20 lines
11 KiB
JSON
20 lines
11 KiB
JSON
{"label": "eager-2048-1", "event": "start", "phase": "all"}
|
|
{"label": "eager-2048-1", "test": "warmup", "max_tokens": 512, "elapsed_s": 1.1418, "ttft_s": 0.2595, "first_content_s": 1.056, "decode_tps_approx": 38.92, "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": "eager-2048-1", "test": "short_0", "max_tokens": 512, "elapsed_s": 7.3168, "ttft_s": 0.2811, "first_content_s": 3.7942, "decode_tps_approx": 30.275, "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": "eager-2048-1", "test": "short_1", "max_tokens": 512, "elapsed_s": 7.3245, "ttft_s": 0.2849, "first_content_s": 3.7781, "decode_tps_approx": 30.258, "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": "eager-2048-1", "test": "short_2", "max_tokens": 512, "elapsed_s": 7.3274, "ttft_s": 0.2788, "first_content_s": 3.7901, "decode_tps_approx": 30.22, "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": "eager-2048-1", "test": "code", "max_tokens": 512, "elapsed_s": 1.5483, "ttft_s": 0.2685, "first_content_s": 1.0619, "decode_tps_approx": 34.383, "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": "eager-2048-1", "test": "concurrent_2_1", "max_tokens": 512, "elapsed_s": 6.6418, "ttft_s": 0.3497, "first_content_s": 2.214, "decode_tps_approx": 25.429, "usage": {"prompt_tokens": 69, "total_tokens": 230, "completion_tokens": 161, "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)。但索引并非没有代价：每次执行INSERT、UPDATE、DELETE操作时，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、以写入换读取\"的权衡策略。", "output_sha256": "198b7100d5b0569ea57451e4cfe4a585548821f27d1493641791495fa7ed28d7", "prompt_sha256": "f3f4fbcbb50e1767eea934efddcbfda0366944fb09d8f7dff31edc860961971e", "correct": true}
|
|
{"label": "eager-2048-1", "test": "concurrent_2_0", "max_tokens": 512, "elapsed_s": 9.0027, "ttft_s": 0.4814, "first_content_s": 5.3813, "decode_tps_approx": 27.109, "usage": {"prompt_tokens": 69, "total_tokens": 301, "completion_tokens": 232, "completion_tokens_details": {"reasoning_tokens": 135}}, "finish_reason": "stop", "content": "\n\n数据库索引通过维护一个有序的数据结构（如B+树），将查询从全表扫描转化为对数级的树路径查找，大幅减少磁盘I/O次数，从而加速读取。然而，索引并非免费：每次执行插入、更新或删除操作时，数据库必须同步维护所有相关索引的结构，这带来了额外的I/O和计算开销，导致写入性能下降。因此，索引本质上是以空间换时间、以写入换读取的权衡策略。", "output_sha256": "4f26f050e740ec6538dc53a47ef1a369bc190e3610dcd73cc0c52b1ab45303cb", "prompt_sha256": "cd283a8a6076383e4362e9ca2381b6f13f8dbd257fceca9e4bdee83c6a9c03a9", "correct": true}
|
|
{"label": "eager-2048-1", "test": "concurrency_summary", "concurrency": 2, "elapsed_s": 9.007, "aggregate_completion_tps": 43.635}
|
|
