GLM / GLM-5.3-Flash
GLM-5.3-Flash Distillation
Estimated until measured during engineering
Before → After
Lancamento baseCheckpoint verificado
320.000B
320.000B
DistillationProfessor / Aluno
Quality retention target
Quality retention target
Artefacto de producaoArtefacto
Relatorio de linhagem e benchmark
Relatorio de linhagem e benchmark
Destino de GPU
Pesos e cache KVEstimated until measured during engineering
Estimated VRAMMeasured and certified results are delivered with the project.
Empacotamento de runtimetransformers, vllm, sglang, tokenspeed, ktransformers