# Cannot see training accuracy, only validation accuracy

**URL:** <https://discuss.huggingface.co/t/cannot-see-training-accuracy-only-validation-accuracy/61807>\
**Category:** 🤗Transformers\
**Created:** [November 10, 2023, 1:37pm UTC](https://discuss.huggingface.co/t/cannot-see-training-accuracy-only-validation-accuracy/61807 "2023-11-10T13:37:01Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![ialuronico](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/ialuronico/32/33223_2.png) [@ialuronico](https://discuss.huggingface.co/u/ialuronico)\
**Post date:** [November 10, 2023, 1:37pm UTC](https://discuss.huggingface.co/t/cannot-see-training-accuracy-only-validation-accuracy/61807/1 "2023-11-10T13:37:01Z")

</div>

I am using this piece of code to fine tune a transformer model for text classification in a Jupyter notebook:

```auto
metric = evaluate.load("accuracy")
def compute_metrics(eval_pred):
    logits, labels = eval_pred
    predictions = np.argmax(logits, axis=-1)
    return metric.compute(predictions=predictions, references=labels)

training_args = TrainingArguments(output_dir = "output_fine_tuning",
                                  evaluation_strategy = "epoch",
                                  logging_strategy="epoch",
                                  num_train_epochs = 3)
trainer = Trainer(
    model=model,
    args=training_args,
    train_dataset=tokenized_datasets['train'],
    eval_dataset=tokenized_datasets['test'],
    compute_metrics=compute_metrics,
)
train = trainer.train()

```

I only see printing out **Training Loss, Validation Loss, Accuracy** , where the last one is validation accuracy.

I tried to use techniques from here: [python - How to get the accuracy per epoch or step for the huggingface.transformers Trainer? - Stack Overflow](https://stackoverflow.com/questions/67457480/how-to-get-the-accuracy-per-epoch-or-step-for-the-huggingface-transformers-train) but nothing really worked.

---

<div class="post-metadata">

**Author:** ![celsowm](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/celsowm/32/50357_2.png) [@celsowm](https://discuss.huggingface.co/u/celsowm)\
**Post date:** [February 19, 2024, 10:48pm UTC](https://discuss.huggingface.co/t/cannot-see-training-accuracy-only-validation-accuracy/61807/2 "2024-02-19T22:48:37Z")

</div>

`results = trainer.evaluate()`

---

<div class="post-metadata">

**Author:** ![nielsr](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/nielsr/32/39617_2.png) [@nielsr](https://discuss.huggingface.co/u/nielsr)\
**Post date:** [February 20, 2024, 8:04am UTC](https://discuss.huggingface.co/t/cannot-see-training-accuracy-only-validation-accuracy/61807/3 "2024-02-20T08:04:51Z")

</div>

Refer to this thread: [Metrics for Training Set in Trainer](https://discuss.huggingface.co/t/metrics-for-training-set-in-trainer/2461)