LLM EXPLORER 56,604 MODELS INDEXED

Model Name by Loomel

By Loomel · 8 downloads
HF Repo ↗ Copy URL

Model Name is an open-source language model by Loomel. Features: 4b LLM, VRAM: 8.1GB, Context: 32K, License: apache-2.0, Quantized, LLM Explorer Score: 0.19.

  4bit   Conversational   En   Endpoints compatible   Quantized   Qwen3   Region:us   Safetensors   Sharded   Tensorflow   Unsloth
Model Card on HF 🤗: https://huggingface.co/Loomel/model-name 

Model Name Parameters and Internals

LLM NameModel Name
Repository 🤗https://huggingface.co/Loomel/model-name 
Base Model(s)  unsloth/qwen3-4b-base-unsloth-bnb-4bit   unsloth/qwen3-4b-base-unsloth-bnb-4bit
Model Size4b
Required VRAM8.1 GB
Updated2026-08-08
MaintainerLoomel
Model Typeqwen3
Model Files  5.0 GB: 1-of-2   3.1 GB: 2-of-2
Supported Languagesen
Quantization Type4bit
Model ArchitectureQwen3ForCausalLM
Licenseapache-2.0
Context Length32768
Model Max Length32768
Transformers Version4.56.1
Tokenizer ClassQwen2Tokenizer
Padding Token<|vision_pad|>
Vocabulary Size151936
Torch Data Typebfloat16
Errorsreplace

Best Alternatives to Model Name

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3 4b Merged Txt256K / 8.1 GB5350
Qwen3.8 4B Distilled256K / 8.1 GB900
...Instruct 2507 Unsloth Bnb 4bit256K / 3.5 GB15596214
...wen3 4B Thinking 2507 MLX 4bit256K / 2.3 GB5322514
...wen3 4B Thinking 2507 MLX 8bit256K / 4.3 GB447318
Qwen3 4B Instruct 2507 4bit256K / 2.3 GB4095511
...wen 3 4b Mythos Finetune 16bit256K / 8.1 GB920
EduMentor Qwen3 4B FP16256K / 8.1 GB751
...4B Cybersecurity Heretic 16bit256K / 8.1 GB4364
Qwen Finetune 4bit256K / 8.1 GB610
Note: green Score (e.g. "73.2") means that the model is better than Loomel/model-name.