Delete assets/basic_inference_llama_2_70b_dolphin.ipynb
Browse files
assets/basic_inference_llama_2_70b_dolphin.ipynb
DELETED
@@ -1,355 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"cells": [
|
3 |
-
{
|
4 |
-
"cell_type": "code",
|
5 |
-
"execution_count": 1,
|
6 |
-
"id": "6f46e840-8a7f-4be2-a082-49b9ebf5a8c5",
|
7 |
-
"metadata": {},
|
8 |
-
"outputs": [
|
9 |
-
{
|
10 |
-
"name": "stdout",
|
11 |
-
"output_type": "stream",
|
12 |
-
"text": [
|
13 |
-
"\n",
|
14 |
-
"\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip is available: \u001b[0m\u001b[31;49m23.1.2\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m23.2.1\u001b[0m\n",
|
15 |
-
"\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49mpython3 -m pip install --upgrade pip\u001b[0m\n"
|
16 |
-
]
|
17 |
-
}
|
18 |
-
],
|
19 |
-
"source": [
|
20 |
-
"!pip install -q -U huggingface_hub peft transformers torch accelerate\n"
|
21 |
-
]
|
22 |
-
},
|
23 |
-
{
|
24 |
-
"cell_type": "code",
|
25 |
-
"execution_count": 2,
|
26 |
-
"id": "2d2918a1-d701-4a66-946c-6f668cb4ac1e",
|
27 |
-
"metadata": {},
|
28 |
-
"outputs": [
|
29 |
-
{
|
30 |
-
"name": "stdout",
|
31 |
-
"output_type": "stream",
|
32 |
-
"text": [
|
33 |
-
"Mon Jul 24 21:41:13 2023 \n",
|
34 |
-
"+-----------------------------------------------------------------------------+\n",
|
35 |
-
"| NVIDIA-SMI 525.105.17 Driver Version: 525.105.17 CUDA Version: 12.0 |\n",
|
36 |
-
"|-------------------------------+----------------------+----------------------+\n",
|
37 |
-
"| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |\n",
|
38 |
-
"| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |\n",
|
39 |
-
"| | | MIG M. |\n",
|
40 |
-
"|===============================+======================+======================|\n",
|
41 |
-
"| 0 NVIDIA H100 PCIe On | 00000000:06:00.0 Off | 0 |\n",
|
42 |
-
"| N/A 39C P0 52W / 350W | 0MiB / 81559MiB | 0% Default |\n",
|
43 |
-
"| | | Disabled |\n",
|
44 |
-
"+-------------------------------+----------------------+----------------------+\n",
|
45 |
-
" \n",
|
46 |
-
"+-----------------------------------------------------------------------------+\n",
|
47 |
-
"| Processes: |\n",
|
48 |
-
"| GPU GI CI PID Type Process name GPU Memory |\n",
|
49 |
-
"| ID ID Usage |\n",
|
50 |
-
"|=============================================================================|\n",
|
51 |
-
"| No running processes found |\n",
|
52 |
-
"+-----------------------------------------------------------------------------+\n"
|
53 |
-
]
|
54 |
-
}
|
55 |
-
],
|
56 |
-
"source": [
|
57 |
-
"!nvidia-smi"
|
58 |
-
]
|
59 |
-
},
|
60 |
-
{
|
61 |
-
"cell_type": "code",
|
62 |
-
"execution_count": 3,
|
63 |
-
"id": "0afdf8a6-ea7d-44ab-a1f9-a19e550e9dbd",
|
64 |
-
"metadata": {},
|
65 |
-
"outputs": [
|
66 |
-
{
|
67 |
-
"name": "stderr",
|
68 |
-
"output_type": "stream",
|
69 |
-
"text": [
|
70 |
-
"/home/ubuntu/.local/lib/python3.8/site-packages/pandas/core/computation/expressions.py:20: UserWarning: Pandas requires version '2.7.3' or newer of 'numexpr' (version '2.7.1' currently installed).\n",
|
71 |
-
" from pandas.core.computation.check import NUMEXPR_INSTALLED\n"
|
72 |
-
