base_model: VAGOsolutions/SauerkrautLM-7b-HerO
inference: false
language:
- en
- de
library_name: transformers
license: apache-2.0
model_creator: VAGO solutions
model_name: SauerkrautLM 7B HerO
model_type: mistral
pipeline_tag: text-generation
prompt_template: |
<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
quantized_by: TheBloke
tags:
- mistral
- finetune
- chatml
- augmentation
- german
TheBloke's LLM work is generously supported by a grant from andreessen horowitz (a16z)
SauerkrautLM 7B HerO - GPTQ
- Model creator: VAGO solutions
- Original model: SauerkrautLM 7B HerO
Description
This repo contains GPTQ model files for VAGO solutions's SauerkrautLM 7B HerO.
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.
These files were quantised using hardware kindly provided by Massed Compute.
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
- VAGO solutions's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions
Prompt template: ChatML
<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
Known compatible clients / servers
These GPTQ models are known to work in the following inference servers/webuis.
This may not be a complete list; if you know of others, please let me know!
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 and Mistral models in 4-bit.
| Branch | Bits | GS | Act Order | Damp % | GPTQ Dataset | Seq Len | Size | ExLlama | Desc |
|---|---|---|---|---|---|---|---|---|---|
| main | 4 | 128 | Yes | 0.1 | German Quad | 4096 | 4.16 GB | Yes | 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy. |
| gptq-4bit-32g-actorder_True | 4 | 32 | Yes | 0.1 | German Quad | 4096 | 4.57 GB | Yes | 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage. |
| gptq-8bit--1g-actorder_True | 8 | None | Yes | 0.1 | German Quad | 4096 | 7.52 GB | No | 8-bit, with Act Order. No group size, to lower VRAM requirements. |
| gptq-8bit-128g-actorder_True | 8 | 128 | Yes | 0.1 | German Quad | 4096 | 7.68 GB | No | 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy. |
| gptq-8bit-32g-actorder_True | 8 | 32 | Yes | 0.1 | German Quad | 4096 | 8.17 GB | No | 8-bit, with group size 32g and Act Order for maximum inference quality. |
| gptq-4bit-64g-actorder_True | 4 | 64 | Yes | 0.1 | German Quad | 4096 | 4.30 GB | Yes | 4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy. |
How to download, including from branches
In text-generation-webui
To download from the main branch, enter TheBloke/SauerkrautLM-7B-HerO-GPTQ in the "Download model" box.
To download from another branch, add :branchname to the end of the download name, eg TheBloke/SauerkrautLM-7B-HerO-GPTQ:gptq-4bit-32g-actorder_True
From the command line
I recommend using the huggingface-hub Python library:
pip3 install huggingface-hub
To download the main branch to a folder called SauerkrautLM-7B-HerO-GPTQ:
mkdir SauerkrautLM-7B-HerO-GPTQ
huggingface-cli download TheBloke/SauerkrautLM-7B-HerO-GPTQ --local-dir SauerkrautLM-7B-HerO-GPTQ --local-dir-use-symlinks False
To download from a different branch, add the --revision parameter:
mkdir SauerkrautLM-7B-HerO-GPTQ
huggingface-cli download TheBloke/SauerkrautLM-7B-HerO-GPTQ --revision gptq-4bit-32g-actorder_True --local-dir SauerkrautLM-7B-HerO-GPTQ --local-dir-use-symlinks False
More advanced huggingface-cli download usage
If you remove the --local-dir-use-symlinks False parameter, the files will instead be stored in the central Hugging Face 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.
For more documentation on downloading with huggingface-cli, please see: HF -> Hub Python Library -> Download files -> Download from the CLI.
To accelerate downloads on fast connections (1Gbit/s or higher), install hf_transfer:
pip3 install hf_transfer
And set environment variable HF_HUB_ENABLE_HF_TRANSFER to 1:
mkdir SauerkrautLM-7B-HerO-GPTQ
HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/SauerkrautLM-7B-HerO-GPTQ --local-dir SauerkrautLM-7B-HerO-GPTQ --local-dir-use-symlinks False
Windows Command Line users: You can set the environment variable by running set HF_HUB_ENABLE_HF_TRANSFER=1 before the download command.
