fix linkcheck & build

_config.yml

@@ -100,8 +100,11 @@ sphinx:
     committers: .
     bibliography: .
     prem_theme: .
+  recursive_update: true  # append to rather than overwrite html_theme_options
   config:
     html_theme: prem_theme
+    html_theme_options:
+      navigation_with_keys: false
     myst_heading_anchors: 4
     html_js_files:
     - [https://analytics.prem.ninja/js/script.js, {defer: defer, data-domain: book.premai.io}]
@@ -134,6 +137,7 @@ sphinx:
     - https://accent.gmu.edu
     - https://www.crcv.ucf.edu/data/UCF101.php
     - https://www.pinecone.io/learn/series/faiss/locality-sensitive-hashing
+    - https://chat.lmsys.org
     linkcheck_allowed_redirects:
       https://doi.org/.*/.*: https://.*
       https://codespaces.new/.*: https://github.com/login.*

references.md

@@ -33,7 +33,7 @@ Couldn't decide which chapter(s) these links are related to. They're mostly abou
 - "LLaMA from scratch (or how to implement a paper without crying)" (misc tips, scaled-down version of LLaMA for training) https://blog.briankitano.com/llama-from-scratch
 - "Swift Transformers: Run On-Device LLMs in Apple Devices" https://huggingface.co/blog/swift-coreml-llm
 - "Why GPT-3.5-turbo is (mostly) cheaper than LLaMA-2" https://cursor.sh/blog/llama-inference#user-content-fn-gpt4-leak
-- http://marble.onl/posts/why_host_your_own_llm.html
+- https://www.marble.onl/posts/why_host_your_own_llm.html
 - https://betterprogramming.pub/you-dont-need-hosted-llms-do-you-1160b2520526
 - "Low-code framework for building custom LLMs, neural networks, and other AI models" https://github.com/ludwig-ai/ludwig
 - "GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers" https://arxiv.org/abs/2210.17323