import logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s %(message)s')
logging.getLogger('imported_module').setLevel(logging.WARNING)
from options.train_options import TrainOptions
from datasets import get_dataloader
from executors import get_executor
from models.parser import Seq2seqParser
from trainer import Trainer
import numpy as np
import random
np.random.seed(42)
random.seed(42)
import clevr_parser
graph_parser = clevr_parser.Parser(backend='spacy', model='en_core_web_sm',
has_spatial=True,
has_matching=True).get_backend(identifier='spacy')
embedder = clevr_parser.Embedder(backend='torch', parser=graph_parser).get_backend(identifier='torch')
opt = TrainOptions().parse()
train_loader = get_dataloader(opt, 'train', graph_parser=graph_parser, embedder=embedder)
val_loader = get_dataloader(opt, 'val', graph_parser=graph_parser, embedder=embedder)
model = Seq2seqParser(opt)
executor = get_executor(opt)
trainer = Trainer(opt, train_loader, val_loader, model, executor)
trainer.train()
def main(args):
"""
Save nx.graph (Gss, Gts,...) and corresponding torch_geometric.data.PairData
(via clevr_parse embedder api).
"""
if (args.input_vocab_json == '') and (args.output_vocab_json == ''):
logger.info(
'Must give one of --input_vocab_json or --output_vocab_json')
return
graph_parser = clevr_parser.Parser(
backend='spacy',
model=args.parser_lm,
has_spatial=True,
has_matching=True).get_backend(identifier='spacy')
embedder = clevr_parser.Embedder(
backend='torch', parser=graph_parser).get_backend(identifier='torch')
is_directed_graph = args.is_directed_graph # Parse graphs as nx.MultiDiGraph
out_dir, out_f_prefix = _get_out_dir_and_file_prefix(args)
checkpoint_dir = f"{out_dir}/checkpoints"
utils.mkdirs(checkpoint_dir)
questions, img_scenes = get_questions_and_parsed_scenes(
args.input_questions_json, args.input_parsed_img_scenes_json)
if args.is_debug:
set_default_level(10)
questions = questions[:
128] # default BSZ is 64 ensuring enought for batch iter
logger.debug(
f"In DEBUG mode, sampling {len(questions)} questions only..")
# Process Vocab #
vocab = _process_vocab(args, questions)
# Encode all questions and programs
logger.info('Encoding data')
questions_encoded, programs_encoded, answers, image_idxs = [], [], [], []
question_families = []
orig_idxs = []
# Graphs and Embeddings #
data_s_list = [] # List [torch_geometric.data.Data]
data_t_list = [] # List [torch_geometric.data.Data]
num_samples = 0 # Counter for keeping track of processed samples
num_skipped = 0 # Counter for tracking num of samples skipped
for orig_idx, q in enumerate(questions):
# First See if Gss, Gts are possible to extract.
# If not (for e.g., some edges cases like plurality, skip data sample
img_idx = q['image_index']
img_fn = q['image_filename']
logger.debug(f"\tProcessing Image - {img_idx}: {img_fn} ...")
# q_idx = q['question_index']
# q_fam_idx = q['question_family_index']
## 1: Ensure both Gs,Gt is parseable for this question sample, o.w. skip
img_scene = list(
filter(lambda x: x['image_index'] == img_idx, img_scenes))[0]
try:
Gt, t_doc = graph_parser.get_doc_from_img_scene(
img_scene, is_directed_graph=is_directed_graph)
X_t, ei_t, e_attr_t = embedder.embed_t(
img_idx, args.input_parsed_img_scenes_json)
except AssertionError as ae:
logger.warning(f"AssertionError Encountered: {ae}")
logger.warning(f"[{img_fn}] Excluding images with > 10 objects")
num_skipped += 1
continue
if Gt is None and ("SKIP" in t_doc):
# If the derendering pipeline failed, then just skip the
# scene, don't process the labels (and text_scenes) for the image
print(f"Got None img_doc at image_index: {img_idx}")
print(f"Skipping all text_scenes for imgage idx: {img_idx}")
num_skipped += 1
continue
s = q['question']
orig_idx = q['question_index']
try:
Gs, s_doc = graph_parser.parse(s,
return_doc=True,
is_directed_graph=is_directed_graph)
X_s, ei_s, e_attr_s = embedder.embed_s(s)
except ValueError as ve:
logger.warning(f"ValueError Encountered: {ve}")
logger.warning(f"Skipping question: {s} for {img_fn}")
num_skipped += 1
continue
if Gs is None and ("SKIP" in s_doc):
logger.warning(
"Got None as Gs and 'SKIP' in Gs_embd. (likely plural with CLEVR_OBJS label) "
)
logger.warning(
f"SKIPPING processing {s} for {img_fn} and at {img_idx}")
num_skipped += 1
continue
# Using ClevrData allows us a debug extension to Data
data_s = ClevrData(x=X_s, edge_index=ei_s, edge_attr=e_attr_s)
data_t = ClevrData(x=X_t, edge_index=ei_t, edge_attr=e_attr_t)
data_s_list.append(data_s)
data_t_list.append(data_t)
question = q['question']
orig_idxs.append(orig_idx)
image_idxs.append(img_idx)
if 'question_family_index' in q:
question_families.append(q['question_family_index'])
question_tokens = preprocess_utils.tokenize(question,
punct_to_keep=[';', ','],
punct_to_remove=['?', '.'])
question_encoded = preprocess_utils.encode(
question_tokens,
vocab['question_token_to_idx'],
allow_unk=args.encode_unk == 1)
questions_encoded.append(question_encoded)
has_prog_seq = 'program' in q
if has_prog_seq:
program = q['program']
program_str = program_to_str(program, args.mode)
program_tokens = preprocess_utils.tokenize(program_str)
program_encoded = preprocess_utils.encode(
program_tokens, vocab['program_token_to_idx'])
programs_encoded.append(program_encoded)
if 'answer' in q:
ans = q['answer']
answers.append(vocab['answer_token_to_idx'][ans])
num_samples += 1
logger.info("-" * 50)
logger.info(f"Samples processed count = {num_samples}")
if has_prog_seq:
logger.info(f"\n[{orig_idx}]: question: {question} \n"
f"\tprog_str: {program_str} \n"
f"\tanswer: {ans}")
logger.info("-" * 50)
# ---- CHECKPOINT ---- #
if num_samples % args.checkpoint_every == 0:
logger.info(f"Checkpointing at {num_samples}")
checkpoint_fn_prefix = f"{out_f_prefix}_{num_samples}"
_out_dir = f"{checkpoint_dir}/{out_f_prefix}_{num_samples}"
utils.mkdirs(_out_dir)
out_fpp = f"{_out_dir}/{checkpoint_fn_prefix}"
# ------------ Checkpoint .H5 ------------#
logger.info(
f"CHECKPOINT: Saving checkpoint files at directory: {out_fpp}")
save_h5(f"{out_fpp}.h5", vocab, questions_encoded, image_idxs,
orig_idxs, programs_encoded, question_families, answers)
# ------------ Checkpoint GRAPH DATA ------------#
save_graph_pairdata(out_fpp,
data_s_list,
data_t_list,
is_directed_graph=is_directed_graph)
logger.info(f"-------------- CHECKPOINT: COMPLETED --------")
if (args.max_sample > 0) and (num_samples >= args.max_sample):
logger.info(f"len(questions_encoded = {len(questions_encoded)}")
logger.info("args.max_sample reached: Completing ... ")
break
logger.debug(f"Total samples skipped = {num_skipped}")
logger.debug(f"Total samples processed = {num_samples}")
out_fpp = f"{out_dir}/{out_f_prefix}"
## SAVE .H5: Baseline {dataset}_h5.h5 file (q,p,ans,img_idx) as usual
logger.info(f"Saving baseline (processed) data in: {out_fpp}.h5")
save_h5(f"{out_fpp}.h5", vocab, questions_encoded, image_idxs, orig_idxs,
programs_encoded, question_families, answers)
## ------------ SAVE GRAPH DATA ------------ ##
## N.b. Ensure the len of theses lists are all equals
save_graph_pairdata(out_fpp,
data_s_list,
data_t_list,
is_directed_graph=is_directed_graph)
logger.info(f"Saved Graph Data in: {out_fpp}_*.[h5|.gpickle|.npz|.pt] ")
def embedder(parser):
embedder = clevr_parser.Embedder(backend='torch', parser=parser).get_backend(identifier='torch')
return embedder
def main(args):
"""
Save nx.graph (Gss, Gts,...) and corresponding torch_geometric.data.PairData
(via clevr_parse embedder api).
