---

Osmosis-Structure-0.6B-GGUF

Original Model

osmosis-ai/Osmosis-Structure-0.6B

Run with LlamaEdge

• LlamaEdge version: v0.21.0 and above

• Prompt template

  • Prompt type: chatml

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

• Context size: 128000

• Run as LlamaEdge service

  wasmedge --dir .:. --nn-preload default:GGML:AUTO:Osmosis-Structure-0.6B-Q5_K_M.gguf \
    llama-api-server.wasm \
    --prompt-template chatml \
    --model-name Osmosis-Structure-0.6B \
    --ctx-size 128000
  

Quantized GGUF Models

Name	Quant method	Bits	Size	Use case
Osmosis-Structure-0.6B-Q2_K.gguf	Q2_K	2	296 MB	smallest, significant quality loss - not recommended for most purposes
Osmosis-Structure-0.6B-Q3_K_L.gguf	Q3_K_L	3	368 MB	small, substantial quality loss
Osmosis-Structure-0.6B-Q3_K_M.gguf	Q3_K_M	3	347 MB	very small, high quality loss
Osmosis-Structure-0.6B-Q3_K_S.gguf	Q3_K_S	3	323 MB	very small, high quality loss
Osmosis-Structure-0.6B-Q4_0.gguf	Q4_0	4	382 MB	legacy; small, very high quality loss - prefer using Q3_K_M
Osmosis-Structure-0.6B-Q4_K_M.gguf	Q4_K_M	4	397 MB	medium, balanced quality - recommended
Osmosis-Structure-0.6B-Q4_K_S.gguf	Q4_K_S	4	383 MB	small, greater quality loss
Osmosis-Structure-0.6B-Q5_0.gguf	Q5_0	5	437 MB	legacy; medium, balanced quality - prefer using Q4_K_M
Osmosis-Structure-0.6B-Q5_K_M.gguf	Q5_K_M	5	444 MB	large, very low quality loss - recommended
Osmosis-Structure-0.6B-Q5_K_S.gguf	Q5_K_S	5	437 MB	large, low quality loss - recommended
Osmosis-Structure-0.6B-Q6_K.gguf	Q6_K	6	495 MB	very large, extremely low quality loss
Osmosis-Structure-0.6B-Q8_0.gguf	Q8_0	8	639 MB	very large, extremely low quality loss - not recommended
Osmosis-Structure-0.6B-f16.gguf	f16	16	1.20 GB

Quantized with llama.cpp b5501