CoolFace
Modelpublic

RichardErkhov/sonthenguyen_-_OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered-gguf

sourceHugging Faceupdated 2y agoView on Hugging Face
0likes5.1kdownloads
Model Card

Quantization made by Richard Erkhov.

Github

Discord

Request more models

OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered - GGUF

  • —Model creator: https://huggingface.co/sonthenguyen/
  • —Original model: https://huggingface.co/sonthenguyen/OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered/
NameQuant methodSize
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q2_K.ggufQ2_K2.53GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.IQ3_XS.ggufIQ3_XS2.81GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.IQ3_S.ggufIQ3_S2.96GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q3_K_S.ggufQ3KS2.95GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.IQ3_M.ggufIQ3_M3.06GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q3_K.ggufQ3_K3.28GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q3_K_M.ggufQ3KM3.28GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q3_K_L.ggufQ3KL3.56GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.IQ4_XS.ggufIQ4_XS3.67GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q4_0.ggufQ4_03.83GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.IQ4_NL.ggufIQ4_NL3.87GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q4_K_S.ggufQ4KS3.86GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q4_K.ggufQ4_K4.07GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q4_K_M.ggufQ4KM4.07GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q4_1.ggufQ4_14.24GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q5_0.ggufQ5_04.65GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q5_K_S.ggufQ5KS4.65GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q5_K.ggufQ5_K4.78GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q5_K_M.ggufQ5KM4.78GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q5_1.ggufQ5_15.07GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q6_K.ggufQ6_K5.53GB
OpenHermes-2.5-Mistral-7B-mt-bench-DPO-recovered.Q8_0.ggufQ8_07.17GB

Original model description: --- license: apache-2.0 ---

Training hyperparameters LoRA: r=16 loraalpha=16 loradropout=0.05 bias="none" tasktype="CAUSALLM" targetmodules=['kproj', 'gateproj', 'vproj', 'upproj', 'qproj', 'oproj', 'downproj']

Training arguments: autofindbatchsize=True gradientcheckpointing=True learningrate=5e-7 lrschedulertype="cosine" maxsteps=3922 optim="pagedadamw32bit" warmup_steps=100

DPOTrainer: beta=0.1 maxpromptlength=1024 max_length=1536

Arxiv link: https://arxiv.org/abs/2403.02745