mostlyai-engine 1.0.0__tar.gz → 1.0.1__tar.gz

{mostlyai_engine-1.0.0 → mostlyai_engine-1.0.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mostlyai-engine
-Version: 1.0.0
+Version: 1.0.1
 Summary: Synthetic Data - Engine
 Project-URL: Homepage, https://github.com/mostly-ai/mostlyai-engine
 Project-URL: Documentation, https://mostly-ai.github.io/mostlyai-engine/
@@ -29,7 +29,7 @@ Requires-Dist: numpy<2,>=1.26.3
 Requires-Dist: opacus<2,>=1.5.2
 Requires-Dist: pandas~=2.2.0
 Requires-Dist: peft<0.12,>=0.11.1
-Requires-Dist: psutil<6,>=5.9.8
+Requires-Dist: psutil<6,>=5.9.5
 Requires-Dist: pyarrow>=16.0.0
 Requires-Dist: rich<14,>=13.9.4
 Requires-Dist: setuptools<76,>=75.0.0
@@ -105,7 +105,10 @@ engine.split(                         # split data as PQT files for `trn` + `val
 )
 engine.analyze(workspace_dir=ws)      # generate column-level statistics to `{ws}/ModelData/tgt-stats/stats.json`
 engine.encode(workspace_dir=ws)       # encode training data to `{ws}/OriginalData/encoded-data`
-engine.train(workspace_dir=ws)        # train model and store to `{ws}/ModelData/model-data`
+engine.train(                         # train model and store to `{ws}/ModelStore/model-data`
+    workspace_dir=ws,
+    max_training_time=1,              # limit TRAIN to 1 minute for demo purposes
+)
 engine.generate(workspace_dir=ws)     # use model to generate synthetic samples to `{ws}/SyntheticData`
 pd.read_parquet(ws / "SyntheticData") # load synthetic data
 ```
@@ -138,7 +141,7 @@ engine.analyze(workspace_dir=ws)      # generate column-level statistics to `{ws
 engine.encode(workspace_dir=ws)       # encode training data to `{ws}/OriginalData/encoded-data`
 engine.train(                         # train model and store to `{ws}/ModelStore/model-data`
     workspace_dir=ws,
-    max_training_time=2,              # limit TRAIN to 1 minute for demo purposes
+    max_training_time=1,              # limit TRAIN to 1 minute for demo purposes
 )
 engine.generate(workspace_dir=ws)     # use model to generate synthetic samples to `{ws}/SyntheticData`
 pd.read_parquet(ws / "SyntheticData") # load synthetic data

{mostlyai_engine-1.0.0 → mostlyai_engine-1.0.1}/README.md

@@ -58,7 +58,10 @@ engine.split(                         # split data as PQT files for `trn` + `val
 )
 engine.analyze(workspace_dir=ws)      # generate column-level statistics to `{ws}/ModelData/tgt-stats/stats.json`
 engine.encode(workspace_dir=ws)       # encode training data to `{ws}/OriginalData/encoded-data`
-engine.train(workspace_dir=ws)        # train model and store to `{ws}/ModelData/model-data`
+engine.train(                         # train model and store to `{ws}/ModelStore/model-data`
+    workspace_dir=ws,
+    max_training_time=1,              # limit TRAIN to 1 minute for demo purposes
+)
 engine.generate(workspace_dir=ws)     # use model to generate synthetic samples to `{ws}/SyntheticData`
 pd.read_parquet(ws / "SyntheticData") # load synthetic data
 ```
@@ -91,7 +94,7 @@ engine.analyze(workspace_dir=ws)      # generate column-level statistics to `{ws
 engine.encode(workspace_dir=ws)       # encode training data to `{ws}/OriginalData/encoded-data`
 engine.train(                         # train model and store to `{ws}/ModelStore/model-data`
     workspace_dir=ws,
-    max_training_time=2,              # limit TRAIN to 1 minute for demo purposes
+    max_training_time=1,              # limit TRAIN to 1 minute for demo purposes
 )
 engine.generate(workspace_dir=ws)     # use model to generate synthetic samples to `{ws}/SyntheticData`
 pd.read_parquet(ws / "SyntheticData") # load synthetic data

{mostlyai_engine-1.0.0 → mostlyai_engine-1.0.1}/mostlyai/engine/__init__.py

@@ -21,7 +21,7 @@ from mostlyai.engine.generation import generate
 
 
 __all__ = ["split", "analyze", "encode", "train", "generate"]
-__version__ = "1.0.0"
+__version__ = "1.0.1"
 
 # suppress specific warning related to os.fork() in multi-threaded processes
 warnings.filterwarnings("ignore", category=DeprecationWarning, message=".*multi-threaded.*fork.*")

{mostlyai_engine-1.0.0 → mostlyai_engine-1.0.1}/mostlyai/engine/_language/training.py

@@ -197,8 +197,8 @@ def _calculate_val_loss(model: PreTrainedModel | GradSampleModule, val_dataloade
 def train(
     *,
     model: str = "MOSTLY_AI/LSTMFromScratch-3m",
-    max_training_time: float = 60.0,
-    max_epochs: float = 100.0,
+    max_training_time: float = 14400.0,  # 10 days
+    max_epochs: float = 100.0,  # 100 epochs
     batch_size: int | None = None,
     gradient_accumulation_steps: int | None = None,
     enable_flexible_generation: bool = True,

{mostlyai_engine-1.0.0 → mostlyai_engine-1.0.1}/mostlyai/engine/_tabular/training.py

@@ -339,8 +339,8 @@ def _calculate_average_trn_loss(trn_sample_losses: list[torch.Tensor], n: int |
 def train(
     *,
     model: str = "MOSTLY_AI/Medium",
-    max_training_time: float = 60.0,
-    max_epochs: float = 100.0,
+    max_training_time: float = 14400.0,  # 10 days
+    max_epochs: float = 100.0,  # 100 epochs
     batch_size: int | None = None,
     gradient_accumulation_steps: int | None = None,
     max_sequence_window: int = 100,

{mostlyai_engine-1.0.0 → mostlyai_engine-1.0.1}/mostlyai/engine/training.py

@@ -26,8 +26,8 @@ from mostlyai.engine._workspace import resolve_model_type
 def train(
     *,
     model: str | None = None,
-    max_training_time: float = 60.0,
-    max_epochs: float = 100.0,
+    max_training_time: float = 14400.0,  # 10 days
+    max_epochs: float = 100.0,  # 100 epochs
     batch_size: int | None = None,
     gradient_accumulation_steps: int | None = None,
     enable_flexible_generation: bool = True,

{mostlyai_engine-1.0.0 → mostlyai_engine-1.0.1}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "mostlyai-engine"
-version = "1.0.0"
+version = "1.0.1"
 description = "Synthetic Data - Engine"
 authors = [{ name = "MOSTLY AI", email = "dev@mostly.ai" }]
 requires-python = ">=3.10,<3.13"
@@ -27,7 +27,7 @@ dependencies = [
     "torch==2.5.1; sys_platform == 'darwin'",
     "pyarrow>=16.0.0",
     "joblib>=1.3.0,<2",
-    "psutil>=5.9.8,<6",
+    "psutil>=5.9.5,<6",  # upgrade when colab psutil is updated
     "tokenizers>=0.20.1,<0.21",
     "transformers>=4.45.2,<5",
     "datasets>=2.20.0,<3",