| coreference-resolution-on-winograd-schema | GPT-2-XL 1.5B | #28 | Accuracy: 70.7 |
| dialogue-state-tracking-on-simmc2-0 | GPT-2 | #2 | Act F1: 94.5Slot F1: 81.7 |
| document-summarization-on-cnn-daily-mail | GPT-2 | #23 | ROUGE-1: 29.34ROUGE-2: 8.27ROUGE-L: 26.58 |
| language-modelling-on-enwiki8 | GPT-2 (48 layers, h=1600) | #1 | Bit per Character (BPC): 0.93Number of params: 1542M |
| language-modelling-on-lambada | GPT-2 1.5B (Zero Shot) | #23 | Accuracy: 63.24Perplexity: 8.63 |
| language-modelling-on-one-billion-word | GPT-2 | #22 | PPL: 42.16Number of params: 1.54B |
| language-modelling-on-penn-treebank-word | GPT-2 | #3 | Test perplexity: 35.76Params: 1542M |
| language-modelling-on-text8 | GPT-2 | #1 | Bit per Character (BPC): 0.98Number of params: 1542M |
| language-modelling-on-wikitext-103 | GPT-2 Full | #25 | Test perplexity: 17.48Number of params: 1542M |
| language-modelling-on-wikitext-103 | GPT-2 Large | #45 | Test perplexity: 22.05Number of params: 774M |
| language-modelling-on-wikitext-103 | GPT-2 Medium | #61 | Test perplexity: 26.37Number of params: 355M |
| language-modelling-on-wikitext-103 | GPT-2 Small | #72 | Test perplexity: 37.50Number of params: 124M |
| language-modelling-on-wikitext-2 | GPT-2 | #6 | Test perplexity: 18.34Number of params: 1542M |
| language-modelling-on-wikitext-2 | GPT-2 (large) | #7 | Test perplexity: 19.93Number of params: 762M |
| language-modelling-on-wikitext-2 | GPT-2 (medium) | #8 | Test perplexity: 22.76Number of params: 345M |
| language-modelling-on-wikitext-2 | GPT-2 (small) | #9 | Test perplexity: 29.41Number of params: 117M |
| question-answering-on-fever | Zero-shot | #7 | EM: 50 |
| question-answering-on-webquestions | Zero-shot | #9 | EM: 43 |
| response-generation-on-simmc2-0 | GPT-2 | #3 | BLEU: 19.2 |
| sentiment-analysis-on-imdb | GPT-2 Finetuned | #29 | Accuracy: 92.36 |
| text-generation-on-openwebtext | GPT2-124M | #2 | eval_loss: 3.12 |