diff --git a/.gitattributes b/.gitattributes
index 957437fca5576e20b087eaa9af60631dc3f71113..d2ed8cef28183ea2f7f821488d8404406e9afee4 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -71,3 +71,177 @@ qwq131k/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T23-40-20.886537.jsonl
 qwq131ktemp/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T23-39-30.477228.jsonl filter=lfs diff=lfs merge=lfs -text
 qwq_agg64temp1/Qwen__QwQ-32B/samples_aime24_nofigures_agg64_2025-03-11T02-54-37.245696.jsonl filter=lfs diff=lfs merge=lfs -text
 qwq_agg64temp1/Qwen__QwQ-32B/samples_aime25_nofigures_agg64_2025-03-11T02-54-37.245696.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/results_2025-04-11T18-27-08.320752.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-06T06-56-03.478011.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-14T09-49-18.086229.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T01-22-22.787541.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/results_2025-04-14T05-36-17.081647.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-03T22-06-15.102588.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-12T01-28-43.330003.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-12T01-28-43.330003.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-04T00-57-01.239273.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-14T14-17-13.604932.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-12T01-28-43.330003.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-12T17-51-44.493746.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/results_2025-04-14T06-51-57.303362.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-12T03-35-43.851203.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-12T17-51-44.493746.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-12T17-43-21.508059.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-12T19-42-30.355594.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/results_2025-04-14T07-51-00.615118.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T07-51-00.615118.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T12-07-55.948557.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/results_2025-04-09T20-56-16.072064.json filter=lfs diff=lfs merge=lfs -text
+DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime25_nofigures_agg64_2025-04-02T20-33-55.445905.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/results_2025-04-04T22-48-19.450896.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-21-17.818231.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/results_2025-04-12T17-51-44.493746.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-12T19-42-30.355594.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/results_2025-04-04T00-57-01.239273.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-07T20-20-02.007830.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T01-22-22.787541.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-10T19-16-27.462978.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-11T18-27-08.320752.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-12T03-35-43.851203.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/results_2025-04-07T20-20-02.007830.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-14T14-29-12.444170.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-04T22-48-19.450896.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-04T02-08-03.821182.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/results_2025-04-04T00-54-02.924841.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/results_2025-04-07T07-14-51.598130.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-10T19-16-27.462978.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-11T18-27-08.320752.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-12T19-42-30.355594.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/results_2025-04-08T23-35-58.097461.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-03T22-00-29.567285.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-04T00-54-02.924841.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-06T08-04-41.184191.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-08T18-31-04.038583.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-14T14-17-13.604932.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-09T20-56-16.072064.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-12T17-43-21.508059.jsonl filter=lfs diff=lfs merge=lfs -text
+DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime24_nofigures_agg64_2025-04-04T22-57-33.728468.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-07T07-14-51.598130.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/results_2025-04-10T19-16-27.462978.json filter=lfs diff=lfs merge=lfs -text
+Qwen3-0.6B/Qwen__Qwen3-0.6B/results_2025-05-01T18-39-51.154214.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/results_2025-04-04T00-41-50.574430.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-06T05-06-14.582869.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/results_2025-04-14T09-49-18.086229.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/results_2025-04-14T12-07-55.948557.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T05-36-17.081647.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-10T03-15-35.955666.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-18-25.309796.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-07T20-31-59.584821.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-20T18-56-57.562977.jsonl filter=lfs diff=lfs merge=lfs -text
+Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime24_nofigures_agg64_2025-05-01T18-39-51.154214.jsonl filter=lfs diff=lfs merge=lfs -text
+Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_gpqa_diamond_openai_agg8_2025-05-01T03-29-26.282856.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-06T06-39-11.509480.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-04T00-41-50.574430.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T23-35-58.097461.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-07T20-31-59.584821.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-08T18-58-51.625344.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/results_2025-04-14T12-31-16.950742.json filter=lfs diff=lfs merge=lfs -text
+DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-04T22-57-33.728468.json filter=lfs diff=lfs merge=lfs -text
+Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime24_nofigures_agg64_2025-05-01T03-29-26.282856.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-03T22-06-15.102588.jsonl filter=lfs diff=lfs merge=lfs -text
+DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime25_nofigures_agg64_2025-04-19T16-26-39.509062.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/results_2025-04-06T08-04-41.184191.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-06T03-57-17.586717.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T08-43-46.180116.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-09T20-56-16.072064.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/results_2025-04-12T17-43-21.508059.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-12T17-43-21.508059.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-04T00-57-01.239273.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-21-17.818231.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/results_2025-04-12T01-28-43.330003.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-06T06-56-03.478011.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/results_2025-04-14T10-31-55.643857.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-12T17-51-44.493746.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/results_2025-04-06T03-57-17.586717.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-29-56.401365.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-07T20-20-02.007830.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_aime24_nofigures_agg64_2025-04-08T18-31-04.038583.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-14T09-49-18.086229.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/results_2025-04-14T01-22-22.787541.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-03T22-00-29.567285.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-04T00-54-02.924841.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-12T03-35-43.851203.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-04T02-08-03.821182.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-08T18-31-04.038583.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T08-43-46.180116.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-14T14-29-12.444170.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-10T19-16-27.462978.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-11T18-27-08.320752.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-09T20-56-16.072064.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/results_2025-04-08T18-37-15.377720.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-09T20-45-21.052863.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-09T20-45-21.052863.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-08T18-37-15.377720.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-08T18-58-51.625344.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-14T10-31-55.643857.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T06-51-57.303362.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-06T05-36-11.778357.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/results_2025-04-07T20-31-59.584821.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/results_2025-04-03T22-06-15.102588.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-06T05-06-14.582869.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-08T18-37-15.377720.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-20T18-56-57.562977.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/results_2025-04-10T03-15-35.955666.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T06-51-57.303362.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/results_2025-04-12T19-42-30.355594.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-09T20-45-21.052863.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-04T00-41-50.574430.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/results_2025-04-14T14-17-13.604932.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/results_2025-04-20T18-56-57.562977.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-20T18-56-57.562977.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/results_2025-04-09T20-45-21.052863.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/results_2025-04-03T22-00-29.567285.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/results_2025-04-06T06-56-03.478011.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/results_2025-04-08T18-58-51.625344.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-07T07-14-51.598130.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-14T14-17-13.604932.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/results_2025-04-14T14-29-12.444170.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T12-07-55.948557.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T05-36-17.081647.jsonl filter=lfs diff=lfs merge=lfs -text
+Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime25_nofigures_agg64_2025-05-01T18-39-51.154214.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/results_2025-04-06T05-06-14.582869.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-18-25.309796.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-10T03-15-35.955666.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/results_2025-04-12T03-35-43.851203.json filter=lfs diff=lfs merge=lfs -text
+DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_gpqa_diamond_openai_agg8_2025-04-02T20-33-55.445905.jsonl filter=lfs diff=lfs merge=lfs -text
+Qwen3-0.6B/Qwen__Qwen3-0.6B/results_2025-05-01T03-29-26.282856.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/results_2025-04-14T08-43-46.180116.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T12-31-16.950742.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T06-51-57.303362.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T07-51-00.615118.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/results_2025-04-04T02-08-03.821182.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-06T05-36-11.778357.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-08T18-37-15.377720.jsonl filter=lfs diff=lfs merge=lfs -text
+Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_gpqa_diamond_openai_agg8_2025-05-01T18-39-51.154214.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-07T20-31-59.584821.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-10T03-15-35.955666.jsonl filter=lfs diff=lfs merge=lfs -text
+DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-19T16-26-39.509062.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-06T08-04-41.184191.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-14T09-49-18.086229.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T01-22-22.787541.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T12-07-55.948557.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T05-36-17.081647.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-06T06-39-11.509480.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-06T05-06-14.582869.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/results_2025-04-08T18-31-04.038583.json filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T08-43-46.180116.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T07-51-00.615118.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-14T14-29-12.444170.jsonl filter=lfs diff=lfs merge=lfs -text
+Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime25_nofigures_agg64_2025-05-01T03-29-26.282856.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-14T10-31-55.643857.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T12-31-16.950742.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-06T03-57-17.586717.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-29-56.401365.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-06T05-36-11.778357.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-07T20-20-02.007830.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-08T18-58-51.625344.jsonl filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-07T07-14-51.598130.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-14T10-31-55.643857.jsonl filter=lfs diff=lfs merge=lfs -text
+1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T12-31-16.950742.jsonl filter=lfs diff=lfs merge=lfs -text
+DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-02T20-33-55.445905.json filter=lfs diff=lfs merge=lfs -text
+s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/results_2025-04-06T06-39-11.509480.json filter=lfs diff=lfs merge=lfs -text
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/results_2025-04-14T08-43-46.180116.json b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/results_2025-04-14T08-43-46.180116.json
new file mode 100644
index 0000000000000000000000000000000000000000..f2b18b30854f6bbfc1261968568a967374cd07ee
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/results_2025-04-14T08-43-46.180116.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10683b25df9e3538b5c958c23454097ab48686b1e0245d4658f49992d2227bb0
+size 75374
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T08-43-46.180116.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T08-43-46.180116.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..d16d2acb870f75313d79acbc8fb4a25e2f6f0d64
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T08-43-46.180116.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ba0d5f818c683d32e7f7823b61804dbf28d01e80475e630c26721fd4d8de6cf
+size 155097702
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T08-43-46.180116.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T08-43-46.180116.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c8352adaa5e945919da0fc9a9973bfe2bd65211a
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T08-43-46.180116.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4d4b869526fd185d6e48c72b1af9b21433c93db3aa6881165a4701711d2538c
+size 157858015
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T08-43-46.180116.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T08-43-46.180116.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..908e77f8d224610307a9d594beeb43f01e47f16a
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T08-43-46.180116.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:417eb42b047556dd27a7722ec12e49d0403e77eb7c5ff168f6d6760b6ca3ab81
+size 92011689
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/results_2025-04-14T09-49-18.086229.json b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/results_2025-04-14T09-49-18.086229.json
new file mode 100644
index 0000000000000000000000000000000000000000..99e9921d9dcd96e736ef8185959cb872e1aca794
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/results_2025-04-14T09-49-18.086229.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59a50a3958943ba2d7b659de0fecdb894d6da2fd8c4d466043c65e7c6ce7c2f7
+size 75252
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-14T09-49-18.086229.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-14T09-49-18.086229.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..8c652cad726e7b78db6acfae1d7cb29c88913a54
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-14T09-49-18.086229.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8c626d6ada67efcb714d7e12c9f85c1681e23e8fb5a275a0a8d3cdece8b54c4
+size 150479471
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-14T09-49-18.086229.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-14T09-49-18.086229.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..80eb335b4081a4f2ea02933bb1afd0ad6e99a27c
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-14T09-49-18.086229.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71c8c8911dc50ef5562cc8d52dc6a337ab4a534092f249ce01224f8aba65909a
+size 155619293
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-14T09-49-18.086229.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-14T09-49-18.086229.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2711c86b2bff1f44287b79a1de9616c9ceb7fb2b
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-14T09-49-18.086229.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adec9a45dcf62773f61f9f12e412b12b0d10956d5ae04c51226b46302234363c
+size 89449606
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/results_2025-04-14T10-31-55.643857.json b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/results_2025-04-14T10-31-55.643857.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9d3121135ab5b4d567dba8b32b32f6ff8f42a36
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/results_2025-04-14T10-31-55.643857.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8f125ae4ab14944887f1eb6e99884811c74125e441f09856b904eb9c9ccb2a3
+size 75134
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-14T10-31-55.643857.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-14T10-31-55.643857.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..854256f1162d66b3761f7e546a8958aa28100fb3
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-14T10-31-55.643857.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:780ca269d37e102fa5be3ff0ac29e8331f404e11d058e499a0039b997bb621d0
+size 149240585
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-14T10-31-55.643857.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-14T10-31-55.643857.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..833573817c966275e69ea2fb0c0195014b1afcf0
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-14T10-31-55.643857.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c231baf97504f6d7ac0fbae2f28f29a63076b5d2c7029926c3497c68ac279d6b
+size 149340528
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-14T10-31-55.643857.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-14T10-31-55.643857.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c34dd9ee7017ba781e77173e1175560867868e49
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-14T10-31-55.643857.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34f9a5eef77d93c9e7b01f6bb6261715b3a79691c1905fe2bcbd95c96385e152
+size 82843773
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/results_2025-04-14T07-51-00.615118.json b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/results_2025-04-14T07-51-00.615118.json
new file mode 100644
index 0000000000000000000000000000000000000000..a244e9bcca47b58b6d4dc0d426783f7deca78424
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/results_2025-04-14T07-51-00.615118.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:114130e88c58286d90921a198e08cf252353ca0a80e084c1ba4dd868fb8adfc3
+size 75216
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T07-51-00.615118.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T07-51-00.615118.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..58f029917a24f001125141ddd445d9811d177628
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T07-51-00.615118.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc78bb8532b53e286c7f141abd7151dc3342f1d8dcdfdcd301ee6f0135bb5c66
+size 172551971
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T07-51-00.615118.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T07-51-00.615118.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..42c3ddba8f7bfaf96251bd7f028653c94135aba9
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T07-51-00.615118.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62047710b7630592012ddce8d6701358c791e1006356f8b1755b1dec5deafb62
+size 163550199
diff --git a/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T07-51-00.615118.jsonl b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T07-51-00.615118.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..9b4af2f760c75b118c85d8a4241e64983f7a8a3a
--- /dev/null
+++ b/1.5B-v10/__data__niklas__s2__open-r1__data__1.5B-v10__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T07-51-00.615118.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b18c7a4de047b9a434e0a20ebf237059eda97118beef4a41dc8918d4636ae3cf
+size 96809040
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/results_2025-04-14T01-22-22.787541.json b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/results_2025-04-14T01-22-22.787541.json
new file mode 100644
index 0000000000000000000000000000000000000000..1fef462f0c4f7db22ac37366b832a76cea809720
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/results_2025-04-14T01-22-22.787541.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0f04bcf0efedbe95907727a246c41bf8ff173f67a239d40271cc2ebcc9e6eb7
+size 75149
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/results_2025-04-14T12-31-16.950742.json b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/results_2025-04-14T12-31-16.950742.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f3b184e237e4cb9ff5214a6cfcc182e24aadcf5
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/results_2025-04-14T12-31-16.950742.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd53c59ff3c6f9837a47618a121132ba1dcac52786701b415627ad103eec873b
+size 75126
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T01-22-22.787541.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T01-22-22.787541.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..47419aa7945df318bbac6a95965d86e2081d121b
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T01-22-22.787541.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1675405fb64a0dbc15382f26ca3eea2237f31fac8bd97ec0a0d085f238399e02
+size 164355569
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T12-31-16.950742.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T12-31-16.950742.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..47419aa7945df318bbac6a95965d86e2081d121b
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-14T12-31-16.950742.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1675405fb64a0dbc15382f26ca3eea2237f31fac8bd97ec0a0d085f238399e02
+size 164355569
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T01-22-22.787541.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T01-22-22.787541.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..1442b412f64883493604698937b243c0ff64a489
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T01-22-22.787541.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43134510401b47107ccef1bf454a8124ef19c575715feb166b85410c0797e351
+size 161280909
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T12-31-16.950742.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T12-31-16.950742.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2f946483956dc484e49691cfe8e9183be16ee889
