Skip to content

Commit

Permalink
Built site for gh-pages
Browse files Browse the repository at this point in the history
  • Loading branch information
@ohmeow
ohmeow committed Jul 7, 2024
1 parent a8af2fe commit decdabc
Show file tree
Hide file tree
Showing 6 changed files with 321 additions and 6 deletions.
2 changes: 1 addition & 1 deletion .nojekyll
View file
Original file line number Diff line number Diff line change
@@ -1 +1 @@
420142e5
e1be6e82
6 changes: 3 additions & 3 deletions index.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -296,7 +296,7 @@ <h5 class="quarto-listing-category-title">Categories</h5><div class="quarto-list
</div>
</div>
<div class="list quarto-listing-default">
<div class="quarto-post image-right" data-index="0" data-categories="LLMs,pydantic,Instructor" data-listing-date-sort="1720249200000" data-listing-file-modified-sort="1720311407204" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="8" data-listing-word-count-sort="1558">
<div class="quarto-post image-right" data-index="0" data-categories="LLMs,pydantic,Instructor" data-listing-date-sort="1720249200000" data-listing-file-modified-sort="1720383108838" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="11" data-listing-word-count-sort="2076">
<div class="thumbnail">
<p><a href="./posts/2024-07-06-llms-and-enums.html" class="no-external"></a></p><a href="./posts/2024-07-06-llms-and-enums.html" class="no-external">
<p><img loading="lazy" src="./posts/images/blog-20240706/header.png" class="thumbnail-image"></p>
Expand Down Expand Up @@ -333,10 +333,10 @@ <h3 class="no-anchor listing-title">
Wayde Gilliam
</div>
<div class="listing-reading-time">
8 min
11 min
</div>
<div class="metadata-value listing-file-modified">
7/6/24, 5:16:47 PM
7/7/24, 1:11:48 PM
</div>
</a>
</div>
Expand Down
248 changes: 248 additions & 0 deletions index.xml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -547,6 +547,254 @@ font-style: inherit;">"</span>,</span>
<li><p>I can use that same beautiful <code>Enum</code> across all parts of my application</p></li>
</ol>
</section>
<section id="v2.0.1-using-enum-and-fuzzywuzzy" class="level2">
<h2 class="anchored" data-anchor-id="v2.0.1-using-enum-and-fuzzywuzzy">v2.0.1: Using <code>Enum</code> and <code>fuzzywuzzy</code></h2>
<p>A suggestion from a Twitter user inspired me to enhance our approach by implementing similarity-based matching rather than relying on exact matches. To make it so, I installed the <code>fuzzywuzzy</code> library and made the necessary modifications to increase the likelihood of delivering high-quality results.</p>
<div class="sourceCode" id="cb4" style="background: #f1f3f5;"><pre class="sourceCode python code-with-copy"><code class="sourceCode python"><span id="cb4-1"><span class="kw" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">class</span> NamedEntityType(<span class="bu" style="color: null;
background-color: null;
font-style: inherit;">str</span>, Enum):</span>
<span id="cb4-2"> <span class="co" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">"""Valid types of named entities to extract."""</span></span>
<span id="cb4-3"></span>
<span id="cb4-4"> PERSON <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"PERSON"</span></span>
<span id="cb4-5"> NORP <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"NORP"</span></span>
<span id="cb4-6"> FAC <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"FAC"</span></span>
<span id="cb4-7"> ORG <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"ORG"</span></span>
<span id="cb4-8"> GPE <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"GPE"</span></span>
<span id="cb4-9"> LOC <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"LOC"</span></span>
<span id="cb4-10"> PRODUCT <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"PRODUCT"</span></span>
<span id="cb4-11"> EVENT <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"EVENT"</span></span>
<span id="cb4-12"> WORK_OF_ART <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"WORK_OF_ART"</span></span>
<span id="cb4-13"> LAW <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"LAW"</span></span>
<span id="cb4-14"> LANGUAGE <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"LANGUAGE"</span></span>
<span id="cb4-15"> DATE <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"DATE"</span></span>
<span id="cb4-16"> TIME <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"TIME"</span></span>
<span id="cb4-17"> PERCENT <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"PERCENT"</span></span>
<span id="cb4-18"> MONEY <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"MONEY"</span></span>
<span id="cb4-19"> QUANTITY <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"QUANTITY"</span></span>
<span id="cb4-20"> ORDINAL <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"ORDINAL"</span></span>
<span id="cb4-21"> CARDINAL <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"CARDINAL"</span></span>
<span id="cb4-22"> OTHER <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> <span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"OTHER"</span></span>
<span id="cb4-23"></span>
<span id="cb4-24"></span>
<span id="cb4-25"><span class="kw" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">class</span> NamedEntity(BaseModel):</span>
<span id="cb4-26"> <span class="co" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">"""A named entity result."""</span></span>
<span id="cb4-27"></span>
<span id="cb4-28"> <span class="kw" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">def</span> convert_str_to_named_entity_type(v: <span class="bu" style="color: null;
