Add files using upload-large-folder tool

.gitattributes

@@ -57,3 +57,5 @@ eval_results_2/vllm/qwen3-1.7B\#sfkl_nnm_lora/nnm0.1_K128_L4_epoch1_lr1e-4_kdr0.
 eval_results_2/vllm/qwen3-1.7B\#sfkl_nnm_lora/nnm_new0.1_K128_L4_epoch2_lr1e-4_kdr1.0/results__qwen3-1.7B\#sfkl_nnm_lora__nnm_new0.1_K128_L4_epoch2_lr1e-4_kdr1.0/samples_gsm_plus_2026-05-21T22-58-09.777308.jsonl filter=lfs diff=lfs merge=lfs -text
 eval_results/vllm/gemma-2-2b-it/nnm0.5_K128_L4_epoch2_lr1e-4_kdr1.0/google__gemma-2-2b-it/samples_gsm8k_2026-05-20T10-45-09.944715.jsonl filter=lfs diff=lfs merge=lfs -text
 eval_results_2/vllm/qwen3-1.7B\#sfkl_nnm_lora/nnm1.0_K128_L4_epoch1_lr1e-4_kdr1.0/results__qwen3-1.7B\#sfkl_nnm_lora__nnm1.0_K128_L4_epoch1_lr1e-4_kdr1.0/samples_gsm_plus_2026-05-21T22-07-38.678553.jsonl filter=lfs diff=lfs merge=lfs -text
+eval_results/vllm/qwen3-1.7B\#sfkl_nnm_lora/nnm0.1_K128_L4_epoch2_lr1e-4_kdr1.0/Qwen__Qwen3-1.7B/samples_gsm_plus_2026-05-20T14-33-31.362188.jsonl filter=lfs diff=lfs merge=lfs -text
+eval_results/vllm/qwen2.5-1.5B-it-amid/results__qwen2.5-1.5B-Instruct\#amid__ab_pr_0.5_0.5_4_1e-4/samples_gsm_plus_2026-05-17T08-22-15.616148.jsonl filter=lfs diff=lfs merge=lfs -text

eval_results/vllm/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results_2026-05-20T04-40-01.886597.json

@@ -0,0 +1,140 @@
+{
+  "results": {
+    "sciq": {
+      "name": "sciq",
+      "alias": "sciq",
+      "sample_len": 1000,
+      "acc,none": 0.863,
+      "acc_stderr,none": 0.010878848714333304,
+      "acc_norm,none": 0.646,
+      "acc_norm_stderr,none": 0.015129868238451773
+    }
+  },
+  "group_subtasks": {},
+  "configs": {
+    "sciq": {
+      "task": "sciq",
+      "dataset_path": "allenai/sciq",
+      "training_split": "train",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "{{support.lstrip()}}\nQuestion: {{question}}\nAnswer:",
+      "doc_to_target": 3,
+      "unsafe_code": false,
+      "doc_to_choice": "{{[distractor1, distractor2, distractor3, correct_answer]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "default",
+        "split": null,
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{support.lstrip()}}\nQuestion: {{question}}\nAnswer:",
+        "doc_to_choice": "{{[distractor1, distractor2, distractor3, correct_answer]}}",
+        "doc_to_target": 3,
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "acc_norm",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{support}} {{question}}",
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/sciq/sciq.yaml"
+      }
+    }
+  },
+  "versions": {
+    "sciq": 1.0
+  },
+  "n-shot": {
+    "sciq": 0
+  },
+  "higher_is_better": {
+    "sciq": {
+      "acc": true,
+      "acc_norm": true
+    }
+  },
+  "n-samples": {
+    "sciq": {
+      "original": 1000,
+      "effective": 1000
+    }
+  },
+  "config": {
+    "model": "vllm",
+    "model_args": {
+      "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+      "tensor_parallel_size": 2,
+      "dtype": "float16",
+      "gpu_memory_utilization": 0.8,
+      "trust_remote_code": true
+    },
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {},
+    "random_seed": 0,
+    "numpy_seed": 1234,
+    "torch_seed": 1234,
+    "fewshot_seed": 1234
+  },
+  "git_hash": "04b98db",
+  "date": 1779251948.917537,
+  "pretty_env_info": "PyTorch version: 2.10.0+cu128\nIs debug build: False\nCUDA used to build PyTorch: 12.8\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 24.04.4 LTS (x86_64)\nGCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0\nClang version: Could not collect\nCMake version: version 3.28.3\nLibc version: glibc-2.39\n\nPython version: 3.10.20 (main, May 10 2026, 19:27:29) [Clang 22.1.3 ] (64-bit runtime)\nPython platform: Linux-6.17.0-1013-aws-x86_64-with-glibc2.39\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: \nGPU models and configuration: \nGPU 0: NVIDIA H200\nGPU 1: NVIDIA H200\nGPU 2: NVIDIA H200\nGPU 3: NVIDIA H200\nGPU 4: NVIDIA H200\nGPU 5: NVIDIA H200\nGPU 6: NVIDIA H200\nGPU 7: NVIDIA H200\n\nNvidia driver version: 580.126.16\ncuDNN version: Probably one of the following:\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_adv.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_cnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_precompiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_runtime_compiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_graph.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_heuristic.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_ops.so.9\nIs XPU available: False\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\nCaching allocator config: N/A\n\nCPU:\nArchitecture:                            x86_64\nCPU op-mode(s):                          32-bit, 64-bit\nAddress sizes:                           48 bits physical, 48 bits virtual\nByte Order:                              Little Endian\nCPU(s):                                  192\nOn-line CPU(s) list:                     0-191\nVendor ID:                               AuthenticAMD\nModel name:                              AMD EPYC 7R13 Processor\nCPU family:                              25\nModel:                                   1\nThread(s) per core:                      2\nCore(s) per socket:                      48\nSocket(s):                               2\nStepping:                                1\nBogoMIPS:                                5299.99\nFlags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf tsc_known_freq pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch topoext perfctr_core ssbd ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 clzero xsaveerptr rdpru wbnoinvd arat npt nrip_save vaes vpclmulqdq rdpid\nHypervisor vendor:                       KVM\nVirtualization type:                     full\nL1d cache:                               3 MiB (96 instances)\nL1i cache:                               3 MiB (96 instances)\nL2 cache:                                48 MiB (96 instances)\nL3 cache:                                384 MiB (12 instances)\nNUMA node(s):                            4\nNUMA node0 CPU(s):                       0-23,96-119\nNUMA node1 CPU(s):                       24-47,120-143\nNUMA node2 CPU(s):                       48-71,144-167\nNUMA node3 CPU(s):                       72-95,168-191\nVulnerability Gather data sampling:      Not affected\nVulnerability Ghostwrite:                Not affected\nVulnerability Indirect target selection: 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 Old microcode:             Not affected\nVulnerability Reg file data sampling:    Not affected\nVulnerability Retbleed:                  Not affected\nVulnerability Spec rstack overflow:      Mitigation; Safe RET\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; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                     Not affected\nVulnerability Tsa:                       Mitigation; Clear CPU buffers\nVulnerability Tsx async abort:           Not affected\nVulnerability Vmscape:                   Not affected\n\nVersions of relevant libraries:\n[pip3] Could not collect\n[conda] numpy                       2.4.5            pypi_0              pypi\n[conda] nvidia-cublas-cu12          12.4.5.8         pypi_0              pypi\n[conda] nvidia-cuda-cupti-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-nvrtc-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-runtime-cu12    12.4.127         pypi_0              pypi\n[conda] nvidia-cudnn-cu12           9.1.0.70         pypi_0              pypi\n[conda] nvidia-cufft-cu12           11.2.1.3         pypi_0              pypi\n[conda] nvidia-curand-cu12          10.3.5.147       pypi_0              pypi\n[conda] nvidia-cusolver-cu12        11.6.1.9         pypi_0              pypi\n[conda] nvidia-cusparse-cu12        12.3.1.170       pypi_0              pypi\n[conda] nvidia-nccl-cu12            2.21.5           pypi_0              pypi\n[conda] nvidia-nvjitlink-cu12       12.4.127         pypi_0              pypi\n[conda] nvidia-nvtx-cu12            12.4.127         pypi_0              pypi\n[conda] torch                       2.5.1+cu124      pypi_0              pypi",
+  "transformers_version": "4.57.3",
+  "lm_eval_version": "0.4.12",
+  "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": {
+    "sciq": "94d7c305efca6804164ebf166179faaee89d73aba3a1539fb62b358258a4b95f"
+  },
+  "model_source": "vllm",
+  "model_name": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "model_name_sanitized": "results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "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",
+  "total_evaluation_time_seconds": "60.88098437001463"
+}

eval_results/vllm/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results_2026-05-20T04-52-34.266695.json

@@ -0,0 +1,1541 @@
+{
+  "results": {
+    "mmlu_abstract_algebra": {
+      "name": "mmlu_abstract_algebra",
+      "alias": "abstract_algebra",
+      "sample_len": 100,
+      "acc,none": 0.23,
+      "acc_stderr,none": 0.04229525846816505
+    },
+    "mmlu_anatomy": {
+      "name": "mmlu_anatomy",
+      "alias": "anatomy",
+      "sample_len": 135,
+      "acc,none": 0.5111111111111111,
+      "acc_stderr,none": 0.04318275491977976
+    },
+    "mmlu_astronomy": {
+      "name": "mmlu_astronomy",
+      "alias": "astronomy",
+      "sample_len": 152,
+      "acc,none": 0.6052631578947368,
+      "acc_stderr,none": 0.039777499346220734
+    },
+    "mmlu_college_biology": {
+      "name": "mmlu_college_biology",
+      "alias": "college_biology",
+      "sample_len": 144,
+      "acc,none": 0.5972222222222222,
+      "acc_stderr,none": 0.04101405519842426
+    },
+    "mmlu_college_chemistry": {
+      "name": "mmlu_college_chemistry",
+      "alias": "college_chemistry",
+      "sample_len": 100,
+      "acc,none": 0.34,
+      "acc_stderr,none": 0.04760952285695235
+    },
+    "mmlu_college_computer_science": {
+      "name": "mmlu_college_computer_science",
+      "alias": "college_computer_science",
+      "sample_len": 100,
+      "acc,none": 0.32,
+      "acc_stderr,none": 0.046882617226215034
+    },
+    "mmlu_college_mathematics": {
+      "name": "mmlu_college_mathematics",
+      "alias": "college_mathematics",
+      "sample_len": 100,
+      "acc,none": 0.3,
+      "acc_stderr,none": 0.046056618647183814
+    },
+    "mmlu_college_physics": {
+      "name": "mmlu_college_physics",
+      "alias": "college_physics",
+      "sample_len": 102,
+      "acc,none": 0.37254901960784315,
+      "acc_stderr,none": 0.04810840148082633
+    },
+    "mmlu_computer_security": {
+      "name": "mmlu_computer_security",
+      "alias": "computer_security",
+      "sample_len": 100,
+      "acc,none": 0.67,
+      "acc_stderr,none": 0.04725815626252607
+    },
+    "mmlu_conceptual_physics": {
+      "name": "mmlu_conceptual_physics",
+      "alias": "conceptual_physics",
+      "sample_len": 235,
+      "acc,none": 0.5702127659574469,
+      "acc_stderr,none": 0.03236214467715564
+    },
+    "mmlu_electrical_engineering": {
+      "name": "mmlu_electrical_engineering",
+      "alias": "electrical_engineering",
+      "sample_len": 145,
+      "acc,none": 0.4689655172413793,
+      "acc_stderr,none": 0.04158632762097828
+    },
+    "mmlu_elementary_mathematics": {
+      "name": "mmlu_elementary_mathematics",
+      "alias": "elementary_mathematics",
+      "sample_len": 378,
+      "acc,none": 0.3888888888888889,
+      "acc_stderr,none": 0.025107425481137282
+    },
+    "mmlu_high_school_biology": {
+      "name": "mmlu_high_school_biology",
+      "alias": "high_school_biology",
+      "sample_len": 310,
+      "acc,none": 0.6516129032258065,
+      "acc_stderr,none": 0.027104826328100944
+    },
+    "mmlu_high_school_chemistry": {
+      "name": "mmlu_high_school_chemistry",
+      "alias": "high_school_chemistry",
+      "sample_len": 203,
+      "acc,none": 0.458128078817734,
+      "acc_stderr,none": 0.03505630140785741
+    },
+    "mmlu_high_school_computer_science": {
+      "name": "mmlu_high_school_computer_science",
+      "alias": "high_school_computer_science",
+      "sample_len": 100,
+      "acc,none": 0.54,
+      "acc_stderr,none": 0.05009082659620331
+    },
+    "mmlu_high_school_mathematics": {
+      "name": "mmlu_high_school_mathematics",
+      "alias": "high_school_mathematics",
+      "sample_len": 270,
+      "acc,none": 0.26666666666666666,
+      "acc_stderr,none": 0.026962424325073828
+    },
+    "mmlu_high_school_physics": {
+      "name": "mmlu_high_school_physics",
+      "alias": "high_school_physics",
+      "sample_len": 151,
+      "acc,none": 0.304635761589404,
+      "acc_stderr,none": 0.03757949922943343
+    },
+    "mmlu_high_school_statistics": {
+      "name": "mmlu_high_school_statistics",
+      "alias": "high_school_statistics",
+      "sample_len": 216,
+      "acc,none": 0.36574074074074076,
+      "acc_stderr,none": 0.03284738857647207
+    },
+    "mmlu_machine_learning": {
+      "name": "mmlu_machine_learning",
+      "alias": "machine_learning",
+      "sample_len": 112,
+      "acc,none": 0.44642857142857145,
+      "acc_stderr,none": 0.047184714852195886
+    },
+    "mmlu_stem": {
+      "alias": "stem",
+      "name": "mmlu_stem",
+      "sample_len": 3153,
+      "acc,none": 0.4490960989533777,
+      "acc_stderr,none": 0.0085671683185063,
+      "sample_count": {
+        "acc,none": 3153
+      }
+    }
+  },
+  "groups": {
+    "mmlu_stem": {
+      "alias": "stem",
+      "name": "mmlu_stem",
+      "sample_len": 3153,
+      "acc,none": 0.4490960989533777,
+      "acc_stderr,none": 0.0085671683185063,
+      "sample_count": {
+        "acc,none": 3153
+      }
+    }
+  },
+  "group_subtasks": {
+    "mmlu_stem": [
+      "mmlu_abstract_algebra",
+      "mmlu_anatomy",
+      "mmlu_astronomy",
+      "mmlu_college_biology",
+      "mmlu_college_chemistry",
+      "mmlu_college_computer_science",
+      "mmlu_college_mathematics",
+      "mmlu_college_physics",
+      "mmlu_computer_security",
+      "mmlu_conceptual_physics",
+      "mmlu_electrical_engineering",
+      "mmlu_elementary_mathematics",
+      "mmlu_high_school_biology",
+      "mmlu_high_school_chemistry",
+      "mmlu_high_school_computer_science",
+      "mmlu_high_school_mathematics",
+      "mmlu_high_school_physics",
+      "mmlu_high_school_statistics",
+      "mmlu_machine_learning"
+    ]
+  },
+  "configs": {
+    "mmlu_abstract_algebra": {
+      "task": "mmlu_abstract_algebra",
+      "task_alias": "abstract_algebra",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "abstract_algebra",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about abstract algebra.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_abstract_algebra.yaml"
+      }
+    },
+    "mmlu_anatomy": {
+      "task": "mmlu_anatomy",
+      "task_alias": "anatomy",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "anatomy",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about anatomy.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_anatomy.yaml"
+      }
+    },
+    "mmlu_astronomy": {
+      "task": "mmlu_astronomy",
+      "task_alias": "astronomy",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "astronomy",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about astronomy.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_astronomy.yaml"
+      }
+    },
+    "mmlu_college_biology": {
+      "task": "mmlu_college_biology",
+      "task_alias": "college_biology",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_biology",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college biology.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_college_biology.yaml"
+      }
+    },
+    "mmlu_college_chemistry": {
+      "task": "mmlu_college_chemistry",
+      "task_alias": "college_chemistry",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_chemistry",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college chemistry.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_college_chemistry.yaml"
+      }
+    },
+    "mmlu_college_computer_science": {
+      "task": "mmlu_college_computer_science",
+      "task_alias": "college_computer_science",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_computer_science",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college computer science.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_college_computer_science.yaml"
+      }
+    },
+    "mmlu_college_mathematics": {
+      "task": "mmlu_college_mathematics",
+      "task_alias": "college_mathematics",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_mathematics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college mathematics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_college_mathematics.yaml"
+      }
+    },
+    "mmlu_college_physics": {
+      "task": "mmlu_college_physics",
+      "task_alias": "college_physics",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_physics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college physics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_college_physics.yaml"
+      }
+    },
+    "mmlu_computer_security": {
+      "task": "mmlu_computer_security",
+      "task_alias": "computer_security",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "computer_security",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about computer security.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_computer_security.yaml"
+      }
+    },
+    "mmlu_conceptual_physics": {
+      "task": "mmlu_conceptual_physics",
+      "task_alias": "conceptual_physics",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "conceptual_physics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about conceptual physics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_conceptual_physics.yaml"
+      }
+    },
+    "mmlu_electrical_engineering": {
+      "task": "mmlu_electrical_engineering",
+      "task_alias": "electrical_engineering",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "electrical_engineering",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about electrical engineering.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_electrical_engineering.yaml"
+      }
+    },
+    "mmlu_elementary_mathematics": {
+      "task": "mmlu_elementary_mathematics",
+      "task_alias": "elementary_mathematics",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "elementary_mathematics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about elementary mathematics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_elementary_mathematics.yaml"
+      }
+    },
+    "mmlu_high_school_biology": {
+      "task": "mmlu_high_school_biology",
+      "task_alias": "high_school_biology",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_biology",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school biology.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_high_school_biology.yaml"
+      }
+    },
+    "mmlu_high_school_chemistry": {
+      "task": "mmlu_high_school_chemistry",
+      "task_alias": "high_school_chemistry",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_chemistry",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school chemistry.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_high_school_chemistry.yaml"
+      }
+    },
+    "mmlu_high_school_computer_science": {
+      "task": "mmlu_high_school_computer_science",
+      "task_alias": "high_school_computer_science",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_computer_science",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school computer science.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_high_school_computer_science.yaml"
+      }
+    },
+    "mmlu_high_school_mathematics": {
+      "task": "mmlu_high_school_mathematics",
+      "task_alias": "high_school_mathematics",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_mathematics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school mathematics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_high_school_mathematics.yaml"
+      }
+    },
+    "mmlu_high_school_physics": {
+      "task": "mmlu_high_school_physics",
+      "task_alias": "high_school_physics",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_physics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school physics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_high_school_physics.yaml"
+      }
+    },
+    "mmlu_high_school_statistics": {
+      "task": "mmlu_high_school_statistics",
+      "task_alias": "high_school_statistics",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_statistics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school statistics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_high_school_statistics.yaml"
+      }
+    },
+    "mmlu_machine_learning": {
+      "task": "mmlu_machine_learning",
+      "task_alias": "machine_learning",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "machine_learning",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about machine learning.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu/default/mmlu_machine_learning.yaml"
+      }
+    }
+  },
+  "versions": {
+    "mmlu_abstract_algebra": 1.0,
+    "mmlu_anatomy": 1.0,
+    "mmlu_astronomy": 1.0,
+    "mmlu_college_biology": 1.0,
+    "mmlu_college_chemistry": 1.0,
+    "mmlu_college_computer_science": 1.0,
+    "mmlu_college_mathematics": 1.0,
+    "mmlu_college_physics": 1.0,
+    "mmlu_computer_security": 1.0,
+    "mmlu_conceptual_physics": 1.0,
+    "mmlu_electrical_engineering": 1.0,
+    "mmlu_elementary_mathematics": 1.0,
+    "mmlu_high_school_biology": 1.0,
+    "mmlu_high_school_chemistry": 1.0,
+    "mmlu_high_school_computer_science": 1.0,
+    "mmlu_high_school_mathematics": 1.0,
+    "mmlu_high_school_physics": 1.0,
+    "mmlu_high_school_statistics": 1.0,
+    "mmlu_machine_learning": 1.0,
+    "mmlu_stem": "2"
+  },
+  "n-shot": {
+    "mmlu_abstract_algebra": 0,
+    "mmlu_anatomy": 0,
+    "mmlu_astronomy": 0,
+    "mmlu_college_biology": 0,
+    "mmlu_college_chemistry": 0,
+    "mmlu_college_computer_science": 0,
+    "mmlu_college_mathematics": 0,
+    "mmlu_college_physics": 0,
+    "mmlu_computer_security": 0,
+    "mmlu_conceptual_physics": 0,
+    "mmlu_electrical_engineering": 0,
+    "mmlu_elementary_mathematics": 0,
+    "mmlu_high_school_biology": 0,
+    "mmlu_high_school_chemistry": 0,
+    "mmlu_high_school_computer_science": 0,
+    "mmlu_high_school_mathematics": 0,
+    "mmlu_high_school_physics": 0,
+    "mmlu_high_school_statistics": 0,
+    "mmlu_machine_learning": 0,
+    "mmlu_stem": 0
+  },
+  "higher_is_better": {
+    "mmlu_abstract_algebra": {
+      "acc": true
+    },
+    "mmlu_anatomy": {
+      "acc": true
+    },
+    "mmlu_astronomy": {
+      "acc": true
+    },
+    "mmlu_college_biology": {
+      "acc": true
+    },
+    "mmlu_college_chemistry": {
+      "acc": true
+    },
+    "mmlu_college_computer_science": {
+      "acc": true
+    },
+    "mmlu_college_mathematics": {
+      "acc": true
+    },
+    "mmlu_college_physics": {
+      "acc": true
+    },
+    "mmlu_computer_security": {
+      "acc": true
+    },
+    "mmlu_conceptual_physics": {
+      "acc": true
+    },
+    "mmlu_electrical_engineering": {
+      "acc": true
+    },
+    "mmlu_elementary_mathematics": {
+      "acc": true
+    },
+    "mmlu_high_school_biology": {
+      "acc": true
+    },
+    "mmlu_high_school_chemistry": {
+      "acc": true
+    },
+    "mmlu_high_school_computer_science": {
+      "acc": true
+    },
+    "mmlu_high_school_mathematics": {
+      "acc": true
+    },
+    "mmlu_high_school_physics": {
+      "acc": true
+    },
+    "mmlu_high_school_statistics": {
+      "acc": true
+    },
+    "mmlu_machine_learning": {
+      "acc": true
+    },
+    "mmlu_stem": {
+      "acc": true
+    }
+  },
+  "n-samples": {
+    "mmlu_abstract_algebra": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_anatomy": {
+      "original": 135,
+      "effective": 135
+    },
+    "mmlu_astronomy": {
+      "original": 152,
+      "effective": 152
+    },
+    "mmlu_college_biology": {
+      "original": 144,
+      "effective": 144
+    },
+    "mmlu_college_chemistry": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_college_computer_science": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_college_mathematics": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_college_physics": {
+      "original": 102,
+      "effective": 102
+    },
+    "mmlu_computer_security": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_conceptual_physics": {
+      "original": 235,
+      "effective": 235
+    },
+    "mmlu_electrical_engineering": {
+      "original": 145,
+      "effective": 145
+    },
+    "mmlu_elementary_mathematics": {
+      "original": 378,
+      "effective": 378
+    },
+    "mmlu_high_school_biology": {
+      "original": 310,
+      "effective": 310
+    },
+    "mmlu_high_school_chemistry": {
+      "original": 203,
+      "effective": 203
+    },
+    "mmlu_high_school_computer_science": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_high_school_mathematics": {
+      "original": 270,
+      "effective": 270
+    },
+    "mmlu_high_school_physics": {
+      "original": 151,
+      "effective": 151
+    },
+    "mmlu_high_school_statistics": {
+      "original": 216,
+      "effective": 216
+    },
+    "mmlu_machine_learning": {
+      "original": 112,
+      "effective": 112
+    }
+  },
+  "config": {
+    "model": "vllm",
+    "model_args": {
+      "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+      "tensor_parallel_size": 2,
+      "dtype": "float16",
+      "gpu_memory_utilization": 0.8,
+      "trust_remote_code": true
+    },
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {},
+    "random_seed": 0,
+    "numpy_seed": 1234,
+    "torch_seed": 1234,
+    "fewshot_seed": 1234
+  },
+  "git_hash": "04b98db",
+  "date": 1779252665.7140193,
+  "pretty_env_info": "PyTorch version: 2.10.0+cu128\nIs debug build: False\nCUDA used to build PyTorch: 12.8\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 24.04.4 LTS (x86_64)\nGCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0\nClang version: Could not collect\nCMake version: version 3.28.3\nLibc version: glibc-2.39\n\nPython version: 3.10.20 (main, May 10 2026, 19:27:29) [Clang 22.1.3 ] (64-bit runtime)\nPython platform: Linux-6.17.0-1013-aws-x86_64-with-glibc2.39\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: \nGPU models and configuration: \nGPU 0: NVIDIA H200\nGPU 1: NVIDIA H200\nGPU 2: NVIDIA H200\nGPU 3: NVIDIA H200\nGPU 4: NVIDIA H200\nGPU 5: NVIDIA H200\nGPU 6: NVIDIA H200\nGPU 7: NVIDIA H200\n\nNvidia driver version: 580.126.16\ncuDNN version: Probably one of the following:\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_adv.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_cnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_precompiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_runtime_compiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_graph.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_heuristic.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_ops.so.9\nIs XPU available: False\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\nCaching allocator config: N/A\n\nCPU:\nArchitecture:                            x86_64\nCPU op-mode(s):                          32-bit, 64-bit\nAddress sizes:                           48 bits physical, 48 bits virtual\nByte Order:                              Little Endian\nCPU(s):                                  192\nOn-line CPU(s) list:                     0-191\nVendor ID:                               AuthenticAMD\nModel name:                              AMD EPYC 7R13 Processor\nCPU family:                              25\nModel:                                   1\nThread(s) per core:                      2\nCore(s) per socket:                      48\nSocket(s):                               2\nStepping:                                1\nBogoMIPS:                                5299.99\nFlags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf tsc_known_freq pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch topoext perfctr_core ssbd ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 clzero xsaveerptr rdpru wbnoinvd arat npt nrip_save vaes vpclmulqdq rdpid\nHypervisor vendor:                       KVM\nVirtualization type:                     full\nL1d cache:                               3 MiB (96 instances)\nL1i cache:                               3 MiB (96 instances)\nL2 cache:                                48 MiB (96 instances)\nL3 cache:                                384 MiB (12 instances)\nNUMA node(s):                            4\nNUMA node0 CPU(s):                       0-23,96-119\nNUMA node1 CPU(s):                       24-47,120-143\nNUMA node2 CPU(s):                       48-71,144-167\nNUMA node3 CPU(s):                       72-95,168-191\nVulnerability Gather data sampling:      Not affected\nVulnerability Ghostwrite:                Not affected\nVulnerability Indirect target selection: 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 Old microcode:             Not affected\nVulnerability Reg file data sampling:    Not affected\nVulnerability Retbleed:                  Not affected\nVulnerability Spec rstack overflow:      Mitigation; Safe RET\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; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                     Not affected\nVulnerability Tsa:                       Mitigation; Clear CPU buffers\nVulnerability Tsx async abort:           Not affected\nVulnerability Vmscape:                   Not affected\n\nVersions of relevant libraries:\n[pip3] Could not collect\n[conda] numpy                       2.4.5            pypi_0              pypi\n[conda] nvidia-cublas-cu12          12.4.5.8         pypi_0              pypi\n[conda] nvidia-cuda-cupti-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-nvrtc-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-runtime-cu12    12.4.127         pypi_0              pypi\n[conda] nvidia-cudnn-cu12           9.1.0.70         pypi_0              pypi\n[conda] nvidia-cufft-cu12           11.2.1.3         pypi_0              pypi\n[conda] nvidia-curand-cu12          10.3.5.147       pypi_0              pypi\n[conda] nvidia-cusolver-cu12        11.6.1.9         pypi_0              pypi\n[conda] nvidia-cusparse-cu12        12.3.1.170       pypi_0              pypi\n[conda] nvidia-nccl-cu12            2.21.5           pypi_0              pypi\n[conda] nvidia-nvjitlink-cu12       12.4.127         pypi_0              pypi\n[conda] nvidia-nvtx-cu12            12.4.127         pypi_0              pypi\n[conda] torch                       2.5.1+cu124      pypi_0              pypi",
+  "transformers_version": "4.57.3",
+  "lm_eval_version": "0.4.12",
+  "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": {
+    "mmlu_abstract_algebra": "1ecc29cafeff483bf80347adc2878dae00d01596988032f438afce6291c8cfe7",
+    "mmlu_anatomy": "51232e78f32967e0dd431bab9309f69be490f9998b4ddb93b641b09ba1f613cf",
+    "mmlu_astronomy": "3e8d7c33bd0431518a3d5d4b1666bdabccf46c4cc4cfe95a3886cb2266743f86",
+    "mmlu_college_biology": "5378a297efa1cd71344228c95dcb29ba54757c7e527261400a4d65fd8ceed92b",
+    "mmlu_college_chemistry": "714e60ea7e340f0344ffde41b247a19b68e6384152a1bdc9acc0ecf24cd0e7cd",
+    "mmlu_college_computer_science": "2b66d244f98e61e6a78e5ed48b728e0ca2ca81c7757ccdd086fe6b0c1cf02da1",
+    "mmlu_college_mathematics": "90b1ce25d7d38abd925181ae3aad095cfd783375fe8de0607a57420f47454f0f",
+    "mmlu_college_physics": "2abddb12a247904f9ba60b39766b19a8967c24fe8eee021022a353b93d73742c",
+    "mmlu_computer_security": "8ffba21cebb9e39d5794c848a22e7e37cddff5c59f7b47a1d89872d9e8704b0c",
+    "mmlu_conceptual_physics": "c0da3219fe7e5292dad2aa7e72790e3eba951711007cc57618b0850d941b85ec",
+    "mmlu_electrical_engineering": "0a02cb16a31ec7debba4190ab1cc31afadb9416764e0b074b9b96553f346b892",
+    "mmlu_elementary_mathematics": "6048f1506468e757010d537a0ceedaadc587c18159a03e7664b5994843afedc2",
+    "mmlu_high_school_biology": "2aabe83bb100a352d24a0136c24e612b68490bbe818c33f06cc6cee5379ecad8",
+    "mmlu_high_school_chemistry": "dd3321fd42c8717094f57dd0eaeab0002d9ed192518a7de7aaef4913796c30f0",
+    "mmlu_high_school_computer_science": "888aabb9a7ac77299e8b2b8429fa44c1fbbb5ac8165f556a0b00b347e51de3b5",
+    "mmlu_high_school_mathematics": "9b9567fc5c7875065ac91eac69621904f201cab822621440a2442a390f62b497",
+    "mmlu_high_school_physics": "53edb93aefb6fc54fba283278bae10a58a7902902ff55edc30cb9c81745dbaa8",
+    "mmlu_high_school_statistics": "e0d652bf0cb6f927b0fc738e58a141f0ba8f9073615f4ddd70bf7a3b97998f21",
+    "mmlu_machine_learning": "13ad1d3650a26d655a2a2eaa3be1f106d990e5b693b0e9006f8ebef221d2ed82"
+  },
+  "model_source": "vllm",
+  "model_name": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "model_name_sanitized": "results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "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",
+  "total_evaluation_time_seconds": "96.21692177001387"
+}

