Fine-Tuning-Qwen3-to-generate-new-Russian-Poems Fine tuning Qwen3 on a dataset including Russian Poems Dataset can be downloaded from here https://huggingface.co/datasets/Georgii/russianPoetry Qwen3 model with 4B parameters chosen and trained for 1 epoch. LoRA rank 128 and re-trained 264M parameters. the training took 1:30 minutes on GPU 4080 Super. you can see the sample of output below:

В тихом свете солнца, в сумраке ночи Тишина. И сладостно мне, как снится, Что уж вновь я в детстве, и что вдруг вижу Тихий мир, где нет ни забот, ни сечи, И в лесу соловьи пели на лугах, И, как в грезах, сонные звери бегут, И на бреге морском ветер свистит, И луна на небе горит.

If you dont want to train the model then just download the Poem_russian (included LoRA and tokenizer).

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Tr3z0r/Fine_Tuned_Qwen3_on_Russian_poems

Base model

Qwen/Qwen3-4B-Base
Finetuned
Qwen/Qwen3-4B
Finetuned
(300)
this model

Dataset used to train Tr3z0r/Fine_Tuned_Qwen3_on_Russian_poems