refactor: simplify inference engine and backend dispatch

- merge _generate_streaming/_generate_non_streaming into single _generate() with stream flag
- delete dead GenerationRequest class and generate_with_request method
- inline _next_token helper into generate_async
- replace flash-attn double-checked locking with functools.lru_cache
- extract _write_and_gather_kv helper shared by TorchNative/FlashAttn backends
- inline _kv_cache_is_contiguous into its sole call site in FlashAttnBackend
- change default backend priority from flash>cuda>torch to cuda>flash>torch
- add ASTR_BACKEND env var to override default backend at resolve time
- add supports_graph() static method to AttentionBackend ABC, override in CudaBackend
- replace isinstance(get_backend(), CudaBackend) with get_backend().supports_graph() in executor
- add torch.cuda.is_available() guard to CudaBackend.supports()

astrai/inference/__init__.py

@@ -5,7 +5,7 @@ Layers:
   - api/:         HTTP orchestration (ProtocolHandler, server)
   - protocols/:   Response builders (OpenAI, Anthropic)
   - transport/:   SSE transport utilities
-  - engine.py:    Facade (InferenceEngine), Value Object (GenerationRequest)
+  - engine.py:    Facade (InferenceEngine)
   - sample.py:    Strategy pattern (TemperatureStrategy, TopKStrategy, TopPStrategy, FrequencyPenaltyStrategy)
 """
 
@@ -42,7 +42,7 @@ from astrai.inference.core import (
     TaskStatus,
     page_hash,
 )
-from astrai.inference.engine import GenerationRequest, InferenceEngine
+from astrai.inference.engine import InferenceEngine
 from astrai.inference.sample import (
     BaseSamplingStrategy,
     FrequencyPenaltyStrategy,
@@ -55,7 +55,6 @@ from astrai.inference.sample import (
 
 __all__ = [
     "InferenceEngine",
-    "GenerationRequest",
     "InferenceScheduler",
     "Executor",
     "STOP",