Databricks' dolly-v2-12b, an instruction-following large language model trained on the Databricks machine learning platform that is licensed for commercial use. Based on pythia-12b, Dolly is trained on ~15k instruction/response fine tuning records databricks-dolly-15k generated by Databricks employees in capability domains from the InstructGPT paper, including brainstorming, classification, closed QA, generation, information extraction, open QA and summarization. dolly-v2-12b is not a state-of-the-art model, but does exhibit surprisingly high quality instruction following behavior not characteristic of the foundation model on which it is based.
For more information -> https://huggingface.co/databricks/dolly-v2-12b
This deployment рассчитан на использоваание on GPUs NVIDIA V100, A100 and H100. After launch container, the application should download the trained model from the project repository with total weight of 24Gb and it may take some time depending on the internet speed of GPU provider.
Unfortunately, I could not achieve a normal display of logs. They do not display an indicator that the trained application models are loading, and they do not display requests to the server. But when manually launched locally, the logs are displayed. Therefore, in this case, in order to find out that the application is ready to work, you will have to constantly update the application page and wait for the interface to appear.
Now the logs look like on this screenshot: