AmberYifan/capsd-marin-8b-base-code_ifd_b16000_s0
The AmberYifan/capsd-marin-8b-base-code_ifd_b16000_s0 model is an 8 billion parameter language model, fine-tuned from marin-community/marin-8b-base. This model was specifically fine-tuned on the capsd_marin-8b-base-n80000-opc__mix_code_ifd_b16000_s0 dataset, indicating an optimization for code-related tasks. It utilizes an 8192 token context length and was trained with a learning rate of 1e-05 over one epoch.
Loading preview...
Model Overview
The AmberYifan/capsd-marin-8b-base-code_ifd_b16000_s0 is an 8 billion parameter language model, derived from the marin-community/marin-8b-base architecture. This model has undergone a specific fine-tuning process, leveraging the capsd_marin-8b-base-n80000-opc__mix_code_ifd_b16000_s0 dataset. The fine-tuning suggests a specialization towards code-related applications or tasks involving mixed code data.
Key Training Details
- Base Model: Fine-tuned from marin-community/marin-8b-base.
- Dataset: Fine-tuned on the
capsd_marin-8b-base-n80000-opc__mix_code_ifd_b16000_s0dataset. - Hyperparameters:
- Learning Rate: 1e-05
- Optimizer: AdamW with betas=(0.9, 0.999) and epsilon=1e-08
- Scheduler: Cosine with 0.03 warmup steps
- Epochs: 1
- Total Batch Size: 64 (across 4 GPUs with 8 gradient accumulation steps)
Potential Use Cases
Given its fine-tuning on a code-centric dataset, this model is likely optimized for tasks such as:
- Code generation
- Code completion
- Code summarization
- Understanding and processing code-related text.