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-14T12-31-16.950742.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6555da62fec2ace4032ac0b50b0624321d6df367dd271d33a38dbe0051ea59d
+size 161282893
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T01-22-22.787541.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T01-22-22.787541.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ed3f1dfa9c5388e17538dc97e7ff9a9046994f6b
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T01-22-22.787541.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06b94017fd8e25b5ad4ff3e92052a8f9ac4f698f5584939f03d37389173ba337
+size 93442896
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T12-31-16.950742.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T12-31-16.950742.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..626731506724728100d4cedef366c04053ff04d8
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-14T12-31-16.950742.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ac5c8bbb9bd2175900343db550d2b0eaf4df0226cac00ca72788f489a9fb78a
+size 93442902
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/results_2025-04-14T14-17-13.604932.json b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/results_2025-04-14T14-17-13.604932.json
new file mode 100644
index 0000000000000000000000000000000000000000..3aff24f049c937199183541eb6a510be95c73be1
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/results_2025-04-14T14-17-13.604932.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b0f5d3b6a6ce8714053e448af12eb824f8a3b6411e029f6b6994ec33d736f5a
+size 75011
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-14T14-17-13.604932.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-14T14-17-13.604932.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..1da0a8281904eae5eee8ebc97cacf67f3b37f332
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-14T14-17-13.604932.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:361cc6b37179bd84846f47cc00ea4de1e07313f89206f711d0b7b63732c62b18
+size 155865300
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-14T14-17-13.604932.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-14T14-17-13.604932.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..7653e8c036a929d26ba92d8d709bf4b5813cb254
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-14T14-17-13.604932.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1b5303aae6742f58c6fe2705c2464a43bb58be566cbcb6e1bfdc5201efeec8f
+size 160326692
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-14T14-17-13.604932.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-14T14-17-13.604932.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c09620196674dda9a960acd64460bb950dd2a7ad
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-14T14-17-13.604932.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66abbf804af585634d002472116cb476180cc5ad5bf94a927576cc4c47cebca6
+size 89449405
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/results_2025-04-14T14-29-12.444170.json b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/results_2025-04-14T14-29-12.444170.json
new file mode 100644
index 0000000000000000000000000000000000000000..a5284fc916aa67e68fa4ac94c293efdc8089b6ee
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/results_2025-04-14T14-29-12.444170.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe2f62fbc9575c6cda751524b3eafed67958deafe728567437c18f59e6e3fe9c
+size 75083
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-14T14-29-12.444170.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-14T14-29-12.444170.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..f70d7bc11979276f98d2a2e8708701f2d8b3bbe4
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-14T14-29-12.444170.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d22728ad9a6f7dfda22fa3c6427472419a8814aa5725af2ca530238df397336f
+size 159665157
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-14T14-29-12.444170.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-14T14-29-12.444170.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..7563c620bf001ef470908c86e70c4cb54a245489
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-14T14-29-12.444170.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b45036313a3a6542417c059d60d09dff122ef6a1da7b73b4b30e405f6635a6a6
+size 145953146
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-14T14-29-12.444170.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-14T14-29-12.444170.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3ade210b5e9e19c781ef0bd54d7bc9668499b823
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-14T14-29-12.444170.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4cf3bd20c16226e4652cf398f80c60f0c7056c9456315f758dcf54b934ff3be
+size 87701021
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/results_2025-04-14T12-07-55.948557.json b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/results_2025-04-14T12-07-55.948557.json
new file mode 100644
index 0000000000000000000000000000000000000000..2000f576ef40c3f47fdb2dfaa9c90680791cedc4
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/results_2025-04-14T12-07-55.948557.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a2d2d8f2b0728424b481114d88005eac9732055c2d05f4ebe275024e0826318
+size 75093
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T12-07-55.948557.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T12-07-55.948557.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..1f562856ae797fc8b1193c48d8e595a7bc020144
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T12-07-55.948557.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c988e487ab00765ba1e53558b955d8ce163ddc152ba59f3515ae3ebe06060739
+size 169089120
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T12-07-55.948557.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T12-07-55.948557.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..0b85c9c9227a9874c2c0b8e63e24e7c4e4beec6a
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T12-07-55.948557.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eaea6990be1d68b7991e96ffaa5566c9196732c860ea52bf843a09da0df2908a
+size 175290251
diff --git a/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T12-07-55.948557.jsonl b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T12-07-55.948557.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..99b2718d6f10bf9a94a7f6ecfaec4b3bbb0da50c
--- /dev/null
+++ b/1.5B-v11/__data__niklas__s2__open-r1__data__1.5B-v11__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T12-07-55.948557.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce1adfa677359392db1413b33337b988e59a3cca32d63e2ccee463a17955fc62
+size 91646959
diff --git a/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/results_2025-04-20T18-56-57.562977.json b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/results_2025-04-20T18-56-57.562977.json
new file mode 100644
index 0000000000000000000000000000000000000000..938ea13fd5b2cebb43cef483db116412db2668d2
--- /dev/null
+++ b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/results_2025-04-20T18-56-57.562977.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:037470bbe24433c01263285dc93f76294ea8b7c1c6fbc58fb30e6daf871d19db
+size 78187
diff --git a/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-20T18-56-57.562977.jsonl b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-20T18-56-57.562977.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..0b20815ef925587e508a854dada30c62a89a0fdd
--- /dev/null
+++ b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-20T18-56-57.562977.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54090a60e18ddc5d22167db24f4e82479c3eb3839e09b55d9d762755272a5542
+size 183035273
diff --git a/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-20T18-56-57.562977.jsonl b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-20T18-56-57.562977.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..4acfee3cc8931c0b5afde3d6712cdf65379f65f6
--- /dev/null
+++ b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-20T18-56-57.562977.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e69f6122b6a272d3e9e315d9a33e99b6af386007b6a5132651eafa383831621
+size 185891775
diff --git a/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-20T18-56-57.562977.jsonl b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-20T18-56-57.562977.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..164980e6122fd9beffbab0ac64c4c107744533cc
--- /dev/null
+++ b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-20T18-56-57.562977.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c06cfea5a0b35a25b3fe6dc6686f6e1fe8a9dd9b1c4116e9b8b632574170f96
+size 119029811
diff --git a/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/results_2025-04-14T05-36-17.081647.json b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/results_2025-04-14T05-36-17.081647.json
new file mode 100644
index 0000000000000000000000000000000000000000..d860394f1b0d01e4b64bec060db2a5f0389ff02f
--- /dev/null
+++ b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/results_2025-04-14T05-36-17.081647.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4d3175eddadd8caff570ea8fee5f6b20e0cb0e42f4843a5173fb4ae99c3735a
+size 75459
diff --git a/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T05-36-17.081647.jsonl b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T05-36-17.081647.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2542fe2ad684b233d5fe8d5b113165f28a1d8a71
--- /dev/null
+++ b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T05-36-17.081647.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e1884abce3f0f302cae297f2e1bff422e772b24e9e0f5035f2e8754b1c430f8
+size 174392211
diff --git a/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T05-36-17.081647.jsonl b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T05-36-17.081647.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..4689888cf15d7a9eb9df8d08781858c3504e56ff
--- /dev/null
+++ b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T05-36-17.081647.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c81b5b95d22523b7d85b70f66e31f2b80bd473d2f67a71e678196fbc92cf50b4
+size 174407215
diff --git a/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T05-36-17.081647.jsonl b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T05-36-17.081647.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..b686bdb5c01c12285bfca619e206a2221bb84f10
--- /dev/null
+++ b/1.5B-v12/__data__niklas__s2__open-r1__data__1.5B-v12__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T05-36-17.081647.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34bc6a9e1e1206c74f4ba3cdbf5a3436085876af5d4345307c54d1798a5a9bfd
+size 111387692
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/results_2025-04-09T20-56-16.072064.json b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/results_2025-04-09T20-56-16.072064.json
new file mode 100644
index 0000000000000000000000000000000000000000..45f2b85e2217d117adf1c6caa6441451627331ab
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/results_2025-04-09T20-56-16.072064.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ae67dd5d14c259baeaea401d40326a83f7c8d482072777aca48b190a2b34643
+size 75256
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-09T20-56-16.072064.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-09T20-56-16.072064.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2de91f1d75a9f8f073f7a4e93e8e4589942d7927
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-09T20-56-16.072064.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ec924c7ca229fff2646c3d0f0bcbca6101ee6079336a4387add9a2fbd501b2f
+size 168288165
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-09T20-56-16.072064.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-09T20-56-16.072064.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..00db79671aa0368233608e3aa15b5a979feb9961
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-09T20-56-16.072064.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c8dd62ca7fdbd8d7290a251f55e954da0d3d3c27245e1c2e3dc2cf3ec4261ac
+size 176163012
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-09T20-56-16.072064.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-09T20-56-16.072064.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..85be04f3eff4505e6c653e970b2554f2fac8d8cc
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-09T20-56-16.072064.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67d90c12f1d33f1c79d66f75744b72fad2305c68656ce907ff68065b487c4b9e
+size 155782285
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/results_2025-04-10T03-15-35.955666.json b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/results_2025-04-10T03-15-35.955666.json
new file mode 100644
index 0000000000000000000000000000000000000000..680d51a85ed4576216d394cfa6798a2a25151eef
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/results_2025-04-10T03-15-35.955666.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16c5ad2893a7d637f0a247c822c1a7b8f9d702e6775b38d2677c94e06ad0deaf
+size 74904
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-10T03-15-35.955666.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-10T03-15-35.955666.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6e564e874109c7078f64e687a40e18959cf2b142
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-10T03-15-35.955666.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c9cb49501b56f887007b2d33d601fac3bcdad73858ab50f34ca15f053e1e164
+size 185416742
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-10T03-15-35.955666.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-10T03-15-35.955666.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..d7bb0acfcbbf5b382c1ad2c174af655697a6b58a
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-10T03-15-35.955666.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2190f8e67c60f8610d1ced83316ed946e8b8367d2c314f775dec75632f5a7247
+size 179498621
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-10T03-15-35.955666.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-10T03-15-35.955666.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..44a7b7c7fe58100f3818c6e0bc23010dcf67435a
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-10T03-15-35.955666.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0bb41f7b1d3a56f09ba4167ee2d4c43a03184462a9d12e735946db475d7c6ad7
+size 131109314
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/results_2025-04-10T19-16-27.462978.json b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/results_2025-04-10T19-16-27.462978.json
new file mode 100644
index 0000000000000000000000000000000000000000..f382f40ea51199fc96227c0a8d8838be64f8d15c
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/results_2025-04-10T19-16-27.462978.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6404b6966cd6265d482db45b2aeb5fbd90fd6b4fd32f098e6daf67ab057bcd45
+size 75128
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-10T19-16-27.462978.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-10T19-16-27.462978.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..12c2504f30a24be2df0359320a39a9a68b026fb8
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-10T19-16-27.462978.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63d055ba33fa7b4b46afca9b7fe5a6c375d400a6bde8cf8238889e65f10f9acc
+size 168750424
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-10T19-16-27.462978.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-10T19-16-27.462978.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..e5512e5823e639a66fa3896bbeeccca30b66a2ab
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-10T19-16-27.462978.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e070373828731dd584c615aad9d0f369b56ad8ce0c6433f9097eec33893b463f
+size 196271174
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-10T19-16-27.462978.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-10T19-16-27.462978.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..a0e5dda6c6691368fffee2fd234594c7f2cdb044
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-10T19-16-27.462978.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d85dc9dc5d36ed08df1411c5aaf941e61db76f81a7cfac52fce11cfa5213d7dc
+size 159444267
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/results_2025-04-09T20-45-21.052863.json b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/results_2025-04-09T20-45-21.052863.json
new file mode 100644
index 0000000000000000000000000000000000000000..9909ba917955fd5eea9f37993b6ceca9e6159415
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/results_2025-04-09T20-45-21.052863.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:051e8e8891d84bb5db221dcfd0b8fd19a797b3be51562df8f5850f93a74b01e6
+size 74970
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-09T20-45-21.052863.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-09T20-45-21.052863.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..82c29daf5310ef8bdd3b4d7a7007328b2a2247a5
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-09T20-45-21.052863.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dc4e3fcbd1866503cd49ba868c7e66a319b17ea0a960be8353cd880e76261af
+size 179751008
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-09T20-45-21.052863.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-09T20-45-21.052863.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..e9237b12f95951f3d2f53c28ad8f230c9adb0d0c
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-09T20-45-21.052863.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90c88fc73e1e9620daf6359e67f6db4f3a7444f7e252148b9fb231a206430505
+size 182815291
diff --git a/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-09T20-45-21.052863.jsonl b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-09T20-45-21.052863.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..5d9a6d96945a0819cf62aee77c8862f3c7558fae
--- /dev/null
+++ b/1.5B-v7/__data__niklas__s2__open-r1__data__1.5B-v7__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-09T20-45-21.052863.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:813ecaaef726f6d897281339f1213fd63cbaa23e6980450858bc7463ca55817e
+size 125419118
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/results_2025-04-11T18-27-08.320752.json b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/results_2025-04-11T18-27-08.320752.json
new file mode 100644
index 0000000000000000000000000000000000000000..629cfbfa3decb52c31bceca2ce02dcaf4c747caa
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/results_2025-04-11T18-27-08.320752.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73a8fd9775157d03b29a021d0c05dfeda60a859991bd81852c03b7d434438828
+size 75253
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-11T18-27-08.320752.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-11T18-27-08.320752.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..836533aaf8c554c006879ba73e7e30436ea53cdc
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-11T18-27-08.320752.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc30fcb1037eaaf459147480b1a080085930acb2e79090f64946fe94273eb9e4
+size 157488219
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-11T18-27-08.320752.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-11T18-27-08.320752.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..69073f10332d6e2ecc66727e93c6be0c6f18ab40
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-11T18-27-08.320752.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abf8fc18f5949b5b8c34b209bc82a76310e78b72b037decc717b5c4a7f3277d1
+size 158541349
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-11T18-27-08.320752.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-11T18-27-08.320752.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..b7f684752786868d47eadf1292be477b354788a5
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-11T18-27-08.320752.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ec1fc05307ce337216afbef3bd8d8adabb51c8ea65da329b10a1992badf5f5c
+size 87746030
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/results_2025-04-12T01-28-43.330003.json b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/results_2025-04-12T01-28-43.330003.json
new file mode 100644
index 0000000000000000000000000000000000000000..02324079fc0a784054374afb6720beb175e0883c
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/results_2025-04-12T01-28-43.330003.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8e210c2140ed8c59131f7ed09182ad17b7c0b82fb0b7b78122bf9d66e76a80f
+size 75271
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-12T01-28-43.330003.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-12T01-28-43.330003.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..49c9e9ed910aab5f503fdc389f2caf4565d4b8e9
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-12T01-28-43.330003.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90a9eebf075392b2618c4880f152d347c5e22dc3ab7547fad4ccbea0c42105c5
+size 155064967
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-12T01-28-43.330003.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-12T01-28-43.330003.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..d14ee34bf2a936fd29d22842cce496cafdf6a626
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-12T01-28-43.330003.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b14f9b28cad5710c8e8313e98b6665ab41ddcb3eb87d1dd0ab537c451c19fd8
+size 159158347
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-12T01-28-43.330003.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-12T01-28-43.330003.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..1614ff28cd611257221f36d4bfd1b493ccd79eb0
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-12T01-28-43.330003.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5f5b3a1c7561cbd7e9410f9b67e187687a5ebcf3a51b5ced926aa96572511d8
+size 95813114
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/results_2025-04-12T03-35-43.851203.json b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/results_2025-04-12T03-35-43.851203.json
new file mode 100644
index 0000000000000000000000000000000000000000..72605f7f651f73452bb3b72b3386e44686f6baaa
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/results_2025-04-12T03-35-43.851203.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28b79bc290bf61f98d6e6ff86927be59d1bcbbe63cd11f1e46564c0ba89c45cb
+size 75013
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-12T03-35-43.851203.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-12T03-35-43.851203.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3d435d8580d2733920470e502c8aa0e0d687e14b
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-12T03-35-43.851203.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d183393f27cb5641c20af734762ba84263ec537aaee701fac4985a4fede8a0f
+size 154786582
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-12T03-35-43.851203.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-12T03-35-43.851203.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6477ea324854250b8c1cc004c71b4a7ffcb61a09
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-12T03-35-43.851203.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce29ccea0f36942c1a28f605e2f9e020d59241e44446060b3d74b6929f9625dc
+size 159077445
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-12T03-35-43.851203.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-12T03-35-43.851203.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..e7d2a86d6b30299372c6db09cf8ea32fcecc6715
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-12T03-35-43.851203.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc7bcba55fb0f8fd33779b96cbef3667168f2bf3e841c27933bbd1088a8c0c5e
+size 94444375
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/results_2025-04-12T17-51-44.493746.json b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/results_2025-04-12T17-51-44.493746.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9d8ef6942d279d3b6d5a5b7121d343adfdd508b
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/results_2025-04-12T17-51-44.493746.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e0a21f3693c718039a13ef2ddef4cfa600b6a108e738fb87deadbe6724388d2
+size 75349
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-12T17-51-44.493746.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-12T17-51-44.493746.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..278d5586b4345bfadefbc5d98556284ed8e84fce
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-12T17-51-44.493746.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91dbfea13f039173eabcdcc9009e00e1de1da27a7855877aef2e397b70ec3017
+size 163125225
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-12T17-51-44.493746.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-12T17-51-44.493746.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..042e7a9a7cb0c9f51494bd397aa6711a5d0cfe54
