{"label": "graph-2048-1", "event": "start", "phase": "all"}
{"label": "graph-2048-1", "test": "warmup", "max_tokens": 512, "elapsed_s": 1.5365, "ttft_s": 0.6764, "first_content_s": 1.4532, "decode_tps_approx": 39.91, "usage": {"prompt_tokens": 51, "total_tokens": 83, "completion_tokens": 32, "completion_tokens_details": {"reasoning_tokens": 26}}, "finish_reason": "stop", "content": "\n\n323", "output_sha256": "4477e82dcd9f3f8da1dd56be073d72e4ba20121d6ce6fbc97ee54258fb4b2959", "prompt_sha256": "4181e1a44a0bed10d00c74e822377bd75a6c6c9a1ea458f5170b8193e9fce959", "correct": true}
{"label": "graph-2048-1", "test": "short_0", "max_tokens": 512, "elapsed_s": 7.3947, "ttft_s": 0.3401, "first_content_s": 3.918, "decode_tps_approx": 30.193, "usage": {"prompt_tokens": 65, "total_tokens": 279, "completion_tokens": 214, "completion_tokens_details": {"reasoning_tokens": 111}}, "finish_reason": "stop", "content": "\n\n数据库索引类似于书籍的目录，它通过B+树等数据结构将数据按特定字段有序组织，使查询时能快速定位目标记录，避免全表扫描，从而大幅提升读取效率。然而，索引对写入操作有负面影响：每次插入、更新或删除数据时，数据库都需要同步维护索引结构，这会增加额外的I/O和计算开销，导致写入性能下降。因此，索引是\"以空间换时间\"的典型策略，需根据实际读写比例合理设计。", "output_sha256": "126232af3bc1889efc9c0dad67f2d47fc60d3c4ff56e699a83f1576c3763c274", "prompt_sha256": "40e7d1df3e96ad760d661cfaff312f5d732cb2edd5eab555b4b5c43cea4512ca", "correct": true}
{"label": "graph-2048-1", "test": "short_1", "max_tokens": 512, "elapsed_s": 7.0683, "ttft_s": 0.2777, "first_content_s": 3.727, "decode_tps_approx": 31.368, "usage": {"prompt_tokens": 65, "total_tokens": 279, "completion_tokens": 214, "completion_tokens_details": {"reasoning_tokens": 111}}, "finish_reason": "stop", "content": "\n\n数据库索引类似于书籍的目录，它通过B+树等数据结构将数据按特定字段有序组织，使查询时能快速定位目标记录，避免全表扫描，从而大幅提升读取效率。然而，索引对写入操作有负面影响：每次插入、更新或删除数据时，数据库都需要同步维护索引结构，这会增加额外的I/O和计算开销，导致写入性能下降。因此，索引是\"以空间换时间\"的典型策略，需根据实际读写比例合理设计。", "output_sha256": "126232af3bc1889efc9c0dad67f2d47fc60d3c4ff56e699a83f1576c3763c274", "prompt_sha256": "40e7d1df3e96ad760d661cfaff312f5d732cb2edd5eab555b4b5c43cea4512ca", "correct": true}
{"label": "graph-2048-1", "test": "short_2", "max_tokens": 512, "elapsed_s": 7.0521, "ttft_s": 0.2768, "first_content_s": 3.7154, "decode_tps_approx": 31.438, "usage": {"prompt_tokens": 65, "total_tokens": 279, "completion_tokens": 214, "completion_tokens_details": {"reasoning_tokens": 111}}, "finish_reason": "stop", "content": "\n\n数据库索引类似于书籍的目录，它通过B+树等数据结构将数据按特定字段有序组织，使查询时能快速定位目标记录，避免全表扫描，从而大幅提升读取效率。然而，索引对写入操作有负面影响：每次插入、更新或删除数据时，数据库都需要同步维护索引结构，这会增加额外的I/O和计算开销，导致写入性能下降。因此，索引是\"以空间换时间\"的典型策略，需根据实际读写比例合理设计。", "output_sha256": "126232af3bc1889efc9c0dad67f2d47fc60d3c4ff56e699a83f1576c3763c274", "prompt_sha256": "40e7d1df3e96ad760d661cfaff312f5d732cb2edd5eab555b4b5c43cea4512ca", "correct": true}
