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.
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.
8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy.
How to download, including from branches
In text-generation-webui
To download from the main branch, enter TheBloke/CaPlatTessDolXaBoros-Yi-34B-200K-DARE-Ties-HighDensity-GPTQ in the "Download model" box.
To download from another branch, add :branchname to the end of the download name, eg TheBloke/CaPlatTessDolXaBoros-Yi-34B-200K-DARE-Ties-HighDensity-GPTQ:gptq-4bit-128g-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 CaPlatTessDolXaBoros-Yi-34B-200K-DARE-Ties-HighDensity-GPTQ:
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.
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.)
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/CaPlatTessDolXaBoros-Yi-34B-200K-DARE-Ties-HighDensity-GPTQ.
To download from a specific branch, enter for example TheBloke/CaPlatTessDolXaBoros-Yi-34B-200K-DARE-Ties-HighDensity-GPTQ:gptq-4bit-128g-actorder_True
see Provided Files above for the list of branches for each option.
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: CaPlatTessDolXaBoros-Yi-34B-200K-DARE-Ties-HighDensity-GPTQ
The 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.
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
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.
shell
1pip3 install --upgrade transformers optimum
2# If using PyTorch 2.1 + CUDA 12.x:3pip3 install --upgrade auto-gptq
4# or, if using PyTorch 2.1 + CUDA 11.x:5pip3 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:
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: Michael Levine, 阿明, Trailburnt, Nikolai Manek, John Detwiler, Randy H, Will Dee, Sebastain Graf, NimbleBox.ai, Eugene Pentland, Emad Mostaque, Ai Maven, Jim Angel, Jeff Scroggin, Michael Davis, Manuel Alberto Morcote, Stephen Murray, Robert, Justin Joy, Luke @flexchar, Brandon Frisco, Elijah Stavena, S_X, Dan Guido, Undi ., Komninos Chatzipapas, Shadi, theTransient, Lone Striker, Raven Klaugh, jjj, Cap'n Zoog, Michel-Marie MAUDET (LINAGORA), Matthew Berman, David, Fen Risland, Omer Bin Jawed, Luke Pendergrass, Kalila, OG, Erik Bjäreholt, Rooh Singh, Joseph William Delisle, Dan Lewis, TL, John Villwock, AzureBlack, Brad, Pedro Madruga, Caitlyn Gatomon, K, jinyuan sun, Mano Prime, Alex, Jeffrey Morgan, Alicia Loh, Illia Dulskyi, Chadd, transmissions 11, fincy, Rainer Wilmers, ReadyPlayerEmma, knownsqashed, Mandus, biorpg, Deo Leter, Brandon Phillips, SuperWojo, Sean Connelly, Iucharbius, Jack West, Harry Royden McLaughlin, Nicholas, terasurfer, Vitor Caleffi, Duane Dunston, Johann-Peter Hartmann, David Ziegler, Olakabola, Ken Nordquist, Trenton Dambrowitz, Tom X Nguyen, Vadim, Ajan Kanaga, Leonard Tan, Clay Pascal, Alexandros Triantafyllidis, JM33133, Xule, vamX, ya boyyy, subjectnull, Talal Aujan, Alps Aficionado, wassieverse, Ari Malik, James Bentley, Woland, Spencer Kim, Michael Dempsey, Fred von Graf, Elle, zynix, William Richards, Stanislav Ovsiannikov, Edmond Seymore, Jonathan Leane, Martin Kemka, usrbinkat, Enrico Ros
Thank you to all my generous patrons and donaters!
And thank you again to a16z for their generous grant.
Original model card: brucethemoose's CaPlatTessDolXaBoros Yi 34B 200K Dare Ties Highdensity
Dolphin-2.2-yi-34b-200k, Nous-Capybara-34B, Tess-M-v1.4, Airoboros-3_1-yi-34b-200k, PlatYi-34B-200K-Q, and Una-xaberius-34b-v1beta merged with a new, experimental implementation of "dare ties" via mergekit. See:
Various densities were tested with perplexity tests and long context prompts. Relatively high densities seem to perform better, contrary to the findings of the Super Mario paper.
This particular version is merged with more than the "recommended" max density of 0.5. It seems to result in even better perplexity, but I'm not sure if this translates to better output.
Weights that add up to 1 seems to be optimal.
Dare Ties is also resulting in seemingly better, lower perplexity merges than a regular ties merge, task arithmetic or a slerp merge.
Xaberuis is not a 200K model, hence it was merged at a very low density to try and preserve Yi 200K's long context performance while still inheriting some of Xaberius's performance.
I chose not to include other finetunes because they aren't trained on the 200K base. If any other 200K finetunes pop up, let me know.
It might recognize ChatML from Dolphin+Xaberius, and Llama-chat from Airoboros.
Sometimes the model "spells out" the stop token as </s> like Capybara, so you may need to add </s> as an additional stopping condition.
Running
Being a Yi model, try disabling the BOS token and/or running a lower temperature with 0.05-0.13 MinP, a little repitition penalty, and no other samplers. Yi tends to run "hot" by default.
24GB GPUs can run Yi-34B-200K models at 45K-75K context with exllamav2. I go into more detail in this post
I recommend exl2 quantizations profiled on data similar to the desired task. It is especially sensitive to the quantization data at low bpw!
To load this in full-context backends like transformers and vllm, you must change max_position_embeddings in config.json to a lower value than 200,000, otherwise you will OOM!
