": input_ids,
|
69 |
+
"streamer": streamer,
|
70 |
+
"max_new_tokens": max_new_tokens,
|
71 |
+
"do_sample": True,
|
72 |
+
"top_p": top_p,
|
73 |
+
"top_k": top_k,
|
74 |
+
"temperature": temperature,
|
75 |
+
"num_beams": 1,
|
76 |
+
"repetition_penalty": repetition_penalty,
|
77 |
+
}
|
78 |
t = Thread(target=model.generate, kwargs=generate_kwargs)
|
79 |
t.start()
|
80 |
|
81 |
outputs = []
|
82 |
for text in streamer:
|
83 |
outputs.append(text)
|
|
|
84 |
|
85 |
+
final_output = "".join(outputs)
|
86 |
|
87 |
+
# If TTS requested, generate speech and return audio file
|
88 |
if is_tts:
|
89 |
+
loop = asyncio.new_event_loop()
|
90 |
+
asyncio.set_event_loop(loop)
|
91 |
+
audio_path = loop.run_until_complete(text_to_speech(final_output))
|
92 |
+
return audio_path # Returning audio file path
|
93 |
|
94 |
+
return final_output # Returning text output
|
95 |
|
96 |
demo = gr.ChatInterface(
|
97 |
fn=generate,
|
|
|
105 |
stop_btn=None,
|
106 |
examples=[
|
107 |
["A train travels 60 kilometers per hour. If it travels for 5 hours, how far will it travel in total?"],
|
108 |
+
["@text What causes rainbows to form?"],
|
109 |
+
["edgetts@tts Explain Newton's third law of motion."],
|
110 |
+
["@text Rewrite the following sentence in passive voice: 'The dog chased the cat.'"],
|
|
|
111 |
],
|
112 |
cache_examples=False,
|
113 |
type="messages",
|
114 |
description=DESCRIPTION,
|
|
|
115 |
fill_height=True,
|
116 |
)
|
117 |
|