Training in progress, epoch 1

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

config.json

@@ -0,0 +1,270 @@
+{
+  "_name_or_path": "FacebookAI/xlm-roberta-large",
+  "architectures": [
+    "XLMRobertaForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "Lo\u1ea1i trang ph\u1ee5c",
+    "1": "O",
+    "2": "Tay \u00e1o",
+    "3": "Ki\u1ec3u khuy",
+    "4": "Gi\u1edbi t\u00ednh",
+    "5": "C\u1ed5 \u00e1o",
+    "6": "Ch\u1ea5t li\u1ec7u",
+    "7": "T\u00ednh n\u0103ng",
+    "8": "Phong c\u00e1ch",
+    "9": "M\u00e0u s\u1eafc",
+    "10": "Size",
+    "11": "Lo\u1ea1i \u0110\u1ed3 L\u00f3t",
+    "12": "Th\u01b0\u01a1ng hi\u1ec7u",
+    "13": "M\u1eabu",
+    "14": "Ki\u1ec3u Bra",
+    "15": "M\u00f9a",
+    "16": "D\u00e1ng ki\u1ec3u \u00e1o",
+    "17": "Lo\u1ea1i trang s\u1ee9c / ph\u1ee5 ki\u1ec7n",
+    "18": "Ch\u1ea5t li\u1ec7u d\u00e2y \u0111eo",
+    "19": "M\u1eb7t \u0111\u1ed3ng h\u1ed3",
+    "20": "H\u1ea1n b\u1ea3o h\u00e0nh",
+    "21": "Lo\u1ea1i k\u1ebft n\u1ed1i",
+    "22": "T\u00edch h\u1ee3p ph\u1ee5 ki\u1ec7n",
+    "23": "M\u1ea1ng kh\u00f4ng d\u00e2y",
+    "24": "Chi\u1ec1u cao",
+    "25": "C\u00e2n n\u1eb7ng",
+    "26": "H\u00ecnh th\u1ee9c s\u1ea3n ph\u1ea9m",
+    "27": "Lo\u1ea1i s\u1ea3n ph\u1ea9m l\u00e0m \u0111\u1eb9p",
+    "28": "Ch\u1ec9 s\u1ed1 ch\u1ed1ng n\u1eafng",
+    "29": "Th\u1ec3 t\u00edch",
+    "30": "V\u00f9ng c\u01a1 th\u1ec3",
+    "31": "Tr\u1ecdng l\u01b0\u1ee3ng",
+    "32": "Lo\u1ea1i th\u1ef1c ph\u1ea9m",
+    "33": "Ch\u1ebf \u0111\u1ed9 \u0103n u\u1ed1ng",
+    "34": "H\u01b0\u01a1ng v\u1ecb",
+    "35": "Th\u00e0nh ph\u1ea7n",
+    "36": "Lo\u1ea1i thu\u1ed1c & th\u1ef1c ph\u1ea9m ch\u1ee9c n\u0103ng",
+    "37": "Th\u00e0nh Ph\u1ea7n Ho\u1ea1t T\u00ednh",
+    "38": "Nh\u00f3m tu\u1ed5i",
+    "39": "\u0110\u1ed9 tu\u1ed5i khuy\u1ebfn ngh\u1ecb",
+    "40": "H\u1ea1n s\u1eed d\u1ee5ng",
+    "41": "H\u1ec7 \u0111i\u1ec1u h\u00e0nh",
+    "42": "K\u00edch th\u01b0\u1edbc m\u00e0n h\u00ecnh",
+    "43": "M\u00e0n h\u00ecnh",
+    "44": "Ki\u1ec3u k\u1ebft n\u1ed1i",
+    "45": "Dung l\u01b0\u1ee3ng",
+    "46": "Lo\u1ea1i thi\u1ebft b\u1ecb v\u00e0 linh ki\u1ec7n \u0111i\u1ec7n t\u1eed",
+    "47": "T\u00ecnh tr\u1ea1ng",
+    "48": "Xu\u1ea5t x\u1ee9",
+    "49": "RAM",
+    "50": "\u0110\u1ed9 ph\u00e2n gi\u1ea3i",
+    "51": "B\u1ed9 x\u1eed l\u00fd",
+    "52": "Lo\u1ea1i ph\u1ee5 ki\u1ec7n / linh ki\u1ec7n xe",
+    "53": "T\u1ed1c \u0111\u1ed9 x\u1eed l\u00fd",
+    "54": "Thi\u1ebft b\u1ecb t\u01b0\u01a1ng th\u00edch",
+    "55": "Lo\u1ea1i da",
+    "56": "Dung t\u00edch",
+    "57": "Lo\u1ea1i d\u1ee5ng c\u1ee5 & thi\u1ebft b\u1ecb s\u1eeda ch\u1eefa",
+    "58": "\u0110\u01b0\u1eddng k\u00ednh",
+    "59": "\u0110\u1ed9 d\u00e0y",
+    "60": "Chi\u1ec1u d\u00e0i",
+    "61": "Lo\u1ea1i v\u1eadt d\u1ee5ng l\u00e0m v\u01b0\u1eddn",
+    "62": "Lo\u1ea1i \u0111\u00e8n",
+    "63": "Ph\u01b0\u01a1ng th\u1ee9c ho\u1ea1t \u0111\u1ed9ng",
+    "64": "Lo\u1ea1i v\u1eadt d\u1ee5ng trong nh\u00e0 kh\u00e1c",
+    "65": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng v\u0103n ph\u00f2ng ph\u1ea9m",
+    "66": "K\u1ebft c\u1ea5u gi\u1ea5y",
+    "67": "K\u00edch th\u01b0\u1edbc gi\u1ea5y",
+    "68": "Tr\u1ecdng l\u01b0\u1ee3ng gi\u1ea5y",
+    "69": "Lo\u1ea1i gi\u00e0y d\u00e9p",
+    "70": "D\u1ecbp",
+    "71": "Lo\u1ea1i m\u0169i gi\u00e0y",
+    "72": "Lo\u1ea1i \u0111\u1ebf gi\u00e0y",
+    "73": "C\u1ed5 gi\u00e0y",
+    "74": "Lo\u1ea1i Kh\u00f3a",
+    "75": "Lo\u1ea1i ph\u1ee5 ki\u1ec7n \u00e2m nh\u1ea1c",
+    "76": "Lo\u1ea1i phi\u00ean b\u1ea3n",
+    "77": "N\u00e9t b\u00fat",
+    "78": "M\u1ef1c v\u00e0 t\u00f4ng m\u00e0u",
+    "79": "Lo\u1ea1i \u0111\u1ed3 ch\u0103m s\u00f3c s\u1ee9c kh\u1ecfe m\u1eb9",
+    "80": "Ki\u1ec3u \u0111\u00f3ng g\u00f3i",
+    "81": "Lo\u1ea1i thi\u1ebft b\u1ecb an to\u00e0n",
+    "82": "Ch\u1ee9c n\u0103ng l\u00e0m \u0111\u1eb9p",
+    "83": "Lo\u1ea1i \u0111\u1ed3 ch\u01a1i",
+    "84": "Handmade",
+    "85": "Lo\u1ea1i voucher & d\u1ecbch v\u1ee5",
+    "86": "Lo\u1ea1i m\u00f4n h\u1ecdc",
+    "87": "Lo\u1ea1i \u0111\u1ed3 v\u1eadt s\u01b0u t\u1ea7m",
+    "88": "Ki\u1ec3u th\u1ea7n t\u01b0\u1ee3ng",
+    "89": "Lo\u1ea1i thi\u1ebft b\u1ecb \u00e2m thanh",
+    "90": "Lo\u1ea1i \u0111\u0129a CD, DVD & Bluray",
+    "91": "Lo\u1ea1i thi\u1ebft b\u1ecb v\u00e0 d\u1ee5ng c\u1ee5 l\u00e0m \u0111\u1eb9p",
+    "92": "Lo\u1ea1i h\u00ecnh th\u1ec3 thao",
+    "93": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng kh\u00e1ch",
+    "94": "Lo\u1ea1i tr\u00f2 ch\u01a1i",
+    "95": "Th\u1ec3 lo\u1ea1i s\u00e1ch truy\u1ec7n",
+    "96": "C\u00f4ng su\u1ea5t",
+    "97": "Lo\u1ea1i thi\u1ebft b\u1ecb y t\u1ebf",
+    "98": "Lo\u1ea1i ph\u1ee5 ki\u1ec7n \u0111i\u1ec7n t\u1eed",
+    "99": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ch\u0103m s\u00f3c b\u00e9",
+    "100": "Chi\u1ec1u r\u1ed9ng",
+    "101": "Lo\u1ea1i th\u1ee9c \u0103n ch\u1ebf bi\u1ebfn s\u1eb5n",
+    "102": "Ki\u1ec3u d\u00e1ng v\u00e1y",
