DeepSpeed
Training Setup
Inference Setup
Training API
Inference API
Model Checkpointing
Affine Shard Transfers
Activation Checkpointing
ZeRO
AutoEP (Automatic Expert Parallelism)
Mixture of Experts (DeepSpeed MoE)
GLM-5.2 (
glm_moe_dsa
)
Transformer Kernels
Pipeline Parallelism
Optimizers
Learning Rate Schedulers
Flops Profiler
Autotuning
Memory Requirements
Host Memory Pinning
Monitoring
DeepSpeed
Search
Please activate JavaScript to enable the search functionality.