hanzceo commited on
Commit
2c4ebeb
Β·
verified Β·
1 Parent(s): 933beea

Upload train.ipynb

Browse files
Files changed (1) hide show
  1. train.ipynb +446 -0
train.ipynb ADDED
@@ -0,0 +1,446 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cells": [
3
+ {
4
+ "cell_type": "code",
5
+ "execution_count": 1,
6
+ "id": "6cc7d98d-23c5-4fb3-86eb-4e0665382bb1",
7
+ "metadata": {},
8
+ "outputs": [
9
+ {
10
+ "name": "stdout",
11
+ "output_type": "stream",
12
+ "text": [
13
+ "\u001b[2mInstalled \u001b[1mPython 3.13.8\u001b[0m \u001b[2min 696ms\u001b[0m\u001b[0m\n",
14
+ " \u001b[32m+\u001b[39m \u001b[1mcpython-3.13.8-linux-x86_64-gnu\u001b[0m (python3.13)\n",
15
+ "\u001b[1m\u001b[33mwarning\u001b[39m\u001b[0m\u001b[1m:\u001b[0m \u001b[1m`\u001b[36m/root/.local/bin\u001b[39m` is not on your PATH. To use installed Python executables, run `\u001b[32mexport PATH=\"/root/.local/bin:$PATH\"\u001b[39m` or `\u001b[32muv python update-shell\u001b[39m`.\u001b[0m\n",
16
+ "Pinned `\u001b[36m.python-version\u001b[39m` to `\u001b[32m3.13\u001b[39m`\n"
17
+ ]
18
+ }
19
+ ],
20
+ "source": [
21
+ "!uv python install 313\n",
22
+ "!uv python pin 313"
23
+ ]
24
+ },
25
+ {
26
+ "cell_type": "code",
27
+ "execution_count": 2,
28
+ "id": "2d741065-01e2-4551-8a7e-70fecca6759e",
29
+ "metadata": {},
30
+ "outputs": [
31
+ {
32
+ "name": "stdout",
33
+ "output_type": "stream",
34
+ "text": [
35
+ "Using CPython \u001b[36m3.13.8\u001b[39m\n",
36
+ "Creating virtual environment at: \u001b[36m.venv\u001b[39m\n",
37
+ "Activate with: \u001b[32msource .venv/bin/activate\u001b[39m\n",
38
+ "\u001b[2K\u001b[2mResolved \u001b[1m30 packages\u001b[0m \u001b[2min 63ms\u001b[0m\u001b[0m \u001b[0m\n",
39
+ "\u001b[2K\u001b[2mPrepared \u001b[1m30 packages\u001b[0m \u001b[2min 703ms\u001b[0m\u001b[0m \n",
40
+ "\u001b[2K\u001b[2mInstalled \u001b[1m30 packages\u001b[0m \u001b[2min 110ms\u001b[0m\u001b[0m \u001b[0m\n",
41
+ " \u001b[32m+\u001b[39m \u001b[1masttokens\u001b[0m\u001b[2m==3.0.2\u001b[0m\n",
42
+ " \u001b[32m+\u001b[39m \u001b[1mcomm\u001b[0m\u001b[2m==0.2.3\u001b[0m\n",
43
+ " \u001b[32m+\u001b[39m \u001b[1mdebugpy\u001b[0m\u001b[2m==1.8.21\u001b[0m\n",
44
+ " \u001b[32m+\u001b[39m \u001b[1mdecorator\u001b[0m\u001b[2m==5.3.1\u001b[0m\n",
45
+ " \u001b[32m+\u001b[39m \u001b[1mexecuting\u001b[0m\u001b[2m==2.2.1\u001b[0m\n",
46
+ " \u001b[32m+\u001b[39m \u001b[1mipykernel\u001b[0m\u001b[2m==7.3.0\u001b[0m\n",
47
+ " \u001b[32m+\u001b[39m \u001b[1mipython\u001b[0m\u001b[2m==9.15.0\u001b[0m\n",
48
+ " \u001b[32m+\u001b[39m \u001b[1mipython-pygments-lexers\u001b[0m\u001b[2m==1.1.1\u001b[0m\n",
49
+ " \u001b[32m+\u001b[39m \u001b[1mjedi\u001b[0m\u001b[2m==0.20.0\u001b[0m\n",
50
+ " \u001b[32m+\u001b[39m \u001b[1mjupyter-client\u001b[0m\u001b[2m==8.9.1\u001b[0m\n",
51
+ " \u001b[32m+\u001b[39m \u001b[1mjupyter-core\u001b[0m\u001b[2m==5.9.1\u001b[0m\n",
52
+ " \u001b[32m+\u001b[39m \u001b[1mmatplotlib-inline\u001b[0m\u001b[2m==0.2.2\u001b[0m\n",
53
+ " \u001b[32m+\u001b[39m \u001b[1mnest-asyncio2\u001b[0m\u001b[2m==1.7.2\u001b[0m\n",
54
