Langmap-TheStack-csharp-100M: C# Code Finetuning Dataset

LangMap-TheStack-csharp-100M is a dataset containing 100 million tokens of C# source code, intended for code finetuning. The data was streamed from the bigcode/the-stack repository and tokenized using the allenai/OLMo-3-1025-7B tokenizer. It was created by MultilingualUnigramLM and last updated on Hugging Face on April 13, 2026.

Use Cases

• Fine-tuning code generation models based on the 100 million tokens of C# source code.
• Training or evaluating models for code completion tasks using sanitized source code examples.
• Studying programming language patterns and syntax specific to C# from a large-scale code corpus.

Strengths

• Contains 100,000,000 tokens of C# code, meeting its stated target size.
• Source code is described as sanitized, which may indicate preprocessing for quality or safety.
• Uses a specific, documented tokenizer (allenai/OLMo-3-1025-7B) for consistency.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and license information are unknown, which may limit suitability assessment.

Provenance

Source

Streamed from the bigcode/the-stack repository on Hugging Face.

Collection Method

Code was collected and tokenized for finetuning purposes.

Freshness

Last updated 2026-04-13 19:17:14; freshness should be verified.