# TypeErrors when passing search spaces to trainable function
**URL:** https://discuss.ray.io/t/typeerrors-when-passing-search-spaces-to-trainable-function/4591
**Category:** Ray Tune
**Created:** [January 5, 2022, 8:30pm UTC](https://discuss.ray.io/t/typeerrors-when-passing-search-spaces-to-trainable-function/4591 "2022-01-05T20:30:17Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@DomByrne](https://discuss.ray.io/u/DomByrne)
#### Post date: [January 5, 2022, 8:30pm UTC](https://discuss.ray.io/t/typeerrors-when-passing-search-spaces-to-trainable-function/4591/1 "2022-01-05T20:30:17Z")
I’m trying to use ray tune to train a pytorch model and I get type errors, suggesting that the whole search space distribution is being passed to the function rather than actual values from the distribution?
Code excerpts:
#search space
config = {
‘hidden\_size’: tune.sample\_from(lambda \_: 2 \*\* np.random.randint(7, 10)),
‘num\_hidden’: tune.sample\_from(lambda \_: np.arange(1,5)),
‘lr’: tune.loguniform(1e-4, 1e-1),
‘batch\_size’: tune.sample\_from(lambda \_: 2 \*\* np.random.randint(7, 10))
}
result = tune.run(
partial(train\_model, config, checkpoint\_dir),
resources\_per\_trial={“cpu”: cpus\_per\_trial, “gpu”: gpus\_per\_trial},
config=config,
num\_samples=num\_resamples,
max\_concurrent\_trials=max\_concurrent,
scheduler=scheduler,
progress\_reporter=reporter,
fail\_fast=‘raise’
)
The most relevant portion of the error message:
File “/Users/dombyrne/phd/synthetic\_genotypes/src/synthetic\_geno\_nn\_cls.py”, line 87, in train\_model
model = MLP(config[‘hidden\_size’], config[‘num\_hidden’], config[‘input\_size’])
File “/Users/dombyrne/phd/synthetic\_genotypes/src/synthetic\_geno\_nn\_cls.py”, line 44, in **init**
l\_sizes = [input\_size] + [hidden\_size] \* int(num\_hidden) + [output\_size]
TypeError: int() argument must be a string, a bytes-like object or a number, not ‘Categorical’
I’ve tried reinstalling tune using pip, which makes no difference. I’m using ray version 1.9.0.
Best wishes,
Dom
---
### Author:  [@matthewdeng](https://discuss.ray.io/u/matthewdeng)
#### Post date: [January 5, 2022, 8:38pm UTC](https://discuss.ray.io/t/typeerrors-when-passing-search-spaces-to-trainable-function/4591/2 "2022-01-05T20:38:12Z")
Hey @DomByrne,
What happens if you just specify `train_model` as the trainable function?
```diff
- partial(train_model, config, checkpoint_dir),
+ train_model,
```
Tune should generate the configuration and pass it into `train_model` during execution.
---
### Author:  [@DomByrne](https://discuss.ray.io/u/DomByrne)
#### Post date: [January 6, 2022, 3:53pm UTC](https://discuss.ray.io/t/typeerrors-when-passing-search-spaces-to-trainable-function/4591/3 "2022-01-06T15:53:41Z")
That works, thank you!