LLM EXPLORER 64,561 MODELS INDEXED

Test 105 7 by master103525

By master103525 · 14 downloads
HF Repo ↗ Copy URL

Test 105 7 is an open-source language model by master103525. Features: 4b LLM, VRAM: 1.1GB, Instruction-Based.

  Arxiv:1910.09700 Base model:adapter:qwen/qwen3-... Base model:qwen/qwen3-4b-instr...   Conversational   Instruct   Lora   Peft   Region:us   Safetensors   Sft   Trl
Model Card on HF 🤗: https://huggingface.co/master103525/test-105-7 

Test 105 7 Parameters and Internals

LLM NameTest 105 7
Repository 🤗https://huggingface.co/master103525/test-105-7 
Base Model(s)  Qwen3 4B Instruct 2507   Qwen/Qwen3-4B-Instruct-2507
Model Size4b
Required VRAM1.1 GB
Updated2026-10-17
Maintainermaster103525
Instruction-BasedYes
Model Files  1.1 GB   0.0 GB
Model ArchitectureAutoModel
Model Max Length1010000
Is Biasednone
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
PEFT TypeLORA
LoRA ModelYes
PEFT Target Modulesup_proj|down_proj|gate_proj|k_proj|v_proj|o_proj|q_proj
LoRA Alpha512
LoRA Dropout0.1
R Param128
Errorsreplace

Best Alternatives to Test 105 7

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3 4B Halo Expert0K / 0.3 GB171
...n3vl 4B Lora No Height R64 Vit0K / 0.3 GB131
... 4B Lora No Parameters R64 Vit0K / 0.3 GB121
Qwen3vl 4B Lora No Fov R64 Vit0K / 0.3 GB111
... 3n 4B It Distill Smollm2 360M0K / 0 GB550
Qwen3 4B Chunky0K / 0.3 GB70
AYA Mistral7B Instruct TR 4B0K / 0.3 GB06
Open Humanizer 4B0K / 0 GB2043
OTSE Qwen3 4B QLoRA0K / 0.3 GB111
QiFu V10K / 0.2 GB641
Note: green Score (e.g. "73.2") means that the model is better than master103525/test-105-7.