Words and The Company They Keep (homework 2)¶
Solution to the second homework for the NPFL067 course.
source_files = [
"CZ.txt",
"EN.txt"
]
dir = "plain"
## load the text files
def load_text(filename, dir) -> str:
with open(f"../texts/{dir}/{filename}", "r", encoding="utf-8") as f:
return f.read()
texts = {file : load_text(file, dir).split('\n') for file in source_files}
for short_text, words in texts.items():
print(f"Text: {short_text}, word-count: {len(words)}, unique-words: {len(set(words))}")
lexicons = {file : set(texts[file]) for file in texts}
alphabets = {file : set("".join(texts[file])) for file in texts}
Text: CZ.txt, word-count: 222413, unique-words: 42827 Text: EN.txt, word-count: 221099, unique-words: 9608
from typing import List, Dict, Set, Tuple, Any
from collections import Counter, defaultdict
import pandas as pd
from math import log2
Best Friends¶
Compute the pointwise mutual information for all the possible word pairs appearing consecutively in the data, disregarding pairs in which one or both words appear less than 10 times in the corpus, and sort the results from the best to the worst (did you get any negative values? Why?) Tabulate the results, and show the best 20 pairs for both data sets.
Do the same now but for distant words, i.e. words which are at least 1 word apart, but not farther than 50 words (both directions). Again, tabulate the results, and show the best 20 pairs for both data sets.
## obtain the count for all word pairs appering consecutively
def count_word_pairs(text : List[str]) -> Counter:
return Counter(zip(text, text[1:]))
def count_word_surrounding_pairs(text : List[str], min_distance : int = 2, max_distance : int = 50) -> Counter:
counter = Counter()
## (both derections)
for dist in range(min_distance, max_distance + 1):
counter.update(zip(text, text[dist:]))
counter.update(zip(text[dist:], text))
return counter
## cut out less frequent word pairs (less than 10)
def cut_non_frequent_words(text : List['str']) -> List['str']:
limit = 10
word_count = Counter(text)
return [word for word in text if word_count[word] >= limit]
## show top 20 most frequent word pairs
def show_top_word_pairs(counter : Counter, n : int = 20) -> None:
for pair, count in counter.most_common(n):
print(f"\t'{pair}' - #{count}")
tables = []
for short_text in texts:
for count_func, shortcut in [
(count_word_pairs, "Consecutive"),
(count_word_surrounding_pairs, "Surrounding")
]:
source_text = cut_non_frequent_words(texts[short_text])
word_pairs : Counter = count_func(source_text)
df = pd.DataFrame(word_pairs.items(), columns=[f"Pair", f"Count"]).nlargest(20, f"Count").reset_index(drop=True)
tables.append(pd.DataFrame([""]*20, columns=[f"({shortcut}) {short_text}"]))
tables.append(df)
print("Top 20 most frequent word pairs - counts")
display(pd.concat(tables, axis=1))
Top 20 most frequent word pairs - counts
(Consecutive) CZ.txt | Pair | Count | (Surrounding) CZ.txt | Pair | Count | (Consecutive) EN.txt | Pair | Count | (Surrounding) EN.txt | Pair | Count | |
---|---|---|---|---|---|---|---|---|---|---|---|---|
0 | (,, že) | 1615 | (,, ,) | 135650 | (of, the) | 2770 | (,, ,) | 106898 | ||||
1 | (., ,) | 834 | (., .) | 118008 | (,, and) | 2004 | (the, ,) | 95306 | ||||
2 | (., .) | 774 | (., ,) | 117613 | (in, the) | 1387 | (,, the) | 95306 | ||||
3 | (,, ,) | 727 | (,, .) | 117613 | (the, same) | 938 | (the, the) | 91746 | ||||
4 | (., ") | 631 | (:, :) | 42618 | (to, the) | 891 | (,, of) | 67540 | ||||
5 | (,, a) | 520 | (:, ,) | 39492 | (;, and) | 769 | (of, ,) | 67540 | ||||
6 | (,, ale) | 452 | (,, :) | 39492 | (,, as) | 767 | (the, of) | 64310 | ||||
7 | ((, )) | 417 | (a, ,) | 39208 | (,, the) | 714 | (of, the) | 64310 | ||||
8 | (:, 1) | 412 | (,, a) | 39208 | (have, been) | 630 | (of, of) | 45186 | ||||
9 | (), ,) | 411 | (a, .) | 37133 | (the, of) | 629 | (., ,) | 41251 | ||||
10 | (., V) | 403 | (., a) | 37133 | (on, the) | 625 | (,, .) | 41251 | ||||
11 | (a, .) | 396 | (v, ,) | 34673 | (that, the) | 567 | (and, ,) | 38733 | ||||
12 | (., se) | 392 | (,, v) | 34673 | (., The) | 528 | (,, and) | 38733 | ||||
13 | (,, který) | 382 | (v, .) | 34066 | (,, which) | 527 | (., the) | 37235 | ||||
14 | (!, !) | 379 | (., v) | 34066 | (,, or) | 464 | (the, .) | 37235 | ||||
15 | (,, které) | 376 | (., :) | 32939 | (,, that) | 462 | (and, the) | 36513 | ||||
16 | (:, 0) | 342 | (:, .) | 32939 | (,, in) | 461 | (the, and) | 36513 | ||||
17 | (., :) | 329 | (,, se) | 30789 | (., ,) | 419 | (,, in) | 33828 | ||||
18 | (1, :) | 327 | (se, ,) | 30789 | (and, the) | 398 | (in, ,) | 33828 | ||||
19 | (", ") | 320 | (., se) | 28024 | (species, ,) | 396 | (,, to) | 32485 |
## compute the pointwise mutual information for the word pairs
results : Dict[str, Dict[str, Dict[Any, float]]] = {}
for short_text in texts:
results[short_text] = {}
for count_func, label in [
(count_word_pairs, "Consecutive"),
(count_word_surrounding_pairs, "Surrounding")
]:
source_text = cut_non_frequent_words(texts[short_text])
text_len = len(source_text)
word_prob = {word : count / text_len for word, count in Counter(source_text).items()}
pairs_prob = defaultdict(float)
word_pairs : Counter = count_func(source_text)
total_pairs = sum(word_pairs.values())
for pair, count in word_pairs.items():
pairs_prob[pair] = count / total_pairs
results[short_text][label] = {pair : log2(pairs_prob[pair] / (word_prob[pair[0]] * word_prob[pair[1]])) for pair in word_pairs}
## show the top n most significant word pairs
n = 20
tables_top = []
tables_worst = []
for short_text in results:
for label in results[short_text]:
df_top = pd.DataFrame(results[short_text][label].items(), columns=[f"Pair", f"PMI"]).nlargest(n, f"PMI").reset_index(drop=True)
tables_top.append(pd.DataFrame([""]*n, columns=[f"({label}) {short_text}"]))
tables_top.append(df_top)
df_worst = pd.DataFrame(results[short_text][label].items(), columns=[f"Pair", f"PMI"]).nsmallest(n, f"PMI").reset_index(drop=True)
tables_worst.append(pd.DataFrame([""]*n, columns=[f"({label}) {short_text}"]))
tables_worst.append(df_worst)
print(f"Top {n} most significant word pairs - PMI")
display(pd.concat(tables_top, axis=1))
print(f"Worst {n} most significant word pairs - PMI")
display(pd.concat(tables_worst, axis=1))
Top 20 most significant word pairs - PMI
(Consecutive) CZ.txt | Pair | PMI | (Surrounding) CZ.txt | Pair | PMI | (Consecutive) EN.txt | Pair | PMI | (Surrounding) EN.txt | Pair | PMI | |
---|---|---|---|---|---|---|---|---|---|---|---|---|
0 | (Hamburger, SV) | 13.712115 | (výher, výher) | 9.571579 | (La, Plata) | 14.042450 | (dried, floated) | 8.690953 | ||||
1 | (Los, Angeles) | 13.485607 | (žel, žel) | 8.974643 | (Asa, Gray) | 13.904947 | (floated, dried) | 8.690953 | ||||
2 | (Johna, Newcomba) | 13.186047 | (Sandžaku, Sandžaku) | 8.834613 | (Fritz, Muller) | 13.235095 | (germinated, dried) | 8.318984 | ||||
3 | (Č., Budějovice) | 13.056763 | (Petrof, Petrof) | 8.723582 | (worth, while) | 13.205949 | (dried, germinated) | 8.318984 | ||||
4 | (série, ATP) | 12.892133 | (CIA, CIA) | 8.628162 | (faced, tumbler) | 13.135560 | (floated, floated) | 8.316887 | ||||
5 | (turnajové, série) | 12.857576 | (IFS, IFS) | 8.401654 | (lowly, organised) | 13.089979 | (dried, dried) | 8.176380 | ||||
6 | (Tomáš, Ježek) | 12.852146 | (13h, 13h) | 8.401654 | (Malay, Archipelago) | 12.983557 | (germinated, floated) | 8.083964 | ||||
7 | (Lidové, noviny) | 12.753087 | (Bělehrad, Benfica) | 8.375181 | (shoulder, stripe) | 12.926973 | (floated, germinated) | 8.083964 | ||||
8 | (Lidových, novin) | 12.694193 | (Benfica, Bělehrad) | 8.375181 | (Great, Britain) | 12.787636 | (clover, clover) | 8.006455 | ||||
9 | (veřejného, mínění) | 12.485607 | (silniční, silniční) | 8.249650 | (THE, THE) | 12.779416 | (heath, heath) | 7.933930 | ||||
10 | (teplota, minus) | 12.404687 | (Divadlo, Divadlo) | 8.170699 | (United, States) | 12.720522 | (days, dried) | 7.927992 | ||||
11 | (Ján, Čarnogurský) | 12.378692 | (IV, výher) | 8.086152 | (deserves, notice) | 12.605045 | (dried, days) | 7.927992 | ||||
12 | (jaderné, zbraně) | 12.378692 | (výher, IV) | 8.086152 | (CHAPTER, THE) | 12.457488 | (floated, days) | 7.893582 | ||||
13 | (Milan, Máčala) | 12.320976 | (km, žel) | 8.064841 | (English, carrier) | 12.398594 | (days, floated) | 7.893582 | ||||
14 | (lidských, práv) | 12.286042 | (žel, km) | 8.064841 | (specially, endowed) | 12.274896 | (vibracula, vibracula) | 7.842956 | ||||
15 | (společném, státě) | 12.131599 | (13h, zataženo) | 7.962163 | (Sir, J) | 12.250443 | (Pacific, eastern) | 7.759900 | ||||
16 | (akciových, společností) | 12.115657 | (zataženo, 13h) | 7.962163 | (branched, off) | 12.250443 | (eastern, Pacific) | 7.759900 | ||||
17 | (Pohár, UEFA) | 12.048543 | (OKD, OKD) | 7.927722 | (de, Candolle) | 12.235095 | (masters, slaves) | 7.758308 | ||||
18 | (privatizačních, projektů) | 12.038842 | (13h, st) | 7.886620 | (mental, qualities) | 12.235095 | (slaves, masters) | 7.758308 | ||||
19 | (George, Bushe) | 12.026175 | (st, 13h) | 7.886620 | (Galapagos, Archipelago) | 12.218022 | (stripe, shoulder) | 7.753014 |
Worst 20 most significant word pairs - PMI
(Consecutive) CZ.txt | Pair | PMI | (Surrounding) CZ.txt | Pair | PMI | (Consecutive) EN.txt | Pair | PMI | (Surrounding) EN.txt | Pair | PMI | |
---|---|---|---|---|---|---|---|---|---|---|---|---|
0 | (., že) | -6.200420 | (jednání, 3) | -6.274230 | (the, a) | -7.684486 | (forms, instincts) | -4.335294 | ||||
1 | (:, že) | -5.287548 | (3, jednání) | -6.274230 | (a, the) | -7.684486 | (instincts, forms) | -4.335294 | ||||
2 | (že, :) | -5.287548 | (že, =) | -6.243805 | (in, in) | -6.806685 | (colour, will) | -4.308060 | ||||
3 | (:, s) | -4.742022 | (=, že) | -6.243805 | (the, the) | -6.311216 | (will, colour) | -4.308060 | ||||
4 | (a, 1) | -4.728166 | (nebo, 0) | -6.109001 | (the, it) | -6.296825 | (fish, all) | -4.235751 | ||||
5 | (,, Na) | -4.609473 | (0, nebo) | -6.109001 | (the, this) | -6.223902 | (all, fish) | -4.235751 | ||||
6 | (Na, .) | -4.516894 | (2, strany) | -5.888557 | (same, the) | -6.053185 | (forms, food) | -4.221338 | ||||
7 | (:, i) | -4.503608 | (strany, 2) | -5.888557 | (their, the) | -5.949668 | (food, forms) | -4.221338 | ||||
8 | (!, .) | -4.478124 | (3, vlády) | -5.842347 | (., be) | -5.852153 | (wax, species) | -4.176287 | ||||
9 | (., co) | -4.432502 | (vlády, 3) | -5.842347 | (., which) | -5.618343 | (species, wax) | -4.176287 | ||||
10 | (a, 2) | -4.360909 | (této, 0) | -5.703182 | (of, we) | -5.613237 | (Europe, selection) | -4.163019 | ||||
11 | ((, () | -4.356932 | (0, této) | -5.703182 | (;, to) | -5.594396 | (selection, Europe) | -4.163019 | ||||
12 | (se, 1) | -4.318905 | (3, státu) | -5.623178 | (of, be) | -5.582921 | (comb, species) | -4.105898 | ||||
13 | ((, že) | -4.313363 | (státu, 3) | -5.623178 | (., by) | -5.569208 | (species, comb) | -4.105898 | ||||
14 | (v, by) | -4.206675 | (2, korun) | -5.585630 | (we, ,) | -5.265299 | (forms, side) | -4.097602 | ||||
15 | (a, V) | -4.197833 | (korun, 2) | -5.585630 | (many, the) | -5.234009 | (side, forms) | -4.097602 | ||||
16 | (:, k) | -4.129483 | (kteří, 4) | -5.572841 | (has, of) | -5.226108 | (flying, species) | -4.031897 | ||||
17 | (kteří, .) | -4.109124 | (4, kteří) | -5.572841 | (this, the) | -5.223902 | (species, flying) | -4.031897 | ||||
18 | (se, je) | -4.044692 | (4, u) | -5.558268 | (the, any) | -5.218822 | (early, no) | -4.027751 | ||||
19 | (., roku) | -4.011659 | (u, 4) | -5.558268 | (at, of) | -5.208164 | (no, early) | -4.027751 |
Intepretation¶
We can see from the obtained results that we got both negative and positive values. That illustrates well the following:
"Good collocation pairs have high PMI because the probability of co-occurrence is only slightly lower than the probabilities of occurrence of each word. Conversely, a pair of words whose probabilities of occurrence are considerably higher than their probability of co-occurrence gets a small PMI score."
Especially, pairs like "Lidové noviny" in Czech or "United States" in English represent very strong collocations used often with the second word (that is why their PMI is very high), while pairs like "na, se" in Czech or "to, and" in English are all but collocations (and that is why their PMI is very low).
We can also see PMI scores closer to 0 for surrounding pairs (compared to the neighbouring pairs) in the TOP chart. This might be because of the fact that the farther away the second word is, the less dependent (relatively to the first word in the text) the occurence of the second word is. And we know that the PMI is equal 0 if both X and Y are independant.
Word Classes¶
Compute a full class hierarchy of words using the first 8,000 words of those data, and only for words occurring 10 times or more (use the same setting for both languages). Ignore the other words for building the classes, but keep them in the data for the bigram counts and all the formulas that use them (including the Mutual Information, the interim sums in the "Tricks", etc.). For details on the algorithm, use the Brown et al. paper available form SIS; some formulas are wrong in the paper, however, so please see the corrections in the slides (formulas for Trick #4). Note the history of the merges, and attach it to your homework. Now run the same algorithm again, but stop when reaching 15 classes. Print out all the members of your 15 classes and attach them too.
Text preparation¶
from typing import List, Dict, Tuple, Set
from collections import Counter
import numpy as np
import math
import itertools
import pandas as pd
separator = '/'
## load the text files
def load_text(filename, dir) -> str:
with open(f"../texts/{dir}/{filename}", "r", encoding="utf-8") as f:
return f.read()
def load_ptg_text(text : str) -> List[str]:
return text.split('\n')
def load_word_and_pos(ptg_text : List[str]) -> List[Tuple[str, str]]:
return [(word.split(separator)[0], "".join(word.split(separator)[1:]))
for word in ptg_text]
def separate_words_and_pos(ptg_text : List[Tuple[str, str]]) -> Tuple[List[str], List[str]]:
return [word for word, _ in ptg_text], [pos for _, pos in ptg_text]
def get_unique_words(text : List[str]) -> Set[str]:
return set(text)
Implementing merging classes with tricks from lecture¶
## TRICK 1 - bigram counts
def get_word_counts(text : List[str]) -> Counter:
return Counter(text)
def get_bigram_counts(text : List[str]) -> Counter:
return Counter(zip(text, text[1:]))
def get_left_unigram_counts(bigram_counter : Counter) -> Counter:
unigram_counter = Counter()
for (word1, word2), count in bigram_counter.items():
unigram_counter[word1] += count
return unigram_counter
def get_right_unigram_counts(bigram_counter : Counter) -> Counter:
unigram_counter = Counter()
for (word1, word2), count in bigram_counter.items():
unigram_counter[word2] += count
return unigram_counter
def total_number_of_words(text : List[str]) -> int:
return len(text)
def total_number_of_unique_words(text : List[str]) -> int:
return len(get_unique_words(text))
def arange_classes(unique_words : Set[str], FREQ_THRESH : int = 1, counter : Counter = None) -> np.ndarray:
if (FREQ_THRESH > 1 and counter is not None):
## only words with frequency higher than K are considered
return np.array([i for i, word in enumerate(unique_words) if counter[word] >= FREQ_THRESH], dtype=np.int64)
else:
## all words are considered
return np.arange(len(unique_words), dtype=np.int64)
def create_bigram_matrix(unique_words : List[str], bigram_counter : Counter, word_to_class_mapping : Dict[str, int]) -> np.ndarray:
C = len(unique_words)
bigram_matrix = np.zeros((C, C), dtype=np.int64)
for (word1, word2), count in bigram_counter.items():
bigram_matrix[word_to_class_mapping[word1], word_to_class_mapping[word2]] = count
return bigram_matrix
def create_unigram_matrix(unique_words : List[str], unigram_counter : Counter, word_to_class_mapping : Dict[str, int]) -> np.ndarray:
C = len(unique_words)
unigram_matrix = np.zeros(C, dtype=np.int64)
for word, count in unigram_counter.items():
unigram_matrix[word_to_class_mapping[word]] = count
return unigram_matrix
def create_word_to_class_mapping(unique_words : List[str]) -> Dict[str, int]:
return {word : i for i, word in enumerate(unique_words)}
def create_class_to_word_mapping(unique_words : List[str]) -> Dict[int, List[str]]:
return {i : [word] for i, word in enumerate(unique_words)}
## TRICK 2 - precompute q and s
def q(c_bi, c_l, c_r, N):
if (c_l == 0 or c_r == 0 or c_bi == 0):
return 0
return (c_bi/N)*np.log2((N*c_bi)/(c_l*c_r))
def precompute_q_and_s(classes, c_bi, c_l, c_r, C, N, verbose : bool = False):
if verbose:
print(f"Computing 'Q'...")
Q = np.zeros((C,C), dtype=float)
for l in range(C):
for r in range(C):
Q[l, r] = q(c_bi[l, r], c_l[l], c_r[r], N)
if verbose:
print(f"Computing 'S'...")
S = np.zeros((C), dtype=float)
for cls in classes:
S[cls] = (np.sum(Q[:, cls]) + np.sum(Q[cls,:]) - Q[cls, cls])
return Q, S
## TRICK 3: sub and add
def sub(S, Q, a : int, b : int):
return (S[a] + S[b] - Q[a, b] - Q[b, a])
def add(counts_current : tuple["np.ndarray", "np.ndarray", "np.ndarray"], N, C, a : int, b : int):
"""'c' is the class into which was 'a' and 'b' merged. The rule is to merge 'b' into 'a'."""
c_bi, c_l, c_r = counts_current
sums = np.sum([
q(c_bi[x, a] + c_bi[x, b], c_l[x], c_r[a] + c_r[b], N) ## q(l, a+b)
+ q(c_bi[a, x] + c_bi[b, x], c_l[a] + c_l[b], c_r[x], N) ## q(a+b, r)
for x in range(C) if (x != a and x != b)
])
merged = q(c_bi[a, a] + c_bi[a, b] + c_bi[b, a] + c_bi[b, b], c_l[a] + c_l[b], c_r[a] + c_r[b], N) ## q(a+b, a+b)
return (sums + merged)
## TRICK 4 - table of losses
def init_l(Q, S, N, C, counts_current : tuple["np.ndarray", "np.ndarray", "np.ndarray"], classes, verbose : bool = False):
if verbose:
print(f"Init of 'l'...")
## create table of losses for each possible pair (upper right triangle)
L = np.zeros((C,C), dtype=float)
for cls_one, cls_two in itertools.combinations(classes, 2):
## calculate the loss of merging these two classes
L[cls_one, cls_two] = sub(S, Q, cls_one, cls_two) - add(counts_current, N, C, cls_one, cls_two)
L[cls_two, cls_one] = L[cls_one, cls_two] ## because of symmetry
return L
def find_best_pair(L, classes):
minimal_loss = math.inf
best_pair = (-1, -1)
for i, j in itertools.combinations(classes, 2):
if L[i, j] < minimal_loss:
minimal_loss = L[i, j]
best_pair = (i, j)
return best_pair, minimal_loss
def mergeCounts(current_counts : tuple["np.ndarray", "np.ndarray", "np.ndarray"], a : int, b : int) -> tuple["np.ndarray", "np.ndarray", "np.ndarray"]:
"""Merges the class 'b' into the class 'a'. Adjusts the counts."""
c_bi, c_l, c_r = current_counts
c_bi_f = c_bi.copy()
c_l_f = c_l.copy()
c_r_f = c_r.copy()
## merge the counts
c_bi_f[a, :] += c_bi_f[b, :]
c_bi_f[:, a] += c_bi_f[:, b]
c_l_f[a] += c_l_f[b]
c_r_f[a] += c_r_f[b]
## zero the merged class
c_bi_f[b, :] = np.zeros(c_bi_f.shape[0])
c_bi_f[:, b] = np.zeros(c_bi_f.shape[0])
c_l_f[b] = 0
c_r_f[b] = 0
return c_bi_f, c_l_f, c_r_f
def update(L, S, counts_current : tuple["np.ndarray", "np.ndarray", "np.ndarray"], counts_future : tuple["np.ndarray", "np.ndarray", "np.ndarray"], classes, N, a : int, b : int):
c_bi, c_l, c_r = counts_current
c_bi_f, c_l_f, c_r_f = counts_future
S_f = S.copy()
for i in classes:
if i == b:
continue
S_f[i] = (S[i]
- q(c_bi[i, a], c_l[i], c_r[a], N) ## q(i, a)
- q(c_bi[a, i], c_l[a], c_r[i], N) ## q(a, i)
- q(c_bi[i, b], c_l[i], c_r[b], N) ## q(i, b)
- q(c_bi[b, i], c_l[b], c_r[i], N) ## q(b, i)
+ q(c_bi_f[a, i], c_l_f[a], c_r_f[i], N) ## q(a, i)
+ q(c_bi_f[i, a], c_l_f[i], c_r_f[a], N)) ## q(i, a)
## treated separately - like init
S_f[a] = (np.sum([q(c_bi_f[x, a], c_l_f[x], c_r_f[a], N) for x in range(c_bi.shape[0])]) + np.sum([q(c_bi_f[a, x], c_l_f[a], c_r_f[x], N) for x in range(c_bi.shape[0])]) - q(c_bi_f[a, a], c_l_f[a], c_r_f[a], N))
L_f = L.copy()
for i, j in itertools.combinations(classes, 2):
if (i == a and j not in [a, b]):
## treated separately - like init
L_f[i, j] = (S_f[i]
+ S_f[j]
- q(c_bi_f[i, j], c_l_f[i], c_r_f[j], N)
- q(c_bi_f[j, i], c_l_f[j], c_r_f[i], N)
- q(c_bi_f[i, i] + c_bi_f[i, j] + c_bi_f[j, i] + c_bi_f[j, j], c_l_f[i] + c_l_f[j], c_r_f[i] + c_r_f[j], N)
- np.sum([q(c_bi_f[x, i] + c_bi_f[x, j], c_l_f[x], c_r_f[i] + c_r_f[j], N) for x in range(c_bi.shape[0]) if x not in [i, j]])
- np.sum([q(c_bi_f[i, x] + c_bi_f[j, x], c_l_f[i] + c_l_f[j], c_r_f[x], N) for x in range(c_bi.shape[0]) if x not in [i, j]]))
else:
L_f[i, j] = (L[i, j]
- S[i]
+ S_f[i]
- S[j]
+ S_f[j]
+ q(c_bi[i, a]+c_bi[j, a], c_l[i]+c_l[j], c_r[a], N)
+ q(c_bi[a, i]+c_bi[a, j], c_l[a], c_r[i]+c_r[j], N)
+ q(c_bi[i, b]+c_bi[j, b], c_l[i]+c_l[j], c_r[b], N)
+ q(c_bi[b, i]+c_bi[b, j], c_l[b], c_r[i]+c_r[j], N)
- q(c_bi_f[i, a]+c_bi_f[j, a], c_l_f[i]+c_l_f[j], c_r_f[a], N)
- q(c_bi_f[a, i]+c_bi_f[a, j], c_l_f[a], c_r_f[i]+c_r_f[j], N))
L_f[j, i] = L_f[i, j] ## because of symmetry
return L_f, S_f
source_files = {
"en" : "EN.txt",
"cz" : "CZ.txt"
}
dir = "ptg"
def execute_merging(
lang : str,
mode : str,
LIMIT : int,
MIN_FREQ : int,
K : int,
verbose : bool = False
) -> List[List[str]]:
'''Returns a list of classes, each class is a list of words that were merged together.'''
if verbose:
print(f"Merging language: {lang}, mode: {mode}")
## INITIALIZATION
text = load_text(source_files[lang], dir)
ptgs = load_ptg_text(text)
## TAKE FIRST K WORDS
ptgs = ptgs[:K]
words_and_pos = load_word_and_pos(ptgs)
if (mode == "words"):
words, pos = separate_words_and_pos(words_and_pos)
elif (mode == "pos"):
pos, words = separate_words_and_pos(words_and_pos)
else:
raise ValueError("Invalid mode")
N = total_number_of_words(words)
C = total_number_of_unique_words(words)
unique_words = get_unique_words(words)
classes = arange_classes(unique_words, MIN_FREQ, get_word_counts(words))
class_to_word = create_class_to_word_mapping(unique_words)
word_to_class = create_word_to_class_mapping(unique_words)
bigram_counts = get_bigram_counts(words)
c_bi = create_bigram_matrix(unique_words, bigram_counts, word_to_class)
c_l = create_unigram_matrix(unique_words, get_left_unigram_counts(bigram_counts), word_to_class)
c_r = create_unigram_matrix(unique_words, get_right_unigram_counts(bigram_counts), word_to_class)
Q, S = precompute_q_and_s(classes, c_bi, c_l, c_r, C, N)
if verbose:
print(f"Sum of the Q: {np.sum(Q)}")
counts_current = (c_bi, c_l, c_r)
## MAIN LOOP
L = init_l(Q, S, N, C, counts_current, classes)
history = []
human_readable_history = []
while len(classes) > LIMIT:
## perform the update step
## find the best pair to merge
best_pair, minimal_loss = find_best_pair(L, classes)
## log
if verbose:
print(f"{class_to_word[best_pair[1]]} into {class_to_word[best_pair[0]]}, loss: {minimal_loss}")
## ajdust the counts
counts_future = mergeCounts(counts_current, best_pair[0], best_pair[1])
## update the loss table, pick the best pair to merge
L, S = update(L, S, counts_current, counts_future, classes, N, best_pair[0], best_pair[1])
## update the history
history.append(best_pair)
human_readable_history.append((class_to_word[best_pair[0]].copy(), class_to_word[best_pair[1]].copy()))
## update the class_to_word mapping
class_to_word[best_pair[0]] += class_to_word[best_pair[1]]
class_to_word.pop(best_pair[1])
## update the word_to_class mapping
for word, cls in word_to_class.items():
if cls == best_pair[1]:
word_to_class[word] = best_pair[0]
## update the classes
classes = np.delete(classes, np.where(classes == best_pair[1]))
## update the counts
counts_current = counts_future
if verbose:
for i, (cls_one, cls_two) in enumerate(human_readable_history):
print(f"{i+1}: {cls_two} into {cls_one}")
return [class_to_word[cls] for cls in classes], human_readable_history
for mode in ["words", "pos"]:
if mode == "words":
LIMIT = 15
MIN_FREQ = 10
K = 8000
elif mode == "pos":
LIMIT = 10
MIN_FREQ = 5
K = 200000
for lang in ["en", "cz"]:
print(f"""Language: {lang}.
Merging {mode} into {LIMIT} classes.
