LLM EXPLORER 61,405 MODELS INDEXED

Llm Jp 3 3.7B Instruct by llm-jp

By llm-jp · 1020 downloads

Llm Jp 3 3.7B Instruct is an open-source language model by llm-jp. Features: 3.7b LLM, VRAM: 7.6GB, Context: 4K, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.15.

  Conversational   Deploy:azure   En   Instruct   Ja   Llama   Region:us   Safetensors   Sharded   Tensorflow

Llm Jp 3 3.7B Instruct Parameters and Internals

Model Type 
text-generation
Additional Notes 
The models released here are in the early stages of R&D and have not been tuned for human intent and safety considerations.
Supported Languages 
en (English), ja (Japanese)
Training Details 
Data Sources:
Wikipedia, Common Crawl, WARP/PDF, WARP/HTML, Kaken, Dolma/CC-head, The Stack, Dolma/C4, Dolma/Reddit, Dolma/PeS2o, Dolma/Gutenberg, Dolma/Wiki
Data Volume:
2.1T tokens
Context Length:
4096
Model Architecture:
Transformer-based
Input Output 
Accepted Modalities:
text
Output Format:
text
LLM NameLlm Jp 3 3.7B Instruct
Repository πŸ€—https://huggingface.co/llm-jp/llm-jp-3-3.7b-instruct 
Model Size3.7b
Required VRAM7.6 GB
Updated2026-07-29
Maintainerllm-jp
Model Typellama
Instruction-BasedYes
Model Files  5.0 GB: 1-of-2   2.6 GB: 2-of-2
Supported Languagesen ja
Model ArchitectureLlamaForCausalLM
Licenseapache-2.0
Context Length4096
Model Max Length4096
Transformers Version4.40.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<PAD|LLM-jp>
Vocabulary Size99584
Torch Data Typebfloat16

Best Alternatives to Llm Jp 3 3.7B Instruct

Best Alternatives
Context / RAM
Downloads
Likes
SJT 3.7B4K / 8.2 GB90
....0 Computer Crime Act Instruct4K / 13.7 GB90
...3 3.7B Instruct EZO Humanities4K / 7.6 GB336
Llm Jp 3 3.7B Instruct4K / 7.6 GB50
Note: green Score (e.g. "73.2") means that the model is better than llm-jp/llm-jp-3-3.7b-instruct.