gpt-tokenizer 2.0.0-beta.1 → 2.0.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +125 -43
- package/cjs/GptEncoding.d.ts +5 -0
- package/cjs/GptEncoding.js +6 -0
- package/cjs/GptEncoding.js.map +1 -1
- package/cjs/GptEncoding.test.js +38 -0
- package/cjs/GptEncoding.test.js.map +1 -1
- package/cjs/encoding/cl100k_base.d.ts +4 -0
- package/cjs/encoding/cl100k_base.js +17 -0
- package/cjs/encoding/cl100k_base.js.map +1 -1
- package/cjs/encoding/p50k_base.d.ts +4 -0
- package/cjs/encoding/p50k_base.js +17 -0
- package/cjs/encoding/p50k_base.js.map +1 -1
- package/cjs/encoding/p50k_edit.d.ts +4 -0
- package/cjs/encoding/p50k_edit.js +17 -0
- package/cjs/encoding/p50k_edit.js.map +1 -1
- package/cjs/encoding/r50k_base.d.ts +4 -0
- package/cjs/encoding/r50k_base.js +17 -0
- package/cjs/encoding/r50k_base.js.map +1 -1
- package/cjs/modelParams.js +12 -16
- package/cjs/modelParams.js.map +1 -1
- package/cjs/specialTokens.d.ts +5 -0
- package/cjs/specialTokens.js +9 -0
- package/cjs/specialTokens.js.map +1 -0
- package/data/TestPlans.txt +1024 -0
- package/dist/cl100k_base.js +1 -1
- package/dist/cl100k_base.js.map +1 -1
- package/dist/p50k_base.js +1 -1
- package/dist/p50k_base.js.map +1 -1
- package/dist/p50k_edit.js +1 -1
- package/dist/p50k_edit.js.map +1 -1
- package/dist/r50k_base.js +1 -1
- package/dist/r50k_base.js.map +1 -1
- package/esm/GptEncoding.d.ts +5 -0
- package/esm/GptEncoding.js +8 -1
- package/esm/GptEncoding.js.map +1 -1
- package/esm/GptEncoding.test.js +35 -0
- package/esm/GptEncoding.test.js.map +1 -1
- package/esm/encoding/cl100k_base.d.ts +4 -0
- package/esm/encoding/cl100k_base.js +3 -0
- package/esm/encoding/cl100k_base.js.map +1 -1
- package/esm/encoding/p50k_base.d.ts +4 -0
- package/esm/encoding/p50k_base.js +3 -0
- package/esm/encoding/p50k_base.js.map +1 -1
- package/esm/encoding/p50k_edit.d.ts +4 -0
- package/esm/encoding/p50k_edit.js +3 -0
- package/esm/encoding/p50k_edit.js.map +1 -1
- package/esm/encoding/r50k_base.d.ts +4 -0
- package/esm/encoding/r50k_base.js +3 -0
- package/esm/encoding/r50k_base.js.map +1 -1
- package/esm/modelParams.js +1 -5
- package/esm/modelParams.js.map +1 -1
- package/esm/specialTokens.d.ts +5 -0
- package/esm/specialTokens.js +6 -0
- package/esm/specialTokens.js.map +1 -0
- package/package.json +5 -5
- package/src/GptEncoding.test.ts +40 -0
- package/src/GptEncoding.ts +13 -0
- package/src/encoding/cl100k_base.ts +4 -0
- package/src/encoding/p50k_base.ts +4 -0
- package/src/encoding/p50k_edit.ts +4 -0
- package/src/encoding/r50k_base.ts +4 -0
- package/src/modelParams.ts +7 -6
- package/src/specialTokens.ts +5 -0
package/cjs/modelParams.js
