Upload README.md

README.md

@@ -1,3 +1,55 @@
 ---
-license: cc-by-nc-4.0
+tags:
+- generated_from_trainer
+model-index:
+- name: som_xlmr
+  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. -->
+
+# som_xlmr
+
+This model is a fine-tuned version of [models/som_xlmr/](https://huggingface.co/models/som_xlmr/) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- eval_loss: 0.9361
+- eval_runtime: 28.4052
+- eval_samples_per_second: 32.177
+- eval_steps_per_second: 4.049
+- step: 0
+
+## 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: 5e-05
+- train_batch_size: 5
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 10
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 3.0
+
+### Framework versions
+
+- Transformers 4.21.1
+- Pytorch 1.7.1+cu110
+- Datasets 1.16.1
+- Tokenizers 0.12.1