+    "103": "M\u1eabu xe t\u01b0\u01a1ng th\u00edch",
+    "104": "M\u00f9i h\u01b0\u01a1ng",
+    "105": "Lo\u1ea1i trang thi\u1ebft b\u1ecb th\u1ec3 thao",
+    "106": "Lo\u1ea1i ph\u1ee5 ki\u1ec7n nh\u00e0 b\u1ebfp",
+    "107": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng t\u1eafm",
+    "108": "Lo\u1ea1i v\u1eadt d\u1ee5ng ch\u0103m s\u00f3c c\u00e1 nh\u00e2n",
+    "109": "Lo\u1ea1i gi\u1ed1ng",
+    "110": "Lo\u1ea1i thi\u1ebft b\u1ecb \u0111i\u1ec7n gia d\u1ee5ng",
+    "111": "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng ng\u1ee7",
+    "112": "Lo\u1ea1i hoa",
+    "113": "Lo\u1ea1i \u0111\u1ed3 gia d\u1ee5ng b\u1ebfp",
+    "114": "Ch\u1ee9c n\u0103ng h\u1ed7 tr\u1ee3 s\u1ee9c kh\u1ecfe",
+    "115": "Lo\u1ea1i \u1ed1p",
+    "116": "Lo\u1ea1i tr\u00e1i c\u00e2y",
+    "117": "Lo\u1ea1i b\u1ed9 \u0111\u1ed3 b\u00e0n \u0103n",
+    "118": "Lo\u1ea1i v\u1eadt d\u1ee5ng h\u1ed3 c\u00e1"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "label2id": {
+    "B\u1ed9 x\u1eed l\u00fd": 51,
+    "Chi\u1ec1u cao": 24,
+    "Chi\u1ec1u d\u00e0i": 60,
+    "Chi\u1ec1u r\u1ed9ng": 100,
+    "Ch\u1ea5t li\u1ec7u": 6,
+    "Ch\u1ea5t li\u1ec7u d\u00e2y \u0111eo": 18,
+    "Ch\u1ebf \u0111\u1ed9 \u0103n u\u1ed1ng": 33,
+    "Ch\u1ec9 s\u1ed1 ch\u1ed1ng n\u1eafng": 28,
+    "Ch\u1ee9c n\u0103ng h\u1ed7 tr\u1ee3 s\u1ee9c kh\u1ecfe": 114,
+    "Ch\u1ee9c n\u0103ng l\u00e0m \u0111\u1eb9p": 82,
+    "C\u00e2n n\u1eb7ng": 25,
+    "C\u00f4ng su\u1ea5t": 96,
+    "C\u1ed5 gi\u00e0y": 73,
+    "C\u1ed5 \u00e1o": 5,
+    "Dung l\u01b0\u1ee3ng": 45,
+    "Dung t\u00edch": 56,
+    "D\u00e1ng ki\u1ec3u \u00e1o": 16,
+    "D\u1ecbp": 70,
+    "Gi\u1edbi t\u00ednh": 4,
+    "Handmade": 84,
+    "H\u00ecnh th\u1ee9c s\u1ea3n ph\u1ea9m": 26,
+    "H\u01b0\u01a1ng v\u1ecb": 34,
+    "H\u1ea1n b\u1ea3o h\u00e0nh": 20,
+    "H\u1ea1n s\u1eed d\u1ee5ng": 40,
+    "H\u1ec7 \u0111i\u1ec1u h\u00e0nh": 41,
+    "Ki\u1ec3u Bra": 14,
+    "Ki\u1ec3u d\u00e1ng v\u00e1y": 102,
+    "Ki\u1ec3u khuy": 3,
+    "Ki\u1ec3u k\u1ebft n\u1ed1i": 44,
+    "Ki\u1ec3u th\u1ea7n t\u01b0\u1ee3ng": 88,
+    "Ki\u1ec3u \u0111\u00f3ng g\u00f3i": 80,
+    "K\u00edch th\u01b0\u1edbc gi\u1ea5y": 67,
+    "K\u00edch th\u01b0\u1edbc m\u00e0n h\u00ecnh": 42,
+    "K\u1ebft c\u1ea5u gi\u1ea5y": 66,
+    "Lo\u1ea1i Kh\u00f3a": 74,
+    "Lo\u1ea1i b\u1ed9 \u0111\u1ed3 b\u00e0n \u0103n": 117,
+    "Lo\u1ea1i da": 55,
+    "Lo\u1ea1i d\u1ee5ng c\u1ee5 & thi\u1ebft b\u1ecb s\u1eeda ch\u1eefa": 57,
