refactor: simplify BaseFactory and separate ModelFactory from AutoModel

- Extract _resolve_base_type and _validate_component as module-level helpers
- Replace ForwardRef._evaluate private API with eval in module namespace
- Remove broad except Exception in __init_subclass__, _component_base always set
- Replace direct _entries mutation in strategy.py with register() call form
- Remove dead TOKENIZER_CLASSES registry from AutoTokenizer
- Extract ModelFactory(BaseFactory[nn.Module]) as pure factory
- AutoModel now inherits only nn.Module, no factory state
- Move @AutoModel.register to @ModelFactory.register in transformer.py and encoder.py

astrai/model/automodel.py

@@ -40,11 +40,12 @@ def _disable_random_init(enable: bool = True):
             setattr(nn.init, n, fn)
 
 
-class AutoModel(BaseFactory["AutoModel"], nn.Module):
-    """
-    Autoregressive language model base class.
-    Provides model loading/saving, registration, and generation.
-    """
+class ModelFactory(BaseFactory[nn.Module]):
+    """Pure factory for model dispatch, separated from nn.Module state."""
+
+
+class AutoModel(nn.Module):
+    """Model base class with loading/saving and generation."""
 
     def __init__(self, config: BaseModelConfig):
         super().__init__()
@@ -68,7 +69,7 @@ class AutoModel(BaseFactory["AutoModel"], nn.Module):
         config = ConfigFactory.load(raw)
         model_type = config.model_type or "autoregressive_lm"
 
-        actual_cls = AutoModel.get_component_class(model_type)
+        actual_cls = ModelFactory.get_component_class(model_type)
 
         with _disable_random_init(enable=disable_random_init):
             model = actual_cls(config)

astrai/model/encoder.py

@@ -5,7 +5,7 @@ import torch.nn as nn
 from torch import Tensor
 
 from astrai.config.model_config import EncoderConfig
-from astrai.model.automodel import AutoModel
+from astrai.model.automodel import AutoModel, ModelFactory
 from astrai.model.components.decoder_block import DecoderBlock
 from astrai.model.components.embedding import Embedding
 from astrai.model.components.norm import RMSNorm
@@ -13,7 +13,7 @@ from astrai.model.components.rope import RotaryEmbedding
 from astrai.model.transformer import process_attention_mask
 
 
-@AutoModel.register("embedding")
+@ModelFactory.register("embedding")
 class EmbeddingEncoder(AutoModel):
     def __init__(self, config: EncoderConfig):
         super().__init__(config)

astrai/model/transformer.py

@@ -6,7 +6,7 @@ from torch import Tensor
 
 from astrai.config.model_config import AutoRegressiveLMConfig
 from astrai.inference.core.cache import CacheView
-from astrai.model.automodel import AutoModel
+from astrai.model.automodel import AutoModel, ModelFactory
 from astrai.model.components.decoder_block import DecoderBlock
 from astrai.model.components.embedding import Embedding
 from astrai.model.components.linear import Linear
@@ -26,7 +26,7 @@ def process_attention_mask(
     return input_mask
 
 
-@AutoModel.register("autoregressive_lm")
+@ModelFactory.register("autoregressive_lm")
 class AutoRegressiveLM(AutoModel):
     """Autoregressive language model with paged KV cache."""