Upload folder using huggingface_hub

ifeval/results_ifeval_thinking_seed1234_2026-05-10T12-43-13.891334.json

@@ -0,0 +1,159 @@
+{
+  "results": {
+    "ifeval": {
+      "name": "ifeval",
+      "alias": "ifeval",
+      "sample_len": 541,
+      "prompt_level_strict_acc,none": 0.8687615526802218,
+      "prompt_level_strict_acc_stderr,none": 0.014530613740502803,
+      "inst_level_strict_acc,none": 0.9112709832134293,
+      "inst_level_strict_acc_stderr,none": "N/A",
+      "prompt_level_loose_acc,none": 0.9038817005545287,
+      "prompt_level_loose_acc_stderr,none": 0.01268416714715648,
+      "inst_level_loose_acc,none": 0.934052757793765,
+      "inst_level_loose_acc_stderr,none": "N/A"
+    }
+  },
+  "group_subtasks": {},
+  "configs": {
+    "ifeval": {
+      "task": "ifeval",
+      "dataset_path": "google/IFEval",
+      "test_split": "train",
+      "doc_to_text": "prompt",
+      "doc_to_target": 0,
+      "unsafe_code": false,
+      "process_results": "def process_results(doc, results):\n    inp = InputExample(\n        key=doc[\"key\"],\n        instruction_id_list=doc[\"instruction_id_list\"],\n        prompt=doc[\"prompt\"],\n        kwargs=doc[\"kwargs\"],\n    )\n    response = results[0]\n\n    out_strict = test_instruction_following_strict(inp, response)\n    out_loose = test_instruction_following_loose(inp, response)\n\n    return {\n        \"prompt_level_strict_acc\": out_strict.follow_all_instructions,\n        \"inst_level_strict_acc\": out_strict.follow_instruction_list,\n        \"prompt_level_loose_acc\": out_loose.follow_all_instructions,\n        \"inst_level_loose_acc\": out_loose.follow_instruction_list,\n    }\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "default",
+        "split": null,
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "prompt",
+        "doc_to_choice": null,
+        "doc_to_target": 0,
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "prompt_level_strict_acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "inst_level_strict_acc",
+          "aggregation": "def agg_inst_level_acc(items):\n    flat_items = [item for sublist in items for item in sublist]\n    inst_level_acc = sum(flat_items) / len(flat_items)\n    return inst_level_acc\n",
+          "higher_is_better": true
+        },
+        {
+          "metric": "prompt_level_loose_acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "inst_level_loose_acc",
+          "aggregation": "def agg_inst_level_acc(items):\n    flat_items = [item for sublist in items for item in sublist]\n    inst_level_acc = sum(flat_items) / len(flat_items)\n    return inst_level_acc\n",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [],
+        "do_sample": true,
+        "temperature": 0.6,
+        "max_gen_toks": 64000,
+        "top_p": 0.95,
+        "top_k": 20,
+        "min_p": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 4.0,
+        "model": "inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+        "max_length": 262144,
+        "base_url": "http://0.0.0.0:8000/v1/chat/completions",
+        "num_concurrent": 64,
+        "max_retries": 3,
+        "tokenized_requests": false,
+        "tokenizer_backend": null,
+        "timeout": 1200,
+        "config_source": "/home/Chibukach/environments/lmeval/lib64/python3.12/site-packages/lm_eval/tasks/ifeval/ifeval.yaml"
+      }
+    }
+  },
+  "versions": {
+    "ifeval": 4.0
+  },
+  "n-shot": {
+    "ifeval": 0
+  },
+  "higher_is_better": {
+    "ifeval": {
+      "prompt_level_strict_acc": true,
+      "inst_level_strict_acc": true,
+      "prompt_level_loose_acc": true,
+      "inst_level_loose_acc": true
+    }
+  },
+  "n-samples": {
+    "ifeval": {
+      "original": 541,
+      "effective": 541
+    }
+  },
+  "config": {
+    "model": "local-chat-completions",
+    "model_args": {
+      "model": "inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "max_length": 262144,
+      "base_url": "http://0.0.0.0:8000/v1/chat/completions",
+      "num_concurrent": 64,
+      "max_retries": 3,
+      "tokenized_requests": false,
+      "tokenizer_backend": null,
+      "timeout": 1200
+    },
+    "batch_size": 1,
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {
+      "do_sample": true,
+      "temperature": 0.6,
+      "top_p": 0.95,
+      "top_k": 20,
+      "min_p": 0.0,
+      "max_gen_toks": 64000
+    },
+    "random_seed": 1234,
+    "numpy_seed": 1234,
+    "torch_seed": 1234,
+    "fewshot_seed": 1234
+  },
+  "git_hash": null,
+  "date": 1778427631.3262064,
+  "pretty_env_info": "N/A (torch not installed)",
+  "transformers_version": "N/A",
+  "lm_eval_version": "0.4.12.dev0",
+  "upper_git_hash": null,
+  "task_hashes": {},
+  "model_source": "local-chat-completions",
+  "model_name": "inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+  "model_name_sanitized": "inference-optimization__Qwen3-235B-A22B-Thinking-2507.w8a8",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "chat_template": "",
+  "chat_template_sha": null,
+  "total_evaluation_time_seconds": "3769.5834214600036"
+}

