metadata

license: apache-2.0
base_model: AmberYifan/mistral-safe-sft-full
tags:
  - alignment-handbook
  - generated_from_trainer
datasets:
  - AmberYifan/sft-spin-kcenter-5k
model-index:
  - name: sft-spin-kcenter-5k
    results: []

sft-spin-kcenter-5k

This model is a fine-tuned version of AmberYifan/mistral-safe-sft-full on the AmberYifan/sft-spin-kcenter-5k dataset. It achieves the following results on the evaluation set:

Loss: 0.9619
Rewards/real: 0.2019
Rewards/generated: -0.1441
Rewards/accuracies: 0.8906
Rewards/margins: 0.3460
Logps/generated: -259.6614
Logps/real: -210.6393
Logits/generated: -2.6617
Logits/real: -2.6095

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

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/real	Rewards/generated	Rewards/accuracies	Rewards/margins	Logps/generated	Logps/real	Logits/generated	Logits/real
0.0	0.6369	100	0.9619	0.2019	-0.1441	0.8906	0.3460	-259.6614	-210.6393	-2.6617	-2.6095

Framework versions

Transformers 4.43.3
Pytorch 2.2.2+cu121
Datasets 2.20.0
Tokenizers 0.19.1