LLM EXPLORER 59,420 MODELS INDEXED

Ahma 3B Instruct by Finnish-NLP

By Finnish-NLP · 14 downloads
HF Repo ↗ Copy URL

Ahma 3B Instruct is an open-source language model by Finnish-NLP. Features: 3b LLM, VRAM: 7.3GB, Context: 2K, License: apache-2.0, Instruction-Based, Merged, LLM Explorer Score: 0.13.

  Merged Model   Arxiv:2302.13971   Arxiv:2402.14526 Base model:finetune:finnish-nl...   Base model:finnish-nlp/ahma-3b   Conversational   Doi:10.57967/hf/4010   Fi   Finnish   Instruct   Llama   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/Finnish-NLP/Ahma-3B-Instruct 

Ahma 3B Instruct Parameters and Internals

LLM NameAhma 3B Instruct
Repository 🤗https://huggingface.co/Finnish-NLP/Ahma-3B-Instruct 
Base Model(s)  Ahma 3B   Finnish-NLP/Ahma-3B
Merged ModelYes
Model Size3b
Required VRAM7.3 GB
Updated2026-08-07
MaintainerFinnish-NLP
Model Typellama
Instruction-BasedYes
Model Files  5.0 GB: 1-of-2   2.3 GB: 2-of-2
Supported Languagesfi
Model ArchitectureLlamaForCausalLM
Licenseapache-2.0
Context Length2048
Model Max Length2048
Transformers Version4.41.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<PAD>
Vocabulary Size64257
Torch Data Typefloat16

Best Alternatives to Ahma 3B Instruct

Best Alternatives
Context / RAM
Downloads
Likes
Llama 3.2 3B Instruct128K / 6.5 GB18295722368
Llama 3.2 3B Instruct128K / 6.5 GB18620796
Llama3.2 3B Manumit V1128K / 6.4 GB1350
Llama 3.2 3B Instruct128K / 6.5 GB9919
Alloma 3B Instruct128K / 6.5 GB554566
... V7 SFT 10voicebot Loopcleaned128K / 6.5 GB2290
Llama 3.2 3B RP Toxic Fuse128K / 6.4 GB142
DeepSeek R1 Distill Llama 3B128K / 6.5 GB24716
Schematron 3B128K / 6.5 GB969336
ReasoningCore 3B T1 1128K / 6.5 GB81
Note: green Score (e.g. "73.2") means that the model is better than Finnish-NLP/Ahma-3B-Instruct.