OpenAssistant-Llama2-13B-Orca-8K-3319-AWQ

<!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end -->

Llama2 13B Orca 8K 3319 - AWQ

• Model creator: OpenAssistant
• Original model: Llama2 13B Orca 8K 3319

<!-- description start -->

Description

This repo contains AWQ model files for OpenAssistant's Llama2 13B Orca 8K 3319.

About AWQ

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.

<!-- description end --> <!-- repositories-available start -->

Repositories available

• AWQ model(s) for GPU inference.
• GPTQ models for GPU inference, with multiple quantisation parameter options.
• 2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference
• OpenAssistant's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions

<!-- repositories-available end --> <!-- prompt-template start -->

Prompt template: OpenAssistant-System

<|system|>{system_message}</s><|prompter|>{prompt}</s><|assistant|>

<!-- prompt-template end --> <!-- licensing start -->

Licensing

The creator of the source model has listed its license as other, and this quantization has therefore used that same license.

As this model is based on Llama 2, it is also subject to the Meta Llama 2 license terms, and the license files for that are additionally included. It should therefore be considered as being claimed to be licensed under both licenses. I contacted Hugging Face for clarification on dual licensing but they do not yet have an official position. Should this change, or should Meta provide any feedback on this situation, I will update this section accordingly.

In the meantime, any questions regarding licensing, and in particular how these two licenses might interact, should be directed to the original model repository: OpenAssistant's Llama2 13B Orca 8K 3319.

<!-- licensing end --> <!-- README_AWQ.md-provided-files start -->

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.

<!-- README_AWQ.md-provided-files end --> <!-- README_AWQ.md-use-from-vllm start -->

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/OpenAssistant-Llama2-13B-Orca-8K-3319-AWQ --quantization awq

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

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/OpenAssistant-Llama2-13B-Orca-8K-3319-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}")

<!-- README_AWQ.md-use-from-vllm start --> <!-- README_AWQ.md-use-from-python start -->

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:

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

You can then try the following example code

from awq import AutoAWQForCausalLM
from transformers import AutoTokenizer

model_name_or_path = "TheBloke/OpenAssistant-Llama2-13B-Orca-8K-3319-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}</s><|prompter|>{prompt}</s><|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'])

<!-- README_AWQ.md-use-from-python end --> <!-- README_AWQ.md-compatibility start -->

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.

<!-- README_AWQ.md-compatibility end --> <!-- footer start --> <!-- 200823 -->

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.

<!-- footer end -->

Original model card: OpenAssistant's Llama2 13B Orca 8K 3319

llama2-13b-orca-8k-3319

Model Description

This model is a fine-tuning of Meta's Llama2 13B model with 8K context size on a long-conversation variant of the Dolphin dataset (orca-chat).

Note: At least Huggingface Transformers 4.31.0 is required to load this model!

Usage

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("OpenAssistant/llama2-13b-orca-8k-3319", use_fast=False)
model = AutoModelForCausalLM.from_pretrained("OpenAssistant/llama2-13b-orca-8k-3319", torch_dtype=torch.float16, low_cpu_mem_usage=True, device_map="auto")

system_message = "You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information."
user_prompt = "Write me a poem please"
prompt = f"""<|system|>{system_message}</s><|prompter|>{user_prompt}</s><|assistant|>"""
inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
output = model.generate(**inputs, do_sample=True, top_p=0.95, top_k=0, max_new_tokens=256)
print(tokenizer.decode(output[0], skip_special_tokens=True))

Model Details

• base model: meta-llama/Llama-2-13b
• License: Llama 2 Community License Agreement
• sampling report: 2023-07-25_OpenAssistant_llama2-13b-orca-8k-3319_sampling_llama2_prompt.json
• wandb: public-sft/runs/2jfazjt9
• checkpoint: 3319 steps
• datatpye: fp16
• sponsored by: Redmond.ai

Long context (RoPE Scaling)

This model was fine-tuned with a context size of 8192 tokens using linear scaling of RoPE embeddings. This feature was recently added to Huggingface transformers. Before loading this model please make sure HF transformers >=4.31.0 is installed (pip install transformers>=4.31.0).

