nathansutton

from PIL import Image
from transformers import BlipForConditionalGeneration, BlipProcessor

# read in the model
processor = BlipProcessor.from_pretrained("nathansutton/generate-cxr")
model = BlipForConditionalGeneration.from_pretrained("nathansutton/generate-cxr")

# your data
my_image = 'my-chest-x-ray.jpg'
my_indication = 'RLL crackles, eval for pneumonia'

nathansutton

flyway_1    | Database: jdbc:postgresql://postgres:5432/athena (PostgreSQL 12.4)
flyway_1    | Successfully validated 54 migrations (execution time 00:00.320s)
flyway_1    | Creating Schema History table "public"."flyway_schema_history" ...
flyway_1    | Current version of schema "public": << Empty Schema >>
flyway_1    | Migrating schema "public" to version 0000 - relationship
flyway_1    | Migrating schema "public" to version 0001 - vocabulary
flyway_1    | Migrating schema "public" to version 0002 - concept ancestor
flyway_1    | Migrating schema "public" to version 0003 - concept
flyway_1    | Migrating schema "public" to version 0004 - concept relationship
flyway_1    | Migrating schema "public" to version 0005 - concept class

nathansutton

docker-compose down

nathansutton

docker-compose up

nathansutton

docker-compose up

nathansutton

{sql connection=connection}
create view vw_model as 
select
  visit_occurrence_id
  ,person_id
  ,1.0 - 1.0 / (1.0 + EXP((-2.95983732216539) + (readmit_030_flg * 0.477723271743372) + (readmit_360_flg * 0.531143494501182) + (icd_history_518 * 0.651130360976662))) as risk
from vw_baseline

nathansutton

{r}

translate <- function(fit){
  
  ### return the sql translation of a h2o.glm model ###
  
  # safety 
  stopifnot(class(fit) %in% c("H2OBinomialModel","H2ORegressionModel"))
  stopifnot(.hasSlot(fit, "algorithm"))
  stopifnot(.hasSlot(fit, "model"))

nathansutton

{python}

def translate (fit) :

    """return the sql translation of a H2OGeneralizedLinearEstimator"""

    # pandas dependency
    try:
        import pandas.DataFrame
    except ImportError: