• 4
• 2
• 1

qianggee/FaultInjection

fault injection for pytorch by torch_distpatch

• 7
• 5
• 4

xiangsen2/KJ600

TensorProbe (code name: kj600) is a LLM pretrain debugger with model's torch module , optimizer status, collective communication tensor collection and aggregation. It also supports rule-based alerts.

• 1
• 0
• 4

xue-yuqing/KJ600 forked from xiangsen2/KJ600

TensorProbe (code name: kj600) is a LLM pretrain debugger with model's torch module , optimizer status, collective communication tensor collection and aggregation. It also supports rule-based alerts.

• 1
• 0
• 0

xue-yuqing/image1