Update 2021-3-4-pytorch-1.8-released.md

Author: jspisak

_posts/2021-3-4-pytorch-1.8-released.md

@@ -33,7 +33,7 @@ Because FX transforms consume and produce nn.Module instances, they can be used
 
 You can read more about FX in the official [documentation](https://pytorch.org/docs/master/fx.html). You can also find several examples of program transformations implemented using ```torch.fx``` [here](https://github.com/pytorch/examples/tree/master/fx). We are constantly improving FX and invite you to share any feedback you have about the toolkit on the [forums](https://discuss.pytorch.org/) or [issue tracker](https://github.com/pytorch/pytorch/issues).
 
-We’d like to acknowledge [TorchScript](https://pytorch.org/docs/stable/jit.html) tracing, [Apache MXNet](https://mxnet.apache.org/versions/1.7.0/) hybridize, and more recently [JAX](https://github.com/google/jax) as influences for program acquisition via tracing. We’d also like to acknowledge [Caffe2](https://github.com/facebookarchive/caffe2), [JAX](https://github.com/google/jax), and [TensorFlow](https://www.tensorflow.org/) as inspiration for the value of simple, directed dataflow graph program representations and transformations over those representations. 
+We’d like to acknowledge [TorchScript](https://pytorch.org/docs/stable/jit.html) tracing, [Apache MXNet](https://mxnet.apache.org/versions/1.7.0/) hybridize, and more recently [JAX](https://github.com/google/jax) as influences for program acquisition via tracing. We’d also like to acknowledge [Caffe2](https://caffe2.ai/), [JAX](https://github.com/google/jax), and [TensorFlow](https://www.tensorflow.org/) as inspiration for the value of simple, directed dataflow graph program representations and transformations over those representations. 
 
 # Distributed Training
 The PyTorch 1.8 release added a number of new features as well as improvements to reliability and usability. Concretely, support for: [Stable level async error/timeout handling](https://pytorch.org/docs/stable/distributed.html?highlight=init_process_group#torch.distributed.init_process_group) was added to improve NCCL reliability; and stable support for [RPC based profiling](https://pytorch.org/docs/stable/rpc.html). Additionally, we have added support for pipeline parallelism as well as gradient compression through the use of communication hooks in DDP. Details are below: