Skip to content

Conversation

@xrsrke
Copy link
Contributor

@xrsrke xrsrke commented Jun 11, 2025

Run nanotron's moe implementation

CUDA_DEVICE_MAX_CONNECTIONS=1 torchrun --nproc_per_node=8 run_train.py --config-file  /fsx/phuc/new_workspace/snippets/experiment_configs/qwen_moe/exp20b1_3A30B_qwen_moe_and_num_experts_128_and_topk12_nanotron_imple.yaml

Run nanotron's TE implementation

/fsx/phuc/new_workspace/snippets/experiment_configs/qwen_moe/exp20b0_3A30B_qwen_moe_and_num_experts_128_and_topk12_te_imple.yaml

@xrsrke xrsrke requested review from NouamaneTazi and zzhhjjj June 11, 2025 13:12
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

7 participants