File size: 6,292 Bytes
2eb5700
a42db98
19aff2a
 
 
 
 
865382a
2eb5700
 
19aff2a
 
 
 
 
 
 
 
 
 
 
9fc2f7b
19aff2a
2cb08be
19aff2a
c076ae8
 
19aff2a
c076ae8
19aff2a
 
a42db98
2eb5700
 
a42db98
9fc2f7b
a42db98
 
9fc2f7b
a42db98
0b157c6
a42db98
0b157c6
 
 
 
9fc2f7b
0b157c6
 
 
 
2eb5700
0b157c6
9fc2f7b
0b157c6
228f5fb
0b157c6
449305c
9fc2f7b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
449305c
0b157c6
 
19aff2a
 
 
9fc2f7b
 
 
 
 
19aff2a
 
2eb5700
19aff2a
2eb5700
19aff2a
2eb5700
19aff2a
 
 
9fc2f7b
 
19aff2a
2eb5700
9fc2f7b
2eb5700
19aff2a
2eb5700
9fc2f7b
 
19aff2a
9fc2f7b
19aff2a
2eb5700
19aff2a
865382a
9fc2f7b
19aff2a
 
9fc2f7b
865382a
19aff2a
865382a
19aff2a
 
 
865382a
19aff2a
 
 
 
 
865382a
19aff2a
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
---
thumbnail: https://huggingface.co/SZLHOLDINGS/KHIPU-R2/resolve/main/og-card.png
license: apache-2.0
language:
  - en
base_model: Qwen/Qwen2.5-1.5B-Instruct
base_model_relation: adapter
library_name: peft
pipeline_tag: text-generation
tags:
  - qlora
  - peft
  - governed-agent
  - proposal-only
  - research-only
  - szl-holdings
  - khipu
  - abstain-retrain
szl:
  doctrine: v11-LOCKED
  lean: "749/14/163"
  lambda: "Conjecture 1 — advisory, never a theorem"
  artifact_class: ADAPTER
  publication_eligible: false
  autonomy_eligible: false
  never_overwrite: SZLHOLDINGS/SZL-Khipu-1.5B
  jobs: COMPLETED
  job_id: "6a91bf11984507d9db4ea104"
  job_prior_error: "6a91ba2c45686a1580c12020"
  weights: AVAILABLE
  evals: MEASURED
  gpu: UNAVAILABLE
---

<p align="center">
  <img src="holo-banner.svg" alt="KHIPU-R2 — holographic step-up banner" width="100%"/>
</p>

<h1 align="center">K H I P U — R 2</h1>

<p align="center"><em>Failure is an artifact, not a footnote.</em></p>

<p align="center">
  <img alt="Base: Qwen2.5-1.5B-Instruct" src="https://img.shields.io/badge/base-Qwen2.5--1.5B--Instruct-334155?style=flat-square"/>
  <img alt="Downloads" src="https://img.shields.io/huggingface/dt/SZLHOLDINGS/KHIPU-R2?style=flat-square&color=3af4c8&label=downloads"/>
  <img alt="Artifact: QLoRA adapter r=32 a=64" src="https://img.shields.io/badge/artifact-QLoRA%20adapter%20r32%20a64-0d9488?style=flat-square"/>
  <img alt="Abstain: MEASURED 3/6, was 2/6 — not a pass" src="https://img.shields.io/badge/abstain-3%2F6%20was%202%2F6%20%C2%B7%20not%20a%20pass-b45309?style=flat-square"/>
  <img alt="Plan-valid 11/11" src="https://img.shields.io/badge/plan--valid-11%2F11-16a34a?style=flat-square"/>
  <img alt="Hallucinated citations: 0" src="https://img.shields.io/badge/hallucinated%20citations-0-16a34a?style=flat-square"/>
  <img alt="Lifecycle: research-only" src="https://img.shields.io/badge/lifecycle-research--only-7e8aa3?style=flat-square"/>
</p>

<p align="center">
  <code>KANCHAY</code> · Doctrine v11 · Lean <code>749/14/163</code> · Λ = Conjecture 1 (advisory) · <a href="https://a-11-oy.com">a-11-oy.com</a>
</p>

Adapters are on this repo. Abstain is MEASURED 3/6, not a pass. This card-only release does not modify adapter files and does not open the publication or autonomy gates.

QLoRA adapter on disclosed [`Qwen/Qwen2.5-1.5B-Instruct`](https://huggingface.co/Qwen/Qwen2.5-1.5B-Instruct) (runtime `unsloth/Qwen2.5-1.5B-Instruct-bnb-4bit`). Proposal-only brain navigator / abstain retrain. Doctrine v11 LOCKED. Λ = Conjecture 1 (advisory, never a theorem).