eval_results/vllm/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results_2026-05-20T04-53-36.641824.json

@@ -0,0 +1,140 @@
+{
+  "results": {
+    "sciq": {
+      "name": "sciq",
+      "alias": "sciq",
+      "sample_len": 1000,
+      "acc,none": 0.863,
+      "acc_stderr,none": 0.010878848714333304,
+      "acc_norm,none": 0.646,
+      "acc_norm_stderr,none": 0.015129868238451773
+    }
+  },
+  "group_subtasks": {},
+  "configs": {
+    "sciq": {
+      "task": "sciq",
+      "dataset_path": "allenai/sciq",
+      "training_split": "train",
+      "validation_split": "validation",
+      "test_split": "test",
+      "doc_to_text": "{{support.lstrip()}}\nQuestion: {{question}}\nAnswer:",
+      "doc_to_target": 3,
+      "unsafe_code": false,
+      "doc_to_choice": "{{[distractor1, distractor2, distractor3, correct_answer]}}",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "default",
+        "split": null,
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{support.lstrip()}}\nQuestion: {{question}}\nAnswer:",
+        "doc_to_choice": "{{[distractor1, distractor2, distractor3, correct_answer]}}",
+        "doc_to_target": 3,
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "acc_norm",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": true,
+      "doc_to_decontamination_query": "{{support}} {{question}}",
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/sciq/sciq.yaml"
+      }
+    }
+  },
+  "versions": {
+    "sciq": 1.0
+  },
+  "n-shot": {
+    "sciq": 0
+  },
+  "higher_is_better": {
+    "sciq": {
+      "acc": true,
+      "acc_norm": true
+    }
+  },
+  "n-samples": {
+    "sciq": {
+      "original": 1000,
+      "effective": 1000
+    }
+  },
+  "config": {
+    "model": "vllm",
+    "model_args": {
+      "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+      "tensor_parallel_size": 2,
+      "dtype": "float16",
+      "gpu_memory_utilization": 0.8,
+      "trust_remote_code": true
+    },
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {},
+    "random_seed": 0,
+    "numpy_seed": 1234,
+    "torch_seed": 1234,
+    "fewshot_seed": 1234
+  },
+  "git_hash": "04b98db",
+  "date": 1779252765.1146324,
+  "pretty_env_info": "PyTorch version: 2.10.0+cu128\nIs debug build: False\nCUDA used to build PyTorch: 12.8\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 24.04.4 LTS (x86_64)\nGCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0\nClang version: Could not collect\nCMake version: version 3.28.3\nLibc version: glibc-2.39\n\nPython version: 3.10.20 (main, May 10 2026, 19:27:29) [Clang 22.1.3 ] (64-bit runtime)\nPython platform: Linux-6.17.0-1013-aws-x86_64-with-glibc2.39\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: \nGPU models and configuration: \nGPU 0: NVIDIA H200\nGPU 1: NVIDIA H200\nGPU 2: NVIDIA H200\nGPU 3: NVIDIA H200\nGPU 4: NVIDIA H200\nGPU 5: NVIDIA H200\nGPU 6: NVIDIA H200\nGPU 7: NVIDIA H200\n\nNvidia driver version: 580.126.16\ncuDNN version: Probably one of the following:\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_adv.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_cnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_precompiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_runtime_compiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_graph.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_heuristic.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_ops.so.9\nIs XPU available: False\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\nCaching allocator config: N/A\n\nCPU:\nArchitecture:                            x86_64\nCPU op-mode(s):                          32-bit, 64-bit\nAddress sizes:                           48 bits physical, 48 bits virtual\nByte Order:                              Little Endian\nCPU(s):                                  192\nOn-line CPU(s) list:                     0-191\nVendor ID:                               AuthenticAMD\nModel name:                              AMD EPYC 7R13 Processor\nCPU family:                              25\nModel:                                   1\nThread(s) per core:                      2\nCore(s) per socket:                      48\nSocket(s):                               2\nStepping:                                1\nBogoMIPS:                                5299.99\nFlags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf tsc_known_freq pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch topoext perfctr_core ssbd ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 clzero xsaveerptr rdpru wbnoinvd arat npt nrip_save vaes vpclmulqdq rdpid\nHypervisor vendor:                       KVM\nVirtualization type:                     full\nL1d cache:                               3 MiB (96 instances)\nL1i cache:                               3 MiB (96 instances)\nL2 cache:                                48 MiB (96 instances)\nL3 cache:                                384 MiB (12 instances)\nNUMA node(s):                            4\nNUMA node0 CPU(s):                       0-23,96-119\nNUMA node1 CPU(s):                       24-47,120-143\nNUMA node2 CPU(s):                       48-71,144-167\nNUMA node3 CPU(s):                       72-95,168-191\nVulnerability Gather data sampling:      Not affected\nVulnerability Ghostwrite:                Not affected\nVulnerability Indirect target selection: 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 Old microcode:             Not affected\nVulnerability Reg file data sampling:    Not affected\nVulnerability Retbleed:                  Not affected\nVulnerability Spec rstack overflow:      Mitigation; Safe RET\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; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                     Not affected\nVulnerability Tsa:                       Mitigation; Clear CPU buffers\nVulnerability Tsx async abort:           Not affected\nVulnerability Vmscape:                   Not affected\n\nVersions of relevant libraries:\n[pip3] Could not collect\n[conda] numpy                       2.4.5            pypi_0              pypi\n[conda] nvidia-cublas-cu12          12.4.5.8         pypi_0              pypi\n[conda] nvidia-cuda-cupti-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-nvrtc-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-runtime-cu12    12.4.127         pypi_0              pypi\n[conda] nvidia-cudnn-cu12           9.1.0.70         pypi_0              pypi\n[conda] nvidia-cufft-cu12           11.2.1.3         pypi_0              pypi\n[conda] nvidia-curand-cu12          10.3.5.147       pypi_0              pypi\n[conda] nvidia-cusolver-cu12        11.6.1.9         pypi_0              pypi\n[conda] nvidia-cusparse-cu12        12.3.1.170       pypi_0              pypi\n[conda] nvidia-nccl-cu12            2.21.5           pypi_0              pypi\n[conda] nvidia-nvjitlink-cu12       12.4.127         pypi_0              pypi\n[conda] nvidia-nvtx-cu12            12.4.127         pypi_0              pypi\n[conda] torch                       2.5.1+cu124      pypi_0              pypi",
+  "transformers_version": "4.57.3",
+  "lm_eval_version": "0.4.12",
+  "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": {
+    "sciq": "94d7c305efca6804164ebf166179faaee89d73aba3a1539fb62b358258a4b95f"
+  },
+  "model_source": "vllm",
+  "model_name": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "model_name_sanitized": "results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "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",
+  "total_evaluation_time_seconds": "59.45884429698344"
+}

eval_results/vllm/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results_2026-05-20T05-14-37.474356.json

