From 9f49c6d7646db371b2d0127579b14c584c4cc7d3 Mon Sep 17 00:00:00 2001
From: Caroline DE POURTALES <cdepourt@montana.irit.fr>
Date: Wed, 11 May 2022 14:29:49 +0200
Subject: [PATCH] it runs, some corrections needed next
---
SuperTagger/Linker/Linker.py | 5 -
.../Linker/__pycache__/utils.cpython-38.pyc | Bin 5906 -> 7029 bytes
SuperTagger/Linker/utils.py | 95 ++++++++++++------
SuperTagger/eval.py | 8 +-
train.py | 2 +
5 files changed, 70 insertions(+), 40 deletions(-)
diff --git a/SuperTagger/Linker/Linker.py b/SuperTagger/Linker/Linker.py
index b8dbf8c..281a7ab 100644
--- a/SuperTagger/Linker/Linker.py
+++ b/SuperTagger/Linker/Linker.py
@@ -61,11 +61,6 @@ class Linker(Module):
LayerNorm(self.dim_embedding_atoms, eps=1e-12)
)
- def make_decoder_mask(self, atoms_batch):
- decoder_attn_mask = torch.ones_like(atoms_batch, dtype=torch.float64)
- decoder_attn_mask[atoms_batch.eq(self.padding_id)] = 0.0
- return decoder_attn_mask.unsqueeze(1).repeat(1, atoms_batch.shape[1], 1)
-
def forward(self, atoms_batch_tokenized, atoms_polarity_batch, sents_embedding):
'''
Parameters :
diff --git a/SuperTagger/Linker/__pycache__/utils.cpython-38.pyc b/SuperTagger/Linker/__pycache__/utils.cpython-38.pyc
index 7f42ae68119786d9252d2c8eac5eb9c7a7083377..b13bf57006d2d3eb78625e40b86218f03e27798c 100644
GIT binary patch
delta 1561
zcmbQF_ti{0l$V!_fq{WRZDw_nwF(2nV-N=!GchnQI503U6t_&&Ze~ki3}(<2+4!iI
ziBWuV0&^Lo_~h%%-xwt)pJOp*l$^}OT4kokz`&5sP{R<b7Q<A_Sj$wy)WOifD9Mn`
zR-^`DEnui&TF98kl)^BPsgNa@L6fnFb@BpMOGdfLw^&m}<UqFZFfcH%F){rUVC4GG
z!pO4On~jf=QD|}$yF5P=*FP2}HU_4@Y^*#?Mf#I>amh|z%Kn~FWAapvd5pS~O*oe`
zs!zVdd6dz3av`ty<egmq7@a5Y;9kn;IXR9ejnQfHQJ&R|{*w!NO_|b|C$Hc&k9MhH
zTELvbwvdsLVFGiWTnEDfrWE#tj1w4(E|st>U@c+mU|7hsfPEoDElUbV3R4S14T}py
zGh;1F4ND5ABts2L3YR1U6GJCsCsPe$3UfBo1jZt*cE&Wu6z<6meB!2iIv5u)rLZhy
zEQYXH7cv%|K-ihbSHo1o(ZQI)+{|RoP|MWLl*X9CGueSp%x(f>F<%E0L^GVt3U-<l
z*Z^jbqoD>crLaK^$Xf!ooEKz-%>>5cB$)khHpGZL3?mp}Mu6SG2Qr|30%P$OBsW0V
z5CaZ?4FCrm+zn7qOkm8L01in0g^UxJi;^$`Qh=y{WGZ6c2lk90$SpP#n2H17o`JDJ
zZec2l0lNh`JWxF&1Tw&G0#orKxMyH&hym-s24M9JG(d$xMu6P#4ap4<HpmF(A|?zY
zP$N|YVgPfo2iy%XHpGAsumQ-PfJbT#Q!s<3=w<_c1xCiy$$<jytZuiMQi>+`3Pv*K
zO#UX2$-a{D7H3Xs+T?UWrOAAPfs89BCkV>$l!NjU3n)J^Fiqwa3{oiaVPIe=VgnH(
zAVQRZfuSBvk!X=0ND5U^>*QL&Ovb*+&jsVL8ZsHp5Cv?8%*0TH#W@Sn6k&DF^2y0U
z35-#bj|f@U`-5By_DvuI14ArGdmN}3V`pI$W0Ya!W9DHLVB}*IV&Y-~@p+i|n8X;B
z7<rfkSRgcvFNC5Jqyi>`u9A<b2xM0>sD$(e*#*L&00UuWkPEWFMPVmH4WkP~tW7Ob
zIzugU4U-E)tXwTi38>UyYzFc8YFTTT7ckbag0ga<(&o9sDU6K!Co_pUF$PX{6;)zn
zo17tP#C?k?u>=xpL6cXB%BpPuSs@0pl!JwlgAs)JKwe~&Vq{`q`d<{vz`&p>JXu6+
s7AMG^VE*KjV#*3K3=9k$j9|#b!^p$P!^FbDA<DtS!Oy|V!N(B{0FO&eD*ylh
delta 1028
zcmexrHc3xAl$V!_fq{WxL334-fiwfdV-N=!voSC*I503U6mOcS-E5L7lp>NMnj)6Y
zl+M`96eXN0k}BNH$jFc?lFE}Jk}3?BV+M()Fa|Sdif`Of%fu))*@(G}QEc)`=5LG=
zlNYcUGfGUp!cwIo$H2gl&QQY;D;C34%UH|Q!O+1d$*_Q-hG`+=M5aQPV1^>L$t|pw
zjIxvWv8M9Kf?UGFz`(%9$h6s<jgOI0aIzb_JU1iPe-=hI2Bv>(ESn411sECCCpU7;
zW7L_fz`2}JZSo1uql`wI`?y3I8677t;a<w<KG};WjnQH9MxND-zLWd7#3z^V$}zPt
zPM*SRp6FJ?w1ByWk%wU+BO}8E#yqPIh6PM1YzrBS9+a>wU@c+mU|7hsfPEoDElUb}
z3R4S14T}pyGh;1F4ND4#Bts2L3a2Ck6GJCsCsPem2}cJb$gE~2bB0=`cBVAO6t2k&
zc*W}vbucYpN?}>ZSPWsYE@Uh^2eBDsZ=L{HFB3v<-U6^YxEC@`U@U6Da0w57m$Wmc
zF{bdQFtu=$Ff3qP$T0Z=AHR1<2P4D{6BvsjY*w)82^j7`@&mfNYM6o<H2F6F6Hs7e
zOr9(+<j!)7DWz!gK@PLYA4HTHvnC%E%4F1>yi!<ZvX?OL<ZU9cd{v-m1f?uC1||lM
