Manufacturer

microsoft

Manufacturer profile and assigned models.

🤗 Hugging Face Filter in leaderboard

1Models

6Benchmarks

6Performance

0Harness

Best token generation

302,0 tok/s

Phi-3.5-mini-instruct

Best prefill

19.405,8 tok/s

Phi-3.5-mini-instruct

∅ Token generation

179,2 tok/s

Average of 6 runs

∅ Prefill

8.523,4 tok/s

Average of 6 runs

Best TTFT

380 ms

Phi-3.5-mini-instruct

Best harness rate

no harness run yet

🏆

Phi-3.5-mini-instruct · fastest model at 302,0 tok/s token generation
Best prefill of this manufacturer: 19.405,8 tok/s (Phi-3.5-mini-instruct)

Top models by token generation

Best measured generation throughput per model (tok/s), published performance benchmarks.

Phi-3.5-mini-instruct

302,0 tok/s

Models

1 assigned

All benchmarks →

Phi-3.5-mini-instruct3.8B6 runs

179,2tok/s Ø

Min 39,3Max 302,0