#Finetune free tokens
1 messages · Page 1 of 1 (latest)
Hi @heady axle. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
The free tokens for fine-tuning are refreshed based on UTC time. The exact time is not specified, but it's likely that the refresh occurs at 00:00 UTC each day. However, please note that this is an assumption and the exact refresh time may vary.
For more details, you can refer to the official response (https://community.openai.com/t/923195) on the OpenAI community forum.
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
@vital seal kapa.ai is still learning and improving, please let us know how it did by reacting below
cheers geez
@drowsy osprey I get this error "Creating this fine-tuning job would exceed your hard limit, please check your plan and billing details. Cost of job ftjob-xyYH7ovex9Et97M5HnBSgOia: USD 24.31. Quota remaining for org-zuEQKM5uAG6RvW7Kt8Oqyei6: USD 20.08."
Does this mean i havent recieved my tokens yet
Let me check the knowledge base for you. One moment! :books:
The error message you're seeing is related to the cost of the fine-tuning job exceeding your available quota, not the availability of tokens. Each fine-tuning job has a cost associated with it, which is deducted from your available quota. If the cost of the job exceeds your remaining quota, you will not be able to start the job.
In your case, the cost of the job is USD 24.31, but your remaining quota is only USD 20.08. This is why you're seeing the error message. You will need to either reduce the cost of your fine-tuning job or increase your quota (by upgrading your plan or waiting for the next billing cycle) to proceed.
For more information, you can refer to the official response (https://community.openai.com/t/928194) on the OpenAI community forum.
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
@vital seal kapa.ai is still learning and improving, please let us know how it did by reacting below
@drowsy osprey yeah but shouldnt this matter because i get 1mil tokens a day and this is less than that 1 mil
Let me check the knowledge base for you. One moment! :books:
The free tokens you receive for fine-tuning are separate from your billing quota. The free tokens are a daily allowance for fine-tuning, while your billing quota is the total amount of money you can spend on OpenAI services.
The error message you're seeing is related to your billing quota, not your free tokens. Even if you have not used all your free tokens, you can still exceed your billing quota if the cost of your fine-tuning jobs (plus any other usage of OpenAI services) exceeds the quota.
In your case, the cost of the fine-tuning job is USD 24.31, but your remaining quota is only USD 20.08. This means that you don't have enough quota to cover the cost of the job, even if you have free tokens left.
If you want to proceed with the job, you will need to either reduce the cost of the job (by using fewer tokens or a cheaper model) or increase your quota (by upgrading your plan or waiting for the next billing cycle).
For more information, you can refer to the official response (https://community.openai.com/t/928194) on the OpenAI community forum.
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
@vital seal kapa.ai is still learning and improving, please let us know how it did by reacting below
@drowsy osprey i have more tokens than the job needs tho
Let me check the knowledge base for you. One moment! :books:
The issue you're encountering is not about the number of tokens, but rather the cost associated with the fine-tuning job. Even though you have enough tokens, the cost of the job (USD 24.31) is exceeding your remaining quota (USD 20.08).
The cost of a fine-tuning job is calculated based on the number of tokens and the cost per token, which varies depending on the model you're using. Even if you're within your daily free token limit, if the cost of the job exceeds your remaining quota, you won't be able to start the job.
To resolve this, you can either:
- Reduce the cost of your job by using fewer tokens or a cheaper model.
- Increase your quota by upgrading your plan or waiting for the next billing cycle.