z$rpu#7&RyB3rjQBO};0Vz<6u2y|{I~4+8^3kqC(JXJBB625Hd)rA&4fMlnViMge9n
zMgc}XMj>VaW)Po;Nr;&TDk8?H#K^-Wz%0O21k#(#1kwk^?4Z=cz`(!^GB6LE{5lzG
z7+n}*9cr1<8ETnpm|PfQm1^Z#QW$GlOBiZcYFL{=LVUGsHOvbbYuG^1Q>eDtMIwcf
zan|IOl1_~Nlix@xF|tqQmogG43SeMhxW$xMQUvl~QQ%}BDOukxkYy4e7jUpJaxg+L
zAISBLQjAOtO#h2Q7#J8{g0f;2UrLdNp-x^wyrE8UyrE_lUrG)#UsH5)ztk*gkiWp1
zqBs*v@^g#h3-WVru@)3%=9Nq~l2%qzVqjq4U<5-Z9!4HU9wruc4p9ys4n7WE4n7Wj
K4q*XC4n_bVv)(EI
diff --git a/SuperTagger/Linker/utils.py b/SuperTagger/Linker/utils.py
index 898a921..3f8e892 100644
--- a/SuperTagger/Linker/utils.py
+++ b/SuperTagger/Linker/utils.py
@@ -9,6 +9,7 @@ from SuperTagger.utils import pad_sequence
regex_categories = r'\w+\(\d+,(?:((?R))|(\w+))*,?(?:((?R))|(\w+))*\)'
+
#########################################################################################
################################ Liste des atoms avc _i########################################
#########################################################################################
@@ -26,17 +27,17 @@ def get_axiom_links(max_atoms_in_one_type, atoms_polarity, batch_axiom_links):
range(len(atoms_batch))]
linking_plus_to_minus = pad_sequence(
- [torch.as_tensor([l_polarity_minus[s_idx].index(x) for i, x in enumerate(l_polarity_plus[s_idx])], dtype=torch.long)
- for s_idx in range(len(atoms_batch))], max_len=max_atoms_in_one_type//2, padding_value=-1)
+ [torch.as_tensor([l_polarity_minus[s_idx].index(x) if x in l_polarity_minus[s_idx] else -1 for i, x in
+ enumerate(l_polarity_plus[s_idx])], dtype=torch.long)
+ for s_idx in range(len(atoms_batch))], max_len=max_atoms_in_one_type // 2, padding_value=-1)
linking_plus_to_minus_all_types.append(linking_plus_to_minus)
return torch.stack(linking_plus_to_minus_all_types)
-
def category_to_atoms_axiom_links(category, categories_to_atoms):
- res = [bool(re.match(r''+atom_type+"_\d+", category)) for atom_type in atom_map.keys()]
+ res = [bool(re.match(r'' + atom_type + "_\d+", category)) for atom_type in atom_map.keys()]
if True in res:
return [category]
else:
@@ -56,13 +57,14 @@ def get_atoms_links_batch(category_batch):
batch.append(categories_to_atoms)
return batch
+
#########################################################################################
################################ Liste des atoms ########################################
#########################################################################################
def category_to_atoms(category, categories_to_atoms):
- res = [bool(re.match(r''+atom_type+"_\d+", category)) for atom_type in atom_map.keys()]
+ res = [bool(re.match(r'' + atom_type + "_\d+", category)) for atom_type in atom_map.keys()]
if True in res:
category = re.match(r'([a-zA-Z|_]+)_\d+', category).group(1)
return [category]
@@ -96,9 +98,9 @@ def category_to_atoms_polarity(category, polarity):
Boolean Tensor of shape max_symbols_in_word, containing 1 for pos indexes and 0 for neg indexes
'''
category_to_polarity = []
- res = [bool(re.match(r''+atom_type+"_\d+", category)) for atom_type in atom_map.keys()]
+ res = [bool(re.match(r'' + atom_type + "_\d+", category)) for atom_type in atom_map.keys()]
if True in res or category.startswith("dia") or category.startswith("box"):
- category_to_polarity.append(not polarity)
+ category_to_polarity.append(False)
else:
# dr = /
if category.startswith("dr"):
@@ -106,15 +108,33 @@ def category_to_atoms_polarity(category, polarity):
category_cut = [cat for cat in category_cut if cat is not None]
left_side, right_side = category_cut[0], category_cut[1]
- # for the left side
- category_to_polarity += category_to_atoms_polarity(left_side, not polarity)
-
- # for the right side
- res = [bool(re.match(r'' + atom_type + "_\d+", right_side)) for atom_type in atom_map.keys()]
- if True in res or right_side.startswith("dia") or right_side.startswith("box"):
- category_to_polarity.append(polarity)
- else :
- category_to_polarity += category_to_atoms_polarity(right_side, not polarity)
+ if polarity == True:
+ # for the left side : normal
+ res = [bool(re.match(r'' + atom_type + "_\d+", left_side)) for atom_type in atom_map.keys()]
+ if True in res or left_side.startswith("dia") or left_side.startswith("box"):
+ category_to_polarity.append(False)
+ else:
+ category_to_polarity += category_to_atoms_polarity(left_side, True)
+ # for the right side : change polarity for next right formula
+ res = [bool(re.match(r'' + atom_type + "_\d+", right_side)) for atom_type in atom_map.keys()]
+ if True in res or right_side.startswith("dia") or right_side.startswith("box"):
+ category_to_polarity.append(True)
+ else:
+ category_to_polarity += category_to_atoms_polarity(right_side, False)
+
+ else:
+ # for the left side
+ res = [bool(re.match(r'' + atom_type + "_\d+", left_side)) for atom_type in atom_map.keys()]
+ if True in res or left_side.startswith("dia") or left_side.startswith("box"):
+ category_to_polarity.append(True)
+ else:
+ category_to_polarity += category_to_atoms_polarity(left_side, False)
+ # for the right side : change polarity for next right formula
+ res = [bool(re.match(r'' + atom_type + "_\d+", right_side)) for atom_type in atom_map.keys()]
+ if True in res or right_side.startswith("dia") or right_side.startswith("box"):
+ category_to_polarity.append(False)
+ else:
+ category_to_polarity += category_to_atoms_polarity(right_side, True)
# dl = \
elif category.startswith("dl"):
@@ -122,15 +142,33 @@ def category_to_atoms_polarity(category, polarity):
category_cut = [cat for cat in category_cut if cat is not None]
left_side, right_side = category_cut[0], category_cut[1]
- # for the left side
- res = [bool(re.match(r'' + atom_type + "_\d+", left_side)) for atom_type in atom_map.keys()]
- if True in res or left_side.startswith("dia") or left_side.startswith("box"):
- category_to_polarity.append(polarity)
- else :
- category_to_polarity += category_to_atoms_polarity(left_side, not polarity)
-
- # for the right side
- category_to_polarity += category_to_atoms_polarity(right_side, not polarity)
+ if polarity == True:
+ # for the left side : change polarity
+ res = [bool(re.match(r'' + atom_type + "_\d+", left_side)) for atom_type in atom_map.keys()]
+ if True in res or left_side.startswith("dia") or left_side.startswith("box"):
+ category_to_polarity.append(True)
+ else:
+ category_to_polarity += category_to_atoms_polarity(left_side, False)
+ # for the right side : normal
+ res = [bool(re.match(r'' + atom_type + "_\d+", right_side)) for atom_type in atom_map.keys()]
