Loggetta

Plan the run. Train the model. Keep the evidence.

Loggetta checks your hardware, chooses a supported configuration for a Mixture-of-Experts (MoE) model, and runs QLoRA fine-tuning. It saves the plan and a JSON run report with memory use, timing, and checks that the selected optimizations actually ran. One install includes the runtime (experts4bit-qlora: https://pypi.org/project/experts4bit-qlora/) and the GPU kernels (grouped-nf4-gemm: https://pypi.org/project/grouped-nf4-gemm/).

Limits. Plans are estimates, not an out-of-memory guarantee. The released training path is single-GPU MoE training on Linux with a supported NVIDIA CUDA GPU. Dense models are planned but not yet supported for training. Dense plans assume identical decoder layers, and a model whose layers differ is refused. For a config that declares head_dim per layer, such as gemma4, the refusal names each difference (the per-layer head_dim, layers that reuse another layer's KV, linears whose shapes change). Dense training: https://github.com/pjordanandrsn/loggetta/blob/main/docs/DENSE.md

pip install loggetta
loggetta inspect
loggetta plan Qwen/Qwen3-30B-A3B --seq 2048

The plan shows where weights will live, estimated memory use, and why alternatives were rejected. It checks the budget before downloading model weights. Estimates can miss; they are not an out-of-memory guarantee.

New in 0.6.1

• gemma4 is refused with its reasons instead of failing. On transformers 5.20, describing google/gemma-4-E2B-it raised; plan now refuses it by name, listing each layer difference the planner cannot price yet.
• A model with no MoE layers is refused by the MoE backend at admission, so the dense backend owns it when it can describe it, and otherwise the plan gives each backend's own reason.

New in 0.6.0

• An AMD GPU is found and refused by name. plan says "AMD GPU detected (…); not yet supported" instead of "no GPU", and inspect shows the gfx arch.
• The grouped_nf4 MoE backward is priced once. With experts4bit-qlora 0.53.0 or later, the runtime's estimate prices it, and Loggetta adds no line of its own.
• Dense training's frozen-parameter digest takes a rank-0 parameter; every existing digest is unchanged.
• A held-out check on the RTX 5090 found one setup over its plan; every over-plan run used a setting the planner does not model (pad buckets off, or grouped-nf4-gemm before 0.42.0) (HO1, with addendum: https://github.com/pjordanandrsn/loggetta/blob/main/evidence/2026-10-10-heldout-cards/README.md).

Full list, the CHANGELOG: https://github.com/pjordanandrsn/loggetta/blob/main/CHANGELOG.md.

Train on your data

loggetta train Qwen/Qwen3-30B-A3B \
  --dataset ./data/train.jsonl --format text \
  --seq 512 --micro-batch 1 --steps 20 --seed 42 \
  --out runs/my-training --adapter-out adapters/my-adapter

Local JSONL, JSON, CSV, Parquet and TXT files, Hub datasets, Alpaca instructions and text-only chats are supported. Data is validated and tokenized before weights load. Chat data trains only the assistant turns by default. Reload the adapter with loggetta.load_adapter("adapters/my-adapter"). See the training guide: https://github.com/pjordanandrsn/loggetta/blob/main/docs/TRAINING.md.

To run a saved decision later: loggetta plan MODEL --out plan.json, then loggetta execute plan.json --out runs/. Pass earlier run reports back with --observations runs/ and later plans use those measurements.

Measured results

The included runtime and kernels do the compute; these are matched training runs, not planner benchmarks.

The comparison used torch 2.12.1+cu130 and transformers 5.5.0 for both frameworks, with matched adapters, initialization and tokens. Loggetta does not predict throughput.

Models

Every row is supported by the included runtime's QLoRA path. Hybrid models (Qwen3.6, Granite-4.0-H, LFM2, Nemotron-H) need --packing concat with chat or Alpaca data. The runtime's capability register is the authority: https://github.com/pjordanandrsn/experts4bit-qlora/blob/main/docs/capabilities.json

Scope

Single-GPU MoE planning and QLoRA training; serving placement is planned, not launched. Not yet: multi-GPU, throughput prediction. GPU runs need Linux, an NVIDIA CUDA GPU and a compatible PyTorch. Pre-1.0.

Dense models are not supported for training yet: their plans are estimates that missed a held-out check (a further reading is pending), and their training runs only behind --allow-development-executor until a 24 GB capacity reading.

GitHub: https://github.com/pjordanandrsn/loggetta · Results: https://github.com/pjordanandrsn/loggetta/blob/main/docs/RESULTS.md · Architecture: https://github.com/pjordanandrsn/loggetta/blob/main/docs/ARCHITECTURE.md · Research and releases: https://cerinamroth.com/ml/