Considering first {K} words with min. freq. {MIN_FREQ}...\n""")
classes, history_of_merges = execute_merging(lang, mode, LIMIT, MIN_FREQ, K, verbose=True)
for i, cls in enumerate(classes):
print(f"Class {i+1}: {cls}")
print("\n")
## save history to a file
with open(f"../results/{lang}_{mode}_{LIMIT}_history.txt", "w", encoding="utf-8") as f:
for cls_one, cls_two in history_of_merges:
f.write(f"{cls_two} into {cls_one}\n")
Language: en. Merging words into 15 classes. Considering first 8000 words with min. freq. 10... Merging language: en, mode: words Sum of the Q: 4.995892538589587 ['case'] into ['subject'], loss: 0.0021965665335756945 ['may'] into ['cannot'], loss: 0.002669139511099379 ['structure'] into ['individuals'], loss: 0.0026748091526128705 ['It'] into ['there'], loss: 0.003479400370452433 ['even'] into ['less'], loss: 0.0036556390622295674 ['variation'] into ['nature'], loss: 0.0036909506202147946 ['see'] into ['shall'], loss: 0.0038977958561413565 ['slight'] into ['short'], loss: 0.0039056390622295677 ['distinct'] into ['certain'], loss: 0.003992156793911794 ['manner'] into ['such'], loss: 0.004241409281117521 ['must'] into ['can'], loss: 0.004262927799019989 ['state'] into ['subject', 'case'], loss: 0.0042766639487693235 ['differ'] into ['what'], loss: 0.0042987949406953965 [')'] into ['cases'], loss: 0.004456435556800399 ['me'] into ['only'], loss: 0.00447802686684627 ['how'] into ['nearly'], loss: 0.004540204221812921 ['domesticated'] into ['domestic'], loss: 0.004563547053784825 ['races'] into ['varieties'], loss: 0.004613365915588918 ['if'] into ['when'], loss: 0.004762832413873117 ['than'] into ['('], loss: 0.0048728052253528276 ['do'] into ['believe'], loss: 0.0049492035054291604 ['great'] into ['my'], loss: 0.005125548031345392 ['could'] into ['will'], loss: 0.005600648386150989 ['plants'] into ['facts'], loss: 0.0056896399601957266 ['between'] into [':'], loss: 0.005755267783733755 ['its'] into ['different'], loss: 0.005811278124459137 ['conditions'] into ['nature', 'variation'], loss: 0.005854956030029314 ['In'] into ['The'], loss: 0.005975681770144466 ['long'] into ['short', 'slight'], loss: 0.0062716868916984116 ['these'] into ['each'], loss: 0.0063407179814485605 ['would'] into ['cannot', 'may'], loss: 0.006573989711723835 ['most'] into ['individuals', 'structure'], loss: 0.006710581835738694 ['very'] into ['any'], loss: 0.006827542130200615 ['so'] into ['often'], loss: 0.007051575444189467 ['we'] into ['they'], loss: 0.007132892772609472 ['breeds'] into ['subject', 'case', 'state'], loss: 0.007212068207649337 ['wild'] into ['certain', 'distinct'], loss: 0.007777689674487167 ['less', 'even'] into ['such', 'manner'], loss: 0.008128809562529237 ['it'] into ['there', 'It'], loss: 0.00814083806206068 ['an'] into ['much'], loss: 0.008244846816464119 ['can', 'must'] into ['cannot', 'may', 'would'], loss: 0.00836379129755873 ['animals'] into ['varieties', 'races'], loss: 0.008558816507676306 ['many'] into ['several'], loss: 0.008909921301797073 ['all'] into ['what', 'differ'], loss: 0.009227791157599648 ['only', 'me'] into ['believe', 'do'], loss: 0.009411342758844692 ['but'] into ['(', 'than'], loss: 0.00956669905415268 ['cases', ')'] into ['nearly', 'how'], loss: 0.00965124016553962 ['other'] into ['one'], loss: 0.009719360382467134 ['at'] into [':', 'between'], loss: 0.01028631175693167 ['some'] into ['their'], loss: 0.010690608987174686 ['more'] into ['shall', 'see'], loss: 0.011028134299625523 ['as'] into ['when', 'if'], loss: 0.011158014327088295 ['nature', 'variation', 'conditions'] into ['individuals', 'structure', 'most'], loss: 0.01133854228576027 ['different', 'its'] into ['much', 'an'], loss: 0.012094766529275517 ['my', 'great'] into ['short', 'slight', 'long'], loss: 0.012354424064353053 ['for'] into ['under'], loss: 0.012517483448205951 ['each', 'these'] into ['their', 'some'], loss: 0.012825038255811619 ['species'] into ['varieties', 'races', 'animals'], loss: 0.012974398391393102 ['will', 'could'] into ['cannot', 'may', 'would', 'can', 'must'], loss: 0.012986465443803719 ['our'] into ['several', 'many'], loss: 0.01348182767925122 ['any', 'very'] into ['what', 'differ', 'all'], loss: 0.013459339217764085 ['facts', 'plants'] into ['such', 'manner', 'less', 'even'], loss: 0.013745183113729156 ['with'] into ['often', 'so'], loss: 0.01406037032451422 ['certain', 'distinct', 'wild'] into ['one', 'other'], loss: 0.015344402398685454 ['I'] into ['they', 'we'], loss: 0.01563365339577806 ['has'] into ['not'], loss: 0.015973041777880122 ['same'] into ['subject', 'case', 'state', 'breeds'], loss: 0.01649447587886521 ['domestic', 'domesticated'] into ['one', 'other', 'certain', 'distinct', 'wild'], loss: 0.016525149034757675 ['nearly', 'how', 'cases', ')'] into ['shall', 'see', 'more'], loss: 0.017458719916070224 ['by'] into ['on'], loss: 0.017699186641382918 ['which'] into ['this'], loss: 0.017779702329765442 ['believe', 'do', 'only', 'me'] into ['(', 'than', 'but'], loss: 0.01828753914667661 ['been'] into [':', 'between', 'at'], loss: 0.020305525538123863 ['or'] into ['under', 'for'], loss: 0.022322604123214074 ['from'] into ['often', 'so', 'with'], loss: 0.022452873285990185 ['is'] into ['are'], loss: 0.02330965397012181 ['varieties', 'races', 'animals', 'species'] into ['such', 'manner', 'less', 'even', 'facts', 'plants'], loss: 0.02332631429369478 ['The', 'In'] into ['there', 'It', 'it'], loss: 0.023791644897939068 ['have'] into ['not', 'has'], loss: 0.024063043420479536 ['much', 'an', 'different', 'its'] into ['their', 'some', 'each', 'these'], loss: 0.02505109884743363 ['short', 'slight', 'long', 'my', 'great'] into ['what', 'differ', 'all', 'any', 'very'], loss: 0.025710589246418065 ['they', 'we', 'I'] into ['there', 'It', 'it', 'The', 'In'], loss: 0.0265967329747308 ['that'] into ['when', 'if', 'as'], loss: 0.0266400300416021 ['a'] into ['several', 'many', 'our'], loss: 0.029531699827928234 ['be'] into ['(', 'than', 'but', 'believe', 'do', 'only', 'me'], loss: 0.03284669379207519 ['shall', 'see', 'more', 'nearly', 'how', 'cases', ')'] into ['what', 'differ', 'all', 'any', 'very', 'short', 'slight', 'long', 'my', 'great'], loss: 0.033760944016454436 [':', 'between', 'at', 'been'] into ['often', 'so', 'with', 'from'], loss: 0.03516458858913665 ['in'] into ['on', 'by'], loss: 0.03601025146728219 ['subject', 'case', 'state', 'breeds', 'same'] into ['individuals', 'structure', 'most', 'nature', 'variation', 'conditions'], loss: 0.03974470674006333 ['this', 'which'] into ['their', 'some', 'each', 'these', 'much', 'an', 'different', 'its'], loss: 0.039877009077549604 ['such', 'manner', 'less', 'even', 'facts', 'plants', 'varieties', 'races', 'animals', 'species'] into ['individuals', 'structure', 'most', 'nature', 'variation', 'conditions', 'subject', 'case', 'state', 'breeds', 'same'], loss: 0.04000412379498797 [';'] into ['cannot', 'may', 'would', 'can', 'must', 'will', 'could'], loss: 0.040100532347277756 ['under', 'for', 'or'] into ['are', 'is'], loss: 0.042847519208385344 ['not', 'has', 'have'] into ['(', 'than', 'but', 'believe', 'do', 'only', 'me', 'be'], loss: 0.04414943404587341 ['one', 'other', 'certain', 'distinct', 'wild', 'domestic', 'domesticated'] into ['several', 'many', 'our', 'a'], loss: 0.04510962278374156 ['and'] into ['when', 'if', 'as', 'that'], loss: 0.053149087467073634 ['what', 'differ', 'all', 'any', 'very', 'short', 'slight', 'long', 'my', 'great', 'shall', 'see', 'more', 'nearly', 'how', 'cases', ')'] into ['their', 'some', 'each', 'these', 'much', 'an', 'different', 'its', 'this', 'which'], loss: 0.06085473265433029 1: ['case'] into ['subject'] 2: ['may'] into ['cannot'] 3: ['structure'] into ['individuals'] 4: ['It'] into ['there'] 5: ['even'] into ['less'] 6: ['variation'] into ['nature'] 7: ['see'] into ['shall'] 8: ['slight'] into ['short'] 9: ['distinct'] into ['certain'] 10: ['manner'] into ['such'] 11: ['must'] into ['can'] 12: ['state'] into ['subject', 'case'] 13: ['differ'] into ['what'] 14: [')'] into ['cases'] 15: ['me'] into ['only'] 16: ['how'] into ['nearly'] 17: ['domesticated'] into ['domestic'] 18: ['races'] into ['varieties'] 19: ['if'] into ['when'] 20: ['than'] into ['('] 21: ['do'] into ['believe'] 22: ['great'] into ['my'] 23: ['could'] into ['will'] 24: ['plants'] into ['facts'] 25: ['between'] into [':'] 26: ['its'] into ['different'] 27: ['conditions'] into ['nature', 'variation'] 28: ['In'] into ['The'] 29: ['long'] into ['short', 'slight'] 30: ['these'] into ['each'] 31: ['would'] into ['cannot', 'may'] 32: ['most'] into ['individuals', 'structure'] 33: ['very'] into ['any'] 34: ['so'] into ['often'] 35: ['we'] into ['they'] 36: ['breeds'] into ['subject', 'case', 'state'] 37: ['wild'] into ['certain', 'distinct'] 38: ['less', 'even'] into ['such', 'manner'] 39: ['it'] into ['there', 'It'] 40: ['an'] into ['much'] 41: ['can', 'must'] into ['cannot', 'may', 'would'] 42: ['animals'] into ['varieties', 'races'] 43: ['many'] into ['several'] 44: ['all'] into ['what', 'differ'] 45: ['only', 'me'] into ['believe', 'do'] 46: ['but'] into ['(', 'than'] 47: ['cases', ')'] into ['nearly', 'how'] 48: ['other'] into ['one'] 49: ['at'] into [':', 'between'] 50: ['some'] into ['their'] 51: ['more'] into ['shall', 'see'] 52: ['as'] into ['when', 'if'] 53: ['nature', 'variation', 'conditions'] into ['individuals', 'structure', 'most'] 54: ['different', 'its'] into ['much', 'an'] 55: ['my', 'great'] into ['short', 'slight', 'long'] 56: ['for'] into ['under'] 57: ['each', 'these'] into ['their', 'some'] 58: ['species'] into ['varieties', 'races', 'animals'] 59: ['will', 'could'] into ['cannot', 'may', 'would', 'can', 'must'] 60: ['our'] into ['several', 'many'] 61: ['any', 'very'] into ['what', 'differ', 'all'] 62: ['facts', 'plants'] into ['such', 'manner', 'less', 'even'] 63: ['with'] into ['often', 'so'] 64: ['certain', 'distinct', 'wild'] into ['one', 'other'] 65: ['I'] into ['they', 'we'] 66: ['has'] into ['not'] 67: ['same'] into ['subject', 'case', 'state', 'breeds'] 68: ['domestic', 'domesticated'] into ['one', 'other', 'certain', 'distinct', 'wild'] 69: ['nearly', 'how', 'cases', ')'] into ['shall', 'see', 'more'] 70: ['by'] into ['on'] 71: ['which'] into ['this'] 72: ['believe', 'do', 'only', 'me'] into ['(', 'than', 'but'] 73: ['been'] into [':', 'between', 'at'] 74: ['or'] into ['under', 'for'] 75: ['from'] into ['often', 'so', 'with'] 76: ['is'] into ['are'] 77: ['varieties', 'races', 'animals', 'species'] into ['such', 'manner', 'less', 'even', 'facts', 'plants'] 78: ['The', 'In'] into ['there', 'It', 'it'] 79: ['have'] into ['not', 'has'] 80: ['much', 'an', 'different', 'its'] into ['their', 'some', 'each', 'these'] 81: ['short', 'slight', 'long', 'my', 'great'] into ['what', 'differ', 'all', 'any', 'very'] 82: ['they', 'we', 'I'] into ['there', 'It', 'it', 'The', 'In'] 83: ['that'] into ['when', 'if', 'as'] 84: ['a'] into ['several', 'many', 'our'] 85: ['be'] into ['(', 'than', 'but', 'believe', 'do', 'only', 'me'] 86: ['shall', 'see', 'more', 'nearly', 'how', 'cases', ')'] into ['what', 'differ', 'all', 'any', 'very', 'short', 'slight', 'long', 'my', 'great'] 87: [':', 'between', 'at', 'been'] into ['often', 'so', 'with', 'from'] 88: ['in'] into ['on', 'by'] 89: ['subject', 'case', 'state', 'breeds', 'same'] into ['individuals', 'structure', 'most', 'nature', 'variation', 'conditions'] 90: ['this', 'which'] into ['their', 'some', 'each', 'these', 'much', 'an', 'different', 'its'] 91: ['such', 'manner', 'less', 'even', 'facts', 'plants', 'varieties', 'races', 'animals', 'species'] into ['individuals', 'structure', 'most', 'nature', 'variation', 'conditions', 'subject', 'case', 'state', 'breeds', 'same'] 92: [';'] into ['cannot', 'may', 'would', 'can', 'must', 'will', 'could'] 93: ['under', 'for', 'or'] into ['are', 'is'] 94: ['not', 'has', 'have'] into ['(', 'than', 'but', 'believe', 'do', 'only', 'me', 'be'] 95: ['one', 'other', 'certain', 'distinct', 'wild', 'domestic', 'domesticated'] into ['several', 'many', 'our', 'a'] 96: ['and'] into ['when', 'if', 'as', 'that'] 97: ['what', 'differ', 'all', 'any', 'very', 'short', 'slight', 'long', 'my', 'great', 'shall', 'see', 'more', 'nearly', 'how', 'cases', ')'] into ['their', 'some', 'each', 'these', 'much', 'an', 'different', 'its', 'this', 'which'] Class 1: ['cannot', 'may', 'would', 'can', 'must', 'will', 'could', ';'] Class 2: ['individuals', 'structure', 'most', 'nature', 'variation', 'conditions', 'subject', 'case', 'state', 'breeds', 'same', 'such', 'manner', 'less', 'even', 'facts', 'plants', 'varieties', 'races', 'animals', 'species'] Class 3: ['when', 'if', 'as', 'that', 'and'] Class 4: ['are', 'is', 'under', 'for', 'or'] Class 5: ['their', 'some', 'each', 'these', 'much', 'an', 'different', 'its', 'this', 'which', 'what', 'differ', 'all', 'any', 'very', 'short', 'slight', 'long', 'my', 'great', 'shall', 'see', 'more', 'nearly', 'how', 'cases', ')'] Class 6: ['several', 'many', 'our', 'a', 'one', 'other', 'certain', 'distinct', 'wild', 'domestic', 'domesticated'] Class 7: ['on', 'by', 'in'] Class 8: ['often', 'so', 'with', 'from', ':', 'between', 'at', 'been'] Class 9: ['of'] Class 10: ['there', 'It', 'it', 'The', 'In', 'they', 'we', 'I'] Class 11: ['(', 'than', 'but', 'believe', 'do', 'only', 'me', 'be', 'not', 'has', 'have'] Class 12: ['to'] Class 13: ['the'] Class 14: ['.'] Class 15: [','] Language: cz. Merging words into 15 classes. Considering first 8000 words with min. freq. 10... Merging language: cz, mode: words Sum of the Q: 7.557199800862522 ['listopadu'] into ['OKD'], loss: 0.003082884270599922 ['který'] into ['které'], loss: 0.0033734316421669547 ['státu'] into ['J'], loss: 0.004024966367527468 ['bude'] into ['musí'], loss: 0.0044216001660482285 ['aby'] into ['ale'], loss: 0.004603782679326707 ['nás'] into ['bylo'], loss: 0.004647048343347073 ['si'] into ['pouze'], loss: 0.004991409281117513 ['už'] into ['mezi'], loss: 0.005000000000000001 ['&slash;'] into ['NATO'], loss: 0.00556127812445914 ['ze'] into ['jeho'], loss: 0.00556426990688624 ['však'] into ['být'], loss: 0.005742784599970021 ['byl'] into ['jsou'], loss: 0.005749999999999991 ['před'] into ['?'], loss: 0.0059999999999999915 ['budou'] into ['jako'], loss: 0.0062500000000000056 ['při'] into ['zákona'], loss: 0.006899419316170906 ['V'] into ['Na'], loss: 0.007058395693542378 ['OKD', 'listopadu'] into ['J', 'státu'], loss: 0.007106558562877181 ['za'] into ['u'], loss: 0.007555812533852423 ['včera'] into ['bylo', 'nás'], loss: 0.008115293978715554 ['ČSFR'] into ['od'], loss: 0.00872995224880984 ['že'] into ['které', 'který'], loss: 0.00919518545632083 ['pouze', 'si'] into ['mezi', 'už'], loss: 0.010057083865413086 ['pro'] into ['po'], loss: 0.010249999999999981 ['musí', 'bude'] into ['?', 'před'], loss: 0.010746796771587318 ['ale', 'aby'] into ['NATO', '&slash;'], loss: 0.011340652565987524 ['k'] into ['jeho', 'ze'], loss: 0.011537365052878796 ['být', 'však'] into ['jako', 'budou'], loss: 0.011652102818357001 ['jsou', 'byl'] into ['zákona', 'při'], loss: 0.012310500446943484 ['ve'] into ['to'], loss: 0.013698431273825466 ['by'] into ['bylo', 'nás', 'včera'], loss: 0.014746339632353445 ['Na', 'V'] into ['J', 'státu', 'OKD', 'listopadu'], loss: 0.015035659754277397 ['do'] into ['u', 'za'], loss: 0.015907961185191034 ['je'] into ['i'], loss: 0.016172407347170403 [':'] into ['od', 'ČSFR'], loss: 0.017749999999999985 ['které', 'který', 'že'] into ['NATO', '&slash;', 'ale', 'aby'], loss: 0.01939815642260082 ['mezi', 'už', 'pouze', 'si'] into ['po', 'pro'], loss: 0.01984223927847203 [')'] into ['z'], loss: 0.02215279953493207 ['('] into ['?', 'před', 'musí', 'bude'], loss: 0.022553603635574095 ['jako', 'budou', 'být', 'však'] into ['to', 've'], loss: 0.022948595627144276 ['jeho', 'ze', 'k'] into ['zákona', 'při', 'jsou', 'byl'], loss: 0.02443383437337742 ['s'] into ['"'], loss: 0.028059336263857856 ['-'] into ['bylo', 'nás', 'včera', 'by'], loss: 0.03350440283263173 ['od', 'ČSFR', ':'] into ['z', ')'], loss: 0.03382149888328662 ['o'] into ['u', 'za', 'do'], loss: 0.03490333247494361 ['J', 'státu', 'OKD', 'listopadu', 'Na', 'V'] into ['i', 'je'], loss: 0.0369042623484128 ['po', 'pro', 'mezi', 'už', 'pouze', 'si'] into ['zákona', 'při', 'jsou', 'byl', 'jeho', 'ze', 'k'], loss: 0.04037380577853852 1: ['listopadu'] into ['OKD'] 2: ['který'] into ['které'] 3: ['státu'] into ['J'] 4: ['bude'] into ['musí'] 5: ['aby'] into ['ale'] 6: ['nás'] into ['bylo'] 7: ['si'] into ['pouze'] 8: ['už'] into ['mezi'] 9: ['&slash;'] into ['NATO'] 10: ['ze'] into ['jeho'] 11: ['však'] into ['být'] 12: ['byl'] into ['jsou'] 13: ['před'] into ['?'] 14: ['budou'] into ['jako'] 15: ['při'] into ['zákona'] 16: ['V'] into ['Na'] 17: ['OKD', 'listopadu'] into ['J', 'státu'] 18: ['za'] into ['u'] 19: ['včera'] into ['bylo', 'nás'] 20: ['ČSFR'] into ['od'] 21: ['že'] into ['které', 'který'] 22: ['pouze', 'si'] into ['mezi', 'už'] 23: ['pro'] into ['po'] 24: ['musí', 'bude'] into ['?', 'před'] 25: ['ale', 'aby'] into ['NATO', '&slash;'] 26: ['k'] into ['jeho', 'ze'] 27: ['být', 'však'] into ['jako', 'budou'] 28: ['jsou', 'byl'] into ['zákona', 'při'] 29: ['ve'] into ['to'] 30: ['by'] into ['bylo', 'nás', 'včera'] 31: ['Na', 'V'] into ['J', 'státu', 'OKD', 'listopadu'] 32: ['do'] into ['u', 'za'] 33: ['je'] into ['i'] 34: [':'] into ['od', 'ČSFR'] 35: ['které', 'který', 'že'] into ['NATO', '&slash;', 'ale', 'aby'] 36: ['mezi', 'už', 'pouze', 'si'] into ['po', 'pro'] 37: [')'] into ['z'] 38: ['('] into ['?', 'před', 'musí', 'bude'] 39: ['jako', 'budou', 'být', 'však'] into ['to', 've'] 40: ['jeho', 'ze', 'k'] into ['zákona', 'při', 'jsou', 'byl'] 41: ['s'] into ['"'] 42: ['-'] into ['bylo', 'nás', 'včera', 'by'] 43: ['od', 'ČSFR', ':'] into ['z', ')'] 44: ['o'] into ['u', 'za', 'do'] 45: ['J', 'státu', 'OKD', 'listopadu', 'Na', 'V'] into ['i', 'je'] 46: ['po', 'pro', 'mezi', 'už', 'pouze', 'si'] into ['zákona', 'při', 'jsou', 'byl', 'jeho', 'ze', 'k'] Class 1: ['bylo', 'nás', 'včera', 'by', '-'] Class 2: ['u', 'za', 'do', 'o'] Class 3: ['z', ')', 'od', 'ČSFR', ':'] Class 4: ['se'] Class 5: ['?', 'před', 'musí', 'bude', '('] Class 6: ['a'] Class 7: ['zákona', 'při', 'jsou', 'byl', 'jeho', 'ze', 'k', 'po', 'pro', 'mezi', 'už', 'pouze', 'si'] Class 8: ['i', 'je', 'J', 'státu', 'OKD', 'listopadu', 'Na', 'V'] Class 9: ['to', 've', 'jako', 'budou', 'být', 'však'] Class 10: ['NATO', '&slash;', 'ale', 'aby', 'které', 'který', 'že'] Class 11: ['na'] Class 12: ['.'] Class 13: ['"', 's'] Class 14: ['v'] Class 15: [','] Language: en. Merging pos into 10 classes. Considering first 200000 words with min. freq. 5... Merging language: en, mode: pos Sum of the Q: 0.8825168031143319 ['WP$'] into ['RBR'], loss: 0.00021477719197586455 ['RBR', 'WP$'] into ['JJR'], loss: 0.00038471950695165237 ['FW'] into ['NNPS'], loss: 0.0009191375937520887 ['NNPS', 'FW'] into ['NNP'], loss: 0.0008363385508621645 ['('] into ['.'], loss: 0.0010989642349956289 ['EX'] into ['PRP'], loss: 0.0011060376076082457 ['"'] into ['WP'], loss: 0.0014244387177662697 ['JJS'] into ['JJ'], loss: 0.001729428258254846 ['RBS'] into ['JJR', 'RBR', 'WP$'], loss: 0.002114096705532564 ['WRB'] into ['WP', '"'], loss: 0.002200261876482527 ['SYM'] into ['VBD'], loss: 0.002760619388733474 ['DT'] into ['PRP$'], loss: 0.0027498495389207487 ['CD'] into ['JJ', 'JJS'], loss: 0.00346420026378344 ['WDT'] into ['WP', '"', 'WRB'], loss: 0.005802251430491217 [':'] into [','], loss: 0.005994340401075265 ['VBN'] into ['VBD', 'SYM'], loss: 0.007252006756299058 ['VBP'] into ['VBZ'], loss: 0.007065800672448878 ['VBG'] into ['JJR', 'RBR', 'WP$', 'RBS'], loss: 0.008937882258500894 ['NNS'] into ['NN'], loss: 0.008985748580892529 ['.', '('] into [',', ':'], loss: 0.014538933148508404 ['JJR', 'RBR', 'WP$', 'RBS', 'VBG'] into ['NNP', 'NNPS', 'FW'], loss: 0.016113465288245567 ['WP', '"', 'WRB', 'WDT'] into ['NNP', 'NNPS', 'FW', 'JJR', 'RBR', 'WP$', 'RBS', 'VBG'], loss: 0.01833150630778286 ['MD'] into ['TO'], loss: 0.026436117447222685 ['RB'] into ['NNP', 'NNPS', 'FW', 'JJR', 'RBR', 'WP$', 'RBS', 'VBG', 'WP', '"', 'WRB', 'WDT'], loss: 0.0276992597896548 ['VB'] into ['VBZ', 'VBP'], loss: 0.030610132237013896 ['PRP', 'EX'] into ['NNP', 'NNPS', 'FW', 'JJR', 'RBR', 'WP$', 'RBS', 'VBG', 'WP', '"', 'WRB', 'WDT', 'RB'], loss: 0.032981376181005884 1: ['WP$'] into ['RBR'] 2: ['RBR', 'WP$'] into ['JJR'] 3: ['FW'] into ['NNPS'] 4: ['NNPS', 'FW'] into ['NNP'] 5: ['('] into ['.'] 6: ['EX'] into ['PRP'] 7: ['"'] into ['WP'] 8: ['JJS'] into ['JJ'] 9: ['RBS'] into ['JJR', 'RBR', 'WP$'] 10: ['WRB'] into ['WP', '"'] 11: ['SYM'] into ['VBD'] 12: ['DT'] into ['PRP$'] 13: ['CD'] into ['JJ', 'JJS'] 14: ['WDT'] into ['WP', '"', 'WRB'] 15: [':'] into [','] 16: ['VBN'] into ['VBD', 'SYM'] 17: ['VBP'] into ['VBZ'] 18: ['VBG'] into ['JJR', 'RBR', 'WP$', 'RBS'] 19: ['NNS'] into ['NN'] 20: ['.', '('] into [',', ':'] 21: ['JJR', 'RBR', 'WP$', 'RBS', 'VBG'] into ['NNP', 'NNPS', 'FW'] 22: ['WP', '"', 'WRB', 'WDT'] into ['NNP', 'NNPS', 'FW', 'JJR', 'RBR', 'WP$', 'RBS', 'VBG'] 23: ['MD'] into ['TO'] 24: ['RB'] into ['NNP', 'NNPS', 'FW', 'JJR', 'RBR', 'WP$', 'RBS', 'VBG', 'WP', '"', 'WRB', 'WDT'] 25: ['VB'] into ['VBZ', 'VBP'] 26: ['PRP', 'EX'] into ['NNP', 'NNPS', 'FW', 'JJR', 'RBR', 'WP$', 'RBS', 'VBG', 'WP', '"', 'WRB', 'WDT', 'RB'] Class 1: ['VBZ', 'VBP', 'VB'] Class 2: ['TO', 'MD'] Class 3: ['NNP', 'NNPS', 'FW', 'JJR', 'RBR', 'WP$', 'RBS', 'VBG', 'WP', '"', 'WRB', 'WDT', 'RB', 'PRP', 'EX'] Class 4: ['VBD', 'SYM', 'VBN'] Class 5: ['PRP$', 'DT'] Class 6: ['CC'] Class 7: ['IN'] Class 8: [',', ':', '.', '('] Class 9: ['JJ', 'JJS', 'CD'] Class 10: ['NN', 'NNS'] Language: cz. Merging pos into 10 classes. Considering first 200000 words with min. freq. 5... Merging language: cz, mode: pos Sum of the Q: 1.6931364783689442 ['AAFS6----3A----'] into ['PDFS6----------'], loss: 6.602429515910335e-06 ['AAFS6----1N----'] into ['PSFS6-P1-------'], loss: 1.7397219203907138e-05 ['PSZS6-P1-------'] into ['CrIS6----------'], loss: 3.127729972883606e-05 ['PLXP6----------'] into ['AGFP6-----A----'], loss: 3.4311009749392676e-05 ['AAIP6----3A----'] into ['AAIP6----1A----'], loss: 3.823113425607377e-05 ['CrFS6----------'] into ['PSFS6-P1-------', 'AAFS6----1N----'], loss: 4.086610136210889e-05 ['AAIS6----1N----'] into ['PSZS6FS3-------'], loss: 4.474247407439927e-05 ['PZXP6----------'] into ['Ca--6----------'], loss: 4.55535946066888e-05 ['CrNS6----------'] into ['AUNS6M---------'], loss: 4.7869141299980694e-05 ['AUFS7M---------'] into ['AAFS7----1N----'], loss: 4.888926530050269e-05 ['PJZS3----------'] into ['P4MS4----------'], loss: 5.452478119021985e-05 ['J^------------8'] into ['VB-S---2P-NA---'], loss: 5.616887333118688e-05 ['AGIS7-----A----'] into ['AAIS7----3A----'], loss: 5.753089974986855e-05 ['ClIS4----------'] into ['CrIS4----------'], loss: 5.805235962353457e-05 ['ClZS6----------'] into ['PSZS6FS3-------', 'AAIS6----1N----'], loss: 5.823773739519248e-05 ['AAFP7----1A----'] into ['AAFP7----2A----'], loss: 6.18338460113034e-05 ['AGFS6-----A----'] into ['P4FS6----------'], loss: 6.295422134917569e-05 ['PSHS1-S1-------'] into ['PSHS1-P1-------'], loss: 6.327731653869742e-05 ['AANS7----1A----'] into ['AANS7----1N----'], loss: 6.757107827307703e-05 ['CrFS2----------'] into ['AAFS2----2A----'], loss: 6.887488872184201e-05 ['AGFS7-----A----'] into ['AAFS7----3A----'], loss: 7.180820790410845e-05 ['PZFS6----------'] into ['ClFS6----------'], loss: 7.216505985532242e-05 ['CrNS2----------'] into ['AUNS2M---------'], loss: 7.242132536893354e-05 ['Cn-P6----------'] into ['PSXP6-P1-------'], loss: 7.355760465339303e-05 ['PZFS1----------'] into ['PWFS1----------'], loss: 7.411536904791975e-05 ['AUFS6M---------'] into ['PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------'], loss: 7.486445036578894e-05 ['AAIP2----2A----'] into ['AAIP2----1N----'], loss: 7.568413021558822e-05 ['P5FS2--3-------'] into ['P5ZS2--3------1'], loss: 7.613910461431246e-05 ['P4FS7----------'] into ['P9ZS7----------'], loss: 7.661663906016161e-05 ['P9XP6----------'] into ['PQ--6----------'], loss: 7.861325782343312e-05 ['PSHP1-P2-------'] into ['CrIP1----------'], loss: 7.8978285247296e-05 ['NNNSX-----A----'] into ['P6-X6----------'], loss: 7.983773783657019e-05 ['AAIP4----1N----'] into ['PZYP4----------'], loss: 8.11819960232483e-05 ['AAIS2----2A----'] into ['AAIS2----3A----'], loss: 8.178547770503939e-05 ['NNISX-----A----'] into ['AAIS1----1A---6'], loss: 8.287676760590035e-05 ['PKM-4----------'] into ['ClHP4----------'], loss: 8.287676760590046e-05 ['AUMS1M---------'] into ['AAMS1----1N----'], loss: 8.506559646197513e-05 ['PQ--2----------'] into ['PKM-3----------'], loss: 8.659745483682976e-05 ['PLIP1----------'] into ['PDIP1----------'], loss: 8.678575894221813e-05 ['AGIS4-----A----'] into ['AAIS4----3A----'], loss: 8.766761376346045e-05 ['PWFS4----------'] into ['PZFS4----------'], loss: 8.813236019583441e-05 ['ACQW------A----'] into ['VB-S---1F-AA---'], loss: 8.839324729066904e-05 ['AGNP2-----A----'] into ['CrNP2----------'], loss: 8.936122710717827e-05 ['AAMS7----3A----'] into ['AAMS7----1A----'], loss: 8.953842816170152e-05 ['CrIS6----------', 'PSZS6-P1-------'] into ['PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------'], loss: 9.08189867067739e-05 ['AAMP1----2A----'] into ['PSMP1-P1-------'], loss: 9.148631924391104e-05 ['PJNS1----------'] into ['PJYS1----------'], loss: 9.363837013045997e-05 ['AAIP1----3A----'] into ['AAIP1----1N----'], loss: 9.366961442515615e-05 ['CrIS7----------'] into ['VeYS------A----'], loss: 9.594545399532067e-05 ['PDFS7----------'] into ['AAFS7----1N----', 'AUFS7M---------'], loss: 9.61859381284748e-05 ['P8XP6----------'] into ['AGFP6-----A----', 'PLXP6----------'], loss: 9.74916285957127e-05 ['AUIS1M---------'] into ['PWYS1----------'], loss: 9.822407687161449e-05 ['P5FS3--3-------'] into ['P5ZS3--3-------'], loss: 9.968393113167777e-05 ['AAIP7----3A----'] into ['NNNP1-----A---2'], loss: 0.00010098784914196288 ['ClMS4----------'] into ['P8MS4----------'], loss: 0.00010114378539174149 ['ACTP------A----'] into ['P6-X3----------'], loss: 0.00010199924914586678 ['NNFS2-----N----'] into ['AGNP6-----A----'], loss: 0.00010403738788900647 ['AGFP1-----A----'] into ['CrFP1----------'], loss: 0.00010452519610807613 ['PP-P7--1-------'] into ['P6-X7----------'], loss: 0.00010570640120894628 ['AAFP2----3A----'] into ['AAFP2----2A----'], loss: 0.00010632893777737457 ['PSYS1-S1-------'] into ['PSYS1-P1-------'], loss: 0.00010707977018945014 ['P4NS4----------'] into ['J,-S---1-------'], loss: 0.00010893409705288871 ['Dg-------3A---1'] into ['Dg-------1A---1'], loss: 0.00011040711547118077 ['AGFP7-----A----'] into ['P9ZS7----------', 'P4FS7----------'], loss: 0.00011078938183790214 ['P5XP6--3-------'] into ['P5FS6--3-------'], loss: 0.00011096184852375226 ['PJXP1----------'] into ['P4XP3----------'], loss: 0.00011219240798770428 ['AAFS3----1A----'] into ['PDFS3----------'], loss: 0.00011356527916667278 ['AGNS2-----A----'] into ['NNNS1-----N----'], loss: 0.00011362587118065118 ['PLZS6----------'] into ['AUIS6M---------'], loss: 0.0001165455797007785 ['CrNS4----------'] into ['AANS4----1N----'], loss: 0.00012024031872400246 ['AUFS4M---------'] into ['AAFS4----3A----'], loss: 0.00012205921732702193 ['AAFS4----1N----'] into ['P8FS4----------'], loss: 0.0001223441458035508 ['P4YP4----------'] into ['P4MS4----------', 'PJZS3----------'], loss: 0.00012268510453151732 ['PZIS4----------'] into ['AAIS4----1N----'], loss: 0.00012460674958518362 ['P8XP7----------'] into ['Cn-P7----------'], loss: 0.00012469188836696757 ['ClFS2----------'] into ['AAFS2----1N----'], loss: 0.00012495138800340901 ['PDXP6----------'] into ['Ca--6----------', 'PZXP6----------'], loss: 0.00012616994047902124 ['AAIS2----1N----'] into ['AUIS2M---------'], loss: 0.0001268898734153995 ['VB-S---1P-NA--1'] into ['VB-P---1F-AA---'], loss: 0.00012745111073816972 ['PP-P1--2-------'] into ['AAIP1----2A----'], loss: 0.00012783419647870598 ['PSZS2-S1-------'] into ['PSZS2FS3-------'], loss: 0.00012820781931459574 ['AAFP4----1N----'] into ['PSXP4-P1-------'], loss: 0.00013097032849878178 ['AAMP1----3A----'] into ['AGMP1-----A----'], loss: 0.00013423749482528897 ['PZZS7----------'] into ['PLNS4---------1'], loss: 0.00013462998512546947 ['PZM-4----------'] into ['VB-S---1F-AA---', 'ACQW------A----'], loss: 0.00013492399981792128 ['AAMS6----1A----'] into ['NNMS6-----A---1'], loss: 0.00013707499822178892 ['Co-------------'] into ['Dg-------1A---3'], loss: 0.00013843708239592043 ['AAFP1----3A----'] into ['PLFP1----------'], loss: 0.00013863727975556316 ['P4XP2----------'] into ['P6-X2----------'], loss: 0.00013914136825358297 ['AANS1----1N----'] into ['AANS1----2A----'], loss: 0.0001392613395586721 ['PQ--7----------'] into ['J^------------1'], loss: 0.00014125222406912298 ['PLXP3----------'] into ['ClXP3----------'], loss: 0.0001422319980771262 ['NNFP1-----N----'] into ['NNMP6-----A----'], loss: 0.00014528698157553407 ['J,-P---1-------'] into ['Vi-P---2--N----'], loss: 0.00014542949535726745 ['AGIS2-----A----'] into ['ClZS3----------'], loss: 0.00014621545865437424 ['Vi-S---2--A----'] into ['VB-S---2P-NA---', 'J^------------8'], loss: 0.00014673469019407348 ['AAFP1----1N----'] into ['AAFP1----2A----'], loss: 0.00014721114610258853 ['NNIP2-----A---1'] into ['NNFS1-----N----'], loss: 0.00014734916831885252 ['P9XP2----------'] into ['P5ZS2--3------1', 'P5FS2--3-------'], loss: 0.00014861570198849744 ['AAXXX----1A---8'] into ['A2--------A----'], loss: 0.00014953583876646656 ['NNFP2-----A---1'] into ['NNIS5-----A----'], loss: 0.0001508327581453083 ['CyFP1----------'] into ['AAIS1----1A---6', 'NNISX-----A----'], loss: 0.00015113982593553405 ['AAMS1----2A----'] into ['CrMS1----------'], loss: 0.00015607721280773987 ['P9ZS6----------'] into ['PQ--6----------', 'P9XP6----------'], loss: 0.00015896876411512426 ['PDXP7----------'] into ['AAIS7----1N----'], loss: 0.00015906220039315863 ['II-------------'] into ['NNFS5-----A----'], loss: 0.0001593807792310604 ['C?