+ if True in res or right_side.startswith("dia") or right_side.startswith("box"):
+ category_to_polarity.append(False)
+ else:
+ category_to_polarity += category_to_atoms_polarity(right_side, True)
+
+ else:
+ # for the left side
+ res = [bool(re.match(r'' + atom_type + "_\d+", left_side)) for atom_type in atom_map.keys()]
+ if True in res or left_side.startswith("dia") or left_side.startswith("box"):
+ category_to_polarity.append(False)
+ else:
+ category_to_polarity += category_to_atoms_polarity(left_side, True)
+ # for the right side
+ res = [bool(re.match(r'' + atom_type + "_\d+", right_side)) for atom_type in atom_map.keys()]
+ if True in res or right_side.startswith("dia") or right_side.startswith("box"):
+ category_to_polarity.append(True)
+ else:
+ category_to_polarity += category_to_atoms_polarity(right_side, False)
return category_to_polarity
@@ -147,13 +185,8 @@ def find_pos_neg_idexes(max_atoms_in_sentence, atoms_batch):
for sentence in atoms_batch:
list_atoms = []
for category in sentence:
- polarity = True
- for at in category_to_atoms_polarity(category, polarity):
+ for at in category_to_atoms_polarity(category, True):
list_atoms.append(at)
list_batch.append(torch.as_tensor(list_atoms))
return pad_sequence([list_batch[i] for i in range(len(list_batch))],
max_len=max_atoms_in_sentence, padding_value=0)
-
-atoms_pol = find_pos_neg_idexes(10, [['dr(1,np_1,s_1)', 'dl(1,np_1,s_1)']])
-print(atoms_pol)
-print(get_axiom_links(10, atoms_pol, [['dr(1,np_1,s_1)', 'dl(1,np_1,s_1)']]))
diff --git a/SuperTagger/eval.py b/SuperTagger/eval.py
index 9bfdc85..b287d4b 100644
--- a/SuperTagger/eval.py
+++ b/SuperTagger/eval.py
@@ -18,16 +18,16 @@ class SinkhornLoss(Module):
for link, perm in zip(predictions, truths))
-def mesure_accuracy(linking_plus_to_minus, axiom_links_pred):
+def mesure_accuracy(batch_true_links, axiom_links_pred):
r"""
batch_axiom_links : (batch_size, ...)
axiom_links_pred : (batch_size, max_atoms_type_polarity)
"""
correct_links = torch.ones(axiom_links_pred.size())
- correct_links[axiom_links_pred != linking_plus_to_minus] = 0
- correct_links[linking_plus_to_minus == -1] = 1
+ correct_links[axiom_links_pred != batch_true_links] = 0
+ correct_links[batch_true_links == -1] = 1
num_correct_links = correct_links.sum().item()
- num_masked_atoms = len(linking_plus_to_minus[linking_plus_to_minus == -1])
+ num_masked_atoms = len(batch_true_links[batch_true_links == -1])
# diviser par nombre de links
return (num_correct_links - num_masked_atoms)/(axiom_links_pred.size()[0] * axiom_links_pred.size()[1] * axiom_links_pred.size()[2] - num_masked_atoms)
diff --git a/train.py b/train.py
index fced23b..05d223f 100644
--- a/train.py
+++ b/train.py
@@ -53,9 +53,11 @@ print("atoms_polarity_batch", atoms_polarity_batch.shape)
truth_links_batch = get_axiom_links(max_atoms_in_one_type, atoms_polarity_batch, df_axiom_links["sub_tree"])
print("truth_links_batch", truth_links_batch.permute(1, 0, 2).shape)
+print(" truth_links_batch example on first sentence class cl_r", truth_links_batch[0][0])
sentences_batch = df_axiom_links["Sentences"]
+# Construction tensor dataset
dataset = TensorDataset(atoms_batch_tokenized, atoms_polarity_batch, truth_links_batch.permute(1, 0, 2))
# Calculate the number of samples to include in each set.
--
GitLab