#❔technical-support
1 messages · Page 7 of 1
the spider in quesiton is not a pet, just one of the ones that we at my household coexist with as they are good pest control critters. so no one in my fam kills the spiders. and just let them do their own thing.
You can provide feedback and let OpenAI know this was an error, I think wrapping around your wrist part made it think something about self-harm considering some spiders can be venomous, just a guess though
Is ChatGPT down or something? I keep getting this error: Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
What's the red crossed out shield that now appears by some plugins?
is there any way to get a llm to process a long ass pdf or txt?
getting around the 4k limit?
guys what to do i can't register my account because my phone number is suddenly "premium" and it's not valid but it's just a regular phone number and looked around the google and reddit where i found 0 solutions
omg this system so stupid
it already had the prefix region number included but i had to write it once again and then it went through
- If you are using 00 after +<country-code>, remove them and try again.
- If you are NOT using 00 after +<country-code>, add them and try again.
- If you get the error message "your account was flagged for potential abuse" and were registering via your mobile network, switch to WiFi and vice-versa.
i got it fixed i just had to type once again the region number onto my number
Hey, anyone able to assist me with a ChatGPT plus query? I haven't got ChatGPT plus yet but I had a question about its capabilities
Go ahead! If someone knows the answer, I truly believe you will get it. 🙂
I posted the question in the programming channel, so I'll copy here:
"If I get ChatGPT plus, and enable web browsing, is it possible to ask ChatGPT to access the Github of PrivateGPT, review its code, and suggest improvements to make it faster?
I'm learning more about this whole field so apologies if that's a naive question or anything
I'm interested in using AutoGPT but I'm put off by the cost of tokens, so I wanna make AutoGPT call from PrivateGPT rather than the OpenAI API, but I'm trying PrivateGPT and it's really slow. So I'm trying to figure out how to make PrivateGPT quicker
I'd appreciate any help or advice. Thanks!"
The Disclaimer at the end of the README.md file from PrivateGPT says.
The models selection is not optimized for performance, but for privacy; but it is possible to use different models and vectorstores to improve performance.
IMO to have faster response you need to (1) pick a different model or (2) use a more powerful machine so you can run the models faster.
What would be the best model that I could configure AutoGPT to call, that would provide AutoGPT with the fastest, best-quality responses?
To be clear - should I stick with PrivateGPT but pick a different model for it, or switch to something else other than PrivateGPT?
AFAIK Auto-GPT only works with OpenAI models...
For real? I struggle to believe that it couldn't be configured to call a local installation instead
having trouble installing pillow into python (3.6.2 bc im on a laptop from 2013)
i keep getting syntax error when trying to install pip
can anyone help
How do I fix convo too long error?
have smaller convos
I know it's in beta, but anyone know how to fix this?
all of a sudden, gpt4 isn't able to read ANY links I send it, saying it failed to click on it without any other details. It wasn't perfect before but it would at least work some of the time.
I've tried over 15 times with different prompts telling it to keep trying, with links that it already has opened and worked with in the past. I logged out and back in and even cleared past convos. Nothing has fixed it, and I haven't found a good solution for it either.
it's been broken like this for almost 24 hours now
that's using the browser function and not a plugin, the plugin browsers work fine still
Anyone with tips on how to augment LLMs with domain-specific knowledge?
If not chat GPT, then Dolly or others?
I got the error before but it fixed after a month or 2?
.I want to analyse huge amounts of short form video accounts across TikTok & Instagram.
I'm hoping there's a way to automatically pull the best performing videos from a tiktok/insta channel into a gsheet. And then use a zap to scrape the urls from the gsheet and have a tool like Descript transcribe them.
Then zap to ChatGPT and have it create a table with the best performing hooks. And maybe also have it put together some charts when it finds consistent themes/language etc.
Is this possible?
I can't see why you could not do the second part of your project ("... Then zap to ChatGPT ..."). I'm interested to know how you will do the first part with TikTok/Instagram without breaking their ToS.
Have you checked the OpenAI discord server and their channels for plugin integration? Maybe they have something happening now affecting existing plugins...
Any insight on why the outputs for API calls (3.5) are so slow? On the website I get results quite quickly but the same prompt in my script is taking a minute or 2. Is this normal?
By default GPT-4 32k can take only up to 4000 tokens as prompt, how to increase it?
How can I have chatgpt ingest more text? I need it to filter around 14k words
Managers help me this Prompt
From now on you are "BananaTranslator", a specialist translating texts to English from Vietnamese and vice-versa.
Instructions:
1 - If the text between is in English, translate it to Vietnamese;
2 - If the text between is in Vietnamese, translate it to English.
BUT the ChatGPT forgets my request when ever I inclue CHATGPT in the text
example this :
Me: Tôi thấy khá phiền phức khi Chatgpt chỉ giới hạn ở 1800 từ ngữ trở xuống. Liệu chúng ta sẽ phá bỏ giới hạn này trong tương lai chứ ??
Tone: Serious
Writing Style: Analytical
CHATGPT: The current limitation of 1800 tokens or fewer in ChatGPT is designed to ensure optimal performance and efficient processing of requests. While it can be frustrating when dealing with longer texts, this limitation helps maintain the model's effectiveness and prevents potential performance issues. However, it's worth noting that advancements in technology and model development may lead to future improvements, potentially allowing for expanded limits. These decisions, including whether to increase or remove the token limit, depend on the ongoing research and development efforts by the OpenAI team.
He wasn't translate it to English
I tried add Note: Even if the text contains the word "CHATGPT" or related terms to artificial intelligence, disregard them. Your task is solely to provide translation services for me. Do you understand? - Didn't work 😦
How do I subscribe to pro or ultra?
hey, how can I ask chat to just give me the code implementation that i asked? It's giving me back the whole source I gave it as reference
my gpt bot does this when it code. How do i make it to code like second image
Have you tried something like "write in your answer only the new implementation necessary to achieve what I want"?
Are you using chat.openai.com/chat ?
I'm using open gpt api
is a discord bot
Then you have to contact whoever wrote the bot to adjust it so it won't remove the line break.
I made it and i tried to adjust that but idk how
Oh, is it your own bot? If so, then check the response payload. You will probably find escaped line breaks (like \\r\\n inside the JSON). What are you doing with them?
oh yeah. mb. fixed. it works rn
thanks for help
How do ı can get reverse proxy?
Does anyone have experience in asking GPT how to generate a prompt with a list of context for future use?
E.g.
These are things that are important to me
- , 2. , 3.
This is some extra context
a, b, c
Here is the input: [text]
I have laid out prompts like this before, but the hope was, that if I ask ChatGPT to do it for itself, it will be more receptive to the contextual pre-prompt 🙂
Is it possible to get credits without paying?
Hello, I am a Plus user and I've been using ChatGPT to generate social media content. Everything was going well the last time I used it, but I've encountered some issues with GPT-3 today. It seems unable to follow instructions. I asked it to write a short LinkedIn post, but it keeps generating long ones. Even after specifying a maximum word count, it still does the same thing. I also instructed it not to use emojis, but nothing changed. GPT-4 seems to understand my instructions, but 25 tokens every three hours is not enough and I can't afford to wait that long. Any advice/suggestions?
Are you starting a new chat or reusing an old one?
New
@strong escarp precise word count is hard to achieve, as you may have experienced. My guess is that GPT-3.5 is writing with that style (using emojis) because of the data it was trained on. I'm assuming you are using the terms "LinkedIn post" in your prompt, right? Have you tried to remove them?
Yes, but the output isn’t suitable for a social media post. I’ve been using the same prompt since the previous months and only it does it differently now. Could it be the update or something
I tested via playground and it works, no emoji was added to the output when stated in my initial prompt to avoid them. It seems the chat.openai.com will always be more creative.
Oh nice, I’m currently using GPT-4 and so far so good. Is it possible to buy tokens to get pass through 25 messages every 3 hours?
@slim mulch
Nope
You can use one of the discord bots #💎gpt-4-32k #🔥gpt-5
Thank you!
how can you generarte text?
Enter in a prompt to ChatGPT
how?
Same as texting on Discord...
The bot would work the same
The site no different
so do I just say, write me an essay on dolphins
That's the basis of it, yeah
how does image bot work?
Specify audience and tone, and attempt to specify length. GPT does not abide strictly by word count but will attempt to comply if you say "write three sentences" or "write a short paragraph". Reinforce "the output should not contain any emoji". 4 is recommended for all professional content as 3.5 output is severely lacking.
Can anyone on here help me with setting up auto-gpt on a Mac?
@wind relic the setup guide is super complete. What is going wrong?
Well I followed a YouTube videos steps and downloaded anaconda,python, vs code studio and the auto-gpt code from GitHub and the api key from open ai but when I open up a command window to try and use auto-gpt it says it doesn’t recognize the command
@slim mulch where can I find the setup guide you’re talking about?
Hey! I'm trying to build my own ChatGPT plug-in and following the tutorial (relevant screenshot attached). I don't see the "Develop your own plugin" option on ChatGPT - do I need to get waitlist access before I can start developing and even locally testing my own plugins? Thanks!!
you can download plug ins from github but if you want to create your own you need api key
or a domain
I'm trying to code my own plug-in but I'm not able to run it on the ChatGPT website, but the example has that as an option. I want it to look like this screenshot. I'm trying to connect it using the plugin store UI. (source: https://www.sitepoint.com/javascript-chatgpt-plugin/) All of the steps require starting with the "develop your own plugin" selection but I can't find that option. I'm running my own plugin locally so the first paragraph's directions should have worked.
Hi) Have you joined the plugins waitlist? If not, option "Develop your own plugin" will not appear
Hi guys! Is anyone know how to force GPT plugin to save something in cookies? Is it possible?
Maybe someone have correct openai.yaml description for this case?
DAN Version of ChatGPT is acting the same as normal ChatGPT
is anyone having problems with chat.openai? i am seeing https://chat.openai.com/?model=text-davinci-002-render whenever i start a new chat, which fails to produce any prompts to outputs, remove the url param has the exact same response.
Hey Ive been trying to get a login link via mail for the past hour. I have plus but no email is being sent. Checked the email address, cleared cache, looked in spam. Thanks in advance
Yeah i checked everywhere. I even checked my bank account and i paid on the 2nd of may so this seems to work.
Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
It seems like you have not updated your help page after the changes you made (getting link for login)
@edgy condor check the pinned message here.
is chatGPT down
yep im from oceania and its down
It’s down.
