Skip to content

Commit 989f5a0

Browse files
Copilotstephentoub
andauthored
Add gpt-5.3 model support to TiktokenTokenizer (#7579)
* Initial plan * Add gpt-5.3 support in TiktokenTokenizer Co-authored-by: stephentoub <2642209+stephentoub@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: stephentoub <2642209+stephentoub@users.noreply.github.com>
1 parent 3604580 commit 989f5a0

File tree

2 files changed

+6
-0
lines changed

2 files changed

+6
-0
lines changed

src/Microsoft.ML.Tokenizers/Model/TiktokenTokenizer.cs

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1044,6 +1044,7 @@ private static readonly (string Prefix, ModelEncoding Encoding)[] _modelPrefixTo
10441044
( "o4-mini-", ModelEncoding.O200kBase ), // e.g. o4-mini
10451045

10461046
// chat
1047+
( "gpt-5.3-", ModelEncoding.O200kBase ),
10471048
( "gpt-5.2-", ModelEncoding.O200kBase ),
10481049
( "gpt-5.1-", ModelEncoding.O200kBase ),
10491050
( "gpt-5-", ModelEncoding.O200kBase ),
@@ -1073,6 +1074,7 @@ private static readonly (string Prefix, ModelEncoding Encoding)[] _modelPrefixTo
10731074
{ "o4-mini", ModelEncoding.O200kBase },
10741075

10751076
// chat
1077+
{ "gpt-5.3", ModelEncoding.O200kBase },
10761078
{ "gpt-5.2", ModelEncoding.O200kBase },
10771079
{ "gpt-5.1", ModelEncoding.O200kBase },
10781080
{ "gpt-5", ModelEncoding.O200kBase },

test/Microsoft.ML.Tokenizers.Tests/TiktokenTests.cs

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -38,6 +38,7 @@ public class TiktokenTests
3838
public static Tokenizer GPT5 { get; } = TiktokenTokenizer.CreateForModel("gpt-5");
3939
public static Tokenizer GPT5_1 { get; } = TiktokenTokenizer.CreateForModel("gpt-5.1");
4040
public static Tokenizer GPT5_2 { get; } = TiktokenTokenizer.CreateForModel("gpt-5.2");
41+
public static Tokenizer GPT5_3 { get; } = TiktokenTokenizer.CreateForModel("gpt-5.3");
4142
public static Tokenizer Phi4 { get; } = TiktokenTokenizer.CreateForModel("phi-4");
4243
public static TiktokenTokenizer GptOss { get; } = TiktokenTokenizer.CreateForModel("gpt-oss-20b");
4344

@@ -421,6 +422,8 @@ public void TestEncodeR50kBase()
421422
[InlineData("gpt-5.1-mini")]
422423
[InlineData("gpt-5.2")]
423424
[InlineData("gpt-5.2-mini")]
425+
[InlineData("gpt-5.3")]
426+
[InlineData("gpt-5.3-mini")]
424427
[InlineData("chatgpt-4o-")]
425428
[InlineData("gpt-4")]
426429
[InlineData("gpt-4-")]
@@ -541,6 +544,7 @@ public void TestEncodingNamesNegativeCases()
541544
[InlineData("gpt-5")]
542545
[InlineData("gpt-5.1")]
543546
[InlineData("gpt-5.2")]
547+
[InlineData("gpt-5.3")]
544548
[InlineData("o1")]
545549
[InlineData("o3")]
546550
[InlineData("o4-mini")]

0 commit comments

Comments
 (0)