DataBricks Dolly 2.0, Commercial Use, TRUE Open Source LLM

DataBricks Dolly 2.0, Commercial Use, TRUE Open Source LLM

Databricks has introduced a new language model called Dolly 2.0, which is an LLM with 12 billion parameters. The model is based on the EleutherAI Pythia model family and has been fine-tuned using a new high-quality dataset of human-generated instructions, which was crowdsourced among Databricks employees. This fine-tuning was done exclusively with an open-source dataset created by Databricks. Additionally, Databricks has made available the PyTorch code implementation of Dolly 2.0 on …

code colab commercial databricks dataset dolly dolly 2.0 eleutherai employees family fine-tuning free generated human implementation language language model llm notebook open source open source llm pythia pytorch quality true