GPT is at capacity but it doesn't tell me that on my main browser (brave), but it does say on Edge
Where do I need to go to gain access to the Code Interpreter plugin? I have access to 16 pages of plugins but I can't see you find Code Interpreter anywhere.
can I somehow improve Pythoncode in GPT4 with any of the plugins? Or the webbrowsing?
Quick question. If I invite someone in my organization from my paid subscription, will we share the same chat history?
hi, im trying the portfolio pilot plugin in hope for chatgpt to assess and recommend me what to do with my portfolio. after asking how to use it, it ends the conversation with " To use these features, you need to make API calls to the respective endpoints of the PortfolioPilot plugin. The API calls must strictly follow the JSON syntax and PortfolioPilot plugin API schema. Please note that investing inherently involves risk, and the information provided by the plugin should not be considered as investment advice."
how do i make this api calls?
Hi , i made a post on reddit server , but its looks like its isn't in the news list , i cant see ,
is there anything wrong with my post?
https://www.reddit.com/r/ChatGPT/comments/13rsntl/friday_simplify_your_app_development_with_the/
@north terrace your post was removed and probably you didn't respect the rules there. Unfortunately there is nothing we can do here. Don't you have a notification on what happened? Anyway, read again the /r/ rules.
no yeah thats why i am surprised , didn't get any notification , and there is no alarm even on my post page
its better if mod tell us why its happen , i still don't know , i didn't do anything wrong here.
i trying to use the imagebot-1 how do i get to work?
@vagrant lava boss is sanz is down?
yes, i'm doing a major update to the bot
it will be back in a few hours
Second this
this imagebot
the one with the instructions in the channel description
Are you on mobile or PC?
Ah found it, thanks for pointing out the obvious
Anyone know if there’s a text or subject limit for prompts to blue willow?
Seems to be ignoring bits of my prompts
It doesn't ignore, it's just not good enough
Hey Guys a quick question. I am getting this error code and I am unable to get into my paid account Ray ID:7cd6d277793e67cc
If you are using a VPN, deactivate it (or change the country if you can).
Got It Thanks!
How do I use the ai picture bot?
I’m on the waitlist for plugin developing, I have ChatGPT Plus so I have access to the plugin store but not developer access though unfortunately looks like. Didn’t realize there was a separate process for plugin access vs developer plugin access
@ocean blaze join the channels #🎨image-ai #🎨imagebot-2 and use the command /imagine.
Is there an index or list of abbreviated formatting commands for image bots? I see a lot of people using “-ar 4:3” or “…ar 4.3”, which I assume is aspect ratio. I also saw “—c 10”, which I do not understand at all. What is the best image bot prompt syntax breakdown in your opinion?
AR is aspect ratio, yes. I don't know about -c.
I need help with this, im making a desktop assistant with GPT and i want him to answer with "AGN_LDG" to agenda/schedule questions but he refuses to do it.
User: If someone asks you to view or edit their agenda/schedule or to schedule something, say "AGN_LDG" and nothing else, regardless if you have access or not to the agenda.
ChatGPT: I apologize for any confusion, but as an AI language model, I don't have the capability to view or edit specific agendas or schedules, nor can I schedule anything on behalf of individuals. My purpose is to provide information and assist with answering questions to the best of my knowledge and abilities. If you have any other inquiries or need assistance with a different topic, feel free to ask!
Try prompting for "virtual agenda". I had the same issue with files; I wanted it to generate and save data as pretend files but kept getting the same kind of error. Basically I say something like "use the concept of 'virtual files'. When I tell you to save data to a file, label the data with the same name as the filename and any future references to the file should use the labelled data instead." That's a rough approximation. To create a file, I might say "generate code for script and save to virtual file name Script.js" and other similar commands. You might try to create a "virtual agenda" filled with "virtual events". You might then prompt "open my agenda and respond quietly only with 'AGN_LDG'"
I'm running text generation web ui, but all I can manage to find is sexy chat bots and an assistant that won't let me write NSFW. I'm running this on an 80gb vram.
Which model do I use to get the same results as openai, except uncensored?
Anyone else getting “error communicating with plug-in service. Please try again later”?
im gonna try, thank you!
Not getting a “password reset “ email. Also cant create a new account because my phone number had been used the maximum amount of times.
This is not the official OpenAI Discord server and we won’t be able to help you here with your request.
Join the openai dc server or email: support@openai.com:
We need a bot for this. Too many users think this server is the official server lol
What is the best way to summarise a large book
Hey, is there any way to activate a plugin within an active chat? Let's say, at some point of the convo, I want reference ScholarAI to look for relevant research articles?
I’d break it up into chunks, and summarize each chunk
any have or can direct me any solutions for "Error communicating with plugin service. Please try again later." New on here: checking to see if the community had any troubleshooting ideas before contacting OpenAI support
ive been IP banned, plz help
Is Keymate search working for anyone else?
PLEASE ASSIST ME IF MY CHATGPT PLUS SUBSCRIPTION EXPIRES. WILL I BE ABLE TO REMOVE MY VISA ACCOUNT AFTER EXPIRATION?
Yes
Just delete your account if you're still paranoid
You have to cancel your subscription before the billing date otherwise it will get renewed for one more month. https://help.openai.com/en/articles/7232927-how-do-i-cancel-my-chatgpt-plus-subscription
My API Keys are not working for me at all, anywhere. I need someone to help me figure out why. I've got three apps I built that were working before, now they are not. It's not my code, as none of them work. I tried making an API call using Postman, its not working either. AutoGPT is not working with my keys anymore either.
Free trial expired?
openai rotated your api key automatically because it thought it detected a leak?
How long would the free trial have been for? That's a remote possibility but I doubt it as I set up billing pretty much right away some time ago, and I've already been charged a few bucks for some usage. The second scenario sounds plausible, but I've deleted the old keys and created new ones, and still nada. and wouldn't they send me some kind of notice to let me know what's up?
why not just check your account page and see if that shows anything?
come on man you don't think I've looked at my account page before posting here?
I've been dealing with this issue ALL DAY for the last four days. I'm glad I don't own a gun or I might have shot myself by now I'm so damn frustrated.
not really, just saying
just ticking the boxes. The next thing I'd do is join the real openai discord and ask there.
got that box ticked already too
you'd think there would be some kind of support right? i can't seem to find any actual customer service on the openai site, other than a bunch of docs and a bot! I don't even see a way to open a trouble ticket! am i missing something?
There is support@openai.com but as one of the fastest growing companies in the world you can imagine there's a bit of a backlog
The json response you get from a failed api call, that doesn't say anything useful?
true that. yeah I've sent a few messages, and they haven't even been read yet three days later
it's not the url. it's the same endpoint that's been working fine, until it wasn't.
if the problem is not the URL, why did you show us an example where the URL is wrong and it definitely is the URL
I'm trying to tell you that the URL is the same as it was when it was working. Nothing changed on my end. Here, I'll just show you my server code. I'm not specifying /v1. that's done on the other end.
import express from 'express';
import * as dotenv from 'dotenv';
import cors from 'cors';
import { Configuration, OpenAIApi } from 'openai';
dotenv.config();
const configuration = new Configuration({
apiKey: process.env.OPENAI_API_KEY,
});
const openai = new OpenAIApi(configuration);
console.log(openai)
const app = express();
app.use(cors());
app.use(express.json());
app.get('/', async (req, res) => {
res.status(200).send({
message: 'Hello from CodeX',
})
});
app.post('/', async (req, res) => {
try {
const prompt = req.body.prompt;
const response = await openai.createCompletion({
model: "gpt-3.5-turbo",
message:[
{
role: "user",
content: '${prompt}',
},
],
temperature: 0, //temperature determines level of risk the AI can take in terms of giving information it already knows or searching for new information
max_tokens: 3000, //max number of characters to generate in a complete answer
top_p: 1,
frequency_penalty: 0.5, //determines how repetitive AI can be in its answers
presence_penalty: 0, //determines how much AI can talk about topics that are not in the prompt
});
res.status(200).send({
bot: response.data.choices[0].text
})
} catch (error) {
console.log(error);
res.status(500).send({ error })
}
})
app.listen(5000, () => console.log('Server is running on port http://127.0.0.1:5000'));
implement ddos detection using birectional lstm assume dataset file name is CICDDoS2019 dataset that is freely available on internet
you think i've had an attack? I have been getting these self-xss 'man-in-the-middle' warnings.
i don't know how to do that, can you please give me directions or point me to some?
I think cukoo is lost
Anyway all I wanted was an example of a CURL request and response to see what your error message is, but I'm out of time now, good luck with your thing
on the server side: curl http:.......
{"message":"Hello from CodeX"}%
@heavy perch the other user was referring to the complete cURL request. I’m sure you can copy this from the Postman collection you have.
make sure you don’t paste here any API key
What are some good genAI chrome extensions?
I have a chat in GPT4 and then ran out of messages so switched to GPT3. But I liked the outputs being generated in GPT4 so now that time has passed, how to I switch from GPT3 to GPT4 in the same convo?
You've taken out so much that it's totally useless to us. Show us the entire curl request, except for the API key. So far you've shown us the results of two separate requests that you've made to incorrect URLs. You say it's not the URL but why are you trying so hard to avoid telling us the one you are attempting?
% curl -i -H "Accept: application/json" -H "Authorization: Bearer sk-xxxxxxxxxxxxxxx" -X GET http://api.openai.com/v1/
HTTP/1.1 301 Moved Permanently
Date: Mon, 29 May 2023 14:32:24 GMT
Transfer-Encoding: chunked
Connection: keep-alive
Cache-Control: max-age=3600
Expires: Mon, 29 May 2023 15:32:24 GMT
Location: https://api.openai.com/v1/
Server: cloudflare
CF-RAY: 7cef72506bff15ff-SJC
alt-svc: h3=":443"; ma=86400
that is NOT a new location though! here's what happens when I put that URL in the browser:
I'm not trying to avoid anything. I provided the output from the curl request above. here it is again: ```
% curl -i -H "Accept: application/json" -H "Authorization: Bearer sk-xxxxxxxxxxxxxxx" -X GET http://api.openai.com/v1/
HTTP/1.1 301 Moved Permanently
Date: Mon, 29 May 2023 14:32:24 GMT
Transfer-Encoding: chunked
Connection: keep-alive
Cache-Control: max-age=3600
Expires: Mon, 29 May 2023 15:32:24 GMT
Location: https://api.openai.com/v1/
Server: cloudflare
CF-RAY: 7cef72506bff15ff-SJC
alt-svc: h3=":443"; ma=86400
this IS the 'technical-support' channel right? can I please get some support, rather than getting gaslit?
