Spaces:
Running
Running
File size: 4,536 Bytes
2fc6c45 f1c3845 2fc6c45 f1c3845 2fc6c45 f1c3845 2fc6c45 964b71d 61df3f6 2fc6c45 |
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 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 |
import gradio as gr
from PIL import Image
from io import BytesIO
import openai
import os
from dotenv import load_dotenv
from image_processor import ImageProcessor
from evaluation_processor import EvaluationProcessor
load_dotenv()
client = openai.OpenAI(
api_key=os.getenv("OPENAI_API_KEY")
)
# 设置OpenAI API密钥
openai.api_key = os.getenv("OPENAI_API_KEY")
engine = "gpt-4o-mini"
# 设置Music API密钥
api_key = 'ddc85b14-bd83-4757-9bc4-8a11194da536'
image_processor = ImageProcessor(api_key)
evaluation_processor = EvaluationProcessor(api_key)
# 定义处理函数
def process_input(text=None, images=None, audio=None, video=None):
# 创建GPT请求的描述
system = "1.你是一个专业的钢琴音乐教师，只能回答音乐知识，回复的内容为普通文本格式,不用任何markdown符号。如果提供的乐谱是abc记谱法,则回复时不要用abc记谱法,需要转换为传统的普通记谱法使用专业词汇进行回答问题2.你将根据下面指令回答问题，但是不能违反第一条指令,也不能在回复中提及。"
messages = [{"role": "system", "content": system}]
prompt = ""
if text:
prompt += f"\nText input: {text}"
if images:
# 使用ImageProcessor处理图像
image_bytes_list = []
for image in images:
img = Image.open(image.name)
image_bytes = BytesIO()
img.save(image_bytes, format="PNG")
image_bytes.seek(0)
image_bytes_list.append(image_bytes.getvalue())
try:
processed_image_result = image_processor.process_images(image_bytes_list)
#prompt += f"\n乐谱的内容如下,这是一首杜维诺伊的曲子，请你根据曲子的曲风回答问题: {processed_image_result}"
prompt += f"\n乐谱的内容如下,请你根据曲子的曲风回答问题: {processed_image_result}"
except Exception as e:
return f"Error processing image: {e}", None
if audio:
try:
# 使用EvaluationProcessor处理音频
audio_path = audio.name
result,title = evaluation_processor.process_evaluation(audio_path, is_video=False)
prompt += f'''如果有曲名{title},请你根据这首歌的名字作者，并且'''
prompt += f'''请你根据
"eva_all":综合得分
"eva_completion":完整性
"eva_note":按键
"eva_stability":稳定性
"eva_tempo_sync":节奏
评价一下评测结果: {result}'''
except Exception as e:
return f"Error processing audio: {e}", None
if video:
try:
# 使用EvaluationProcessor处理视频
video_path = video.name
result,title = evaluation_processor.process_evaluation(video_path, is_video=True)
prompt += f'''如果有曲名{title},请你根据这首歌的名字作者，并且'''
prompt += f'''请你根据
"eva_all":综合得分
"eva_completion":完整性
"eva_note":按键
"eva_stability":稳定性
"eva_tempo_sync":节奏
评价一下评测结果: {result}'''
except Exception as e:
return f"Error processing video: {e}", None
# 使用GPT API进行处理
try:
messages.append({"role": "user", "content": prompt})
response = client.chat.completions.create(
model=engine,
messages=messages,
temperature=0.2,
max_tokens=4096,
top_p=0.95,
frequency_penalty=0,
presence_penalty=0,
stop=None
)
print(response.choices[0].message.content)
return response.choices[0].message.content
except Exception as e:
return f"Error: {e}", None
# 创建Gradio接口
iface = gr.Interface(
fn=process_input,
inputs=[
gr.Textbox(label="Input Text", placeholder="我是音乐分析大模型，您可以上传曲谱，演奏的音/视频", lines=2), # 文本输入
gr.File(label="Input Images", file_count="multiple", type="filepath"), # 多文件上传
gr.File(label="Input Audio", type="filepath"), # 音频文件上传
gr.File(label="Input Video", type="filepath") # 视频文件上传
],
outputs=[
gr.Textbox(label="Output Text")
],
live=False,
)
# 启动Gradio应用
iface.launch()
|