With git (not recommended)
To clone a specific branch with git, use a command like this:
git clone --single-branch --branch gptq-4bit-32g-actorder_True https://huggingface.co/TheBloke/SauerkrautLM-7B-HerO-GPTQ
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.)
How to easily download and use this model in text-generation-webui
Please make sure you're using the latest version of text-generation-webui.
It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install.
Click the Model tab.
Under Download custom model or LoRA, enter
TheBloke/SauerkrautLM-7B-HerO-GPTQ.- To download from a specific branch, enter for example
TheBloke/SauerkrautLM-7B-HerO-GPTQ:gptq-4bit-32g-actorder_True - see Provided Files above for the list of branches for each option.
- To download from a specific branch, enter for example
Click Download.
The model will start downloading. Once it's finished it will say "Done".
In the top left, click the refresh icon next to Model.
In the Model dropdown, choose the model you just downloaded:
SauerkrautLM-7B-HerO-GPTQThe model will automatically load, and is now ready for use!
If you want any custom settings, set them and then click Save settings for this model followed by Reload the Model in the top right.
- Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file
quantize_config.json.
- Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file
Once you're ready, click the Text Generation tab and enter a prompt to get started!
Serving this model from Text Generation Inference (TGI)
It's recommended to use TGI version 1.1.0 or later. The official Docker container is: ghcr.io/huggingface/text-generation-inference:1.1.0
Example Docker parameters:
--model-id TheBloke/SauerkrautLM-7B-HerO-GPTQ --port 3000 --quantize gptq --max-input-length 3696 --max-total-tokens 4096 --max-batch-prefill-tokens 4096
Example Python code for interfacing with TGI (requires huggingface-hub 0.17.0 or later):
pip3 install huggingface-hub
from huggingface_hub import InferenceClient
endpoint_url = "https://your-endpoint-url-here"
prompt = "Tell me about AI"
prompt_template=f'''<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
'''
client = InferenceClient(endpoint_url)
response = client.text_generation(prompt,
max_new_tokens=128,
do_sample=True,
temperature=0.7,
top_p=0.95,
top_k=40,
repetition_penalty=1.1)
print(f"Model output: {response}")
Python code example: inference from this GPTQ model
Install the necessary packages
Requires: Transformers 4.33.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later.
pip3 install --upgrade transformers optimum
# If using PyTorch 2.1 + CUDA 12.x:
pip3 install --upgrade auto-gptq
# or, if using PyTorch 2.1 + CUDA 11.x:
pip3 install --upgrade auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/
If you are using PyTorch 2.0, you will need to install AutoGPTQ from source. Likewise if you have problems with the pre-built wheels, you should try building from source:
pip3 uninstall -y auto-gptq
git clone https://github.com/PanQiWei/AutoGPTQ
cd AutoGPTQ
git checkout v0.5.1
pip3 install .
Example Python code
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
model_name_or_path = "TheBloke/SauerkrautLM-7B-HerO-GPTQ"
# To use a different branch, change revision
# For example: revision="gptq-4bit-32g-actorder_True"
model = AutoModelForCausalLM.from_pretrained(model_name_or_path,
device_map="auto",
trust_remote_code=False,
revision="main")
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True)
prompt = "Tell me about AI"
prompt_template=f'''<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant
'''
print("\n\n*** Generate:")
input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda()
output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512)
print(tokenizer.decode(output[0]))
# Inference can also be done using transformers' 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 Transformers. For non-Mistral models, AutoGPTQ can also be used directly.
ExLlama is compatible with Llama and Mistral models in 4-bit. Please see the Provided Files table above for per-file compatibility.
For a list of clients/servers, please see "Known compatible clients / servers", above.