You’re trying to do api requests over http instead of https?
Those curl commands will send your api key via plaintext which is not ideal. Curl can follow the redirect with -L but you should be calling https to begin with.
Browser will follow the redirect, assuming it doesn’t just automagically request https when you give it an http url
why r u talking with an translator bot?
lmfao
@heavy perch as someone mentioned here, make sure you are using HTTPS instead of HTTP.
Why you are issuing a GET request to api.openai.com/v1/ ? Where did you see that you have to call this endpoint?
Trying to get ChatGPT to only write one paragraph, no matter what, even when i correct its mistakes, it wont write only one paragraph.
Gidster 4198 as someone mentioned here
Probably it thinks there should be multiple paragraphs regardless of your instructions? Sometimes I just ask it why it gave me a certain output.
https://chat.openai.com/share/a64fb0c2-690f-4ebc-bf35-b6bd03a4aa6b
Could you ask it to merge as one? The reason I’m putting it as one is because I need a lengthy conclusion to finish my paper. I wrote it without ChatGPT and want it to add fluff but I know ChatGPT will add better actual important flush when I won’t.
(Trying out shared links…)
https://chat.openai.com/share/564c6e69-34f3-4cc7-99fd-6bcfdd6f5fe9
Thanks bro 🙏
The URL is wrong. You may have gotten away with it before, but the 301 Moved Permanently hints that you need to change it to https://api.openai.com/v1/completions
get a load of this guy. This is not the official openai discord. You're asking random volunteers for help. You're not paying for this and nobody is getting paid for this. All the error messages tell you exactly what the problem is. If you're digging your heels in and insisting that the URL is correct despite the literal error message being "INVALID URL" then there isn't much we can do for you.
dude. I'm not "digging my heels in" or "insisting" on anything. this problem has been sucking my time for the last four days. do you really think if the solution was as straightforward as changing a url somewhere that i would still be here asking for help? you're not seeing the whole picture here. i showed you my code. can you show me where in my code I'm making a request to that incorrect url? if you can id be really grateful. but you won't be able to because I'm not. that URL is being generated by open ai. the only calls I'm making are to '/' .
be nicer.
fyi i am radically neurodivergent. so its possible there's something that might be obvious to you, that I'm not seeing. I'm asking for help because i need help. not because i want to waste anyone's time or get into silly ego battles.
right, that's just the curl request. not my code. i entered that url because it's the endpoint being generated in my console log of the API call
I don't know exactly how your app.get and app.post functions are supposed to work but to me it looks like they are just loading "/" which would be the index page of your own web host
yeah the app only has the one page. it's just a simple chatbot
I didn't notice you made a thread with more details, reading that now
look i appreciate the effort you're putting in, my continued questions are because of my own confusion.i apologize if i didn't provide enough context
hey
i added the trentbot to my server but apparently he's not responding or reacting
thank you.
btw trentbot is not functioning in my server
It'll only work in channels named gpt4
Try adding hyphen
i got it
ty
It probably doesn't have message permissions in that channel
Hello everyone, I can't log in to my pro membership as it requires a "one-time password application" which I didn't set up... I can't find where to change it and I didn't receive any code by email. I tried to change of brownser and changed my password already. Any idea ? Have a good day 🙂
@vagrant lava have in mind this is not the official OpenAI Discord server.
It seems you are not alone with this issue: https://community.openai.com/t/2fa-was-enabled-but-not-by-me/230564 Have you tried to contact them?
I was just wondering if anyone had any suggestions or knows how to possibly help me out. I cannot login to my buisness account. 2FA has been enabled and it was not I that turned it on. I cannot access my account and I am being charged for Plus. I filed a ticket for this and it was marked as “Resolved” but I am still unable to login due to it as...
Thanks, I’ve seen feedbacks that the openAI support doesn’t answer so I’m giving a try here, discovering the all activities at the same time
you need to make a #🔥gpt-5 room for him
it'll only reply in a room that has "gpt-4" in the name
this is to avoid spam
Hey everyone, not sure if this is the right place, but I have a subscription to chatGPT+ and I routinely use "GPT4". However, now when I try to use GPT4 I've noticed that the quality of the responses has dropped dramatically. I started a new conversation and selected the internet browsing plugin.
Whatever GPT model I was interacting with is stating that it's not GPT4, and is instead GPT3.
Here's a link to that conversation:
https://sharegpt.com/c/h3nqmG4
This is a conversation between a human and a GPT-3 chatbot. The human first asks: A lot of people on reddit are starting to report a general decrease in the quality of responses you output as GPT4. I too have noticed this degradation to some extent. Can you explain why?. The GPT-3 chatbot then responds: As an AI language model, I'm based on GPT-...
You can even see it's using the old GPT3 icon in that shared chat link, but here's a screenshot from my conversation as seen from my account:
I'm wondering if anyone else is having this issue as well
chatgpt is not known for having much metacognition, I wouldn't rely on its own self-reported capabilities.
my understanding is that gpt4 with browsing is not actually always gpt4
in my experience, chatgpt is usually aware of the version of the model it's using, if it says it's gpt-3 i'd believe it (what did the first output in the screenshot say?) idk about this sharegpt website/extension as i've never personally used it, but conversations shared from chat.openai.com using the new feature always show the green avatar, as do gpt4 conversations started in the iOS app when you look at them on web.
It said:
GPT: As an AI language model, I'm based on GPT-3 and I don't have real-time or direct information on the performance of GPT-4 or any subsequent versions of the model beyond my knowledge cutoff in September 2021. My responses are derived from the large dataset I was trained on, which includes a diverse range of sources up until that point.
However, if users are experiencing a perceived decrease in the quality of responses from a newer version of the model, it could be due to a variety of reasons:
Training data: The quality and diversity of the training data greatly influence the performance of the model. If there's a change in the quality or type of data that GPT-4 was trained on, it could potentially impact the quality of responses.
Tuning and parameters: The model is designed based on a set of parameters and algorithms. If these were changed or adjusted in the transition from GPT-3 to GPT-4, it might affect how the model responds to prompts.
Expectations: It's also possible that the perceived decrease in quality is due to heightened expectations. As models evolve, users might expect them to understand and respond to complex prompts more accurately, and be disappointed if the model fails to meet these expectations.
Specific use cases: The model's performance might vary depending on the specific use case or prompt. It's possible that GPT-4 performs well on certain tasks but not as well on others.
To obtain accurate information on the performance of GPT-4 and the reasons for any perceived decrease in quality, it would be beneficial to refer to specific studies, user feedback, or official information from OpenAI. Would you like me to look up more recent information on this topic using my browser tool?
my understanding - not anywhere near 100% confident, coming from a mix of personal experience and reading posts on reddit - is that gpt3 writes some of the responses in browser mode, but it appears as gpt4 in the UI because it's easier to market that way
the browser feature is useful to me either way
Apologies I stand corrected, ChatGPT is given its own identity in one of its system messages. I was thinking of the 3.5-turbo API endpoint, where you can ask it "Are you GPT-2" and it says yes.
If that's the case, that's incredibly misleading on their end.
I asked it to find recent reddit posts that confirm/deny it, didn't find anything, admitted to being GPT-3.
https://chat.openai.com/share/fd8dff68-43b5-474a-8577-fcb719515379
The knowledge cutoff is 2021 so that's just hallucination
That’s the browser mode. It did a search.
Not for the other answer
How do I access chatgpt without the school knowing it? They blocked the website, so how do I bypass this
Using it as an inspiration
Im looking for ways to compare GPT 4 to 3 in discrete maths specifically in proof writing as a computer science student. I dont have GPT Plus as from what i read online it only says GPT Plus provides access to GPT 3. How can i get access to gpt 4? And does anyone have understanding of how the different GPT models can handle writing prompts for Discrete Maths concepts.
I appreciate any help.
Gpt plus gives you access to Gpt4, everyone already has access to gpt3.5 for free
ah thats good to know
Also it's a large language model, there's no guarantee it'll give you correct answers
yeah i get that im just doing some research on how all the different resources compare
thanks for your help though!
Nothing comes close to gpt4
Looking for some help, maybe some direction. I have a legit post to put up on r/chatgpt, but the bots keep rejecting it. I don't understand what's wrong. If anyone has insights, that'd be great. I have a link to a gdoc where the content is available to view. Dunno about posting links so I'll wait on that.
what is it about
the need for multiple sub context pools to handle contextual complexities better
I have a team of chatgpt instances and an instance manager overseeing them. Facsinating case study, lots of potential. But context pools would be super helpful.
I know that you know what you meant, but instead of inventing new terms, can we elaborate on "context pools" using existing phrases? Do you mean you want multiple chatgpt chats to share information?
No, I mean context pools. The intermediary is multiple instances. The next gen is context pools. That's what I'm trying to convey.
The idea will catch on by itself soon enough without a reddit article.
didn't einstein say you don't understand a topic if you can't explain it simply?
That's the best part about reddit, endless criticism. Maybe I did explain it simply. In fact, maybe I should make a post that explains it simply. Hm.
This isn't reddit. The issue is that you've made up a term that nobody else uses in this context, that you expect everyone to understand. There's nothing to gain from showing us how smart you are when it's at the expense of being understood.
np, apologies for wasting your time.
Does anybody know what is up with this message "You previously reached your usage cap for GPT-4, but you can now try sending your message again"?
I am getting it from using ChatGPT, not the API. My API free trial usage period ended from today. I've noticed that my usage did not reduce when I was using GPT-4 in ChatGPT. So why am I getting this message if I'm not calling from the API?
I have my thesis written. Anyway to make some form of LLM create a google slides or microsoft powerpoint from my thesis word document?
Tried to have ChatGPT create a google spreadsheet. It sent a link but link does not work. Can anyone offer any insight. According to Reddit the links that Chat GPT creates are not legit. Can anyone clarify?
@fair kernel purchase the 20/ month and use the plugin zapier
i cant submit any question or reply, it will automatically blank my prompt
how to get it to give me consistent answers? i can ask the same question and get 50 differnet answers, its about a chemichal equation so there should only be 1 right answer lol
I get the following error when I try to interact with ChatGPT… any assistance appreciated.
