EasyCommand

EasyCommand generates a Bash command locally, displays it and asks before execution.

The Linux x86_64 bundle embeds llama.cpp and the Bash parser; no separate llama binary, Python inference environment, GPU or network is needed after downloading a compatible model. The bundle README lists exact CPU/system requirements and installation steps. Source builds are available for other supported Linux systems.

Models and teaching data are distributed separately on Hugging Face:

• 1.5B A3 Q4_K_M: 212/300 ALFA-updated and 1119/1320 internal.
• 0.6B A1 Q4_K_M / Q8_0: 165/300 and 174/300 ALFA-updated respectively.
• Training dataset: 401,975 distinct request/answer pairs.

Trainable BF16 checkpoints and original adapters are also available. See the model cards and research post for serving profiles, correlated development-panel limitations and benchmark protocol differences. Verify the binary archive with SHA256SUMS.