Skip to content
Build Docker Images - CUDA

FlagScale Train Upgrade: Synchronization with Megatron-LM Core0.17.0 #50

Sign in to view logs

FlagScale Train Upgrade: Synchronization with Megatron-LM Core0.17.0

FlagScale Train Upgrade: Synchronization with Megatron-LM Core0.17.0 #50

Run time

Learn about OS pricing on GitHub Actions
Job Run time
Prepare build matrix
4s
Build train
3m 51s
Build inference
4m 10s
Build all
4m 30s
Build summary
4s
Load and push images
4m 7s
Run CUDA tests / checkout_and_config
7s
Run CUDA tests / cli_validation / functional_test_cli
32s
Run CUDA tests / unit_tests (a100) / unit-a100
1m 59s
Run CUDA tests / functional_tests_benchmark / functional_test_benchmark (a100, benchmark, qwen3, tp2_pp2)
2m 51s
Run CUDA tests / functional_tests_serve / functional_test_serve (a100, serve, qwen2_5, 0.5b)
1m 10s
Run CUDA tests / functional_tests_hetero_train / functional_test_hetero_train (a100, hetero_train, aquila, dp2dp4_shared_embedding)
3m 23s
Run CUDA tests / functional_tests_hetero_train / functional_test_hetero_train (a100, hetero_train, aquila, tp2pp1_tp4pp1_tp2pp1)
6m 11s
Run CUDA tests / functional_tests_hetero_train / functional_test_hetero_train (a100, hetero_train, aquila, tp2dp1pp1_tp2dp2pp1_tp1dp2pp1)
5m 40s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, deepseek, tp2_pp2_ep2_engram)
3m 6s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, aquila, tp2_pp2)
2m 52s
Run CUDA tests / functional_tests_inference / functional_test_inference (a100, inference, qwen3, 4b_tp2)
59s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, mixtral, tp2_pp1_ep2)
3m 3s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, deepseek, tp2_pp2_ep2)
3m 7s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, mixtral, tp4_pp1_ep2)
3m 16s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, aquila, tp4_pp2)
3m 5s
Run CUDA tests / all_tests_complete
2s
Prepare build matrix
4s
Build inference
4m 10s
Build train
3m 43s
Build all
4m 20s
Build summary
2s
Load and push images
4m 8s
Run CUDA tests / checkout_and_config
8s
Run CUDA tests / cli_validation / functional_test_cli
35s
Run CUDA tests / unit_tests (a100) / unit-a100
1m 59s
Run CUDA tests / functional_tests_serve / functional_test_serve (a100, serve, qwen2_5, 0.5b)
1m 12s
Run CUDA tests / functional_tests_benchmark / functional_test_benchmark (a100, benchmark, qwen3, tp2_pp2)
3m 4s
Run CUDA tests / functional_tests_inference / functional_test_inference (a100, inference, qwen3, 4b_tp2)
1m 2s
Run CUDA tests / functional_tests_hetero_train / functional_test_hetero_train (a100, hetero_train, aquila, tp2pp1_tp4pp1_tp2pp1)
5m 54s
Run CUDA tests / functional_tests_hetero_train / functional_test_hetero_train (a100, hetero_train, aquila, dp2dp4_shared_embedding)
3m 23s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, mixtral, tp2_pp1_ep2)
3m 5s
Run CUDA tests / functional_tests_hetero_train / functional_test_hetero_train (a100, hetero_train, aquila, tp2dp1pp1_tp2dp2pp1_tp1dp2pp1)
5m 53s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, aquila, tp4_pp2)
3m 18s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, mixtral, tp4_pp1_ep2)
3m 25s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, aquila, tp2_pp2)
3m 4s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, deepseek, tp2_pp2_ep2)
2m 59s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, deepseek, tp2_pp2_ep2_engram)
2m 57s
Run CUDA tests / all_tests_complete
4s
Prepare build matrix
4s
Build inference
3m 53s
Build train
3m 36s
Build all
3m 51s
Build summary
3s
Load and push images
3m 3s
Run CUDA tests / checkout_and_config
7s
Run CUDA tests / cli_validation / functional_test_cli
34s
Run CUDA tests / unit_tests (a100) / unit-a100
2m 7s
Run CUDA tests / functional_tests_serve / functional_test_serve (a100, serve, qwen2_5, 0.5b)
1m 10s
Run CUDA tests / functional_tests_inference / functional_test_inference (a100, inference, qwen3, 4b_tp2)
1m 6s
Run CUDA tests / functional_tests_hetero_train / functional_test_hetero_train (a100, hetero_train, aquila, dp2dp4_shared_embedding)
5m 24s
Run CUDA tests / functional_tests_benchmark / functional_test_benchmark (a100, benchmark, qwen3, tp2_pp2)
3m 1s
Run CUDA tests / functional_tests_hetero_train / functional_test_hetero_train (a100, hetero_train, aquila, tp2pp1_tp4pp1_tp2pp1)
5m 52s
Run CUDA tests / functional_tests_hetero_train / functional_test_hetero_train (a100, hetero_train, aquila, tp2dp1pp1_tp2dp2pp1_tp1dp2pp1)
5m 51s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, mixtral, tp2_pp1_ep2)
3m 17s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, mixtral, tp4_pp1_ep2)
3m 26s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, deepseek, tp2_pp2_ep2_engram)
3m 1s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, deepseek, tp2_pp2_ep2)
4m 58s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, aquila, tp2_pp2)
3m 0s
Run CUDA tests / functional_tests_train / functional_test_train (a100, train, aquila, tp4_pp2)
3m 26s
Run CUDA tests / all_tests_complete
2s
2h 57m 30s