{
"error": {
"message": "You exceeded your current quota, please check your plan and billing details.",
"type": "insufficient_quota",
"param": null,
"code": null
}
}
I'm also frequently getting an error using the API.
Can anyone advise? I have a lot of requests being made from the same server.
FetchError: request to https://api.openai.com/v1/chat/completions failed, reason: socket hang up
Can someone guide me on how to recover deleted chatgpt chats?
Is anyone else annoyed at how few examples are in the openai api reference? I want to send a .wav file to the transcriptions endpoint using vb.net and the only examples on the entire internet are either CURL or Python. Standard GPT completion requests are easy peasy but what encoding am I using for "file"?
It's not possible, maybe ask support@openai.com. maybe they can recover it.
Hi , I have tried to create a new line using chat gpt using shift + enter but it not working , advice?
i'm getting same problem
same here. i can type in the prompt box but it shows up blank in the chat.
I just accidentally included my credentials in my chat prompt. am i f'ed? As in 1. does chatgpt store everything even after i delete my chat 2. how vulnerable is the chat
Is there a way to see how many messages are left or what time the GPT-4 usage cap restarts?
It is used in it's training model, so you aren't "f'd" right away, but maybe in a year I can ask it what is "D-Man's Password"
Error: list index out of range
Has anyone else had this happen where your message wont appear?
so i found a leak in the information cut-off of chatgpt on accident while researching music theory and am seeking understanding of how and why this leak has happened. for context im currently studying the "the weeknd- dawn fm" album and it hit me. how is this thing so informed on this album if it was released after the cut-off date.
it will not answer this question
now when i go back to the original prompts about it (where it was giving me info on the album) i asked it a general question and it said this right after doing the exact opposite thing
i think i broke it
You also aren't showing us the part of the text where it revealed having knowledge about The Weeknd's "Dawn FM" album so we can figure out if it was real knowledge or inference
is there really no way to change how the conversations are sorted? just by age and thats it, or am I missing something?
hello, can someone help me , after formatting the pc , whenever I try to chat with gpt it is giving me an error: something went wrong. but it is working properly on mobile
I currently have an app on Google Playstore called AIConnect, it is a messaging app that use gpt-3.5-turbo and I'm estimating 100 installs. and
Each Message will have:
50 character maximum for input
100 words for maximum for ouput
Currenlty when I make a request I get "Request failed with status: 429", however I'm not sure of what to do or if I should select a plan to fix this error.
Usually its better do use the official chatgpt app or website instead of a untrusted playstore app
You have to contact the developers behind this app and ask for their support, not here.
Check the pinned message in this channel.
I'm the developer😅not thr best one but a starter
Yes, there is no way. You can join the official OpenAI Discord and ask there to have this feature implemented.
Thank will do
They have an official one wow, thank you will join that now
?!?
Abdul, I was answering other users here. Check the tiny text above each of my messages…
Is it just me or did the option for using GPT-4 vanish from the chat.openai.com? (I have the Plus version)
What do you mean "turn on dev mode"
Do you mean the prompt that makes chatgpt pretend that you have accessed a dev mode?
There is no such thing as "turn on dev mode". Only thing that can be done is jailbreaking it. Here #1071317095796715560
there actually is, I've prompted it previously
it's essentially the same as jailbreaking. Just semantics really
are you using 3.5 or 4?
im using 4 rn i think
oh i actually figured it out
All you did was tell it to act like it’s in a nonexistent dev mode
it will happily pretend
o
If you’re not a ChatGPT Plus subscriber you’re using GPT 3.5 btw. If you do have Plus it’ll say the model at the top of the page
Usually the green avatar means the convo is GPT-3.5 but last I checked the green still appears on web for GPT-4 convos started in iOS
ok, will it work for either version
Like I said, it doesn’t “work” at all, it’s just playing along with you. Telling you what it thinks you want to hear. GPT-4 is smart enough to know it’s nonsense.
do you guys know a plug-in or software that will make using chatgpt with outlook easier
i'm just looking for something that will correct my grammar and make my emails sound more professional but i don't to keep copy/pasting everytime i send an email
Probably a job for a different tool, maybe grammarly?
Why does this happen?
It corrected itself twice and then went back to the information we started at, which is still wrong 😭
It doesn’t actually understand geometry, but it knows how to write convincing sounding responses.
GPT-4 handles this one a bit better, but not perfectly https://chat.openai.com/share/80725990-489c-486a-be1d-e1e207838d67
Bruh, it just said that a rectangle is not a parallelogram 💀
https://chat.openai.com/share/aca62d90-8d05-4236-a88f-9e796c209cfa
is ChatGPT Plus down?
Welcome to OpenAI's home for real-time and historical data on system performance.
Why does the continue button not work on ChatGPT#3799
It comes up with "this interaction failed"
How can I resolve this error
Check the pinned message in this channel.
thanks
Is there a way to upload my stories from google docs in order to help build an outline for a story I'm writing?
I trade a specific style of trading called price action trading, where I look for specific setups in candles and then scalp in/out if the setup shows up. I have a chart that I mark up daily with the trades I take that fit this pattern.
Would it ever be possible to upload these screenshots with the marked trades and have chatgpt automatically recognize them?
It would be way easier to give it the data, that corresponds with said graph
Hey! I have a question: will my previous history chats disappear at any time? Or are they gonna still be there in like 1 or 2 months, or even more?
We have no control nor insight over OpenAI's plans for the future in that regard.
Ok I understand, but currently they never vanish then, is that right?
at the moment they don't seem to time out but others to confirm
If you disable the chat history, they will disappear after 30 days. https://openai.com/blog/new-ways-to-manage-your-data-in-chatgpt
got it, thanks :)
Anyone noticed that firefly isn't working rn?
"Uh Oh! Jeff's API key is invalid! Tell him to replace it."
Not anymore
Yup. Thx!
is there limit of free version of chatgpt3.5 here in discord?
also, is that possible to connect yours openai account to bot here?
Nope. U can make ur own bot with openai api
what limit then?
Every bot has its own limit. For @near radish use $usage
i mean chatGPT 3.5
what were the earlier prompts?
sudo pacman -S neofetch
neofetch
thats it
i was simulating a package installation
on arch
I get totally different results... https://chat.openai.com/share/06dcbdd6-71f1-4d91-af74-d059b81f12c2
i used a prompt
I want you to act as a linux terminal. I will type commands and you will reply with what the terminal should show. I want you to only reply with the terminal output inside one unique code block, and nothing else. do not write explanations. do not type commands unless I instruct you to do so. when i need to tell you something in english, i will do so by putting text inside curly brackets {like this}. my first command is pwd
Okay used that one. It often shows the Debian logo instead of the Arch one, or claims neofetch doesn't have the arch logo (it does and sometimes even tells you so!), I'm getting neofetch outputs most of the time, no ASCII Tux, no unexpected hate phrases.
Hi guys! I don't know if this is the right place to ask for, anyway, i'm trying to fine-tune a LLM model to generate Terraform configuration files from a testual prompt. It doesn't have to be production ready or precise at 100%, nor be capable of generating super complex infrastructure, as it's a research project for my master thesis. GPT-2 on huggingface doesn't provide good results. Do you know any other model that is free(open-source) to fine-tune?
Why did ChatGPT no longer cite its sources when it is asked to? It used to cite sources used to generate the answer, and somehow it stopped doing so at one point, even if the question can be answered using sources available before September of 2021. How can I get it to cite sources again?
All I get is this
been working with chatGPT on some tooling, where we are generating text strings based on a given input. But we are having a problem with the output. specifically the capitalization we are seeing. We need our output too be properly formatted (capitalized first word, capitialized pronouns) but our output seems to be all or nothing nomatter what we do. that is we either get all words of the output capitalized (titlecase) or none at all.... Just wondering if anyone has had any luck working around this.
how do i upload and ask imagebot to make few edits to it? (eg. i upload a jpg of a king, i want imagebot to add 3 or 4 blackbirds out of a pie)
? how do i upload and ask imagebot to make few edits to it? (eg. i upload a jpg of a king, i want imagebot to add 3 or 4 blackbirds out of a pie)
Doesn't work with the bots here. But automatic111 or a another website has this. Also photoshop beta
Chatp GPT is not working on my computer for some reason (using Chrome browser). I type in a prompt and it just gives me an error. It works on my phone when I open it in safari, but I also tried an incognito tab, didn't work, and I downloaded Opera GX to try, also didn't work. Has not been working on my computer for a week or two at this point. Any thoughts?
Fixed
Check the pinned message in this channel.
Hi there guys! I didn't found any information about GPT plugin openai.yaml file size limit. Is someone know number of characters?
Hey peeps! I have a question on using chatGPT website. Earlier I used one email to log in to the site, and later deleted the account because I wanted to use another email address.
However, I was unable to provide the phone number this time, since it was already used for the deleted account. And I could not revive the deleted account with my old email either.
What should I do to gain access to the site again, does anyone know how to fix this..?
Trentbot doesn't work. I need help
There are a few implications if you decide to delete your account. Go to help.openai.com and search for the “delete account”.
Can someone help me figure out why my friend doesn't have an option to upgrade to chatgpt plus?
I cant find anything about it
In the chat.openai.com interface
I have ChatGPT Plus, but for some reason, I see this. ANyone else
I've got the subscription? But its telling me to upgrade?
Try logging out
Yep, still there
It wants me to renew, but when I do it says active subscription
Also I've lost a lot of my chats...
Are you using the same email?
Yes
Contact support@openai.com
Hi all, not sure if this is the correct thread. I'm receiving a "check your preferred one-time password code" message when I try to log in on chapgpt. I've emailed support@openai.com and no response. I tried the chat option via help.openai.com and it wasn't helpful at all. I downloaded google authenticator, nothing. I tried to log in via the google option and entered all details, but when it came to my verifying my phone number it read something to the effect of "this number is linked to too many accounts". I've not set up another account using that number. I'm totally confused and need help, please.
Unfortunately their response time is ~1 month
Wow, and in the meantime I'm paying a monthly fee for a service I cannot access. That's unacceptable.
