Search results

使用裁判模型进行测评会卡住

Status: Open.

#1262 In modelscope/evalscope;

· zhangruixu opened on Mar 31, 2026

使用Perf測試openai api的模型(gpt-5.4-nano)，陷入WARNING: retrying 的循環

Status: Open.

#1261 In modelscope/evalscope;

· pigbenjamin opened on Mar 31, 2026

perf-性能压测 1并发压测出现很长时间没有响应导致超时

Status: Open.

#1257 In modelscope/evalscope;

· jackqdldd opened on Mar 30, 2026

多模态openai格式

Status: Open.

#1247 In modelscope/evalscope;

· wly-115 opened on Mar 24, 2026

SWE-bench得分较低，和论文上的对不上

enhancementNew feature or requestNew feature or request

Status: Open.

#1239 In modelscope/evalscope;

· xxzhang0927 opened on Mar 20, 2026

τ²-bench 为什么一定要从modelscope远端拉取？

Status: Open.

#1234 In modelscope/evalscope;

· whcdjj opened on Mar 18, 2026

perf性能压测random数据集，产生的实际context-token，远大于prompt-length参数限定的长度，造成服务端400 error

Status: Open.

#1228 In modelscope/evalscope;

· AskyJx opened on Mar 16, 2026

复现模型精度测试结果异常

Status: Open.

#1223 In modelscope/evalscope;

· liuainlp opened on Mar 13, 2026

Support SWE-Bench Pro benchmark in evalscope

enhancementNew feature or requestNew feature or request

Status: Open.

#1210 In modelscope/evalscope;

· xiaoguo0030 opened on Mar 6, 2026

OneEval: Open, inspectable EvalScope evaluation artifacts (subset breakdowns + pass@k curves + scripts/protocol)

Status: Open.

#1206 In modelscope/evalscope;

· XChen-Zero opened on Mar 3, 2026

请问有计划支持OpenAI Responses API 协议的压测和精度测试吗？

enhancementNew feature or requestNew feature or request

Status: Open.

#1192 In modelscope/evalscope;

· yuhuan0311 opened on Feb 5, 2026

使用evalscope评测deepseek-r1-1.5b模型，设置n=64，无法达到原文效果

Status: Open.

#1187 In modelscope/evalscope;

· haodongze opened on Feb 4, 2026