]
|
73 |
-
}
|
74 |
-
],
|
75 |
-
"source": [
|
76 |
-
"import torch\n",
|
77 |
-
"from peft import PeftModel, PeftConfig\n",
|
78 |
-
"from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig"
|
79 |
-
]
|
80 |
-
},
|
81 |
-
{
|
82 |
-
"cell_type": "code",
|
83 |
-
"execution_count": 4,
|
84 |
-
"id": "adfcd11e-8d98-4cf3-abf4-e9fa933eb0d6",
|
85 |
-
"metadata": {},
|
86 |
-
"outputs": [
|
87 |
-
{
|
88 |
-
"data": {
|
89 |
-
"application/vnd.jupyter.widget-view+json": {
|
90 |
-
"model_id": "7dc80313fdcd41a5a7ee168956df3dd9",
|
91 |
-
"version_major": 2,
|
92 |
-
"version_minor": 0
|
93 |
-
},
|
94 |
-
"text/plain": [
|
95 |
-
"VBox(children=(HTML(value='<center> <img\\nsrc=https://huggingface.co/front/assets/huggingface_logo-noborder.sv…"
|
96 |
-
]
|
97 |
-
},
|
98 |
-
"metadata": {},
|
99 |
-
"output_type": "display_data"
|
100 |
-
}
|
101 |
-
],
|
102 |
-
"source": [
|
103 |
-
"from huggingface_hub import notebook_login\n",
|
104 |
-
"\n",
|
105 |
-
"notebook_login()"
|
106 |
-
]
|
107 |
-
},
|
108 |
-
{
|
109 |
-
"cell_type": "code",
|
110 |
-
"execution_count": 5,
|
111 |
-
"id": "82cfa4fb-af16-4927-82c4-1fbf0fa84bfa",
|
112 |
-
"metadata": {},
|
113 |
-
"outputs": [
|
114 |
-
{
|
115 |
-
"name": "stderr",
|
116 |
-
"output_type": "stream",
|
117 |
-
"text": [
|
118 |
-
"/home/ubuntu/.local/lib/python3.8/site-packages/transformers/modeling_utils.py:2193: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers.\n",
|
119 |
-
" warnings.warn(\n"
|
120 |
-
]
|
121 |
-
},
|
122 |
-
{
|
123 |
-
"data": {
|
124 |
-
"application/vnd.jupyter.widget-view+json": {
|
125 |
-
"model_id": "d0f18088e32f4d4b857d2de5430528d4",
|
126 |
-
"version_major": 2,
|
127 |
-
"version_minor": 0
|
128 |
-
},
|
129 |
-
"text/plain": [
|
130 |
-
"Loading checkpoint shards: 0%| | 0/15 [00:00<?, ?it/s]"
|
131 |
-
]
|
132 |
-
},
|
133 |
-
"metadata": {},
|
134 |
-
"output_type": "display_data"
|
135 |
-
}
|
136 |
-
],
|
137 |
-
"source": [
|
138 |
-
"# peft_model_id = \"results/checkpoint-12500\"\n",
|
139 |
-
"peft_model_id = \"dfurman/llama-2-70b-dolphin-peft\"\n",
|
140 |
-
"config = PeftConfig.from_pretrained(peft_model_id)\n",
|
141 |
-
"\n",
|
142 |
-
"bnb_config = BitsAndBytesConfig(\n",
|
143 |
-
" load_in_4bit=True,\n",
|
144 |
-
" bnb_4bit_quant_type=\"nf4\",\n",
|
145 |
-
" bnb_4bit_compute_dtype=torch.bfloat16,\n",
|
146 |
-
")\n",
|
147 |
-
"\n",
|
148 |
-
"model = AutoModelForCausalLM.from_pretrained(\n",
|
149 |
-
" config.base_model_name_or_path,\n",
|
150 |
-
" quantization_config=bnb_config,\n",
|
151 |
-
" use_auth_token=True,\n",
|
152 |
-
" torch_dtype=torch.bfloat16,\n",
|
153 |
-
" device_map=\"auto\",\n",
|
154 |
-
")\n",
|
155 |
-
"tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)\n",
|
156 |
-
"tokenizer.pad_token = tokenizer.eos_token\n",
|
157 |
-
"\n",
|
158 |
-
"# Load the Lora model\n",
|
159 |
-
"model = PeftModel.from_pretrained(model, peft_model_id)"
|
160 |
-
]
|
161 |
-
},
|
162 |
-
{
|
163 |
-
"cell_type": "code",
|
164 |
-
"execution_count": 6,
|
165 |
-
"id": "d86f6a79-95f2-4e05-9bc7-3cbcbbbc9552",
|
166 |
-
"metadata": {},
|
167 |
-
"outputs": [],
|
168 |
-
"source": [
|
169 |
-
