---
thumbnail: https://huggingface.co/SZLHOLDINGS/SZL-Khipu-1.5B/resolve/main/og-card.png
license: apache-2.0
language:
  - en
base_model: Qwen/Qwen3.5-0.8B
base_model_relation: adapter
library_name: peft
pipeline_tag: text-generation
tags:
  - lora
  - peft
  - governed-ai
  - proposal-only
  - szl-holdings
  - chaski
  - cutting
szl:
  doctrine: v11-LOCKED
  lean: "749/14/163"
  lambda: "Conjecture 1 — advisory, never a theorem"
  artifact_class: ADAPTER
  originality: FINETUNE_DISCLOSED_BASE
  jobs: local-5050
  job_id: local-5050
  weights: AVAILABLE
  evals: none-this-run
  publication_eligible: false
  autonomy_eligible: false
  never_overwrite: SZLHOLDINGS/chaski
  copied_live_chaski_weights: false
  train_loss: 2.228136855544466
  train_loss_label: MEASURED
  adapter_sha256: 620b3488fac2ebc6518090424de5b3c6a182293cf52dfd5bd9f886f54aef0df5
---

# Chaski-5050

<p align="center">
  <img src="https://huggingface.co/SZLHOLDINGS/SZL-Khipu-1.5B/resolve/main/og-card.png" alt="CHASKI-5050" width="100%"/>
</p>

`KANCHAY` · Doctrine v11 · Lean `749/14/163` · Λ = Conjecture 1 (advisory) · [a-11-oy.com](https://a-11-oy.com)


Adapters are on this repo. Evals none-this-run. Not MEASURED.

Owner-GPU recut on an RTX 5050 Laptop (bf16 LoRA, not QLoRA). Original SZL cut of disclosed Apache [`Qwen/Qwen3.5-0.8B`](https://huggingface.co/Qwen/Qwen3.5-0.8B). Not a republish of Qwen tensors. Not an Unsloth-default card. CUTTING.



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

Mix-ratio is usually a blog footnote. We named the model after the mix. 50/50 is the cut.

Curriculum as identity. The filename is the experiment.

### Silhouette → leave → SZL

| Leader | Take, then tweak |
|---|---|
| Anthropic | Balanced helpful/harmless mix, as a named checkpoint. |
| NVIDIA | Recipe variant, published. |
| Unsloth | LoRA on Qwen3.5-0.8B, cutting tag. |

Nobody else ships this combination. That is the point of a one-of-one.

## Intended use

Ablation sibling of chaski.

## Limitations

- proposal-only
- No signed mix-ablation in this atelier.

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

| | |
|---|---|
| **Artifact** | `adapter_model.safetensors` 25,587,104 bytes **AVAILABLE** (sha256 `620b3488fac2ebc6518090424de5b3c6a182293cf52dfd5bd9f886f54aef0df5`) |
| **Job** | `local-5050` (owner metal, **not** an HF Job) |
| **Does NOT overwrite** | [`SZLHOLDINGS/chaski`](https://huggingface.co/SZLHOLDINGS/chaski) |
| **Dataset** | [`SZLHOLDINGS/szl-1-doctrine-sft`](https://huggingface.co/datasets/SZLHOLDINGS/szl-1-doctrine-sft) · 41 rows · jsonl sha256 `ddc5594b…0611a243` |
| **Publication / autonomy** | false / false |
| **License** | Apache-2.0 |

## Evaluation

**Status: none-this-run.** No JSON/refusal gate ran. Not 5/5. Not 6/6. Do not load this ID into the Khipu lab.

`train_loss` MEASURED `2.228136855544466` is a **train metric**, not an eval (method: Unsloth trainer log, N=41 rows, 3 epochs, 33 steps, `train_runtime` 883.2224s, 2026-08-28 17:56 UTC). File: `training_receipt.json`.

## Training (MEASURED this run)

- Recipe: `train_chaski_bf16_5050.py` · Unsloth 2026.7.2 · transformers 5.5.0 · torch 2.10.0+cu128
- GPU: NVIDIA GeForce RTX 5050 Laptop, 7.96 GB
- LoRA r=16 α=16, bf16, batch 1, ga 4, lr 2e-4, adamw_8bit, seed 11, max_seq 2048
- `copied_live_chaski_weights: false`

## What this is NOT

- Not live Chaski
- Not a Qwen rehost
- Not a GGUF of this adapter. Mini GGUFs are LIVE on [`A11OY-MINI`](https://huggingface.co/SZLHOLDINGS/A11OY-MINI) (evals none-this-run; they do not inherit this card)
- Not production. Lab load forbidden.

## Load

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

base_id = "Qwen/Qwen3.5-0.8B"
tok = AutoTokenizer.from_pretrained(base_id)
base = AutoModelForCausalLM.from_pretrained(base_id)
model = PeftModel.from_pretrained(base, "SZLHOLDINGS/chaski-5050")
```

Doctrine v11 LOCKED. Λ = Conjecture 1 (advisory, never a theorem). Owner: Stephen Lutar / SZL Holdings.