fix bugs with model token

app.py

@@ -100,6 +100,7 @@ def validate_request(request: GenerationRequest) -> GenerationRequest:
         request.input_dataset_split = f"{request.input_dataset_split}[:{request.num_output_examples}]"
 
 
+
     if request.num_output_examples > MAX_SAMPLES:
         raise Exception(f"Requested number of output examples {request.num_output_examples} exceeds the max limit of {MAX_SAMPLES}.")
 
@@ -107,12 +108,28 @@ def validate_request(request: GenerationRequest) -> GenerationRequest:
     if request.prompt_column not in input_dataset_info.features:
         raise Exception(f"Prompt column {request.prompt_column} does not exist in dataset {request.input_dataset_name}. Available columns: {list(input_dataset_info.features.keys())}")
 
+    # This is currently not supported, the output dataset will be created under the org 'synthetic-data-universe'
+    # check output_dataset name is valid
+    if request.output_dataset_name.count("/") != 1:
+        raise Exception("Output dataset name must be in the format 'dataset_name', e.g., 'my-dataset'. The dataset will be created under the org 'synthetic-data-universe/my-dataset'.")
+        
+    # check the output dataset is valid and accessible with the provided token
+    try:
+        output_dataset_info = get_dataset_infos(request.output_dataset_name, token=request.output_dataset_token)
+        raise Exception(f"Output dataset {request.output_dataset_name} already exists. Please choose a different name.")
+    except Exception as e:
+        pass  # dataset does not exist, which is expected
+
     # check the models exists
     try:
-        model_config = AutoConfig.from_pretrained(request.model_name_or_path, revision=request.model_revision, token=request.model_token)
+        model_config = AutoConfig.from_pretrained(request.model_name_or_path, 
+                                                  revision=request.model_revision,     
+                                                  force_download=True,
+                                                  token=False
+        )
     except Exception as e:
         print(e)
-        raise Exception(f"Model {request.model_name_or_path} revision {request.model_revision} does not exist or cannot be accessed with the provided token.")
+        raise Exception(f"Model {request.model_name_or_path} revision {request.model_revision} does not exist or cannot be accessed. The model may be private or gated, which is not supported at this time.")
 
     # check the model max position embeddings is greater than the requested max tokens and less than MAX_TOKENS    
     if model_config.max_position_embeddings < request.max_tokens:
@@ -265,7 +282,7 @@ def main():
                                 max_tok, temp, top_k_val, top_p_val, email_addr, num_output_samples):
 
                 MASTER_ORG = "synthetic-data-universe/"
-                model_token = None # This is currently not supported
+                model_token = False # This is currently not supported
                 input_dataset_token = None # This is currently not supported
                 output_dataset_token = os.getenv("OUTPUT_DATASET_TOKEN")
                 
@@ -281,7 +298,7 @@ def main():
                         prompt_column=prompt_col,
                         model_name_or_path=model_name,
                         model_revision=model_rev,
-                        model_token=model_token if model_token else None,
+                        model_token=model_token,
                         system_prompt=sys_prompt if sys_prompt else None,
                         max_tokens=int(max_tok),
                         temperature=temp,
@@ -319,10 +336,12 @@ def main():
                     "To unlock this and many other cool stuff, please consider upgrading your account.\n\n"
                     "### [**Become a PRO Today!**](http://huggingface.co/subscribe/pro?source=synthetic-data-universe)"
                 )
-                return gr.update(visible=False), gr.update(visible=True, value=message)    
-        
-        demo.load(control_access, inputs=None, outputs=[main_interface, pro_message]) 
-        demo.queue(max_size=None, default_concurrency_limit=None).launch(show_error=True)  
+                return gr.update(visible=False), gr.update(visible=True, value=message)
+
+        login_button = gr.LoginButton() # this is required or AUTH will not work
+
+        demo.load(control_access, inputs=None, outputs=[main_interface, pro_message])
+        demo.queue(max_size=None, default_concurrency_limit=None).launch(show_error=True)
 
 if __name__ == "__main__":
     main()