Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Flow Matching for Generative Modeling
Authors: Yaron Lipman, Ricky T. Q. Chen, Heli Ben-Hamu, Maximilian Nickel, Matthew Le
ICLR 2023 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We explore the empirical benefits of using Flow Matching on the image datasets of CIFAR10 (Krizhevsky et al., 2009) and Image Net at resolutions 32, 64, and 128 (Chrabaszcz et al., 2017; Deng et al., 2009). We find that we can easily train models to achieve favorable performance in both likelihood estimation and sample quality amongst competing diffusion-based methods. |
| Researcher Affiliation | Collaboration | 1Meta AI (FAIR) 2Weizmann Institute of Science |
| Pseudocode | No | The paper describes methods using mathematical equations and prose but does not include any clearly labeled pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not contain an explicit statement about releasing its own source code for the methodology described, nor does it provide a direct link to its specific repository. It only refers to third-party libraries and evaluation scripts. |
| Open Datasets | Yes | We explore the empirical benefits of using Flow Matching on the image datasets of CIFAR10 (Krizhevsky et al., 2009) and Image Net at resolutions 32, 64, and 128 (Chrabaszcz et al., 2017; Deng et al., 2009). |
| Dataset Splits | Yes | Image super-resolution on the Image Net validation set. |
| Hardware Specification | No | The paper mentions using 'full 32 bit-precision' and '16-bit mixed precision' for training and lists the number of GPUs used (e.g., 'GPUs 2, 4, 16, 32' in Table 3), but it does not specify the exact models of GPUs, CPUs, or other hardware components used for experiments. |
| Software Dependencies | No | Additionally, we acknowledge the Python community (Van Rossum & Drake Jr, 1995; Oliphant, 2007) for developing the core set of tools that enabled this work, including Py Torch (Paszke et al., 2019), Py Torch Lightning (Falcon & team, 2019), Hydra (Yadan, 2019), Jupyter (Kluyver et al., 2016), Matplotlib (Hunter, 2007), seaborn (Waskom et al., 2018), numpy (Oliphant, 2006; Van Der Walt et al., 2011), Sci Py (Jones et al., 2014), and torchdiffeq (Chen, 2018). These are citations to papers or authors, not specific software version numbers (e.g., PyTorch 1.9). |
| Experiment Setup | Yes | Table 3: Hyper-parameters used for training each model. All models are trained using the Adam optimizer with the following parameters: β1 = 0.9, β2 = 0.999, weight decay = 0.0, and ϵ = 1e 8. |