From 5ce8f41398c0c9dab2134fb3b04ae3c61d62300d Mon Sep 17 00:00:00 2001 From: Jahir Islam Date: Tue, 6 Nov 2018 00:56:41 +0600 Subject: [PATCH] Updated making mmap-able model part of README.md In the making mmap-able model part of the REAMDE.md file, there is no mention of providing `convert_graphdef_memmapped_format` as the artifact name for `taskcluster.py` file. Without this argument `taskcluster.py` file tries to download the default `native_client.tar.xz` artifact and the code receives a 404 error. --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 9ed85feb..e10dd094 100644 --- a/README.md +++ b/README.md @@ -327,7 +327,7 @@ If you want to experiment with the TF Lite engine, you need to export a model th The `output_graph.pb` model file generated in the above step will be loaded in memory to be dealt with when running inference. This will result in extra loading time and memory consumption. One way to avoid this is to directly read data from the disk. -TensorFlow has tooling to achieve this: it requires building the target `//tensorflow/contrib/util:convert_graphdef_memmapped_format` (binaries are produced by our TaskCluster for some systems including Linux/amd64 and macOS/amd64), use `util/taskcluster.py` tool to download, specifying `tensorflow` as a source. +TensorFlow has tooling to achieve this: it requires building the target `//tensorflow/contrib/util:convert_graphdef_memmapped_format` (binaries are produced by our TaskCluster for some systems including Linux/amd64 and macOS/amd64), use `util/taskcluster.py` tool to download, specifying `tensorflow` as a source and `convert_graphdef_memmapped_format` as artifact. Producing a mmap-able model is as simple as: ``` $ convert_graphdef_memmapped_format --in_graph=output_graph.pb --out_graph=output_graph.pbmm