ifeval/results_ifeval_thinking_seed4158_2026-05-13T14-19-27.010769.json

@@ -0,0 +1,159 @@
+{
+  "results": {
+    "ifeval": {
+      "name": "ifeval",
+      "alias": "ifeval",
+      "sample_len": 541,
+      "prompt_level_strict_acc,none": 0.8502772643253235,
+      "prompt_level_strict_acc_stderr,none": 0.01535420307378572,
+      "inst_level_strict_acc,none": 0.9016786570743405,
+      "inst_level_strict_acc_stderr,none": "N/A",
+      "prompt_level_loose_acc,none": 0.9038817005545287,
+      "prompt_level_loose_acc_stderr,none": 0.01268416714715648,
+      "inst_level_loose_acc,none": 0.9364508393285371,
+      "inst_level_loose_acc_stderr,none": "N/A"
+    }
+  },
+  "group_subtasks": {},
+  "configs": {
+    "ifeval": {
+      "task": "ifeval",
+      "dataset_path": "google/IFEval",
+      "test_split": "train",
+      "doc_to_text": "prompt",
+      "doc_to_target": 0,
+      "unsafe_code": false,
+      "process_results": "def process_results(doc, results):\n    inp = InputExample(\n        key=doc[\"key\"],\n        instruction_id_list=doc[\"instruction_id_list\"],\n        prompt=doc[\"prompt\"],\n        kwargs=doc[\"kwargs\"],\n    )\n    response = results[0]\n\n    out_strict = test_instruction_following_strict(inp, response)\n    out_loose = test_instruction_following_loose(inp, response)\n\n    return {\n        \"prompt_level_strict_acc\": out_strict.follow_all_instructions,\n        \"inst_level_strict_acc\": out_strict.follow_instruction_list,\n        \"prompt_level_loose_acc\": out_loose.follow_all_instructions,\n        \"inst_level_loose_acc\": out_loose.follow_instruction_list,\n    }\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "default",
+        "split": null,
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "prompt",
+        "doc_to_choice": null,
+        "doc_to_target": 0,
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "prompt_level_strict_acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "inst_level_strict_acc",
+          "aggregation": "def agg_inst_level_acc(items):\n    flat_items = [item for sublist in items for item in sublist]\n    inst_level_acc = sum(flat_items) / len(flat_items)\n    return inst_level_acc\n",
+          "higher_is_better": true
+        },
+        {
+          "metric": "prompt_level_loose_acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "inst_level_loose_acc",
+          "aggregation": "def agg_inst_level_acc(items):\n    flat_items = [item for sublist in items for item in sublist]\n    inst_level_acc = sum(flat_items) / len(flat_items)\n    return inst_level_acc\n",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [],
+        "do_sample": true,
+        "temperature": 0.6,
+        "max_gen_toks": 64000,
+        "top_p": 0.95,
+        "top_k": 20,
+        "min_p": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 4.0,
+        "model": "inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+        "max_length": 262144,
+        "base_url": "http://0.0.0.0:8000/v1/chat/completions",
+        "num_concurrent": 64,
+        "max_retries": 3,
+        "tokenized_requests": false,
+        "tokenizer_backend": null,
+        "timeout": 1200,
+        "config_source": "/home/Chibukach/environments/lmeval/lib64/python3.12/site-packages/lm_eval/tasks/ifeval/ifeval.yaml"
+      }
+    }
+  },
+  "versions": {
+    "ifeval": 4.0
+  },
+  "n-shot": {
+    "ifeval": 0
+  },
+  "higher_is_better": {
+    "ifeval": {
+      "prompt_level_strict_acc": true,
+      "inst_level_strict_acc": true,
+      "prompt_level_loose_acc": true,
+      "inst_level_loose_acc": true
+    }
+  },
+  "n-samples": {
+    "ifeval": {
+      "original": 541,
+      "effective": 541
+    }
+  },
+  "config": {
+    "model": "local-chat-completions",
+    "model_args": {
+      "model": "inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "max_length": 262144,
+      "base_url": "http://0.0.0.0:8000/v1/chat/completions",
+      "num_concurrent": 64,
+      "max_retries": 3,
+      "tokenized_requests": false,
+      "tokenizer_backend": null,
+      "timeout": 1200
+    },
+    "batch_size": 1,
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {
+      "do_sample": true,
+      "temperature": 0.6,
+      "top_p": 0.95,
+      "top_k": 20,
+      "min_p": 0.0,
+      "max_gen_toks": 64000
+    },
+    "random_seed": 4158,
+    "numpy_seed": 4158,
+    "torch_seed": 4158,
+    "fewshot_seed": 4158
+  },
+  "git_hash": null,
+  "date": 1778695427.4129255,
+  "pretty_env_info": "N/A (torch not installed)",
+  "transformers_version": "N/A",
+  "lm_eval_version": "0.4.12.dev0",
+  "upper_git_hash": null,
+  "task_hashes": {},
+  "model_source": "local-chat-completions",
+  "model_name": "inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+  "model_name_sanitized": "inference-optimization__Qwen3-235B-A22B-Thinking-2507.w8a8",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "chat_template": "",
+  "chat_template_sha": null,
+  "total_evaluation_time_seconds": "946.7284039759543"
+}

ifeval/results_ifeval_thinking_seed42_2026-05-10T11-40-22.746423.json

@@ -0,0 +1,159 @@
+{
+  "results": {
+    "ifeval": {
+      "name": "ifeval",
+      "alias": "ifeval",
+      "sample_len": 541,
+      "prompt_level_strict_acc,none": 0.844731977818854,
+      "prompt_level_strict_acc_stderr,none": 0.01558488485853813,
+      "inst_level_strict_acc,none": 0.89568345323741,
+      "inst_level_strict_acc_stderr,none": "N/A",
+      "prompt_level_loose_acc,none": 0.8817005545286506,
+      "prompt_level_loose_acc_stderr,none": 0.013898087176706528,
+      "inst_level_loose_acc,none": 0.919664268585132,
+      "inst_level_loose_acc_stderr,none": "N/A"
+    }
+  },
+  "group_subtasks": {},
+  "configs": {
+    "ifeval": {
+      "task": "ifeval",
+      "dataset_path": "google/IFEval",
+      "test_split": "train",
+      "doc_to_text": "prompt",
+      "doc_to_target": 0,
+      "unsafe_code": false,
+      "process_results": "def process_results(doc, results):\n    inp = InputExample(\n        key=doc[\"key\"],\n        instruction_id_list=doc[\"instruction_id_list\"],\n        prompt=doc[\"prompt\"],\n        kwargs=doc[\"kwargs\"],\n    )\n    response = results[0]\n\n    out_strict = test_instruction_following_strict(inp, response)\n    out_loose = test_instruction_following_loose(inp, response)\n\n    return {\n        \"prompt_level_strict_acc\": out_strict.follow_all_instructions,\n        \"inst_level_strict_acc\": out_strict.follow_instruction_list,\n        \"prompt_level_loose_acc\": out_loose.follow_all_instructions,\n        \"inst_level_loose_acc\": out_loose.follow_instruction_list,\n    }\n",
+      "description": "",
+      "target_delimiter": " ",
+      "fewshot_delimiter": "\n\n",
+      "fewshot_config": {
+        "sampler": "default",
+        "split": null,
+        "process_docs": null,
+        "fewshot_indices": null,
+        "samples": null,
+        "doc_to_text": "prompt",
+        "doc_to_choice": null,
+        "doc_to_target": 0,
+        "gen_prefix": null,
+        "fewshot_delimiter": "\n\n",
+        "target_delimiter": " "
+      },
+      "num_fewshot": 0,
+      "metric_list": [
+        {
+          "metric": "prompt_level_strict_acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "inst_level_strict_acc",
+          "aggregation": "def agg_inst_level_acc(items):\n    flat_items = [item for sublist in items for item in sublist]\n    inst_level_acc = sum(flat_items) / len(flat_items)\n    return inst_level_acc\n",
+          "higher_is_better": true
+        },
+        {
+          "metric": "prompt_level_loose_acc",
+          "aggregation": "mean",
+          "higher_is_better": true
+        },
+        {
+          "metric": "inst_level_loose_acc",
+          "aggregation": "def agg_inst_level_acc(items):\n    flat_items = [item for sublist in items for item in sublist]\n    inst_level_acc = sum(flat_items) / len(flat_items)\n    return inst_level_acc\n",
+          "higher_is_better": true
+        }
+      ],
+      "output_type": "generate_until",
+      "generation_kwargs": {
+        "until": [],
+        "do_sample": true,
+        "temperature": 0.6,
+        "max_gen_toks": 64000,
+        "top_p": 0.95,
+        "top_k": 20,
+        "min_p": 0.0
+      },
+      "repeats": 1,
+      "should_decontaminate": false,
+      "metadata": {
+        "version": 4.0,
+        "model": "inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+        "max_length": 262144,
+        "base_url": "http://0.0.0.0:8000/v1/chat/completions",
+        "num_concurrent": 64,
+        "max_retries": 3,
+        "tokenized_requests": false,
+        "tokenizer_backend": null,
+        "timeout": 1200,
+        "config_source": "/home/Chibukach/environments/lmeval/lib64/python3.12/site-packages/lm_eval/tasks/ifeval/ifeval.yaml"
+      }
+    }
+  },
+  "versions": {
+    "ifeval": 4.0
+  },
+  "n-shot": {
+    "ifeval": 0
+  },
+  "higher_is_better": {
+    "ifeval": {
+      "prompt_level_strict_acc": true,
+      "inst_level_strict_acc": true,
+      "prompt_level_loose_acc": true,
+      "inst_level_loose_acc": true
+    }
+  },
+  "n-samples": {
+    "ifeval": {
+      "original": 541,
+      "effective": 541
+    }
+  },
+  "config": {
+    "model": "local-chat-completions",
+    "model_args": {
+      "model": "inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "max_length": 262144,
+      "base_url": "http://0.0.0.0:8000/v1/chat/completions",
+      "num_concurrent": 64,
+      "max_retries": 3,
+      "tokenized_requests": false,
+      "tokenizer_backend": null,
+      "timeout": 1200
+    },
+    "batch_size": 1,
+    "batch_sizes": [],
+    "device": "cuda:0",
+    "use_cache": null,
+    "limit": null,
+    "bootstrap_iters": 100000,
+    "gen_kwargs": {
+      "do_sample": true,
+      "temperature": 0.6,
+      "top_p": 0.95,
+      "top_k": 20,
+      "min_p": 0.0,
+      "max_gen_toks": 64000
+    },
+    "random_seed": 42,
+    "numpy_seed": 42,
+    "torch_seed": 42,
+    "fewshot_seed": 42
+  },
+  "git_hash": null,
+  "date": 1778423974.1340694,
+  "pretty_env_info": "N/A (torch not installed)",
+  "transformers_version": "N/A",
+  "lm_eval_version": "0.4.12.dev0",
+  "upper_git_hash": null,
+  "task_hashes": {},
+  "model_source": "local-chat-completions",
+  "model_name": "inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+  "model_name_sanitized": "inference-optimization__Qwen3-235B-A22B-Thinking-2507.w8a8",
+  "system_instruction": null,
+  "system_instruction_sha": null,
+  "fewshot_as_multiturn": true,
+  "chat_template": "",
+  "chat_template_sha": null,
+  "total_evaluation_time_seconds": "3655.6129989860347"
+}