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-12T17-51-44.493746.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58de2ef1b4e2ee73e697deedf01a6eb07a50bc08eecf887166ddf68cf88cf310
+size 159111885
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-12T17-51-44.493746.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-12T17-51-44.493746.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..fbb9eb7a249782bca7a19c9ecb466d87209b1415
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-12T17-51-44.493746.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5ae8e4517ecf5ef1ff2c7821329d1e18c2054c07a56867cac9a39e4d6eef80
+size 96015368
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/results_2025-04-14T06-51-57.303362.json b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/results_2025-04-14T06-51-57.303362.json
new file mode 100644
index 0000000000000000000000000000000000000000..f8b97ccbbbca5365d4993cb0c810a038e8c1f486
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/results_2025-04-14T06-51-57.303362.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27a9a1e8382a6f46a4279cf7d8b77be8f0fc6c69477999428224ed584539d545
+size 74986
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T06-51-57.303362.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T06-51-57.303362.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..09f976db03fd28fa882cc6eec886b09712bc9ed3
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-14T06-51-57.303362.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca76e68f9a023f23465853e5270986b0d7079c8496160c54d257ece186247e83
+size 172028791
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T06-51-57.303362.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T06-51-57.303362.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2d24c92fcdf960cd31c5d577b67e05d73e7cbdea
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-14T06-51-57.303362.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:580125eae95b57b3c95c5118c86a0e1776a2e13a08c4bbb4f0763780ca60d510
+size 161374745
diff --git a/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T06-51-57.303362.jsonl b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T06-51-57.303362.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..8a6beddc66822faeb4ee0f57612f05f19b13897d
--- /dev/null
+++ b/1.5B-v8/__data__niklas__s2__open-r1__data__1.5B-v8__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-14T06-51-57.303362.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c334c26d3b8a5fc571d48d07283ff632282c259cbd179d13d06d08a51c1ca5d0
+size 94227392
diff --git a/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/results_2025-04-12T17-43-21.508059.json b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/results_2025-04-12T17-43-21.508059.json
new file mode 100644
index 0000000000000000000000000000000000000000..f496ccae2b7f9113ec8cfddc9bf3fedfd627ac5c
--- /dev/null
+++ b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/results_2025-04-12T17-43-21.508059.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1550337b38cc9dc5f49b2225bafe44c95a827a6f39adb31ee48924d6b0ebfec1
+size 75324
diff --git a/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-12T17-43-21.508059.jsonl b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-12T17-43-21.508059.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..4ccfe1490494707c2479e3a6bd5be8b182d96d4f
--- /dev/null
+++ b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-12T17-43-21.508059.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8239c8b0590439272a2fa04c109cf68b5d1231a936f532903a4352eff3cea96a
+size 161200560
diff --git a/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-12T17-43-21.508059.jsonl b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-12T17-43-21.508059.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..aef0fb0480a2e499de7bd79bfd9b4c83e30124db
--- /dev/null
+++ b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-12T17-43-21.508059.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04b92718cabefe8b59466eafe53e0130911f8e17fbf0d311e52ecee5b535e588
+size 164368575
diff --git a/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-12T17-43-21.508059.jsonl b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-12T17-43-21.508059.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..54bb4b70747784c12273f57221dd2877b84dc3de
--- /dev/null
+++ b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-12T17-43-21.508059.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bb0821894b242a20eb26a83737e652c814a101d89870cba7ed3064bb2274504
+size 89460528
diff --git a/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/results_2025-04-12T19-42-30.355594.json b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/results_2025-04-12T19-42-30.355594.json
new file mode 100644
index 0000000000000000000000000000000000000000..fffd40e1e3c1cdc13f3aae668cbde1da65689b6f
--- /dev/null
+++ b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/results_2025-04-12T19-42-30.355594.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:860d746ef91ff956c9fb6f4fd259ab637ae8c2872dd175fde227c845bebc26f2
+size 75230
diff --git a/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-12T19-42-30.355594.jsonl b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-12T19-42-30.355594.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..77590195c67e78240d61b3c559d65b941d0e586d
--- /dev/null
+++ b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-12T19-42-30.355594.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2350980964452da1a7a59f2f6e4e0a867b9e877885b268059abee0c483fbf6c
+size 162253259
diff --git a/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-12T19-42-30.355594.jsonl b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-12T19-42-30.355594.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6b14688cc3f4aec0f2d9d439f81c8387ed67d4b4
--- /dev/null
+++ b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-12T19-42-30.355594.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c27100f117428eae3ea296fe27a1c1bc13900112cc7cea29161d24efa91836b0
+size 169490938
diff --git a/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-12T19-42-30.355594.jsonl b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-12T19-42-30.355594.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ea1d4c0b30254327c56c3c366c11ae2dbeebbc62
--- /dev/null
+++ b/1.5B-v9/__data__niklas__s2__open-r1__data__1.5B-v9__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-12T19-42-30.355594.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46d32ca2c08e30b48dcd97967dee86ba776e812fd24eeec020c318142e4bd4a9
+size 96221904
diff --git a/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-02T20-33-55.445905.json b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-02T20-33-55.445905.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b93359630aed980620f8da708019ac4f1036a8c
--- /dev/null
+++ b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-02T20-33-55.445905.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88d5728297d7eb6da403fdee459d4af73f7184b7167bcc0b00270f3b6dcd8cf6
+size 55441
diff --git a/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-04T22-57-33.728468.json b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-04T22-57-33.728468.json
new file mode 100644
index 0000000000000000000000000000000000000000..7ed557b737ece1fad9372da1fb0019e9fd1f48eb
--- /dev/null
+++ b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-04T22-57-33.728468.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:867687ab648f983a1275bec504d92f01f5e1a322c67ec1bfd6ab639171863e5b
+size 27736
diff --git a/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-19T16-26-39.509062.json b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-19T16-26-39.509062.json
new file mode 100644
index 0000000000000000000000000000000000000000..f58e237ee3ce51c4936068b1fe42700caf00989d
--- /dev/null
+++ b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/results_2025-04-19T16-26-39.509062.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac0e77ed92f9f6eacad0bdc743e143181808fbebb4aa87cdf18062be3776808c
+size 29351
diff --git a/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime24_nofigures_agg64_2025-04-04T22-57-33.728468.jsonl b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime24_nofigures_agg64_2025-04-04T22-57-33.728468.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2e8ecd06a9cab52e748fb3c4a3c9cec5fa52c24c
--- /dev/null
+++ b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime24_nofigures_agg64_2025-04-04T22-57-33.728468.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5f257eb8d264fc4d9463f9aad3c0c72ee3484c92a24d2233da183e57a149b77
+size 170599692
diff --git a/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime25_nofigures_agg64_2025-04-02T20-33-55.445905.jsonl b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime25_nofigures_agg64_2025-04-02T20-33-55.445905.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..e64c03b1f0b69d4fbbfd3ee6992cb7a4be030b39
--- /dev/null
+++ b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime25_nofigures_agg64_2025-04-02T20-33-55.445905.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d380314027de7f194a81d2b783686a8fe89e71f0e971344132a82efd84781e9
+size 186979912
diff --git a/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime25_nofigures_agg64_2025-04-19T16-26-39.509062.jsonl b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime25_nofigures_agg64_2025-04-19T16-26-39.509062.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ea2d64efe08dcaebd62985d2ecf9fa9d36aed755
--- /dev/null
+++ b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_aime25_nofigures_agg64_2025-04-19T16-26-39.509062.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9935607e5ad7327f8b107a1bd4af0dc449c0449f5f01f26215d9fca435bbe2cb
+size 174141053
diff --git a/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_gpqa_diamond_openai_agg8_2025-04-02T20-33-55.445905.jsonl b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_gpqa_diamond_openai_agg8_2025-04-02T20-33-55.445905.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..c8200aedd88854653874eac5c1476d24b23f0fda
--- /dev/null
+++ b/DeepSeek-R1-Distill-Qwen-1.5B/deepseek-ai__DeepSeek-R1-Distill-Qwen-1.5B/samples_gpqa_diamond_openai_agg8_2025-04-02T20-33-55.445905.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a56d8697afffefbc0aab719583086d63e1e7e0431ddc2f2fe3187c08a8f1f562
+size 115794649
diff --git a/Qwen3-0.6B/Qwen__Qwen3-0.6B/results_2025-05-01T03-29-26.282856.json b/Qwen3-0.6B/Qwen__Qwen3-0.6B/results_2025-05-01T03-29-26.282856.json
new file mode 100644
index 0000000000000000000000000000000000000000..1f3c48c7e3f0aa1a29f1d877f472cc33335dd40f
--- /dev/null
+++ b/Qwen3-0.6B/Qwen__Qwen3-0.6B/results_2025-05-01T03-29-26.282856.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:090942ff8f28207abbd10533a03e76eab122accb41caf75d954e6b95c93fc9e9
+size 57608
diff --git a/Qwen3-0.6B/Qwen__Qwen3-0.6B/results_2025-05-01T18-39-51.154214.json b/Qwen3-0.6B/Qwen__Qwen3-0.6B/results_2025-05-01T18-39-51.154214.json
new file mode 100644
index 0000000000000000000000000000000000000000..f34970a1a9251d2bb1b6baae7190efe9989d8681
--- /dev/null
+++ b/Qwen3-0.6B/Qwen__Qwen3-0.6B/results_2025-05-01T18-39-51.154214.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:047f25df6cd665fdd208ea89a4cfe1eb578303013fd382e6815b2226f522662c
+size 79755
diff --git a/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime24_nofigures_agg64_2025-05-01T03-29-26.282856.jsonl b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime24_nofigures_agg64_2025-05-01T03-29-26.282856.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..36484c4e1041df483cf9ecbbdb628c6c527efd70
--- /dev/null
+++ b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime24_nofigures_agg64_2025-05-01T03-29-26.282856.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:817065cedd1e9a430fcb204c42679a700fc06baf7f9e2acfc70991c0b2feadef
+size 5294646
diff --git a/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime24_nofigures_agg64_2025-05-01T18-39-51.154214.jsonl b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime24_nofigures_agg64_2025-05-01T18-39-51.154214.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..4b5fe63152348b60616e63a2c7f9e05494934e7b
--- /dev/null
+++ b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime24_nofigures_agg64_2025-05-01T18-39-51.154214.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38bea1caee1ce07739d24c148056bbe5eacf2f54ab0e30aa951c4fae177517c6
+size 194888128
diff --git a/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime25_nofigures_agg64_2025-05-01T03-29-26.282856.jsonl b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime25_nofigures_agg64_2025-05-01T03-29-26.282856.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ecae661c09ae6dc8935e30e4097f499925b8e2d4
--- /dev/null
+++ b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime25_nofigures_agg64_2025-05-01T03-29-26.282856.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a738a226967cdb1441d97c2b41796bc95ffad6c44b7553c01356118d776eabb7
+size 1358192
diff --git a/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime25_nofigures_agg64_2025-05-01T18-39-51.154214.jsonl b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime25_nofigures_agg64_2025-05-01T18-39-51.154214.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..a42877bfc290bd571efbd7d0c1a80ee5baa24b7e
--- /dev/null
+++ b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_aime25_nofigures_agg64_2025-05-01T18-39-51.154214.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c7014841a5fbfc6fbbd84254c5a1950e6e972b92f62822203c1f2f23e0109d3
+size 176441127
diff --git a/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_gpqa_diamond_openai_agg8_2025-05-01T03-29-26.282856.jsonl b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_gpqa_diamond_openai_agg8_2025-05-01T03-29-26.282856.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..8eb6734da88f4bfa69a9dc299c0f8b92e471894b
--- /dev/null
+++ b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_gpqa_diamond_openai_agg8_2025-05-01T03-29-26.282856.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc54a55ac8a3148fe8d174272da472d0c96d3f3095c458c90be73f4f66e8e5d5
+size 341793
diff --git a/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_gpqa_diamond_openai_agg8_2025-05-01T18-39-51.154214.jsonl b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_gpqa_diamond_openai_agg8_2025-05-01T18-39-51.154214.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..54ecbea7be38cb25e0fcaebd4b4ace6ec803345c
--- /dev/null
+++ b/Qwen3-0.6B/Qwen__Qwen3-0.6B/samples_gpqa_diamond_openai_agg8_2025-05-01T18-39-51.154214.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3920a644977545ffaef4b92a0826630413b7c40bc56a5d60573bd1bf55f3f9f1
+size 80929137
diff --git a/qwen_agg64temp1/Qwen__Qwen2.5-32B-Instruct/results_2025-03-10T00-40-46.825589.json b/qwen_agg64temp1/Qwen__Qwen2.5-32B-Instruct/results_2025-03-10T00-40-46.825589.json
deleted file mode 100644
index 1ab0f46bdadb9aa3e9b6d37623a409a446b8b50c..0000000000000000000000000000000000000000
--- a/qwen_agg64temp1/Qwen__Qwen2.5-32B-Instruct/results_2025-03-10T00-40-46.825589.json
+++ /dev/null
@@ -1,418 +0,0 @@
-{
-  "results": {
-    "aime24_nofigures_agg64": {
-      "alias": "aime24_nofigures_agg64",
-      "exact_match,all": 0.2,
-      "exact_match_stderr,all": "N/A",
-      "extracted_answers,all": -1,
-      "extracted_answers_stderr,all": "N/A",
-      "exact_matches,all": -1,
-      "exact_matches_stderr,all": "N/A",
-      "cov@2,all": 0.23333333333333334,
-      "cov@2_stderr,all": "N/A",
-      "cov@4,all": 0.3,
-      "cov@4_stderr,all": "N/A",
-      "cov@8,all": 0.43333333333333335,
-      "cov@8_stderr,all": "N/A",
-      "cov@16,all": 0.5,
-      "cov@16_stderr,all": "N/A",
-      "cov@32,all": 0.5,
-      "cov@32_stderr,all": "N/A",
-      "cov@64,all": 0.5333333333333333,
-      "cov@64_stderr,all": "N/A",
-      "maj@2,all": 0.2,
-      "maj@2_stderr,all": "N/A",
-      "maj@4,all": 0.2,
-      "maj@4_stderr,all": "N/A",
-      "maj@8,all": 0.16666666666666666,
-      "maj@8_stderr,all": "N/A",
-      "maj@16,all": 0.16666666666666666,
-      "maj@16_stderr,all": "N/A",
-      "maj@32,all": 0.23333333333333334,
-      "maj@32_stderr,all": "N/A",
-      "maj@64,all": 0.23333333333333334,
-      "maj@64_stderr,all": "N/A"
-    },
-    "aime25_nofigures_agg64": {
-      "alias": "aime25_nofigures_agg64",
-      "exact_match,all": 0.06666666666666667,
-      "exact_match_stderr,all": "N/A",
-      "extracted_answers,all": -1,
-      "extracted_answers_stderr,all": "N/A",
-      "exact_matches,all": -1,
-      "exact_matches_stderr,all": "N/A",
-      "cov@2,all": 0.1,
-      "cov@2_stderr,all": "N/A",
-      "cov@4,all": 0.16666666666666666,
-      "cov@4_stderr,all": "N/A",
-      "cov@8,all": 0.23333333333333334,
-      "cov@8_stderr,all": "N/A",
-      "cov@16,all": 0.36666666666666664,
-      "cov@16_stderr,all": "N/A",
-      "cov@32,all": 0.4,
-      "cov@32_stderr,all": "N/A",
-      "cov@64,all": 0.4,
-      "cov@64_stderr,all": "N/A",
-      "maj@2,all": 0.06666666666666667,
-      "maj@2_stderr,all": "N/A",
-      "maj@4,all": 0.1,
-      "maj@4_stderr,all": "N/A",
-      "maj@8,all": 0.13333333333333333,
-      "maj@8_stderr,all": "N/A",
-      "maj@16,all": 0.13333333333333333,
-      "maj@16_stderr,all": "N/A",
-      "maj@32,all": 0.2,
-      "maj@32_stderr,all": "N/A",
-      "maj@64,all": 0.23333333333333334,
-      "maj@64_stderr,all": "N/A"
-    }
-  },
-  "group_subtasks": {
-    "aime24_nofigures_agg64": [],
-    "aime25_nofigures_agg64": []
-  },
-  "configs": {
-    "aime24_nofigures_agg64": {
-      "task": "aime24_nofigures_agg64",
-      "tag": [
-        "math_word_problems"
-      ],
-      "dataset_path": "simplescaling/aime24_nofigures",
-      "dataset_name": "default",
-      "test_split": "train",
-      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        solution = doc.get(\"solution\", doc.get(\"orig_solution\", doc.get(\"orig_orig_solution\")))\n        problem = doc.get(\"problem\", doc.get(\"question\"))\n        answer = doc.get(\"answer\", doc.get(\"orig_answer\", doc.get(\"orig_orig_answer\")))\n        if solution is None:\n            print(\"Warning: No solution found; DOC:\", doc)\n        out_doc = {\n            \"problem\": problem,\n            \"solution\": solution,\n            \"answer\": answer,\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n    return dataset.map(_process_doc)\n",
-      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return QUERY_TEMPLATE.format(Question=doc.get(\"problem\", doc.get(\"question\")))\n",
-      "doc_to_target": "answer",
-      "unsafe_code": false,
-      "process_results": "def process_results(doc: dict, results: List[str]) -> Dict[str, int]:\n    metrics = {\"exact_match\": None, \"extracted_answers\": []}\n    # bp()\n    # Multiple results -> we are measuring cov/maj etc\n    if isinstance(results[0], list):\n        results = results[0]\n        n_res = len(results) # e.g. 64\n        n_res_list = [2**i for i in range(1, int(n_res.bit_length()))] # e.g. [2, 4, 8, 16, 32, 64]\n        metrics = {\n            **metrics,\n            \"exact_matches\": [],\n            **{f\"cov@{n}\": -1 for n in n_res_list},\n            **{f\"maj@{n}\": -1 for n in n_res_list},\n        }\n\n    if os.getenv(\"PROCESSOR\", \"\") == \"gpt-4o-mini\":\n        sampler = ChatCompletionSampler(model=\"gpt-4o-mini\")\n    else:\n        print(f\"Unknown processor: {os.getenv('PROCESSOR')}; set 'PROCESSOR=gpt-4o-mini' and 'OPENAI_API_KEY=YOUR_KEY' for best results.\")\n        sampler = None\n\n    if isinstance(doc[\"answer\"], str) and doc[\"answer\"].isdigit():\n        gt = str(int(doc[\"answer\"])) # 023 -> 23\n    else:\n        gt = str(doc[\"answer\"])\n    split_tokens = [\"<|im_start|>answer\\n\", \"<|im_start|>\"]\n\n    for i, a in enumerate(results, start=1):\n        if split_tokens[0] in a:\n            a = a.split(split_tokens[0])[-1]\n        elif split_tokens[1] in a:\n            a = a.split(split_tokens[1])[-1]\n            if \"\\n\" in a:\n                a = \"\\n\".join(a.split(\"\\n\")[1:])\n\n        if (box := last_boxed_only_string(a)) is not None:\n            a = remove_boxed(box)\n        # re.DOTALL is key such that newlines are included e.g. if it does `Answer: Here is the solution:\\n\\n10`\n        elif (matches := re.findall(ANSWER_PATTERN, a, re.DOTALL)) != []:\n            a = matches[-1]  # Get the last match\n\n        # AIME answers are from 000 to 999 so often it is a digit anyways\n        if (a.isdigit()) and (gt.isdigit()):\n            a = str(int(a)) # 023 -> 23\n        elif sampler is not None:\n            options = [gt] + list(set(metrics[\"extracted_answers\"]) - {gt})\n            if len(options) > 7:\n                # Could switch back to exact returning like in AIME in that case\n                # Problem with exact returning is that it sometimes messes up small things like a dollar sign\n                print(\"Warning: Lots of options which may harm indexing performance:\", options)            \n            # This ensures that if doc['answer'] is \\text{Evelyn} it is represented as such and not \\\\text{Evelyn}\n            options_str = \"[\" + \", \".join([\"'\" + str(o) + \"'\" for o in options]) + \"]\"\n            # a = extract_answer(sampler, options, a)\n            idx = extract_answer_idx(sampler, options_str, a)\n            if idx != \"-1\":\n                if idx.isdigit():\n                    idx = int(idx) - 1\n                    if len(options) > idx >= 0:\n                        a = options[idx]\n                    else:\n                        print(\"Warning: Index out of bounds; leaving answer unchanged\\n\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n                else:\n                    print(\"Warning: Processing did not produce integer index\\na\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n        else:\n            pass # TODO: Maybe add back legacy processing\n\n        metrics[\"extracted_answers\"].append(a)\n        a = int(a == gt)\n        if not(a): # Optional logging\n            print(\"Marked incorrect\\na \" + metrics[\"extracted_answers\"][-1] + \"\\ndoc['answer'] \" + gt)\n        if i == 1:\n            metrics[\"exact_match\"] = a\n            if \"exact_matches\" in metrics:\n                metrics[\"exact_matches\"].append(a)\n        elif i > 1:\n            metrics[\"exact_matches\"].append(a)\n            if i in n_res_list:\n                metrics[f\"cov@{i}\"] = int(1 in metrics[\"exact_matches\"])\n                metrics[f\"maj@{i}\"] = int(gt == Counter(metrics[\"extracted_answers\"]).most_common(1)[0][0])\n\n    return metrics\n",
-      "description": "",
-      "target_delimiter": " ",
-      "fewshot_delimiter": "\n\n",
-      "num_fewshot": 0,
-      "metric_list": [
-        {
-          "metric": "exact_match",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@64",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@32",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@16",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@8",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@4",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@2",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@64",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@32",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@16",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@8",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@4",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@2",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "extracted_answers",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        },
-        {
-          "metric": "exact_matches",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        }
-      ],
-      "output_type": "generate_until",
-      "generation_kwargs": {
-        "until": [],
-        "do_sample": false,
-        "temperature": 1,
-        "max_gen_toks": 32768
-      },
-      "repeats": 64,
-      "filter_list": [
-        {
-          "name": "all",
-          "filter": [
-            {
-              "function": "take_first_k",
-              "k": 64
-            }
-          ]
-        }
-      ],
-      "should_decontaminate": false,
-      "metadata": {
-        "version": 1.0
-      }
-    },
-    "aime25_nofigures_agg64": {
-      "task": "aime25_nofigures_agg64",
-      "tag": [
-        "math_word_problems"
-      ],
-      "dataset_path": "simplescaling/aime25_nofigures",
-      "dataset_name": "default",
-      "test_split": "train",
