tiktoken_ruby 0.0.6-x86_64-darwin → 0.0.7-x86_64-darwin
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/Gemfile +5 -9
- data/Gemfile.lock +44 -33
- data/lib/tiktoken_ruby/3.0/tiktoken_ruby.bundle +0 -0
- data/lib/tiktoken_ruby/3.1/tiktoken_ruby.bundle +0 -0
- data/lib/tiktoken_ruby/3.2/tiktoken_ruby.bundle +0 -0
- data/lib/tiktoken_ruby/3.3/tiktoken_ruby.bundle +0 -0
- data/lib/tiktoken_ruby/version.rb +1 -1
- data/lib/tiktoken_ruby.rb +36 -19
- metadata +4 -3
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 6bbdd6164228f0fa953d7988326bec5899f29b87cbe4c6b19e3146b6a98a6f35
|
4
|
+
data.tar.gz: 5a53fdc3cd7eb9289cffcf2dbe88328d1387c777bdf8312571c9e7c15d8ec2fc
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: d895625940b103633790071183b00db0a0ee4c3df83a047ed34383de1ce54803c5b7caa7e2f71badd3d099197961c050abb425f40da9b1a57a0284e5e26e4d77
|
7
|
+
data.tar.gz: fddc06e751328c8cc773df5da02949ae021ad2f9cc1039e6be6affd7f954e5ca96284e3981c25a923455ef106c4f0d836076b58e0e41f55f2b2f265eaae13162
|
data/Gemfile
CHANGED
@@ -2,15 +2,11 @@
|
|
2
2
|
|
3
3
|
source "https://rubygems.org"
|
4
4
|
|
5
|
-
# Specify your gem's dependencies in tiktoken_ruby.gemspec
|
6
5
|
gemspec
|
7
6
|
|
8
|
-
gem "rake"
|
9
|
-
|
7
|
+
gem "rake"
|
10
8
|
gem "rake-compiler"
|
11
|
-
|
12
|
-
gem "
|
13
|
-
|
14
|
-
gem "
|
15
|
-
|
16
|
-
gem "yard-doctest", "~> 0.1.17"
|
9
|
+
gem "rspec"
|
10
|
+
gem "standard"
|
11
|
+
gem "yard-doctest"
|
12
|
+
gem "racc"
|
data/Gemfile.lock
CHANGED
@@ -1,64 +1,74 @@
|
|
1
1
|
PATH
|
2
2
|
remote: .
|
3
3
|
specs:
|
4
|
-
tiktoken_ruby (0.0.
|
5
|
-
rb_sys (
|
4
|
+
tiktoken_ruby (0.0.7)
|
5
|
+
rb_sys (>= 0.9.86)
|
6
6
|
|
7
7
|
GEM
|
8
8
|
remote: https://rubygems.org/
|
9
9
|
specs:
|
10
10
|
ast (2.4.2)
|
11
11
|
diff-lcs (1.5.0)
|
12
|
-
json (2.
|
12
|
+
json (2.7.1)
|
13
13
|
language_server-protocol (3.17.0.3)
|
14
|
-
|
15
|
-
|
16
|
-
|
14
|
+
lint_roller (1.1.0)
|
15
|
+
minitest (5.21.2)
|
16
|
+
parallel (1.24.0)
|
17
|
+
parser (3.3.0.4)
|
17
18
|
ast (~> 2.4.1)
|
19
|
+
racc
|
20
|
+
racc (1.7.3)
|
18
21
|
rainbow (3.1.1)
|
19
|
-
rake (13.0
|
20
|
-
rake-compiler (1.2.
|
22
|
+
rake (13.1.0)
|
23
|
+
rake-compiler (1.2.5)
|
21
24
|
rake
|
22
|
-
rb_sys (0.9.
|
23
|
-
regexp_parser (2.
|
24
|
-
rexml (3.2.
|
25
|
+
rb_sys (0.9.86)
|
26
|
+
regexp_parser (2.9.0)
|
27
|
+
rexml (3.2.6)
|
25
28
|
rspec (3.12.0)
|
26
29
|
rspec-core (~> 3.12.0)
|
27
30
|
rspec-expectations (~> 3.12.0)
|
28
31
|
rspec-mocks (~> 3.12.0)
|
29
|
-
rspec-core (3.12.
|
32
|
+
rspec-core (3.12.2)
|
30
33
|
rspec-support (~> 3.12.0)
|
31
|
-
rspec-expectations (3.12.
|
34
|
+
rspec-expectations (3.12.3)
|
32
35
|
diff-lcs (>= 1.2.0, < 2.0)
|
33
36
|
rspec-support (~> 3.12.0)
|
34
|
-
rspec-mocks (3.12.
|
37
|
+
rspec-mocks (3.12.6)
|
35
38
|
diff-lcs (>= 1.2.0, < 2.0)
|
36
39
|
rspec-support (~> 3.12.0)
|
37
|
-
rspec-support (3.12.
|
38
|
-
rubocop (1.
