Adding ONNX file of this model

README.md

@@ -4,6 +4,7 @@ tags:
 - donut
 - image-to-text
 - vision
+- onnx
 ---
 
 # Donut (base-sized model, pre-trained only) 

onnx/config.json

@@ -0,0 +1,76 @@
+{
+  "_attn_implementation_autoset": true,
+  "architectures": [
+    "VisionEncoderDecoderModel"
+  ],
+  "decoder": {
+    "activation_dropout": 0.0,
+    "activation_function": "gelu",
+    "add_cross_attention": true,
+    "add_final_layer_norm": true,
+    "attention_dropout": 0.0,
+    "classifier_dropout": 0.0,
+    "d_model": 1024,
+    "decoder_attention_heads": 16,
+    "decoder_ffn_dim": 4096,
+    "decoder_layerdrop": 0.0,
+    "decoder_layers": 4,
+    "dropout": 0.1,
+    "encoder_attention_heads": 16,
+    "encoder_ffn_dim": 4096,
+    "encoder_layerdrop": 0.0,
+    "encoder_layers": 12,
+    "init_std": 0.02,
+    "is_decoder": true,
+    "is_encoder_decoder": false,
+    "max_position_embeddings": 1536,
+    "model_type": "mbart",
+    "num_hidden_layers": 12,
+    "scale_embedding": true,
+    "torch_dtype": "float32",
+    "use_cache": true,
+    "vocab_size": 57525
+  },
+  "encoder": {
+    "attention_probs_dropout_prob": 0.0,
+    "depths": [
+      2,
+      2,
+      14,
+      2
+    ],
+    "drop_path_rate": 0.1,
+    "embed_dim": 128,
+    "hidden_act": "gelu",
+    "hidden_dropout_prob": 0.0,
+    "hidden_size": 1024,
+    "image_size": [
+      2560,
+      1920
+    ],
+    "initializer_range": 0.02,
+    "layer_norm_eps": 1e-05,
+    "mlp_ratio": 4.0,
+    "model_type": "donut-swin",
+    "num_channels": 3,
+    "num_heads": [
+      4,
+      8,
+      16,
+      32
+    ],
+    "num_layers": 4,
+    "patch_size": 4,
+    "path_norm": true,
+    "qkv_bias": true,
+    "torch_dtype": "float32",
+    "use_absolute_embeddings": false,
+    "window_size": 10
+  },
+  "is_encoder_decoder": true,
+  "model_type": "vision-encoder-decoder",
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.51.3",
+  "use_cache": true
+}

onnx/decoder_model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd37227502254300630a953b3569abd626cb2d297ac94fc36d1886de9f3ea397
+size 746470076

onnx/decoder_model_merged.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f3bafd3645a19b8af415dc6df2c610f28ea54d5923a152e6c0acfe1ad4cc22e
+size 746611433

onnx/decoder_with_past_model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fecb41ab42c3993d42f5db6677d0202ad6ab643276f61ad9f320f943b256fd0
+size 712860789

onnx/encoder_model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db85ca838c171661796504ff208bd2e0fb1040ed06f82cc55017297cfe22260e
+size 311378378

onnx/generation_config.json

@@ -0,0 +1,8 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 2,
+  "forced_eos_token_id": 2,
+  "pad_token_id": 1,
+  "transformers_version": "4.51.3"
+}

onnx/preprocessor_config.json

@@ -0,0 +1,26 @@
+{
+  "do_align_long_axis": true,
+  "do_normalize": true,
+  "do_pad": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "do_thumbnail": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "DonutImageProcessor",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "processor_class": "DonutProcessor",
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 2560,
+    "width": 1920
+  }
+}