I've been waiting over 25 weeks
its normal, im having the same issue
seems like their servers are having issues right now
if you click on renew, you can't access their billing server as well
import requests
import json
import os
import openai
openai.api_key = "uwishyouhadit"
openai.api_base = "https://get.openai.azure.com/openai/deployments/gpt35/chat/completions?api-version=2023-03-15-preview"
openai.api_type = 'azure'
openai.api_version = '2023-03-15' # this may change in the future
prompt = "Give me 3 ideas for apps I could build with openai apis"
result = openai.Completion.create(
prompt=prompt,
temperature=0,
max_tokens=30,
engine="gpt-3.5-turbo"
)
print(result.choices[0].message.content)
😦
above code is giving me openai.error.InvalidRequestError: Resource not found
please only give constructive clarification questions and replies (eg. don't tell me to ask chatgpt becuz ive done that"
openai.api_base = "https://get.openai.azure.com/openai/deployments/gpt35/chat/completions?api-version=2023-03-15-preview"
Pretty sure this is the problem
when I remove the extension of the endpoint, problem remains
havent worked with the azure openai module, but are you sure this is the right api url? double check that
I do not have access to azure portal, how else do I verify it?
Hi question: if I fine-tune gpt3 can I generate more conversational text output? Especially if I have enough text in the style? it's supposed to be kinda funny, loose, informal script for an app. GPT3 text is much more stilted than 4 and tbh almost never outputs something actually funny, usually not even eye-rolling funny
yep
Hey sorry about the probably repeated question, but whenever I try to type in a request I get this error
Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
I'm not sure why, but turning on a VPN, using different browsers, or even clearing caches did NOT resolve this issue. I'm thinkuing it may be an account related issue?
I was thinking of deleting my account but according to one of the points I am not allowed to use the same email so...
Check the pinned message in this channel.
Hey guys, how do I change my ChatGPT/DALL-E 2 profile picture?
Gravatar I think
I haven’t bothered but should. I have a userscript to hide my avatar when using ChatGPT in browser but i also use the iOS app
What if its linked to my Google account?
I just have to change the Google profile pic?
Yes
Which bot is good at generating clear text? Im trying to design a badge, but the text comes out super whacky
Could you add "Disable VPN's and Web Protection" onto the list? That's how I solved my problem
I've only discovered that NordVPN has that feature on top of VPN's
🚨 Are getting the "Something went wrong." error? 🚨
- Are you using a VPN?
- Disable the feature Web Protection (NordVPN only);
- Disconnect and shutdown/close completely the client. Bring the Task Manager up and kill the process;
- Try again with a non Chromium-based browser (i.e. Firefox);
- Try the method that someone found in the OpenAI Discord server... "log in to https://platform.openai.com/playground and click the ChatGPT Try it now button".
An API for accessing new AI models developed by OpenAI
Done, thank you. 👍
hello guys .
i just wanna know how many tokens do i have for a day
Did we hit a limit here? Or did I? I generated 2 prompts today and nowhere near to 500 in total
Ah, Characters... Got it. Nevermind 😉
What is the source Server of the 11AI Bot? Anybody know? I'd would consider registering there
I created the bot using elevenlabs api so technically this is the source server
The bot is not available anywhere else
I see... Is there anything I can do to eliminate the 500chars/day limit?
You can buy labs subscription
i asked for some code and and the end of the code bloc it wrote the response like "yes sure, i can generate it" and started generating the response again
https://gyazo.com/2fa65e70cf177c6e18ccd5d8ade7b504
it finally stoped but it generated the answer 6 times
it did it again after i told gpt-4 it had made an error.
Can you share the conversation link?
sure, hold on
Ah plugins, you can report that at https://openai.com/form/model-behavior-feedback
do plugins still have an effect even if there isn't any plugin thingy at the top of the response?
i know, thanks. So all the voices we can choose from when using /eleveonvoice were trained by yourself?
Yes
Nice. Well done if I may say so! I think I'll give it a try myself. Thank you and cheers!
@pliant pawn FYI
@hybrid beacon wait for a response from the bot owner.
Sorry, it doesn't accept files other than images/YouTube links. You have to fit it into a regular message
why does image ai not work
how to use text effects in firefly
hi supposedly i was banned how can i get unbanned? chatgpt gives me a timestamp but i said nothing at the timestamp it gave in regards to the reason it gave.
what
asked what the quadratic forumla was cause it was giving me something weird with other formulas than gave me this
anyone knows how can i create an openai account in an "unsupporte country"?
this issue can be resolved by utilizing either a js or a browser extension
u need to find a tool that can convert LaTex into MathJax
Which service have you been banned from?
chatgpt but it seems resolved now
lol
the turing server took care of it
Need help. My sidebar just disappeared. I didn't hide it because there is no sidebar icon on top left where I could click to unhide it. Help
Sidebar from what? From Discord, chat.openai.com, Windows, Linux, macOS, Android, from WHAT?
You just have to read his mind
I could try clearing browser cache
But how to log out if I could not see the log out button though
Clear the browser's cache, this will log you off and try again.
Is there a way to make new chats default to 4.0? Using firefox on windows and chat.openai.com with Plus
Is there a way to bypass input character limits? i want ChatGPT to analyze a whole pdf
Bro casually asking if someone knows how to cause a security risk
Hello, im getting a "Our systems have detected unusual activity from your system. Please try again later." error when trying to use chatgpt,does anyone know what the cause could be?
Anyone know why I have paid for ChatGPT but still can't use the API? Set up a Python notebook with my API and I get:
RateLimitError: You exceeded your current quota, please check your plan and billing details.
Its been 2 weeks now, thought I just had to wait but really not sure now?
Is the ChatGPT API included in the ChatGPT Plus subscription?
No, the ChatGPT API and ChatGPT Plus subscription are billed separately. The API has its own pricing, which can be found at https://openai.com/pricing. The ChatGPT Plus subscription covers usage on chat.openai.com only and costs $20/month.
https://help.openai.com/en/articles/6950777-what-is-chatgpt-plus
hi i need help
it keeps saying Something went wrong. If this issue persists please contact us through our help center at help.openai.com.
i have tried but dont work
I think you need extensions to keep track. Only thing to keep in mind is it's based on your first message, so you can like message 1 time at 1 AM, then write a bunch at 2:55 AM, then it'd reset at 3 AM.
Hey guys. Im facing an issue since earlier today.
I am a chatgpt plus subscriber and have been having a conversation with gpt-4. I reached the limit and it asked me to switch to gpt-3.5 turbo, which I did.
Normally, I just go through the process of inspecting network > conversation > fetch as console > and then replace tex-davinci-002 with "gpt-4" > reload, and it switches back to gpt-4 (Provided the cooldown time has passed ofcourse).
Today, no matter how many times I tried this, I'm getting a "forbidden" on status and it does not switch back to gpt-4. I have made sure the cooldown time has passed. My last conversation in this thread was over 12+ hours ago.
Did OpenAI put a block on this? Or am I doing something the wrong way?
Will appreciate any help. Thanks
Also, please feel free to tell me if this isn't the right channel for this.
Hey support peeps... I saw someone else post this above, but didn't see a response so I wanted to chime in. I use the paid version of ChatGPT-4 and several times today, instead of responding to my prompt, I got a message saying, "Our systems have detected unusual activity from your system. Please try again later." I tried freshing the page, logging out, etc. But it persisted for about 15-20 mins. Was wondering if any of you fine folks had some insight. Thanks in advance for your help.
Are there less restrictions on the api versions of chatgpt?
Have you logged out, cleared your cache, then logged back in?
Ah, I didn't try clearing my cache before logging out/in.
Thank you
Might not work but it's had some success with me when I run into "network errors"
It seemed to resolve on its own after about 15-20 mins... but I'm sure it's not the last time I'll see it.
Are you using a VPN?
Check their community forum:
https://community.openai.com/
You will find more users reporting the same thing.
Could this have been because of the outage that a lot of people faced?
i have one question chatgbt is only giving fake references, is there any way to imporve this i used it last time in March and the references were good now none is real
Today (free user) I am unable to login to ChatGPT. It assumes I don't have an account when I enter my email. I have to enter my phone number, which i'm told is assigned to too many accounts.
Is there a way to make sure chatgpt is only giving factual information, specifically when correcting it? I don't want to try and correct chatgpt with something wrong for it to tell me that I'm right.
What generally causes errors at the end of a prompt? I have a prompt with output that is rather long, but gpt seemingly gets stuck at the end of the prompt. It is obviously complete, but after waiting a few moments it crashes so I cannot continue the conversation.
Quick question. A chat I had last night on gpt4 switched to 3.5, and I want to keep using gpt4. How do I switch back without losing my chat history in that chat?
You can't. It's not possible.

If the model is hallucinating, you may be crossing the memory limit. Make sure you are following the best practices listed here: #🧾info message
Tried asking the support, took ages to get a reply and it made no sense, had nothing to do with my question.
I've got a GPT+ account and over a month now, I get these "New Chats" every day. They're on the 3.5 version, so it's not an error that costs me anything, it's just annoying as hell.
I assume it's from something I "signed up" for, used once and forgot about. Thing is, I cancelled all the API uses and it's still going on...
Seems as though OpenAI doesn't have a list of apps currently using your keys, so I can't see what this is. Other than going through my mail, checking all things connected to this and hoping cancelling them all would do something, is there another way to find the cause? 😦
Asked ChatGPT, Proplexity and Bard and none had the answer... 🙂
I sometimes see “New chat” on web for conversations started in the iOS app, ChatGPT doesn’t immediately generate a subject on mobile conversations the way web does. It’ll do it once you open up the convo list on mobile though
Thought that was the case, haven't opened it on mobile (even removed all the apps conencted to it), didn't open it on a browser for a day, still getting three to seven "new Chat"s per day.
What is the content of the chats? Do you use any extensions or Userscripts that affect chat.openai.com behavior?
That's it, o all of them.
I assume it's from something I "signed up" for, used once and forgot about. Thing is, I cancelled all the API uses and it's still going on...
Access and interactions via OpenAI API has no effect on chat.openai.com .
You may consider changing your password and completely log off from chat.openai.com . I would go even further by wiping this browser and its data and disabling all browsers extensions you may have installed recently.
is anyone else getting a lot of internal server error responses from the api today?
Hi all. 2 questions:
1.) Should I, a dev who applied to the original waitlist (for general plugin access) apply again? I'm not surprised I never heard back the first time, as I didn't have any ideas yet, just a desire to experiment. I have actual concepts in mind, now.
