|
12 | 12 | "\n",
|
13 | 13 | "This notebook contains the following information and examples:\n",
|
14 | 14 | "\n",
|
15 |
| - "1. Setup LLaVA \n", |
16 |
| - " - Option 1: Use API calls from `Replicate`\n", |
17 |
| - " - Option 2: Setup LLaVA locally (requires GPU)\n", |
18 |
| - "2. Application 1: Image Chat\n", |
19 |
| - "3. Application 2: Figure Creator" |
| 15 | + "1. Install [LLaVA package](#install)\n", |
| 16 | + "2. Setup LLaVA Model\n", |
| 17 | + " - Option 1: Use [API calls from `Replicate`](#replicate)\n", |
| 18 | + " - Option 2: Setup [LLaVA locally (requires GPU)](#local)\n", |
| 19 | + "2. Application 1: [Image Chat](#app-1)\n", |
| 20 | + "3. Application 2: [Figure Creator](#app-2)" |
20 | 21 | ]
|
21 | 22 | },
|
22 | 23 | {
|
|
49 | 50 | "id": "d64154f0",
|
50 | 51 | "metadata": {},
|
51 | 52 | "source": [
|
| 53 | + "<a id=\"install\"></a>\n", |
52 | 54 | "## Install the LLaVA library\n",
|
53 | 55 | "\n",
|
54 | 56 | "Please follow the LLaVA GitHub [page](https://github.com/haotian-liu/LLaVA/) to install LLaVA.\n",
|
|
108 | 110 | "id": "acc4703b",
|
109 | 111 | "metadata": {},
|
110 | 112 | "source": [
|
111 |
| - "# (Option 1, preferred) Use API Calls from Replicate [Remote]\n", |
| 113 | + "<a id=\"replicate\"></a>\n", |
| 114 | + "## (Option 1, preferred) Use API Calls from Replicate [Remote]\n", |
112 | 115 | "We can also use [Replicate](https://replicate.com/yorickvp/llava-13b/api) to use LLaVA directly, which will host the model for you.\n",
|
113 | 116 | "\n",
|
114 | 117 | "1. Run `pip install replicate` to install the package\n",
|
|
148 | 151 | "id": "1805e4bd",
|
149 | 152 | "metadata": {},
|
150 | 153 | "source": [
|
| 154 | + "<a id=\"local\"></a>\n", |
151 | 155 | "## [Option 2] Setup LLaVA Locally\n",
|
152 | 156 | "\n",
|
153 | 157 | "\n",
|
|
537 | 541 | "id": "7e4faf59",
|
538 | 542 | "metadata": {},
|
539 | 543 | "source": [
|
| 544 | + "<a id=\"app-1\"></a>\n", |
540 | 545 | "## Application 1: Image Chat\n",
|
541 | 546 | "\n",
|
542 | 547 | "In this section, we present a straightforward dual-agent architecture to enable user to chat with a multimodal agent.\n",
|
|
859 | 864 | "id": "0c40d0eb",
|
860 | 865 | "metadata": {},
|
861 | 866 | "source": [
|
862 |
| - "# Application 2: Figure Creator\n", |
| 867 | + "<a id=\"app-2\"></a>\n", |
| 868 | + "## Application 2: Figure Creator\n", |
863 | 869 | "\n",
|
864 | 870 | "Here, we define a `FigureCreator` agent, which contains three child agents: commander, coder, and critics.\n",
|
865 | 871 | "\n",
|
|
0 commit comments