Mark as superseded, link the Kisoku 1.6B release and 0arch.io/kisoku
Browse files
README.md
CHANGED
|
@@ -54,6 +54,8 @@ model-index:
|
|
| 54 |
value: 0.4925
|
| 55 |
---
|
| 56 |
|
|
|
|
|
|
|
| 57 |
# Kisoku 3B Base
|
| 58 |
|
| 59 |
A 3B parameter language model trained **entirely from scratch** on Google Cloud TPUs using [MaxText](https://github.com/AI-Hypercomputer/maxtext) (JAX), supported by [Google's TPU Research Cloud (TRC)](https://sites.research.google/trc/).
|
|
|
|
| 54 |
value: 0.4925
|
| 55 |
---
|
| 56 |
|
| 57 |
+
> **Superseded.** This is Kisoku v1, an earlier 3B model. The current release is [Kisoku 1.6B](https://huggingface.co/0arch-io/kisoku-1.6b). Training log, benchmark charts and run instructions: [0arch.io/kisoku](https://0arch.io/kisoku).
|
| 58 |
+
|
| 59 |
# Kisoku 3B Base
|
| 60 |
|
| 61 |
A 3B parameter language model trained **entirely from scratch** on Google Cloud TPUs using [MaxText](https://github.com/AI-Hypercomputer/maxtext) (JAX), supported by [Google's TPU Research Cloud (TRC)](https://sites.research.google/trc/).
|