gpt-tokenizer 2.0.0 → 2.1.0
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 +54 -8
- package/cjs/GptEncoding.d.ts +33 -6
- package/cjs/GptEncoding.js +79 -34
- package/cjs/GptEncoding.js.map +1 -1
- package/cjs/GptEncoding.test.js +54 -0
- package/cjs/GptEncoding.test.js.map +1 -1
- package/cjs/encoding/cl100k_base.d.ts +3 -3
- package/cjs/encoding/cl100k_base.js +6 -4
- package/cjs/encoding/cl100k_base.js.map +1 -1
- package/cjs/encoding/p50k_base.d.ts +2 -2
- package/cjs/encoding/p50k_base.js +4 -4
- package/cjs/encoding/p50k_base.js.map +1 -1
- package/cjs/encoding/p50k_edit.d.ts +2 -2
- package/cjs/encoding/p50k_edit.js +4 -4
- package/cjs/encoding/p50k_edit.js.map +1 -1
- package/cjs/encoding/r50k_base.d.ts +2 -2
- package/cjs/encoding/r50k_base.js +4 -4
- package/cjs/encoding/r50k_base.js.map +1 -1
- package/cjs/mapping.d.ts +37 -0
- package/cjs/mapping.js +34 -2
- package/cjs/mapping.js.map +1 -1
- package/cjs/model/gpt-3.5-turbo-0301.d.ts +6 -0
- package/cjs/model/gpt-3.5-turbo-0301.js +38 -0
- package/cjs/model/gpt-3.5-turbo-0301.js.map +1 -0
- package/cjs/model/gpt-3.5-turbo.d.ts +6 -1
- package/cjs/model/gpt-3.5-turbo.js +21 -1
- package/cjs/model/gpt-3.5-turbo.js.map +1 -1
- package/cjs/model/gpt-4-0314.d.ts +6 -0
- package/cjs/model/gpt-4-0314.js +38 -0
- package/cjs/model/gpt-4-0314.js.map +1 -0
- package/cjs/model/gpt-4-32k-0314.d.ts +6 -0
- package/cjs/model/gpt-4-32k-0314.js +38 -0
- package/cjs/model/gpt-4-32k-0314.js.map +1 -0
- package/cjs/model/gpt-4-32k.d.ts +6 -0
- package/cjs/model/gpt-4-32k.js +38 -0
- package/cjs/model/gpt-4-32k.js.map +1 -0
- package/cjs/model/gpt-4.d.ts +6 -1
- package/cjs/model/gpt-4.js +21 -1
- package/cjs/model/gpt-4.js.map +1 -1
- package/cjs/modelParams.d.ts +5 -4
- package/cjs/modelParams.js +7 -4
- package/cjs/modelParams.js.map +1 -1
- package/cjs/specialTokens.d.ts +3 -0
- package/cjs/specialTokens.js +4 -1
- package/cjs/specialTokens.js.map +1 -1
- package/cjs/util.d.ts +2 -0
- package/cjs/util.js +19 -0
- package/cjs/util.js.map +1 -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 +33 -6
- package/esm/GptEncoding.js +81 -36
- package/esm/GptEncoding.js.map +1 -1
- package/esm/GptEncoding.test.js +55 -1
- package/esm/GptEncoding.test.js.map +1 -1
- package/esm/codegen/generateByModel.js +19 -0
- package/esm/codegen/generateByModel.js.map +1 -0
- package/esm/{generateJsModels.js → codegen/generateJsEncodings.js} +4 -4
- package/esm/codegen/generateJsEncodings.js.map +1 -0
- package/esm/encoding/cl100k_base.d.ts +3 -3
- package/esm/encoding/cl100k_base.js +3 -3
- package/esm/encoding/cl100k_base.js.map +1 -1
- package/esm/encoding/p50k_base.d.ts +2 -2
- package/esm/encoding/p50k_base.js +1 -1
