Clear and fix metrics computation

[hiyuchang]

Description

As the title says.

Now it looks like

Step 0: {'eval/gsm8k-eval/finished_task_count': 1319, 'eval/gsm8k-eval/accuracy/mean@4': 0.34154662623199394, 'eval/gsm8k-eval/accuracy/std@4': 0.2833624429507763, 'eval/gsm8k-eval/accuracy/best@2/mean': 0.46970053070507956, 'eval/gsm8k-eval/accuracy/best@2/std': 0.2601951934227904, 'eval/gsm8k-eval/accuracy/worst@2/mean': 0.21339878695981807, 'eval/gsm8k-eval/accuracy/worst@2/std': 0.22732260946377386, 'eval/gsm8k-eval/accuracy/best@4/mean': 0.582379833206975, 'eval/gsm8k-eval/accuracy/best@4/std': 0.187753964514041, 'eval/gsm8k-eval/accuracy/worst@4/mean': 0.12428809704321456, 'eval/gsm8k-eval/accuracy/worst@4/std': 0.13311398785210515, 'eval/gsm8k-eval/format_score/mean@4': 0.04791508718726308, 'eval/gsm8k-eval/format_score/std@4': 0.060430579521288934, 'eval/gsm8k-eval/format_score/best@2/mean': 0.07519408642911297, 'eval/gsm8k-eval/format_score/best@2/std': 0.043823375514568795, 'eval/gsm8k-eval/format_score/worst@2/mean': 0.020637604245640644, 'eval/gsm8k-eval/format_score/worst@2/std': 0.06012139017838808, 'eval/gsm8k-eval/format_score/best@4/mean': 0.09088036391205459, 'eval/gsm8k-eval/format_score/best@4/std': 0.02081875469787524, 'eval/gsm8k-eval/format_score/worst@4/mean': -0.006707808946171337, 'eval/gsm8k-eval/format_score/worst@4/std': 0.04779745171526119, 'eval/gsm8k-eval/time/run_execution/mean': 1.8486472846342814, 'eval/gsm8k-eval/time/run_execution/std': 0.7957768482777444, 'eval/gsm8k-eval/time/task_execution/mean': 2.0765281061587504, 'eval/gsm8k-eval/time/task_execution/std': 0.8317681867952945, 'time/eval': 77.17239356040955}

Checklist

Please check the following items before code is ready to be reviewed.

• Code has passed all tests
• Docstrings have been added/updated in Google Style
• Documentation has been updated
• Code is ready for review

[pan-x-c]

/unittest-module-trainer

[github-actions]

Summary

Tests 📝	Passed ✅	Failed ❌	Skipped ⏭️	Other ❓	Flaky 🍂	Duration ⏱️
26	18	3	5	0	0	37m 27s

Failed Tests

Failed Tests ❌	Fail Message
❌ tests/trainer/trainer_test.py::TestTrainerCountdown_0_fsdp::test_trainer	The test failed in the call phase
❌ tests/trainer/trainer_test.py::TestTrainerCountdown_1_megatron::test_trainer	The test failed in the call phase
❌ tests/trainer/trainer_test.py::TestTrainerLoRA::test_trainer	The test failed in the call phase

Skipped

Tests	Status
tests/trainer/trainer_test.py::TestMultiModalGRPO::test_trainer	skipped ⏭️
tests/trainer/trainer_test.py::TestMultiModalSFT::test_trainer	skipped ⏭️
tests/trainer/trainer_test.py::TestTinkerTrainer::test_trainer	skipped ⏭️
tests/trainer/trainer_test.py::TestTinkerTrainer::test_trainer_class	skipped ⏭️
tests/trainer/trainer_test.py::AgentScopeTunerTest::test_agentscope_tuner	skipped ⏭️

Tests

Test Name	Status	Flaky	Duration
tests/trainer/trainer_test.py::TestTrainerCountdown_0_fsdp::test_trainer	❌		3m 27s
tests/trainer/trainer_test.py::TestTrainerCountdown_1_megatron::test_trainer	❌		3m 43s
tests/trainer/trainer_test.py::TestStepAheadAsyncRL::test_trainer	✅		1m 40s
tests/trainer/trainer_test.py::TestTrainerGSM8K_0_fsdp::test_trainer	✅		1m 19s
tests/trainer/trainer_test.py::TestTrainerGSM8K_1_fsdp2::test_trainer	✅		47.8s
tests/trainer/trainer_test.py::TestTrainerGSM8K_2_fsdp::test_trainer	✅		54.6s
tests/trainer/trainer_test.py::TestTrainerGSM8K_3_fsdp2::test_trainer	✅		59.0s
tests/trainer/trainer_test.py::TestTrainerSFTWarmupGSM8K::test_trainer	✅		1m 58s
tests/trainer/trainer_test.py::TestTrainerDPO::test_trainer	✅		34.8s
tests/trainer/trainer_test.py::TestTrainerSFT::test_trainer	✅		31.7s
tests/trainer/trainer_test.py::TestTrainerToolsSFT::test_trainer_tools	✅		30.0s
tests/trainer/trainer_test.py::TestFullyAsyncMode_0_fsdp::test_fully_async_mode	✅		1m 31s
tests/trainer/trainer_test.py::TestFullyAsyncMode_1_fsdp::test_fully_async_mode	✅		1m 29s
tests/trainer/trainer_test.py::TestFullyAsyncMode_2_megatron::test_fully_async_mode	✅		2m 20s
tests/trainer/trainer_test.py::TestTrainerCheckpointSave_0_fsdp::test_trainer	✅		2m 4s
tests/trainer/trainer_test.py::TestTrainerCheckpointSave_1_megatron::test_trainer	✅		5m 13s
tests/trainer/trainer_test.py::TestTrainerMIX::test_trainer	✅		1m 33s
tests/trainer/trainer_test.py::TestServeWithTrainer::test_serve_with_trainer	✅		1m 47s
tests/trainer/trainer_test.py::TestMultiModalGRPO::test_trainer	⏭️		554ms
tests/trainer/trainer_test.py::TestMultiModalSFT::test_trainer	⏭️		548ms
tests/trainer/trainer_test.py::TestTrainerLoRA::test_trainer	❌		2m 50s
tests/trainer/trainer_test.py::TestOverRollout::test_trainer	✅		48.9s
tests/trainer/trainer_test.py::TestTrainerPromptTruncation::test_trainer	✅		1m 13s
tests/trainer/trainer_test.py::TestTinkerTrainer::test_trainer	⏭️		1ms
tests/trainer/trainer_test.py::TestTinkerTrainer::test_trainer_class	⏭️		1ms
tests/trainer/trainer_test.py::AgentScopeTunerTest::test_agentscope_tuner	⏭️		1ms

Github Test Reporter by CTRF 💚