Merge pull request #5 from BoardWare-Genius/config

feat: configuration

README.md

@@ -9,8 +9,49 @@
 | python | uvicorn | https://www.uvicorn.org/ | pip install "uvicorn[standard]" | 
 | python | SpeechRecognition |  https://pypi.org/project/SpeechRecognition/ |  pip install SpeechRecognition |
 | python | gtts | https://pypi.org/project/gTTS/ | pip install gTTS |
+| python | PyYAML | https://pypi.org/project/PyYAML/ | pip install PyYAML |
+| python | injector | https://github.com/python-injector/injector | pip install injector |
+
 ## Start
 Dev rh 
 ```bash
 uvicorn main:app --reload
 ```
+
+## Configuration
+```yaml
+tesou:
+  url: http://120.196.116.194:48891/chat/
+
+TokenIDConverter:
+  token_path: src/asr/resources/models/token_list.pkl
+  unk_symbol: <unk>
+
+CharTokenizer:
+  symbol_value:
+  space_symbol: <space>
+  remove_non_linguistic_symbols: false
+
+WavFrontend:
+  cmvn_file: src/asr/resources/models/am.mvn
+  frontend_conf:
+    fs: 16000
+    window: hamming
+    n_mels: 80
+    frame_length: 25
+    frame_shift: 10
+    lfr_m: 7
+    lfr_n: 6
+    filter_length_max: -.inf
+    dither: 0.0
+
+Model:
+  model_path: src/asr/resources/models/model.onnx
+  use_cuda: false
+  CUDAExecutionProvider:
+      device_id: 0
+      arena_extend_strategy: kNextPowerOfTwo
+      cudnn_conv_algo_search: EXHAUSTIVE
+      do_copy_in_default_stream: true
+  batch_size: 3
+```

install-dependencies.sh

@@ -4,3 +4,5 @@ pip install fastapi
 pip install python-multipart
 pip install "uvicorn[standard]"
 pip install SpeechRecognition
+pip install PyYAML
+pip install injector

main.py

@@ -1,59 +1,4 @@
-from typing import Annotated, Union
-
-from fastapi import FastAPI, Request, status, Form
-from fastapi.responses import JSONResponse
-
-from src.dotchain.runtime.interpreter import program_parser
-from src.dotchain.runtime.tokenizer import Tokenizer
-from src.dotchain.runtime.runtime import Runtime
-from src.blackbox.blackbox_factory import BlackboxFactory
 import uvicorn
-from fastapi.middleware.cors import CORSMiddleware
-
-app = FastAPI()
-
-app.add_middleware(
-    CORSMiddleware,
-    allow_origins=["*"],
-    allow_credentials=True,
-    allow_methods=["*"],
-    allow_headers=["*"],
-)
-
-
-blackbox_factory = BlackboxFactory()
-
-@app.post("/")
-async def blackbox(blackbox_name: Union[str, None] = None, request: Request = None):
-    if not blackbox_name:
-        return await JSONResponse(content={"error": "blackbox_name is required"}, status_code=status.HTTP_400_BAD_REQUEST)
-    try:
-        box = blackbox_factory.create_blackbox(blackbox_name)
-    except ValueError:
-        return await JSONResponse(content={"error": "value error"}, status_code=status.HTTP_400_BAD_REQUEST)
-    return await box.fast_api_handler(request)
-
-def read_form_image(request: Request):
-    async def inner(field: str):
-        print(field)
-        return "image"
-    return inner
-
-def read_form_text(request: Request):
-    def inner(field: str):
-        print(field)
-        return "text"
-    return inner
-
-@app.post("/workflows")
-async def workflows(script: Annotated[str, Form()], request: Request=None):
-    dsl_runtime = Runtime(exteral_fun={"print": print, 
-                                       'read_form_image': read_form_image(request),
-                                       "read_form_text": read_form_text(request)})
-    t = Tokenizer()
-    t.init(script)
-    ast = program_parser(t)
-    ast.exec(dsl_runtime)
 
 if __name__ == "__main__":
-    uvicorn.run("main:app", host="0.0.0.0", port=8000, log_level="info")
+    uvicorn.run("router:app", host="0.0.0.0", port=8000, log_level="info")

router.py

@@ -0,0 +1,35 @@
+from typing import Annotated, Union
+
+from fastapi import FastAPI, Request, status, Form
+from fastapi.responses import JSONResponse
+
+from src.dotchain.runtime.interpreter import program_parser
+from src.dotchain.runtime.tokenizer import Tokenizer
+from src.dotchain.runtime.runtime import Runtime
+from src.blackbox.blackbox_factory import BlackboxFactory
+import uvicorn
+from fastapi.middleware.cors import CORSMiddleware
+from injector import Injector
+
+app = FastAPI()
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+injector = Injector()
+blackbox_factory = injector.get(BlackboxFactory)
+
+@app.post("/")
+async def blackbox(blackbox_name: Union[str, None] = None, request: Request = None):
+    if not blackbox_name:
+        return await JSONResponse(content={"error": "blackbox_name is required"}, status_code=status.HTTP_400_BAD_REQUEST)
+    try:
+        box = blackbox_factory.call_blackbox(blackbox_name)
+    except ValueError:
+        return await JSONResponse(content={"error": "value error"}, status_code=status.HTTP_400_BAD_REQUEST)
+    return await box.fast_api_handler(request)

