lint

Signed-off-by: Jennifer Chen <jennifchen@nvidia.com>

Author: jenchen13

modelopt/torch/quantization/config.py

@@ -163,12 +163,12 @@
 }
 
 _mamba_moe_disabled_quantizer_cfg = {
-    "*fc1_latent_proj*": {"enable": False}, # Skip Latent MOE
-    "*fc2_latent_proj*": {"enable": False}, # Skip Latent MOE
-    "*q_proj*": {"enable": False}, # Skip QKV Linear
-    "*k_proj*": {"enable": False}, # Skip QKV Linear
-    "*v_proj*": {"enable": False}, # Skip QKV Linear
-    "*o_proj*": {"enable": False}, # Skip QKV Output Projection
+    "*fc1_latent_proj*": {"enable": False},  # Skip Latent MOE
+    "*fc2_latent_proj*": {"enable": False},  # Skip Latent MOE
+    "*q_proj*": {"enable": False},  # Skip QKV Linear
+    "*k_proj*": {"enable": False},  # Skip QKV Linear
+    "*v_proj*": {"enable": False},  # Skip QKV Linear
+    "*o_proj*": {"enable": False},  # Skip QKV Output Projection
 }
 
 INT8_DEFAULT_CFG = {