From 03c817412c638c905dea6c1a1967691f3ced57b8 Mon Sep 17 00:00:00 2001
From: Marcin Chrzanowski <m@m-chrzan.xyz>
Date: Thu, 27 May 2021 19:59:03 +0200
Subject: Add configurable run script

---
 experiment/__init__.py   |  0
 experiment/__main__.py   | 17 +++++++++++++++++
 experiment/experiment.py | 32 ++++++++++++++++++++++++++++++++
 util/parse_config.py     | 31 +++++++++++++++++++++++++++++++
 4 files changed, 80 insertions(+)
 create mode 100644 experiment/__init__.py
 create mode 100644 experiment/__main__.py
 create mode 100644 experiment/experiment.py
 create mode 100644 util/parse_config.py

diff --git a/experiment/__init__.py b/experiment/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/experiment/__main__.py b/experiment/__main__.py
new file mode 100644
index 0000000..bf4de94
--- /dev/null
+++ b/experiment/__main__.py
@@ -0,0 +1,17 @@
+import sys
+
+import torch
+
+from experiment.experiment import Experiment
+
+file = None
+if len(sys.argv) < 2:
+    print('Provide YAML configuration file as argument')
+    exit(1)
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+file = sys.argv[1]
+prefix = sys.argv[2] if len(sys.argv) > 2 else ''
+
+experiment = Experiment(file, prefix, device)
+experiment.run()
diff --git a/experiment/experiment.py b/experiment/experiment.py
new file mode 100644
index 0000000..600732f
--- /dev/null
+++ b/experiment/experiment.py
@@ -0,0 +1,32 @@
+import subprocess
+import os
+import time
+
+from train.train import train_model
+from util.parse_config import parse_file
+from model.encoder import EncoderModel
+
+class Experiment:
+    def __init__(self, file, prefix, device):
+        self.file = file
+        self.device = device
+        self.make_dir(prefix)
+        self.copy_config(file)
+
+    def run(self):
+        model_config, train_config = parse_file(self.file)
+        model = EncoderModel(device=self.device, **model_config).to(self.device)
+        train_model(model, device=self.device, **train_config)
+
+    def dir_path(self, file):
+        return '{}/{}'.format(self.dirname, file)
+
+    def make_dir(self, prefix):
+        time_string = time.strftime('%Y%m%d%H%M%S')
+        prefix = '' if prefix == '' else '{}-'.format(prefix)
+        dirname = 'outputs/{}{}'.format(prefix, time_string)
+        self.dirname = dirname
+        os.mkdir(dirname)
+
+    def copy_config(self, file):
+        subprocess.run(['cp', file, '{}/config.yaml'.format(self.dirname)])
diff --git a/util/parse_config.py b/util/parse_config.py
new file mode 100644
index 0000000..00a4fd4
--- /dev/null
+++ b/util/parse_config.py
@@ -0,0 +1,31 @@
+import yaml
+
+def parse_file(file):
+    with open(file) as file:
+        return parse_yaml(yaml.safe_load(file))
+
+def parse_model_config(yaml):
+    required_verbatim_params = ['hidden_dim', 'd_ff', 'n_layers', 'num_heads']
+    config = { key: yaml[key] for key in required_verbatim_params }
+
+    config['input_dim'] = yaml['n_tokens']
+    config['output_dim'] = yaml['max_count'] + 1
+
+    if 'use_positional' in yaml:
+        config['use_positional'] = yaml['use_positional']
+    if 'use_feedforward' in yaml:
+        config['use_feedforward'] = yaml['use_feedforward']
+
+    return config
+
+def parse_train_config(yaml):
+    required_verbatim_params = [
+        'lr', 'num_steps', 'batch_size', 'n_tokens', 'seqlen', 'max_count'
+    ]
+    config = { key: yaml[key] for key in required_verbatim_params }
+    return config
+
+def parse_yaml(yaml):
+    model_config = parse_model_config(yaml)
+    train_config = parse_train_config(yaml)
+    return model_config, train_config
-- 
cgit v1.2.3