Fixing link to RAT, and minor mods to OpenFunctions

ShishirPatil · Nov 28, 2023 · ac5bd97 · ac5bd97
1 parent d0c954f
commit ac5bd97
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 3 deletions.
diff --git a/blog.html b/blog.html
@@ -285,7 +285,7 @@ <h2 class="blog-title">Reality Bytes: How to Measure Hallucinations in LLMs</h2>
 
         <!-- Blog Post 3 -->
         <div class="blog-post">
-            <a href="blogs/3_retriever_ware_training.html">
+            <a href="blogs/3_retriever_aware_training.html">
                 <div class="blog-content">
                     <h2 class="blog-title">Retriever-Aware Training (RAT): Are LLMs memorizing or understanding?</h2>
                     <div class="author-date">

diff --git a/blogs/4_open_functions.html b/blogs/4_open_functions.html
@@ -97,7 +97,7 @@ <h2 class="blog-title">Gorilla OpenFunctions</h2>
                         <ul>
                             <li>Colab Notebook: <a href="https://colab.research.google.com/drive/16M5J2H9F8YQora_W2PDnp120slZH-Mqd?usp=sharing">OpenFunctions</a></li>
                             <li>Dataset: <a href="https://github.com/ShishirPatil/gorilla/tree/main/openfunctions">GitHub</a></li>
-                            <li>Models on HuggingFace: <a href="https://huggingface.co/gorilla-llm/gorilla-openfunctions-v1">gorilla-llm/gorilla-openfunctions-v1</a></li>
+                            <li>Models on HuggingFace: <a href="https://huggingface.co/gorilla-llm/gorilla-openfunctions-v0">gorilla-llm/gorilla-openfunctions-v0</a></li>
                         </ul>
                     </p>
                 </div>
@@ -214,7 +214,7 @@ <h4 id="benchmarking">OpenFunctions Performance Benchmarking</h4>
                         </i>
                         <br>
                     <p> When calling to OpenAI's function calling models, if the required parameters are not supplied within the instruction, this leads the function calling models to output "follow-up" questions requesting the required parameters. This results in "incomplete" status as displayed in above graph. 
-                        We treat "incomplete" execution as "success" during our accuracy calculation because the model recognized the missing parameters successfully. 
+                        We treat "incomplete" execution as "success" during our accuracy calculation because the model recognized the missing parameters successfully. Note that this is consistent across all evaluations. 
                         Our OpenFunctions model, as well as regular GPT-4, due to its chat completion nature, will fill up the required parameters with placeholders or default values, allowing undisturbed generation. </p>
 
                 </div>