← Back to browse
Editor score 57
Tools
gpt-neox
High-performance, model-parallel autoregressive transformers on GPUs; implementation based on Megatron-LM and DeepSpeed for large language models.
Why use it
High-performance, model-parallel autoregressive transformers on GPUs; implementation based on Megatron-LM and DeepSpeed for large language models.
Best for
beginner
Deployment options
compose · docker
Resource requirements
server
Alternative to
No mapping yet
Common setup stack
Reverse proxy · HTTPS certs · auth gateway · backups