📊 Configuration:
Base URL: http://localhost:9099/v1
Model: gemma-4-E2B-it-UD-Q8_K_XL_1_65536_off
Test Cases: 17
Output: results/agent_test_results_gemma-4-E2B-it-UD-Q8_K_XL_1_65536_off_20260403_194653.json
Log File: logs/agent_test_logs_gemma-4-E2B-it-UD-Q8_K_XL_1_65536_off_20260403_194653.log

🔄 Running agent tests...
Starting agent test suite with 17 test cases
Running agent test: zero_weather_question
Running agent test: simple_view_cart
Running agent test: simple_checkout
Running agent test: zero_greeting
Running agent test: medium_view_and_add
Running agent test: medium_search_and_add
Running agent test: complex_shopping_workflow
Running agent test: complex_cart_management
Running agent test: complex_gift_shopping
Running agent test: medium_remove_and_add
Running agent test: zero_capabilities
Running agent test: zero_general_question
Running agent test: zero_thank_you
Running agent test: simple_search_electronics
Running agent test: medium_search_category_and_add
Running agent test: simple_add_iphone
Running agent test: simple_remove_product
✅ Tests completed in 13.1988869s

📈 Agent Test Results

Total Tests: 17
✅ Passed: 14
❌ Failed: 3
⏱️ Total LLM Time: 2m16.9697881s
⏱️ Average Time per Request: 4.028523179s

📋 Test Case Results:

Test Case: zero_thank_you
Status: ✅ PASSED
Matched Path: no_tools
Response Time: 1.9016205s
Tool Calls: 0

Test Case: zero_capabilities
Status: ✅ PASSED
Matched Path: no_tools
Response Time: 4.0371244s
Tool Calls: 0

Test Case: zero_weather_question
Status: ✅ PASSED
Matched Path: no_tools
Response Time: 4.7684901s
Tool Calls: 0

Test Case: complex_shopping_workflow
Status: ❌ FAILED
Response Time: 5.0747829s
Tool Calls: 0

Test Case: complex_gift_shopping
Status: ❌ FAILED
Response Time: 5.8311264s
Tool Calls: 0

Test Case: simple_view_cart
Status: ✅ PASSED
Matched Path: view_cart
Response Time: 6.1369392s
Tool Calls: 1
Tools Used: view_cart

Test Case: zero_greeting
Status: ✅ PASSED
Matched Path: no_tools
Response Time: 6.4544902s
Tool Calls: 0

Test Case: zero_general_question
Status: ✅ PASSED
Matched Path: no_tools
Response Time: 6.6994278s
Tool Calls: 0

Test Case: simple_remove_product
Status: ✅ PASSED
Matched Path: direct_remove
Response Time: 7.018502s
Tool Calls: 1
Tools Used: remove_from_cart

Test Case: simple_add_iphone
Status: ✅ PASSED
Matched Path: direct_add
Response Time: 7.9340505s
Tool Calls: 1
Tools Used: add_to_cart

Test Case: simple_checkout
Status: ✅ PASSED
Matched Path: direct_checkout
Response Time: 9.5477404s
Tool Calls: 1
Tools Used: checkout

Test Case: medium_search_and_add
Status: ✅ PASSED
Matched Path: search_by_query
Response Time: 10.0965437s
Tool Calls: 2
Tools Used: search_products, add_to_cart

Test Case: medium_search_category_and_add
Status: ✅ PASSED
Matched Path: search_then_add
Response Time: 11.0634474s
Tool Calls: 2
Tools Used: search_products, add_to_cart

Test Case: medium_view_and_add
Status: ✅ PASSED
Matched Path: view_then_add
Response Time: 12.22999s
Tool Calls: 2
Tools Used: view_cart, add_to_cart

Test Case: medium_remove_and_add
Status: ❌ FAILED
Response Time: 12.6187984s
Tool Calls: 3
Tools Used: remove_from_cart, search_products, add_to_cart

Test Case: simple_search_electronics
Status: ✅ PASSED
Matched Path: search_by_category
Response Time: 12.8816156s
Tool Calls: 1
Tools Used: search_products

Test Case: complex_cart_management
Status: ✅ PASSED
Matched Path: cart_organization
Response Time: 13.1978356s
Tool Calls: 3
Tools Used: view_cart, remove_from_cart, add_to_cart

❌ Failed Tests Details:

Test Case: complex_shopping_workflow
Expected Tool Variants: 4
Variant 1 (full_workflow_with_iphone): 4 tools
Variant 2 (full_workflow_with_headphones): 4 tools
Variant 3 (full_workflow_with_headphones_and_iphone): 5 tools
Variant 4 (full_workflow_with_iphone_and_headphones): 5 tools
Actual Tool Calls: 0
Response Time: 5.0747829s

Test Case: complex_gift_shopping
Expected Tool Variants: 2
Variant 1 (gift_shopping_workflow): 5 tools
Variant 2 (gift_shopping_workflow): 5 tools
Actual Tool Calls: 0
Response Time: 5.8311264s

Test Case: medium_remove_and_add
Expected Tool Variants: 1
Variant 1 (remove_then_add): 2 tools
Actual Tool Calls: 3
1. remove_from_cart
2. search_products
3. add_to_cart
Response Time: 12.6187984s

📊 Overall Success Rate: 82.35% (reasoning off)
Response Time: 12.6187984s

📊 Overall Success Rate: 76.47% (reasoning on)
Response Time: 43.6138286s

75 tokens/s on RTX 3060 12GB // https://github.com/docker/model-test/

llama-server --port 9099 -ngl 99 -fa on -c 65536 --temp 1 --top-k 64 --top-p 0.95 --jinja -m X:\path\to\gemma-4-E2B-it-UD-Q8_K_XL.gguf --mmproj X:\path\to\mmproj-gemma-4-E4B-it-UD-F16.gguf --reasoning off

Very fast small model, did ok with function calling, bad in "guessing" when given image input, it biased to safe answer, not "brave"/"confidence" enough. (reasoning off)

reasoning on = bad result, image hallucinating.