+    "Lo\u1ea1i gi\u00e0y d\u00e9p": 69,
+    "Lo\u1ea1i gi\u1ed1ng": 109,
+    "Lo\u1ea1i hoa": 112,
+    "Lo\u1ea1i h\u00ecnh th\u1ec3 thao": 92,
+    "Lo\u1ea1i k\u1ebft n\u1ed1i": 21,
+    "Lo\u1ea1i m\u00f4n h\u1ecdc": 86,
+    "Lo\u1ea1i m\u0169i gi\u00e0y": 71,
+    "Lo\u1ea1i phi\u00ean b\u1ea3n": 76,
+    "Lo\u1ea1i ph\u1ee5 ki\u1ec7n / linh ki\u1ec7n xe": 52,
+    "Lo\u1ea1i ph\u1ee5 ki\u1ec7n nh\u00e0 b\u1ebfp": 106,
+    "Lo\u1ea1i ph\u1ee5 ki\u1ec7n \u00e2m nh\u1ea1c": 75,
+    "Lo\u1ea1i ph\u1ee5 ki\u1ec7n \u0111i\u1ec7n t\u1eed": 98,
+    "Lo\u1ea1i s\u1ea3n ph\u1ea9m l\u00e0m \u0111\u1eb9p": 27,
+    "Lo\u1ea1i thi\u1ebft b\u1ecb an to\u00e0n": 81,
+    "Lo\u1ea1i thi\u1ebft b\u1ecb v\u00e0 d\u1ee5ng c\u1ee5 l\u00e0m \u0111\u1eb9p": 91,
+    "Lo\u1ea1i thi\u1ebft b\u1ecb v\u00e0 linh ki\u1ec7n \u0111i\u1ec7n t\u1eed": 46,
+    "Lo\u1ea1i thi\u1ebft b\u1ecb y t\u1ebf": 97,
+    "Lo\u1ea1i thi\u1ebft b\u1ecb \u00e2m thanh": 89,
+    "Lo\u1ea1i thi\u1ebft b\u1ecb \u0111i\u1ec7n gia d\u1ee5ng": 110,
+    "Lo\u1ea1i thu\u1ed1c & th\u1ef1c ph\u1ea9m ch\u1ee9c n\u0103ng": 36,
+    "Lo\u1ea1i th\u1ee9c \u0103n ch\u1ebf bi\u1ebfn s\u1eb5n": 101,
+    "Lo\u1ea1i th\u1ef1c ph\u1ea9m": 32,
+    "Lo\u1ea1i trang ph\u1ee5c": 0,
+    "Lo\u1ea1i trang s\u1ee9c / ph\u1ee5 ki\u1ec7n": 17,
+    "Lo\u1ea1i trang thi\u1ebft b\u1ecb th\u1ec3 thao": 105,
+    "Lo\u1ea1i tr\u00e1i c\u00e2y": 116,
+    "Lo\u1ea1i tr\u00f2 ch\u01a1i": 94,
+    "Lo\u1ea1i voucher & d\u1ecbch v\u1ee5": 85,
+    "Lo\u1ea1i v\u1eadt d\u1ee5ng ch\u0103m s\u00f3c c\u00e1 nh\u00e2n": 108,
+    "Lo\u1ea1i v\u1eadt d\u1ee5ng h\u1ed3 c\u00e1": 118,
+    "Lo\u1ea1i v\u1eadt d\u1ee5ng l\u00e0m v\u01b0\u1eddn": 61,
+    "Lo\u1ea1i v\u1eadt d\u1ee5ng trong nh\u00e0 kh\u00e1c": 64,
+    "Lo\u1ea1i \u0110\u1ed3 L\u00f3t": 11,
+    "Lo\u1ea1i \u0111\u00e8n": 62,
+    "Lo\u1ea1i \u0111\u0129a CD, DVD & Bluray": 90,
+    "Lo\u1ea1i \u0111\u1ebf gi\u00e0y": 72,
+    "Lo\u1ea1i \u0111\u1ed3 ch\u0103m s\u00f3c s\u1ee9c kh\u1ecfe m\u1eb9": 79,
+    "Lo\u1ea1i \u0111\u1ed3 ch\u01a1i": 83,
+    "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ch\u0103m s\u00f3c b\u00e9": 99,
+    "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng kh\u00e1ch": 93,
+    "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng ng\u1ee7": 111,
+    "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng ph\u00f2ng t\u1eafm": 107,
+    "Lo\u1ea1i \u0111\u1ed3 d\u00f9ng v\u0103n ph\u00f2ng ph\u1ea9m": 65,
