Model Anatomy See what's inside open-weight AI models

Can I run LFM3-1.7B-GGUF on an NVIDIA H200?

Yes

Its Q8_0 file (1.85 GB) fits in the 127 GB an NVIDIA H200 can use, with room for ≈ 256K tokens of context.

Full analysis of LFM3-1.7B-GGUF Everything that runs on NVIDIA H200 Get the Q8_0 file on Hugging Face
ollama run hf.co/AtomicChat/LFM3-1.7B-GGUF:Q8_0

Files and context on NVIDIA H200

Each file of LFM3-1.7B-GGUF (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.

FileSizeOn NVIDIA H200 Context (16-bit)Context (8-bit)
Q8_01.85 GBFits256K256K
Q6_K1.59 GBFits256K256K
Q5_K_M1.31 GBFits256K256K
Q4_K_M1.13 GBFits256K256K
IQ4_XS1.09 GBFits256K256K
IQ3_S922 MBFits256K256K

LFM3-1.7B-GGUF on other hardware

Popular models that run on NVIDIA H200

Assumes 90% of the card's memory is usable and at least 4K tokens of context for one conversation, with the context cache in 16-bit unless noted. Fits leaves at least 20% of that memory for context; Tight leaves less.