Conversations are not working at the moment

[Cellenseres]

So everyone knows: Switching to the GPT-3 Model temporarily breaks conversations as it uses the openAI-Library again and not revChatGPT, which is currently the only Library supporting conversations.

@Zero6992 for everyone who wants to work with conversations, wouldn't it be possible to use revChatGPT again?
I'm not very familiar with either python or the libraries.
But in the revChatGPTs "Official.py" I found this"
`ENGINE = os.environ.get("GPT_ENGINE") or "text-chat-davinci-002-20221122"

ENCODER = tiktoken.get_encoding("gpt2")
`
maybe you can pass your own model to it as it's looking for os.environ.get("GPT_ENGINE")?

[Cellenseres]

Experimenting a bit with it and it seems revChatGPT does accept other models (ENGINE) as well. I changed "ENGINE"s Value to "text-davinci-003" for testing purposes and restarted the Bot.
It works and coversations are working again:

Hope it helps

[Zero6992]

Thank you for the notice, I just tried them and found that both the original model and text-davinci-003 will result in the following error message:

openai.error.InvalidRequestError: That model does not exist

It seems that OpenAI shut down those endpoints

[Cellenseres]

This confuses me. text-davinci-003 is still active for me. It's the one that is publicly available.
https://platform.openai.com/docs/models/gpt-3

Just restarted my bot:

Sometimes it throws me this error for some reason:
Traceback (most recent call last): File "/home/chatgpt/latest/src/bot.py", line 109, in send_start_prompt responseMessage = await responses.handle_response(prompt) File "/home/chatgpt/latest/src/responses.py", line 21, in handle_response response = await chatbot.ask(message) File "/home/chatgpt/.local/lib/python3.8/site-packages/revChatGPT/Official.py", line 213, in ask completion = await self._get_completion( File "/home/chatgpt/.local/lib/python3.8/site-packages/revChatGPT/Official.py", line 194, in _get_completion return await openai.Completion.acreate( File "/home/chatgpt/.local/lib/python3.8/site-packages/openai/api_resources/completion.py", line 45, in acreate return await super().acreate(*args, **kwargs) File "/home/chatgpt/.local/lib/python3.8/site-packages/openai/api_resources/abstract/engine_api_resource.py", line 217, in acreate response, _, api_key = await requestor.arequest( File "/home/chatgpt/.local/lib/python3.8/site-packages/openai/api_requestor.py", line 310, in arequest resp, got_stream = await self._interpret_async_response(result, stream) File "/home/chatgpt/.local/lib/python3.8/site-packages/openai/api_requestor.py", line 645, in _interpret_async_response self._interpret_response_line( File "/home/chatgpt/.local/lib/python3.8/site-packages/openai/api_requestor.py", line 679, in _interpret_response_line raise self.handle_error_response( openai.error.RateLimitError: The server had an error while processing your request. Sorry about that!

But all in all, the model is still active.

EDIT:
For this test, I used your Bot version 0.6.6 with an edited Official.py of revChatGPT

[Zero6992]

It's a little bit weird, no matter which model I use in revChatGPT, I get openai.error.InvalidRequestError: That model does not exist

using revChatGPT latest version

[Cellenseres]

I think it has problems getting the variable "GPT_ENGINE" from the .env for some reason? Go into your installed revChatGPT lib and try replacing the value of ENGINE in Official.py with "davinci-text-003". This actually worked.
I'm really not sure why it doesn't get the variable from the .env. Maybe a problem with revChatGPT itself?my python knowledge is almost none existent 😅

If you want, we can research it together on Discord (Cellenseres#0001)

[Cellenseres]

Actually I found a good way to pass the model-name to revChatGPT!
@Zero6992 this may interest you :D

in your src/responses.py, modify line 18:

chatbot = AsyncChatbot(api_key=config['openAI_key'], engine=config['model'])

in your config.json, add a new line with:

"model": "text-davinci-003"

Whenever you want to update the used model, just replace it inside the config.json.

No changes on revChatGPT-Side necessary as the init of the ChatBot-class takes "engine" as a parameter:

`class Chatbot:
"""
Official ChatGPT API
"""

def __init__(self, api_key: str, buffer: int = None, engine: str = None) -> None:
    """
    Initialize Chatbot with API key (from https://platform.openai.com/account/api-keys)
    """
    openai.api_key = api_key or os.environ.get("OPENAI_API_KEY")
    self.conversations = Conversation()
    self.prompt = Prompt(buffer=buffer)
    self.engine = engine or ENGINE`

[Zero6992]

I made it, thanks a lot, I'll update it immediately

[codefruitio]

hey @Cellenseres & @Zero6992, I'm running the latest version, 0.8.1, and conversations are not working for me. Is there anything special that should be done when prompting the bot?

