jayr014 commited on
Commit
20cd870
1 Parent(s): 63bd8a0

updating one line and adding in human eval to GPT4

Browse files
README.md CHANGED
@@ -48,7 +48,7 @@ Bloom chat should NOT be used for:
48
  - Mission-critical applications
49
  - Applications that involve the safety of others
50
  - Making highly important decisions
51
- - Important automated pipelines/decisions
52
 
53
  This model is still in early development and can be prone to mistakes and hallucinations, there is still room for improvement. This model is intended to provide the community with a good baseline.
54
 
@@ -162,9 +162,15 @@ In the nucleus of atoms, protons and neutrons are bound together in a structure
162
  </figure>
163
  <br>
164
 
 
 
 
 
 
 
165
  ![Multilingual evaluation](images/Multilingual_capabilities_comparison.png)
166
  <figure style="text-align:center;">
167
- <figcaption><b>BLOOMChat surpasses other Bloom variants and state-of-the-art open-source chat models in translation setting</b></figcaption>
168
  </figure>
169
  <br>
170
 
 
48
  - Mission-critical applications
49
  - Applications that involve the safety of others
50
  - Making highly important decisions
51
+ - Important automated pipelines
52
 
53
  This model is still in early development and can be prone to mistakes and hallucinations, there is still room for improvement. This model is intended to provide the community with a good baseline.
54
 
 
162
  </figure>
163
  <br>
164
 
165
+ ![Human evaluation against GPT4](images/Human_evaluation_gpt4.png)
166
+ <figure style="text-align:center;">
167
+ <figcaption><b>BLOOMChat vs GPT-4 in Human Preference Ranking</b></figcaption>
168
+ </figure>
169
+ <br>
170
+
171
  ![Multilingual evaluation](images/Multilingual_capabilities_comparison.png)
172
  <figure style="text-align:center;">
173
+ <figcaption><b>BLOOMChat surpasses other Bloom variants and state-of-the-art open-source chat models in translation tasks</b></figcaption>
174
  </figure>
175
  <br>
176
 
images/Human_evaluation.png CHANGED
images/Human_evaluation_gpt4.png ADDED