Numbat

The 772 shards total ~7.72 TB of tensor data. Each shard is safetensors format, under 10.6 GB per file. configurationnumbat.py and modelingnumbat.py are included: a complete reference implementation of the checkpoint. Its module tree has an exact slot for every tensor in every shard, so the full checkpoint loads with loadstatedict(strict=True) (and, since the classes derive from PreTrainedModel, also via AutoModelForCausalLM.frompretrained(..., trustremotecode=True)). NumbatForCausalLM.forward runs the trunk and NumbatForCausalLM.draft runs the MTP layer.