Merge pull request #443 from sthaha/estimator-improvements

Estimator improvements

Author: Sunil Thaha

pyproject.toml

@@ -44,6 +44,7 @@ dependencies = [
   "boto3==1.34.155",
   "pymarkdownlnt==0.9.22",
   "yamllint==1.35.1",
+  "requests-file==2.1.0",
 ]
 
 [project.scripts]
@@ -67,8 +68,6 @@ extra-dependencies = [
   "coverage[toml]>=6.5",
 	"ipdb",
 	"ipython",
-  "ipdb",
-  "ipython",
   "pytest",
 ]
 

src/kepler_model/estimate/archived_model.py

@@ -1,10 +1,15 @@
+import logging
+
 import requests
+from requests_file import FileAdapter
 
 from kepler_model.estimate.model_server_connector import unpack
 from kepler_model.util.config import get_init_model_url
 from kepler_model.util.loader import load_metadata
 from kepler_model.util.train_types import ModelOutputType
 
+logger = logging.getLogger(__name__)
+
 failed_list = []
 
 FILTER_ITEM_DELIMIT = ";"
@@ -38,19 +43,21 @@ def valid_metrics(metrics, features):
 def is_valid_model(metrics, metadata, filters):
     if not valid_metrics(metrics, metadata["features"]):
         return False
+
     for attrb, val in filters.items():
         if not hasattr(metadata, attrb) or getattr(metadata, attrb) is None:
-            print("{} has no {}".format(metadata["model_name"], attrb))
+            logger.warning(f"{metadata['model_name']} has no {attrb}")
             return False
-        else:
-            cmp_val = getattr(metadata, attrb)
-            val = float(val)
-            if attrb == "abs_max_corr":  # higher is better
-                valid = cmp_val >= val
-            else:  # lower is better
-                valid = cmp_val <= val
-            if not valid:
-                return False
+
+        cmp_val = getattr(metadata, attrb)
+        val = float(val)
+        if attrb == "abs_max_corr":  # higher is better
+            valid = cmp_val >= val
+        else:  # lower is better
+            valid = cmp_val <= val
+        if not valid:
+            return False
+
     return True
 
 
@@ -60,21 +67,25 @@ def reset_failed_list():
 
 
 def get_achived_model(power_request):
-    print("get archived model")
     global failed_list
     output_type_name = power_request.output_type
     if output_type_name in failed_list:
         return None
     output_type = ModelOutputType[power_request.output_type]
     url = get_init_model_url(power_request.energy_source, output_type_name)
     if url == "":
-        print("no URL set for ", output_type_name, power_request.energy_source)
+        logger.warning(f"no URL set for {output_type_name}, {power_request.energy_source}")
         return None
-    print(f"try getting archieved model from URL: {url} for {output_type_name}")
-    response = requests.get(url)
-    print(response)
+    logger.info(f"try getting archieved model from URL: {url} for {output_type_name}")
+
+    s = requests.Session()
+    s.mount("file://", FileAdapter())
+    response = s.get(url)
+    logger.debug(f"response: {response}")
+
     if response.status_code != 200:
         return None
+
     output_path = unpack(power_request.energy_source, output_type, response, replace=False)
     if output_path is not None:
         metadata = load_metadata(output_path)
@@ -83,7 +94,8 @@ def get_achived_model(power_request):
             if not is_valid_model(power_request.metrics, metadata, filters):
                 failed_list += [output_type_name]
                 return None
-        except:
-            print("cannot validate the archived model")
+        except Exception as e:
+            logger.warning(f"cannot validate the archived model: {e}")
             return None
+
     return output_path

src/kepler_model/estimate/estimator.py

@@ -13,7 +13,7 @@
 from kepler_model.estimate.model.model import load_downloaded_model
 from kepler_model.estimate.model_server_connector import is_model_server_enabled, make_request
 from kepler_model.train.profiler.node_type_index import NodeTypeSpec, discover_spec_values, get_machine_spec
-from kepler_model.util.config import CONFIG_PATH, SERVE_SOCKET, download_path, set_env_from_model_config, set_config_dir
+from kepler_model.util.config import CONFIG_PATH, SERVE_SOCKET, download_path, set_config_dir, set_env_from_model_config
 from kepler_model.util.loader import get_download_output_path, load_metadata
 from kepler_model.util.train_types import ModelOutputType, convert_enery_source, is_output_type_supported
 

src/kepler_model/util/config.py

@@ -12,8 +12,8 @@
 #
 #################################################
 
-import os
 import logging
+import os
 
 import requests
 
@@ -165,7 +165,7 @@ def get_init_model_url(energy_source, output_type, model_topurl=model_topurl):
     for prefix in modelConfigPrefix:
         if get_energy_source(prefix) == energy_source:
             modelURL = get_init_url(prefix)
-            logger.info("get init url", modelURL)
+            logger.info(f"get init url: {modelURL}")
             url = get_url(
                 feature_group=FeatureGroup.BPFOnly,
                 output_type=ModelOutputType[output_type],