Skip to content

Commit

Permalink
redeploy
Browse files Browse the repository at this point in the history
  • Loading branch information
acganesh committed Nov 14, 2023
1 parent 68b404f commit 9f99c4f
Show file tree
Hide file tree
Showing 3 changed files with 4 additions and 4 deletions.
2 changes: 1 addition & 1 deletion docs/posts/speculative_decoding/index.html
Original file line number Diff line number Diff line change
Expand Up @@ -71,7 +71,7 @@ <h1 id="precursor-llms-are-bound-by-memory-bandwidth-at-inference-time">
Precursor: LLMs are bound by memory-bandwidth at inference time
<a href="#precursor-llms-are-bound-by-memory-bandwidth-at-inference-time" class="heading-anchor">#</a>
</h1>
<p>Below is the hierarchy of memory on a system with a CPU and A100 GPU. <label for="marginnote-1" class="margin-toggle marginnote-ind"></label>
<p>Below is the hierarchy of memory on a system with a CPU and A100 GPU. <label for="marginnote-1" class="margin-toggle marginnote-ind">💬</label>
<input type="checkbox" id="marginnote-1" class="margin-toggle"/>
<span class="marginnote">
Source: The <a href="https://arxiv.org/pdf/2205.14135.pdf">FlashAttention paper</a>.
Expand Down
4 changes: 2 additions & 2 deletions site/config.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -12,14 +12,14 @@ params:
codeblocksdark: false
# Customize the indicator for margin notes
# Some suggestions: ⊕, 💬, 💭, 📑, 🏷 , ✍ , 💡, 🧐, 📎, 📌
marginNoteInd: ""
marginNoteInd: "💬"
# Your name or the name of you company
# copyright: Copyright 2023
# copyrightHolder: Copyright Holder
# Show the "Powered by Hugo-Tufte and Hugo."
showPoweredBy: false
# Site wide kill switch for date in pages
hidedate: false
hidedate: true
# Site wide kill switch for post summary on home page
showSummary: true
# Site wide kill switch for LaTeX support
Expand Down
2 changes: 1 addition & 1 deletion site/public/posts/speculative_decoding/index.html
Original file line number Diff line number Diff line change
Expand Up @@ -71,7 +71,7 @@ <h1 id="precursor-llms-are-bound-by-memory-bandwidth-at-inference-time">
Precursor: LLMs are bound by memory-bandwidth at inference time
<a href="#precursor-llms-are-bound-by-memory-bandwidth-at-inference-time" class="heading-anchor">#</a>
</h1>
<p>Below is the hierarchy of memory on a system with a CPU and A100 GPU. <label for="marginnote-1" class="margin-toggle marginnote-ind"></label>
<p>Below is the hierarchy of memory on a system with a CPU and A100 GPU. <label for="marginnote-1" class="margin-toggle marginnote-ind">💬</label>
<input type="checkbox" id="marginnote-1" class="margin-toggle"/>
<span class="marginnote">
Source: The <a href="https://arxiv.org/pdf/2205.14135.pdf">FlashAttention paper</a>.
Expand Down

0 comments on commit 9f99c4f

Please sign in to comment.