Loading…
Loading…
Gemma3-SWA mimarisiyle eğitilmiş 1256M parametreli dil modeli.
| Parametre | 1,256,312,832 |
| Katman | 34 |
| Hidden | 1792 |
| Head (Q/KV) | 28 / 4 |
| Context | 8192 |
| Vocab | 65,536 |
| Eğitim token | 9.339 B |
| En iyi val loss | 2.8047 |
from transformers import AutoModelForCausalLM, AutoTokenizer
tok = AutoTokenizer.from_pretrained("bcckfdn/cevher3-test-5")
m = AutoModelForCausalLM.from_pretrained("bcckfdn/cevher3-test-5")