Discord
For further support, and discussions on these models and AI in general, join us at:
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: Brandon Frisco, LangChain4j, Spiking Neurons AB, transmissions 11, Joseph William Delisle, Nitin Borwankar, Willem Michiel, Michael Dempsey, vamX, Jeffrey Morgan, zynix, jjj, Omer Bin Jawed, Sean Connelly, jinyuan sun, Jeromy Smith, Shadi, Pawan Osman, Chadd, Elijah Stavena, Illia Dulskyi, Sebastain Graf, Stephen Murray, terasurfer, Edmond Seymore, Celu Ramasamy, Mandus, Alex, biorpg, Ajan Kanaga, Clay Pascal, Raven Klaugh, 阿明, K, ya boyyy, usrbinkat, Alicia Loh, John Villwock, ReadyPlayerEmma, Chris Smitley, Cap'n Zoog, fincy, GodLy, S_X, sidney chen, Cory Kujawski, OG, Mano Prime, AzureBlack, Pieter, Kalila, Spencer Kim, Tom X Nguyen, Stanislav Ovsiannikov, Michael Levine, Andrey, Trailburnt, Vadim, Enrico Ros, Talal Aujan, Brandon Phillips, Jack West, Eugene Pentland, Michael Davis, Will Dee, webtim, Jonathan Leane, Alps Aficionado, Rooh Singh, Tiffany J. Kim, theTransient, Luke @flexchar, Elle, Caitlyn Gatomon, Ari Malik, subjectnull, Johann-Peter Hartmann, Trenton Dambrowitz, Imad Khwaja, Asp the Wyvern, Emad Mostaque, Rainer Wilmers, Alexandros Triantafyllidis, Nicholas, Pedro Madruga, SuperWojo, Harry Royden McLaughlin, James Bentley, Olakabola, David Ziegler, Ai Maven, Jeff Scroggin, Nikolai Manek, Deo Leter, Matthew Berman, Fen Risland, Ken Nordquist, Manuel Alberto Morcote, Luke Pendergrass, TL, Fred von Graf, Randy H, Dan Guido, NimbleBox.ai, Vitor Caleffi, Gabriel Tamborski, knownsqashed, Lone Striker, Erik Bjäreholt, John Detwiler, Leonard Tan, Iucharbius
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 7B HerO
VAGO solutions SauerkrautLM-7b-HerO
Introducing SauerkrautLM-7b-HerO – the pinnacle of German language model technology! Crafted through the merging of Teknium's OpenHermes-2.5-Mistral-7B and Open-Orca's Mistral-7B-OpenOrca and uniquely fine-tuned with the Sauerkraut dataset. SauerkrautLM-7b-HerO represents a breakthrough in language modeling, achieving an optimal balance between extensive German data and essential international sources. This ensures the model not only excels in understanding the nuances of the German language but also retains its global capabilities. Harnessing the innovative power of the gradient SLERP method from MergeKit, we've achieved a groundbreaking fusion of two of the most best performing 7B models based on the Mistral framework. This merge has allowed us to combine the best features of both models, creating an unparalleled synergy. Coupled with the German Sauerkraut dataset, which consists of a mix of augmented and translated data, we have successfully taught the English-speaking merged model the intricacies of the German language. This was achieved without the typical loss of core competencies often associated with fine-tuning in another language of models previously trained mainly in English. Our approach ensures that the model retains its original strengths while acquiring a profound understanding of German, setting a new benchmark in bilingual language model proficiency.
Table of Contents
- Overview of all Her0 models
- Model Details
- Evaluation
- Disclaimer
- Contact
- Collaborations
- Acknowledgement
All HerO Models
| Model | HF | GPTQ | GGUF | AWQ |
|---|---|---|---|---|
| SauerkrautLM-7b-HerO | Link | coming soon | coming soon | coming soon |
Model Details
SauerkrautLM-7b-HerO
- Model Type: SauerkrautLM-7b-HerO is an auto-regressive language model based on the transformer architecture
- Language(s): English, German
- License: APACHE 2.0
- Contact: Website David Golchinfar
Training Dataset:
SauerkrautLM-7b-HerO 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.
Merge Procedure:
SauerkrautLM-7b-HerO was merged on 1 A100 with mergekit. The merged model contains OpenHermes-2.5-Mistral-7B and Open-Orca/Mistral-7B-OpenOrca. We applied the gradient SLERP method.
