Skip to content

A LLM trained only on data from certain time periods to reduce modern bias

Python

Update README.md

Hayk Grigorian4a9cae9da48b
diff --git a/README.md b/README.md
index 7f55759..93c2b55 100644
--- a/README.md
+++ b/README.md
@@ -213,10 +213,10 @@ GPU: H100 SXM rented
## Related Work
- **[Violet 1.4B](https://huggingface.co/zakarth/violet-1b4-chat)**
- — 1.4B parameter LLM trained on 1800-1899 texts with Q&A. There is also a [160M](https://huggingface.co/zakarth/violet-160m-chat) version.
+ - 1.4B parameter LLM trained on 1800-1899 texts with Q&A. There is also a [160M](https://huggingface.co/zakarth/violet-160m-chat) version.
- **[Ranke-4B](https://github.com/DGoettlich/history-llms/blob/main/ranke-4b/prerelease_notes.md)**
- - a family of 4B parameter LLMs based on the Qwen3 architecture trained from scratch on 80B tokens of historical data up to knowledge-cutoffs 1913,1929,1933,1939,1946
+ - a family of 4B parameter LLMs based on the Qwen3 architecture trained from scratch on 80B tokens of historical data up to knowledge-cutoffs 1913,1929,1933,1939,1946.
- **[MondadGPT](https://huggingface.co/Pclanglais/MonadGPT)**
- MonadGPT is a finetune of Mistral-Hermes 2 on 11,000 early modern texts in English, French and Latin, mostly coming from EEBO and Gallica.