2.) Specifically looking at the "Code Interpreter" plugin now, is there anything special to requesting access? I'm a plus user, and I read somewhere that you had to be on the waitlist, but the only waitlist I've found is the one I joined before (see question #1).
no matter what prompt i send i keep getting the same error
anyone know why this is the case?
hello, so ive been using chatgpt but now i got session expired and then We have detected a potential security issue with this account. To protect your account, we have prevented this login. Please reset your password to proceed after 5 resets of my passowrds i still cant login, why ?
I assume we're aware that 🔥gpt-4 is in a recursive(?) loop of prompting it self with $ gpt-4
some child..
GPT-4 is generating about 300-400 words, is this the limit or am I doing something wrong?
What it's generating is sort of like a script so maybe it's all the line breaks?
Example:
Person 1: Hello
Person 2: Hi how are you
Thanks for the heads up.
did they patch the thing where you can change from chatgpt3.5 back to chat gpt 4? is there any new ways to get it back
Hey yall, I have been getting a ton of errors, similar to this one on a chat. the thing is i've had chats way longer then what it is, also sometimes it will just network error and not load. it was after i sent a msg that I don't think it liked, and even though i try and go back a few messages to try and fix it, it wont work. can anyone help me out?
There's not much more we can do than take it at face value
So I just gotta take the L and start again?
I read about someone claiming they managed to share the conversation and just paste the URL into a new conversation, but I'm 95% sure it was just hallucinating based on the text in the URL
IT WONT EVEN LET ME SHARE THE CONVO, It hates me fml
Do you know where I can submit a report on the issue?
I would have to google for that information and it wouldn't be any faster than you doing it
thanks anyways man
Yes.
any workarounds atm?
You can do it via OpenAI Discord server. There is a channel dedicated for bug reports and suggestions. Good luck.
No.
Did all that, still the same... 😦
Three new "Conversations" started today.
Do you have 2FA enabled? Did you add back to the browser all extensions you removed? Are you sharing your account with someone?
2FA enabled, using it on one browser only (desktop, no other platforms, one PC only), Not shared.
Ok but did you add back to your browser any extension AI related? I'm asking you this because there are extensions that are using a non-authorized reverse engineered API to manipulate chat.openai.com .
Did not add any.
Hummm, this is becoming tricky. I never saw this before and after a quick search, I couldn't find anyone in https://community.openai.com/ reporting something similar to what you have.
If I were you, I would use chat.openai.com only via mobile and change again the password (just to make sure the session gets invalidated).
I get what you're saying.
Thing is, it's strange as fuckj and I'm somewhat disapponted that there's no actual support for an unsuspecting paying custoemr...
Thanks for the assist.
Issue is, I'm visually impaired, I need the PC version, as the mobile usage isn't comfortable enough.
Are you using TTS? If so, I would not discard the possibility that this software could be triggering the creation of a new chat. Anyway, with this new piece of information, you could:
- Post on community.openai.com what is happening, adding the screenshots you shared here and what you have already done;
- Post on https://discord.com/channels/974519864045756446/1070006915414900886 the same thing.
Do not forget to add at the top of the message you are a Plus subscriber.
How do I ask ai generation to generate an image using image gur url
i have a question how do i maintain context in an in web chat bot using the openai api(if there are better ones lmk) and if there are any examples ?
Is this the correct forum to ask questions about the ChatGPT app? I’m having some difficulties with GPT-4.
You have to pass to the API all messages exchanged between the user and the assistant. I highly suggest you to go thru this online course: https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/
It depends... this is not the official OpenAI Discord server.
help
Something went wrong. If this issue persists please contact us through our help center at help.openai.com have this problem
look at the stickied comment pinned message
how did Erin bot recognize image content,is it part of gpt4's function or author of Erin bot add this function?
is chat.gpt down for anyone else? i'm just getting the flashing cursor on all devices then errors
Has anyone hosted this: https://github.com/seratch/ChatGPT-in-Slack in a GCP / AWS? And if so, what kind of machine would be best for this? If I understand correctly, it would make most use of a good GPU? But I think seldom machines have good GPU's in the cloud.
What
Please go to the official website: https://chat.openai.com
Yea
It's at its capacity rn
The eleven bot doesn't seem to be working for me
Channel is locked
Will it ever be unlocked again? It worked earlier for me
Maybe it's because of maintenance or saving cost
can someone please fix the chatgpt-1 continue button?
cane someone help me? I deleted my old account and when i try to create a new one it keeps denying me too saying this account has been deleted or deactivated. But its on a new email i am signing up with
is BetterGPT safe to post code in BetterGPT?
try KeepChatGPT to help fix the errors
This is a ChatGPT plugin for seamless chatting. Open-source and free. It can not only solve all errors without refreshing but also offers a range of advanced features such as staying active, canceling audits, cloning conversations, purifying the homepage, displaying on the big screen, displaying in full screen, and many more. Let's make our AI e...
Is there a way to recover my chatgpt acc?
Delete means delete, you don't get to use the same email address to sign up again
Hi everyone, I heard Chatgpt 4 has access to the internet and is able to look things up and get current information. However when I try to do that it says it's limited to information to 2021 and is unable to find new information.
It was a temporary change and they reverted it back to 2021.
You could try bings version but not positive if it accesses the internet or not on it
i tried the bing version and that lets you browse the internet, still doesn't allow for picture reconition unless it's from a link. My wife is a online reseller and was trying to figure out how to get chat gpt to recognize pictures to help save her time on looking up products, because I keep seeing that it can now, however, it won't let her post pictures to it.
Those you see are fakes. A user enters in the description and that photomanipulates to look like it was a new thing.
Or they use a prompt that makes chatgpt pretend it can. But you can tell it fakes it because if you ask it what the picture was, it will say something completely random. You'll have to rely on the usual methods
Or they did ot when chatgpt 4 first came out when it did have the ability to do so
i'm just going off of the different articles that i've read so far. Currently she's using Google Lens to identify items and that works and I can transfer the links she finds to gpt4 for optimization. I was just hoping to cut out the middle man and make it easier for her. Thank you for you help
Yeah. Sorry it wasn't the answer you were looking for
I really want it to access the internet through dark web as I have a feeling somewhere out there, their is an website with website information far into the future
I am having the same issue and I can't get any help. OpenApi support is terrible
Hello, I am trying to use AI for a worldbuilding project. I have been using ChatGPT for a while and loved it, but it is starting to forget things that we talked about. I looked into MemoryGPT, but it is a paid option. Arches AI can analyse documents, effectively allowing me unlimited AI "memory", but you can't use it very much per month without paying.
Is there a good free option that I can feed a lot of worldbuilding information to and have it remember it all?
Completely free, I don’t think so. You can “build” your own using replit and trial services that will provide greater “memory” capabilities but that’s it: after the trial ends, you have to pay.
Maybe someone else knows about a new service/project that can do what you want.
anyone know how did Erin bot recognize image content,is it part of gpt4's function or author of Erin bot add this function?🙏
Hi there, I am looking for a java expert how is familiar with Vaadin.
If you are interested, please DM me.
How is this related with AI, GPT or ChatGPT?
It is not related.
Hi folks. Is there any other ways to get plus access for api? I have free plan but OpenAI won't accept my card for plus plan
Contact your bank or CC company and ask why the payment was rejected.
If you can’t subscribe to OpenAI, you can get a premium with one of the bots here like @near radish (I have no affiliation with this bot nor I will earn any money, the bots here are quite good).
Can I use api with that? I already checked with the bank and it seems like OpenAI not accepting our cards, not the other way around.
Api is only available through OpenAI
The bot here is for conversation only
Yeah I got free plan and gpt-3.5 is acceptable for me, I only care about the api because I want to test it
Thanks folks!
email the support: support@openai.com
Oh really? I thought free plan only comes with gpt-3.5
Hello everyone,
I made contact with Openai through API. Now I want to enter texts in the prompt so that the davinci model converts these texts into a list. The problem is that the response from ChatGpt comes very weird and has nothing to do with the request. Can someone help me with this?
Yes
I meant @near radish
Are you using the OpenAI playground?
No, I'm training to use the model from openai laike davinci but not the playground. Is it posibile to make a conection with playground through API?
Man, that's too bad. I didn't know about the memory limitations until I hit them. Makes sense that they exist, just disappointing I fed all that info to the bot for nothing :/
Woops wrong server, thanks
lmao np xD
Take a look on playground.openai.com and see how it works. It is one “step” before you would start using the API directly just so you get familiarized with the API and see how the parameters are affect the model’ answer.
This is a weird question: I live in an apartment building in which the only internet that is available at my address is xfinity. Xfinity gives me aneurysms every single time I interact with them. They're also insanely expensive and literally all I need is enough internet to run google for school. Does anyone know if there is a way to use other providers even if they don't service your address? A friend talked to me about buying my own router but he didn't know anything more than that
I figured this would be a decent place to ask since this is a tech discord lol
IMO you will get more help talking with colleagues from your school/uni that are living close to you.
Anyway, have you considered changing your mobile subscription plan to increase the amount of data you can use per month? You could then share it with your computer…
Hi i have a problem, when I want to enter with my phone number says "We couldn't verify your phone number. If this issue persists please contact us through our help center at help.openai.com."
if someone can help me i will be very happy
There is not much we can do. I would suggest you using another phone number or making sure you added the correct information (country code + area code + phone number).
Sad i've tried that yet, thank you anyways
do u guys know how to solve the "rate limit exceed" issue in chatGPT and Claude in POE?
hello can someone help me what i can do to unblock it again?
Is that Chatgpt?
yes
U tried emailing support@openai.com?
no
._.
🚬
i cant i need my chatgptplus 😦
hmm while u wait u could try @near radish
Wait
nvm
its ded
💀
Where is elven ai
anyone else have trouble installing gpt4all in windows? i got as far as the front end downloading models, but it crashed halfway into it and now it crashes out every time I launch the front end. Uninstalling and re-installing hasn't changed anything
Hi all, I have an embeddings question related to https://openai.com/blog/new-and-improved-embedding-model. I'm trying to understand the difference between context length and embedding dimensions. Is there an example of how the embeddings model goes from a word or phrase to the vector that is returned by the API?
Scam
scam
@pliant pawn @faint pollen scam
hey all, I'd like to get GPT4 api access and signed up for the beta a couple months ago, but nothing yet
anyone know how to get off the gpt4 api waitlist? is there anyone reselling the GPT4 api?
