| datasets: | |
| - wikimedia/wikipedia | |
| - nthngdy/oscar-small | |
| language: | |
| - pl | |
| base_model: | |
| - distilbert/distilgpt2 | |
| license: apache-2.0 | |
| distilgpt2 with new tokenizer, trained from scratch with polish datasets. | |
| Needs more training, however it's able to generate correct polish sentences. |