Corrected the readme to be compliant with the standard HF MM pipeline

README.md

@@ -159,20 +159,14 @@ pixel_values = [pixel_values]
 
 # generate output
 with torch.inference_mode():
-    gen_kwargs = dict(
-        max_new_tokens=1024,
-        do_sample=False,
-        top_p=None,
-        top_k=None,
-        temperature=None,
-        repetition_penalty=None,
-        eos_token_id=model.generation_config.eos_token_id,
-        pad_token_id=text_tokenizer.pad_token_id,
-        use_cache=True
-    )
-    output_ids = model.generate(input_ids, pixel_values=pixel_values, attention_mask=attention_mask, **gen_kwargs)[0]
+    if inputs['pixel_values'] is not None:
+        inputs['pixel_values'] = [pix.to(model.dtype).to(model.device) for pix in inputs['pixel_values']]
+    inputs = inputs.to('cuda')
+
+    output_ids = model.generate(inputs =inputs.pop('input_ids'), **inputs)[0]
     output = text_tokenizer.decode(output_ids, skip_special_tokens=True)
     print(f'Output:\n{output}')
+
 ```
 
 <details>