I encountered a problem while creating a model fine-tuning task: when the output of a certain data sample is too long, it prompts an error, indicating that it cannot exceed 5000 characters. Additionally, when I use the fine-tuned model for text generation, the output stops at a maximum of 1024 tokens. I wonder if these limitations are because I am using a free project? Would these restrictions still apply if I had a paid account?
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
How to overcome gemini 1.0 fine tuning limit in Google AI Studio? | 1 | 73 | September 23, 2024 | |
Error: Tuned model creation failed | 3 | 467 | September 5, 2024 | |
New tuned model is not working? | 5 | 220 | June 24, 2024 | |
Size limit on tuning data? | 1 | 103 | May 27, 2024 | |
Tuning doesn't work even with Gemini Pro 1.0 for datasets containing outputs with more than 5000 characters | 0 | 34 | October 3, 2024 |