Picture for Urmish Thakker

Urmish Thakker

Distributed LLMs and Multimodal Large Language Models: A Survey on Advances, Challenges, and Future Directions

Add code
Mar 20, 2025
Viaarxiv icon

LLMs Know What to Drop: Self-Attention Guided KV Cache Eviction for Efficient Long-Context Inference

Add code
Mar 11, 2025
Viaarxiv icon

Training Domain Draft Models for Speculative Decoding: Best Practices and Insights

Add code
Mar 10, 2025
Viaarxiv icon

SubgoalXL: Subgoal-based Expert Learning for Theorem Proving

Add code
Aug 20, 2024
Figure 1 for SubgoalXL: Subgoal-based Expert Learning for Theorem Proving
Figure 2 for SubgoalXL: Subgoal-based Expert Learning for Theorem Proving
Figure 3 for SubgoalXL: Subgoal-based Expert Learning for Theorem Proving
Figure 4 for SubgoalXL: Subgoal-based Expert Learning for Theorem Proving
Viaarxiv icon

SambaNova SN40L: Scaling the AI Memory Wall with Dataflow and Composition of Experts

Add code
May 13, 2024
Figure 1 for SambaNova SN40L: Scaling the AI Memory Wall with Dataflow and Composition of Experts
Figure 2 for SambaNova SN40L: Scaling the AI Memory Wall with Dataflow and Composition of Experts
Figure 3 for SambaNova SN40L: Scaling the AI Memory Wall with Dataflow and Composition of Experts
Figure 4 for SambaNova SN40L: Scaling the AI Memory Wall with Dataflow and Composition of Experts
Viaarxiv icon

SambaLingo: Teaching Large Language Models New Languages

Add code
Apr 08, 2024
Figure 1 for SambaLingo: Teaching Large Language Models New Languages
Figure 2 for SambaLingo: Teaching Large Language Models New Languages
Figure 3 for SambaLingo: Teaching Large Language Models New Languages
Figure 4 for SambaLingo: Teaching Large Language Models New Languages
Viaarxiv icon

Efficiently Adapting Pretrained Language Models To New Languages

Add code
Nov 09, 2023
Figure 1 for Efficiently Adapting Pretrained Language Models To New Languages
Figure 2 for Efficiently Adapting Pretrained Language Models To New Languages
Figure 3 for Efficiently Adapting Pretrained Language Models To New Languages
Figure 4 for Efficiently Adapting Pretrained Language Models To New Languages
Viaarxiv icon

Training Large Language Models Efficiently with Sparsity and Dataflow

Add code
Apr 11, 2023
Viaarxiv icon

BLOOM: A 176B-Parameter Open-Access Multilingual Language Model

Add code
Nov 09, 2022
Viaarxiv icon

PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts

Add code
Feb 02, 2022
Figure 1 for PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts
Figure 2 for PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts
Figure 3 for PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts
Figure 4 for PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts
Viaarxiv icon