| Age | Commit message (Collapse) | Author | Files | Lines | |
|---|---|---|---|---|---|
| 2022-02-16 | Add a new communism scenario. | Rose Hogenson | 1 | -17/+28 | |
| 2022-02-15 | Add a doctor scenario. | Rose Hogenson | 1 | -0/+12 | |
| 2022-02-15 | Use a better sentence detection strategy. | Rose Hogenson | 1 | -8/+14 | |
| It's actually pretty easy to see whether a snippet is a complete sentence, just append an extra word and see if it tokenizes into two sentences or just one. This should be a lot more reliable. | |||||
| 2022-02-15 | Wrap the prologue. | Rose Hogenson | 1 | -2/+2 | |
| Sometimes the AI generates really long lines. | |||||
| 2022-02-15 | Use gpt2 by default. | Rose Hogenson | 1 | -1/+1 | |
| I'm actually getting way better results from gpt2. Now gpt2 is my new best friend! | |||||
| 2022-02-15 | Increase the temperature to 0.9. | Rose Hogenson | 1 | -1/+1 | |
| 2022-02-15 | Use real terminal width instead of assuming 72. | Rose Hogenson | 1 | -3/+5 | |
| 2022-02-15 | Auto-download nltk packages if they're missing. | Rose Hogenson | 1 | -1/+6 | |
| 2022-02-15 | Use an nltk tokenizer to trim sentence output. | Rose Hogenson | 1 | -7/+8 | |
| It kind of feels bad to depend on nltk and hugging face, but we're doing the best we can. Too bad that nltk isn't easier to install, too. | |||||
| 2022-02-15 | Use a naive regexp to strip extra output. | Rose Hogenson | 1 | -7/+10 | |
| The sentence tokenizer wasn't correct because not all sentences end in periods. The current regexp doesn't handle punctuation inside words such as apostrophes or abbreviations, but we're trying it. Maybe a tokenizer would be better? | |||||
| 2022-02-14 | Format with black. | Rose Hogenson | 1 | -60/+57 | |
| 2022-02-14 | Lint the code and clean up a bit. | Rose Hogenson | 1 | -0/+146 | |
 |
index : gpt-adventure | |
| This is a clone of AI dungeon |
| summaryrefslogtreecommitdiffstats |
path: root/ai_dungeon.py