Downloads · 30 days
6
16% of all-time downloads
scrubster/dr-stein-stage25-qwen15-doc-verifier-r16-augmented-v2
dr-stein-stage25-qwen15-doc-verifier-r16-augmented-v2 is a machine learning model from scrubster. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for peft.
PEFT/LoRA adapter for Qwen/Qwen2.5-1.5B-Instruct, fine-tuned on the slm-learning GAD-tool translation track.
Downloads · 30 days
6
16% of all-time downloads
All-time downloads
38
Public
Repo size
85.3 MB
Likes
0
Public
Click a slice to open those files.
.safetensors73.9 MB · 87%
From the Hugging Face model README
PEFT/LoRA adapter for Qwen/Qwen2.5-1.5B-Instruct, fine-tuned on the
slm-learning GAD-tool translation track.
(instruction, gad CLI command) pairsstage25_qwen15_doc_verifier_r16_augmented_v2local-cuda-0from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer
base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-1.5B-Instruct")
tok = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-1.5B-Instruct")
model = PeftModel.from_pretrained(base, "scrubster/dr-stein-stage25-qwen15-doc-verifier-r16-augmented-v2")
Doc-verifier rank=16 retrain on the v2 augmented corpus — same fake- parent-prefix synthesis but with 150 distinct reason phrasings drawn from a pool of 20 haiku-paraphrased templates (vs the v1 attempt's single uniform template). The v1 attempt regressed: F1 0.720 -> 0.599, unknown F1 0.462 -> 0.167. Root cause per slm-learning-093: model overfit to the uniform reason template "basename X matches elsewhere; literal path Y not found" (used 150x identically). The 20-template pool now produces 150 unique reason strings — model cannot memorize the template, must learn the underlying structural concept. Same recipe as r=16 baseline (LoRA r=16, alpha=32, 3 epochs, lr=2e-4). Per slm-learning-051: candidate-only. Promotion gate per benchmarks/README.md: f1_macro >= 0.85 + json validity >= 0.90. Falsification criterion: if f1_macro >= 0.720 AND unknown_f1 >= 0.462 the diversity hypothesis is at least partially supported (improvement over r=16 baseline). If still regressed, the data-imbalance lane is exhausted — the next move is a different angle (r=32 on original corpus, or DIFFERENT minority-class signal entirely).