AWQ is an efficient, accurate and blazing-fast low-bit weight quantization method, currently supporting 4-bit quantization. Compared to GPTQ, it offers faster Transformers-based inference.

It is also now supported by continuous batching server vLLM, allowing use of AWQ models for high-throughput concurrent inference in multi-user server scenarios. Note that, at the time of writing, overall throughput is still lower than running vLLM with unquantised models, however using AWQ enables using much smaller GPUs which can lead to easier deployment and overall cost savings. For example, a 70B model can be run on 1 x 48GB GPU instead of 2 x 80GB.

Repositories available

Prompt template: Orca-Vicuna

SYSTEM: {system_message}
USER: {prompt}
ASSISTANT:

Provided files and AWQ parameters

For my first release of AWQ models, I am releasing 128g models only. I will consider adding 32g as well if there is interest, and once I have done perplexity and evaluation comparisons, but at this time 32g models are still not fully tested with AutoAWQ and vLLM.

Models are released as sharded safetensors files.

Branch	Bits	GS	AWQ Dataset	Seq Len	Size
main	4	128	wikitext	4096	7.25 GB

Serving this model from vLLM

Documentation on installing and using vLLM can be found here.

When using vLLM as a server, pass the --quantization awq parameter, for example:

python3 python -m vllm.entrypoints.api_server --model TheBloke/Dolphin-Llama-13B-AWQ --quantization awq

When using vLLM from Python code, pass the quantization=awq parameter, for example:

python

from vllm import LLM, SamplingParams

prompts = [
    "Hello, my name is",
    "The president of the United States is",
    "The capital of France is",
    "The future of AI is",
]
sampling_params = SamplingParams(temperature=0.8, top_p=0.95)

llm = LLM(model="TheBloke/Dolphin-Llama-13B-AWQ", quantization="awq")

outputs = llm.generate(prompts, sampling_params)

# Print the outputs.
for output in outputs:
    prompt = output.prompt
    generated_text = output.outputs[0].text
    print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")

How to use this AWQ model from Python code

Install the necessary packages

Requires: AutoAWQ 0.0.2 or later

pip3 install autoawq

If you have problems installing AutoAWQ using the pre-built wheels, install it from source instead:

shell

pip3 uninstall -y autoawq
git clone https://github.com/casper-hansen/AutoAWQ
cd AutoAWQ
pip3 install .

You can then try the following example code

python

from awq import AutoAWQForCausalLM
from transformers import AutoTokenizer

model_name_or_path = "TheBloke/Dolphin-Llama-13B-AWQ"

# Load model
model = AutoAWQForCausalLM.from_quantized(model_name_or_path, fuse_layers=True,
                                          trust_remote_code=False, safetensors=True)
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=False)

prompt = "Tell me about AI"
prompt_template=f'''SYSTEM: {system_message}
USER: {prompt}
ASSISTANT:

'''

print("\n\n*** Generate:")

tokens = tokenizer(
    prompt_template,
    return_tensors='pt'
).input_ids.cuda()

# Generate output
generation_output = model.generate(
    tokens,
    do_sample=True,
    temperature=0.7,
    top_p=0.95,
    top_k=40,
    max_new_tokens=512
)

print("Output: ", tokenizer.decode(generation_output[0]))

# Inference can also be done using transformers' pipeline
from transformers import pipeline

print("*** Pipeline:")
pipe = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    max_new_tokens=512,
    do_sample=True,
    temperature=0.7,
    top_p=0.95,
    top_k=40,
    repetition_penalty=1.1
)

print(pipe(prompt_template)[0]['generated_text'])

Compatibility

The files provided are tested to work with AutoAWQ, and vLLM.

Huggingface Text Generation Inference (TGI) is not yet compatible with AWQ, but a PR is open which should bring support soon: TGI PR #781.

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!

Thanks to Clay from gpus.llm-utils.org!

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: Aemon Algiz.

