Edit model card

Llamacpp imatrix Quantizations of DeepSeek-V2-Chat-0628

Using llama.cpp release b3389 for quantization.

Original model: https://huggingface.co/deepseek-ai/DeepSeek-V2-Chat-0628

All quants made using imatrix option with dataset from here

Run them in LM Studio

Prompt format

<|begin▁of▁sentence|>{system_prompt}

<|User|>{prompt}<|Assistant|><|end▁of▁sentence|><|Assistant|>

Download a file (not the whole branch) from below:

Filename Quant type File Size Split Description
DeepSeek-V2-Chat-0628-Q4_K_M.gguf Q4_K_M 142.45GB true Good quality, default size for must use cases, recommended.
DeepSeek-V2-Chat-0628-IQ4_XS.gguf IQ4_XS 125.56GB true Decent quality, smaller than Q4_K_S with similar performance, recommended.
DeepSeek-V2-Chat-0628-Q3_K_M.gguf Q3_K_M 112.67GB true Low quality.
DeepSeek-V2-Chat-0628-IQ3_M.gguf IQ3_M 103.37GB true Medium-low quality, new method with decent performance comparable to Q3_K_M.
DeepSeek-V2-Chat-0628-Q3_K_S.gguf Q3_K_S 101.68GB true Low quality, not recommended.
DeepSeek-V2-Chat-0628-IQ3_XXS.gguf IQ3_XXS 90.85GB true Lower quality, new method with decent performance, comparable to Q3 quants.
DeepSeek-V2-Chat-0628-Q2_K.gguf Q2_K 85.95GB true Very low quality but surprisingly usable.
DeepSeek-V2-Chat-0628-IQ2_M.gguf IQ2_M 76.92GB true Relatively low quality, uses SOTA techniques to be surprisingly usable.
DeepSeek-V2-Chat-0628-IQ2_XS.gguf IQ2_XS 68.71GB true Low quality, uses SOTA techniques to be usable.
DeepSeek-V2-Chat-0628-IQ2_XXS.gguf IQ2_XXS 61.50GB true Very low quality, uses SOTA techniques to be usable.
DeepSeek-V2-Chat-0628-IQ1_M.gguf IQ1_M 52.68GB true Extremely low quality, not recommended.

Credits

Thank you kalomaze and Dampf for assistance in creating the imatrix calibration dataset

Thank you ZeroWw for the inspiration to experiment with embed/output

Downloading using huggingface-cli

First, make sure you have hugginface-cli installed:

pip install -U "huggingface_hub[cli]"

Then, you can target the specific file you want:

huggingface-cli download bartowski/DeepSeek-V2-Chat-0628-GGUF --include "DeepSeek-V2-Chat-0628-Q4_K_M.gguf" --local-dir ./

If the model is bigger than 50GB, it will have been split into multiple files. In order to download them all to a local folder, run:

huggingface-cli download bartowski/DeepSeek-V2-Chat-0628-GGUF --include "DeepSeek-V2-Chat-0628-Q8_0.gguf/*" --local-dir DeepSeek-V2-Chat-0628-Q8_0

You can either specify a new local-dir (DeepSeek-V2-Chat-0628-Q8_0) or download them all in place (./)

Which file should I choose?

A great write up with charts showing various performances is provided by Artefact2 here

The first thing to figure out is how big a model you can run. To do this, you'll need to figure out how much RAM and/or VRAM you have.

If you want your model running as FAST as possible, you'll want to fit the whole thing on your GPU's VRAM. Aim for a quant with a file size 1-2GB smaller than your GPU's total VRAM.

If you want the absolute maximum quality, add both your system RAM and your GPU's VRAM together, then similarly grab a quant with a file size 1-2GB Smaller than that total.

Next, you'll need to decide if you want to use an 'I-quant' or a 'K-quant'.

If you don't want to think too much, grab one of the K-quants. These are in format 'QX_K_X', like Q5_K_M.

If you want to get more into the weeds, you can check out this extremely useful feature chart:

llama.cpp feature matrix

But basically, if you're aiming for below Q4, and you're running cuBLAS (Nvidia) or rocBLAS (AMD), you should look towards the I-quants. These are in format IQX_X, like IQ3_M. These are newer and offer better performance for their size.

These I-quants can also be used on CPU and Apple Metal, but will be slower than their K-quant equivalent, so speed vs performance is a tradeoff you'll have to decide.

The I-quants are not compatible with Vulcan, which is also AMD, so if you have an AMD card double check if you're using the rocBLAS build or the Vulcan build. At the time of writing this, LM Studio has a preview with ROCm support, and other inference engines have specific builds for ROCm.

Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski

Downloads last month
212
GGUF
Model size
236B params
Architecture
deepseek2

1-bit

2-bit

3-bit

4-bit

Inference Examples
Unable to determine this model's library. Check the docs .

Model tree for bartowski/DeepSeek-V2-Chat-0628-GGUF

Quantized
(3)
this model