@@ -0,0 +1,176 @@
+{
+  "results": {
+    "gsm_plus": {
+      "name": "gsm_plus",
+      "alias": "gsm_plus",
+      "sample_len": 10552,
+      "exact_match,strict-match": 0.3302691432903715,
+      "exact_match_stderr,strict-match": 0.004578648906188317,
+      "exact_match,flexible-extract": 0.4431387414708112,
+      "exact_match_stderr,flexible-extract": 0.0048361142334929315
+    }
+  },
+  "group_subtasks": {},
+  "configs": {
+    "gsm_plus": {
+      "task": "gsm_plus",
+      "dataset_path": "qintongli/GSM-Plus",
+      "training_split": "test",
+      "test_split": "test",
+      "fewshot_split": "test",
+      "doc_to_text": "Question: {{question}}\nAnswer:",
+      "doc_to_target": "{{solution}}",
+      "unsafe_code": false,
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "default",
+        "split": "test",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "Question: {{question}}\nAnswer:",
+        "doc_to_choice": null,
+        "doc_to_target": "{{solution}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 5,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true,
+          "ignore_case": true,
+          "ignore_punctuation": false,
+          "regexes_to_ignore": [
+            ",",
+            "\\$",
+            "(?s).*#### ",
+            "\\.$"
+          ]
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Question:",
+          "</s>",
+          "<|im_end|>"
+        ],
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "filter_list": [
+        {
+          "name": "strict-match",
+          "filter": [
+            {
+              "function": "regex",
+              "regex_pattern": "#### (\\-?[0-9\\.\\,]+)"
+            },
+            {
+              "function": "take_first"
+            }
+          ]
+        },
+        {
+          "name": "flexible-extract",
+          "filter": [
+            {
+              "function": "regex",
+              "group_select": -1,
+              "regex_pattern": "(-?[$0-9.,]{2,})|(-?[0-9]+)"
+            },
+            {
+              "function": "take_first"
+            }
+          ]
+        }
+      ],
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/gsm_plus/gsm_plus.yaml"
+      }
+    }
+  },
+  "versions": {
+    "gsm_plus": 1.0
+  },
+  "n-shot": {
+    "gsm_plus": 5
+  },
+  "higher_is_better": {
+    "gsm_plus": {
+      "exact_match": true
+    }
+  },
+  "n-samples": {
+    "gsm_plus": {
+      "original": 10552,
+      "effective": 10552
+    }
+  },
+  "config": {
+    "model": "vllm",
+    "model_args": {
+      "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+      "tensor_parallel_size": 2,
+      "dtype": "float16",
+      "gpu_memory_utilization": 0.8,
+      "trust_remote_code": true
+    },
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {},
+    "random_seed": 0,
+    "numpy_seed": 1234,
+    "torch_seed": 1234,
+    "fewshot_seed": 1234
+  },
+  "git_hash": "04b98db",
+  "date": 1779253911.668041,
+  "pretty_env_info": "PyTorch version: 2.10.0+cu128\nIs debug build: False\nCUDA used to build PyTorch: 12.8\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 24.04.4 LTS (x86_64)\nGCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0\nClang version: Could not collect\nCMake version: version 3.28.3\nLibc version: glibc-2.39\n\nPython version: 3.10.20 (main, May 10 2026, 19:27:29) [Clang 22.1.3 ] (64-bit runtime)\nPython platform: Linux-6.17.0-1013-aws-x86_64-with-glibc2.39\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: \nGPU models and configuration: \nGPU 0: NVIDIA H200\nGPU 1: NVIDIA H200\nGPU 2: NVIDIA H200\nGPU 3: NVIDIA H200\nGPU 4: NVIDIA H200\nGPU 5: NVIDIA H200\nGPU 6: NVIDIA H200\nGPU 7: NVIDIA H200\n\nNvidia driver version: 580.126.16\ncuDNN version: Probably one of the following:\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_adv.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_cnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_precompiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_runtime_compiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_graph.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_heuristic.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_ops.so.9\nIs XPU available: False\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\nCaching allocator config: N/A\n\nCPU:\nArchitecture:                            x86_64\nCPU op-mode(s):                          32-bit, 64-bit\nAddress sizes:                           48 bits physical, 48 bits virtual\nByte Order:                              Little Endian\nCPU(s):                                  192\nOn-line CPU(s) list:                     0-191\nVendor ID:                               AuthenticAMD\nModel name:                              AMD EPYC 7R13 Processor\nCPU family:                              25\nModel:                                   1\nThread(s) per core:                      2\nCore(s) per socket:                      48\nSocket(s):                               2\nStepping:                                1\nBogoMIPS:                                5299.99\nFlags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf tsc_known_freq pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch topoext perfctr_core ssbd ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 clzero xsaveerptr rdpru wbnoinvd arat npt nrip_save vaes vpclmulqdq rdpid\nHypervisor vendor:                       KVM\nVirtualization type:                     full\nL1d cache:                               3 MiB (96 instances)\nL1i cache:                               3 MiB (96 instances)\nL2 cache:                                48 MiB (96 instances)\nL3 cache:                                384 MiB (12 instances)\nNUMA node(s):                            4\nNUMA node0 CPU(s):                       0-23,96-119\nNUMA node1 CPU(s):                       24-47,120-143\nNUMA node2 CPU(s):                       48-71,144-167\nNUMA node3 CPU(s):                       72-95,168-191\nVulnerability Gather data sampling:      Not affected\nVulnerability Ghostwrite:                Not affected\nVulnerability Indirect target selection: 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 Old microcode:             Not affected\nVulnerability Reg file data sampling:    Not affected\nVulnerability Retbleed:                  Not affected\nVulnerability Spec rstack overflow:      Mitigation; Safe RET\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; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                     Not affected\nVulnerability Tsa:                       Mitigation; Clear CPU buffers\nVulnerability Tsx async abort:           Not affected\nVulnerability Vmscape:                   Not affected\n\nVersions of relevant libraries:\n[pip3] Could not collect\n[conda] numpy                       2.4.5            pypi_0              pypi\n[conda] nvidia-cublas-cu12          12.4.5.8         pypi_0              pypi\n[conda] nvidia-cuda-cupti-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-nvrtc-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-runtime-cu12    12.4.127         pypi_0              pypi\n[conda] nvidia-cudnn-cu12           9.1.0.70         pypi_0              pypi\n[conda] nvidia-cufft-cu12           11.2.1.3         pypi_0              pypi\n[conda] nvidia-curand-cu12          10.3.5.147       pypi_0              pypi\n[conda] nvidia-cusolver-cu12        11.6.1.9         pypi_0              pypi\n[conda] nvidia-cusparse-cu12        12.3.1.170       pypi_0              pypi\n[conda] nvidia-nccl-cu12            2.21.5           pypi_0              pypi\n[conda] nvidia-nvjitlink-cu12       12.4.127         pypi_0              pypi\n[conda] nvidia-nvtx-cu12            12.4.127         pypi_0              pypi\n[conda] torch                       2.5.1+cu124      pypi_0              pypi",
+  "transformers_version": "4.57.3",
+  "lm_eval_version": "0.4.12",
+  "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": {
+    "gsm_plus": "9fc5cf479857a1eb1a143016f6217e0f6f7a6a3251c77540c70b68178e0daa41"
+  },
+  "model_source": "vllm",
+  "model_name": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "model_name_sanitized": "results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "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",
+  "total_evaluation_time_seconds": "173.83287550101522"
+}

eval_results/vllm/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results_2026-05-20T05-15-51.468213.json

@@ -0,0 +1,155 @@
+{
+  "results": {
+    "mmlu_pro_math": {
+      "name": "mmlu_pro_math",
+      "alias": "math",
+      "sample_len": 1351,
+      "exact_match,custom-extract": 0.4219096965210955,
+      "exact_match_stderr,custom-extract": 0.013441282580517128
+    }
+  },
+  "group_subtasks": {},
+  "configs": {
+    "mmlu_pro_math": {
+      "task": "mmlu_pro_math",
+      "task_alias": "math",
+      "dataset_path": "TIGER-Lab/MMLU-Pro",
+      "test_split": "test",
+      "fewshot_split": "validation",
+      "process_docs": "functools.partial(<function process_docs at 0x77cafd3e7640>, subject='math')",
+      "doc_to_text": "functools.partial(<function format_cot_example at 0x77cafd3e7370>, including_answer=False)",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "description": "The following are multiple choice questions (with answers) about math. Think step by step and then finish your answer with \"the answer is (X)\" where X is the correct letter choice.\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "validation",
+        "process_docs": "functools.partial(<function process_docs at 0x77cafd3e7640>, subject='math')",
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "functools.partial(<function format_cot_example at 0x77cafd3e7370>, including_answer=False)",
+        "doc_to_choice": null,
+        "doc_to_target": "functools.partial(<function format_cot_target at 0x77cafd3e77f0>, including_answer=True)",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 5,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true,
+          "ignore_case": true,
+          "ignore_punctuation": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Question:"
+        ],
+        "max_gen_toks": 2048,
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "filter_list": [
+        {
+          "name": "custom-extract",
+          "filter": [
+            {
+              "function": "regex",
+              "regex_pattern": "answer is \\(?([ABCDEFGHIJ])\\)?"
+            },
+            {
+              "function": "take_first"
+            }
+          ]
+        }
+      ],
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.1,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true,
+        "config_source": "/home/ubuntu/AMiD/.venv/lib/python3.10/site-packages/lm_eval/tasks/mmlu_pro/mmlu_pro_math.yaml"
+      }
+    }
+  },
+  "versions": {
+    "mmlu_pro_math": 3.1
+  },
+  "n-shot": {
+    "mmlu_pro_math": 5
+  },
+  "higher_is_better": {
+    "mmlu_pro_math": {
+      "exact_match": true
+    }
+  },
+  "n-samples": {
+    "mmlu_pro_math": {
+      "original": 1351,
+      "effective": 1351
+    }
+  },
+  "config": {
+    "model": "vllm",
+    "model_args": {
+      "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+      "tensor_parallel_size": 2,
+      "dtype": "float16",
+      "gpu_memory_utilization": 0.8,
+      "trust_remote_code": true
+    },
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {},
+    "random_seed": 0,
+    "numpy_seed": 1234,
+    "torch_seed": 1234,
+    "fewshot_seed": 1234
+  },
+  "git_hash": "04b98db",
+  "date": 1779254089.912131,
+  "pretty_env_info": "PyTorch version: 2.10.0+cu128\nIs debug build: False\nCUDA used to build PyTorch: 12.8\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 24.04.4 LTS (x86_64)\nGCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0\nClang version: Could not collect\nCMake version: version 3.28.3\nLibc version: glibc-2.39\n\nPython version: 3.10.20 (main, May 10 2026, 19:27:29) [Clang 22.1.3 ] (64-bit runtime)\nPython platform: Linux-6.17.0-1013-aws-x86_64-with-glibc2.39\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: \nGPU models and configuration: \nGPU 0: NVIDIA H200\nGPU 1: NVIDIA H200\nGPU 2: NVIDIA H200\nGPU 3: NVIDIA H200\nGPU 4: NVIDIA H200\nGPU 5: NVIDIA H200\nGPU 6: NVIDIA H200\nGPU 7: NVIDIA H200\n\nNvidia driver version: 580.126.16\ncuDNN version: Probably one of the following:\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_adv.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_cnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_precompiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_runtime_compiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_graph.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_heuristic.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_ops.so.9\nIs XPU available: False\nHIP runtime version: N/A\nMIOpen runtime version: N/A\nIs XNNPACK available: True\nCaching allocator config: N/A\n\nCPU:\nArchitecture:                            x86_64\nCPU op-mode(s):                          32-bit, 64-bit\nAddress sizes:                           48 bits physical, 48 bits virtual\nByte Order:                              Little Endian\nCPU(s):                                  192\nOn-line CPU(s) list:                     0-191\nVendor ID:                               AuthenticAMD\nModel name:                              AMD EPYC 7R13 Processor\nCPU family:                              25\nModel:                                   1\nThread(s) per core:                      2\nCore(s) per socket:                      48\nSocket(s):                               2\nStepping:                                1\nBogoMIPS:                                5299.99\nFlags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf tsc_known_freq pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch topoext perfctr_core ssbd ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 clzero xsaveerptr rdpru wbnoinvd arat npt nrip_save vaes vpclmulqdq rdpid\nHypervisor vendor:                       KVM\nVirtualization type:                     full\nL1d cache:                               3 MiB (96 instances)\nL1i cache:                               3 MiB (96 instances)\nL2 cache:                                48 MiB (96 instances)\nL3 cache:                                384 MiB (12 instances)\nNUMA node(s):                            4\nNUMA node0 CPU(s):                       0-23,96-119\nNUMA node1 CPU(s):                       24-47,120-143\nNUMA node2 CPU(s):                       48-71,144-167\nNUMA node3 CPU(s):                       72-95,168-191\nVulnerability Gather data sampling:      Not affected\nVulnerability Ghostwrite:                Not affected\nVulnerability Indirect target selection: 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 Old microcode:             Not affected\nVulnerability Reg file data sampling:    Not affected\nVulnerability Retbleed:                  Not affected\nVulnerability Spec rstack overflow:      Mitigation; Safe RET\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; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                     Not affected\nVulnerability Tsa:                       Mitigation; Clear CPU buffers\nVulnerability Tsx async abort:           Not affected\nVulnerability Vmscape:                   Not affected\n\nVersions of relevant libraries:\n[pip3] Could not collect\n[conda] numpy                       2.4.5            pypi_0              pypi\n[conda] nvidia-cublas-cu12          12.4.5.8         pypi_0              pypi\n[conda] nvidia-cuda-cupti-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-nvrtc-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-runtime-cu12    12.4.127         pypi_0              pypi\n[conda] nvidia-cudnn-cu12           9.1.0.70         pypi_0              pypi\n[conda] nvidia-cufft-cu12           11.2.1.3         pypi_0              pypi\n[conda] nvidia-curand-cu12          10.3.5.147       pypi_0              pypi\n[conda] nvidia-cusolver-cu12        11.6.1.9         pypi_0              pypi\n[conda] nvidia-cusparse-cu12        12.3.1.170       pypi_0              pypi\n[conda] nvidia-nccl-cu12            2.21.5           pypi_0              pypi\n[conda] nvidia-nvjitlink-cu12       12.4.127         pypi_0              pypi\n[conda] nvidia-nvtx-cu12            12.4.127         pypi_0              pypi\n[conda] torch                       2.5.1+cu124      pypi_0              pypi",
+  "transformers_version": "4.57.3",
+  "lm_eval_version": "0.4.12",
+  "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": {
+    "mmlu_pro_math": "2ddae5317c3ae5b3b96f4724d31f9877fdaf30fb4e532897de01a2e284b82493"
+  },
+  "model_source": "vllm",
+  "model_name": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "model_name_sanitized": "results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "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",
+  "total_evaluation_time_seconds": "69.59606047498528"
+}

eval_results/vllm/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results_2026-05-20T09-42-55.394719.json