Patreon special mentions: Alicia Loh, Stephen Murray, K, Ajan Kanaga, RoA, Magnesian, Deo Leter, Olakabola, Eugene Pentland, zynix, Deep Realms, Raymond Fosdick, Elijah Stavena, Iucharbius, Erik Bjäreholt, Luis Javier Navarrete Lozano, Nicholas, theTransient, John Detwiler, alfie_i, knownsqashed, Mano Prime, Willem Michiel, Enrico Ros, LangChain4j, OG, Michael Dempsey, Pierre Kircher, Pedro Madruga, James Bentley, Thomas Belote, Luke @flexchar, Leonard Tan, Johann-Peter Hartmann, Illia Dulskyi, Fen Risland, Chadd, S_X, Jeff Scroggin, Ken Nordquist, Sean Connelly, Artur Olbinski, Swaroop Kallakuri, Jack West, Ai Maven, David Ziegler, Russ Johnson, transmissions 11, John Villwock, Alps Aficionado, Clay Pascal, Viktor Bowallius, Subspace Studios, Rainer Wilmers, Trenton Dambrowitz, vamX, Michael Levine, 준교 김, Brandon Frisco, Kalila, Trailburnt, Randy H, Talal Aujan, Nathan Dryer, Vadim, 阿明, ReadyPlayerEmma, Tiffany J. Kim, George Stoitzev, Spencer Kim, Jerry Meng, Gabriel Tamborski, Cory Kujawski, Jeffrey Morgan, Spiking Neurons AB, Edmond Seymore, Alexandros Triantafyllidis, Lone Striker, Cap'n Zoog, Nikolai Manek, danny, ya boyyy, Derek Yates, usrbinkat, Mandus, TL, Nathan LeClaire, subjectnull, Imad Khwaja, webtim, Raven Klaugh, Asp the Wyvern, Gabriel Puliatti, Caitlyn Gatomon, Joseph William Delisle, Jonathan Leane, Luke Pendergrass, SuperWojo, Sebastain Graf, Will Dee, Fred von Graf, Andrey, Dan Guido, Daniel P. Andersen, Nitin Borwankar, Elle, Vitor Caleffi, biorpg, jjj, NimbleBox.ai, Pieter, Matthew Berman, terasurfer, Michael Davis, Alex, Stanislav Ovsiannikov

Thank you to all my generous patrons and donaters!

And thank you again to a16z for their generous grant.

Original model card: Eric Hartford's Dolphin Llama 13B

Dolphin 🐬 https://erichartford.com/dolphin

This model is based on llama1, so it is for non-commercial use only. Future versions will be trained on llama2 and other open models that are suitable for commercial use.

This model is uncensored. I have filtered the dataset to remove alignment and bias. This makes the model compliant to any requests. You are advised to implement your own alignment layer before exposing the model as a service. It will be highly compliant to any requests, even unethical ones. Please read my blog post about uncensored models. https://erichartford.com/uncensored-models You are responsible for any content you create using this model. Enjoy responsibly.

Dataset

This dataset is an open source implementation of Microsoft's Orca

After uncensoring, deduping, and cleaning, our dataset consists of:

842,610 instructions of FLANv2 augmented with GPT-4 completions
2,625,353 instructions of FLANv2 augmented with GPT-3.5 completions

We followed the submix and system prompt distribution outlined in the Orca paper. With a few exceptions. We included all 75k of CoT in the FLAN-1m dataset rather than sampling that. Also, we found that many items were duplicated, so we removed duplicates.

Then we filtered out instances of alignment, refusal, avoidance, and bias, in order to produce an uncensored model upon which can be layered your personalized alignment LoRA.

We also filtered out duplicates and cleaned the data.

Training

We trained with the flan5m (gpt3.5 completions) dataset in its entirety for 3 epochs at a learning rate of 2e-5 before we stopped training to avoid overfit. We trained with the flan1m (gpt4 completions) dataset in its entirety for 2.5 epochs at a learning rate of 1e-5 before we stopped training to avoid overfit. It took about 600 hours to train on 8x H100s We used a prompt format similar to Vicuna, but we added the SYSTEM: field.

