# Jev-style spam LoRA decision model by Abubakar Abid

> Jev-style spam LoRA adapts Qwen3.5-2B for forum-spam decisions. Read its reported test results, synthetic-data limits and base requirements.

Updated 2026-11-07 (Australia/Adelaide)

## Quick answers

### What is Jev-style spam LoRA by Abubakar Abid?

[Jev-style spam LoRA](<https://decisionmodels.ai/models/jev-spam>): A Qwen3.5-2B LoRA for one forum-spam question.

[Jev-style spam LoRA facts](<https://decisionmodels.ai/models/jev-spam>)

[Source](<https://huggingface.co/abidlabs/jev-spam-decision-2b>)

### Who makes Jev-style spam LoRA?

[Jev-style spam LoRA](<https://decisionmodels.ai/models/jev-spam>): Abubakar Abid

[Jev-style spam LoRA facts](<https://decisionmodels.ai/models/jev-spam>)

[Source](<https://huggingface.co/abidlabs/jev-spam-decision-2b>)

By Abubakar Abid · Hugging Face package release: 2026-11-05 · OPEN

A Qwen3.5-2B LoRA for one forum-spam question. It reads yes/no token logits as a spam probability after supervised training on the dataset’s training split. It is independent of TypeSafe’s hosted [Jev](<https://decisionmodels.ai/models/jev>) and needs the separate Qwen base.

## Facts

- Task: Binary forum-spam detection; text truncated to 1,500 characters in evaluation
- Package: PEFT LoRA adapter; separate Qwen3.5-2B base
- Licence: Apache-2.0

## Claims and sources

- **VENDOR-REPORTED**: The maker reports 2,723 of 2,725 test rows correct, F1 0.9993 and ECE 0.001, with two false positives and no false negatives. Spam examples are largely synthetic promotions. The fine-tune saw this dataset’s training domain while the compared open models were general-purpose; no result against TypeSafe’s [Jev](<https://decisionmodels.ai/models/jev>) is claimed. [source](<https://huggingface.co/abidlabs/jev-spam-decision-2b>)

## Links

- [Source: Card and adapter](<https://huggingface.co/abidlabs/jev-spam-decision-2b>)

## Related projects

No related projects are listed in the registry.