Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- summarize_from_feedback
|
| 7 |
+
metrics:
|
| 8 |
+
- rouge
|
| 9 |
+
model-index:
|
| 10 |
+
- name: flan-t5-large-finetuned-openai-summarize_from_feedback
|
| 11 |
+
results:
|
| 12 |
+
- task:
|
| 13 |
+
name: Sequence-to-sequence Language Modeling
|
| 14 |
+
type: text2text-generation
|
| 15 |
+
dataset:
|
| 16 |
+
name: summarize_from_feedback
|
| 17 |
+
type: summarize_from_feedback
|
| 18 |
+
config: comparisons
|
| 19 |
+
split: train
|
| 20 |
+
args: comparisons
|
| 21 |
+
metrics:
|
| 22 |
+
- name: Rouge1
|
| 23 |
+
type: rouge
|
| 24 |
+
value: 30.2401
|
| 25 |
+
- name: Rouge2
|
| 26 |
+
type: rouge
|
| 27 |
+
value: 11.4916
|
| 28 |
+
- name: RougeL
|
| 29 |
+
type: rouge
|
| 30 |
+
value: 24.6485
|
| 31 |
+
- name: RougeLSum
|
| 32 |
+
type: rouge
|
| 33 |
+
value: 26.1801
|
| 34 |
+
pipeline_tag: summarization
|
| 35 |
+
---
|
| 36 |
+
|
| 37 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 38 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 39 |
+
|
| 40 |
+
# flan-t5-large-finetuned-openai-summarize_from_feedback
|
| 41 |
+
|
| 42 |
+
This model is a fine-tuned version of [google/flan-t5-large](https://huggingface.co/google/flan-t5-large) on the summarize_from_feedback dataset.
|
| 43 |
+
It achieves the following results on the evaluation set:
|
| 44 |
+
- Loss: 2.3118
|
| 45 |
+
- Rouge1: 30.2401
|
| 46 |
+
- Rouge2: 11.4916
|
| 47 |
+
- Rougel: 24.6485
|
| 48 |
+
- Rougelsum: 26.1801
|
| 49 |
+
- Gen Len: 18.8428
|
| 50 |
+
|
| 51 |
+
## Model description
|
| 52 |
+
|
| 53 |
+
More information needed
|
| 54 |
+
|
| 55 |
+
## Intended uses & limitations
|
| 56 |
+
|
| 57 |
+
More information needed
|
| 58 |
+
|
| 59 |
+
## Training and evaluation data
|
| 60 |
+
|
| 61 |
+
More information needed
|
| 62 |
+
|
| 63 |
+
## Training procedure
|
| 64 |
+
|
| 65 |
+
### Training hyperparameters
|
| 66 |
+
|
| 67 |
+
The following hyperparameters were used during training:
|
| 68 |
+
- learning_rate: 5e-05
|
| 69 |
+
- train_batch_size: 16
|
| 70 |
+
- eval_batch_size: 32
|
| 71 |
+
- seed: 42
|
| 72 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 73 |
+
- lr_scheduler_type: linear
|
| 74 |
+
- num_epochs: 6
|
| 75 |
+
|
| 76 |
+
### Training results
|
| 77 |
+
|
| 78 |
+
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|
| 79 |
+
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:|
|
| 80 |
+
| 1.7678 | 1.0 | 5804 | 1.8833 | 29.3494 | 10.9406 | 23.9907 | 25.461 | 18.9265 |
|
| 81 |
+
| 1.5839 | 2.0 | 11608 | 1.8992 | 29.6239 | 11.1795 | 24.2927 | 25.7183 | 18.9358 |
|
| 82 |
+
| 1.4812 | 3.0 | 17412 | 1.8929 | 29.8899 | 11.2855 | 24.4193 | 25.9219 | 18.9189 |
|
| 83 |
+
| 1.4198 | 4.0 | 23216 | 1.8939 | 29.8897 | 11.2606 | 24.3262 | 25.8642 | 18.9309 |
|
| 84 |
+
| 1.3612 | 5.0 | 29020 | 1.9105 | 29.8469 | 11.2112 | 24.2483 | 25.7884 | 18.9396 |
|
| 85 |
+
| 1.3279 | 6.0 | 34824 | 1.9170 | 30.038 | 11.3426 | 24.4385 | 25.9675 | 18.9328 |
|
| 86 |
+
|
| 87 |
+
|
| 88 |
+
### Framework versions
|
| 89 |
+
|
| 90 |
+
- Transformers 4.25.1
|
| 91 |
+
- Pytorch 1.13.0+cu116
|
| 92 |
+
- Datasets 2.8.0
|
| 93 |
+
- Tokenizers 0.13.2
|