src/blackbox/asr.py

@@ -7,11 +7,13 @@ from fastapi.responses import JSONResponse
 from ..asr.rapid_paraformer.utils import read_yaml
 from ..asr.rapid_paraformer import RapidParaformer
 from .blackbox import Blackbox
+from injector import singleton
 
+@singleton
 class ASR(Blackbox):
 
-    def __init__(self, *args, **kwargs) -> None:
-        config = read_yaml(args[0])
+    def __init__(self, path = ".env.yaml") -> None:
+        config = read_yaml(path)
         self.paraformer = RapidParaformer(config)
     
     def __call__(self, *args, **kwargs):

src/blackbox/audio_chat.py

@@ -1,11 +1,18 @@
 from fastapi import Request, Response,status
 from fastapi.responses import JSONResponse
+from injector import inject, singleton
+
+from .asr import ASR
+from .tesou import Tesou
+from .tts import TTS
 
 from .blackbox import Blackbox
 
+@singleton
 class AudioChat(Blackbox):
 
-    def __init__(self, asr, gpt, tts):
+    @inject
+    def __init__(self, asr: ASR, gpt: Tesou, tts: TTS):
         self.asr = asr
         self.gpt = gpt
         self.tts = tts

src/blackbox/audio_to_text.py

@@ -3,9 +3,10 @@ from fastapi.responses import JSONResponse
 import speech_recognition as sr
 import filetype
 import io
-
+from injector import singleton
 from .blackbox import Blackbox
 
+@singleton
 class AudioToText(Blackbox):
 
     def __call__(self, *args, **kwargs):

src/blackbox/blackbox_factory.py

@@ -9,43 +9,40 @@ from .tesou import Tesou
 from .fastchat import Fastchat
 from .g2e import G2E
 from .text_and_image import TextAndImage
+from injector import inject, singleton
 
+@singleton
 class BlackboxFactory:
+    models = {}
 
-    def __init__(self) -> None:
-        self.tts = TTS()
-        self.asr = ASR(".env.yaml")
-        self.sentiment = Sentiment()
-        self.audio_to_text = AudioToText()
-        self.text_to_audio = TextToAudio()
-        self.tesou = Tesou()
-        self.fastchat = Fastchat()
-        self.audio_chat = AudioChat(self.asr, self.tesou, self.tts)
-        self.g2e = G2E()
-        self.text_and_image = TextAndImage()
+    @inject
+    def __init__(self, 
+                 audio_to_text: AudioToText,
+                 text_to_audio: TextToAudio,
+                 asr: ASR,
+                 tts: TTS, 
+                 sentiment_engine: Sentiment, 
+                 tesou: Tesou, 
+                 fastchat: Fastchat,
+                 audio_chat: AudioChat,
+                 g2e: G2E,
+                 text_and_image:TextAndImage ) -> None:
+        self.models["audio_to_text"] = audio_to_text
+        self.models["text_to_audio"] = text_to_audio
+        self.models["asr"] = asr
+        self.models["tts"] = tts
+        self.models["sentiment_engine"] = sentiment_engine
+        self.models["tesou"] = tesou
+        self.models["fastchat"] = fastchat
+        self.models["audio_chat"] = audio_chat
+        self.models["g2e"] = g2e
+        self.models["text_and_image"] = text_and_image
 
     def __call__(self, *args, **kwargs):
         return self.processing(*args, **kwargs)
 
-    def create_blackbox(self, blackbox_name: str) -> Blackbox:
-        if blackbox_name == "audio_to_text":
-            return self.audio_to_text
-        if blackbox_name == "text_to_audio":
-            return self.text_to_audio
-        if blackbox_name == "asr":
-            return self.asr
-        if blackbox_name == "tts":
-            return self.tts
-        if blackbox_name == "sentiment_engine":
-            return self.sentiment
-        if blackbox_name == "tesou":
-            return self.tesou
-        if blackbox_name == "fastchat":
-            return self.fastchat
-        if blackbox_name == "audio_chat":
-            return self.audio_chat
-        if blackbox_name == "g2e":
-            return self.g2e
-        if blackbox_name == 'text_and_image':
-            return self.text_and_image
+    def call_blackbox(self, blackbox_name: str) -> Blackbox:
+        model = self.models.get(blackbox_name)
+        if model is None:
             raise ValueError("Invalid blockbox type")
+        return model

src/blackbox/emotion.py

@@ -6,8 +6,10 @@ from fastapi.responses import JSONResponse
 from .blackbox import Blackbox
 from lagent.llms.lmdepoly_wrapper import LMDeployClient
 from lagent.llms.meta_template import INTERNLM2_META as META
+from injector import singleton
 