"# text generation function\n",
|
170 |
-
"\n",
|
171 |
-
"\n",
|
172 |
-
"def llama_generate(\n",
|
173 |
-
" model: AutoModelForCausalLM,\n",
|
174 |
-
" tokenizer: AutoTokenizer,\n",
|
175 |
-
" prompt: str,\n",
|
176 |
-
" max_new_tokens: int = 128,\n",
|
177 |
-
" temperature: int = 1.0,\n",
|
178 |
-
") -> str:\n",
|
179 |
-
" \"\"\"\n",
|
180 |
-
" Initialize the pipeline\n",
|
181 |
-
" Uses Hugging Face GenerationConfig defaults\n",
|
182 |
-
" https://huggingface.co/docs/transformers/v4.29.1/en/main_classes/text_generation#transformers.GenerationConfig\n",
|
183 |
-
" Args:\n",
|
184 |
-
" model (transformers.AutoModelForCausalLM): Falcon model for text generation\n",
|
185 |
-
" tokenizer (transformers.AutoTokenizer): Tokenizer for model\n",
|
186 |
-
" prompt (str): Prompt for text generation\n",
|
187 |
-
" max_new_tokens (int, optional): Max new tokens after the prompt to generate. Defaults to 128.\n",
|
188 |
-
" temperature (float, optional): The value used to modulate the next token probabilities.\n",
|
189 |
-
" Defaults to 1.0\n",
|
190 |
-
" \"\"\"\n",
|
191 |
-
" device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
|
192 |
-
"\n",
|
193 |
-
" inputs = tokenizer(\n",
|
194 |
-
" [prompt],\n",
|
195 |
-
" return_tensors=\"pt\",\n",
|
196 |
-
" return_token_type_ids=False,\n",
|
197 |
-
" ).to(\n",
|
198 |
-
" device\n",
|
199 |
-
" ) # tokenize inputs, load on device\n",
|
200 |
-
"\n",
|
201 |
-
" # when running Torch modules in lower precision, it is best practice to use the torch.autocast context manager.\n",
|
202 |
-
" with torch.autocast(\"cuda\", dtype=torch.bfloat16):\n",
|
203 |
-
" response = model.generate(\n",
|
204 |
-
" **inputs,\n",
|
205 |
-
" max_new_tokens=max_new_tokens,\n",
|
206 |
-
" temperature=temperature,\n",
|
207 |
-
" return_dict_in_generate=True,\n",
|
208 |
-
" eos_token_id=tokenizer.eos_token_id,\n",
|
209 |
-
" pad_token_id=tokenizer.pad_token_id,\n",
|
210 |
-
" )\n",
|
211 |
-
"\n",
|
212 |
-
" decoded_output = tokenizer.decode(\n",
|
213 |
-
" response[\"sequences\"][0],\n",
|
214 |
-
" skip_special_tokens=True,\n",
|
215 |
-
" ) # grab output in natural language\n",
|
216 |
-
"\n",
|
217 |
-
" return decoded_output[len(prompt) :] # remove prompt from output"
|
218 |
-
]
|
219 |
-
},
|
220 |
-
{
|
221 |
-
"cell_type": "code",
|
222 |
-
"execution_count": 7,
|
223 |
-
"id": "28be263a-dd15-419f-a67e-7ca05b27435f",
|
224 |
-
"metadata": {},
|
225 |
-
"outputs": [
|
226 |
-
{
|
227 |
-
"name": "stdout",
|
228 |
-
"output_type": "stream",
|
229 |
-
"text": [
|
230 |
-
"Sure! Here's a delicious and easy vegan banana bread recipe:\n",
|
231 |
-
"\n",
|
232 |
-
"Ingredients:\n",
|
233 |
-
"- 2 cups all-purpose flour\n",
|
234 |
-
"- 1/2 cup sugar\n",
|
235 |
-
"- 1/2 cup vegan butter (such as Earth Balance)\n",
|
236 |
-
"- 1/2 cup vegan milk (such as almond milk)\n",
|
237 |
-
"- 1/2 cup unsweetened applesauce\n",
|
238 |
-
"- 1/2 cup mashed ripe bananas (about 2 medium bananas)\n",
|
239 |
-
"- 1 teaspoon baking soda\n",
|
240 |
-
"- 1/2 teaspoon salt\n",
|
241 |
-
"- 1/2 teaspoon ground cinnamon\n",
|
242 |
-
"- 1/2 teaspoon ground nutmeg\n",
|
243 |
