Put PROVIDER_MODEL_DICT into DEFAULT_API_CONFIG, working on choosing supporting file formats feature on llamaindex

Author: yjg30737

pyqt_openai/__init__.py

@@ -401,40 +401,15 @@ def move_bin(filename, dst_dir):
 
 STT_MODEL = "whisper-1"
 
-# Endpoint
-# https://platform.openai.com/docs/models/model-endpoint-compatibility
-OPENAI_ENDPOINT_DICT = {
-    "/v1/chat/completions": ["gpt-4o", "gpt-4o-mini", "o1-preview", "o1-mini"],
-    "/v1/completions": [
-        "text-davinci-003",
-        "text-davinci-002",
-        "text-curie-001",
-        "text-babbage-001",
-        "text-ada-001",
-        "davinci",
-        "curie",
-        "babbage",
-        "ada",
-    ],
-    "/v1/edits": ["text-davinci-edit-001", "code-davinci-edit-001"],
-    "/v1/audio/transcriptions": ["whisper-1"],
-    "/v1/audio/translations": ["whisper-1"],
-    "/v1/fine-tunes": ["davinci", "curie", "babbage", "ada"],
-    "/v1/embeddings": ["text-embedding-ada-002", "text-search-ada-doc-001"],
-    "/vi/moderations": ["text-moderation-stable", "text-moderation-latest"],
-}
-
 DEFAULT_TOKEN_CHUNK_SIZE = 1024
 
 # This doesn't need endpoint
 OPENAI_DEFAULT_IMAGE_MODEL = "dall-e-3"
 
 DEFAULT_DATETIME_FORMAT = "%Y-%m-%d %H:%M:%S"
 
-OPENAI_CHAT_ENDPOINT = "/v1/chat/completions"
-
-# Other models' configuration data
-DEFAULT_GEMINI_MODEL = "gemini/gemini-1.5-flash"
+# This has to be managed separately since some of the arguments are different with usual models
+O1_MODELS = ["o1-preview", "o1-mini"]
 
 # Overall API configuration data
 DEFAULT_API_CONFIGS = [
@@ -444,6 +419,7 @@ def move_bin(filename, dst_dir):
         "env_var_name": "OPENAI_API_KEY",
         "api_key": "",
         "manual_url": HOW_TO_GET_OPENAI_API_KEY_URL,
+        "model_list": ["gpt-4o", "gpt-4o-mini"] + O1_MODELS
     },
     # Azure
     {
@@ -503,13 +479,15 @@ def move_bin(filename, dst_dir):
         "api_key": "",
         "manual_url": HOW_TO_GET_GEMINI_API_KEY_URL,
         "prefix": "gemini",
+        "model_list": ["gemini/gemini-1.5-flash", "gemini/gemini-1.5-pro"]
     },
     # Anthropic
     {
         "display_name": "Anthropic",
         "env_var_name": "ANTHROPIC_API_KEY",
         "api_key": "",
         "manual_url": HOW_TO_GET_CLAUDE_API_KEY_URL,
+        "model_list": ["claude-3-5-sonnet-20240620"]
     },
     # AWS Sagemaker
     {
@@ -913,10 +891,6 @@ def move_bin(filename, dst_dir):
     },
 ]
 
-
-# This has to be managed separately since some of the arguments are different with usual models
-O1_MODELS = ["o1-preview", "o1-mini"]
-
 DEFAULT_LLM = "gpt-4o"
 
 G4F_PROVIDER_DEFAULT = "Auto"
@@ -925,13 +899,6 @@ def move_bin(filename, dst_dir):
 
 G4F_DEFAULT_IMAGE_MODEL = "flux"
 
-# Dictionary that stores the platform and model pairs
-PROVIDER_MODEL_DICT = {
-    "OpenAI": ["gpt-4o", "gpt-4o-mini"] + O1_MODELS,
-    "Gemini": ["gemini/gemini-1.5-flash", "gemini/gemini-1.5-pro"],
-    "Anthropic": ["claude-3-5-sonnet-20240620"],
-}
-
 # Constants related to the number of messages LLM will store
 MAXIMUM_MESSAGES_IN_PARAMETER = 40
 MAXIMUM_MESSAGES_IN_PARAMETER_RANGE = 2, 1000

pyqt_openai/chat_widget/right_sidebar/llama_widget/listWidget.py renamed to pyqt_openai/chat_widget/right_sidebar/llama_widget/filesWidget.py

@@ -18,7 +18,7 @@
 from pyqt_openai.util.common import getSeparator
 
 
-class FileListWidget(QWidget):
+class FilesWidget(QWidget):
     itemUpdate = Signal(bool)
     onDirectorySelected = Signal()
     clicked = Signal(str)

pyqt_openai/chat_widget/right_sidebar/llama_widget/llamaPage.py

@@ -5,7 +5,8 @@
 
 from pyqt_openai import SMALL_LABEL_PARAM
 from pyqt_openai.config_loader import CONFIG_MANAGER
-from pyqt_openai.chat_widget.right_sidebar.llama_widget.listWidget import FileListWidget
+from pyqt_openai.chat_widget.right_sidebar.llama_widget.filesWidget import FilesWidget
+from pyqt_openai.chat_widget.right_sidebar.llama_widget.supportedFileFormatsWidget import SupportedFileFormatsWidget
 from pyqt_openai.lang.translations import LangClass
 
