A LLM trained only on data from certain time periods to reduce modern bias
Python
Update README.md
Hayk Grigorian4e991f54cb9d
diff --git a/README.md b/README.mdindex 1a65f88..ad6e6a3 100644--- a/README.md+++ b/README.md@@ -53,7 +53,7 @@ I started training on a 435MB (108 M tokens) corpus, it's going pretty smooth ri# V0 Model Behavior & Limitations-Early prompts show the model responding with 1800's language and behavior. For example, I prompted it with "Who art Henry?" and it replied "I know that man, I have did not a black, the storm." and yeah that sentences makes no sense but the LLM is recognizing I'm asking about a person. I should have around 1-2GB of data+Early prompts show the model responding with 1800's language and behavior. For example, I prompted it with "Who art Henry?" and it replied "I know that man, I have did not a black, the storm." and yeah that sentence makes no sense but the LLM is recognizing I'm asking about a person.