-class Sentiment(Blackbox):
+@singleton
+class Emotion(Blackbox):
 
     def __init__(self, model_name, model_url) -> None:
         self.model = LMDeployClient(

src/blackbox/fastchat.py

@@ -7,6 +7,8 @@ from .blackbox import Blackbox
 import requests
 import json
 
+from injector import singleton
+@singleton
 class Fastchat(Blackbox):
     
     def __call__(self, *args, **kwargs):

src/blackbox/g2e.py

@@ -7,7 +7,8 @@ from .blackbox import Blackbox
 import requests
 import json
 from openai import OpenAI
-
+from injector import singleton
+@singleton
 class G2E(Blackbox):
     
     def __call__(self, *args, **kwargs):

src/blackbox/sentiment.py

@@ -6,7 +6,8 @@ from fastapi.responses import JSONResponse
 from ..sentiment_engine.sentiment_engine import SentimentEngine
 from .blackbox import Blackbox
 
-
+from injector import singleton
+@singleton
 class Sentiment(Blackbox):
 
     def __init__(self) -> None:

src/blackbox/tesou.py

@@ -2,11 +2,20 @@ from typing import Any, Coroutine
 
 from fastapi import Request, Response, status
 from fastapi.responses import JSONResponse
+from injector import inject
+from ..configuration import TesouConf
 from .blackbox import Blackbox
 
 import requests
+from injector import singleton
 
+@singleton
 class Tesou(Blackbox):
+    url: str
+
+    @inject
+    def __init__(self, tesou_config: TesouConf):
+        self.url = tesou_config.url
     
     def __call__(self, *args, **kwargs):
         return self.processing(*args, **kwargs)
@@ -17,12 +26,11 @@ class Tesou(Blackbox):
 
     # 用户输入的数据格式为：[{"id": "123", "prompt": "叉烧饭，帮我查询叉烧饭的介绍"}]
     def processing(self, id, prompt)  -> str:
-        url = 'http://120.196.116.194:48891/chat/'
         message = {
             "user_id": id,
             "prompt": prompt,
         }
-        response = requests.post(url, json=message)
+        response = requests.post(self.url, json=message)
         return response.json()
     
     async def fast_api_handler(self, request: Request) -> Response:

src/blackbox/text_and_image.py

@@ -4,7 +4,8 @@ from fastapi import Request, Response, status
 from fastapi.responses import JSONResponse
 from .blackbox import Blackbox
 
-
+from injector import singleton
+@singleton
 class TextAndImage(Blackbox):
     
     def __call__(self, *args, **kwargs):

src/blackbox/text_to_audio.py

@@ -3,7 +3,8 @@ from fastapi.responses import JSONResponse
 from .blackbox import Blackbox
 from gtts import gTTS
 from io import BytesIO
-
+from injector import singleton
+@singleton
 class TextToAudio(Blackbox):
     
     def __call__(self, *args, **kwargs):

src/blackbox/tts.py

@@ -6,7 +6,9 @@ from fastapi import Request, Response, status
 from fastapi.responses import JSONResponse
 from .blackbox import Blackbox
 from ..tts.tts_service import TTService
+from injector import singleton
 
+@singleton
 class TTS(Blackbox):
 
     def __init__(self, *args, **kwargs) -> None:

src/configuration.py

@@ -0,0 +1,47 @@
+
+from dataclasses import dataclass
+from injector import Injector, inject
+import yaml
+import sys
+
+class Configuration():
+
+    @inject
+    def __init__(self) -> None:
+        config_file_path = ""
+        try:
+            config_file_path = sys.argv[1]
+        except:
+            config_file_path = ".env.yaml"
+        with open(config_file_path) as f:
+            cfg = yaml.load(f, Loader=yaml.FullLoader)
+            self.cfg = cfg
+    
+    def getDict(self):
+        return self.cfg
+
+    """
+    # yaml 檔中的路徑 get("aaa.bbb.ccc")
+    aaa:
+      bbb:
+         ccc: "hello world"
+    """
+    def get(self, path: str | list[str], cfg: dict = None):
+        if isinstance(path, str):
+            if cfg is None:
+                cfg = self.cfg
+            return self.get(path.split("."), cfg)
+        lenght = len(path)
+        if lenght == 0  or not isinstance(cfg, dict):
+            return None
+        if lenght == 1:
+            return cfg.get(path[0])
+        return self.get(path[1:], cfg.get(path[0]))
+        
+        
+class TesouConf():
+    url: str
+    
+    @inject
+    def __init__(self,config: Configuration) -> None:
+        self.url = config.get("tesou.url")