| { |
| "format": "5bit", |
| "strict_reload": true, |
| "finite_prefill_decode_logits": true, |
| "cache_offsets_correct": true, |
| "source_sidecars_identical": true, |
| "native_thinking_template_preserved": true, |
| "temperature": 0.0, |
| "max_tokens": 512, |
| "smoke_tests": [ |
| { |
| "task": "arithmetic", |
| "prompt": "What is 17 multiplied by 23? Give the answer.", |
| "text": "<think>1. The user asks for 17 multiplied by 23 and to give the answer.\n2. Calculation: 17 * 23 = 391.\n3. Provide the answer directly and clearly.</think>17 multiplied by 23 is **391**.", |
| "answer": "17 multiplied by 23 is **391**.", |
| "checks": { |
| "expected_391": true |
| }, |
| "generated_tokens": 55, |
| "seconds": 0.5447727499995381 |
| }, |
| { |
| "task": "retrieval", |
| "prompt": "The passcode is 941731. Repeat the passcode exactly.", |
| "text": "<think>The user is asking me to repeat a passcode. This is a simple request. I should just repeat it exactly as they provided it.\n\nThe passcode is 941731.\n\nI should repeat it exactly. No need for extra commentary unless it seems helpful, but the instruction is clear: \"Repeat the passcode exactly.\"\n\nSo my response should just be: 941731\n\nLet me make sure I don't add anything extra that might be unwanted. The user wants me to repeat it exactly, so I should just output the number.</think>941731", |
| "answer": "941731", |
| "checks": { |
| "expected_passcode": true |
| }, |
| "generated_tokens": 109, |
| "seconds": 1.0578467911109328 |
| }, |
| { |
| "task": "code", |
| "prompt": "Write a Python function add(a, b) that returns a + b.", |
| "text": "<think>We need to write a Python function add(a, b) that returns a + b. This is straightforward. The user likely expects a simple implementation.\n\nWe should provide the code, maybe with a brief explanation.\n\nThe instruction: \"Write a Python function add(a, b) that returns a + b.\"\n\nSo we can just output:\n\ndef add(a, b):\n return a + b\n\nWe might also include a docstring or example usage, but the core is that.\n\nSince the user asked to write the function, we should just provide the code. Possibly in a code block.\n\nWe'll respond with the function definition and maybe a short explanation.\n\nNo need for extra complexity.\n\nLet's craft the response.</think>Here's a simple Python function that adds two numbers:\n\n```python\ndef add(a, b):\n return a + b\n```\n\nYou can use it like this:\n\n```python\nresult = add(3, 5) # result will be 8\n```\n\nThe function works with any types that support the `+` operator, such as integers, floats, strings, lists, etc.", |
| "answer": "Here's a simple Python function that adds two numbers:\n\n```python\ndef add(a, b):\n return a + b\n```\n\nYou can use it like this:\n\n```python\nresult = add(3, 5) # result will be 8\n```\n\nThe function works with any types that support the `+` operator, such as integers, floats, strings, lists, etc.", |
| "checks": { |
| "add_function": true |
| }, |
| "generated_tokens": 221, |
| "seconds": 1.9357592919841409 |
| } |
| ], |
| "all_smoke_checks_pass": true, |
| "device": { |
| "device_name": "Apple M5 Max", |
| "max_recommended_working_set_size": 115448725504, |
| "memory_size": 137438953472, |
| "architecture": "applegpu_g17s", |
| "max_buffer_length": 86586540032, |
| "resource_limit": 499000 |
| }, |
| "mlx_version": "0.32.4", |
| "mlx_vlm_version": "0.7.10", |
| "converter_revision": "6d9ba328abbd6129c7461eda86280aa9fbfd9986", |
| "python": "3.12.15", |
| "peak_active_gb": 27.436328788, |
| "seconds": 5.28359774989076, |
| "swap_before": "vm.swapusage: total = 10240.00M used = 8736.50M free = 1503.50M (encrypted)", |
| "swap_after": "vm.swapusage: total = 10240.00M used = 8736.50M free = 1503.50M (encrypted)", |
| "limits": "Three synthetic smoke prompts and finite/cache checks; not a quality benchmark or proof of parity to BF16/Transformers." |
| } |
