File size: 3,206 Bytes
13a81c2 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 | {
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"colab": {
"provenance": [],
"gpuType": "A100"
},
"kernelspec": {
"name": "python3",
"display_name": "Python 3"
},
"language_info": {
"name": "python"
},
"accelerator": "GPU"
},
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "aCl-IzLoDr2H"
},
"outputs": [],
"source": [
"!pip install -U transformers mamba-ssm"
]
},
{
"cell_type": "markdown",
"source": [
"# Load Models"
],
"metadata": {
"id": "SpRo_KJIRsxv"
}
},
{
"cell_type": "code",
"source": [
"import torch\n",
"from transformers import AutoTokenizer, AutoModelForCausalLM\n",
"\n",
"# Load tokenizer and model\n",
"tokenizer = AutoTokenizer.from_pretrained(\"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16\")\n",
"model = AutoModelForCausalLM.from_pretrained(\n",
" \"nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16\",\n",
" torch_dtype=torch.bfloat16,\n",
" trust_remote_code=True,\n",
" device_map=\"auto\"\n",
")\n"
],
"metadata": {
"id": "waveliieEI1n"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"source": [
"# Define Input with Tools"
],
"metadata": {
"id": "xjVkqaSdRx0_"
}
},
{
"cell_type": "code",
"source": [
"from transformers.utils import get_json_schema\n",
"\n",
"def multiply(a: float, b: float):\n",
" \"\"\"\n",
" A function that multiplies two numbers\n",
"\n",
" Args:\n",
" a: The first number to multiply\n",
" b: The second number to multiply\n",
" \"\"\"\n",
" return a * b\n",
"\n",
"messages = [\n",
" {\"role\": \"user\", \"content\": \"what is 2.0909090923 x 0.897987987\"},\n",
"]\n",
"\n",
"tokenized_chat = tokenizer.apply_chat_template(\n",
" messages,\n",
" tools=[\n",
" multiply\n",
" ],\n",
" tokenize=True,\n",
" add_generation_prompt=True,\n",
" return_tensors=\"pt\"\n",
").to(model.device)\n"
],
"metadata": {
"id": "zxZZ7iMZETsw"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"source": [
"# Inference"
],
"metadata": {
"id": "SVBAG3dLRw4v"
}
},
{
"cell_type": "code",
"source": [
"outputs = model.generate(\n",
" tokenized_chat,\n",
" max_new_tokens=1024,\n",
" temperature=1.0,\n",
" top_p=1.0,\n",
" eos_token_id=tokenizer.eos_token_id\n",
")\n",
"print(tokenizer.decode(outputs[0]))"
],
"metadata": {
"id": "BKYqPT5ORDx3"
},
"execution_count": null,
"outputs": []
}
]
} |