A benchmark dataset capturing test results for agentic coding tasks performed by local LLMs on consumer hardware. The dataset likely contains performance metrics from tests run on an NVIDIA RTX 4060 Ti 8GB GPU and Intel i7-14700F CPU system. It was created by witcheer and uploaded to Hugging Face in May 2026.
Use Cases
- Benchmarking local LLM performance on agentic coding tasks based on the described hardware specifications
- Comparing the effectiveness of different agent frameworks like Hermes Agent based on the test results
- Evaluating the feasibility of multi-turn tool calling and debugging on limited VRAM based on the dataset's purpose
Strengths
- Test results are based on a specific hardware configuration (NVIDIA RTX 4060 Ti 8GB, Intel i7-14700F)
- Focuses on a concrete, emerging application area: agentic coding with local LLMs
Limitations
- Column-level documentation is absent; field semantics must be inferred after download
- Row count is unknown, which may limit suitability assessment
Provenance
- Source
- witcheer
- Collection Method
- Likely contains real test results from running agentic coding benchmarks.
- Freshness
- Last updated 2026-05-21 10:48:08; freshness should be verified