Skip to content

Commit a076842

Browse files
authored
Update quantization config for small_bf16_linear model
1 parent 2de61f6 commit a076842

File tree

1 file changed

+2
-2
lines changed

1 file changed

+2
-2
lines changed

benchmarks/dashboard/microbenchmark_quantization_config.yml

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -11,10 +11,10 @@ model_params:
1111
- name: "small_bf16_linear"
1212
matrix_shapes:
1313
- name: "small_sweep"
14-
min_power: 10
14+
min_power: 14
1515
max_power: 15
1616
high_precision_dtype: "torch.bfloat16"
17-
torch_compile_mode: "max-autotune"
17+
torch_compile_mode: "default"
1818
device: "cuda"
1919
model_type: "linear"
2020
enable_memory_profiler: true

0 commit comments

Comments
 (0)