fix import errors

README.md

@@ -60,9 +60,10 @@ import json
 import PIL
 import IPython.display as display
 import torch
-model = AutoModelForVision2Seq.from_pretrained("./", trust_remote_code=True)
-tokenizer = AutoTokenizer.from_pretrained("./", trust_remote_code=True, use_fast=True, legacy=False)
-image_processor = AutoImageProcessor.from_pretrained("./", trust_remote_code=True)
+model_name_or_path = "Salesforce/xgen-mm-phi3-mini-base-r-v1"
+model = AutoModelForVision2Seq.from_pretrained(model_name_or_path, trust_remote_code=True)
+tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=True, use_fast=True, legacy=False)
+image_processor = AutoImageProcessor.from_pretrained(model_name_or_path, trust_remote_code=True)
 tokenizer = model.update_special_tokens(tokenizer)
 
 model = model.to('cuda')

modeling_xgenmm.py

@@ -3,8 +3,8 @@ import torch
 import open_clip
 from typing import List, Optional, Tuple, Union
 from utils import check_embedding_fns
-from vlm import PerceiverResampler, Kosmos
-from configuration_xgenmm import XGenMMVisionEncoderConfig, XGenMMVisionTokenizerConfig, XGenMMConfig
+from .vlm import PerceiverResampler, Kosmos
+from .configuration_xgenmm import XGenMMVisionEncoderConfig, XGenMMVisionTokenizerConfig, XGenMMConfig
 
 class XGenMMVisionEncoder(PreTrainedModel):
     main_input_name = "pixel_values"

vlm.py

@@ -11,7 +11,7 @@ from dataclasses import dataclass
 from transformers import CLIPVisionModel
 import transformers
 
-from utils import num_params, getattr_recursive, stack_with_padding, get_anyres_image_grid_shape, unpad_image
+from .utils import num_params, getattr_recursive, stack_with_padding, get_anyres_image_grid_shape, unpad_image
 
 
 class VisionTokenizer(nn.Module):