|
40
|
+
rspec-support (3.12.1)
|
41
|
+
rubocop (1.59.0)
|
39
42
|
json (~> 2.3)
|
43
|
+
language_server-protocol (>= 3.17.0)
|
40
44
|
parallel (~> 1.10)
|
41
|
-
parser (>= 3.2.
|
45
|
+
parser (>= 3.2.2.4)
|
42
46
|
rainbow (>= 2.2.2, < 4.0)
|
43
47
|
regexp_parser (>= 1.8, < 3.0)
|
44
48
|
rexml (>= 3.2.5, < 4.0)
|
45
|
-
rubocop-ast (>= 1.
|
49
|
+
rubocop-ast (>= 1.30.0, < 2.0)
|
46
50
|
ruby-progressbar (~> 1.7)
|
47
51
|
unicode-display_width (>= 2.4.0, < 3.0)
|
48
|
-
rubocop-ast (1.
|
52
|
+
rubocop-ast (1.30.0)
|
49
53
|
parser (>= 3.2.1.0)
|
50
|
-
rubocop-performance (1.
|
51
|
-
rubocop (>= 1.
|
52
|
-
rubocop-ast (>=
|
54
|
+
rubocop-performance (1.20.2)
|
55
|
+
rubocop (>= 1.48.1, < 2.0)
|
56
|
+
rubocop-ast (>= 1.30.0, < 2.0)
|
53
57
|
ruby-progressbar (1.13.0)
|
54
|
-
standard (1.
|
58
|
+
standard (1.33.0)
|
55
59
|
language_server-protocol (~> 3.17.0.2)
|
56
|
-
|
57
|
-
rubocop
|
58
|
-
|
59
|
-
|
60
|
-
|
61
|
-
|
60
|
+
lint_roller (~> 1.0)
|
61
|
+
rubocop (~> 1.59.0)
|
62
|
+
standard-custom (~> 1.0.0)
|
63
|
+
standard-performance (~> 1.3)
|
64
|
+
standard-custom (1.0.2)
|
65
|
+
lint_roller (~> 1.0)
|
66
|
+
rubocop (~> 1.50)
|
67
|
+
standard-performance (1.3.1)
|
68
|
+
lint_roller (~> 1.1)
|
69
|
+
rubocop-performance (~> 1.20.2)
|
70
|
+
unicode-display_width (2.5.0)
|
71
|
+
yard (0.9.34)
|
62
72
|
yard-doctest (0.1.17)
|
63
73
|
minitest
|
64
74
|
yard
|
@@ -70,12 +80,13 @@ PLATFORMS
|
|
70
80
|
x86_64-linux
|
71
81
|
|
72
82
|
DEPENDENCIES
|
73
|
-
|
83
|
+
racc
|
84
|
+
rake
|
74
85
|
rake-compiler
|
75
|
-
rspec
|
76
|
-
standard
|
86
|
+
rspec
|
87
|
+
standard
|
77
88
|
tiktoken_ruby!
|
78
|
-
yard-doctest
|
89
|
+
yard-doctest
|
79
90
|
|
80
91
|
BUNDLED WITH
|
81
92
|
2.4.6
|
Binary file
|
Binary file
|
Binary file
|
Binary file
|
data/lib/tiktoken_ruby.rb
CHANGED
@@ -33,17 +33,15 @@ module Tiktoken
|
|
33
33
|
# enc = Tiktoken.encoding_for_model("gpt-4")
|
34
34
|
# enc.encode("hello world").length #=> 2
|
35
35
|
def encoding_for_model(model_name)
|
36
|
-
|
37
|
-
|
38
|
-
model_name = prefix
|
39
|
-
break
|
40
|
-
end
|
36
|
+
if MODEL_TO_ENCODING_NAME.key?(model_name.to_sym)
|
37
|
+
return get_encoding(MODEL_TO_ENCODING_NAME[model_name.to_sym])
|
41
38
|
end
|
42
39
|
|
43
|
-
|
44
|
-
|
45
|
-
|
46
|
-
|
40
|
+
MODEL_PREFIX_TO_ENCODING.each do |prefix, encoding|
|
41
|
+
if model_name.start_with?(prefix.to_s)
|
42
|
+
return get_encoding(encoding)
|
43
|
+
end
|
44
|
+
end
|
47
45
|
end
|
48
46
|
|
49
47
|
# Lists all the encodings that are supported
|
@@ -67,12 +65,22 @@ module Tiktoken
|
|
67
65
|
:cl100k_base
|
68
66
|
]
|
69
67
|
|
70
|
-
# taken from the python library here https://github.com/openai/tiktoken/blob/main/tiktoken/model.py
|
68
|
+
# taken from the python library here https://github.com/openai/tiktoken/blob/main/tiktoken/model.py
|
71
69
|
# that is also MIT licensed but by OpenAI
|
72
70
|
MODEL_TO_ENCODING_NAME = {
|
71
|
+
# chat
|
73
72
|
"gpt-4": "cl100k_base",
|
74
73
|
"gpt-3.5-turbo": "cl100k_base",
|
75
|
-
#
|
74
|
+
"gpt-35-turbo": "cl100k_base", # Azure deployment name
|
75
|
+
# base
|
76
|
+
"davinci-002": "cl100k_base",
|
77
|
+
"babbage-002": "cl100k_base",
|
78
|
+
# embeddings
|
79
|
+
"text-embedding-ada-002": "cl100k_base",
|
80
|
+