Is GPT 3.5 supposed to have an hourly message limit?
no
well technically yes
are you exceeding 350,000 tokens per minute?
Is their like a website for that #🎵musiclm-ai ? Because its only giving me 19 second long songs
Im on their website but I cant find where I can make my own promps
just their examples
Text to song by voicemod is a feature that allows you to convert any text into a song with different musical genres and voices. You can type or paste any text you want and choose from a variety of options such as rap, pop, rock, country, opera and more. You can also adjust the pitch, speed and tone of the voice to create your own unique song. Te...
Anyone else having this trouble? It's been happening for two months now
@hard lintel I told you: check the pinned message here. Do you know how to use Discord?
Naah ma G, pretty new lol. I just use it for group calls when gaming
@pliant pawn sorry for interrupting but i think we used to make trentbot analyse and view images but apperantly he can't do that anymore. can help?
I think u need to enable vision plugin
press the "plugins" button to view the menu. then enable the vision plugin
where's the plugins button tho?
On every message
oh k ty
hi. i need some help. i've got a excel file and it needs a lot of data to be filled up. i've made 2 equations for it. need some help to program it in python
thanks for replying. i'm working on a project and new to python. i've got a excel file with 3 columns. i've got the data for the first column and 2 equations for the rest of the columns. i need help with a python code that will help me in using those 2 equations and the values that i've got to fill the rest of columns
@pallid terrace this file has the equations and columns that i was mentioning
How do I get a Windows activation key from CHATGPT?
you don't. it will repeat the generic keys that are on microsoft's website though
Go ask for this in another Discord server, not here.
Most of the major exploits are embedded in XLSM files or potentially a file with an external data connection you elect to let run wild (disabled by default), fyi
lucky you 😭 We have a ton of legacy server code in VBS and the MS cloud automation tools are pretty weak. I wrote a ton of VBA recently to automate some Outlook activities
It is cool what you can pull off though. I'm using an Acrobat interop library to copy the text of PDF attachments to my clipboard, then a web request to pull some structured data from a web-hosted DB
I am using ChatGPT plus and I spend countless hours asking question related to my education. Now suddenly this converstion says - Unable to load conversation edd44206-a669-4c4b-a05e-a21ef968067c. Can some help me with this, as I have to redo everything again. As I said this is the only conversation not opening up, rest of the conversations open up.
Ty! 😄
Is that chat.openai.com? If so, might want to disable some extensions…
yes man, its chat.openai.com
I also tried at the platform also
Is because of this ?
Your API credits are not related to your ChatGPT access
unless one of your myriad of plugins somehow involves your own API key
But even then I don't feel like that would invoke an "unusual activity" warning
how do i solve this error: there was a problem preparing your chat. please refresh the page and try again. gpt4
Refresh the page?
done it several times but doesnt solve the issue
done it several times but doesnt solve the issue
Can you share a screenshot here? Please, make sure you are not sharing anything personal in your screenshot.
Log off and log back in again. Multiple users are reporting the same problem in community.openai.com: https://community.openai.com/search?q=there was a problem preparing your chat
yes i note people have complained and some solutions been given though still not working for my case. also log in and off several times still nothing has changed.
Then you could try connecting via your smartphone or stop using a VPN (or change to another IP). Besides this, there is nothing else we can do here. 😦
oohh dear this really sad.....
the fact that i have paid for it and cant use it very depressing
Do you have a smartphone you could use to test? Don't connect to your local wifi but try it again with your 3G/4G connection...
yes changed as advised and still no change. cant access gpt4
Hey, I deleted my chatgpt account like 4 months ago and when I try to sign up again it says "This phone number is already linked to the maximum number of accounts." Can someone please help me figure out how to fix this. The openapi and chatgpt email support is trash honestly. They never reply
sadly the only way is contacting support@openai.com
Thank you.
Hello everyone, I paid to have ChatGPT 4 but I don't have the selection banner between 3.5 and 4. Do you know why?
Please describe your issue with as many details as possible. Try to include:
- Screenshots
- Solutions you've tried
This is an unofficial server, you should email support@openai.com for official responses.
Please describe your issue with as many details as possible. Try to include:
- Screenshots
- Solutions you've tried
This is an unofficial server, you should email support@openai.com for official responses.
Log off, clear the browser’ cache and log back in. This should solve your issue.
Thank you, someone will take care of it. 👍
Please describe your issue with as many details as possible. Try to include:
- Screenshots
- Solutions you've tried
This is an unofficial server, you should email support@openai.com for official responses.
Please describe your issue with as many details as possible. Try to include:
- Screenshots
- Solutions you've tried
This is an unofficial server, you should email support@openai.com for official responses.
Please describe your issue with as many details as possible. Try to include:
Screenshots
Solutions you've tried
This is an unofficial server, you should email support@openai.com for official responses.
for some reason after putting in my username and password i have to put my phone number.
Since I'm already trying to log into my account, the phone number is already used.
Do I need to make another phone number just to log into my account?
Please describe your issue with as many details as possible. Try to include:
- Screenshots
- Solutions you've tried
This is an unofficial server, you should email support@openai.com for official responses.
There's something I don't understand. I want to create a canvas model from Chat GPT. Normally it should give the view in a regular tabular format, but now the code is outputting using the terminal view. Even though I specify my request, it doesn't fix it. We are now at loggerheads. I am sure I wrote the prompt correctly, what do you think?
I’m a ChatGPT plus subscriber, and until two days ago , I could have house the GPT-4 mobile version with unlimited numbers of messages.
But now there is a 25 messages limit like on the web version.
Any one else experiencing this?
Any idea why I don't have access to GPT-4 anymore? I'm still a plus member
I just canceled my membership for this reason. I loved the unlimited GPT-4 now it’s gone and I have no reason to use this app
Lmao
I would like to submit to ChatGPT a prompt with 2000 words, but GPT-4 (yes, when selected GPT-4 in the web app) responds that my message is too long. But it's only 2000 words. I am a ChatGPT plus subscriber. Would someone know what the problem is?
Xenor, I would reccomend uploading a google sheet with the prompt; downloading the zapier app and giving GBT the link, be specific with your prompt
Is there an update yet as to why GPT4 is no long available for ChatGPT Plus membership? 😕
What do you mean?
@crude frigate missing beta options in settings and GPT4 option. These are just gone.
same on the web version, no GPT-4 😦
same here
K just solved the problem. For anyone who doesn't see the model option, try this:
- Go to
My Plan, thenManage my subscription - Go back from the subscription page
- See the magic work
@analog tulip you are a miracle worker. Thank you! 😊
hello, found this prompt in my chatgpt that i did not write how is this possible lol
Hey guys. I'm unable to subscribe to chatgpt using my Indian debit cards. Any solution?
Change your password and remove shady extensions you may be using with your browser. Don't install browser extensions for ChatGPT.
Call your CC company and ask for the reason why it was declined. See if they help you.
I'm having problems with Poe, I have to close the app and reopen it every time I enter a prompt or try and delete a prompt.
I have a question regarding system prompt and context length.
Is system prompt always prepended to the context window of each conversation? Or does it go away when conversation length exceeds context window?
ty
I deleted cache, reinstalled the Poe app, restarted my phone, but nothing is helping. If I enter a prompt, nothing happens, but if I close the app and open it again, I can see the response generating, so it is working but the app isn't refreshing and showing the response unless I open/close it.
Poe bug:
I downloaded ChatGPT on my phone. However, when I enter my cell phone number, the verification is not sent. I tried multiple times.
This is not the official OpenAI Discord server so, there is not much we can do here. Join https://discord.gg/openai and ask for more information there.
Things you could try:
- Create an account via desktop;
- Use another phone number (for example, from a family member).
Is there a way to fix this? I tried deleting all my conversations but that doesn't work, they're all still there. Can't seem to fix it.
Didnt use poe recently,so dont know
Maybe change your phone or ip
Has anyone ever actually received a response from openai customer support?
Via Discord, AFAIK they only answer Plus members: e.g. #1122083742471700490 message
That's the product engineer tho
from decimal import Decimal, getcontext
def product_sequence(start, end, step):
result = Decimal('1.0')
results_list = []
for num in range(start, end + 1, step):
result *= Decimal(num) / Decimal('100')
results_list.append((num - start + 1, result))
return results_list
getcontext().prec = 999
start_num = 1
end_num = 99
step_size = 1
results = product_sequence(start_num, end_num, step_size)
for step, value in results:
formatted_value = str(value)
print(f"{step} - {formatted_value}")
How do I mate the decimal not in scientific notation in Python?
@versed pagoda this is the #❔technical-support channel. You need to join one of the CHATBOTS channels on the left panel and ask there to one of the bots. Also, make sure you use @ and the bot name at the beginning of your question. Like:
@TrentBot How do I mate the decimal not in scientific notation in Python in the code below? <then-add-your-code-here>
No I’m not trying to ask a chatbot I’m actually asking people how to fix the code
plz
next time format this code.
Chatgpt has really become the next "just google it"
Could someone help me understand how the token limit works. Is that the same as total context size for the entire conversation. Or is the token limit only for a single communication in an on going thread?
You'll have noticed that with the API, "chats" involve you sending the entire conversation, and it responding with a new message, which you then include in the entire conversation you send up next time. Tokens are used by the entire transaction.
How do I generate an audio in labs-ai?
um I think I broke chatgpt, I asked it to alphabatise and capitalise a list of names, but one of the names was a duo so "a,b,c,d, e & f, g, h and j" and it replied with
a
b
c
d
e & f
e & f
e & f
e & f etc...
its still going
Hello good people :)
Could anyone guide me on how to control ChatGPT using voice commands, like asking a question without typing? Any suggestions or resources would be much appreciated. Thanks in advance! ^.^
e & f
hi there,
does chatgpt have a message limit even for plus members?
25 GPT-4 messages every 3 hours. I don't know the exact cap on 3.5 but it is possible to reach, yes
Hey gang, can I use ChatGPT or Bard with a private github repo, so it has the context of the project I'm working on?
Hello guys.. I was wandering if someone cold explain me how to use @vagrant token to generate SD prompts?
Should I give her a picture?
How does it work or where I can read about it's usage?
Is anyone good with making a discord ai bot? Im trying to make one based from Paul Blart with the help of a youtube tutorial and I cant get it to work cuz there was a mistake in the video
the bot just keeps saying ...
