liveinsoccer
/

distilbert-base-multilingual-cased-finetuned-psi-classification-gpu

Generated from Trainer

Model card Files Files and versions Community

liveinsoccer commited on Aug 18

Commit

299cb59

•

1 Parent(s): cbcd74d

training completed

Files changed (1) hide show

README.md +14 -19

README.md CHANGED Viewed

@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
 This model is a fine-tuned version of [distilbert-base-multilingual-cased](https://huggingface.co/distilbert-base-multilingual-cased) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9416
-- Accuracy: 0.8130
-- F1: 0.8172
 ## Model description
@@ -45,27 +45,22 @@ The following hyperparameters were used during training:
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 15
 ### Training results
 | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
-| 1.2344        | 1.0   | 120  | 0.9301          | 0.7214   | 0.6924 |
-| 0.8454        | 2.0   | 240  | 0.7082          | 0.7672   | 0.7448 |
-| 0.6587        | 3.0   | 360  | 0.6435          | 0.8053   | 0.8080 |
-| 0.5137        | 4.0   | 480  | 0.6696          | 0.8015   | 0.8038 |
-| 0.3836        | 5.0   | 600  | 0.6537          | 0.8359   | 0.8326 |
-| 0.3019        | 6.0   | 720  | 0.7467          | 0.8053   | 0.8095 |
-| 0.2521        | 7.0   | 840  | 0.7103          | 0.8282   | 0.8319 |
-| 0.1759        | 8.0   | 960  | 0.7775          | 0.8092   | 0.8188 |
-| 0.1588        | 9.0   | 1080 | 0.8922          | 0.7977   | 0.8106 |
-| 0.1254        | 10.0  | 1200 | 0.8947          | 0.8053   | 0.8111 |
-| 0.095         | 11.0  | 1320 | 0.8608          | 0.8244   | 0.8282 |
-| 0.0849        | 12.0  | 1440 | 0.8891          | 0.8168   | 0.8176 |
-| 0.0719        | 13.0  | 1560 | 0.9622          | 0.7977   | 0.8101 |
-| 0.0644        | 14.0  | 1680 | 0.9314          | 0.8168   | 0.8209 |
-| 0.0525        | 15.0  | 1800 | 0.9416          | 0.8130   | 0.8172 |
 ### Framework versions

 This model is a fine-tuned version of [distilbert-base-multilingual-cased](https://huggingface.co/distilbert-base-multilingual-cased) on the None dataset.
 It achieves the following results on the evaluation set:
+- Loss: 0.7609
+- Accuracy: 0.8321
+- F1: 0.8279
 ## Model description
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
+- num_epochs: 10
 ### Training results
 | Training Loss | Epoch | Step | Validation Loss | Accuracy | F1     |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|
+| 1.2028        | 1.0   | 120  | 0.9207          | 0.7137   | 0.6561 |
+| 0.8459        | 2.0   | 240  | 0.7827          | 0.7443   | 0.7268 |
+| 0.685         | 3.0   | 360  | 0.6843          | 0.7901   | 0.7792 |
+| 0.5474        | 4.0   | 480  | 0.6593          | 0.8015   | 0.8075 |
+| 0.4037        | 5.0   | 600  | 0.6393          | 0.8244   | 0.8211 |
+| 0.3282        | 6.0   | 720  | 0.6804          | 0.8015   | 0.8060 |
+| 0.2728        | 7.0   | 840  | 0.6991          | 0.8321   | 0.8289 |
+| 0.1979        | 8.0   | 960  | 0.7389          | 0.8206   | 0.8196 |
+| 0.1795        | 9.0   | 1080 | 0.7615          | 0.8244   | 0.8231 |
+| 0.1676        | 10.0  | 1200 | 0.7609          | 0.8321   | 0.8279 |
 ### Framework versions