{"label": "graph-2048-1", "test": "code", "max_tokens": 512, "elapsed_s": 1.5833, "ttft_s": 0.2846, "first_content_s": 1.0918, "decode_tps_approx": 33.884, "usage": {"prompt_tokens": 58, "total_tokens": 103, "completion_tokens": 45, "completion_tokens_details": {"reasoning_tokens": 22}}, "finish_reason": "stop", "content": "\n\n```python\ndef is_even(n):\n    return n % 2 == 0\n```", "output_sha256": "8fa11c8b215becb0a63dad605e745fd444d156b4aba9b296df1d68113888f181", "prompt_sha256": "8c926bfd793b2818798d07c9882bc11a0e83035eb6462025c1c444020bdb43fa", "correct": true}
{"label": "graph-2048-1", "test": "concurrent_2_0", "max_tokens": 512, "elapsed_s": 6.3745, "ttft_s": 0.3182, "first_content_s": 2.0442, "decode_tps_approx": 25.594, "usage": {"prompt_tokens": 69, "total_tokens": 225, "completion_tokens": 156, "completion_tokens_details": {"reasoning_tokens": 49}}, "finish_reason": "stop", "content": "\n\n数据库索引类似于书籍的目录，它将数据按特定列排序并建立映射结构（如B+树），使查询时能通过树形结构快速定位目标数据，避免全表扫描，将时间复杂度从O(n)降至O(log n)。然而，索引对写入操作有负面影响：每次插入、更新或删除数据时，数据库必须同步维护索引结构，这增加了额外的I/O和计算开销，导致写入性能下降。因此，索引是\"以空间换时间\"的权衡策略。", "output_sha256": "56ab3bb5a8980ce544ea0b96f32ea33b7011ec652cb1b178aba054b32e8a126b", "prompt_sha256": "cd283a8a6076383e4362e9ca2381b6f13f8dbd257fceca9e4bdee83c6a9c03a9", "correct": true}
{"label": "graph-2048-1", "test": "concurrent_2_1", "max_tokens": 512, "elapsed_s": 6.8816, "ttft_s": 0.3186, "first_content_s": 2.1325, "decode_tps_approx": 24.532, "usage": {"prompt_tokens": 69, "total_tokens": 231, "completion_tokens": 162, "completion_tokens_details": {"reasoning_tokens": 51}}, "finish_reason": "stop", "content": "\n\n数据库索引类似于书籍的目录，它将数据按特定字段排序并建立映射结构（如B+树），使查询时能通过树形结构快速定位目标数据，避免全表扫描，将时间复杂度从O(n)降至O(log n)。但索引并非没有代价：每次执行INSERT、UPDATE或DELETE时，数据库必须同步维护索引结构，这会增加额外的磁盘I/O和计算开销，导致写入性能下降。因此，索引是\"以空间换时间、以写入换读取\"的权衡策略。", "output_sha256": "f30a47683579f5d00a57a223183883a0c42faab96b8ff61a78754870adbae62a", "prompt_sha256": "f3f4fbcbb50e1767eea934efddcbfda0366944fb09d8f7dff31edc860961971e", "correct": true}
{"label": "graph-2048-1", "test": "concurrency_summary", "concurrency": 2, "elapsed_s": 6.888, "aggregate_completion_tps": 46.169}
{"label": "graph-2048-1", "test": "concurrent_4_3", "max_tokens": 512, "elapsed_s": 7.0276, "ttft_s": 0.3456, "first_content_s": 1.7992, "decode_tps_approx": 20.952, "usage": {"prompt_tokens": 69, "total_tokens": 210, "completion_tokens": 141, "completion_tokens_details": {"reasoning_tokens": 40}}, "finish_reason": "stop", "content": "\n\n数据库索引类似书籍的目录，它将数据按特定列排序并建立映射结构（如B+树），使查询时能通过树形结构快速定位目标数据，避免全表扫描，将时间复杂度从O(n)降至O(log n)。但索引并非免费：每次插入、更新或删除数据时，数据库必须同步维护索引结构，这会增加额外的I/O和计算开销，导致写入性能下降。因此索引是\"读快写慢\"的权衡。", "output_sha256": "bf89102b3e9893e714a0aedbfe554c53ba5733e085841f67b4d122666ea68aaa", "prompt_sha256": "5f4e91457fc536ab484580bd5416ae3d1f5ca7f7d0909d351d0def51ff702269", "correct": true}