lighteval_thinking/results_2026-05-15T04-28-37.363609.json

@@ -0,0 +1,263 @@
+{
+  "config_general": {
+    "lighteval_sha": "6f0f351abfd7c06004bf0dcdaeab14c8f289b86c",
+    "num_fewshot_seeds": 1,
+    "max_samples": null,
+    "job_id": "0",
+    "start_time": 7796486.065836329,
+    "end_time": 7799696.230554569,
+    "total_evaluation_time_secondes": "3210.164718239568",
+    "model_config": {
+      "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "generation_parameters": {
+        "num_blocks": null,
+        "block_size": null,
+        "early_stopping": null,
+        "repetition_penalty": null,
+        "frequency_penalty": null,
+        "length_penalty": null,
+        "presence_penalty": null,
+        "max_new_tokens": 32000,
+        "min_new_tokens": null,
+        "seed": 4158,
+        "stop_tokens": null,
+        "temperature": 0.6,
+        "top_k": 20,
+        "min_p": 0.0,
+        "top_p": 0.95,
+        "truncate_prompt": null,
+        "cache_implementation": null,
+        "response_format": null
+      },
+      "system_prompt": null,
+      "cache_dir": "~/.cache/huggingface/lighteval",
+      "provider": "hosted_vllm",
+      "base_url": "http://0.0.0.0:8000/v1",
+      "api_key": "",
+      "concurrent_requests": 64,
+      "verbose": false,
+      "max_model_length": null,
+      "api_max_retry": 8,
+      "api_retry_sleep": 1.0,
+      "api_retry_multiplier": 2.0,
+      "timeout": 1200.0
+    },
+    "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8"
+  },
+  "results": {
+    "gpqa:diamond|0": {
+      "gpqa_pass@k:k=1": 0.8080808080808081,
+      "gpqa_pass@k:k=1_stderr": 0.02805779167298901
+    },
+    "math_500|0": {
+      "pass@k:k=1&n=1": 0.9,
+      "pass@k:k=1&n=1_stderr": 0.013429844431075347
+    },
+    "aime25|0": {
+      "pass@k:k=1&n=1": 0.8666666666666667,
+      "pass@k:k=1&n=1_stderr": 0.06312427686319994,
+      "avg@n:n=1": 0.8666666666666667,
+      "avg@n:n=1_stderr": 0.06312427686319994
+    },
+    "all": {
+      "gpqa_pass@k:k=1": 0.8080808080808081,
+      "gpqa_pass@k:k=1_stderr": 0.02805779167298901,
+      "pass@k:k=1&n=1": 0.8833333333333333,
+      "pass@k:k=1&n=1_stderr": 0.038277060647137644,
+      "avg@n:n=1": 0.8666666666666667,
+      "avg@n:n=1_stderr": 0.06312427686319994
+    }
+  },
+  "versions": {},
+  "config_tasks": {
+    "gpqa:diamond|0": {
+      "name": "gpqa:diamond",
+      "prompt_function": "gpqa_instruct_prompt",
+      "hf_repo": "Idavidrein/gpqa",
+      "hf_subset": "gpqa_diamond",
+      "metrics": [
+        {
+          "metric_name": "gpqa_pass@k:k=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": "sample_to_fewshot",
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 1
+    },
+    "math_500|0": {
+      "name": "math_500",
+      "prompt_function": "math_500_prompt",
+      "hf_repo": "HuggingFaceH4/MATH-500",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "test"
+      ],
+      "evaluation_splits": [
+        "test"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "aime25|0": {
+      "name": "aime25",
+      "prompt_function": "aime_prompt",
+      "hf_repo": "yentinglin/aime_2025",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        },
+        {
+          "metric_name": "avg@n:n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "AvgAtN(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, n=1, attribute_must_be_set=['n'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": null,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    }
+  },
+  "summary_tasks": {
+    "gpqa:diamond|0": {
+      "hashes": {
+        "hash_examples": "7e6ab78504360639",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "42cba4b3cd185311",
+        "hash_cont_tokens": "42cba4b3cd185311"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "math_500|0": {
+      "hashes": {
+        "hash_examples": "9f076bfe85016b9f",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "6e2146e74832427e",
+        "hash_cont_tokens": "6e2146e74832427e"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "aime25|0": {
+      "hashes": {
+        "hash_examples": "8df1bb48b826e206",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "f04b89be03da9f73",
+        "hash_cont_tokens": "f04b89be03da9f73"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    }
+  },
+  "summary_general": {
+    "hashes": {
+      "hash_examples": "9f27d13c4db0f7b7",
+      "hash_full_prompts": "15a9db8cd5f71ff9",
+      "hash_input_tokens": "a81faa1610a8e7f6",
+      "hash_cont_tokens": "a81faa1610a8e7f6"
+    },
+    "truncated": 0,
+    "non_truncated": 0,
+    "padded": 0,
+    "non_padded": 0
+  }
+}

