update model card README.md

README.md

@@ -0,0 +1,59 @@
+---
+license: apache-2.0
+tags:
+- generated_from_trainer
+datasets:
+- squad
+model-index:
+- name: distilbert-base-uncased-finetuned-triviaqa-finetuned-squad
+  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. -->
+
+# distilbert-base-uncased-finetuned-triviaqa-finetuned-squad
+
+This model is a fine-tuned version of [FabianWillner/distilbert-base-uncased-finetuned-triviaqa](https://huggingface.co/FabianWillner/distilbert-base-uncased-finetuned-triviaqa) on the squad dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.1417
+
+## 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: 16
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 2
+
+### Training results
+
+| Training Loss | Epoch | Step  | Validation Loss |
+|:-------------:|:-----:|:-----:|:---------------:|
+| 1.2153        | 1.0   | 5533  | 1.1555          |
+| 0.9614        | 2.0   | 11066 | 1.1417          |
+
+
+### Framework versions
+
+- Transformers 4.19.2
+- Pytorch 1.11.0+cu113
+- Datasets 2.2.2
+- Tokenizers 0.12.1