-      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        solution = doc.get(\"solution\", doc.get(\"orig_solution\", doc.get(\"orig_orig_solution\")))\n        problem = doc.get(\"problem\", doc.get(\"question\"))\n        answer = doc.get(\"answer\", doc.get(\"orig_answer\", doc.get(\"orig_orig_answer\")))\n        if solution is None:\n            print(\"Warning: No solution found; DOC:\", doc)\n        out_doc = {\n            \"problem\": problem,\n            \"solution\": solution,\n            \"answer\": answer,\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n    return dataset.map(_process_doc)\n",
-      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return QUERY_TEMPLATE.format(Question=doc.get(\"problem\", doc.get(\"question\")))\n",
-      "doc_to_target": "answer",
-      "unsafe_code": false,
-      "process_results": "def process_results(doc: dict, results: List[str]) -> Dict[str, int]:\n    metrics = {\"exact_match\": None, \"extracted_answers\": []}\n    # bp()\n    # Multiple results -> we are measuring cov/maj etc\n    if isinstance(results[0], list):\n        results = results[0]\n        n_res = len(results) # e.g. 64\n        n_res_list = [2**i for i in range(1, int(n_res.bit_length()))] # e.g. [2, 4, 8, 16, 32, 64]\n        metrics = {\n            **metrics,\n            \"exact_matches\": [],\n            **{f\"cov@{n}\": -1 for n in n_res_list},\n            **{f\"maj@{n}\": -1 for n in n_res_list},\n        }\n\n    if os.getenv(\"PROCESSOR\", \"\") == \"gpt-4o-mini\":\n        sampler = ChatCompletionSampler(model=\"gpt-4o-mini\")\n    else:\n        print(f\"Unknown processor: {os.getenv('PROCESSOR')}; set 'PROCESSOR=gpt-4o-mini' and 'OPENAI_API_KEY=YOUR_KEY' for best results.\")\n        sampler = None\n\n    if isinstance(doc[\"answer\"], str) and doc[\"answer\"].isdigit():\n        gt = str(int(doc[\"answer\"])) # 023 -> 23\n    else:\n        gt = str(doc[\"answer\"])\n    split_tokens = [\"<|im_start|>answer\\n\", \"<|im_start|>\"]\n\n    for i, a in enumerate(results, start=1):\n        if split_tokens[0] in a:\n            a = a.split(split_tokens[0])[-1]\n        elif split_tokens[1] in a:\n            a = a.split(split_tokens[1])[-1]\n            if \"\\n\" in a:\n                a = \"\\n\".join(a.split(\"\\n\")[1:])\n\n        if (box := last_boxed_only_string(a)) is not None:\n            a = remove_boxed(box)\n        # re.DOTALL is key such that newlines are included e.g. if it does `Answer: Here is the solution:\\n\\n10`\n        elif (matches := re.findall(ANSWER_PATTERN, a, re.DOTALL)) != []:\n            a = matches[-1]  # Get the last match\n\n        # AIME answers are from 000 to 999 so often it is a digit anyways\n        if (a.isdigit()) and (gt.isdigit()):\n            a = str(int(a)) # 023 -> 23\n        elif sampler is not None:\n            options = [gt] + list(set(metrics[\"extracted_answers\"]) - {gt})\n            if len(options) > 7:\n                # Could switch back to exact returning like in AIME in that case\n                # Problem with exact returning is that it sometimes messes up small things like a dollar sign\n                print(\"Warning: Lots of options which may harm indexing performance:\", options)            \n            # This ensures that if doc['answer'] is \\text{Evelyn} it is represented as such and not \\\\text{Evelyn}\n            options_str = \"[\" + \", \".join([\"'\" + str(o) + \"'\" for o in options]) + \"]\"\n            # a = extract_answer(sampler, options, a)\n            idx = extract_answer_idx(sampler, options_str, a)\n            if idx != \"-1\":\n                if idx.isdigit():\n                    idx = int(idx) - 1\n                    if len(options) > idx >= 0:\n                        a = options[idx]\n                    else:\n                        print(\"Warning: Index out of bounds; leaving answer unchanged\\n\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n                else:\n                    print(\"Warning: Processing did not produce integer index\\na\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n        else:\n            pass # TODO: Maybe add back legacy processing\n\n        metrics[\"extracted_answers\"].append(a)\n        a = int(a == gt)\n        if not(a): # Optional logging\n            print(\"Marked incorrect\\na \" + metrics[\"extracted_answers\"][-1] + \"\\ndoc['answer'] \" + gt)\n        if i == 1:\n            metrics[\"exact_match\"] = a\n            if \"exact_matches\" in metrics:\n                metrics[\"exact_matches\"].append(a)\n        elif i > 1:\n            metrics[\"exact_matches\"].append(a)\n            if i in n_res_list:\n                metrics[f\"cov@{i}\"] = int(1 in metrics[\"exact_matches\"])\n                metrics[f\"maj@{i}\"] = int(gt == Counter(metrics[\"extracted_answers\"]).most_common(1)[0][0])\n\n    return metrics\n",
-      "description": "",
-      "target_delimiter": " ",
-      "fewshot_delimiter": "\n\n",
-      "num_fewshot": 0,
-      "metric_list": [
-        {
-          "metric": "exact_match",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@64",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@32",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@16",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@8",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@4",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@2",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@64",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@32",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@16",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@8",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@4",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@2",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "extracted_answers",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        },
-        {
-          "metric": "exact_matches",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        }
-      ],
-      "output_type": "generate_until",
-      "generation_kwargs": {
-        "until": [],
-        "do_sample": false,
-        "temperature": 1,
-        "max_gen_toks": 32768
-      },
-      "repeats": 64,
-      "filter_list": [
-        {
-          "name": "all",
-          "filter": [
-            {
-              "function": "take_first_k",
-              "k": 64
-            }
-          ]
-        }
-      ],
-      "should_decontaminate": false,
-      "metadata": {
-        "version": 1.0
-      }
-    }
-  },
-  "versions": {
-    "aime24_nofigures_agg64": 1.0,
-    "aime25_nofigures_agg64": 1.0
-  },
-  "n-shot": {
-    "aime24_nofigures_agg64": 0,
-    "aime25_nofigures_agg64": 0
-  },
-  "higher_is_better": {
-    "aime24_nofigures_agg64": {
-      "exact_match": true,
-      "cov@64": true,
-      "cov@32": true,
-      "cov@16": true,
-      "cov@8": true,
-      "cov@4": true,
-      "cov@2": true,
-      "maj@64": true,
-      "maj@32": true,
-      "maj@16": true,
-      "maj@8": true,
-      "maj@4": true,
-      "maj@2": true,
-      "extracted_answers": true,
-      "exact_matches": true
-    },
-    "aime25_nofigures_agg64": {
-      "exact_match": true,
-      "cov@64": true,
-      "cov@32": true,
-      "cov@16": true,
-      "cov@8": true,
-      "cov@4": true,
-      "cov@2": true,
-      "maj@64": true,
-      "maj@32": true,
-      "maj@16": true,
-      "maj@8": true,
-      "maj@4": true,
-      "maj@2": true,
-      "extracted_answers": true,
-      "exact_matches": true
-    }
-  },
-  "n-samples": {
-    "aime25_nofigures_agg64": {
-      "original": 30,
-      "effective": 30
-    },
-    "aime24_nofigures_agg64": {
-      "original": 30,
-      "effective": 30
-    }
-  },
-  "config": {
-    "model": "vllm",
-    "model_args": "pretrained=Qwen/Qwen2.5-32B-Instruct,dtype=auto,tensor_parallel_size=2",
-    "batch_size": "auto",
-    "batch_sizes": [],
-    "device": null,
-    "use_cache": null,
-    "limit": null,
-    "bootstrap_iters": 0,
-    "gen_kwargs": {
-      "max_gen_toks": 32768,
-      "temperature": 1
-    },
-    "random_seed": 0,
-    "numpy_seed": 1234,
-    "torch_seed": 1234,
-    "fewshot_seed": 1234
-  },
-  "git_hash": "250c17c",
-  "date": 1741566388.9928138,
-  "pretty_env_info": "PyTorch version: 2.5.1+cu124\nIs debug build: False\nCUDA used to build PyTorch: 12.4\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 22.04.5 LTS (x86_64)\nGCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0\nClang version: 14.0.0-1ubuntu1.1\nCMake version: Could not collect\nLibc version: glibc-2.35\n\nPython version: 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:36:39) [GCC 12.3.0] (64-bit runtime)\nPython platform: Linux-6.8.0-1004-gcp-tcpx-x86_64-with-glibc2.35\nIs CUDA available: True\nCUDA runtime version: 12.4.99\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: \nGPU 0: NVIDIA H100 80GB HBM3\nGPU 1: NVIDIA H100 80GB HBM3\nGPU 2: NVIDIA H100 80GB HBM3\nGPU 3: NVIDIA H100 80GB HBM3\nGPU 4: NVIDIA H100 80GB HBM3\nGPU 5: NVIDIA H100 80GB HBM3\nGPU 6: NVIDIA H100 80GB HBM3\nGPU 7: NVIDIA H100 80GB HBM3\n\nNvidia driver version: 550.127.08\ncuDNN version: Could not collect\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\n\nCPU:\nArchitecture:                         x86_64\nCPU op-mode(s):                       32-bit, 64-bit\nAddress sizes:                        52 bits physical, 57 bits virtual\nByte Order:                           Little Endian\nCPU(s):                               208\nOn-line CPU(s) list:                  0-207\nVendor ID:                            GenuineIntel\nModel name:                           Intel(R) Xeon(R) Platinum 8481C CPU @ 2.70GHz\nCPU family:                           6\nModel:                                143\nThread(s) per core:                   2\nCore(s) per socket:                   52\nSocket(s):                            2\nStepping:                             8\nBogoMIPS:                             5399.99\nFlags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch ssbd ibrs ibpb stibp ibrs_enhanced fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rtm avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves avx_vnni avx512_bf16 arat avx512vbmi umip avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid cldemote movdiri movdir64b fsrm md_clear serialize amx_bf16 avx512_fp16 amx_tile amx_int8 arch_capabilities\nHypervisor vendor:                    KVM\nVirtualization type:                  full\nL1d cache:                            4.9 MiB (104 instances)\nL1i cache:                            3.3 MiB (104 instances)\nL2 cache:                             208 MiB (104 instances)\nL3 cache:                             210 MiB (2 instances)\nNUMA node(s):                         2\nNUMA node0 CPU(s):                    0-51,104-155\nNUMA node1 CPU(s):                    52-103,156-207\nVulnerability Gather data sampling:   Not affected\nVulnerability Itlb multihit:          Not affected\nVulnerability L1tf:                   Not affected\nVulnerability Mds:                    Not affected\nVulnerability Meltdown:               Not affected\nVulnerability Mmio stale data:        Not affected\nVulnerability Reg file data sampling: Not affected\nVulnerability Retbleed:               Not affected\nVulnerability Spec rstack overflow:   Not affected\nVulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl\nVulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization\nVulnerability Spectre v2:             Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI SW loop, KVM SW loop\nVulnerability Srbds:                  Not affected\nVulnerability Tsx async abort:        Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==1.26.4\n[pip3] torch==2.5.1\n[pip3] torchaudio==2.5.1\n[pip3] torchvision==0.20.1\n[pip3] triton==3.1.0\n[conda] numpy                     1.26.4                   pypi_0    pypi\n[conda] torch                     2.5.1                    pypi_0    pypi\n[conda] torchaudio                2.5.1                    pypi_0    pypi\n[conda] torchvision               0.20.1                   pypi_0    pypi\n[conda] triton                    3.1.0                    pypi_0    pypi",
-  "transformers_version": "4.49.0",
-  "upper_git_hash": null,
-  "tokenizer_pad_token": [
-    "<|endoftext|>",
-    "151643"
-  ],
-  "tokenizer_eos_token": [
-    "<|im_end|>",
-    "151645"
-  ],
-  "tokenizer_bos_token": [
-    null,
-    "None"
-  ],
-  "eot_token_id": 151645,
-  "max_length": 32768,
-  "task_hashes": {
-    "aime25_nofigures_agg64": "62013a4d0ca159910322dea9c00f72c99530d7ecd42bdeea9044f8a11bf8406e",
-    "aime24_nofigures_agg64": "3eb5fb976b3f4dea4e4e2a2caf5efa2cfea98aa3ae68cd0f3bfa8a3f197b0e2d"
-  },
-  "model_source": "vllm",
-  "model_name": "Qwen/Qwen2.5-32B-Instruct",
-  "model_name_sanitized": "Qwen__Qwen2.5-32B-Instruct",
-  "system_instruction": null,
-  "system_instruction_sha": null,
-  "fewshot_as_multiturn": false,
-  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
-  "chat_template_sha": "cd8e9439f0570856fd70470bf8889ebd8b5d1107207f67a5efb46e342330527f",
-  "start_time": 2010343.602167882,
-  "end_time": 2011216.285952138,
-  "total_evaluation_time_seconds": "872.6837842559908"
-}
\ No newline at end of file
diff --git a/qwen_agg64temp1/Qwen__Qwen2.5-32B-Instruct/samples_aime24_nofigures_agg64_2025-03-10T00-40-46.825589.jsonl b/qwen_agg64temp1/Qwen__Qwen2.5-32B-Instruct/samples_aime24_nofigures_agg64_2025-03-10T00-40-46.825589.jsonl
deleted file mode 100644
index 80553a75d895e7d60b9a6348decd08825ed81290..0000000000000000000000000000000000000000
--- a/qwen_agg64temp1/Qwen__Qwen2.5-32B-Instruct/samples_aime24_nofigures_agg64_2025-03-10T00-40-46.825589.jsonl
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:6887999776d48c0f918edee22a2a0c770bbd6faaa722fdb4fcdaf1a4834f5915
-size 9362555
diff --git a/qwen_agg64temp1/Qwen__Qwen2.5-32B-Instruct/samples_aime25_nofigures_agg64_2025-03-10T00-40-46.825589.jsonl b/qwen_agg64temp1/Qwen__Qwen2.5-32B-Instruct/samples_aime25_nofigures_agg64_2025-03-10T00-40-46.825589.jsonl
deleted file mode 100644
index a22b83af41f2552f3359dfdb58842cc94297c54b..0000000000000000000000000000000000000000
--- a/qwen_agg64temp1/Qwen__Qwen2.5-32B-Instruct/samples_aime25_nofigures_agg64_2025-03-10T00-40-46.825589.jsonl
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f0ece190152dd728e908d6efc2597b4ee7ef56b2acea5fcd03cf29ac2be5a777
-size 9107791
diff --git a/qwq/Qwen__QwQ-32B/results_2025-03-09T22-25-20.807425.json b/qwq/Qwen__QwQ-32B/results_2025-03-09T22-25-20.807425.json
deleted file mode 100644
index 3c71e62ce921fea3dc085cb66aa3af547b029cd3..0000000000000000000000000000000000000000
--- a/qwq/Qwen__QwQ-32B/results_2025-03-09T22-25-20.807425.json
+++ /dev/null
@@ -1,126 +0,0 @@
-{
-  "results": {
-    "aime25_nofigures": {
-      "alias": "aime25_nofigures",
-      "exact_match,none": 0.6666666666666666,
-      "exact_match_stderr,none": "N/A",
-      "extracted_answers,none": -1,
-      "extracted_answers_stderr,none": "N/A"
-    }
-  },
-  "group_subtasks": {
-    "aime25_nofigures": []
-  },
-  "configs": {
-    "aime25_nofigures": {
-      "task": "aime25_nofigures",
-      "tag": [
-        "math_word_problems"
-      ],
-      "dataset_path": "simplescaling/aime25_nofigures",
-      "dataset_name": "default",
-      "test_split": "train",
-      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        solution = doc.get(\"solution\", doc.get(\"orig_solution\", doc.get(\"orig_orig_solution\")))\n        problem = doc.get(\"problem\", doc.get(\"question\"))\n        answer = doc.get(\"answer\", doc.get(\"orig_answer\", doc.get(\"orig_orig_answer\")))\n        if solution is None:\n            print(\"Warning: No solution found; DOC:\", doc)\n        out_doc = {\n            \"problem\": problem,\n            \"solution\": solution,\n            \"answer\": answer,\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n    return dataset.map(_process_doc)\n",
-      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return QUERY_TEMPLATE.format(Question=doc.get(\"problem\", doc.get(\"question\")))\n",
-      "doc_to_target": "answer",
-      "unsafe_code": false,
-      "process_results": "def process_results(doc: dict, results: List[str]) -> Dict[str, int]:\n    metrics = {\"exact_match\": None, \"extracted_answers\": []}\n    # bp()\n    # Multiple results -> we are measuring cov/maj etc\n    if isinstance(results[0], list):\n        results = results[0]\n        n_res = len(results) # e.g. 64\n        n_res_list = [2**i for i in range(1, int(n_res.bit_length()))] # e.g. [2, 4, 8, 16, 32, 64]\n        metrics = {\n            **metrics,\n            \"exact_matches\": [],\n            **{f\"cov@{n}\": -1 for n in n_res_list},\n            **{f\"maj@{n}\": -1 for n in n_res_list},\n        }\n\n    if os.getenv(\"PROCESSOR\", \"\") == \"gpt-4o-mini\":\n        sampler = ChatCompletionSampler(model=\"gpt-4o-mini\")\n    else:\n        print(f\"Unknown processor: {os.getenv('PROCESSOR')}; set 'PROCESSOR=gpt-4o-mini' and 'OPENAI_API_KEY=YOUR_KEY' for best results.\")\n        sampler = None\n\n    if isinstance(doc[\"answer\"], str) and doc[\"answer\"].isdigit():\n        gt = str(int(doc[\"answer\"])) # 023 -> 23\n    else:\n        gt = str(doc[\"answer\"])\n    split_tokens = [\"<|im_start|>answer\\n\", \"<|im_start|>\"]\n\n    for i, a in enumerate(results, start=1):\n        if split_tokens[0] in a:\n            a = a.split(split_tokens[0])[-1]\n        elif split_tokens[1] in a:\n            a = a.split(split_tokens[1])[-1]\n            if \"\\n\" in a:\n                a = \"\\n\".join(a.split(\"\\n\")[1:])\n\n        if (box := last_boxed_only_string(a)) is not None:\n            a = remove_boxed(box)\n        # re.DOTALL is key such that newlines are included e.g. if it does `Answer: Here is the solution:\\n\\n10`\n        elif (matches := re.findall(ANSWER_PATTERN, a, re.DOTALL)) != []:\n            a = matches[-1]  # Get the last match\n\n        # AIME answers are from 000 to 999 so often it is a digit anyways\n        if (a.isdigit()) and (gt.isdigit()):\n            a = str(int(a)) # 023 -> 23\n        elif sampler is not None:\n            options = [gt] + list(set(metrics[\"extracted_answers\"]) - {gt})\n            if len(options) > 7:\n                # Could switch back to exact returning like in AIME in that case\n                # Problem with exact returning is that it sometimes messes up small things like a dollar sign\n                print(\"Warning: Lots of options which may harm indexing performance:\", options)            \n            # This ensures that if doc['answer'] is \\text{Evelyn} it is represented as such and not \\\\text{Evelyn}\n            options_str = \"[\" + \", \".join([\"'\" + str(o) + \"'\" for o in options]) + \"]\"\n            # a = extract_answer(sampler, options, a)\n            idx = extract_answer_idx(sampler, options_str, a)\n            if idx != \"-1\":\n                if idx.isdigit():\n                    idx = int(idx) - 1\n                    if len(options) > idx >= 0:\n                        a = options[idx]\n                    else:\n                        print(\"Warning: Index out of bounds; leaving answer unchanged\\n\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n                else:\n                    print(\"Warning: Processing did not produce integer index\\na\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n        else:\n            pass # TODO: Maybe add back legacy processing\n\n        metrics[\"extracted_answers\"].append(a)\n        a = int(a == gt)\n        if not(a): # Optional logging\n            print(\"Marked incorrect\\na \" + metrics[\"extracted_answers\"][-1] + \"\\ndoc['answer'] \" + gt)\n        if i == 1:\n            metrics[\"exact_match\"] = a\n            if \"exact_matches\" in metrics:\n                metrics[\"exact_matches\"].append(a)\n        elif i > 1:\n            metrics[\"exact_matches\"].append(a)\n            if i in n_res_list:\n                metrics[f\"cov@{i}\"] = int(1 in metrics[\"exact_matches\"])\n                metrics[f\"maj@{i}\"] = int(gt == Counter(metrics[\"extracted_answers\"]).most_common(1)[0][0])\n\n    return metrics\n",
-      "description": "",
-      "target_delimiter": " ",
-      "fewshot_delimiter": "\n\n",
-      "num_fewshot": 0,
-      "metric_list": [
-        {
-          "metric": "exact_match",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "extracted_answers",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        }
-      ],
-      "output_type": "generate_until",
-      "generation_kwargs": {
-        "until": [],
-        "do_sample": false,
-        "temperature": 0.0,
-        "max_gen_toks": 32768
-      },
-      "repeats": 1,
-      "should_decontaminate": false,
-      "metadata": {
-        "version": 1.0
-      }
-    }
-  },
-  "versions": {
-    "aime25_nofigures": 1.0
-  },
-  "n-shot": {
-    "aime25_nofigures": 0
-  },
-  "higher_is_better": {
-    "aime25_nofigures": {
-      "exact_match": true,
-      "extracted_answers": true
-    }
-  },
-  "n-samples": {
-    "aime25_nofigures": {
-      "original": 30,
-      "effective": 30
-    }
-  },
-  "config": {
-    "model": "vllm",
-    "model_args": "pretrained=Qwen/QwQ-32B,dtype=auto,tensor_parallel_size=2",
-    "batch_size": "auto",
-    "batch_sizes": [],
-    "device": null,
-    "use_cache": null,
-    "limit": null,
-    "bootstrap_iters": 0,
-    "gen_kwargs": {
-      "max_gen_toks": 32768
-    },
-    "random_seed": 0,
-    "numpy_seed": 1234,
-    "torch_seed": 1234,
-    "fewshot_seed": 1234
-  },
-  "git_hash": "9d173236",
-  "date": 1741558017.2585754,
-  "pretty_env_info": "PyTorch version: 2.5.1+cu124\nIs debug build: False\nCUDA used to build PyTorch: 12.4\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 22.04.5 LTS (x86_64)\nGCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0\nClang version: 14.0.0-1ubuntu1.1\nCMake version: Could not collect\nLibc version: glibc-2.35\n\nPython version: 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:36:39) [GCC 12.3.0] (64-bit runtime)\nPython platform: Linux-6.8.0-1004-gcp-tcpx-x86_64-with-glibc2.35\nIs CUDA available: True\nCUDA runtime version: 12.4.99\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: \nGPU 0: NVIDIA H100 80GB HBM3\nGPU 1: NVIDIA H100 80GB HBM3\nGPU 2: NVIDIA H100 80GB HBM3\nGPU 3: NVIDIA H100 80GB HBM3\nGPU 4: NVIDIA H100 80GB HBM3\nGPU 5: NVIDIA H100 80GB HBM3\nGPU 6: NVIDIA H100 80GB HBM3\nGPU 7: NVIDIA H100 80GB HBM3\n\nNvidia driver version: 550.127.08\ncuDNN version: Could not collect\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\n\nCPU:\nArchitecture:                         x86_64\nCPU op-mode(s):                       32-bit, 64-bit\nAddress sizes:                        52 bits physical, 57 bits virtual\nByte Order:                           Little Endian\nCPU(s):                               208\nOn-line CPU(s) list:                  0-207\nVendor ID:                            GenuineIntel\nModel name:                           Intel(R) Xeon(R) Platinum 8481C CPU @ 2.70GHz\nCPU family:                           6\nModel:                                143\nThread(s) per core:                   2\nCore(s) per socket:                   52\nSocket(s):                            2\nStepping:                             8\nBogoMIPS:                             5399.99\nFlags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch ssbd ibrs ibpb stibp ibrs_enhanced fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rtm avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves avx_vnni avx512_bf16 arat avx512vbmi umip avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid cldemote movdiri movdir64b fsrm md_clear serialize amx_bf16 avx512_fp16 amx_tile amx_int8 arch_capabilities\nHypervisor vendor:                    KVM\nVirtualization type:                  full\nL1d cache:                            4.9 MiB (104 instances)\nL1i cache:                            3.3 MiB (104 instances)\nL2 cache:                             208 MiB (104 instances)\nL3 cache:                             210 MiB (2 instances)\nNUMA node(s):                         2\nNUMA node0 CPU(s):                    0-51,104-155\nNUMA node1 CPU(s):                    52-103,156-207\nVulnerability Gather data sampling:   Not affected\nVulnerability Itlb multihit:          Not affected\nVulnerability L1tf:                   Not affected\nVulnerability Mds:                    Not affected\nVulnerability Meltdown:               Not affected\nVulnerability Mmio stale data:        Not affected\nVulnerability Reg file data sampling: Not affected\nVulnerability Retbleed:               Not affected\nVulnerability Spec rstack overflow:   Not affected\nVulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl\nVulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization\nVulnerability Spectre v2:             Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI SW loop, KVM SW loop\nVulnerability Srbds:                  Not affected\nVulnerability Tsx async abort:        Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==1.26.4\n[pip3] torch==2.5.1\n[pip3] torchaudio==2.5.1\n[pip3] torchvision==0.20.1\n[pip3] triton==3.1.0\n[conda] numpy                     1.26.4                   pypi_0    pypi\n[conda] torch                     2.5.1                    pypi_0    pypi\n[conda] torchaudio                2.5.1                    pypi_0    pypi\n[conda] torchvision               0.20.1                   pypi_0    pypi\n[conda] triton                    3.1.0                    pypi_0    pypi",
