mostlyai-engine 1.4.7__tar.gz → 1.4.8__tar.gz

{mostlyai_engine-1.4.7 → mostlyai_engine-1.4.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mostlyai-engine
-Version: 1.4.7
+Version: 1.4.8
 Summary: Synthetic Data Engine
 Project-URL: homepage, https://github.com/mostly-ai/mostlyai-engine
 Project-URL: repository, https://github.com/mostly-ai/mostlyai-engine

{mostlyai_engine-1.4.7 → mostlyai_engine-1.4.8}/mostlyai/engine/__init__.py

@@ -22,7 +22,7 @@ from mostlyai.engine.splitting import split
 from mostlyai.engine.training import train
 
 __all__ = ["split", "analyze", "encode", "train", "generate", "init_logging", "set_random_state"]
-__version__ = "1.4.7"
+__version__ = "1.4.8"
 
 # suppress specific warning related to os.fork() in multi-threaded processes
 warnings.filterwarnings("ignore", category=DeprecationWarning, message=".*multi-threaded.*fork.*")

{mostlyai_engine-1.4.7 → mostlyai_engine-1.4.8}/mostlyai/engine/_encoding_types/tabular/character.py

@@ -155,7 +155,7 @@ def decode_character(df_encoded: pd.DataFrame, stats: dict) -> pd.Series:
         values = values.apply(lambda item: item.replace(UNKNOWN_TOKEN, "")).str.rstrip()
     else:
         # handle de-generate case, where no tokens were stored
-        values = pd.Series(pd.NA).repeat(df_encoded.shape[0])
+        values = pd.Series(pd.NA, index=range(df_encoded.shape[0]))
     if stats["has_nan"]:
         values[df_encoded["nan"] == 1] = pd.NA
     return values

{mostlyai_engine-1.4.7 → mostlyai_engine-1.4.8}/mostlyai/engine/_language/generation.py

@@ -256,9 +256,6 @@ def generate(
         _LOG.info(f"{max_new_tokens=}")
 
         t0 = time.time()
-        # use MOSTLY_HUGGING_FACE_TOKEN if available, otherwise HF_TOKEN should be unset or with a pre-set value as is
-        if os.getenv("MOSTLY_HUGGING_FACE_TOKEN"):
-            os.environ["HF_TOKEN"] = os.environ["MOSTLY_HUGGING_FACE_TOKEN"]
 
         is_peft_adapter = (workspace.model_path / "adapter_config.json").exists()
         is_vllm_available = importlib.util.find_spec("vllm") is not None

{mostlyai_engine-1.4.7 → mostlyai_engine-1.4.8}/mostlyai/engine/_language/training.py

@@ -16,7 +16,6 @@ import gc
 import json
 import logging
 import math
-import os
 import time
 import warnings
 from collections.abc import Callable
@@ -465,10 +464,6 @@ def train(
                 model_config = LSTMFromScratchConfig(vocab_size=len(tokenizer), with_dp=with_dp)
                 model = LSTMFromScratchLMHeadModel(model_config).to(device)
         else:
-            # use MOSTLY_HUGGING_FACE_TOKEN if available
-            if os.getenv("MOSTLY_HUGGING_FACE_TOKEN"):
-                os.environ["HF_TOKEN"] = os.environ["MOSTLY_HUGGING_FACE_TOKEN"]
-
             model, model_config = load_base_model_and_config(
                 model_id_or_path,
                 device=device,

{mostlyai_engine-1.4.7 → mostlyai_engine-1.4.8}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "mostlyai-engine"
-version = "1.4.7"
+version = "1.4.8"
 description = "Synthetic Data Engine"
 authors = [{ name = "MOSTLY AI", email = "dev@mostly.ai" }]
 requires-python = ">=3.10"