+ " \u001b[32m+\u001b[39m \u001b[1mpackaging\u001b[0m\u001b[2m==26.2\u001b[0m\n",
55
+ " \u001b[32m+\u001b[39m \u001b[1mparso\u001b[0m\u001b[2m==0.8.7\u001b[0m\n",
56
+ " \u001b[32m+\u001b[39m \u001b[1mpexpect\u001b[0m\u001b[2m==4.9.0\u001b[0m\n",
57
+ " \u001b[32m+\u001b[39m \u001b[1mplatformdirs\u001b[0m\u001b[2m==4.11.0\u001b[0m\n",
58
+ " \u001b[32m+\u001b[39m \u001b[1mprompt-toolkit\u001b[0m\u001b[2m==3.0.53\u001b[0m\n",
59
+ " \u001b[32m+\u001b[39m \u001b[1mpsutil\u001b[0m\u001b[2m==7.2.2\u001b[0m\n",
60
+ " \u001b[32m+\u001b[39m \u001b[1mptyprocess\u001b[0m\u001b[2m==0.7.0\u001b[0m\n",
61
+ " \u001b[32m+\u001b[39m \u001b[1mpure-eval\u001b[0m\u001b[2m==0.2.3\u001b[0m\n",
62
+ " \u001b[32m+\u001b[39m \u001b[1mpygments\u001b[0m\u001b[2m==2.20.0\u001b[0m\n",
63
+ " \u001b[32m+\u001b[39m \u001b[1mpython-dateutil\u001b[0m\u001b[2m==2.9.0.post0\u001b[0m\n",
64
+ " \u001b[32m+\u001b[39m \u001b[1mpyzmq\u001b[0m\u001b[2m==27.1.0\u001b[0m\n",
65
+ " \u001b[32m+\u001b[39m \u001b[1msix\u001b[0m\u001b[2m==1.17.0\u001b[0m\n",
66
+ " \u001b[32m+\u001b[39m \u001b[1mstack-data\u001b[0m\u001b[2m==0.6.3\u001b[0m\n",
67
+ " \u001b[32m+\u001b[39m \u001b[1mtornado\u001b[0m\u001b[2m==6.5.7\u001b[0m\n",
68
+ " \u001b[32m+\u001b[39m \u001b[1mtraitlets\u001b[0m\u001b[2m==5.15.1\u001b[0m\n",
69
+ " \u001b[32m+\u001b[39m \u001b[1mtyping-extensions\u001b[0m\u001b[2m==4.16.0\u001b[0m\n",
70
+ " \u001b[32m+\u001b[39m \u001b[1mwcwidth\u001b[0m\u001b[2m==0.8.2\u001b[0m\n",
71
+ "Installed kernelspec uv-kernel in /root/.local/share/jupyter/kernels/uv-kernel\n"
72
+ ]
73
+ }
74
+ ],
75
+ "source": [
76
+ "!uv venv --clear\n",
77
+ "!uv pip install ipykernel\n",
78
+ "!uv run python -m ipykernel install --user --name=uv-kernel --display-name \"Python (uv venv)\""
79
+ ]
80
+ },
81
+ {
82
+ "cell_type": "code",
83
+ "execution_count": 4,
84
+ "id": "ae35bf68-b4b1-435e-8a79-61bd140bc77e",
85
+ "metadata": {},
86
+ "outputs": [
87
+ {
88
+ "name": "stdout",
89
+ "output_type": "stream",
90
+ "text": [
91
+ "\u001b[2mAudited \u001b[1m10 packages\u001b[0m \u001b[2min 10ms\u001b[0m\u001b[0m\n"
92
+ ]
93
+ }
94
+ ],
95
+ "source": [
96
+ "!uv pip install transformers huggingface_hub datasets fastrand torch sentencepiece protobuf evaluate scikit-learn accelerate>1.1.0"
97
+ ]
98
+ },
99
+ {
100
+ "cell_type": "code",
101
+ "execution_count": 1,
102
+ "id": "345615b8-726e-4c9a-ab56-deb322632c82",
103
+ "metadata": {},
104
+ "outputs": [
105
+ {
106
+ "name": "stderr",
107
+ "output_type": "stream",
108
+ "text": [
109
+ "/workspace/.venv/lib/python3.13/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
110
+ " from .autonotebook import tqdm as notebook_tqdm\n"
111
+ ]
112
+ }
113
+ ],
114
+ "source": [
115
+ "import numpy as np\n",
116
+ "import torch\n",
117
+ "from datasets import load_dataset\n",
118
+ "from transformers import (\n",
119
+ " AutoTokenizer, \n",
120
+ " AutoModelForSequenceClassification, \n",
121
+ " TrainingArguments, \n",
122
+ " Trainer,\n",
123
+ " DataCollatorWithPadding\n",
124
+ ")\n",
125
+ "from huggingface_hub import notebook_login\n",
126
+ "import evaluate"
127
+ ]
128
+ },
129
+ {
130
+ "cell_type": "code",
131
+ "execution_count": 2,
132
+ "id": "033f1b47-2771-48ff-b89f-fcc3d93ab47d",
133
+ "metadata": {},
134
+ "outputs": [],
135
+ "source": [
136
+ "notebook_login()"
137
+ ]
138
+ },
139
+ {
140
+ "cell_type": "code",
141
+ "execution_count": 3,
142
+ "id": "f0ff7440-f91c-4705-9d32-d37ac1089239",
143
+ "metadata": {},
144
+ "outputs": [
145
+ {
146
+ "name": "stderr",
147
+ "output_type": "stream",
148
+ "text": [
149
+ "[transformers] `torch_dtype` is deprecated! Use `dtype` instead!\n",
150
+ "Loading weights: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 172/172 [00:00<00:00, 1366.64it/s]\n",
151
+ "[transformers] \u001b[1mModernBertForSequenceClassification LOAD REPORT\u001b[0m from: jhu-clsp/ettin-encoder-1b\n",
152
+ "Key | Status | \n",
153
+ "------------------+------------+-\n",
154
+ "decoder.bias | UNEXPECTED | \n",
155
+ "decoder.weight | UNEXPECTED | \n",
156
+ "classifier.weight | MISSING | \n",
157
+ "classifier.bias | MISSING | \n",
158
+ "\n",
159
+ "Notes:\n",
160
+ "- UNEXPECTED:\tcan be ignored when loading from different task/architecture; not ok if you expect identical arch.\n",
161
+ "- MISSING:\tthose params were newly initialized because missing from the checkpoint. Consider training on your downstream task.\n"
162
+ ]
163
+ }
164
+ ],
165
+ "source": [
166
+ "# 1. Load tokenizer and model for classification\n",
167
+ "model_name = \"jhu-clsp/ettin-encoder-1b\"\n",
168
+ "tokenizer = AutoTokenizer.from_pretrained(model_name)\n",
169
+ "model = AutoModelForSequenceClassification.from_pretrained(\n",
170
+ " model_name, num_labels=2,\n",
171
+ " torch_dtype=torch.bfloat16 if torch.cuda.is_bf16_supported() else torch.float16\n",
172
+ ") # Adjust num_labels"
173
+ ]
174
+ },
175
+ {
176
+ "cell_type": "code",
177
+ "execution_count": 4,
178
+ "id": "46b5f750-1b36-4437-b7cb-26331d92c595",
179
+ "metadata": {},
180
+ "outputs": [],
181
+ "source": [
182
+ "# 2. Load and tokenize dataset (example using IMDb for binary classification)\n",
183
+ "dataset = load_dataset(\"hanzceo/JOSS-data-L2\")[\"train\"]"
184
+ ]
185
+ },
186
+ {
187
+ "cell_type": "code",
188
+ "execution_count": 5,
189
+ "id": "37f80b16-499f-4002-93c5-a534c4f7b289",
190
+ "metadata": {},
191
+ "outputs": [],
192
+ "source": [
193
+ "dataset = dataset.train_test_split(test_size=0.2)"
194
+ ]
195
+ },
196
+ {
197
+ "cell_type": "code",
198
+ "execution_count": 6,
199
+ "id": "4a8f8b10-e197-4d58-a818-2f5e8ba77d0e",
200
+ "metadata": {},
201
+ "outputs": [
202
+ {
203
+ "data": {
204
+ "text/plain": [
205
+ "DatasetDict({\n",
206
+ " train: Dataset({\n",
207
+ " features: ['sentence1', 'score'],\n",
208
+ " num_rows: 1600\n",
209
+ " })\n",
210
+ " test: Dataset({\n",
211
+ " features: ['sentence1', 'score'],\n",
212
+ " num_rows: 400\n",
213
+ " })\n",
214
+ "})"
215
+ ]
216
+ },
217
+ "execution_count": 6,
218
+ "metadata": {},
219
+ "output_type": "execute_result"
220
+ }
221
+ ],
222
+ "source": [
223
+ "dataset"
224
+ ]
225
+ },
226
+ {
227
+ "cell_type": "code",
228
+ "execution_count": 7,
229
+ "id": "fa7253f7-05a7-40a4-9727-d107adac325e",
230
+ "metadata": {},
231
+ "outputs": [
232
+ {
233
+ "name": "stderr",
234
+ "output_type": "stream",
235
+ "text": [
236
+ "Map: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1600/1600 [00:01<00:00, 1399.51 examples/s]\n",
237
+ "Map: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 400/400 [00:00<00:00, 1632.19 examples/s]\n"
238
+ ]
239
+ }
240
+ ],
241
+ "source": [
242
+ "def tokenize_function(examples):\n",
243
+ " return tokenizer(examples[\"sentence1\"], truncation=True)\n",
244
+ "tokenized_datasets = dataset.map(tokenize_function, batched=True)\n",
245
+ "data_collator = DataCollatorWithPadding(tokenizer=tokenizer)"
246
+ ]
247
+ },
248
+ {
249
+ "cell_type": "code",
250
+ "execution_count": 8,
251
+ "id": "49744185-47ff-460a-9a1b-90f33f2bc991",
252
+ "metadata": {},
253
+ "outputs": [
254
+ {
255
+ "data": {
256
+ "text/plain": [
257
+ "DatasetDict({\n",
258
+ " train: Dataset({\n",
259
+ " features: ['sentence1', 'labels', 'input_ids', 'attention_mask'],\n",
260
+ " num_rows: 1600\n",
261
+ " })\n",
262
+ " test: Dataset({\n",
263
+ " features: ['sentence1', 'labels', 'input_ids', 'attention_mask'],\n",
264
+ " num_rows: 400\n",
265
+ " })\n",
266
+ "})"
267
+ ]
268
+ },
269
+ "execution_count": 8,
270
+ "metadata": {},
271
+ "output_type": "execute_result"
272
+ }
273
+ ],
274
+ "source": [
275
+ "tokenized_datasets = tokenized_datasets.rename_column(\"score\", \"labels\")\n",
276
+ "\n",
277
+ "tokenized_datasets"
278
+ ]
279
+ },
280
+ {
281
+ "cell_type": "code",
282
+ "execution_count": 9,
283
+ "id": "93531513-4e7f-4d16-8292-8cecb44e0aa3",
284
+ "metadata": {},
285
+ "outputs": [],
286
+ "source": [
287
+ "# 3. Setup evaluation metrics\n",
288
+ "metric = evaluate.load(\"accuracy\")\n",
289
+ "def compute_metrics(eval_pred):\n",
290
+ " logits, labels = eval_pred\n",
291
+ " predictions = np.argmax(logits, axis=-1)\n",
292
+ " return metric.compute(predictions=predictions, references=labels)"
293
+ ]
294
+ },
295
+ {
296
+ "cell_type": "code",
297
+ "execution_count": 10,
298
+ "id": "b663bbee-15cb-44a9-9393-a78cf051c0fc",
299
+ "metadata": {},
300
+ "outputs": [],
301
+ "source": [
302
+ "# 4. Define Training Arguments\n",
303
+ "training_args = TrainingArguments(\n",
304
+ " output_dir=\"models/ettin-classification\",\n",
305
+ " eval_strategy=\"epoch\",\n",
306
+ " learning_rate=2e-5,\n",
307
+ " per_device_train_batch_size=2,\n",
308
+ " per_device_eval_batch_size=2,\n",
309
+ " num_train_epochs=3,\n",
310
+ " weight_decay=0.01,\n",
311
+ " bf16=True\n",
312
+ ")"
313
+ ]
314
+ },
315
+ {
316
+ "cell_type": "code",
317
+ "execution_count": null,
318
+ "id": "685b3f75-3a68-4965-bf28-ba3e70b97618",
319
+ "metadata": {},
320
+ "outputs": [
321
+ {
322
+ "data": {
323
+ "text/html": [
324
+ "\n",
325
+ " <div>\n",
326
+ " \n",
327
+ " <progress value='2401' max='2400' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
328
+ " [2400/2400 17:50, Epoch 3/3]\n",
329
+ " </div>\n",
330
+ " <table border=\"1\" class=\"dataframe\">\n",
