polly-scbe-7b-v2

<!-- scbe-status -->

Status: superseded. Canonical Polly model: polly-scbe-7b-v2-merged. Kept as research history.

Polly SCBE 7B v2 — LoRA Adapter

QLoRA adapter for Qwen/Qwen2.5-7B-Instruct, fine-tuned on the SCBE-AETHERMOORE Sacred Tongues governance framework.

For the merged full-weight model, see issdandavis/polly-scbe-7b-v2-merged.

Usage (adapter)

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

base = AutoModelForCausalLM.from_pretrained(
    "Qwen/Qwen2.5-7B-Instruct",
    device_map="auto",
    load_in_4bit=True,
)
model = PeftModel.from_pretrained(base, "issdandavis/polly-scbe-7b-v2")
tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-7B-Instruct")

About Sacred Tongues

The Sacred Tongues tokenizer encodes text through 6 language-archetype dimensions (Kor'aelin, Avali, Runethic, Cassisivadan, Umbroth, Draumric), seeded from 528 pages of Everweave RPG session logs. Each dimension maps to a primary coding language and cognitive style, giving the vocabulary semantic density beyond frequency statistics.

See issdandavis/polly-scbe-7b-v2-merged for full documentation.

• Developer: Issac Davis — solo builder, Port Angeles WA