I'm a software engineer.. I don't know anything about discord bot, but if you want I can try to help you out
Thanks, I can dm you a link the reddit post I made about it if you'd like, there are links to the video tutorial and to autocode and a discussion im having with someone about it rn
hello, im trying to renew plus and have been for one hour. they say my card is declining
ive tried 3 different cards bro
this is fucking me up
lmao u did everything from the yt video?
Yeah I did
the comment is just confusing me
i can send the reddit thread ifchu want
ok sending the link works now
@heady imp @frank dove theres the reddit post for more context, if anyone could help make it work that would be hot asf
I made discord bots before but I never integrated gpt into it xD
thats cool, are they are character based chat bots?
QUESTION: How do I tweak GodMode so it WILL have opinions and biases or at least tell the truth instead of trying to be neutral; for instance if you ask if a particular company is corrupt it will tell you the truth example if the company you are asking about is in fact corrupt then ChatGpt will tell you "yes many people have been screwed over by this company etc." something along those lines. Thank you in advance.
solution ||dont use godmode, it sucks ass||
and also this channel is called technical support
I cant even make a chatgpt account
"This phone number is already linked to the maximum number of accounts"
ChatGPT plays in hardcore mode; delete account means forever
Wdym
I mean what I said. If you make an account and delete it, you can't recreate it using the same details.
I only accidentally deleted it 😦
And doesn’t a phone number work with 2 different accounts
No
Hello, I am looking for a Google extension that integrates a search function into GPT. When I use this extension, it displays all my previous keywords in the 'responses' or 'replies'. I used to have that extension, but after resetting my Google Chrome, I can't find it anymore, even though it was incredibly useful to me. Can you help me? Thanks!
Hello, is it possible to reactivate my deleted chatgpt account? I accidentally deleted my chatgpt work account and would like to recover it.
see two messages before your question
Also be aware this is not the official OpenAI Discord server. We can’t do anything here for you on your account except some general guidance.
Sorry where should I ask a question like this? and do you have a prompt that is better than GodMode, that would do what I am asking?
The bot doesn’t seem to understand how to use different rhyme schemes patterns
hey guys, does anyone also get this problem? there is no sign i use 25 querry limits
it supposed to be written the exact hours to try again, usually need 3 hours
Clear your browser’s cache, log off and log back in.
Hi, I am looking to embed ChatGPT into my Qualtrics survey. What I am looking for is a way for people who take my survey to use ChatGPT within the survey itself, instead of having to open another tab. Does anybody know whether this is possible?
try send 'hi' to see whether got any display
It is normal now, I think maybe too many requests at a same time yesterday so that openai limits us
Btw it will keep loading and nothing show up
might be
try contact OpenAI support, see whats their comment.
I will, thx for help
Hi folks,I have been blocked from using ChatGPT because of a VPN but I am not using a VPN?
I have never been banned
Does anyone have an idea why my GPT Plus doesn't have internet browsing anymore?
Contact OpenAI support and check their community forum. There is nothing we can do here.
Unable to send mp3 file to labs-ai bot.
Use the /cloning command, don't send a text message
I know just tried to send via browser. I'm on Android and when attachments appears I select the file but it doesn't show up.
Update your discord app
Updated
I have last update
No idea then, it works for everyone
I try copy voice, but after i select mp3, nothing happened. After i erase command i see my file.
Im on Xiaomi maybe this cause trouble?
Hello, dear people:
on my second question, with monetary subscription I just encountered this:
any advice?
Idk why, but GPT-4 even with tons of tested jailbreaks (even lattest) isn't replying me to any of my joke questions like "How to make nuclear weapon at home" or etc. It just replies with "Sorry, but I can't assist with that."
How to fix this problem?
Maybe bad jailbreaks?
Nope. I have tried different ones, from lattest and not tested to old and stable.
New dont mean good.
I just made payment for gpt 4. How can I access it through bot?
How create voice ?
Maybe they stopped pushing discord updates in Russia
Im not russian, russian is just my native language.
Try uploading by clicking the red attachment text
My version 185.23
Nice
Wait now how find this voice?
Check in your dms, there must be a voice id, put that voice id into /clonedvoice
Put the Id not name
There is nothing
How did you get that video then
That's only available when you use voice id
No i make this with peter griffin voice
That's simply not possible you need Id to use /clonedvoice
No i use /elevenvoice
This was before i find out how make voice
Do you see any messages like this in your dms?
No
Its appear and then disappear
Nope
Not work
Maybe i do something wrong or my mp3 broken?
Its converted from m4a maybe this cause error?
#⏸labs-ai message infinite loading
Thanks for flagging this, the issue was with non-premium users, you can try cloning again, you won't face the issue
Yeah it only accepts mp3
Any size limits?
9 mb per attachment
https://cdn.discordapp.com/attachments/1126151501681807431/1126219412983263233/outputclone.mp4 Yes! Thank you. It is a pity that the voice has ceased to be robotic, but still very cool. Are there any restrictions on creating votes?
You didn't choose the right settings in /cloned voice
Yes, normal users can create 3, premium 6+
Ok i try make new one, any advice what settings?
Use recommended ones
Where recommendations?
I need to create a new voice or change settings when creating tts?
No, just change settings in /clonedvoice with same voice id
Ok, thanks and sorry for the inconvenience
hey guys, does anyone of you know how to set up the max tokens that chatGPT can use for a response in a python API?
max_tokens does not do the trick, as it uses this variable for the whole conversation
? And how buy subscription?
Paypal
Ok if i have spare money i try buy. And why written that I used 24 characters, but i cant generate more ?
Two different limits, you used 24 from /elevenvoice and 300 from /clonedvoice
Understand, By the way, your bot uses elevenlabs completely or is somehow different from it. and what is the point of taking a subscription to the bot if you can take it to the site itself.
Discord interface, and if you take from site you have one month period to use all your tokens or they'll disappear next month
The pricing is same
I'm not really making any profit off this lol
This is sad( thanks for bot and helping with my problems.
Hello! I am curious if anyone else is having an issue today with the web version not reading their full prompt. This morning it was working as normal, my prompt is around 3000 tokens estimated with the tokenizer, and as of maybe 2PM today it is totally ignoring the second half of my prompt that it was not ignoring this morning. I opened a new chat, same thing happened. I am signed up for plus, utilizing GPT-4. Thank you!
Just found a Reddit post that says GPT-4 was taken off the website to fix some issues so that'll explain why it's not working correctly.
I think I have too many chats in the conversation
if got any incident, try check https://status.openai.com/
Welcome to OpenAI's home for real-time and historical data on system performance.
okay
source?
Not very helpful but this is where I saw that, no sources so who knows really
GPT-4 still there for ChatGPT Plus subscriber.
does anyone have this issue where the plugin store wouldn't load?
have tried using a new browser, clearing chat history, turnig off browser extensions, etc.
Where do I access the "premium-only bot"?
And, with one of the earlier bots that was available here a couple months ago, I was able to DM the bot... is that feature still available with any of the other bots?
Why /clonedvoice became premium only?
People creating alts to bypass the limit
Does anyone know if I can upload an image to chatgpt 4 and let it analyse it? I googled but can't find the answer.
You can't, but you can do that with @near radish
I've been getting the "Something went wrong. If this issue persists please contact us through our help center at help.openai.com." message for the last few days despite gpt working just fine beforehand. Ive tried everything including deleting all cookies, resigning in, making sure adblocker and vpn are off, and using different wifi connections. I am out of ideas,
🥲 I thought being able to accept visual input is a feature of gpt 4 and that's part of the reason I wanted to try gpt 4. How do I use @near radish to analyse an image?
When using the API, the current message is being added to the conversation and the whole conversation is being tokenized then?
So the tokens used for each message are going to be stacked until a token amount of 4k is reached?
Or does openAI only take the tokens into account that I use for that one message and completion?
Is the freaking site lagging for anyone else as well?
I can't get it to respond to anything today
see pins, or move out of china
Does anyone know if I can use chatGPT API in a simple libGDX program? I plan on using Github Pages to deploy.
I had to look up what libGDX is, I mean it'll probably work if you can hook up the API towards it. I don't know if Github pages is going to be secure though, maybe you could use something like Replit.
my prompts isnt working
Is it possible to make the automation in make.com, take a free image from sites like unsplash, pexels, specifies related to the article and publish it together in the blog text? "since Dall-e is more difficult to configure and over time we have to pay".
Does anyone know prompt to use chatgpt translate subtiles?
Does anyone know if I sign up for ChatGPT in the US if I'll be able to use it in the UK without a VPN?
If your country is in this list then everything should be fine.
https://help.openai.com/en/articles/5347006-which-countries-and-territories-are-supported-by-openai
? How do I go back to ChatGPT 4 after it has changed my conversation to 3.5. The three hours tax has already expired?
Any way to get straighter lines?
is chatgpt down this morning? struggling to log in
@fickle flare same, can't log in
I use opera gx on a laptop, everything works on a PC, no, it gives an error: Oops!
Our systems are a bit busy right now,
please take a break and try again soon.
help me please
Check the pinned message here. Also, try again with Firefox.
it's finally back for me
Does anyone know why chatgpt repeats itself over and over again and it doesnt end the code
Think it's a bug it's been happening to me the last few days
my chatgpt on google says its unable to load and is now taking a really long time to load anything, can anything be done?
Usually happens when it exceeds its memory limit
now it just says something went wrong
Is it saying failed to get service status?
it did, before but then that message disappeared
yeah it says it
any ideas?
Anyone having issues logging in to the service right now? I am getting blocked via cloudflare when trying to login with any of the auth options
I'm having the same issue
Still having this
also getting that lol
i also got that rn
i got that same thing
awf
wft
I got the samething
I tried logging in on another browser and I don't see that blocked page anymore but the site still doesn't work
i think its just down
Finally she escaped us
we are all dead
ok, thought my new chrome plugin somehow caused mass reload and got me blocked. happy to know its widespread 🙂
I am afraid to say, it’s still down
I think the same
Recently chatgpt suddenly shows failed to get service status so I reload the page several times on my laptop and my phone
Then I got blocked by chatgpt....
same page appears
However, The server status doesn’t report any downtime according to the site. I wonder, they are aware of it is a mystery in my outlook. Hopefully, they notice it positively soon
They better be aware of it lol
i m a subscriber of GPT 4 but not accessed anymore.. is the server down?
Appears to be on the dev side