Public KG record
LFM2
Recorded overview
LFM2.5 family model with DSpark draft checkpoint supporting speculative decoding
An evidence-backed entity from the project knowledge graph; detailed tool metadata is still being added.
Recorded details
- Benchmarks: 22.6% on IFStruct benchmark
- Benchmarks: 29.7% on IFStruct benchmark
- Benchmarks: AIME25
- Benchmarks: GPQA Diamond
- Benchmarks: GSM8K
- Benchmarks: IFStruct
- Benchmarks: MMLU-Pro
- Capabilities: 提供预训练语言能力并生成流畅文本
Evidence coverage
The KG currently records 12 known information dimensions and has 7 undocumented dimensions, linked to 25 public source articles. These counts describe record coverage, not model performance or verified fit.
Related public records
- Liquid AI — develops
- nvidia/Nemotron-RL-instruction_following-structured_outputs — trained on
- GRPO — applied to
- AIME25 — evaluated on
- GPQA Diamond — evaluated on