import requests, zipfile, io
import os
def download_files(dir_name):
if not os.path.exists(dir_name):
# 28 training examples from the SQA training set + table csv data
urls = ["https://www.dropbox.com/s/2p6ez9xro357i63/sqa_train_set_28_examples.zip?dl=1",
"https://www.dropbox.com/s/abhum8ssuow87h6/table_csv.zip?dl=1"
]
for url in urls:
r = requests.get(url)
z = zipfile.ZipFile(io.BytesIO(r.content))
z.extractall()
dir_name = "sqa_data"
download_files(dir_name)
import pandas as pd
data = pd.read_excel("sqa_train_set_28_examples.xlsx")
data.head()
id
annotator
position
question
table_file
answer_coordinates
answer_text
0
nt-639
0
0
where are the players from?
table_csv/203_149.csv
['(0, 4)', '(1, 4)', '(2, 4)', '(3, 4)', '(4, ...
['Louisiana State University', 'Valley HS (Las...
1
nt-639
0
1
which player went to louisiana state university?
table_csv/203_149.csv
['(0, 1)']
['Ben McDonald']
2
nt-639
1
0
who are the players?
table_csv/203_149.csv
['(0, 1)', '(1, 1)', '(2, 1)', '(3, 1)', '(4, ...
['Ben McDonald', 'Tyler Houston', 'Roger Salke...
3
nt-639
1
1
which ones are in the top 26 picks?
table_csv/203_149.csv
['(0, 1)', '(1, 1)', '(2, 1)', '(3, 1)', '(4, ...
['Ben McDonald', 'Tyler Houston', 'Roger Salke...
4
nt-639
1
2
and of those, who is from louisiana state univ...
table_csv/203_149.csv
['(0, 1)']
['Ben McDonald']
import ast
def _parse_answer_coordinates(answer_coordinate_str):
"""Parses the answer_coordinates of a question.
Args:
answer_coordinate_str: A string representation of a Python list of tuple
strings.
For example: "['(1, 4)','(1, 3)', ...]"
"""
try:
answer_coordinates = []
# make a list of strings
coords = ast.literal_eval(answer_coordinate_str)
# parse each string as a tuple
for row_index, column_index in sorted(
ast.literal_eval(coord) for coord in coords):
answer_coordinates.append((row_index, column_index))
except SyntaxError:
raise ValueError('Unable to evaluate %s' % answer_coordinate_str)
return answer_coordinates
def _parse_answer_text(answer_text):
"""Populates the answer_texts field of `answer` by parsing `answer_text`.
Args:
answer_text: A string representation of a Python list of strings.
For example: "[u'test', u'hello', ...]"
answer: an Answer object.
"""
try:
answer = []
for value in ast.literal_eval(answer_text):
answer.append(value)
except SyntaxError:
raise ValueError('Unable to evaluate %s' % answer_text)
return answer
data['answer_coordinates'] = data['answer_coordinates'].apply(lambda coords_str: _parse_answer_coordinates(coords_str))
data['answer_text'] = data['answer_text'].apply(lambda txt: _parse_answer_text(txt))
data.head(10)
id
annotator
position
question
table_file
answer_coordinates
answer_text
0
nt-639
0
0
where are the players from?
table_csv/203_149.csv
[(0, 4), (1, 4), (2, 4), (3, 4), (4, 4), (5, 4...
[Louisiana State University, Valley HS (Las Ve...
1
nt-639
0
1
which player went to louisiana state university?
table_csv/203_149.csv
[(0, 1)]
[Ben McDonald]
2
nt-639
1
0
who are the players?
table_csv/203_149.csv
[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, 1...
[Ben McDonald, Tyler Houston, Roger Salkeld, J...
3
nt-639
1
1
which ones are in the top 26 picks?
table_csv/203_149.csv
[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, 1...
[Ben McDonald, Tyler Houston, Roger Salkeld, J...
4
nt-639
1
2
and of those, who is from louisiana state univ...
table_csv/203_149.csv
[(0, 1)]
[Ben McDonald]
5
nt-639
2
0
who are the players in the top 26?
table_csv/203_149.csv
[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, 1...
[Ben McDonald, Tyler Houston, Roger Salkeld, J...
6
nt-639
2
1
of those, which one was from louisiana state u...
table_csv/203_149.csv
[(0, 1)]
[Ben McDonald]
7
nt-11649
0
0
what are all the names of the teams?
table_csv/204_135.csv
[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, 1...
[Cordoba CF, CD Malaga, Granada CF, UD Las Pal...
8
nt-11649
0
1
of these, which teams had any losses?
table_csv/204_135.csv
[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, 1...
[Cordoba CF, CD Malaga, Granada CF, UD Las Pal...
9
nt-11649
0
2
of these teams, which had more than 21 losses?
table_csv/204_135.csv
[(15, 1)]
[CD Villarrobledo]
def get_sequence_id(example_id, annotator):
if "-" in str(annotator):
raise ValueError('"-" not allowed in annotator.')
return f"{example_id}-{annotator}"
data['sequence_id'] = data.apply(lambda x: get_sequence_id(x.id, x.annotator), axis=1)
data.head()
id
annotator
position
question
table_file
answer_coordinates
answer_text
sequence_id
0
nt-639
0
0
where are the players from?
table_csv/203_149.csv
[(0, 4), (1, 4), (2, 4), (3, 4), (4, 4), (5, 4...
[Louisiana State University, Valley HS (Las Ve...
nt-639-0
1
nt-639
0
1
which player went to louisiana state university?
table_csv/203_149.csv
[(0, 1)]
[Ben McDonald]
nt-639-0
2
nt-639
1
0
who are the players?
table_csv/203_149.csv
[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, 1...
[Ben McDonald, Tyler Houston, Roger Salkeld, J...
nt-639-1
3
nt-639
1
1
which ones are in the top 26 picks?
table_csv/203_149.csv
[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, 1...
[Ben McDonald, Tyler Houston, Roger Salkeld, J...
nt-639-1
4
nt-639
1
2
and of those, who is from louisiana state univ...
table_csv/203_149.csv
[(0, 1)]
[Ben McDonald]
nt-639-1
# let's group table-question pairs by sequence id, and remove some columns we don't need
grouped = data.groupby(by='sequence_id').agg(lambda x: x.tolist())
grouped = grouped.drop(columns=['id', 'annotator', 'position'])
grouped['table_file'] = grouped['table_file'].apply(lambda x: x[0])
grouped.head(10)
question
table_file
answer_coordinates
answer_text
sequence_id
ns-1292-0
[who are all the athletes?, where are they fro...
table_csv/204_521.csv
[[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, ...
[[Tommy Green, Janis Dalins, Ugo Frigerio, Kar...
nt-10730-0
[what was the production numbers of each revol...
table_csv/203_253.csv
[[(0, 4), (1, 4), (2, 4), (3, 4), (4, 4), (5, ...
[[1,900 (estimated), 14,500 (estimated), 6,000...
nt-10730-1
[what three revolver models had the least amou...
table_csv/203_253.csv
[[(0, 0), (6, 0), (7, 0)], [(0, 0)]]
[[Remington-Beals Army Model Revolver, New Mod...
nt-10730-2
[what are all of the remington models?, how ma...
table_csv/203_253.csv
[[(0, 0), (1, 0), (2, 0), (3, 0), (4, 0), (5, ...
[[Remington-Beals Army Model Revolver, Remingt...
nt-11649-0
[what are all the names of the teams?, of thes...
table_csv/204_135.csv
[[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, ...
[[Cordoba CF, CD Malaga, Granada CF, UD Las Pa...
nt-11649-1
[what are the losses?, what team had more than...
table_csv/204_135.csv
[[(0, 6), (1, 6), (2, 6), (3, 6), (4, 6), (5, ...
[[6, 6, 9, 10, 10, 12, 12, 11, 13, 14, 15, 14,...
nt-11649-2
[what were all the teams?, what were the loss ...
table_csv/204_135.csv
[[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, ...
[[Cordoba CF, CD Malaga, Granada CF, UD Las Pa...
nt-639-0
[where are the players from?, which player wen...
table_csv/203_149.csv
[[(0, 4), (1, 4), (2, 4), (3, 4), (4, 4), (5, ...
[[Louisiana State University, Valley HS (Las V...
nt-639-1
[who are the players?, which ones are in the t...
table_csv/203_149.csv
[[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, ...
[[Ben McDonald, Tyler Houston, Roger Salkeld, ...
nt-639-2
[who are the players in the top 26?, of those,...
table_csv/203_149.csv
[[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, ...
[[Ben McDonald, Tyler Houston, Roger Salkeld, ...
# path to the directory containing all csv files
table_csv_path = "table_csv"
item = grouped.iloc[0]
table = pd.read_csv(table_csv_path + item.table_file[9:]).astype(str)
display(table)
print("")
print(item.question)
Rank
Name
Nationality
Time (hand)
Notes
0
nan
Tommy Green
Great Britain
4:50:10
OR
1
nan
Janis Dalins
Latvia
4:57:20
nan
2
nan
Ugo Frigerio
Italy
4:59:06
nan
3
4.0
Karl Hahnel
Germany
5:06:06
nan
4
5.0
Ettore Rivolta
Italy
5:07:39
nan
5
6.0
Paul Sievert
Germany
5:16:41
nan
6
7.0
Henri Quintric
France
5:27:25
nan
7
8.0
Ernie Crosbie
United States
5:28:02
nan
8
9.0
Bill Chisholm
United States
5:51:00
nan
9
10.0
Alfred Maasik
Estonia
6:19:00
nan
10
nan
Henry Cieman
Canada
nan
DNF
11
nan
John Moralis
Greece
nan
DNF
12
nan
Francesco Pretti
Italy
nan
DNF
13
nan
Arthur Tell Schwab
Switzerland
nan
DNF
14
nan
Harry Hinkel
United States
nan
DNF
['who are all the athletes?', 'where are they from?', 'along with paul sievert, which athlete is from germany?']
import tensorflow as tf
from transformers import TapasTokenizer
# initialize the tokenizer
tokenizer = TapasTokenizer.from_pretrained("google/tapas-base")
class TableDataset:
def __init__(self, df, tokenizer):
self.df = df
self.tokenizer = tokenizer
def __iter__(self):
for idx in range(self.__len__()):
item = self.df.iloc[idx]
table = pd.read_csv(table_csv_path + item.table_file[9:]).astype(str) # TapasTokenizer expects the table data to be text only
if item.position != 0:
# use the previous table-question pair to correctly set the prev_labels token type ids
previous_item = self.df.iloc[idx-1]
encoding = self.tokenizer(table=table,
queries=[previous_item.question, item.question],
answer_coordinates=[previous_item.answer_coordinates, item.answer_coordinates],
answer_text=[previous_item.answer_text, item.answer_text],
padding="max_length",
truncation=True,
return_tensors="tf"
)
# use encodings of second table-question pair in the batch
encoding = {key: val[-1] for key, val in encoding.items()}
else:
# this means it's the first table-question pair in a sequence
encoding = self.tokenizer(table=table,
queries=item.question,
answer_coordinates=item.answer_coordinates,
answer_text=item.answer_text,
padding="max_length",
truncation=True,
return_tensors="tf"
)
# remove the batch dimension which the tokenizer adds
encoding = {key: tf.squeeze(val,0) for key, val in encoding.items()}
yield encoding['input_ids'],encoding['attention_mask'],encoding['numeric_values'], \
encoding['numeric_values_scale'], encoding['token_type_ids'], encoding['labels']
def __len__(self):
return len(self.df)
__call__ = __iter__
train_dataset = TableDataset(df=data, tokenizer=tokenizer)
