dl-experiment-code 1.0.0__py3-none-any.whl

dl_experiment_code-1.0.0.dist-info/METADATA

@@ -0,0 +1,146 @@
+Metadata-Version: 2.4
+Name: dl-experiment-code
+Version: 1.0.0
+Summary: Ready-to-use Deep Learning experiment code for experiments 5-10 and a NumPy RNN from scratch
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+
+# Deep Learning Experiment Code Package
+
+This package provides ready-to-use code for:
+
+- Experiment 5: CNN image classification
+- Experiment 6: CNN hyperparameter tuning
+- Experiment 7: Autoencoder dimensionality reduction
+- Experiment 8: Educational single-object detection
+- Experiment 9: LSTM sentiment analysis
+- Experiment 10: LSTM hyperparameter tuning
+- Additional: Vanilla RNN implemented from scratch with NumPy
+
+For experiments 5-10 there are two versions:
+
+1. `local` - uses a local CSV/image/Pascal VOC dataset.
+2. `online` - uses a Keras dataset or an online TensorFlow Datasets dataset.
+
+The code follows the structure and concepts of the supplied original experiments.
+
+## Installation
+
+From this folder:
+
+```bash
+pip install .
+```
+
+For development/editable installation:
+
+```bash
+pip install -e .
+```
+
+## List experiments
+
+```bash
+dlab-code list
+```
+
+## Get code
+
+Online/Keras dataset version:
+
+```bash
+dlab-code get 5 --dataset online
+```
+
+Local dataset version:
+
+```bash
+dlab-code get 5 --dataset local
+```
+
+Experiments 6-10 work the same way:
+
+```bash
+dlab-code get 6 --dataset online
+dlab-code get 7 --dataset local
+dlab-code get 8 --dataset online
+dlab-code get 9 --dataset local
+dlab-code get 10 --dataset online
+```
+
+## Save directly to a Python file
+
+```bash
+dlab-code get 5 --dataset online --save exp5.py
+```
+
+Then:
+
+```bash
+python exp5.py
+```
+
+## RNN from scratch
+
+```bash
+dlab-code get rnn-scratch
+```
+
+Or save it:
+
+```bash
+dlab-code get rnn-scratch --save rnn_from_scratch.py
+```
+
+The RNN uses NumPy only for the recurrent computation and manually implements:
+
+- hidden-state calculation
+- output calculation
+- loss
+- backpropagation through time
+- gradient clipping
+- parameter updates
+
+## Dataset notes
+
+### Experiment 5
+- Online: Keras CIFAR-10
+- Local: image directory with one folder per class
+
+### Experiment 6
+- Online: Keras CIFAR-10
+- Local: image directory with one folder per class
+
+### Experiment 7
+- Online: Keras MNIST
+- Local: numeric CSV
+
+### Experiment 8
+- Online: Pascal VOC through TensorFlow Datasets
+- Local: Pascal VOC-style images + XML annotations
+
+The object detector is deliberately educational and predicts the first object in an image. It is not a production multi-object detector.
+
+### Experiment 9
+- Online: Keras IMDB
+- Local: `text,label` CSV
+
+### Experiment 10
+- Online: Keras IMDB
+- Local: `text,label` CSV
+
+## Dependencies for running generated code
+
+Install the common ML stack:
+
+```bash
+pip install tensorflow scikit-learn pandas numpy matplotlib pillow
+```
+
+Experiment 8 online additionally needs:
+
+```bash
+pip install tensorflow-datasets
+```
+
+The package itself has no heavy ML dependency because it only generates/prints the code.

dl_experiment_code-1.0.0.dist-info/RECORD

@@ -0,0 +1,20 @@
+dl_experiment_code/__init__.py,sha256=J-j-u0itpEFT6irdmWmixQqYMadNl1X91TxUmoiLHMI,22
+dl_experiment_code/cli.py,sha256=Oghpi-MT0Q2Rc9RLftF6AGGCWlgvecNbp9KWyn0WYI4,3230
+dl_experiment_code/templates/exp10_local.py,sha256=_lTNiafsJpUsgK3aBveAac0sKukI6XxXFBy3vkFKFxw,3264
+dl_experiment_code/templates/exp10_online.py,sha256=dwgWOgH_RZYvMaIGi6zdG1yJDlqx5IsRWXppnoln5XA,2704
+dl_experiment_code/templates/exp5_local.py,sha256=H4C7DjU2xsrKMfobrWWbTdFVey-J3hFYRBXYK2pDxC0,2055
+dl_experiment_code/templates/exp5_online.py,sha256=dL2YpaD6ioXGrtbJGSh0KU6Fz1uYVK8Zz20VkoqmzTc,1547
+dl_experiment_code/templates/exp6_local.py,sha256=sthjUPUZdvCw-jVi4kUB0IaC6rAtyGhU1YM0dMqDGY0,2797
+dl_experiment_code/templates/exp6_online.py,sha256=f96d7Lyo9uvnURvpe8r8X8tSc2YSzrC1bz46rD58IYM,2221
+dl_experiment_code/templates/exp7_local.py,sha256=MbainAAi3nb97IT_2Sl9BLp28PMjK04R6uAZEOFG8hs,1666
+dl_experiment_code/templates/exp7_online.py,sha256=jhqjdUu0qbOGbfzaH_-4QcJ4JfV9-CCdD8c-r_6e4iU,1626
+dl_experiment_code/templates/exp8_local.py,sha256=Jdux8z9ru7Af_oK7qRlteU4LGMR11km0XwLlxIr5j3s,3528
+dl_experiment_code/templates/exp8_online.py,sha256=CSATt_FSUYdxAx20Jmw6l6_uQzvo6_vmMwDmPr73YQ4,3196
+dl_experiment_code/templates/exp9_local.py,sha256=lrZDHKi2XH6gyoabiD7rYxCbtmq_Imyfkl4-3WOZo_Q,1896
+dl_experiment_code/templates/exp9_online.py,sha256=wxwgpidkp7YMUK5KUzMetbLf84p7thyeq3d4AcAt1SQ,1050
+dl_experiment_code/templates/rnn_scratch.py,sha256=mTp3MSO44-CHnaq7URNwMNyTUlEVMiWAIrPSwbsti24,3104
+dl_experiment_code-1.0.0.dist-info/METADATA,sha256=wNwD20L7EycMJ7zI4EjNsZOeP1kgOCQ5OU5sZxVvjO0,3074
+dl_experiment_code-1.0.0.dist-info/WHEEL,sha256=YVMoNqKzERt-wjUZwJ33xBGAwnFl-4cqbYkTtWa4itE,91
+dl_experiment_code-1.0.0.dist-info/entry_points.txt,sha256=9SDDec93jHV_2WevqFkaDJnP4bmxGca6Tvf5Mrdifso,58
+dl_experiment_code-1.0.0.dist-info/top_level.txt,sha256=nHVCBncv79oF6DPbCGidx1VRH77ACC4ePE-8uwG_hdg,19
+dl_experiment_code-1.0.0.dist-info/RECORD,,

dl_experiment_code-1.0.0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (84.0.0)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

dl_experiment_code-1.0.0.dist-info/entry_points.txt

@@ -0,0 +1,2 @@
+[console_scripts]
+dlab-code = dl_experiment_code.cli:main

dl_experiment_code-1.0.0.dist-info/top_level.txt

@@ -0,0 +1 @@
+dl_experiment_code