
1 Epoch of specialization with 2 experts Active instead of ALL 3 Experts Active for the MoE Research Experiment now that its performing well on the dataset without completing a single epoch yet.
6b411d0
verified
- SHA256:
- ddba2c3fcd660853ad62bd343cb0f587a260afd34aeeb1190350f800c27410f1
- Pointer size:
- 135 Bytes
- Size of remote file:
- 5 GB
- Xet backed hash:
- a4e3557bc923ed4723171cfc8c8fa3b4adb9ca867d1e051dfce56a4229f98c2e
·
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.