Merge pull request #250 from jschrier/my-pages

minor link updates, etc.:AM lab and fine tuning

_posts/2024-06-29-LLM-Finetuning-Notes.md

@@ -39,6 +39,8 @@ What is frustrating (*circa late June/early July 2024*) is that you can't see wh
 
 You've got to use the web-interface; as of late-June 2024, the API for fine-tuning is still in beta.
 
+**(update: Nov 2024:)** I created a [short tutorial example](https://github.com/jschrier/KRICT_hackathon_phosphors/) of fine-tuning with various models on a chemistry prediction task using OpenPipe for a hackathon project. This shows off the convenience of the new 
+
 
 # Predibase
 
@@ -128,6 +130,6 @@ Unsloth is a library for performance tunning LLMs and they have a worked example
 
 # Other things seen on the internet
 
-- Alex Strick van Linschoten has been following the same LLM fine-tuning course as me, and has written blog posts about his experiences doing a [local (4-bit quantized) fine-tune using Axolotl](https://mlops.systems/posts/2024-06-15-isafpr-first-finetune.html) and comparing [one-click fine-tuning on Predibase, OpenPipe, and OpenAI](https://mlops.systems/posts/2024-06-17-one-click-finetuning.html) and [head to head comparisons and evals of his fine-tuned models against GPT-4](https://mlops.systems/posts/2024-07-01-full-finetuned-model-evaluation.html)
+- Alex Strick van Linschoten followed the same LLM fine-tuning course as me, and has written blog posts about his experiences doing a [local (4-bit quantized) fine-tune using Axolotl](https://mlops.systems/posts/2024-06-15-isafpr-first-finetune.html) and comparing [one-click fine-tuning on Predibase, OpenPipe, and OpenAI](https://mlops.systems/posts/2024-06-17-one-click-finetuning.html) and [head to head comparisons and evals of his fine-tuned models against GPT-4](https://mlops.systems/posts/2024-07-01-full-finetuned-model-evaluation.html)
 - [Distributed fine tuning Gemma 2 - 27B with Keras](https://developers.googleblog.com/zh-hans/fine-tuning-gemma-2-with-keras-hugging-face-update/) guide from Goog
 

_posts/2024-08-05-Imaginary-Syllabi:-Additive-Manufacturing-Lab.md

@@ -20,7 +20,7 @@ tags: 3dprinting diy art science teaching imaginary-syllabi
 
 # First principles and foundations
 
-- *Conceptually comprehensive:* Include all of the major areas of materials (polymers, metals, ceramics) and try to include many (but not all) additive modalities.  Aim for illustrating basic transferable principles, that have a connection to underlying fundamental chemistry...sort of like the lab complement to [Shackelford's intro materials science textbook](https://amzn.to/3JNUFbz)   
+- *Conceptually comprehensive:* Include the major areas of materials (polymers, metals, ceramics) and try to include many (but not all) additive modalities.  Aim for illustrating basic transferable principles, that have a connection to underlying fundamental chemistry...sort of like the lab complement to [Shackelford's intro materials science textbook](https://amzn.to/3JNUFbz)   
 
 - *Safety:* We don't want powder explosion hazards, etc.  Focus on relatively safe methods
 
@@ -29,7 +29,7 @@ tags: 3dprinting diy art science teaching imaginary-syllabi
 
 # Polymer
 
-- *Fused deposition modeling:* Obviously a must have, but also really easy.  Get a [Prusa MK4](https://www.prusa3d.com/product/original-prusa-mk4-2) ($1100 assembled or $700 DIY) or [XL $2K](https://www.prusa3d.com/product/original-prusa-xl/) and be done with it. I've personally found the MK3S+ super reliable. MK4 has easy nozzle changes, which will be important for doing some of the sinter-based metal products
+- *Fused deposition modeling:* Obviously a must have, but also really easy.  Get a [Prusa MK4](https://www.prusa3d.com/product/original-prusa-mk4-2) ($1100 assembled or $700 DIY) or [XL $2K](https://www.prusa3d.com/product/original-prusa-xl/) or the new [Core 1](https://www.prusa3d.com/product/prusa-core-one/) ($1200 assembled or $900) and be done with it. I've personally found the MK3S+ super reliable. Easy nozzle changes, which will be important for doing some of the sinter-based metal products
     - Learning goals:  Polymer structure/property relations (PLA, PETG, TPU, ABS, etc.)
     - Basic tricks of the trade: modeling, 3d file formats, slicers
     - Good for introducing ideas about: anisotropy (of strength under compression and tension, which is more pronounced because of layering, design for manufacture, processing conditions (temperature, humidity, curing (in the case of PLA+)))