"""
if args.is_debug:
set_default_level(10)
is_directed_graph = args.is_directed_graph
logger.debug(f"Parser flag is_directed_graph = {is_directed_graph}")
graph_parser = clevr_parser.Parser(
backend="spacy",
model='en_core_web_sm',
has_spatial=True,
has_matching=True).get_backend(identifier='spacy')
embedder = clevr_parser.Embedder(
backend='torch', parser=graph_parser).get_backend(identifier='torch')
raw_questions, img_scenes = get_questions_and_parsed_scenes(
args.input_questions_json, args.input_parsed_img_scenes_json)
logger.info('| importing questions from %s' % args.input_question_h5)
input_questions = h5py.File(args.input_question_h5, 'r')
#N = len(input_questions['questions'])
# Baseline Entities #
questions, programs, answers, question_families, orig_idxs, img_idxs = [], [], [], [], [], []
family_count = np.zeros(90)
# Graphs and Embeddings #
data_s_list = [] # List [torch_geometric.data.Data]
data_t_list = [] # List [torch_geometric.data.Data]
filename = get_output_filename(args)
__all_question_families: np.ndarray = input_questions['question_families'][
()]
__all_enc_questions: np.ndarray = input_questions['questions'][()]
__all_img_indices: np.ndarray = input_questions['image_idxs'][()]
logger.debug(f"__all_question_families len {len(__all_question_families)}")
# Sample N items for each 90 families #
fam2indices = get_question_fam_to_indices(args)
M = len(fam2indices.keys()) # 90
N = args.n_questions_per_family # 50
max_sample = N * M # 90 * 50 = 4500
family_count = np.zeros(M) # family_count = Counter()
# TODO: accumulating values here need to be parallelized, and joined write ex-post
num_skipped = 0 # Counter for tracking num of samples skipped
for fam_idx, i_samples in enumerate(fam2indices):
all_fam_samples = fam2indices[fam_idx]
logger.debug(
f"Question_family {fam_idx} has {len(all_fam_samples)} samples to choose {N} samples"
)
N_question_sample_indices = np.random.choice(
all_fam_samples, N, replace=False) # N.b seed is fixed
assert len(N_question_sample_indices) == N
# TODO: parallelize this iteration loop
for i in N_question_sample_indices:
try:
img_idx = __all_img_indices[i]
logger.debug(
f"\tProcessing Image - {img_idx} from fam_idx {fam_idx}: {i} of {i_samples}"
)
img_scene = list(
filter(lambda x: x['image_index'] == img_idx,
img_scenes))[0]
except IndexError as ie:
logger.warning(f"For {img_idx}: {ie}")
num_skipped += 1
continue
try:
Gt, t_doc = graph_parser.get_doc_from_img_scene(
img_scene, is_directed_graph=is_directed_graph)
X_t, ei_t, e_attr_t = embedder.embed_t(
img_idx, args.input_parsed_img_scenes_json)
except AssertionError as ae:
logger.warning(f"AssertionError Encountered: {ae}")
logger.warning(
f"[{img_idx}] Excluding images with > 10 objects")
num_skipped += 1
continue
if Gt is None and ("SKIP" in t_doc):
# If the derendering pipeline failed, then just skip the
# scene, don't process the labels (and text_scenes) for the image
logger.warning(f"Got None img_doc at image_index: {img_idx}")
print(f"Skipping all text_scenes for imgage idx: {img_idx}")
num_skipped += 1
continue
q_idx = input_questions['orig_idxs'][i]
q_obj = list(
filter(lambda x: x['question_index'] == q_idx,
raw_questions))[0]
assert q_obj['image_index'] == img_idx
s = q_obj['question']
try:
Gs, s_doc = graph_parser.parse(
s, return_doc=True, is_directed_graph=is_directed_graph)
X_s, ei_s, e_attr_s = embedder.embed_s(s)
except ValueError as ve:
logger.warning(f"ValueError Encountered: {ve}")
logger.warning(f"Skipping question: {s} for {img_fn}")
num_skipped += 1
continue
if Gs is None and ("SKIP" in s_doc):
logger.warning(
"Got None as Gs and 'SKIP' in Gs_embd. (likely plural with CLEVR_OBJS label) "
)
logger.warning(f"SKIPPING processing {s} at {img_idx}")
num_skipped += 1
continue
data_s = ClevrData(x=X_s, edge_index=ei_s, edge_attr=e_attr_s)
data_t = ClevrData(x=X_t, edge_index=ei_t, edge_attr=e_attr_t)
data_s_list.append(data_s)
data_t_list.append(data_t)
family_count[fam_idx] += 1
questions.append(input_questions['questions'][i])
programs.append(input_questions['programs'][i])
answers.append(input_questions['answers'][i])
question_families.append(input_questions['question_families'][i])
orig_idxs.append(input_questions['orig_idxs'][i])
img_idxs.append(img_idx)
logger.info(f"\nCount = {family_count.sum()}\n")
if family_count.sum() >= max_sample:
break
logger.debug(
f"Total samples skipped (due to errors/exceptions) = {num_skipped}")
# ---------------------------------------------------------------------------#
## SAVE .H5
if not os.path.isdir(args.output_dir):
os.mkdir(args.output_dir)
output_file = os.path.join(args.output_dir, filename)
out_dir = args.output_dir
out_f_prefix = filename.split('.')[0]
out_fpp = f"{out_dir}/{out_f_prefix}"
logger.debug(f"out_fpp = {out_fpp}")
print('sampled question family distribution')
print(family_count)
print('| saving output file to %s' % output_file)
with h5py.File(output_file, 'w') as f:
f.create_dataset('questions',
data=np.asarray(questions, dtype=np.int32))
f.create_dataset('programs', data=np.asarray(programs, dtype=np.int32))
f.create_dataset('answers', data=np.asarray(answers))
f.create_dataset('image_idxs', data=np.asarray(img_idxs))
f.create_dataset('orig_idxs', data=np.asarray(orig_idxs))
f.create_dataset('question_families',
data=np.asarray(question_families))
## ------------ SAVE GRAPH DATA ------------ ##
save_graph_pairdata(out_fpp,
data_s_list,
data_t_list,
is_directed_graph=is_directed_graph)
logger.info(f"Saved Graph Data in: {out_fpp}_*.[h5|.gpickle|.npz|.pt] ")
print('| done')