# Trainer log my custom metrics at training step
**URL:**
**Author:**  [@bossalex](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fbossalex)%5C
**Post date:** [September 16, 2023, 4:41am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Ftrainer-log-my-custom-metrics-at-training-step%2F55204%2F1 "2023-09-16T04:41:38Z")
Hey there.
I’m using the Huggingface Trainer to finetune my model, and use tensorboard to display the mertics.
I find that the trainer only logs the train\_loss which is return by the model\_ouput. However, I wonder if there is a way for me to have more information logged during the train\_step, such as my own loss which is part the trian\_loss.
I check the trainer code
```auto
def _maybe_log_save_evaluate(self, tr_loss, model, trial, epoch, ignore_keys_for_eval):
if self.control.should_log:
if is_torch_tpu_available():
xm.mark_step()
logs: Dict[str, float] = {}
# all_gather + mean() to get average loss over all processes
tr_loss_scalar = self._nested_gather(tr_loss).mean().item()
# reset tr_loss to zero
tr_loss -= tr_loss
logs["loss"] = round(tr_loss_scalar / (self.state.global_step - self._globalstep_last_logged), 4)
logs["learning_rate"] = self._get_learning_rate()
self._total_loss_scalar += tr_loss_scalar
self._globalstep_last_logged = self.state.global_step
self.store_flos()
self.log(logs)
```
I think this code has defined the log metrics which I may not have the change to customize my own metric during the training step.
I would like to know if there are some ways to log.
Cheers & Thanks for an tips!
---
**Author:**  [@HaohuaLv](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FHaohuaLv)%5C
**Post date:** [December 18, 2023, 5:18pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Ftrainer-log-my-custom-metrics-at-training-step%2F55204%2F2 "2023-12-18T17:18:27Z")
I have the same problem. Have u solve it ?
---
**Author:**  [@ch272h](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fch272h)%5C
**Post date:** [January 24, 2024, 9:14pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Ftrainer-log-my-custom-metrics-at-training-step%2F55204%2F3 "2024-01-24T21:14:20Z")
If you don’t use gradient accumulation, then I usually just hack by overwriting `Trainer.compute_loss` and tucking in one line of `self.log(compute_my_metric(output)`
If you use gradient accumulation, one alternative is to trigger a CustomCallback per [Metrics for Training Set in Trainer - #7 by Kaveri](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fmetrics-for-training-set-in-trainer%2F2461%2F7). For example, you can do one forward pass on the entire train set on\_epoch\_end or on\_evaluate. It will be repeated work, slow and coarse.
And let me know if you figured out an easy way to log custom loss!
---
**Author:**  [@zipzou](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fzipzou)%5C
**Post date:** [July 11, 2024, 2:19am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Ftrainer-log-my-custom-metrics-at-training-step%2F55204%2F4 "2024-07-11T02:19:23Z")
[https://github.com/zipzou/hf-multitask-trainer](https://github.com/zipzou/hf-multitask-trainer)
You can try to use this to track different losses.