metadata

license: other
license_name: tongyi-qianwen
license_link: https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE
model_name: Qwen1.5-110B-Chat
arxiv: 2307.09288
base_model: Qwen/Qwen1.5-110B-Chat
inference: false
model_creator: Qwen
pipeline_tag: text-generation
quantized_by: Second State Inc.
language:
  - en
tags:
  - chat

Qwen1.5-110B-Chat-GGUF

Original Model

Qwen/Qwen1.5-110B-Chat

Run with LlamaEdge

LlamaEdge version: coming soon

Prompt template

Prompt type: chatml

Prompt string

<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

Context size: 8192

Run as LlamaEdge service

wasmedge --dir .:. --nn-preload default:GGML:AUTO:Qwen1.5-110B-Chat-Q2_K.gguf \
  llama-api-server.wasm \
  --prompt-template chatml \
  --ctx-size 8192 \
  --model-name qwen1.5-110b-chat

Run as LlamaEdge command app

wasmedge --dir .:. --nn-preload default:GGML:AUTO:Qwen1.5-110B-Chat-Q2_K.gguf \
  llama-chat.wasm \
  --prompt-template chatml

Quantized GGUF Models

Name	Quant method	Bits	Size	Use case
Qwen1.5-110B-Chat-Q2_K.gguf	Q2_K	2	41.2 GB	smallest, significant quality loss - not recommended for most purposes
Qwen1.5-110B-Chat-Q3_K_M-00001-of-00002.gguf	Q3_K_M	3	32.2 GB	very small, high quality loss
Qwen1.5-110B-Chat-Q3_K_M-00002-of-00002.gguf	Q3_K_M	3	21.5 GB	very small, high quality loss
Qwen1.5-110B-Chat-Q4_K_M-00001-of-00003.gguf	Q4_K_M	4	32 GB	medium, balanced quality - recommended
Qwen1.5-110B-Chat-Q4_K_M-00002-of-00003.gguf	Q4_K_M	4	32.1 GB	medium, balanced quality - recommended
Qwen1.5-110B-Chat-Q4_K_M-00003-of-00003.gguf	Q4_K_M	4	3.09 GB	medium, balanced quality - recommended
Qwen1.5-110B-Chat-Q5_K_M-00001-of-00003.gguf	Q5_K_M	5	32.1 GB	large, very low quality loss - recommended
Qwen1.5-110B-Chat-Q5_K_M-00002-of-00003.gguf	Q5_K_M	5	32 GB	large, very low quality loss - recommended
Qwen1.5-110B-Chat-Q5_K_M-00003-of-00003.gguf	Q5_K_M	5	14.8 GB	large, very low quality loss - recommended

Quantized with llama.cpp b2824.