ridger's picture
Update README.md
4039295
|
raw
history blame
272 Bytes
metadata
license: apache-2.0
datasets:
  - the_pile_openwebtext2
language:
  - en

Model Description

SpikeGPT-OpenWebText-216M is a L18-D768 SpikeGPT model trained on OpenWebText. See https://github.com/ridgerchu/SpikeGPT for details.

ctx_len = 1024 n_layer = 18 n_embd = 768