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.
Files changed (133) hide show
  1. package/README.md +54 -8
  2. package/cjs/GptEncoding.d.ts +33 -6
  3. package/cjs/GptEncoding.js +79 -34
  4. package/cjs/GptEncoding.js.map +1 -1
  5. package/cjs/GptEncoding.test.js +54 -0
  6. package/cjs/GptEncoding.test.js.map +1 -1
  7. package/cjs/encoding/cl100k_base.d.ts +3 -3
  8. package/cjs/encoding/cl100k_base.js +6 -4
  9. package/cjs/encoding/cl100k_base.js.map +1 -1
  10. package/cjs/encoding/p50k_base.d.ts +2 -2
  11. package/cjs/encoding/p50k_base.js +4 -4
  12. package/cjs/encoding/p50k_base.js.map +1 -1
  13. package/cjs/encoding/p50k_edit.d.ts +2 -2
  14. package/cjs/encoding/p50k_edit.js +4 -4
  15. package/cjs/encoding/p50k_edit.js.map +1 -1
  16. package/cjs/encoding/r50k_base.d.ts +2 -2
  17. package/cjs/encoding/r50k_base.js +4 -4
  18. package/cjs/encoding/r50k_base.js.map +1 -1
  19. package/cjs/mapping.d.ts +37 -0
  20. package/cjs/mapping.js +34 -2
  21. package/cjs/mapping.js.map +1 -1
  22. package/cjs/model/gpt-3.5-turbo-0301.d.ts +6 -0
  23. package/cjs/model/gpt-3.5-turbo-0301.js +38 -0
  24. package/cjs/model/gpt-3.5-turbo-0301.js.map +1 -0
  25. package/cjs/model/gpt-3.5-turbo.d.ts +6 -1
  26. package/cjs/model/gpt-3.5-turbo.js +21 -1
  27. package/cjs/model/gpt-3.5-turbo.js.map +1 -1
  28. package/cjs/model/gpt-4-0314.d.ts +6 -0
  29. package/cjs/model/gpt-4-0314.js +38 -0
  30. package/cjs/model/gpt-4-0314.js.map +1 -0
  31. package/cjs/model/gpt-4-32k-0314.d.ts +6 -0
  32. package/cjs/model/gpt-4-32k-0314.js +38 -0
  33. package/cjs/model/gpt-4-32k-0314.js.map +1 -0
  34. package/cjs/model/gpt-4-32k.d.ts +6 -0
  35. package/cjs/model/gpt-4-32k.js +38 -0
  36. package/cjs/model/gpt-4-32k.js.map +1 -0
  37. package/cjs/model/gpt-4.d.ts +6 -1
  38. package/cjs/model/gpt-4.js +21 -1
  39. package/cjs/model/gpt-4.js.map +1 -1
  40. package/cjs/modelParams.d.ts +5 -4
  41. package/cjs/modelParams.js +7 -4
  42. package/cjs/modelParams.js.map +1 -1
  43. package/cjs/specialTokens.d.ts +3 -0
  44. package/cjs/specialTokens.js +4 -1
  45. package/cjs/specialTokens.js.map +1 -1
  46. package/cjs/util.d.ts +2 -0
  47. package/cjs/util.js +19 -0
  48. package/cjs/util.js.map +1 -0
  49. package/dist/cl100k_base.js +1 -1
  50. package/dist/cl100k_base.js.map +1 -1
  51. package/dist/p50k_base.js +1 -1
  52. package/dist/p50k_base.js.map +1 -1
  53. package/dist/p50k_edit.js +1 -1
  54. package/dist/p50k_edit.js.map +1 -1
  55. package/dist/r50k_base.js +1 -1
  56. package/dist/r50k_base.js.map +1 -1
  57. package/esm/GptEncoding.d.ts +33 -6
  58. package/esm/GptEncoding.js +81 -36
  59. package/esm/GptEncoding.js.map +1 -1
  60. package/esm/GptEncoding.test.js +55 -1
  61. package/esm/GptEncoding.test.js.map +1 -1
  62. package/esm/codegen/generateByModel.js +19 -0
  63. package/esm/codegen/generateByModel.js.map +1 -0
  64. package/esm/{generateJsModels.js → codegen/generateJsEncodings.js} +4 -4
  65. package/esm/codegen/generateJsEncodings.js.map +1 -0
  66. package/esm/encoding/cl100k_base.d.ts +3 -3
  67. package/esm/encoding/cl100k_base.js +3 -3
  68. package/esm/encoding/cl100k_base.js.map +1 -1
  69. package/esm/encoding/p50k_base.d.ts +2 -2
  70. package/esm/encoding/p50k_base.js +1 -1
  71. package/esm/encoding/p50k_base.js.map +1 -1