@@ -0,0 +1,658 @@
+{
+  "results": {
+    "minerva_math": {
+      "math_verify,none": 0.3274,
+      "math_verify_stderr,none": 0.0061821901048786584,
+      "exact_match,none": 0.154,
+      "exact_match_stderr,none": 0.004921159117892714,
+      "alias": "minerva_math"
+    },
+    "minerva_math_algebra": {
+      "alias": " - minerva_math_algebra",
+      "exact_match,none": 0.22156697556866048,
+      "exact_match_stderr,none": 0.012059263822234172,
+      "math_verify,none": 0.485256950294861,
+      "math_verify_stderr,none": 0.014512384734975407
+    },
+    "minerva_math_counting_and_prob": {
+      "alias": " - minerva_math_counting_and_prob",
+      "exact_match,none": 0.1371308016877637,
+      "exact_match_stderr,none": 0.01581646484131674,
+      "math_verify,none": 0.25949367088607594,
+      "math_verify_stderr,none": 0.020155668268063687
+    },
+    "minerva_math_geometry": {
+      "alias": " - minerva_math_geometry",
+      "exact_match,none": 0.1315240083507307,
+      "exact_match_stderr,none": 0.015458504556847504,
+      "math_verify,none": 0.29018789144050106,
+      "math_verify_stderr,none": 0.02075856828195949
+    },
+    "minerva_math_intermediate_algebra": {
+      "alias": " - minerva_math_intermediate_algebra",
+      "exact_match,none": 0.04318936877076412,
+      "exact_match_stderr,none": 0.006768589184759893,
+      "math_verify,none": 0.10852713178294573,
+      "math_verify_stderr,none": 0.010356665722569117
+    },
+    "minerva_math_num_theory": {
+      "alias": " - minerva_math_num_theory",
+      "exact_match,none": 0.07037037037037037,
+      "exact_match_stderr,none": 0.01101680124762625,
+      "math_verify,none": 0.2388888888888889,
+      "math_verify_stderr,none": 0.018366551616528413
+    },
+    "minerva_math_prealgebra": {
+      "alias": " - minerva_math_prealgebra",
+      "exact_match,none": 0.3099885189437428,
+      "exact_match_stderr,none": 0.01567982953031626,
+      "math_verify,none": 0.5648679678530425,
+      "math_verify_stderr,none": 0.01680832260828408
+    },
+    "minerva_math_precalc": {
+      "alias": " - minerva_math_precalc",
+      "exact_match,none": 0.05860805860805861,
+      "exact_match_stderr,none": 0.010061567725278804,
+      "math_verify,none": 0.14652014652014653,
+      "math_verify_stderr,none": 0.015147712649192267
+    }
+  },
+  "groups": {
+    "minerva_math": {
+      "math_verify,none": 0.3274,
+      "math_verify_stderr,none": 0.0061821901048786584,
+      "exact_match,none": 0.154,
+      "exact_match_stderr,none": 0.004921159117892714,
+      "alias": "minerva_math"
+    }
+  },
+  "group_subtasks": {
+    "minerva_math": [
+      "minerva_math_algebra",
+      "minerva_math_counting_and_prob",
+      "minerva_math_geometry",
+      "minerva_math_intermediate_algebra",
+      "minerva_math_num_theory",
+      "minerva_math_prealgebra",
+      "minerva_math_precalc"
+    ]
+  },
+  "configs": {
+    "minerva_math_algebra": {
+      "task": "minerva_math_algebra",
+      "tag": [
+        "math_word_problems"
+      ],
+      "dataset_path": "EleutherAI/hendrycks_math",
+      "dataset_name": "algebra",
+      "training_split": "train",
+      "test_split": "test",
+      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        out_doc = {\n            \"problem\": doc[\"problem\"],\n            \"solution\": doc[\"solution\"],\n            \"answer\": normalize_final_answer(\n                remove_boxed(last_boxed_only_string(doc[\"solution\"]))\n            ),\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n\n    return dataset.map(_process_doc)\n",
+      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return \"Problem:\" + \"\\n\" + doc[\"problem\"] + \"\\n\\n\" + \"Solution:\"\n",
+      "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+      "unsafe_code": false,
+      "process_results": "def process_results(doc: dict, results: list[str]) -> dict[str, int]:\n    candidates = results[0]\n\n    unnormalized_answer = get_unnormalized_answer(candidates)\n    answer = normalize_final_answer(unnormalized_answer)\n\n    if is_equiv(answer, doc[\"answer\"]):\n        retval = 1\n    else:\n        retval = 0\n\n    # math_verify\n    _mvres = verify(\n        gold=parse(doc[\"solution\"]),\n        target=parse(candidates),\n    )\n    mathval = 1 if _mvres else 0\n\n    res = {\n        \"exact_match\": retval,\n        \"math_verify\": mathval,\n    }\n    return res\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": null,
+        "process_docs": "<function process_docs at 0x77e7dc04c790>",
+        "fewshot_indices": null,
+        "samples": "<function list_fewshot_samples at 0x77e7dc04e830>",
+        "doc_to_text": "<function doc_to_text at 0x77e7dc04d1b0>",
+        "doc_to_choice": null,
+        "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 4,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "math_verify",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Problem:"
+        ],
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "minerva_math_counting_and_prob": {
+      "task": "minerva_math_counting_and_prob",
+      "tag": [
+        "math_word_problems"
+      ],
+      "dataset_path": "EleutherAI/hendrycks_math",
+      "dataset_name": "counting_and_probability",
+      "training_split": "train",
+      "test_split": "test",
+      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        out_doc = {\n            \"problem\": doc[\"problem\"],\n            \"solution\": doc[\"solution\"],\n            \"answer\": normalize_final_answer(\n                remove_boxed(last_boxed_only_string(doc[\"solution\"]))\n            ),\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n\n    return dataset.map(_process_doc)\n",
+      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return \"Problem:\" + \"\\n\" + doc[\"problem\"] + \"\\n\\n\" + \"Solution:\"\n",
+      "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+      "unsafe_code": false,
+      "process_results": "def process_results(doc: dict, results: list[str]) -> dict[str, int]:\n    candidates = results[0]\n\n    unnormalized_answer = get_unnormalized_answer(candidates)\n    answer = normalize_final_answer(unnormalized_answer)\n\n    if is_equiv(answer, doc[\"answer\"]):\n        retval = 1\n    else:\n        retval = 0\n\n    # math_verify\n    _mvres = verify(\n        gold=parse(doc[\"solution\"]),\n        target=parse(candidates),\n    )\n    mathval = 1 if _mvres else 0\n\n    res = {\n        \"exact_match\": retval,\n        \"math_verify\": mathval,\n    }\n    return res\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": null,
+        "process_docs": "<function process_docs at 0x77e7dc02a050>",
+        "fewshot_indices": null,
+        "samples": "<function list_fewshot_samples at 0x77e7dc04c160>",
+        "doc_to_text": "<function doc_to_text at 0x77e7dc02aa70>",
+        "doc_to_choice": null,
+        "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 4,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "math_verify",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Problem:"
+        ],
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "minerva_math_geometry": {
+      "task": "minerva_math_geometry",
+      "tag": [
+        "math_word_problems"
+      ],
+      "dataset_path": "EleutherAI/hendrycks_math",
+      "dataset_name": "geometry",
+      "training_split": "train",
+      "test_split": "test",
+      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        out_doc = {\n            \"problem\": doc[\"problem\"],\n            \"solution\": doc[\"solution\"],\n            \"answer\": normalize_final_answer(\n                remove_boxed(last_boxed_only_string(doc[\"solution\"]))\n            ),\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n\n    return dataset.map(_process_doc)\n",
+      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return \"Problem:\" + \"\\n\" + doc[\"problem\"] + \"\\n\\n\" + \"Solution:\"\n",
+      "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+      "unsafe_code": false,
+      "process_results": "def process_results(doc: dict, results: list[str]) -> dict[str, int]:\n    candidates = results[0]\n\n    unnormalized_answer = get_unnormalized_answer(candidates)\n    answer = normalize_final_answer(unnormalized_answer)\n\n    if is_equiv(answer, doc[\"answer\"]):\n        retval = 1\n    else:\n        retval = 0\n\n    # math_verify\n    _mvres = verify(\n        gold=parse(doc[\"solution\"]),\n        target=parse(candidates),\n    )\n    mathval = 1 if _mvres else 0\n\n    res = {\n        \"exact_match\": retval,\n        \"math_verify\": mathval,\n    }\n    return res\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": null,
+        "process_docs": "<function process_docs at 0x77e7dcddf6d0>",
+        "fewshot_indices": null,
+        "samples": "<function list_fewshot_samples at 0x77e7dcddf130>",
+        "doc_to_text": "<function doc_to_text at 0x77e7dc028160>",
+        "doc_to_choice": null,
+        "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 4,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "math_verify",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Problem:"
+        ],
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "minerva_math_intermediate_algebra": {
+      "task": "minerva_math_intermediate_algebra",
+      "tag": [
+        "math_word_problems"
+      ],
+      "dataset_path": "EleutherAI/hendrycks_math",
+      "dataset_name": "intermediate_algebra",
+      "training_split": "train",
+      "test_split": "test",
+      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        out_doc = {\n            \"problem\": doc[\"problem\"],\n            \"solution\": doc[\"solution\"],\n            \"answer\": normalize_final_answer(\n                remove_boxed(last_boxed_only_string(doc[\"solution\"]))\n            ),\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n\n    return dataset.map(_process_doc)\n",
+      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return \"Problem:\" + \"\\n\" + doc[\"problem\"] + \"\\n\\n\" + \"Solution:\"\n",
+      "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+      "unsafe_code": false,
+      "process_results": "def process_results(doc: dict, results: list[str]) -> dict[str, int]:\n    candidates = results[0]\n\n    unnormalized_answer = get_unnormalized_answer(candidates)\n    answer = normalize_final_answer(unnormalized_answer)\n\n    if is_equiv(answer, doc[\"answer\"]):\n        retval = 1\n    else:\n        retval = 0\n\n    # math_verify\n    _mvres = verify(\n        gold=parse(doc[\"solution\"]),\n        target=parse(candidates),\n    )\n    mathval = 1 if _mvres else 0\n\n    res = {\n        \"exact_match\": retval,\n        \"math_verify\": mathval,\n    }\n    return res\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": null,
+        "process_docs": "<function process_docs at 0x77e7dcddd240>",
+        "fewshot_indices": null,
+        "samples": "<function list_fewshot_samples at 0x77e7dcda9cf0>",
+        "doc_to_text": "<function doc_to_text at 0x77e7dcdddc60>",
+        "doc_to_choice": null,
+        "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 4,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "math_verify",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Problem:"
+        ],
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "minerva_math_num_theory": {
+      "task": "minerva_math_num_theory",
+      "tag": [
+        "math_word_problems"
+      ],
+      "dataset_path": "EleutherAI/hendrycks_math",
+      "dataset_name": "number_theory",
+      "training_split": "train",
+      "test_split": "test",
+      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        out_doc = {\n            \"problem\": doc[\"problem\"],\n            \"solution\": doc[\"solution\"],\n            \"answer\": normalize_final_answer(\n                remove_boxed(last_boxed_only_string(doc[\"solution\"]))\n            ),\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n\n    return dataset.map(_process_doc)\n",
+      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return \"Problem:\" + \"\\n\" + doc[\"problem\"] + \"\\n\\n\" + \"Solution:\"\n",
+      "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+      "unsafe_code": false,
+      "process_results": "def process_results(doc: dict, results: list[str]) -> dict[str, int]:\n    candidates = results[0]\n\n    unnormalized_answer = get_unnormalized_answer(candidates)\n    answer = normalize_final_answer(unnormalized_answer)\n\n    if is_equiv(answer, doc[\"answer\"]):\n        retval = 1\n    else:\n        retval = 0\n\n    # math_verify\n    _mvres = verify(\n        gold=parse(doc[\"solution\"]),\n        target=parse(candidates),\n    )\n    mathval = 1 if _mvres else 0\n\n    res = {\n        \"exact_match\": retval,\n        \"math_verify\": mathval,\n    }\n    return res\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": null,
+        "process_docs": "<function process_docs at 0x77e7dcdaa950>",
+        "fewshot_indices": null,
+        "samples": "<function list_fewshot_samples at 0x77e7dcdaa560>",
