# Model training problem

**URL:** <https://discuss.huggingface.co/t/model-training-problem/105635>\
**Category:** Beginners\
**Created:** [September 6, 2024, 1:46pm UTC](https://discuss.huggingface.co/t/model-training-problem/105635 "2024-09-06T13:46:04Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Yasirpbs](https://avatars.discourse-cdn.com/v4/letter/y/e99b99/32.png) [@Yasirpbs](https://discuss.huggingface.co/u/Yasirpbs)\
**Post date:** [September 6, 2024, 1:46pm UTC](https://discuss.huggingface.co/t/model-training-problem/105635/1 "2024-09-06T13:46:04Z")

</div>

Hi community  
I have problem with training of t5small model. I have trained the model with 21 SQL queries and respective text in natural language. But I tested the trained model, it gives me incorrect results, even select and from statements were missing in so called generated query

---

<div class="post-metadata">

**Author:** ![mahmutc](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/mahmutc/32/52583_2.png) [@mahmutc](https://discuss.huggingface.co/u/mahmutc)\
**Post date:** [September 6, 2024, 7:13pm UTC](https://discuss.huggingface.co/t/model-training-problem/105635/2 "2024-09-06T19:13:47Z")

</div>

hi @Yasirpbs  
Don’t you think 21 examples are not enough to fine-tune the model?

You might check [Fine tuning a T5 small model to generate SQL from natural language with 92.3% accuracy | by Martin Keywood | Medium](https://medium.com/@martinkeywood/fine-tuning-a-t5-small-model-to-generate-sql-from-natural-language-with-92-3-accuracy-fb29e062c638)  
or [mrm8488/t5-base-finetuned-wikiSQL · Hugging Face](https://huggingface.co/mrm8488/t5-base-finetuned-wikiSQL) to see more details.