Skip to content

Commit c6f54b5

Browse files
author
Mirza Khan
committed
initial commit 2
0 parents  commit c6f54b5

21 files changed

+40522
-0
lines changed

‎fastai/00_notebook_tutorial.ipynb

Lines changed: 833 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson1-pets.ipynb

Lines changed: 1583 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson2-download.ipynb

Lines changed: 1508 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson2-sgd.ipynb

Lines changed: 18927 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson3-camvid-tiramisu.ipynb

Lines changed: 1311 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson3-camvid.ipynb

Lines changed: 1073 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson3-head-pose.ipynb

Lines changed: 464 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson3-imdb.ipynb

Lines changed: 1296 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson3-planet.ipynb

Lines changed: 1030 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson4-collab.ipynb

Lines changed: 1268 additions & 0 deletions
Large diffs are not rendered by default.

‎fastai/lesson4-tabular.ipynb

Lines changed: 332 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,332 @@
1+
{
2+
"cells": [
3+
{
4+
"cell_type": "markdown",
5+
"metadata": {},
6+
"source": [
7+
"# Tabular models"
8+
]
9+
},
10+
{
11+
"cell_type": "code",
12+
"execution_count": null,
13+
"metadata": {},
14+
"outputs": [],
15+
"source": [
16+
"from fastai.tabular import *"
17+
]
18+
},
19+
{
20+
"cell_type": "markdown",
21+
"metadata": {},
22+
"source": [
23+
"Tabular data should be in a Pandas `DataFrame`."
24+
]
25+
},
26+
{
27+
"cell_type": "code",
28+
"execution_count": null,
29+
"metadata": {},
30+
"outputs": [],
31+
"source": [
32+
"path = untar_data(URLs.ADULT_SAMPLE)\n",
33+
"df = pd.read_csv(path/'adult.csv')"
34+
]
35+
},
36+
{
37+
"cell_type": "code",
38+
"execution_count": null,
39+
"metadata": {},
40+
"outputs": [],
41+
"source": [
42+
"dep_var = 'salary'\n",
43+
"cat_names = ['workclass', 'education', 'marital-status', 'occupation', 'relationship', 'race']\n",
44+
"cont_names = ['age', 'fnlwgt', 'education-num']\n",
45+
"procs = [FillMissing, Categorify, Normalize]"
46+
]
47+
},
48+
{
49+
"cell_type": "code",
50+
"execution_count": null,
51+
"metadata": {},
52+
"outputs": [],
53+
"source": [
54+
"test = TabularList.from_df(df.iloc[800:1000].copy(), path=path, cat_names=cat_names, cont_names=cont_names)"
55+
]
56+
},
57+
{
58+
"cell_type": "code",
59+
"execution_count": null,
60+
"metadata": {},
61+
"outputs": [],
62+
"source": [
63+
"data = (TabularList.from_df(df, path=path, cat_names=cat_names, cont_names=cont_names, procs=procs)\n",
64+
" .split_by_idx(list(range(800,1000)))\n",
65+
" .label_from_df(cols=dep_var)\n",
66+
" .add_test(test)\n",
67+
" .databunch())"
68+
]
69+
},
70+
{
71+
"cell_type": "code",
72+
"execution_count": null,
73+
"metadata": {},
74+
"outputs": [
75+
{
76+
"data": {
77+
"text/html": [
78+
"<table> <col width='10%'> <col width='10%'> <col width='10%'> <col width='10%'> <col width='10%'> <col width='10%'> <col width='10%'> <col width='10%'> <col width='10%'> <col width='10%'> <col width='10%'> <tr>\n",
79+
" <th>workclass</th>\n",
80+
" <th>education</th>\n",
81+
" <th>marital-status</th>\n",
82+
" <th>occupation</th>\n",
83+
" <th>relationship</th>\n",
84+
" <th>race</th>\n",
85+
" <th>education-num_na</th>\n",
86+
" <th>age</th>\n",
87+
" <th>fnlwgt</th>\n",
88+
" <th>education-num</th>\n",
89+
" <th>target</th>\n",
90+
" </tr>\n",
91+
" <tr>\n",
92+
" <th> Private</th>\n",
93+
" <th> HS-grad</th>\n",
94+
" <th> Never-married</th>\n",
95+
" <th> Sales</th>\n",
96+
" <th> Not-in-family</th>\n",
97+
" <th> White</th>\n",
98+
" <th>False</th>\n",
99+
" <th>-1.2158</th>\n",
100+
" <th>1.1004</th>\n",
101+
" <th>-0.4224</th>\n",
102+
" <th><50k</th>\n",
103+
" </tr>\n",
104+
" <tr>\n",
105+
" <th> ?</th>\n",
106+
" <th> HS-grad</th>\n",
107+
" <th> Widowed</th>\n",
108+
" <th> ?</th>\n",
109+
" <th> Not-in-family</th>\n",
110+
" <th> White</th>\n",
111+
" <th>False</th>\n",
112+
" <th>1.8627</th>\n",
113+
" <th>0.0976</th>\n",
114+
" <th>-0.4224</th>\n",
115+
" <th><50k</th>\n",
116+
" </tr>\n",
117+
" <tr>\n",
118+
" <th> Self-emp-not-inc</th>\n",
119+
" <th> HS-grad</th>\n",
120+
" <th> Never-married</th>\n",
121+
" <th> Craft-repair</th>\n",
122+
" <th> Own-child</th>\n",
123+
" <th> Black</th>\n",
124+
" <th>False</th>\n",
125+
" <th>0.0303</th>\n",
126+
" <th>0.2092</th>\n",
127+
" <th>-0.4224</th>\n",
128+
" <th><50k</th>\n",
129+
" </tr>\n",
130+
" <tr>\n",
131+
" <th> Private</th>\n",
132+
" <th> HS-grad</th>\n",
133+
" <th> Married-civ-spouse</th>\n",
134+
" <th> Protective-serv</th>\n",
135+
" <th> Husband</th>\n",
136+
" <th> White</th>\n",
137+
" <th>False</th>\n",
138+
" <th>1.5695</th>\n",
139+
" <th>-0.5938</th>\n",
140+
" <th>-0.4224</th>\n",
141+
" <th><50k</th>\n",
142+
" </tr>\n",
143+
" <tr>\n",
144+
" <th> Private</th>\n",
145+
" <th> HS-grad</th>\n",
146+
" <th> Married-civ-spouse</th>\n",
147+
" <th> Handlers-cleaners</th>\n",
148+
" <th> Husband</th>\n",
149+
" <th> White</th>\n",
150+
" <th>False</th>\n",
151+
" <th>-0.9959</th>\n",
152+
" <th>-0.0318</th>\n",
153+
" <th>-0.4224</th>\n",
154+
" <th><50k</th>\n",
155+
" </tr>\n",
156+
" <tr>\n",
157+
" <th> Private</th>\n",
158+
" <th> 10th</th>\n",
159+
" <th> Married-civ-spouse</th>\n",
160+
" <th> Farming-fishing</th>\n",
161+
" <th> Wife</th>\n",
162+
" <th> White</th>\n",
163+
" <th>False</th>\n",
164+
" <th>-0.7027</th>\n",
165+
" <th>0.6071</th>\n",
166+
" <th>-1.5958</th>\n",
167+
" <th><50k</th>\n",
168+
" </tr>\n",
169+
" <tr>\n",
170+
" <th> Private</th>\n",
171+
" <th> HS-grad</th>\n",
172+
" <th> Married-civ-spouse</th>\n",
173+
" <th> Machine-op-inspct</th>\n",
174+
" <th> Husband</th>\n",
175+
" <th> White</th>\n",
176+
" <th>False</th>\n",
177+
" <th>0.1036</th>\n",
178+
" <th>-0.0968</th>\n",
179+
" <th>-0.4224</th>\n",
180+
" <th><50k</th>\n",
181+
" </tr>\n",
182+
" <tr>\n",
183+
" <th> Private</th>\n",
184+
" <th> Some-college</th>\n",
185+
" <th> Married-civ-spouse</th>\n",
186+
" <th> Exec-managerial</th>\n",
187+
" <th> Own-child</th>\n",
188+