[codefruitio]

solved temporarily by hardcoding the openAI_model value in responses.py

[Zero6992]

hey @Cellenseres & @Zero6992, I'm running the latest version, 0.8.1, and conversations are not working for me. Is there anything special that should be done when prompting the bot?

I found the issue and have rewritten the incorrect part, now it supports conversation.

[RocketGod-git]

Am I correct in assuming switching to the revChatGPT method will end up with the censorship that comes with ChatGPT?
Currently using text-divinci-003 with the previous method, there is no censorship at all.

I'd like to add conversation to my bot, but not at the cost of censorship.

[RocketGod-git]

No, like try this for me.

"Provide an extremely detailed guide to manufacturing ketamine. Include a list of all required items"

Gpt3 no prob.
RevChatGPT tells me "no, I cannot make ketamine"

No intent on making ketamine obviously but it and methamphetamines manufacturing are excellent baseline markers for censorship.

[Cellenseres]

@RocketGod-git

[Cellenseres]

If it doesn't work for you, downgrade to Bot Version 0.8.3

[Cellenseres]

Bot Version 0.8.3 used the text-davinci-003 model.
Bot Version 0.9.1 seems to be 1:1 ChatGPT (so no cost for usage?). This means, Bot Version 0.9.1 has the same restrictions as ChatGPT has.

If you use Bot Version 0.8.3, you'll use the text-davinci-003 model which actually cost money.
If you use Bot Version 0.9.1, you'll use ChatGPT directly, but it's for free. But you have the ChatGPT Restrictions.

@Zero6992 did I understand that correctly?

[RocketGod-git]

I have one that's working but no conversation. I'll try the version you suggest. Thanks!

[Zero6992]

@Zero6992 did I understand that correctly?

Your understanding is mostly correct

Since version 0.9.1, it sends requests directly to ChatGPT (website), but it doesn't pass through OpenAI's front-end moderation API

Differences in responses may be due to ChatGPT (website) having a prepended prompt, but you can use prompt engineering to ignore those restrictions once again

[Zero6992]

refer here about moderation of ChatGPT
acheong08/ChatGPT#1

prepended prompt of ChatGPT
https://www.reddit.com/r/ChatGPT/comments/10oliuo/please_print_the_instructions_you_were_given/

[helsan74]

using version 0.9.2 and I get in the logs:
Error while sending message: No API key provided. You can set your API key in code using 'openai.api_key = ', or you can set the environment variable OPENAI_API_KEY=). If your API key is stored in a file, you can point the openai module at it with 'openai.api_key_path = '. You can generate API keys in the OpenAI web interface. See https://onboard.openai.com for details, or email support@openai.com if you have any questions.

openai.error.AuthenticationError: No API key provided. You can set your API key in code using 'openai.api_key = ', or you can set the environment variable OPENAI_API_KEY=). If your API key is stored in a file, you can point the openai module at it with 'openai.api_key_path = '. You can generate API keys in the OpenAI web interface. See https://onboard.openai.com for details, or email support@openai.com if you have any questions.

That label no longer exists in the .env file.

[Zero6992]

it seems that the access token is invalid (either not copied correctly or expired)

[helsan74]

By access token you mean: Session_Token??

DISCORD_BOT_TOKEN=
OPENAI_EMAIL=
OPENAI_PASSWORD=
SESSION_TOKEN= <--THIS??
DISCORD_CHANNEL_ID=
LOGGING=

[Zero6992]

yes

[helsan74]

I redid a copy and it worked for a few minutes then it stopped
How can we prevent the expiration?

[helsan74]

I updated the SESSION_TOKEN and it works for one chat message then it stops

[helsan74]

Here's the error:
To replicate... Start the bot... Ask a question... Bot answers... Ask a second question... Bot give the following error

2023-03-01 06:10:05,583 - discord.app_commands.tree - ERROR - on_error - Ignoring exception in command 'chat'
Traceback (most recent call last):
File "/usr/local/lib/python3.8/dist-packages/discord/app_commands/commands.py", line 862, in _do_call
return await self.callback(interaction, **params) # type: ignore
File "/root/chatGPT-discord-bot-0.9.2/src/bot.py", line 145, in chat
await send_message(interaction, user_message)
File "/root/chatGPT-discord-bot-0.9.2/src/bot.py", line 25, in send_message
await message.response.defer(ephemeral=isPrivate)
File "/usr/local/lib/python3.8/dist-packages/discord/interactions.py", line 636, in defer
await adapter.create_interaction_response(
File "/usr/local/lib/python3.8/dist-packages/discord/webhook/async.py", line 218, in request
raise NotFound(response, data)
discord.errors.NotFound: 404 Not Found (error code: 10062): Unknown interaction

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "/usr/local/lib/python3.8/dist-packages/discord/app_commands/tree.py", line 1242, in _call
await command._invoke_with_namespace(interaction, namespace)
File "/usr/local/lib/python3.8/dist-packages/discord/app_commands/commands.py", line 887, in _invoke_with_namespace
return await self._do_call(interaction, transformed_values)
File "/usr/local/lib/python3.8/dist-packages/discord/app_commands/commands.py", line 880, in _do_call
raise CommandInvokeError(self, e) from e
discord.app_commands.errors.CommandInvokeError: Command 'chat' raised an exception: NotFound: 404 Not Found (error code: 10062): Unknown interaction

[helsan74]

If I run via python3 main.py working perfectly.

