File size: 7,053 Bytes
c607184 8e20010 c607184 8e20010 c607184 bba9479 c607184 bba9479 8e20010 c607184 8e20010 c607184 8e20010 5d9d806 8e20010 c607184 8e20010 c607184 8e20010 c607184 8e20010 c607184 0985eef c607184 5d9d806 8e20010 c607184 8e20010 c607184 8e20010 c607184 8e20010 0985eef 5d9d806 0985eef 8e20010 c607184 5d9d806 0985eef 5d9d806 bba9479 5d9d806 bba9479 0985eef 5d9d806 c607184 8e20010 c607184 8e20010 0985eef 5d9d806 0985eef 8e20010 c607184 5d9d806 0985eef 5d9d806 bba9479 5d9d806 bba9479 0985eef 5d9d806 c607184 8e20010 c607184 0985eef 5d9d806 0985eef 8e20010 5d9d806 0985eef 5d9d806 bba9479 5d9d806 bba9479 0985eef 5d9d806 8e20010 c607184 8e20010 c607184 8e20010 c607184 5d9d806 0985eef bba9479 5d9d806 bba9479 c607184 0985eef 5d9d806 0985eef 8e20010 c607184 5d9d806 c607184 bba9479 c607184 bba9479 5d9d806 8e20010 5d9d806 c607184 8e20010 c607184 8e20010 c607184 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 |
{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"id": "U8RTc2PmnX-v"
},
"source": [
"Initial setup"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "kGW7vfRkrqHe"
},
"outputs": [],
"source": [
"!pip install -r https://huggingface.co/flunardelli/llm-metaeval/raw/main/requirements.txt"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "2I850FIsCVNw"
},
"outputs": [],
"source": [
"from datetime import datetime\n",
"import os\n",
"from huggingface_hub import login, upload_folder\n",
"from google.colab import userdata\n",
"import shutil\n",
"\n",
"HF_TOKEN = userdata.get('HF_TOKEN')\n",
"login(HF_TOKEN, True)\n",
"BASE_DATASET='mmlu'\n",
"REPO_ID='flunardelli/llm-metaeval'\n",
"BASE_FOLDER=f\"/content/{BASE_DATASET}/\"#{datetime.now().strftime('%Y-%m-%dT%H-%M-%S')}\n",
"OUTPUT_FOLDER=os.path.join(BASE_FOLDER,'output')\n",
"TASK_FOLDER=os.path.join(BASE_FOLDER,'tasks')\n",
"#shutil.rmtree(BASE_FOLDER)\n",
"os.makedirs(OUTPUT_FOLDER)\n",
"os.makedirs(TASK_FOLDER)\n",
"os.environ['HF_TOKEN'] = HF_TOKEN\n",
"os.environ['OUTPUT_FOLDER'] = OUTPUT_FOLDER\n",
"os.environ['TASK_FOLDER'] = TASK_FOLDER\n",
"\n",
"def hf_upload_folder(folder_path):\n",
" upload_folder(\n",
" folder_path=folder_path,\n",
" path_in_repo=\"evals/\",\n",
" repo_id=REPO_ID,\n",
" token=HF_TOKEN,\n",
" repo_type=\"dataset\"\n",
" )\n",
"\n",
"def create_task(content, filename):\n",
" filename_path = os.path.join(TASK_FOLDER,filename)\n",
" with open(filename_path, \"w\") as f:\n",
" f.write(content)"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "Jd2JwKZaPkNS"
},
"source": [
"Create task for MMLU all datasets"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "xP0cC_sHih7C"
},
"outputs": [],
"source": [
"YAML_mmlu_en_us_string = \"\"\"\n",
"task: mmlu_all\n",
"dataset_path: cais/mmlu\n",
"dataset_name: all\n",
"description: \"MMLU dataset\"\n",
"test_split: test\n",
"fewshot_split: dev\n",
"fewshot_config:\n",
" sampler: first_n\n",
"num_fewshot: 5\n",
"output_type: multiple_choice\n",
"doc_to_text: \"{{question.strip()}}\\nA. {{choices[0]}}\\nB. {{choices[1]}}\\nC. {{choices[2]}}\\nD. {{choices[3]}}\\nAnswer:\"\n",
