perf(mining): batch per-chunk upserts and add optional GPU acceleration

The miner upserted one drawer per ChromaDB call, paying tokenizer +
ONNX session setup per chunk. The embedding device was CPU-only because
no EmbeddingFunction was ever wired through the backend.

Two changes, each a speedup in its own right; stacked they give ~10x
end-to-end on a medium corpus (20 files, 568 drawers):

1. Batched upsert. `process_file` and `_file_chunks_locked` now collect
   all chunks of a file into a single `collection.upsert(...)` so the
   embedding model runs one forward pass per file instead of N.

2. Hardware-accelerated embedding function. New `mempalace/embedding.py`
   wraps `ONNXMiniLM_L6_V2` with configurable `preferred_providers`.
   `MEMPALACE_EMBEDDING_DEVICE` (or `embedding_device` in config.json)
   selects auto / cpu / cuda / coreml / dml. Unavailable accelerators
   log a warning and fall back to CPU.

   The factory subclasses `ONNXMiniLM_L6_V2` and spoofs its `name()` to
   `"default"` so the persisted EF identity matches existing palaces
   created with ChromaDB's bare `DefaultEmbeddingFunction` -- same
   model, same 384-dim vectors, no rebuild needed when turning GPU on.

   `ChromaBackend.get_collection` / `create_collection` now pass the
   resolved EF on every call so miner writes and searcher reads agree.

Benchmarks (i9-12900KF + RTX 3090, medium scenario, 568 drawers):

  per-chunk + CPU   19.77s ·  29 drw/s   (baseline)
  batched   + CPU    8.07s ·  70 drw/s   (2.4x)
  batched   + CUDA   2.15s · 264 drw/s   (9.2x)

Reproducible via `benchmarks/mine_bench.py`.

Install paths:
  pip install mempalace[gpu]       # NVIDIA CUDA
  pip install mempalace[dml]       # DirectML (Windows)
  pip install mempalace[coreml]    # macOS Neural Engine

Mine header now prints `Device: cpu|cuda|...` so users can confirm the
accelerator engaged.

mempalace/backends/chroma.py

@@ -405,6 +405,23 @@ class ChromaBackend(BaseBackend):
         self._freshness: dict[str, tuple[int, float]] = {}
         self._closed = False
 
+    @staticmethod
+    def _resolve_embedding_function():
+        """Return the EF for the user's ``embedding_device`` setting.
+
+        Both ``get_collection`` and ``get_or_create_collection`` must receive
+        the EF explicitly — ChromaDB 1.x does not persist it with the
+        collection, so a reader that omits the argument silently gets the
+        library default and its queries won't match the writer's vectors.
+        """
+        try:
+            from ..embedding import get_embedding_function
+
+            return get_embedding_function()
+        except Exception:
+            logger.exception("Failed to build embedding function; using chromadb default")
+            return None
+
     # ------------------------------------------------------------------
     # Internal helpers
     # ------------------------------------------------------------------
@@ -532,12 +549,15 @@ class ChromaBackend(BaseBackend):
         if options and isinstance(options, dict):
             hnsw_space = options.get("hnsw_space", hnsw_space)
 
+        ef = self._resolve_embedding_function()
+        ef_kwargs = {"embedding_function": ef} if ef is not None else {}
+
         if create:
             collection = client.get_or_create_collection(
-                collection_name, metadata={"hnsw:space": hnsw_space}
+                collection_name, metadata={"hnsw:space": hnsw_space}, **ef_kwargs
             )
         else:
-            collection = client.get_collection(collection_name)
+            collection = client.get_collection(collection_name, **ef_kwargs)
         return ChromaCollection(collection)
 
     def close_palace(self, palace) -> None:
@@ -578,8 +598,10 @@ class ChromaBackend(BaseBackend):
         self, palace_path: str, collection_name: str, hnsw_space: str = "cosine"
     ) -> ChromaCollection:
         """Create (not get-or-create) ``collection_name`` with the given HNSW space."""
+        ef = self._resolve_embedding_function()
+        ef_kwargs = {"embedding_function": ef} if ef is not None else {}
         collection = self._client(palace_path).create_collection(
-            collection_name, metadata={"hnsw:space": hnsw_space}
+            collection_name, metadata={"hnsw:space": hnsw_space}, **ef_kwargs
         )
         return ChromaCollection(collection)