 
@@ -20,9 +21,11 @@ def __initUi(self):
         self.__apiCheckPreviewLbl = QLabel()
         self.__apiCheckPreviewLbl.setFont(QFont(*SMALL_LABEL_PARAM))
 
-        self.__listWidget = FileListWidget()
-        self.__listWidget.clicked.connect(self.__setTextInBrowser)
-        self.__listWidget.onDirectorySelected.connect(self.__onDirectorySelected)
+        self.__filesWidget = FilesWidget()
+        self.__filesWidget.clicked.connect(self.__setTextInBrowser)
+        self.__filesWidget.onDirectorySelected.connect(self.__onDirectorySelected)
+
+        self.__supportedFileFormatsWidget = SupportedFileFormatsWidget()
 
         self.__txtBrowser = QTextBrowser()
         self.__txtBrowser.setPlaceholderText(
@@ -33,7 +36,8 @@ def __initUi(self):
         self.__txtBrowser.setMaximumHeight(150)
 
         lay = QVBoxLayout()
-        lay.addWidget(self.__listWidget)
+        lay.addWidget(self.__supportedFileFormatsWidget)
+        lay.addWidget(self.__filesWidget)
         lay.addWidget(self.__txtBrowser)
         lay.setAlignment(Qt.AlignmentFlag.AlignTop)
 
@@ -42,7 +46,7 @@ def __initUi(self):
         self.setDirectory()
 
     def __onDirectorySelected(self):
-        selected_dirname = self.__listWidget.getDirectory()
+        selected_dirname = self.__filesWidget.getDirectory()
         self.onDirectorySelected.emit(selected_dirname)
 
     def __setTextInBrowser(self, txt_file):
@@ -51,4 +55,4 @@ def __setTextInBrowser(self, txt_file):
 
     def setDirectory(self):
         directory = CONFIG_MANAGER.get_general_property("llama_index_directory")
-        self.__listWidget.setDirectory(directory, called_from_btn=False)
+        self.__filesWidget.setDirectory(directory, called_from_btn=False)

pyqt_openai/chat_widget/right_sidebar/llama_widget/supportedFileFormatsWidget.py

@@ -0,0 +1,26 @@
+from PySide6.QtWidgets import QWidget, QVBoxLayout, QLabel
+
+from pyqt_openai.widgets.checkBoxListWidget import CheckBoxListWidget
+
+
+class SupportedFileFormatsWidget(QWidget):
+    def __init__(self, parent=None):
+        super().__init__(parent)
+        self.__initUi()
+
+    def __initUi(self):
+        self.__listWidget = CheckBoxListWidget()
+        self.__listWidget.checkedSignal.connect(self.__sendCheckedSignal)
+
+        ext_lst = ['.txt', '.docx', '.xlsx', '.md', '.pdf', '.html']
+
+        self.__listWidget.addItems(ext_lst, checked=True)
+
+        lay = QVBoxLayout()
+        # TODO LANGUAGE
+        lay.addWidget(QLabel('Supported File Formats'))
+        lay.addWidget(self.__listWidget)
+        self.setLayout(lay)
+
+    def __sendCheckedSignal(self, r_idx, state):
+        print(r_idx, state)

pyqt_openai/util/common.py

@@ -7,6 +7,7 @@
 
 import asyncio
 import base64
+import csv
 import json
 import os
 import random
@@ -19,8 +20,6 @@
 import traceback
 import wave
 import zipfile
-import csv
-
 from datetime import datetime
 from io import BytesIO
 from pathlib import Path
@@ -61,10 +60,7 @@
     AUTOSTART_REGISTRY_KEY,
     is_frozen,
     G4F_PROVIDER_DEFAULT,
-    PROVIDER_MODEL_DICT,
     O1_MODELS,
-    OPENAI_ENDPOINT_DICT,
-    OPENAI_CHAT_ENDPOINT,
     STT_MODEL,
     DEFAULT_DATETIME_FORMAT,
     DEFAULT_TOKEN_CHUNK_SIZE, DEFAULT_API_CONFIGS, INDENT_SIZE,
@@ -580,9 +576,9 @@ def get_chat_model(is_g4f=False):
     if is_g4f:
         return get_g4f_models()
     else:
-        all_models = [
-            model for models in PROVIDER_MODEL_DICT.values() for model in models
-        ]
+        all_models = []
+        for obj in DEFAULT_API_CONFIGS:
+            all_models.extend(obj["model_list"])
         return all_models
 
 def get_gemini_argument(model, system, messages, cur_text, stream, images):
@@ -659,18 +655,6 @@ def set_api_key(env_var_name, api_key):
     # Set environment variables dynamically
     os.environ[env_var_name] = api_key
 
-
-def get_openai_model_endpoint(model):
-    for k, v in OPENAI_ENDPOINT_DICT.items():
-        endpoint_group = list(v)
-        if model in endpoint_group:
-            return k
-
-
-def get_openai_chat_model():
-    return OPENAI_ENDPOINT_DICT[OPENAI_CHAT_ENDPOINT]
-
-
 def get_image_url_from_local(image, is_openai=False):
     """
     Image is bytes, this function converts it to base64 and returns the image url
@@ -693,9 +677,9 @@ def get_message_obj(role, content):
 
 # Check which provider a specific model belongs to
 def get_provider_from_model(model):
-    for provider, models in PROVIDER_MODEL_DICT.items():
-        if model in models:
-            return provider
+    for obj in DEFAULT_API_CONFIGS:
+        if model in obj["model_list"]:
+            return obj["display_name"]
     return None