CodeRM-8B

Model overview

CodeRM-8B is published by KAKA22 as a Llama text-generation model. The captured configuration identifies LlamaForCausalLM with model type llama and Safetensors metadata reports 8,030,261,248 parameters. According to the model card, it is built on Llama3.1-8B-Instruct for Python unit-test generation.

Recorded capabilities

Use cases in the source record

• Python unit-test generation workflows that follow the card's function-format prompt and detailed-comment guidance.
• Code reward-modeling research that reuses the openly released CodeRM-UnitTest data and the publisher's evaluation framing.

Limitations and unknowns

• According to the model card, evaluation figures are publisher-reported reward-setting results; no independent Ethen evaluation was extracted.
• No context-window value was extracted from this record.
• According to the model card, full training detail is deferred to the publisher paper Dynamic Scaling of Unit Tests for Code Reward Modeling.
• Provider state is historical snapshot data, not independently refreshed current availability.

Source and provenance

Source: KAKA22/CodeRM-8B

Captured: Unknown. Processed: 2026-09-07T19:34:32.402121+00:00.

Introduction CodeRM-8B is a small yet powerful model designed to enable efficient and high-quality unit test generation. It is trained on a dataset of 60k high-quality synthetic Python unit tests using Llama3.1-70B-Instruct. These unit tests are synthesized based on two well-regarded code instruction tuning datasets: CodeFeedback-Filtered-Instruction and the training set of TACO . The training dataset used for unit test generation is openly available under CodeRM-UnitTest . For further information and details of training, refer to our paper: "Dynamic Scaling of Unit Tests for Code Reward Modeling" available on arXiv . You can also v…