output_signature = (
tf.TensorSpec(shape=(512,), dtype=tf.int32),
tf.TensorSpec(shape=(512,), dtype=tf.int32),
tf.TensorSpec(shape=(512,), dtype=tf.float32),
tf.TensorSpec(shape=(512,), dtype=tf.float32),
tf.TensorSpec(shape=(512,7), dtype=tf.int32),
tf.TensorSpec(shape=(512,), dtype=tf.int32) )
train_dataloader = tf.data.Dataset.from_generator(train_dataset,output_signature=output_signature).batch(2)
batch = next(iter(train_dataloader))
batch[0].shape
TensorShape([2, 512])
batch[4].shape
TensorShape([2, 512, 7])
batch[2].shape
TensorShape([2, 512])
tokenizer.decode(batch[0][1])
'[CLS] which player went to louisiana state university? [SEP] pick player team position school 1 ben mcdonald baltimore orioles rhp louisiana state university 2 tyler houston atlanta braves c valley hs ( las vegas, nv ) 3 roger salkeld seattle mariners rhp saugus ( ca ) hs 4 jeff jackson philadelphia phillies of simeon hs ( chicago, il ) 5 donald harris texas rangers of texas tech university 6 paul coleman saint louis cardinals of frankston ( tx ) hs 7 frank thomas chicago white sox 1b auburn university 8 earl cunningham chicago cubs of lancaster ( sc ) hs 9 kyle abbott california angels lhp long beach state university 10 charles johnson montreal expos c westwood hs ( fort pierce, fl ) 11 calvin murray cleveland indians 3b w. t. white high school ( dallas, tx ) 12 jeff juden houston astros rhp salem ( ma ) hs 13 brent mayne kansas city royals c cal state fullerton 14 steve hosey san francisco giants of fresno state university 15 kiki jones los angeles dodgers rhp hillsborough hs ( tampa, fl ) 16 greg blosser boston red sox of sarasota ( fl ) hs 17 cal eldred milwaukee brewers rhp university of iowa 18 willie greene pittsburgh pirates ss jones county hs ( gray, ga ) 19 eddie zosky toronto blue jays ss fresno state university 20 scott bryant cincinnati reds of university of texas 21 greg gohr detroit tigers rhp santa clara university 22 tom goodwin los angeles dodgers of fresno state university 23 mo vaughn boston red sox 1b seton hall university 24 alan zinter new york mets c university of arizona 25 chuck knoblauch minnesota twins 2b texas a & m university 26 scott burrell seattle mariners rhp hamden ( ct ) hs [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD] [PAD]'
assert sum(batch[4][0][:,3]) == 0
print(sum(batch[4][1][:,3]))
tf.Tensor(132, shape=(), dtype=int32)
for id, prev_label in zip(batch[0][1], batch[4][1][:,3]):
if id != 0:
print(tokenizer.decode([id]), prev_label.numpy().item())
[CLS] 0
which 0
player 0
went 0
to 0
louisiana 0
state 0
university 0
? 0
[SEP] 0
pick 0
player 0
team 0
position 0
school 0
1 0
ben 0
mcdonald 0
baltimore 0
orioles 0
r 0
##hp 0
louisiana 1
state 1
university 1
2 0
tyler 0
houston 0
atlanta 0
braves 0
c 0
valley 1
hs 1
( 1
las 1
vegas 1
, 1
n 1
##v 1
) 1
3 0
roger 0
sal 0
##kel 0
##d 0
seattle 0
mariners 0
r 0
##hp 0
sa 1
##ug 1
##us 1
( 1
ca 1
) 1
hs 1
4 0
jeff 0
jackson 0
philadelphia 0
phillies 0
of 0
simeon 1
hs 1
( 1
chicago 1
, 1
il 1
) 1
5 0
donald 0
harris 0
texas 0
rangers 0
of 0
texas 1
tech 1
university 1
6 0
paul 0
coleman 0
saint 0
louis 0
cardinals 0
of 0
franks 1
##ton 1
( 1
tx 1
) 1
hs 1
7 0
frank 0
thomas 0
chicago 0
white 0
sox 0
1b 0
auburn 1
university 1
8 0
earl 0
cunningham 0
chicago 0
cubs 0
of 0
lancaster 1
( 1
sc 1
) 1
hs 1
9 0
kyle 0
abbott 0
california 0
angels 0
l 0
##hp 0
long 1
beach 1
state 1
university 1
10 0
charles 0
johnson 0
montreal 0
expo 0
##s 0
c 0
westwood 1
hs 1
( 1
fort 1
pierce 1
, 1
fl 1
) 1
11 0
calvin 0
murray 0
cleveland 0
indians 0
3 0
##b 0
w 1
. 1
t 1
. 1
white 1
high 1
school 1
( 1
dallas 1
, 1
tx 1
) 1
12 0
jeff 0
jude 0
##n 0
houston 0
astros 0
r 0
##hp 0
salem 1
( 1
ma 1
) 1
hs 1
13 0
brent 0
may 0
##ne 0
kansas 0
city 0
royals 0
c 0
cal 1
state 1
fuller 1
##ton 1
14 0
steve 0
hose 0
##y 0
san 0
francisco 0
giants 0
of 0
fresno 1
state 1
university 1
15 0
ki 0
##ki 0
jones 0
los 0
angeles 0
dodgers 0
r 0
##hp 0
hillsborough 1
hs 1
( 1
tampa 1
, 1
fl 1
) 1
16 0
greg 0
b 0
##los 0
##ser 0
boston 0
red 0
sox 0
of 0
sara 1
##so 1
##ta 1
( 1
fl 1
) 1
hs 1
17 0
cal 0
el 0
##dre 0
##d 0
milwaukee 0
brewers 0
r 0
##hp 0
university 1
of 1
iowa 1
18 0
willie 0
greene 0
pittsburgh 0
pirates 0
ss 0
jones 1
county 1
hs 1
( 1
gray 1
, 1
ga 1
) 1
19 0
eddie 0
z 0
##os 0
##ky 0
toronto 0
blue 0
jays 0
ss 0
fresno 1
state 1
university 1
20 0
scott 0
bryant 0
cincinnati 0
reds 0
of 0
university 1
of 1
texas 1
21 0
greg 0
go 0
##hr 0
detroit 0
tigers 0
r 0
##hp 0
santa 1
clara 1
university 1
22 0
tom 0
goodwin 0
los 0
angeles 0
dodgers 0
of 0
fresno 1
state 1
university 1
23 0
mo 0
vaughn 0
boston 0
red 0
sox 0
1b 0
seton 1
hall 1
university 1
24 0
alan 0
z 0
##int 0
##er 0
new 0
york 0
mets 0
c 0
university 1
of 1
arizona 1
25 0
chuck 0
knob 0
##lau 0
##ch 0
minnesota 0
twins 0
2 0
##b 0
texas 1
a 1
& 1
m 1
university 1
26 0
scott 0
burr 0
##ell 0
seattle 0
mariners 0
r 0
##hp 0
ham 1
##den 1
( 1
ct 1
) 1
hs 1
from transformers import TFTapasForQuestionAnswering
model = TFTapasForQuestionAnswering.from_pretrained("google/tapas-base")
Downloading: 0%| | 0.00/422M [00:00<?, ?B/s]
All model checkpoint layers were used when initializing TFTapasForQuestionAnswering.
Some layers of TFTapasForQuestionAnswering were not initialized from the model checkpoint at google/tapas-base and are newly initialized: ['compute_column_logits', 'dropout_37', 'compute_token_logits']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
model.config.num_aggregation_labels
0
optimizer = tf.keras.optimizers.Adam(learning_rate=5e-5)
for epoch in range(10): # loop over the dataset multiple times
print("Epoch:", epoch)
for idx, batch in enumerate(train_dataloader):
# get the inputs;
input_ids = batch[0]
attention_mask = batch[1]
token_type_ids = batch[4]
labels = batch[-1]
with tf.GradientTape() as tape:
outputs = model(input_ids=input_ids, attention_mask=attention_mask, token_type_ids=token_type_ids,
labels=labels, training=True)
print("loss: ",outputs.loss.numpy().item())
grads = tape.gradient(outputs.loss, model.trainable_weights)
optimizer.apply_gradients(zip(grads, model.trainable_weights))
Epoch: 0
loss: 2.2713990211486816
WARNING:tensorflow:Gradients do not exist for variables ['tf_tapas_for_question_answering/tapas/pooler/dense/kernel:0', 'tf_tapas_for_question_answering/tapas/pooler/dense/bias:0'] when minimizing the loss. If you're using `model.compile()`, did you forget to provide a `loss`argument?
loss: 2.0187385082244873
WARNING:tensorflow:Gradients do not exist for variables ['tf_tapas_for_question_answering/tapas/pooler/dense/kernel:0', 'tf_tapas_for_question_answering/tapas/pooler/dense/bias:0'] when minimizing the loss. If you're using `model.compile()`, did you forget to provide a `loss`argument?
loss: 1.3549939393997192
---------------------------------------------------------------------------
KeyboardInterrupt Traceback (most recent call last)
<ipython-input-33-dbbf9094cd94> in <module>()
16
17 print("loss: ",outputs.loss.numpy().item())
---> 18 grads = tape.gradient(outputs.loss, model.trainable_weights)
19 optimizer.apply_gradients(zip(grads, model.trainable_weights))
/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/backprop.py in gradient(self, target, sources, output_gradients, unconnected_gradients)
1088 output_gradients=output_gradients,
1089 sources_raw=flat_sources_raw,
-> 1090 unconnected_gradients=unconnected_gradients)
1091
1092 if not self._persistent:
/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/imperative_grad.py in imperative_grad(tape, target, sources, output_gradients, sources_raw, unconnected_gradients)
75 output_gradients,
76 sources_raw,
---> 77 compat.as_str(unconnected_gradients.value))
/usr/local/lib/python3.7/dist-packages/tensorflow/python/eager/backprop.py in _gradient_function(op_name, attr_tuple, num_inputs, inputs, outputs, out_grads, skip_input_indices, forward_pass_name_scope)
157 gradient_name_scope += forward_pass_name_scope + "/"
158 with ops.name_scope(gradient_name_scope):
--> 159 return grad_fn(mock_op, *out_grads)
160 else:
161 return grad_fn(mock_op, *out_grads)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/ops/math_grad.py in _BatchMatMulV2(op, grad)
1866 if not adj_x:
1867 if not adj_y:
-> 1868 grad_x = math_ops.matmul(grad, y, adjoint_a=False, adjoint_b=True)
1869 grad_y = math_ops.matmul(x, grad, adjoint_a=True, adjoint_b=False)
1870 else:
/usr/local/lib/python3.7/dist-packages/tensorflow/python/util/traceback_utils.py in error_handler(*args, **kwargs)
148 filtered_tb = None
149 try:
--> 150 return fn(*args, **kwargs)
151 except Exception as e:
152 filtered_tb = _process_traceback_frames(e.__traceback__)
/usr/local/lib/python3.7/dist-packages/tensorflow/python/util/dispatch.py in op_dispatch_handler(*args, **kwargs)
1094 # Fallback dispatch system (dispatch v1):
1095 try:
-> 1096 return dispatch_target(*args, **kwargs)
1097 except (TypeError, ValueError):
1098 # Note: convert_to_eager_tensor currently raises a ValueError, not a
/usr/local/lib/python3.7/dist-packages/tensorflow/python/ops/math_ops.py in matmul(a, b, transpose_a, transpose_b, adjoint_a, adjoint_b, a_is_sparse, b_is_sparse, output_type, name)
3652 else:
3653 return gen_math_ops.batch_mat_mul_v2(
-> 3654 a, b, adj_x=adjoint_a, adj_y=adjoint_b, name=name)
3655
3656 # Neither matmul nor sparse_matmul support adjoint, so we conjugate
/usr/local/lib/python3.7/dist-packages/tensorflow/python/ops/gen_math_ops.py in batch_mat_mul_v2(x, y, adj_x, adj_y, name)
1570 try:
1571 _result = pywrap_tfe.TFE_Py_FastPathExecute(
-> 1572 _ctx, "BatchMatMulV2", name, x, y, "adj_x", adj_x, "adj_y", adj_y)
1573 return _result
1574 except _core._NotOkStatusException as e:
KeyboardInterrupt:
import tensorflow_probability as tfp
import collections
import numpy as np
def compute_prediction_sequence(model, data):
"""Computes predictions using model's answers to the previous questions."""