--4----------'] into ['NNNS1-----N----', 'AGNS2-----A----'], loss: 0.00016584349122706005 ['CrNP1----------'] into ['AANP1----1A----'], loss: 0.00016606226380311197 ['CrIP1----------', 'PSHP1-P2-------'] into ['AAIP1----1N----', 'AAIP1----3A----'], loss: 0.0001665291446984363 ['P8FP4---------1'] into ['PLFP4----------'], loss: 0.00016681379923999453 ['VB-P---3F-NA---'] into ['VB-P---3P-NA--1'], loss: 0.0001692071897158036 ['PLNS1---------1'] into ['ClNS1----------'], loss: 0.00016990451942833999 ['PSZS1FS3-------'] into ['PZYS1----------'], loss: 0.00017087322714851526 ['P6-X4----------'] into ['P5ZS4--3-------'], loss: 0.00017156485440534847 ['NNFS2-----A---1'] into ['AGNP6-----A----', 'NNFS2-----N----'], loss: 0.0001720042265890269 ['AAFS7----3A----', 'AGFS7-----A----'] into ['AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------'], loss: 0.0001743127751912309 ['PDMS4----------'] into ['VB-P---2P-NA---'], loss: 0.00017478198242740326 ['PSFS4-P1-------'] into ['PZFS4----------', 'PWFS4----------'], loss: 0.00017482692246999416 ['CrFS4----------'] into ['AAFS4----3A----', 'AUFS4M---------'], loss: 0.00017495957586905505 ['AAIS1----1N----'] into ['AAIS1----2A----'], loss: 0.00017617852748895392 ['PSXP2-P1-------'] into ['AAFP2----2A----', 'AAFP2----3A----'], loss: 0.000177744110711148 ['P5XP7--3-------'] into ['P5ZS7--3-------'], loss: 0.00017937870682232203 ['PDIP4----------'] into ['PZYP4----------', 'AAIP4----1N----'], loss: 0.00017940732058412904 ['P5ZS6--3-------'] into ['P9FS6----------'], loss: 0.00018073855084608896 ['PPXP2--3-------'] into ['PPZS7--3-------'], loss: 0.00018144545932726906 ['ClXP6----------'] into ['PSXP6-P1-------', 'Cn-P6----------'], loss: 0.0001816870020965788 ['PP-P2--1-------'] into ['P5XP2--3-------'], loss: 0.0001823718156486508 ['AGIP2-----A----'] into ['AAIP2----1N----', 'AAIP2----2A----'], loss: 0.00018289236305195052 ['P4ZS6----------'] into ['P8ZS6----------'], loss: 0.00018317711043890897 ['PZFP1----------'] into ['PSHP1-P1-------'], loss: 0.00018387965791479737 ['PZMP1----------'] into ['PPMP1--3-------'], loss: 0.0001852795735940271 ['AAMP2----1N----'] into ['AGMP2-----A----'], loss: 0.00018627294405387557 ['PZFP4----------'] into ['PWFP4----------'], loss: 0.00018716701076192045 ['AANS1----3A----'] into ['PLNS4----------'], loss: 0.00018821089164023763 ['P5FS4--3-------'] into ['P5XP4--3-------'], loss: 0.00019170156619993791 ['RR--X----------'] into ['RR--1----------'], loss: 0.00019638406637867683 ['P4ZS7----------'] into ['PKM-3----------', 'PQ--2----------'], loss: 0.00019639055573759533 ['ClFS4----------'] into ['AAFS4----2A----'], loss: 0.00019924067790500649 ['P5ZS2--3-------'] into ['P6-X2----------', 'P4XP2----------'], loss: 0.0002010416221262818 ['AANS2----1N----'] into ['AUNS2M---------', 'CrNS2----------'], loss: 0.00020122622909928673 ['AUFS2M---------'] into ['AAFS2----2A----', 'CrFS2----------'], loss: 0.00020279016885173338 ['Vf--------A---1'] into ['NNNP1-----A---2', 'AAIP7----3A----'], loss: 0.0002028166002103161 ['PJYS1----------', 'PJNS1----------'] into ['J,-S---1-------', 'P4NS4----------'], loss: 0.00020373751003689602 ['AAIS1----3A----'] into ['CrIS1----------'], loss: 0.00020412171537973812 ['AGIS6-----A----'] into ['ClFS6----------', 'PZFS6----------'], loss: 0.0002060984192502088 ['PSHS1-P1-------', 'PSHS1-S1-------'] into ['PWFS1----------', 'PZFS1----------'], loss: 0.0002064044606313823 ['Vi-P---1--A----'] into ['Vi-P---2--A----'], loss: 0.00020643483801385787 ['ClZS2----------'] into ['CrIS2----------'], loss: 0.0002084604997200577 ['P6-X6----------', 'NNNSX-----A----'] into ['P5FS6--3-------', 'P5XP6--3-------'], loss: 0.000210025720627161 ['ClXP7----------'] into ['Cn-P7----------', 'P8XP7----------'], loss: 0.00021167363742975514 ['AGMS1-----A----'] into ['AAMS1----1N----', 'AUMS1M---------'], loss: 0.00021198644396826975 ['PLFS1----------'] into ['P6-X3----------', 'ACTP------A----'], loss: 0.00021251103041784898 ['PDMP4----------'] into ['PLNS4---------1', 'PZZS7----------'], loss: 0.0002144170404838699 ['PLXP7----------'] into ['NNMP2-----A---1'], loss: 0.00021488390096222033 ['PWYS1----------', 'AUIS1M---------'] into ['PZYS1----------', 'PSZS1FS3-------'], loss: 0.0002195148540697191 ['NNIXX-----A---8'] into ['NNMXX-----A---8'], loss: 0.00021982108299129204 ['P4FP1----------'] into ['P4IP1----------'], loss: 0.0002203726399170082 ['P8FS6---------1'] into ['P4FS6----------', 'AGFS6-----A----'], loss: 0.00022066675710559936 ['AUIS6M---------', 'PLZS6----------'] into ['P8ZS6----------', 'P4ZS6----------'], loss: 0.0002207468287509875 ['AAIS7----3A----', 'AGIS7-----A----'] into ['AAIS7----1N----', 'PDXP7----------'], loss: 0.0002224065508947806 ['PPNS4--3-------'] into ['P8ZS3----------'], loss: 0.0002246285940855873 ['AGFS2-----A----'] into ['PSFS2-P1-------'], loss: 0.00023024442018551865 ['AAFS6----1A----'] into ['PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------', 'AUFS6M---------'], loss: 0.00023244046421189157 ['P4FS4----------'] into ['P4MS4----------', 'PJZS3----------', 'P4YP4----------'], loss: 0.00023525723414344834 ['AAMP3----1A----'] into ['P8XP3----------'], loss: 0.00023566745197148731 ['ClFS1----------'] into ['AAFS1----3A----'], loss: 0.00023673095631175178 ['P8YP4---------1'] into ['PLYP4----------'], loss: 0.00023676185077836577 ['AAIS2----3A----', 'AAIS2----2A----'] into ['CrIS2----------', 'ClZS2----------'], loss: 0.0002409702870763543 ['AUFS1M---------'] into ['NNFS1-----A---1'], loss: 0.00024426151875370575 ['PSMP1-P1-------', 'AAMP1----2A----'] into ['AGMP1-----A----', 'AAMP1----3A----'], loss: 0.00024484033094982914 ['AAFP2----1N----'] into ['AGFP2-----A----'], loss: 0.0002451343320812255 ['AAMP1----1N----'] into ['CrFP1----------', 'AGFP1-----A----'], loss: 0.00024788853929374566 ['AAMP2----3A----'] into ['Ca--2----------'], loss: 0.0002542378967566631 ['PDFP4----------'] into ['ClHP4----------', 'PKM-4----------'], loss: 0.00025724867473463043 ['P1XXXXP3-------'] into ['P1XXXZS3-------'], loss: 0.0002588935045554231 ['Dg-------1A---1', 'Dg-------3A---1'] into ['Dg-------1A---3', 'Co-------------'], loss: 0.0002589550998341334 ['ClZS7----------'] into ['ClFS7----------'], loss: 0.0002639107342328079 ['Cn-P2----------'] into ['CrNP2----------', 'AGNP2-----A----'], loss: 0.0002649749511074971 ['PP-P4--2-------'] into ['PP-S4--1-------'], loss: 0.000268647439089649 ['P4IS4----------'] into ['P4FP4----------'], loss: 0.00027013929906276194 ['PP-P1--1-------'] into ['AAIP1----2A----', 'PP-P1--2-------'], loss: 0.0002763471492169589 ['AAMS4----1A----'] into ['P8MS4----------', 'ClMS4----------'], loss: 0.00027644904385682725 ['P8FS4----------', 'AAFS4----1N----'] into ['AAFS4----2A----', 'ClFS4----------'], loss: 0.00027714217757024667 ['AUNS6M---------', 'CrNS6----------'] into ['PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------'], loss: 0.0002833672109680205 ['AAFS1----2A----'] into ['AAFS1----1N----'], loss: 0.00028385536798410744 ['CrNS1----------'] into ['AUNS1M---------'], loss: 0.0002841033468206095 ['P8NS4---------1'] into ['AANS4----1N----', 'CrNS4----------'], loss: 0.0002847826421111708 ['NNFS4-----N----'] into ['AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1'], loss: 0.00028787490310984735 ['NNIS2-----A---1'] into ['ClNS1----------', 'PLNS1---------1'], loss: 0.00028884412054119304 ['P8IS4----------'] into ['AAIS4----3A----', 'AGIS4-----A----'], loss: 0.000290818911179244 ['P8XP2----------'] into ['PZXP2----------'], loss: 0.0002910759948502109 ['PJYS2----------'] into ['Vi-P---2--N----', 'J,-P---1-------'], loss: 0.0002921140232921668 ['VpTP---XR-AA--1'] into ['VpNS---XR-AA--1'], loss: 0.00029911558921482274 ['PDXP3----------'] into ['ClXP3----------', 'PLXP3----------'], loss: 0.0002997726438863338 ['P4FS1----------'] into ['PE--1----------'], loss: 0.0003008320878797009 ['PDIP1----------', 'PLIP1----------'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------'], loss: 0.000304231086467576 ['VB-P---1F-AA---', 'VB-S---1P-NA--1'] into ['VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------'], loss: 0.0003061793502050869 ['PPYS1--3-------'] into ['AGIP1-----A----'], loss: 0.00030851448339539226 ['AANP7----1A----'] into ['NNIS5-----A----', 'NNFP2-----A---1'], loss: 0.00031213826968988186 ['P8FS2---------1'] into ['AAFS2----1N----', 'ClFS2----------'], loss: 0.00031493407391819037 ['VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----'] into ['NNFS5-----A----', 'II-------------'], loss: 0.0003174523242206826 ['ClZS3----------', 'AGIS2-----A----'] into ['VB-P---2P-NA---', 'PDMS4----------'], loss: 0.00031848083668677206 ['P4NS1----------'] into ['P4XP3----------', 'PJXP1----------'], loss: 0.00032714185466686547 ['PPFS3--3-------'] into ['PZ--1----------'], loss: 0.00033184553275016483 ['CrMS1----------', 'AAMS1----2A----'] into ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----'], loss: 0.0003326616795199225 ['CrFS1----------'] into ['PSFSXFS3-------'], loss: 0.0003327983565674221 ['P9FS6----------', 'P5ZS6--3-------'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----'], loss: 0.00033371520435819535 ['PSXP4-P1-------', 'AAFP4----1N----'] into ['PLNS4---------1', 'PZZS7----------', 'PDMP4----------'], loss: 0.0003359470335769679 ['PDIS4----------'] into ['CrIS4----------', 'ClIS4----------'], loss: 0.0003401485603056163 ['P8FS7---------1'] into ['AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----'], loss: 0.0003419849608371473 ['NNNP3-----A----'] into ['NNFS1-----N----', 'NNIP2-----A---1'], loss: 0.0003453450841095695 ['PWFP4----------', 'PZFP4----------'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------'], loss: 0.0003455456800750222 ['ClHP1----------'] into ['PLFP1----------', 'AAFP1----3A----'], loss: 0.0003473558692622784 ['NNMS6-----A---1', 'AAMS6----1A----'] into ['NNMP6-----A----', 'NNFP1-----N----'], loss: 0.00034815103931413213 ['VB-S---2P-AA---'] into ['AGNS1-----A----'], loss: 0.0003485906053483591 ['PSZS2-P1-------'] into ['P8ZS2----------'], loss: 0.0003488404636092586 ['PSZS2FS3-------', 'PSZS2-S1-------'] into ['P8ZS2----------', 'PSZS2-P1-------'], loss: 0.00028167265826382253 ['VeYS------A----', 'CrIS7----------'] into ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----'], loss: 0.0003548361387977095 ['AAIS2----1A----'] into ['AUIS2M---------', 'AAIS2----1N----'], loss: 0.0003609439276236123 ['P8FS4---------1'] into ['PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------'], loss: 0.00036112443514467766 ['AAIS4----1N----', 'PZIS4----------'] into ['AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------'], loss: 0.00036635204344666675 ['ClYP4----------'] into ['ClXP4----------'], loss: 0.00036706565100136585 ['PLNS4----------', 'AANS1----3A----'] into ['AGNS1-----A----', 'VB-S---2P-AA---'], loss: 0.00036707748228823143 ['ClYS1----------'] into ['AGIS1-----A----'], loss: 0.00037002922220758286 ['PW--1----------'] into ['PW--4----------'], loss: 0.00037223841660657194 ['NNMS3-----A---1'] into ['AAMS3----1A----'], loss: 0.0003723114757928749 ['VB-S---3F-NA---'] into ['J^------------1', 'PQ--7----------'], loss: 0.0003724035334773572 ['P5ZS3--3-------', 'P5FS3--3-------'] into ['P6-X3----------', 'ACTP------A----', 'PLFS1----------'], loss: 0.00037794169786199347 ['Vc-P---1-------'] into ['ACMP------A----'], loss: 0.0003800867254184037 ['PZM-1----------'] into ['PP-P3--2-------'], loss: 0.0003881583819167302 ['P4YS1----------'] into ['J,-S---1-------', 'P4NS4----------', 'PJYS1----------', 'PJNS1----------'], loss: 0.0003930253142851066 ['VB-P---1P-NA---'] into ['VB-S---1P-AA--1'], loss: 0.0003953433978250661 ['PPMP1--3-------', 'PZMP1----------'] into ['AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----'], loss: 0.00039909434835052465 ['RF-------------'] into ['NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------'], loss: 0.000401458052038866 ['P5ZS4--3-------', 'P6-X4----------'] into ['P5XP4--3-------', 'P5FS4--3-------'], loss: 0.0004037776259040245 ['VpTP---XR-NA---'] into ['VB-P---3P-NA--1', 'VB-P---3F-NA---'], loss: 0.0004046001536796113 ['PPZS7--3-------', 'PPXP2--3-------'] into ['P8ZS3----------', 'PPNS4--3-------'], loss: 0.0004092492513763831 ['CrIS1----------', 'AAIS1----3A----'] into ['PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------'], loss: 0.00040970326699290953 ['PQ--1----------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------'], loss: 0.0004124207482654181 ['A2--------A----', 'AAXXX----1A---8'] into ['NNMXX-----A---8', 'NNIXX-----A---8'], loss: 0.0004149139345738088 ['AANS2----1A----'] into ['AUNS2M---------', 'CrNS2----------', 'AANS2----1N----'], loss: 0.00041576859281495376 ['NNMS5-----A----'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------'], loss: 0.0004246467763391417 ['NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1'] into ['CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----'], loss: 0.0004280756591562735 ['PDFS1----------'] into ['PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------'], loss: 0.00043474027441732857 ['PSYS1-P1-------', 'PSYS1-S1-------'] into ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----'], loss: 0.00043499652687142756 ['ClXP2----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------'], loss: 0.00043578802043221614 ['PDFS2----------'] into ['PSFS2-P1-------', 'AGFS2-----A----'], loss: 0.0004369551481315506 ['PKM-1----------'] into ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------'], loss: 0.0004376604141117077 ['Ca--2----------', 'AAMP2----3A----'] into ['PZXP2----------', 'P8XP2----------'], loss: 0.0004428430914211407 ['PH-S4--1-------'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1'], loss: 0.0004432304414115737 ['AAIP2----1A----'] into ['AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----'], loss: 0.0004446777604874691 ['AANP6----1A----'] into ['PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------'], loss: 0.00044571297602563273 ['Ca--1----------'] into ['Ca--4----------'], loss: 0.0004485041820442622 ['PDFS4----------'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------'], loss: 0.000451320259733851 ['AAIP4----1A----'] into ['PZYP4----------', 'AAIP4----1N----', 'PDIP4----------'], loss: 0.0004605769068917846 ['NNMS6-----A----'] into ['NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----'], loss: 0.00046401985387540005 ['PZ--4----------'] into ['ACYS------A----'], loss: 0.00047961092802832096 ['Vf--------N----'] into ['AANS1----2A----', 'AANS1----1N----'], loss: 0.00048298850764411224 ['P8ZS7----------'] into ['AANP3----1A----'], loss: 0.0004871403553058573 ['AAMP7----1A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----'], loss: 0.0005030576878570095 ['PP-P4--1-------'] into ['PP-S4--1-------', 'PP-P4--2-------'], loss: 0.0005042513569848831 ['RR--1----------', 'RR--X----------'] into ['NNFS1-----A---1', 'AUFS1M---------'], loss: 0.0005072496345855284 ['AGFS1-----A----'] into ['AAFS1----3A----', 'ClFS1----------'], loss: 0.0005091139825422382 ['PSHP1-P1-------', 'PZFP1----------'] into ['PLFP1----------', 'AAFP1----3A----', 'ClHP1----------'], loss: 0.0005104549573868638 ['VpQW---XR-NA---'] into ['VpQW---XR-AA--1'], loss: 0.0005105955088162892 ['PQ--6----------', 'P9XP6----------', 'P9ZS6----------'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------'], loss: 0.0005220893249769072 ['AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------'], loss: 0.0005255893429910495 ['PPFS4--3-------'] into ['PHZS4--3-------'], loss: 0.000525953834143153 ['ClXP1----------'] into ['AAFP1----2A----', 'AAFP1----1N----'], loss: 0.0005300479197926648 ['NNFSX-----A----'] into ['NNIPX-----A----'], loss: 0.000530080612474559 ['NNMP2-----A---1', 'PLXP7----------'] into ['AGIP1-----A----', 'PPYS1--3-------'], loss: 0.000531265705439928 ['PDMP1----------'] into ['PLMP1----------'], loss: 0.0005319391418313899 ['PDXP2----------'] into ['AGFP2-----A----', 'AAFP2----1N----'], loss: 0.0005322017709938525 ['PWM-1----------'] into ['PLNS1----------'], loss: 0.000540604911257465 ['AAFS7----1A----'] into ['AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1'], loss: 0.0005474023285434483 ['PLFP4----------', 'P8FP4---------1'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------'], loss: 0.0005480233720124073 ['PP-S1--1-------'] into ['AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------'], loss: 0.0005497474311855156 ['PPXP4--3-------'] into ['PPZS4--3------2'], loss: 0.0005502797931046697 ['VpMP---XR-AA--1'] into ['VpMP---XR-NA---'], loss: 0.0005521782027592512 ['PLXP2----------'] into ['PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----'], loss: 0.0005531428213135189 ['VB-P---3P-AA--1'] into ['VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1'], loss: 0.0005537818575130785 ['AANP4----1A----'] into ['PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----'], loss: 0.0005588926176024263 ['AANS4----1A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1'], loss: 0.0005601759243670741 ['Vi-P---2--A----', 'Vi-P---1--A----'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----'], loss: 0.000575822707220917 ['PSXXXXP3-------'] into ['P1XXXZS3-------', 'P1XXXXP3-------'], loss: 0.000579795243742538 ['AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1'] into ['PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------'], loss: 0.000584519090265391 ['AAMP2----1A----'] into ['AGMP2-----A----', 'AAMP2----1N----'], loss: 0.0005951677157307982 ['NNNP7-----A----'] into ['P6-X7----------', 'PP-P7--1-------'], loss: 0.0005962249134426291 ['P5ZS7--3-------', 'P5XP7--3-------'] into ['P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----'], loss: 0.0005897804390823753 ['Db------------8'] into ['NNFXX-----A---8'], loss: 0.0005963660423994998 ['Dg-------3A----'] into ['ClFS7----------', 'ClZS7----------'], loss: 0.0006083206967914185 ['P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------'] into ['Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------'], loss: 0.0006106331948039364 ['P4FP4----------', 'P4IS4----------'] into ['Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------', 'P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------'], loss: 0.0005872867140389662 ['AGFS4-----A----'] into ['VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----'], loss: 0.0006199942254642194 ['PHZS3--3-------'] into ['PZ--1----------', 'PPFS3--3-------'], loss: 0.0006239881523732995 ['Cn-P7----------', 'P8XP7----------', 'ClXP7----------'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----'], loss: 0.0006273828521090154 ['PPXP3--3-------'] into ['PP-P3--2-------', 'PZM-1----------'], loss: 0.0006296357880983048 ['P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1'] into ['P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------'], loss: 0.0006333290990911274 ['ClFS6----------', 'PZFS6----------', 'AGIS6-----A----'] into ['P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1'], loss: 0.0006263619565683022 ['AAIP6----1A----', 'AAIP6----3A----'] into ['Ca--6----------', 'PZXP6----------', 'PDXP6----------'], loss: 0.0006336693077611706 ['P4MP1----------'] into ['P4IP1----------', 'P4FP1----------'], loss: 0.0006380145473267127 ['VsMP---XX-AP---'] into ['ACMP------A----', 'Vc-P---1-------'], loss: 0.0006411226378871652 ['AAFP6----1A----'] into ['AGFP6-----A----', 'PLXP6----------', 'P8XP6----------'], loss: 0.0006514733867610915 ['P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------'] into ['ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1'], loss: 0.0006550263730965035 ['ClXP3----------', 'PLXP3----------', 'PDXP3----------'] into ['P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------'], loss: 0.0006592274241572849 ['AAIS4----1A----'] into ['AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------'], loss: 0.0006649276231363854 ['C}-------------'] into ['NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8'], loss: 0.0006669066563271707 ['AAIS1----2A----', 'AAIS1----1N----'] into ['AGIS1-----A----', 'ClYS1----------'], loss: 0.0006787611994773894 ['P5XP2--3-------', 'PP-P2--1-------'] into ['ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------'], loss: 0.0006796269759717288 ['ClYP1----------'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------'], loss: 0.0006796773221613146 ['AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------'] into ['PLMP1----------', 'PDMP1----------'], loss: 0.000681926578723245 ['PLYS1----------'] into ['PDYS1----------'], loss: 0.0006849175510178647 ['PDFP1----------'] into ['PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------'], loss: 0.0006987169761761103 ['VB-P---2P-AA---'] into ['VB-S---1P-NA---'], loss: 0.0006999086709006141 ['PSFSXFS3-------', 'CrFS1----------'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----'], loss: 0.000702608746563072 ['Vc-S---1-------'] into ['VB-S---1P-AA---'], loss: 0.0007047303753939232 ['VpNS---XR-AA--1', 'VpTP---XR-AA--1'] into ['VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---'], loss: 0.0007114893772403401 ['PP-P3--1-------'] into ['PH-S3--1-------'], loss: 0.0007121887454635005 ['NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---'], loss: 0.0007145520609939992 ['AANS6----1A----'] into ['PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------'], loss: 0.0007151456072016539 ['PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----'] into ['PDYS1----------', 'PLYS1----------'], loss: 0.0007228456501253088 ['PDZS7----------'] into ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------'], loss: 0.0007235311062447052 ['AANS1----1A----'] into ['AUNS1M---------', 'CrNS1----------'], loss: 0.0007324702340073067 ['AAMP4----1A----'] into ['PLYP4----------', 'P8YP4---------1'], loss: 0.0007334007106947794 ['CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----'] into ['AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----'], loss: 0.0007411117891758807 ['Dg-------2A---1'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------'], loss: 0.0007675954489430932 ['Db------------1'] into ['Cv-------------'], loss: 0.0007724283853681238 ['P8XP3----------', 'AAMP3----1A----'] into ['P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------'], loss: 0.0007829677730353791 ['NNFPX-----A----'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----'], loss: 0.000785682299382591 ['ACNS------A----'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----'], loss: 0.0007884705329255261 ['VB-S---1P-AA--1', 'VB-P---1P-NA---'] into ['VB-S---1P-NA---', 'VB-P---2P-AA---'], loss: 0.0007927509719976223 ['CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------'], loss: 0.000801198866963955 ['AAFS1----1N----', 'AAFS1----2A----'] into ['AANS1----2A----', 'AANS1----1N----', 'Vf--------N----'], loss: 0.0008088697255468915 ['Cn-S1----------'] into ['Ca--4----------', 'Ca--1----------'], loss: 0.0008233528556061231 ['CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1'] into ['AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------'], loss: 0.0008288299828685312 ['PDZS2----------'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------'], loss: 0.0008436384949763263 ['PE--1----------', 'P4FS1----------'] into ['J,-S---1-------', 'P4NS4----------', 'PJYS1----------', 'PJNS1----------', 'P4YS1----------'], loss: 0.0008461908271523971 ['P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------'] into ['PPZS4--3------2', 'PPXP4--3-------'], loss: 0.0008517332081007891 ['AAIP3----1A----'] into ['AANP3----1A----', 'P8ZS7----------'], loss: 0.0008816974828067528 ['PW--4----------', 'PW--1----------'] into ['PLNS1----------', 'PWM-1----------'], loss: 0.0008826879932773077 ['NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------'] into ['AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------'], loss: 0.0008981738855807385 ['Dg-------1N----'] into ['ACYS------A----', 'PZ--4----------'], loss: 0.0009049432454271188 ['NNMS3-----A----'] into ['AAMS3----1A----', 'NNMS3-----A---1'], loss: 0.000908241324715659 ['AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----'] into ['VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----'], loss: 0.00022346334040751177 ['VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----'] into ['AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----'], loss: 0.0006149557620880788 ['AAMS1----1A----'] into ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------'], loss: 0.0009091035763646531 ['PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----'], loss: 0.0009131077280775487 ['NNIS6-----A----'] into ['NNIS6-----A---1'], loss: 0.0009300286860455947 ['NNMP1-----A----'] into ['NNMP1-----A---1'], loss: 0.0009326844847098577 ['AAMP1----1A----'] into ['PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------'], loss: 0.0009270809523082175 ['PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----'] into ['AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----'], loss: 0.0009471533147338215 ['NNFP6-----A----'] into ['NNNP6-----A----'], loss: 0.0008076195170246304 ['ClXP4----------', 'ClYP4----------'] into ['PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----'], loss: 0.0009677592271756449 ['PQ--4----------'] into ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------', 'PKM-1----------'], loss: 0.0009678295910635392 ['RV--3----------'] into ['RR--3----------'], loss: 0.0009720246510990481 ['VpYS---XR-AA--1'] into ['VpYS---XR-AA---'], loss: 0.0009720827141094402 ['AAFP3----1A----'] into ['NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----'], loss: 0.0009863086792256063 ['AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------'], loss: 0.001028480803557828 ['VB-P---3P-NA---'] into ['VpMP---XR-NA---', 'VpMP---XR-AA--1'], loss: 0.0010347891979745867 ['VpTP---XR-AA---'] into ['VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1'], loss: 0.0010454767695581492 ['ClXS2----------'] into ['PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------'], loss: 0.0010584317386930983 ['Cn-S4----------'] into ['Ca--4----------', 'Ca--1----------', 'Cn-S1----------'], loss: 0.0010572540924380024 ['PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------'] into ['PH-S3--1-------', 'PP-P3--1-------'], loss: 0.0010606584189547799 ['NNIPX-----A----', 'NNFSX-----A----'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----'], loss: 0.0010613189461292204 ['PHZS4--3-------', 'PPFS4--3-------'] into ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------'], loss: 0.001071331787781754 ['VB-P---3F-AA---'] into ['VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1'], loss: 0.0010715906786546224 ['NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----'], loss: 0.0010835893698258648 ['Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------', 'P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------', 'P4FP4----------', 'P4IS4----------'] into ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------', 'PKM-1----------', 'PQ--4----------'], loss: 0.0010925690102154889 ['RR--4----------'] into ['RV--4----------'], loss: 0.001091856102494451 ['AAFP4----1A----'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1'], loss: 0.0010787855607888148 ['PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------'], loss: 0.0010948987048702687 ['VpNS---XR-NA---'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----'], loss: 0.0010976803197157842 ['NNFXX-----A---8', 'Db------------8'] into ['NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------'], loss: 0.0010991340184958454 ['PDZS6----------'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------'], loss: 0.0011118594208968647 ['NNNP4-----A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------'], loss: 0.0011222186708015267 ['NNMP7-----A----'] into ['P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------'], loss: 0.0011391617719325843 ['PDZS3----------'] into ['AAIS3----1A----'], loss: 0.0011478163950671295 ['NNMS4-----A----'] into ['P8MS4----------', 'ClMS4----------', 'AAMS4----1A----'], loss: 0.0011495274990832938 ['Dg-------2A----'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1'], loss: 0.001208213008683829 ['PDNS4----------'] into ['PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------'], loss: 0.0012135337520150671 ['AAFP7----2A----', 'AAFP7----1A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------'], loss: 0.0012395555713032406 ['P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------'], loss: 0.0012437291554775647 ['RV--2----------'] into ['AAFS2----2A----', 'CrFS2----------', 'AUFS2M---------'], loss: 0.0012910226430475071 ['J,-X---3-------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------', 'PQ--1----------'], loss: 0.001297213133604938 ['AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------'] into ['AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------'], loss: 0.0013082914260193125 ['NNFP3-----A----'] into ['NNIP3-----A----'], loss: 0.0013278617870135512 ['NNIP6-----A----'] into ['NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----', 'NNMS6-----A----'], loss: 0.001331602242482039 ['ClFS7----------', 'ClZS7----------', 'Dg-------3A----'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----'], loss: 0.0013331746744274784 ['NNFP7-----A----'] into ['P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----'], loss: 0.0013340080057473464 ['AAFS4----1A----'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1'], loss: 0.0013512390984049827 ['VsYS---XX-AP---'] into ['AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----'], loss: 0.0013716335403436957 ['PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------'] into ['AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1'], loss: 0.0014110146924557147 ['AAIP1----1A----'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------'], loss: 0.001411294796742799 ['RV--7----------'] into ['RR--7----------'], loss: 0.0014301735744251046 ['P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----'], loss: 0.0014267693093746618 ['NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----'], loss: 0.0013973237713253584 ['NNMP3-----A----'] into ['NNIP3-----A----', 'NNFP3-----A----'], loss: 0.0014376503637783472 ['VB-S---3F-AA---'] into ['VpQW---XR-AA--1', 'VpQW---XR-NA---'], loss: 0.001443511721998469 ['VB-S---3P-NA---'] into ['VpYS---XR-NA---'], loss: 0.001466076376956317 ['ACYS------A----', 'PZ--4----------', 'Dg-------1N----'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----'], loss: 0.0014756253365247167 ['Db-------------'] into ['Cv-------------', 'Db------------1'], loss: 0.0014568770490205889 ['PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------'] into ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------'], loss: 0.001491775704238096 ['PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------'] into ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------'], loss: 0.0014091573140008855 ['VsTP---XX-AP---'] into ['ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---'], loss: 0.0014869651192212121 ['AANS7----1N----', 'AANS7----1A----'] into ['AAMS7----1A----', 'AAMS7----3A----'], loss: 0.0014922318651327373 ['XX-------------'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------'], loss: 0.001512015671686637 ['PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----'], loss: 0.0015503766251371027 ['VB-P---1P-AA---'] into ['VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---'], loss: 0.0015628283859096049 ['VpMP---XR-AA---'] into ['VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---'], loss: 0.0015568230038479204 ['AAIS7----1A----'] into ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------'], loss: 0.001589652404592011 ['VsQW---XX-AP---'] into ['VsNS---XX-AP---'], loss: 0.0015916369539163988 ['PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----'] into ['PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------'], loss: 0.0016087858883503184 ['NNIP4-----A----'] into ['NNMP4-----A----'], loss: 0.0012874379529257789 ['NNNXX-----A----'] into ['NNIXX-----A----'], loss: 0.0016483399476322279 ['PDFS6----------', 'AAFS6----3A----'] into ['Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----'], loss: 0.0016551534144433732 ['NNNS3-----A----'] into ['AANS3----1A----'], loss: 0.0016702177818335457 ['AAIP7----1A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----'], loss: 0.0016859543126698363 ['P4IP1----------', 'P4FP1----------', 'P4MP1----------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------', 'PQ--1----------', 'J,-X---3-------'], loss: 0.0017225721488248258 ['AAXXX----1A----'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----'], loss: 0.0017406687203390218 ['AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------'], loss: 0.0017446691137504942 ['AANP1----1A----', 'CrNP1----------'] into ['CrIS4----------', 'ClIS4----------', 'PDIS4----------'], loss: 0.0017408085304985686 ['VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---'] into ['VpYS---XR-NA---', 'VB-S---3P-NA---'], loss: 0.0018758592295739182 ['VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---'], loss: 0.0019146421343032968 ['NNMS2-----A----'] into ['AAMS2----1A----'], loss: 0.001957230268844726 ['AAMS2----1A----', 'NNMS2-----A----'] into ['PDFS3----------', 'AAFS3----1A----'], loss: -0.00804188814726723 ['PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------'], loss: -0.010406776425865733 ['P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----'], loss: -0.008911437946749276 ['P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----'], loss: -0.01026459046954874 ['Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----'], loss: -0.007629447183640893 ['AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----'], loss: 5.7883248249381014e-05 ['NNNP6-----A----', 'NNFP6-----A----'] into ['NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----', 'NNMS6-----A----', 'NNIP6-----A----'], loss: 0.0011663564248341196 ['RV--6----------'] into ['RR--6----------'], loss: 0.0019047440579213247 ['RR--2----------'] into ['AAFS2----2A----', 'CrFS2----------', 'AUFS2M---------', 'RV--2----------'], loss: 0.0019799047105345693 ['ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------'], loss: 0.0020028443367202326 ['J,-S---1-------', 'P4NS4----------', 'PJYS1----------', 'PJNS1----------', 'P4YS1----------', 'PE--1----------', 'P4FS1----------'] into ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------', 'PKM-1----------', 'PQ--4----------', 'Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------', 'P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------', 'P4FP4----------', 'P4IS4----------'], loss: 0.00200313036417176 ['AAFP1----1A----'] into ['AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------'], loss: 0.0021046427834323436 ['Xx-------------'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----'], loss: 0.0021111332534291594 ['AAIS1----1A----'] into ['PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----'], loss: 0.0021225545558653655 ['TT-------------'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----'], loss: 0.002138270823056438 ['AANP2----1A----'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------'], loss: 0.002144593858427937 ['PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----'], loss: 0.0016540799332079182 ['X@-------------'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------', 'XX-------------'], loss: 0.002221545064182193 ['VB-P---3P-AA---'] into ['VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---'], loss: 0.002363733209273179 ['NNFS3-----A----'] into ['NNIP3-----A----', 'NNFP3-----A----', 'NNMP3-----A----'], loss: 0.0023918863503050156 ['Dg-------1A----'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----'], loss: 0.0024218556088530393 ['AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----'] into ['CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------'], loss: 0.002434361203902464 ['AAFS1----1A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----'], loss: 0.0025174246017960283 ['P7-X4----------'] into ['P7-X3----------'], loss: 0.002539340066858116 ['NNNS7-----A----'] into ['NNMS7-----A----'], loss: 0.0025909236978939884 ['NNNP1-----A----'] into ['NNFP1-----A----'], loss: 0.0026402216063716073 ['AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----'] into ['NNIP7-----A----'], loss: 0.002651228900769199 ['AAIS3----1A----', 'PDZS3----------'] into ['AANS3----1A----', 'NNNS3-----A----'], loss: 0.0026941076748845934 ['P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----'], loss: 0.0027814373308090844 ['AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---'] into ['VsNS---XX-AP---', 'VsQW---XX-AP---'], loss: 0.0028207486544391215 ['NNFXX-----A----'] into ['NNIXX-----A----', 'NNNXX-----A----'], loss: 0.0028517832737261764 ['VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---'], loss: 0.002869407565843583 ['PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------'] into ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------'], loss: 0.0028738946647087924 ['AAIS6----1A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------'], loss: 0.0028910280042056136 ['NNIS3-----A----'] into ['AANS3----1A----', 'NNNS3-----A----', 'AAIS3----1A----', 'PDZS3----------'], loss: 0.002920633954280995 ['NNNP2-----A----'] into ['NNMP2-----A----'], loss: 0.00292776059253616 ['AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------'], loss: 0.0029433270588885138 ['Ca--4----------', 'Ca--1----------', 'Cn-S1----------', 'Cn-S4----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----', 'PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------'], loss: 0.002996828570055904 ['Cv-------------', 'Db------------1', 'Db-------------'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------'], loss: 0.003084239923887796 ['Vf--------A----'] into ['ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---'], loss: 0.003177510923646168 ['VpNS---XR-AA---'] into ['VpQW---XR-AA---'], loss: 0.0033277573775163153 ['PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------'], loss: 0.003349431266080769 ['P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----'], loss: 0.0034034502774638645 ['Vc-X---3-------'] into ['VB-S---1P-AA---', 'Vc-S---1-------'], loss: 0.0036265218164639795 ['PDNS1----------'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------'], loss: 0.0036396589673149166 ['NNFP4-----A----'] into ['NNMP4-----A----', 'NNIP4-----A----'], loss: 0.003708267388660934 ['AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----'], loss: 0.003772852648616518 ['NNMS7-----A----', 'NNNS7-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----', 'P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----'], loss: 0.00345299100323309 ['NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------', 'NNFXX-----A---8', 'Db------------8'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----', 'Xx-------------'], loss: 0.003860495333398277 ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------', 'XX-------------', 'X@-------------'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----', 'Xx-------------', 'NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------', 'NNFXX-----A---8', 'Db------------8'], loss: 0.0031507994832887555 ['VB-S---3P-AA---'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---'], loss: 0.0038025972029711867 ['VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'] into ['VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------'], loss: 0.0038326308515695046 ['PSXXXZS3-------'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1'], loss: 0.0038632069490340654 ['PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----'], loss: 0.0039062044126886917 ['J,-------------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------', 'PQ--1----------', 'J,-X---3-------', 'P4IP1----------', 'P4FP1----------', 'P4MP1----------'], loss: 0.003944845783473494 ['NNNS4-----A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----'], loss: 0.00402806841086859 ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------', 'PKM-1----------', 'PQ--4----------', 'Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------', 'P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------', 'P4FP4----------', 'P4IS4----------', 'J,-S---1-------', 'P4NS4----------', 'PJYS1----------', 'PJNS1----------', 'P4YS1----------', 'PE--1----------', 'P4FS1----------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------', 'PQ--1----------', 'J,-X---3-------', 'P4IP1----------', 'P4FP1----------', 'P4MP1----------', 'J,-------------'], loss: 0.004128475598082032 ['AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----', 'AAIP2----1A----'] into ['AGMP2-----A----', 'AAMP2----1N----', 'AAMP2----1A----'], loss: 0.004167584994181479 ['NNIP2-----A----'] into ['NNMP2-----A----', 'NNNP2-----A----'], loss: 0.0022591868825844463 ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----', 'PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------', 'Ca--4----------', 'Ca--1----------', 'Cn-S1----------', 'Cn-S4----------'] into ['AGMP2-----A----', 'AAMP2----1N----', 'AAMP2----1A----', 'AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----', 'AAIP2----1A----'], loss: 0.003721809233251873 ['NNNS6-----A----'] into ['NNIS6-----A---1', 'NNIS6-----A----'], loss: 0.004142942700799372 ['PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------', 'AANS6----1A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------', 'AAIS6----1A----'], loss: 0.0014645681962276657 ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------'], loss: 0.004134441800701419 ['NNFS7-----A----'] into ['NNIS7-----A----'], loss: 0.004364474442160471 ['NNMS1-----A----'] into ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----'], loss: 0.004586127437324053 ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----'] into ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----'], loss: -0.0039883901140692415 ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----', 'AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----', 'AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----'], loss: -0.011550171667215305 ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----', 'AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----', 'AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----', 'AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----', 'P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----', 'NNMS7-----A----', 'NNNS7-----A----'], loss: -0.008934883084231324 ['NNIXX-----A----', 'NNNXX-----A----', 'NNFXX-----A----'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----', 'Xx-------------', 'NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------', 'NNFXX-----A---8', 'Db------------8', 'AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------', 'XX-------------', 'X@-------------'], loss: 0.0046303039209365515 ['AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----'], loss: 0.004656173386433508 ['NNIP3-----A----', 'NNFP3-----A----', 'NNMP3-----A----', 'NNFS3-----A----'] into ['AANS3----1A----', 'NNNS3-----A----', 'AAIS3----1A----', 'PDZS3----------', 'NNIS3-----A----'], loss: 0.0047029881575612765 ['PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----'], loss: 0.004877333086881201 ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------'], loss: 0.00515065552577383 ['VpYS---XR-AA---', 'VpYS---XR-AA--1'] into ['VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---'], loss: 0.005399945511679554 ['NNIP1-----A----'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----'], loss: 0.005814647418408689 ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----'], loss: 0.00213146516251325 ['AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------'], loss: 0.005968599684499783 ['VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---'] into ['VpQW---XR-AA---', 'VpNS---XR-AA---'], loss: 0.006077094996580996 ['AUNS1M---------', 'CrNS1----------', 'AANS1----1A----'] into ['VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---'], loss: 0.006276290998991909 ['NNNS1-----A----'] into ['NNMP1-----A---1', 'NNMP1-----A----'], loss: 0.006494455113920047 ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---'] into ['VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---'], loss: 0.006237120943834685 ['NNIS2-----A----'] into ['NNNS2-----A----'], loss: 0.006493621379351295 ['NNFP2-----A----'] into ['NNMP2-----A----', 'NNNP2-----A----', 'NNIP2-----A----'], loss: 0.0066410293766830195 ['AAFP2----1A----'] into ['AGMP2-----A----', 'AAMP2----1N----', 'AAMP2----1A----', 'AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----', 'AAIP2----1A----', 'AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----', 'PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------', 'Ca--4----------', 'Ca--1----------', 'Cn-S1----------', 'Cn-S4----------'], loss: 0.0013400990645433502 ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------'], loss: 0.006654161283370457 ['NNIS4-----A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----'], loss: 0.00702189846250178 ['CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----'], loss: 0.00799569906598508 ['ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----'] into ['VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----'], loss: 0.007977778186216422 ['NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----', 'NNMS6-----A----', 'NNIP6-----A----', 'NNNP6-----A----', 'NNFP6-----A----'] into ['NNIS6-----A---1', 'NNIS6-----A----', 'NNNS6-----A----'], loss: 0.00801353870834221 ['VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----'], loss: 0.008202285323553793 ['NNFP1-----A----', 'NNNP1-----A----'] into ['NNMP1-----A---1', 'NNMP1-----A----', 'NNNS1-----A----'], loss: 0.008338979431132945 ['AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----'], loss: 0.00724492951290308 ['NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----'], loss: 0.009695361773092732 ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----'], loss: -0.00026655805131180557 ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['NNNS2-----A----', 'NNIS2-----A----'], loss: -0.005252071569042942 ['NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----'], loss: 0.009296354263374159 ['VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----'], loss: 0.00992904682477062 ['NNIS7-----A----', 'NNFS7-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----', 'P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----', 'NNMS7-----A----', 'NNNS7-----A----', 'P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----', 'AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----', 'AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----', 'AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----'], loss: 0.009486509218816793 ['PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------', 'AUFS6M---------', 'AAFS6----1A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------', 'AAIS6----1A----', 'PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------', 'AANS6----1A----'], loss: 0.01009637654691997 ['RR--3----------', 'RV--3----------'] into ['RR--7----------', 'RV--7----------'], loss: 0.010563693497178698 ['AANS3----1A----', 'NNNS3-----A----', 'AAIS3----1A----', 'PDZS3----------', 'NNIS3-----A----', 'NNIP3-----A----', 'NNFP3-----A----', 'NNMP3-----A----', 'NNFS3-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----', 'P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----', 'NNMS7-----A----', 'NNNS7-----A----', 'P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----', 'AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----', 'AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----', 'AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----', 'NNIS7-----A----', 'NNFS7-----A----'], loss: 0.010315021337354069 ['NNFS1-----A----'] into ['NNIS1-----A----'], loss: 0.011958125126148404 ['PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----'], loss: 0.003227865282079747 ['C=-------------'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----', 'Xx-------------', 'NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------', 'NNFXX-----A---8', 'Db------------8', 'AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------', 'XX-------------', 'X@-------------', 'NNIXX-----A----', 'NNNXX-----A----', 'NNFXX-----A----'], loss: 0.011764928005260918 ['RV--4----------', 'RR--4----------'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1', 'AAFS4----1A----'], loss: 0.013272332397128817 ['NNMP1-----A---1', 'NNMP1-----A----', 'NNNS1-----A----', 'NNFP1-----A----', 'NNNP1-----A----'] into ['NNIS1-----A----', 'NNFS1-----A----'], loss: 0.014548344718174013 ['NNFS2-----A----'] into ['AUIS2M---------', 'AAIS2----1N----', 'AAIS2----1A----'], loss: 0.015582456763485072 ['J^-------------'] into ['P7-X3----------', 'P7-X4----------'], loss: 0.016350975646489033 ['VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1'] into ['P7-X3----------', 'P7-X4----------', 'J^-------------'], loss: 0.0161603278774853 ['VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---'] into ['P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1'], loss: 0.01182389683268497 ['NNFS6-----A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------', 'AAIS6----1A----', 'PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------', 'AANS6----1A----', 'PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------', 'AUFS6M---------', 'AAFS6----1A----'], loss: 0.016702297363430107 ['NNIS6-----A---1', 'NNIS6-----A----', 'NNNS6-----A----', 'NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----', 'NNMS6-----A----', 'NNIP6-----A----', 'NNNP6-----A----', 'NNFP6-----A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------', 'AAIS6----1A----', 'PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------', 'AANS6----1A----', 'PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------', 'AUFS6M---------', 'AAFS6----1A----', 'NNFS6-----A----'], loss: 0.010917748865734248 ['NNFS4-----A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'], loss: 0.019925709511248012 ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----'], loss: 0.015464270984066511 ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'] into ['P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---'], loss: 0.022196162305820416 ['P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---', 'PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----', 'AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----'], loss: -0.017212197393538534 ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1', 'AAFS4----1A----', 'RV--4----------', 'RR--4----------'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----', 'AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----', 'P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---', 'PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'], loss: 0.014063921010342928 ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['AAFS2----1A----'], loss: 0.021425650217452838 ['AAFS2----1A----', 'AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----', 'PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----'], loss: 0.004693733206278717 ['AUIS2M---------', 'AAIS2----1N----', 'AAIS2----1A----', 'NNFS2-----A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----', 'PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----', 'AAFS2----1A----', 'AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'], loss: 0.015971563910243278 ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----', 'PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----', 'AAFS2----1A----', 'AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----', 'AUIS2M---------', 'AAIS2----1N----', 'AAIS2----1A----', 'NNFS2-----A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----', 'AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----', 'P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---', 'PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---', 'AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1', 'AAFS4----1A----', 'RV--4----------', 'RR--4----------'], loss: -0.0014891800435382317 ['AAFS2----2A----', 'CrFS2----------', 'AUFS2M---------', 'RV--2----------', 'RR--2----------'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----', 'AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----', 'P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---', 'PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---', 'AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1', 'AAFS4----1A----', 'RV--4----------', 'RR--4----------', 'AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----', 'PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----', 'AAFS2----1A----', 'AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----', 'AUIS2M---------', 'AAIS2----1N----', 'AAIS2----1A----', 'NNFS2-----A----'], loss: 0.01619700027927127 ['NNMP2-----A----', 'NNNP2-----A----', 'NNIP2-----A----', 'NNFP2-----A----'] into ['AGMP2-----A----', 'AAMP2----1N----', 'AAMP2----1A----', 'AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----', 'AAIP2----1A----', 'AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----', 'PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------', 'Ca--4----------', 'Ca--1----------', 'Cn-S1----------', 'Cn-S4----------', 'AAFP2----1A----'], loss: 0.0228845129938635 1: ['AAFS6----3A----'] into ['PDFS6----------'] 2: ['AAFS6----1N----'] into ['PSFS6-P1-------'] 3: ['PSZS6-P1-------'] into ['CrIS6----------'] 4: ['PLXP6----------'] into ['AGFP6-----A----'] 5: ['AAIP6----3A----'] into ['AAIP6----1A----'] 6: ['CrFS6----------'] into ['PSFS6-P1-------', 'AAFS6----1N----'] 7: ['AAIS6----1N----'] into ['PSZS6FS3-------'] 8: ['PZXP6----------'] into ['Ca--6----------'] 9: ['CrNS6----------'] into ['AUNS6M---------'] 10: ['AUFS7M---------'] into ['AAFS7----1N----'] 11: ['PJZS3----------'] into ['P4MS4----------'] 12: ['J^------------8'] into ['VB-S---2P-NA---'] 13: ['AGIS7-----A----'] into ['AAIS7----3A----'] 14: ['ClIS4----------'] into ['CrIS4----------'] 15: ['ClZS6----------'] into ['PSZS6FS3-------', 'AAIS6----1N----'] 16: ['AAFP7----1A----'] into ['AAFP7----2A----'] 17: ['AGFS6-----A----'] into ['P4FS6----------'] 18: ['PSHS1-S1-------'] into ['PSHS1-P1-------'] 19: ['AANS7----1A----'] into ['AANS7----1N----'] 20: ['CrFS2----------'] into ['AAFS2----2A----'] 21: ['AGFS7-----A----'] into ['AAFS7----3A----'] 22: ['PZFS6----------'] into ['ClFS6----------'] 23: ['CrNS2----------'] into ['AUNS2M---------'] 24: ['Cn-P6----------'] into ['PSXP6-P1-------'] 25: ['PZFS1----------'] into ['PWFS1----------'] 26: ['AUFS6M---------'] into ['PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------'] 27: ['AAIP2----2A----'] into ['AAIP2----1N----'] 28: ['P5FS2--3-------'] into ['P5ZS2--3------1'] 29: ['P4FS7----------'] into ['P9ZS7----------'] 30: ['P9XP6----------'] into ['PQ--6----------'] 31: ['PSHP1-P2-------'] into ['CrIP1----------'] 32: ['NNNSX-----A----'] into ['P6-X6----------'] 33: ['AAIP4----1N----'] into ['PZYP4----------'] 34: ['AAIS2----2A----'] into ['AAIS2----3A----'] 35: ['NNISX-----A----'] into ['AAIS1----1A---6'] 36: ['PKM-4----------'] into ['ClHP4----------'] 37: ['AUMS1M---------'] into ['AAMS1----1N----'] 38: ['PQ--2----------'] into ['PKM-3----------'] 39: ['PLIP1----------'] into ['PDIP1----------'] 40: ['AGIS4-----A----'] into ['AAIS4----3A----'] 41: ['PWFS4----------'] into ['PZFS4----------'] 42: ['ACQW------A----'] into ['VB-S---1F-AA---'] 43: ['AGNP2-----A----'] into ['CrNP2----------'] 44: ['AAMS7----3A----'] into ['AAMS7----1A----'] 45: ['CrIS6----------', 'PSZS6-P1-------'] into ['PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------'] 46: ['AAMP1----2A----'] into ['PSMP1-P1-------'] 47: ['PJNS1----------'] into ['PJYS1----------'] 48: ['AAIP1----3A----'] into ['AAIP1----1N----'] 49: ['CrIS7----------'] into ['VeYS------A----'] 50: ['PDFS7----------'] into ['AAFS7----1N----', 'AUFS7M---------'] 51: ['P8XP6----------'] into ['AGFP6-----A----', 'PLXP6----------'] 52: ['AUIS1M---------'] into ['PWYS1----------'] 53: ['P5FS3--3-------'] into ['P5ZS3--3-------'] 54: ['AAIP7----3A----'] into ['NNNP1-----A---2'] 55: ['ClMS4----------'] into ['P8MS4----------'] 56: ['ACTP------A----'] into ['P6-X3----------'] 57: ['NNFS2-----N----'] into ['AGNP6-----A----'] 58: ['AGFP1-----A----'] into ['CrFP1----------'] 59: ['PP-P7--1-------'] into ['P6-X7----------'] 60: ['AAFP2----3A----'] into ['AAFP2----2A----'] 61: ['PSYS1-S1-------'] into ['PSYS1-P1-------'] 62: ['P4NS4----------'] into ['J,-S---1-------'] 63: ['Dg-------3A---1'] into ['Dg-------1A---1'] 64: ['AGFP7-----A----'] into ['P9ZS7----------', 'P4FS7----------'] 65: ['P5XP6--3-------'] into ['P5FS6--3-------'] 66: ['PJXP1----------'] into ['P4XP3----------'] 67: ['AAFS3----1A----'] into ['PDFS3----------'] 68: ['AGNS2-----A----'] into ['NNNS1-----N----'] 69: ['PLZS6----------'] into ['AUIS6M---------'] 70: ['CrNS4----------'] into ['AANS4----1N----'] 71: ['AUFS4M---------'] into ['AAFS4----3A----'] 72: ['AAFS4----1N----'] into ['P8FS4----------'] 73: ['P4YP4----------'] into ['P4MS4----------', 'PJZS3----------'] 74: ['PZIS4----------'] into ['AAIS4----1N----'] 75: ['P8XP7----------'] into ['Cn-P7----------'] 76: ['ClFS2----------'] into ['AAFS2----1N----'] 77: ['PDXP6----------'] into ['Ca--6----------', 'PZXP6----------'] 78: ['AAIS2----1N----'] into ['AUIS2M---------'] 79: ['VB-S---1P-NA--1'] into ['VB-P---1F-AA---'] 80: ['PP-P1--2-------'] into ['AAIP1----2A----'] 81: ['PSZS2-S1-------'] into ['PSZS2FS3-------'] 82: ['AAFP4----1N----'] into ['PSXP4-P1-------'] 83: ['AAMP1----3A----'] into ['AGMP1-----A----'] 84: ['PZZS7----------'] into ['PLNS4---------1'] 85: ['PZM-4----------'] into ['VB-S---1F-AA---', 'ACQW------A----'] 86: ['AAMS6----1A----'] into ['NNMS6-----A---1'] 87: ['Co-------------'] into ['Dg-------1A---3'] 88: ['AAFP1----3A----'] into ['PLFP1----------'] 89: ['P4XP2----------'] into ['P6-X2----------'] 90: ['AANS1----1N----'] into ['AANS1----2A----'] 91: ['PQ--7----------'] into ['J^------------1'] 92: ['PLXP3----------'] into ['ClXP3----------'] 93: ['NNFP1-----N----'] into ['NNMP6-----A----'] 94: ['J,-P---1-------'] into ['Vi-P---2--N----'] 95: ['AGIS2-----A----'] into ['ClZS3----------'] 96: ['Vi-S---2--A----'] into ['VB-S---2P-NA---', 'J^------------8'] 97: ['AAFP1----1N----'] into ['AAFP1----2A----'] 98: ['NNIP2-----A---1'] into ['NNFS1-----N----'] 99: ['P9XP2----------'] into ['P5ZS2--3------1', 'P5FS2--3-------'] 100: ['AAXXX----1A---8'] into ['A2--------A----'] 101: ['NNFP2-----A---1'] into ['NNIS5-----A----'] 102: ['CyFP1----------'] into ['AAIS1----1A---6', 'NNISX-----A----'] 103: ['AAMS1----2A----'] into ['CrMS1----------'] 104: ['P9ZS6----------'] into ['PQ--6----------', 'P9XP6----------'] 105: ['PDXP7----------'] into ['AAIS7----1N----'] 106: ['II-------------'] into ['NNFS5-----A----'] 107: ['C?