- package/esm/encoding/p50k_base.js.map +1 -1
- package/esm/encoding/p50k_edit.d.ts +2 -2
- package/esm/encoding/p50k_edit.js +1 -1
- package/esm/encoding/p50k_edit.js.map +1 -1
- package/esm/encoding/r50k_base.d.ts +2 -2
- package/esm/encoding/r50k_base.js +1 -1
- package/esm/encoding/r50k_base.js.map +1 -1
- package/esm/mapping.d.ts +37 -0
- package/esm/mapping.js +32 -0
- package/esm/mapping.js.map +1 -1
- package/esm/model/gpt-3.5-turbo-0301.d.ts +6 -0
- package/esm/model/gpt-3.5-turbo-0301.js +11 -0
- package/esm/model/gpt-3.5-turbo-0301.js.map +1 -0
- package/esm/model/gpt-3.5-turbo.d.ts +6 -1
- package/esm/model/gpt-3.5-turbo.js +10 -1
- package/esm/model/gpt-3.5-turbo.js.map +1 -1
- package/esm/model/gpt-4-0314.d.ts +6 -0
- package/esm/model/gpt-4-0314.js +11 -0
- package/esm/model/gpt-4-0314.js.map +1 -0
- package/esm/model/gpt-4-32k-0314.d.ts +6 -0
- package/esm/model/gpt-4-32k-0314.js +11 -0
- package/esm/model/gpt-4-32k-0314.js.map +1 -0
- package/esm/model/gpt-4-32k.d.ts +6 -0
- package/esm/model/gpt-4-32k.js +11 -0
- package/esm/model/gpt-4-32k.js.map +1 -0
- package/esm/model/gpt-4.d.ts +6 -1
- package/esm/model/gpt-4.js +10 -1
- package/esm/model/gpt-4.js.map +1 -1
- package/esm/modelParams.d.ts +5 -4
- package/esm/modelParams.js +6 -3
- package/esm/modelParams.js.map +1 -1
- package/esm/specialTokens.d.ts +3 -0
- package/esm/specialTokens.js +3 -0
- package/esm/specialTokens.js.map +1 -1
- package/esm/util.d.ts +2 -0
- package/esm/util.js +14 -0
- package/esm/util.js.map +1 -0
- package/package.json +8 -6
- package/src/GptEncoding.test.ts +78 -1
- package/src/GptEncoding.ts +125 -44
- package/src/__snapshots__/GptEncoding.test.ts.snap +811 -0
- package/src/codegen/generateByModel.ts +35 -0
- package/src/{generateJsModels.ts → codegen/generateJsEncodings.ts} +3 -3
- package/src/encoding/cl100k_base.ts +5 -1
- package/src/encoding/p50k_base.ts +1 -1
- package/src/encoding/p50k_edit.ts +1 -1
- package/src/encoding/r50k_base.ts +1 -1
- package/src/mapping.ts +42 -0
- package/src/model/gpt-3.5-turbo-0301.ts +32 -0
- package/src/model/gpt-3.5-turbo.ts +32 -1
- package/src/model/gpt-4-0314.ts +32 -0
- package/src/model/gpt-4-32k-0314.ts +32 -0
- package/src/model/gpt-4-32k.ts +32 -0
- package/src/model/gpt-4.ts +32 -1
- package/src/modelParams.ts +17 -10
- package/src/specialTokens.ts +3 -0
- package/src/util.ts +15 -0
- package/esm/generateByModel.js +0 -13
- package/esm/generateByModel.js.map +0 -1
- package/esm/generateJsModels.js.map +0 -1
- package/src/generateByModel.ts +0 -20
- /package/esm/{generateByModel.d.ts → codegen/generateByModel.d.ts} +0 -0
- /package/esm/{generateJsModels.d.ts → codegen/generateJsEncodings.d.ts} +0 -0
