Jarvis isn't passing text response to the next step

[neilbGH]

I'm trying to get jarvis to speak the current weather. All I get is the question converted to audio, not the answer jarvis fetched from the internet.

eg: please read the result of “get the current weather”

{"input": "please read the result of "get the current weather"", "task": "[{"task": "text2text-generation", "id": 0, "dep": [-1], "args": {"text": "get the current weather" }}, {"task": "text-to-speech", "id": 1, "dep": [0], "args": {"text": "<GENERATED>-0" }}]", "results": {"0": {"task": {"task": "text2text-generation", "id": 0, "dep": [-1], "args": {"text": "get the current weather"}}, "inference result": {"response": "The current weather is mostly sunny with temperatures in the high 80s, a slight breeze, and a chance of rain showers later in the afternoon."}, "choose model result": {"id": "ChatGPT", "reason": "ChatGPT is the best model for this task."}}, "1": {"task": {"task": "text-to-speech", "id": 1, "dep": [0], "args": {"text": "get the current weather"}}, "inference result": {"generated audio": "/audios/e275.flac"}, "choose model result": {"id": "facebook/fastspeech2-en-ljspeech", "reason": "This model has the highest number of likes (121) and has the most potential to solve the user request as it has a local inference endpoint for speed and stability"}}}, "response": "The current weather is mostly sunny with temperatures in the high 80s, a slight breeze, and a chance of rain showers later in the afternoon. To answer your request, I used two models: ChatGPT and facebook/fastspeech2-en-ljspeech. ChatGPT was used to generate the response to your request, and facebook/fastspeech2-en-ljspeech was used to generate an audio file of the response. The audio file is located at /audios/e275.flac. Is there anything else I can help you with?", "during": 16.613928079605103, "op": "response"}

[StillKeepTry]

Thanks. The current version of JARVIS is to connect AI models from HugigngFace, rather than the Internet.

[neilbGH]

What I'm looking for is a method to feed text into the pipeline so it can be manipulated like images, audio or video are.