Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them.
Ein Chat zwischen einem Benutzer und einem KI-Assistenten. Der KI-Assistent gibt hilfreiche, detaillierte und höfliche Antworten.
User: {prompt}
Assistant:
Provided files, and GPTQ parameters
Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements.
Each separate quant is in a different branch. See below for instructions on fetching from different branches.
Most GPTQ files are made with AutoGPTQ. Mistral models are currently made with Transformers.
Explanation of GPTQ parameters
Bits: The bit size of the quantised model.
GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value.
Act Order: True or False. Also known as desc_act. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now.
Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy.
GPTQ dataset: The calibration dataset used during quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ calibration dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s).
Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences.
ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama models in 4-bit.
If you remove the --local-dir-use-symlinks False parameter, the files will instead be stored in the central Huggingface cache directory (default location on Linux is: ~/.cache/huggingface), and symlinks will be added to the specified --local-dir, pointing to their real location in the cache. This allows for interrupted downloads to be resumed, and allows you to quickly clone the repo to multiple places on disk without triggering a download again. The downside, and the reason why I don't list that as the default option, is that the files are then hidden away in a cache folder and it's harder to know where your disk space is being used, and to clear it up if/when you want to remove a download model.
The cache location can be changed with the HF_HOME environment variable, and/or the --cache-dir parameter to huggingface-cli.
Note that using Git with HF repos is strongly discouraged. It will be much slower than using huggingface-hub, and will use twice as much disk space as it has to store the model files twice (it stores every byte both in the intended target folder, and again in the .git folder as a blob.)
Example Python code for interfacing with TGI (requires huggingface-hub 0.17.0 or later):
pip3 install huggingface-hub
python
1from huggingface_hub import InferenceClient
23endpoint_url ="https://your-endpoint-url-here"45prompt ="Tell me about AI"6prompt_template=f'''Ein Chat zwischen einem Benutzer und einem KI-Assistenten. Der KI-Assistent gibt hilfreiche, detaillierte und höfliche Antworten.
7User: {prompt}8Assistant:
9'''1011client = InferenceClient(endpoint_url)12response = client.text_generation(prompt,13 max_new_tokens=128,14 do_sample=True,15 temperature=0.7,16 top_p=0.95,17 top_k=40,18 repetition_penalty=1.1)1920print(f"Model output: {response}")
How to use this GPTQ model from Python code
Install the necessary packages
Requires: Transformers 4.33.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later.
shell
1pip3 install transformers optimum
2pip3 install auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/ # Use cu117 if on CUDA 11.7
If you have problems installing AutoGPTQ using the pre-built wheels, install it from source instead:
1from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
23model_name_or_path ="TheBloke/SauerkrautLM-13B-v1-GPTQ"4# To use a different branch, change revision5# For example: revision="gptq-4bit-32g-actorder_True"6model = AutoModelForCausalLM.from_pretrained(model_name_or_path,7 device_map="auto",8 trust_remote_code=False,9 revision="main")1011tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)1213prompt ="Tell me about AI"14prompt_template=f'''Ein Chat zwischen einem Benutzer und einem KI-Assistenten. Der KI-Assistent gibt hilfreiche, detaillierte und höfliche Antworten.
15User: {prompt}16Assistant:
17'''1819print("\n\n*** Generate:")2021input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda()22output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512)23print(tokenizer.decode(output[0]))2425# Inference can also be done using transformers' pipeline2627print("*** Pipeline:")28pipe = pipeline(29"text-generation",30 model=model,31 tokenizer=tokenizer,32 max_new_tokens=512,33 do_sample=True,34 temperature=0.7,35 top_p=0.95,36 top_k=40,37 repetition_penalty=1.138)3940print(pipe(prompt_template)[0]['generated_text'])
Compatibility
The files provided are tested to work with AutoGPTQ, both via Transformers and using AutoGPTQ directly. They should also work with Occ4m's GPTQ-for-LLaMa fork.
ExLlama is compatible with Llama and Mistral models in 4-bit. Please see the Provided Files table above for per-file compatibility.
