change-log.txt

# train_sdss.py
- changed z_max from 0.5 to 2.0 
- changed tags in data loader functions to train and valid. 
```
trainloader = SDSS.get_data_loader(args.dir, tag="train", which="train",
validloader = SDSS.get_data_loader(args.dir, tag="valid", which="valid",
```

# util.py
- 

# sdss.py 
- temporarily named classname to BOSS 
```
classname = cls.__mro__[0].__name__
filename = f"{classname}{tag}_*.pkl"
```

- changed how `batch_files` is found
```
#batch_files = glob.glob(dir + "/" + filename)
batch_files = []

for pickle_file in os.listdir(dir):
    if pickle_file.split('_')[0] == f'{classname}{tag}' and pickle_file.split('.')[-1] = '.pkl':
        batch_files.append(os.path.join(dir, pickle_file))
```

- changed it just to return batches 

```
accelerator = Accelerator(mixed_precision='fp16', cpu=True) # changed cpu to True? --> i reverted to having cpu=False
```

## working on ginsburg version

- commented out nflows in install requires because it threw this error on ginsberg installation: 
```
  File "/burg/home/tjk2147/.local/lib/python3.11/site-packages/nflows/nn/nde/__init__.py", line 1, in <module>
    from nflows.nn.nde.made import MixtureOfGaussiansMADE

ImportError: /lib64/libstdc++.so.6: version `GLIBCXX_3.4.29' not found (required by /burg/opt/anaconda3-2023.09/lib/python3.11/site-packages/matplotlib/_path.cpython-311-x86_64-linux-gnu.so)
slurmstepd: error: _cgroup_procs_check: failed on path /sys/fs/cgroup/memory/slurm/uid_555096/job_12815040/step_batch/cgroup.procs: No such file or directory
slurmstepd: error: unable to read '/sys/fs/cgroup/memory/slurm/uid_555096/job_12815040/step_batch/cgroup.procs'
```
- also edited line of 4 of __init__.py for spender to comment out the .flow import NeuralDensityEstimator 

-  also commented out the function "load_flow_model(filename, n_latent, **kwargs)" in __init__.py

- this will be issue for normalizing flow outlier estimation later, but i think it might be okay for just training the model itself