About the role
The Speculative Decoding team develops algorithms and models to accelerate large language model generation. The research intern investigates new approaches to speculative decoding, including model architectures and training methods. This involves training models on GPU clusters and evaluating their performance under realistic workloads.
Our summary of the posting; the original is on the Together AI careers page.
What the posting requires
- Required
- Machine Learning
- Distributed training
- Mixture-of-Experts
- Deep Learning
- Megatron-LM
- Hybrid attention
- FSDP
- Master
- PhD
- Computer Science
- Preferred
- Knowledge distillation
- LLM serving
- Open-source
- Speculative decoding
- Train
- Serving engines
These tags are our reading of the posting. They can miss something; the original posting is the reference.