Language Modeling on nanochat (val)
11.42Perplexity (val)Parcae
Evaluation Results
| Method | Links | |
|---|---|---|
| ParcaeNumber of Parameters=1.3B, T=82026.04 | 11.42 | |
| TransformerNumber of Parameters=1.3B, T=-2026.04 | 11.95 | |
| ParcaeNumber of Parameters=770M, T=82026.04 | 12.49 | |
| TransformerNumber of Parameters=770M, T=-2026.04 | 13.08 | |
| ParcaeNumber of Parameters=370M, T=82026.04 | 14.49 | |
| TransformerNumber of Parameters=370M, T=-2026.04 | 15.79 | |
| ParcaeNumber of Parameters=140M, T=82026.04 | 19.06 | |
| TransformerNumber of Parameters=140M, T=-2026.04 | 21.48 |