LLM EXPLORER 59,420 MODELS INDEXED

Pretrained Llama3.2 3B NCKH by baesad

By baesad · 6 downloads
HF Repo ↗ Copy URL

Pretrained Llama3.2 3B NCKH is an open-source language model by baesad. Features: 3b LLM, VRAM: 7.2GB, Context: 128K, Quantized, LLM Explorer Score: 0.16.

  Arxiv:1910.09700   4bit   Endpoints compatible   Llama   Pytorch   Quantized   Region:us   Safetensors   Sft   Sharded   Tensorflow   Trl   Unsloth
Model Card on HF 🤗: https://huggingface.co/baesad/Pretrained-Llama3.2-3B-NCKH 

Pretrained Llama3.2 3B NCKH Parameters and Internals

LLM NamePretrained Llama3.2 3B NCKH
Repository 🤗https://huggingface.co/baesad/Pretrained-Llama3.2-3B-NCKH 
Model Size3b
Required VRAM7.2 GB
Updated2026-07-11
Maintainerbaesad
Model Typellama
Model Files  5.0 GB: 1-of-2   2.2 GB: 2-of-2   5.0 GB: 1-of-2   2.2 GB: 2-of-2
Quantization Type4bit
Model ArchitectureLlamaForCausalLM
Context Length131072
Model Max Length131072
Transformers Version4.49.0
Tokenizer ClassPreTrainedTokenizer
Padding Token<|finetune_right_pad_id|>
Vocabulary Size128256
Torch Data Typebfloat16

Best Alternatives to Pretrained Llama3.2 3B NCKH

Best Alternatives
Context / RAM
Downloads
Likes
Nanbeige4.1 3B 8bit256K / 4.2 GB18614
Nanbeige4.1 3B 4bit256K / 2.2 GB991
Nanbeige4.1 3B Heretic 8bit256K / 4.2 GB392
...Llama 3.2 3B Legal Chatbot Sft128K / 6.5 GB8170
Legal Sft Llama3.2 3B128K / 6.5 GB11131
Legal Llama3.2 3B Sft Id128K / 6.5 GB6800
Llama 3.2 3B Indo Alpaca PGABL128K / 6.5 GB8620
...gabl Legal Chatbot Llama3.2 3B128K / 6.5 GB5280
Llama 3.2 3B Alpaca Id Sft128K / 6.5 GB4620
Hypa Orpheus 3B TTS VC128K / 6.6 GB1860
Note: green Score (e.g. "73.2") means that the model is better than baesad/Pretrained-Llama3.2-3B-NCKH.