Figure 7 - uploaded by Shachar Mirkin
Content may be subject to copyright.
Download
View publication
Performance of various LLMs on subset of tasks from SuperGLUE benchmark in zero-and one-shot prompt-based setting.

Performance of various LLMs on subset of tasks from SuperGLUE benchmark in zero-and one-shot prompt-based setting.

Source publication
Figure 1: Organization of BigScience working groups.
Figure 2: Creation Pipeline of the ROOTS Corpus. The purple-colored...
Figure 5: The BLOOM architecture. The k head slope parameters for ALIBI...
Figure 6: DP+PP+TP combination leads to 3D parallelism.
Figure 7: Performance of various LLMs on subset of tasks from SuperGLUE...
BLOOM: A 176B-Parameter Open-Access Multilingual Language Model Major Contributors Prompt Engineering BigScience Workshop Architecture and Objective Engineering Evaluation and Interpretability Broader Impacts
Preprint
Full-text available
  • Mar 2023
Large language models (LLMs) have been shown to be able to perform new tasks based on a few demonstrations or natural language instructions. While these capabilities have led to widespread adoption, most LLMs are developed by resource-rich organizations and are frequently kept from the public. As a step towards democratizing this powerful technolog...
Cite
Download full-text

Context in source publication

Context 1
... mGPT ( Shliazhko et al., 2022), GPT-style models trained on 60 languages from Wikipedia and Common Crawl Figure 7 shows zero-and one-shot performance on SuperGLUE. In both settings, on entailment tasks (BoolQ and CB), performance is well above random chance for BLOOM, T0, OPT, and GPT-J. ...

Similar publications

Figure 1. Proposed architecture diagram
Figure 2. Exact extracted location on folium map with blue marker
Hyper parameter tuning results
Headline results
A multimodal machine learning approach to generate news articles from geo-tagged images
Article
Full-text available
  • Jun 2024
Classical machine learning algorithms typically operate on unimodal data and hence it can analyze and make predictions based on data from a single source (modality). Whereas multimodal machine learning algorithm, learns from information across multiple modalities, such as text, images, audio, and sensor data. The paper leverages the functionalities...
View