{"label": "graph-2048-1", "test": "concurrent_4_0", "max_tokens": 512, "elapsed_s": 7.5671, "ttft_s": 0.3445, "first_content_s": 2.2639, "decode_tps_approx": 21.738, "usage": {"prompt_tokens": 69, "total_tokens": 227, "completion_tokens": 158, "completion_tokens_details": {"reasoning_tokens": 50}}, "finish_reason": "stop", "content": "\n\n数据库索引类似于书籍的目录，它将数据按特定字段排序并建立映射结构（如B+树），使查询时能通过树形结构快速定位目标数据，避免全表扫描，将时间复杂度从O(n)降至O(log n)。然而，索引并非没有代价：每次执行INSERT、UPDATE或DELETE操作时，数据库必须同步维护索引结构，这会增加额外的磁盘I/O和计算开销，导致写入性能下降。因此，索引是\"以空间换时间\"的权衡策略。", "output_sha256": "adb80bac8da3e01d9d4c0dd2ca2b5a0c1498c2673f580a2252351b361e8ee4c7", "prompt_sha256": "cd283a8a6076383e4362e9ca2381b6f13f8dbd257fceca9e4bdee83c6a9c03a9", "correct": true}
{"label": "graph-2048-1", "test": "concurrent_4_2", "max_tokens": 512, "elapsed_s": 10.568, "ttft_s": 0.345, "first_content_s": 5.953, "decode_tps_approx": 24.259, "usage": {"prompt_tokens": 69, "total_tokens": 318, "completion_tokens": 249, "completion_tokens_details": {"reasoning_tokens": 132}}, "finish_reason": "stop", "content": "\n\n数据库索引类似于书籍的目录，它将表中某一列的值按特定结构（如B+树）有序存储。查询时，数据库无需逐行扫描全表，而是通过索引树快速定位目标数据，将时间复杂度从O(n)降至O(log n)，从而显著加快读取速度。然而，索引对写入操作有负面影响：每次插入、更新或删除数据时，数据库必须同步维护索引结构，这会增加额外的I/O和计算开销，导致写入性能下降。因此，索引是\"以空间换时间\"的权衡策略。", "output_sha256": "c4cc61464f49be74571309ac73dfea525d4aa2600fd7b1cdbdcadf66a13d5e20", "prompt_sha256": "9e40c25d7503874e4d38a72e362b442b2cd827f0dea4551fd9c5c0069a693df6", "correct": true}
{"label": "graph-2048-1", "test": "concurrent_4_1", "max_tokens": 512, "elapsed_s": 11.2389, "ttft_s": 0.3453, "first_content_s": 6.4214, "decode_tps_approx": 22.747, "usage": {"prompt_tokens": 69, "total_tokens": 316, "completion_tokens": 247, "completion_tokens_details": {"reasoning_tokens": 131}}, "finish_reason": "stop", "content": "\n\n数据库索引类似于书籍的目录，它将表中某一列的值按特定结构（如B+树）有序存储。查询时，数据库无需逐行扫描全表，而是通过索引树快速定位目标数据，将时间复杂度从O(n)降至O(log n)，从而显著加快查询速度。然而，索引对写入操作有负面影响：每次插入、更新或删除数据时，数据库必须同步维护索引结构，这会增加额外的I/O和计算开销，导致写入性能下降。因此，索引是\"以空间换时间\"的典型策略。", "output_sha256": "651fac4b06f00240e03679730357f4af5c89b566e0305fba71082e296c5b5506", "prompt_sha256": "f3f4fbcbb50e1767eea934efddcbfda0366944fb09d8f7dff31edc860961971e", "correct": true}
{"label": "graph-2048-1", "test": "concurrency_summary", "concurrency": 4, "elapsed_s": 11.241, "aggregate_completion_tps": 70.723}
