metadata

inference: false
license: other

Want to contribute? TheBloke's Patreon page

Falcon 40B-Instruct GGML GGML

These files are GGML format model files for Falcon 40B-Instruct GGML.

GGML files are for CPU + GPU inference using llama.cpp and libraries and UIs which support this format, such as:

Repositories available

Compatibility

Original llama.cpp quant methods: `q4_0, q4_1, q5_0, q5_1, q8_0`

I have quantized these 'original' quantisation methods using an older version of llama.cpp so that they remain compatible with llama.cpp as of May 19th, commit 2d5db48.

They should be compatible with all current UIs and libraries that use llama.cpp, such as those listed at the top of this README.

New k-quant methods: `q2_K, q3_K_S, q3_K_M, q3_K_L, q4_K_S, q4_K_M, q5_K_S, q6_K`

These new quantisation methods are only compatible with llama.cpp as of June 6th, commit 2d43387.

They will NOT be compatible with koboldcpp, text-generation-ui, and other UIs and libraries yet. Support is expected to come over the next few days.

Explanation of the new k-quant methods

The new methods available are:

GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw)
GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw.
GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw.
GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw
GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw
GGML_TYPE_Q8_K - "type-0" 8-bit quantization. Only used for quantizing intermediate results. The difference to the existing Q8_0 is that the block size is 256. All 2-6 bit dot products are implemented for this quantization type.

Refer to the Provided Files table below to see what files use which methods, and how.

Provided files

Name	Quant method	Bits	Size	Max RAM required	Use case
Falcon-40b-Instruct.ggmlv3.q4_0.bin	q4_0	4	23.54 GB	26.04 GB	Original llama.cpp quant method, 4-bit.
Falcon-40b-Instruct.ggmlv3.q4_1.bin	q4_1	4	26.15 GB	28.65 GB	Original llama.cpp quant method, 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models.
Falcon-40b-Instruct.ggmlv3.q5_0.bin	q5_0	5	28.77 GB	31.27 GB	Original llama.cpp quant method, 5-bit. Higher accuracy, higher resource usage and slower inference.
Falcon-40b-Instruct.ggmlv3.q5_1.bin	q5_1	5	31.38 GB	33.88 GB	Original llama.cpp quant method, 5-bit. Even higher accuracy, resource usage and slower inference.

Note: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.

How to run in `llama.cpp`

I use the following command line; adjust for your tastes and needs:

./main -t 10 -ngl 32 -m wizardcoder-15b.ggmlv3.q5_0.bin --color -c 2048 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "### Instruction: Write a story about llamas\n### Response:"

Change -t 10 to the number of physical CPU cores you have. For example if your system has 8 cores/16 threads, use -t 8.

Change -ngl 32 to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.

If you want to have a chat-style conversation, replace the -p <PROMPT> argument with -i -ins

How to run in `text-generation-webui`

Further instructions here: text-generation-webui/docs/llama.cpp-models.md.

Discord

For further support, and discussions on these models and AI in general, join us at:

TheBloke AI's Discord server

Thanks, and how to contribute.

Thanks to the chirper.ai team!

I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training.

If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects.

Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits.

Patreon: https://patreon.com/TheBlokeAI
Ko-Fi: https://ko-fi.com/TheBlokeAI

Special thanks to: Luke from CarbonQuill, Aemon Algiz, Dmitriy Samsonov.

Patreon special mentions: Oscar Rangel, Eugene Pentland, Talal Aujan, Cory Kujawski, Luke, Asp the Wyvern, Ai Maven, Pyrater, Alps Aficionado, senxiiz, Willem Michiel, Junyu Yang, trip7s trip, Sebastain Graf, Joseph William Delisle, Lone Striker, Jonathan Leane, Johann-Peter Hartmann, David Flickinger, Spiking Neurons AB, Kevin Schuppel, Mano Prime, Dmitriy Samsonov, Sean Connelly, Nathan LeClaire, Alain Rossmann, Fen Risland, Derek Yates, Luke Pendergrass, Nikolai Manek, Khalefa Al-Ahmad, Artur Olbinski, John Detwiler, Ajan Kanaga, Imad Khwaja, Trenton Dambrowitz, Kalila, vamX, webtim, Illia Dulskyi.

