CoolFace
Modelpublic

RichardErkhov/athirdpath_-_Llama-3.1-Techne-RP-8b-v1-gguf

sourceHugging Faceupdated 2y agoView on Hugging Face
0likes787downloads
Model Card

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Llama-3.1-Techne-RP-8b-v1 - GGUF

  • —Model creator: https://huggingface.co/athirdpath/
  • —Original model: https://huggingface.co/athirdpath/Llama-3.1-Techne-RP-8b-v1/
NameQuant methodSize
Llama-3.1-Techne-RP-8b-v1.Q2_K.ggufQ2_K2.96GB
Llama-3.1-Techne-RP-8b-v1.IQ3_XS.ggufIQ3_XS3.28GB
Llama-3.1-Techne-RP-8b-v1.IQ3_S.ggufIQ3_S3.43GB
Llama-3.1-Techne-RP-8b-v1.Q3_K_S.ggufQ3KS3.41GB
Llama-3.1-Techne-RP-8b-v1.IQ3_M.ggufIQ3_M3.52GB
Llama-3.1-Techne-RP-8b-v1.Q3_K.ggufQ3_K3.74GB
Llama-3.1-Techne-RP-8b-v1.Q3_K_M.ggufQ3KM3.74GB
Llama-3.1-Techne-RP-8b-v1.Q3_K_L.ggufQ3KL4.03GB
Llama-3.1-Techne-RP-8b-v1.IQ4_XS.ggufIQ4_XS4.18GB
Llama-3.1-Techne-RP-8b-v1.Q4_0.ggufQ4_04.34GB
Llama-3.1-Techne-RP-8b-v1.IQ4_NL.ggufIQ4_NL4.38GB
Llama-3.1-Techne-RP-8b-v1.Q4_K_S.ggufQ4KS4.37GB
Llama-3.1-Techne-RP-8b-v1.Q4_K.ggufQ4_K4.58GB
Llama-3.1-Techne-RP-8b-v1.Q4_K_M.ggufQ4KM4.58GB
Llama-3.1-Techne-RP-8b-v1.Q4_1.ggufQ4_14.78GB
Llama-3.1-Techne-RP-8b-v1.Q5_0.ggufQ5_05.21GB
Llama-3.1-Techne-RP-8b-v1.Q5_K_S.ggufQ5KS5.21GB
Llama-3.1-Techne-RP-8b-v1.Q5_K.ggufQ5_K5.34GB
Llama-3.1-Techne-RP-8b-v1.Q5_K_M.ggufQ5KM5.34GB
Llama-3.1-Techne-RP-8b-v1.Q5_1.ggufQ5_15.65GB
Llama-3.1-Techne-RP-8b-v1.Q6_K.ggufQ6_K6.14GB
Llama-3.1-Techne-RP-8b-v1.Q8_0.ggufQ8_07.95GB

Original model description: --- license: llama3.1 tags:

  • —not-for-all-audiences --- <p align="center"><img src="https://cdn-uploads.huggingface.co/production/uploads/633a809fa4a8f33508dce32c/BMdwgJ6cHZWbiGL48Q-Wq.png"/><font size="6"><b>Techne-RP-8b</b></font></p> <p align="center"><font size="4"><b>Trained with Llama 3 prompt formatting, Alpaca works too</b></font></p> <p align="center"><img src="https://cdn-uploads.huggingface.co/production/uploads/633a809fa4a8f33508dce32c/wNVjJ1u0kLR0cL-9Bwk4P.png"/></p>

<p align="center"><font size="5"> <b>Assistant Example @ q5km</b> </font></p> <p align="center"><img src="https://cdn-uploads.huggingface.co/production/uploads/633a809fa4a8f33508dce32c/dN45v5YHdIVyOacRx4xSc.png"/></p> <p align="center"><img src="https://cdn-uploads.huggingface.co/production/uploads/633a809fa4a8f33508dce32c/8qE0ikdtibgFtMZ-SVH1P.png"/></p>

<p align="center"><font size="5"> <b>NSFW Writing Example @ q5km</b> </font></p> <p align="center"><img src="https://cdn-uploads.huggingface.co/production/uploads/633a809fa4a8f33508dce32c/exle9vh1IFoKnAcIL1D64.png"/></p>

<p align="center"><font size="5"> <b>Training Methodology</b> </font></p>

athirdpath/Llama-3.1-InstructNSFW-pretrainede1-plus_reddit was further trained in the order below:

SFT

  • —Doctor-Shotgun/no-robots-sharegpt
  • —grimulkan/LimaRP-augmented
  • —Inv/c2-logs-cleaned-deslopped

DPO

  • —jondurbin/truthy-dpo-v0.1
  • —Undi95/Weyaxi-humanish-dpo-project-noemoji
  • —athirdpath/DPO_Pairs-Roleplay-Llama3-NSFW