Training in progress, step 395

config.json

@@ -0,0 +1,189 @@
+{
+  "_name_or_path": "dandelin/vilt-b32-finetuned-vqa",
+  "architectures": [
+    "ViltForQuestionAnswering"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "Tunics",
+    "1": "Watches",
+    "2": "Winter",
+    "3": "Belts",
+    "4": "Heels",
+    "5": "Formal",
+    "6": "Pink",
+    "7": "Red",
+    "8": "White",
+    "9": "Shorts",
+    "10": "Grey",
+    "11": "Copper",
+    "12": "Sports Sandals",
+    "13": "Tops",
+    "14": "Sweatshirts",
+    "15": "Mustard",
+    "16": "Ethnic",
+    "17": "Sports Shoes",
+    "18": "Taupe",
+    "19": "Sports",
+    "20": "Multi",
+    "21": "Jeans",
+    "22": "Socks",
+    "23": "Green",
+    "24": "Bracelet",
+    "25": "Men",
+    "26": "Trousers",
+    "27": "Maroon",
+    "28": "Boys",
+    "29": "Unisex",
+    "30": "Sunglasses",
+    "31": "Yellow",
+    "32": "Steel",
+    "33": "Purple",
+    "34": "Peach",
+    "35": "Formal Shoes",
+    "36": "Bra",
+    "37": "Magenta",
+    "38": "Spring",
+    "39": "Ties",
+    "40": "Necklace and Chains",
+    "41": "Beige",
+    "42": "Briefs",
+    "43": "Flip Flops",
+    "44": "Casual",
+    "45": "Scarves",
+    "46": "Handbags",
+    "47": "Nightdress",
+    "48": "Gold",
+    "49": "Shirts",
+    "50": "Khaki",
+    "51": "Women",
+    "52": "Sweaters",
+    "53": "Gloves",
+    "54": "Navy Blue",
+    "55": "Brown",
+    "56": "Flats",
+    "57": "Cream",
+    "58": "Tan",
+    "59": "Tshirts",
+    "60": "Black",
+    "61": "Blue",
+    "62": "Track Pants",
+    "63": "Fall",
+    "64": "Sandals",
+    "65": "Jumpsuit",
+    "66": "Jeggings",
+    "67": "Summer",
+    "68": "Night suits",
+    "69": "Casual Shoes",
+    "70": "Off White",
+    "71": "Charcoal",
+    "72": "Innerwear Vests",
+    "73": "Stockings",
+    "74": "Silver",
+    "75": "Orange",
+    "76": "Girls",
+    "77": "Caps"
+  },
+  "image_size": 384,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "Beige": 41,
+    "Belts": 3,
+    "Black": 60,
+    "Blue": 61,
+    "Boys": 28,
+    "Bra": 36,
+    "Bracelet": 24,
+    "Briefs": 42,
+    "Brown": 55,
+    "Caps": 77,
+    "Casual": 44,
+    "Casual Shoes": 69,
+    "Charcoal": 71,
+    "Copper": 11,
+    "Cream": 57,
+    "Ethnic": 16,
+    "Fall": 63,
+    "Flats": 56,
+    "Flip Flops": 43,
+    "Formal": 5,
+    "Formal Shoes": 35,
+    "Girls": 76,
+    "Gloves": 53,
+    "Gold": 48,
+    "Green": 23,
+    "Grey": 10,
+    "Handbags": 46,
+    "Heels": 4,
+    "Innerwear Vests": 72,
+    "Jeans": 21,
+    "Jeggings": 66,
+    "Jumpsuit": 65,
+    "Khaki": 50,
+    "Magenta": 37,
+    "Maroon": 27,
+    "Men": 25,
+    "Multi": 20,
+    "Mustard": 15,
+    "Navy Blue": 54,
+    "Necklace and Chains": 40,
+    "Night suits": 68,
+    "Nightdress": 47,
+    "Off White": 70,
+    "Orange": 75,
+    "Peach": 34,
+    "Pink": 6,
+    "Purple": 33,
+    "Red": 7,
+    "Sandals": 64,
+    "Scarves": 45,
+    "Shirts": 49,
+    "Shorts": 9,
+    "Silver": 74,
+    "Socks": 22,
+    "Sports": 19,
+    "Sports Sandals": 12,
+    "Sports Shoes": 17,
+    "Spring": 38,
+    "Steel": 32,
+    "Stockings": 73,
+    "Summer": 67,
+    "Sunglasses": 30,
+    "Sweaters": 52,
+    "Sweatshirts": 14,
+    "Tan": 58,
+    "Taupe": 18,
+    "Ties": 39,
+    "Tops": 13,
+    "Track Pants": 62,
+    "Trousers": 26,
+    "Tshirts": 59,
+    "Tunics": 0,
+    "Unisex": 29,
+    "Watches": 1,
+    "White": 8,
+    "Winter": 2,
+    "Women": 51,
+    "Yellow": 31
+  },
+  "layer_norm_eps": 1e-12,
+  "max_image_length": -1,
+  "max_position_embeddings": 40,
+  "modality_type_vocab_size": 2,
+  "model_type": "vilt",
+  "num_attention_heads": 12,
+  "num_channels": 3,
+  "num_hidden_layers": 12,
+  "num_images": -1,
+  "patch_size": 32,
+  "qkv_bias": true,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.47.1",
+  "type_vocab_size": 2,
+  "vocab_size": 30522
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a01d60e112be35609927bbbb27c083b9857e4b510f08086505136450cc997a0d
+size 451621416

preprocessor_config.json

@@ -0,0 +1,24 @@
+{
+  "do_normalize": true,
+  "do_pad": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "ViltImageProcessor",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "processor_class": "ViltProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "shortest_edge": 384
+  },
+  "size_divisor": 32
+}

special_tokens_map.json

@@ -0,0 +1,37 @@
+{
+  "cls_token": {
+    "content": "[CLS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "[MASK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "[SEP]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,57 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "extra_special_tokens": {},
+  "mask_token": "[MASK]",
+  "model_max_length": 40,
+  "pad_token": "[PAD]",
+  "processor_class": "ViltProcessor",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14d27c84772e1532c0474dee4bb4100a2a6dcd2c967074730c2a59cfd18762c1
+size 5368