Personal Claude Code Conversation Logs with 549 Sessions

549 sessions of conversation logs between a user and various Claude AI models for coding tasks, exported with DataClaw. The dataset includes 15.1 billion input tokens and 4.6 million output tokens, with the most recent update in February 2026. It was authored by misterkerns and hosted on Hugging Face.

Use Cases

• Analyze patterns in human-AI coding interactions based on the conversation logs.
• Study the distribution of model usage and token consumption for different coding tasks.
• Train or fine-tune models for code generation based on the documented assistant interactions.

Strengths

• 549 recorded sessions provide a substantial interaction volume.
• 15.1 billion input tokens indicate a large-scale conversational corpus.
• Includes metadata on model usage across five different Claude versions.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect the personal usage bias inherent to a single user's logs.

Provenance

Source

misterkerns (personal user)

Collection Method

Exported with DataClaw from personal interactions with Claude AI.

Time Range

Temporal coverage is not specified.

Freshness

Last updated 2026-02-25 06:33:34; freshness should be verified.

Geography

Spatial coverage is not specified.