Downloads · 30 days
11
28% of all-time downloads
mr-mc/flowjudge-dialam-qwen3-0.6b-v1-n2048
flowjudge-dialam-qwen3-0.6b-v1-n2048 is a text generation model from mr-mc. Use it when you need the model to write or continue text. It is set up for peft. The card lists the license as other.
This is the QLoRA adapter for the narrow FlowJudge behavior: given one new proposition and a complete block of earlier propositions, emit one bare JSON object containing only direct SUPPORT, ATTACK, or REPHRASE edges…
Downloads · 30 days
11
28% of all-time downloads
All-time downloads
39
Public
Repo size
51.8 MB
Likes
0
Public
Click a slice to open those files.
.safetensors40.4 MB · 72%
From the Hugging Face model README
This is the QLoRA adapter for the narrow FlowJudge behavior: given one new proposition and a complete block of earlier propositions, emit one bare JSON object containing only direct SUPPORT, ATTACK, or REPHRASE edges to supplied IDs, or an empty relation list.
The adapter uses Qwen/Qwen3-0.6B with QLoRA r=16, alpha=32, three epochs, seed 20260823, and 2,048 private transformed QT30 training examples. It is the submission checkpoint because it had the best exact-patch result in the fixed v1 data-efficiency curve, while the subsequent hard-negative v2 experiment increased false edges and failed its preregistered material-improvement test. The training text is not redistributed. Consumers must obtain the official corpus and appropriate permission separately.
| Metric | Selected v1 n=2048 | Rejected v2 n=2048 |
|---|---|---|
| Exact patch accuracy | 26.7% | 23.3% |
| Edge precision | 16.7% | 18.8% |
| Edge recall | 16.7% | 25.0% |
| Edge F1 | 16.7% | 21.4% |
| Relation macro-F1 | 16.7% | 18.8% |
| False edges/update | 0.667 | 0.867 |
No tested training size cleared the preregistered reliability threshold. This checkpoint is published to make the assignment evidence reproducible, not as a claim that the behavior is reliable. The included manifests record the frozen eval/rubric hashes, deterministic metrics, data mix, fixed training config, adapter hash, and private-artifact hashes. The evaluation is small and corpus-specific; this is not evidence of general argument-mining reliability.
Load this PEFT adapter with a compatible Transformers/PEFT or Unsloth runtime.
The training runtime used the 4-bit unsloth/qwen3-0.6b-unsloth-bnb-4bit
mirror of the canonical Qwen/Qwen3-0.6B base.