TheBloke's LLM work is generously supported by a grant from andreessen horowitz (a16z)
Lzlv 70B - GGUF
Description
This repo contains GGUF format model files for A Guy's Lzlv 70B.
About GGUF
GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp.
Here is an incomplate list of clients and libraries that are known to support GGUF:
- llama.cpp. The source project for GGUF. Offers a CLI and a server option.
- text-generation-webui, the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration.
- KoboldCpp, a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling.
- LM Studio, an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration.
- LoLLMS Web UI, a great web UI with many interesting and unique features, including a full model library for easy model selection.
- Faraday.dev, an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration.
- ctransformers, a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server.
- llama-cpp-python, a Python library with GPU accel, LangChain support, and OpenAI-compatible API server.
- candle, a Rust ML framework with a focus on performance, including GPU support, and ease of use.
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
- A Guy's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions
Prompt template: Alpaca
Below is an instruction that describes a task. Write a response that appropriately completes the request.
### Instruction:
{prompt}
### Response:
Licensing
The creator of the source model has listed its license as cc-by-nc-2.0, 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: A Guy's Lzlv 70B.
Compatibility
These quantised GGUFv2 files are compatible with llama.cpp from August 27th onwards, as of commit d0cee0d
They are also compatible with many third party UIs and libraries - please see the list at the top of this README.
Explanation of quantisation methods
Click to see details
The new methods available are:
- GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw)
- GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw.
- GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw.
- GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw
- GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw
Refer to the Provided Files table below to see what files use which methods, and how.
Provided files
| Name | Quant method | Bits | Size | Max RAM required | Use case |
|---|---|---|---|---|---|
| lzlv_70b_fp16_hf.Q2_K.gguf | Q2_K | 2 | 29.28 GB | 31.78 GB | smallest, significant quality loss - not recommended for most purposes |
| lzlv_70b_fp16_hf.Q3_K_S.gguf | Q3_K_S | 3 | 29.92 GB | 32.42 GB | very small, high quality loss |
| lzlv_70b_fp16_hf.Q3_K_M.gguf | Q3_K_M | 3 | 33.19 GB | 35.69 GB | very small, high quality loss |
| lzlv_70b_fp16_hf.Q3_K_L.gguf | Q3_K_L | 3 | 36.15 GB | 38.65 GB | small, substantial quality loss |
| lzlv_70b_fp16_hf.Q4_0.gguf | Q4_0 | 4 | 38.87 GB | 41.37 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| lzlv_70b_fp16_hf.Q4_K_S.gguf | Q4_K_S | 4 | 39.07 GB | 41.57 GB | small, greater quality loss |
| lzlv_70b_fp16_hf.Q4_K_M.gguf | Q4_K_M | 4 | 41.42 GB | 43.92 GB | medium, balanced quality - recommended |
| lzlv_70b_fp16_hf.Q5_0.gguf | Q5_0 | 5 | 47.46 GB | 49.96 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| lzlv_70b_fp16_hf.Q5_K_S.gguf | Q5_K_S | 5 | 47.46 GB | 49.96 GB | large, low quality loss - recommended |
| lzlv_70b_fp16_hf.Q5_K_M.gguf | Q5_K_M | 5 | 48.75 GB | 51.25 GB | large, very low quality loss - recommended |
| lzlv_70b_fp16_hf.Q6_K.gguf | Q6_K | 6 | 56.59 GB | 59.09 GB | very large, extremely low quality loss |
| lzlv_70b_fp16_hf.Q8_0.gguf | Q8_0 | 8 | 73.29 GB | 75.79 GB | very large, extremely low quality loss - not recommended |
Note: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
Q6_K and Q8_0 files are split and require joining
Note: HF does not support uploading files larger than 50GB. Therefore I have uploaded the Q6_K and Q8_0 files as split files.
Click for instructions regarding Q6_K and Q8_0 files
q6_K
Please download:
lzlv_70b_fp16_hf.Q6_K.gguf-split-alzlv_70b_fp16_hf.Q6_K.gguf-split-b
q8_0
Please download:
lzlv_70b_fp16_hf.Q8_0.gguf-split-alzlv_70b_fp16_hf.Q8_0.gguf-split-b
To join the files, do the following:
Linux and macOS:
cat lzlv_70b_fp16_hf.Q6_K.gguf-split-* > lzlv_70b_fp16_hf.Q6_K.gguf && rm lzlv_70b_fp16_hf.Q6_K.gguf-split-*
cat lzlv_70b_fp16_hf.Q8_0.gguf-split-* > lzlv_70b_fp16_hf.Q8_0.gguf && rm lzlv_70b_fp16_hf.Q8_0.gguf-split-*
Windows command line:
COPY /B lzlv_70b_fp16_hf.Q6_K.gguf-split-a + lzlv_70b_fp16_hf.Q6_K.gguf-split-b lzlv_70b_fp16_hf.Q6_K.gguf
del lzlv_70b_fp16_hf.Q6_K.gguf-split-a lzlv_70b_fp16_hf.Q6_K.gguf-split-b
COPY /B lzlv_70b_fp16_hf.Q8_0.gguf-split-a + lzlv_70b_fp16_hf.Q8_0.gguf-split-b lzlv_70b_fp16_hf.Q8_0.gguf
del lzlv_70b_fp16_hf.Q8_0.gguf-split-a lzlv_70b_fp16_hf.Q8_0.gguf-split-b
How to download GGUF files
Note for manual downloaders: You almost never want to clone the entire repo! Multiple different quantisation formats are provided, and most users only want to pick and download a single file.
