specmem-hardwicksoftware 3.7.12 → 3.7.13

package/embedding-sandbox/frankenstein-embeddings.py

@@ -189,6 +189,13 @@ PROJECT_PATH = os.environ.get('SPECMEM_PROJECT_PATH', 'default')
 
 SPECMEM_HOME = os.environ.get('SPECMEM_HOME', os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 SPECMEM_RUN_DIR = os.environ.get('SPECMEM_RUN_DIR', os.path.join(SPECMEM_HOME, 'run'))
+
+# Bundled model: shipped with the npm package, no download needed
+_SCRIPT_DIR = os.path.dirname(os.path.abspath(__file__))
+_BUNDLED_MODEL_DIR = os.path.join(_SCRIPT_DIR, 'models', 'all-MiniLM-L6-v2')
+BUNDLED_MODEL_PATH = _BUNDLED_MODEL_DIR if os.path.isfile(os.path.join(_BUNDLED_MODEL_DIR, 'onnx', 'model_quint8_avx2.onnx')) else None
+if BUNDLED_MODEL_PATH:
+    print(f"📦 Bundled model found: {BUNDLED_MODEL_PATH}", file=sys.stderr)
 # Socket directory: {PROJECT}/specmem/sockets/ - matches config.ts expectations
 # This is the ONLY location config.ts checks for per-project sockets
 def _get_socket_dir():
@@ -821,7 +828,8 @@ class LayerOffloadingTransformer:
     """
 
     def __init__(self, model_name: str, cache_dir: Path):
-        self.model_name = model_name
+        # Use bundled model if available
+        self.model_name = BUNDLED_MODEL_PATH if BUNDLED_MODEL_PATH else model_name
         self.cache_dir = cache_dir
         self.tokenizer = None
         self.model_config = None
@@ -2026,7 +2034,8 @@ class FrankensteinEmbeddings:
         self.ram_guard = RAMGuard()
 
         # Store model name for lazy-loading
-        self.base_model = base_model
+        # Use bundled model path if available (no network download needed)
+        self.base_model = BUNDLED_MODEL_PATH if BUNDLED_MODEL_PATH else base_model
 
         # Track request time for idle cleanup
         self.last_request_time = time.time()
@@ -2049,10 +2058,10 @@ class FrankensteinEmbeddings:
             self.dim_config.native_dims = 384  # MiniLM-L6-v2 is always 384
         else:
             # EAGER MODE: Load model immediately (for high-RAM or heavyOps)
-            print(f"Loading model: {base_model} ({_BEST_ONNX_FILE})", file=sys.stderr)
+            print(f"Loading model: {self.base_model} ({_BEST_ONNX_FILE})", file=sys.stderr)
             # NOTE: backend='onnx' is REQUIRED for model_kwargs file_name to work
             self.model = SentenceTransformer(
-                base_model,
+                self.base_model,
                 device='cpu',
                 backend='onnx',
                 cache_folder=str(self.cache_dir),

package/embedding-sandbox/models/all-MiniLM-L6-v2/1_Pooling/config.json

@@ -0,0 +1,7 @@
+{
+  "word_embedding_dimension": 384,
+  "pooling_mode_cls_token": false,
+  "pooling_mode_mean_tokens": true,
+  "pooling_mode_max_tokens": false,
+  "pooling_mode_mean_sqrt_len_tokens": false
+}

package/embedding-sandbox/models/all-MiniLM-L6-v2/config.json

@@ -0,0 +1,24 @@
+{
+  "_name_or_path": "nreimers/MiniLM-L6-H384-uncased",
+  "architectures": [
+    "BertModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 384,
+  "initializer_range": 0.02,
+  "intermediate_size": 1536,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 6,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.8.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

package/embedding-sandbox/models/all-MiniLM-L6-v2/config_sentence_transformers.json

@@ -0,0 +1,7 @@
+{
+  "__version__": {
+    "sentence_transformers": "2.0.0",
+    "transformers": "4.6.1",
+    "pytorch": "1.8.1"
+  }
+}

package/embedding-sandbox/models/all-MiniLM-L6-v2/modules.json

@@ -0,0 +1,20 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.models.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_Pooling",
+    "type": "sentence_transformers.models.Pooling"
+  },
+  {
+    "idx": 2,
+    "name": "2",
+    "path": "2_Normalize",
+    "type": "sentence_transformers.models.Normalize"
+  }
+]

package/embedding-sandbox/models/all-MiniLM-L6-v2/sentence_bert_config.json

@@ -0,0 +1,4 @@
+{
+  "max_seq_length": 256,
+  "do_lower_case": false
+}

package/embedding-sandbox/models/all-MiniLM-L6-v2/special_tokens_map.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}