CHANGED
|
@@ -1,18 +1,14 @@
|
|
|
1
1
|
"use strict";
|
|
2
2
|
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
3
|
exports.getModelParamsAsync = exports.getModelParams = void 0;
|
|
4
|
-
const
|
|
5
|
-
const FimPrefix = '<|fim_prefix|>';
|
|
6
|
-
const FimMiddle = '<|fim_middle|>';
|
|
7
|
-
const FimSuffix = '<|fim_suffix|>';
|
|
8
|
-
const EndOfPrompt = '<|endofprompt|>';
|
|
4
|
+
const specialTokens_js_1 = require("./specialTokens.js");
|
|
9
5
|
const tokenSplitRegex = /'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+/gu;
|
|
10
6
|
function R50KBase(mergeableBytePairRanks) {
|
|
11
7
|
return {
|
|
12
8
|
expectedVocabularySize: 50257,
|
|
13
9
|
tokenSplitRegex,
|
|
14
10
|
mergeableBytePairRanks,
|
|
15
|
-
specialTokenMapping: new Map([[EndOfText, 50256]]),
|
|
11
|
+
specialTokenMapping: new Map([[specialTokens_js_1.EndOfText, 50256]]),
|
|
16
12
|
};
|
|
17
13
|
}
|
|
18
14
|
function P50KBase(mergeableBytePairRanks) {
|
|
@@ -20,15 +16,15 @@ function P50KBase(mergeableBytePairRanks) {
|
|
|
20
16
|
expectedVocabularySize: 50281,
|
|
21
17
|
tokenSplitRegex,
|
|
22
18
|
mergeableBytePairRanks,
|
|
23
|
-
specialTokenMapping: new Map([[EndOfText, 50256]]),
|
|
19
|
+
specialTokenMapping: new Map([[specialTokens_js_1.EndOfText, 50256]]),
|
|
24
20
|
};
|
|
25
21
|
}
|
|
26
22
|
function P50KEdit(mergeableBytePairRanks) {
|
|
27
23
|
const specialTokenMapping = new Map([
|
|
28
|
-
[EndOfText, 50256],
|
|
29
|
-
[FimPrefix, 50281],
|
|
30
|
-
[FimMiddle, 50282],
|
|
31
|
-
[FimSuffix, 50283],
|
|
24
|
+
[specialTokens_js_1.EndOfText, 50256],
|
|
25
|
+
[specialTokens_js_1.FimPrefix, 50281],
|
|
26
|
+
[specialTokens_js_1.FimMiddle, 50282],
|
|
27
|
+
[specialTokens_js_1.FimSuffix, 50283],
|
|
32
28
|
]);
|
|
33
29
|
return {
|
|
34
30
|
tokenSplitRegex,
|
|
@@ -38,11 +34,11 @@ function P50KEdit(mergeableBytePairRanks) {
|
|
|
38
34
|
}
|
|
39
35
|
function Cl100KBase(mergeableBytePairRanks) {
|
|
40
36
|
const specialTokenMapping = new Map([
|
|
41
|
-
[EndOfText, 100257],
|
|
42
|
-
[FimPrefix, 100258],
|
|
43
|
-
[FimMiddle, 100259],
|
|
44
|
-
[FimSuffix, 100260],
|
|
45
|
-
[EndOfPrompt, 100276],
|
|
37
|
+
[specialTokens_js_1.EndOfText, 100257],
|
|
38
|
+
[specialTokens_js_1.FimPrefix, 100258],
|
|
39
|
+
[specialTokens_js_1.FimMiddle, 100259],
|
|
40
|
+
[specialTokens_js_1.FimSuffix, 100260],
|
|
41
|
+
[specialTokens_js_1.EndOfPrompt, 100276],
|
|
46
42
|
]);
|
|
47
43
|
return {
|
|
48
44
|
tokenSplitRegex: /(?:'s|'t|'re|'ve|'m|'ll|'d)|[^\r\n\p{L}\p{N}]?\p{L}+|\p{N}{1,3}| ?[^\s\p{L}\p{N}]+[\r\n]*|\s*[\r\n]+|\s+(?!\S)|\s+/giu,
|
