dvmazur

4 ms·

1. 1.▲
   by dvmazur 3mo ago

   AsyncResoning[1] does a trick of that sort to give agents concurrent cache views. You basically have two agents look at the same cache under different views. Say agent_0 gets [a_1, a_0] and agent_1 gets [a_0, a_1]. They also write to this c
2. 2.▲
   by dvmazur 11mo ago

   GitHub Pages are affected too
3. 3.▲
   by dvmazur 3y ago

   You should check out Hivemind[1]. It is very similar to what you described except it used MoE for "fragmentation". They have a couple of examples of pre-training in their repo. Hivemind was also used to build Petals[2] but it only
4. 4.▲
   C++ Attribute: Likely, Unlikely(en.cppreference.com)

   88 points by dvmazur 3y ago|63 comments