bart-abs-1509-0313-lr-0.0003-bs-4-maxep-6

roequitz/bart-abs-1509-0313-lr-0.0003-bs-4-maxep-6

bart-abs-1509-0313-lr-0.0003-bs-4-maxep-6 is a machine learning model from roequitz. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

16% of all-time downloads

All-time downloads

Public

Parameters

306M

1.2 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors1.2 GB · 100%

At a glance

License: apache-2.0
Model type: bart
Access: Public
Created: Sep 15, 2024
Updated: Sep 15, 2024
SHA: ed1fbca7

Base models

sshleifer/distilbart-xsum-12-6

Model card

From the Hugging Face model README

bart-abs-1509-0313-lr-0.0003-bs-4-maxep-6

This model is a fine-tuned version of sshleifer/distilbart-xsum-12-6 on the None dataset. It achieves the following results on the evaluation set:

Loss: 3.8541
Rouge/rouge1: 0.4315
Rouge/rouge2: 0.1861
Rouge/rougel: 0.3638
Rouge/rougelsum: 0.3654
Bertscore/bertscore-precision: 0.8936
Bertscore/bertscore-recall: 0.8875
Bertscore/bertscore-f1: 0.8904
Meteor: 0.3814
Gen Len: 35.4273

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: 0.0003
train_batch_size: 4
eval_batch_size: 4
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 6
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge/rouge1	Rouge/rouge2	Rouge/rougel	Rouge/rougelsum	Bertscore/bertscore-precision	Bertscore/bertscore-recall	Bertscore/bertscore-f1	Meteor	Gen Len
2.1071	1.0	217	2.5496	0.4238	0.1753	0.3521	0.3541	0.8943	0.8836	0.8888	0.3359	32.1
1.5392	2.0	434	2.5867	0.4296	0.1807	0.3556	0.3551	0.8921	0.886	0.8889	0.3598	35.0909
1.0328	3.0	651	2.6952	0.4096	0.1667	0.3444	0.3453	0.8919	0.8826	0.8871	0.3519	33.8818
0.62	4.0	868	2.9126	0.4104	0.16	0.3478	0.3487	0.8904	0.8815	0.8858	0.3524	33.4273
0.3251	5.0	1085	3.3250	0.43	0.1771	0.3591	0.3598	0.8935	0.8861	0.8896	0.3744	34.9636
0.1503	6.0	1302	3.8541	0.4315	0.1861	0.3638	0.3654	0.8936	0.8875	0.8904	0.3814	35.4273

Framework versions

Transformers 4.44.0
Pytorch 2.4.0
Datasets 2.21.0
Tokenizers 0.19.1

Type: bart
License: apache-2.0
Created: Sep 15, 2024
Updated: Sep 15, 2024