# from util_tf import profile # m = Transformer.new().data() # forcing = m.forcing(trainable= False) # autoreg = m.autoreg(trainable= False) # feed = {m.src_: src_train[:batch_size], m.tgt_: tgt_train[:batch_size]} # with tf.Session() as sess: # tf.global_variables_initializer().run() # with tf.summary.FileWriter(join(logdir, "graph"), sess.graph) as wtr: # profile(sess, wtr, forcing.loss, feed, tag= 'forcing') # profile(sess, wtr, autoreg.loss, feed, tag= 'autoreg') #################### # validation model # #################### model = Transformer.new() model_valid = model.data(*batch((src_valid, tgt_valid), batch_size), len_cap) forcing_valid = model_valid.forcing(trainable= False) autoreg_valid = model_valid.autoreg(trainable= False) idx_tgt = PointedIndex(np.load("trial/data/index_tgt.npy").item()) def trans(path, m= autoreg_valid, src= src_valid, idx= idx_tgt, len_cap= len_cap, batch_size= batch_size): rng = range(0, len(src) + batch_size, batch_size) with open(path, 'w') as f: for i, j in zip(rng, rng[1:]): for p in m.pred.eval({m.src: src[i:j], m.tgt: src[i:j,:1], m.len_tgt: len_cap}): print(decode(idx, p), file= f) # from util_io import encode # idx_src = PointedIndex(np.load("trial/data/index_src.npy").item()) # def auto(s, m= autoreg_valid, idx_src= idx_src, idx_tgt= idx_tgt, len_cap= len_cap):
names = names.astype(np.str) x = vpack(map(load, names), complex('(nan+nanj)'), 1, 1) # x = vpack(map(comp(load, path), names), complex('(nan+nanj)'), 1, 1) x[:, 0] = 0j x = c2r(x) _, t, d = x.shape assert t <= len_cap assert d == dim_tgt return x #################### # validation model # #################### model = Transformer.new(dim_src=len(index), dim_tgt=dim_tgt) model_valid = model.data(texts[:split], load_batch(names[:split]), len_cap) forcing_valid = model_valid.forcing(trainable=False) autoreg_valid = model_valid.autoreg(trainable=False) # # for profiling # from util_tf import profile # with tf.Session() as sess: # tf.global_variables_initializer().run() # with tf.summary.FileWriter(join(logdir, "graph"), sess.graph) as wtr: # profile(sess, wtr, forcing_valid.loss, tag= 'forcing') # profile(sess, wtr, autoreg_valid.loss, tag= 'autoreg') # ' according to their categories or crimes.\n' src, tgt = texts[978:979, :42], load_batch(names[978:979])[:, :-1] synth_forcing = {forcing_valid.src: src, forcing_valid.tgt: tgt}