Canonical train 20 epochs with 18 batch size

README.md

@@ -13,18 +13,18 @@ should probably proofread and complete it, then remove this comment. -->
 
 # bert2bert-extabs-canonicalcleandata-lr-5e-05-batchsize-4-encmaxlen-512-decmaxlen-256
 
-This model is a fine-tuned version of [](https://huggingface.co/) on the id_liputan6 dataset.
+This model was trained from scratch on the id_liputan6 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 8.0476
-- R1 Precision: 0.0
-- R1 Recall: 0.0
-- R1 Fmeasure: 0.0
-- R2 Precision: 0.0
-- R2 Recall: 0.0
-- R2 Fmeasure: 0.0
-- Rl Precision: 0.0
-- Rl Recall: 0.0
-- Rl Fmeasure: 0.0
+- Loss: 0.2026
+- R1 Precision: 0.5676
+- R1 Recall: 0.2496
+- R1 Fmeasure: 0.3412
+- R2 Precision: 0.4787
+- R2 Recall: 0.1992
+- R2 Fmeasure: 0.2767
+- Rl Precision: 0.5419
+- Rl Recall: 0.2376
+- Rl Fmeasure: 0.3251
 
 ## Model description
 
@@ -44,25 +44,33 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 5e-05
-- train_batch_size: 1
-- eval_batch_size: 1
+- train_batch_size: 18
+- eval_batch_size: 18
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 2
+- num_epochs: 10
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | R1 Precision | R1 Recall | R1 Fmeasure | R2 Precision | R2 Recall | R2 Fmeasure | Rl Precision | Rl Recall | Rl Fmeasure |
-|:-------------:|:-----:|:----:|:---------------:|:------------:|:---------:|:-----------:|:------------:|:---------:|:-----------:|:------------:|:---------:|:-----------:|
-| No log        | 1.0   | 8    | 8.3802          | 0.0          | 0.0       | 0.0         | 0.0          | 0.0       | 0.0         | 0.0          | 0.0       | 0.0         |
-| No log        | 2.0   | 16   | 8.0476          | 0.0          | 0.0       | 0.0         | 0.0          | 0.0       | 0.0         | 0.0          | 0.0       | 0.0         |
+| Training Loss | Epoch | Step   | Validation Loss | R1 Precision | R1 Recall | R1 Fmeasure | R2 Precision | R2 Recall | R2 Fmeasure | Rl Precision | Rl Recall | Rl Fmeasure |
+|:-------------:|:-----:|:------:|:---------------:|:------------:|:---------:|:-----------:|:------------:|:---------:|:-----------:|:------------:|:---------:|:-----------:|
+| 0.9444        | 1.0   | 10772  | 0.1834          | 0.5328       | 0.2367    | 0.3226      | 0.4379       | 0.1842    | 0.2551      | 0.5062       | 0.2243    | 0.3059      |
+| 0.1703        | 2.0   | 21544  | 0.1605          | 0.565        | 0.2474    | 0.3386      | 0.4764       | 0.1974    | 0.2745      | 0.5403       | 0.236     | 0.3232      |
+| 0.1407        | 3.0   | 32316  | 0.1567          | 0.5553       | 0.242     | 0.3318      | 0.4652       | 0.1914    | 0.2669      | 0.5302       | 0.2303    | 0.3161      |
+| 0.1199        | 4.0   | 43088  | 0.1565          | 0.5651       | 0.2446    | 0.3361      | 0.4774       | 0.1954    | 0.2728      | 0.5408       | 0.2333    | 0.3209      |
+| 0.1022        | 5.0   | 53860  | 0.1578          | 0.5681       | 0.2449    | 0.3368      | 0.4818       | 0.1962    | 0.2744      | 0.5442       | 0.2338    | 0.322       |
+| 0.0864        | 6.0   | 64632  | 0.1614          | 0.5657       | 0.2457    | 0.3373      | 0.4779       | 0.196     | 0.2735      | 0.5409       | 0.2341    | 0.3217      |
+| 0.0709        | 7.0   | 75404  | 0.1712          | 0.5669       | 0.2473    | 0.339       | 0.4782       | 0.1974    | 0.275       | 0.5418       | 0.2356    | 0.3233      |
+| 0.0579        | 8.0   | 86176  | 0.1830          | 0.5629       | 0.2499    | 0.3407      | 0.472        | 0.1987    | 0.2749      | 0.5366       | 0.2377    | 0.3242      |
+| 0.0458        | 9.0   | 96948  | 0.1955          | 0.5708       | 0.2513    | 0.3435      | 0.4817       | 0.2009    | 0.2789      | 0.5451       | 0.2393    | 0.3273      |
+| 0.0364        | 10.0  | 107720 | 0.2026          | 0.5676       | 0.2496    | 0.3412      | 0.4787       | 0.1992    | 0.2767      | 0.5419       | 0.2376    | 0.3251      |
 
 
 ### Framework versions
 
-- Transformers 4.38.2
-- Pytorch 2.2.1+cu121
+- Transformers 4.39.3
+- Pytorch 2.2.1
 - Datasets 2.18.0
 - Tokenizers 0.15.2

generation_config.json

@@ -1,5 +1,5 @@
 {
   "bos_token_id": 0,
   "pad_token_id": 0,
-  "transformers_version": "4.38.2"
+  "transformers_version": "4.39.3"
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a2b24955e0750787db0a98b3d1f334b68021368b4b58d4ea5790ca884df28b06
+oid sha256:8748c352c242bed75265cc4efeb80d194fd94089c85feb8c64dfcc222b1ff779
 size 998132132

runs/Apr16_07-18-52_728561d7d15f/events.out.tfevents.1713251933.728561d7d15f.381.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:456b72d974ac9a6c853b5187494d189b7b9f38e4d945c99ab417f522aab8a202
-size 18338
+oid sha256:fa93f73e9a7fc19e312be559f59bfdb65102216debde96420863353b260462c3
+size 19475