laion/bugs-r2egym-stackseq
laion/bugs-r2egym-stackseq is an 8 billion parameter language model fine-tuned from Qwen/Qwen3-8B. It is specifically trained on datasets related to inferred bugs, R2EGYM, and StackExchange overflow sandboxes. This model is optimized for tasks involving code analysis, bug detection, and understanding programming-related discussions.
Loading preview...
Model Overview
laion/bugs-r2egym-stackseq is an 8 billion parameter language model derived from the Qwen/Qwen3-8B architecture. This model has undergone specialized fine-tuning to enhance its performance on specific programming-related tasks.
Key Capabilities
This model's training regimen suggests a focus on:
- Bug Detection and Analysis: Fine-tuned on the
penfever/GLM-4.6-inferredbugs-32eps-65kdataset, indicating proficiency in identifying and understanding software bugs. - Programming Environment Interaction: Training with the
penfever/glm-4.6-r2egym-32ep-32kdataset implies capabilities related to interacting with or analyzing programming environments. - Code-related Q&A: The inclusion of the
penfever/GLM-4.6-stackexchange-overflow-sandboxes-32eps-65kdataset suggests an ability to process and generate content relevant to programming questions and answers, similar to those found on platforms like Stack Overflow.
Training Details
The model was trained with a learning rate of 4e-05 over 7 epochs, utilizing a total batch size of 16 across 8 GPUs. The optimizer used was ADAMW_TORCH_FUSED with a cosine learning rate scheduler and a warmup ratio of 0.1.