+    "Lo\u1ea1i \u0111\u1ed3 gia d\u1ee5ng b\u1ebfp": 113,
+    "Lo\u1ea1i \u0111\u1ed3 v\u1eadt s\u01b0u t\u1ea7m": 87,
+    "Lo\u1ea1i \u1ed1p": 115,
+    "M\u00e0n h\u00ecnh": 43,
+    "M\u00e0u s\u1eafc": 9,
+    "M\u00f9a": 15,
+    "M\u00f9i h\u01b0\u01a1ng": 104,
+    "M\u1ea1ng kh\u00f4ng d\u00e2y": 23,
+    "M\u1eabu": 13,
+    "M\u1eabu xe t\u01b0\u01a1ng th\u00edch": 103,
+    "M\u1eb7t \u0111\u1ed3ng h\u1ed3": 19,
+    "M\u1ef1c v\u00e0 t\u00f4ng m\u00e0u": 78,
+    "Nh\u00f3m tu\u1ed5i": 38,
+    "N\u00e9t b\u00fat": 77,
+    "O": 1,
+    "Phong c\u00e1ch": 8,
+    "Ph\u01b0\u01a1ng th\u1ee9c ho\u1ea1t \u0111\u1ed9ng": 63,
+    "RAM": 49,
+    "Size": 10,
+    "Tay \u00e1o": 2,
+    "Thi\u1ebft b\u1ecb t\u01b0\u01a1ng th\u00edch": 54,
+    "Th\u00e0nh Ph\u1ea7n Ho\u1ea1t T\u00ednh": 37,
+    "Th\u00e0nh ph\u1ea7n": 35,
+    "Th\u01b0\u01a1ng hi\u1ec7u": 12,
+    "Th\u1ec3 lo\u1ea1i s\u00e1ch truy\u1ec7n": 95,
+    "Th\u1ec3 t\u00edch": 29,
+    "Tr\u1ecdng l\u01b0\u1ee3ng": 31,
+    "Tr\u1ecdng l\u01b0\u1ee3ng gi\u1ea5y": 68,
+    "T\u00ecnh tr\u1ea1ng": 47,
+    "T\u00edch h\u1ee3p ph\u1ee5 ki\u1ec7n": 22,
+    "T\u00ednh n\u0103ng": 7,
+    "T\u1ed1c \u0111\u1ed9 x\u1eed l\u00fd": 53,
+    "V\u00f9ng c\u01a1 th\u1ec3": 30,
+    "Xu\u1ea5t x\u1ee9": 48,
+    "\u0110\u01b0\u1eddng k\u00ednh": 58,
+    "\u0110\u1ed9 d\u00e0y": 59,
+    "\u0110\u1ed9 ph\u00e2n gi\u1ea3i": 50,
+    "\u0110\u1ed9 tu\u1ed5i khuy\u1ebfn ngh\u1ecb": 39
+  },
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "xlm-roberta",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "output_past": true,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.44.2",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 250002
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96413b98b048d66871dbdb0c112716b8f2e94d0592167a75e43daf4877318065
+size 2235899764

special_tokens_map.json

@@ -0,0 +1,15 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": {
+    "content": "<mask>",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8373f9cd3d27591e1924426bcc1c8799bc5a9affc4fc857982c5d66668dd1f41
+size 17082832

tokenizer_config.json

@@ -0,0 +1,54 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "250001": {
+      "content": "<mask>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": "<mask>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "XLMRobertaTokenizer",
+  "unk_token": "<unk>"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26a455b527db209fea43fabc5d9b265be4b2758dc1be97be8c467c514238c06f
+size 5240