LLM EXPLORER 61,918 MODELS INDEXED

Qwen3.5 2B Auto Optimized by islamassanov

By islamassanov · 31 downloads
HF Repo ↗ Copy URL

Qwen3.5 2B Auto Optimized is an open-source language model by islamassanov. Features: 2b LLM, VRAM: 4.5GB, License: apache-2.0, LLM Explorer Score: 0.28.

Base model:finetune:qwen/qwen3...   Base model:qwen/qwen3.5-2b   Conversational   Cuda-graphs   Endpoints compatible   H100   Image-text-to-text   Inference-optimization   Qwen3.5   Qwen3 5   Region:us   Safetensors   Sharded   Speculative-decoding   Tensorflow   Triton
Model Card on HF 🤗: https://huggingface.co/islamassanov/Qwen3.5-2B-auto-optimized 

Qwen3.5 2B Auto Optimized Parameters and Internals

LLM NameQwen3.5 2B Auto Optimized
Repository 🤗https://huggingface.co/islamassanov/Qwen3.5-2B-auto-optimized 
Base Model(s)  Qwen/Qwen3.5-2B   Qwen/Qwen3.5-2B
Model Size2b
Required VRAM4.5 GB
Updated2026-09-22
Maintainerislamassanov
Model Typeqwen3_5
Model Files  4.5 GB: 1-of-1
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version4.57.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to Qwen3.5 2B Auto Optimized

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3.5 2B0K / 4.5 GB2633311352
WeMM Embedding 2B0K / 5.4 GB543368
Qwen3.5 2B Base0K / 4.5 GB80959182
Qwen3.8 2B Distill0K / 4.5 GB205624
Decider 2B Vision0K / 4.4 GB1055
Qwen3.5 2B0K / 4.5 GB10348718
Qwen3.5 2B Manumit V20K / 4.4 GB4310
Qwen3.5 2B Vi SFT0K / 4.5 GB1911
Qwen3.8 2B0K / 4.5 GB31913
JOSIE 2 2B OSS0K / 4.5 GB4848
Note: green Score (e.g. "73.2") means that the model is better than islamassanov/Qwen3.5-2B-auto-optimized.