75 lines
4.0 KiB
Markdown
75 lines
4.0 KiB
Markdown
---
|
||
base_model: deepseek-ai/deepseek-llm-7b-chat
|
||
inference: false
|
||
license: other
|
||
license_link: LICENSE
|
||
license_name: deepseek
|
||
model_creator: DeepSeek
|
||
model_name: Deepseek LLM 7B Chat
|
||
model_type: deepseek
|
||
quantized_by: Second State Inc.
|
||
---
|
||
|
||
<!-- header start -->
|
||
<!-- 200823 -->
|
||
<div style="width: auto; margin-left: auto; margin-right: auto">
|
||
<img src="https://github.com/LlamaEdge/LlamaEdge/raw/dev/assets/logo.svg" style="width: 100%; min-width: 400px; display: block; margin: auto;">
|
||
</div>
|
||
<hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
|
||
<!-- header end -->
|
||
|
||
# Deepseek-LLM-7B-Chat-GGUF
|
||
|
||
## Original Model
|
||
|
||
[deepseek-ai/deepseek-llm-7b-chat](https://huggingface.co/deepseek-ai/deepseek-llm-7b-chat)
|
||
|
||
## Run with LlamaEdge
|
||
|
||
- LlamaEdge version: [v0.2.8](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.2.8) and above
|
||
|
||
- Prompt template
|
||
|
||
- Prompt type: `deepseek-chat`
|
||
|
||
- Prompt string
|
||
|
||
```text
|
||
User: {user_message_1}
|
||
|
||
Assistant: {assistant_message_1}<｜end▁of▁sentence｜>User: {user_message_2}
|
||
|
||
Assistant:
|
||
```
|
||
|
||
- Context size: `4096`
|
||
|
||
- Run as LlamaEdge service
|
||
|
||
```bash
|
||
wasmedge --dir .:. --nn-preload default:GGML:AUTO:deepseek-llm-7b-chat-Q5_K_M.gguf llama-api-server.wasm -p deepseek-chat
|
||
```
|
||
|
||
- Run as LlamaEdge command app
|
||
|
||
```bash
|
||
wasmedge --dir .:. --nn-preload default:GGML:AUTO:deepseek-llm-7b-chat-Q5_K_M.gguf llama-chat.wasm -p deepseek-chat
|
||
```
|
||
|
||
## Quantized GGUF Models
|
||
|
||
| Name | Quant method | Bits | Size | Use case |
|
||
| ---- | ---- | ---- | ---- | ----- |
|
||
| [deepseek-llm-7b-chat-Q2_K.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q2_K.gguf) | Q2_K | 2 | 2.72 GB| smallest, significant quality loss - not recommended for most purposes |
|
||
| [deepseek-llm-7b-chat-Q3_K_L.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q3_K_L.gguf) | Q3_K_L | 3 | 3.75 GB| small, substantial quality loss |
|
||
| [deepseek-llm-7b-chat-Q3_K_M.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q3_K_M.gguf) | Q3_K_M | 3 | 3.46 GB| very small, high quality loss |
|
||
| [deepseek-llm-7b-chat-Q3_K_S.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q3_K_S.gguf) | Q3_K_S | 3 | 3.14 GB| very small, high quality loss |
|
||
| [deepseek-llm-7b-chat-Q4_0.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q4_0.gguf) | Q4_0 | 4 | 4.00 GB| legacy; small, very high quality loss - prefer using Q3_K_M |
|
||
| [deepseek-llm-7b-chat-Q4_K_M.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q4_K_M.gguf) | Q4_K_M | 4 | 4.22 GB| medium, balanced quality - recommended |
|
||
| [deepseek-llm-7b-chat-Q4_K_S.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q4_K_S.gguf) | Q4_K_S | 4 | 4.03 GB| small, greater quality loss |
|
||
| [deepseek-llm-7b-chat-Q5_0.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q5_0.gguf) | Q5_0 | 5 | 4.81 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
|
||
| [deepseek-llm-7b-chat-Q5_K_M.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q5_K_M.gguf) | Q5_K_M | 5 | 4.93 GB| large, very low quality loss - recommended |
|
||
| [deepseek-llm-7b-chat-Q5_K_S.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q5_K_S.gguf) | Q5_K_S | 5 | 4.81 GB| large, low quality loss - recommended |
|
||
| [deepseek-llm-7b-chat-Q6_K.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q6_K.gguf) | Q6_K | 6 | 5.67 GB| very large, extremely low quality loss |
|
||
| [deepseek-llm-7b-chat-Q8_0.gguf](https://huggingface.co/second-state/Deepseek-LLM-7B-Chat-GGUF/blob/main/deepseek-llm-7b-chat-Q8_0.gguf) | Q8_0 | 8 | 7.35 GB| very large, extremely low quality loss - not recommended |
|