-  "transformers_version": "4.49.0",
-  "upper_git_hash": "250c17c32540b954735b0284ac42a202ff910005",
-  "tokenizer_pad_token": [
-    "<|endoftext|>",
-    "151643"
-  ],
-  "tokenizer_eos_token": [
-    "<|im_end|>",
-    "151645"
-  ],
-  "tokenizer_bos_token": [
-    null,
-    "None"
-  ],
-  "eot_token_id": 151645,
-  "max_length": 131072,
-  "task_hashes": {
-    "aime25_nofigures": "6c6c9c546466598a7d7e6baa470c8ed21d835e4d4f78a9bffe965e1125db2916"
-  },
-  "model_source": "vllm",
-  "model_name": "Qwen/QwQ-32B",
-  "model_name_sanitized": "Qwen__QwQ-32B",
-  "system_instruction": null,
-  "system_instruction_sha": null,
-  "fewshot_as_multiturn": false,
-  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- '' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n  {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" and not message.tool_calls %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n<think>\\n' }}\n{%- endif %}\n",
-  "chat_template_sha": "61433aea8393589935ac6fdc49e937f991c796925e27241bd2b1d1080d9cf9af",
-  "start_time": 2001972.53406931,
-  "end_time": 2003090.262468118,
-  "total_evaluation_time_seconds": "1117.7283988080453"
-}
\ No newline at end of file
diff --git a/qwq/Qwen__QwQ-32B/results_2025-03-09T22-51-59.932168.json b/qwq/Qwen__QwQ-32B/results_2025-03-09T22-51-59.932168.json
deleted file mode 100644
index c6e07a2923e426dbeef0c486c8c655773a673381..0000000000000000000000000000000000000000
--- a/qwq/Qwen__QwQ-32B/results_2025-03-09T22-51-59.932168.json
+++ /dev/null
@@ -1,126 +0,0 @@
-{
-  "results": {
-    "aime24_nofigures": {
-      "alias": "aime24_nofigures",
-      "exact_match,none": 0.7666666666666667,
-      "exact_match_stderr,none": "N/A",
-      "extracted_answers,none": -1,
-      "extracted_answers_stderr,none": "N/A"
-    }
-  },
-  "group_subtasks": {
-    "aime24_nofigures": []
-  },
-  "configs": {
-    "aime24_nofigures": {
-      "task": "aime24_nofigures",
-      "tag": [
-        "math_word_problems"
-      ],
-      "dataset_path": "simplescaling/aime24_nofigures",
-      "dataset_name": "default",
-      "test_split": "train",
-      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        solution = doc.get(\"solution\", doc.get(\"orig_solution\", doc.get(\"orig_orig_solution\")))\n        problem = doc.get(\"problem\", doc.get(\"question\"))\n        answer = doc.get(\"answer\", doc.get(\"orig_answer\", doc.get(\"orig_orig_answer\")))\n        if solution is None:\n            print(\"Warning: No solution found; DOC:\", doc)\n        out_doc = {\n            \"problem\": problem,\n            \"solution\": solution,\n            \"answer\": answer,\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n    return dataset.map(_process_doc)\n",
-      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return QUERY_TEMPLATE.format(Question=doc.get(\"problem\", doc.get(\"question\")))\n",
-      "doc_to_target": "answer",
-      "unsafe_code": false,
-      "process_results": "def process_results(doc: dict, results: List[str]) -> Dict[str, int]:\n    metrics = {\"exact_match\": None, \"extracted_answers\": []}\n    # bp()\n    # Multiple results -> we are measuring cov/maj etc\n    if isinstance(results[0], list):\n        results = results[0]\n        n_res = len(results) # e.g. 64\n        n_res_list = [2**i for i in range(1, int(n_res.bit_length()))] # e.g. [2, 4, 8, 16, 32, 64]\n        metrics = {\n            **metrics,\n            \"exact_matches\": [],\n            **{f\"cov@{n}\": -1 for n in n_res_list},\n            **{f\"maj@{n}\": -1 for n in n_res_list},\n        }\n\n    if os.getenv(\"PROCESSOR\", \"\") == \"gpt-4o-mini\":\n        sampler = ChatCompletionSampler(model=\"gpt-4o-mini\")\n    else:\n        print(f\"Unknown processor: {os.getenv('PROCESSOR')}; set 'PROCESSOR=gpt-4o-mini' and 'OPENAI_API_KEY=YOUR_KEY' for best results.\")\n        sampler = None\n\n    if isinstance(doc[\"answer\"], str) and doc[\"answer\"].isdigit():\n        gt = str(int(doc[\"answer\"])) # 023 -> 23\n    else:\n        gt = str(doc[\"answer\"])\n    split_tokens = [\"<|im_start|>answer\\n\", \"<|im_start|>\"]\n\n    for i, a in enumerate(results, start=1):\n        if split_tokens[0] in a:\n            a = a.split(split_tokens[0])[-1]\n        elif split_tokens[1] in a:\n            a = a.split(split_tokens[1])[-1]\n            if \"\\n\" in a:\n                a = \"\\n\".join(a.split(\"\\n\")[1:])\n\n        if (box := last_boxed_only_string(a)) is not None:\n            a = remove_boxed(box)\n        # re.DOTALL is key such that newlines are included e.g. if it does `Answer: Here is the solution:\\n\\n10`\n        elif (matches := re.findall(ANSWER_PATTERN, a, re.DOTALL)) != []:\n            a = matches[-1]  # Get the last match\n\n        # AIME answers are from 000 to 999 so often it is a digit anyways\n        if (a.isdigit()) and (gt.isdigit()):\n            a = str(int(a)) # 023 -> 23\n        elif sampler is not None:\n            options = [gt] + list(set(metrics[\"extracted_answers\"]) - {gt})\n            if len(options) > 7:\n                # Could switch back to exact returning like in AIME in that case\n                # Problem with exact returning is that it sometimes messes up small things like a dollar sign\n                print(\"Warning: Lots of options which may harm indexing performance:\", options)            \n            # This ensures that if doc['answer'] is \\text{Evelyn} it is represented as such and not \\\\text{Evelyn}\n            options_str = \"[\" + \", \".join([\"'\" + str(o) + \"'\" for o in options]) + \"]\"\n            # a = extract_answer(sampler, options, a)\n            idx = extract_answer_idx(sampler, options_str, a)\n            if idx != \"-1\":\n                if idx.isdigit():\n                    idx = int(idx) - 1\n                    if len(options) > idx >= 0:\n                        a = options[idx]\n                    else:\n                        print(\"Warning: Index out of bounds; leaving answer unchanged\\n\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n                else:\n                    print(\"Warning: Processing did not produce integer index\\na\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n        else:\n            pass # TODO: Maybe add back legacy processing\n\n        metrics[\"extracted_answers\"].append(a)\n        a = int(a == gt)\n        if not(a): # Optional logging\n            print(\"Marked incorrect\\na \" + metrics[\"extracted_answers\"][-1] + \"\\ndoc['answer'] \" + gt)\n        if i == 1:\n            metrics[\"exact_match\"] = a\n            if \"exact_matches\" in metrics:\n                metrics[\"exact_matches\"].append(a)\n        elif i > 1:\n            metrics[\"exact_matches\"].append(a)\n            if i in n_res_list:\n                metrics[f\"cov@{i}\"] = int(1 in metrics[\"exact_matches\"])\n                metrics[f\"maj@{i}\"] = int(gt == Counter(metrics[\"extracted_answers\"]).most_common(1)[0][0])\n\n    return metrics\n",
-      "description": "",
-      "target_delimiter": " ",
-      "fewshot_delimiter": "\n\n",
-      "num_fewshot": 0,
-      "metric_list": [
-        {
-          "metric": "exact_match",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "extracted_answers",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        }
-      ],
-      "output_type": "generate_until",
-      "generation_kwargs": {
-        "until": [],
-        "do_sample": false,
-        "temperature": 0.0,
-        "max_gen_toks": 32768
-      },
-      "repeats": 1,
-      "should_decontaminate": false,
-      "metadata": {
-        "version": 1.0
-      }
-    }
-  },
-  "versions": {
-    "aime24_nofigures": 1.0
-  },
-  "n-shot": {
-    "aime24_nofigures": 0
-  },
-  "higher_is_better": {
-    "aime24_nofigures": {
-      "exact_match": true,
-      "extracted_answers": true
-    }
-  },
-  "n-samples": {
-    "aime24_nofigures": {
-      "original": 30,
-      "effective": 30
-    }
-  },
-  "config": {
-    "model": "vllm",
-    "model_args": "pretrained=Qwen/QwQ-32B,dtype=auto,tensor_parallel_size=2",
-    "batch_size": "auto",
-    "batch_sizes": [],
-    "device": null,
-    "use_cache": null,
-    "limit": null,
-    "bootstrap_iters": 0,
-    "gen_kwargs": {
-      "max_gen_toks": 32768
-    },
-    "random_seed": 0,
-    "numpy_seed": 1234,
-    "torch_seed": 1234,
-    "fewshot_seed": 1234
-  },
-  "git_hash": "250c17c",
-  "date": 1741559722.8094802,
-  "pretty_env_info": "PyTorch version: 2.5.1+cu124\nIs debug build: False\nCUDA used to build PyTorch: 12.4\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 22.04.5 LTS (x86_64)\nGCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0\nClang version: 14.0.0-1ubuntu1.1\nCMake version: Could not collect\nLibc version: glibc-2.35\n\nPython version: 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:36:39) [GCC 12.3.0] (64-bit runtime)\nPython platform: Linux-6.8.0-1004-gcp-tcpx-x86_64-with-glibc2.35\nIs CUDA available: True\nCUDA runtime version: 12.4.99\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: \nGPU 0: NVIDIA H100 80GB HBM3\nGPU 1: NVIDIA H100 80GB HBM3\nGPU 2: NVIDIA H100 80GB HBM3\nGPU 3: NVIDIA H100 80GB HBM3\nGPU 4: NVIDIA H100 80GB HBM3\nGPU 5: NVIDIA H100 80GB HBM3\nGPU 6: NVIDIA H100 80GB HBM3\nGPU 7: NVIDIA H100 80GB HBM3\n\nNvidia driver version: 550.127.08\ncuDNN version: Could not collect\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\n\nCPU:\nArchitecture:                         x86_64\nCPU op-mode(s):                       32-bit, 64-bit\nAddress sizes:                        52 bits physical, 57 bits virtual\nByte Order:                           Little Endian\nCPU(s):                               208\nOn-line CPU(s) list:                  0-207\nVendor ID:                            GenuineIntel\nModel name:                           Intel(R) Xeon(R) Platinum 8481C CPU @ 2.70GHz\nCPU family:                           6\nModel:                                143\nThread(s) per core:                   2\nCore(s) per socket:                   52\nSocket(s):                            2\nStepping:                             8\nBogoMIPS:                             5399.99\nFlags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch ssbd ibrs ibpb stibp ibrs_enhanced fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rtm avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves avx_vnni avx512_bf16 arat avx512vbmi umip avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid cldemote movdiri movdir64b fsrm md_clear serialize amx_bf16 avx512_fp16 amx_tile amx_int8 arch_capabilities\nHypervisor vendor:                    KVM\nVirtualization type:                  full\nL1d cache:                            4.9 MiB (104 instances)\nL1i cache:                            3.3 MiB (104 instances)\nL2 cache:                             208 MiB (104 instances)\nL3 cache:                             210 MiB (2 instances)\nNUMA node(s):                         2\nNUMA node0 CPU(s):                    0-51,104-155\nNUMA node1 CPU(s):                    52-103,156-207\nVulnerability Gather data sampling:   Not affected\nVulnerability Itlb multihit:          Not affected\nVulnerability L1tf:                   Not affected\nVulnerability Mds:                    Not affected\nVulnerability Meltdown:               Not affected\nVulnerability Mmio stale data:        Not affected\nVulnerability Reg file data sampling: Not affected\nVulnerability Retbleed:               Not affected\nVulnerability Spec rstack overflow:   Not affected\nVulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl\nVulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization\nVulnerability Spectre v2:             Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI SW loop, KVM SW loop\nVulnerability Srbds:                  Not affected\nVulnerability Tsx async abort:        Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==1.26.4\n[pip3] torch==2.5.1\n[pip3] torchaudio==2.5.1\n[pip3] torchvision==0.20.1\n[pip3] triton==3.1.0\n[conda] numpy                     1.26.4                   pypi_0    pypi\n[conda] torch                     2.5.1                    pypi_0    pypi\n[conda] torchaudio                2.5.1                    pypi_0    pypi\n[conda] torchvision               0.20.1                   pypi_0    pypi\n[conda] triton                    3.1.0                    pypi_0    pypi",
-  "transformers_version": "4.49.0",
-  "upper_git_hash": null,
-  "tokenizer_pad_token": [
-    "<|endoftext|>",
-    "151643"
-  ],
-  "tokenizer_eos_token": [
-    "<|im_end|>",
-    "151645"
-  ],
-  "tokenizer_bos_token": [
-    null,
-    "None"
-  ],
-  "eot_token_id": 151645,
-  "max_length": 131072,
-  "task_hashes": {
-    "aime24_nofigures": "5559473020237fa3e670d94426cd56a532b8f02b3cba30c936a326b25f86b169"
-  },
-  "model_source": "vllm",
-  "model_name": "Qwen/QwQ-32B",
-  "model_name_sanitized": "Qwen__QwQ-32B",
-  "system_instruction": null,
-  "system_instruction_sha": null,
-  "fewshot_as_multiturn": false,
-  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- '' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n  {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" and not message.tool_calls %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n<think>\\n' }}\n{%- endif %}\n",
-  "chat_template_sha": "61433aea8393589935ac6fdc49e937f991c796925e27241bd2b1d1080d9cf9af",
-  "start_time": 2003677.752713673,
-  "end_time": 2004689.402279533,
-  "total_evaluation_time_seconds": "1011.6495658599306"
-}
\ No newline at end of file
diff --git a/qwq/Qwen__QwQ-32B/samples_aime24_nofigures_2025-03-09T22-51-59.932168.jsonl b/qwq/Qwen__QwQ-32B/samples_aime24_nofigures_2025-03-09T22-51-59.932168.jsonl
deleted file mode 100644
index 0a83127f89a721a21ec60a3e0911088db89d3003..0000000000000000000000000000000000000000
--- a/qwq/Qwen__QwQ-32B/samples_aime24_nofigures_2025-03-09T22-51-59.932168.jsonl
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0d58df1826452b3e029f65aa4e4f410630ec5007f66916756a5f52d3a3f30d07
-size 2615066
diff --git a/qwq/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T22-25-20.807425.jsonl b/qwq/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T22-25-20.807425.jsonl
deleted file mode 100644
index 82429c9c8304db2b88ad9c4c2fc13cea9d6bf35a..0000000000000000000000000000000000000000
--- a/qwq/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T22-25-20.807425.jsonl
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:006d5d06a14a7238901bf46114b3f7878e056adb633bb02fce113cd2e7c9f8d5
-size 2986401
diff --git a/qwq131k/Qwen__QwQ-32B/results_2025-03-09T23-40-20.886537.json b/qwq131k/Qwen__QwQ-32B/results_2025-03-09T23-40-20.886537.json
deleted file mode 100644
index b479ce0ffe481a29c4578a1a1361fca711e90c0f..0000000000000000000000000000000000000000
--- a/qwq131k/Qwen__QwQ-32B/results_2025-03-09T23-40-20.886537.json
+++ /dev/null
@@ -1,126 +0,0 @@
-{
-  "results": {
-    "aime25_nofigures": {
-      "alias": "aime25_nofigures",
-      "exact_match,none": 0.6666666666666666,
-      "exact_match_stderr,none": "N/A",
-      "extracted_answers,none": -1,
-      "extracted_answers_stderr,none": "N/A"
-    }
-  },
-  "group_subtasks": {
-    "aime25_nofigures": []
-  },
-  "configs": {
-    "aime25_nofigures": {
-      "task": "aime25_nofigures",
-      "tag": [
-        "math_word_problems"
-      ],
-      "dataset_path": "simplescaling/aime25_nofigures",
-      "dataset_name": "default",
-      "test_split": "train",
-      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        solution = doc.get(\"solution\", doc.get(\"orig_solution\", doc.get(\"orig_orig_solution\")))\n        problem = doc.get(\"problem\", doc.get(\"question\"))\n        answer = doc.get(\"answer\", doc.get(\"orig_answer\", doc.get(\"orig_orig_answer\")))\n        if solution is None:\n            print(\"Warning: No solution found; DOC:\", doc)\n        out_doc = {\n            \"problem\": problem,\n            \"solution\": solution,\n            \"answer\": answer,\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n    return dataset.map(_process_doc)\n",
-      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return QUERY_TEMPLATE.format(Question=doc.get(\"problem\", doc.get(\"question\")))\n",
-      "doc_to_target": "answer",
-      "unsafe_code": false,
-      "process_results": "def process_results(doc: dict, results: List[str]) -> Dict[str, int]:\n    metrics = {\"exact_match\": None, \"extracted_answers\": []}\n    # bp()\n    # Multiple results -> we are measuring cov/maj etc\n    if isinstance(results[0], list):\n        results = results[0]\n        n_res = len(results) # e.g. 64\n        n_res_list = [2**i for i in range(1, int(n_res.bit_length()))] # e.g. [2, 4, 8, 16, 32, 64]\n        metrics = {\n            **metrics,\n            \"exact_matches\": [],\n            **{f\"cov@{n}\": -1 for n in n_res_list},\n            **{f\"maj@{n}\": -1 for n in n_res_list},\n        }\n\n    if os.getenv(\"PROCESSOR\", \"\") == \"gpt-4o-mini\":\n        sampler = ChatCompletionSampler(model=\"gpt-4o-mini\")\n    else:\n        print(f\"Unknown processor: {os.getenv('PROCESSOR')}; set 'PROCESSOR=gpt-4o-mini' and 'OPENAI_API_KEY=YOUR_KEY' for best results.\")\n        sampler = None\n\n    if isinstance(doc[\"answer\"], str) and doc[\"answer\"].isdigit():\n        gt = str(int(doc[\"answer\"])) # 023 -> 23\n    else:\n        gt = str(doc[\"answer\"])\n    split_tokens = [\"<|im_start|>answer\\n\", \"<|im_start|>\"]\n\n    for i, a in enumerate(results, start=1):\n        if split_tokens[0] in a:\n            a = a.split(split_tokens[0])[-1]\n        elif split_tokens[1] in a:\n            a = a.split(split_tokens[1])[-1]\n            if \"\\n\" in a:\n                a = \"\\n\".join(a.split(\"\\n\")[1:])\n\n        if (box := last_boxed_only_string(a)) is not None:\n            a = remove_boxed(box)\n        # re.DOTALL is key such that newlines are included e.g. if it does `Answer: Here is the solution:\\n\\n10`\n        elif (matches := re.findall(ANSWER_PATTERN, a, re.DOTALL)) != []:\n            a = matches[-1]  # Get the last match\n\n        # AIME answers are from 000 to 999 so often it is a digit anyways\n        if (a.isdigit()) and (gt.isdigit()):\n            a = str(int(a)) # 023 -> 23\n        elif sampler is not None:\n            options = [gt] + list(set(metrics[\"extracted_answers\"]) - {gt})\n            if len(options) > 7:\n                # Could switch back to exact returning like in AIME in that case\n                # Problem with exact returning is that it sometimes messes up small things like a dollar sign\n                print(\"Warning: Lots of options which may harm indexing performance:\", options)            \n            # This ensures that if doc['answer'] is \\text{Evelyn} it is represented as such and not \\\\text{Evelyn}\n            options_str = \"[\" + \", \".join([\"'\" + str(o) + \"'\" for o in options]) + \"]\"\n            # a = extract_answer(sampler, options, a)\n            idx = extract_answer_idx(sampler, options_str, a)\n            if idx != \"-1\":\n                if idx.isdigit():\n                    idx = int(idx) - 1\n                    if len(options) > idx >= 0:\n                        a = options[idx]\n                    else:\n                        print(\"Warning: Index out of bounds; leaving answer unchanged\\n\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n                else:\n                    print(\"Warning: Processing did not produce integer index\\na\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n        else:\n            pass # TODO: Maybe add back legacy processing\n\n        metrics[\"extracted_answers\"].append(a)\n        a = int(a == gt)\n        if not(a): # Optional logging\n            print(\"Marked incorrect\\na \" + metrics[\"extracted_answers\"][-1] + \"\\ndoc['answer'] \" + gt)\n        if i == 1:\n            metrics[\"exact_match\"] = a\n            if \"exact_matches\" in metrics:\n                metrics[\"exact_matches\"].append(a)\n        elif i > 1:\n            metrics[\"exact_matches\"].append(a)\n            if i in n_res_list:\n                metrics[f\"cov@{i}\"] = int(1 in metrics[\"exact_matches\"])\n                metrics[f\"maj@{i}\"] = int(gt == Counter(metrics[\"extracted_answers\"]).most_common(1)[0][0])\n\n    return metrics\n",
-      "description": "",
-      "target_delimiter": " ",
-      "fewshot_delimiter": "\n\n",
-      "num_fewshot": 0,
-      "metric_list": [
-        {
-          "metric": "exact_match",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "extracted_answers",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        }
-      ],
-      "output_type": "generate_until",
-      "generation_kwargs": {
-        "until": [],
-        "do_sample": false,
-        "temperature": 0.0,
-        "max_gen_toks": 131072
-      },
-      "repeats": 1,
-      "should_decontaminate": false,
-      "metadata": {
-        "version": 1.0
-      }
-    }
-  },
-  "versions": {
-    "aime25_nofigures": 1.0
-  },
-  "n-shot": {
-    "aime25_nofigures": 0
-  },
-  "higher_is_better": {
-    "aime25_nofigures": {
-      "exact_match": true,
-      "extracted_answers": true
-    }
-  },
-  "n-samples": {
-    "aime25_nofigures": {
-      "original": 30,
-      "effective": 30
-    }
-  },
-  "config": {
-    "model": "vllm",
-    "model_args": "pretrained=Qwen/QwQ-32B,dtype=auto,tensor_parallel_size=2",
-    "batch_size": "auto",
-    "batch_sizes": [],
-    "device": null,
-    "use_cache": null,
-    "limit": null,
-    "bootstrap_iters": 0,
-    "gen_kwargs": {
-      "max_gen_toks": 131072
-    },
-    "random_seed": 0,
-    "numpy_seed": 1234,
-    "torch_seed": 1234,
-    "fewshot_seed": 1234
-  },
-  "git_hash": "250c17c",
-  "date": 1741562381.0680323,
-  "pretty_env_info": "PyTorch version: 2.5.1+cu124\nIs debug build: False\nCUDA used to build PyTorch: 12.4\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 22.04.5 LTS (x86_64)\nGCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0\nClang version: 14.0.0-1ubuntu1.1\nCMake version: Could not collect\nLibc version: glibc-2.35\n\nPython version: 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:36:39) [GCC 12.3.0] (64-bit runtime)\nPython platform: Linux-6.8.0-1004-gcp-tcpx-x86_64-with-glibc2.35\nIs CUDA available: True\nCUDA runtime version: 12.4.99\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: \nGPU 0: NVIDIA H100 80GB HBM3\nGPU 1: NVIDIA H100 80GB HBM3\nGPU 2: NVIDIA H100 80GB HBM3\nGPU 3: NVIDIA H100 80GB HBM3\nGPU 4: NVIDIA H100 80GB HBM3\nGPU 5: NVIDIA H100 80GB HBM3\nGPU 6: NVIDIA H100 80GB HBM3\nGPU 7: NVIDIA H100 80GB HBM3\n\nNvidia driver version: 550.127.08\ncuDNN version: Could not collect\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\n\nCPU:\nArchitecture:                         x86_64\nCPU op-mode(s):                       32-bit, 64-bit\nAddress sizes:                        52 bits physical, 57 bits virtual\nByte Order:                           Little Endian\nCPU(s):                               208\nOn-line CPU(s) list:                  0-207\nVendor ID:                            GenuineIntel\nModel name:                           Intel(R) Xeon(R) Platinum 8481C CPU @ 2.70GHz\nCPU family:                           6\nModel:                                143\nThread(s) per core:                   2\nCore(s) per socket:                   52\nSocket(s):                            2\nStepping:                             8\nBogoMIPS:                             5399.99\nFlags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch ssbd ibrs ibpb stibp ibrs_enhanced fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rtm avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves avx_vnni avx512_bf16 arat avx512vbmi umip avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid cldemote movdiri movdir64b fsrm md_clear serialize amx_bf16 avx512_fp16 amx_tile amx_int8 arch_capabilities\nHypervisor vendor:                    KVM\nVirtualization type:                  full\nL1d cache:                            4.9 MiB (104 instances)\nL1i cache:                            3.3 MiB (104 instances)\nL2 cache:                             208 MiB (104 instances)\nL3 cache:                             210 MiB (2 instances)\nNUMA node(s):                         2\nNUMA node0 CPU(s):                    0-51,104-155\nNUMA node1 CPU(s):                    52-103,156-207\nVulnerability Gather data sampling:   Not affected\nVulnerability Itlb multihit:          Not affected\nVulnerability L1tf:                   Not affected\nVulnerability Mds:                    Not affected\nVulnerability Meltdown:               Not affected\nVulnerability Mmio stale data:        Not affected\nVulnerability Reg file data sampling: Not affected\nVulnerability Retbleed:               Not affected\nVulnerability Spec rstack overflow:   Not affected\nVulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl\nVulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization\nVulnerability Spectre v2:             Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI SW loop, KVM SW loop\nVulnerability Srbds:                  Not affected\nVulnerability Tsx async abort:        Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==1.26.4\n[pip3] torch==2.5.1\n[pip3] torchaudio==2.5.1\n[pip3] torchvision==0.20.1\n[pip3] triton==3.1.0\n[conda] numpy                     1.26.4                   pypi_0    pypi\n[conda] torch                     2.5.1                    pypi_0    pypi\n[conda] torchaudio                2.5.1                    pypi_0    pypi\n[conda] torchvision               0.20.1                   pypi_0    pypi\n[conda] triton                    3.1.0                    pypi_0    pypi",
