update model card README.md

README.md

@@ -0,0 +1,64 @@
+---
+tags:
+- generated_from_trainer
+model-index:
+- name: BertjeWDialDataALL04
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# BertjeWDialDataALL04
+
+This model is a fine-tuned version of [GroNLP/bert-base-dutch-cased](https://huggingface.co/GroNLP/bert-base-dutch-cased) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.0160
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 2e-05
+- train_batch_size: 16
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 64
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 8.0
+
+### Training results
+
+| Training Loss | Epoch | Step  | Validation Loss |
+|:-------------:|:-----:|:-----:|:---------------:|
+| 2.4695        | 1.0   | 1542  | 2.0689          |
+| 2.3622        | 2.0   | 3084  | 2.0509          |
+| 2.3303        | 3.0   | 4626  | 2.0796          |
+| 2.2865        | 4.0   | 6168  | 1.9921          |
+| 2.2598        | 5.0   | 7710  | 1.9844          |
+| 2.2544        | 6.0   | 9252  | 2.0077          |
+| 2.2128        | 7.0   | 10794 | 1.9914          |
+| 2.1963        | 8.0   | 12336 | 1.9882          |
+
+
+### Framework versions
+
+- Transformers 4.13.0.dev0
+- Pytorch 1.10.0
+- Datasets 1.16.1
+- Tokenizers 0.10.3