AmberYifan/capsd-marin-8b-base-code_qurating_b12000_s0

TEXT GENERATIONPricing:Input $0.37 / Cached $0.074 / Output $0.38Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:8kTool Calling:SupportedPublished:Aug 5, 2026License:otherArchitecture:Transformer Featherless Exclusive Cold

The AmberYifan/capsd-marin-8b-base-code_qurating_b12000_s0 model is an 8 billion parameter language model, fine-tuned from marin-community/marin-8b-base. This model is specifically adapted using the capsd_marin-8b-base-n80000-opc__mix_code_qurating_b12000_s0 dataset, suggesting an optimization for code-related tasks. With an 8192 token context length, it is designed for applications requiring processing of substantial code snippets or technical documentation. Its fine-tuning on a specialized dataset indicates a focus on improving performance in specific coding or technical domains.

Loading preview...

Overview

This model, AmberYifan/capsd-marin-8b-base-code_qurating_b12000_s0, is an 8 billion parameter language model derived from the marin-community/marin-8b-base architecture. It has been fine-tuned on a specific dataset, capsd_marin-8b-base-n80000-opc__mix_code_qurating_b12000_s0, which implies a specialization towards code-related tasks or technical content processing. The model supports a context length of 8192 tokens.

Training Details

The fine-tuning process involved the following key hyperparameters:

  • Learning Rate: 1e-05
  • Batch Sizes: train_batch_size of 2, eval_batch_size of 8, leading to a total_train_batch_size of 64 and total_eval_batch_size of 32 (with 4 GPUs and 8 gradient accumulation steps).
  • Optimizer: ADAMW_TORCH with default betas and epsilon.
  • LR Scheduler: Cosine type with 0.03 warmup steps.
  • Epochs: Trained for 1 epoch.

Intended Uses

While specific intended uses and limitations are not detailed in the original model card, the fine-tuning on a dataset with "code_qurating" in its name suggests its potential application in:

  • Code generation and completion.
  • Code analysis and understanding.
  • Technical documentation processing.
  • Tasks requiring contextual understanding of programming languages.

Finetunes
1 models