---
license: apache-2.0
base_model: facebook/bart-large
tags:
- generated_from_trainer
datasets:
- learn3r/gov_report_memsum_oracle
metrics:
- rouge
model-index:
- name: bart_large_gov
  results:
  - task:
      name: Summarization
      type: summarization
    dataset:
      name: learn3r/gov_report_memsum_oracle
      type: learn3r/gov_report_memsum_oracle
    metrics:
    - name: Rouge1
      type: rouge
      value: 56.2783
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# bart_large_gov

This model is a fine-tuned version of [facebook/bart-large](https://huggingface.co/facebook/bart-large) on the learn3r/gov_report_memsum_oracle dataset.
It achieves the following results on the evaluation set:
- Loss: 1.4450
- Rouge1: 56.2783
- Rouge2: 31.1387
- Rougel: 39.2121
- Rougelsum: 51.8068
- Gen Len: 128.5062

## 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-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20.0

### Training results

| Training Loss | Epoch | Step | Validation Loss | Rouge1  | Rouge2  | Rougel  | Rougelsum | Gen Len  |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:--------:|
| 1.6694        | 1.0   | 136  | 1.5338          | 54.2061 | 29.3577 | 37.2911 | 49.8337   | 139.7253 |
| 1.5178        | 1.99  | 272  | 1.4698          | 55.6621 | 30.6254 | 38.7491 | 51.2934   | 128.9475 |
| 1.4208        | 3.0   | 409  | 1.4487          | 55.4905 | 30.4201 | 38.412  | 51.1108   | 129.5658 |
| 1.3399        | 3.99  | 545  | 1.4450          | 56.2783 | 31.1387 | 39.2121 | 51.8068   | 128.5062 |
| 1.2326        | 5.0   | 682  | 1.4478          | 56.0182 | 30.7104 | 38.8337 | 51.6162   | 129.1358 |
| 1.1784        | 6.0   | 818  | 1.4533          | 56.4333 | 31.4483 | 39.5546 | 52.1347   | 128.7315 |
| 1.1739        | 7.0   | 955  | 1.4607          | 56.3636 | 31.1125 | 39.4055 | 51.9709   | 128.8241 |
| 1.1585        | 8.0   | 1091 | 1.4774          | 55.9356 | 30.7012 | 38.7824 | 51.5664   | 128.9640 |
| 1.0297        | 8.99  | 1227 | 1.4939          | 56.7487 | 31.552  | 39.6461 | 52.411    | 128.6553 |
| 1.0085        | 10.0  | 1364 | 1.5075          | 56.3918 | 31.2201 | 39.4213 | 51.9449   | 128.6265 |
| 0.9738        | 10.99 | 1500 | 1.5237          | 56.3041 | 30.9239 | 39.2625 | 51.8217   | 128.8282 |
| 0.9583        | 12.0  | 1637 | 1.5444          | 55.6539 | 30.2395 | 38.5901 | 51.1518   | 128.9136 |
| 0.9601        | 12.99 | 1773 | 1.5516          | 55.9154 | 30.5471 | 38.8607 | 51.2856   | 128.9784 |
| 0.8882        | 14.0  | 1910 | 1.5736          | 56.3282 | 30.9807 | 39.2351 | 51.8022   | 128.5206 |
| 0.851         | 15.0  | 2046 | 1.5891          | 56.0531 | 30.6748 | 38.8847 | 51.5739   | 128.7623 |
| 0.8825        | 16.0  | 2183 | 1.5978          | 56.0084 | 30.7943 | 38.9692 | 51.5587   | 128.7798 |
| 0.8169        | 17.0  | 2319 | 1.6076          | 55.8274 | 30.41   | 38.6258 | 51.3009   | 128.8632 |
| 0.8194        | 17.99 | 2455 | 1.6177          | 56.3214 | 30.9896 | 39.4754 | 51.9525   | 128.6461 |
| 0.8441        | 19.0  | 2592 | 1.6260          | 55.9842 | 30.6332 | 38.999  | 51.5685   | 128.8241 |
| 0.792         | 19.94 | 2720 | 1.6328          | 55.8983 | 30.5018 | 38.7764 | 51.3611   | 128.7407 |


### Framework versions

- Transformers 4.37.0.dev0
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.15.0