File size: 1,307 Bytes
d38e956
3b11cb7
 
 
 
dc024f2
3b11cb7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
76a5ccd
 
 
 
 
3b11cb7
 
76a5ccd
 
3b11cb7
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
import gradio as gr
from gtts import gTTS
from io import BytesIO
import IPython.display as ipd

# Load model
codellama_model = gr.Interface.load("models/meta-llama/CodeLlama-7b-Python-hf")
deepseek_model = gr.Interface.load("models/deepseek-ai/deepseek-coder-1.3b-instruct")

def process_text_codellama(input_text):
    return codellama_model.predict(input_text)

def process_speech_codellama(audio):
    response = codellama_model.predict(audio)
    tts = gTTS(text=response, lang='en')
    fp = BytesIO()
    tts.write_to_fp(fp)
    fp.seek(0)
    return ipd.Audio(fp.read(), autoplay=True)

def process_text_deepseek(input_text):
    return deepseek_model.predict(input_text)

def process_speech_deepseek(audio):
    response = deepseek_model.predict(audio)
    tts = gTTS(text=response, lang='en')
    fp = BytesIO()
    tts.write_to_fp(fp)
    fp.seek(0)
    return ipd.Audio(fp.read(), autoplay=True)

def main(input_text):
    if input_text[1]:  
        return process_text_deepseek(input_text[0]), process_speech_deepseek(input_text[0])
    else:
        return process_text_codellama(input_text[0]), process_speech_codellama(input_text[0])

gr.Interface(
    fn=main,
    inputs=["text", "checkbox"],  # ["input text", "enable voice input"]
    outputs=["text", "audio"],
    live=True
).launch()