easy_ml 0.2.0.pre.rc97 → 0.2.0.pre.rc99

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: a7de56b964119aa83ec62b64cf28fddc4f83ae3947d99a656643fe7355662ac5
-  data.tar.gz: fbdf0cdf54e96c128c37a5998cd2cf20135229e25320e1cce5b003defa448a52
+  metadata.gz: c65153b704330d5ef94c4844bbe505c91ff3bf3d6a67b21375f5178679a4bf68
+  data.tar.gz: 7ffbe174d6be8d6fdd43767ab383a186c211a279aa2e07eb258372ab67b88d27
 SHA512:
-  metadata.gz: 202db609a33bdd29437450364bb945e4a51cfe18f6e6b075f343ec504489a88f254b90c5e14e87d0098c917d372a6adc1ad51800627acc08de57fbc4269fb375
-  data.tar.gz: fcb1ab6d584f6b55d68ffc5a4118807443d61e26dd85cd3968bcd57cec4aa1d1183c2703956f06f38bf549ec7ef705aecb094014149685a761d7b1e4766720ee
+  metadata.gz: 318d8ddafdf1de61cfad9c9afa64b4c29ac27bf3745d37249bccee94dc80886293fe63744a8e3a3d049c75c8134748d061b5c7087dcc534f18990b63ed2030bf
+  data.tar.gz: 3f329b087f27044b9da1711716208b551161affe1c6f9c96edf2d5f729c30d3de9bd40c2931eef09a126f8826924615f3331fca71478f13d2659e83cdc484ff1

data/app/models/easy_ml/cleaner.rb

@@ -29,8 +29,6 @@ module EasyML
       end
     end
 
-    private
-
     def files_to_keep_for_dir(dir)
       files_to_keep.map(&:to_s).select { |f| f.start_with?(dir.to_s) }
     end
@@ -55,8 +53,12 @@ module EasyML
         end
     end
 
+    def test_mode?
+      Rails.env.test?
+    end
+
     def model_files_to_keep
-      if Rails.env.test?
+      if test_mode?
         []
       else
         active_models.map(&:model_file).compact.map(&:full_path).uniq
@@ -64,15 +66,15 @@ module EasyML
     end
 
     def dataset_files_to_keep
-      if Rails.env.test?
+      if test_mode?
         []
       else
-        EasyML::Dataset.all.flat_map(&:files).uniq
+        EasyML::Dataset.all.flat_map(&:files).uniq + active_models.map(&:dataset).flat_map(&:files).uniq
       end
     end
 
     def datasource_files_to_keep
-      if Rails.env.test?
+      if test_mode?
         Dir.glob(EasyML::Engine.root_dir.glob("datasources/**/*.{csv,parquet}")).uniq
       else
         EasyML::Datasource.all.flat_map(&:files).uniq

data/app/models/easy_ml/dataset.rb

@@ -667,8 +667,6 @@ module EasyML
     end
 
     def upload_remote_files
-      return if !needs_refresh?
-
       processed.upload.tap do
         features.each(&:upload_remote_files)
         features.each(&:save)

data/app/models/easy_ml/deploy.rb

@@ -65,8 +65,14 @@ module EasyML
 
           EasyML::Deploy.transaction do
             update(model_history_id: self.model_version.id, snapshot_id: self.model_version.snapshot_id, status: :success)
-            model.retraining_runs.where(status: :deployed).update_all(status: :success)
-            retraining_run.update(model_history_id: self.model_version.id, snapshot_id: self.model_version.snapshot_id, deploy_id: id, status: :deployed,)
+            model.retraining_runs.where(status: :deployed).update_all(status: :success, is_deploying: false)
+            retraining_run.update(
+              model_history_id: self.model_version.id,
+              snapshot_id: self.model_version.snapshot_id,
+              deploy_id: id,
+              status: :deployed,
+              is_deploying: false
+            )
           end
 
           model_version.tap do

data/app/models/easy_ml/model.rb

@@ -559,7 +559,6 @@ module EasyML
       dataset.upload_remote_files
       model_snapshot = snapshot
 