package/cjs/specialTokens.js
CHANGED
|
@@ -1,9 +1,12 @@
|
|
|
1
1
|
"use strict";
|
|
2
2
|
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
-
exports.EndOfPrompt = exports.FimSuffix = exports.FimMiddle = exports.FimPrefix = exports.EndOfText = void 0;
|
|
3
|
+
exports.EndOfPrompt = exports.ImSep = exports.ImEnd = exports.ImStart = exports.FimSuffix = exports.FimMiddle = exports.FimPrefix = exports.EndOfText = void 0;
|
|
4
4
|
exports.EndOfText = '<|endoftext|>';
|
|
5
5
|
exports.FimPrefix = '<|fim_prefix|>';
|
|
6
6
|
exports.FimMiddle = '<|fim_middle|>';
|
|
7
7
|
exports.FimSuffix = '<|fim_suffix|>';
|
|
8
|
+
exports.ImStart = '<|im_start|>'; // 100264
|
|
9
|
+
exports.ImEnd = '<|im_end|>'; // 100265
|
|
10
|
+
exports.ImSep = '<|im_sep|>'; // 100266
|
|
8
11
|
exports.EndOfPrompt = '<|endofprompt|>';
|
|
9
12
|
//# sourceMappingURL=specialTokens.js.map
|
package/cjs/specialTokens.js.map
CHANGED
|
@@ -1 +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"}
|
|
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,OAAO,GAAG,cAAc,CAAA,CAAC,SAAS;AAClC,QAAA,KAAK,GAAG,YAAY,CAAA,CAAC,SAAS;AAC9B,QAAA,KAAK,GAAG,YAAY,CAAA,CAAC,SAAS;AAC9B,QAAA,WAAW,GAAG,iBAAiB,CAAA"}
|
package/cjs/util.d.ts
ADDED
package/cjs/util.js
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
3
|
+
exports.getSpecialTokenRegex = exports.getMaxValueFromMap = void 0;
|
|
4
|
+
const escapeRegExp_js_1 = require("./escapeRegExp.js");
|
|
5
|
+
function getMaxValueFromMap(map) {
|
|
6
|
+
let max = 0;
|
|
7
|
+
map.forEach((val) => {
|
|
8
|
+
max = Math.max(max, val);
|
|
9
|
+
});
|
|
10
|
+
return max;
|
|
11
|
+
}
|
|
12
|
+
exports.getMaxValueFromMap = getMaxValueFromMap;
|
|
13
|
+
function getSpecialTokenRegex(tokens) {
|
|
14
|
+
const escapedTokens = [...tokens].map(escapeRegExp_js_1.escapeRegExp);
|
|
15
|
+
const inner = escapedTokens.join('|');
|
|
16
|
+
return new RegExp(`(${inner})`);
|
|
17
|
+
}
|
|
18
|
+
exports.getSpecialTokenRegex = getSpecialTokenRegex;
|
|
19
|
+
//# sourceMappingURL=util.js.map
|
package/cjs/util.js.map
ADDED
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"util.js","sourceRoot":"","sources":["../src/util.ts"],"names":[],"mappings":";;;AAAA,uDAAgD;AAEhD,SAAgB,kBAAkB,CAAC,GAAyB;IAC1D,IAAI,GAAG,GAAG,CAAC,CAAA;IACX,GAAG,CAAC,OAAO,CAAC,CAAC,GAAG,EAAE,EAAE;QAClB,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,EAAE,GAAG,CAAC,CAAA;IAC1B,CAAC,CAAC,CAAA;IACF,OAAO,GAAG,CAAA;AACZ,CAAC;AAND,gDAMC;AAED,SAAgB,oBAAoB,CAAC,MAAmB;IACtD,MAAM,aAAa,GAAG,CAAC,GAAG,MAAM,CAAC,CAAC,GAAG,CAAC,8BAAY,CAAC,CAAA;IACnD,MAAM,KAAK,GAAG,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;IACrC,OAAO,IAAI,MAAM,CAAC,IAAI,KAAK,GAAG,CAAC,CAAA;AACjC,CAAC;AAJD,oDAIC"}
|