Search for a command to run...
Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning