Picture for Denis Dimitrov

Denis Dimitrov

Sber AI

Your Transformer is Secretly Linear

Add code
May 19, 2024
Figure 1 for Your Transformer is Secretly Linear
Figure 2 for Your Transformer is Secretly Linear
Figure 3 for Your Transformer is Secretly Linear
Figure 4 for Your Transformer is Secretly Linear
Viaarxiv icon

OmniFusion Technical Report

Add code
Apr 09, 2024
Figure 1 for OmniFusion Technical Report
Figure 2 for OmniFusion Technical Report
Figure 3 for OmniFusion Technical Report
Figure 4 for OmniFusion Technical Report
Viaarxiv icon

MERA: A Comprehensive LLM Evaluation in Russian

Add code
Jan 12, 2024
Figure 1 for MERA: A Comprehensive LLM Evaluation in Russian
Figure 2 for MERA: A Comprehensive LLM Evaluation in Russian
Figure 3 for MERA: A Comprehensive LLM Evaluation in Russian
Figure 4 for MERA: A Comprehensive LLM Evaluation in Russian
Viaarxiv icon

Kandinsky 3.0 Technical Report

Add code
Dec 11, 2023
Figure 1 for Kandinsky 3.0 Technical Report
Figure 2 for Kandinsky 3.0 Technical Report
Figure 3 for Kandinsky 3.0 Technical Report
Figure 4 for Kandinsky 3.0 Technical Report
Viaarxiv icon

FusionFrames: Efficient Architectural Aspects for Text-to-Video Generation Pipeline

Add code
Nov 22, 2023
Figure 1 for FusionFrames: Efficient Architectural Aspects for Text-to-Video Generation Pipeline
Figure 2 for FusionFrames: Efficient Architectural Aspects for Text-to-Video Generation Pipeline
Figure 3 for FusionFrames: Efficient Architectural Aspects for Text-to-Video Generation Pipeline
Figure 4 for FusionFrames: Efficient Architectural Aspects for Text-to-Video Generation Pipeline
Viaarxiv icon

The Shape of Learning: Anisotropy and Intrinsic Dimensions in Transformer-Based Models

Add code
Nov 10, 2023
Figure 1 for The Shape of Learning: Anisotropy and Intrinsic Dimensions in Transformer-Based Models
Figure 2 for The Shape of Learning: Anisotropy and Intrinsic Dimensions in Transformer-Based Models
Figure 3 for The Shape of Learning: Anisotropy and Intrinsic Dimensions in Transformer-Based Models
Figure 4 for The Shape of Learning: Anisotropy and Intrinsic Dimensions in Transformer-Based Models
Viaarxiv icon

Kandinsky: an Improved Text-to-Image Synthesis with Image Prior and Latent Diffusion

Add code
Oct 05, 2023
Figure 1 for Kandinsky: an Improved Text-to-Image Synthesis with Image Prior and Latent Diffusion
Figure 2 for Kandinsky: an Improved Text-to-Image Synthesis with Image Prior and Latent Diffusion
Figure 3 for Kandinsky: an Improved Text-to-Image Synthesis with Image Prior and Latent Diffusion
Figure 4 for Kandinsky: an Improved Text-to-Image Synthesis with Image Prior and Latent Diffusion
Viaarxiv icon

RusTitW: Russian Language Text Dataset for Visual Text in-the-Wild Recognition

Add code
Mar 29, 2023
Figure 1 for RusTitW: Russian Language Text Dataset for Visual Text in-the-Wild Recognition
Figure 2 for RusTitW: Russian Language Text Dataset for Visual Text in-the-Wild Recognition
Figure 3 for RusTitW: Russian Language Text Dataset for Visual Text in-the-Wild Recognition
Figure 4 for RusTitW: Russian Language Text Dataset for Visual Text in-the-Wild Recognition
Viaarxiv icon

Eco2AI: carbon emissions tracking of machine learning models as the first step towards sustainable AI

Add code
Aug 03, 2022
Figure 1 for Eco2AI: carbon emissions tracking of machine learning models as the first step towards sustainable AI
Figure 2 for Eco2AI: carbon emissions tracking of machine learning models as the first step towards sustainable AI
Figure 3 for Eco2AI: carbon emissions tracking of machine learning models as the first step towards sustainable AI
Figure 4 for Eco2AI: carbon emissions tracking of machine learning models as the first step towards sustainable AI
Viaarxiv icon

RuCLIP -- new models and experiments: a technical report

Add code
Feb 22, 2022
Figure 1 for RuCLIP -- new models and experiments: a technical report
Figure 2 for RuCLIP -- new models and experiments: a technical report
Figure 3 for RuCLIP -- new models and experiments: a technical report
Figure 4 for RuCLIP -- new models and experiments: a technical report
Viaarxiv icon