What is StableLM?
StableLM is an open source AI language model released by Stability AI which is trained to predict and generate subsequent words in text based on preceding context, learned patterns and relationships within the training data.
Language models will form the backbone of our digital economy, and we want everyone to have a voice in their design. Models like StableLM demonstrate our commitment to AI technology that is transparent, accessible, and supportive
What StableLM models are being released?
We are releasing two StableLM-Alpha models at 3B & 7B parameters:
StableLM-Base-Alpha - Trained on the new pile-v2 dataset, a new dataset which contains 1.5 trillion tokens, roughly 3x the size of the older pile-v1 dataset.
StableLM-Alpaca-Alpha - Fine-tuned the model using Stanford Alpaca's procedure using a combination of Stanford's Alpaca, Nomic-AI's gpt4all, and RyokoAI's ShareGPT52K datasets.
Despite its modest size of 3 to 7 billion parameters, StableLM displays remarkable performance in conversational and coding tasks, due to the richness of the underlying dataset.
How do I use StableLM?
StableLM is open sourced! You can find the checkpoints on our Github:
GH:
https://github.com/stability-AI/stableLM/
What license is StableLM under?
StableLM Base models are released under CC BY-SA-4.0.
https://creativecommons.org/licenses/by-sa/4.0
🔥🔥 will anyone try it today?
Couldn't make it work, the space was way overcrowded
Official blog post here