metadata

license: mit
base_model: facebook/bart-large-cnn
tags:
  - generated_from_trainer
metrics:
  - rouge
model-index:
  - name: bart-large-cnn-finetuned-scope-summarization
    results: []

bart-large-cnn-finetuned-scope-summarization

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: 0.0552
Rouge1: 49.8374
Rouge2: 38.0885
Rougel: 42.6985
Rougelsum: 42.4809

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: 5.6e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
0.6831	1.0	43	0.3928	40.6965	25.3494	30.1716	29.9938
0.3578	2.0	86	0.3598	43.284	27.9071	32.9941	32.9077
0.3302	3.0	129	0.3362	45.2375	30.4709	34.8733	34.6801
0.309	4.0	172	0.3136	44.928	30.8601	34.7804	34.6754
0.2948	5.0	215	0.2919	44.5169	30.2429	34.5979	34.4672
0.2841	6.0	258	0.2755	45.7172	31.6555	34.9668	34.9069
0.268	7.0	301	0.2618	46.4085	32.782	35.804	35.6348
0.252	8.0	344	0.2424	47.8634	33.6728	36.9559	36.9081
0.2405	9.0	387	0.2286	46.8182	34.4363	37.7534	37.6356
0.2255	10.0	430	0.2276	46.8516	33.3166	37.6246	37.5024
0.2125	11.0	473	0.1946	47.6772	33.9627	37.8554	37.7735
0.1918	12.0	516	0.1682	46.851	33.6098	38.2906	38.24
0.1726	13.0	559	0.1442	48.8833	36.4235	39.4263	39.1955
0.152	14.0	602	0.1305	50.5835	39.2008	43.3793	43.1671
0.1344	15.0	645	0.1109	47.3517	35.4446	38.0845	38.0578
0.116	16.0	688	0.0842	48.9774	37.6705	41.6306	41.4792
0.1007	17.0	731	0.0762	49.9775	38.4186	42.647	42.4334
0.0899	18.0	774	0.0623	50.1358	38.9943	43.4025	43.1603
0.0805	19.0	817	0.0571	51.5974	40.1928	44.1821	43.9354
0.0753	20.0	860	0.0552	49.8374	38.0885	42.6985	42.4809

Framework versions

Transformers 4.38.2
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.15.2