agile slate
data: '<html>\r\n' +
'<head><title>502 Bad Gateway</title></head>\r\n' +
'<body>\r\n' +
'<center><h1>502 Bad Gateway</h1></center>\r\n' +
'<hr><center>nginx</center>\r\n' +
'</body>\r\n' +
'</html>\r\n'
I hope the API will be stable in the near future, I have a project in production that relies heavily on this API 😦
warm herald
```
data: '<html>\r\n' +
'<head><title>502 Bad Gateway</title></head>\r\n'...
Are you still running into this issue?
static jay
same error to after some request 502 gateway error
agile slate
Are you still running into this issue?
I just wrote an error handler in my code that can handle this, I will rerun the code in production and tell you soon if I have caught 502 errors again
agile slate
Can confirm, there's still 502 errors sometime ...
upper mirage
Can confirm this is still happening
still wolf
Can confirm too. API takes too many time to answer, and sometimes it returns that 502 error. For me it is not reliable at all, at least 1/4 of my requests are not answered.
upper mirage
Same here, on average about 1/4 or the requests encounter this issue. I have reported the bug to OpenAI support
grand shard
I get 502 and 500 all the time too. I wonder if there's anything more stable than OpenAI 
agile slate
Yea, if they can't scale their server when there are big demands, it won't be reliable for production, it's really a shame
I hope that other companies will get into this field soon, OpenAI has almost a monopoly for chat completion of this quality
static jay
yeah i hope google
half beacon
Been getting 502 errors for 2 days straight too
upper mirage
I have a feeling this has only been happening since the outage yesterday, is that the same for other too?
agile slate
I have a feeling this has only been happening since the outage yesterday, is tha...
Indeed since yesterday also
agile slate
Erros again and again last night but with this code this time:
hasty nebula
data: {
error: {
message: 'The server had an error while processing your request. Sorry about that!',
type: 'server_error',
param: null,
code: null
}
}
EDIT: This time this is a 429 error for "Too Many Request", i perform only one request every 30 seconds
agile slate
Everything seems to be alright for status bar 😦
I can imagine that the devs must be working hard, I really hope so, but I would like to have a more concrete feedback on this situation. I can imagine that the devs must be working hard, I really hope so, but I would like to have a more concrete feedback on this situation
upbeat idol
openai.error.RateLimitError: The server had an error while processing your request. Sorry about that!
agile slate
For me it seems to be better since this morning, no error for the moment
For gpt-3.5
static jay
For me it seems to be better since this morning, no error for the moment
what is your rpm mine is TPM: 2181 RPM: 22 getting 429 errors
agile slate
what is your rpm mine is TPM: 2181 RPM: 22 getting 429 errors
Where do you get theses numbers? But i think for me it's less of yours. I have read somewhere that the API limit users to 60rpm
static jay
Where do you get theses numbers? But i think for me it's less of yours. I have r...
im calculating it self
startcheck = Date.now();
tokens = tokens.filter(token => (Date.now() - token.time) <= 60000);
response = await openai.createChatCompletion(prompt);
msg = response.data.choices[0].message.content
endcheck = Date.now();
console.log(`Execution time: ${endcheck - startcheck} ms`);
tokens.push({tokens:response.data.usage.total_tokens,time:Date.now()})
const totalTokens = tokens.reduce((total, token) => total + token.tokens, 0);
console.log('TPM: ' + totalTokens+' RPM: '+tokens.length);