-  "transformers_version": "4.49.0",
-  "upper_git_hash": null,
-  "tokenizer_pad_token": [
-    "<|endoftext|>",
-    "151643"
-  ],
-  "tokenizer_eos_token": [
-    "<|im_end|>",
-    "151645"
-  ],
-  "tokenizer_bos_token": [
-    null,
-    "None"
-  ],
-  "eot_token_id": 151645,
-  "max_length": 131072,
-  "task_hashes": {
-    "aime25_nofigures": "6c6c9c546466598a7d7e6baa470c8ed21d835e4d4f78a9bffe965e1125db2916"
-  },
-  "model_source": "vllm",
-  "model_name": "Qwen/QwQ-32B",
-  "model_name_sanitized": "Qwen__QwQ-32B",
-  "system_instruction": null,
-  "system_instruction_sha": null,
-  "fewshot_as_multiturn": false,
-  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- '' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n  {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" and not message.tool_calls %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n<think>\\n' }}\n{%- endif %}\n",
-  "chat_template_sha": "61433aea8393589935ac6fdc49e937f991c796925e27241bd2b1d1080d9cf9af",
-  "start_time": 2006336.707330244,
-  "end_time": 2007590.32072118,
-  "total_evaluation_time_seconds": "1253.6133909360506"
-}
\ No newline at end of file
diff --git a/qwq131k/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T23-40-20.886537.jsonl b/qwq131k/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T23-40-20.886537.jsonl
deleted file mode 100644
index e38e8a1e568a350199621a43a6c2f8e667e00a1a..0000000000000000000000000000000000000000
--- a/qwq131k/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T23-40-20.886537.jsonl
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:51447483fb3859af32e249473b7e7db7861fd55eae543532def58553564e72ef
-size 2942433
diff --git a/qwq131ktemp/Qwen__QwQ-32B/results_2025-03-09T23-39-30.477228.json b/qwq131ktemp/Qwen__QwQ-32B/results_2025-03-09T23-39-30.477228.json
deleted file mode 100644
index bf3c341fa2a513442fc1241717e7703d96b7dfc8..0000000000000000000000000000000000000000
--- a/qwq131ktemp/Qwen__QwQ-32B/results_2025-03-09T23-39-30.477228.json
+++ /dev/null
@@ -1,129 +0,0 @@
-{
-  "results": {
-    "aime25_nofigures": {
-      "alias": "aime25_nofigures",
-      "exact_match,none": 0.6,
-      "exact_match_stderr,none": "N/A",
-      "extracted_answers,none": -1,
-      "extracted_answers_stderr,none": "N/A"
-    }
-  },
-  "group_subtasks": {
-    "aime25_nofigures": []
-  },
-  "configs": {
-    "aime25_nofigures": {
-      "task": "aime25_nofigures",
-      "tag": [
-        "math_word_problems"
-      ],
-      "dataset_path": "simplescaling/aime25_nofigures",
-      "dataset_name": "default",
-      "test_split": "train",
-      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        solution = doc.get(\"solution\", doc.get(\"orig_solution\", doc.get(\"orig_orig_solution\")))\n        problem = doc.get(\"problem\", doc.get(\"question\"))\n        answer = doc.get(\"answer\", doc.get(\"orig_answer\", doc.get(\"orig_orig_answer\")))\n        if solution is None:\n            print(\"Warning: No solution found; DOC:\", doc)\n        out_doc = {\n            \"problem\": problem,\n            \"solution\": solution,\n            \"answer\": answer,\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n    return dataset.map(_process_doc)\n",
-      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return QUERY_TEMPLATE.format(Question=doc.get(\"problem\", doc.get(\"question\")))\n",
-      "doc_to_target": "answer",
-      "unsafe_code": false,
-      "process_results": "def process_results(doc: dict, results: List[str]) -> Dict[str, int]:\n    metrics = {\"exact_match\": None, \"extracted_answers\": []}\n    # bp()\n    # Multiple results -> we are measuring cov/maj etc\n    if isinstance(results[0], list):\n        results = results[0]\n        n_res = len(results) # e.g. 64\n        n_res_list = [2**i for i in range(1, int(n_res.bit_length()))] # e.g. [2, 4, 8, 16, 32, 64]\n        metrics = {\n            **metrics,\n            \"exact_matches\": [],\n            **{f\"cov@{n}\": -1 for n in n_res_list},\n            **{f\"maj@{n}\": -1 for n in n_res_list},\n        }\n\n    if os.getenv(\"PROCESSOR\", \"\") == \"gpt-4o-mini\":\n        sampler = ChatCompletionSampler(model=\"gpt-4o-mini\")\n    else:\n        print(f\"Unknown processor: {os.getenv('PROCESSOR')}; set 'PROCESSOR=gpt-4o-mini' and 'OPENAI_API_KEY=YOUR_KEY' for best results.\")\n        sampler = None\n\n    if isinstance(doc[\"answer\"], str) and doc[\"answer\"].isdigit():\n        gt = str(int(doc[\"answer\"])) # 023 -> 23\n    else:\n        gt = str(doc[\"answer\"])\n    split_tokens = [\"<|im_start|>answer\\n\", \"<|im_start|>\"]\n\n    for i, a in enumerate(results, start=1):\n        if split_tokens[0] in a:\n            a = a.split(split_tokens[0])[-1]\n        elif split_tokens[1] in a:\n            a = a.split(split_tokens[1])[-1]\n            if \"\\n\" in a:\n                a = \"\\n\".join(a.split(\"\\n\")[1:])\n\n        if (box := last_boxed_only_string(a)) is not None:\n            a = remove_boxed(box)\n        # re.DOTALL is key such that newlines are included e.g. if it does `Answer: Here is the solution:\\n\\n10`\n        elif (matches := re.findall(ANSWER_PATTERN, a, re.DOTALL)) != []:\n            a = matches[-1]  # Get the last match\n\n        # AIME answers are from 000 to 999 so often it is a digit anyways\n        if (a.isdigit()) and (gt.isdigit()):\n            a = str(int(a)) # 023 -> 23\n        elif sampler is not None:\n            options = [gt] + list(set(metrics[\"extracted_answers\"]) - {gt})\n            if len(options) > 7:\n                # Could switch back to exact returning like in AIME in that case\n                # Problem with exact returning is that it sometimes messes up small things like a dollar sign\n                print(\"Warning: Lots of options which may harm indexing performance:\", options)            \n            # This ensures that if doc['answer'] is \\text{Evelyn} it is represented as such and not \\\\text{Evelyn}\n            options_str = \"[\" + \", \".join([\"'\" + str(o) + \"'\" for o in options]) + \"]\"\n            # a = extract_answer(sampler, options, a)\n            idx = extract_answer_idx(sampler, options_str, a)\n            if idx != \"-1\":\n                if idx.isdigit():\n                    idx = int(idx) - 1\n                    if len(options) > idx >= 0:\n                        a = options[idx]\n                    else:\n                        print(\"Warning: Index out of bounds; leaving answer unchanged\\n\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n                else:\n                    print(\"Warning: Processing did not produce integer index\\na\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n        else:\n            pass # TODO: Maybe add back legacy processing\n\n        metrics[\"extracted_answers\"].append(a)\n        a = int(a == gt)\n        if not(a): # Optional logging\n            print(\"Marked incorrect\\na \" + metrics[\"extracted_answers\"][-1] + \"\\ndoc['answer'] \" + gt)\n        if i == 1:\n            metrics[\"exact_match\"] = a\n            if \"exact_matches\" in metrics:\n                metrics[\"exact_matches\"].append(a)\n        elif i > 1:\n            metrics[\"exact_matches\"].append(a)\n            if i in n_res_list:\n                metrics[f\"cov@{i}\"] = int(1 in metrics[\"exact_matches\"])\n                metrics[f\"maj@{i}\"] = int(gt == Counter(metrics[\"extracted_answers\"]).most_common(1)[0][0])\n\n    return metrics\n",
-      "description": "",
-      "target_delimiter": " ",
-      "fewshot_delimiter": "\n\n",
-      "num_fewshot": 0,
-      "metric_list": [
-        {
-          "metric": "exact_match",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "extracted_answers",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        }
-      ],
-      "output_type": "generate_until",
-      "generation_kwargs": {
-        "until": [],
-        "do_sample": false,
-        "temperature": 0.6,
-        "max_gen_toks": 131072,
-        "top_p": 0.95
-      },
-      "repeats": 1,
-      "should_decontaminate": false,
-      "metadata": {
-        "version": 1.0
-      }
-    }
-  },
-  "versions": {
-    "aime25_nofigures": 1.0
-  },
-  "n-shot": {
-    "aime25_nofigures": 0
-  },
-  "higher_is_better": {
-    "aime25_nofigures": {
-      "exact_match": true,
-      "extracted_answers": true
-    }
-  },
-  "n-samples": {
-    "aime25_nofigures": {
-      "original": 30,
-      "effective": 30
-    }
-  },
-  "config": {
-    "model": "vllm",
-    "model_args": "pretrained=Qwen/QwQ-32B,dtype=auto,tensor_parallel_size=2",
-    "batch_size": "auto",
-    "batch_sizes": [],
-    "device": null,
-    "use_cache": null,
-    "limit": null,
-    "bootstrap_iters": 0,
-    "gen_kwargs": {
-      "max_gen_toks": 131072,
-      "temperature": 0.6,
-      "top_p": 0.95
-    },
-    "random_seed": 0,
-    "numpy_seed": 1234,
-    "torch_seed": 1234,
-    "fewshot_seed": 1234
-  },
-  "git_hash": "250c17c",
-  "date": 1741562400.9670637,
-  "pretty_env_info": "PyTorch version: 2.5.1+cu124\nIs debug build: False\nCUDA used to build PyTorch: 12.4\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 22.04.5 LTS (x86_64)\nGCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0\nClang version: 14.0.0-1ubuntu1.1\nCMake version: Could not collect\nLibc version: glibc-2.35\n\nPython version: 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:36:39) [GCC 12.3.0] (64-bit runtime)\nPython platform: Linux-6.8.0-1004-gcp-tcpx-x86_64-with-glibc2.35\nIs CUDA available: True\nCUDA runtime version: 12.4.99\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: \nGPU 0: NVIDIA H100 80GB HBM3\nGPU 1: NVIDIA H100 80GB HBM3\nGPU 2: NVIDIA H100 80GB HBM3\nGPU 3: NVIDIA H100 80GB HBM3\nGPU 4: NVIDIA H100 80GB HBM3\nGPU 5: NVIDIA H100 80GB HBM3\nGPU 6: NVIDIA H100 80GB HBM3\nGPU 7: NVIDIA H100 80GB HBM3\n\nNvidia driver version: 550.127.08\ncuDNN version: Could not collect\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\n\nCPU:\nArchitecture:                         x86_64\nCPU op-mode(s):                       32-bit, 64-bit\nAddress sizes:                        52 bits physical, 57 bits virtual\nByte Order:                           Little Endian\nCPU(s):                               208\nOn-line CPU(s) list:                  0-207\nVendor ID:                            GenuineIntel\nModel name:                           Intel(R) Xeon(R) Platinum 8481C CPU @ 2.70GHz\nCPU family:                           6\nModel:                                143\nThread(s) per core:                   2\nCore(s) per socket:                   52\nSocket(s):                            2\nStepping:                             8\nBogoMIPS:                             5399.99\nFlags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch ssbd ibrs ibpb stibp ibrs_enhanced fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rtm avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves avx_vnni avx512_bf16 arat avx512vbmi umip avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid cldemote movdiri movdir64b fsrm md_clear serialize amx_bf16 avx512_fp16 amx_tile amx_int8 arch_capabilities\nHypervisor vendor:                    KVM\nVirtualization type:                  full\nL1d cache:                            4.9 MiB (104 instances)\nL1i cache:                            3.3 MiB (104 instances)\nL2 cache:                             208 MiB (104 instances)\nL3 cache:                             210 MiB (2 instances)\nNUMA node(s):                         2\nNUMA node0 CPU(s):                    0-51,104-155\nNUMA node1 CPU(s):                    52-103,156-207\nVulnerability Gather data sampling:   Not affected\nVulnerability Itlb multihit:          Not affected\nVulnerability L1tf:                   Not affected\nVulnerability Mds:                    Not affected\nVulnerability Meltdown:               Not affected\nVulnerability Mmio stale data:        Not affected\nVulnerability Reg file data sampling: Not affected\nVulnerability Retbleed:               Not affected\nVulnerability Spec rstack overflow:   Not affected\nVulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl\nVulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization\nVulnerability Spectre v2:             Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI SW loop, KVM SW loop\nVulnerability Srbds:                  Not affected\nVulnerability Tsx async abort:        Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==1.26.4\n[pip3] torch==2.5.1\n[pip3] torchaudio==2.5.1\n[pip3] torchvision==0.20.1\n[pip3] triton==3.1.0\n[conda] numpy                     1.26.4                   pypi_0    pypi\n[conda] torch                     2.5.1                    pypi_0    pypi\n[conda] torchaudio                2.5.1                    pypi_0    pypi\n[conda] torchvision               0.20.1                   pypi_0    pypi\n[conda] triton                    3.1.0                    pypi_0    pypi",
-  "transformers_version": "4.49.0",
-  "upper_git_hash": null,
-  "tokenizer_pad_token": [
-    "<|endoftext|>",
-    "151643"
-  ],
-  "tokenizer_eos_token": [
-    "<|im_end|>",
-    "151645"
-  ],
-  "tokenizer_bos_token": [
-    null,
-    "None"
-  ],
-  "eot_token_id": 151645,
-  "max_length": 131072,
-  "task_hashes": {
-    "aime25_nofigures": "6c6c9c546466598a7d7e6baa470c8ed21d835e4d4f78a9bffe965e1125db2916"
-  },
-  "model_source": "vllm",
-  "model_name": "Qwen/QwQ-32B",
-  "model_name_sanitized": "Qwen__QwQ-32B",
-  "system_instruction": null,
-  "system_instruction_sha": null,
-  "fewshot_as_multiturn": false,
-  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- '' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n  {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" and not message.tool_calls %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n<think>\\n' }}\n{%- endif %}\n",
-  "chat_template_sha": "61433aea8393589935ac6fdc49e937f991c796925e27241bd2b1d1080d9cf9af",
-  "start_time": 2006351.846069493,
-  "end_time": 2007543.018514211,
-  "total_evaluation_time_seconds": "1191.1724447179586"
-}
\ No newline at end of file
diff --git a/qwq131ktemp/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T23-39-30.477228.jsonl b/qwq131ktemp/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T23-39-30.477228.jsonl
deleted file mode 100644
index 100abef0a68b376941a1a69a750da64a7cb9c2d2..0000000000000000000000000000000000000000
--- a/qwq131ktemp/Qwen__QwQ-32B/samples_aime25_nofigures_2025-03-09T23-39-30.477228.jsonl
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:323e8593eb6211fcba499d168ab394c7a42c11da6fd854a861cdf4806503e073
-size 2926073
diff --git a/qwq_agg64temp1/Qwen__QwQ-32B/results_2025-03-11T02-54-37.245696.json b/qwq_agg64temp1/Qwen__QwQ-32B/results_2025-03-11T02-54-37.245696.json
deleted file mode 100644
index 7d8b5ac0fe43651d8e896e1f693b0e6bab7d1dcd..0000000000000000000000000000000000000000
--- a/qwq_agg64temp1/Qwen__QwQ-32B/results_2025-03-11T02-54-37.245696.json
+++ /dev/null
@@ -1,418 +0,0 @@
-{
-  "results": {
-    "aime24_nofigures_agg64": {
-      "alias": "aime24_nofigures_agg64",
-      "exact_match,all": 0.7666666666666667,
-      "exact_match_stderr,all": "N/A",
-      "extracted_answers,all": -1,
-      "extracted_answers_stderr,all": "N/A",
-      "exact_matches,all": -1,
-      "exact_matches_stderr,all": "N/A",
-      "cov@2,all": 0.7666666666666667,
-      "cov@2_stderr,all": "N/A",
-      "cov@4,all": 0.8333333333333334,
-      "cov@4_stderr,all": "N/A",
-      "cov@8,all": 0.8666666666666667,
-      "cov@8_stderr,all": "N/A",
-      "cov@16,all": 0.8666666666666667,
-      "cov@16_stderr,all": "N/A",
-      "cov@32,all": 0.9333333333333333,
-      "cov@32_stderr,all": "N/A",
-      "cov@64,all": 0.9333333333333333,
-      "cov@64_stderr,all": "N/A",
-      "maj@2,all": 0.7666666666666667,
-      "maj@2_stderr,all": "N/A",
-      "maj@4,all": 0.7666666666666667,
-      "maj@4_stderr,all": "N/A",
-      "maj@8,all": 0.8333333333333334,
-      "maj@8_stderr,all": "N/A",
-      "maj@16,all": 0.8333333333333334,
-      "maj@16_stderr,all": "N/A",
-      "maj@32,all": 0.8,
-      "maj@32_stderr,all": "N/A",
-      "maj@64,all": 0.8,
-      "maj@64_stderr,all": "N/A"
-    },
-    "aime25_nofigures_agg64": {
-      "alias": "aime25_nofigures_agg64",
-      "exact_match,all": 0.6333333333333333,
-      "exact_match_stderr,all": "N/A",
-      "extracted_answers,all": -1,
-      "extracted_answers_stderr,all": "N/A",
-      "exact_matches,all": -1,
-      "exact_matches_stderr,all": "N/A",
-      "cov@2,all": 0.7,
-      "cov@2_stderr,all": "N/A",
-      "cov@4,all": 0.8,
-      "cov@4_stderr,all": "N/A",
-      "cov@8,all": 0.8333333333333334,
-      "cov@8_stderr,all": "N/A",
-      "cov@16,all": 0.8333333333333334,
-      "cov@16_stderr,all": "N/A",
-      "cov@32,all": 0.9,
-      "cov@32_stderr,all": "N/A",
-      "cov@64,all": 0.9333333333333333,
-      "cov@64_stderr,all": "N/A",
-      "maj@2,all": 0.6333333333333333,
-      "maj@2_stderr,all": "N/A",
-      "maj@4,all": 0.7333333333333333,
-      "maj@4_stderr,all": "N/A",
-      "maj@8,all": 0.7333333333333333,
-      "maj@8_stderr,all": "N/A",
-      "maj@16,all": 0.8,
-      "maj@16_stderr,all": "N/A",
-      "maj@32,all": 0.8,
-      "maj@32_stderr,all": "N/A",
-      "maj@64,all": 0.8,
-      "maj@64_stderr,all": "N/A"
-    }
-  },
-  "group_subtasks": {
-    "aime24_nofigures_agg64": [],
-    "aime25_nofigures_agg64": []
-  },
-  "configs": {
-    "aime24_nofigures_agg64": {
-      "task": "aime24_nofigures_agg64",
-      "tag": [
-        "math_word_problems"
-      ],
-      "dataset_path": "simplescaling/aime24_nofigures",
-      "dataset_name": "default",
-      "test_split": "train",
-      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        solution = doc.get(\"solution\", doc.get(\"orig_solution\", doc.get(\"orig_orig_solution\")))\n        problem = doc.get(\"problem\", doc.get(\"question\"))\n        answer = doc.get(\"answer\", doc.get(\"orig_answer\", doc.get(\"orig_orig_answer\")))\n        if solution is None:\n            print(\"Warning: No solution found; DOC:\", doc)\n        out_doc = {\n            \"problem\": problem,\n            \"solution\": solution,\n            \"answer\": answer,\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n    return dataset.map(_process_doc)\n",
-      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return QUERY_TEMPLATE.format(Question=doc.get(\"problem\", doc.get(\"question\")))\n",
-      "doc_to_target": "answer",
-      "unsafe_code": false,
-      "process_results": "def process_results(doc: dict, results: List[str]) -> Dict[str, int]:\n    metrics = {\"exact_match\": None, \"extracted_answers\": []}\n    # bp()\n    # Multiple results -> we are measuring cov/maj etc\n    if isinstance(results[0], list):\n        results = results[0]\n        n_res = len(results) # e.g. 64\n        n_res_list = [2**i for i in range(1, int(n_res.bit_length()))] # e.g. [2, 4, 8, 16, 32, 64]\n        metrics = {\n            **metrics,\n            \"exact_matches\": [],\n            **{f\"cov@{n}\": -1 for n in n_res_list},\n            **{f\"maj@{n}\": -1 for n in n_res_list},\n        }\n\n    if os.getenv(\"PROCESSOR\", \"\") == \"gpt-4o-mini\":\n        sampler = ChatCompletionSampler(model=\"gpt-4o-mini\")\n    else:\n        print(f\"Unknown processor: {os.getenv('PROCESSOR')}; set 'PROCESSOR=gpt-4o-mini' and 'OPENAI_API_KEY=YOUR_KEY' for best results.\")\n        sampler = None\n\n    if isinstance(doc[\"answer\"], str) and doc[\"answer\"].isdigit():\n        gt = str(int(doc[\"answer\"])) # 023 -> 23\n    else:\n        gt = str(doc[\"answer\"])\n    split_tokens = [\"<|im_start|>answer\\n\", \"<|im_start|>\"]\n\n    for i, a in enumerate(results, start=1):\n        if split_tokens[0] in a:\n            a = a.split(split_tokens[0])[-1]\n        elif split_tokens[1] in a:\n            a = a.split(split_tokens[1])[-1]\n            if \"\\n\" in a:\n                a = \"\\n\".join(a.split(\"\\n\")[1:])\n\n        if (box := last_boxed_only_string(a)) is not None:\n            a = remove_boxed(box)\n        # re.DOTALL is key such that newlines are included e.g. if it does `Answer: Here is the solution:\\n\\n10`\n        elif (matches := re.findall(ANSWER_PATTERN, a, re.DOTALL)) != []:\n            a = matches[-1]  # Get the last match\n\n        # AIME answers are from 000 to 999 so often it is a digit anyways\n        if (a.isdigit()) and (gt.isdigit()):\n            a = str(int(a)) # 023 -> 23\n        elif sampler is not None:\n            options = [gt] + list(set(metrics[\"extracted_answers\"]) - {gt})\n            if len(options) > 7:\n                # Could switch back to exact returning like in AIME in that case\n                # Problem with exact returning is that it sometimes messes up small things like a dollar sign\n                print(\"Warning: Lots of options which may harm indexing performance:\", options)            \n            # This ensures that if doc['answer'] is \\text{Evelyn} it is represented as such and not \\\\text{Evelyn}\n            options_str = \"[\" + \", \".join([\"'\" + str(o) + \"'\" for o in options]) + \"]\"\n            # a = extract_answer(sampler, options, a)\n            idx = extract_answer_idx(sampler, options_str, a)\n            if idx != \"-1\":\n                if idx.isdigit():\n                    idx = int(idx) - 1\n                    if len(options) > idx >= 0:\n                        a = options[idx]\n                    else:\n                        print(\"Warning: Index out of bounds; leaving answer unchanged\\n\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n                else:\n                    print(\"Warning: Processing did not produce integer index\\na\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n        else:\n            pass # TODO: Maybe add back legacy processing\n\n        metrics[\"extracted_answers\"].append(a)\n        a = int(a == gt)\n        if not(a): # Optional logging\n            print(\"Marked incorrect\\na \" + metrics[\"extracted_answers\"][-1] + \"\\ndoc['answer'] \" + gt)\n        if i == 1:\n            metrics[\"exact_match\"] = a\n            if \"exact_matches\" in metrics:\n                metrics[\"exact_matches\"].append(a)\n        elif i > 1:\n            metrics[\"exact_matches\"].append(a)\n            if i in n_res_list:\n                metrics[f\"cov@{i}\"] = int(1 in metrics[\"exact_matches\"])\n                metrics[f\"maj@{i}\"] = int(gt == Counter(metrics[\"extracted_answers\"]).most_common(1)[0][0])\n\n    return metrics\n",
-      "description": "",
-      "target_delimiter": " ",
-      "fewshot_delimiter": "\n\n",
-      "num_fewshot": 0,
-      "metric_list": [
-        {
-          "metric": "exact_match",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@64",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@32",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@16",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@8",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@4",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@2",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@64",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@32",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@16",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@8",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@4",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@2",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "extracted_answers",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        },
