Improve user experience when api key is not valid

app/transcribe/app_utils.py

@@ -10,7 +10,7 @@
 from sdk import transcriber_models as tm  # noqa: E402 pylint: disable=C0413
 
 
-def create_responder(provider_name: str, config, convo, save_to_file: bool, 
+def create_responder(provider_name: str, config, convo, save_to_file: bool,
                      response_file_name: str):
     """Creates a responder / Inference provider object based on input parameters
     """

app/transcribe/args.py

@@ -7,6 +7,7 @@
 import interactions  # noqa: E402 pylint: disable=C0413
 from tsutils import utilities, duration, configuration  # noqa: E402 pylint: disable=C0413
 from sdk import audio_recorder as ar  # noqa: E402 pylint: disable=C0413
+import openai
 
 
 def create_args() -> argparse.Namespace:
@@ -35,6 +36,9 @@ def create_args() -> argparse.Namespace:
                           help='Save the API key for accessing OpenAI APIs to override.yaml file.\
                             \nSubsequent invocations of the program will not require API key on command line.\
                             \nTo not persist the API key use the -k option.')
+    cmd_args.add_argument('-vk', '--validate_api_key', action='store', default=None,
+                          help='Validate that it is a valid functioning api_key.\
+                           \nWithout the API Key only transcription works.')
     cmd_args.add_argument('-t', '--transcribe', action='store', default=None,
                           help='Transcribe the given audio file to generate text.\
                             \nThis option respects the -m (model) option.\
@@ -64,7 +68,7 @@ def create_args() -> argparse.Namespace:
     return args
 
 
-def handle_args_batch_tasks(args: argparse.Namespace, global_vars: TranscriptionGlobals):
+def handle_args_batch_tasks(args: argparse.Namespace, global_vars: TranscriptionGlobals, config: dict):
     """Handle batch tasks, after which the program will exit."""
     interactions.params(args)
 
@@ -77,6 +81,26 @@ def handle_args_batch_tasks(args: argparse.Namespace, global_vars: Transcription
         save_api_key(args)
         sys.exit(0)
 
+    if args.validate_api_key is not None:
+        chat_inference_provider = config['General']['chat_inference_provider']
+        if chat_inference_provider == 'openai':
+            base_url = config['OpenAI']['base_url']
+        elif chat_inference_provider == 'together':
+            base_url = config['Together']['base_url']
+
+        if utilities.is_api_key_valid(api_key=args.validate_api_key, base_url=base_url):
+            print('The api_key is valid')
+            base_url = config['OpenAI']['base_url']
+            client = openai.OpenAI(api_key=args.validate_api_key, base_url=base_url)
+            models = utilities.get_available_models(client=client)
+            print('Available models: ')
+            for model in models:
+                print(f'    {model}')
+            client.close()
+        else:
+            print('The api_key is not valid')
+        sys.exit(0)
+
     if args.transcribe is not None:
         with duration.Duration(name='Transcription', log=False, screen=True):
             output_file = args.output_file if args.output_file is not None else "transcription.txt"

app/transcribe/main.py

@@ -57,7 +57,7 @@ def main():
     # transcription. This order of initialization results in initialization of Mic, Speaker
     # as well which is not necessary for some batch tasks.
     # This does not have any side effects.
-    handle_args_batch_tasks(args, global_vars)
+    handle_args_batch_tasks(args, global_vars, config)
 
     # Initiate logging
     log_listener = al.initiate_log(config=config)

app/transcribe/ui.py

@@ -441,8 +441,8 @@ def create_ui_components(root, config: dict):
 
     response_enabled = bool(config['General']['continuous_response'])
     b_text = "Suggest Responses Continuously" if not response_enabled else "Do Not Suggest Responses Continuously"
-    freeze_button = ctk.CTkButton(root, text=b_text, command=None)
-    freeze_button.grid(row=1, column=1, padx=10, pady=3, sticky="nsew")
+    continuous_response_button = ctk.CTkButton(root, text=b_text, command=None)
+    continuous_response_button.grid(row=1, column=1, padx=10, pady=3, sticky="nsew")
 
     response_now_button = ctk.CTkButton(root, text="Suggest Response Now", command=None)
     response_now_button.grid(row=2, column=1, padx=10, pady=3, sticky="nsew")
@@ -483,6 +483,22 @@ def create_ui_components(root, config: dict):
     m.add_separator()
     m.add_command(label="Quit", command=root.quit)
 
