added card

app.py

@@ -8,8 +8,7 @@ import gradio as gr
 from huggingface_hub import HfApi
 from huggingface_hub import ModelCard
 
-from gradio_huggingfacehub_search import HuggingfaceHubSearch
-from apscheduler.schedulers.background import BackgroundScheduler
+from textwrap import dedent
 
 HF_PATH = "https://huggingface.co/"
 
@@ -92,12 +91,44 @@ def button_click(hf_model_id, conv_template, quantization, oauth_token: gr.OAuth
     
     # push to HF
     user_name = api.whoami()["name"]
-    api.create_repo(repo_id=f"{user_name}/{mlc_model_name}", private=True)
+    created_repo_url = api.create_repo(repo_id=f"{user_name}/{mlc_model_name}", private=True)
+    created_repo_id = created_repo_url.repo_id
 
     api.upload_large_folder(folder_path=f"./dist/{mlc_model_name}",
-                  repo_id=f"{user_name}/{mlc_model_name}",
-                  repo_type="model")
+                            repo_id=f"{user_name}/{mlc_model_name}",
+                            repo_type="model")
     
+    # push model card to HF
+    card = ModelCard.load(hf_model_id, token=oauth_token.token)
+    if not card.data.tags:
+        card.data.tags = []
+    card.data.tags.append("mlc-ai")
+    card.data.tags.append("MLC-Weight-Conversion")
+    card.data.base_model = hf_model_id
+
+    card.text = dedent(
+        f"""
+        # {created_repo_id}
+        This model was compiled using MLC-LLM with {quantization} quantization from [{hf_model_id}]({HF_PATH}{hf_model_id}).
+        The conversion was done using the [MLC-Weight-Conversion](https://huggingface.co/spaces/mlc-ai/MLC-Weight-Conversion) space.
+
+        To run this model, please first install [MLC-LLM](https://llm.mlc.ai/docs/install/mlc_llm.html#install-mlc-packages).
+
+        To chat with the model on your terminal:
+        ```bash
+        mlc_llm chat HF://{created_repo_id}
+        ```
+
+        For more information on how to use MLC-LLM, please visit the MLC-LLM [documentation](https://llm.mlc.ai/docs/index.html).
+        """
+    )
+    card.save("./dist/README.md")
+
+    api.upload_file(path_or_fileobj="./dist/README.md",
+                    path_in_repo="README.md",
+                    repo_id=created_repo_id,
+                    repo_type="model")
+
     os.system("rm -rf dist/")
     
     return "Successful"