-        {
-          "metric": "exact_matches",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        }
-      ],
-      "output_type": "generate_until",
-      "generation_kwargs": {
-        "until": [],
-        "do_sample": false,
-        "temperature": 1,
-        "max_gen_toks": 32768
-      },
-      "repeats": 64,
-      "filter_list": [
-        {
-          "name": "all",
-          "filter": [
-            {
-              "function": "take_first_k",
-              "k": 64
-            }
-          ]
-        }
-      ],
-      "should_decontaminate": false,
-      "metadata": {
-        "version": 1.0
-      }
-    },
-    "aime25_nofigures_agg64": {
-      "task": "aime25_nofigures_agg64",
-      "tag": [
-        "math_word_problems"
-      ],
-      "dataset_path": "simplescaling/aime25_nofigures",
-      "dataset_name": "default",
-      "test_split": "train",
-      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        solution = doc.get(\"solution\", doc.get(\"orig_solution\", doc.get(\"orig_orig_solution\")))\n        problem = doc.get(\"problem\", doc.get(\"question\"))\n        answer = doc.get(\"answer\", doc.get(\"orig_answer\", doc.get(\"orig_orig_answer\")))\n        if solution is None:\n            print(\"Warning: No solution found; DOC:\", doc)\n        out_doc = {\n            \"problem\": problem,\n            \"solution\": solution,\n            \"answer\": answer,\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n    return dataset.map(_process_doc)\n",
-      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return QUERY_TEMPLATE.format(Question=doc.get(\"problem\", doc.get(\"question\")))\n",
-      "doc_to_target": "answer",
-      "unsafe_code": false,
-      "process_results": "def process_results(doc: dict, results: List[str]) -> Dict[str, int]:\n    metrics = {\"exact_match\": None, \"extracted_answers\": []}\n    # bp()\n    # Multiple results -> we are measuring cov/maj etc\n    if isinstance(results[0], list):\n        results = results[0]\n        n_res = len(results) # e.g. 64\n        n_res_list = [2**i for i in range(1, int(n_res.bit_length()))] # e.g. [2, 4, 8, 16, 32, 64]\n        metrics = {\n            **metrics,\n            \"exact_matches\": [],\n            **{f\"cov@{n}\": -1 for n in n_res_list},\n            **{f\"maj@{n}\": -1 for n in n_res_list},\n        }\n\n    if os.getenv(\"PROCESSOR\", \"\") == \"gpt-4o-mini\":\n        sampler = ChatCompletionSampler(model=\"gpt-4o-mini\")\n    else:\n        print(f\"Unknown processor: {os.getenv('PROCESSOR')}; set 'PROCESSOR=gpt-4o-mini' and 'OPENAI_API_KEY=YOUR_KEY' for best results.\")\n        sampler = None\n\n    if isinstance(doc[\"answer\"], str) and doc[\"answer\"].isdigit():\n        gt = str(int(doc[\"answer\"])) # 023 -> 23\n    else:\n        gt = str(doc[\"answer\"])\n    split_tokens = [\"<|im_start|>answer\\n\", \"<|im_start|>\"]\n\n    for i, a in enumerate(results, start=1):\n        if split_tokens[0] in a:\n            a = a.split(split_tokens[0])[-1]\n        elif split_tokens[1] in a:\n            a = a.split(split_tokens[1])[-1]\n            if \"\\n\" in a:\n                a = \"\\n\".join(a.split(\"\\n\")[1:])\n\n        if (box := last_boxed_only_string(a)) is not None:\n            a = remove_boxed(box)\n        # re.DOTALL is key such that newlines are included e.g. if it does `Answer: Here is the solution:\\n\\n10`\n        elif (matches := re.findall(ANSWER_PATTERN, a, re.DOTALL)) != []:\n            a = matches[-1]  # Get the last match\n\n        # AIME answers are from 000 to 999 so often it is a digit anyways\n        if (a.isdigit()) and (gt.isdigit()):\n            a = str(int(a)) # 023 -> 23\n        elif sampler is not None:\n            options = [gt] + list(set(metrics[\"extracted_answers\"]) - {gt})\n            if len(options) > 7:\n                # Could switch back to exact returning like in AIME in that case\n                # Problem with exact returning is that it sometimes messes up small things like a dollar sign\n                print(\"Warning: Lots of options which may harm indexing performance:\", options)            \n            # This ensures that if doc['answer'] is \\text{Evelyn} it is represented as such and not \\\\text{Evelyn}\n            options_str = \"[\" + \", \".join([\"'\" + str(o) + \"'\" for o in options]) + \"]\"\n            # a = extract_answer(sampler, options, a)\n            idx = extract_answer_idx(sampler, options_str, a)\n            if idx != \"-1\":\n                if idx.isdigit():\n                    idx = int(idx) - 1\n                    if len(options) > idx >= 0:\n                        a = options[idx]\n                    else:\n                        print(\"Warning: Index out of bounds; leaving answer unchanged\\n\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n                else:\n                    print(\"Warning: Processing did not produce integer index\\na\", a, \"\\noptions\", options_str, \"\\ndoc['answer']\", gt, \"\\nidx\", idx)\n        else:\n            pass # TODO: Maybe add back legacy processing\n\n        metrics[\"extracted_answers\"].append(a)\n        a = int(a == gt)\n        if not(a): # Optional logging\n            print(\"Marked incorrect\\na \" + metrics[\"extracted_answers\"][-1] + \"\\ndoc['answer'] \" + gt)\n        if i == 1:\n            metrics[\"exact_match\"] = a\n            if \"exact_matches\" in metrics:\n                metrics[\"exact_matches\"].append(a)\n        elif i > 1:\n            metrics[\"exact_matches\"].append(a)\n            if i in n_res_list:\n                metrics[f\"cov@{i}\"] = int(1 in metrics[\"exact_matches\"])\n                metrics[f\"maj@{i}\"] = int(gt == Counter(metrics[\"extracted_answers\"]).most_common(1)[0][0])\n\n    return metrics\n",
-      "description": "",
-      "target_delimiter": " ",
-      "fewshot_delimiter": "\n\n",
-      "num_fewshot": 0,
-      "metric_list": [
-        {
-          "metric": "exact_match",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@64",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@32",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@16",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@8",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@4",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "cov@2",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@64",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@32",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@16",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@8",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@4",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "maj@2",
-          "aggregation": "mean",
-          "higher_is_better": true
-        },
-        {
-          "metric": "extracted_answers",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        },
-        {
-          "metric": "exact_matches",
-          "aggregation": "bypass",
-          "higher_is_better": true
-        }
-      ],
-      "output_type": "generate_until",
-      "generation_kwargs": {
-        "until": [],
-        "do_sample": false,
-        "temperature": 1,
-        "max_gen_toks": 32768
-      },
-      "repeats": 64,
-      "filter_list": [
-        {
-          "name": "all",
-          "filter": [
-            {
-              "function": "take_first_k",
-              "k": 64
-            }
-          ]
-        }
-      ],
-      "should_decontaminate": false,
-      "metadata": {
-        "version": 1.0
-      }
-    }
-  },
-  "versions": {
-    "aime24_nofigures_agg64": 1.0,
-    "aime25_nofigures_agg64": 1.0
-  },
-  "n-shot": {
-    "aime24_nofigures_agg64": 0,
-    "aime25_nofigures_agg64": 0
-  },
-  "higher_is_better": {
-    "aime24_nofigures_agg64": {
-      "exact_match": true,
-      "cov@64": true,
-      "cov@32": true,
-      "cov@16": true,
-      "cov@8": true,
-      "cov@4": true,
-      "cov@2": true,
-      "maj@64": true,
-      "maj@32": true,
-      "maj@16": true,
-      "maj@8": true,
-      "maj@4": true,
-      "maj@2": true,
-      "extracted_answers": true,
-      "exact_matches": true
-    },
-    "aime25_nofigures_agg64": {
-      "exact_match": true,
-      "cov@64": true,
-      "cov@32": true,
-      "cov@16": true,
-      "cov@8": true,
-      "cov@4": true,
-      "cov@2": true,
-      "maj@64": true,
-      "maj@32": true,
-      "maj@16": true,
-      "maj@8": true,
-      "maj@4": true,
-      "maj@2": true,
-      "extracted_answers": true,
-      "exact_matches": true
-    }
-  },
-  "n-samples": {
-    "aime25_nofigures_agg64": {
-      "original": 30,
-      "effective": 30
-    },
-    "aime24_nofigures_agg64": {
-      "original": 30,
-      "effective": 30
-    }
-  },
-  "config": {
-    "model": "vllm",
-    "model_args": "pretrained=Qwen/QwQ-32B,dtype=auto,tensor_parallel_size=2",
-    "batch_size": "auto",
-    "batch_sizes": [],
-    "device": null,
-    "use_cache": null,
-    "limit": null,
-    "bootstrap_iters": 0,
-    "gen_kwargs": {
-      "max_gen_toks": 32768,
-      "temperature": 1
-    },
-    "random_seed": 0,
-    "numpy_seed": 1234,
-    "torch_seed": 1234,
-    "fewshot_seed": 1234
-  },
-  "git_hash": "250c17c",
-  "date": 1741566445.8850312,
-  "pretty_env_info": "PyTorch version: 2.5.1+cu124\nIs debug build: False\nCUDA used to build PyTorch: 12.4\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 22.04.5 LTS (x86_64)\nGCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0\nClang version: 14.0.0-1ubuntu1.1\nCMake version: Could not collect\nLibc version: glibc-2.35\n\nPython version: 3.10.13 | packaged by conda-forge | (main, Dec 23 2023, 15:36:39) [GCC 12.3.0] (64-bit runtime)\nPython platform: Linux-6.8.0-1004-gcp-tcpx-x86_64-with-glibc2.35\nIs CUDA available: True\nCUDA runtime version: 12.4.99\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: \nGPU 0: NVIDIA H100 80GB HBM3\nGPU 1: NVIDIA H100 80GB HBM3\nGPU 2: NVIDIA H100 80GB HBM3\nGPU 3: NVIDIA H100 80GB HBM3\nGPU 4: NVIDIA H100 80GB HBM3\nGPU 5: NVIDIA H100 80GB HBM3\nGPU 6: NVIDIA H100 80GB HBM3\nGPU 7: NVIDIA H100 80GB HBM3\n\nNvidia driver version: 550.127.08\ncuDNN version: Could not collect\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\n\nCPU:\nArchitecture:                         x86_64\nCPU op-mode(s):                       32-bit, 64-bit\nAddress sizes:                        52 bits physical, 57 bits virtual\nByte Order:                           Little Endian\nCPU(s):                               208\nOn-line CPU(s) list:                  0-207\nVendor ID:                            GenuineIntel\nModel name:                           Intel(R) Xeon(R) Platinum 8481C CPU @ 2.70GHz\nCPU family:                           6\nModel:                                143\nThread(s) per core:                   2\nCore(s) per socket:                   52\nSocket(s):                            2\nStepping:                             8\nBogoMIPS:                             5399.99\nFlags:                                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch ssbd ibrs ibpb stibp ibrs_enhanced fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rtm avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves avx_vnni avx512_bf16 arat avx512vbmi umip avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid cldemote movdiri movdir64b fsrm md_clear serialize amx_bf16 avx512_fp16 amx_tile amx_int8 arch_capabilities\nHypervisor vendor:                    KVM\nVirtualization type:                  full\nL1d cache:                            4.9 MiB (104 instances)\nL1i cache:                            3.3 MiB (104 instances)\nL2 cache:                             208 MiB (104 instances)\nL3 cache:                             210 MiB (2 instances)\nNUMA node(s):                         2\nNUMA node0 CPU(s):                    0-51,104-155\nNUMA node1 CPU(s):                    52-103,156-207\nVulnerability Gather data sampling:   Not affected\nVulnerability Itlb multihit:          Not affected\nVulnerability L1tf:                   Not affected\nVulnerability Mds:                    Not affected\nVulnerability Meltdown:               Not affected\nVulnerability Mmio stale data:        Not affected\nVulnerability Reg file data sampling: Not affected\nVulnerability Retbleed:               Not affected\nVulnerability Spec rstack overflow:   Not affected\nVulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl\nVulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization\nVulnerability Spectre v2:             Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI SW loop, KVM SW loop\nVulnerability Srbds:                  Not affected\nVulnerability Tsx async abort:        Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==1.26.4\n[pip3] torch==2.5.1\n[pip3] torchaudio==2.5.1\n[pip3] torchvision==0.20.1\n[pip3] triton==3.1.0\n[conda] numpy                     1.26.4                   pypi_0    pypi\n[conda] torch                     2.5.1                    pypi_0    pypi\n[conda] torchaudio                2.5.1                    pypi_0    pypi\n[conda] torchvision               0.20.1                   pypi_0    pypi\n[conda] triton                    3.1.0                    pypi_0    pypi",
-  "transformers_version": "4.49.0",
-  "upper_git_hash": null,
-  "tokenizer_pad_token": [
-    "<|endoftext|>",
-    "151643"
-  ],
-  "tokenizer_eos_token": [
-    "<|im_end|>",
-    "151645"
-  ],
-  "tokenizer_bos_token": [
-    null,
-    "None"
-  ],
-  "eot_token_id": 151645,
-  "max_length": 131072,
-  "task_hashes": {
-    "aime25_nofigures_agg64": "6c6c9c546466598a7d7e6baa470c8ed21d835e4d4f78a9bffe965e1125db2916",
-    "aime24_nofigures_agg64": "5559473020237fa3e670d94426cd56a532b8f02b3cba30c936a326b25f86b169"
-  },
-  "model_source": "vllm",
-  "model_name": "Qwen/QwQ-32B",
-  "model_name_sanitized": "Qwen__QwQ-32B",
-  "system_instruction": null,
-  "system_instruction_sha": null,
-  "fewshot_as_multiturn": false,
-  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- '' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n  {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" and not message.tool_calls %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n<think>\\n' }}\n{%- endif %}\n",
-  "chat_template_sha": "61433aea8393589935ac6fdc49e937f991c796925e27241bd2b1d1080d9cf9af",
-  "start_time": 2010403.261026473,
-  "end_time": 2105649.786818585,
-  "total_evaluation_time_seconds": "95246.52579211188"
-}
\ No newline at end of file
diff --git a/qwq_agg64temp1/Qwen__QwQ-32B/samples_aime24_nofigures_agg64_2025-03-11T02-54-37.245696.jsonl b/qwq_agg64temp1/Qwen__QwQ-32B/samples_aime24_nofigures_agg64_2025-03-11T02-54-37.245696.jsonl
deleted file mode 100644
index b9f0321b63862b8fa6bf1d890a0e70b65e38c29a..0000000000000000000000000000000000000000
--- a/qwq_agg64temp1/Qwen__QwQ-32B/samples_aime24_nofigures_agg64_2025-03-11T02-54-37.245696.jsonl
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c0878b0489bac8d6cfbf0567daea25040e5eccba11721e0866b5373ac61d6fd1
-size 183322234
diff --git a/qwq_agg64temp1/Qwen__QwQ-32B/samples_aime25_nofigures_agg64_2025-03-11T02-54-37.245696.jsonl b/qwq_agg64temp1/Qwen__QwQ-32B/samples_aime25_nofigures_agg64_2025-03-11T02-54-37.245696.jsonl
deleted file mode 100644
index 2e678b2a890fb36ea54fcf917aabd1304a5153cc..0000000000000000000000000000000000000000
--- a/qwq_agg64temp1/Qwen__QwQ-32B/samples_aime25_nofigures_agg64_2025-03-11T02-54-37.245696.jsonl
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:47e0a9f76570c1a6b42dc51ae6d7aa8df1cc2fe602cccad622708e670fadd26e
-size 212237535
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/results_2025-04-03T22-06-15.102588.json b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/results_2025-04-03T22-06-15.102588.json
new file mode 100644
index 0000000000000000000000000000000000000000..bd7006f9e51e6b27b97c693447ce0e1d5064c3e6
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/results_2025-04-03T22-06-15.102588.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b8fb6f8b34800e9bee79dafbc5d179d5fa2cbd96df5e51b92896482a69cf228
+size 55562
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-03T22-06-15.102588.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-03T22-06-15.102588.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2e8c64d6acae1c6a1406df06416a56b3d0cfe11c
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-03T22-06-15.102588.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9e91a0e8482c55e5e8c18a6b91d8ea472571e7b7256d440fafee06391d83104
+size 170940368
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-03T22-06-15.102588.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-03T22-06-15.102588.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..e6f74434c95e07166a4c2b1afb9f5523322c0291
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-03T22-06-15.102588.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3814edda42cf68aaf8c2e94497961b2a75831192a6ebfd9be2f01a244776c93
+size 118846156
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/results_2025-04-03T22-00-29.567285.json b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/results_2025-04-03T22-00-29.567285.json
new file mode 100644
index 0000000000000000000000000000000000000000..ceaacf5069e3ad55bfb83e2725abb72343fbb8b5
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/results_2025-04-03T22-00-29.567285.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6d2583a419f4372c85f2ac67c61b57de1f26ffa550eab0e3f2ab3c24d80e620
+size 55358
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-03T22-00-29.567285.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-03T22-00-29.567285.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..02b9312bc4bf12a9c41b4bca2ad2fe9e58ad66b7
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-03T22-00-29.567285.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66c2edfa99b4c7651999a6899b50ddcf506680b98e134f411e6a62135895cbe6
+size 165353624
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-03T22-00-29.567285.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-03T22-00-29.567285.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..2a08668f4d5f0860b03af9e1d50645592006d307
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-03T22-00-29.567285.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c48df1453d75b54cd64056bdb9d6a079fad118c1c9015a0dcb73c2d5b690f177
+size 115676739
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/results_2025-04-04T00-57-01.239273.json b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/results_2025-04-04T00-57-01.239273.json
new file mode 100644
index 0000000000000000000000000000000000000000..370c5ef7bcfda07031c11bee6345ab8f2a7d6733
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/results_2025-04-04T00-57-01.239273.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c144a1d60139ea6c0033d701209c94db5846e6a239cfe6150a80b568d710b751
+size 55340
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-04T00-57-01.239273.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-04T00-57-01.239273.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..08fb16f89d3fa1c27cec76eacb27a5caed002386
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-04T00-57-01.239273.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a11708ae0dae4b9e43ba00c9c1790ab7519dc40fe2cfb3582b6d7fb12289e1b
+size 178793629
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-04T00-57-01.239273.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-04T00-57-01.239273.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3b9f4d81735c075abc3c1e216d39e221dba04897
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-04T00-57-01.239273.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbcdd4f2320d0e614c17b75bfdf89f7dd96e9a321738f4b7be38551a8c640374
+size 135153458
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/results_2025-04-04T00-41-50.574430.json b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/results_2025-04-04T00-41-50.574430.json
new file mode 100644
index 0000000000000000000000000000000000000000..7ea11a6c53b8f6561bd74629144bd115295f57c6
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/results_2025-04-04T00-41-50.574430.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ad062a8e1f43015a34d53e1db0d9d8e9b64711f5f748716b59e6ddd53c7552f
+size 55303
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-04T00-41-50.574430.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-04T00-41-50.574430.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..24d37fa8feb9182125b68dd56a82c06cdb0eb920
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-04T00-41-50.574430.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f9f1f850194eca5c42582ce86b859bc97885b9822729262a8f3203c3f2632fd
+size 168584234
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-04T00-41-50.574430.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-04T00-41-50.574430.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..5601334545ce1adde0f79e292c668e113e1b5660
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-04T00-41-50.574430.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07b9ed30b7870862657fa6595d8f784eed6802ba629c5ac63882e9603d92b976
+size 122892520
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/results_2025-04-04T02-08-03.821182.json b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/results_2025-04-04T02-08-03.821182.json
new file mode 100644
index 0000000000000000000000000000000000000000..be3b09583090fce2331fea74874a1609587fe37d
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/results_2025-04-04T02-08-03.821182.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c1c3d7a1eaff146892bde5db4d8e5cc724a7748f2201f3d3c86cf9fef31c445
+size 55591
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/results_2025-04-04T22-48-19.450896.json b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/results_2025-04-04T22-48-19.450896.json
new file mode 100644
index 0000000000000000000000000000000000000000..fa4c915d4d15054b8757067991a89d8aadcf7865
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/results_2025-04-04T22-48-19.450896.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af6e9f59934f6a845391756b05685cdb89d54f0a0651b2bcf98027fef67ee06e
+size 27872
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-04T22-48-19.450896.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-04T22-48-19.450896.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..df142e0c8ba2c392f9647e01d46d8ecfd4b3385e
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-04T22-48-19.450896.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9f4cf9c0d3914d9b17b0760cf646cb636f949648aa1b00453407ab07802289e
+size 158936585
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-04T02-08-03.821182.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-04T02-08-03.821182.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..9de462305e12862622771cb6330b59df27f50477
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-04T02-08-03.821182.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f42b891330d471fc59131b596018a086e6961d3bda6902972d44bc675708fbd
+size 170869980
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-04T02-08-03.821182.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-04T02-08-03.821182.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..0321dd33c2589239737daeba018ef50c0176b7ec
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-04T02-08-03.821182.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:51889f0926ee1e9fe1461f069eca4e82a15ada131766175634efda47e6ac2e74
+size 122309040
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/results_2025-04-04T00-54-02.924841.json b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/results_2025-04-04T00-54-02.924841.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ec2194d848b002467cfe3bc29a83a7f0af002cc
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/results_2025-04-04T00-54-02.924841.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:683db82c0e9a5abd3f75f2689594bd44d40082a4f3c5d879ee6581f16eeec22a
+size 55447
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/results_2025-04-08T23-35-58.097461.json b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/results_2025-04-08T23-35-58.097461.json
new file mode 100644
index 0000000000000000000000000000000000000000..21a8f2ce33a53b29b95bb5a726e27c94a41e9fe6
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/results_2025-04-08T23-35-58.097461.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac7331308f8486501e99cd2c2a617124e33242e64ff9f13a9b319d0f6dd8efc8
