Updated compression_config to quantization_config

config.json

@@ -10,46 +10,6 @@
     "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM"
   },
   "bos_token_id": 1,
-  "compression_config": {
-    "config_groups": {
-      "group_0": {
-        "input_activations": {
-          "block_structure": null,
-          "dynamic": true,
-          "group_size": null,
-          "num_bits": 8,
-          "observer": "memoryless",
-          "observer_kwargs": {},
-          "strategy": "token",
-          "symmetric": true,
-          "type": "int"
-        },
-        "output_activations": null,
-        "targets": [
-          "Linear"
-        ],
-        "weights": {
-          "block_structure": null,
-          "dynamic": false,
-          "group_size": null,
-          "num_bits": 8,
-          "observer": "minmax",
-          "observer_kwargs": {},
-          "strategy": "channel",
-          "symmetric": true,
-          "type": "int"
-        }
-      }
-    },
-    "format": "int-quantized",
-    "global_compression_ratio": 1.1652744252495255,
-    "ignore": [
-      "lm_head"
-    ],
-    "kv_cache_scheme": null,
-    "quant_method": "compressed-tensors",
-    "quantization_status": "frozen"
-  },
   "embd_pdrop": 0.0,
   "eos_token_id": 32000,
   "hidden_act": "silu",
@@ -206,5 +166,45 @@
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.1",
   "use_cache": true,
-  "vocab_size": 32064
+  "vocab_size": 32064,
+  "quantization_config": {
+    "config_groups": {
+      "group_0": {
+        "input_activations": {
+          "block_structure": null,
+          "dynamic": true,
+          "group_size": null,
+          "num_bits": 8,
+          "observer": "memoryless",
+          "observer_kwargs": {},
+          "strategy": "token",
+          "symmetric": true,
+          "type": "int"
+        },
+        "output_activations": null,
+        "targets": [
+          "Linear"
+        ],
+        "weights": {
+          "block_structure": null,
+          "dynamic": false,
+          "group_size": null,
+          "num_bits": 8,
+          "observer": "minmax",
+          "observer_kwargs": {},
+          "strategy": "channel",
+          "symmetric": true,
+          "type": "int"
+        }
+      }
+    },
+    "format": "int-quantized",
+    "global_compression_ratio": 1.1652744252495255,
+    "ignore": [
+      "lm_head"
+    ],
+    "kv_cache_scheme": null,
+    "quant_method": "compressed-tensors",
+    "quantization_status": "frozen"
+  }
 }