Phi-3.5-MoE-instruct is Microsoft's 16x3.8B MoE with 6.6B active parameters with 128K context and MIT-licensed weights.
Phi-3.5-MoE-instruct is Microsoft's 16x3.8B MoE with 6.6B active parameters with 128K context and MIT-licensed weights.
Phi-3.5-MoE-instruct is a Microsoft Phi-3.5 checkpoint released in August 2024. It is a 16×3.8B MoE with 6.6B active parameters.
Microsoft documents a 128K context window, MIT licensing and a static knowledge cutoff of October 2023.
It is best suited to multilingual reasoning|coding|long-context assistants.
It is a static model, can produce inaccurate outputs and should be evaluated against newer Phi models before a new production deployment.
Summarize this long technical document.Explain this code and identify bugs.Answer this multilingual question with a concise explanation.MIT-licensed downloadable checkpoint; no fixed Microsoft per-token price applies to the model weights.
A well-documented Microsoft checkpoint for multilingual reasoning|coding|long-context assistants, especially when permissive licensing and self-hosting matter.
G2Gemma 2 2B IT is Google's compact instruction-tuned Gemma 2 checkpoint for text generation, summarization and question answering,…
L3Llama 3.2 3B Instruct is Meta's compact multilingual instruction model with a 128K context window and December 2023…
P3Phi-3.5-mini-instruct is Microsoft's 3.8B dense model with 128K context and MIT-licensed weights.
P3Phi-3 is Microsoft's open small-model family spanning Mini, Small, Medium and Vision variants for local, cloud and edge…