Merge pull request #78 from Luodian/cliangyu-patch-1

Cliangyu patch 1 Former-commit-id: ce541ab05729eb8c79922bb950cdbe7793dca591
Luodian · Apr 30, 2023 · d52e6e2 · d52e6e2
2 parents 676a748 + 5594b7d
commit d52e6e2
Show file tree

Hide file tree

Showing 3 changed files with 4 additions and 4 deletions.
diff --git a/README.md b/README.md
@@ -101,7 +101,7 @@ Our Otter model is also developed in this way and it's deployed on the 🤗 Hugg
 
 ## 🗄 Dataset Preparation
 
-### Multi-model instruction tuning dataset with in-context examples (ICI)
+### Multi-modal instruction tuning dataset with in-context examples (ICI)
 
 The pre-training process for the OpenFlamingo model employs the MMC4 interleaved multimodality dataset to endow the model with in-context few-shot learning capabilities. The development of our instruction-following dataset adheres to the guiding principles of MMC4, which dictate that the instruction and image examples incorporated into the context should exhibit semantic pertinence to the query instruction and image.
 

diff --git a/docs/model_card.md b/docs/model_card.md
@@ -8,7 +8,7 @@ datasets:
 
 [Code](https://github.com/Luodian/PET-VLM) | [Demo](https://otter.cliangyu.com/)
 
-Otter is an instruction-following large multi-model model built upon Open-Flamingo-9B. Through in-context instruction following, Otter is able to perform tasks more aligned with human preferences and more accurate.
+Otter is an instruction-following large multi-modal model built upon Open-Flamingo-9B. Through in-context instruction following, Otter is able to perform tasks more aligned with human preferences and more accurate.
 
 ## Model Details
 

diff --git a/pipeline/serve/gradio_web_server.py b/pipeline/serve/gradio_web_server.py
@@ -458,7 +458,7 @@ def http_bot(
 
 <h3><a href="https://github.com/Luodian/otter"><img src="https://upload.wikimedia.org/wikipedia/commons/9/91/Octicons-mark-github.svg" style="height: 15px; display:inline;" class="icon" alt="github">GitHub</a>
 &nbsp;&nbsp;&nbsp;
-<a href="https://youtu.be/PYHEhHSZFg8"><img src="https://www.svgrepo.com/show/13671/youtube.svg" style="height: 15px; display:inline;" class="icon" alt="video demo">Video Demo</a>
+<a href="https://youtu.be/ys4Hwh86vM0"><img src="https://www.svgrepo.com/show/13671/youtube.svg" style="height: 15px; display:inline;" class="icon" alt="video demo">Video Demo</a>
 </h3>
 
 | Choose a model to chat with | |
@@ -625,7 +625,7 @@ def build_demo(embed_mode):
         gr.Examples(
             examples=[
                 [f"{cur_dir}/examples/cat.jpg", "An image of", "two cats.", f"{cur_dir}/examples/bathroom.jpg", "An image of", "a bathroom sink.", f"{cur_dir}/examples/dinner.jpg", "An image of"],
-                [f"{cur_dir}/examples/baseball.jpg", "What is the potential danger of this sport?", "bumped into others while chasing ball.", f"{cur_dir}/examples/tennis.jpg", "What is the potential danger of this sport?", "hitted by the tennis ball.", f"{cur_dir}/examples/soccer.png", "What is the potential danger of this sport?"],
+                [f"{cur_dir}/examples/tennis.jpg", "What is the danger of this sport?", "The player may get hitted by the tennis ball.", f"{cur_dir}/examples/baseball.jpg", "What is the danger of this sport?", "While chasing the baseball, the player may inadvertently collide with other players.",  f"{cur_dir}/examples/soccer.png", "What is the danger of this sport?"],
             ],
             inputs=[
                 imagebox_demo_1,