Update README.md
Browse files
README.md
CHANGED
@@ -48,15 +48,19 @@ model-index:
|
|
48 |
|
49 |
# flan-t5-large-stacked-samsum-1024
|
50 |
|
|
|
|
|
|
|
|
|
51 |
This model is a fine-tuned version of [google/flan-t5-large](https://huggingface.co/google/flan-t5-large) on the `stacked-summaries/stacked-samsum-1024` dataset.
|
52 |
|
53 |
It achieves the following results on the evaluation set:
|
54 |
-
- Loss: 2.
|
55 |
-
- Rouge1:
|
56 |
-
- Rouge2:
|
57 |
-
- Rougel: 44.
|
58 |
-
- Rougelsum: 54.
|
59 |
-
- Gen Len: 122.
|
60 |
|
61 |
## Model description
|
62 |
|
@@ -65,37 +69,38 @@ More information needed
|
|
65 |
## Intended uses & limitations
|
66 |
|
67 |
- max input/output is 1024 tokens
|
68 |
-
- this is mostly a test because `samsum` is not exactly the best dataset for general
|
69 |
|
70 |
## Training and evaluation data
|
71 |
|
72 |
-
|
73 |
|
74 |
## Training procedure
|
75 |
|
76 |
### Training hyperparameters
|
77 |
|
78 |
The following hyperparameters were used during training:
|
79 |
-
- learning_rate: 0.
|
80 |
-
- train_batch_size:
|
81 |
-
- eval_batch_size:
|
82 |
-
- seed:
|
83 |
- distributed_type: multi-GPU
|
84 |
-
- num_devices: 2
|
85 |
- gradient_accumulation_steps: 32
|
86 |
- total_train_batch_size: 256
|
87 |
-
- total_eval_batch_size: 4
|
88 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
89 |
- lr_scheduler_type: cosine
|
90 |
- lr_scheduler_warmup_ratio: 0.02
|
91 |
-
- num_epochs:
|
92 |
|
93 |
### Training results
|
94 |
|
95 |
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|
96 |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:|
|
97 |
-
| 0.
|
98 |
-
| 0.
|
|
|
|
|
|
|
99 |
|
100 |
|
101 |
### Framework versions
|
@@ -103,4 +108,4 @@ The following hyperparameters were used during training:
|
|
103 |
- Transformers 4.26.0.dev0
|
104 |
- Pytorch 1.13.0+cu117
|
105 |
- Datasets 2.6.1
|
106 |
-
- Tokenizers 0.13.1
|
|
|
48 |
|
49 |
# flan-t5-large-stacked-samsum-1024
|
50 |
|
51 |
+
<a href="https://colab.research.google.com/gist/pszemraj/a4bf61f593ebda9a8db6dc58839d9de4/brief-demo-flan-t5-stacked-samsum.ipynb">
|
52 |
+
<img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab"/>
|
53 |
+
</a>
|
54 |
+
|
55 |
This model is a fine-tuned version of [google/flan-t5-large](https://huggingface.co/google/flan-t5-large) on the `stacked-summaries/stacked-samsum-1024` dataset.
|
56 |
|
57 |
It achieves the following results on the evaluation set:
|
58 |
+
- Loss: 2.1846
|
59 |
+
- Rouge1: 57.9637
|
60 |
+
- Rouge2: 28.7446
|
61 |
+
- Rougel: 44.3826
|
62 |
+
- Rougelsum: 54.0399
|
63 |
+
- Gen Len: 122.77
|
64 |
|
65 |
## Model description
|
66 |
|
|
|
69 |
## Intended uses & limitations
|
70 |
|
71 |
- max input/output is 1024 tokens
|
72 |
+
- this is mostly a test because `samsum` is not exactly the best dataset for general-purpose summarization
|
73 |
|
74 |
## Training and evaluation data
|
75 |
|
76 |
+
See the dataset card linked on this page for info
|
77 |
|
78 |
## Training procedure
|
79 |
|
80 |
### Training hyperparameters
|
81 |
|
82 |
The following hyperparameters were used during training:
|
83 |
+
- learning_rate: 0.0001
|
84 |
+
- train_batch_size: 8
|
85 |
+
- eval_batch_size: 4
|
86 |
+
- seed: 24915
|
87 |
- distributed_type: multi-GPU
|
|
|
88 |
- gradient_accumulation_steps: 32
|
89 |
- total_train_batch_size: 256
|
|
|
90 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
91 |
- lr_scheduler_type: cosine
|
92 |
- lr_scheduler_warmup_ratio: 0.02
|
93 |
+
- num_epochs: 1.0
|
94 |
|
95 |
### Training results
|
96 |
|
97 |
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|
98 |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:|
|
99 |
+
| 0.1195 | 0.17 | 20 | 2.0635 | 57.8829 | 28.7887 | 44.4256 | 54.1299 | 121.8 |
|
100 |
+
| 0.1084 | 0.35 | 40 | 2.1178 | 58.0416 | 28.6487 | 44.3905 | 54.1557 | 122.893 |
|
101 |
+
| 0.1019 | 0.52 | 60 | 2.1576 | 57.816 | 28.7069 | 44.4242 | 53.9598 | 120.524 |
|
102 |
+
| 0.0975 | 0.7 | 80 | 2.1821 | 57.9597 | 28.8178 | 44.4854 | 54.068 | 121.793 |
|
103 |
+
| 0.0947 | 0.87 | 100 | 2.1846 | 57.9637 | 28.7446 | 44.3826 | 54.0399 | 122.77 |
|
104 |
|
105 |
|
106 |
### Framework versions
|
|
|
108 |
- Transformers 4.26.0.dev0
|
109 |
- Pytorch 1.13.0+cu117
|
110 |
- Datasets 2.6.1
|
111 |
+
- Tokenizers 0.13.1
|