--4----------'] into ['NNNS1-----N----', 'AGNS2-----A----'] 108: ['CrNP1----------'] into ['AANP1----1A----'] 109: ['CrIP1----------', 'PSHP1-P2-------'] into ['AAIP1----1N----', 'AAIP1----3A----'] 110: ['P8FP4---------1'] into ['PLFP4----------'] 111: ['VB-P---3F-NA---'] into ['VB-P---3P-NA--1'] 112: ['PLNS1---------1'] into ['ClNS1----------'] 113: ['PSZS1FS3-------'] into ['PZYS1----------'] 114: ['P6-X4----------'] into ['P5ZS4--3-------'] 115: ['NNFS2-----A---1'] into ['AGNP6-----A----', 'NNFS2-----N----'] 116: ['AAFS7----3A----', 'AGFS7-----A----'] into ['AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------'] 117: ['PDMS4----------'] into ['VB-P---2P-NA---'] 118: ['PSFS4-P1-------'] into ['PZFS4----------', 'PWFS4----------'] 119: ['CrFS4----------'] into ['AAFS4----3A----', 'AUFS4M---------'] 120: ['AAIS1----1N----'] into ['AAIS1----2A----'] 121: ['PSXP2-P1-------'] into ['AAFP2----2A----', 'AAFP2----3A----'] 122: ['P5XP7--3-------'] into ['P5ZS7--3-------'] 123: ['PDIP4----------'] into ['PZYP4----------', 'AAIP4----1N----'] 124: ['P5ZS6--3-------'] into ['P9FS6----------'] 125: ['PPXP2--3-------'] into ['PPZS7--3-------'] 126: ['ClXP6----------'] into ['PSXP6-P1-------', 'Cn-P6----------'] 127: ['PP-P2--1-------'] into ['P5XP2--3-------'] 128: ['AGIP2-----A----'] into ['AAIP2----1N----', 'AAIP2----2A----'] 129: ['P4ZS6----------'] into ['P8ZS6----------'] 130: ['PZFP1----------'] into ['PSHP1-P1-------'] 131: ['PZMP1----------'] into ['PPMP1--3-------'] 132: ['AAMP2----1N----'] into ['AGMP2-----A----'] 133: ['PZFP4----------'] into ['PWFP4----------'] 134: ['AANS1----3A----'] into ['PLNS4----------'] 135: ['P5FS4--3-------'] into ['P5XP4--3-------'] 136: ['RR--X----------'] into ['RR--1----------'] 137: ['P4ZS7----------'] into ['PKM-3----------', 'PQ--2----------'] 138: ['ClFS4----------'] into ['AAFS4----2A----'] 139: ['P5ZS2--3-------'] into ['P6-X2----------', 'P4XP2----------'] 140: ['AANS2----1N----'] into ['AUNS2M---------', 'CrNS2----------'] 141: ['AUFS2M---------'] into ['AAFS2----2A----', 'CrFS2----------'] 142: ['Vf--------A---1'] into ['NNNP1-----A---2', 'AAIP7----3A----'] 143: ['PJYS1----------', 'PJNS1----------'] into ['J,-S---1-------', 'P4NS4----------'] 144: ['AAIS1----3A----'] into ['CrIS1----------'] 145: ['AGIS6-----A----'] into ['ClFS6----------', 'PZFS6----------'] 146: ['PSHS1-P1-------', 'PSHS1-S1-------'] into ['PWFS1----------', 'PZFS1----------'] 147: ['Vi-P---1--A----'] into ['Vi-P---2--A----'] 148: ['ClZS2----------'] into ['CrIS2----------'] 149: ['P6-X6----------', 'NNNSX-----A----'] into ['P5FS6--3-------', 'P5XP6--3-------'] 150: ['ClXP7----------'] into ['Cn-P7----------', 'P8XP7----------'] 151: ['AGMS1-----A----'] into ['AAMS1----1N----', 'AUMS1M---------'] 152: ['PLFS1----------'] into ['P6-X3----------', 'ACTP------A----'] 153: ['PDMP4----------'] into ['PLNS4---------1', 'PZZS7----------'] 154: ['PLXP7----------'] into ['NNMP2-----A---1'] 155: ['PWYS1----------', 'AUIS1M---------'] into ['PZYS1----------', 'PSZS1FS3-------'] 156: ['NNIXX-----A---8'] into ['NNMXX-----A---8'] 157: ['P4FP1----------'] into ['P4IP1----------'] 158: ['P8FS6---------1'] into ['P4FS6----------', 'AGFS6-----A----'] 159: ['AUIS6M---------', 'PLZS6----------'] into ['P8ZS6----------', 'P4ZS6----------'] 160: ['AAIS7----3A----', 'AGIS7-----A----'] into ['AAIS7----1N----', 'PDXP7----------'] 161: ['PPNS4--3-------'] into ['P8ZS3----------'] 162: ['AGFS2-----A----'] into ['PSFS2-P1-------'] 163: ['AAFS6----1A----'] into ['PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------', 'AUFS6M---------'] 164: ['P4FS4----------'] into ['P4MS4----------', 'PJZS3----------', 'P4YP4----------'] 165: ['AAMP3----1A----'] into ['P8XP3----------'] 166: ['ClFS1----------'] into ['AAFS1----3A----'] 167: ['P8YP4---------1'] into ['PLYP4----------'] 168: ['AAIS2----3A----', 'AAIS2----2A----'] into ['CrIS2----------', 'ClZS2----------'] 169: ['AUFS1M---------'] into ['NNFS1-----A---1'] 170: ['PSMP1-P1-------', 'AAMP1----2A----'] into ['AGMP1-----A----', 'AAMP1----3A----'] 171: ['AAFP2----1N----'] into ['AGFP2-----A----'] 172: ['AAMP1----1N----'] into ['CrFP1----------', 'AGFP1-----A----'] 173: ['AAMP2----3A----'] into ['Ca--2----------'] 174: ['PDFP4----------'] into ['ClHP4----------', 'PKM-4----------'] 175: ['P1XXXXP3-------'] into ['P1XXXZS3-------'] 176: ['Dg-------1A---1', 'Dg-------3A---1'] into ['Dg-------1A---3', 'Co-------------'] 177: ['ClZS7----------'] into ['ClFS7----------'] 178: ['Cn-P2----------'] into ['CrNP2----------', 'AGNP2-----A----'] 179: ['PP-P4--2-------'] into ['PP-S4--1-------'] 180: ['P4IS4----------'] into ['P4FP4----------'] 181: ['PP-P1--1-------'] into ['AAIP1----2A----', 'PP-P1--2-------'] 182: ['AAMS4----1A----'] into ['P8MS4----------', 'ClMS4----------'] 183: ['P8FS4----------', 'AAFS4----1N----'] into ['AAFS4----2A----', 'ClFS4----------'] 184: ['AUNS6M---------', 'CrNS6----------'] into ['PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------'] 185: ['AAFS1----2A----'] into ['AAFS1----1N----'] 186: ['CrNS1----------'] into ['AUNS1M---------'] 187: ['P8NS4---------1'] into ['AANS4----1N----', 'CrNS4----------'] 188: ['NNFS4-----N----'] into ['AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1'] 189: ['NNIS2-----A---1'] into ['ClNS1----------', 'PLNS1---------1'] 190: ['P8IS4----------'] into ['AAIS4----3A----', 'AGIS4-----A----'] 191: ['P8XP2----------'] into ['PZXP2----------'] 192: ['PJYS2----------'] into ['Vi-P---2--N----', 'J,-P---1-------'] 193: ['VpTP---XR-AA--1'] into ['VpNS---XR-AA--1'] 194: ['PDXP3----------'] into ['ClXP3----------', 'PLXP3----------'] 195: ['P4FS1----------'] into ['PE--1----------'] 196: ['PDIP1----------', 'PLIP1----------'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------'] 197: ['VB-P---1F-AA---', 'VB-S---1P-NA--1'] into ['VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------'] 198: ['PPYS1--3-------'] into ['AGIP1-----A----'] 199: ['AANP7----1A----'] into ['NNIS5-----A----', 'NNFP2-----A---1'] 200: ['P8FS2---------1'] into ['AAFS2----1N----', 'ClFS2----------'] 201: ['VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----'] into ['NNFS5-----A----', 'II-------------'] 202: ['ClZS3----------', 'AGIS2-----A----'] into ['VB-P---2P-NA---', 'PDMS4----------'] 203: ['P4NS1----------'] into ['P4XP3----------', 'PJXP1----------'] 204: ['PPFS3--3-------'] into ['PZ--1----------'] 205: ['CrMS1----------', 'AAMS1----2A----'] into ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----'] 206: ['CrFS1----------'] into ['PSFSXFS3-------'] 207: ['P9FS6----------', 'P5ZS6--3-------'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----'] 208: ['PSXP4-P1-------', 'AAFP4----1N----'] into ['PLNS4---------1', 'PZZS7----------', 'PDMP4----------'] 209: ['PDIS4----------'] into ['CrIS4----------', 'ClIS4----------'] 210: ['P8FS7---------1'] into ['AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----'] 211: ['NNNP3-----A----'] into ['NNFS1-----N----', 'NNIP2-----A---1'] 212: ['PWFP4----------', 'PZFP4----------'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------'] 213: ['ClHP1----------'] into ['PLFP1----------', 'AAFP1----3A----'] 214: ['NNMS6-----A---1', 'AAMS6----1A----'] into ['NNMP6-----A----', 'NNFP1-----N----'] 215: ['VB-S---2P-AA---'] into ['AGNS1-----A----'] 216: ['PSZS2-P1-------'] into ['P8ZS2----------'] 217: ['PSZS2FS3-------', 'PSZS2-S1-------'] into ['P8ZS2----------', 'PSZS2-P1-------'] 218: ['VeYS------A----', 'CrIS7----------'] into ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----'] 219: ['AAIS2----1A----'] into ['AUIS2M---------', 'AAIS2----1N----'] 220: ['P8FS4---------1'] into ['PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------'] 221: ['AAIS4----1N----', 'PZIS4----------'] into ['AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------'] 222: ['ClYP4----------'] into ['ClXP4----------'] 223: ['PLNS4----------', 'AANS1----3A----'] into ['AGNS1-----A----', 'VB-S---2P-AA---'] 224: ['ClYS1----------'] into ['AGIS1-----A----'] 225: ['PW--1----------'] into ['PW--4----------'] 226: ['NNMS3-----A---1'] into ['AAMS3----1A----'] 227: ['VB-S---3F-NA---'] into ['J^------------1', 'PQ--7----------'] 228: ['P5ZS3--3-------', 'P5FS3--3-------'] into ['P6-X3----------', 'ACTP------A----', 'PLFS1----------'] 229: ['Vc-P---1-------'] into ['ACMP------A----'] 230: ['PZM-1----------'] into ['PP-P3--2-------'] 231: ['P4YS1----------'] into ['J,-S---1-------', 'P4NS4----------', 'PJYS1----------', 'PJNS1----------'] 232: ['VB-P---1P-NA---'] into ['VB-S---1P-AA--1'] 233: ['PPMP1--3-------', 'PZMP1----------'] into ['AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----'] 234: ['RF-------------'] into ['NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------'] 235: ['P5ZS4--3-------', 'P6-X4----------'] into ['P5XP4--3-------', 'P5FS4--3-------'] 236: ['VpTP---XR-NA---'] into ['VB-P---3P-NA--1', 'VB-P---3F-NA---'] 237: ['PPZS7--3-------', 'PPXP2--3-------'] into ['P8ZS3----------', 'PPNS4--3-------'] 238: ['CrIS1----------', 'AAIS1----3A----'] into ['PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------'] 239: ['PQ--1----------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------'] 240: ['A2--------A----', 'AAXXX----1A---8'] into ['NNMXX-----A---8', 'NNIXX-----A---8'] 241: ['AANS2----1A----'] into ['AUNS2M---------', 'CrNS2----------', 'AANS2----1N----'] 242: ['NNMS5-----A----'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------'] 243: ['NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1'] into ['CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----'] 244: ['PDFS1----------'] into ['PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------'] 245: ['PSYS1-P1-------', 'PSYS1-S1-------'] into ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----'] 246: ['ClXP2----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------'] 247: ['PDFS2----------'] into ['PSFS2-P1-------', 'AGFS2-----A----'] 248: ['PKM-1----------'] into ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------'] 249: ['Ca--2----------', 'AAMP2----3A----'] into ['PZXP2----------', 'P8XP2----------'] 250: ['PH-S4--1-------'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1'] 251: ['AAIP2----1A----'] into ['AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----'] 252: ['AANP6----1A----'] into ['PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------'] 253: ['Ca--1----------'] into ['Ca--4----------'] 254: ['PDFS4----------'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------'] 255: ['AAIP4----1A----'] into ['PZYP4----------', 'AAIP4----1N----', 'PDIP4----------'] 256: ['NNMS6-----A----'] into ['NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----'] 257: ['PZ--4----------'] into ['ACYS------A----'] 258: ['Vf--------N----'] into ['AANS1----2A----', 'AANS1----1N----'] 259: ['P8ZS7----------'] into ['AANP3----1A----'] 260: ['AAMP7----1A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----'] 261: ['PP-P4--1-------'] into ['PP-S4--1-------', 'PP-P4--2-------'] 262: ['RR--1----------', 'RR--X----------'] into ['NNFS1-----A---1', 'AUFS1M---------'] 263: ['AGFS1-----A----'] into ['AAFS1----3A----', 'ClFS1----------'] 264: ['PSHP1-P1-------', 'PZFP1----------'] into ['PLFP1----------', 'AAFP1----3A----', 'ClHP1----------'] 265: ['VpQW---XR-NA---'] into ['VpQW---XR-AA--1'] 266: ['PQ--6----------', 'P9XP6----------', 'P9ZS6----------'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------'] 267: ['AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------'] 268: ['PPFS4--3-------'] into ['PHZS4--3-------'] 269: ['ClXP1----------'] into ['AAFP1----2A----', 'AAFP1----1N----'] 270: ['NNFSX-----A----'] into ['NNIPX-----A----'] 271: ['NNMP2-----A---1', 'PLXP7----------'] into ['AGIP1-----A----', 'PPYS1--3-------'] 272: ['PDMP1----------'] into ['PLMP1----------'] 273: ['PDXP2----------'] into ['AGFP2-----A----', 'AAFP2----1N----'] 274: ['PWM-1----------'] into ['PLNS1----------'] 275: ['AAFS7----1A----'] into ['AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1'] 276: ['PLFP4----------', 'P8FP4---------1'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------'] 277: ['PP-S1--1-------'] into ['AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------'] 278: ['PPXP4--3-------'] into ['PPZS4--3------2'] 279: ['VpMP---XR-AA--1'] into ['VpMP---XR-NA---'] 280: ['PLXP2----------'] into ['PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----'] 281: ['VB-P---3P-AA--1'] into ['VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1'] 282: ['AANP4----1A----'] into ['PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----'] 283: ['AANS4----1A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1'] 284: ['Vi-P---2--A----', 'Vi-P---1--A----'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----'] 285: ['PSXXXXP3-------'] into ['P1XXXZS3-------', 'P1XXXXP3-------'] 286: ['AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1'] into ['PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------'] 287: ['AAMP2----1A----'] into ['AGMP2-----A----', 'AAMP2----1N----'] 288: ['NNNP7-----A----'] into ['P6-X7----------', 'PP-P7--1-------'] 289: ['P5ZS7--3-------', 'P5XP7--3-------'] into ['P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----'] 290: ['Db------------8'] into ['NNFXX-----A---8'] 291: ['Dg-------3A----'] into ['ClFS7----------', 'ClZS7----------'] 292: ['P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------'] into ['Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------'] 293: ['P4FP4----------', 'P4IS4----------'] into ['Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------', 'P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------'] 294: ['AGFS4-----A----'] into ['VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----'] 295: ['PHZS3--3-------'] into ['PZ--1----------', 'PPFS3--3-------'] 296: ['Cn-P7----------', 'P8XP7----------', 'ClXP7----------'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----'] 297: ['PPXP3--3-------'] into ['PP-P3--2-------', 'PZM-1----------'] 298: ['P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1'] into ['P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------'] 299: ['ClFS6----------', 'PZFS6----------', 'AGIS6-----A----'] into ['P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1'] 300: ['AAIP6----1A----', 'AAIP6----3A----'] into ['Ca--6----------', 'PZXP6----------', 'PDXP6----------'] 301: ['P4MP1----------'] into ['P4IP1----------', 'P4FP1----------'] 302: ['VsMP---XX-AP---'] into ['ACMP------A----', 'Vc-P---1-------'] 303: ['AAFP6----1A----'] into ['AGFP6-----A----', 'PLXP6----------', 'P8XP6----------'] 304: ['P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------'] into ['ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1'] 305: ['ClXP3----------', 'PLXP3----------', 'PDXP3----------'] into ['P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------'] 306: ['AAIS4----1A----'] into ['AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------'] 307: ['C}-------------'] into ['NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8'] 308: ['AAIS1----2A----', 'AAIS1----1N----'] into ['AGIS1-----A----', 'ClYS1----------'] 309: ['P5XP2--3-------', 'PP-P2--1-------'] into ['ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------'] 310: ['ClYP1----------'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------'] 311: ['AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------'] into ['PLMP1----------', 'PDMP1----------'] 312: ['PLYS1----------'] into ['PDYS1----------'] 313: ['PDFP1----------'] into ['PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------'] 314: ['VB-P---2P-AA---'] into ['VB-S---1P-NA---'] 315: ['PSFSXFS3-------', 'CrFS1----------'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----'] 316: ['Vc-S---1-------'] into ['VB-S---1P-AA---'] 317: ['VpNS---XR-AA--1', 'VpTP---XR-AA--1'] into ['VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---'] 318: ['PP-P3--1-------'] into ['PH-S3--1-------'] 319: ['NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---'] 320: ['AANS6----1A----'] into ['PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------'] 321: ['PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----'] into ['PDYS1----------', 'PLYS1----------'] 322: ['PDZS7----------'] into ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------'] 323: ['AANS1----1A----'] into ['AUNS1M---------', 'CrNS1----------'] 324: ['AAMP4----1A----'] into ['PLYP4----------', 'P8YP4---------1'] 325: ['CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----'] into ['AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----'] 326: ['Dg-------2A---1'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------'] 327: ['Db------------1'] into ['Cv-------------'] 328: ['P8XP3----------', 'AAMP3----1A----'] into ['P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------'] 329: ['NNFPX-----A----'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----'] 330: ['ACNS------A----'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----'] 331: ['VB-S---1P-AA--1', 'VB-P---1P-NA---'] into ['VB-S---1P-NA---', 'VB-P---2P-AA---'] 332: ['CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------'] 333: ['AAFS1----1N----', 'AAFS1----2A----'] into ['AANS1----2A----', 'AANS1----1N----', 'Vf--------N----'] 334: ['Cn-S1----------'] into ['Ca--4----------', 'Ca--1----------'] 335: ['CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1'] into ['AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------'] 336: ['PDZS2----------'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------'] 337: ['PE--1----------', 'P4FS1----------'] into ['J,-S---1-------', 'P4NS4----------', 'PJYS1----------', 'PJNS1----------', 'P4YS1----------'] 338: ['P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------'] into ['PPZS4--3------2', 'PPXP4--3-------'] 339: ['AAIP3----1A----'] into ['AANP3----1A----', 'P8ZS7----------'] 340: ['PW--4----------', 'PW--1----------'] into ['PLNS1----------', 'PWM-1----------'] 341: ['NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------'] into ['AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------'] 342: ['Dg-------1N----'] into ['ACYS------A----', 'PZ--4----------'] 343: ['NNMS3-----A----'] into ['AAMS3----1A----', 'NNMS3-----A---1'] 344: ['AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----'] into ['VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----'] 345: ['VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----'] into ['AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----'] 346: ['AAMS1----1A----'] into ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------'] 347: ['PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----'] 348: ['NNIS6-----A----'] into ['NNIS6-----A---1'] 349: ['NNMP1-----A----'] into ['NNMP1-----A---1'] 350: ['AAMP1----1A----'] into ['PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------'] 351: ['PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----'] into ['AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----'] 352: ['NNFP6-----A----'] into ['NNNP6-----A----'] 353: ['ClXP4----------', 'ClYP4----------'] into ['PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----'] 354: ['PQ--4----------'] into ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------', 'PKM-1----------'] 355: ['RV--3----------'] into ['RR--3----------'] 356: ['VpYS---XR-AA--1'] into ['VpYS---XR-AA---'] 357: ['AAFP3----1A----'] into ['NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----'] 358: ['AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------'] 359: ['VB-P---3P-NA---'] into ['VpMP---XR-NA---', 'VpMP---XR-AA--1'] 360: ['VpTP---XR-AA---'] into ['VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1'] 361: ['ClXS2----------'] into ['PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------'] 362: ['Cn-S4----------'] into ['Ca--4----------', 'Ca--1----------', 'Cn-S1----------'] 363: ['PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------'] into ['PH-S3--1-------', 'PP-P3--1-------'] 364: ['NNIPX-----A----', 'NNFSX-----A----'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----'] 365: ['PHZS4--3-------', 'PPFS4--3-------'] into ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------'] 366: ['VB-P---3F-AA---'] into ['VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1'] 367: ['NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----'] 368: ['Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------', 'P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------', 'P4FP4----------', 'P4IS4----------'] into ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------', 'PKM-1----------', 'PQ--4----------'] 369: ['RR--4----------'] into ['RV--4----------'] 370: ['AAFP4----1A----'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1'] 371: ['PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------'] 372: ['VpNS---XR-NA---'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----'] 373: ['NNFXX-----A---8', 'Db------------8'] into ['NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------'] 374: ['PDZS6----------'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------'] 375: ['NNNP4-----A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------'] 376: ['NNMP7-----A----'] into ['P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------'] 377: ['PDZS3----------'] into ['AAIS3----1A----'] 378: ['NNMS4-----A----'] into ['P8MS4----------', 'ClMS4----------', 'AAMS4----1A----'] 379: ['Dg-------2A----'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1'] 380: ['PDNS4----------'] into ['PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------'] 381: ['AAFP7----2A----', 'AAFP7----1A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------'] 382: ['P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------'] 383: ['RV--2----------'] into ['AAFS2----2A----', 'CrFS2----------', 'AUFS2M---------'] 384: ['J,-X---3-------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------', 'PQ--1----------'] 385: ['AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------'] into ['AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------'] 386: ['NNFP3-----A----'] into ['NNIP3-----A----'] 387: ['NNIP6-----A----'] into ['NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----', 'NNMS6-----A----'] 388: ['ClFS7----------', 'ClZS7----------', 'Dg-------3A----'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----'] 389: ['NNFP7-----A----'] into ['P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----'] 390: ['AAFS4----1A----'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1'] 391: ['VsYS---XX-AP---'] into ['AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----'] 392: ['PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------'] into ['AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1'] 393: ['AAIP1----1A----'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------'] 394: ['RV--7----------'] into ['RR--7----------'] 395: ['P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----'] 396: ['NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----'] 397: ['NNMP3-----A----'] into ['NNIP3-----A----', 'NNFP3-----A----'] 398: ['VB-S---3F-AA---'] into ['VpQW---XR-AA--1', 'VpQW---XR-NA---'] 399: ['VB-S---3P-NA---'] into ['VpYS---XR-NA---'] 400: ['ACYS------A----', 'PZ--4----------', 'Dg-------1N----'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----'] 401: ['Db-------------'] into ['Cv-------------', 'Db------------1'] 402: ['PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------'] into ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------'] 403: ['PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------'] into ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------'] 404: ['VsTP---XX-AP---'] into ['ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---'] 405: ['AANS7----1N----', 'AANS7----1A----'] into ['AAMS7----1A----', 'AAMS7----3A----'] 406: ['XX-------------'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------'] 407: ['PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----'] 408: ['VB-P---1P-AA---'] into ['VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---'] 409: ['VpMP---XR-AA---'] into ['VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---'] 410: ['AAIS7----1A----'] into ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------'] 411: ['VsQW---XX-AP---'] into ['VsNS---XX-AP---'] 412: ['PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----'] into ['PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------'] 413: ['NNIP4-----A----'] into ['NNMP4-----A----'] 414: ['NNNXX-----A----'] into ['NNIXX-----A----'] 415: ['PDFS6----------', 'AAFS6----3A----'] into ['Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----'] 416: ['NNNS3-----A----'] into ['AANS3----1A----'] 417: ['AAIP7----1A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----'] 418: ['P4IP1----------', 'P4FP1----------', 'P4MP1----------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------', 'PQ--1----------', 'J,-X---3-------'] 419: ['AAXXX----1A----'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----'] 420: ['AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------'] 421: ['AANP1----1A----', 'CrNP1----------'] into ['CrIS4----------', 'ClIS4----------', 'PDIS4----------'] 422: ['VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---'] into ['VpYS---XR-NA---', 'VB-S---3P-NA---'] 423: ['VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---'] 424: ['NNMS2-----A----'] into ['AAMS2----1A----'] 425: ['AAMS2----1A----', 'NNMS2-----A----'] into ['PDFS3----------', 'AAFS3----1A----'] 426: ['PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------'] 427: ['P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----'] 428: ['P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----'] 429: ['Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----'] 430: ['AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----'] 431: ['NNNP6-----A----', 'NNFP6-----A----'] into ['NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----', 'NNMS6-----A----', 'NNIP6-----A----'] 432: ['RV--6----------'] into ['RR--6----------'] 433: ['RR--2----------'] into ['AAFS2----2A----', 'CrFS2----------', 'AUFS2M---------', 'RV--2----------'] 434: ['ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------'] 435: ['J,-S---1-------', 'P4NS4----------', 'PJYS1----------', 'PJNS1----------', 'P4YS1----------', 'PE--1----------', 'P4FS1----------'] into ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------', 'PKM-1----------', 'PQ--4----------', 'Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------', 'P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------', 'P4FP4----------', 'P4IS4----------'] 436: ['AAFP1----1A----'] into ['AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------'] 437: ['Xx-------------'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----'] 438: ['AAIS1----1A----'] into ['PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----'] 439: ['TT-------------'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----'] 440: ['AANP2----1A----'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------'] 441: ['PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----'] 442: ['X@-------------'] into ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------', 'XX-------------'] 443: ['VB-P---3P-AA---'] into ['VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---'] 444: ['NNFS3-----A----'] into ['NNIP3-----A----', 'NNFP3-----A----', 'NNMP3-----A----'] 445: ['Dg-------1A----'] into ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----'] 446: ['AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----'] into ['CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------'] 447: ['AAFS1----1A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----'] 448: ['P7-X4----------'] into ['P7-X3----------'] 449: ['NNNS7-----A----'] into ['NNMS7-----A----'] 450: ['NNNP1-----A----'] into ['NNFP1-----A----'] 451: ['AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----'] into ['NNIP7-----A----'] 452: ['AAIS3----1A----', 'PDZS3----------'] into ['AANS3----1A----', 'NNNS3-----A----'] 453: ['P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----'] 454: ['AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---'] into ['VsNS---XX-AP---', 'VsQW---XX-AP---'] 455: ['NNFXX-----A----'] into ['NNIXX-----A----', 'NNNXX-----A----'] 456: ['VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---'] 457: ['PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------'] into ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------'] 458: ['AAIS6----1A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------'] 459: ['NNIS3-----A----'] into ['AANS3----1A----', 'NNNS3-----A----', 'AAIS3----1A----', 'PDZS3----------'] 460: ['NNNP2-----A----'] into ['NNMP2-----A----'] 461: ['AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------'] 462: ['Ca--4----------', 'Ca--1----------', 'Cn-S1----------', 'Cn-S4----------'] into ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----', 'PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------'] 463: ['Cv-------------', 'Db------------1', 'Db-------------'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------'] 464: ['Vf--------A----'] into ['ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---'] 465: ['VpNS---XR-AA---'] into ['VpQW---XR-AA---'] 466: ['PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------'] 467: ['P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----'] 468: ['Vc-X---3-------'] into ['VB-S---1P-AA---', 'Vc-S---1-------'] 469: ['PDNS1----------'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------'] 470: ['NNFP4-----A----'] into ['NNMP4-----A----', 'NNIP4-----A----'] 471: ['AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----'] 472: ['NNMS7-----A----', 'NNNS7-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----', 'P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----'] 473: ['NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------', 'NNFXX-----A---8', 'Db------------8'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----', 'Xx-------------'] 474: ['AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------', 'XX-------------', 'X@-------------'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----', 'Xx-------------', 'NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------', 'NNFXX-----A---8', 'Db------------8'] 475: ['VB-S---3P-AA---'] into ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---'] 476: ['VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'] into ['VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------'] 477: ['PSXXXZS3-------'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1'] 478: ['PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----'] 479: ['J,-------------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------', 'PQ--1----------', 'J,-X---3-------', 'P4IP1----------', 'P4FP1----------', 'P4MP1----------'] 480: ['NNNS4-----A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----'] 481: ['PKM-3----------', 'PQ--2----------', 'P4ZS7----------', 'PKM-1----------', 'PQ--4----------', 'Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------', 'P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------', 'P4FP4----------', 'P4IS4----------', 'J,-S---1-------', 'P4NS4----------', 'PJYS1----------', 'PJNS1----------', 'P4YS1----------', 'PE--1----------', 'P4FS1----------'] into ['P4XP3----------', 'PJXP1----------', 'P4NS1----------', 'PQ--1----------', 'J,-X---3-------', 'P4IP1----------', 'P4FP1----------', 'P4MP1----------', 'J,-------------'] 482: ['AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----', 'AAIP2----1A----'] into ['AGMP2-----A----', 'AAMP2----1N----', 'AAMP2----1A----'] 483: ['NNIP2-----A----'] into ['NNMP2-----A----', 'NNNP2-----A----'] 484: ['AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----', 'PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------', 'Ca--4----------', 'Ca--1----------', 'Cn-S1----------', 'Cn-S4----------'] into ['AGMP2-----A----', 'AAMP2----1N----', 'AAMP2----1A----', 'AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----', 'AAIP2----1A----'] 485: ['NNNS6-----A----'] into ['NNIS6-----A---1', 'NNIS6-----A----'] 486: ['PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------', 'AANS6----1A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------', 'AAIS6----1A----'] 487: ['Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----'] into ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------'] 488: ['NNFS7-----A----'] into ['NNIS7-----A----'] 489: ['NNMS1-----A----'] into ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----'] 490: ['AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----'] into ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----'] 491: ['AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----', 'AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----'] into ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----', 'AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----'] 492: ['P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----', 'AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----', 'AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----', 'AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----', 'P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----', 'NNMS7-----A----', 'NNNS7-----A----'] 493: ['NNIXX-----A----', 'NNNXX-----A----', 'NNFXX-----A----'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----', 'Xx-------------', 'NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------', 'NNFXX-----A---8', 'Db------------8', 'AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------', 'XX-------------', 'X@-------------'] 494: ['AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----'] 495: ['NNIP3-----A----', 'NNFP3-----A----', 'NNMP3-----A----', 'NNFS3-----A----'] into ['AANS3----1A----', 'NNNS3-----A----', 'AAIS3----1A----', 'PDZS3----------', 'NNIS3-----A----'] 496: ['PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----'] 497: ['PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------'] 498: ['VpYS---XR-AA---', 'VpYS---XR-AA--1'] into ['VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---'] 499: ['NNIP1-----A----'] into ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----'] 500: ['AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----'] 501: ['AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------'] 502: ['VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---'] into ['VpQW---XR-AA---', 'VpNS---XR-AA---'] 503: ['AUNS1M---------', 'CrNS1----------', 'AANS1----1A----'] into ['VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---'] 504: ['NNNS1-----A----'] into ['NNMP1-----A---1', 'NNMP1-----A----'] 505: ['J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---'] into ['VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---'] 506: ['NNIS2-----A----'] into ['NNNS2-----A----'] 507: ['NNFP2-----A----'] into ['NNMP2-----A----', 'NNNP2-----A----', 'NNIP2-----A----'] 508: ['AAFP2----1A----'] into ['AGMP2-----A----', 'AAMP2----1N----', 'AAMP2----1A----', 'AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----', 'AAIP2----1A----', 'AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----', 'PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------', 'Ca--4----------', 'Ca--1----------', 'Cn-S1----------', 'Cn-S4----------'] 509: ['AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------'] 510: ['NNIS4-----A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----'] 511: ['CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----'] into ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----'] 512: ['ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----'] into ['VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----'] 513: ['NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----', 'NNMS6-----A----', 'NNIP6-----A----', 'NNNP6-----A----', 'NNFP6-----A----'] into ['NNIS6-----A---1', 'NNIS6-----A----', 'NNNS6-----A----'] 514: ['VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'] into ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----'] 515: ['NNFP1-----A----', 'NNNP1-----A----'] into ['NNMP1-----A---1', 'NNMP1-----A----', 'NNNS1-----A----'] 516: ['AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----'] 517: ['NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----'] 518: ['ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----'] 519: ['P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['NNNS2-----A----', 'NNIS2-----A----'] 520: ['NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----'] 521: ['VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----'] 522: ['NNIS7-----A----', 'NNFS7-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----', 'P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----', 'NNMS7-----A----', 'NNNS7-----A----', 'P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----', 'AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----', 'AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----', 'AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----'] 523: ['PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------', 'AUFS6M---------', 'AAFS6----1A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------', 'AAIS6----1A----', 'PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------', 'AANS6----1A----'] 524: ['RR--3----------', 'RV--3----------'] into ['RR--7----------', 'RV--7----------'] 525: ['AANS3----1A----', 'NNNS3-----A----', 'AAIS3----1A----', 'PDZS3----------', 'NNIS3-----A----', 'NNIP3-----A----', 'NNFP3-----A----', 'NNMP3-----A----', 'NNFS3-----A----'] into ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----', 'P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----', 'NNMS7-----A----', 'NNNS7-----A----', 'P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----', 'AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----', 'AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----', 'AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----', 'NNIS7-----A----', 'NNFS7-----A----'] 526: ['NNFS1-----A----'] into ['NNIS1-----A----'] 527: ['PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----'] 528: ['C=-------------'] into ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----', 'Xx-------------', 'NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------', 'NNFXX-----A---8', 'Db------------8', 'AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------', 'XX-------------', 'X@-------------', 'NNIXX-----A----', 'NNNXX-----A----', 'NNFXX-----A----'] 529: ['RV--4----------', 'RR--4----------'] into ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1', 'AAFS4----1A----'] 530: ['NNMP1-----A---1', 'NNMP1-----A----', 'NNNS1-----A----', 'NNFP1-----A----', 'NNNP1-----A----'] into ['NNIS1-----A----', 'NNFS1-----A----'] 531: ['NNFS2-----A----'] into ['AUIS2M---------', 'AAIS2----1N----', 'AAIS2----1A----'] 532: ['J^-------------'] into ['P7-X3----------', 'P7-X4----------'] 533: ['VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1'] into ['P7-X3----------', 'P7-X4----------', 'J^-------------'] 534: ['VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---'] into ['P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1'] 535: ['NNFS6-----A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------', 'AAIS6----1A----', 'PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------', 'AANS6----1A----', 'PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------', 'AUFS6M---------', 'AAFS6----1A----'] 536: ['NNIS6-----A---1', 'NNIS6-----A----', 'NNNS6-----A----', 'NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----', 'NNMS6-----A----', 'NNIP6-----A----', 'NNNP6-----A----', 'NNFP6-----A----'] into ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------', 'AAIS6----1A----', 'PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------', 'AANS6----1A----', 'PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------', 'AUFS6M---------', 'AAFS6----1A----', 'NNFS6-----A----'] 537: ['NNFS4-----A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----'] 538: ['AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----'] 539: ['PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'] into ['P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---'] 540: ['P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---', 'PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----', 'AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----'] 541: ['AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1', 'AAFS4----1A----', 'RV--4----------', 'RR--4----------'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----', 'AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----', 'P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---', 'PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---'] 542: ['AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['AAFS2----1A----'] 543: ['AAFS2----1A----', 'AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----', 'PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----'] 544: ['AUIS2M---------', 'AAIS2----1N----', 'AAIS2----1A----', 'NNFS2-----A----'] into ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----', 'PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----', 'AAFS2----1A----', 'AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----'] 545: ['AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----', 'PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----', 'AAFS2----1A----', 'AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----', 'AUIS2M---------', 'AAIS2----1N----', 'AAIS2----1A----', 'NNFS2-----A----'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----', 'AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----', 'P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---', 'PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---', 'AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1', 'AAFS4----1A----', 'RV--4----------', 'RR--4----------'] 546: ['AAFS2----2A----', 'CrFS2----------', 'AUFS2M---------', 'RV--2----------', 'RR--2----------'] into ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----', 'AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----', 'P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---', 'PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---', 'AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1', 'AAFS4----1A----', 'RV--4----------', 'RR--4----------', 'AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----', 'PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----', 'AAFS2----1A----', 'AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----', 'AUIS2M---------', 'AAIS2----1N----', 'AAIS2----1A----', 'NNFS2-----A----'] 547: ['NNMP2-----A----', 'NNNP2-----A----', 'NNIP2-----A----', 'NNFP2-----A----'] into ['AGMP2-----A----', 'AAMP2----1N----', 'AAMP2----1A----', 'AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----', 'AAIP2----1A----', 'AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----', 'PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------', 'Ca--4----------', 'Ca--1----------', 'Cn-S1----------', 'Cn-S4----------', 'AAFP2----1A----'] Class 1: ['P5XP4--3-------', 'P5FS4--3-------', 'P5ZS4--3-------', 'P6-X4----------', 'NNNP4-----A----', 'PLNS4---------1', 'PZZS7----------', 'PDMP4----------', 'PSXP4-P1-------', 'AAFP4----1N----', 'AANP4----1A----', 'AGFP6-----A----', 'PLXP6----------', 'P8XP6----------', 'AAFP6----1A----', 'PSXP6-P1-------', 'Cn-P6----------', 'ClXP6----------', 'AANP6----1A----', 'Ca--6----------', 'PZXP6----------', 'PDXP6----------', 'AAIP6----1A----', 'AAIP6----3A----', 'PDFS6----------', 'AAFS6----3A----', 'P8ZS6----------', 'P4ZS6----------', 'AUIS6M---------', 'PLZS6----------', 'P4FS6----------', 'AGFS6-----A----', 'P8FS6---------1', 'ClFS6----------', 'PZFS6----------', 'AGIS6-----A----', 'P1XXXZS3-------', 'P1XXXXP3-------', 'PSXXXXP3-------', 'PDFS3----------', 'AAFS3----1A----', 'AAMS2----1A----', 'NNMS2-----A----', 'NNFS4-----A----', 'AANS4----1N----', 'CrNS4----------', 'P8NS4---------1', 'AANS4----1A----', 'P8MS4----------', 'ClMS4----------', 'AAMS4----1A----', 'NNMS4-----A----', 'NNNS4-----A----', 'NNIS4-----A----', 'CrIS4----------', 'ClIS4----------', 'PDIS4----------', 'AANP1----1A----', 'CrNP1----------', 'AAIS4----3A----', 'AGIS4-----A----', 'P8IS4----------', 'AAIS4----1N----', 'PZIS4----------', 'AAIS4----1A----', 'P7-X3----------', 'P7-X4----------', 'J^-------------', 'VpMP---XR-NA---', 'VpMP---XR-AA--1', 'VB-P---3P-NA---', 'VpMP---XR-AA---', 'VpYS---XR-AA---', 'VpYS---XR-AA--1', 'VpQW---XR-AA---', 'VpNS---XR-AA---', 'VB-P---3P-NA--1', 'VB-P---3F-NA---', 'VpTP---XR-NA---', 'VpNS---XR-AA--1', 'VpTP---XR-AA--1', 'VpTP---XR-AA---', 'VB-P---3P-AA---', 'J^------------1', 'PQ--7----------', 'VB-S---3F-NA---', 'NNIS5-----A----', 'NNFP2-----A---1', 'AANP7----1A----', 'VpNS---XR-NA---', 'VB-S---1F-AA---', 'ACQW------A----', 'PZM-4----------', 'VB-P---1F-AA---', 'VB-S---1P-NA--1', 'VB-P---3P-AA--1', 'VB-P---3F-AA---', 'VpYS---XR-NA---', 'VB-S---3P-NA---', 'VpQW---XR-AA--1', 'VpQW---XR-NA---', 'VB-S---3F-AA---', 'VB-S---3P-AA---', 'PLNS1----------', 'PWM-1----------', 'PW--4----------', 'PW--1----------', 'ClNS1----------', 'PLNS1---------1', 'NNIS2-----A---1', 'P6-X2----------', 'P4XP2----------', 'P5ZS2--3-------', 'P5XP2--3-------', 'PP-P2--1-------', 'AGIP1-----A----', 'PPYS1--3-------', 'NNMP2-----A---1', 'PLXP7----------', 'NNNS1-----N----', 'AGNS2-----A----', 'C?--4----------', 'RF-------------', 'AAIP1----2A----', 'PP-P1--2-------', 'PP-P1--1-------', 'PP-S1--1-------', 'PDNS1----------', 'PPZS4--3------2', 'PPXP4--3-------', 'P8ZS3----------', 'PPNS4--3-------', 'PPZS7--3-------', 'PPXP2--3-------', 'PHZS4--3-------', 'PPFS4--3-------', 'PP-P3--2-------', 'PZM-1----------', 'PPXP3--3-------', 'PH-S3--1-------', 'PP-P3--1-------', 'PZ--1----------', 'PPFS3--3-------', 'PHZS3--3-------', 'PP-S4--1-------', 'PP-P4--2-------', 'PP-P4--1-------', 'PDNS4----------', 'AGNS1-----A----', 'VB-S---2P-AA---', 'PLNS4----------', 'AANS1----3A----', 'ACNS------A----', 'ClFS7----------', 'ClZS7----------', 'Dg-------3A----', 'TT-------------', 'Cv-------------', 'Db------------1', 'Db-------------', 'Dg-------1A---3', 'Co-------------', 'Dg-------1A---1', 'Dg-------3A---1', 'PH-S4--1-------', 'Dg-------2A---1', 'Dg-------2A----', 'ACYS------A----', 'PZ--4----------', 'Dg-------1N----', 'Dg-------1A----', 'VB-S---1P-AA---', 'Vc-S---1-------', 'Vc-X---3-------', 'VB-S---1P-NA---', 'VB-P---2P-AA---', 'VB-S---1P-AA--1', 'VB-P---1P-NA---', 'VB-P---1P-AA---', 'AAFS4----3A----', 'AUFS4M---------', 'CrFS4----------', 'PDFS4----------', 'AAFS4----2A----', 'ClFS4----------', 'P8FS4----------', 'AAFS4----1N----', 'PZFS4----------', 'PWFS4----------', 'PSFS4-P1-------', 'P8FS4---------1', 'AAFS4----1A----', 'RV--4----------', 'RR--4----------', 'AAFS1----3A----', 'ClFS1----------', 'AGFS1-----A----', 'PSFSXFS3-------', 'CrFS1----------', 'PWFS1----------', 'PZFS1----------', 'PSHS1-P1-------', 'PSHS1-S1-------', 'PDFS1----------', 'AANS1----2A----', 'AANS1----1N----', 'Vf--------N----', 'AAFS1----1N----', 'AAFS1----2A----', 'AAFS1----1A----', 'PDYS1----------', 'PLYS1----------', 'PZYS1----------', 'PSZS1FS3-------', 'PWYS1----------', 'AUIS1M---------', 'CrIS1----------', 'AAIS1----3A----', 'AAIS1----1A----', 'AAFS2----1A----', 'AANP3----1A----', 'P8ZS7----------', 'AAIP3----1A----', 'P6-X3----------', 'ACTP------A----', 'PLFS1----------', 'P5ZS3--3-------', 'P5FS3--3-------', 'ClXP3----------', 'PLXP3----------', 'PDXP3----------', 'P8XP3----------', 'AAMP3----1A----', 'NNFS1-----N----', 'NNIP2-----A---1', 'NNNP3-----A----', 'AAFP3----1A----', 'AAFS7----1N----', 'AUFS7M---------', 'PDFS7----------', 'AAFS7----3A----', 'AGFS7-----A----', 'P8FS7---------1', 'AAFS7----1A----', 'AAIP1----1N----', 'AAIP1----3A----', 'CrIP1----------', 'PSHP1-P2-------', 'PDIP1----------', 'PLIP1----------', 'ClYP1----------', 'AAIP1----1A----', 'PLMP1----------', 'PDMP1----------', 'AGMP1-----A----', 'AAMP1----3A----', 'PSMP1-P1-------', 'AAMP1----2A----', 'PPMP1--3-------', 'PZMP1----------', 'AAMP1----1A----', 'NNIP1-----A----', 'AAFP1----2A----', 'AAFP1----1N----', 'ClXP1----------', 'CrFP1----------', 'AGFP1-----A----', 'AAMP1----1N----', 'NNNP1-----A---2', 'AAIP7----3A----', 'Vf--------A---1', 'PLFP1----------', 'AAFP1----3A----', 'ClHP1----------', 'PSHP1-P1-------', 'PZFP1----------', 'PDFP1----------', 'AAFP1----1A----', 'VsNS---XX-AP---', 'VsQW---XX-AP---', 'AGIS1-----A----', 'ClYS1----------', 'AAIS1----2A----', 'AAIS1----1N----', 'VsYS---XX-AP---', 'AUNS1M---------', 'CrNS1----------', 'AANS1----1A----', 'ACMP------A----', 'Vc-P---1-------', 'VsMP---XX-AP---', 'VsTP---XX-AP---', 'Vf--------A----', 'NNNS2-----A----', 'NNIS2-----A----', 'P5ZS2--3------1', 'P5FS2--3-------', 'P9XP2----------', 'PDZS2----------', 'P8ZS2----------', 'PSZS2-P1-------', 'PSZS2FS3-------', 'PSZS2-S1-------', 'PSFS2-P1-------', 'AGFS2-----A----', 'PDFS2----------', 'AAFS2----1N----', 'ClFS2----------', 'P8FS2---------1', 'PSXXXZS3-------', 'AUNS2M---------', 'CrNS2----------', 'AANS2----1N----', 'AANS2----1A----', 'ClHP4----------', 'PKM-4----------', 'PDFP4----------', 'PWFP4----------', 'PZFP4----------', 'PLFP4----------', 'P8FP4---------1', 'AAFP4----1A----', 'PZYP4----------', 'AAIP4----1N----', 'PDIP4----------', 'AAIP4----1A----', 'ClXP4----------', 'ClYP4----------', 'PLYP4----------', 'P8YP4---------1', 'AAMP4----1A----', 'NNMP4-----A----', 'NNIP4-----A----', 'NNFP4-----A----', 'AUIS2M---------', 'AAIS2----1N----', 'AAIS2----1A----', 'NNFS2-----A----', 'AAFS2----2A----', 'CrFS2----------', 'AUFS2M---------', 'RV--2----------', 'RR--2----------'] Class 2: ['NNIP7-----A----', 'AGNP6-----A----', 'NNFS2-----N----', 'NNFS2-----A---1', 'NNFS4-----N----', 'CrIS2----------', 'ClZS2----------', 'AAIS2----3A----', 'AAIS2----2A----', 'VB-P---2P-NA---', 'PDMS4----------', 'ClZS3----------', 'AGIS2-----A----', 'AGFS4-----A----', 'AAMS3----1A----', 'NNMS3-----A---1', 'NNMS3-----A----', 'P6-X7----------', 'PP-P7--1-------', 'NNNP7-----A----', 'P5ZS7--3-------', 'P5XP7--3-------', 'NNMP7-----A----', 'NNFP7-----A----', 'NNMS7-----A----', 'NNNS7-----A----', 'P9ZS7----------', 'P4FS7----------', 'AGFP7-----A----', 'AAMP7----1A----', 'Cn-P7----------', 'P8XP7----------', 'ClXP7----------', 'AAFP7----2A----', 'AAFP7----1A----', 'AAIP7----1A----', 'AAMS7----1A----', 'AAMS7----3A----', 'AANS7----1N----', 'AANS7----1A----', 'AAIS7----1N----', 'PDXP7----------', 'AAIS7----3A----', 'AGIS7-----A----', 'VeYS------A----', 'CrIS7----------', 'PDZS7----------', 'AAIS7----1A----', 'AAMS1----1N----', 'AUMS1M---------', 'AGMS1-----A----', 'CrMS1----------', 'AAMS1----2A----', 'PSYS1-P1-------', 'PSYS1-S1-------', 'AAMS1----1A----', 'NNMS1-----A----', 'NNIS7-----A----', 'NNFS7-----A----', 'AANS3----1A----', 'NNNS3-----A----', 'AAIS3----1A----', 'PDZS3----------', 'NNIS3-----A----', 'NNIP3-----A----', 'NNFP3-----A----', 'NNMP3-----A----', 'NNFS3-----A----'] Class 3: ['P5FS6--3-------', 'P5XP6--3-------', 'P6-X6----------', 'NNNSX-----A----', 'P9FS6----------', 'P5ZS6--3-------', 'PQ--6----------', 'P9XP6----------', 'P9ZS6----------', 'PDZS6----------', 'AAIS6----1A----', 'PSZS6FS3-------', 'AAIS6----1N----', 'ClZS6----------', 'CrIS6----------', 'PSZS6-P1-------', 'AUNS6M---------', 'CrNS6----------', 'AANS6----1A----', 'PSFS6-P1-------', 'AAFS6----1N----', 'CrFS6----------', 'AUFS6M---------', 'AAFS6----1A----', 'NNFS6-----A----', 'NNIS6-----A---1', 'NNIS6-----A----', 'NNNS6-----A----', 'NNMP6-----A----', 'NNFP1-----N----', 'NNMS6-----A---1', 'AAMS6----1A----', 'NNMS6-----A----', 'NNIP6-----A----', 'NNNP6-----A----', 'NNFP6-----A----'] Class 4: ['AGMP2-----A----', 'AAMP2----1N----', 'AAMP2----1A----', 'AAIP2----1N----', 'AAIP2----2A----', 'AGIP2-----A----', 'AAIP2----1A----', 'AAFP2----2A----', 'AAFP2----3A----', 'PSXP2-P1-------', 'ClXP2----------', 'CrNP2----------', 'AGNP2-----A----', 'Cn-P2----------', 'AGFP2-----A----', 'AAFP2----1N----', 'PDXP2----------', 'AANP2----1A----', 'PZXP2----------', 'P8XP2----------', 'Ca--2----------', 'AAMP2----3A----', 'PLXP2----------', 'ClXS2----------', 'Ca--4----------', 'Ca--1----------', 'Cn-S1----------', 'Cn-S4----------', 'AAFP2----1A----', 'NNMP2-----A----', 'NNNP2-----A----', 'NNIP2-----A----', 'NNFP2-----A----'] Class 5: ['P4XP3----------', 'PJXP1----------', 'P4NS1----------', 'PQ--1----------', 'J,-X---3-------', 'P4IP1----------', 'P4FP1----------', 'P4MP1----------', 'J,-------------', 'PKM-3----------', 'PQ--2----------', 'P4ZS7----------', 'PKM-1----------', 'PQ--4----------', 'Vi-P---2--N----', 'J,-P---1-------', 'PJYS2----------', 'P4MS4----------', 'PJZS3----------', 'P4YP4----------', 'P4FS4----------', 'P4FP4----------', 'P4IS4----------', 'J,-S---1-------', 'P4NS4----------', 'PJYS1----------', 'PJNS1----------', 'P4YS1----------', 'PE--1----------', 'P4FS1----------'] Class 6: ['NNFS5-----A----', 'II-------------', 'VB-S---2P-NA---', 'J^------------8', 'Vi-S---2--A----', 'Vi-P---2--A----', 'Vi-P---1--A----', 'NNIPX-----A----', 'NNFSX-----A----', 'AAXXX----1A----', 'Xx-------------', 'NNMXX-----A---8', 'NNIXX-----A---8', 'A2--------A----', 'AAXXX----1A---8', 'C}-------------', 'NNFXX-----A---8', 'Db------------8', 'AAIS1----1A---6', 'NNISX-----A----', 'CyFP1----------', 'NNMS5-----A----', 'NNFPX-----A----', 'NNFS1-----A---1', 'AUFS1M---------', 'RR--1----------', 'RR--X----------', 'XX-------------', 'X@-------------', 'NNIXX-----A----', 'NNNXX-----A----', 'NNFXX-----A----', 'C=-------------'] Class 7: ['Z:-------------'] Class 8: ['NNIS1-----A----', 'NNFS1-----A----', 'NNMP1-----A---1', 'NNMP1-----A----', 'NNNS1-----A----', 'NNFP1-----A----', 'NNNP1-----A----'] Class 9: ['RR--7----------', 'RV--7----------', 'RR--3----------', 'RV--3----------'] Class 10: ['RR--6----------', 'RV--6----------']
for mode in ["words", "pos"]:
if mode == "words":
LIMIT = 1
MIN_FREQ = 10
K = 8000
elif mode == "pos":
LIMIT = 1
MIN_FREQ = 5
K = 200000
for lang in ["en", "cz"]:
print(f"""Language: {lang}.
