7 PhD Students Train Open-Source 7B Model ZGCM-1 in One Summer

ZGCM-1 is a fully open-source 7B dense model trained from scratch over one summer by 7 PhD students, targeting math reasoning and agentic search with 256K context. The team released all training data, recipes, code, checkpoints and logs.

2026-09-14 ~ 2026-09-15 · 3 related posts

• 7 PhD students train a 7B model from scratch in one summer using hundreds of agents — 量子位 · 2026-09-14
• ZGCM-1: A fully open 7B foundation model for math and agentic search — zgcagi · 2026-09-15
• ZGCM-1: fully open 7B model matches much larger frontier models on math and agentic search — _reachsumit · 2026-09-15