lighteval_thinking/results_2026-05-15T05-23-55.938792.json

@@ -0,0 +1,263 @@
+{
+  "config_general": {
+    "lighteval_sha": "6f0f351abfd7c06004bf0dcdaeab14c8f289b86c",
+    "num_fewshot_seeds": 1,
+    "max_samples": null,
+    "job_id": "0",
+    "start_time": 7799704.391019128,
+    "end_time": 7803014.806037306,
+    "total_evaluation_time_secondes": "3310.4150181775913",
+    "model_config": {
+      "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "generation_parameters": {
+        "num_blocks": null,
+        "block_size": null,
+        "early_stopping": null,
+        "repetition_penalty": null,
+        "frequency_penalty": null,
+        "length_penalty": null,
+        "presence_penalty": null,
+        "max_new_tokens": 32000,
+        "min_new_tokens": null,
+        "seed": 5322,
+        "stop_tokens": null,
+        "temperature": 0.6,
+        "top_k": 20,
+        "min_p": 0.0,
+        "top_p": 0.95,
+        "truncate_prompt": null,
+        "cache_implementation": null,
+        "response_format": null
+      },
+      "system_prompt": null,
+      "cache_dir": "~/.cache/huggingface/lighteval",
+      "provider": "hosted_vllm",
+      "base_url": "http://0.0.0.0:8000/v1",
+      "api_key": "",
+      "concurrent_requests": 64,
+      "verbose": false,
+      "max_model_length": null,
+      "api_max_retry": 8,
+      "api_retry_sleep": 1.0,
+      "api_retry_multiplier": 2.0,
+      "timeout": 1200.0
+    },
+    "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8"
+  },
+  "results": {
+    "aime25|0": {
+      "pass@k:k=1&n=1": 0.8333333333333334,
+      "pass@k:k=1&n=1_stderr": 0.06920456654478328,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478328
+    },
+    "math_500|0": {
+      "pass@k:k=1&n=1": 0.916,
+      "pass@k:k=1&n=1_stderr": 0.012417584015643692
+    },
+    "gpqa:diamond|0": {
+      "gpqa_pass@k:k=1": 0.8434343434343434,
+      "gpqa_pass@k:k=1_stderr": 0.025890520358141454
+    },
+    "all": {
+      "pass@k:k=1&n=1": 0.8746666666666667,
+      "pass@k:k=1&n=1_stderr": 0.04081107528021349,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478328,
+      "gpqa_pass@k:k=1": 0.8434343434343434,
+      "gpqa_pass@k:k=1_stderr": 0.025890520358141454
+    }
+  },
+  "versions": {},
+  "config_tasks": {
+    "aime25|0": {
+      "name": "aime25",
+      "prompt_function": "aime_prompt",
+      "hf_repo": "yentinglin/aime_2025",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        },
+        {
+          "metric_name": "avg@n:n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "AvgAtN(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, n=1, attribute_must_be_set=['n'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": null,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "math_500|0": {
+      "name": "math_500",
+      "prompt_function": "math_500_prompt",
+      "hf_repo": "HuggingFaceH4/MATH-500",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "test"
+      ],
+      "evaluation_splits": [
+        "test"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "gpqa:diamond|0": {
+      "name": "gpqa:diamond",
+      "prompt_function": "gpqa_instruct_prompt",
+      "hf_repo": "Idavidrein/gpqa",
+      "hf_subset": "gpqa_diamond",
+      "metrics": [
+        {
+          "metric_name": "gpqa_pass@k:k=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": "sample_to_fewshot",
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 1
+    }
+  },
+  "summary_tasks": {
+    "aime25|0": {
+      "hashes": {
+        "hash_examples": "8df1bb48b826e206",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "f04b89be03da9f73",
+        "hash_cont_tokens": "f04b89be03da9f73"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "math_500|0": {
+      "hashes": {
+        "hash_examples": "9f076bfe85016b9f",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "6e2146e74832427e",
+        "hash_cont_tokens": "6e2146e74832427e"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "gpqa:diamond|0": {
+      "hashes": {
+        "hash_examples": "7e6ab78504360639",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "42cba4b3cd185311",
+        "hash_cont_tokens": "42cba4b3cd185311"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    }
+  },
+  "summary_general": {
+    "hashes": {
+      "hash_examples": "9f27d13c4db0f7b7",
+      "hash_full_prompts": "15a9db8cd5f71ff9",
+      "hash_input_tokens": "a81faa1610a8e7f6",
+      "hash_cont_tokens": "a81faa1610a8e7f6"
+    },
+    "truncated": 0,
+    "non_truncated": 0,
+    "padded": 0,
+    "non_padded": 0
+  }
+}

