{ "model_id": "m1", "model_tag": "qwen3.5:9b", "box_class": "cpu-mini", "started_utc": "2026-08-25T14:09:19Z", "nproc": 16, "num_predict": 256, "prompt_sha256": "90eedd0c53f9554ae3837674504fcb7090013d9a432a653352183c0f25a7ce5c", "nvidia_before": "no nvidia-smi (no GPU on this box)", "ollama_version": "0.32.15", "runs": [ { "ttft_ms": 210.65, "first_token_was_thinking": false, "wall_s": 28.3115, "eval_count": 256, "eval_duration_ns": 28098036000, "decode_tok_s": 9.111, "prompt_eval_count": 537, "prompt_eval_duration_ns": 205508000, "prefill_tok_s": 2613.037, "load_duration_ns": 834593, "total_duration_ns": 28308415784, "streamed_chunks": 239, "response_chars": 1040, "thinking_chars": 0, "done_reason": "length", "run": 1 }, { "ttft_ms": 209.46, "first_token_was_thinking": false, "wall_s": 28.3276, "eval_count": 256, "eval_duration_ns": 28115215000, "decode_tok_s": 9.105, "prompt_eval_count": 537, "prompt_eval_duration_ns": 204314000, "prefill_tok_s": 2628.307, "load_duration_ns": 859845, "total_duration_ns": 28324344866, "streamed_chunks": 239, "response_chars": 1040, "thinking_chars": 0, "done_reason": "length", "run": 2 }, { "ttft_ms": 210.78, "first_token_was_thinking": false, "wall_s": 28.3122, "eval_count": 256, "eval_duration_ns": 28098598000, "decode_tok_s": 9.111, "prompt_eval_count": 537, "prompt_eval_duration_ns": 205391000, "prefill_tok_s": 2614.525, "load_duration_ns": 892551, "total_duration_ns": 28308962932, "streamed_chunks": 239, "response_chars": 1040, "thinking_chars": 0, "done_reason": "length", "run": 3 }, { "ttft_ms": 212.93, "first_token_was_thinking": false, "wall_s": 28.8847, "eval_count": 256, "eval_duration_ns": 28669041000, "decode_tok_s": 8.929, "prompt_eval_count": 537, "prompt_eval_duration_ns": 208197000, "prefill_tok_s": 2579.288, "load_duration_ns": 908675, "total_duration_ns": 28881459054, "streamed_chunks": 239, "response_chars": 1040, "thinking_chars": 0, "done_reason": "length", "run": 4 }, { "ttft_ms": 213.15, "first_token_was_thinking": false, "wall_s": 28.4482, "eval_count": 256, "eval_duration_ns": 28232212000, "decode_tok_s": 9.068, "prompt_eval_count": 537, "prompt_eval_duration_ns": 208016000, "prefill_tok_s": 2581.532, "load_duration_ns": 820510, "total_duration_ns": 28444945126, "streamed_chunks": 239, "response_chars": 1040, "thinking_chars": 0, "done_reason": "length", "run": 5 } ], "errors": [], "quantization": "Q4_K_M", "param_size": "9.7B", "family": "qwen35", "context_length": 262144, "think_mode": false, "think_note": "think:false accepted", "warmup_discarded": { "ttft_ms": 209.47, "first_token_was_thinking": false, "wall_s": 28.56, "eval_count": 256, "eval_duration_ns": 28347714000, "decode_tok_s": 9.031, "prompt_eval_count": 537, "prompt_eval_duration_ns": 203857000, "prefill_tok_s": 2634.199, "load_duration_ns": 911873, "total_duration_ns": 28556799331, "streamed_chunks": 239, "response_chars": 1040, "thinking_chars": 0, "done_reason": "length" }, "ps_after_load": [ { "name": "qwen3.5:9b", "model": "qwen3.5:9b", "size": 6071601067, "digest": "6488c96fa5faab64bb65cbd30d4289e20e6130ef535a93ef9a49f42eda893ea7", "details": { "parent_model": "", "format": "gguf", "family": "qwen35", "families": [ "qwen35" ], "parameter_size": "9.7B", "quantization_level": "Q4_K_M" }, "expires_at": "2026-08-25T14:20:09.582038125Z", "size_vram": 0, "context_length": 4096 } ], "size_vram": 0, "size_total": 6071601067, "cpu_only_receipt": true, "peak_runner_rss_kb": 6964916, "peak_runner_rss_gib": 6.64, "peak_runner_threads": 23, "rss_samples": 280, "derived": { "decode_tok_s": { "min": 8.929, "median": 9.105, "max": 9.111, "mean": 9.065, "n": 5 }, "ttft_ms": { "min": 209.46, "median": 210.78, "max": 213.15, "mean": 211.394, "n": 5 }, "wall_s": { "min": 28.3115, "median": 28.3276, "max": 28.8847, "mean": 28.457, "n": 5 }, "prefill_tok_s": { "min": 2579.288, "median": 2613.037, "max": 2628.307, "mean": 2603.338, "n": 5 }, "eval_count": { "min": 256, "median": 256, "max": 256, "mean": 256.0, "n": 5 } }, "cold_load": [ { "ttft_ms": 20318.95, "first_token_was_thinking": false, "wall_s": 20.3213, "eval_count": 1, "eval_duration_ns": 1000, "decode_tok_s": 1000000.0, "prompt_eval_count": 537, "prompt_eval_duration_ns": 15966654000, "prefill_tok_s": 33.633, "load_duration_ns": 4347426673, "total_duration_ns": 20318493475, "streamed_chunks": 1, "response_chars": 5, "thinking_chars": 0, "done_reason": "length", "wall_to_first_token_s": 20.3214, "ps_confirmed_empty_before": true, "load_duration_s": 4.3474 }, { "ttft_ms": 20431.54, "first_token_was_thinking": false, "wall_s": 20.434, "eval_count": 1, "eval_duration_ns": 1000, "decode_tok_s": 1000000.0, "prompt_eval_count": 537, "prompt_eval_duration_ns": 16085466000, "prefill_tok_s": 33.384, "load_duration_ns": 4340963088, "total_duration_ns": 20431060133, "streamed_chunks": 1, "response_chars": 5, "thinking_chars": 0, "done_reason": "length", "wall_to_first_token_s": 20.4341, "ps_confirmed_empty_before": true, "load_duration_s": 4.341 } ], "cold_load_note": "PAGE-CACHE-WARM: no sudo to drop caches, so this is warm-disk reload, not a true cold read from NVMe.", "nvidia_after": "no nvidia-smi (no GPU on this box)", "ps_final": [ { "name": "qwen3.5:9b", "model": "qwen3.5:9b", "size": 6071601067, "digest": "6488c96fa5faab64bb65cbd30d4289e20e6130ef535a93ef9a49f42eda893ea7", "details": { "parent_model": "", "format": "gguf", "family": "qwen35", "families": [ "qwen35" ], "parameter_size": "9.7B", "quantization_level": "Q4_K_M" }, "expires_at": "2026-08-25T14:23:13.383450448Z", "size_vram": 0, "context_length": 4096 } ], "finished_utc": "2026-08-25T14:13:13Z" }