SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2306.00288 · ACL · 2023

Training-free Neural Architecture Search for RNNs and Transformers

Jugal Kalita, Aaron Serianni

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 5 functions out of this paper's own repositories and ran 0 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
aaronserianni/training-free-nas canonical 0 of 5
FunctionStatusWhere it lives
create_int_feature Not yet run aaronserianni/training-free-nas/tensorflow-model/build_pretraining_dataset.py
code served (permissive licence) · get_code("b03d45af664a39e3")
dct Not yet run aaronserianni/training-free-nas/modeling_electra.py
code served (permissive licence) · get_code("0e1b2cabd3823f3f")
dct_2d Not yet run aaronserianni/training-free-nas/modeling_electra.py
code served (permissive licence) · get_code("6cacf2c0ab871ae9")
get_softmax_output Not yet run aaronserianni/training-free-nas/tensorflow-model/run_pretraining.py
code served (permissive licence) · get_code("170c7c6b569e1aa2")
secs_to_str Not yet run aaronserianni/training-free-nas/tensorflow-model/util/training_utils.py
code served (permissive licence) · get_code("9c127cf70c5ee0db")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Neural architecture search (NAS) has allowed for the automatic creation of new and effective neural network architectures, offering an alternative to the laborious process of manually designing complex architectures. However, traditional NAS algorithms are slow and require immense amounts of computing power. Recent research has investigated training-free NAS metrics for image classification architectures, drastically speeding up search algorithms. In this paper, we investigate trainingfree NAS metrics for recurrent neural network (RNN) and BERT-based transformer architectures, targeted towards language modeling tasks. First, we develop a new trainingfree metric, named hidden covariance, that predicts the trained performance of an RNN architecture and significantly outperforms existing training-free metrics. We experimentally evaluate the effectiveness of the hidden covariance metric on the NAS-Bench-NLP benchmark. Second, we find that the current search space paradigm for transformer architectures is not optimized for training-free neural architecture search. Instead, a simple qualitative analysis can effectively shrink the search space to the best performing architectures. This conclusion is based on our investigation of existing training-free metrics and new metrics developed from recent transformer pruning literature, evaluated on our own benchmark of trained BERT architectures. Ultimately, our analysis shows that the architecture search space and the training-free metric must be developed together in order to achieve effective results. Our source code is available at https://github. com/aaronserianni/training-free-nas.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2306.00288")
get_code_for_paper("2306.00288")
have("2306.00288")

Connect an agent — have() is free.