-
"- 1/2 teaspoon ground cloves\n",
|
244 |
-
"- 1/2 cup chopped walnuts (optional)\n",
|
245 |
-
"\n",
|
246 |
-
"Instructions:\n",
|
247 |
-
"1. Preheat the oven to 350°F (175°C). Grease a 9x5-inch loaf pan with vegan butter or cooking spray.\n",
|
248 |
-
"2. In a large bowl, mix together the flour, sugar, vegan butter, vegan milk, applesauce, bananas, baking soda, salt, cinnamon, nutmeg, and cloves. Stir until well combined.\n",
|
249 |
-
"3. Fold in the chopped walnuts, if using.\n",
|
250 |
-
"4. Pour the batter into the prepared loaf pan.\n",
|
251 |
-
"5. Bake for 50-60 minutes, or until a toothpick inserted into the center of the bread comes out clean.\n",
|
252 |
-
"6. Let the bread cool in the pan for 10 minutes before transferring it to a wire rack to cool completely.\n",
|
253 |
-
"7. Slice and enjoy!\n",
|
254 |
-
"\n",
|
255 |
-
"Note: You can also add chocolate chips, dried fruit, or other mix-ins to the batter for extra flavor and texture. Enjoy your vegan banana bread!\n",
|
256 |
-
"\n",
|
257 |
-
"\n",
|
258 |
-
"\n",
|
259 |
-
"\n",
|
260 |
-
"\n",
|
261 |
-
"\n",
|
262 |
-
"\n",
|
263 |
-
"\n",
|
264 |
-
"\n",
|
265 |
-
"\n",
|
266 |
-
"\n",
|
267 |
-
"\n",
|
268 |
-
"\n",
|
269 |
-
"\n",
|
270 |
-
"\n",
|
271 |
-
"\n",
|
272 |
-
"\n",
|
273 |
-
"\n",
|
274 |
-
"\n",
|
275 |
-
"\n",
|
276 |
-
"\n",
|
277 |
-
"\n",
|
278 |
-
"\n",
|
279 |
-
"\n",
|
280 |
-
"\n",
|
281 |
-
"\n",
|
282 |
-
"\n",
|
283 |
-
"\n",
|
284 |
-
"\n",
|
285 |
-
"\n",
|
286 |
-
"\n",
|
287 |
-
"\n",
|
288 |
-
"\n",
|
289 |
-
"\n",
|
290 |
-
"\n",
|
291 |
-
"\n",
|
292 |
-
"\n",
|
293 |
-
"\n",
|
294 |
-
"\n",
|
295 |
-
"\n",
|
296 |
-
"\n",
|
297 |
-
"\n",
|
298 |
-
"\n",
|
299 |
-
"\n",
|
300 |
-
"\n",
|
301 |
-
"\n",
|
302 |
-
"\n",
|
303 |
-
"\n",
|
304 |
-
"\n",
|
305 |
-
"\n",
|
306 |
-
"\n",
|
307 |
-
"\n"
|
308 |
-
]
|
309 |
-
}
|
310 |
-
],
|
311 |
-
"source": [
|
312 |
-
"prompt = \"You are a helpful assistant. Tell me a recipe for vegan banana bread.\\n\"\n",
|
313 |
-
"\n",
|
314 |
-
"response = llama_generate(\n",
|
315 |
-
" model,\n",
|
316 |
-
" tokenizer,\n",
|
317 |
-
" prompt,\n",
|
318 |
-
" max_new_tokens=500,\n",
|
319 |
-
" temperature=0.92,\n",
|
320 |
-
")\n",
|
321 |
-
"\n",
|
322 |
-
"print(response)"
|
323 |
-
]
|
324 |
-
},
|
325 |
-
{
|
326 |
-
"cell_type": "code",
|
327 |
-
"execution_count": null,
|
328 |
-
"id": "3625b3ff-6467-43ea-8557-9541934539ec",
|
329 |
-
"metadata": {},
|
330 |
-
"outputs": [],
|
331 |
-
"source": []
|
332 |
-
}
|
333 |
-
],
|
334 |
-
"metadata": {
|
335 |
-
"kernelspec": {
|
336 |
-
"display_name": "Python 3",
|
337 |
-
"language": "python",
|
338 |
-
"name": "python3"
|
339 |
-
},
|
340 |
-
"language_info": {
|
341 |
-
"codemirror_mode": {
|
342 |
-
"name": "ipython",
|
343 |
-
"version": 3
|
344 |
-
},
|
345 |
-
"file_extension": ".py",
|
346 |
-
"mimetype": "text/x-python",
|
347 |
-
"name": "python",
|
348 |
-
"nbconvert_exporter": "python",
|
349 |
-
"pygments_lexer": "ipython3",
|
350 |
-
"version": "3.8.10"
|
351 |
-
}
|
352 |
-
},
|
353 |
-
"nbformat": 4,
|
354 |
-
"nbformat_minor": 5
|
355 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|