metadata

base_model: 01-ai/Yi-1.5-6B-Chat
inference: false
license: other
license_link: LICENSE
license_name: yi-license
model_creator: 01-ai
model_name: Yi-1.5-6B-Chat
model_type: yi
pipeline_tag: text-generation
quantized_by: Second State Inc.

Yi-1.5-6B-Chat-GGUF

Original Model

01-ai/Yi-1.5-6B-Chat

Run with LlamaEdge

LlamaEdge version: v0.10.0 and above

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: 4096

Run as LlamaEdge service

wasmedge --dir .:. --nn-preload default:GGML:AUTO:Yi-1.5-6B-Chat-Q5_K_M.gguf \
  llama-api-server.wasm \
  --prompt-template chatml \
  --ctx-size 4096 \
  --model-name Yi-1.5-6B-Chat

Run as LlamaEdge command app

wasmedge --dir .:. --nn-preload default:GGML:AUTO:Yi-1.5-6B-Chat-Q5_K_M.gguf \
  llama-chat.wasm \
  --prompt-template chatml \
  --ctx-size 4096

Quantized GGUF Models

Name	Quant method	Bits	Size	Use case
Yi-1.5-6B-Chat-Q2_K.gguf	Q2_K	2	2.34 GB	smallest, significant quality loss - not recommended for most purposes
Yi-1.5-6B-Chat-Q3_K_L.gguf	Q3_K_L	3	3.24 GB	small, substantial quality loss
Yi-1.5-6B-Chat-Q3_K_M.gguf	Q3_K_M	3	2.99 GB	very small, high quality loss
Yi-1.5-6B-Chat-Q3_K_S.gguf	Q3_K_S	3	2.71 GB	very small, high quality loss
Yi-1.5-6B-Chat-Q4_0.gguf	Q4_0	4	3.48 GB	legacy; small, very high quality loss - prefer using Q3_K_M
Yi-1.5-6B-Chat-Q4_K_M.gguf	Q4_K_M	4	3.67 GB	medium, balanced quality - recommended
Yi-1.5-6B-Chat-Q4_K_S.gguf	Q4_K_S	4	3.5 GB	small, greater quality loss
Yi-1.5-6B-Chat-Q5_0.gguf	Q5_0	5	4.2 GB	legacy; medium, balanced quality - prefer using Q4_K_M
Yi-1.5-6B-Chat-Q5_K_M.gguf	Q5_K_M	5	4.3 GB	large, very low quality loss - recommended
Yi-1.5-6B-Chat-Q5_K_S.gguf	Q5_K_S	5	4.2 GB	large, low quality loss - recommended
Yi-1.5-6B-Chat-Q6_K.gguf	Q6_K	6	4.97 GB	very large, extremely low quality loss
Yi-1.5-6B-Chat-Q8_0.gguf	Q8_0	8	6.44 GB	very large, extremely low quality loss - not recommended
Yi-1.5-6B-Chat-f16.gguf	f16	16	12.1 GB

Quantized with llama.cpp b2824