Update configs for most recent transformers package

config.json

@@ -4,7 +4,9 @@
   ],
   "do_image_splitting": true,
   "downsample_factor": 2,
+  "dtype": "bfloat16",
   "encoder_patch_size": 16,
+  "image_token_id": 396,
   "image_token_index": 396,
   "max_image_tokens": 256,
   "max_num_patches": 1024,
@@ -36,6 +38,7 @@
     "conv_dim": 2048,
     "conv_dim_out": 2048,
     "conv_use_xavier_init": true,
+    "dtype": "bfloat16",
     "eos_token_id": 7,
     "hidden_size": 2048,
     "initializer_range": 0.02,
@@ -66,18 +69,17 @@
     "num_hidden_layers": 16,
     "num_key_value_heads": 8,
     "rope_theta": 1000000.0,
-    "torch_dtype": "bfloat16",
     "use_cache": true,
     "use_pos_enc": true,
     "vocab_size": 65536
   },
   "tile_size": 512,
-  "torch_dtype": "bfloat16",
-  "transformers_version": "4.55.0",
+  "transformers_version": "4.57.1",
   "use_image_special_tokens": true,
   "use_thumbnail": true,
   "vision_config": {
     "attention_dropout": 0.0,
+    "dtype": "bfloat16",
     "hidden_act": "gelu_pytorch_tanh",
     "hidden_size": 1152,
     "intermediate_size": 4304,
@@ -88,7 +90,6 @@
     "num_hidden_layers": 26,
     "num_patches": 256,
     "patch_size": 16,
-    "torch_dtype": "bfloat16",
     "vision_use_head": false
   }
 }

preprocessor_config.json

@@ -1,11 +1,18 @@
 {
-  "auto_map": {
-    "AutoProcessor": "processing_lfm2_vl.Lfm2VlProcessor"
-  },
+  "crop_size": null,
+  "data_format": "channels_first",
+  "default_to_square": true,
+  "device": null,
+  "disable_grouping": null,
+  "do_center_crop": null,
   "do_convert_rgb": null,
+  "do_image_splitting": true,
   "do_normalize": true,
+  "do_pad": true,
   "do_rescale": true,
   "do_resize": false,
+  "downsample_factor": 2,
+  "encoder_patch_size": 16,
   "image_mean": [
     0.5,
     0.5,
@@ -17,10 +24,24 @@
     0.5,
     0.5
   ],
+  "input_data_format": null,
+  "max_image_tokens": 256,
   "max_num_patches": 1024,
+  "max_pixels_tolerance": 2.0,
+  "max_tiles": 10,
+  "min_image_tokens": 64,
+  "min_tiles": 2,
+  "pad_size": null,
   "patch_size": 16,
   "processor_class": "Lfm2VlProcessor",
   "resample": 2,
   "rescale_factor": 0.00392156862745098,
-  "return_tensors": "pt"
+  "return_row_col_info": true,
+  "return_tensors": null,
+  "size": {
+    "height": 512,
+    "width": 512
+  },
+  "tile_size": 512,
+  "use_thumbnail": true
 }

processor_config.json

@@ -1,18 +1,4 @@
 {
-  "auto_map": {
-    "AutoProcessor": "processing_lfm2_vl.Lfm2VlProcessor"
-  },
-  "do_image_splitting": true,
-  "downsample_factor": 2,
-  "encoder_patch_size": 16,
-  "max_image_tokens": 256,
-  "max_num_patches": 1024,
-  "max_pixels_tolerance": 1.5,
-  "max_tiles": 10,
-  "min_image_tokens": 64,
-  "min_tiles": 2,
   "processor_class": "Lfm2VlProcessor",
-  "tile_size": 512,
-  "use_image_special_tokens": true,
-  "use_thumbnail": true
+  "use_image_special_tokens": true
 }