+        "doc_to_text": "<function doc_to_text at 0x77e7dcdab370>",
+        "doc_to_choice": null,
+        "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 4,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "math_verify",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Problem:"
+        ],
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "minerva_math_prealgebra": {
+      "task": "minerva_math_prealgebra",
+      "tag": [
+        "math_word_problems"
+      ],
+      "dataset_path": "EleutherAI/hendrycks_math",
+      "dataset_name": "prealgebra",
+      "training_split": "train",
+      "test_split": "test",
+      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        out_doc = {\n            \"problem\": doc[\"problem\"],\n            \"solution\": doc[\"solution\"],\n            \"answer\": normalize_final_answer(\n                remove_boxed(last_boxed_only_string(doc[\"solution\"]))\n            ),\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n\n    return dataset.map(_process_doc)\n",
+      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return \"Problem:\" + \"\\n\" + doc[\"problem\"] + \"\\n\\n\" + \"Solution:\"\n",
+      "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+      "unsafe_code": false,
+      "process_results": "def process_results(doc: dict, results: list[str]) -> dict[str, int]:\n    candidates = results[0]\n\n    unnormalized_answer = get_unnormalized_answer(candidates)\n    answer = normalize_final_answer(unnormalized_answer)\n\n    if is_equiv(answer, doc[\"answer\"]):\n        retval = 1\n    else:\n        retval = 0\n\n    # math_verify\n    _mvres = verify(\n        gold=parse(doc[\"solution\"]),\n        target=parse(candidates),\n    )\n    mathval = 1 if _mvres else 0\n\n    res = {\n        \"exact_match\": retval,\n        \"math_verify\": mathval,\n    }\n    return res\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": null,
+        "process_docs": "<function process_docs at 0x77e7dcd727a0>",
+        "fewshot_indices": null,
+        "samples": "<function list_fewshot_samples at 0x77e7dcd73880>",
+        "doc_to_text": "<function doc_to_text at 0x77e7dcd71bd0>",
+        "doc_to_choice": null,
+        "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 4,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "math_verify",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Problem:"
+        ],
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "minerva_math_precalc": {
+      "task": "minerva_math_precalc",
+      "tag": [
+        "math_word_problems"
+      ],
+      "dataset_path": "EleutherAI/hendrycks_math",
+      "dataset_name": "precalculus",
+      "training_split": "train",
+      "test_split": "test",
+      "process_docs": "def process_docs(dataset: datasets.Dataset) -> datasets.Dataset:\n    def _process_doc(doc: dict) -> dict:\n        out_doc = {\n            \"problem\": doc[\"problem\"],\n            \"solution\": doc[\"solution\"],\n            \"answer\": normalize_final_answer(\n                remove_boxed(last_boxed_only_string(doc[\"solution\"]))\n            ),\n        }\n        if getattr(doc, \"few_shot\", None) is not None:\n            out_doc[\"few_shot\"] = True\n        return out_doc\n\n    return dataset.map(_process_doc)\n",
+      "doc_to_text": "def doc_to_text(doc: dict) -> str:\n    return \"Problem:\" + \"\\n\" + doc[\"problem\"] + \"\\n\\n\" + \"Solution:\"\n",
+      "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+      "unsafe_code": false,
+      "process_results": "def process_results(doc: dict, results: list[str]) -> dict[str, int]:\n    candidates = results[0]\n\n    unnormalized_answer = get_unnormalized_answer(candidates)\n    answer = normalize_final_answer(unnormalized_answer)\n\n    if is_equiv(answer, doc[\"answer\"]):\n        retval = 1\n    else:\n        retval = 0\n\n    # math_verify\n    _mvres = verify(\n        gold=parse(doc[\"solution\"]),\n        target=parse(candidates),\n    )\n    mathval = 1 if _mvres else 0\n\n    res = {\n        \"exact_match\": retval,\n        \"math_verify\": mathval,\n    }\n    return res\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": null,
+        "process_docs": "<function process_docs at 0x77e7dd195a20>",
+        "fewshot_indices": null,
+        "samples": "<function list_fewshot_samples at 0x77e7dd197490>",
+        "doc_to_text": "<function doc_to_text at 0x77e7dd194ee0>",
+        "doc_to_choice": null,
+        "doc_to_target": "{{answer if few_shot is undefined else solution}}",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 4,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "math_verify",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Problem:"
+        ],
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    }
+  },
+  "versions": {
+    "minerva_math": 1.0,
+    "minerva_math_algebra": 3.0,
+    "minerva_math_counting_and_prob": 3.0,
+    "minerva_math_geometry": 3.0,
+    "minerva_math_intermediate_algebra": 3.0,
+    "minerva_math_num_theory": 3.0,
+    "minerva_math_prealgebra": 3.0,
+    "minerva_math_precalc": 3.0
+  },
+  "n-shot": {
+    "minerva_math_algebra": 4,
+    "minerva_math_counting_and_prob": 4,
+    "minerva_math_geometry": 4,
+    "minerva_math_intermediate_algebra": 4,
+    "minerva_math_num_theory": 4,
+    "minerva_math_prealgebra": 4,
+    "minerva_math_precalc": 4
+  },
+  "higher_is_better": {
+    "minerva_math": {
+      "exact_match": true,
+      "math_verify": true
+    },
+    "minerva_math_algebra": {
+      "exact_match": true,
+      "math_verify": true
+    },
+    "minerva_math_counting_and_prob": {
+      "exact_match": true,
+      "math_verify": true
+    },
+    "minerva_math_geometry": {
+      "exact_match": true,
+      "math_verify": true
+    },
+    "minerva_math_intermediate_algebra": {
+      "exact_match": true,
+      "math_verify": true
+    },
+    "minerva_math_num_theory": {
+      "exact_match": true,
+      "math_verify": true
+    },
+    "minerva_math_prealgebra": {
+      "exact_match": true,
+      "math_verify": true
+    },
+    "minerva_math_precalc": {
+      "exact_match": true,
+      "math_verify": true
+    }
+  },
+  "n-samples": {
+    "minerva_math_algebra": {
+      "original": 1187,
+      "effective": 1187
+    },
+    "minerva_math_counting_and_prob": {
+      "original": 474,
+      "effective": 474
+    },
+    "minerva_math_geometry": {
+      "original": 479,
+      "effective": 479
+    },
+    "minerva_math_intermediate_algebra": {
+      "original": 903,
+      "effective": 903
+    },
+    "minerva_math_num_theory": {
+      "original": 540,
+      "effective": 540
+    },
+    "minerva_math_prealgebra": {
+      "original": 871,
+      "effective": 871
+    },
+    "minerva_math_precalc": {
+      "original": 546,
+      "effective": 546
+    }
+  },
+  "config": {
+    "model": "vllm",
+    "model_args": {
+      "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+      "tensor_parallel_size": 2,
+      "dtype": "float16",
+      "gpu_memory_utilization": 0.8,
+      "trust_remote_code": true
+    },
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {},
+    "random_seed": 0,
+    "numpy_seed": 1234,
+    "torch_seed": 1234,
+    "fewshot_seed": 1234
+  },
+  "git_hash": "10cc1bd",
+  "date": 1779269874.648554,
+  "pretty_env_info": "PyTorch version: 2.9.1+cu128\nIs debug build: False\nCUDA used to build PyTorch: 12.8\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 24.04.4 LTS (x86_64)\nGCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0\nClang version: Could not collect\nCMake version: version 3.28.3\nLibc version: glibc-2.39\n\nPython version: 3.10.20 (main, May 10 2026, 19:27:29) [Clang 22.1.3 ] (64-bit runtime)\nPython platform: Linux-6.17.0-1013-aws-x86_64-with-glibc2.39\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: \nGPU models and configuration: \nGPU 0: NVIDIA H200\nGPU 1: NVIDIA H200\nGPU 2: NVIDIA H200\nGPU 3: NVIDIA H200\nGPU 4: NVIDIA H200\nGPU 5: NVIDIA H200\nGPU 6: NVIDIA H200\nGPU 7: NVIDIA H200\n\nNvidia driver version: 580.126.16\ncuDNN version: Probably one of the following:\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_adv.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_cnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_precompiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_runtime_compiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_graph.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_heuristic.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_ops.so.9\nIs XPU available: False\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:                           48 bits physical, 48 bits virtual\nByte Order:                              Little Endian\nCPU(s):                                  192\nOn-line CPU(s) list:                     0-191\nVendor ID:                               AuthenticAMD\nModel name:                              AMD EPYC 7R13 Processor\nCPU family:                              25\nModel:                                   1\nThread(s) per core:                      2\nCore(s) per socket:                      48\nSocket(s):                               2\nStepping:                                1\nBogoMIPS:                                5299.99\nFlags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf tsc_known_freq pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch topoext perfctr_core ssbd ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 clzero xsaveerptr rdpru wbnoinvd arat npt nrip_save vaes vpclmulqdq rdpid\nHypervisor vendor:                       KVM\nVirtualization type:                     full\nL1d cache:                               3 MiB (96 instances)\nL1i cache:                               3 MiB (96 instances)\nL2 cache:                                48 MiB (96 instances)\nL3 cache:                                384 MiB (12 instances)\nNUMA node(s):                            4\nNUMA node0 CPU(s):                       0-23,96-119\nNUMA node1 CPU(s):                       24-47,120-143\nNUMA node2 CPU(s):                       48-71,144-167\nNUMA node3 CPU(s):                       72-95,168-191\nVulnerability Gather data sampling:      Not affected\nVulnerability Ghostwrite:                Not affected\nVulnerability Indirect target selection: 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 Old microcode:             Not affected\nVulnerability Reg file data sampling:    Not affected\nVulnerability Retbleed:                  Not affected\nVulnerability Spec rstack overflow:      Mitigation; Safe RET\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; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                     Not affected\nVulnerability Tsa:                       Mitigation; Clear CPU buffers\nVulnerability Tsx async abort:           Not affected\nVulnerability Vmscape:                   Not affected\n\nVersions of relevant libraries:\n[pip3] Could not collect\n[conda] numpy                       2.4.5            pypi_0              pypi\n[conda] nvidia-cublas-cu12          12.4.5.8         pypi_0              pypi\n[conda] nvidia-cuda-cupti-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-nvrtc-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-runtime-cu12    12.4.127         pypi_0              pypi\n[conda] nvidia-cudnn-cu12           9.1.0.70         pypi_0              pypi\n[conda] nvidia-cufft-cu12           11.2.1.3         pypi_0              pypi\n[conda] nvidia-curand-cu12          10.3.5.147       pypi_0              pypi\n[conda] nvidia-cusolver-cu12        11.6.1.9         pypi_0              pypi\n[conda] nvidia-cusparse-cu12        12.3.1.170       pypi_0              pypi\n[conda] nvidia-nccl-cu12            2.21.5           pypi_0              pypi\n[conda] nvidia-nvjitlink-cu12       12.4.127         pypi_0              pypi\n[conda] nvidia-nvtx-cu12            12.4.127         pypi_0              pypi\n[conda] torch                       2.5.1+cu124      pypi_0              pypi",
+  "transformers_version": "4.57.3",
+  "lm_eval_version": "0.4.11",
+  "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": {
+    "minerva_math_algebra": "a5597b7619af3e15e50469ae895189e416c0276a0b4efb7ac598bacb74ef7c86",
+    "minerva_math_counting_and_prob": "3c9aea6323941f5250896f019716ca268b53ea203877e263e5eb435df5110bde",
+    "minerva_math_geometry": "51d9457ad25bf15988ea75e6dc804dd80b437e5b50f9cb5fb211a7ee7256b75e",
+    "minerva_math_intermediate_algebra": "a94c1fc928d616f136f55bd50382516a109bf506696f75bb904b3e4cfb4a210e",
+    "minerva_math_num_theory": "6a9e327184475c4aabeba734be9a2c5a1435f0a138687194585e201e23aac3b0",
+    "minerva_math_prealgebra": "1ad14186c0a46b9c62b4e248ebb2bb5c2e3daaee2f501fcfb14efa4654ddb56d",
+    "minerva_math_precalc": "4d9f2cff781009dce40c0acd4942b199a05a6a576a3e1fb5e26123db28c35d0d"
+  },
+  "model_source": "vllm",
+  "model_name": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "model_name_sanitized": "results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "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",
+  "total_evaluation_time_seconds": "304.4746452749241"
+}

