Model Card for DenseFormer Mistral-7B-v0.1
This is a DenseFormer implementation of Mistral-7B-v0.1. The details about DenseFormer are in the paper.
You will need to use trust_remote_code=True to load this model.
This model WILL require additional pretraining to see the improvements indicated by the DenseFormer paper. Per the paper:
Once deciding to go to the valley where DWA weights are not used, it is difficult to recover and ultimately benefit from newly added DWA weights. We believe this phenomenon could be mitigated using a better learning rate scheduler. We leave this investigation as future work.
The Mistral-7B-v0.1 Large Language Model (LLM) is a pretrained generative text model with 7 billion parameters. Mistral-7B-v0.1 outperforms Llama 2 13B on all benchmarks we tested.
For full details of this model please read our paper and release blog post.
Model Architecture
Mistral-7B-v0.1 is a transformer model, with the following architecture choices:
- Grouped-Query Attention
- Sliding-Window Attention
- Byte-fallback BPE tokenizer
Troubleshooting
- If you see the following error:
KeyError: 'mistral'
- Or:
NotImplementedError: Cannot copy out of meta tensor; no data!
Ensure you are utilizing a stable version of Transformers, 4.34.0 or newer.
Notice
Mistral 7B is a pretrained base model and therefore does not have any moderation mechanisms.
The Mistral AI Team
Albert Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix, William El Sayed.