Merging {mode} into {LIMIT} classes.
Considering first {K} words with min. freq. {MIN_FREQ}...\n""")
classes, history_of_merges = execute_merging(lang, mode, LIMIT, MIN_FREQ, K, verbose=True)
for i, cls in enumerate(classes):
print(f"Class {i+1}: {cls}")
print("\n")
## save history to a file
with open(f"../results/{lang}_{mode}_{LIMIT}_history.txt", "w", encoding="utf-8") as f:
for cls_one, cls_two in history_of_merges:
f.write(f"{cls_two} into {cls_one}\n")
Interpretation¶
Merging words¶
The merging algorithm is having (optically) better results in english compared to czech. By 'better' we mean that words with similar properties are being grouped together more likely then in czech. In english, for example, even the 50th merge was:
50: ['more'] into ['even', 'less', 'much'], grouping words that clearly describe quantity.
Whereas in czech, already the 35th merge was:
35: ['státu', 'J', 'ČSFR'] into ['mezi', 'už', 'bylo', 'nás'], where it is harder to see their common property.
This applies to all later merges. Reason for that might be the following counts of the source texts:
Text File | Words | Unique Words |
---|---|---|
CZ.txt | 222,413 | 42,827 |
EN.txt | 221,099 | 9,608 |
While having roughly the same length, there are ~4 times more unique words in czech relatively to english. This diversification and (on average) globally lower bigram counts makes it more difficult for the merging algorithm to (find and) merge 'similar' words into one class. We could experiment and try to improve the merging by introducing lemmas and merging lemmas of the words instead of the exact words, or conversaly keep merging words, but introduce more than just first 8000 words in czech. We have therefore tried to run the algorithm on 4x longer text to hopefully get 'optically' better results.
mode = "words"
lang = "cz"
LIMIT = 15
MIN_FREQ = 10
K = 32000
print(f"""Language: {lang}.
Merging {mode} into {LIMIT} classes.
Considering first {K} words with min. freq. {MIN_FREQ}...\n""")
classes, history_of_merges = execute_merging(lang, mode, LIMIT, MIN_FREQ, K, verbose=True)
for i, cls in enumerate(classes):
print(f"Class {i+1}: {cls}")
print("\n")
## save history to a file
with open(f"../results/{lang}_{mode}_{LIMIT}_history.txt", "w", encoding="utf-8") as f:
for cls_one, cls_two in history_of_merges:
f.write(f"{cls_two} into {cls_one}\n")
Language: cz. Merging words into 15 classes. Considering first 32000 words with min. freq. 10... Merging language: cz, mode: words Sum of the Q: 7.293523347895157 ['C'] into ['I'], loss: 0.0005384192482320643 ['Ale'] into ['Je'], loss: 0.000721102553055495 ['To'] into ['Jak'], loss: 0.0007409856393401377 ['J'] into ['P'], loss: 0.0007795989272068385 ['OKD'] into ['kolo'], loss: 0.0008028958453116571 ['protože'] into ['kterou'], loss: 0.000905530147429931 ['pokud'] into ['místo'], loss: 0.0009104544614628209 ['ES'] into ['KSČ'], loss: 0.0009117620858367708 ['otázku'] into ['druhé'], loss: 0.000913909765557391 ['Moskva'] into ['Sparta'], loss: 0.0009226531966366703 ['parlamentu'] into ['podniků'], loss: 0.0009292292966721748 ['toho'] into ['těch'], loss: 0.0009341058050920184 ['18'] into ['1991'], loss: 0.0009366370680159043 ['třeba'] into ['později'], loss: 0.0009414902759546625 ['Z'] into ['Když'], loss: 0.0009535448166031399 ['měla'] into ['mohl'], loss: 0.0009698336655218874 ['dolarů'] into ['korun'], loss: 0.0009729264261012192 ['přece'] into ['prý'], loss: 0.000982532789164596 ['informace'] into ['smlouvy'], loss: 0.001026711030383539 ['8'] into ['9'], loss: 0.00102930829567095 ['sebe'] into ['včerejší'], loss: 0.0010344813452425085 ['často'] into ['dokonce'], loss: 0.0010357889696164559 ['25'] into ['programu'], loss: 0.0010389097655573894 ['organizace'] into ['války'], loss: 0.0010389097655573928 ['lidí'] into ['prostředí'], loss: 0.0010433475074317616 ['případě'] into ['roce'], loss: 0.0010528417118531343 ['všechno'] into ['teď'], loss: 0.0010756718513941638 ['jeden'] into ['poslední'], loss: 0.0010778195311147848 ['nich'] into ['české'], loss: 0.0010828218085466024 ['IFS'] into ['nám'], loss: 0.0010900805103972733 ['den'] into ['SNR'], loss: 0.001093139062229566 ['tuto'] into ['této'], loss: 0.0010934109707946803 ['tato'] into ['její'], loss: 0.0011014097655573894 ['armády'] into ['stavu'], loss: 0.0011014097655573912 ['Za'] into ["'"], loss: 0.0011085208374529889 ['svých'] into ['dvou'], loss: 0.001116729296672175 ['České'] into ['celé'], loss: 0.0011321110718955987 ['zájem'] into ['%'], loss: 0.0011403195311147831 ['12'] into ['10'], loss: 0.0011432075490481527 ['proto'] into ['lidé'], loss: 0.0011458270958998734 ['dnech'] into ['poslanců'], loss: 0.001146442554721988 ['miliónů'] into ['tisíc'], loss: 0.0011575205437204562 ['K'] into ['Kčs'], loss: 0.0011621282112493573 ['tomto'] into ['úterý'], loss: 0.0011634676065215357 ['majetku'] into ['ČNR'], loss: 0.0011761085007312398 ['O'] into ['Po'], loss: 0.001177329636761825 ['která'] into ['kde'], loss: 0.001179353129567011 ['mít'] into ['některé'], loss: 0.00118535232027938 ['chce'] into ['ceny'], loss: 0.00118535232027938 ['může'] into ['sám'], loss: 0.0011874999999999993 ['zatím'] into ['ten'], loss: 0.0011874999999999993 ['skupiny'] into ['státní'], loss: 0.001187500000000001 ['soudu'] into ['zahraničí'], loss: 0.0011884731162203135 ['Praha'] into ['SSSR'], loss: 0.001243231658450781 ['vůbec'] into ['mne'], loss: 0.0012478523202793827 ['jenom'] into ['li'], loss: 0.0012767534522681188 ['M'] into ['11'], loss: 0.0012816011987000966 ['asi'] into ['zcela'], loss: 0.0012869489640353407 ['Branky'] into ['kolo', 'OKD'], loss: 0.0013028219907537447 ['především'] into ['zápas'], loss: 0.0013042915407808142 ['svou'] into ['zase'], loss: 0.0013124999999999977 ['ministr'] into ['vláda'], loss: 0.0013341294329205552 ['Praze'] into ['době'], loss: 0.0013445499657345874 ['kdy'] into ['kterou', 'protože'], loss: 0.001349505245968892 ['Jan'] into ['prezident'], loss: 0.0013730391984779471 ['Do'] into ['Podle'], loss: 0.0013863005917115045 ['naší'] into ['sociální'], loss: 0.0013982538715143888 ['5'] into ['7'], loss: 0.0014233355439576389 ['přes'] into ['bez'], loss: 0.0014528195311147851 ['listopadu'] into ['I', 'C'], loss: 0.0014884596631834987 ['4'] into ['6'], loss: 0.001534306804530615 ['mu'] into ['dosud'], loss: 0.0015560539092281831 ['tomu'] into ['tím'], loss: 0.0015706982779089057 ['které'] into ['kteří'], loss: 0.0015725560158424377 ['státu'] into ['privatizace'], loss: 0.0016071509689703986 ['všech'] into ['tohoto'], loss: 0.0016180571096380053 ['dva'] into ['tento'], loss: 0.0016578940755229577 ['ČR'] into ['&slash;'], loss: 0.001672987704790311 ['měl'] into ['mohl', 'měla'], loss: 0.001715509125982959 ['A'] into ['Je', 'Ale'], loss: 0.001716512552222207 ['FS'] into ['nic'], loss: 0.001720770563345919 ['již'] into ['NATO'], loss: 0.0017245186237007833 ['jak'] into ['když'], loss: 0.0017280600113803654 ['návrh'] into ['nové'], loss: 0.0017874551835489784 ['tu'] into ['dnes'], loss: 0.0017905222142359347 ['20'] into ['13'], loss: 0.0018034707099778904 ['nad'] into ['tam'], loss: 0.001819391646208218 ['tedy'] into ['ovšem'], loss: 0.0018506945430876503 ['3'] into ['2'], loss: 0.0018552540832437459 ['h'] into ['%', 'zájem'], loss: 0.0018791367416811974 ['vlastní'] into ['tom'], loss: 0.0019014350043542252 ['Jak', 'To'] into ['Když', 'Z'], loss: 0.001910405036593739 ['ke'] into ['proti'], loss: 0.0019139097655573936 ['strany'] into ['podniků', 'parlamentu'], loss: 0.0019218879806178236 ['druhé', 'otázku'] into ['včerejší', 'sebe'], loss: 0.0019632050947216498 ['ani'] into ['teď', 'všechno'], loss: 0.001972240404854938 ['ho'] into ['pouze'], loss: 0.0020153195311147882 ['1'] into ['0'], loss: 0.0020223970120349447 ['KSČ', 'ES'] into ['ČNR', 'majetku'], loss: 0.002024736323118854 ['mají'] into ['jim'], loss: 0.002037691124737481 ['místo', 'pokud'] into ['později', 'třeba'], loss: 0.0020673210325877333 ['1991', '18'] into ['10', '12'], loss: 0.0020825370164485463 ['prý', 'přece'] into ['dokonce', 'často'], loss: 0.002083768634729254 ['P', 'J'] into ['11', 'M'], loss: 0.0020955576043704665 ['*'] into ["'", 'Za'], loss: 0.0021106639025538837 ['korun', 'dolarů'] into ['prostředí', 'lidí'], loss: 0.002127574621266941 ['vlády'] into ['zahraničí', 'soudu'], loss: 0.002128546272704592 ['jednání'] into ['roku'], loss: 0.0021529581516380753 ['smlouvy', 'informace'] into ['poslanců', 'dnech'], loss: 0.0021628100413849036 ['další'] into ['ceny', 'chce'], loss: 0.002179229296672174 ['federální'] into ['její', 'tato'], loss: 0.00218634386765327 ['PRAHA'] into ['státní', 'skupiny'], loss: 0.0022005546477683637 ['který'] into ['kde', 'která'], loss: 0.002212449790603168 ['let'] into ['Sparta', 'Moskva'], loss: 0.0022256015978940396 ['roce', 'případě'] into ['době', 'Praze'], loss: 0.002246117768133918 ['nebo'] into ['byla'], loss: 0.0022693720511780917 ['st'] into ['nám', 'IFS'], loss: 0.002270056260488552 ['války', 'organizace'] into ['SNR', 'den'], loss: 0.002272492847119026 ['byly'] into ['ten', 'zatím'], loss: 0.002288473592426658 ['české', 'nich'] into ['celé', 'České'], loss: 0.0022982933751332334 ['co'] into ['než'], loss: 0.0023335544104494104 ['programu', '25'] into ['mne', 'vůbec'], loss: 0.0023349905102701216 ['není'] into ['li', 'jenom'], loss: 0.002341145571922235 ['této', 'tuto'] into ['úterý', 'tomto'], loss: 0.002380316263194765 ['První'] into ['Po', 'O'], loss: 0.0023890501922489843 ['stavu', 'armády'] into ['zápas', 'především'], loss: 0.0024332046405587597 ['těch', 'toho'] into ['lidé', 'proto'], loss: 0.002440635244398505 ['poslední', 'jeden'] into ['některé', 'mít'], loss: 0.002475361200795871 ['mi'] into ['zase', 'svou'], loss: 0.0024836785148541866 ['má'] into ['musí'], loss: 0.0024836854819853416 ['také'] into ['řekl'], loss: 0.0025000000000000014 ['Kčs', 'K'] into ['I', 'C', 'listopadu'], loss: 0.0025591181329135786 ['9', '8'] into ['7', '5'], loss: 0.0025791297045510177 ['své'] into ['dvou', 'svých'], loss: 0.002582168631398502 ['jsme'] into ['sám', 'může'], loss: 0.002686498045968969 ['aby'] into ['kterou', 'protože', 'kdy'], loss: 0.0026945488277869524 ['nás'] into ['zákona'], loss: 0.0027848688196145405 ['vláda', 'ministr'] into ['prezident', 'Jan'], loss: 0.002812499999999999 ['pak'] into ['tak'], loss: 0.0028285670774004373 ['Na'] into ['Podle', 'Do'], loss: 0.002841520398666003 ['#'] into ['či'], loss: 0.002874999999999999 ['podle'] into ['zcela', 'asi'], loss: 0.0029923084529004766 ['první'] into ['SSSR', 'Praha'], loss: 0.003081496340133173 ['při'] into ['tisíc', 'miliónů'], loss: 0.003110557621313204 ['ČSFR'] into ['&slash;', 'ČR'], loss: 0.003198068689439117 ['žen'] into ['privatizace', 'státu'], loss: 0.0031999048303329204 ['bez', 'přes'] into ['tohoto', 'všech'], loss: 0.003250000000000001 ['sociální', 'naší'] into ['tento', 'dva'], loss: 0.0032606980528989728 ['jejich'] into ['tom', 'vlastní'], loss: 0.0032759087023037647 ['dosud', 'mu'] into ['jim', 'mají'], loss: 0.003325686046597299 ['budou'] into ['nic', 'FS'], loss: 0.0034368063692512938 ['mezi'] into ['proti', 'ke'], loss: 0.0034643783854885375 ['='] into ["'", 'Za', '*'], loss: 0.003572048714886117 ['NATO', 'již'] into ['tam', 'nad'], loss: 0.003621115783858103 ['ovšem', 'tedy'] into ['li', 'jenom', 'není'], loss: 0.0036925891408396676 ['až'] into ['bylo'], loss: 0.0036975874695105142 ['když', 'jak'] into ['později', 'třeba', 'místo', 'pokud'], loss: 0.0037618637569854004 ['tím', 'tomu'] into ['řekl', 'také'], loss: 0.0037955665827846645 ['teď', 'všechno', 'ani'] into ['dokonce', 'často', 'prý', 'přece'], loss: 0.0038807089931848122 ['kolo', 'OKD', 'Branky'] into ['Když', 'Z', 'Jak', 'To'], loss: 0.0039197737909314415 ['ještě'] into ['být'], loss: 0.003939529111445961 ['utkání'] into ['nové', 'návrh'], loss: 0.003989063791411937 ['10', '12', '1991', '18'] into ['7', '5', '9', '8'], loss: 0.004007448852547103 ['však'] into ['dnes', 'tu'], loss: 0.004029196140738805 ['kteří', 'které'] into ['než', 'co'], loss: 0.004064294100056022 ['pouze', 'ho'] into ['zápas', 'především', 'stavu', 'armády'], loss: 0.004125448003320774 ['podniků', 'parlamentu', 'strany'] into ['ČNR', 'majetku', 'KSČ', 'ES'], loss: 0.004211594545384369 ['mohl', 'měla', 'měl'] into ['musí', 'má'], loss: 0.004272317958557526 ['ceny', 'chce', 'další'] into ['SNR', 'den', 'války', 'organizace'], loss: 0.0043244147772177974 ['Je', 'Ale', 'A'] into ['Po', 'O', 'První'], loss: 0.004380505874064937 ['!'] into ['%', 'zájem', 'h'], loss: 0.004390974581898702 ['13', '20'] into ['Sparta', 'Moskva', 'let'], loss: 0.004305709786259467 ['roku', 'jednání'] into ['poslanců', 'dnech', 'smlouvy', 'informace'], loss: 0.004401178431686491 ['už'] into ['ten', 'zatím', 'byly'], loss: 0.004485810154793851 ['ze'] into [';'], loss: 0.004504381542021978 ['prostředí', 'lidí', 'korun', 'dolarů'] into ['nám', 'IFS', 'st'], loss: 0.004507146340000337 ['byl'] into ['byla', 'nebo'], loss: 0.004530639062229575 ['včerejší', 'sebe', 'druhé', 'otázku'] into ['zákona', 'nás'], loss: 0.0045596289985178964 ['její', 'tato', 'federální'] into ['některé', 'mít', 'poslední', 'jeden'], loss: 0.0045728614010567825 ['mne', 'vůbec', 'programu', '25'] into ['lidé', 'proto', 'těch', 'toho'], loss: 0.004748937435083653 ['včera'] into ['bude'], loss: 0.00477066148213461 ['USA'] into ['SSSR', 'Praha', 'první'], loss: 0.004881459688664964 ['u'] into ['jeho'], loss: 0.004905639062229577 ['před'] into ['zase', 'svou', 'mi'], loss: 0.005147348011163736 ['státní', 'skupiny', 'PRAHA'] into ['dvou', 'svých', 'své'], loss: 0.005156229027441242 ['době', 'Praze', 'roce', 'případě'] into ['celé', 'České', 'české', 'nich'], loss: 0.005248800582776482 ['jsem'] into ['sám', 'může', 'jsme'], loss: 0.00531076512416718 ['jsou'] into ['zahraničí', 'soudu', 'vlády'], loss: 0.00538976639768567 ['tak', 'pak'] into ['zcela', 'asi', 'podle'], loss: 0.005519516752857589 ['ale'] into ['kterou', 'protože', 'kdy', 'aby'], loss: 0.005554489771549268 ['?'] into ['prezident', 'Jan', 'vláda', 'ministr'], loss: 0.005600914324596355 ['2', '3'] into ['7', '5', '9', '8', '10', '12', '1991', '18'], loss: 0.005832669813534676 ['jen'] into ['či', '#'], loss: 0.0059238677664323084 ['V'] into ['I', 'C', 'listopadu', 'Kčs', 'K'], loss: 0.005986487691836354 ['že'] into ['kde', 'která', 'který'], loss: 0.0061651741129960524 ['Podle', 'Do', 'Na'] into ['Když', 'Z', 'Jak', 'To', 'kolo', 'OKD', 'Branky'], loss: 0.006234922659799358 ['od'] into ['tisíc', 'miliónů', 'při'], loss: 0.0062872295783028415 ['úterý', 'tomto', 'této', 'tuto'] into ['celé', 'České', 'české', 'nich', 'době', 'Praze', 'roce', 'případě'], loss: 0.006349276116390923 ['si'] into ['jim', 'mají', 'dosud', 'mu'], loss: 0.006408269572065707 ['nic', 'FS', 'budou'] into ['dnes', 'tu', 'však'], loss: 0.006358295193085626 ['6', '4'] into ['Sparta', 'Moskva', 'let', '13', '20'], loss: 0.00642369195787644 ['tohoto', 'všech', 'bez', 'přes'] into ['tento', 'dva', 'sociální', 'naší'], loss: 0.0064544095628858685 ['privatizace', 'státu', 'žen'] into ['tam', 'nad', 'NATO', 'již'], loss: 0.0067043694757753 ['&slash;', 'ČR', 'ČSFR'] into ['ČNR', 'majetku', 'KSČ', 'ES', 'podniků', 'parlamentu', 'strany'], loss: 0.007022159823427105 ['jako'] into ['proti', 'ke', 'mezi'], loss: 0.007099931528193331 ['li', 'jenom', 'není', 'ovšem', 'tedy'] into ['bylo', 'až'], loss: 0.007194926685810704 ['tom', 'vlastní', 'jejich'] into ['řekl', 'také', 'tím', 'tomu'], loss: 0.007353394626230476 ['0', '1'] into ['7', '5', '9', '8', '10', '12', '1991', '18', '2', '3'], loss: 0.007692804369738099 ['nové', 'návrh', 'utkání'] into [';', 'ze'], loss: 0.007426159012418433 ['poslanců', 'dnech', 'smlouvy', 'informace', 'roku', 'jednání'] into ['zákona', 'nás', 'včerejší', 'sebe', 'druhé', 'otázku'], loss: 0.007787237212970951 ['než', 'co', 'kteří', 'které'] into ['kterou', 'protože', 'kdy', 'aby', 'ale'], loss: 0.008036086658854122 ['by'] into ['později', 'třeba', 'místo', 'pokud', 'když', 'jak'], loss: 0.008238247810362686 ['za'] into ['být', 'ještě'], loss: 0.008434252877951954 ["'", 'Za', '*', '='] into ['%', 'zájem', 'h', '!'], loss: 0.00861886171779707 ['SNR', 'den', 'války', 'organizace', 'ceny', 'chce', 'další'] into ['zahraničí', 'soudu', 'vlády', 'jsou'], loss: 0.008760138873449282 ['pro'] into ['zápas', 'především', 'stavu', 'armády', 'pouze', 'ho'], loss: 0.008975021669750794 ['byla', 'nebo', 'byl'] into ['některé', 'mít', 'poslední', 'jeden', 'její', 'tato', 'federální'], loss: 0.00896437368103599 ['ten', 'zatím', 'byly', 'už'] into ['bude', 'včera'], loss: 0.008954397905722129 ['dokonce', 'často', 'prý', 'přece', 'teď', 'všechno', 'ani'] into ['zase', 'svou', 'mi', 'před'], loss: 0.008970100646232286 ['ve'] into ['po'], loss: 0.009005051977045125 ['Po', 'O', 'První', 'Je', 'Ale', 'A'] into ['Když', 'Z', 'Jak', 'To', 'kolo', 'OKD', 'Branky', 'Podle', 'Do', 'Na'], loss: 0.009298121255074787 ['11', 'M', 'P', 'J'] into ['Sparta', 'Moskva', 'let', '13', '20', '6', '4'], loss: 0.009681760451150025 ['musí', 'má', 'mohl', 'měla', 'měl'] into ['sám', 'může', 'jsme', 'jsem'], loss: 0.009818447253199405 ['lidé', 'proto', 'těch', 'toho', 'mne', 'vůbec', 'programu', '25'] into ['jeho', 'u'], loss: 0.009992553249331947 ['dvou', 'svých', 'své', 'státní', 'skupiny', 'PRAHA'] into ['prezident', 'Jan', 'vláda', 'ministr', '?'], loss: 0.010497776232287294 ['nám', 'IFS', 'st', 'prostředí', 'lidí', 'korun', 'dolarů'] into ['tam', 'nad', 'NATO', 'již', 'privatizace', 'státu', 'žen'], loss: 0.01066070296083086 ['zcela', 'asi', 'podle', 'tak', 'pak'] into ['dnes', 'tu', 'však', 'nic', 'FS', 'budou'], loss: 0.01083856083390548 ['SSSR', 'Praha', 'první', 'USA'] into ['ČNR', 'majetku', 'KSČ', 'ES', 'podniků', 'parlamentu', 'strany', '&slash;', 'ČR', 'ČSFR'], loss: 0.011179118962051914 ['k'] into ['tisíc', 'miliónů', 'při', 'od'], loss: 0.011270399834148511 ['to'] into ['řekl', 'také', 'tím', 'tomu', 'tom', 'vlastní', 'jejich'], loss: 0.011767899265340743 ['či', '#', 'jen'] into ['bylo', 'až', 'li', 'jenom', 'není', 'ovšem', 'tedy'], loss: 0.012159785666876009 ['i'] into ['jim', 'mají', 'dosud', 'mu', 'si'], loss: 0.012430601628855549 ['z'] into ['tento', 'dva', 'sociální', 'naší', 'tohoto', 'všech', 'bez', 'přes'], loss: 0.013879357773197236 ['I', 'C', 'listopadu', 'Kčs', 'K', 'V'] into ['Sparta', 'Moskva', 'let', '13', '20', '6', '4', '11', 'M', 'P', 'J'], loss: 0.01391457361069619 ['je'] into ['později', 'třeba', 'místo', 'pokud', 'když', 'jak', 'by'], loss: 0.014102668905790323 ['proti', 'ke', 'mezi', 'jako'] into ['být', 'ještě', 'za'], loss: 0.0142646325516133 ['bude', 'včera', 'ten', 'zatím', 'byly', 'už'] into ['zase', 'svou', 'mi', 'před', 'dokonce', 'často', 'prý', 'přece', 'teď', 'všechno', 'ani'], loss: 0.015050366376280376 ['celé', 'České', 'české', 'nich', 'době', 'Praze', 'roce', 'případě', 'úterý', 'tomto', 'této', 'tuto'] into ['zahraničí', 'soudu', 'vlády', 'jsou', 'SNR', 'den', 'války', 'organizace', 'ceny', 'chce', 'další'], loss: 0.015959838978526125 ['do'] into [';', 'ze', 'nové', 'návrh', 'utkání'], loss: 0.01615810372473548 ['některé', 'mít', 'poslední', 'jeden', 'její', 'tato', 'federální', 'byla', 'nebo', 'byl'] into ['jeho', 'u', 'lidé', 'proto', 'těch', 'toho', 'mne', 'vůbec', 'programu', '25'], loss: 0.016458803243523872 ['zápas', 'především', 'stavu', 'armády', 'pouze', 'ho', 'pro'] into ['sám', 'může', 'jsme', 'jsem', 'musí', 'má', 'mohl', 'měla', 'měl'], loss: 0.016829705071551262 ['s'] into ['po', 've'], loss: 0.017263915207179747 ['zákona', 'nás', 'včerejší', 'sebe', 'druhé', 'otázku', 'poslanců', 'dnech', 'smlouvy', 'informace', 'roku', 'jednání'] into ['%', 'zájem', 'h', '!', "'", 'Za', '*', '='], loss: 0.018178601896546787 [')'] into ['tam', 'nad', 'NATO', 'již', 'privatizace', 'státu', 'žen', 'nám', 'IFS', 'st', 'prostředí', 'lidí', 'korun', 'dolarů'], loss: 0.019597243864837143 ['dnes', 'tu', 'však', 'nic', 'FS', 'budou', 'zcela', 'asi', 'podle', 'tak', 'pak'] into ['řekl', 'také', 'tím', 'tomu', 'tom', 'vlastní', 'jejich', 'to'], loss: 0.02019180581431627 ['prezident', 'Jan', 'vláda', 'ministr', '?', 'dvou', 'svých', 'své', 'státní', 'skupiny', 'PRAHA'] into ['tisíc', 'miliónů', 'při', 'od', 'k'], loss: 0.02132934213964802 ['ČNR', 'majetku', 'KSČ', 'ES', 'podniků', 'parlamentu', 'strany', '&slash;', 'ČR', 'ČSFR', 'SSSR', 'Praha', 'první', 'USA'] into ['bylo', 'až', 'li', 'jenom', 'není', 'ovšem', 'tedy', 'či', '#', 'jen'], loss: 0.022629925937956243 ['kterou', 'protože', 'kdy', 'aby', 'ale', 'než', 'co', 'kteří', 