What are the hardware requirements for WILLAY?

WILLAY requires approximately 2 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed WILLAY and how large is it?

WILLAY is developed by SZLHOLDINGS, a model with 0.5b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate WILLAY?

WILLAY is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

WILLAY by SZLHOLDINGS — VRAM 2GB

Name: WILLAY
Author: SZLHOLDINGS

WILLAY is an open-source language model by SZLHOLDINGS. Features: 0.5b LLM, VRAM: 2GB, License: apache-2.0, Instruction-Based.

Adapter Base model:adapter:qwen/qwen2.... Base model:qwen/qwen2.5-0.5b-i... Conversational Dataset:szlholdings/doctrine-v... Dataset:szlholdings/szl-1-doct... Dataset:szlholdings/why-we-lea... Doctrine-v11 En Finetuned Governed-ai Identity Instruct Lora Peft Proposal-only Qwen2 Region:us Safetensors Sft Szl-holdings Willay

Model Card on HF 🤗: https://huggingface.co/SZLHOLDINGS/WILLAY

WILLAY Parameters and Internals

LLM Name	WILLAY
Repository 🤗	https://huggingface.co/SZLHOLDINGS/WILLAY
Base Model(s)	Qwen/Qwen2.5-0.5B-Instruct Qwen/Qwen2.5-0.5B-Instruct
Model Size	0.5b
Required VRAM	2 GB
Updated	2026-11-06
Maintainer	SZLHOLDINGS
Instruction-Based	Yes
Model Files	0.0 GB 2.0 GB
Supported Languages	en
Model Architecture	Adapter
License	apache-2.0
Model Max Length	131072
Is Biased	none
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
PEFT Type	LORA
LoRA Model	Yes
PEFT Target Modules	q_proj\|v_proj
LoRA Alpha	32
LoRA Dropout	0.05
R Param	16
Errors	replace

Rank the WILLAY capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to WILLAY

Best Alternatives	Context / RAM	Downloads	Likes
Miha Qwen2.5 0.5B Lora	0K / 0 GB	47	1
...2.5 0.5B Banking77 Intent Lora	0K / 0 GB	57	1
...B Instruct Opensec Triage Lora	0K / 0 GB	15	1
...mato Risk Reasoner V0.1.7 Lora	0K / 0 GB	42	0
...tion Risk Reasoner V0.1.8 Lora	0K / 0 GB	21	0
Olo Qwen2.5 0.5B Hotfolder	0K / 0 GB	32	1

Note: green Score (e.g. "73.2") means that the model is better than SZLHOLDINGS/WILLAY.