From 63de39d21c904395b3536eb6f2c7d089985a7195 Mon Sep 17 00:00:00 2001 From: Tomasz Pietruszka Date: Tue, 25 Feb 2020 13:32:16 +0100 Subject: [PATCH 1/3] readme install - no longer specifying tensorflow-probability-gpu --- README.md | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/README.md b/README.md index d1f82b3..9f5ec67 100644 --- a/README.md +++ b/README.md @@ -72,11 +72,7 @@ For example: conda create -n tfp3.7 python=3.7 numpy scipy conda activate tfp3.7 - conda install cudatoolkit - conda install cudnn - conda install tensorflow-gpu=1.15 - conda install tensorflow-probability-gpu=1.15 - conda install tensorflow-probability + conda install cudatoolkit cudnn tensorflow-gpu=1.15 tensorflow-probability pip install -e From 2321e2dc48736dc71352268adbeb5ca3e3f5b5be Mon Sep 17 00:00:00 2001 From: Tomasz Pietruszka Date: Tue, 25 Feb 2020 13:41:32 +0100 Subject: [PATCH 2/3] fix: readme links --- README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index 9f5ec67..7ad0707 100644 --- a/README.md +++ b/README.md @@ -6,18 +6,18 @@ was used in the following papers (alse see [citations](#citations)): - [On the state of the art of evaluation in neural language models](https://arxiv.org/abs/1707.05589) - See [./experiment/on-the-state/README.md](./experiment/on-the-state/README.md) + See [./lamb/experiment/on-the-state/README.md](./lamb/experiment/on-the-state/README.md) for more. - [Pushing the bounds of dropout](https://arxiv.org/abs/1805.09208) See - [./experiment/pushing-the-bounds/README.md](./experiment/pushing-the-bounds/README.md) + [./lamb/experiment/pushing-the-bounds/README.md](./lamb/experiment/pushing-the-bounds/README.md) for more. - [Mogrifier LSTM](https://arxiv.org/abs/1909.01792) - See [./experiment/mogrifier/README.md](./experiment/mogrifier/README.md) for + See [./lamb/experiment/mogrifier/README.md](./lamb/experiment/mogrifier/README.md) for more. From 46efefcb54e09433238be588fa75e530c9e7f4bf Mon Sep 17 00:00:00 2001 From: Tomasz Pietruszka Date: Tue, 25 Feb 2020 13:57:21 +0000 Subject: [PATCH 3/3] Link fixes in paper-specific readmes --- lamb/experiment/mogrifier/README.md | 2 +- lamb/experiment/on-the-state/README.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/lamb/experiment/mogrifier/README.md b/lamb/experiment/mogrifier/README.md index 178bea8..8b6ecb9 100644 --- a/lamb/experiment/mogrifier/README.md +++ b/lamb/experiment/mogrifier/README.md @@ -2,7 +2,7 @@ This directory contains saved configuration files for tuned models from the [Mogrifier LSTM](https://arxiv.org/abs/1909.01792) paper. Model weights are not included. -Don't forget to [set up the data](../../README.md). +Don't forget to [set up the data](../../../README.md). For example, to train a Mogrifier LSTM with 24M parameters on PTB with tuned hyperparameters (see the paper above): diff --git a/lamb/experiment/on-the-state/README.md b/lamb/experiment/on-the-state/README.md index 1aa71ae..00bed67 100644 --- a/lamb/experiment/on-the-state/README.md +++ b/lamb/experiment/on-the-state/README.md @@ -2,7 +2,7 @@ This directory contains saved configuration files for tuned models from the [On the state of the art of evaluation in neural language models](https://arxiv.org/abs/1707.05589) paper. Model weights are not included. -Don't forget to [set up the data](../../README.md). +Don't forget to [set up the data](../../../README.md). To train the 1 layer LSTM model of 10m weights on PTB with tuned hyperparameters (see the paper above):