taskwatch logs: sdp-10.165.117.28 /home/sdp/lkk/humaneval_plus/results/qwen38-mxfp4-bf16ssm_humaneval_indep_suite 2026-11-08T12:34:34Z
Browse files
machines/sdp-10.165.117.28/logs/qwen38-mxfp4-bf16ssm_humaneval_indep-9456aeb7/run07.log
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[INFO] Waiting up to 300s for http://sc09super22-b200.sc.intel.com:8005/health
|
| 2 |
+
[INFO] vLLM is ready
|
| 3 |
+
curl: /home/sdp/miniforge3/lib/libcurl.so.4: no version information available (required by curl)
|
| 4 |
+
[INFO] Running HumanEval+ with model qwen38-mxfp4-bf16ssm (samples=1, temperature=0.6)
|
| 5 |
+
[INFO] EvalPlus log: /home/sdp/lkk/humaneval_plus/results/qwen38-mxfp4-bf16ssm_humaneval_indep_suite/runs/qwen38-mxfp4-bf16ssm_humaneval_indep_run07/evalplus.log
|
| 6 |
+
# HumanEval+ suite: qwen38-mxfp4-bf16ssm_humaneval_indep_run07
|
| 7 |
+
|
| 8 |
+
- Model: qwen38-mxfp4-bf16ssm
|
| 9 |
+
- Temperature: 0.6
|
| 10 |
+
- Samples per problem: 1
|
| 11 |
+
- Tasks: 164
|
| 12 |
+
|
| 13 |
+
| Sample | Status | Evaluated | Passed | pass@1 | base pass@1 |
|
| 14 |
+
|---:|---|---:|---:|---:|---:|
|
| 15 |
+
| 1 | completed | 164 | 147 | 89.63% | 92.07% |
|
| 16 |
+
|
| 17 |
+
**Mean pass@1: 89.63% ± 0.00%** (min 89.63%, max 89.63%, n=1)
|
| 18 |
+
**Mean base pass@1: 92.07% ± 0.00%**
|
| 19 |
+
**Overall pass@1: 89.63%** (147/164)
|
| 20 |
+
- Official EvalPlus pass@1 (base + extra): 89.60%
|
| 21 |
+
- Official EvalPlus base pass@1: 92.10%
|
| 22 |
+
- Flaky tasks: 0; always failed: 17
|
| 23 |
+
[INFO] Summary: /home/sdp/lkk/humaneval_plus/results/qwen38-mxfp4-bf16ssm_humaneval_indep_suite/runs/qwen38-mxfp4-bf16ssm_humaneval_indep_run07/summary.md
|
| 24 |
+
[INFO] Summary JSON: /home/sdp/lkk/humaneval_plus/results/qwen38-mxfp4-bf16ssm_humaneval_indep_suite/runs/qwen38-mxfp4-bf16ssm_humaneval_indep_run07/summary.json
|
| 25 |
+
[INFO] Samples: /home/sdp/lkk/humaneval_plus/results/qwen38-mxfp4-bf16ssm_humaneval_indep_suite/runs/qwen38-mxfp4-bf16ssm_humaneval_indep_run07/evalplus/humaneval/qwen38-mxfp4-bf16ssm_openai_temp_0.6.jsonl
|
| 26 |
+
[INFO] EvalPlus log: /home/sdp/lkk/humaneval_plus/results/qwen38-mxfp4-bf16ssm_humaneval_indep_suite/runs/qwen38-mxfp4-bf16ssm_humaneval_indep_run07/evalplus.log
|