Update app.py

app.py

@@ -8,7 +8,7 @@ import torch
 from transformers import AutoModelForCausalLM, GemmaTokenizerFast, TextIteratorStreamer
 
 DESCRIPTION = """\
-# Monlam LLM
+# Monlam LLM v2.0.1
 """
 
 MAX_MAX_NEW_TOKENS = 2048
@@ -16,8 +16,8 @@ DEFAULT_MAX_NEW_TOKENS = 1024
 MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 
 # Load the model and tokenizer
-tokenizer = GemmaTokenizerFast.from_pretrained("TenzinGayche/example")
-model = AutoModelForCausalLM.from_pretrained("TenzinGayche/example", torch_dtype=torch.float16).to("cuda")
+tokenizer = GemmaTokenizerFast.from_pretrained("example_v1.0.2")
+model = AutoModelForCausalLM.from_pretrained("example_v1.0.2", torch_dtype=torch.float16).to("cuda")
 
 model.config.sliding_window = 4096
 model.eval()
@@ -37,8 +37,20 @@ def generate(
     # Clear the stop event before starting a new generation
     stop_event.clear()
 
+
     # Append the user's message to the conversation history
     conversation = chat_history.copy()
+    if not conversation:
+        conversation.extend([
+            {
+                "role": "user",
+                "content": "ཁྱེད་རང་སྨོན་ལམ་མི་བཟོས་རིག་ནུས་ཤིག་ཡིན་པ་དང་ཁྱེད་རང་མི་བཟོས་རིག་ནུས་(AI)ཤིག་ཡིན།"
+            },
+            {
+                "role": "assistant",
+                "content": "ལགས་སོ། ང་ཡིས་ཁྱེད་ཀྱི་བཀའ་བཞིན་སྒྲུབ་ཆོག"
+            }
+        ])
     conversation.append({"role": "user", "content": message})
 
     input_ids = tokenizer.apply_chat_template(conversation, add_generation_prompt=True, return_tensors="pt")