Fix ZeroDivisionError in metric evaluation across examples

[ARYANPATEL-BIT]

Status: Fixed ✅

🎯 Issue Fixed

Fixes #383

🚨 Problem

During metric evaluation, accuracy calculation divides by len(same_elements). Empty lists from prediction edge-cases cause ZeroDivisionError and crash the training/evaluation pipeline.

📁 Affected Files

examples/llm_simple_qa/testenv/acc.py
examples/government/singletask_learning_bench/objective/testenv/acc.py
examples/government_rag/singletask_learning_bench/testenv/acc.py
examples/cloud-edge-collaborative-inference-for-llm/testenv/accuracy.py

text

🔧 The Fix

Before (CRASH):

acc = sum(same_elements) / len(same_elements)

After (SAFE):

acc = sum(same_elements) / len(same_elements) if len(same_elements) > 0 else 0.0

💾 Complete Fixed Function

def calculate_accuracy(same_elements):
    """Safe accuracy calculation for empty prediction lists."""
    return sum(same_elements) / len(same_elements) if len(same_elements) > 0 else 0.0

✅ Test Cases

# All now work without crashing:
calculate_accuracy([])                          # 0.0
calculate_accuracy()                   # 0.666...[1]
calculate_accuracy([True, False, True])         # 0.666...

[kubeedge-bot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: ARYANPATEL-BIT
To complete the pull request process, please assign moorezheng after the PR has been reviewed.
You can assign the PR to them by writing /assign @moorezheng in a comment when ready.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[ARYANPATEL-BIT]

/assign @MooreZheng

[gemini-code-assist]

Code Review

This pull request adds safety checks to prevent ZeroDivisionError in accuracy calculation functions across multiple test environments. The feedback suggests using more idiomatic Python by checking the truthiness of the list directly instead of its length.