eval_results/vllm/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results_2026-05-20T09-44-39.890673.json

@@ -0,0 +1,1492 @@
+{
+  "results": {
+    "mmlu_stem": {
+      "acc,none": 0.4490960989533777,
+      "acc_stderr,none": 0.00856810876320342,
+      "alias": "stem"
+    },
+    "mmlu_abstract_algebra": {
+      "alias": " - abstract_algebra",
+      "acc,none": 0.23,
+      "acc_stderr,none": 0.04229525846816505
+    },
+    "mmlu_anatomy": {
+      "alias": " - anatomy",
+      "acc,none": 0.5111111111111111,
+      "acc_stderr,none": 0.04318275491977976
+    },
+    "mmlu_astronomy": {
+      "alias": " - astronomy",
+      "acc,none": 0.6052631578947368,
+      "acc_stderr,none": 0.039777499346220734
+    },
+    "mmlu_college_biology": {
+      "alias": " - college_biology",
+      "acc,none": 0.5972222222222222,
+      "acc_stderr,none": 0.04101405519842426
+    },
+    "mmlu_college_chemistry": {
+      "alias": " - college_chemistry",
+      "acc,none": 0.34,
+      "acc_stderr,none": 0.04760952285695235
+    },
+    "mmlu_college_computer_science": {
+      "alias": " - college_computer_science",
+      "acc,none": 0.32,
+      "acc_stderr,none": 0.046882617226215034
+    },
+    "mmlu_college_mathematics": {
+      "alias": " - college_mathematics",
+      "acc,none": 0.3,
+      "acc_stderr,none": 0.046056618647183814
+    },
+    "mmlu_college_physics": {
+      "alias": " - college_physics",
+      "acc,none": 0.37254901960784315,
+      "acc_stderr,none": 0.04810840148082633
+    },
+    "mmlu_computer_security": {
+      "alias": " - computer_security",
+      "acc,none": 0.67,
+      "acc_stderr,none": 0.04725815626252607
+    },
+    "mmlu_conceptual_physics": {
+      "alias": " - conceptual_physics",
+      "acc,none": 0.5702127659574469,
+      "acc_stderr,none": 0.03236214467715564
+    },
+    "mmlu_electrical_engineering": {
+      "alias": " - electrical_engineering",
+      "acc,none": 0.4689655172413793,
+      "acc_stderr,none": 0.04158632762097828
+    },
+    "mmlu_elementary_mathematics": {
+      "alias": " - elementary_mathematics",
+      "acc,none": 0.3862433862433862,
+      "acc_stderr,none": 0.02507598176760168
+    },
+    "mmlu_high_school_biology": {
+      "alias": " - high_school_biology",
+      "acc,none": 0.6516129032258065,
+      "acc_stderr,none": 0.027104826328100944
+    },
+    "mmlu_high_school_chemistry": {
+      "alias": " - high_school_chemistry",
+      "acc,none": 0.458128078817734,
+      "acc_stderr,none": 0.03505630140785741
+    },
+    "mmlu_high_school_computer_science": {
+      "alias": " - high_school_computer_science",
+      "acc,none": 0.54,
+      "acc_stderr,none": 0.05009082659620331
+    },
+    "mmlu_high_school_mathematics": {
+      "alias": " - high_school_mathematics",
+      "acc,none": 0.26666666666666666,
+      "acc_stderr,none": 0.026962424325073828
+    },
+    "mmlu_high_school_physics": {
+      "alias": " - high_school_physics",
+      "acc,none": 0.31125827814569534,
+      "acc_stderr,none": 0.03780445850526732
+    },
+    "mmlu_high_school_statistics": {
+      "alias": " - high_school_statistics",
+      "acc,none": 0.36574074074074076,
+      "acc_stderr,none": 0.03284738857647207
+    },
+    "mmlu_machine_learning": {
+      "alias": " - machine_learning",
+      "acc,none": 0.44642857142857145,
+      "acc_stderr,none": 0.047184714852195886
+    }
+  },
+  "groups": {
+    "mmlu_stem": {
+      "acc,none": 0.4490960989533777,
+      "acc_stderr,none": 0.00856810876320342,
+      "alias": "stem"
+    }
+  },
+  "group_subtasks": {
+    "mmlu_stem": [
+      "mmlu_abstract_algebra",
+      "mmlu_anatomy",
+      "mmlu_astronomy",
+      "mmlu_college_biology",
+      "mmlu_college_chemistry",
+      "mmlu_college_computer_science",
+      "mmlu_college_mathematics",
+      "mmlu_college_physics",
+      "mmlu_computer_security",
+      "mmlu_conceptual_physics",
+      "mmlu_electrical_engineering",
+      "mmlu_elementary_mathematics",
+      "mmlu_high_school_biology",
+      "mmlu_high_school_chemistry",
+      "mmlu_high_school_computer_science",
+      "mmlu_high_school_mathematics",
+      "mmlu_high_school_physics",
+      "mmlu_high_school_statistics",
+      "mmlu_machine_learning"
+    ]
+  },
+  "configs": {
+    "mmlu_abstract_algebra": {
+      "task": "mmlu_abstract_algebra",
+      "task_alias": "abstract_algebra",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "abstract_algebra",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about abstract algebra.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_anatomy": {
+      "task": "mmlu_anatomy",
+      "task_alias": "anatomy",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "anatomy",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about anatomy.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_astronomy": {
+      "task": "mmlu_astronomy",
+      "task_alias": "astronomy",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "astronomy",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about astronomy.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_college_biology": {
+      "task": "mmlu_college_biology",
+      "task_alias": "college_biology",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_biology",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college biology.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_college_chemistry": {
+      "task": "mmlu_college_chemistry",
+      "task_alias": "college_chemistry",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_chemistry",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college chemistry.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_college_computer_science": {
+      "task": "mmlu_college_computer_science",
+      "task_alias": "college_computer_science",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_computer_science",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college computer science.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_college_mathematics": {
+      "task": "mmlu_college_mathematics",
+      "task_alias": "college_mathematics",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_mathematics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college mathematics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_college_physics": {
+      "task": "mmlu_college_physics",
+      "task_alias": "college_physics",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "college_physics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about college physics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_computer_security": {
+      "task": "mmlu_computer_security",
+      "task_alias": "computer_security",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "computer_security",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about computer security.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_conceptual_physics": {
+      "task": "mmlu_conceptual_physics",
+      "task_alias": "conceptual_physics",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "conceptual_physics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about conceptual physics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_electrical_engineering": {
+      "task": "mmlu_electrical_engineering",
+      "task_alias": "electrical_engineering",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "electrical_engineering",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about electrical engineering.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_elementary_mathematics": {
+      "task": "mmlu_elementary_mathematics",
+      "task_alias": "elementary_mathematics",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "elementary_mathematics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about elementary mathematics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_high_school_biology": {
+      "task": "mmlu_high_school_biology",
+      "task_alias": "high_school_biology",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_biology",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school biology.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_high_school_chemistry": {
+      "task": "mmlu_high_school_chemistry",
+      "task_alias": "high_school_chemistry",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_chemistry",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school chemistry.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_high_school_computer_science": {
+      "task": "mmlu_high_school_computer_science",
+      "task_alias": "high_school_computer_science",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_computer_science",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school computer science.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_high_school_mathematics": {
+      "task": "mmlu_high_school_mathematics",
+      "task_alias": "high_school_mathematics",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_mathematics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school mathematics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_high_school_physics": {
+      "task": "mmlu_high_school_physics",
+      "task_alias": "high_school_physics",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_physics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school physics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_high_school_statistics": {
+      "task": "mmlu_high_school_statistics",
+      "task_alias": "high_school_statistics",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "high_school_statistics",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about high school statistics.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    },
+    "mmlu_machine_learning": {
+      "task": "mmlu_machine_learning",
+      "task_alias": "machine_learning",
+      "tag": "mmlu_stem_tasks",
+      "dataset_path": "cais/mmlu",
+      "dataset_name": "machine_learning",
+      "test_split": "test",
+      "fewshot_split": "dev",
+      "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+      "doc_to_target": "answer",
+      "unsafe_code": false,
+      "doc_to_choice": [
+        "A",
+        "B",
+        "C",
+        "D"
+      ],
+      "description": "The following are multiple choice questions (with answers) about machine learning.\n\n",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "first_n",
+        "split": "dev",
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "{{question.strip()}}\nA. {{choices[0]}}\nB. {{choices[1]}}\nC. {{choices[2]}}\nD. {{choices[3]}}\nAnswer:",
+        "doc_to_choice": [
+          "A",
+          "B",
+          "C",
+          "D"
+        ],
+        "doc_to_target": "answer",
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "multiple_choice",
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 1.0,
+        "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+        "tensor_parallel_size": 2,
+        "dtype": "float16",
+        "gpu_memory_utilization": 0.8,
+        "trust_remote_code": true
+      }
+    }
+  },
+  "versions": {
+    "mmlu_abstract_algebra": 1.0,
+    "mmlu_anatomy": 1.0,
+    "mmlu_astronomy": 1.0,
+    "mmlu_college_biology": 1.0,
+    "mmlu_college_chemistry": 1.0,
+    "mmlu_college_computer_science": 1.0,
+    "mmlu_college_mathematics": 1.0,
+    "mmlu_college_physics": 1.0,
+    "mmlu_computer_security": 1.0,
+    "mmlu_conceptual_physics": 1.0,
+    "mmlu_electrical_engineering": 1.0,
+    "mmlu_elementary_mathematics": 1.0,
+    "mmlu_high_school_biology": 1.0,
+    "mmlu_high_school_chemistry": 1.0,
+    "mmlu_high_school_computer_science": 1.0,
+    "mmlu_high_school_mathematics": 1.0,
+    "mmlu_high_school_physics": 1.0,
+    "mmlu_high_school_statistics": 1.0,
+    "mmlu_machine_learning": 1.0,
+    "mmlu_stem": 2
+  },
+  "n-shot": {
+    "mmlu_abstract_algebra": 0,
+    "mmlu_anatomy": 0,
+    "mmlu_astronomy": 0,
+    "mmlu_college_biology": 0,
+    "mmlu_college_chemistry": 0,
+    "mmlu_college_computer_science": 0,
+    "mmlu_college_mathematics": 0,
+    "mmlu_college_physics": 0,
+    "mmlu_computer_security": 0,
+    "mmlu_conceptual_physics": 0,
+    "mmlu_electrical_engineering": 0,
+    "mmlu_elementary_mathematics": 0,
+    "mmlu_high_school_biology": 0,
+    "mmlu_high_school_chemistry": 0,
+    "mmlu_high_school_computer_science": 0,
+    "mmlu_high_school_mathematics": 0,
+    "mmlu_high_school_physics": 0,
+    "mmlu_high_school_statistics": 0,
+    "mmlu_machine_learning": 0
+  },
+  "higher_is_better": {
+    "mmlu_abstract_algebra": {
+      "acc": true
+    },
+    "mmlu_anatomy": {
+      "acc": true
+    },
+    "mmlu_astronomy": {
+      "acc": true
+    },
+    "mmlu_college_biology": {
+      "acc": true
+    },
+    "mmlu_college_chemistry": {
+      "acc": true
+    },
+    "mmlu_college_computer_science": {
+      "acc": true
+    },
+    "mmlu_college_mathematics": {
+      "acc": true
+    },
+    "mmlu_college_physics": {
+      "acc": true
+    },
+    "mmlu_computer_security": {
+      "acc": true
+    },
+    "mmlu_conceptual_physics": {
+      "acc": true
+    },
+    "mmlu_electrical_engineering": {
+      "acc": true
+    },
+    "mmlu_elementary_mathematics": {
+      "acc": true
+    },
+    "mmlu_high_school_biology": {
+      "acc": true
+    },
+    "mmlu_high_school_chemistry": {
+      "acc": true
+    },
+    "mmlu_high_school_computer_science": {
+      "acc": true
+    },
+    "mmlu_high_school_mathematics": {
+      "acc": true
+    },
+    "mmlu_high_school_physics": {
+      "acc": true
+    },
+    "mmlu_high_school_statistics": {
+      "acc": true
+    },
+    "mmlu_machine_learning": {
+      "acc": true
+    },
+    "mmlu_stem": {
+      "acc": true
+    }
+  },
+  "n-samples": {
+    "mmlu_abstract_algebra": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_anatomy": {
+      "original": 135,
+      "effective": 135
+    },
+    "mmlu_astronomy": {
+      "original": 152,
+      "effective": 152
+    },
+    "mmlu_college_biology": {
+      "original": 144,
+      "effective": 144
+    },
+    "mmlu_college_chemistry": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_college_computer_science": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_college_mathematics": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_college_physics": {
+      "original": 102,
+      "effective": 102
+    },
+    "mmlu_computer_security": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_conceptual_physics": {
+      "original": 235,
+      "effective": 235
+    },
+    "mmlu_electrical_engineering": {
+      "original": 145,
+      "effective": 145
+    },
+    "mmlu_elementary_mathematics": {
+      "original": 378,
+      "effective": 378
+    },
+    "mmlu_high_school_biology": {
+      "original": 310,
+      "effective": 310
+    },
+    "mmlu_high_school_chemistry": {
+      "original": 203,
+      "effective": 203
+    },
+    "mmlu_high_school_computer_science": {
+      "original": 100,
+      "effective": 100
+    },
+    "mmlu_high_school_mathematics": {
+      "original": 270,
+      "effective": 270
+    },
+    "mmlu_high_school_physics": {
+      "original": 151,
+      "effective": 151
+    },
+    "mmlu_high_school_statistics": {
+      "original": 216,
+      "effective": 216
+    },
+    "mmlu_machine_learning": {
+      "original": 112,
+      "effective": 112
+    }
+  },
+  "config": {
+    "model": "vllm",
+    "model_args": {
+      "pretrained": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+      "tensor_parallel_size": 2,
+      "dtype": "float16",
+      "gpu_memory_utilization": 0.8,
+      "trust_remote_code": true
+    },
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {},
+    "random_seed": 0,
+    "numpy_seed": 1234,
+    "torch_seed": 1234,
+    "fewshot_seed": 1234
+  },
+  "git_hash": "10cc1bd",
+  "date": 1779270187.7350535,
+  "pretty_env_info": "PyTorch version: 2.9.1+cu128\nIs debug build: False\nCUDA used to build PyTorch: 12.8\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 24.04.4 LTS (x86_64)\nGCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0\nClang version: Could not collect\nCMake version: version 3.28.3\nLibc version: glibc-2.39\n\nPython version: 3.10.20 (main, May 10 2026, 19:27:29) [Clang 22.1.3 ] (64-bit runtime)\nPython platform: Linux-6.17.0-1013-aws-x86_64-with-glibc2.39\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: \nGPU models and configuration: \nGPU 0: NVIDIA H200\nGPU 1: NVIDIA H200\nGPU 2: NVIDIA H200\nGPU 3: NVIDIA H200\nGPU 4: NVIDIA H200\nGPU 5: NVIDIA H200\nGPU 6: NVIDIA H200\nGPU 7: NVIDIA H200\n\nNvidia driver version: 580.126.16\ncuDNN version: Probably one of the following:\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_adv.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_cnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_precompiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_runtime_compiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_graph.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_heuristic.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_ops.so.9\nIs XPU available: False\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:                           48 bits physical, 48 bits virtual\nByte Order:                              Little Endian\nCPU(s):                                  192\nOn-line CPU(s) list:                     0-191\nVendor ID:                               AuthenticAMD\nModel name:                              AMD EPYC 7R13 Processor\nCPU family:                              25\nModel:                                   1\nThread(s) per core:                      2\nCore(s) per socket:                      48\nSocket(s):                               2\nStepping:                                1\nBogoMIPS:                                5299.99\nFlags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf tsc_known_freq pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch topoext perfctr_core ssbd ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 clzero xsaveerptr rdpru wbnoinvd arat npt nrip_save vaes vpclmulqdq rdpid\nHypervisor vendor:                       KVM\nVirtualization type:                     full\nL1d cache:                               3 MiB (96 instances)\nL1i cache:                               3 MiB (96 instances)\nL2 cache:                                48 MiB (96 instances)\nL3 cache:                                384 MiB (12 instances)\nNUMA node(s):                            4\nNUMA node0 CPU(s):                       0-23,96-119\nNUMA node1 CPU(s):                       24-47,120-143\nNUMA node2 CPU(s):                       48-71,144-167\nNUMA node3 CPU(s):                       72-95,168-191\nVulnerability Gather data sampling:      Not affected\nVulnerability Ghostwrite:                Not affected\nVulnerability Indirect target selection: 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 Old microcode:             Not affected\nVulnerability Reg file data sampling:    Not affected\nVulnerability Retbleed:                  Not affected\nVulnerability Spec rstack overflow:      Mitigation; Safe RET\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; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                     Not affected\nVulnerability Tsa:                       Mitigation; Clear CPU buffers\nVulnerability Tsx async abort:           Not affected\nVulnerability Vmscape:                   Not affected\n\nVersions of relevant libraries:\n[pip3] Could not collect\n[conda] numpy                       2.4.5            pypi_0              pypi\n[conda] nvidia-cublas-cu12          12.4.5.8         pypi_0              pypi\n[conda] nvidia-cuda-cupti-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-nvrtc-cu12      12.4.127         pypi_0              pypi\n[conda] nvidia-cuda-runtime-cu12    12.4.127         pypi_0              pypi\n[conda] nvidia-cudnn-cu12           9.1.0.70         pypi_0              pypi\n[conda] nvidia-cufft-cu12           11.2.1.3         pypi_0              pypi\n[conda] nvidia-curand-cu12          10.3.5.147       pypi_0              pypi\n[conda] nvidia-cusolver-cu12        11.6.1.9         pypi_0              pypi\n[conda] nvidia-cusparse-cu12        12.3.1.170       pypi_0              pypi\n[conda] nvidia-nccl-cu12            2.21.5           pypi_0              pypi\n[conda] nvidia-nvjitlink-cu12       12.4.127         pypi_0              pypi\n[conda] nvidia-nvtx-cu12            12.4.127         pypi_0              pypi\n[conda] torch                       2.5.1+cu124      pypi_0              pypi",
+  "transformers_version": "4.57.3",
+  "lm_eval_version": "0.4.11",
+  "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": {
+    "mmlu_abstract_algebra": "1ecc29cafeff483bf80347adc2878dae00d01596988032f438afce6291c8cfe7",
+    "mmlu_anatomy": "51232e78f32967e0dd431bab9309f69be490f9998b4ddb93b641b09ba1f613cf",
+    "mmlu_astronomy": "3e8d7c33bd0431518a3d5d4b1666bdabccf46c4cc4cfe95a3886cb2266743f86",
+    "mmlu_college_biology": "5378a297efa1cd71344228c95dcb29ba54757c7e527261400a4d65fd8ceed92b",
+    "mmlu_college_chemistry": "714e60ea7e340f0344ffde41b247a19b68e6384152a1bdc9acc0ecf24cd0e7cd",
+    "mmlu_college_computer_science": "2b66d244f98e61e6a78e5ed48b728e0ca2ca81c7757ccdd086fe6b0c1cf02da1",
+    "mmlu_college_mathematics": "90b1ce25d7d38abd925181ae3aad095cfd783375fe8de0607a57420f47454f0f",
+    "mmlu_college_physics": "2abddb12a247904f9ba60b39766b19a8967c24fe8eee021022a353b93d73742c",
+    "mmlu_computer_security": "8ffba21cebb9e39d5794c848a22e7e37cddff5c59f7b47a1d89872d9e8704b0c",
+    "mmlu_conceptual_physics": "c0da3219fe7e5292dad2aa7e72790e3eba951711007cc57618b0850d941b85ec",
+    "mmlu_electrical_engineering": "0a02cb16a31ec7debba4190ab1cc31afadb9416764e0b074b9b96553f346b892",
+    "mmlu_elementary_mathematics": "6048f1506468e757010d537a0ceedaadc587c18159a03e7664b5994843afedc2",
+    "mmlu_high_school_biology": "2aabe83bb100a352d24a0136c24e612b68490bbe818c33f06cc6cee5379ecad8",
+    "mmlu_high_school_chemistry": "dd3321fd42c8717094f57dd0eaeab0002d9ed192518a7de7aaef4913796c30f0",
+    "mmlu_high_school_computer_science": "888aabb9a7ac77299e8b2b8429fa44c1fbbb5ac8165f556a0b00b347e51de3b5",
+    "mmlu_high_school_mathematics": "9b9567fc5c7875065ac91eac69621904f201cab822621440a2442a390f62b497",
+    "mmlu_high_school_physics": "53edb93aefb6fc54fba283278bae10a58a7902902ff55edc30cb9c81745dbaa8",
+    "mmlu_high_school_statistics": "e0d652bf0cb6f927b0fc738e58a141f0ba8f9073615f4ddd70bf7a3b97998f21",
+    "mmlu_machine_learning": "13ad1d3650a26d655a2a2eaa3be1f106d990e5b693b0e9006f8ebef221d2ed82"
+  },
+  "model_source": "vllm",
+  "model_name": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "model_name_sanitized": "results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "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",
+  "total_evaluation_time_seconds": "95.85217410896439"
+}