"text-embedding-3-small": "cl100k_base",
|
81
|
+
"text-embedding-3-large": "cl100k_base",
|
82
|
+
# DEPRECATED MODELS
|
83
|
+
# text (DEPRECATED)
|
76
84
|
"text-davinci-003": "p50k_base",
|
77
85
|
"text-davinci-002": "p50k_base",
|
78
86
|
"text-davinci-001": "r50k_base",
|
@@ -83,19 +91,17 @@ module Tiktoken
|
|
83
91
|
curie: "r50k_base",
|
84
92
|
babbage: "r50k_base",
|
85
93
|
ada: "r50k_base",
|
86
|
-
# code
|
94
|
+
# code (DEPRECATED)
|
87
95
|
"code-davinci-002": "p50k_base",
|
88
96
|
"code-davinci-001": "p50k_base",
|
89
97
|
"code-cushman-002": "p50k_base",
|
90
98
|
"code-cushman-001": "p50k_base",
|
91
99
|
"davinci-codex": "p50k_base",
|
92
100
|
"cushman-codex": "p50k_base",
|
93
|
-
# edit
|
101
|
+
# edit (DEPRECATED)
|
94
102
|
"text-davinci-edit-001": "p50k_edit",
|
95
103
|
"code-davinci-edit-001": "p50k_edit",
|
96
|
-
# embeddings
|
97
|
-
"text-embedding-ada-002": "cl100k_base",
|
98
|
-
# old embeddings
|
104
|
+
# old embeddings (DEPRECATED)
|
99
105
|
"text-similarity-davinci-001": "r50k_base",
|
100
106
|
"text-similarity-curie-001": "r50k_base",
|
101
107
|
"text-similarity-babbage-001": "r50k_base",
|
@@ -105,10 +111,21 @@ module Tiktoken
|
|
105
111
|
"text-search-babbage-doc-001": "r50k_base",
|
106
112
|
"text-search-ada-doc-001": "r50k_base",
|
107
113
|
"code-search-babbage-code-001": "r50k_base",
|
108
|
-
"code-search-ada-code-001": "r50k_base"
|
114
|
+
"code-search-ada-code-001": "r50k_base",
|
115
|
+
# open source
|
116
|
+
gpt2: "gpt2"
|
109
117
|
}
|
110
118
|
|
111
|
-
|
112
|
-
|
119
|
+
MODEL_PREFIX_TO_ENCODING = {
|
120
|
+
# chat
|
121
|
+
"gpt-4-": "cl100k_base", # e.g., gpt-4-0314, etc., plus gpt-4-32k
|
122
|
+
"gpt-3.5-turbo-": "cl100k_base", # e.g, gpt-3.5-turbo-0301, -0401, etc.
|
123
|
+
"gpt-35-turbo-": "cl100k_base", # Azure deployment name
|
124
|
+
# fine-tuned
|
125
|
+
"ft:gpt-4": "cl100k_base",
|
126
|
+
"ft:gpt-3.5-turbo": "cl100k_base",
|
127
|
+
"ft:davinci-002": "cl100k_base",
|
128
|
+
"ft:babbage-002": "cl100k_base"
|
129
|
+
}
|
113
130
|
end
|
114
131
|
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: tiktoken_ruby
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.0.
|
4
|
+
version: 0.0.7
|
5
5
|
platform: x86_64-darwin
|
6
6
|
authors:
|
7
7
|
- IAPark
|
8
8
|
autorequire:
|
9
9
|
bindir: exe
|
10
10
|
cert_chain: []
|
11
|
-
date:
|
11
|
+
date: 2024-02-14 00:00:00.000000000 Z
|
12
12
|
dependencies: []
|
13
13
|
description: An unofficial Ruby wrapper for Tiktoken, a BPE tokenizer written by and
|
14
14
|
used by OpenAI. It can be used to count the number of tokens in text before sending
|
@@ -31,6 +31,7 @@ files:
|
|
31
31
|
- lib/tiktoken_ruby/3.0/tiktoken_ruby.bundle
|
32
32
|
- lib/tiktoken_ruby/3.1/tiktoken_ruby.bundle
|
33
33
|
- lib/tiktoken_ruby/3.2/tiktoken_ruby.bundle
|
34
|
+
- lib/tiktoken_ruby/3.3/tiktoken_ruby.bundle
|
34
35
|
- lib/tiktoken_ruby/encoding.rb
|
35
36
|
- lib/tiktoken_ruby/version.rb
|
36
37
|
- sig/tiktoken_ruby.rbs
|
@@ -52,7 +53,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
|
|
52
53
|
version: '3.0'
|
53
54
|
- - "<"
|
54
55
|
- !ruby/object:Gem::Version
|
55
|
-
version: 3.
|
56
|
+
version: 3.4.dev
|
56
57
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
57
58
|
requirements:
|
58
59
|
- - ">="
|