From 8191fe5744863a2d093e9ffd038d08d917235ebf Mon Sep 17 00:00:00 2001
From: Juan Rodriguez <juand-r@utexas.edu>
Date: Tue, 27 Aug 2024 10:57:16 -0500
Subject: [PATCH 1/2] added dolch word list

---
 vocab/dolch-220-95.txt | 315 +++++++++++++++++++++++++++++++++++++++++
 vocab/notes/note.md    |   3 +-
 2 files changed, 317 insertions(+), 1 deletion(-)
 create mode 100644 vocab/dolch-220-95.txt

diff --git a/vocab/dolch-220-95.txt b/vocab/dolch-220-95.txt
new file mode 100644
index 0000000..efbd459
--- /dev/null
+++ b/vocab/dolch-220-95.txt
@@ -0,0 +1,315 @@
+and
+as
+because
+but
+if
+or
+about
+after
+at
+by
+down
+for
+from
+in
+into
+of
+on
+over
+to
+under
+upon
+with
+he
+her
+him
+his
+i
+it
+its
+me
+my
+myself
+our
+she
+that
+their
+them
+these
+they
+this
+those
+us
+we
+what
+which
+who
+you
+your
+again
+always
+around
+away
+before
+far
+fast
+first
+here
+how
+just
+much
+never
+no
+not
+now
+off
+once
+only
+out
+so
+soon
+then
+there
+today
+together
+too
+up
+very
+well
+when
+where
+why
+yes
+a
+all
+an
+any
+best
+better
+big
+black
+blue
+both
+brown
+clean
+cold
+eight
+every
+five
+four
+full
+funny
+good
+green
+hat
+kind
+light
+little
+long
+many
+new
+old
+one
+own
+pretty
+red
+right
+round
+seven
+six
+small
+some
+nine
+ten
+the
+three
+two
+warm
+white
+yellow
+am
+are
+ask
+ate
+be
+been
+bring
+buy
+call
+came
+can
+carry
+come
+could
+cut
+did
+do
+does
+done
+don't
+draw
+drink
+eat
+fall
+find
+fly
+found
+gave
+get
+give
+go
+goes
+going
+got
+grow
+had
+has
+have
+help
+hold
+hurt
+is
+jump
+keep
+know
+laugh
+let
+like
+live
+look
+made
+make
+may
+must
+open
+pick
+play
+please
+pull
+put
+rani
+read
+ride
+run
+said
+saw
+say
+see
+shall
+show
+sing
+sit
+sleep
+start
+stop
+take
+tell
+thank
+think
+try
+use
+walk
+want
+was
+wash
+went
+were
+will
+wish
+work
+would
+write
+apple
+baby
+back
+ball
+bear
+bed
+bell
+bird
+birthday
+boat
+box
+boy
+bread
+brother
+cake
+car
+cat
+chair
+chicken
+children
+christmas
+coat
+corn
+cow
+day
+dog
+doll
+door
+duck
+egg
+eye
+farm
+farmer
+father
+feet
+fire
+fish
+floor
+flower
+game
+garden
+girl
+goodbye
+grass
+ground
+hand
+head
+hill
+home
+horse
+house
+kitty
+leg
+letter
+man
+men
+milk
+money
+morning
+mother
+name
+nest
+night
+paper
+party
+picture
+pig
+rabbit
+rain
+ring
+robin
+school
+seed
+sheep
+shoe
+sister
+snow
+song
+squirrel
+stock
+street
+sun
+table
+thing
+time
+top
+toy
+tree
+watch
+water
+way
+wind
+window
+wood
diff --git a/vocab/notes/note.md b/vocab/notes/note.md
index 060a3bb..248dbf7 100644
--- a/vocab/notes/note.md
+++ b/vocab/notes/note.md
@@ -11,4 +11,5 @@ CH_HF:
     https://github.com/lpmi-13/machine_readable_wordlists/tree/master
     List originally from Macalister and Webb, "Can L1 children's literature be used in the English language classroom?
     High frequency words in writing for children"
-
+Dolch:
+    220 words (verbs, conjunctions, prepositions, pronouns) and 95 nouns from Dolch, Edward W. "A basic sight vocabulary." The Elementary School Journal 36.6 (1936): 456-460.

From 08a32226ac82598c4ceac1dfc10e5f31c3a01466 Mon Sep 17 00:00:00 2001
From: Juan Rodriguez <juand-r@utexas.edu>
Date: Tue, 27 Aug 2024 19:52:47 -0500
Subject: [PATCH 2/2] added a few extra characters to regex (colon, semicolon,
 dash, asterisk)

---
 make_hf_tokenizer.py | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/make_hf_tokenizer.py b/make_hf_tokenizer.py
index 3a691b2..950e1f9 100644
--- a/make_hf_tokenizer.py
+++ b/make_hf_tokenizer.py
@@ -36,7 +36,7 @@
 from tokenizers.pre_tokenizers import PreTokenizer
 
 #NOTE: this function from https://github.com/noanabeshima/tiny_tokenizer/tree/main
-single_character_toks = re.compile(r'[ ,."\'0-9?!()\n-]')
+single_character_toks = re.compile(r'[ ,—\*;:\."\'0-9?!\(\)\n-]')
 
 def word_tokenize(text, known_toks=None, add_bos=True):
     """The function used for pre-tokenization."""
@@ -121,6 +121,8 @@ def make_tokenizer(vocab_files):
             normalizers.Replace("“", '"'),
             normalizers.Replace("”", '"'),
             normalizers.Replace("’", "'"),
+            normalizers.Replace("‘", "'"),
+            normalizers.Replace("–", "—"),
             normalizers.NFKD(),
             normalizers.Lowercase(),
             normalizers.StripAccents(),