Prompt Template:
<|im_start|>system
Du bist Sauerkraut-HerO, ein großes Sprachmodell, das höflich und kompetent antwortet. Schreibe deine Gedanken Schritt für Schritt auf, um Probleme sinnvoll zu lösen.<|im_end|>
<|im_start|>user
Wie geht es dir?<|im_end|>
<|im_start|>assistant
Mir geht es gut!<|im_end|>
<|im_start|>user
Bitte erkläre mir, wie die Zusammenführung von Modellen durch bestehende Spitzenmodelle profitieren kann.<|im_end|>
<|im_start|>assistant
Evaluation
GPT4ALL:
Compared to relevant German Closed and Open Source models
Language Model evaluation Harness:
Compared to Aleph Alpha Luminous Models
*performed with newest Language Model Evaluation Harness
Big Bench:
*performed with newest Language Model Evaluation Harness
MMLU:
Compared to Big Boy LLMs (Grok0,Grok1,GPT3.5,GPT4)
TruthfulQA:
Compared to OpenAI Models (GPT3.5,GPT4)
MT-Bench (German):
########## First turn ##########
score
model turn
SauerkrautLM-70b-v1 1 7.25000
SauerkrautLM-7b-HerO <--- 1 6.96875
SauerkrautLM-7b-v1-mistral 1 6.30625
leo-hessianai-13b-chat 1 6.18750
SauerkrautLM-13b-v1 1 6.16250
leo-mistral-hessianai-7b-chat 1 6.15625
Llama-2-70b-chat-hf 1 6.03750
vicuna-13b-v1.5 1 5.80000
SauerkrautLM-7b-v1 1 5.65000
leo-hessianai-7b-chat 1 5.52500
vicuna-7b-v1.5 1 5.42500
Mistral-7B-v0.1 1 5.37500
SauerkrautLM-3b-v1 1 3.17500
Llama-2-7b 1 1.28750
open_llama_3b_v2 1 1.68750
########## Second turn ##########
score
model turn
SauerkrautLM-70b-v1 2 6.83125
SauerkrautLM-7b-HerO <--- 2 6.30625
vicuna-13b-v1.5 2 5.63125
SauerkrautLM-13b-v1 2 5.34375
SauerkrautLM-7b-v1-mistral 2 5.26250
leo-mistral-hessianai-7b-chat 2 4.99375
SauerkrautLM-7b-v1 2 4.73750
leo-hessianai-13b-chat 2 4.71250
vicuna-7b-v1.5 2 4.67500
Llama-2-70b-chat-hf 2 4.66250
Mistral-7B-v0.1 2 4.53750
leo-hessianai-7b-chat 2 2.65000
SauerkrautLM-3b-v1 2 1.98750
open_llama_3b_v2 2 1.22500
Llama-2-7b 2 1.07500
########## Average ##########
score
model
SauerkrautLM-70b-v1 7.040625
SauerkrautLM-7b-HerO <--- 6.637500
SauerkrautLM-7b-v1-mistral 5.784375
SauerkrautLM-13b-v1 5.753125
vicuna-13b-v1.5 5.715625
leo-mistral-hessianai-7b-chat 5.575000
leo-hessianai-13b-chat 5.450000
Llama-2-70b-chat-hf 5.350000
SauerkrautLM-v1-7b 5.193750
vicuna-7b-v1.5 5.050000
Mistral-7B-v0.1 4.956250
leo-hessianai-7b-chat 4.087500
SauerkrautLM-3b-v1 2.581250
open_llama_3b_v2 1.456250
Llama-2-7b 1.181250
*performed with the newest FastChat Version
MT-Bench (English):
########## First turn ##########
score
model turn
OpenHermes-2.5-Mistral-7B 1 8.21875
SauerkrautLM-7b-HerO <--- 1 8.03125
Mistral-7B-OpenOrca 1 7.65625
neural-chat-7b-v3-1 1 7.22500
########## Second turn ##########
score
model turn
OpenHermes-2.5-Mistral-7B 2 7.1000
SauerkrautLM-7b-HerO <--- 2 6.7875
neural-chat-7b-v3-1 2 6.4000
Mistral-7B-OpenOrca 2 6.1750
########## Average ##########
score
model
OpenHermes-2.5-Mistral-7B 7.659375
SauerkrautLM-7b-HerO <--- 7.409375
Mistral-7B-OpenOrca 6.915625
neural-chat-7b-v3-1 6.812500
*performed with the newest FastChat Version
Additional German Benchmark results:
*performed with newest Language Model Evaluation Harness
Disclaimer
We must inform users that despite our best efforts in data cleansing, the possibility of uncensored content slipping through cannot be entirely ruled out. However, we cannot 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 Apache 2.0 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.
Acknowledgement
Many thanks to OpenOrca and teknium for providing such valuable models to the Open-Source community.
