bart_large_cnn_samsum_model_10epoch

This model is a fine-tuned version of facebook/bart-large-cnn on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 2.5260
  • Model Preparation Time: 0.0066
  • Rouge1: 0.4165
  • Rouge2: 0.1911
  • Rougel: 0.3142
  • Rougelsum: 0.3143
  • Gen Len: 60.615

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: 2e-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 16
  • optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • num_epochs: 10
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Model Preparation Time Rouge1 Rouge2 Rougel Rougelsum Gen Len
No log 1.0 200 1.4282 0.0066 0.4109 0.2008 0.3084 0.3085 59.755
No log 2.0 400 1.5080 0.0066 0.4214 0.2027 0.3175 0.3175 59.3862
1.2171 3.0 600 1.5348 0.0066 0.4093 0.1949 0.3071 0.307 60.2062
1.2171 4.0 800 1.7114 0.0066 0.4092 0.1928 0.3067 0.3066 60.38
0.6518 5.0 1000 1.8757 0.0066 0.4149 0.1935 0.3118 0.3117 59.5
0.6518 6.0 1200 2.0521 0.0066 0.4126 0.1902 0.3107 0.3108 60.335
0.6518 7.0 1400 2.1551 0.0066 0.4138 0.1917 0.3117 0.3115 60.1888
0.3371 8.0 1600 2.4051 0.0066 0.4132 0.1913 0.3116 0.3116 60.28
0.3371 9.0 1800 2.4850 0.0066 0.4146 0.1897 0.3129 0.3131 60.7375
0.2072 10.0 2000 2.5260 0.0066 0.4165 0.1911 0.3142 0.3143 60.615

Framework versions

  • Transformers 4.46.3
  • Pytorch 2.5.1+cu121
  • Datasets 3.1.0
  • Tokenizers 0.20.3
Downloads last month
46
Safetensors
Model size
406M params
Tensor type
F32
ยท
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for Jiraheya/pegasus_xsum_samsum_model_10epoch

Finetuned
(313)
this model

Space using Jiraheya/pegasus_xsum_samsum_model_10epoch 1