Upgrade to transformers==4.36.1; q_align==1.1.0

Author: haoning.wu

README.md

@@ -59,16 +59,33 @@
 </div>
 </div> 
 
-## [Important Note!]
+## [Important Note!] v1.1 <-> transformers==4.36.1
 
+This LLaVA-style repository has been built on `transformers==4.31.0`, which is incompatible with many new models available on Hugging Face. This requires to build a separate environment for the MLLM/LMM repository, which is somewhat troublesome for this **visual scoring** model, as we expect the **Q-Align/OneAlign** to effectively boost other disciplines (image/video generation, *etc*). Both the repository and the **AutoModel** (as follows) are updated to the newest version. 
 
-Thus, we have modified respective code for mPLUG-Owl2 to adapt it to the newest transformer version, i.e. `transformers==4.36.1`, so that you do not need to create a separate outdated environment while using it alongside other projects. The updated code is no longer compatible with the old-version Q-Align (v1.0.1/v1.0.0, and before), please update to the newest version via the following scripts:
+To this end, we have modified respective code for mPLUG-Owl2 to adapt it to the newest transformer version, i.e. `transformers==4.36.1`, so that you do not need to create a separate outdated environment while using it alongside other projects. The updated code is no longer compatible with the old-version Q-Align (v1.0.1/v1.0.0, and before), please update to the newest version via the following scripts:
 
 ```shell
 git pull
 pip install -e .
 ```
 
+## [Installation Free!] Quicker Start with HuggingFace AutoModel
+
+No need to install this GitHub repo. 
+
+```python
+import requests
+import torch
+from transformers import AutoModelForCausalLM
+
+model = AutoModelForCausalLM.from_pretrained("q-future/one-align", trust_remote_code=True, 
+                                             torch_dtype=torch.float16, device_map="auto")
+
+from PIL import Image
+model.score([Image.open(requests.get("https://raw.githubusercontent.com/Q-Future/Q-Align/main/fig/singapore_flyer.jpg",
+                                     stream=True).raw)], task_="quality", input_="image") # task_ : quality | aesthetics; # input_: image | video
+```
 
 ## Installation
 
@@ -88,22 +105,7 @@ pip install flash_attn --no-build-isolation
 ```
 
 
-## Quicker Start with HuggingFace AutoModel
 
-No need to install this GitHub repo. 
-
-```python
-import requests
-import torch
-from transformers import AutoModelForCausalLM
-
-model = AutoModelForCausalLM.from_pretrained("q-future/one-align", trust_remote_code=True, 
-                                             torch_dtype=torch.float16, device_map="auto")
-
-from PIL import Image
-model.score([Image.open(requests.get("https://raw.githubusercontent.com/Q-Future/Q-Align/main/fig/singapore_flyer.jpg",
-                                     stream=True).raw)], task_="quality", input_="image") # task_ : quality | aesthetics; # input_: image | video
-```
 ## Quick Start
 
 We have fixed the multi-GPU inference problem.