[CI] Cleanup run_tests.sh logs (#700)

HabanaAI · Jan 17, 2025 · 7eea2df · 7eea2df
1 parent a685225
commit 7eea2df
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 3 deletions.
diff --git a/.jenkins/lm-eval-harness/run-tests.sh b/.jenkins/lm-eval-harness/run-tests.sh
@@ -42,6 +42,7 @@ do
     export LM_EVAL_TP_SIZE=$TP_SIZE
     export PT_HPU_ENABLE_LAZY_COLLECTIVES=true
     export VLLM_SKIP_WARMUP=true
+    export TQDM_BAR_FORMAT="{desc}: {percentage:3.0f}% {bar:10} | {n_fmt}/{total_fmt} [{elapsed}<{remaining}]" 
     RANDOM_SUFFIX=$(tr -dc A-Za-z0-9 </dev/urandom | head -c 4; echo)
     JUNIT_FAMILY=""
     JUNIT_XML=""

diff --git a/.jenkins/lm-eval-harness/test_lm_eval_correctness.py b/.jenkins/lm-eval-harness/test_lm_eval_correctness.py
@@ -24,6 +24,9 @@
     "LM_EVAL_TEST_DATA_FILE",
     ".jenkins/lm-eval-harness/configs/Meta-Llama-3-8B-Instruct.yaml")
 
+REPORT_PERFORMANCE = os.environ.get("LM_EVAL_REPORT_PERFORMANCE",
+                                    "false") in ['1', 'true']
+
 TP_SIZE = os.environ.get("LM_EVAL_TP_SIZE", 1)
 
 
@@ -170,9 +173,10 @@ def test_lm_eval_correctness(record_xml_attribute, record_property):
                                 x['resps'])))['input_ids'])) for x in samples
             ]
             tokenized_outputs_lens = [len(x) for x in tokenized_outputs]
-            report_performance(task['name'], tokenized_inputs_lens,
-                               tokenized_outputs_lens, total_time,
-                               record_property)
+            if REPORT_PERFORMANCE:
+                report_performance(task['name'], tokenized_inputs_lens,
+                                   tokenized_outputs_lens, total_time,
+                                   record_property)
 
             for metric in task["metrics"]:
                 ground_truth = metric["value"]