I'm seeking a freelancer to assist with the implementation of a small scale natural language processing (NLP) pretraining experiment. We want to experiment with pretraining SLMs on clustered data.
Required skills:
- Transformers and language models
- MOE architecture, including upcycling from dense models
- SLM pretraining, using commmon frameworks
- large text dataset clustering
- Huggingface
Your task will be to guide and implement the chosen experiment professionally and effectively. I look forward to your bids.