Language Modelling On Openwebtext
평가 지표
eval_loss
eval_perplexity
parameters
평가 결과
이 벤치마크에서 각 모델의 성능 결과
모델 이름 | eval_loss | eval_perplexity | parameters | Paper Title | Repository |
|---|---|---|---|---|---|
| GPT2-Tropical | 2.92 | 18.64 | 124M | Polynomial, trigonometric, and tropical activations | |
| GPT2-GELU | 2.95 | 19.24 | 124M | Polynomial, trigonometric, and tropical activations | |
| GPT2-Fourier | 2.93 | 18.72 | 124M | Polynomial, trigonometric, and tropical activations | |
| GPT2-Hermite | 2.91 | 18.39 | 124M | Polynomial, trigonometric, and tropical activations |
0 of 4 row(s) selected.