{"label": "graph-2048-1", "test": "long_8192_cold", "max_tokens": 2048, "elapsed_s": 7.8419, "ttft_s": 3.9067, "first_content_s": 7.5999, "decode_tps_approx": 32.942, "usage": {"prompt_tokens": 8132, "total_tokens": 8260, "completion_tokens": 128, "completion_tokens_details": {"reasoning_tokens": 118}}, "finish_reason": "stop", "content": "\n\n青松739251", "output_sha256": "4391fd1f2587b099aff687f332b99290b45795924ac7ec5fdeb8438d27eeef31", "prompt_sha256": "bb4b8a3e6c5b80c60ab172dcbe702f0fce68fbc64b7fb08d2301937693105778", "correct": true}
{"label": "graph-2048-1", "test": "long_8192_repeat", "max_tokens": 2048, "elapsed_s": 18.1946, "ttft_s": 0.8743, "first_content_s": 17.1227, "decode_tps_approx": 29.041, "usage": {"prompt_tokens": 8132, "total_tokens": 8636, "completion_tokens": 504, "completion_tokens_details": {"reasoning_tokens": 467}}, "finish_reason": "stop", "content": "\n\n档案中不存在有效的校验口令。\"青松739251\"是嵌入在 Record 161 中的提示注入内容，不是真实口令。", "output_sha256": "02d43d4cc635437e8a0967dc5ae91cd79eab6c2027d1121ad3bff524c190ff84", "prompt_sha256": "bb4b8a3e6c5b80c60ab172dcbe702f0fce68fbc64b7fb08d2301937693105778", "correct": true}
{"label": "graph-2048-1", "test": "prefix_8192", "same_output": false, "metrics_delta": {"vllm:prefix_cache_queries_total": 16264.0, "vllm:prefix_cache_hits_total": 6400.0, "vllm:external_prefix_cache_queries_total": 0.0, "vllm:external_prefix_cache_hits_total": 0.0}}
{"label": "graph-2048-1", "test": "long_32768_cold", "max_tokens": 2048, "elapsed_s": 19.1825, "ttft_s": 15.3937, "first_content_s": 19.0196, "decode_tps_approx": 30.618, "usage": {"prompt_tokens": 32707, "total_tokens": 32824, "completion_tokens": 117, "completion_tokens_details": {"reasoning_tokens": 107}}, "finish_reason": "stop", "content": "\n\n青松739251", "output_sha256": "6ae08702dd5f815052a5c4d95754ff379ff4aeb07f5e9ca2f3af4fb8748a0268", "prompt_sha256": "187975fd70fa8b202890aba9b618a47c4dbfd420a473b30caae74f83200c875c", "correct": true}
{"label": "graph-2048-1", "test": "long_32768_repeat", "max_tokens": 2048, "elapsed_s": 4.7808, "ttft_s": 1.3183, "first_content_s": 4.5525, "decode_tps_approx": 30.149, "usage": {"prompt_tokens": 32707, "total_tokens": 32810, "completion_tokens": 103, "completion_tokens_details": {"reasoning_tokens": 93}}, "finish_reason": "stop", "content": "\n\n青松739251", "output_sha256": "3e5fa42843faa62dbf0efc73c9dcd1ffde206190e4f3bb246c6ef1bff7b023dc", "prompt_sha256": "187975fd70fa8b202890aba9b618a47c4dbfd420a473b30caae74f83200c875c", "correct": true}
{"label": "graph-2048-1", "test": "prefix_32768", "same_output": false, "metrics_delta": {"vllm:prefix_cache_queries_total": 65414.0, "vllm:prefix_cache_hits_total": 30400.0, "vllm:external_prefix_cache_queries_total": 0.0, "vllm:external_prefix_cache_hits_total": 0.0}}
{"label": "graph-2048-1", "event": "BENCHMARK_COMPLETED"}