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 special mentions: Pierre Kircher, Stanislav Ovsiannikov, Michael Levine, Eugene Pentland, Andrey, 준교 김, Randy H, Fred von Graf, Artur Olbinski, Caitlyn Gatomon, terasurfer, Jeff Scroggin, James Bentley, Vadim, Gabriel Puliatti, Harry Royden McLaughlin, Sean Connelly, Dan Guido, Edmond Seymore, Alicia Loh, subjectnull, AzureBlack, Manuel Alberto Morcote, Thomas Belote, Lone Striker, Chris Smitley, Vitor Caleffi, Johann-Peter Hartmann, Clay Pascal, biorpg, Brandon Frisco, sidney chen, transmissions 11, Pedro Madruga, jinyuan sun, Ajan Kanaga, Emad Mostaque, Trenton Dambrowitz, Jonathan Leane, Iucharbius, usrbinkat, vamX, George Stoitzev, Luke Pendergrass, theTransient, Olakabola, Swaroop Kallakuri, Cap'n Zoog, Brandon Phillips, Michael Dempsey, Nikolai Manek, danny, Matthew Berman, Gabriel Tamborski, alfie_i, Raymond Fosdick, Tom X Nguyen, Raven Klaugh, LangChain4j, Magnesian, Illia Dulskyi, David Ziegler, Mano Prime, Luis Javier Navarrete Lozano, Erik Bjäreholt, 阿明, Nathan Dryer, Alex, Rainer Wilmers, zynix, TL, Joseph William Delisle, John Villwock, Nathan LeClaire, Willem Michiel, Joguhyik, GodLy, OG, Alps Aficionado, Jeffrey Morgan, ReadyPlayerEmma, Tiffany J. Kim, Sebastain Graf, Spencer Kim, Michael Davis, webtim, Talal Aujan, knownsqashed, John Detwiler, Imad Khwaja, Deo Leter, Jerry Meng, Elijah Stavena, Rooh Singh, Pieter, SuperWojo, Alexandros Triantafyllidis, Stephen Murray, Ai Maven, ya boyyy, Enrico Ros, Ken Nordquist, Deep Realms, Nicholas, Spiking Neurons AB, Elle, Will Dee, Jack West, RoA, Luke @flexchar, Viktor Bowallius, Derek Yates, Subspace Studios, jjj, Toran Billups, Asp the Wyvern, Fen Risland, Ilya, NimbleBox.ai, Chadd, Nitin Borwankar, Emre, Mandus, Leonard Tan, Kalila, K, Trailburnt, S_X, Cory Kujawski
Thank you to all my generous patrons and donaters!
And thank you again to a16z for their generous grant.
Original model card: VAGO solutions's SauerkrautLM 13B v1
SauerkrautLM
VAGO solutions SauerkrautLM
Introducing SauerkrautLM-v1 - Your German Language Powerhouse!
We are thrilled to unveil our very first release, SauerkrautLM-v1. This remarkable creation marks a significant milestone as it is specifically tailored for the German-speaking community. In a landscape where German language models are scarce, we are proud to offer a solution that fills this void.
What sets SauerkrautLM-v1 apart is its versatility. Whether you are an individual looking to harness its capabilities for personal use or a business seeking to integrate it into your projects, our model is designed to accommodate all. It operates under the LLAMA 2 License, providing you with the freedom to explore its potential in both private and commercial applications.
Performance is at the heart of SauerkrautLM-v1. We put it to the test using a customized version of MT-Bench for the German language, and the results speak volumes. It currently stands as the most robust German Language Model on Hugging Face (based on german mt-bench results), showcasing its exceptional capabilities. Rest assured, this model is here to shine and set new standards. And the best thing is it comes in three different sizes (3B, 7B, 13B) to address your individual needs.
Our model's journey began with meticulous training using an augmented dataset within the QLoRA approach. This is just the beginning of our model series, promising even more innovative and powerful solutions in the future.
Join us on this exciting adventure as we redefine the possibilities of language modeling for the German-speaking world.
SauerkrautLM-v1 is here to empower your language-related endeavors like never before.
SauerkrautLM was trained with mix of German data augmentation and translated data.
We found, that only a simple translation of training data can lead to unnatural German phrasings.
Data augmentation techniques were used to grant grammatical, syntactical correctness and a more natural German wording in our training data.
Training Procedure:
SauerkrautLM-13b-v1 was fine-tuned using QLoRA on 1 A100 80GB with Axolotl.
Trained by: SauerkrautLM-v1 trained by VAGO solutions
Model Type: SauerkrautLM-v1 is an auto-regressive language model based on the transformer architecture
Ein Chat zwischen einem Benutzer und einem KI-Assistenten. Der KI-Assistent gibt hilfreiche, detaillierte und höfliche Antworten.
User: {prompt}
Assistant:
Our models have been meticulously trained on extensive datasets. While we have made diligent efforts to thoroughly screen and eliminate any instances of coarse or inappropriate language from our data, we must inform users that despite our best efforts in data cleansing, the possibility of some such content slipping through cannot be entirely ruled out.
Furthermore, it is important to note that we have implemented filters within our models; however, we cannot always guarantee consistently appropriate behavior. Therefore, if you encounter any issues or come across inappropriate content, we kindly request that you inform us through the contact information provided.
Additionally, it is essential to understand that the licensing of these models does not constitute legal advice. We are not held responsible for the actions of third parties who utilize our models. These models may be employed for commercial purposes, and the original Llama2 license remains applicable and is included with the model files.
Contact
If you are interested in customized LLMs for business applications, please get in contact with us via our website or contact us at Dr. Daryoush Vaziri. We are also grateful for your feedback and suggestions.
Collaborations
We are also keenly seeking support and investment for our startup, VAGO solutions, where we continuously advance the development of robust language models designed to address a diverse range of purposes and requirements. If the prospect of collaboratively navigating future challenges excites you, we warmly invite you to reach out to us.
