-
Notifications
You must be signed in to change notification settings - Fork 2
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
[BUG]:
flash_decoding_unpaged_forwardin theDeFTAttentiondoes the dummy process, e.g., the attention functionmemory_efficient_attention_forwardis commented and undefined.bugSomething isn't workingSomething isn't workingStatus: Open.#117 In LINs-lab/DeFT;[FEAT]: Add dynamic grouping indicator for GQA/MQA
enhancementNew feature or requestNew feature or requestpriority-0priority of the issuepriority of the issueStatus: Open.#116 In LINs-lab/DeFT;[ORG]: Add pytest and CI
enhancementNew feature or requestNew feature or requestpriority-0priority of the issuepriority of the issueStatus: Open.#115 In LINs-lab/DeFT;[BUG]: Llama3.1-8B to fix
bugSomething isn't workingSomething isn't workingpriority-0priority of the issuepriority of the issueStatus: Open.#114 In LINs-lab/DeFT;[FEAT]: code structure of kernel
enhancementNew feature or requestNew feature or requestpriority-0priority of the issuepriority of the issueStatus: Open.#17 In LINs-lab/DeFT;[FEAT]: implement real speculative decoding(SD) and multi-step reasoning(MS)
enhancementNew feature or requestNew feature or requestpriority-1priority of the issuepriority of the issueStatus: Open.[FEAT]: implement KV cache manager
enhancementNew feature or requestNew feature or requestpriority-0priority of the issuepriority of the issueStatus: Open.[FEAT]: memory management for deft kernel tests
enhancementNew feature or requestNew feature or requestpriority-0priority of the issuepriority of the issueStatus: Open.[FEAT]: cuda kernel of deft
enhancementNew feature or requestNew feature or requestpriority-0priority of the issuepriority of the issueStatus: Open.