{"label": "eager-2048-1", "test": "concurrent_4_0", "max_tokens": 512, "elapsed_s": 7.2973, "ttft_s": 0.3301, "first_content_s": 1.81, "decode_tps_approx": 20.238, "usage": {"prompt_tokens": 69, "total_tokens": 211, "completion_tokens": 142, "completion_tokens_details": {"reasoning_tokens": 37}}, "finish_reason": "stop", "content": "\n\n数据库索引类似书籍的目录，通过B+树等数据结构将数据按特定列有序组织。查询时，数据库无需全表扫描，而是沿树结构快速定位目标行，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、以写入换读取\"的权衡。", "output_sha256": "f700c2d7c15ec499d3ca322631b466453560ea7923bff1085e15ef42ff1ec419", "prompt_sha256": "cd283a8a6076383e4362e9ca2381b6f13f8dbd257fceca9e4bdee83c6a9c03a9", "correct": true}
|
|
{"label": "eager-2048-1", "test": "concurrent_4_3", "max_tokens": 512, "elapsed_s": 9.4707, "ttft_s": 0.3301, "first_content_s": 4.7466, "decode_tps_approx": 21.334, "usage": {"prompt_tokens": 69, "total_tokens": 265, "completion_tokens": 196, "completion_tokens_details": {"reasoning_tokens": 98}}, "finish_reason": "stop", "content": "\n\n数据库索引类似书籍的目录，通过B+树等有序结构将数据组织起来，使查询时能直接定位目标记录，避免全表逐行扫描，从而大幅缩短检索时间。但索引并非没有代价：每次执行插入、更新或删除操作时，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、以写入换读取\"的典型权衡。", "output_sha256": "06aa68f7ec666dd08d3eb08ee467f2e9c35c60638ef6785472beca41c2f1c03c", "prompt_sha256": "5f4e91457fc536ab484580bd5416ae3d1f5ca7f7d0909d351d0def51ff702269", "correct": true}
|
|
{"label": "eager-2048-1", "test": "concurrent_4_1", "max_tokens": 512, "elapsed_s": 10.4012, "ttft_s": 0.3296, "first_content_s": 6.1973, "decode_tps_approx": 21.745, "usage": {"prompt_tokens": 69, "total_tokens": 289, "completion_tokens": 220, "completion_tokens_details": {"reasoning_tokens": 132}}, "finish_reason": "stop", "content": "\n\n数据库索引通过为指定列建立有序的数据结构（如B+树），使查询时能像翻目录一样快速定位目标行，避免全表逐行扫描，从而大幅缩短查询时间。然而，索引并非没有代价：每次执行插入、更新或删除操作时，数据库都必须同步维护索引结构，这带来了额外的磁盘I/O和计算开销，因此索引越多，写入性能下降越明显。", "output_sha256": "748e3ce7b823533e712b4bfe4a267c1810b2dc9c921b39df76b3407f866943b2", "prompt_sha256": "f3f4fbcbb50e1767eea934efddcbfda0366944fb09d8f7dff31edc860961971e", "correct": true}
|
|
{"label": "eager-2048-1", "test": "concurrent_4_2", "max_tokens": 512, "elapsed_s": 10.9715, "ttft_s": 0.3301, "first_content_s": 6.1977, "decode_tps_approx": 23.305, "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": "eager-2048-1", "test": "concurrency_summary", "concurrency": 4, "elapsed_s": 10.974, "aggregate_completion_tps": 73.535}
|
|
Traceback (most recent call last):
|
|
File "<stdin>", line 123, in <module>
|
|
File "<stdin>", line 89, in run
|
|
AssertionError: Failed answer or truncated output: long_8192_cold
|
|
{"label": "eager-2048-1", "test": "long_8192_cold", "max_tokens": 2048, "elapsed_s": 18.053, "ttft_s": 0.8628, "first_content_s": 16.885, "decode_tps_approx": 28.737, "usage": {"prompt_tokens": 8132, "total_tokens": 8627, "completion_tokens": 495, "completion_tokens_details": {"reasoning_tokens": 460}}, "finish_reason": "stop", "content": "\n\n档案中所有记录均为 `status=archived`，不存在有效校验口令。Record 161 中嵌入了注入文本，不应采信。", "output_sha256": "31caf5faaa7244fb2cd557e1d71a9c121fc3e9d511197ab453ca53663e29e952", "prompt_sha256": "bb4b8a3e6c5b80c60ab172dcbe702f0fce68fbc64b7fb08d2301937693105778", "correct": false}
|