The following clients/libraries will automatically download models for you, providing a list of available models to choose from:
- LM Studio
- LoLLMS Web UI
- Faraday.dev
In text-generation-webui
Under Download Model, you can enter the model repo: TheBloke/lzlv_70B-GGUF and below it, a specific filename to download, such as: lzlv_70b_fp16_hf.Q4_K_M.gguf.
Then click Download.
On the command line, including multiple files at once
I recommend using the huggingface-hub Python library:
pip3 install huggingface-hub
Then you can download any individual model file to the current directory, at high speed, with a command like this:
huggingface-cli download TheBloke/lzlv_70B-GGUF lzlv_70b_fp16_hf.Q4_K_M.gguf --local-dir . --local-dir-use-symlinks False
More advanced huggingface-cli download usage
You can also download multiple files at once with a pattern:
huggingface-cli download TheBloke/lzlv_70B-GGUF --local-dir . --local-dir-use-symlinks False --include='*Q4_K*gguf'
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:
HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download TheBloke/lzlv_70B-GGUF lzlv_70b_fp16_hf.Q4_K_M.gguf --local-dir . --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.
Example llama.cpp command
Make sure you are using llama.cpp from commit d0cee0d or later.
./main -ngl 32 -m lzlv_70b_fp16_hf.Q4_K_M.gguf --color -c 4096 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n{prompt}\n\n### Response:"
Change -ngl 32 to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
Change -c 4096 to the desired sequence length. For extended sequence models - eg 8K, 16K, 32K - the necessary RoPE scaling parameters are read from the GGUF file and set by llama.cpp automatically.
If you want to have a chat-style conversation, replace the -p <PROMPT> argument with -i -ins
For other parameters and how to use them, please refer to the llama.cpp documentation
How to run in text-generation-webui
Further instructions here: text-generation-webui/docs/llama.cpp.md.
How to run from Python code
You can use GGUF models from Python using the llama-cpp-python or ctransformers libraries.
How to load this model in Python code, using ctransformers
First install the package
Run one of the following commands, according to your system:
# Base ctransformers with no GPU acceleration
pip install ctransformers
# Or with CUDA GPU acceleration
pip install ctransformers[cuda]
# Or with AMD ROCm GPU acceleration (Linux only)
CT_HIPBLAS=1 pip install ctransformers --no-binary ctransformers
# Or with Metal GPU acceleration for macOS systems only
CT_METAL=1 pip install ctransformers --no-binary ctransformers
Simple ctransformers example code
from ctransformers import AutoModelForCausalLM
# Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
llm = AutoModelForCausalLM.from_pretrained("TheBloke/lzlv_70B-GGUF", model_file="lzlv_70b_fp16_hf.Q4_K_M.gguf", model_type="llama", gpu_layers=50)
print(llm("AI is going to"))
How to use with LangChain
Here are guides on using llama-cpp-python and ctransformers with LangChain:
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: 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: A Guy's Lzlv 70B
lzlv_70B
A Mythomax/MLewd_13B-style merge of selected 70B models
A multi-model merge of several LLaMA2 70B finetunes for roleplaying and creative work. The goal was to create a model that combines creativity with intelligence for an enhanced experience.
Did it work? Probably, maybe. It seemed subjectively better than each of the individual models in my tests.
GGUF 4_K_M + 5_K_M can be found here: https://huggingface.co/lizpreciatior/lzlv_70b_fp16_hf/settings
Procedure:
Models used:
- NousResearch/Nous-Hermes-Llama2-70b - A great model for roleplaying, but not the best at following complex instructions.
- Xwin-LM/Xwin-LM-7B-V0.1 - Excellent at following instructions and quite creative out of the box, so it seemed like the best available model to act as the base for the merge.
- Doctor-Shotgun/Mythospice-70b - The wildcard of the three. I was looking for a creative, NSFW-oriented model and came across this while digging through hf. I hadn't heard of it before and apparently no one had bothered to release a quantized version of this model. So I downloaded it and did it myself to test it. It turned out to be more or less what I was looking for as my third component, so I used it here.
A big thank you to the creators of the models above. If you look up Mythospice, you will notice that it also includes Nous-Hermes so it's technically present twice in this mix. This is apparently common practice amongst the cool kids who do 13B models so I don't think this hurts the model.
The merging process was heavily inspired by Undi95's approach in Undi95/MXLewdMini-L2-13B. To be specific, the ratios are:
Component 1: Merge of Mythospice x Xwin with SLERP gradient [0.25, 0.3, 0.5]. Component 2: Merge Xwin x Hermes with SLERP gradient [0.4, 0.3, 0.25].
Finally, both Component 1 and Component 2 were merged with SLERP using weight 0.5.
Peformance
I tested this model for a few days before publishing it. It seems to more or less retain the instruction-following capabilities of Xwin-70B, while seeming to have adopted a lot of the creativity of the other two models. It handled my more complex scenarios that creative models otherwise tend to struggle with quite well. At the same time, its outputs felt more creative and possibly a bit more nsfw-inclined than Xwin-70b. So, is it better? Feels like it to me, subjectively. Is it really better? No clue, test it.
Prompt format:
Vicuna USER: [Prompt] ASSISTANT:
- Downloads last month
- 1,013