LLM EXPLORER 59,420 MODELS INDEXED

Falcon Mamba 7B Instruct by tiiuae

By tiiuae · 18925 downloads
HF Repo ↗ Copy URL

Falcon Mamba 7B Instruct is an open-source language model by tiiuae. Features: 7b LLM, VRAM: 14.6GB, License: other, Instruction-Based, LLM Explorer Score: 0.18.

  Arxiv:2312.00752   Arxiv:2410.05355 Base model:finetune:tiiuae/fal... Base model:tiiuae/falcon-mamba...   Conversational Dataset:huggingfacefw/fineweb-... Dataset:tiiuae/falcon-refinedw...   En   Endpoints compatible   Falcon mamba   Instruct   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/tiiuae/falcon-mamba-7b-instruct 

Falcon Mamba 7B Instruct Parameters and Internals

Model Type 
Causal decoder-only
Supported Languages 
en (Main)
Training Details 
Data Sources:
tiiuae/falcon-refinedweb, HuggingFaceFW/fineweb-edu
Data Volume:
~ 5,500 GT
Methodology:
Multi-stage training strategy increasing context length and Curriculum Learning techniques
Context Length:
8192
Training Time:
Roughly two months
Hardware Used:
256 H100 80GB GPUs
Model Architecture:
Mamba
LLM NameFalcon Mamba 7B Instruct
Repository 🤗https://huggingface.co/tiiuae/falcon-mamba-7b-instruct 
Base Model(s)  Falcon Mamba 7B   tiiuae/falcon-mamba-7b
Model Size7b
Required VRAM14.6 GB
Updated2026-07-25
Maintainertiiuae
Model Typefalcon_mamba
Instruction-BasedYes
Model Files  5.0 GB: 1-of-3   5.0 GB: 2-of-3   4.6 GB: 3-of-3
Supported Languagesen
Model ArchitectureFalconMambaForCausalLM
Licenseother
Transformers Version4.44.0.dev0
Tokenizer ClassPreTrainedTokenizerFast
Padding Token>>TITLE<<
Vocabulary Size65024
Torch Data Typebfloat16

Best Alternatives to Falcon Mamba 7B Instruct

Best Alternatives
Context / RAM
Downloads
Likes
Falcon3 Mamba 7B Instruct0K / 14.6 GB93633
Falcon3 Mamba R1 V00K / 14.6 GB1112
Note: green Score (e.g. "73.2") means that the model is better than tiiuae/falcon-mamba-7b-instruct.