7b CodeLlama model finetuned on HQ + self-play augmented data to optimize C++ programs. For more details on how the data was collected and training was done, refer to this webpage.

Downloads last month
7
Safetensors
Model size
6.74B params
Tensor type
BF16
·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.

Model tree for LearningOpt/pie-hq-selfplay-7b

Quantizations
2 models