[Qwen3-TTS/ASR]update quantization comments of model conversion (#3414)

Author: openvino-dev-samples

notebooks/qwen3-asr/qwen3-asr.ipynb

@@ -236,7 +236,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": null,
    "id": "0c1a380f",
    "metadata": {},
    "outputs": [
@@ -251,6 +251,8 @@
    "source": [
     "from qwen_3_asr_helper import convert_qwen3_asr_model\n",
     "\n",
+    "# from nncf import CompressWeightsMode\n",
+    "\n",
     "model_id = model_selector.value\n",
     "model_name = model_id.split(\"/\")[-1]\n",
     "ov_model_dir = Path(f\"{model_name}-OV\")\n",
@@ -260,7 +262,7 @@
     "convert_qwen3_asr_model(\n",
     "    model_id=model_id,\n",
     "    output_dir=ov_model_dir,\n",
-    "    quantization_config=None,  # Set to {\"mode\": \"INT8_SYM\"} for INT8 quantization\n",
+    "    quantization_config=None,  # Set to {\"mode\": CompressWeightsMode.INT8_SYM} for INT8 quantization\n",
     ")"
    ]
   },

notebooks/qwen3-tts/qwen3-tts.ipynb

@@ -237,7 +237,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -381,6 +381,8 @@
    "source": [
     "from qwen_3_tts_helper import convert_qwen3_tts_model\n",
     "\n",
+    "# from nncf import CompressWeightsMode\n",
+    "\n",
     "model_name = model_selector.value\n",
     "model_id = model_options[model_name]\n",
     "ov_model_dir = Path(f\"{model_name}-OV\")\n",
@@ -390,7 +392,7 @@
     "convert_qwen3_tts_model(\n",
     "    model_id=model_id,\n",
     "    output_dir=ov_model_dir,\n",
-    "    quantization_config=None,  # Set to {\"mode\": \"INT8_SYM\"} for INT8 quantization\n",
+    "    quantization_config=None,  # Set to {\"mode\": CompressWeightsMode.INT8_SYM} for INT8 quantization\n",
     ")"
    ]
   },

notebooks/qwen3-tts/qwen_3_tts_helper.py

@@ -866,10 +866,10 @@ def forward_wrap_code_predictor(
         patch_stateful(ov_model, 2)
         print("✅ Talker Code Predictor model successfully converted")
 
-        if quantization_config is not None:
-            print(f"⌛ Weights compression with {quantization_config['mode']} mode started")
-            ov_model = nncf.compress_weights(ov_model, **quantization_config)
-            print("✅ Weights compression finished")
+        # if quantization_config is not None:
+        #     print(f"⌛ Weights compression with {quantization_config['mode']} mode started")
+        #     ov_model = nncf.compress_weights(ov_model, **quantization_config)
+        #     print("✅ Weights compression finished")
 
         ov.save_model(ov_model, talker_code_predictor_path)
         del ov_model