Local LLM testing (16GB ram, Intel N150, no GPU)

results.txt

(venv) jeff@echelon3:~/ollama_tests$ python structure.py 
---
Testing: llama3.2:latest ...
Using prompt of: Tell me about Canada.
name='Canada' capital='Ottawa' languages=['English', 'French']
Execution time: 7.5 seconds
---

---
Testing: llama3.2:latest ...
Using prompt of: Tell me about Australia.
name='Australia' capital='Canberra' languages=['English', 'Australian']
Execution time: 4.1 seconds
---

---
Testing: llama3.2:latest ...
Using prompt of: Tell me about China.
name='China' capital='Beijing' languages=['Mandarin Chinese', 'Cantonese', 'Mandarin Taiwanese', 'other languages, such as Wu and Min Nan']
Execution time: 6.0 seconds
---

TOTAL llama3.2:latest MODEL EXECUTION TIME: 17.7 seconds

---
Testing: phi3:latest ...
Using prompt of: Tell me about Canada.
name='Canada' capital='Ottawa' languages=['English', 'French']
Execution time: 8.7 seconds
---

---
Testing: phi3:latest ...
Using prompt of: Tell me about Australia.
name='Australia' capital='Canberra' languages=['English', 'Aboriginal languages (over 250)']
Execution time: 5.5 seconds
---

---
Testing: phi3:latest ...
Using prompt of: Tell me about China.
name='China' capital='Beijing' languages=['Mandarin Chinese (Putonghua)', 'Cantonese']
Execution time: 6.4 seconds
---

TOTAL phi3:latest MODEL EXECUTION TIME: 20.6 seconds

---
Testing: gemma3:latest ...
Using prompt of: Tell me about Canada.
name='Canada' capital='Ottawa' languages=['English', 'French']
Execution time: 10.1 seconds
---

---
Testing: gemma3:latest ...
Using prompt of: Tell me about Australia.
name='Australia' capital='Canberra' languages=['English (de facto)']
Execution time: 7.7 seconds
---

---
Testing: gemma3:latest ...
Using prompt of: Tell me about China.
name='China' capital='Beijing' languages=['Mandarin Chinese (official)', 'Cantonese', 'Wu', 'Min', 'Hakka', 'Yue', 'and many others']
Execution time: 10.5 seconds
---

TOTAL gemma3:latest MODEL EXECUTION TIME: 28.3 seconds

---
Testing: deepseek-r1:latest ...
Using prompt of: Tell me about Canada.
name='Canada' capital='Ottawa' languages=['English', 'French']
Execution time: 14.7 seconds
---

---
Testing: deepseek-r1:latest ...
Using prompt of: Tell me about Australia.
name='Australia' capital='Canberra' languages=['English', 'Aboriginal languages', 'sign language', 'other minority languages like Chinese, Vietnamese, Italian etc.']
Execution time: 16.8 seconds
---

---
Testing: deepseek-r1:latest ...
Using prompt of: Tell me about China.
name='China' capital='Beijing' languages=['Chinese']
Execution time: 8.3 seconds
---

TOTAL deepseek-r1:latest MODEL EXECUTION TIME: 39.8 seconds

---
Testing: gemma3n:latest ...
Using prompt of: Tell me about Canada.
name='Canada' capital='Ottawa' languages=['English', 'French', 'Other (Indigenous languages)']
Execution time: 17.7 seconds
---

---
Testing: gemma3n:latest ...
Using prompt of: Tell me about Australia.
name='Australia' capital='Canberra' languages=['English']
Execution time: 10.3 seconds
---

---
Testing: gemma3n:latest ...
Using prompt of: Tell me about China.
name='China' capital='Beijing' languages=['Mandarin Chinese (official)', 'Wu', 'Yue (Cantonese)', 'Min', 'Xiang', 'Gan', 'Hakka']
Execution time: 18.6 seconds
---

TOTAL gemma3n:latest MODEL EXECUTION TIME: 46.6 seconds

---
Testing: qwen3:latest ...
Using prompt of: Tell me about Canada.
name='Canada' capital='Ottawa' languages=['English', 'French', 'Official Languages Act, 1982 (French and English)']
Execution time: 20.9 seconds
---

---
Testing: qwen3:latest ...
Using prompt of: Tell me about Australia.
name='Australia' capital='Canberra' languages=['English', 'Aboriginal languages', 'other languages such as Chinese, Japanese, Korean, etc.']
Execution time: 11.8 seconds
---

---
Testing: qwen3:latest ...
Using prompt of: Tell me about China.
name='China' capital='Beijing' languages=['Mandarin', 'Cantonese', 'other languages of China', 'English (widely taught)']
Execution time: 13.4 seconds
---

TOTAL qwen3:latest MODEL EXECUTION TIME: 46.1 seconds