diff --git a/README.md b/README.md
index 4821221..4d35179 100644
--- a/README.md
+++ b/README.md
@@ -3,26 +3,7 @@
 ensemble nets are a single pass way of representing M models in a neural
 net ensemble. see this [blog post](http://matpalm.com/blog/ensemble_nets)
 
+note: blog post was based on a vmap version run on my local single GPU machine.
+the code for this first version is under tag v1.
 
-```
-# to reproduce the base line tuning
-python3 tune_with_ax.py --mode siso
-```
-
-```
-# to reproduce the single_input case
-python3 tune_with_ax.py --mode simo
-```
-
-```
-# to reproduce the multi_input case
-python3 tune_with_ax.py --mode mimo
-```
-
-```
-# to reproduce the single_input case with logit dropout
-python3 tune_with_ax.py --mode simo_ld
-```
-
-see the notebooks under the blog/ folder to reproduce additional
-figures in the blog post
+the code as is now is a port of the model to run on a pod tpu slice using haiku
\ No newline at end of file
diff --git a/requirements.txt b/requirements.txt
index 8193d49..a7f8cfd 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,6 +1,6 @@
 tensorflow-gpu
 tensorflow-datasets
 jax
-objax
+haiku
 wandb
-ax-platform
\ No newline at end of file
+ax-platform