This model checkpoint is not finished yet. We're still polishing the model generalization.

luau-coder-preview-28B-A3B

This model is build on top of REAPED Qwen3.6-35B-A3B model with Continual Pretraining (CPT), it has 205 total experts with 8 active experts.

Uploaded finetuned model

  • Developed by: khtsly
  • License: apache-2.0
  • Finetuned from model : 0xSero/Qwen3.6-28B-REAP

This qwen3_5_moe_text model was trained 2x faster with Unsloth and Huggingface's TRL library.

Downloads last month
226
Safetensors
Model size
28B params
Tensor type
BF16
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for khtsly/luau-coder-preview-28B-A3B-noft

Finetuned
(1)
this model

Datasets used to train khtsly/luau-coder-preview-28B-A3B-noft