background-color: null;
font-style: inherit;">str</span> <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">|</span> NamedEntityType) <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">-&gt;</span> NamedEntityType:</span>
<span id="cb4-29"> <span class="co" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">"""Ensure entity type is a valid enum."""</span></span>
<span id="cb4-30"> <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">if</span> <span class="bu" style="color: null;
background-color: null;
font-style: inherit;">isinstance</span>(v, NamedEntityType):</span>
<span id="cb4-31"> <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">return</span> v</span>
<span id="cb4-32"> <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">else</span>:</span>
<span id="cb4-33"> <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">try</span>:</span>
<span id="cb4-34"> match, score <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> fuzzy_process.extractOne(v.upper(), [e.value <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">for</span> e <span class="kw" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">in</span> <span class="bu" style="color: null;
background-color: null;
font-style: inherit;">list</span>(NamedEntityType)])</span>
<span id="cb4-35"> <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">return</span> NamedEntityType(match) <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">if</span> score <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">&gt;=</span> <span class="dv" style="color: #AD0000;
background-color: null;
font-style: inherit;">60</span> <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">else</span> NamedEntityType.OTHER</span>
<span id="cb4-36"> <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">except</span> <span class="pp" style="color: #AD0000;
background-color: null;
font-style: inherit;">ValueError</span>:</span>
<span id="cb4-37"> <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">return</span> NamedEntityType.OTHER</span>
<span id="cb4-38"></span>
<span id="cb4-39"> entity_type: Annotated[<span class="bu" style="color: null;
background-color: null;
font-style: inherit;">str</span>, BeforeValidator(convert_str_to_named_entity_type)]</span>
<span id="cb4-40"> entity_mention: <span class="bu" style="color: null;
background-color: null;
font-style: inherit;">str</span> <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> Field(..., description<span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span><span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">"The named entity recognized."</span>)</span>
<span id="cb4-41"></span>
<span id="cb4-42"></span>
<span id="cb4-43"><span class="kw" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">class</span> DocumentNERTask(BaseModel):</span>
<span id="cb4-44"> <span class="co" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">"""Extracts the named entities found in the document.</span></span>
<span id="cb4-45"></span>
<span id="cb4-46"><span class="co" style="color: #5E5E5E;
background-color: null;
font-style: inherit;"> This tool should be used anytime the user asks for named entity recognition (NER)</span></span>
<span id="cb4-47"><span class="co" style="color: #5E5E5E;
background-color: null;
font-style: inherit;"> or wants to identify named entities.</span></span>
<span id="cb4-48"><span class="co" style="color: #5E5E5E;
background-color: null;
font-style: inherit;"> """</span></span>
<span id="cb4-49"></span>
<span id="cb4-50"> named_entities: <span class="bu" style="color: null;
background-color: null;
font-style: inherit;">list</span>[NamedEntity] <span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span> Field(</span>
<span id="cb4-51"> ...,</span>
<span id="cb4-52"> description<span class="op" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">=</span><span class="ss" style="color: #20794D;
background-color: null;
font-style: inherit;">f"Perform Named Entity Recognition that finds the following entities: </span><span class="sc" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">{</span><span class="st" style="color: #20794D;
background-color: null;
font-style: inherit;">', '</span><span class="sc" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">.</span>join([x.name <span class="cf" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">for</span> x <span class="kw" style="color: #003B4F;
background-color: null;
font-weight: bold;
font-style: inherit;">in</span> NamedEntityType])<span class="sc" style="color: #5E5E5E;
background-color: null;
font-style: inherit;">}</span><span class="ss" style="color: #20794D;
background-color: null;
font-style: inherit;">"</span>,</span>
<span id="cb4-53"> )</span></code></pre></div>
<p>This improves those cases where, for example, the LLM wants to define the entity type as “ORGANIZATION” but it is defined in the <code>Enum</code> as “ORG”.</p>
<p>Another option potentially worth exploring is to use the <code>llm_validator</code> function to make a call out to the LLM when exceptions happen and prompt it to coerce the value into something in the <code>Enum</code>. This could hike up your costs a bit but I imagine using a cheap model like GPT-3.5-Turbo could do the job just fine, and would likely you give an addtional robustness in quality results.</p>
</section>
<section id="conclusion" class="level2">
<h2 class="anchored" data-anchor-id="conclusion">Conclusion</h2>
<p>That’s it.</p>
Expand Down
Loading

0 comments on commit decdabc

Please sign in to comment.