+size 27933
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T23-35-58.097461.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T23-35-58.097461.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..bb5aee74c37f77cd095c804b10307885f5629a1c
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T23-35-58.097461.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a45adfee31df83e9fe3af01e7975b51574f0560410c9037b3f3b802cb048fc2f
+size 176822105
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-04T00-54-02.924841.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-04T00-54-02.924841.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..20c47bfe4439876c341ad4222a44b91ee99f6f78
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-04T00-54-02.924841.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2516b9ae746bc8a66e28633bcf0829c32b82480892b8a61d1aec67f0a26e7571
+size 174874953
diff --git a/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-04T00-54-02.924841.jsonl b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-04T00-54-02.924841.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..bba0fadb80ccd374448a52cd8f5a086f88584e74
--- /dev/null
+++ b/s2-1.5B-v4/__data__niklas__s2__open-r1__data__s2-1.5B-v4__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-04T00-54-02.924841.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1a15fefe020d9c90d6022b8b8044d0c19c49e6823206f8451f3077ad18e8e05
+size 124166355
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/results_2025-04-06T06-39-11.509480.json b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/results_2025-04-06T06-39-11.509480.json
new file mode 100644
index 0000000000000000000000000000000000000000..56378f03d5af1fc4d6c00c2f058bd0fcb64c325e
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/results_2025-04-06T06-39-11.509480.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d32d156d7282e3ffc2557d9fc16756d17a8a1fa16b5e3fcf8d172b954d26de78
+size 55497
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-06T06-39-11.509480.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-06T06-39-11.509480.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..20e3e4f861399401fcdee16e07e35978b90a1693
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-06T06-39-11.509480.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1da83c26849b8349d77eebdddf72dbfac22e9f858e4fbdc640d94116cfb49cc1
+size 182260749
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-06T06-39-11.509480.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-06T06-39-11.509480.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3bf3e9c7ea54961bc38d3471cb5071785bbeaaf3
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-06T06-39-11.509480.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9bf60fab38bfffae8578b165de4cd9076ebd5604e1a7ce7c5a8b89b6cfced737
+size 111938727
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/results_2025-04-06T06-56-03.478011.json b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/results_2025-04-06T06-56-03.478011.json
new file mode 100644
index 0000000000000000000000000000000000000000..52d26e160d7c85cdbf26af9ff7cef308aca465d2
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/results_2025-04-06T06-56-03.478011.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90f77182b4b68d8a5bd60d425ea704b2ec23b8f28dc08dfe0cbc75bf019821ce
+size 55582
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-06T06-56-03.478011.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-06T06-56-03.478011.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..8ba2e640ccea9a7254af0cada7f66f63abee2f03
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-06T06-56-03.478011.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f490cded89ec0a3a9eeb756e2aa38a2f61535ae0b14e43be29fd852c3c8d3a95
+size 178070166
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-06T06-56-03.478011.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-06T06-56-03.478011.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..395768903284a52696efc2f4ddb0f0fdb6d96492
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-06T06-56-03.478011.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:658fcaa4bf837ee10a73e740a26e4613f95f126b99001b18fc2f518ccc1332d8
+size 111227940
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/results_2025-04-06T08-04-41.184191.json b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/results_2025-04-06T08-04-41.184191.json
new file mode 100644
index 0000000000000000000000000000000000000000..81f97b5472945478ab70f00310b07ba6937ff347
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/results_2025-04-06T08-04-41.184191.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:235906374e60b692d44f8d0f1ba11d406cb432ea91dca7ce2c4463a5d9fcd75c
+size 55427
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-06T08-04-41.184191.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-06T08-04-41.184191.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..70fe96e9178308ecbb1f9cc84848e0fd10ae7874
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-06T08-04-41.184191.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8874e6f101fc7d2700db072b13a8acc67ed7f724a32d1b1d0b895ca6f3f382cf
+size 174242818
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-06T08-04-41.184191.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-06T08-04-41.184191.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..5e9e9a1a569568579efa696c2b13a6bb51189a00
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-06T08-04-41.184191.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f27f5738ca68aaac34715815895b295c69629c251568cc89d10ea78232a43f1
+size 108809789
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/results_2025-04-06T03-57-17.586717.json b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/results_2025-04-06T03-57-17.586717.json
new file mode 100644
index 0000000000000000000000000000000000000000..6465fcc8888441794de05ba75ccf19141b76e613
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/results_2025-04-06T03-57-17.586717.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e07cf40eb2c5b08bf82141a7b9eb88e0ce86cb68d4973ab80b2f1276645b0fb
+size 55636
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-06T03-57-17.586717.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-06T03-57-17.586717.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ddeebec870d5b6f7d533f8530b71a8ae9c39d836
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-06T03-57-17.586717.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ddda5cafa65e1e4ec53f1f201c1c91c7529623a2cc53b4c27618a107f3648875
+size 167090284
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-06T03-57-17.586717.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-06T03-57-17.586717.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..be47c530895e03c7660bcfd7ced7d036a5aff6d2
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-06T03-57-17.586717.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e7475f5ca226b44e796e01980d58d0596110816ac773b61d314472668883c7c
+size 117068098
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/results_2025-04-06T05-06-14.582869.json b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/results_2025-04-06T05-06-14.582869.json
new file mode 100644
index 0000000000000000000000000000000000000000..88279b40efdfd7f21aabd50a1acc292f6a27eb1e
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/results_2025-04-06T05-06-14.582869.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:81f485b0523de9fb6695658b2d0b6eb285051f6b6c1b13fb219019a238629641
+size 74986
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-06T05-06-14.582869.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-06T05-06-14.582869.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..d8280fa4835387905d141f5fc02b8aeb1e55678f
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-06T05-06-14.582869.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ff429eaa40760d0317d667446dc3e1f787a6fc3ccb62805486cd1f4d105b627
+size 175260595
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-06T05-06-14.582869.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-06T05-06-14.582869.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..753b01fe97bbbe60fb9c1590b1beb4d48d520728
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-06T05-06-14.582869.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0fc25fa27bde520486e1cad6e1076532ee9a26950915800c03efebf82f613ead
+size 183353007
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-06T05-06-14.582869.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-06T05-06-14.582869.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..8f0d5440c5676efc436128887f622d257e3fb477
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-06T05-06-14.582869.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b49b54d2d7236a9042561b6a6a790397f5199dfaeb0c7fc513f27c6a0a7dd899
+size 115849850
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-06T05-36-11.778357.json b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-06T05-36-11.778357.json
new file mode 100644
index 0000000000000000000000000000000000000000..841a7865eb46229a333f7415aef43046d2e21b06
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-06T05-36-11.778357.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:190d7c7d6767ec1d98a3e04ff8faad0e158a0f5347822e6c2ad14cb34d2f8518
+size 55425
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-18-25.309796.json b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-18-25.309796.json
new file mode 100644
index 0000000000000000000000000000000000000000..be186a2250296d29961a8b590f27ca571062a196
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-18-25.309796.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:616b7d0e689172134fd58b8d9a0ddae440b2e4992f4a8809b000ee65f8bda1a8
+size 27946
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-21-17.818231.json b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-21-17.818231.json
new file mode 100644
index 0000000000000000000000000000000000000000..c18fc4b6b6f13930b4c08a982906ea857e842370
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-21-17.818231.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b79e1e452111b3784a7ca72c5744ae5417095329d7a8d3c23d8be89e2cb42ee0
+size 27945
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-29-56.401365.json b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-29-56.401365.json
new file mode 100644
index 0000000000000000000000000000000000000000..03e255c05621917bf3fd2f239fea20e2a94d590a
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/results_2025-04-08T20-29-56.401365.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97c0d6655eddf03500d824f2cb0e819bc3a2eccef3a9dc86e7240032ea7692d4
+size 27947
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-18-25.309796.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-18-25.309796.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6c01a19075086bdcd4a9ab4b505b0109932c809a
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-18-25.309796.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bca416b8a90c6c1a6668673f51e43e5e23f62b76a629e33dbc3d045fa71e7c7
+size 181733174
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-21-17.818231.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-21-17.818231.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6c01a19075086bdcd4a9ab4b505b0109932c809a
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-21-17.818231.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bca416b8a90c6c1a6668673f51e43e5e23f62b76a629e33dbc3d045fa71e7c7
+size 181733174
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-29-56.401365.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-29-56.401365.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..6c01a19075086bdcd4a9ab4b505b0109932c809a
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-08T20-29-56.401365.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bca416b8a90c6c1a6668673f51e43e5e23f62b76a629e33dbc3d045fa71e7c7
+size 181733174
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-06T05-36-11.778357.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-06T05-36-11.778357.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..f1d38a49522fea770bec957896f75f701694d62b
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-06T05-36-11.778357.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b5b0c260ffc7900bc4dbe6e35ef9d432b3ed69ddb53f745eefced64282ab68c
+size 182997242
diff --git a/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-06T05-36-11.778357.jsonl b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-06T05-36-11.778357.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..99eb05a5747b5d72a9b3ec4203f1f6f0be1b7acc
--- /dev/null
+++ b/s2-1.5B-v5/__data__niklas__s2__open-r1__data__s2-1.5B-v5__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-06T05-36-11.778357.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:493074b6968ddd6f5d9b11fc0e6d104540f443dcfce04190ad578b082e3646d9
+size 105553813
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/results_2025-04-07T20-31-59.584821.json b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/results_2025-04-07T20-31-59.584821.json
new file mode 100644
index 0000000000000000000000000000000000000000..8b478047e4b4ed04ff343212dfc7fb276b822fdd
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/results_2025-04-07T20-31-59.584821.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7720ab540938430d1324c89322f99fe702199b0242f8b1713b946c44bc62b4cc
+size 75086
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-07T20-31-59.584821.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-07T20-31-59.584821.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3fe53ecb28e38035a212045be7953c098461c870
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_aime24_nofigures_agg64_2025-04-07T20-31-59.584821.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3657235d867fab9220935d198c49a535697fec7139ac8ff79550276c5e47dd04
+size 170367132
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-07T20-31-59.584821.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-07T20-31-59.584821.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..380d5a1554a601549312d53b9c3a25138fd1bde4
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_aime25_nofigures_agg64_2025-04-07T20-31-59.584821.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d36165b8e1e96475218799c9f3bd2d1b341caad58baf7c975b31d344c0297d0
+size 165580638
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-07T20-31-59.584821.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-07T20-31-59.584821.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..a9878bb659fcbc301d0670637359a7386c779195
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-100/samples_gpqa_diamond_openai_agg8_2025-04-07T20-31-59.584821.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90fd4c6a7cec360dd9a33d72471aea44c8752f6317c9d98c18d473cb611b70fb
+size 108890419
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/results_2025-04-07T20-20-02.007830.json b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/results_2025-04-07T20-20-02.007830.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0bf44d08269ba8169b93f21389878c3bd7fd012
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/results_2025-04-07T20-20-02.007830.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffb1552eafeb0e22109eb21177ffa900850d66c697edb9b68a7c65424c20f84a
+size 75045
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-07T20-20-02.007830.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-07T20-20-02.007830.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..b4320b50f69a9a9205e7d0a78bdd6fb0943cd30a
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_aime24_nofigures_agg64_2025-04-07T20-20-02.007830.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f521d9f2a0a62c4fee01ec2d8ffde42ad4083f6138afa8b9fa6b58d9eaae163
+size 162708959
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-07T20-20-02.007830.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-07T20-20-02.007830.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3a7d49c7a21d8ee0f064df589f743645021bc26e
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_aime25_nofigures_agg64_2025-04-07T20-20-02.007830.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc4977240b128d0201375b18271958e2a4664a5aa8c9b27ce9a62f41f8ad9a0b
+size 159084745
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-07T20-20-02.007830.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-07T20-20-02.007830.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..cd9880590bd61a1a638337a9f050c6376d837517
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-150/samples_gpqa_diamond_openai_agg8_2025-04-07T20-20-02.007830.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ff504512c3e1741a90ec09f16bc357f4cb7923342a983b96959d23b02f7c17c
+size 109308861
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/results_2025-04-08T18-37-15.377720.json b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/results_2025-04-08T18-37-15.377720.json
new file mode 100644
index 0000000000000000000000000000000000000000..2edb3395078930b05cb2d442255007701ef834da
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/results_2025-04-08T18-37-15.377720.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a2f96fe4dee42f37c706d797f475e8a9df95f13a0aac5f193402f5f6bb74799
+size 75044
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-08T18-37-15.377720.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-08T18-37-15.377720.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..0061ac34a9b3939a45044bcc7678578b4a2f656e
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_aime24_nofigures_agg64_2025-04-08T18-37-15.377720.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c5c68c3d7a45bb68665c55cedd69c068616cd055bb726b5003ee9b8d6dd06de
+size 155155651
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-08T18-37-15.377720.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-08T18-37-15.377720.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..a13c385dbe8a27c1c6a6156721a95dc4b1eedf0d
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_aime25_nofigures_agg64_2025-04-08T18-37-15.377720.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36aedc11c8a0b5e10980dc29edb7da7a4f61ff1a4e3f6a21a8b17c92ba94ed40
+size 163723434
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-08T18-37-15.377720.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-08T18-37-15.377720.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..ab10ff259cdb3aa9b494c6959253cdb26cb263ad
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-200/samples_gpqa_diamond_openai_agg8_2025-04-08T18-37-15.377720.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a81a47152ec91e18a61a825a7b70f0bfe383116410e42093098074a421472041
+size 106229889
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/results_2025-04-08T18-31-04.038583.json b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/results_2025-04-08T18-31-04.038583.json
new file mode 100644
index 0000000000000000000000000000000000000000..23a1e49c9fa21a37beb213d07e0c4ca7f16d754e
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/results_2025-04-08T18-31-04.038583.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2566a967ef4374a588395a134340ab3ca488981862fc75dc123fe0352c822625
+size 75320
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_aime24_nofigures_agg64_2025-04-08T18-31-04.038583.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_aime24_nofigures_agg64_2025-04-08T18-31-04.038583.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..cc0a9468a553c64bee6239512d984d2913128d40
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_aime24_nofigures_agg64_2025-04-08T18-31-04.038583.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1df280c7995191547f545e32c3a0d9f553d21c99b2a5774fde9f45347312b9f3
+size 147899851
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-08T18-31-04.038583.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-08T18-31-04.038583.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..f03ef92a13b2cd01ef44f641750cd4586e5a2e69
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_aime25_nofigures_agg64_2025-04-08T18-31-04.038583.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7572bc714808f8df4e5c178a03c40fc662bbae85226cae0b3e19c8cd9b4f50e4
+size 144228044
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-08T18-31-04.038583.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-08T18-31-04.038583.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..8f6ec5aa1d2ab91270a11564c048845abbade3a9
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-250/samples_gpqa_diamond_openai_agg8_2025-04-08T18-31-04.038583.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43b2d7dd00f11a4dbd149c200dccc8307cfb610b2779dc72d6ee0f6ff785c1a6
+size 116714048
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/results_2025-04-08T18-58-51.625344.json b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/results_2025-04-08T18-58-51.625344.json
new file mode 100644
index 0000000000000000000000000000000000000000..d58b33ce03ea2e02666f54d0c7e780d7f8b9ff34
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/results_2025-04-08T18-58-51.625344.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90a9bd84102fb87fb9f3019a56223c281bdb25b9f6348c70e10110ce4d2592f3
+size 74895
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-08T18-58-51.625344.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-08T18-58-51.625344.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..627bcd7f1971df8b766f9407279666f1ef82003c
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_aime24_nofigures_agg64_2025-04-08T18-58-51.625344.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4d53e66936d95f38390f1eaf3699b52521f783ea63ec84a30ffc2bd405dde0e
+size 149719433
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-08T18-58-51.625344.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-08T18-58-51.625344.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..3f1242cdcd2899613d68a17063df7d409a7224e2
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_aime25_nofigures_agg64_2025-04-08T18-58-51.625344.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5156c32ea4086759ce1ecda6f09bdc98bf353e3da51dc7e011fcebfa7723a350
+size 145418655
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-08T18-58-51.625344.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-08T18-58-51.625344.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..f9dc1d1e6d17c64654e7dba0cf6876f0a767eeb4
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-300/samples_gpqa_diamond_openai_agg8_2025-04-08T18-58-51.625344.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0d74674d73b470e93172f8bd7fd391965cd7225df5502ba404c0de777daa8cf
+size 123561024
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/results_2025-04-07T07-14-51.598130.json b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/results_2025-04-07T07-14-51.598130.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a6b690e75be76c0a9cd41746aaf817318b91642
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/results_2025-04-07T07-14-51.598130.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de80cd8ad6b83034a6f43c174ff12be5f77ec88af2b6d2cbe374a36af7e64c69
+size 75206
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-07T07-14-51.598130.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-07T07-14-51.598130.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..59eaf143e476b61fcbe1d75c2e7d1b45a4635447
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_aime24_nofigures_agg64_2025-04-07T07-14-51.598130.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:540ccafdad64363d46de1d6157d145d9404412db9c4142910014134824b80103
+size 177715260
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-07T07-14-51.598130.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-07T07-14-51.598130.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..03463258e2c3c2cfd4e6b6ca083960d6e6fac8c7
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_aime25_nofigures_agg64_2025-04-07T07-14-51.598130.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6887c04d0c9e4b1ecaebe41969bca28f6b07351b579d2a5fd62ca3b3cf9ac8b2
+size 179168299
diff --git a/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-07T07-14-51.598130.jsonl b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-07T07-14-51.598130.jsonl
new file mode 100644
index 0000000000000000000000000000000000000000..043b88f37a26d13c812f856e449448b317566d93
--- /dev/null
+++ b/s2-1.5B-v6/__data__niklas__s2__open-r1__data__s2-1.5B-v6__checkpoint-50/samples_gpqa_diamond_openai_agg8_2025-04-07T07-14-51.598130.jsonl
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4895343b0b696d76e6376ba677dcafc377010d7d59715c9f87b8ceb2856e742c
+size 120027058