Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions scripts/performance/perf_plugins.py
Original file line number Diff line number Diff line change
Expand Up @@ -250,7 +250,7 @@ def _set_model_specific_environment_variables(
):
"""Set model-specific environment variables"""
if (
model_family_name in ["llama31"]
model_family_name in ["llama"]
and model_recipe_name in ["llama31_405b"]
and train_task == "pretrain"
and gpu in ["gb200"]
Expand All @@ -259,16 +259,16 @@ def _set_model_specific_environment_variables(
executor.env_vars["PYTORCH_CUDA_ALLOC_CONF"] = "expandable_segments:True"
del_cudnn_ln = True
if gpu in ["h100"]:
if model_family_name == "llama3" and model_recipe_name == "llama3_8b" and train_task == "pretrain":
if model_family_name == "llama" and model_recipe_name == "llama3_8b" and train_task == "pretrain":
if compute_dtype == "fp8_cs":
# executor.env_vars["NCCL_NVLS_ENABLE"] = "1" # This causes OOM; worked fine with NeMo2 and 25.09
executor.env_vars["NCCL_CTA_POLICY"] = "1"
del_cudnn_ln = False
if gpu in ["gb200", "gb300"]:
if model_family_name == "llama3" and model_recipe_name == "llama3_70b" and train_task == "pretrain":
if model_family_name == "llama" and model_recipe_name == "llama3_70b" and train_task == "pretrain":
if compute_dtype == "bf16" or (compute_dtype == "fp8_cs"):
del_cudnn_ln = False
if model_family_name == "llama31" and model_recipe_name == "llama31_405b" and train_task == "pretrain":
if model_family_name == "llama" and model_recipe_name == "llama31_405b" and train_task == "pretrain":
if compute_dtype == "fp8_cs":
del_cudnn_ln = False
if del_cudnn_ln:
Expand Down
Loading