Modelpublic

second-state/Hermes-2-Pro-Llama-3-8B-GGUF

sourceHugging Faceupdated 2y agoView on Hugging Face

2likes1.3kdownloads

Model Card

<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;">

Hermes-2-Pro-Llama-3-8B-GGUF

Original Model

NousResearch/Hermes-2-Pro-Llama-3-8B

Run with LlamaEdge

—LlamaEdge version: v0.11.3 and above

—Prompt template

—Prompt type: chatml

—Prompt string

text

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

—Context size: 4096

—Run as LlamaEdge service

—Chat

bash

    wasmedge --dir .:. --nn-preload default:GGML:AUTO:Hermes-2-Pro-Llama-3-8B-Q5_K_M.gguf \
      llama-api-server.wasm \
      --prompt-template chatml \
      --ctx-size 4096 \
      --model-name hermes-2-pro-llama-3-8b

—Tool use

bash

    wasmedge --dir .:. --nn-preload default:GGML:AUTO:Hermes-2-Pro-Llama-3-8B-Q5_K_M.gguf \
      llama-api-server.wasm \
      --prompt-template chatml-tool \
      --ctx-size 4096 \
      --model-name hermes-2-pro-llama-3-8b

—Run as LlamaEdge command app

bash

  wasmedge --dir .:. --nn-preload default:GGML:AUTO:Hermes-2-Pro-Llama-3-8B-Q5_K_M.gguf \
    llama-chat.wasm \
    --prompt-template chatml \
    --ctx-size 4096

Quantized GGUF Models

Name	Quant method	Bits	Size	Use case
Hermes-2-Pro-Llama-3-8B-Q2_K.gguf	Q2_K	2	3.18 GB	smallest, significant quality loss - not recommended for most purposes
Hermes-2-Pro-Llama-3-8B-Q3_K_L.gguf	Q3KL	3	4.32 GB	small, substantial quality loss
Hermes-2-Pro-Llama-3-8B-Q3_K_M.gguf	Q3KM	3	4.02 GB	very small, high quality loss
Hermes-2-Pro-Llama-3-8B-Q3_K_S.gguf	Q3KS	3	3.66 GB	very small, high quality loss
Hermes-2-Pro-Llama-3-8B-Q4_0.gguf	Q4_0	4	4.66 GB	legacy; small, very high quality loss - prefer using Q3KM
Hermes-2-Pro-Llama-3-8B-Q4_K_M.gguf	Q4KM	4	4.92 GB	medium, balanced quality - recommended
Hermes-2-Pro-Llama-3-8B-Q4_K_S.gguf	Q4KS	4	4.69 GB	small, greater quality loss
Hermes-2-Pro-Llama-3-8B-Q5_0.gguf	Q5_0	5	5.6 GB	legacy; medium, balanced quality - prefer using Q4KM
Hermes-2-Pro-Llama-3-8B-Q5_K_M.gguf	Q5KM	5	5.73 GB	large, very low quality loss - recommended
Hermes-2-Pro-Llama-3-8B-Q5_K_S.gguf	Q5KS	5	5.6 GB	large, low quality loss - recommended
Hermes-2-Pro-Llama-3-8B-Q6_K.gguf	Q6_K	6	6.6 GB	very large, extremely low quality loss
Hermes-2-Pro-Llama-3-8B-Q8_0.gguf	Q8_0	8	8.54 GB	very large, extremely low quality loss - not recommended
Hermes-2-Pro-Llama-3-8B-f16.gguf	f16	16	16.1 GB

Quantized with llama.cpp b3135.