Skip to content

Actions: deepspeedai/DeepSpeed

Actions

cpu-torch-latest

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
cpu-torch-latest.yml
2,810 workflow runs
2,810 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

Add sequential allgather optimization for ZeRO-3 cpu-torch-latest #7010: Pull request #7661 synchronize by aeeeeeep
22m 33s aeeeeeep:allgather_single_param
aeeeeeep:allgather_single_param
22m 33s
cpu-torch-latest cpu-torch-latest #7009: Scheduled
22m 54s master
master
22m 54s
Add full test suite workflow cpu-torch-latest #7008: Pull request #7795 opened by tohtana
23m 33s tohtana:tohtana/add_full_test_workflow_v2
tohtana:tohtana/add_full_test_workflow_v2
23m 33s
Fix bf16 dtype mismatch in ZeRO-3 with zero_quantized_weights cpu-torch-latest #7007: Pull request #7792 synchronize by juyterman1000
24m 1s juyterman1000:fix/bf16-zero3-quantized-weights
juyterman1000:fix/bf16-zero3-quantized-weights
24m 1s
Fix AutoTP test numerical tolerance with rtol cpu-torch-latest #7006: Pull request #7794 synchronize by tohtana
24m 2s tohtana:tohtana/autotp_test_tolerance
tohtana:tohtana/autotp_test_tolerance
24m 2s
Fix AutoTP test numerical tolerance with rtol cpu-torch-latest #7005: Pull request #7794 opened by tohtana
9m 53s tohtana:tohtana/autotp_test_tolerance
tohtana:tohtana/autotp_test_tolerance
9m 53s
Fix bf16 dtype mismatch in ZeRO-3 with zero_quantized_weights cpu-torch-latest #7004: Pull request #7792 synchronize by juyterman1000
22m 38s juyterman1000:fix/bf16-zero3-quantized-weights
juyterman1000:fix/bf16-zero3-quantized-weights
22m 38s
Fix bf16 dtype mismatch in ZeRO-3 with zero_quantized_weights cpu-torch-latest #7003: Pull request #7792 synchronize by juyterman1000
Action required juyterman1000:fix/bf16-zero3-quantized-weights
juyterman1000:fix/bf16-zero3-quantized-weights
Action required
Fix bf16 dtype mismatch in ZeRO-3 with zero_quantized_weights cpu-torch-latest #7002: Pull request #7792 synchronize by juyterman1000
21m 15s juyterman1000:fix/bf16-zero3-quantized-weights
juyterman1000:fix/bf16-zero3-quantized-weights
21m 15s
Fix bf16 dtype mismatch in ZeRO-3 with zero_quantized_weights cpu-torch-latest #7001: Pull request #7792 synchronize by juyterman1000
Action required juyterman1000:fix/bf16-zero3-quantized-weights
juyterman1000:fix/bf16-zero3-quantized-weights
Action required
Fix bf16 dtype mismatch in ZeRO-3 with zero_quantized_weights cpu-torch-latest #7000: Pull request #7792 synchronize by juyterman1000
-1s juyterman1000:fix/bf16-zero3-quantized-weights
juyterman1000:fix/bf16-zero3-quantized-weights
-1s
Fix bf16 dtype mismatch in ZeRO-3 with zero_quantized_weights cpu-torch-latest #6999: Pull request #7792 synchronize by juyterman1000
Action required juyterman1000:fix/bf16-zero3-quantized-weights
juyterman1000:fix/bf16-zero3-quantized-weights
Action required
Fix Evoformer compilation cpu-torch-latest #6998: Pull request #7760 synchronize by sdvillal
23m 0s sdvillal:improve-evoformer-compilation
sdvillal:improve-evoformer-compilation
23m 0s
Add sequential allgather optimization for ZeRO-3 cpu-torch-latest #6997: Pull request #7661 synchronize by aeeeeeep
Action required aeeeeeep:allgather_single_param
aeeeeeep:allgather_single_param
Action required
Fix loss scaling and backward call of ZenFlow cpu-torch-latest #6996: Pull request #7793 opened by tohtana
24m 11s tohtana:tohtana/fix_zenfulow_fp16_loss_scaling
tohtana:tohtana/fix_zenfulow_fp16_loss_scaling
24m 11s
Fix bf16 dtype mismatch in ZeRO-3 with zero_quantized_weights cpu-torch-latest #6995: Pull request #7792 synchronize by juyterman1000
Action required juyterman1000:fix/bf16-zero3-quantized-weights
juyterman1000:fix/bf16-zero3-quantized-weights
Action required
Fix bf16 dtype mismatch in ZeRO-3 with zero_quantized_weights cpu-torch-latest #6994: Pull request #7792 opened by juyterman1000
Action required juyterman1000:fix/bf16-zero3-quantized-weights
juyterman1000:fix/bf16-zero3-quantized-weights
Action required
Add bf16 model with fp32 grad_accum to supported configs cpu-torch-latest #6993: Pull request #7790 opened by tohtana
22m 18s tohtana:tohtana/fix_ds_initialize_test_bf16_fp32
tohtana:tohtana/fix_ds_initialize_test_bf16_fp32
22m 18s
Skip empty parameters in gradient reduction cpu-torch-latest #6992: Pull request #7789 opened by tohtana
23m 20s tohtana:tohtana/fix_onebit_lamb_empty_params
tohtana:tohtana/fix_onebit_lamb_empty_params
23m 20s
Fix issue with BF16 optimizer selection cpu-torch-latest #6991: Pull request #7788 opened by tohtana
22m 27s tohtana:tohtana/fix_bf16_optimizer_select
tohtana:tohtana/fix_bf16_optimizer_select
22m 27s
cpu-torch-latest cpu-torch-latest #6990: Scheduled
22m 26s master
master
22m 26s
Fix backward for pipeline engine cpu-torch-latest #6989: Pull request #7787 opened by tohtana
21m 21s tohtana:tohtana/fix_pp_engine_backward
tohtana:tohtana/fix_pp_engine_backward
21m 21s
Fix Evoformer compilation cpu-torch-latest #6988: Pull request #7760 synchronize by tohtana
23m 24s sdvillal:improve-evoformer-compilation
sdvillal:improve-evoformer-compilation
23m 24s
Fix Evoformer compilation cpu-torch-latest #6986: Pull request #7760 synchronize by sdvillal
Action required sdvillal:improve-evoformer-compilation
sdvillal:improve-evoformer-compilation
Action required
fix checkpointing/loading of z0+bf16 cpu-torch-latest #6985: Pull request #7786 opened by tohtana
22m 20s tohtana:tohtana/fix_z0_bf16_checkpoint_opt_state
tohtana:tohtana/fix_z0_bf16_checkpoint_opt_state
22m 20s