Instructions to use modrill/MT11-HIGH-L45 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use modrill/MT11-HIGH-L45 with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-4B-Base") model = PeftModel.from_pretrained(base_model, "modrill/MT11-HIGH-L45") - Notebooks
- Google Colab
- Kaggle
File size: 5,686 Bytes
0a5b3cd | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 | {
"schema": "mt11-high-heavy-l45-primary-selection/1",
"status": "FINAL_FROZEN",
"date": "2026-08-16",
"selection": {
"primary": "HIGH_HEAVY_L45",
"downgraded": "MEDCONF_MED34_HIGH30_CENTERED_S2",
"downgrade_evidence": {
"fresh_correct": 29,
"fresh_n": 240,
"vs_fresh_base_delta_correct": 9,
"question_cluster_bootstrap_95ci_delta_correct": [
-3,
22
],
"significantly_above_fresh_base": false
},
"owner_override": {
"applied": true,
"prior_report_verdict": "DO_NOT_STOP_TRAINING_MIXED_OR_UNSTABLE",
"reason": "The prior report verdict combined HIGH and MEDCONF. The owner had predeclared that fresh-seed reproduction of the 34-point result ends training; HIGH exactly reproduced 34/240 and remained significantly above Fresh Base."
}
},
"high_complete_evidence": {
"no_cherry_pick": true,
"all_seeds_included": [
42,
43,
44,
45,
46,
47,
48,
49
],
"pairing_note": "Different eval seeds are not sample-paired; comparisons use question aggregation. Combined health metrics are exhaustive sample totals.",
"original_block": {
"seeds": [
42,
43,
44,
45
],
"n": 240,
"correct": 34,
"correct_by_seed": {
"42": 8,
"43": 9,
"44": 6,
"45": 11
},
"cap": {
"n": 91,
"pct": 37.917
},
"decoded_complete_box": {
"n": 150,
"pct": 62.5
},
"natural_stop": {
"n": 149,
"pct": 62.083
}
},
"fresh_block": {
"seeds": [
46,
47,
48,
49
],
"n": 240,
"correct": 34,
"correct_by_seed": {
"46": 8,
"47": 9,
"48": 9,
"49": 8
},
"cap": {
"n": 84,
"pct": 35.0
},
"decoded_complete_box": {
"n": 151,
"pct": 62.917
},
"natural_stop": {
"n": 156,
"pct": 65.0
},
"vs_fresh_base": {
"fresh_base_correct": 20,
"fresh_base_n": 240,
"delta_correct": 14,
"question_cluster_bootstrap_95ci_delta_correct": [
2,
27
],
"significantly_above": true
}
},
"combined_block": {
"seeds": [
42,
43,
44,
45,
46,
47,
48,
49
],
"n": 480,
"correct": {
"n": 68,
"pct": 14.167
},
"cap": {
"n": 175,
"pct": 36.458
},
"decoded_complete_box": {
"n": 301,
"pct": 62.708
},
"natural_stop": {
"n": 305,
"pct": 63.542
}
}
},
"frozen_identity": {
"adapter_path": "/workspace/DDC-v3.1/runs/mt11_r2_384k96_execution_20260814/arms_90r6pp/HIGH_HEAVY_L45/train/checkpoint-update-00000024-tokens-000000409131/adapter",
"adapter_sha256": "628a5b4c40584c524479b80773166f49b5c49d7223b34d4cf254064d6b65bd29",
"base_path": "/workspace/MATH-NOTHINK-AIME-P1-v1.0/model/mn8_eos_r1_2/Qwen3-4B-Base-906bfd4",
"tokenizer_path": "/workspace/MATH-NOTHINK-AIME-P1-v1.0/model/mn8_eos_r1_2/Qwen3-4B-Base-906bfd4",
"dynamic_lora": true,
"base_adapter_separate": true,
"merge": false,
"bake": false
},
"standard_think_evalscope_contract": {
"mode": "think",
"enable_thinking": true,
"framework": "EvalScope",
"score_authority": "EvalScope reviews",
"suites": [
"AIME24",
"AIME25"
],
"independent_questions": 60,
"samples_per_four_seed_block": 240,
"sampling": {
"temperature": 0.6,
"top_p": 0.95,
"top_k": 20
},
"max_model_len": 32768,
"reserve_tokens": 64,
"max_new_tokens_rule": "32768 - exact_upstream_tokenize_prompt_count - 64",
"statistics": "question-clustered; do not treat 240 samples as 240 independent questions",
"serving": "vLLM dynamic LoRA mount; no merge or bake"
},
"route_freeze": {
"stop_all_new_mt11_training": true,
"short_cot": "FAIL",
"med34_score_uplift_r1_family_gate": "FAIL",
"automatic_continuation_forbidden": true,
"automatic_arm_expansion_forbidden": true,
"automatic_warm_start_forbidden": true,
"automatic_lora_combination_forbidden": true,
"historical_result_cards_modified": false,
"training_started_by_this_action": false,
"evaluation_started_by_this_action": false
},
"mn9_crossprotocol_eval": {
"approved_eval_only": true,
"must_not_stop_or_interfere": true,
"may_not_trigger_training_automatically": true,
"may_not_trigger_lora_combination_automatically": true,
"separate_approval_required_for_any_followup": true
},
"supervisor_read_only_check": {
"mt11_training_running": false,
"action_taken": "NONE",
"running_approved_eval_only_services": [
"ddc-mt11-mn9-crossprotocol-controller",
"ddc-mt11-mn9-crossprotocol-eval-disk-guard",
"ddc-mt11-mn9-crossprotocol-eval-gpu0",
"ddc-mt11-mn9-crossprotocol-eval-gpu1",
"ddc-mt11-mn9-crossprotocol-eval-gpu2",
"ddc-mt11-mn9-crossprotocol-eval-gpu3"
]
},
"authoritative_sources": [
"/workspace/DDC-v3.1/runs/mt11_two34_freshseed_reval_20260815/MT11_TWO34_FRESHSEED_REVAL_FINAL.md",
"/workspace/DDC-v3.1/runs/mt11_two34_freshseed_reval_20260815/MT11_TWO34_FRESHSEED_REVAL_FINAL.json",
"/workspace/DDC-v3.1/runs/mt11_r2_384k96_execution_20260814/RESULT_CARD_HIGH_HEAVY_L45.json",
"/workspace/DDC-v3.1/runs/MED34_SCORE_UPLIFT_CANARY_V1R1_SHARED_PREFIX/MED34_SCORE_UPLIFT_REPORT.md"
]
}
|