Is it compatible with Ubuntu 22.04?

[fahdmirza]

Hi,
I have installed this tool in Ubuntu where my Ollama is also running with llama3 model on default port at local host. This program runs in background but how do you use it then? I am remotely accessing the GUI of my Ubuntu server from windows and there is no command key there. If it only works with MAC, can you please clearly mention that in your Read Me?

Also how do you integrate this with local Ollama?

regards

[olyaiy]

hi! just type /llama3 , and you should see a message saying "llama3 selected", then you'll be able to use the llama3:8b version.

if you want me to add another model, lmk and I will add them on :)
Or if you prefer to do it yourself, to head to api_calls.py and modify or duplicate the function to make the api call to the endpoint of your chosen model. Follow the ollama documentation https://github.com/ollama/ollama/blob/main/docs/api.md :

def query_llama3(text, image_base64=None):
    url = "http://localhost:11434/api/chat"
    
    payload = {
        "model": "llama3",
        "messages": [
            {
                "role": "system",
                "content": system_prompt
            },
            {
                "role": "user",
                "content": text
            }
        ],
        "stream": False
    }
    
    if image_base64:
        payload["messages"][1]["images"] = [image_base64]
    
    try:
        response = requests.post(url, json=payload)
        response.raise_for_status()
        result = response.json()
        return result['message']['content']
    except requests.exceptions.RequestException as e:
        return f"Error in Ollama API call: {str(e)}"

[fahdmirza]

As I mentioned in my question, it doesn't do anything when I type /llama3 anywhere such as browser search bar or in notepad or in vscode in Ubuntu.

What is alternative of command key in windows or Ubuntu?

[olyaiy]

oh i see, when you run run.sh in your terminal, whats the message you're getting? you should get a message telling you that its running.

[fahdmirza]

Sorry, as I mentioned earlier in my comment. It is definitely running. No error. It says application running in background.

But then when I try to use it in any text box in any windowwith /ask or with /llama3, it doesnt work.

[olyaiy]

ah okay i see. Im sorry to hear that. is your keyboard layout a mac keyboard with the cmmd button? if not, this might be a compatibility issue I'll have to fix.

[fahdmirza]

its windows keyboard. There is no command button on it.

[olyaiy]

ah i see :/ ok ill look into releasing a fix asap. (1-2 days), ill send a message here so you can test it. In the meantime, if you look in the TypeGPT.py file and are able to locate the issue feel free to make a pull request :) all contributions are welcome

[fahdmirza]

Thanks. I am just looking to review it on my channel once its done at youtube.com/@fahdmirza

[olyaiy]

Thank you! Thats very kind of you :) ill work on it asap and get it working for you.