如何只仿真像nccl-test/perftest这样的非模型数据？
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 1.2k
- Forks
- 184
- Avg merge
- 13h 4m
- Merged PRs (30d)
- 1
Description
按照我的理解是不是可以构建特殊的workload文件，格式如下：
MICRO model_parallel_NPU_group: 8 ep: 1 pp: 1 ga: 2048 all_gpus: 16 checkpoints: 0 checkpoint_initiates: 0
30
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
embedding_layer -1 0 ALLREDUCE 8388608 1 NONE 0 0 NONE 0 100
以上是不是就是相当于用16卡做nclltest的all-reduce测试 通行size是8MB 迭代次数是30
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
No repository file, test, or entry point is named. Start by locating the workload-file parser and any existing MICRO workload examples, then compare this proposed configuration with the simulator's handling of nccl-test/perftest-style traffic. Done means documenting whether it represents a 16-GPU, 8 MB all-reduce with 30 iterations and explaining any required format changes.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- distributed-systems
- Issue type
- Documentation
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100