metadata

base_model: HuggingFaceTB/cosmo2-350M-webinst-sc2
tags:
  - alignment-handbook
  - trl
  - dpo
  - generated_from_trainer
  - trl
  - dpo
  - generated_from_trainer
datasets:
  - argilla/dpo-mix-7k
model-index:
  - name: cosmo2-350M-webinst-sc2-dpo-argilla-ep1
    results: []

cosmo2-350M-webinst-sc2-dpo-argilla-ep1

This model is a fine-tuned version of HuggingFaceTB/cosmo2-350M-webinst-sc2 on the argilla/dpo-mix-7k dataset. It achieves the following results on the evaluation set:

Loss: 0.6834
Rewards/chosen: -0.0086
Rewards/rejected: -0.0304
Rewards/accuracies: 0.5938
Rewards/margins: 0.0218
Logps/rejected: -418.5675
Logps/chosen: -442.4709
Logits/rejected: -0.7106
Logits/chosen: -0.5211
IFEval prompt loose 17.01
IFEval prompt strict 14.05

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-06
train_batch_size: 2
eval_batch_size: 4
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 8
total_train_batch_size: 128
total_eval_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 1

Training results

Framework versions

Transformers 4.42.3
Pytorch 2.1.2
Datasets 2.20.0
Tokenizers 0.19.1