We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent 4121b67 commit 41901d1Copy full SHA for 41901d1
tensorrt_llm/_torch/autotuner.py
@@ -1141,7 +1141,10 @@ def _optimization_profiles(
1141
# Add the current input value as one of the opt values
1142
opt_shapes = set(opt_shapes)
1143
opt_shapes.add(
1144
- base_profile.shapes[spec.input_idx][spec.dim_idx].val)
+ min(
1145
+ tuning_config.tune_max_num_tokens,
1146
+ base_profile.shapes[spec.input_idx][spec.dim_idx].val,
1147
+ ))
1148
opt_shapes = sorted(list(opt_shapes))
1149
opt_shapes_max = tuple(opt_shapes[1:]) + (float('inf'), )
1150
opt_shapes_max = {
0 commit comments