dffarr/gemma-3-27b-pt-orbax

Gemma 3 27B PT – Orbax/TensorStore (OCDBT)

This repository contains the Orbax/TensorStore (OCDBT) checkpoint converted from google/gemma-3-27b-pt Hugging Face safetensors. The conversion stacks transformer layer weights along the depth axis for efficient JAX/Orbax loading.

Conversion script can be found at on GitHub: https://github.com/dominik3141/hf-to-orbax

What’s included

• Orbax/TensorStore checkpoint files (OCDBT)
• LICENSE and NOTICE per Gemma Terms of Use

Conversion details

• Source: google/gemma-3-27b-pt
• Format change only: no weight changes beyond layout/format
• Layer weights use the same HF key with the numeric layer index removed and are stacked on axis 0. Example: model.layers.12.self_attn.q_proj.weight → model.layers.self_attn.q_proj.weight

Loading example (JAX/Orbax)

from huggingface_hub import snapshot_download
import orbax.checkpoint as ocp

path = snapshot_download("dffarr/gemma-3-27b-pt-orbax")
ckpt = ocp.StandardCheckpointer().restore(path)

# Example access
embed_tokens = ckpt["model.embed_tokens.weight"]
q_proj = ckpt["model.layers.self_attn.q_proj.weight"]  # stacked: [num_layers, ...]

License & Use

This repository is provided under the Gemma Terms of Use. Please read LICENSE and comply with the Gemma Prohibited Use Policy: https://ai.google.dev/gemma/terms

Conversion script

The conversion script used to generate this checkpoint is included as convert.py.