LLM EXPLORER 60,519 MODELS INDEXED

Phi3 Wbn by adhirajpandey

By adhirajpandey · 14 downloads

Phi3 Wbn is an open-source language model by adhirajpandey. Features: 3.8b LLM, VRAM: 7.7GB, Context: 128K, Quantized, Instruction-Based, LLM Explorer Score: 0.13.

  Arxiv:1910.09700   Conversational   Custom code   Endpoints compatible   Gguf   Instruct   Phi3   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF πŸ€—: https://huggingface.co/adhirajpandey/phi3-wbn 

Phi3 Wbn Parameters and Internals

LLM NamePhi3 Wbn
Repository πŸ€—https://huggingface.co/adhirajpandey/phi3-wbn 
Model Size3.8b
Required VRAM7.7 GB
Updated2026-08-08
Maintaineradhirajpandey
Model Typephi3
Instruction-BasedYes
Model Files  5.0 GB: 1-of-2   2.7 GB: 2-of-2   2.4 GB   7.6 GB
GGUF QuantizationYes
Quantization Typegguf
Model ArchitecturePhi3ForCausalLM
Context Length131072
Model Max Length131072
Transformers Version4.41.2
Tokenizer ClassLlamaTokenizer
Padding Token<|im_end|>
Vocabulary Size32013
Torch Data Typefloat16

Best Alternatives to Phi3 Wbn

Best Alternatives
Context / RAM
Downloads
Likes
...URA Eve Tactical Instruct 3.8B128K / 2.5 GB1770
Lexi Coder V4.3128K / 7.7 GB3091
Phi 4 Mini Instruct Desyced128K / 7.7 GB4720
Sign Language 20250823 190451128K / 7.7 GB180
...nstruct Cinder With 16bit GGUF4K / 7.7 GB571
Ph3 Q8128K / 4 GB100
...hi 3 Mini 4K Instruct Sft 4bit4K / 2.3 GB100
...hi 3 Mini 4K Instruct Sft 8bit4K / 4.1 GB50
Phi 4 Mini Instruct128K / 7.7 GB514603791
Phi 3.5 Mini Instruct128K / 7.7 GB1164079999
Note: green Score (e.g. "73.2") means that the model is better than adhirajpandey/phi3-wbn.