'n_layers': 1,
'hidden_size': nli_args.hidden_size,
'bidirectional': nli_args.bidirectional,
'embedding_size': 300,
'fix_emb': True,
'dp_ratio': 0.3,
'd_out': 2, # 2 classes
'mlp_classif_hidden_size_list': [512, 512],
'cuda': torch.cuda.is_available(),
})
squad_state = {k: v for k, v in squad_args.items()}
if __name__ == "__main__":
print(nli_args)
nli_dm = wrangle.DataManager(nli_args)
nli_args.n_embed = nli_dm.vocab.n_words
if nli_args.type == 'siamese':
nli_model = siamese_pytorch.SiameseClassifier(config=nli_args)
squad_dm = SquadDataManager(squad_args)
squad_args.n_embed = squad_dm.vocab.n_words
squad_model = squad_pytorch.SquadClassifier(
config=squad_args,
encoder=nli_model.encoder,
)
else:
raise Exception('model type not supported')
print("number of trainable parameters found {}".format(
sum(param.nelement()
'embedding_size': 300,
'fix_emb': True,
'dp_ratio': 0.3,
'd_out': 2, # 2 classes
'mlp_classif_hidden_size_list': [512, 512],
'cuda': torch.cuda.is_available(),
})
squad_state = {k: v for k, v in squad_args.items()}
if __name__ == "__main__":
print(args)
checkpoint = sys.argv[1]
print('found checkpoint dir {}'.format(checkpoint))
dm = wrangle.DataManager(args)
if args.add_squad: # add squad to vocab to match checkpoint
squad_dm = SquadDataManager(squad_args, vocab=dm.vocab)
args.n_embed = dm.vocab.n_words
if args.type == 'siamese':
model = siamese_pytorch.SiameseClassifier(config=args)
model.embed.weight.data = load_embeddings.load_embeddings(
dm.vocab, constants.EMBED_DATA_PATH, args.embedding_size)
elif args.type == 'decomposable':
model = decomposable_pytorch.SNLIClassifier(config=args)
model.encoder.embedding.weight.data = load_embeddings.load_embeddings(
dm.vocab, constants.EMBED_DATA_PATH, args.embedding_size)
else:
model = Seq2SeqPytorch(args=args, vocab=dm.vocab)
model.encoder.embedding.weight.data = load_embeddings.load_embeddings(
dm.vocab, constants.EMBED_DATA_PATH, args.embedding_size)