+    chat_inference_provider = config['General']['chat_inference_provider']
+    if chat_inference_provider == 'openai':
+        api_key = config['OpenAI']['api_key']
+        base_url = config['OpenAI']['base_url']
+    elif chat_inference_provider == 'together':
+        api_key = config['Together']['api_key']
+        base_url = config['Together']['base_url']
+
+    if not utilities.is_api_key_valid(api_key=api_key, base_url=base_url):
+        # Disable buttons that interact with backend services
+        continuous_response_button.configure(state='disabled')
+        response_now_button.configure(state='disabled')
+        continuous_response_button.configure(state='disabled')
+        read_response_now_button.configure(state='disabled')
+        summarize_button.configure(state='disabled')
+
     def show_context_menu(event):
         try:
             m.tk_popup(event.x_root, event.y_root)
@@ -494,6 +510,6 @@ def show_context_menu(event):
     # Order of returned components is important.
     # Add new components to the end
     return [transcript_textbox, response_textbox, update_interval_slider,
-            update_interval_slider_label, freeze_button, lang_combobox,
+            update_interval_slider_label, continuous_response_button, lang_combobox,
             filemenu, response_now_button, read_response_now_button, editmenu,
             github_link, issue_link, summarize_button]

sdk/audio_recorder.py

@@ -159,6 +159,9 @@ def write_wav_data_to_file(self) -> str:
         if self.audio_file_name is None:
             return
 
+        if not os.path.exists(self.audio_file_name+'.bak'):
+            return
+
         frame_rate = self.source.SAMPLE_RATE
         sample_width = self.source.SAMPLE_WIDTH
         channels = self.source.channels
@@ -172,8 +175,8 @@ def write_wav_data_to_file(self) -> str:
             wf.setsampwidth(sample_width)    # pylint: disable=E1101
             wf.setframerate(frame_rate)    # pylint: disable=E1101
             wf.writeframes(data)    # pylint: disable=E1101
-            print(f'datasize: {len(data)}')
-        print(f'filesize: {os.path.getsize(self.audio_file_name)}')
+            # print(f'datasize: {len(data)}')
+        # print(f'filesize: {os.path.getsize(self.audio_file_name)}')
 
 
 class MicRecorder(BaseRecorder):

tsutils/utilities.py

@@ -3,6 +3,7 @@
 import copy
 import subprocess
 import zipfile
+import openai
 
 
 def merge(first: dict, second: dict, path=[]):
@@ -189,3 +190,35 @@ def zip_files_in_folder(folder_path: str, zip_file_name: str,
             if skip_zip_files and file.endswith(".zip"):
                 continue
             my_zip.write(f'{folder_path}/{file}')
+
+
+def get_available_models(client: openai.OpenAI) -> list:
+    """Get the list of available models from the provider.
+    """
+    try:
+        models = client.models.list()
+        return_val = []
+        for model in models.data:
+            return_val.append(model.id)
+    except openai.AuthenticationError as e:
+        print(e)
+        return None
+
+    return sorted(return_val)
+
+
+def is_api_key_valid(api_key: str, base_url) -> bool:
+    """Check if it is valid openai compatible openai key for the provider
+    """
+    openai.api_key = api_key
+
+    client = openai.OpenAI(api_key=api_key, base_url=base_url)
+
+    try:
+        client.models.list()
+        client.close()
+    except openai.AuthenticationError as e:
+        print(e)
+        return False
+
+    return True