# mila-iqia/blocks

A Theano framework for building and training neural networks

Repository: https://github.com/mila-iqia/blocks
Canonical: https://ross.abutalabs.com/products/mila-iqia-blocks
Language: Python
License: NOASSERTION
License Family: other
Last push: 2019-02-19T12:41:38+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 4350, "days_push": 2752, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1152, forks 348 (observed 2026-08-28T04:03:46.901726+00:00)

## What it is
Blocks is a Python framework for building and training neural network models on top of Theano. It provides parametrized operations called 'bricks', model selection via pattern matching, optimization algorithms, training monitoring, and graph transformations like dropout.

## Use cases
- build neural network models on top of Theano
- train deep learning models with monitoring and checkpointing
- apply graph transformations like dropout to models
- select and manipulate variables in large Theano models
- save and resume training runs

## When to choose
- you are maintaining or reproducing legacy research code built on Theano and Blocks

## When to avoid
- you are starting a new deep learning project (use PyTorch, TensorFlow, or JAX instead)
- you need actively maintained tooling or modern hardware acceleration
- you do not want to depend on Theano, which is itself unmaintained

## Facets
- artifact type: framework
- maturity: abandoned
- function: deep-learning, machine-learning
- domain: deep-learning, machine-learning
- platform: python
- tags: theano, neural-networks, bricks, training-loop, graph-transformations

## Member repositories
- mila-iqia/blocks (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:46.901726+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:33:33.789028+00:00, confidence not recorded.
  - readme: https://github.com/mila-iqia/blocks (fetched 2026-08-28T04:03:46.901726+00:00, sha a39a9cedcd41)
- Data as of 2026-08-30T08:39:29.467469+00:00.