HTTPError: 401 Client Error: Unauthorized for url: https://api.openai.com/v1/completions

[TouchstoneTheDev]

╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ /home/c/.local/lib/python3.10/site-packages/sgpt.py:150 in main │
│ │
│ 147 │ │ max_tokens = 1024 │
│ 148 │ if editor: │
│ 149 │ │ prompt = get_edited_prompt() │
│ ❱ 150 │ response_text = openai_request(prompt, model, max_tokens, api_key, temperature, top_ │
│ 151 │ # For some reason OpenAI returns several leading/trailing white spaces. │
│ 152 │ response_text = response_text.strip() │
│ 153 │ typer_writer(response_text, code, shell, animation) │
│ │
│ ╭─────────────────────────────────────────── locals ───────────────────────────────────────────╮ │
│ │ animation = True │ │
│ │ api_key = :) │ │
│ │ code = False │ │
│ │ editor = False │ │
│ │ execute = False │ │
│ │ max_tokens = 2048 │ │
│ │ model = <Model.davinci: 'text-davinci-003'> │ │
│ │ prompt = 'nginx default config file location' │ │
│ │ shell = False │ │
│ │ spinner = True │ │
│ │ temperature = 1.0 │ │
│ │ top_probability = 1.0 │ │
│ ╰──────────────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/c/.local/lib/python3.10/site-packages/sgpt.py:69 in wrapper │
│ │
│ 66 │ │ text = TextColumn("[green]Requesting OpenAI...") │
│ 67 │ │ with Progress(SpinnerColumn(), text, transient=True) as progress: │
│ 68 │ │ │ progress.add_task("request") │
│ ❱ 69 │ │ │ return func(*args, **kwargs) │
│ 70 │ │
│ 71 │ return wrapper │
│ 72 │
│ │
│ ╭─────────────────────────────────────────── locals ───────────────────────────────────────────╮ │
│ │ args = ( │ │
│ │ │ 'nginx default config file location', │ │
│ │ │ <Model.davinci: 'text-davinci-003'>, │ │
│ │ │ 2048, │ │
│ │ │ │ │
│ │ ' :)
│ │
│ │ │ 1.0, │ │
│ │ │ 1.0 │ │
│ │ ) │ │
│ │ func = <function openai_request at 0x7f39b3a579a0> │ │
│ │ kwargs = {} │ │
│ │ progress = <rich.progress.Progress object at 0x7f39b3a32fe0> │ │
│ │ text = <rich.progress.TextColumn object at 0x7f39b3a33370> │ │
│ ╰──────────────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/c/.local/lib/python3.10/site-packages/sgpt.py:101 in openai_request │
│ │
│ 98 │ │ "top_p": top_p, │
│ 99 │ } │
│ 100 │ response = requests.post(API_URL, headers=headers, json=data, timeout=180) │
│ ❱ 101 │ response.raise_for_status() │
│ 102 │ return response.json()["choices"][0]["text"] │
│ 103 │
│ 104 │
│ │
│ ╭─────────────────────────────────────────── locals ───────────────────────────────────────────╮ │
│ │ api_key = ' :) │ │
│ │ data = { │ │
│ │ │ 'prompt': 'nginx default config file location', │ │
│ │ │ 'model': <Model.davinci: 'text-davinci-003'>, │ │
│ │ │ 'max_tokens': 2048, │ │
│ │ │ 'temperature': 1.0, │ │
│ │ │ 'top_p': 1.0 │ │
│ │ } │ │
│ │ headers = { │ │
│ │ │ 'Content-Type': 'application/json', │ │
│ │ │ 'Authorization': 'Bearer │ │
│ │ :)
│ │
│ │ } │ │
│ │ max_tokens = 2048 │ │
│ │ model = <Model.davinci: 'text-davinci-003'> │ │
│ │ prompt = 'nginx default config file location' │ │
│ │ response = <Response [401]> │ │
│ │ temperature = 1.0 │ │
│ │ top_p = 1.0 │ │
│ ╰──────────────────────────────────────────────────────────────────────────────────────────────╯ │
│ │
│ /home/c/.local/lib/python3.10/site-packages/requests/models.py:1021 in raise_for_status │
│ │
│ 1018 │ │ │ ) │
│ 1019 │ │ │
│ 1020 │ │ if http_error_msg: │
│ ❱ 1021 │ │ │ raise HTTPError(http_error_msg, response=self) │
│ 1022 │ │
│ 1023 │ def close(self): │
│ 1024 │ │ """Releases the connection back to the pool. Once this method has been │
│ │
│ ╭─────────────────────────────────────────── locals ───────────────────────────────────────────╮ │
│ │ http_error_msg = '401 Client Error: Unauthorized for url: │ │
│ │ https://api.openai.com/v1/completions' │ │
│ │ reason = 'Unauthorized' │ │
│ │ self = <Response [401]> │ │
│ ╰──────────────────────────────────────────────────────────────────────────────────────────────╯ │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯

[TheR1D]

This happens since you provided ":)" as your API key. Follow these steps to change your API key.

[TouchstoneTheDev]

I don't want to expose my keys so i just changed it in this post to :)

[TouchstoneTheDev]

btw thenks for above link/reference will try

[TheR1D]

Since you didn't mention it, I thought this is a real value 😀. Just make sure you have right API key in ~/.config/shell-gpt/api_key.txt

[TouchstoneTheDev]

Now I cannot access chatgpt
it's just checking my browser
cannot go further with it !

[kid-gorgeous]

Hiya, I've ran into the same problem, and maybe I can help with the situation?

I had to make sure one I messed this up that I removed all traces of Shell GPT.
I removed all related packages from (they are labeled accordingly)

/Users/<user>/Library/Python/3.11/lib/python/site-packages
^ I also noticed and api_key.txt here, but I think its been removed..

Then, I noticed after a clean installation, that the same api string was still being used. After some looking around, I found my .config file.. You must either edit this, or completely delete it and start fresh :)

If you would like to remove it, use ls -a, and open your .config directory. It is a hidden directory. From here you will see the dir shell-gpt/. Then you've found the right file! rm api_key.txt
Since you are starting fresh Please think about creating a new api key, just to eliminate anymore confusion.

Otherwise ... this will also work ~/.config/shell-gpt/api_key.txt, but just incase you'd like to know where everything is at, please use the prior. Either way, edit this file once more! :D

If you're still having anymore trouble, double check your path extension. I notice that it might not automatically be added everytime, but it's seriously such an easy fix. The extension can be added by using the following:

export "/Users/<user>/Library/Python/3.11/bin/:$PATH"
echo "PATH=$PATH:/Users/<user>/Library/Python/3.11/bin/" >> "$HOME/.$(basename $SHELL)rc"

The later choice might be your best option. After I did this, it worked. Every instance in one terminal, in different shells, in two sperate terminals, etc. I have noticed a 500 Error, but these, and other related issues are because of the OpenAI API.

Kudos to the creator for this. I can imagine how overwhelming these can get, so hopefully this helps @tanmay-veer?

[TouchstoneTheDev]

Ig no problem now using shellg bur I am not able to access chatgpt :()

[TouchstoneTheDev]

is there any guide for prompt engineering or doc for text-davinci-003

[kid-gorgeous]

@tanmay-veer yeah, there might be more around the internet refer to this website for the help with completions, and refer to this website for models. I would recommend still using davinci because of its token cost.