CoolFace
Modelpublic

legraphista/DeepSeek-V2-Lite-Chat-IMat-GGUF

sourceHugging Faceupdated 2y agoView on Hugging Face
12likes3.7kdownloads
Model Card

DeepSeek-V2-Lite-Chat-IMat-GGUF

Llama.cpp imatrix quantization of deepseek-ai/DeepSeek-V2-Lite-Chat

Original Model: deepseek-ai/DeepSeek-V2-Lite-Chat Original dtype: BF16 (bfloat16) Quantized by: llama.cpp fork PR 7519 IMatrix dataset: here

Files

IMatrix

Status: ✅ Available Link: here

Common Quants

FilenameQuant typeFile SizeStatusUses IMatrixIs Split
DeepSeek-V2-Lite-Chat.Q8_0.ggufQ8_016.70GB✅ Available⚪ No📦 No
DeepSeek-V2-Lite-Chat.Q6_K.ggufQ6_K14.07GB✅ Available⚪ No📦 No
DeepSeek-V2-Lite-Chat.Q4_K.ggufQ4_K10.36GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.Q3_K.ggufQ3_K8.13GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.Q2_K.ggufQ2_K6.43GB✅ Available🟢 Yes📦 No

All Quants

FilenameQuant typeFile SizeStatusUses IMatrixIs Split
DeepSeek-V2-Lite-Chat.FP16.ggufF1631.42GB✅ Available⚪ No📦 No
DeepSeek-V2-Lite-Chat.BF16.ggufBF1631.42GB✅ Available⚪ No📦 No
DeepSeek-V2-Lite-Chat.Q5_K.ggufQ5_K11.85GB✅ Available⚪ No📦 No
DeepSeek-V2-Lite-Chat.Q5_K_S.ggufQ5KS11.14GB✅ Available⚪ No📦 No
DeepSeek-V2-Lite-Chat.Q4_K_S.ggufQ4KS9.53GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.Q3_K_L.ggufQ3KL8.46GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.Q3_K_S.ggufQ3KS7.49GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.Q2_K_S.ggufQ2KS6.46GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ4_NL.ggufIQ4_NL8.91GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ4_XS.ggufIQ4_XS8.57GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ3_M.ggufIQ3_M7.55GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ3_S.ggufIQ3_S7.49GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ3_XS.ggufIQ3_XS7.12GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ3_XXS.ggufIQ3_XXS6.96GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ2_M.ggufIQ2_M6.33GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ2_S.ggufIQ2_S6.01GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ2_XS.ggufIQ2_XS5.97GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ2_XXS.ggufIQ2_XXS5.64GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ1_M.ggufIQ1_M5.24GB✅ Available🟢 Yes📦 No
DeepSeek-V2-Lite-Chat.IQ1_S.ggufIQ1_S4.99GB✅ Available🟢 Yes📦 No

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 legraphista/DeepSeek-V2-Lite-Chat-IMat-GGUF --include "DeepSeek-V2-Lite-Chat.Q8_0.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 legraphista/DeepSeek-V2-Lite-Chat-IMat-GGUF --include "DeepSeek-V2-Lite-Chat.Q8_0/*" --local-dir DeepSeek-V2-Lite-Chat.Q8_0
# see FAQ for merging GGUF's

Inference

Simple chat template

<｜begin▁of▁sentence｜>User: {user_message_1}

Assistant: {assistant_message_1}<｜end▁of▁sentence｜>User: {user_message_2}

Assistant:

Chat template with system prompt

<｜begin▁of▁sentence｜>{system_message}

User: {user_message_1}

Assistant: {assistant_message_1}<｜end▁of▁sentence｜>User: {user_message_2}

Assistant:

Llama.cpp

llama.cpp/main -m DeepSeek-V2-Lite-Chat.Q8_0.gguf --color -i -p "prompt here (according to the chat template)"

FAQ

Why is the IMatrix not applied everywhere?

According to this investigation, it appears that lower quantizations are the only ones that benefit from the imatrix input (as per hellaswag results).

How do I merge a split GGUF?

  1. 1.Make sure you have gguf-split available
  2. 2.To get hold of gguf-split, navigate to https://github.com/ggerganov/llama.cpp/releases
  3. 3.Download the appropriate zip for your system from the latest release
  4. 4.Unzip the archive and you should be able to find gguf-split
  5. 5.Locate your GGUF chunks folder (ex: DeepSeek-V2-Lite-Chat.Q8_0)
  6. 6.Run gguf-split --merge DeepSeek-V2-Lite-Chat.Q8_0/DeepSeek-V2-Lite-Chat.Q8_0-00001-of-XXXXX.gguf DeepSeek-V2-Lite-Chat.Q8_0.gguf
  7. 7.Make sure to point gguf-split to the first chunk of the split.

Got a suggestion? Ping me @legraphista!