Skip to content
Back to pull request #651

parallelise model loading #466

Sign in to view logs

parallelise model loading

parallelise model loading #466

Run time

Learn about OS pricing on GitHub Actions
Job Run time
single-m4-pro (llama-3.2-1b) / generate-matrix
1s
single-m4-pro (llama-3.2-3b) / generate-matrix
1s
single-m4-pro (llama-3.1-8b) / generate-matrix
1s
three-m4-pro-cluster (llama-3.2-1b) / generate-matrix
1s
three-m4-pro-cluster (llama-3.1-8b) / generate-matrix
1s
three-m4-pro-cluster (llama-3.2-3b) / generate-matrix
1s
three-m4-pro-cluster (llama-3.3-70b) / generate-matrix
1s
two-m4-pro-cluster (llama-3.1-8b) / generate-matrix
1s
two-m4-pro-cluster (llama-3.2-1b) / generate-matrix
1s
two-m4-pro-cluster (llama-3.2-3b) / generate-matrix
1s
single-m4-pro (llama-3.2-3b) / run-distributed-job (M4PRO_GPU16_24GB)
56s
single-m4-pro (llama-3.1-8b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 0s
three-m4-pro-cluster (llama-3.2-1b) / run-distributed-job (M4PRO_GPU16_24GB)
2m 6s
three-m4-pro-cluster (llama-3.2-1b) / run-distributed-job (M4PRO_GPU16_24GB)
2m 6s
three-m4-pro-cluster (llama-3.2-1b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 26s
single-m4-pro (llama-3.2-1b) / run-distributed-job (M4PRO_GPU16_24GB)
56s
three-m4-pro-cluster (llama-3.3-70b) / run-distributed-job (M4PRO_GPU16_24GB)
2m 25s
three-m4-pro-cluster (llama-3.3-70b) / run-distributed-job (M4PRO_GPU16_24GB)
3m 25s
three-m4-pro-cluster (llama-3.3-70b) / run-distributed-job (M4PRO_GPU16_24GB)
3m 24s
three-m4-pro-cluster (llama-3.2-3b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 12s
three-m4-pro-cluster (llama-3.2-3b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 46s
three-m4-pro-cluster (llama-3.2-3b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 30s
two-m4-pro-cluster (llama-3.1-8b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 25s
two-m4-pro-cluster (llama-3.1-8b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 26s
three-m4-pro-cluster (llama-3.1-8b) / run-distributed-job (M4PRO_GPU16_24GB)
2m 42s
three-m4-pro-cluster (llama-3.1-8b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 47s
three-m4-pro-cluster (llama-3.1-8b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 31s
two-m4-pro-cluster (llama-3.2-3b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 25s
two-m4-pro-cluster (llama-3.2-3b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 24s
two-m4-pro-cluster (llama-3.2-1b) / run-distributed-job (M4PRO_GPU16_24GB)
2m 13s
two-m4-pro-cluster (llama-3.2-1b) / run-distributed-job (M4PRO_GPU16_24GB)
1m 15s
37m 30s