Scalable Bayesian Meta-Learning through Generalized Implicit Gradients

Authors: Yilang Zhang, Bingcong Li, Shijian Gao, Georgios B. Giannakis

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive numerical tests are also carried out to empirically validate the performance of the proposed method.
Researcher Affiliation Academia Dept. of ECE, University of Minnesota, Minneapolis, MN, USA {zhan7453,lixx5599,gao00379,georgios}@umn.edu
Pseudocode Yes Algorithm 1: Implicit Bayesian meta-learning (i Ba ML)
Open Source Code Yes Our implementation relies on the Py Torch (Paszke et al. 2019), and codes are available at https://github.com/zhangyilang/i Ba ML.
Open Datasets Yes We consider one of the most widely used few-shot dataset for classification mini Image Net (Vinyals et al. 2016).
Dataset Splits Yes We adopt the dataset splitting suggested by (Ravi and Larochelle 2017), where 64, 16 and 20 disjoint classes are used for meta-training, meta-validation and meta-testing, respectively.
Hardware Specification No No specific hardware details (like GPU/CPU models, memory, or cloud instances) used for experiments were mentioned.
Software Dependencies No Our implementation relies on the Py Torch (Paszke et al. 2019).
Experiment Setup Yes The model is a standard 4-layer 32-channel convolutional neural network, and the chosen baseline algorithms are MAML (Finn, Abbeel, and Levine 2017) and ABML (Ravi and Beatson 2019); see also the Appendix for alternative setups. Due to the large number of training tasks, it is impractical to compute the exact meta-training loss. As an alternative, we adopt the test nll (averaged over 1, 000 test tasks) as our metric, and also report their corresponding accuracy. For fairness, we set L = 5 when implementing the implicit gradients so that the time complexity is similar to explicit one with K = 5.