lighteval_thinking/results_2026-05-15T06-19-07.509781.json

@@ -0,0 +1,263 @@
+{
+  "config_general": {
+    "lighteval_sha": "6f0f351abfd7c06004bf0dcdaeab14c8f289b86c",
+    "num_fewshot_seeds": 1,
+    "max_samples": null,
+    "job_id": "0",
+    "start_time": 7803022.871390088,
+    "end_time": 7806326.37657224,
+    "total_evaluation_time_secondes": "3303.5051821516827",
+    "model_config": {
+      "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "generation_parameters": {
+        "num_blocks": null,
+        "block_size": null,
+        "early_stopping": null,
+        "repetition_penalty": null,
+        "frequency_penalty": null,
+        "length_penalty": null,
+        "presence_penalty": null,
+        "max_new_tokens": 32000,
+        "min_new_tokens": null,
+        "seed": 1356,
+        "stop_tokens": null,
+        "temperature": 0.6,
+        "top_k": 20,
+        "min_p": 0.0,
+        "top_p": 0.95,
+        "truncate_prompt": null,
+        "cache_implementation": null,
+        "response_format": null
+      },
+      "system_prompt": null,
+      "cache_dir": "~/.cache/huggingface/lighteval",
+      "provider": "hosted_vllm",
+      "base_url": "http://0.0.0.0:8000/v1",
+      "api_key": "",
+      "concurrent_requests": 64,
+      "verbose": false,
+      "max_model_length": null,
+      "api_max_retry": 8,
+      "api_retry_sleep": 1.0,
+      "api_retry_multiplier": 2.0,
+      "timeout": 1200.0
+    },
+    "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8"
+  },
+  "results": {
+    "gpqa:diamond|0": {
+      "gpqa_pass@k:k=1": 0.8282828282828283,
+      "gpqa_pass@k:k=1_stderr": 0.02686971618742992
+    },
+    "math_500|0": {
+      "pass@k:k=1&n=1": 0.91,
+      "pass@k:k=1&n=1_stderr": 0.012811255071733791
+    },
+    "aime25|0": {
+      "pass@k:k=1&n=1": 0.8333333333333334,
+      "pass@k:k=1&n=1_stderr": 0.06920456654478328,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478328
+    },
+    "all": {
+      "gpqa_pass@k:k=1": 0.8282828282828283,
+      "gpqa_pass@k:k=1_stderr": 0.02686971618742992,
+      "pass@k:k=1&n=1": 0.8716666666666667,
+      "pass@k:k=1&n=1_stderr": 0.04100791080825854,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478328
+    }
+  },
+  "versions": {},
+  "config_tasks": {
+    "gpqa:diamond|0": {
+      "name": "gpqa:diamond",
+      "prompt_function": "gpqa_instruct_prompt",
+      "hf_repo": "Idavidrein/gpqa",
+      "hf_subset": "gpqa_diamond",
+      "metrics": [
+        {
+          "metric_name": "gpqa_pass@k:k=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": "sample_to_fewshot",
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 1
+    },
+    "math_500|0": {
+      "name": "math_500",
+      "prompt_function": "math_500_prompt",
+      "hf_repo": "HuggingFaceH4/MATH-500",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "test"
+      ],
+      "evaluation_splits": [
+        "test"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "aime25|0": {
+      "name": "aime25",
+      "prompt_function": "aime_prompt",
+      "hf_repo": "yentinglin/aime_2025",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        },
+        {
+          "metric_name": "avg@n:n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "AvgAtN(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, n=1, attribute_must_be_set=['n'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": null,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    }
+  },
+  "summary_tasks": {
+    "gpqa:diamond|0": {
+      "hashes": {
+        "hash_examples": "7e6ab78504360639",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "42cba4b3cd185311",
+        "hash_cont_tokens": "42cba4b3cd185311"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "math_500|0": {
+      "hashes": {
+        "hash_examples": "9f076bfe85016b9f",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "6e2146e74832427e",
+        "hash_cont_tokens": "6e2146e74832427e"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "aime25|0": {
+      "hashes": {
+        "hash_examples": "8df1bb48b826e206",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "f04b89be03da9f73",
+        "hash_cont_tokens": "f04b89be03da9f73"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    }
+  },
+  "summary_general": {
+    "hashes": {
+      "hash_examples": "9f27d13c4db0f7b7",
+      "hash_full_prompts": "15a9db8cd5f71ff9",
+      "hash_input_tokens": "a81faa1610a8e7f6",
+      "hash_cont_tokens": "a81faa1610a8e7f6"
+    },
+    "truncated": 0,
+    "non_truncated": 0,
+    "padded": 0,
+    "non_padded": 0
+  }
+}

lighteval_thinking/results_2026-05-15T07-16-39.718051.json

@@ -0,0 +1,263 @@
+{
+  "config_general": {
+    "lighteval_sha": "6f0f351abfd7c06004bf0dcdaeab14c8f289b86c",
+    "num_fewshot_seeds": 1,
+    "max_samples": null,
+    "job_id": "0",
+    "start_time": 7806334.481675618,
+    "end_time": 7809778.585685903,
+    "total_evaluation_time_secondes": "3444.1040102848783",
+    "model_config": {
+      "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "generation_parameters": {
+        "num_blocks": null,
+        "block_size": null,
+        "early_stopping": null,
+        "repetition_penalty": null,
+        "frequency_penalty": null,
+        "length_penalty": null,
+        "presence_penalty": null,
+        "max_new_tokens": 32000,
+        "min_new_tokens": null,
+        "seed": 9843,
+        "stop_tokens": null,
+        "temperature": 0.6,
+        "top_k": 20,
+        "min_p": 0.0,
+        "top_p": 0.95,
+        "truncate_prompt": null,
+        "cache_implementation": null,
+        "response_format": null
+      },
+      "system_prompt": null,
+      "cache_dir": "~/.cache/huggingface/lighteval",
+      "provider": "hosted_vllm",
+      "base_url": "http://0.0.0.0:8000/v1",
+      "api_key": "",
+      "concurrent_requests": 64,
+      "verbose": false,
+      "max_model_length": null,
+      "api_max_retry": 8,
+      "api_retry_sleep": 1.0,
+      "api_retry_multiplier": 2.0,
+      "timeout": 1200.0
+    },
+    "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8"
+  },
+  "results": {
+    "math_500|0": {
+      "pass@k:k=1&n=1": 0.908,
+      "pass@k:k=1&n=1_stderr": 0.012938578501027533
+    },
+    "aime25|0": {
+      "pass@k:k=1&n=1": 0.8666666666666667,
+      "pass@k:k=1&n=1_stderr": 0.06312427686319994,
+      "avg@n:n=1": 0.8666666666666667,
+      "avg@n:n=1_stderr": 0.06312427686319994
+    },
+    "gpqa:diamond|0": {
+      "gpqa_pass@k:k=1": 0.8383838383838383,
+      "gpqa_pass@k:k=1_stderr": 0.02622591986362928
+    },
+    "all": {
+      "pass@k:k=1&n=1": 0.8873333333333333,
+      "pass@k:k=1&n=1_stderr": 0.03803142768211373,
+      "avg@n:n=1": 0.8666666666666667,
+      "avg@n:n=1_stderr": 0.06312427686319994,
+      "gpqa_pass@k:k=1": 0.8383838383838383,
+      "gpqa_pass@k:k=1_stderr": 0.02622591986362928
+    }
+  },
+  "versions": {},
+  "config_tasks": {
+    "math_500|0": {
+      "name": "math_500",
+      "prompt_function": "math_500_prompt",
+      "hf_repo": "HuggingFaceH4/MATH-500",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "test"
+      ],
+      "evaluation_splits": [
+        "test"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "aime25|0": {
+      "name": "aime25",
+      "prompt_function": "aime_prompt",
+      "hf_repo": "yentinglin/aime_2025",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        },
+        {
+          "metric_name": "avg@n:n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "AvgAtN(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, n=1, attribute_must_be_set=['n'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": null,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "gpqa:diamond|0": {
+      "name": "gpqa:diamond",
+      "prompt_function": "gpqa_instruct_prompt",
+      "hf_repo": "Idavidrein/gpqa",
+      "hf_subset": "gpqa_diamond",
+      "metrics": [
+        {
+          "metric_name": "gpqa_pass@k:k=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": "sample_to_fewshot",
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 1
+    }
+  },
+  "summary_tasks": {
+    "math_500|0": {
+      "hashes": {
+        "hash_examples": "9f076bfe85016b9f",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "6e2146e74832427e",
+        "hash_cont_tokens": "6e2146e74832427e"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "aime25|0": {
+      "hashes": {
+        "hash_examples": "8df1bb48b826e206",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "f04b89be03da9f73",
+        "hash_cont_tokens": "f04b89be03da9f73"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "gpqa:diamond|0": {
+      "hashes": {
+        "hash_examples": "7e6ab78504360639",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "42cba4b3cd185311",
+        "hash_cont_tokens": "42cba4b3cd185311"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    }
+  },
+  "summary_general": {
+    "hashes": {
+      "hash_examples": "9f27d13c4db0f7b7",
+      "hash_full_prompts": "15a9db8cd5f71ff9",
+      "hash_input_tokens": "a81faa1610a8e7f6",
+      "hash_cont_tokens": "a81faa1610a8e7f6"
+    },
+    "truncated": 0,
+    "non_truncated": 0,
+    "padded": 0,
+    "non_padded": 0
+  }
+}