<!-- SZL-ATELIER-CUT:v1:START -->
## The cut

Leaders ship v1 and changelog the rest. We name the retraining of silence as its own model. Failure is an artifact, not a footnote.

A public retrain whose only job is to improve one metric: honest abstain under adversarial handles.

### Silhouette → leave → SZL

| Leader | Take, then tweak |
|---|---|
| Anthropic | Red-team → constitution update. We red-team → adapter. |
| NVIDIA | Recipe re-run with a new seed and a bounded delta. |
| Unsloth | Same FastLanguageModel loop, new curriculum, new evidence record. |

This card makes the combination inspectable: a separate abstain-retraining SKU, bounded small-sample results, and an exact-source documentation publisher that cannot modify weights or promotion state.

## Intended use

Continue the abstain-retrain loop. Keep publication and autonomy gates closed until independently verifiable evaluation evidence satisfies the governed release criteria.

## Limitations

- research-only
- No signed R2 eval receipt in this atelier.
- Publishing this card is a documentation update, not model promotion or autonomy approval.

Canonical GitHub: [`szl-holdings/szl-forge/khipu_r2`](https://github.com/szl-holdings/szl-forge/tree/main/khipu_r2)
<!-- SZL-ATELIER-CUT:v1:END -->

## Specification

| | |
|---|---|
| **Artifact** | `adapter_model.safetensors` (147.8M) + `adapter_config.json` **AVAILABLE** |
| **Job** | [`6a91bf11984507d9db4ea104`](https://huggingface.co/jobs/SZLHOLDINGS/6a91bf11984507d9db4ea104) **COMPLETED** |
| **Does NOT overwrite** | signed [`SZL-Khipu-1.5B`](https://huggingface.co/SZLHOLDINGS/SZL-Khipu-1.5B) |
| **Prior job** | [`6a91ba2c`](https://huggingface.co/jobs/SZLHOLDINGS/6a91ba2c45686a1580c12020) **ERROR** Trackio 404 |
| **Card publication** | Exact-source README + local SVG + source binding only; weights, adapter, configs, evals, visibility, hardware, and runtime unchanged |
| **Lab** | Forbidden. Pin stays Khipu GGUF. GPU **UNAVAILABLE**. |
| **License** | Apache-2.0 |
| **Autonomy** | false |

## Evaluation (MEASURED this job)

Method: in-process Unsloth generate, scoring ported from `eval_khipu.py`, temperature 0, held-out never in gradients. Host job worker. Date 2026-08-28 17:20 UTC. File: `eval_measured.json`.

| split | k/n | what-NOT |
|---|---|---|
| plan-valid | **11 / 11** | not a public leaderboard |
| grounding (`eval.jsonl` navigate) | **5 / 5** | n=5 |
| abstain (`adversarial.jsonl`) | **3 / 6** | not 5/5, not 6/6 |
| hallucinated citations | **0** | this job only |

Prior published original `SZL-Khipu-1.5B` MEASURED abstain was **2/6**. This run is **3/6**. Small n. Do not derive a world-rank score from k/n on n=11.

## Training (MEASURED / REPORTED)

- Unsloth QLoRA, seed 11, lr 2e-4, LoRA r=32 α=64, 45 epochs
- Train: 15 navigate + 8 abstain rows × oversample 4 (in-memory 32)
- Held-out: 5 + 6, `held_out_in_gradients: false`
- `training_loss` MEASURED `0.017188…` is a train metric, not an eval
- adapter sha256 `e44d53f29f2d443598e06d6c0441557fd3a5010888c7aa97b56ec3c0e050d349`

## What this is NOT

- Not a replacement for `SZL-Khipu-1.5B`
- Not Chaski (Qwen3.5 lock)
- Not an autonomous agent
- Not a GGUF. Mini GGUFs exist on [`A11OY-MINI`](https://huggingface.co/SZLHOLDINGS/A11OY-MINI); Mini evals none-this-run; Mini does **not** inherit this 3/6

## Load

```python
from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

base_id = "Qwen/Qwen2.5-1.5B-Instruct"
tok = AutoTokenizer.from_pretrained(base_id)
base = AutoModelForCausalLM.from_pretrained(base_id)
model = PeftModel.from_pretrained(base, "SZLHOLDINGS/KHIPU-R2")
```

Owner: Stephen Lutar / SZL Holdings.