package/cjs/modelParams.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"modelParams.js","sourceRoot":"","sources":["../src/modelParams.ts"],"names":[],"mappings":";;;
|
|
1
|
+
{"version":3,"file":"modelParams.js","sourceRoot":"","sources":["../src/modelParams.ts"],"names":[],"mappings":";;;AAGA,yDAM2B;AAmB3B,MAAM,eAAe,GACnB,8EAA8E,CAAA;AAEhF,SAAS,QAAQ,CAAC,sBAAkC;IAClD,OAAO;QACL,sBAAsB,EAAE,KAAM;QAC9B,eAAe;QACf,sBAAsB;QACtB,mBAAmB,EAAE,IAAI,GAAG,CAAiB,CAAC,CAAC,4BAAS,EAAE,KAAM,CAAC,CAAC,CAAC;KACpE,CAAA;AACH,CAAC;AAED,SAAS,QAAQ,CAAC,sBAAkC;IAClD,OAAO;QACL,sBAAsB,EAAE,KAAM;QAC9B,eAAe;QACf,sBAAsB;QACtB,mBAAmB,EAAE,IAAI,GAAG,CAAiB,CAAC,CAAC,4BAAS,EAAE,KAAM,CAAC,CAAC,CAAC;KACpE,CAAA;AACH,CAAC;AAED,SAAS,QAAQ,CAAC,sBAAkC;IAClD,MAAM,mBAAmB,GAAG,IAAI,GAAG,CAAiB;QAClD,CAAC,4BAAS,EAAE,KAAM,CAAC;QACnB,CAAC,4BAAS,EAAE,KAAM,CAAC;QACnB,CAAC,4BAAS,EAAE,KAAM,CAAC;QACnB,CAAC,4BAAS,EAAE,KAAM,CAAC;KACpB,CAAC,CAAA;IAEF,OAAO;QACL,eAAe;QACf,sBAAsB;QACtB,mBAAmB;KACpB,CAAA;AACH,CAAC;AAED,SAAS,UAAU,CAAC,sBAAkC;IACpD,MAAM,mBAAmB,GAAG,IAAI,GAAG,CAAiB;QAClD,CAAC,4BAAS,EAAE,MAAO,CAAC;QACpB,CAAC,4BAAS,EAAE,MAAO,CAAC;QACpB,CAAC,4BAAS,EAAE,MAAO,CAAC;QACpB,CAAC,4BAAS,EAAE,MAAO,CAAC;QACpB,CAAC,8BAAW,EAAE,MAAO,CAAC;KACvB,CAAC,CAAA;IAEF,OAAO;QACL,eAAe,EACb,uHAAuH;QACzH,sBAAsB;QACtB,mBAAmB;KACpB,CAAA;AACH,CAAC;AAOD,SAAgB,cAAc,CAC5B,YAA0B,EAC1B,iBAAsC;IAEtC,MAAM,sBAAsB,GAAG,iBAAiB,CAAC,YAAY,CAAC,CAAA;IAC9D,QAAQ,YAAY,CAAC,WAAW,EAAE,EAAE;QAClC,KAAK,WAAW;YACd,OAAO,QAAQ,CAAC,sBAAsB,CAAC,CAAA;QAEzC,KAAK,WAAW;YACd,OAAO,QAAQ,CAAC,sBAAsB,CAAC,CAAA;QAEzC,KAAK,WAAW;YACd,OAAO,QAAQ,CAAC,sBAAsB,CAAC,CAAA;QAEzC,KAAK,aAAa;YAChB,OAAO,UAAU,CAAC,sBAAsB,CAAC,CAAA;QAE3C;YACE,MAAM,IAAI,KAAK,CAAC,0BAA0B,YAAY,EAAE,CAAC,CAAA;KAC5D;AACH,CAAC;AArBD,wCAqBC;AAEM,KAAK,UAAU,mBAAmB,CACvC,YAA0B,EAC1B,iBAA2C;IAE3C,MAAM,sBAAsB,GAAG,MAAM,iBAAiB,CAAC,YAAY,CAAC,CAAA;IACpE,OAAO,cAAc,CAAC,YAAY,EAAE,GAAG,EAAE,CAAC,sBAAsB,CAAC,CAAA;AACnE,CAAC;AAND,kDAMC"}
|
|
@@ -0,0 +1,9 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.EndOfPrompt = exports.FimSuffix = exports.FimMiddle = exports.FimPrefix = exports.EndOfText = void 0;
|
|
4
|
+
exports.EndOfText = '<|endoftext|>';
|
|
5
|
+
exports.FimPrefix = '<|fim_prefix|>';
|
|
6
|
+
exports.FimMiddle = '<|fim_middle|>';
|
|
7
|
+
exports.FimSuffix = '<|fim_suffix|>';
|
|
8
|
+
exports.EndOfPrompt = '<|endofprompt|>';
|
|
9
|
+
//# sourceMappingURL=specialTokens.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"specialTokens.js","sourceRoot":"","sources":["../src/specialTokens.ts"],"names":[],"mappings":";;;AAAa,QAAA,SAAS,GAAG,eAAe,CAAA;AAC3B,QAAA,SAAS,GAAG,gBAAgB,CAAA;AAC5B,QAAA,SAAS,GAAG,gBAAgB,CAAA;AAC5B,QAAA,SAAS,GAAG,gBAAgB,CAAA;AAC5B,QAAA,WAAW,GAAG,iBAAiB,CAAA"}
|