Update crash step in ft launcher test

Signed-off-by: Shriya Palsamudram <[email protected]>
NVIDIA · Dec 3, 2024 · 6deffa9 · 6deffa9
1 parent fa6a773
commit 6deffa9
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/.github/workflows/cicd-main.yml b/.github/workflows/cicd-main.yml
@@ -3934,7 +3934,7 @@ jobs:
         export FAULT_TOL_FINISHED_FLAG_FILE="/tmp/llm_tests/llama_pretrain_results/sample_job_finished_flag"; \
         python tests/collections/llm/test_fault_nvrx.py \
         --devices=2 \
-        --crash-step=4 \
+        --crash-step=16 \
         --experiment-dir=/tmp/llm_tests/llama_pretrain_results \
         --data-path=/home/TestData/nlp/megatron_llama/data/rp2_sample_sentencepiece_preproc_text_document \
         --tokenizer-path=/home/TestData/nlp/megatron_llama/tokenizer.model \