CoolFace
Modelpublic

RichardErkhov/ENERGY-DRINK-LOVE_-_deepnoid_DPOv3-gguf

sourceHugging Faceupdated 2y agoView on Hugging Face
0likes297downloads
Model Card

Quantization made by Richard Erkhov.

Github

Discord

Request more models

deepnoid_DPOv3 - GGUF

  • —Model creator: https://huggingface.co/ENERGY-DRINK-LOVE/
  • —Original model: https://huggingface.co/ENERGY-DRINK-LOVE/deepnoid_DPOv3/
NameQuant methodSize
deepnoid_DPOv3.Q2_K.ggufQ2_K3.73GB
deepnoid_DPOv3.Q3_K_S.ggufQ3KS4.34GB
deepnoid_DPOv3.Q3_K.ggufQ3_K4.84GB
deepnoid_DPOv3.Q3_K_M.ggufQ3KM4.84GB
deepnoid_DPOv3.Q3_K_L.ggufQ3KL5.26GB
deepnoid_DPOv3.IQ4_XS.ggufIQ4_XS5.43GB
deepnoid_DPOv3.Q4_0.ggufQ4_05.66GB
deepnoid_DPOv3.IQ4_NL.ggufIQ4_NL5.72GB
deepnoid_DPOv3.Q4_K_S.ggufQ4KS5.7GB
deepnoid_DPOv3.Q4_K.ggufQ4_K6.02GB
deepnoid_DPOv3.Q4_K_M.ggufQ4KM6.02GB
deepnoid_DPOv3.Q4_1.ggufQ4_16.27GB
deepnoid_DPOv3.Q5_0.ggufQ5_06.89GB
deepnoid_DPOv3.Q5_K_S.ggufQ5KS6.89GB
deepnoid_DPOv3.Q5_K.ggufQ5_K7.08GB
deepnoid_DPOv3.Q5_K_M.ggufQ5KM7.08GB
deepnoid_DPOv3.Q5_1.ggufQ5_17.51GB
deepnoid_DPOv3.Q6_K.ggufQ6_K8.2GB
deepnoid_DPOv3.Q8_0.ggufQ8_010.62GB

Original model description: --- license: apache-2.0 basemodel: Deepnoid/mergekitv2 tags:

  • —trl
  • —dpo
  • —generatedfromtrainer model-index:
  • —name: nhndpov3mergekitv2_DPO results: [] ---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

nhndpov3mergekitv2_DPO

This model is a fine-tuned version of Deepnoid/mergekit_v2 on an unknown dataset.

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • —learning_rate: 5e-07
  • —trainbatchsize: 1
  • —evalbatchsize: 8
  • —seed: 42
  • —distributed_type: multi-GPU
  • —num_devices: 6
  • —gradientaccumulationsteps: 8
  • —totaltrainbatch_size: 48
  • —totalevalbatch_size: 48
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: cosine
  • —lrschedulerwarmup_ratio: 0.1
  • —num_epochs: 1

Training results

Framework versions

  • —Transformers 4.36.2
  • —Pytorch 2.0.1+cu117
  • —Datasets 2.16.1
  • —Tokenizers 0.15.1