Parameter descriptions:

Base Model Data

Are datasources for training the base model comprehensively documented and made available? In case a distinction between base (foundation) and end (user) model is not applicable, this mirrors the end model data entries.

End User Model Data

Are datasources for training the model that the end user interacts with comprehensively documented and made available?

Base Model Weights

Are the weights of the base models made freely available? In case a distinction between base (foundation) and end (user) model is not applicable, this mirrors the end model data entries.

End User Model Weights

Are the weights of the model that the end user interacts with made freely available?

Training Code

Is the source code of dataset processing, model training and tuning comprehensively made available?

Code Documentation

Is the source code of datasource processing, model training and tuning comprehensively documented?

Hardware Architecture

Is the hardware architecture used for datasource processing and model training comprehensively documented?

Preprint

Are archived preprint(s) are available that detail all major parts of the system including datasource processing, model training and tuning steps?

Paper

Are peer-reviewed scientific publications available that detail all major parts of the system including datasource processing, model training and tuning steps?

Modelcard

Is a model card available in standardized format that provides comprehensive insight on model architecture, training, fine-tuning, and evaluation?

Datasheet

Is a datasheet as defined in "Datasheets for Datasets" (Gebru et al. 2021) available?

Package

Is a packaged release of the model available on a software repository (e.g. a Python Package Index, Homebrew)?

API and Meta Prompts

Is an API available that provides unrestricted access to the model (other than security and CDN restrictions)? If applicable, this entry also collects information on the use and availability of meta prompts.

Licenses

Is the project fully covered by Open Source Initiative (OSI)-approved licenses, including all data sources and training pipeline code?

Falcon

by Technology Innovation Institute

About the model:

Open foundation model.

Model type:

Text

Model performance class:

Full

Link to the model:

https://huggingface.co/tiiuae/Falcon3-10B-Instruct

Base models:

Falcon3-10B-Base

End model:

Falcon3-10B-Instruct

End model license:

Falcon-LLM-License

About the organisation:

The Technology Innovation Institute, a research center. Part of the Abu Dhabi Government’s Advanced Technology Research Council (ATRC).

Link to the organisation:

https://falconllm.tii.ae

Model release date:

May 2023

Availability

Base Model Data

From the documentation 'The key ingredient for the high quality of the Falcon models is their training data, predominantly based (>80%) on RefinedWeb — a novel massive web dataset based on CommonCrawl' (https://huggingface.co/blog/falcon). However, only a small sample is made available.

https://huggingface.co/datasets/tiiuae/falcon-refinedweb

End User Model Data

RL data inherited from Baize but provenance not well-documented. From the documentation 'Falcon-40B-Instruct was finetuned on a 150M tokens from Baize mixed with 5% of RefinedWeb data.'

https://github.com/project-baize/baize-chatbot

Base Model Weights

Model weights available through HuggingFace library

https://huggingface.co/tiiuae/falcon-40b-instruct/tree/main

End User Model Weights

No RL weights or checkpoints made available

https://github.com/project-baize/baize-chatbot#v1

Training Code

No source code shared, even though the term "open source" is used.

https://huggingface.co/tiiuae/falcon-40b-instruct

Documentation

Code Documentation

No source code found, therefore no documentation found.

Hardware Architecture

Architecture sketched on HuggingFace as "Falcon-40B-Instruct is a 40B parameters causal decoder-only model built by TII based on Falcon-40B and finetuned on a mixture of Baize."

Preprint

First preprint covers the creation and curation of RefinedWeb dataset, but not other aspects of the model. The second preprint provides more details about the model architecture, implementation, evaluation results, and limitations.

https://arxiv.org/abs/2306.01116 https://arxiv.org/abs/2311.16867

Paper

No peer-reviewed paper known.

Modelcard

Model card on HuggingFace is mostly used to advertise the model, not to document its training and evaluation details.

https://huggingface.co/tiiuae/falcon-40b-instruct

Datasheet

The preprint on RefinedWeb dataset presents a datasheet in a section named "RefinedWeb Datasheet", which goes into quite a bit of detail on collection, preprocessing etc.

https://arxiv.org/abs/2306.01116

Access

Package

Model available on Ollama.

https://ollama.com/library/falcon3

API and Meta Prompts

There is no API, and HuggingFace inference API is disabled for this model.

https://huggingface.co/tiiuae/falcon-40b-instruct

Licenses

First release came with a legally murky license that was swiftly criticised and now generates a 404. Current documentation 'Falcon-40B-Instruct is made available under the Apache 2.0 license.'

Is this information not up to date?

Contribute here ->

Supported by the Centre for Language Studies and the Dutch Research Council. Website design & development © 2024 by BSTN. This version of the index generated 13 Aug 2025, website content last updated 15 Aug 2025.