refactor: assemble inference engines through a shared composition root

- add build_engine() to astrai.inference.engine as the single load-place-wire path for InferenceEngine, accepting a checkpoint path or live model/tokenizer plus passthrough engine kwargs
- migrate the server lifespan, generate CLI, humaneval/ifeval evals, and all three demos to build_engine; app._create_engine collapses into a direct call
- export build_engine from astrai and astrai.inference
- parameterize the autoregressive demo with --prompt one-shot continuation plus model path and sampling knobs, exiting cleanly on !exit or EOF
- cover the composition root with unit tests for live-object assembly, kwargs passthrough, and argument validation

scripts/demo/stream_chat.py

@@ -1,11 +1,7 @@
 from argparse import ArgumentParser
 from pathlib import Path
 
-import torch
-
-from astrai import InferenceEngine
-from astrai.model import AutoModel
-from astrai.tokenize import AutoTokenizer
+from astrai import build_engine
 
 PROJECT_ROOT = Path(__file__).resolve().parents[2]
 
@@ -68,10 +64,8 @@ def chat():
     args = parse_args()
     model_path = args.model_path
 
-    model = AutoModel.from_pretrained(model_path)
-    tokenizer = AutoTokenizer.from_pretrained(model_path)
-    model.to(device="cuda", dtype=torch.bfloat16)
-    engine = InferenceEngine(model=model, tokenizer=tokenizer)
+    engine = build_engine(model_path)
+    tokenizer = engine.tokenizer
 
     while True:
         query = input(">> ")