Deepseek V2 Chat GGUF

🚀 DeepSeek-V2-Chat-GGUF

This is a quantized model from DeepSeek-V2-Chat, offering various quantization options for different usage scenarios.

Quantizised from https://huggingface.co/deepseek-ai/DeepSeek-V2-Chat

Using llama.cpp b3026 for quantizisation. Given the rapid release of llama.cpp builds, this will likely change over time.

⚠️ Important Note

Please set the metadata KV overrides below.

🚀 Quick Start

✨ Features

• Quantized from the original DeepSeek-V2-Chat model.
• Supports multiple quantization options with different quality and size trade - offs.
• Can be run in command - line chat mode or through an OpenAI compatible server using llama.cpp.

📦 Installation

Downloading the bf16

• Find the relevant directory.
• Download all files.
• Run merge.py.
• The merged GGUF should appear.

Downloading the quantizations

• Find the relevant directory.
• Download all files.
• Point to the first split (most programs should load all the splits automatically now).

💻 Usage Examples

Basic Usage

To start in command line chat mode (chat completion):

main -m DeepSeek-V2-Chat.{quant}.gguf -c {context length} --color -c (-i)

Advanced Usage

To use llama.cpp's OpenAI compatible server:

server \
  -m DeepSeek-V2-Chat.{quant}.gguf \
  -c {context_length} \
  (--color [recommended: colored output in supported terminals]) \
  (-i [note: interactive mode]) \
  (--mlock [note: avoid using swap]) \
  (--verbose) \
  (--log-disable [note: disable logging to file, may be useful for prod]) \
  (--metrics [note: prometheus compatible monitoring endpoint]) \
  (--api-key [string]) \
  (--port [int]) \
  (--flash-attn [note: must be fully offloaded to supported GPU])

Making an importance matrix:

imatrix \
  -m DeepSeek-V2-Chat.{quant}.gguf \
  -f groups_merged.txt \
  --verbosity [0, 1, 2] \
  -ngl {GPU offloading; must build with CUDA} \
  --ofreq {recommended: 1}

Making a quant:

quantize \
  DeepSeek-V2-Chat.bf16.gguf \
  DeepSeek-V2-Chat.{quant}.gguf \
  {quant} \
  (--imatrix [file])

⚠️ Important Note

Use iMatrix quants only if you can fully offload to GPU, otherwise speed will be affected negatively.

📚 Documentation

Quants

Planned Quants (weighted/iMatrix)

Metadata KV overrides (pass them using --override-kv, can be specified multiple times):

deepseek2.attention.q_lora_rank=int:1536
deepseek2.attention.kv_lora_rank=int:512
deepseek2.expert_shared_count=int:2
deepseek2.expert_feed_forward_length=int:1536
deepseek2.expert_weights_scale=float:16
deepseek2.leading_dense_block_count=int:1
deepseek2.rope.scaling.yarn_log_multiplier=float:0.0707

🔧 Technical Details

• Performance: ~1.5t/s with Ryzen 3 3700x (96gb 3200mhz) [Q2_K].
• iMatrix: Find imatrix.dat in the root of this repo, made with a Q2_K quant containing 62 chunks (see here for info: https://github.com/ggerganov/llama.cpp/issues/5153#issuecomment-1913185693). Using groups_merged.txt, find it here: https://github.com/ggerganov/llama.cpp/discussions/5263#discussioncomment-8395384.

📄 License

• DeepSeek license for model weights, which can be found in the LICENSE file in the root of this repo.
• MIT license for any repo code.

Censorship

This model is a bit censored, finetuning on toxic DPO might help.