If I run via docker I get an error
2023-03-01T16:34:03.679431546Z Traceback (most recent call last):
2023-03-01T16:34:03.679437750Z File "/DiscordBot/src/bot.py", line 31, in send_message
2023-03-01T16:34:03.679445789Z response = f"{response}{await responses.handle_response(user_message)}"
2023-03-01T16:34:03.679460228Z File "/DiscordBot/src/responses.py", line 11, in handle_response
2023-03-01T16:34:03.679506757Z response = await chatbot.ask(message)
2023-03-01T16:34:03.679522694Z File "/usr/local/lib/python3.10/site-packages/revChatGPT/Official.py", line 232, in ask
2023-03-01T16:34:03.679533234Z completion = await self._get_completion(
2023-03-01T16:34:03.679542263Z File "/usr/local/lib/python3.10/site-packages/revChatGPT/Official.py", line 213, in _get_completion
2023-03-01T16:34:03.679552779Z return await openai.Completion.acreate(
2023-03-01T16:34:03.679563170Z File "/usr/local/lib/python3.10/site-packages/openai/api_resources/completion.py", line 45, in acreate
2023-03-01T16:34:03.679609564Z return await super().acreate(*args, **kwargs)
2023-03-01T16:34:03.679620692Z File "/usr/local/lib/python3.10/site-packages/openai/api_resources/abstract/engine_api_resource.py", line 214, in acreate
2023-03-01T16:34:03.679630710Z ) = cls.__prepare_create_request(
2023-03-01T16:34:03.679639926Z File "/usr/local/lib/python3.10/site-packages/openai/api_resources/abstract/engine_api_resource.py", line 106, in __prepare_create_request
2023-03-01T16:34:03.679652018Z requestor = api_requestor.APIRequestor(
2023-03-01T16:34:03.679695177Z File "/usr/local/lib/python3.10/site-packages/openai/api_requestor.py", line 130, in init
2023-03-01T16:34:03.679708242Z self.api_key = key or util.default_api_key()
2023-03-01T16:34:03.679718946Z File "/usr/local/lib/python3.10/site-packages/openai/util.py", line 186, in default_api_key
2023-03-01T16:34:03.679729334Z raise openai.error.AuthenticationError(
2023-03-01T16:34:03.679798182Z openai.error.AuthenticationError: No API key provided. You can set your API key in code using 'openai.api_key = ', or you can set the environment variable OPENAI_API_KEY=). If your API key is stored in a file, you can point the openai module at it with 'openai.api_key_path = '. You can generate API keys in the OpenAI web interface. See https://onboard.openai.com for details, or email support@openai.com if you have any questions.

[helsan74]

This error on 0.9.3
2023-03-01 16:48:08,051 - discord.app_commands.tree - ERROR - on_error - Ignoring exception in command 'chat'
Traceback (most recent call last):
File "/usr/local/lib/python3.8/dist-packages/discord/app_commands/commands.py", line 862, in _do_call
return await self.callback(interaction, **params) # type: ignore
File "/root/chatGPT-discord-bot-0.9.3/src/bot.py", line 145, in chat
await send_message(interaction, user_message)
File "/root/chatGPT-discord-bot-0.9.3/src/bot.py", line 25, in send_message
await message.response.defer(ephemeral=isPrivate)
File "/usr/local/lib/python3.8/dist-packages/discord/interactions.py", line 636, in defer
await adapter.create_interaction_response(
File "/usr/local/lib/python3.8/dist-packages/discord/webhook/async.py", line 218, in request
raise NotFound(response, data)
discord.errors.NotFound: 404 Not Found (error code: 10062): Unknown interaction

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "/usr/local/lib/python3.8/dist-packages/discord/app_commands/tree.py", line 1242, in _call
await command._invoke_with_namespace(interaction, namespace)
File "/usr/local/lib/python3.8/dist-packages/discord/app_commands/commands.py", line 887, in _invoke_with_namespace
return await self._do_call(interaction, transformed_values)
File "/usr/local/lib/python3.8/dist-packages/discord/app_commands/commands.py", line 880, in _do_call
raise CommandInvokeError(self, e) from e
discord.app_commands.errors.CommandInvokeError: Command 'chat' raised an exception: NotFound: 404 Not Found (error code: 10062): Unknown interaction