ChatGPT
ChatGPT copied to clipboard
acheong08
InvalidRequestError: That model does not exist
@acheong08 FYI If you add them in a list, GitHub will auto-expand to show all the titles. Makes it nicer to see/read:
- https://github.com/acheong08/ChatGPT/issues/517
- https://github.com/acheong08/ChatGPT/issues/523
- https://github.com/acheong08/ChatGPT/issues/532
- https://github.com/acheong08/ChatGPT/issues/530
Anyone with Azure's OpenAI service might be able to find the new model name.
Refer to
- https://learn.microsoft.com/en-us/rest/api/cognitiveservices/azureopenaistable/models/list?tabs=HTTP
@acheong08 https://github.com/waylaidwanderer/node-chatgpt-api/issues/19#issuecomment-1414759800
Found without access to Azure's OpenAI
import openai
import concurrent.futures
from revChatGPT.Official import Chatbot
defaultModel = "text-chat-davinci-002-XXXXXXXX"
successfulModels = []
def test_model(model):
try:
bot = Chatbot(api_key="", engine=model)
response = bot.ask("what is the capitol of the unitd states of america?")
if response:
successfulModels.append(model)
except openai.InvalidRequestError:
print("Model [" + model + "] failed to load. (Invalid Request Error)")
with concurrent.futures.ThreadPoolExecutor(max_workers=16) as executor:
for year in range(2020, 2031):
for month in range(1, 13):
for day in range(1, 32):
formatted_year = "{:04d}".format(year)
formatted_month = "{:02d}".format(month)
formatted_day = "{:02d}".format(day)
model = defaultModel.replace("XXXXXXXX", formatted_year + formatted_month + formatted_day)
executor.submit(test_model, model)
print("Successful Models: " + str(successfulModels))
Slightly modified the module to take in a "model" property
@canfam man!!!! you are soooooo smart!!! thankS! thank you!!!!!!!!
https://github.com/acheong08/ChatGPT/releases/tag/1.1.1
offical version if the interface is encapsulated how to control the context？
offical version if the interface is encapsulated how to control the context？
Conversations and stuff are stored locally (in memory or file)
You can take a read through the code to see how that was done (in the main function)
Confirmed it works and doesn't use credit on an account with $0. It still charges the same amount as davinci on account with $
Are you saying it still has the same date and just different name?
I already said that in discord Yeah the new model has the same date with a new name, its looks like just renamed not removed
I've just been playing around with updating some of my scripts/research/etc in my PoC ChatGPT repo this morning:
- https://github.com/0xdevalias/poc-chatgpt
- https://github.com/0xdevalias/poc-chatgpt/commits/main
And if you look at this commit, you can see that when I extracted/sorted the output of
/v1/modelsby created date, there is one at the top which curiously seems to share a date (20230126) with the original leaked model (text-chat-davinci-002-20230126):2023-01-26T23:36:25Z audio-transcribe-deprecatedThough if I try and use it with the completions API, I get an error:
⇒ curl https://api.openai.com/v1/completions -H 'Content-Type: application/json' -H "Authorization: Bearer $OPENAI_API_KEY" -d '{ "model": "audio-transcribe-deprecated", "prompt": "Say this is a test", "max_tokens": 7, "temperature": 0 }' { "error": { "message": "The server had an error processing your request. Sorry about that! You can retry your request, or contact [email protected] if you keep seeing this error. (Please include the request ID 316fc3215fa0a3278824c35138d3665a in your email.)", "type": "server_error", "param": null, "code": null } }Could be a coincidence.. but figured I would include the info here in case it's helpful to anyone.
Originally posted by @0xdevalias in https://github.com/acheong08/ChatGPT/issues/517#issuecomment-1416911061
You're using the wrong model
The error is back, "That model does not exist"
sorry, if I'm being dumby.
but how do you upgrade it?
should I manually modify, "model": "text-davinci-002-render in ChatGPT.py to the model name you posted?,
I tryied it, but it didn't work.
Please read the README for usage. You are meant to be using revChatGPT.Official
Example usage: https://github.com/acheong08/ChatGPT/blob/be9307d72d1d0baf86d757950edf0bd3648a5d4d/src/revChatGPT/main.py#L1-L123
My account was banned after frequently using the official API. Getting message "openai.error.RateLimitError: Your access was terminated due to violation of our policies, please check your email for more information. If you believe this is in error and would like to appeal, please contact [email protected]."
Is there a hard limit on how many requests we can send to the model before the account gets banned?
This is new. You seem to be the first one affected. Were you generating any NSFW or disinformation against their TOS?
Do you mind sharing the suspended API key? I would like to document the JSON error message
Or maybe can you share the response from the completions endpoint when suspended?
API key: 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
I was asking the model to generate the prompts that were used to train ChatGPT and I think this might be the reason for breaking TOS.
Interesting fact: The second you share an API key on GitHub, it is instantly revoked.
The cooperation between OpenAI and Microsoft runs deep