lighteval_thinking/results_2026-05-15T08-16-35.990299.json

@@ -0,0 +1,263 @@
+{
+  "config_general": {
+    "lighteval_sha": "6f0f351abfd7c06004bf0dcdaeab14c8f289b86c",
+    "num_fewshot_seeds": 1,
+    "max_samples": null,
+    "job_id": "0",
+    "start_time": 7809786.699125677,
+    "end_time": 7813374.857749364,
+    "total_evaluation_time_secondes": "3588.1586236869916",
+    "model_config": {
+      "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "generation_parameters": {
+        "num_blocks": null,
+        "block_size": null,
+        "early_stopping": null,
+        "repetition_penalty": null,
+        "frequency_penalty": null,
+        "length_penalty": null,
+        "presence_penalty": null,
+        "max_new_tokens": 32000,
+        "min_new_tokens": null,
+        "seed": 3344,
+        "stop_tokens": null,
+        "temperature": 0.6,
+        "top_k": 20,
+        "min_p": 0.0,
+        "top_p": 0.95,
+        "truncate_prompt": null,
+        "cache_implementation": null,
+        "response_format": null
+      },
+      "system_prompt": null,
+      "cache_dir": "~/.cache/huggingface/lighteval",
+      "provider": "hosted_vllm",
+      "base_url": "http://0.0.0.0:8000/v1",
+      "api_key": "",
+      "concurrent_requests": 64,
+      "verbose": false,
+      "max_model_length": null,
+      "api_max_retry": 8,
+      "api_retry_sleep": 1.0,
+      "api_retry_multiplier": 2.0,
+      "timeout": 1200.0
+    },
+    "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8"
+  },
+  "results": {
+    "aime25|0": {
+      "pass@k:k=1&n=1": 0.8333333333333334,
+      "pass@k:k=1&n=1_stderr": 0.06920456654478331,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478331
+    },
+    "math_500|0": {
+      "pass@k:k=1&n=1": 0.902,
+      "pass@k:k=1&n=1_stderr": 0.013309631572104932
+    },
+    "gpqa:diamond|0": {
+      "gpqa_pass@k:k=1": 0.8333333333333334,
+      "gpqa_pass@k:k=1_stderr": 0.026552207828215293
+    },
+    "all": {
+      "pass@k:k=1&n=1": 0.8676666666666667,
+      "pass@k:k=1&n=1_stderr": 0.04125709905844412,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478331,
+      "gpqa_pass@k:k=1": 0.8333333333333334,
+      "gpqa_pass@k:k=1_stderr": 0.026552207828215293
+    }
+  },
+  "versions": {},
+  "config_tasks": {
+    "aime25|0": {
+      "name": "aime25",
+      "prompt_function": "aime_prompt",
+      "hf_repo": "yentinglin/aime_2025",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        },
+        {
+          "metric_name": "avg@n:n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "AvgAtN(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, n=1, attribute_must_be_set=['n'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": null,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "math_500|0": {
+      "name": "math_500",
+      "prompt_function": "math_500_prompt",
+      "hf_repo": "HuggingFaceH4/MATH-500",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "test"
+      ],
+      "evaluation_splits": [
+        "test"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "gpqa:diamond|0": {
+      "name": "gpqa:diamond",
+      "prompt_function": "gpqa_instruct_prompt",
+      "hf_repo": "Idavidrein/gpqa",
+      "hf_subset": "gpqa_diamond",
+      "metrics": [
+        {
+          "metric_name": "gpqa_pass@k:k=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": "sample_to_fewshot",
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 1
+    }
+  },
+  "summary_tasks": {
+    "aime25|0": {
+      "hashes": {
+        "hash_examples": "8df1bb48b826e206",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "f04b89be03da9f73",
+        "hash_cont_tokens": "f04b89be03da9f73"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "math_500|0": {
+      "hashes": {
+        "hash_examples": "9f076bfe85016b9f",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "6e2146e74832427e",
+        "hash_cont_tokens": "6e2146e74832427e"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "gpqa:diamond|0": {
+      "hashes": {
+        "hash_examples": "7e6ab78504360639",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "42cba4b3cd185311",
+        "hash_cont_tokens": "42cba4b3cd185311"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    }
+  },
+  "summary_general": {
+    "hashes": {
+      "hash_examples": "9f27d13c4db0f7b7",
+      "hash_full_prompts": "15a9db8cd5f71ff9",
+      "hash_input_tokens": "a81faa1610a8e7f6",
+      "hash_cont_tokens": "a81faa1610a8e7f6"
+    },
+    "truncated": 0,
+    "non_truncated": 0,
+    "padded": 0,
+    "non_padded": 0
+  }
+}