Thank you to all my generous patrons and donaters!

Original model card: Falcon 40B-Instruct GGML

✨ Falcon-40B-Instruct

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. 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!

Why use Falcon-40B-Instruct?

You are looking for a ready-to-use chat/instruct model based on Falcon-40B.
Falcon-40B is the best open-source model available. It outperforms LLaMA, StableLM, RedPajama, MPT, etc. See the OpenLLM Leaderboard.
It features an architecture optimized for inference, with FlashAttention (Dao et al., 2022) and multiquery (Shazeer et al., 2019).

💬 This is an instruct model, which may not be ideal for further finetuning. If you are interested in building your own instruct/chat model, we recommend starting from Falcon-40B.

💸 Looking for a smaller, less expensive model? Falcon-7B-Instruct is Falcon-40B-Instruct's little brother!

from transformers import AutoTokenizer, AutoModelForCausalLM
import transformers
import torch

model = "tiiuae/falcon-40b-instruct"

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    torch_dtype=torch.bfloat16,
    trust_remote_code=True,
    device_map="auto",
)
sequences = pipeline(
   "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:",
    max_length=200,
    do_sample=True,
    top_k=10,
    num_return_sequences=1,
    eos_token_id=tokenizer.eos_token_id,
)
for seq in sequences:
    print(f"Result: {seq['generated_text']}")

For fast inference with Falcon, check-out Text Generation Inference! Read more in this blogpost.

You will need at least 85-100GB of memory to swiftly run inference with Falcon-40B.

Model Card for Falcon-40B-Instruct

Model Details

Model Description

Developed by: https://www.tii.ae;
Model type: Causal decoder-only;
Language(s) (NLP): English and French;
License: Apache 2.0;
Finetuned from model: Falcon-40B.

Model Source

Paper: coming soon.

Uses

Direct Use

Falcon-40B-Instruct has been finetuned on a chat dataset.

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-Instruct is mostly trained on English data, and 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-Instruct to develop guardrails and to take appropriate precautions for any production use.

How to Get Started with the Model

from transformers import AutoTokenizer, AutoModelForCausalLM
import transformers
import torch

model = "tiiuae/falcon-40b-instruct"

tokenizer = AutoTokenizer.from_pretrained(model)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    torch_dtype=torch.bfloat16,
    trust_remote_code=True,
    device_map="auto",
)
sequences = pipeline(
   "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:",
    max_length=200,
    do_sample=True,
    top_k=10,
    num_return_sequences=1,
    eos_token_id=tokenizer.eos_token_id,
)
for seq in sequences:
    print(f"Result: {seq['generated_text']}")

Training Details

Training Data

Falcon-40B-Instruct was finetuned on a 150M tokens from Bai ze mixed with 5% of RefinedWeb data.

The data was tokenized with the Falcon-7B/40B tokenizer.

Evaluation

Paper coming soon.

See the OpenLLM Leaderboard for early results.

Technical Specifications

For more information about pretraining, see Falcon-40B.

Model Architecture and Objective

Falcon-40B is a causal decoder-only model trained on a causal language modeling task (i.e., predict the next token).

The architecture is broadly adapted from the GPT-3 paper (Brown et al., 2020), with the following differences:

Positionnal embeddings: rotary (Su et al., 2021);
Attention: multiquery (Shazeer et al., 2019) and FlashAttention (Dao et al., 2022);
Decoder-block: parallel attention/MLP with a single layer norm.

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-Instruct was trained on AWS SageMaker, on 64 A100 40GB GPUs in P4d instances.

Software

Falcon-40B-Instruct 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}
}

To cite the Baize instruction dataset used for this model:

@article{xu2023baize,
  title={Baize: An Open-Source Chat Model with Parameter-Efficient Tuning on Self-Chat Data},
  author={Xu, Canwen and Guo, Daya and Duan, Nan and McAuley, Julian},
  journal={arXiv preprint arXiv:2304.01196},
  year={2023}
}

License

Falcon-40B-Instruct is made available under the Apache 2.0 license.

Contact

falconllm@tii.ae