Merge branch 'vijay/suppress_ninja_output' into 'main'

suppress kernel loading output See merge request ADLR/megatron-lm!155

Merge branch 'vijay/suppress_ninja_output' into 'main'
suppress kernel loading output See merge request ADLR/megatron-lm!155
5753e8f1 · Mohammad Shoeybi · 538f0e05 · 8fca9b49 · 5753e8f1
Commit 5753e8f1 authored Oct 12, 2020 by Mohammad Shoeybi
Hide whitespace changes
Inline Side-by-side

Showing with 2 additions and 4 deletions

megatron/fused_kernels/__init__.py megatron/fused_kernels/__init__.py +2 -4

No files found.
--- a/megatron/fused_kernels/__init__.py
+++ b/megatron/fused_kernels/__init__.py
@@ -69,8 +69,7 @@ def load_scaled_upper_triang_masked_softmax_fusion_kernel():
                           '-U__CUDA_NO_HALF_CONVERSIONS__',
                           '--expt-relaxed-constexpr',
                           '--expt-extended-lambda',
-                           '--use_fast_math'] + cc_flag,
-        verbose=True)
+                           '--use_fast_math'] + cc_flag)

 def load_scaled_masked_softmax_fusion_kernel():

@@ -98,5 +97,4 @@ def load_scaled_masked_softmax_fusion_kernel():
                           '-U__CUDA_NO_HALF_CONVERSIONS__',
                           '--expt-relaxed-constexpr',
                           '--expt-extended-lambda',
-                           '--use_fast_math'] + cc_flag,
-        verbose=True)
+                           '--use_fast_math'] + cc_flag)