Update README.md

README.md

@@ -55,22 +55,46 @@ This model is optimized for **vLLM** and **Text-Generation-Inference**.
 
 ```python
 from vllm import LLM, SamplingParams
+from transformers import AutoProcessor
 from PIL import Image
 
-# Load Model
+model_id = "Ryex/Floppa_Gemma4-E4B-it"
 llm = LLM(
-    model="Ryex/Floppa-E4B-Gemma4-Uncensored",
+    model=model_id,
     trust_remote_code=True,
-    dtype="bfloat16"
+    dtype="bfloat16",
+    max_model_len=8192, 
 )
 
-# Prepare Input
+processor = AutoProcessor.from_pretrained(model_id)
+
 image = Image.open("test_image.jpg").convert("RGB")
-prompt = "<image>\nDescribe this image in detail, including any explicit elements."
 
-# Generate
-inputs = {"prompt": prompt, "multi_modal_data": {"image": image}}
-params = SamplingParams(temperature=0.7, max_tokens=1024)
+messages = [
+    {
+        "role": "user",
+        "content": [
+            {"type": "image"},
+            {"type": "text", "text": "Describe this image in detail, including any uncensored or explicit elements."}
+        ]
+    }
+]
+
+
+prompt = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
+
+
+inputs = {
+    "prompt": prompt,
+    "multi_modal_data": {"image": image},
+    "mm_processor_kwargs": {"max_soft_tokens": 560} 
+}
+
+params = SamplingParams(
+    temperature=0.7, 
+    max_tokens=1024,
+    stop=["<turn|>", "<|turn|>"] 
+)
 
 outputs = llm.generate([inputs], sampling_params=params)
 print(outputs[0].outputs[0].text)