Mistobaan

Mistobaan

Mistobaan

Mistobaan

Mistobaan

with models.DAG() as dag:
    dataset = 'BREATHE'
    
    # create a task for each table in the BREATHE dataset
    for table_name in list_breathe_tables():  
      destination_uri = 'gs://' + '/'.join((dest_bucket,
                                  '{{ ds_nodash }}', # one day is enough for the full snapshot
                                  table_name,
                                  table_name + '-{{ ds_nodash }}-*.jsonl'))
      

Mistobaan

DECLARE SAMPLE_SIZE INT64 DEFAULT 5;

SELECT *
FROM 
    BREATHE.ncbi_xml 
WHERE
    RAND() < SAMPLE_SIZE/(SELECT COUNT(*) FROM `BREATHE.ncbi_xml`)

Mistobaan

I Purchased this mini keyboard to extend my editor with combinations of keys that I will never remember. Sadly installing the App was problematic as it does not work right away.

1. The App is written in Adobe air (!). Download and install their library
2. Install and run this command to overcome Catalina problems (Steve we miss you) (from this thread)[https://community.adobe.com/t5/air/adobe-air-error-message-macosx-catalina/td-p/10683302?page=1]:

sudo xattr -r -d com.apple.quarantine /Library/Frameworks/Adobe\ AIR.framework

1. Downloaded the online software from Koolertron as of today (2020 June 20) is this link

Mistobaan

pip install streamlit
pip install spacy
python -m spacy download en_core_web_sm
python -m spacy download en_core_web_md
python -m spacy download de_core_news_sm

Mistobaan

# how to make sure to restart the jupyter notebook 
try: 
  %pip install --user --quiet --upgrade my_module
  import my_module
except ModuleNotFoundError:
  import os
  os.kill(os.getpid(), 9)

Mistobaan

NUM_TRAINING_EXAMPLES = 1000 * 1000
STOP_POINT = 14.3
TOTAL_TRAINING_EXAMPLES = int(STOP_POINT * NUM_TRAINING_EXAMPLES)
BATCH_SIZE = 100
NUM_CHECKPOINTS = 15
steps_per_epoch = (TOTAL_TRAINING_EXAMPLES // 
                   (BATCH_SIZE*NUM_CHECKPOINTS))
cp_callback = tf.keras.callbacks.ModelCheckpoint(...)
history = model.fit(trainds, 
                    validation_data=evalds,