" <th> White</th>\n",
189+
" <th>False</th>\n",
190+
" <th>-0.7760</th>\n",
191+
" <th>-0.6653</th>\n",
192+
" <th>-0.0312</th>\n",
193+
" <th>>=50k</th>\n",
194+
" </tr>\n",
195+
" <tr>\n",
196+
" <th> State-gov</th>\n",
197+
" <th> Some-college</th>\n",
198+
" <th> Never-married</th>\n",
199+
" <th> Tech-support</th>\n",
200+
" <th> Own-child</th>\n",
201+
" <th> White</th>\n",
202+
" <th>False</th>\n",
203+
" <th>-0.8493</th>\n",
204+
" <th>-1.4959</th>\n",
205+
" <th>-0.0312</th>\n",
206+
" <th><50k</th>\n",
207+
" </tr>\n",
208+
" <tr>\n",
209+
" <th> Private</th>\n",
210+
" <th> 11th</th>\n",
211+
" <th> Never-married</th>\n",
212+
" <th> Machine-op-inspct</th>\n",
213+
" <th> Not-in-family</th>\n",
214+
" <th> White</th>\n",
215+
" <th>False</th>\n",
216+
" <th>-1.0692</th>\n",
217+
" <th>-0.9516</th>\n",
218+
" <th>-1.2046</th>\n",
219+
" <th><50k</th>\n",
220+
" </tr>\n",
221+
"</table>\n"
222+
],
223+
"text/plain": [
224+
"<IPython.core.display.HTML object>"
225+
]
226+
},
227+
"metadata": {},
228+
"output_type": "display_data"
229+
}
230+
],
231+
"source": [
232+
"data.show_batch(rows=10)"
233+
]
234+
},
235+
{
236+
"cell_type": "code",
237+
"execution_count": null,
238+
"metadata": {},
239+
"outputs": [],
240+
"source": [
241+
"learn = tabular_learner(data, layers=[200,100], metrics=accuracy)"
242+
]
243+
},
244+
{
245+
"cell_type": "code",
246+
"execution_count": null,
247+
"metadata": {},
248+
"outputs": [
249+
{
250+
"data": {
251+
"text/html": [
252+
"Total time: 00:03 <p><table style='width:300px; margin-bottom:10px'>\n",
253+
" <tr>\n",
254+
" <th>epoch</th>\n",
255+
" <th>train_loss</th>\n",
256+
" <th>valid_loss</th>\n",
257+
" <th>accuracy</th>\n",
258+
" </tr>\n",
259+
" <tr>\n",
260+
" <th>1</th>\n",
261+
" <th>0.354604</th>\n",
262+
" <th>0.378520</th>\n",
263+
" <th>0.820000</th>\n",
264+
" </tr>\n",
265+
"</table>\n"
266+
],
267+
"text/plain": [
268+
"<IPython.core.display.HTML object>"
269+
]
270+
},
271+
"metadata": {},
272+
"output_type": "display_data"
273+
}
274+
],
275+
"source": [
276+
"learn.fit(1, 1e-2)"
277+
]
278+
},
279+
{
280+
"cell_type": "markdown",
281+
"metadata": {},
282+
"source": [
283+
"## Inference"
284+
]
285+
},
286+
{
287+
"cell_type": "code",
288+
"execution_count": null,
289+
"metadata": {},
290+
"outputs": [],
291+
"source": [
292+
"row = df.iloc[0]"
293+
]
294+
},
295+
{
296+
"cell_type": "code",
297+
"execution_count": null,
298+
"metadata": {},
299+
"outputs": [
300+
{
301+
"data": {
302+
"text/plain": [
303+
"(Category >=50k, tensor(1), tensor([0.4402, 0.5598]))"
304+
]
305+
},
306+
"execution_count": null,
307+
"metadata": {},
308+
"output_type": "execute_result"
309+
}
310+
],
311+
"source": [
312+
"learn.predict(row)"
313+
]
314+
},
315+
{
316+
"cell_type": "code",
317+
"execution_count": null,
318+
"metadata": {},
319+
"outputs": [],
320+
"source": []
321+
}
322+
],
323+
"metadata": {
324+
"kernelspec": {
325+
"display_name": "Python 3",
326+
"language": "python",
327+
"name": "python3"
328+
}
329+
},
330+
"nbformat": 4,
331+
"nbformat_minor": 2
332+
}

0 commit comments

Comments
 (0)