-      Thread.current[:deploying] = true
       # Prepare the model to be retrained (reset values so they don't conflict with our snapshotted version)
       bump_version(force: true)
       dataset.bump_versions(version)

data/app/models/easy_ml/model_file.rb

@@ -36,6 +36,10 @@ module EasyML
       )
     end
 
+    def s3_prefix
+      EasyML::Configuration.model_s3_path(model.name)
+    end
+
     def root_dir
       Pathname.new(model.root_dir)
     end

data/app/models/easy_ml/settings.rb

@@ -21,6 +21,16 @@ module EasyML
 
     validates :storage, inclusion: { in: %w[file s3] }, if: -> { storage.present? }
 
+    after_initialize :set_defaults, if: -> { new_record? }
+
+    def set_defaults
+      self.s3_prefix ||= "easy_ml"
+    end
+
+    def s3_prefix
+      read_attribute(:s3_prefix) || "easy_ml"
+    end
+
     TIMEZONES = [
       { value: "America/New_York", label: "Eastern Time" },
       { value: "America/Chicago", label: "Central Time" },

data/lib/easy_ml/configuration.rb

@@ -60,5 +60,17 @@ module EasyML
     def db_settings
       @db_settings ||= EasyML::Settings.first_or_create
     end
+
+    def self.s3_path_for(type, name)
+      File.join(instance.s3_prefix, type.to_s, name.to_s)
+    end
+
+    def self.model_s3_path(model_name)
+      s3_path_for("models", model_name.parameterize.gsub("-", "_"))
+    end
+
+    def self.dataset_s3_path(relative_path)
+      s3_path_for("datasets", relative_path)
+    end
   end
 end

data/lib/easy_ml/data/dataset_manager.rb

@@ -125,6 +125,10 @@ module EasyML
         synced_directory.download
       end
 
+      def remote_files
+        synced_directory.remote_files
+      end
+
       private
 
       def root

data/lib/easy_ml/data/splits/file_split.rb

@@ -46,7 +46,9 @@ module EasyML
         end
 
         def s3_prefix
-          File.join("datasets", dir.split("datasets").last)
+          EasyML::Configuration.dataset_s3_path(
+            dir.split("datasets").last
+          )
         end
 
         def synced_directory

data/lib/easy_ml/feature_store.rb

@@ -36,6 +36,19 @@ module EasyML
       )
     end
 
+    # Expose S3 information for testing and verification
+    def s3_prefix
+      EasyML::Configuration.dataset_s3_path(
+        File.join(feature_dir.split("datasets").last, "compacted")
+      )
+    end
+
+    def s3_key(filename = nil)
+      filename ||= File.basename(files.first) if files.any?
+      return nil unless filename
+      File.join(s3_prefix, filename)
+    end
+
     private
 
     def batch_size
@@ -49,9 +62,5 @@ module EasyML
         feature&.name&.parameterize&.gsub("-", "_")
       )
     end
-
-    def s3_prefix
-      File.join("datasets", feature_dir.split("datasets").last)
-    end
   end
 end

data/lib/easy_ml/support/synced_file.rb

@@ -52,6 +52,10 @@ module EasyML
         full_path
       end
 
+      def s3_key
+        s3_prefix.present? ? File.join(s3_prefix, filename) : filename
+      end
+
       def path
         File.join(root_dir, s3_prefix)
       end

data/lib/easy_ml/version.rb

@@ -1,7 +1,7 @@
 # frozen_string_literal: true
 
 module EasyML
-  VERSION = "0.2.0-rc97"
+  VERSION = "0.2.0-rc99"
 
   module Version
   end

metadata

@@ -1,14 +1,13 @@
 --- !ruby/object:Gem::Specification
 name: easy_ml
 version: !ruby/object:Gem::Version
-  version: 0.2.0.pre.rc97
+  version: 0.2.0.pre.rc99
 platform: ruby
 authors:
 - Brett Shollenberger
-autorequire:
 bindir: exe
 cert_chain: []
-date: 2025-03-06 00:00:00.000000000 Z
+date: 2025-03-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activerecord
@@ -865,7 +864,6 @@ metadata:
   homepage_uri: https://github.com/brettshollenberger/easy_ml
   source_code_uri: https://github.com/brettshollenberger/easy_ml
   changelog_uri: https://github.com/brettshollenberger/easy_ml
-post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -876,12 +874,11 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: 3.0.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">"
+  - - ">="
     - !ruby/object:Gem::Version
-      version: 1.3.1
+      version: '0'
 requirements: []
-rubygems_version: 3.4.10
-signing_key:
+rubygems_version: 3.6.5
 specification_version: 4
 summary: Effortless Machine Learning in Ruby
 test_files: []