⚠️ ARCHIVED / LEGACY MODEL NOTICE
This repository is part of a legacy collection quantized around 2023. To manage storage quotas and maintain active community projects, some rarely used quantization formats (e.g., Q2_K, Q3_K, Q4_1, Q5_1) have been permanently removed.
Only the most popular and stable formats (Q4_0, Q4_K_M, Q5_K_M, Q6_K, and Q8_0) remain available.
💡 Looking for something modern?
If you are starting a new project, we highly recommend using newer architectures (like Llama 3, Mistral, or Qwen) provided by official maintainers or active community members (e.g., Bartowski, TheBloke legacy files, or official organization handles).
⚠️ This repository is no longer actively maintained. Existing files are provided "as is" for archival and legacy hardware purposes.
New releases of Llama.cpp now support K-quantization for previously incompatible models, in particular all Falcon 7B models (While Falcon 40b is and always has been fully compatible with K-Quantisation). This is achieved by employing a fallback solution for model layers that cannot be quantized with real K-quants.
For Falcon 7B models, although only a quarter of the layers can be quantized with true K-quants, this approach still benefits from utilizing different legacy quantization types Q4_0, Q4_1, Q5_0, and Q5_1. As a result, it offers better quality at the same file size or smaller file sizes with comparable performance.
So this solution ensures improved performance and efficiency over legacy Q4_0, Q4_1, Q5_0 and Q5_1 Quantizations.
Brief
These are gguf quantized models of the riginal Falcon 40B Model by tiiuae.
Falcon is a foundational large language model coming in different sizes: 7b, 40b and 180b.
Sadly, as the Falcon 180b Models are note really free models, I do not provide quantized versions here.
About GGUF format
gguf is the current file format used by the ggml library.
A growing list of Software is using it and can therefore use this model.
The core project making use of the ggml library is the llama.cpp project by Georgi Gerganov
Quantization variants
There is a bunch of quantized files available to cater to your specific needs. Here's how to choose the best option for you:
Legacy quants
Q4_0, Q4_1, Q5_0, Q5_1 and Q8 are legacy quantization types.
Nevertheless, they are fully supported, as there are several circumstances that cause certain model not to be compatible with the modern K-quants.
Note:
Now there's a new option to use K-quants even for previously 'incompatible' models, although this involves some fallback solution that makes them not real K-quants. More details can be found in affected model descriptions.
(This mainly refers to Falcon 7b and Starcoder models)
K-quants
K-quants are designed with the idea that different levels of quantization in specific parts of the model can optimize performance, file size, and memory load.
So, if possible, use K-quants.
With a Q6_K, you'll likely find it challenging to discern a quality difference from the original model - ask your model two times the same question and you may encounter bigger quality differences.
Original Model Card:
🚀 Falcon-40B
Falcon-40B is a 40B parameters causal decoder-only model built by TII and trained on 1,000B tokens of RefinedWeb enhanced with curated corpora. It is made available under the Apache 2.0 license.
Paper coming soon 😊.
🤗 To get started with Falcon (inference, finetuning, quantization, etc.), we recommend reading this great blogpost fron HF!
It is made available under a permissive Apache 2.0 license allowing for commercial use, without any royalties or restrictions.
⚠️ This is a raw, pretrained model, which should be further finetuned for most usecases. If you are looking for a version better suited to taking generic instructions in a chat format, we recommend taking a look at Falcon-40B-Instruct.
💸 Looking for a smaller, less expensive model?Falcon-7B is Falcon-40B's little brother!
python
1from transformers import AutoTokenizer, AutoModelForCausalLM
2import transformers
3import torch
45model ="tiiuae/falcon-40b"67tokenizer = AutoTokenizer.from_pretrained(model)8pipeline = transformers.pipeline(9"text-generation",10 model=model,11 tokenizer=tokenizer,12 torch_dtype=torch.bfloat16,13 trust_remote_code=True,14 device_map="auto",15)16sequences = pipeline(17"Girafatron is obsessed with giraffes, the most glorious animal on the face of this Earth. Giraftron believes all other animals are irrelevant when compared to the glorious majesty of the giraffe.\nDaniel: Hello, Girafatron!\nGirafatron:",18 max_length=200,19 do_sample=True,20 top_k=10,21 num_return_sequences=1,22 eos_token_id=tokenizer.eos_token_id,23)24for seq in sequences:25print(f"Result: {seq['generated_text']}")26
💥 Falcon LLMs require PyTorch 2.0 for use with transformers!