"doc_to_choice: [\"A\", \"B\", \"C\", \"D\"]\n",
"doc_to_target: answer\n",
"metric_list:\n",
" - metric: acc\n",
" aggregation: mean\n",
" higher_is_better: true\n",
"\"\"\"\n",
"create_task(YAML_mmlu_en_us_string, 'mmlu_en_us.yaml')\n",
"os.environ['TASKS'] = 'mmlu_all'\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "mJjo_A5tP-Td"
},
"source": [
"Llama Models"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "IzP5nyP0Gwk8"
},
"outputs": [],
"source": [
"!lm_eval \\\n",
"--model hf --model_args pretrained=meta-llama/Llama-3.2-1B-Instruct,revision=d0a2081ed47e20ce524e8bc5d132f3fad2f69ff0,trust_remote_code=False,dtype=bfloat16,parallelize=True \\\n",
"--tasks $TASKS \\\n",
"--include_path $TASK_FOLDER/. --output $OUTPUT_FOLDER --log_samples \\\n",
"--batch_size auto &> run.log\n",
"#--limit 10 \\"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "uMoitxJkHerH"
},
"outputs": [],
"source": [
"hf_upload_folder(BASE_FOLDER)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "oIACOAhDW5ow"
},
"outputs": [],
"source": [
"!lm_eval \\\n",
"--model hf --model_args pretrained=meta-llama/Llama-3.2-3B-Instruct,revision=392a143b624368100f77a3eafaa4a2468ba50a72,trust_remote_code=False,dtype=bfloat16,parallelize=True \\\n",
"--tasks $TASKS \\\n",
"--include_path $TASK_FOLDER/. --output $OUTPUT_FOLDER --log_samples \\\n",
"--batch_size auto &> run.log\n",
"#--limit 10 \\"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "eIUOqu5sHfkM"
},
"outputs": [],
"source": [
"hf_upload_folder(BASE_FOLDER)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "cFFYPzBIYGf7"
},
"outputs": [],
"source": [
"!lm_eval \\\n",
"--model hf --model_args pretrained=meta-llama/Meta-Llama-3-8B,revision=62bd457b6fe961a42a631306577e622c83876cb6,trust_remote_code=False,dtype=bfloat16,parallelize=True \\\n",
"--tasks $TASKS \\\n",
"--include_path $TASK_FOLDER/. --output $OUTPUT_FOLDER --log_samples \\\n",
"--batch_size auto &> run.log\n",
"#--limit 10 \\"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "xsL82Q4SHgMn"
},
"outputs": [],
"source": [
"hf_upload_folder(BASE_FOLDER)"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "1fEX-49hQ-Be"
},
"source": [
"Mistral Models"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": true,
"id": "ilu9_ulWTy3p"
},
"outputs": [],
"source": [
"!lm_eval \\\n",
"--model hf --model_args pretrained=mistralai/Mixtral-8x7B-Instruct-v0.1,revision=41bd4c9e7e4fb318ca40e721131d4933966c2cc1,trust_remote_code=False,dtype=bfloat16,parallelize=True \\\n",
"--tasks $TASKS \\\n",
"--include_path $TASK_FOLDER/. --output $OUTPUT_FOLDER --log_samples \\\n",
"--batch_size auto &> run.log\n",
"#--limit 10 \\"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "jE5r8gVDHhAz"
},
"outputs": [],
"source": [
"hf_upload_folder(BASE_FOLDER)"
]
}
],
"metadata": {
"accelerator": "GPU",
"colab": {
"gpuType": "L4",
"machine_shape": "hm",
"provenance": []
},
"kernelspec": {
"display_name": "Python 3",
"name": "python3"
},
"language_info": {
"name": "python"
}
},
"nbformat": 4,
"nbformat_minor": 0
} |