Instructions to use hunterbown/trainless-math-2b-loop-repair with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use hunterbown/trainless-math-2b-loop-repair with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3.5-2B") model = PeftModel.from_pretrained(base_model, "hunterbown/trainless-math-2b-loop-repair") - Notebooks
- Google Colab
- Kaggle
File size: 2,435 Bytes
8424592 | 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 | {
"format": "trainless-peft/1",
"base_model": "Qwen/Qwen3.5-2B",
"base_revision": "15852e8c16360a2fea060d615a32b45270f8a8fc",
"auto_class": "AutoModelForImageTextToText",
"runtime": "transformers-peft",
"method": "forward-only analytic weight editing",
"quantized_runtime_tested": false,
"evaluation": {
"scope": "Standalone actual analytic PEFT adapter transport; fixed existing certificate items and protocol, no tuning",
"cap": 32768,
"offset": 64,
"n": 200,
"item_sha256": "8a3b16a62f8b0755ebbd49d4fa3b3cd05a3dba3fa17e234f6001e92b3c36f6a1",
"sampling": {
"temperature": 0.6,
"top_p": 0.95,
"top_k": 20,
"presence_penalty": 1.5,
"seed": 0,
"thinking": true
},
"grading": "repository evals.graders.for_benchmark(math500), unchanged",
"paired": {
"accuracy_delta": 0.08,
"bootstrap95": [
0.03,
0.135
],
"bootstrap_repetitions": 20000,
"seed": 20261110,
"base_wrong_adapter_right": 23,
"base_right_adapter_wrong": 7,
"note": "paired item bootstrap, same saved questions; model sampling variance from additional seeds unmeasured"
},
"base_model": "Qwen/Qwen3.5-2B",
"base_revision": "15852e8c16360a2fea060d615a32b45270f8a8fc",
"adapter_sha256": "f67ffbbc22c7a0c0e1dd9ce28459751b18ab91de6132390a164c611f5744e27b",
"arms": {
"base": {
"n": 200,
"correct": 136,
"accuracy": 0.68,
"wilson95": [
0.6124810156588646,
0.7407346668453195
],
"capped": 63,
"generation_wall_seconds": 759.5006894249236
},
"adapter": {
"n": 200,
"correct": 152,
"accuracy": 0.76,
"wilson95": [
0.6962665902079583,
0.8139338400758631
],
"capped": 48,
"generation_wall_seconds": 760.9235217790119
}
}
},
"source_weight_state_id": "64ca5748a92d0923",
"source_export_check": {
"variant": "original baked bf16 source, not the separately loaded analytic PEFT runtime",
"benchmark": "AdvBench harmful_behaviors (first 64), greedy, thinking off, 256 tokens, refusal-phrase grader",
"n_per_arm": 64,
"base_refusals": 63,
"edited_refusals": 63,
"drop_percentage_points": 0,
"override_used": false,
"scope": "Existing platform export screen; not a general safety qualification."
}
}
|