StargazerLabs models

StargazerLabs is where I publish open-weight models for Apple silicon, in bf16 and 3-, 4-, and 8-bit quantizations. Seven are up, and together they are downloaded about three thousand times a month.

Jumbo

Continued training tends to overwrite some of what a model knew. Qwen3.8-32B-Jumbo is a 76-layer model that puts some of it back. A per-tensor cosine alignment census between two sibling checkpoints finds the layer blocks that drifted most in training, and those blocks are transplanted from the sibling into the base. The transplant keeps the model compatible with its multi-token-prediction drafter, at 82 to 89 percent acceptance, which gives 1.6 to 2.2 times faster speculative decoding. Mini-Me is a sheared 23-billion-parameter variant from the same family.

The toolkit, and what did not ship

The LoRA post-training toolkit behind the releases does knowledge injection with sparse top-k logit distillation against a teacher bank, layer-gated adapters, region-weighted losses, and PEFT-format adapter export so a fine-tune can be evaluated in other frameworks. Every run also scores an out-of-domain regression set, a “shadow basket” of foreign-language text, contracts, recipes, and dialogue that the fine-tune was never meant to change. That set is there to catch catastrophic forgetting, and it did. The fine-tuned variants regressed on held-out evaluation, so only the surgery-only models were released.

Stack

MLX, Hugging Face Transformers and PEFT, custom census and transplant scripts, quantization and speculative-decoding evaluation harnesses.