Conversation Template

For the initial response use (e.g. the llama2 default system prompt works well):

<|system|>system message</s><|prompter|>user prompt</s><|assistant|>

For multi-turn conversations use:

<|system|>system message</s><|prompter|>Q1</s><|assistant|>A1</s><|prompter|>Q2</s><|assistant|>

The model was trained with the following 15 system messages used to generate the training examples (see ORCA paper):

1. You are an AI assistant. Provide a detailed answer so user don’t need to search outside to understand the answer.
2. You are an AI assistant. You will be given a task. You must generate a detailed and long answer.
3. You are a helpful assistant, who always provide explanation. Think like you are answering to a five year old.
4. You are an AI assistant that follows instruction extremely well. Help as much as you can.
5. You are an AI assistant that helps people find information. Provide a detailed answer so user don’t need to search outside to understand the answer.
6. You are an AI assistant. User will you give you a task. Your goal is to complete the task as faithfully as you can. While performing the task think step-by-step and justify your steps.
7. You should describe the task and explain your answer. While answering a multiple choice question, first output the correct answer(s). Then explain why other answers are wrong. Think like you are answering to a five year old.
8. Explain how you used the definition to come up with the answer.
9. You are an AI assistant. You should describe the task and explain your answer. While answering a multiple choice question, first output the correct answer(s). Then explain why other answers are wrong. You might need to use additional knowledge to answer the question.
10. You are an AI assistant that helps people find information. User will you give you a question. Your task is to answer as faithfully as you can. While answering think step-by- step and justify your answer.
11. User will you give you a task with some instruction. Your job is follow the instructions as faithfully as you can. While answering think step-by-step and justify your answer.
12. You are a teacher. Given a task, you explain in simple steps what the task is asking, any guidelines it provides and how to use those guidelines to find the answer.
13. You are an AI assistant, who knows every language and how to translate one language to another. Given a task, you explain in simple steps what the task is asking, any guidelines that it provides. You solve the task and show how you used the guidelines to solve the task.
14. Given a definition of a task and a sample input, break the definition into small parts. Each of those parts will have some instruction. Explain their meaning by showing an example that meets the criteria in the instruction. Use the following format: Part #: a key part of the definition. Usage: Sample response that meets the criteria from the key part. Explain why you think it meets the criteria.
15. You are an AI assistant that helps people find information.

Datasets: Orca-Chat/Dolphin, RedPajama1T & FanFics

This model was trained on:

• shahules786/orca-chat
• togethercomputer/RedPajama-Data-1T-Sample
• atom-in-the-universe/fanfics-10k-50k

Dataset Composition:
    Tain (sampled):
       orca-chat: 188842 (100%)
       fanfics: 47760 (100%)
       red_pajama: 188262 (25%)
    Valid:
       orca-chat: 5000
       fanfics: 1000
       red_pajama: 1000

The dataset shahules786/orca-chat combines similar examples of the GPT-4 subset of ehartford/dolphin to form longer conversations to improve long-context training.

Additionally, RedPajama and FanFics were used for classic language modelling as an auxiliary task to improve the RoPE scaling for the 8k context size.

Model Configuration

llama2_13b_orca_8k:
  rng_seed: 0xe1291f1a
  use_custom_sampler: true
  sort_by_length: false
  dtype: fp16
  log_dir: "llama2_log_13b_orca_8k"
  learning_rate: 1e-5
  model_name: /mnt/data/llama2/Llama-2-13b-hf/
  output_dir: llama2_13b_orca_8k
  deepspeed_config: configs/zero_config_pretrain.json
  weight_decay: 0.0
  max_length: 8192
  warmup_steps: 100
  use_flash_attention: true
  gradient_checkpointing: true
  gradient_accumulation_steps: 8
  per_device_train_batch_size: 2
  per_device_eval_batch_size: 1
  residual_dropout: 0.0
  eval_steps: 200
  save_steps: 1000  # (total steps: 3319)
  num_train_epochs: 1
  save_total_limit: 4
  superhot: true
  superhot_config:
    type: linear
    scale: 2
  datasets:
    - orca-chat:
        max_val_set: 5000
    - fanfics:
        max_chunk_size: 65535
        max_val_set: 1000
    - red_pajama:
        fraction: 0.25
        max_val_set: 1000
        max_chunk_size: 65535
  peft_model: false

Developers

• shahules786
• jordiclive
• andreaskoepf

Special Thanks

We want to especially thank Eric Hartford who spared no expense in replicating ORCA and making it available at ehartford/dolphin! Also, shoutout to the whole team working on LLongMA-2-13b & the scaled-rope repository for their awesome work: bloc97, jquesnelle & conceptofmind!

The whole Open-Assistant team is very grateful for the continued support of Redmond.ai who sponsored the training compute required for this model.

License

• Llama 2 is licensed under the LLAMA 2 Community License, Copyright © Meta Platforms, Inc. All Rights Reserved.
• Your use of the Llama Materials must comply with applicable laws and regulations (including trade compliance laws and regulations) and adhere to the Acceptable Use Policy for the Llama Materials.