# prepare data
all_logits = []
prev_answers = None
batch_size = inputs["input_ids"].shape[0]
input_ids = inputs["input_ids"]
attention_mask = inputs["attention_mask"]
token_type_ids = inputs["token_type_ids"]
token_type_ids_example = None
for index in range(batch_size):
# If sequences have already been processed, the token type IDs will be created according to the previous
# answer.
if prev_answers is not None:
prev_labels_example = token_type_ids_example[:, 3] # shape (seq_len,)
model_labels = np.zeros_like(prev_labels_example, dtype=np.int32) # shape (seq_len,)
token_type_ids_example = token_type_ids[index].numpy() # shape (seq_len, 7)
for i in range(model_labels.shape[0]):
segment_id = token_type_ids_example[:, 0].tolist()[i]
col_id = token_type_ids_example[:, 1].tolist()[i] - 1
row_id = token_type_ids_example[:, 2].tolist()[i] - 1
if row_id >= 0 and col_id >= 0 and segment_id == 1:
model_labels[i] = int(prev_answers[(col_id, row_id)])
token_type_ids_example[:, 3] = model_labels
input_ids_example = input_ids[index]
attention_mask_example = attention_mask[index] # shape (seq_len,)
token_type_ids_example = token_type_ids[index] # shape (seq_len, 7)
outputs = model(
input_ids=np.expand_dims(input_ids_example, axis=0),
attention_mask=np.expand_dims(attention_mask_example, axis=0),
token_type_ids=np.expand_dims(token_type_ids_example, axis=0),
)
logits = outputs.logits
all_logits.append(logits)
dist_per_token = tfp.distributions.Bernoulli(logits=logits)
probabilities = dist_per_token.probs_parameter() * tf.cast(attention_mask_example, tf.float32)
coords_to_probs = collections.defaultdict(list)
token_type_ids_example = token_type_ids_example.numpy()
for i, p in enumerate(tf.squeeze(probabilities).numpy().tolist()):
segment_id = token_type_ids_example[:, 0].tolist()[i]
col = token_type_ids_example[:, 1].tolist()[i] - 1
row = token_type_ids_example[:, 2].tolist()[i] - 1
if col >= 0 and row >= 0 and segment_id == 1:
coords_to_probs[(col, row)].append(p)
prev_answers = {key: np.array(coords_to_probs[key]).mean() > 0.5 for key in coords_to_probs}
logits_batch = tf.concat(tuple(all_logits), 0)
return logits_batch
data = {'Actors': ["Brad Pitt", "Leonardo Di Caprio", "George Clooney"],
'Age': ["56", "45", "59"],
'Number of movies': ["87", "53", "69"],
'Date of birth': ["7 february 1967", "10 june 1996", "28 november 1967"]}
queries = ["How many movies has George Clooney played in?", "How old is he?", "What's his date of birth?"]
table = pd.DataFrame.from_dict(data)
inputs = tokenizer(table=table, queries=queries, padding='max_length', return_tensors="tf")
logits = compute_prediction_sequence(model, inputs)
# handy helper function in case inference on Pandas dataframe
answers = []
for coordinates in predicted_answer_coordinates:
if len(coordinates) == 1:
# only a single cell:
answers.append(table.iat[coordinates[0]])
else:
# multiple cells
cell_values = []
for coordinate in coordinates:
cell_values.append(table.iat[coordinate])
answers.append(", ".join(cell_values))
display(table)
print("")
for query, answer in zip(queries, answers):
print(query)
print("Predicted answer: " + answer)
Actors
Age
Number of movies
Date of birth
0
Brad Pitt
56
87
7 february 1967
1
Leonardo Di Caprio
45
53
10 june 1996
2
George Clooney
59
69
28 november 1967
How many movies has George Clooney played in?
Predicted answer: 7 february 1967, 10 june 1996, 28 november 1967
How old is he?
Predicted answer: 7 february 1967, 10 june 1996, 28 november 1967
What's his date of birth?
Predicted answer: 7 february 1967, 10 june 1996, 28 november 1967
