{ "model_id": "m1", "model_tag": "qwen3.5:9b", "box_class": "cpu-server", "started_utc": "2026-08-25T14:32:03Z", "nproc": 64, "num_predict": 256, "prompt_sha256": "90eedd0c53f9554ae3837674504fcb7090013d9a432a653352183c0f25a7ce5c", "nvidia_before": "0, 38764 MiB, 0 %", "ollama_version": "0.32.15", "runs": [ { "ttft_ms": 149.39, "first_token_was_thinking": false, "wall_s": 19.4388, "eval_count": 256, "eval_duration_ns": 19285220000, "decode_tok_s": 13.274, "prompt_eval_count": 537, "prompt_eval_duration_ns": 140843000, "prefill_tok_s": 3812.756, "load_duration_ns": 1534863, "total_duration_ns": 19434115796, "streamed_chunks": 237, "response_chars": 1021, "thinking_chars": 0, "done_reason": "length", "run": 1 }, { "ttft_ms": 140.3, "first_token_was_thinking": false, "wall_s": 19.0737, "eval_count": 256, "eval_duration_ns": 18929117000, "decode_tok_s": 13.524, "prompt_eval_count": 537, "prompt_eval_duration_ns": 132406000, "prefill_tok_s": 4055.707, "load_duration_ns": 1422804, "total_duration_ns": 19069210541, "streamed_chunks": 237, "response_chars": 1021, "thinking_chars": 0, "done_reason": "length", "run": 2 }, { "ttft_ms": 141.31, "first_token_was_thinking": false, "wall_s": 19.0707, "eval_count": 256, "eval_duration_ns": 18924986000, "decode_tok_s": 13.527, "prompt_eval_count": 537, "prompt_eval_duration_ns": 132923000, "prefill_tok_s": 4039.933, "load_duration_ns": 1217245, "total_duration_ns": 19065826927, "streamed_chunks": 237, "response_chars": 1021, "thinking_chars": 0, "done_reason": "length", "run": 3 }, { "ttft_ms": 140.85, "first_token_was_thinking": false, "wall_s": 19.0806, "eval_count": 256, "eval_duration_ns": 18935623000, "decode_tok_s": 13.519, "prompt_eval_count": 537, "prompt_eval_duration_ns": 133064000, "prefill_tok_s": 4035.652, "load_duration_ns": 1499084, "total_duration_ns": 19076105493, "streamed_chunks": 237, "response_chars": 1021, "thinking_chars": 0, "done_reason": "length", "run": 4 }, { "ttft_ms": 139.75, "first_token_was_thinking": false, "wall_s": 19.1388, "eval_count": 256, "eval_duration_ns": 18994311000, "decode_tok_s": 13.478, "prompt_eval_count": 537, "prompt_eval_duration_ns": 131549000, "prefill_tok_s": 4082.129, "load_duration_ns": 1937912, "total_duration_ns": 19134024062, "streamed_chunks": 237, "response_chars": 1021, "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": 135.73, "first_token_was_thinking": false, "wall_s": 19.1377, "eval_count": 256, "eval_duration_ns": 18997747000, "decode_tok_s": 13.475, "prompt_eval_count": 537, "prompt_eval_duration_ns": 127214000, "prefill_tok_s": 4221.234, "load_duration_ns": 1458974, "total_duration_ns": 19133052357, "streamed_chunks": 237, "response_chars": 1021, "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": "2026-08-25T14:42:37.925201669Z", "size_vram": 0, "context_length": 4096 } ], "size_vram": 0, "size_total": 6071601067, "cpu_only_receipt": true, "peak_runner_rss_kb": 6977544, "peak_runner_rss_gib": 6.65, "peak_runner_threads": 103, "rss_samples": 175, "derived": { "decode_tok_s": { "min": 13.274, "median": 13.519, "max": 13.527, "mean": 13.464, "n": 5 }, "ttft_ms": { "min": 139.75, "median": 140.85, "max": 149.39, "mean": 142.32, "n": 5 }, "wall_s": { "min": 19.0707, "median": 19.0806, "max": 19.4388, "mean": 19.161, "n": 5 }, "prefill_tok_s": { "min": 3812.756, "median": 4039.933, "max": 4082.129, "mean": 4005.235, "n": 5 }, "eval_count": { "min": 256, "median": 256, "max": 256, "mean": 256.0, "n": 5 } }, "cold_load": [ { "ttft_ms": 14530.12, "first_token_was_thinking": false, "wall_s": 14.5331, "eval_count": 1, "eval_duration_ns": 1000, "decode_tok_s": 1000000.0, "prompt_eval_count": 537, "prompt_eval_duration_ns": 6023641000, "prefill_tok_s": 89.149, "load_duration_ns": 8496397187, "total_duration_ns": 14529374409, "streamed_chunks": 1, "response_chars": 5, "thinking_chars": 0, "done_reason": "length", "wall_to_first_token_s": 14.5333, "ps_confirmed_empty_before": true, "load_duration_s": 8.4964 }, { "ttft_ms": 14533.24, "first_token_was_thinking": false, "wall_s": 14.5362, "eval_count": 1, "eval_duration_ns": 1000, "decode_tok_s": 1000000.0, "prompt_eval_count": 537, "prompt_eval_duration_ns": 6026516000, "prefill_tok_s": 89.106, "load_duration_ns": 8497205584, "total_duration_ns": 14532220227, "streamed_chunks": 1, "response_chars": 5, "thinking_chars": 0, "done_reason": "length", "wall_to_first_token_s": 14.5364, "ps_confirmed_empty_before": true, "load_duration_s": 8.4972 } ], "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": "0, 46719 MiB, 0 %", "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": "2026-08-25T14:44:44.830249343Z", "size_vram": 0, "context_length": 4096 } ], "finished_utc": "2026-08-25T14:34:45Z" }