Update README.md

README.md

@@ -42,8 +42,8 @@ base_model = AutoModelForCausalLM.from_pretrained(
 )
 
 # Load LoRA adapters
-model = PeftModel.from_pretrained(base_model, "coolAI/precis-granite")
-tokenizer = AutoTokenizer.from_pretrained("coolAI/precis-granite")
+model = PeftModel.from_pretrained(base_model, "cernis-intelligence/precis")
+tokenizer = AutoTokenizer.from_pretrained("cernis-intelligence/precis")
 
 # Generate summary
 document = """Your long document here..."""
@@ -77,7 +77,7 @@ print(summary)
 from unsloth import FastLanguageModel
 
 model, tokenizer = FastLanguageModel.from_pretrained(
-    model_name="coolAI/precis-granite",
+    model_name="cernis-intelligence/precis",
     max_seq_length=2048,
     load_in_4bit=True,  # For lower memory usage
 )
@@ -117,7 +117,7 @@ llm = LLM(
 lora_request = LoRARequest(
     "precis-granite",
     1,
-    "coolAI/precis-granite"
+    "cernis-intelligence/precis"
 )
 
 # Sampling parameters