lighteval_thinking/results_2026-05-15T09-09-55.466784.json

@@ -0,0 +1,263 @@
+{
+  "config_general": {
+    "lighteval_sha": "6f0f351abfd7c06004bf0dcdaeab14c8f289b86c",
+    "num_fewshot_seeds": 1,
+    "max_samples": null,
+    "job_id": "0",
+    "start_time": 7813382.9037439,
+    "end_time": 7816574.333411877,
+    "total_evaluation_time_secondes": "3191.429667976685",
+    "model_config": {
+      "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "generation_parameters": {
+        "num_blocks": null,
+        "block_size": null,
+        "early_stopping": null,
+        "repetition_penalty": null,
+        "frequency_penalty": null,
+        "length_penalty": null,
+        "presence_penalty": null,
+        "max_new_tokens": 32000,
+        "min_new_tokens": null,
+        "seed": 5678,
+        "stop_tokens": null,
+        "temperature": 0.6,
+        "top_k": 20,
+        "min_p": 0.0,
+        "top_p": 0.95,
+        "truncate_prompt": null,
+        "cache_implementation": null,
+        "response_format": null
+      },
+      "system_prompt": null,
+      "cache_dir": "~/.cache/huggingface/lighteval",
+      "provider": "hosted_vllm",
+      "base_url": "http://0.0.0.0:8000/v1",
+      "api_key": "",
+      "concurrent_requests": 64,
+      "verbose": false,
+      "max_model_length": null,
+      "api_max_retry": 8,
+      "api_retry_sleep": 1.0,
+      "api_retry_multiplier": 2.0,
+      "timeout": 1200.0
+    },
+    "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8"
+  },
+  "results": {
+    "gpqa:diamond|0": {
+      "gpqa_pass@k:k=1": 0.7878787878787878,
+      "gpqa_pass@k:k=1_stderr": 0.0291265228345868
+    },
+    "aime25|0": {
+      "pass@k:k=1&n=1": 0.8333333333333334,
+      "pass@k:k=1&n=1_stderr": 0.06920456654478331,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478331
+    },
+    "math_500|0": {
+      "pass@k:k=1&n=1": 0.9,
+      "pass@k:k=1&n=1_stderr": 0.013429844431075352
+    },
+    "all": {
+      "gpqa_pass@k:k=1": 0.7878787878787878,
+      "gpqa_pass@k:k=1_stderr": 0.0291265228345868,
+      "pass@k:k=1&n=1": 0.8666666666666667,
+      "pass@k:k=1&n=1_stderr": 0.04131720548792933,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478331
+    }
+  },
+  "versions": {},
+  "config_tasks": {
+    "gpqa:diamond|0": {
+      "name": "gpqa:diamond",
+      "prompt_function": "gpqa_instruct_prompt",
+      "hf_repo": "Idavidrein/gpqa",
+      "hf_subset": "gpqa_diamond",
+      "metrics": [
+        {
+          "metric_name": "gpqa_pass@k:k=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": "sample_to_fewshot",
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 1
+    },
+    "aime25|0": {
+      "name": "aime25",
+      "prompt_function": "aime_prompt",
+      "hf_repo": "yentinglin/aime_2025",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        },
+        {
+          "metric_name": "avg@n:n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "AvgAtN(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, n=1, attribute_must_be_set=['n'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": null,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "math_500|0": {
+      "name": "math_500",
+      "prompt_function": "math_500_prompt",
+      "hf_repo": "HuggingFaceH4/MATH-500",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "test"
+      ],
+      "evaluation_splits": [
+        "test"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    }
+  },
+  "summary_tasks": {
+    "gpqa:diamond|0": {
+      "hashes": {
+        "hash_examples": "7e6ab78504360639",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "42cba4b3cd185311",
+        "hash_cont_tokens": "42cba4b3cd185311"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "aime25|0": {
+      "hashes": {
+        "hash_examples": "8df1bb48b826e206",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "f04b89be03da9f73",
+        "hash_cont_tokens": "f04b89be03da9f73"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "math_500|0": {
+      "hashes": {
+        "hash_examples": "9f076bfe85016b9f",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "6e2146e74832427e",
+        "hash_cont_tokens": "6e2146e74832427e"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    }
+  },
+  "summary_general": {
+    "hashes": {
+      "hash_examples": "9f27d13c4db0f7b7",
+      "hash_full_prompts": "15a9db8cd5f71ff9",
+      "hash_input_tokens": "a81faa1610a8e7f6",
+      "hash_cont_tokens": "a81faa1610a8e7f6"
+    },
+    "truncated": 0,
+    "non_truncated": 0,
+    "padded": 0,
+    "non_padded": 0
+  }
+}

lighteval_thinking/results_2026-05-15T10-56-12.189451.json

@@ -0,0 +1,263 @@
+{
+  "config_general": {
+    "lighteval_sha": "6f0f351abfd7c06004bf0dcdaeab14c8f289b86c",
+    "num_fewshot_seeds": 1,
+    "max_samples": null,
+    "job_id": "0",
+    "start_time": 7819732.474486429,
+    "end_time": 7822951.05722643,
+    "total_evaluation_time_secondes": "3218.582740000449",
+    "model_config": {
+      "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "generation_parameters": {
+        "num_blocks": null,
+        "block_size": null,
+        "early_stopping": null,
+        "repetition_penalty": null,
+        "frequency_penalty": null,
+        "length_penalty": null,
+        "presence_penalty": null,
+        "max_new_tokens": 32000,
+        "min_new_tokens": null,
+        "seed": 42,
+        "stop_tokens": null,
+        "temperature": 0.6,
+        "top_k": 20,
+        "min_p": 0.0,
+        "top_p": 0.95,
+        "truncate_prompt": null,
+        "cache_implementation": null,
+        "response_format": null
+      },
+      "system_prompt": null,
+      "cache_dir": "~/.cache/huggingface/lighteval",
+      "provider": "hosted_vllm",
+      "base_url": "http://0.0.0.0:8000/v1",
+      "api_key": "",
+      "concurrent_requests": 64,
+      "verbose": false,
+      "max_model_length": null,
+      "api_max_retry": 8,
+      "api_retry_sleep": 1.0,
+      "api_retry_multiplier": 2.0,
+      "timeout": 1200.0
+    },
+    "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8"
+  },
+  "results": {
+    "gpqa:diamond|0": {
+      "gpqa_pass@k:k=1": 0.8181818181818182,
+      "gpqa_pass@k:k=1_stderr": 0.027479603010538794
+    },
+    "math_500|0": {
+      "pass@k:k=1&n=1": 0.91,
+      "pass@k:k=1&n=1_stderr": 0.012811255071733795
+    },
+    "aime25|0": {
+      "pass@k:k=1&n=1": 0.8,
+      "pass@k:k=1&n=1_stderr": 0.07427813527082075,
+      "avg@n:n=1": 0.8,
+      "avg@n:n=1_stderr": 0.07427813527082075
+    },
+    "all": {
+      "gpqa_pass@k:k=1": 0.8181818181818182,
+      "gpqa_pass@k:k=1_stderr": 0.027479603010538794,
+      "pass@k:k=1&n=1": 0.855,
+      "pass@k:k=1&n=1_stderr": 0.04354469517127727,
+      "avg@n:n=1": 0.8,
+      "avg@n:n=1_stderr": 0.07427813527082075
+    }
+  },
+  "versions": {},
+  "config_tasks": {
+    "gpqa:diamond|0": {
+      "name": "gpqa:diamond",
+      "prompt_function": "gpqa_instruct_prompt",
+      "hf_repo": "Idavidrein/gpqa",
+      "hf_subset": "gpqa_diamond",
+      "metrics": [
+        {
+          "metric_name": "gpqa_pass@k:k=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": "sample_to_fewshot",
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 1
+    },
+    "math_500|0": {
+      "name": "math_500",
+      "prompt_function": "math_500_prompt",
+      "hf_repo": "HuggingFaceH4/MATH-500",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "test"
+      ],
+      "evaluation_splits": [
+        "test"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "aime25|0": {
+      "name": "aime25",
+      "prompt_function": "aime_prompt",
+      "hf_repo": "yentinglin/aime_2025",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        },
+        {
+          "metric_name": "avg@n:n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "AvgAtN(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, n=1, attribute_must_be_set=['n'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": null,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    }
+  },
+  "summary_tasks": {
+    "gpqa:diamond|0": {
+      "hashes": {
+        "hash_examples": "7e6ab78504360639",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "42cba4b3cd185311",
+        "hash_cont_tokens": "42cba4b3cd185311"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "math_500|0": {
+      "hashes": {
+        "hash_examples": "9f076bfe85016b9f",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "6e2146e74832427e",
+        "hash_cont_tokens": "6e2146e74832427e"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "aime25|0": {
+      "hashes": {
+        "hash_examples": "8df1bb48b826e206",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "f04b89be03da9f73",
+        "hash_cont_tokens": "f04b89be03da9f73"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    }
+  },
+  "summary_general": {
+    "hashes": {
+      "hash_examples": "9f27d13c4db0f7b7",
+      "hash_full_prompts": "15a9db8cd5f71ff9",
+      "hash_input_tokens": "a81faa1610a8e7f6",
+      "hash_cont_tokens": "a81faa1610a8e7f6"
+    },
+    "truncated": 0,
+    "non_truncated": 0,
+    "padded": 0,
+    "non_padded": 0
+  }
+}

