From a94db2d83679db463445a00623ebe874bda3e44e Mon Sep 17 00:00:00 2001 From: Kajol Tanesh Shah Date: Sun, 21 Sep 2025 19:52:59 -0400 Subject: [PATCH 1/3] Fix spelling in flowchart and update images --- chapters/en/chapter1/4.mdx | 4 +- chapters/en/images/finetuning_darkupdated.svg | 178 ++++++++++++++++++ chapters/en/images/finetuning_updated.svg | 156 +++++++++++++++ 3 files changed, 336 insertions(+), 2 deletions(-) create mode 100644 chapters/en/images/finetuning_darkupdated.svg create mode 100644 chapters/en/images/finetuning_updated.svg diff --git a/chapters/en/chapter1/4.mdx b/chapters/en/chapter1/4.mdx index 3870b541f..fd0b3c926 100644 --- a/chapters/en/chapter1/4.mdx +++ b/chapters/en/chapter1/4.mdx @@ -121,8 +121,8 @@ This pretraining is usually done on very large amounts of data. Therefore, it re For example, one could leverage a pretrained model trained on the English language and then fine-tune it on an arXiv corpus, resulting in a science/research-based model. The fine-tuning will only require a limited amount of data: the knowledge the pretrained model has acquired is "transferred," hence the term *transfer learning*.
-The fine-tuning of a language model is cheaper than pretraining in both time and money. - +The fine-tuning of a language model is cheaper than pretraining in both time and money. +
Fine-tuning a model therefore has lower time, data, financial, and environmental costs. It is also quicker and easier to iterate over different fine-tuning schemes, as the training is less constraining than a full pretraining. diff --git a/chapters/en/images/finetuning_darkupdated.svg b/chapters/en/images/finetuning_darkupdated.svg new file mode 100644 index 000000000..87c989f48 --- /dev/null +++ b/chapters/en/images/finetuning_darkupdated.svg @@ -0,0 +1,178 @@ + + + + + + + + + + + + + + + + + + + + + + + + + Easily Reproducible + diff --git a/chapters/en/images/finetuning_updated.svg b/chapters/en/images/finetuning_updated.svg new file mode 100644 index 000000000..093a9f8d9 --- /dev/null +++ b/chapters/en/images/finetuning_updated.svg @@ -0,0 +1,156 @@ + + + + + + + + + + + + + + + + + + + + + + + Easily Reproducible + From bca911c6053a5f7aba74948f6bcf30befca5b75d Mon Sep 17 00:00:00 2001 From: Kajol Shah <37287505+kajolshah310@users.noreply.github.com> Date: Sun, 21 Sep 2025 20:02:12 -0400 Subject: [PATCH 2/3] Update 4.mdx --- chapters/en/chapter1/4.mdx | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/chapters/en/chapter1/4.mdx b/chapters/en/chapter1/4.mdx index fd0b3c926..71c8b727d 100644 --- a/chapters/en/chapter1/4.mdx +++ b/chapters/en/chapter1/4.mdx @@ -121,7 +121,9 @@ This pretraining is usually done on very large amounts of data. Therefore, it re For example, one could leverage a pretrained model trained on the English language and then fine-tune it on an arXiv corpus, resulting in a science/research-based model. The fine-tuning will only require a limited amount of data: the knowledge the pretrained model has acquired is "transferred," hence the term *transfer learning*.
-The fine-tuning of a language model is cheaper than pretraining in both time and money. + +The fine-tuning of a language model is cheaper than pretraining in both time and money. +
From 15457b24270112b69c1a576be8f22ec0b2d5573c Mon Sep 17 00:00:00 2001 From: Kajol Tanesh Shah Date: Mon, 22 Sep 2025 18:25:05 -0400 Subject: [PATCH 3/3] images updated --- chapters/en/images/finetuning_darkupdated.svg | 18 ++++++++--------- chapters/en/images/finetuning_updated.svg | 20 +++++++++---------- 2 files changed, 19 insertions(+), 19 deletions(-) diff --git a/chapters/en/images/finetuning_darkupdated.svg b/chapters/en/images/finetuning_darkupdated.svg index 87c989f48..d82109c76 100644 --- a/chapters/en/images/finetuning_darkupdated.svg +++ b/chapters/en/images/finetuning_darkupdated.svg @@ -24,12 +24,12 @@ inkscape:pagecheckerboard="0" inkscape:deskcolor="#d1d1d1" inkscape:zoom="0.50686018" - inkscape:cx="914.45338" - inkscape:cy="413.32898" + inkscape:cx="914.45337" + inkscape:cy="414.31544" inkscape:window-width="1440" - inkscape:window-height="872" + inkscape:window-height="791" inkscape:window-x="0" - inkscape:window-y="28" + inkscape:window-y="25" inkscape:window-maximized="0" inkscape:current-layer="svg12" /> Easily Reproducible + x="1124.5704" + y="495.2056" + style="font-family:sans-serif;font-size:33.3333px;text-align:center;font-weight:300;text-anchor:middle">Easily Reproducible diff --git a/chapters/en/images/finetuning_updated.svg b/chapters/en/images/finetuning_updated.svg index 093a9f8d9..8df4d3adb 100644 --- a/chapters/en/images/finetuning_updated.svg +++ b/chapters/en/images/finetuning_updated.svg @@ -24,12 +24,12 @@ inkscape:pagecheckerboard="0" inkscape:deskcolor="#d1d1d1" inkscape:zoom="0.54260239" - inkscape:cx="945.44367" - inkscape:cy="312.38344" + inkscape:cx="944.52219" + inkscape:cy="314.22641" inkscape:window-width="1440" - inkscape:window-height="900" + inkscape:window-height="872" inkscape:window-x="0" - inkscape:window-y="0" + inkscape:window-y="28" inkscape:window-maximized="0" inkscape:current-layer="svg10" /> Easily Reproducible + x="1121.7754" + y="445.99878" + style="font-family:sans-serif;font-size:33.3333px;text-align:center;font-weight:300;text-anchor:middle">Easily Reproducible