Sable proves bit-exact Qwen inference on one GPU
On one RTX PRO 6000, Sable proves Qwen2.5-32B in 92.5 seconds for 4,096 tokens, about 6.4 cents per request.
1 stories, newest first · one card per story
On one RTX PRO 6000, Sable proves Qwen2.5-32B in 92.5 seconds for 4,096 tokens, about 6.4 cents per request.
Today 680 readers · 799 views · 262 reads · all time 187K views