Clarify standard MLX memory handling in validation evidence

The stock batch generator uses Apple's recommended GPU wired limit. No manual override or macOS memory setting was changed. Measured results, test script, runtime patch and weights are unchanged.

FULL_MAC_VALIDATION.md

@@ -3,7 +3,7 @@
 Both complete, existing Swift checkpoints passed native Metal text generation
 and two cached follow-up requests on an AWS M4 Pro Mac with 48 GiB RAM.
 All weight shards were verified against their recorded SHA-256 hashes before
-loading. No quantization, tensor edits, context changes or memory-limit overrides
+loading. No quantization, tensor edits, context changes or manual memory-limit overrides
 were performed. Hub offline mode was enabled during testing.
 
 The input is our own synthetic repeated-record conversation, followed by a short
@@ -45,8 +45,12 @@ HF_HUB_OFFLINE=1 TRANSFORMERS_OFFLINE=1 python \
 ```
 
 Output directories must be new. The harness uses the real server and complete
-released weights, without building or quantizing any synthetic model. It disables
-the server's optional wired-limit adjustment and reports the actual Metal device.
+released weights, without building or quantizing any synthetic model. It suppresses
+the server startup helper's wired-limit call. The unchanged stock BatchGenerator
+still calls MLX's set_wired_limit with Apple's recommended working-set size;
+macOS memory settings are not changed. The legacy result field
+`memory_limit_overrides=false` denotes no manual override, not suppression of this
+normal batch-generator behavior. The actual Metal device is recorded in the results.
 
 These results establish the tested text workload on the stated 48 GiB machine.
 They do not establish 24 GiB operation, arbitrary 262k-token workloads, GUI/plugin

UPLOAD_MANIFEST.json

@@ -8,8 +8,8 @@
     },
     {
       "path": "FULL_MAC_VALIDATION.md",
-      "bytes": 3246,
-      "sha256": "a3a164e3affcfbe8a506b0205f4d4e63a69807ea56f4a8baa06b2201b20b1faa"
+      "bytes": 3562,
+      "sha256": "92cf16c5549a39fd8176028053bf57181eaf2c5fe31b142b36b961d0a5f6bf4a"
     },
     {
       "path": "LICENSE",
@@ -75,13 +75,13 @@
     },
     {
       "path": "compatibility/cache-tests/full-mac-4bit-results.json",
-      "bytes": 2699,
-      "sha256": "6087feec8158cd3f1df6e7c3394a51a5091ee96649066ad875f6b6ac866dd76b"
+      "bytes": 3127,
+      "sha256": "57c1c0924fadfec93f70a0787e31af02c36e8c996e30529ebd318b3a680b9f6f"
     },
     {
       "path": "compatibility/cache-tests/full-mac-5bit-results.json",
-      "bytes": 2695,
-      "sha256": "274b30cdac2d75d0258ff79a91ec53ce9b7029f36d1d64edf8783037a9c67459"
+      "bytes": 3123,
+      "sha256": "92f539012ef169ada9ab9970b077425664c80c7369f70c0faa58c927c6005951"
     },
     {
       "path": "compatibility/cache-tests/offline-4bit-results.json",

compatibility/cache-tests/full-mac-4bit-results.json

@@ -93,5 +93,12 @@
     "retained_limit_observed_in_all_samples": true
   },
   "model_repository": "ukisai/Swift-1.5-4bit-MLX",
-  "tested_revision": "c4765c752caee282928f62685722e11000ec64cf"
+  "tested_revision": "c4765c752caee282928f62685722e11000ec64cf",
+  "memory_limit_scope": {
+    "manual_overrides": false,
+    "macos_settings_changed": false,
+    "server_startup_wired_limit_helper": "suppressed by the HTTP harness",
+    "batch_generator_wired_limit": "unchanged stock MLX-LM behavior: uses Apple's recommended working-set size",
+    "legacy_boolean_meaning": "memory_limit_overrides=false records no manual override; it does not suppress stock batch-generator behavior"
+  }
 }

compatibility/cache-tests/full-mac-5bit-results.json

@@ -93,5 +93,12 @@
     "retained_limit_observed_in_all_samples": true
   },
   "model_repository": "ukisai/Swift-1.5-5bit-MLX",
-  "tested_revision": "5359607ded7c88e3c7379a0adfb879eaf483cd98"
+  "tested_revision": "5359607ded7c88e3c7379a0adfb879eaf483cd98",
+  "memory_limit_scope": {
+    "manual_overrides": false,
+    "macos_settings_changed": false,
+    "server_startup_wired_limit_helper": "suppressed by the HTTP harness",
+    "batch_generator_wired_limit": "unchanged stock MLX-LM behavior: uses Apple's recommended working-set size",
+    "legacy_boolean_meaning": "memory_limit_overrides=false records no manual override; it does not suppress stock batch-generator behavior"
+  }
 }