Commit
•
69d3286
1
Parent(s):
c002b4a
update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,87 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- un_multi
|
| 7 |
+
metrics:
|
| 8 |
+
- bleu
|
| 9 |
+
model-index:
|
| 10 |
+
- name: m2m100_418M-evaluated-en-to-ar-2000instancesUNMULTI-leaningRate2e-05-batchSize8-regu1
|
| 11 |
+
results:
|
| 12 |
+
- task:
|
| 13 |
+
name: Sequence-to-sequence Language Modeling
|
| 14 |
+
type: text2text-generation
|
| 15 |
+
dataset:
|
| 16 |
+
name: un_multi
|
| 17 |
+
type: un_multi
|
| 18 |
+
args: ar-en
|
| 19 |
+
metrics:
|
| 20 |
+
- name: Bleu
|
| 21 |
+
type: bleu
|
| 22 |
+
value: 41.8577
|
| 23 |
+
---
|
| 24 |
+
|
| 25 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 26 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 27 |
+
|
| 28 |
+
# m2m100_418M-evaluated-en-to-ar-2000instancesUNMULTI-leaningRate2e-05-batchSize8-regu1
|
| 29 |
+
|
| 30 |
+
This model is a fine-tuned version of [facebook/m2m100_418M](https://huggingface.co/facebook/m2m100_418M) on the un_multi dataset.
|
| 31 |
+
It achieves the following results on the evaluation set:
|
| 32 |
+
- Loss: 0.3603
|
| 33 |
+
- Bleu: 41.8577
|
| 34 |
+
- Meteor: 0.4199
|
| 35 |
+
- Gen Len: 41.9
|
| 36 |
+
|
| 37 |
+
## Model description
|
| 38 |
+
|
| 39 |
+
More information needed
|
| 40 |
+
|
| 41 |
+
## Intended uses & limitations
|
| 42 |
+
|
| 43 |
+
More information needed
|
| 44 |
+
|
| 45 |
+
## Training and evaluation data
|
| 46 |
+
|
| 47 |
+
More information needed
|
| 48 |
+
|
| 49 |
+
## Training procedure
|
| 50 |
+
|
| 51 |
+
### Training hyperparameters
|
| 52 |
+
|
| 53 |
+
The following hyperparameters were used during training:
|
| 54 |
+
- learning_rate: 2e-05
|
| 55 |
+
- train_batch_size: 8
|
| 56 |
+
- eval_batch_size: 8
|
| 57 |
+
- seed: 42
|
| 58 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 59 |
+
- lr_scheduler_type: linear
|
| 60 |
+
- num_epochs: 11
|
| 61 |
+
- mixed_precision_training: Native AMP
|
| 62 |
+
|
| 63 |
+
### Training results
|
| 64 |
+
|
| 65 |
+
| Training Loss | Epoch | Step | Validation Loss | Bleu | Meteor | Gen Len |
|
| 66 |
+
|:-------------:|:-----:|:----:|:---------------:|:-------:|:------:|:-------:|
|
| 67 |
+
| 5.111 | 0.5 | 100 | 3.2467 | 29.5017 | 0.3371 | 42.425 |
|
| 68 |
+
| 2.1491 | 1.0 | 200 | 1.0018 | 33.0563 | 0.3593 | 41.205 |
|
| 69 |
+
| 0.5911 | 1.5 | 300 | 0.4159 | 34.5818 | 0.3705 | 42.0625 |
|
| 70 |
+
| 0.3546 | 2.0 | 400 | 0.3723 | 36.6179 | 0.3823 | 40.925 |
|
| 71 |
+
| 0.2487 | 2.5 | 500 | 0.3595 | 39.0331 | 0.3956 | 41.56 |
|
| 72 |
+
| 0.2365 | 3.0 | 600 | 0.3485 | 39.5188 | 0.4023 | 41.6425 |
|
| 73 |
+
| 0.1687 | 3.5 | 700 | 0.3542 | 40.1728 | 0.4043 | 42.61 |
|
| 74 |
+
| 0.1791 | 4.0 | 800 | 0.3466 | 40.4858 | 0.4101 | 41.5575 |
|
| 75 |
+
| 0.1196 | 4.5 | 900 | 0.3493 | 41.2457 | 0.4123 | 41.755 |
|
| 76 |
+
| 0.1394 | 5.0 | 1000 | 0.3486 | 40.5606 | 0.4114 | 41.78 |
|
| 77 |
+
| 0.0958 | 5.5 | 1100 | 0.3568 | 41.1873 | 0.4157 | 41.7275 |
|
| 78 |
+
| 0.1043 | 6.0 | 1200 | 0.3557 | 41.2749 | 0.4165 | 41.935 |
|
| 79 |
+
| 0.073 | 6.5 | 1300 | 0.3603 | 41.8577 | 0.4199 | 41.9 |
|
| 80 |
+
|
| 81 |
+
|
| 82 |
+
### Framework versions
|
| 83 |
+
|
| 84 |
+
- Transformers 4.20.1
|
| 85 |
+
- Pytorch 1.11.0
|
| 86 |
+
- Datasets 2.1.0
|
| 87 |
+
- Tokenizers 0.12.1
|