gboard isnt working you should improve your appp :D

I have a oppo watch4pro based on Android 11,

really wanna try it.

Hi,

I know one cannot spend too much time on those hobby projects, but do you find the time to add a simple complication for direct voice input or at least a better method to input spoken text? Currently, you are using a text input, but the best practice would be to have the menu to choose between input methods as in other apps (see documentation). Currently, it's at least 4 clicks to access a new chat and speech input, which is unhandy if used as an assistant, which is the primary purpose since you cannot provide long text for more elaborate tasks.

As a side note, did you have a look at Assistants API? This could be a powerful tool to have on the wrist.

Cheers!

I figured it would be best to create a separate issue for this (from #12 side discussion). Adding support to use a custom OpenAI Assistant instead of using the normal model endpoints might be less work than the complications, depending on the current setup (didn't have a detailed look yet :)). The Java API wrapper seems to already support Threads and Assistants or they are working on it.

Is it normal that I see start screen (Welcome to WristAssist) and API key request on every launch?

I want to ask a quick question, but it takes a lot of time with all screens :)

I have a custom api server, can it be a reason?

As something very far in the future, it is possible to add a feature to, for example, take the GPS coordinates of the smartwtch and create an interface that communicates, for example, with the OpenStreetMap API and process the output for questions such as:
"What restaurants are close to where I am?"

There's an option called "hands-free input" in the app settings, but I don't see what enabling it changes. Is this a way to use the app without needing to manually start and stop STT recordings?

I use the Galaxy Watch 6 classic. When I click on new chat I get a black screen. The app is still running, I can swipe back.

Ich nutze die Galaxy Watch 6 classic. Wenn ich auf Neuer Chat klicke bekomme ich ein schwarzen Bildschirm. Die app läuft noch, ich kann zurück wischen.

Hi I'm wondering if you might expose the ability to change the OpenAPI endpoint url so that users can integrate with other OpenApi compatible instances. My specific use case is a is an internal OpenApi compatible LLM that has ingested my companies private data. We currently use a web portal for general access but it would be nice to have access via our managed wearos devices for addhoc queries.

I'd like to suggest to allow pre-entering prompt templates.

This could work similarly to custom instructions or GPTs in ChatGPT. You pre-enter a system prompt and save them in a list. When creating a new chat, you can either start from scratch or select a template from the pre-defined list.

In a next step, I could also imagine allowing placeholders like %currentTime or %gpsLocation which are then replaced when starting the new chat.

I found that expressions can be recognized. So, if the uploaded image recognition function is added to the input interface, can it also be used?

There is a bug in the TTS function module. The answer can only read the sentence before the punctuation mark, and sometimes it will crash. Also, can you add a voice conversation function?

Hi I am Anshu Mishra, Currently running 2 startups in India and they are profitable, I am quite interested in doing one with you on WristAssist Application. Its quite unique and we can reach a good number of people if you take my business advices, I am ready to be your partner, You handle the tech and I the business, but first we need to talk so here's one of my Linkedin profile Link or you can mail me at - [email protected]

Hello, can you add the Azure OpenAI interface? In this way, you can use the API KEY of Azure OpenAI. The KEY of Azure OpenAI includes three functions: language APIS, Dall-e APIS, and Whisper APIS. Can you also add the model option of GPT4-32K? Thank you.

I'm looking at your ChatActivity.java file and it appears you are pointing to an older version of ChatGPT which will be discontinued. The way you select your version of ChatGPT also isn't futureproof as you will have to respecify a new version number of the model every time it updates.

        ChatCompletionRequest ccr = ChatCompletionRequest.builder()
                .model("gpt-3.5-turbo-0301")
                .messages(chatAdapter.getChatMessages())
                .build();

Instead of using "gpt-3.5-turbo-0301" with a version number you can just use "gpt-3.5-turbo" as the API will automatically use the most up to date version of ChatGPT.

As quoted from the OpenAI API Documentation:

The following models are the temporary snapshots, we will announce their deprecation dates once updated versions are available. If you want to use the latest model version, use the standard model names like gpt-4 or gpt-3.5-turbo.

Hello,

I've finally written the API key in the app. I've triple checked it's correct, and also I've ensured that I can use that API key using it in a python script on my laptop. But on the Smartwatch all I see when sending a prompt is Connection to OpenAI API failed: Please check your internet connection. I have it synced with my phone via bluetooth. Any idea?

From my understanding the google-services.json is being used to use firebase crashlytics but it is missing. it is also listed in the gitignore. What should I do as it does not let me build without it.