1 build of LFM2.5 1.2B measured on real machines — different packagers, precisions and engines, each a different file with its own numbers. The best each engine proved is first; every build has its own page with the machines and signatures behind it.
trained 125K · proved 4K context proved by unsloth Q4_K_M
16.1 tok/s peak, by unsloth Q4_K_M
proved by at least one build: Tool calling · Tool loop finished · Faithful to tool results · Chained tool calls · Structured output · Sustained output · Writes code · Reasons about code
| build | publisher | precision | engine | whose work | context proved | best tok/s | machines |
|---|---|---|---|---|---|---|---|
| hf.co/unsloth/LFM2.5-1.2B-Instruct-GGUF:Q4_K_M | unsloth | Q4_K_M | ollama | quantizer | 4,096 | 16.1 | 1 |
Fine-tunes, merges and abliterations of this member, each its own model with its own page. None of their numbers are counted above.