'které'] into ['později', 'třeba', 'místo', 'pokud', 'když', 'jak', 'by', 'je'], loss: 0.022996237778382708 ['kde', 'která', 'který', 'že'] into ['později', 'třeba', 'místo', 'pokud', 'když', 'jak', 'by', 'je', 'kterou', 'protože', 'kdy', 'aby', 'ale', 'než', 'co', 'kteří', 'které'], loss: 0.02133642361451281 ['jim', 'mají', 'dosud', 'mu', 'si', 'i'] into ['být', 'ještě', 'za', 'proti', 'ke', 'mezi', 'jako'], loss: 0.024187697600952917 ['o'] into ['tento', 'dva', 'sociální', 'naší', 'tohoto', 'všech', 'bez', 'přes', 'z'], loss: 0.024565154887369715 ['7', '5', '9', '8', '10', '12', '1991', '18', '2', '3', '0', '1'] into ['Sparta', 'Moskva', 'let', '13', '20', '6', '4', '11', 'M', 'P', 'J', 'I', 'C', 'listopadu', 'Kčs', 'K', 'V'], loss: 0.02517990273816964 [':'] into ['('], loss: 0.026286911014724034 ['"'] into ['Když', 'Z', 'Jak', 'To', 'kolo', 'OKD', 'Branky', 'Podle', 'Do', 'Na', 'Po', 'O', 'První', 'Je', 'Ale', 'A'], loss: 0.026936736442399353 ['zahraničí', 'soudu', 'vlády', 'jsou', 'SNR', 'den', 'války', 'organizace', 'ceny', 'chce', 'další', 'celé', 'České', 'české', 'nich', 'době', 'Praze', 'roce', 'případě', 'úterý', 'tomto', 'této', 'tuto'] into ['%', 'zájem', 'h', '!', "'", 'Za', '*', '=', 'zákona', 'nás', 'včerejší', 'sebe', 'druhé', 'otázku', 'poslanců', 'dnech', 'smlouvy', 'informace', 'roku', 'jednání'], loss: 0.02742165251266529 ['zase', 'svou', 'mi', 'před', 'dokonce', 'často', 'prý', 'přece', 'teď', 'všechno', 'ani', 'bude', 'včera', 'ten', 'zatím', 'byly', 'už'] into ['sám', 'může', 'jsme', 'jsem', 'musí', 'má', 'mohl', 'měla', 'měl', 'zápas', 'především', 'stavu', 'armády', 'pouze', 'ho', 'pro'], loss: 0.027985776249456966 ['jeho', 'u', 'lidé', 'proto', 'těch', 'toho', 'mne', 'vůbec', 'programu', '25', 'některé', 'mít', 'poslední', 'jeden', 'její', 'tato', 'federální', 'byla', 'nebo', 'byl'] into ['řekl', 'také', 'tím', 'tomu', 'tom', 'vlastní', 'jejich', 'to', 'dnes', 'tu', 'však', 'nic', 'FS', 'budou', 'zcela', 'asi', 'podle', 'tak', 'pak'], loss: 0.02920638102938234 [';', 'ze', 'nové', 'návrh', 'utkání', 'do'] into ['po', 've', 's'], loss: 0.03126716129684685 ['na'] into ['tisíc', 'miliónů', 'při', 'od', 'k', 'prezident', 'Jan', 'vláda', 'ministr', '?', 'dvou', 'svých', 'své', 'státní', 'skupiny', 'PRAHA'], loss: 0.03698270723703551 ['tam', 'nad', 'NATO', 'již', 'privatizace', 'státu', 'žen', 'nám', 'IFS', 'st', 'prostředí', 'lidí', 'korun', 'dolarů', ')'] into ['bylo', 'až', 'li', 'jenom', 'není', 'ovšem', 'tedy', 'či', '#', 'jen', 'ČNR', 'majetku', 'KSČ', 'ES', 'podniků', 'parlamentu', 'strany', '&slash;', 'ČR', 'ČSFR', 'SSSR', 'Praha', 'první', 'USA'], loss: 0.03914143532004303 ['se'] into ['být', 'ještě', 'za', 'proti', 'ke', 'mezi', 'jako', 'jim', 'mají', 'dosud', 'mu', 'si', 'i'], loss: 0.042315502042312256 ['v'] into ['tento', 'dva', 'sociální', 'naší', 'tohoto', 'všech', 'bez', 'přes', 'z', 'o'], loss: 0.04965256852182998 ['-'] into ['(', ':'], loss: 0.050608285188534705 1: ['C'] into ['I'] 2: ['Ale'] into ['Je'] 3: ['To'] into ['Jak'] 4: ['J'] into ['P'] 5: ['OKD'] into ['kolo'] 6: ['protože'] into ['kterou'] 7: ['pokud'] into ['místo'] 8: ['ES'] into ['KSČ'] 9: ['otázku'] into ['druhé'] 10: ['Moskva'] into ['Sparta'] 11: ['parlamentu'] into ['podniků'] 12: ['toho'] into ['těch'] 13: ['18'] into ['1991'] 14: ['třeba'] into ['později'] 15: ['Z'] into ['Když'] 16: ['měla'] into ['mohl'] 17: ['dolarů'] into ['korun'] 18: ['přece'] into ['prý'] 19: ['informace'] into ['smlouvy'] 20: ['8'] into ['9'] 21: ['sebe'] into ['včerejší'] 22: ['často'] into ['dokonce'] 23: ['25'] into ['programu'] 24: ['organizace'] into ['války'] 25: ['lidí'] into ['prostředí'] 26: ['případě'] into ['roce'] 27: ['všechno'] into ['teď'] 28: ['jeden'] into ['poslední'] 29: ['nich'] into ['české'] 30: ['IFS'] into ['nám'] 31: ['den'] into ['SNR'] 32: ['tuto'] into ['této'] 33: ['tato'] into ['její'] 34: ['armády'] into ['stavu'] 35: ['Za'] into ["'"] 36: ['svých'] into ['dvou'] 37: ['České'] into ['celé'] 38: ['zájem'] into ['%'] 39: ['12'] into ['10'] 40: ['proto'] into ['lidé'] 41: ['dnech'] into ['poslanců'] 42: ['miliónů'] into ['tisíc'] 43: ['K'] into ['Kčs'] 44: ['tomto'] into ['úterý'] 45: ['majetku'] into ['ČNR'] 46: ['O'] into ['Po'] 47: ['která'] into ['kde'] 48: ['mít'] into ['některé'] 49: ['chce'] into ['ceny'] 50: ['může'] into ['sám'] 51: ['zatím'] into ['ten'] 52: ['skupiny'] into ['státní'] 53: ['soudu'] into ['zahraničí'] 54: ['Praha'] into ['SSSR'] 55: ['vůbec'] into ['mne'] 56: ['jenom'] into ['li'] 57: ['M'] into ['11'] 58: ['asi'] into ['zcela'] 59: ['Branky'] into ['kolo', 'OKD'] 60: ['především'] into ['zápas'] 61: ['svou'] into ['zase'] 62: ['ministr'] into ['vláda'] 63: ['Praze'] into ['době'] 64: ['kdy'] into ['kterou', 'protože'] 65: ['Jan'] into ['prezident'] 66: ['Do'] into ['Podle'] 67: ['naší'] into ['sociální'] 68: ['5'] into ['7'] 69: ['přes'] into ['bez'] 70: ['listopadu'] into ['I', 'C'] 71: ['4'] into ['6'] 72: ['mu'] into ['dosud'] 73: ['tomu'] into ['tím'] 74: ['které'] into ['kteří'] 75: ['státu'] into ['privatizace'] 76: ['všech'] into ['tohoto'] 77: ['dva'] into ['tento'] 78: ['ČR'] into ['&slash;'] 79: ['měl'] into ['mohl', 'měla'] 80: ['A'] into ['Je', 'Ale'] 81: ['FS'] into ['nic'] 82: ['již'] into ['NATO'] 83: ['jak'] into ['když'] 84: ['návrh'] into ['nové'] 85: ['tu'] into ['dnes'] 86: ['20'] into ['13'] 87: ['nad'] into ['tam'] 88: ['tedy'] into ['ovšem'] 89: ['3'] into ['2'] 90: ['h'] into ['%', 'zájem'] 91: ['vlastní'] into ['tom'] 92: ['Jak', 'To'] into ['Když', 'Z'] 93: ['ke'] into ['proti'] 94: ['strany'] into ['podniků', 'parlamentu'] 95: ['druhé', 'otázku'] into ['včerejší', 'sebe'] 96: ['ani'] into ['teď', 'všechno'] 97: ['ho'] into ['pouze'] 98: ['1'] into ['0'] 99: ['KSČ', 'ES'] into ['ČNR', 'majetku'] 100: ['mají'] into ['jim'] 101: ['místo', 'pokud'] into ['později', 'třeba'] 102: ['1991', '18'] into ['10', '12'] 103: ['prý', 'přece'] into ['dokonce', 'často'] 104: ['P', 'J'] into ['11', 'M'] 105: ['*'] into ["'", 'Za'] 106: ['korun', 'dolarů'] into ['prostředí', 'lidí'] 107: ['vlády'] into ['zahraničí', 'soudu'] 108: ['jednání'] into ['roku'] 109: ['smlouvy', 'informace'] into ['poslanců', 'dnech'] 110: ['další'] into ['ceny', 'chce'] 111: ['federální'] into ['její', 'tato'] 112: ['PRAHA'] into ['státní', 'skupiny'] 113: ['který'] into ['kde', 'která'] 114: ['let'] into ['Sparta', 'Moskva'] 115: ['roce', 'případě'] into ['době', 'Praze'] 116: ['nebo'] into ['byla'] 117: ['st'] into ['nám', 'IFS'] 118: ['války', 'organizace'] into ['SNR', 'den'] 119: ['byly'] into ['ten', 'zatím'] 120: ['české', 'nich'] into ['celé', 'České'] 121: ['co'] into ['než'] 122: ['programu', '25'] into ['mne', 'vůbec'] 123: ['není'] into ['li', 'jenom'] 124: ['této', 'tuto'] into ['úterý', 'tomto'] 125: ['První'] into ['Po', 'O'] 126: ['stavu', 'armády'] into ['zápas', 'především'] 127: ['těch', 'toho'] into ['lidé', 'proto'] 128: ['poslední', 'jeden'] into ['některé', 'mít'] 129: ['mi'] into ['zase', 'svou'] 130: ['má'] into ['musí'] 131: ['také'] into ['řekl'] 132: ['Kčs', 'K'] into ['I', 'C', 'listopadu'] 133: ['9', '8'] into ['7', '5'] 134: ['své'] into ['dvou', 'svých'] 135: ['jsme'] into ['sám', 'může'] 136: ['aby'] into ['kterou', 'protože', 'kdy'] 137: ['nás'] into ['zákona'] 138: ['vláda', 'ministr'] into ['prezident', 'Jan'] 139: ['pak'] into ['tak'] 140: ['Na'] into ['Podle', 'Do'] 141: ['#'] into ['či'] 142: ['podle'] into ['zcela', 'asi'] 143: ['první'] into ['SSSR', 'Praha'] 144: ['při'] into ['tisíc', 'miliónů'] 145: ['ČSFR'] into ['&slash;', 'ČR'] 146: ['žen'] into ['privatizace', 'státu'] 147: ['bez', 'přes'] into ['tohoto', 'všech'] 148: ['sociální', 'naší'] into ['tento', 'dva'] 149: ['jejich'] into ['tom', 'vlastní'] 150: ['dosud', 'mu'] into ['jim', 'mají'] 151: ['budou'] into ['nic', 'FS'] 152: ['mezi'] into ['proti', 'ke'] 153: ['='] into ["'", 'Za', '*'] 154: ['NATO', 'již'] into ['tam', 'nad'] 155: ['ovšem', 'tedy'] into ['li', 'jenom', 'není'] 156: ['až'] into ['bylo'] 157: ['když', 'jak'] into ['později', 'třeba', 'místo', 'pokud'] 158: ['tím', 'tomu'] into ['řekl', 'také'] 159: ['teď', 'všechno', 'ani'] into ['dokonce', 'často', 'prý', 'přece'] 160: ['kolo', 'OKD', 'Branky'] into ['Když', 'Z', 'Jak', 'To'] 161: ['ještě'] into ['být'] 162: ['utkání'] into ['nové', 'návrh'] 163: ['10', '12', '1991', '18'] into ['7', '5', '9', '8'] 164: ['však'] into ['dnes', 'tu'] 165: ['kteří', 'které'] into ['než', 'co'] 166: ['pouze', 'ho'] into ['zápas', 'především', 'stavu', 'armády'] 167: ['podniků', 'parlamentu', 'strany'] into ['ČNR', 'majetku', 'KSČ', 'ES'] 168: ['mohl', 'měla', 'měl'] into ['musí', 'má'] 169: ['ceny', 'chce', 'další'] into ['SNR', 'den', 'války', 'organizace'] 170: ['Je', 'Ale', 'A'] into ['Po', 'O', 'První'] 171: ['!'] into ['%', 'zájem', 'h'] 172: ['13', '20'] into ['Sparta', 'Moskva', 'let'] 173: ['roku', 'jednání'] into ['poslanců', 'dnech', 'smlouvy', 'informace'] 174: ['už'] into ['ten', 'zatím', 'byly'] 175: ['ze'] into [';'] 176: ['prostředí', 'lidí', 'korun', 'dolarů'] into ['nám', 'IFS', 'st'] 177: ['byl'] into ['byla', 'nebo'] 178: ['včerejší', 'sebe', 'druhé', 'otázku'] into ['zákona', 'nás'] 179: ['její', 'tato', 'federální'] into ['některé', 'mít', 'poslední', 'jeden'] 180: ['mne', 'vůbec', 'programu', '25'] into ['lidé', 'proto', 'těch', 'toho'] 181: ['včera'] into ['bude'] 182: ['USA'] into ['SSSR', 'Praha', 'první'] 183: ['u'] into ['jeho'] 184: ['před'] into ['zase', 'svou', 'mi'] 185: ['státní', 'skupiny', 'PRAHA'] into ['dvou', 'svých', 'své'] 186: ['době', 'Praze', 'roce', 'případě'] into ['celé', 'České', 'české', 'nich'] 187: ['jsem'] into ['sám', 'může', 'jsme'] 188: ['jsou'] into ['zahraničí', 'soudu', 'vlády'] 189: ['tak', 'pak'] into ['zcela', 'asi', 'podle'] 190: ['ale'] into ['kterou', 'protože', 'kdy', 'aby'] 191: ['?'] into ['prezident', 'Jan', 'vláda', 'ministr'] 192: ['2', '3'] into ['7', '5', '9', '8', '10', '12', '1991', '18'] 193: ['jen'] into ['či', '#'] 194: ['V'] into ['I', 'C', 'listopadu', 'Kčs', 'K'] 195: ['že'] into ['kde', 'která', 'který'] 196: ['Podle', 'Do', 'Na'] into ['Když', 'Z', 'Jak', 'To', 'kolo', 'OKD', 'Branky'] 197: ['od'] into ['tisíc', 'miliónů', 'při'] 198: ['úterý', 'tomto', 'této', 'tuto'] into ['celé', 'České', 'české', 'nich', 'době', 'Praze', 'roce', 'případě'] 199: ['si'] into ['jim', 'mají', 'dosud', 'mu'] 200: ['nic', 'FS', 'budou'] into ['dnes', 'tu', 'však'] 201: ['6', '4'] into ['Sparta', 'Moskva', 'let', '13', '20'] 202: ['tohoto', 'všech', 'bez', 'přes'] into ['tento', 'dva', 'sociální', 'naší'] 203: ['privatizace', 'státu', 'žen'] into ['tam', 'nad', 'NATO', 'již'] 204: ['&slash;', 'ČR', 'ČSFR'] into ['ČNR', 'majetku', 'KSČ', 'ES', 'podniků', 'parlamentu', 'strany'] 205: ['jako'] into ['proti', 'ke', 'mezi'] 206: ['li', 'jenom', 'není', 'ovšem', 'tedy'] into ['bylo', 'až'] 207: ['tom', 'vlastní', 'jejich'] into ['řekl', 'také', 'tím', 'tomu'] 208: ['0', '1'] into ['7', '5', '9', '8', '10', '12', '1991', '18', '2', '3'] 209: ['nové', 'návrh', 'utkání'] into [';', 'ze'] 210: ['poslanců', 'dnech', 'smlouvy', 'informace', 'roku', 'jednání'] into ['zákona', 'nás', 'včerejší', 'sebe', 'druhé', 'otázku'] 211: ['než', 'co', 'kteří', 'které'] into ['kterou', 'protože', 'kdy', 'aby', 'ale'] 212: ['by'] into ['později', 'třeba', 'místo', 'pokud', 'když', 'jak'] 213: ['za'] into ['být', 'ještě'] 214: ["'", 'Za', '*', '='] into ['%', 'zájem', 'h', '!'] 215: ['SNR', 'den', 'války', 'organizace', 'ceny', 'chce', 'další'] into ['zahraničí', 'soudu', 'vlády', 'jsou'] 216: ['pro'] into ['zápas', 'především', 'stavu', 'armády', 'pouze', 'ho'] 217: ['byla', 'nebo', 'byl'] into ['některé', 'mít', 'poslední', 'jeden', 'její', 'tato', 'federální'] 218: ['ten', 'zatím', 'byly', 'už'] into ['bude', 'včera'] 219: ['dokonce', 'často', 'prý', 'přece', 'teď', 'všechno', 'ani'] into ['zase', 'svou', 'mi', 'před'] 220: ['ve'] into ['po'] 221: ['Po', 'O', 'První', 'Je', 'Ale', 'A'] into ['Když', 'Z', 'Jak', 'To', 'kolo', 'OKD', 'Branky', 'Podle', 'Do', 'Na'] 222: ['11', 'M', 'P', 'J'] into ['Sparta', 'Moskva', 'let', '13', '20', '6', '4'] 223: ['musí', 'má', 'mohl', 'měla', 'měl'] into ['sám', 'může', 'jsme', 'jsem'] 224: ['lidé', 'proto', 'těch', 'toho', 'mne', 'vůbec', 'programu', '25'] into ['jeho', 'u'] 225: ['dvou', 'svých', 'své', 'státní', 'skupiny', 'PRAHA'] into ['prezident', 'Jan', 'vláda', 'ministr', '?'] 226: ['nám', 'IFS', 'st', 'prostředí', 'lidí', 'korun', 'dolarů'] into ['tam', 'nad', 'NATO', 'již', 'privatizace', 'státu', 'žen'] 227: ['zcela', 'asi', 'podle', 'tak', 'pak'] into ['dnes', 'tu', 'však', 'nic', 'FS', 'budou'] 228: ['SSSR', 'Praha', 'první', 'USA'] into ['ČNR', 'majetku', 'KSČ', 'ES', 'podniků', 'parlamentu', 'strany', '&slash;', 'ČR', 'ČSFR'] 229: ['k'] into ['tisíc', 'miliónů', 'při', 'od'] 230: ['to'] into ['řekl', 'také', 'tím', 'tomu', 'tom', 'vlastní', 'jejich'] 231: ['či', '#', 'jen'] into ['bylo', 'až', 'li', 'jenom', 'není', 'ovšem', 'tedy'] 232: ['i'] into ['jim', 'mají', 'dosud', 'mu', 'si'] 233: ['z'] into ['tento', 'dva', 'sociální', 'naší', 'tohoto', 'všech', 'bez', 'přes'] 234: ['I', 'C', 'listopadu', 'Kčs', 'K', 'V'] into ['Sparta', 'Moskva', 'let', '13', '20', '6', '4', '11', 'M', 'P', 'J'] 235: ['je'] into ['později', 'třeba', 'místo', 'pokud', 'když', 'jak', 'by'] 236: ['proti', 'ke', 'mezi', 'jako'] into ['být', 'ještě', 'za'] 237: ['bude', 'včera', 'ten', 'zatím', 'byly', 'už'] into ['zase', 'svou', 'mi', 'před', 'dokonce', 'často', 'prý', 'přece', 'teď', 'všechno', 'ani'] 238: ['celé', 'České', 'české', 'nich', 'době', 'Praze', 'roce', 'případě', 'úterý', 'tomto', 'této', 'tuto'] into ['zahraničí', 'soudu', 'vlády', 'jsou', 'SNR', 'den', 'války', 'organizace', 'ceny', 'chce', 'další'] 239: ['do'] into [';', 'ze', 'nové', 'návrh', 'utkání'] 240: ['některé', 'mít', 'poslední', 'jeden', 'její', 'tato', 'federální', 'byla', 'nebo', 'byl'] into ['jeho', 'u', 'lidé', 'proto', 'těch', 'toho', 'mne', 'vůbec', 'programu', '25'] 241: ['zápas', 'především', 'stavu', 'armády', 'pouze', 'ho', 'pro'] into ['sám', 'může', 'jsme', 'jsem', 'musí', 'má', 'mohl', 'měla', 'měl'] 242: ['s'] into ['po', 've'] 243: ['zákona', 'nás', 'včerejší', 'sebe', 'druhé', 'otázku', 'poslanců', 'dnech', 'smlouvy', 'informace', 'roku', 'jednání'] into ['%', 'zájem', 'h', '!', "'", 'Za', '*', '='] 244: [')'] into ['tam', 'nad', 'NATO', 'již', 'privatizace', 'státu', 'žen', 'nám', 'IFS', 'st', 'prostředí', 'lidí', 'korun', 'dolarů'] 245: ['dnes', 'tu', 'však', 'nic', 'FS', 'budou', 'zcela', 'asi', 'podle', 'tak', 'pak'] into ['řekl', 'také', 'tím', 'tomu', 'tom', 'vlastní', 'jejich', 'to'] 246: ['prezident', 'Jan', 'vláda', 'ministr', '?', 'dvou', 'svých', 'své', 'státní', 'skupiny', 'PRAHA'] into ['tisíc', 'miliónů', 'při', 'od', 'k'] 247: ['ČNR', 'majetku', 'KSČ', 'ES', 'podniků', 'parlamentu', 'strany', '&slash;', 'ČR', 'ČSFR', 'SSSR', 'Praha', 'první', 'USA'] into ['bylo', 'až', 'li', 'jenom', 'není', 'ovšem', 'tedy', 'či', '#', 'jen'] 248: ['kterou', 'protože', 'kdy', 'aby', 'ale', 'než', 'co', 'kteří', 'které'] into ['později', 'třeba', 'místo', 'pokud', 'když', 'jak', 'by', 'je'] 249: ['kde', 'která', 'který', 'že'] into ['později', 'třeba', 'místo', 'pokud', 'když', 'jak', 'by', 'je', 'kterou', 'protože', 'kdy', 'aby', 'ale', 'než', 'co', 'kteří', 'které'] 250: ['jim', 'mají', 'dosud', 'mu', 'si', 'i'] into ['být', 'ještě', 'za', 'proti', 'ke', 'mezi', 'jako'] 251: ['o'] into ['tento', 'dva', 'sociální', 'naší', 'tohoto', 'všech', 'bez', 'přes', 'z'] 252: ['7', '5', '9', '8', '10', '12', '1991', '18', '2', '3', '0', '1'] into ['Sparta', 'Moskva', 'let', '13', '20', '6', '4', '11', 'M', 'P', 'J', 'I', 'C', 'listopadu', 'Kčs', 'K', 'V'] 253: [':'] into ['('] 254: ['"'] into ['Když', 'Z', 'Jak', 'To', 'kolo', 'OKD', 'Branky', 'Podle', 'Do', 'Na', 'Po', 'O', 'První', 'Je', 'Ale', 'A'] 255: ['zahraničí', 'soudu', 'vlády', 'jsou', 'SNR', 'den', 'války', 'organizace', 'ceny', 'chce', 'další', 'celé', 'České', 'české', 'nich', 'době', 'Praze', 'roce', 'případě', 'úterý', 'tomto', 'této', 'tuto'] into ['%', 'zájem', 'h', '!', "'", 'Za', '*', '=', 'zákona', 'nás', 'včerejší', 'sebe', 'druhé', 'otázku', 'poslanců', 'dnech', 'smlouvy', 'informace', 'roku', 'jednání'] 256: ['zase', 'svou', 'mi', 'před', 'dokonce', 'často', 'prý', 'přece', 'teď', 'všechno', 'ani', 'bude', 'včera', 'ten', 'zatím', 'byly', 'už'] into ['sám', 'může', 'jsme', 'jsem', 'musí', 'má', 'mohl', 'měla', 'měl', 'zápas', 'především', 'stavu', 'armády', 'pouze', 'ho', 'pro'] 257: ['jeho', 'u', 'lidé', 'proto', 'těch', 'toho', 'mne', 'vůbec', 'programu', '25', 'některé', 'mít', 'poslední', 'jeden', 'její', 'tato', 'federální', 'byla', 'nebo', 'byl'] into ['řekl', 'také', 'tím', 'tomu', 'tom', 'vlastní', 'jejich', 'to', 'dnes', 'tu', 'však', 'nic', 'FS', 'budou', 'zcela', 'asi', 'podle', 'tak', 'pak'] 258: [';', 'ze', 'nové', 'návrh', 'utkání', 'do'] into ['po', 've', 's'] 259: ['na'] into ['tisíc', 'miliónů', 'při', 'od', 'k', 'prezident', 'Jan', 'vláda', 'ministr', '?', 'dvou', 'svých', 'své', 'státní', 'skupiny', 'PRAHA'] 260: ['tam', 'nad', 'NATO', 'již', 'privatizace', 'státu', 'žen', 'nám', 'IFS', 'st', 'prostředí', 'lidí', 'korun', 'dolarů', ')'] into ['bylo', 'až', 'li', 'jenom', 'není', 'ovšem', 'tedy', 'či', '#', 'jen', 'ČNR', 'majetku', 'KSČ', 'ES', 'podniků', 'parlamentu', 'strany', '&slash;', 'ČR', 'ČSFR', 'SSSR', 'Praha', 'první', 'USA'] 261: ['se'] into ['být', 'ještě', 'za', 'proti', 'ke', 'mezi', 'jako', 'jim', 'mají', 'dosud', 'mu', 'si', 'i'] 262: ['v'] into ['tento', 'dva', 'sociální', 'naší', 'tohoto', 'všech', 'bez', 'přes', 'z', 'o'] 263: ['-'] into ['(', ':'] Class 1: ['bylo', 'až', 'li', 'jenom', 'není', 'ovšem', 'tedy', 'či', '#', 'jen', 'ČNR', 'majetku', 'KSČ', 'ES', 'podniků', 'parlamentu', 'strany', '&slash;', 'ČR', 'ČSFR', 'SSSR', 'Praha', 'první', 'USA', 'tam', 'nad', 'NATO', 'již', 'privatizace', 'státu', 'žen', 'nám', 'IFS', 'st', 'prostředí', 'lidí', 'korun', 'dolarů', ')'] Class 2: ['později', 'třeba', 'místo', 'pokud', 'když', 'jak', 'by', 'je', 'kterou', 'protože', 'kdy', 'aby', 'ale', 'než', 'co', 'kteří', 'které', 'kde', 'která', 'který', 'že'] Class 3: ['Když', 'Z', 'Jak', 'To', 'kolo', 'OKD', 'Branky', 'Podle', 'Do', 'Na', 'Po', 'O', 'První', 'Je', 'Ale', 'A', '"'] Class 4: ['Sparta', 'Moskva', 'let', '13', '20', '6', '4', '11', 'M', 'P', 'J', 'I', 'C', 'listopadu', 'Kčs', 'K', 'V', '7', '5', '9', '8', '10', '12', '1991', '18', '2', '3', '0', '1'] Class 5: ['%', 'zájem', 'h', '!', "'", 'Za', '*', '=', 'zákona', 'nás', 'včerejší', 'sebe', 'druhé', 'otázku', 'poslanců', 'dnech', 'smlouvy', 'informace', 'roku', 'jednání', 'zahraničí', 'soudu', 'vlády', 'jsou', 'SNR', 'den', 'války', 'organizace', 'ceny', 'chce', 'další', 'celé', 'České', 'české', 'nich', 'době', 'Praze', 'roce', 'případě', 'úterý', 'tomto', 'této', 'tuto'] Class 6: ['tisíc', 'miliónů', 'při', 'od', 'k', 'prezident', 'Jan', 'vláda', 'ministr', '?', 'dvou', 'svých', 'své', 'státní', 'skupiny', 'PRAHA', 'na'] Class 7: ['tento', 'dva', 'sociální', 'naší', 'tohoto', 'všech', 'bez', 'přes', 'z', 'o', 'v'] Class 8: ['sám', 'může', 'jsme', 'jsem', 'musí', 'má', 'mohl', 'měla', 'měl', 'zápas', 'především', 'stavu', 'armády', 'pouze', 'ho', 'pro', 'zase', 'svou', 'mi', 'před', 'dokonce', 'často', 'prý', 'přece', 'teď', 'všechno', 'ani', 'bude', 'včera', 'ten', 'zatím', 'byly', 'už'] Class 9: ['po', 've', 's', ';', 'ze', 'nové', 'návrh', 'utkání', 'do'] Class 10: ['řekl', 'také', 'tím', 'tomu', 'tom', 'vlastní', 'jejich', 'to', 'dnes', 'tu', 'však', 'nic', 'FS', 'budou', 'zcela', 'asi', 'podle', 'tak', 'pak', 'jeho', 'u', 'lidé', 'proto', 'těch', 'toho', 'mne', 'vůbec', 'programu', '25', 'některé', 'mít', 'poslední', 'jeden', 'její', 'tato', 'federální', 'byla', 'nebo', 'byl'] Class 11: ['(', ':', '-'] Class 12: ['být', 'ještě', 'za', 'proti', 'ke', 'mezi', 'jako', 'jim', 'mají', 'dosud', 'mu', 'si', 'i', 'se'] Class 13: ['a'] Class 14: ['.'] Class 15: [',']
Even though it is very hard to measure whether the merging is performing better then before, it optically seems little more promising, especially at the beginning (first 20 merges).
Merging tags¶
Since there are less tags than words, the algorithm works far quicker and allows to run on longer texts. Some interesting final classes are:
- Class 1: ['VBZ', 'VBP', 'VB'] - representing present tense verbs
- Class 8: [',', ':', '.', '('] - representing all functional characters
- Class 2: ['TO', 'MD'] - representing modal verbs and "to" ('to go'...)