  72. package/esm/encoding/p50k_edit.d.ts +2 -2
  73. package/esm/encoding/p50k_edit.js +1 -1
  74. package/esm/encoding/p50k_edit.js.map +1 -1
  75. package/esm/encoding/r50k_base.d.ts +2 -2
  76. package/esm/encoding/r50k_base.js +1 -1
  77. package/esm/encoding/r50k_base.js.map +1 -1
  78. package/esm/mapping.d.ts +37 -0
  79. package/esm/mapping.js +32 -0
  80. package/esm/mapping.js.map +1 -1
  81. package/esm/model/gpt-3.5-turbo-0301.d.ts +6 -0
  82. package/esm/model/gpt-3.5-turbo-0301.js +11 -0
  83. package/esm/model/gpt-3.5-turbo-0301.js.map +1 -0
  84. package/esm/model/gpt-3.5-turbo.d.ts +6 -1
  85. package/esm/model/gpt-3.5-turbo.js +10 -1
  86. package/esm/model/gpt-3.5-turbo.js.map +1 -1
  87. package/esm/model/gpt-4-0314.d.ts +6 -0
  88. package/esm/model/gpt-4-0314.js +11 -0
  89. package/esm/model/gpt-4-0314.js.map +1 -0
  90. package/esm/model/gpt-4-32k-0314.d.ts +6 -0
  91. package/esm/model/gpt-4-32k-0314.js +11 -0
  92. package/esm/model/gpt-4-32k-0314.js.map +1 -0
  93. package/esm/model/gpt-4-32k.d.ts +6 -0
  94. package/esm/model/gpt-4-32k.js +11 -0
  95. package/esm/model/gpt-4-32k.js.map +1 -0
  96. package/esm/model/gpt-4.d.ts +6 -1
  97. package/esm/model/gpt-4.js +10 -1
  98. package/esm/model/gpt-4.js.map +1 -1
  99. package/esm/modelParams.d.ts +5 -4
  100. package/esm/modelParams.js +6 -3
  101. package/esm/modelParams.js.map +1 -1
  102. package/esm/specialTokens.d.ts +3 -0
  103. package/esm/specialTokens.js +3 -0
  104. package/esm/specialTokens.js.map +1 -1
  105. package/esm/util.d.ts +2 -0
  106. package/esm/util.js +14 -0
  107. package/esm/util.js.map +1 -0
  108. package/package.json +8 -6
  109. package/src/GptEncoding.test.ts +78 -1
  110. package/src/GptEncoding.ts +125 -44
  111. package/src/__snapshots__/GptEncoding.test.ts.snap +811 -0
  112. package/src/codegen/generateByModel.ts +35 -0
  113. package/src/{generateJsModels.ts → codegen/generateJsEncodings.ts} +3 -3
  114. package/src/encoding/cl100k_base.ts +5 -1
  115. package/src/encoding/p50k_base.ts +1 -1
  116. package/src/encoding/p50k_edit.ts +1 -1
  117. package/src/encoding/r50k_base.ts +1 -1
  118. package/src/mapping.ts +42 -0
  119. package/src/model/gpt-3.5-turbo-0301.ts +32 -0
  120. package/src/model/gpt-3.5-turbo.ts +32 -1
  121. package/src/model/gpt-4-0314.ts +32 -0
  122. package/src/model/gpt-4-32k-0314.ts +32 -0
  123. package/src/model/gpt-4-32k.ts +32 -0
  124. package/src/model/gpt-4.ts +32 -1
  125. package/src/modelParams.ts +17 -10
  126. package/src/specialTokens.ts +3 -0
  127. package/src/util.ts +15 -0
  128. package/esm/generateByModel.js +0 -13
  129. package/esm/generateByModel.js.map +0 -1
  130. package/esm/generateJsModels.js.map +0 -1
  131. package/src/generateByModel.ts +0 -20
  132. /package/esm/{generateByModel.d.ts → codegen/generateByModel.d.ts} +0 -0
  133. /package/esm/{generateJsModels.d.ts → codegen/generateJsEncodings.d.ts} +0 -0
@@ -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
@@ -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
@@ -0,0 +1,2 @@
1
+ export declare function getMaxValueFromMap(map: Map<unknown, number>): number;
2
+ export declare function getSpecialTokenRegex(tokens: Set<string>): RegExp;
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
@@ -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"}