RichardErkhov/chasedreaminf_-_gpt2-dpo-dutch-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

gpt2-dpo-dutch - GGUF

• —Model creator: https://huggingface.co/chasedreaminf/
• —Original model: https://huggingface.co/chasedreaminf/gpt2-dpo-dutch/

Original model description: --- libraryname: transformers license: apache-2.0 basemodel: pkarypis/gpt2-sft-dutch tags:

• —alignment-handbook
• —trl
• —dpo
• —generatedfromtrainer
• —trl
• —dpo
• —generatedfromtrainer datasets:
• —BramVanroy/ultrafeedbackdutch model-index:
• —name: gpt2-dpo-dutch 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. -->

gpt2-dpo-dutch

This model is a fine-tuned version of pkarypis/gpt2-sft-dutch on the BramVanroy/ultrafeedbackdutch dataset. It achieves the following results on the evaluation set:

• —Loss: 0.6936
• —Rewards/chosen: 0.0022
• —Rewards/rejected: -0.0016
• —Rewards/accuracies: 0.5172
• —Rewards/margins: 0.0038
• —Logps/rejected: -1243.3655
• —Logps/chosen: -1947.5537
• —Logits/rejected: -3.7675
• —Logits/chosen: -3.4617

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: 2
• —evalbatchsize: 2
• —seed: 42
• —distributed_type: multi-GPU
• —num_devices: 4
• —totaltrainbatch_size: 8
• —totalevalbatch_size: 8
• —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 4.44.2
• —Pytorch 2.1.2
• —Datasets 3.0.0
• —Tokenizers 0.19.1