diff --git a/nlp/bert/pytorch/README.md b/nlp/bert/pytorch/README.md
index d2e3d9243..bbf00d7ac 100644
--- a/nlp/bert/pytorch/README.md
+++ b/nlp/bert/pytorch/README.md
@@ -340,23 +340,6 @@ In order to re-generate the sample dataset, run the following script:
 python3 third_party/create_pretraining_data.py --input-file data/sample_text.txt --output-file data/sample_text.tfrecord --sequence-length 128 --mask-tokens 20 --duplication-factor 4 --do-lower-case --model bert-base-uncased
 ```
 
-
-### Running and benchmarking
-
-To run a tested and optimised configuration and to reproduce the performance shown on our [performance results page](https://www.graphcore.ai/performance-results), please follow the setup instructions in this README to setup the environment, and then use the `examples_utils` module (installed automatically as part of the environment setup) to run one or more benchmarks. For example:
-
-```python
-python3 -m examples_utils benchmark --spec <path to benchmarks.yml file>
-```
-
-Or to run a specific benchmark in the `benchmarks.yml` file provided:
-
-```python
-python3 -m examples_utils benchmark --spec <path to benchmarks.yml file> --benchmark <name of benchmark>
-```
-
-For more information on using the examples-utils benchmarking module, please refer to [the README](https://github.com/graphcore/examples-utils/blob/master/examples_utils/benchmarks/README.md).
-
 ### Troubleshooting
 
 If Triton server tests fails with such error:
diff --git a/vision/cnns/pytorch/makefile b/vision/cnns/pytorch/makefile
index 2e2a5c189..263cc8cb7 100644
--- a/vision/cnns/pytorch/makefile
+++ b/vision/cnns/pytorch/makefile
@@ -3,9 +3,10 @@ all: install get-data install-turbojpeg
 install-turbojpeg: get-turbojpeg build-turbojpeg
 
 install:
-	pip install -r requirements.txt
-	pip uninstall pillow -y
-	CC="cc -mavx2" pip install --no-cache-dir -U --force-reinstall pillow-simd
+	pip3 install -r requirements.txt
+	pip3 uninstall pillow -y
+	CC="cc -mavx2" pip3 install --no-cache-dir -U --force-reinstall pillow-simd
+	HOROVOD_WITH_PYTORCH=1 pip3 install --no-cache-dir 'horovod==0.24.3'
 
 get-turbojpeg:
 	cd datasets && rm -rf libjpeg-turbo && git clone https://github.com/libjpeg-turbo/libjpeg-turbo.git
@@ -20,5 +21,5 @@ get-data:
 	cd datasets && ./get_images.sh
 
 test:
-	cd tests && pytest -n 10 --forked
-	cd tests_serial && pytest
+	cd tests && python3 -m pytest -n 10 --forked
+	cd tests_serial && python3 -m pytest
diff --git a/vision/cnns/pytorch/requirements.txt b/vision/cnns/pytorch/requirements.txt
index e21f4b456..68d66e14a 100644
--- a/vision/cnns/pytorch/requirements.txt
+++ b/vision/cnns/pytorch/requirements.txt
@@ -13,7 +13,6 @@ timm==0.5.4
 wandb==0.12.8
 pyyaml==5.4.1
 checksumdir==1.2.0
-horovod==0.24.3
 tritonclient[grpc]==2.16.0
 git+https://github.com/lilohuang/PyTurboJPEG.git@8706665787c7290397859075ae2f0bf35afeb41a
 git+https://github.com/graphcore/examples-utils@0f3024827451878955ed495daf2e1b552eaf749f#egg=examples-utils