Prompt format:

SYSTEM: {system}
USER: {prompt}
ASSISTANT:

Example:

SYSTEM: you are an expert marine biologist.
USER: Please list 10 ways that dolphins are superior to orcas.
ASSISTANT:

Evaluation

Evaluation will be coming soon.

Team

The core Dolphin Team includes:

Eric "Faldore" Hartford
Pankaj Mathur
Rob "Rohan" O'Callahan
Tom "TheBloke" Jobbins

Gratitude

Thank you to Microsoft for authoring the Orca paper and inspiring this work.
Special thanks to WingLian, NanoBit, Teknium for helpful advice
Special thanks to EdenCoder and chirper.ai for mentorship and financial sponsorship.
Special thanks to Kilkonie for his very valued mentorship.
Thank you to Catto.
Thank you to Nicolai Schleifer, financial sponsor.
Thank you to Eric Fleming, financial sponsor.
Thank you to all the other people in the Open Source AI community who have taught me and helped me along the way.
And biggest thanks and love to my wife Hope and my babies, who have endured two months of my absenteeism while working long days and nights on this project.

Example Output

SYSTEM: you are an expert marine biologist.
USER: Please list 10 ways that dolphins are superior to orcas.
ASSISTANT:

Social structure: Dolphins live in complex social groups called pods, which consist of extended family units and can include hundreds of individuals. Orcas, on the other hand, typically form smaller, more temporary groupings known as clans or matrilineal kinship groups.
Communication: Dolphins have a highly developed system of vocalizations, body language, and even telepathy to communicate with each other for various purposes such as coordinating cooperative hunting strategies, sharing information about food sources, and maintaining social bonds. While orcas also exhibit some communication abilities, their methods are generally less sophisticated compared to those of dolphins.
Intelligence: Dolphins are widely regarded as one of the most intelligent animal species, showcasing advanced problem-solving skills, self-awareness, tool use (such as using sea sponges as tools), and learning capabilities. Some dolphin species have been observed demonstrating altruism and understanding of human emotions. In contrast, while orcas are also smart, they do not display the same level of cognitive complexity as dolphins.
Playfulness: Dolphins are well-known for their playful behavior, often engaging in acrobatic displays, chasing each other's tails, and playing with objects like balls or seaweed. This playful nature is thought to be essential for their social bonding and mental stimulation. Although orcas may occasionally interact playfully, it is not as prevalent or diverse as among dolphins.
Agility and speed: Dolphins are incredibly agile swimmers, capable of reaching speeds up to 30 miles per hour (48 km/h) when pursuing prey. They can change direction rapidly and maneuver through tight spaces with ease. While orcas are powerful predators and fast swimmers themselves, they cannot match the sheer acceleration and agility of dolphins.
Feeding habits: Both dolphins and orcas are carnivorous, feeding primarily on fish and cephalopods. However, dolphins employ a wider variety of hunting techniques, including cooperative hunting strategies involving herding, encircling, and driving schools of fish into shallow waters where they become disoriented and easier to catch. Orcas mainly rely on ambushing their prey from below and using their powerful jaws to strike quickly.
Reproduction: Female dolphins usually give birth to a single calf after a gestation period of around 12 months. The mother plays a crucial role in raising and nurturing her offspring until they become independent. On the other hand, female orcas tend to breed every three to five years, giving birth to multiple offspring at once, and relying heavily on their mothers for protection and teaching.
Vocal repertoire: Dolphins possess a wide range of sounds, including clicks, whistles, and body sounds, which allow them to convey different messages and establish unique identities within their pods. Orcas, too, produce vocalizations but lack the diversity found in dolphin calls.
Conservation status: Many dolphin species face threats due to habitat loss, pollution, overfishing, and climate change. As a result, several populations are considered vulnerable or endangered by the International Union for Conservation of Nature (IUCN). In comparison, orca conservation status varies depending on the region; however, no species is currently listed as critically endangered or facing extinction.
Adaptability: Dolphins have adapted to various aquatic environments, ranging from warm coastal waters to colder open oceans. Their versatility allows them to thrive in both tropical and temperate climates. Orcas, conversely, are adapted to specific habitats such as cooler coastal waters and are mostly found in the Northern Hemisphere.