lighteval_thinking/results_2026-05-15T11-54-13.927603.json

@@ -0,0 +1,263 @@
+{
+  "config_general": {
+    "lighteval_sha": "6f0f351abfd7c06004bf0dcdaeab14c8f289b86c",
+    "num_fewshot_seeds": 1,
+    "max_samples": null,
+    "job_id": "0",
+    "start_time": 7822959.392426659,
+    "end_time": 7826432.79482352,
+    "total_evaluation_time_secondes": "3473.4023968605325",
+    "model_config": {
+      "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8",
+      "generation_parameters": {
+        "num_blocks": null,
+        "block_size": null,
+        "early_stopping": null,
+        "repetition_penalty": null,
+        "frequency_penalty": null,
+        "length_penalty": null,
+        "presence_penalty": null,
+        "max_new_tokens": 32000,
+        "min_new_tokens": null,
+        "seed": 1234,
+        "stop_tokens": null,
+        "temperature": 0.6,
+        "top_k": 20,
+        "min_p": 0.0,
+        "top_p": 0.95,
+        "truncate_prompt": null,
+        "cache_implementation": null,
+        "response_format": null
+      },
+      "system_prompt": null,
+      "cache_dir": "~/.cache/huggingface/lighteval",
+      "provider": "hosted_vllm",
+      "base_url": "http://0.0.0.0:8000/v1",
+      "api_key": "",
+      "concurrent_requests": 64,
+      "verbose": false,
+      "max_model_length": null,
+      "api_max_retry": 8,
+      "api_retry_sleep": 1.0,
+      "api_retry_multiplier": 2.0,
+      "timeout": 1200.0
+    },
+    "model_name": "hosted_vllm/inference-optimization/Qwen3-235B-A22B-Thinking-2507.w8a8"
+  },
+  "results": {
+    "aime25|0": {
+      "pass@k:k=1&n=1": 0.8333333333333334,
+      "pass@k:k=1&n=1_stderr": 0.06920456654478328,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478328
+    },
+    "math_500|0": {
+      "pass@k:k=1&n=1": 0.91,
+      "pass@k:k=1&n=1_stderr": 0.012811255071733797
+    },
+    "gpqa:diamond|0": {
+      "gpqa_pass@k:k=1": 0.803030303030303,
+      "gpqa_pass@k:k=1_stderr": 0.028335609732463362
+    },
+    "all": {
+      "pass@k:k=1&n=1": 0.8716666666666667,
+      "pass@k:k=1&n=1_stderr": 0.04100791080825854,
+      "avg@n:n=1": 0.8333333333333334,
+      "avg@n:n=1_stderr": 0.06920456654478328,
+      "gpqa_pass@k:k=1": 0.803030303030303,
+      "gpqa_pass@k:k=1_stderr": 0.028335609732463362
+    }
+  },
+  "versions": {},
+  "config_tasks": {
+    "aime25|0": {
+      "name": "aime25",
+      "prompt_function": "aime_prompt",
+      "hf_repo": "yentinglin/aime_2025",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        },
+        {
+          "metric_name": "avg@n:n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "AvgAtN(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, n=1, attribute_must_be_set=['n'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": null,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "math_500|0": {
+      "name": "math_500",
+      "prompt_function": "math_500_prompt",
+      "hf_repo": "HuggingFaceH4/MATH-500",
+      "hf_subset": "default",
+      "metrics": [
+        {
+          "metric_name": "pass@k:k=1&n=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=True, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve",
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": null,
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "test"
+      ],
+      "evaluation_splits": [
+        "test"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 2
+    },
+    "gpqa:diamond|0": {
+      "name": "gpqa:diamond",
+      "prompt_function": "gpqa_instruct_prompt",
+      "hf_repo": "Idavidrein/gpqa",
+      "hf_subset": "gpqa_diamond",
+      "metrics": [
+        {
+          "metric_name": "gpqa_pass@k:k=1",
+          "higher_is_better": true,
+          "category": "GENERATIVE",
+          "sample_level_fn": "PassAtK(normalize=None, strip_strings=False, compute_score=compute, type_exact_match=None, k=1, n=1, attribute_must_be_set=['k'])",
+          "corpus_level_fn": "mean",
+          "batched_compute": false
+        }
+      ],
+      "hf_data_files": null,
+      "solver": [
+        "solve"
+      ],
+      "scorer": "score",
+      "sample_fields": "record_to_sample",
+      "sample_to_fewshot": "sample_to_fewshot",
+      "filter": null,
+      "hf_revision": null,
+      "hf_filter": null,
+      "hf_avail_splits": [
+        "train"
+      ],
+      "evaluation_splits": [
+        "train"
+      ],
+      "few_shots_split": null,
+      "few_shots_select": null,
+      "generation_size": 32768,
+      "generation_grammar": null,
+      "stop_sequence": [],
+      "num_samples": null,
+      "original_num_docs": -1,
+      "effective_num_docs": -1,
+      "must_remove_duplicate_docs": false,
+      "num_fewshots": 0,
+      "version": 1
+    }
+  },
+  "summary_tasks": {
+    "aime25|0": {
+      "hashes": {
+        "hash_examples": "8df1bb48b826e206",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "f04b89be03da9f73",
+        "hash_cont_tokens": "f04b89be03da9f73"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "math_500|0": {
+      "hashes": {
+        "hash_examples": "9f076bfe85016b9f",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "6e2146e74832427e",
+        "hash_cont_tokens": "6e2146e74832427e"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    },
+    "gpqa:diamond|0": {
+      "hashes": {
+        "hash_examples": "7e6ab78504360639",
+        "hash_full_prompts": "ef46db3751d8e999",
+        "hash_input_tokens": "42cba4b3cd185311",
+        "hash_cont_tokens": "42cba4b3cd185311"
+      },
+      "truncated": 0,
+      "non_truncated": 0,
+      "padded": 0,
+      "non_padded": 0
+    }
+  },
+  "summary_general": {
+    "hashes": {
+      "hash_examples": "9f27d13c4db0f7b7",
+      "hash_full_prompts": "15a9db8cd5f71ff9",
+      "hash_input_tokens": "a81faa1610a8e7f6",
+      "hash_cont_tokens": "a81faa1610a8e7f6"
+    },
+    "truncated": 0,
+    "non_truncated": 0,
+    "padded": 0,
+    "non_padded": 0
+  }
+}