331
+ " <thead>\n",
332
+ " <tr style=\"text-align: left;\">\n",
333
+ " <th>Epoch</th>\n",
334
+ " <th>Training Loss</th>\n",
335
+ " <th>Validation Loss</th>\n",
336
+ " <th>Accuracy</th>\n",
337
+ " </tr>\n",
338
+ " </thead>\n",
339
+ " <tbody>\n",
340
+ " <tr>\n",
341
+ " <td>1</td>\n",
342
+ " <td>0.423433</td>\n",
343
+ " <td>0.226460</td>\n",
344
+ " <td>0.947500</td>\n",
345
+ " </tr>\n",
346
+ " <tr>\n",
347
+ " <td>2</td>\n",
348
+ " <td>0.217008</td>\n",
349
+ " <td>0.211006</td>\n",
350
+ " <td>0.947500</td>\n",
351
+ " </tr>\n",
352
+ " </tbody>\n",
353
+ "</table><p>\n",
354
+ " <div>\n",
355
+ " \n",
356
+ " <progress value='119' max='200' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
357
+ " [119/200 00:14 < 00:09, 8.22 it/s]\n",
358
+ " </div>\n",
359
+ " "
360
+ ],
361
+ "text/plain": [
362
+ "<IPython.core.display.HTML object>"
363
+ ]
364
+ },
365
+ "metadata": {},
366
+ "output_type": "display_data"
367
+ },
368
+ {
369
+ "name": "stderr",
370
+ "output_type": "stream",
371
+ "text": [
372
+ "Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 1.30it/s]\n",
373
+ "Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 1.85it/s]\n",
374
+ "Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 1.28it/s]\n",
375
+ "Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 1.39it/s]\n",
376
+ "Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 1.28it/s]\n"
377
+ ]
378
+ }
379
+ ],
380
+ "source": [
381
+ "# 5. Train\n",
382
+ "trainer = Trainer(\n",
383
+ " model=model,\n",
384
+ " args=training_args,\n",
385
+ " data_collator=data_collator,\n",
386
+ " train_dataset=tokenized_datasets[\"train\"].shuffle(seed=42),\n",
387
+ " eval_dataset=tokenized_datasets[\"test\"].shuffle(seed=42),\n",
388
+ " compute_metrics=compute_metrics,\n",
389
+ ")\n",
390
+ "\n",
391
+ "trainer.train()"
392
+ ]
393
+ },
394
+ {
395
+ "cell_type": "code",
396
+ "execution_count": null,
397
+ "id": "250248f5-021e-4247-8b2d-1d1098cf8a31",
398
+ "metadata": {},
399
+ "outputs": [],
400
+ "source": [
401
+ "model"
402
+ ]
403
+ },
404
+ {
405
+ "cell_type": "code",
406
+ "execution_count": null,
407
+ "id": "91374ca9-04ca-4ab7-a136-a5b329071426",
408
+ "metadata": {},
409
+ "outputs": [],
410
+ "source": [
411
+ "model.push_to_hub(\"hanzceo/JOSS-dev1\")"
412
+ ]
413
+ },
414
+ {
415
+ "cell_type": "code",
416
+ "execution_count": null,
417
+ "id": "ed10d1ca-ada0-43db-ac78-357a865d2b4a",
418
+ "metadata": {},
419
+ "outputs": [],
420
+ "source": [
421
+ "tokenizer.push_to_hub(\"hanzceo/JOSS-dev1\")"
422
+ ]
423
+ }
424
+ ],
425
+ "metadata": {
426
+ "kernelspec": {
427
+ "display_name": "Python (uv venv)",
428
+ "language": "python",
429
+ "name": "uv-kernel"
430
+ },
431
+ "language_info": {
432
+ "codemirror_mode": {
433
+ "name": "ipython",
434
+ "version": 3
435
+ },
436
+ "file_extension": ".py",
437
+ "mimetype": "text/x-python",
438
+ "name": "python",
439
+ "nbconvert_exporter": "python",
440
+ "pygments_lexer": "ipython3",
441
+ "version": "3.13.8"
442
+ }
443
+ },
444
+ "nbformat": 4,
445
+ "nbformat_minor": 5
446
+ }