eval_results/vllm/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0/results_2026-05-20T10-00-28.593701.json

@@ -0,0 +1,156 @@
+{
+  "results": {
+    "gsm8k": {
+      "alias": "gsm8k",
+      "exact_match,strict-match": 0.30856709628506446,
+      "exact_match_stderr,strict-match": 0.012723076049815894,
+      "exact_match,flexible-extract": 0.6095526914329037,
+      "exact_match_stderr,flexible-extract": 0.01343782986466858
+    }
+  },
+  "group_subtasks": {
+    "gsm8k": []
+  },
+  "configs": {
+    "gsm8k": {
+      "task": "gsm8k",
+      "tag": [
+        "math_word_problems"
+      ],
+      "dataset_path": "gsm8k",
+      "dataset_name": "main",
+      "training_split": "train",
+      "test_split": "test",
+      "fewshot_split": "train",
+      "doc_to_text": "Question: {{question}}\nAnswer:",
+      "doc_to_target": "{{answer}}",
+      "unsafe_code": false,
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "num_fewshot": 5,
+      "metric_list": [
+        {
+          "metric": "exact_match",
+          "aggregation": "mean",
+          "higher_is_better": true,
+          "ignore_case": true,
+          "ignore_punctuation": false,
+          "regexes_to_ignore": [
+            ",",
+            "\\$",
+            "(?s).*#### ",
+            "\\.$"
+          ]
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [
+          "Question:",
+          "</s>",
+          "<|im_end|>"
+        ],
+        "do_sample": false,
+        "temperature": 0.0
+      },
+      "repeats": 1,
+      "filter_list": [
+        {
+          "name": "strict-match",
+          "filter": [
+            {
+              "function": "regex",
+              "regex_pattern": "#### (\\-?[0-9\\.\\,]+)"
+            },
+            {
+              "function": "take_first"
+            }
+          ]
+        },
+        {
+          "name": "flexible-extract",
+          "filter": [
+            {
+              "function": "regex",
+              "group_select": -1,
+              "regex_pattern": "(-?[$0-9.,]{2,})|(-?[0-9]+)"
+            },
+            {
+              "function": "take_first"
+            }
+          ]
+        }
+      ],
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 3.0
+      }
+    }
+  },
+  "versions": {
+    "gsm8k": 3.0
+  },
+  "n-shot": {
+    "gsm8k": 5
+  },
+  "higher_is_better": {
+    "gsm8k": {
+      "exact_match": true
+    }
+  },
+  "n-samples": {
+    "gsm8k": {
+      "original": 1319,
+      "effective": 1319
+    }
+  },
+  "config": {
+    "model": "vllm",
+    "model_args": "pretrained=results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0,tensor_parallel_size=2,dtype=float16,gpu_memory_utilization=0.8,trust_remote_code=True",
+    "batch_size": "auto",
+    "batch_sizes": [],
+    "device": null,
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": null,
+    "random_seed": 0,
+    "numpy_seed": 1234,
+    "torch_seed": 1234,
+    "fewshot_seed": 1234
+  },
+  "git_hash": "0966638",
+  "date": 1779271138.300138,
+  "pretty_env_info": "PyTorch version: 2.7.0+cu126\nIs debug build: False\nCUDA used to build PyTorch: 12.6\nROCM used to build PyTorch: N/A\n\nOS: Ubuntu 24.04.4 LTS (x86_64)\nGCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0\nClang version: Could not collect\nCMake version: version 3.28.3\nLibc version: glibc-2.39\n\nPython version: 3.10.20 (main, Mar 11 2026, 17:46:40) [GCC 14.3.0] (64-bit runtime)\nPython platform: Linux-6.17.0-1013-aws-x86_64-with-glibc2.39\nIs CUDA available: True\nCUDA runtime version: Could not collect\nCUDA_MODULE_LOADING set to: LAZY\nGPU models and configuration: \nGPU 0: NVIDIA H200\nGPU 1: NVIDIA H200\nGPU 2: NVIDIA H200\nGPU 3: NVIDIA H200\nGPU 4: NVIDIA H200\nGPU 5: NVIDIA H200\nGPU 6: NVIDIA H200\nGPU 7: NVIDIA H200\n\nNvidia driver version: 580.126.16\ncuDNN version: Probably one of the following:\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_adv.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_cnn.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_precompiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_engines_runtime_compiled.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_graph.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_heuristic.so.9\n/opt/pytorch/lib/python3.13/site-packages/nvidia/cu13/lib/libcudnn_ops.so.9\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:                           48 bits physical, 48 bits virtual\nByte Order:                              Little Endian\nCPU(s):                                  192\nOn-line CPU(s) list:                     0-191\nVendor ID:                               AuthenticAMD\nModel name:                              AMD EPYC 7R13 Processor\nCPU family:                              25\nModel:                                   1\nThread(s) per core:                      2\nCore(s) per socket:                      48\nSocket(s):                               2\nStepping:                                1\nBogoMIPS:                                5299.99\nFlags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf tsc_known_freq pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch topoext perfctr_core ssbd ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 clzero xsaveerptr rdpru wbnoinvd arat npt nrip_save vaes vpclmulqdq rdpid\nHypervisor vendor:                       KVM\nVirtualization type:                     full\nL1d cache:                               3 MiB (96 instances)\nL1i cache:                               3 MiB (96 instances)\nL2 cache:                                48 MiB (96 instances)\nL3 cache:                                384 MiB (12 instances)\nNUMA node(s):                            4\nNUMA node0 CPU(s):                       0-23,96-119\nNUMA node1 CPU(s):                       24-47,120-143\nNUMA node2 CPU(s):                       48-71,144-167\nNUMA node3 CPU(s):                       72-95,168-191\nVulnerability Gather data sampling:      Not affected\nVulnerability Ghostwrite:                Not affected\nVulnerability Indirect target selection: 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 Old microcode:             Not affected\nVulnerability Reg file data sampling:    Not affected\nVulnerability Retbleed:                  Not affected\nVulnerability Spec rstack overflow:      Mitigation; Safe RET\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; Retpolines; IBPB conditional; IBRS_FW; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected\nVulnerability Srbds:                     Not affected\nVulnerability Tsa:                       Mitigation; Clear CPU buffers\nVulnerability Tsx async abort:           Not affected\nVulnerability Vmscape:                   Not affected\n\nVersions of relevant libraries:\n[pip3] numpy==2.2.6\n[pip3] nvidia-cublas==13.1.0.3\n[pip3] nvidia-cublas-cu12==12.6.4.1\n[pip3] nvidia-cuda-cupti==13.0.85\n[pip3] nvidia-cuda-cupti-cu12==12.6.80\n[pip3] nvidia-cuda-nvrtc==13.0.88\n[pip3] nvidia-cuda-nvrtc-cu12==12.6.77\n[pip3] nvidia-cuda-runtime==13.0.96\n[pip3] nvidia-cuda-runtime-cu12==12.6.77\n[pip3] nvidia-cudnn-cu12==9.5.1.17\n[pip3] nvidia-cudnn-cu13==9.19.0.56\n[pip3] nvidia-cudnn-frontend==1.18.0\n[pip3] nvidia-cufft==12.0.0.61\n[pip3] nvidia-cufft-cu12==11.3.0.4\n[pip3] nvidia-curand==10.4.0.35\n[pip3] nvidia-curand-cu12==10.3.7.77\n[pip3] nvidia-cusolver==12.0.4.66\n[pip3] nvidia-cusolver-cu12==11.7.1.2\n[pip3] nvidia-cusparse==12.6.3.3\n[pip3] nvidia-cusparse-cu12==12.5.4.2\n[pip3] nvidia-cusparselt-cu12==0.6.3\n[pip3] nvidia-cusparselt-cu13==0.8.0\n[pip3] nvidia-nccl-cu12==2.26.2\n[pip3] nvidia-nccl-cu13==2.28.9\n[pip3] nvidia-nvjitlink==13.0.88\n[pip3] nvidia-nvjitlink-cu12==12.6.85\n[pip3] nvidia-nvtx==13.0.85\n[pip3] nvidia-nvtx-cu12==12.6.77\n[pip3] tokenspeed-triton==3.7.10.post20260505\n[pip3] torch==2.7.0\n[pip3] torch_c_dlpack_ext==0.1.5\n[pip3] torchaudio==2.7.0\n[pip3] torchvision==0.22.0\n[pip3] triton==3.3.0\n[conda] numpy                                       2.2.6                pypi_0           pypi\n[conda] nvidia-cublas                               13.1.0.3             pypi_0           pypi\n[conda] nvidia-cublas-cu12                          12.6.4.1             pypi_0           pypi\n[conda] nvidia-cuda-cupti                           13.0.85              pypi_0           pypi\n[conda] nvidia-cuda-cupti-cu12                      12.6.80              pypi_0           pypi\n[conda] nvidia-cuda-nvrtc                           13.0.88              pypi_0           pypi\n[conda] nvidia-cuda-nvrtc-cu12                      12.6.77              pypi_0           pypi\n[conda] nvidia-cuda-runtime                         13.0.96              pypi_0           pypi\n[conda] nvidia-cuda-runtime-cu12                    12.6.77              pypi_0           pypi\n[conda] nvidia-cudnn-cu12                           9.5.1.17             pypi_0           pypi\n[conda] nvidia-cudnn-cu13                           9.19.0.56            pypi_0           pypi\n[conda] nvidia-cudnn-frontend                       1.18.0               pypi_0           pypi\n[conda] nvidia-cufft                                12.0.0.61            pypi_0           pypi\n[conda] nvidia-cufft-cu12                           11.3.0.4             pypi_0           pypi\n[conda] nvidia-curand                               10.4.0.35            pypi_0           pypi\n[conda] nvidia-curand-cu12                          10.3.7.77            pypi_0           pypi\n[conda] nvidia-cusolver                             12.0.4.66            pypi_0           pypi\n[conda] nvidia-cusolver-cu12                        11.7.1.2             pypi_0           pypi\n[conda] nvidia-cusparse                             12.6.3.3             pypi_0           pypi\n[conda] nvidia-cusparse-cu12                        12.5.4.2             pypi_0           pypi\n[conda] nvidia-cusparselt-cu12                      0.6.3                pypi_0           pypi\n[conda] nvidia-cusparselt-cu13                      0.8.0                pypi_0           pypi\n[conda] nvidia-nccl-cu12                            2.26.2               pypi_0           pypi\n[conda] nvidia-nccl-cu13                            2.28.9               pypi_0           pypi\n[conda] nvidia-nvjitlink                            13.0.88              pypi_0           pypi\n[conda] nvidia-nvjitlink-cu12                       12.6.85              pypi_0           pypi\n[conda] nvidia-nvtx                                 13.0.85              pypi_0           pypi\n[conda] nvidia-nvtx-cu12                            12.6.77              pypi_0           pypi\n[conda] tokenspeed-triton                           3.7.10.post20260505  pypi_0           pypi\n[conda] torch                                       2.7.0                pypi_0           pypi\n[conda] torch-c-dlpack-ext                          0.1.5                pypi_0           pypi\n[conda] torchaudio                                  2.7.0                pypi_0           pypi\n[conda] torchvision                                 0.22.0               pypi_0           pypi\n[conda] triton                                      3.3.0                pypi_0           pypi",
+  "transformers_version": "4.53.3",
+  "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": {
+    "gsm8k": "97e5cd2f1232fe6955b69546f32e60f943926d78c43ec00d61f5b3b09fd80d1e"
+  },
+  "model_source": "vllm",
+  "model_name": "results/qwen2.5-1.5B-Instruct#sfkl_nnm_lora/nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "model_name_sanitized": "results__qwen2.5-1.5B-Instruct#sfkl_nnm_lora__nnm0.7_K128_L4_epoch2_lr1e-4_kdr1.0",
+  "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": 673492.503243923,
+  "end_time": 673594.283504519,
+  "total_evaluation_time_seconds": "101.7802605960751"
+}