fix: update vto and product recontext params (#2431)

Author: katiemn

vision/getting-started/imagen_product_recontext.ipynb

@@ -375,7 +375,7 @@
         "\n",
         "In each call, you can specify the following parameters in addition to `product_images`:\n",
         " - **Prompt:** Describe the setting you'd like the product to appear in.\n",
-        " - **Base steps:** This integer controls image generation. Higher values increase quality but also increase latency.\n",
+        " - **Enhance Prompt:** This model offers the option to enhance your provided prompt, which is enabled by default. If you'd like to disable this feature, set `enhance_prompt` to False.\n",
         " - **Number of images:** 1 - 4"
       ]
     },
@@ -398,10 +398,10 @@
         "        ],\n",
         "    ),\n",
         "    config=RecontextImageConfig(\n",
-        "        base_steps=32,\n",
+        "        enhance_prompt=True,\n",
+        "        output_mime_type=\"image/jpeg\",\n",
         "        number_of_images=1,\n",
         "        safety_filter_level=\"BLOCK_LOW_AND_ABOVE\",\n",
-        "        person_generation=\"ALLOW_ADULT\",\n",
         "    ),\n",
         ")\n",
         "\n",
@@ -417,12 +417,8 @@
         "You can also recontextualize products without setting a `prompt` in the request. The model will generate an appropriate background based on the product image.\n",
         "\n",
         "\n",
-        "When generating images with people you can also set the `safety_filter_level` and `person_generation` parameters accordingly:\n",
+        "When generating images you can also set the `safety_filter_level` accordingly:\n",
         "\n",
-        "- `person_generation`\n",
-        "  - `DONT_ALLOW`\n",
-        "  - `ALLOW_ADULT`\n",
-        "  - `ALLOW_ALL`\n",
         "- `safety_filter_level`\n",
         "  - `BLOCK_LOW_AND_ABOVE`\n",
         "  - `BLOCK_MEDIUM_AND_ABOVE`\n",
@@ -446,10 +442,9 @@
         "        ],\n",
         "    ),\n",
         "    config=RecontextImageConfig(\n",
-        "        base_steps=32,\n",
+        "        output_mime_type=\"image/jpeg\",\n",
         "        number_of_images=1,\n",
         "        safety_filter_level=\"BLOCK_LOW_AND_ABOVE\",\n",
-        "        person_generation=\"ALLOW_ADULT\",\n",
         "    ),\n",
         ")\n",
         "\n",
@@ -519,8 +514,10 @@
       },
       "outputs": [],
       "source": [
+        "# fmt: off\n",
         "prompt = \"on a running track\"  # @param {type: 'string'}\n",
         "gcs_uri = \"gs://cloud-samples-data/generative-ai/image/shoes.jpg\"  # @param {type: 'string'}\n",
+        "# fmt: on\n",
         "\n",
         "\n",
         "response = client.models.recontext_image(\n",
@@ -530,10 +527,10 @@
         "        product_images=[ProductImage(product_image=Image(gcs_uri=gcs_uri))],\n",
         "    ),\n",
         "    config=RecontextImageConfig(\n",
-        "        base_steps=32,\n",
+        "        output_mime_type=\"image/jpeg\",\n",
+        "        enhance_prompt=True,\n",
         "        number_of_images=1,\n",
         "        safety_filter_level=\"BLOCK_LOW_AND_ABOVE\",\n",
-        "        person_generation=\"ALLOW_ADULT\",\n",
         "    ),\n",
         ")\n",
         "\n",

vision/getting-started/virtual_try_on.ipynb

@@ -108,7 +108,7 @@
         "- Try-on multiple clothing items from locally stored images\n",
         "- Try-on a clothing item in Cloud Storage with an Imagen generated person\n",
         "\n",
-        "Learn more about [quotas](https://cloud.google.com/vertex-ai/generative-ai/docs/models/imagen/virtual-try-on-preview-08-04) and [pricing](https://cloud.google.com/vertex-ai/generative-ai/pricing#imagen-models) for Virtual Try-On in the product documentation. "
+        "Learn more about [quotas](https://cloud.google.com/vertex-ai/generative-ai/docs/models/imagen/virtual-try-on-preview-08-04) and [pricing](https://cloud.google.com/vertex-ai/generative-ai/pricing#imagen-models) for Virtual Try-On in the product documentation."
       ]
     },
     {
@@ -418,7 +418,6 @@
         "### Send the request\n",
         "\n",
         "With the Virtual Try-On model, you can only specify one clothing item to try on at a time. Since this example has two clothing items, you'll need to make two separate requests. In each call, you can specify the following parameters in addition to the `person_image` and `product_images`:\n",
-        " - **Base steps:** An integer that controls image generation, with higher steps trading higher quality for increased latency.\n",
         " - **Number of images:** 1 - 4\n",
         "\n",
         "You'll save the output image locally so that it can be referenced in the next step."
@@ -441,10 +440,9 @@
         "        ],\n",
         "    ),\n",
         "    config=RecontextImageConfig(\n",
-        "        base_steps=32,\n",
+        "        output_mime_type=\"image/jpeg\",\n",
         "        number_of_images=1,\n",
         "        safety_filter_level=\"BLOCK_LOW_AND_ABOVE\",\n",
-        "        person_generation=\"ALLOW_ADULT\",\n",
         "    ),\n",
         ")\n",
         "\n",
@@ -458,12 +456,7 @@
         "id": "cP_gkDUVx-3Z"
       },
       "source": [
-        "When generating images of people you can also set the `safety_filter_level` and `person_generation` parameters accordingly:\n",
-        "\n",
-        "- `person_generation`\n",
-        "  - `DONT_ALLOW`\n",
-        "  - `ALLOW_ADULT`\n",
-        "  - `ALLOW_ALL`\n",
+        "When generating images you can also set the `safety_filter_level` parameter accordingly:\n",
         "- `safety_filter_level`\n",
         "  - `BLOCK_LOW_AND_ABOVE`\n",
         "  - `BLOCK_MEDIUM_AND_ABOVE`\n",
@@ -488,10 +481,9 @@
         "        ],\n",
         "    ),\n",
         "    config=RecontextImageConfig(\n",
-        "        base_steps=32,\n",
+        "        output_mime_type=\"image/jpeg\",\n",
         "        number_of_images=1,\n",
         "        safety_filter_level=\"BLOCK_LOW_AND_ABOVE\",\n",
-        "        person_generation=\"ALLOW_ADULT\",\n",
         "    ),\n",
         ")\n",
         "display_image(response.generated_images[0].image)"
@@ -535,10 +527,10 @@
         "    model=image_generation,\n",
         "    prompt=prompt,\n",
         "    config=GenerateImagesConfig(\n",
+        "        output_mime_type=\"image/jpeg\",\n",
         "        number_of_images=1,\n",
         "        image_size=\"2K\",\n",
         "        safety_filter_level=\"BLOCK_MEDIUM_AND_ABOVE\",\n",
-        "        person_generation=\"ALLOW_ADULT\",\n",
         "    ),\n",
         ")\n",
         "display_image(image.generated_images[0].image)"
@@ -609,10 +601,9 @@
         "        ],\n",
         "    ),\n",
         "    config=RecontextImageConfig(\n",
-        "        base_steps=32,\n",
+        "        output_mime_type=\"image/jpeg\",\n",
         "        number_of_images=1,\n",
         "        safety_filter_level=\"BLOCK_LOW_AND_ABOVE\",\n",
-        "        person_generation=\"ALLOW_ADULT\",\n",
         "    ),\n",
         ")\n",
         "display_image(response.generated_images[0].image)"