Flaky test: EnglishRobertaTests.TestTokenizerEncoding

Author: svickCreated Sep 16, 2026Updated Sep 19, 2026

Labelsuntriagedblocking-clean-ciKnown Build Error

Build Information

Build: https://dev.azure.com/dnceng-public/cbb18261-c48f-4abb-8651-8cdcb5474649/_build/results?buildId=1599215 Build error leg or test failing: Microsoft.ML.Tokenizers.Tests.EnglishRobertaTests.TestTokenizerEncoding Pull request: https://github.com/dotnet/machinelearning/pull/7704

Error Message

DO NOT USE JSON BELOW IF THIS IS A BUILD BREAK otherwise build analysis will allow pull requests to merge that break the build worse. For a build break, do not use this issue form. Make a regular new issue.

Fill the error message using step by step known issues guidance.

json

{
  "ErrorMessage": ["EnglishRobertaTests.TestTokenizerEncoding", "ArgumentOutOfRangeException : value must be non-negative"],
  "BuildRetry": false,
  "ExcludeConsoleLog": false
}

Microsoft.ML.Tokenizers.Tests.EnglishRobertaTests.TestTokenizerEncoding(text: \"the brown fox jumped over the lazy dog!\", expectedTokens: [\"the\", \"Ġbrown\", \"Ġfox\", \"Ġjumped\", \"Ġover\", ···], expectedOffsets: [Tuple (0, 3), Tuple (3, 6), Tuple (9, 4), Tuple (13, 7), Tuple (20, 5), ···], expectedIds: [1169, 7586, 21831, 11687, 625, ···])
Exception Message
System.ArgumentOutOfRangeException : value must be non-negative
Parameter name: value
Stack Trace
   at System.Index.ThrowValueArgumentOutOfRange_NeedNonNegNumException()
   at Microsoft.ML.Tokenizers.EnglishRobertaTokenizer.ModifyTokenListOffsets(IReadOnlyList`1 tokens, Span`1 indexMapping) in D:\a\_work\1\s\src\Microsoft.ML.Tokenizers\Model\EnglishRobertaTokenizer.cs:line 913
   at Microsoft.ML.Tokenizers.EnglishRobertaTokenizer.EncodeInternal(ReadOnlySpan`1 text) in D:\a\_work\1\s\src\Microsoft.ML.Tokenizers\Model\EnglishRobertaTokenizer.cs:line 397
   at Microsoft.ML.Tokenizers.EnglishRobertaTokenizer.EncodeToTokens(String text, ReadOnlySpan`1 textSpan, EncodeSettings settings) in D:\a\_work\1\s\src\Microsoft.ML.Tokenizers\Model\EnglishRobertaTokenizer.cs:line 337
   at Microsoft.ML.Tokenizers.Tokenizer.EncodeToTokens(ReadOnlySpan`1 text, String& normalizedText, Boolean considerPreTokenization, Boolean considerNormalization) in D:\a\_work\1\s\src\Microsoft.ML.Tokenizers\Tokenizer.cs:line 166
   at Microsoft.ML.Tokenizers.Tests.EnglishRobertaTests.TestTokenizerEncoding(String text, String[] expectedTokens, ValueTuple`2[] expectedOffsets, Int32[] expectedIds) in D:\a\_work\1\s\test\Microsoft.ML.Tokenizers.Tests\EnglishRobertaTests.cs:line 179

Known issue validation

Build: :mag_right: https://dev.azure.com/dnceng-public/public/_build/results?buildId=1599215 Error message validated: [EnglishRobertaTests.TestTokenizerEncoding ArgumentOutOfRangeException : value must be non-negative] Result validation: :white_check_mark: Known issue matched with the provided build. Validation performed at: 9/16/2026 2:17:25 PM UTC

Report

Build	Repository	Test	Pull Request
1599215	dotnet/machinelearning	Microsoft.ML.Tokenizers.Tests.EnglishRobertaTests.TestTokenizerEncoding	dotnet/machinelearning#7704

Summary

24-Hour Hit Count	7-Day Hit Count	1-Month Count
0	1	1

Source: dotnet/machinelearning

View original on GitHub View discussion on GitHub