Language(s) (NLP): English, German, Spanish, French (and limited capabilities in Italian, Portuguese, Polish, Dutch, Romanian, Czech, Swedish);
License: Apache 2.0 license.
Model Source
Paper:coming soon.
Uses
Direct Use
Research on large language models; as a foundation for further specialization and finetuning for specific usecases (e.g., summarization, text generation, chatbot, etc.)
Out-of-Scope Use
Production use without adequate assessment of risks and mitigation; any use cases which may be considered irresponsible or harmful.
Bias, Risks, and Limitations
Falcon-40B is trained mostly on English, German, Spanish, French, with limited capabilities also in in Italian, Portuguese, Polish, Dutch, Romanian, Czech, Swedish. It will not generalize appropriately to other languages. Furthermore, as it is trained on a large-scale corpora representative of the web, it will carry the stereotypes and biases commonly encountered online.
Recommendations
We recommend users of Falcon-40B to consider finetuning it for the specific set of tasks of interest, and for guardrails and appropriate precautions to be taken for any production use.
How to Get Started with the Model
python
1from transformers import AutoTokenizer, AutoModelForCausalLM
2import transformers
3import torch
45model ="tiiuae/falcon-40b"67tokenizer = AutoTokenizer.from_pretrained(model)8pipeline = transformers.pipeline(9"text-generation",10 model=model,11 tokenizer=tokenizer,12 torch_dtype=torch.bfloat16,13 trust_remote_code=True,14 device_map="auto",15)16sequences = pipeline(17"Girafatron is obsessed with giraffes, the most glorious animal on the face of this Earth. Giraftron believes all other animals are irrelevant when compared to the glorious majesty of the giraffe.\nDaniel: Hello, Girafatron!\nGirafatron:",18 max_length=200,19 do_sample=True,20 top_k=10,21 num_return_sequences=1,22 eos_token_id=tokenizer.eos_token_id,23)24for seq in sequences:25print(f"Result: {seq['generated_text']}")26
Training Details
Training Data
Falcon-40B was trained on 1,000B tokens of RefinedWeb, a high-quality filtered and deduplicated web dataset which we enhanced with curated corpora. Significant components from our curated copora were inspired by The Pile (Gao et al., 2020).
Decoder-block: parallel attention/MLP with a two layer norms.
For multiquery, we are using an internal variant which uses independent key and values per tensor parallel degree.
Hyperparameter
Value
Comment
Layers
60
d_model
8192
head_dim
64
Reduced to optimise for FlashAttention
Vocabulary
65024
Sequence length
2048
Compute Infrastructure
Hardware
Falcon-40B was trained on AWS SageMaker, on 384 A100 40GB GPUs in P4d instances.
Software
Falcon-40B was trained a custom distributed training codebase, Gigatron. It uses a 3D parallelism approach combined with ZeRO and high-performance Triton kernels (FlashAttention, etc.)
Citation
Paper coming soon 😊. In the meanwhile, you can use the following information to cite:
@article{falcon40b,
title={{Falcon-40B}: an open large language model with state-of-the-art performance},
author={Almazrouei, Ebtesam and Alobeidli, Hamza and Alshamsi, Abdulaziz and Cappelli, Alessandro and Cojocaru, Ruxandra and Debbah, Merouane and Goffinet, Etienne and Heslow, Daniel and Launay, Julien and Malartic, Quentin and Noune, Badreddine and Pannier, Baptiste and Penedo, Guilherme},
year={2023}
}
To learn more about the pretraining dataset, see the 📓 RefinedWeb paper.
@article{refinedweb,
title={The {R}efined{W}eb dataset for {F}alcon {LLM}: outperforming curated corpora with web data, and web data only},
author={Guilherme Penedo and Quentin Malartic and Daniel Hesslow and Ruxandra Cojocaru and Alessandro Cappelli and Hamza Alobeidli and Baptiste Pannier and Ebtesam Almazrouei and Julien Launay},
journal={arXiv preprint arXiv:2306.01116},
eprint={2306.01116},
eprinttype = {arXiv},
url={https://arxiv.org/abs/2306.01116},
year={2023}
}
License
Falcon-40B is made available under the Apache 2.0 license.
Coming Soon: I'm in the process of launching a sponsorship/crowdfunding campaign for my work. I'm evaluating Kickstarter, Patreon, or the new GitHub Sponsors platform, and I am hoping for some support and contribution to the continued availability of these kind of models. Your support will enable me to provide even more valuable resources and maintain the models you rely on. Your patience and ongoing support are greatly appreciated as I work to make this page an even more valuable resource for the community.
