Local Code Arena: MBPP Benchmark Results for StarCoder 7B

Raw evaluation metrics, execution telemetry logs, and structural syntax outputs from running the Mostly Basic Python Problems (MBPP) benchmark against the StarCoder 7B base model. The dataset was created by ShahzebKhoso and last updated on May 28, 2026. It is intended to document model behavior within automated benchmarking pipelines.

Use Cases

• Analyzing model performance on basic Python problems based on the MBPP benchmark.
• Studying execution telemetry and runtime behavior of code generation models.
• Comparing foundational model weights using structured syntax outputs from automated pipelines.

Strengths

• Captures multiple data types including raw metrics, telemetry logs, and syntax outputs.
• Focuses on a specific, established benchmark (MBPP) for code generation.
• Documents the behavior of a specific model version (StarCoder 7B base).

Limitations

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

Provenance

Source

huggingface

Collection Method

Execution of the MBPP benchmark against the StarCoder 7B model.

Freshness

Last updated 2026-05-28 11:48:54; freshness should be verified.