🔥 prog-love/gemma3-270m-beast
Fine-tuned google/gemma-3-270m-it with Unsloth on Single GPU.
Training
| Datasets | Glaive-Code 60k · MetaMathQA 60k · ShareGPT 3k |
| LoRA rank / alpha | 32 / 64 |
| Optimizer | AdamW (torch) |
| Final loss | 2.4420 |
Fine-tuned google/gemma-3-270m-it with Unsloth on Single GPU.
| Datasets | Glaive-Code 60k · MetaMathQA 60k · ShareGPT 3k |
| LoRA rank / alpha | 32 / 64 |
| Optimizer | AdamW (torch) |
| Final loss | 2.4420 |