Upload 7 files

added_tokens.json

@@ -0,0 +1,303 @@
+{
+  "[MASK]": 128000,
+  "\\Bbb": 128053,
+  "\\Big": 128280,
+  "\\Bigl": 128262,
+  "\\Bigr": 128264,
+  "\\Box": 128285,
+  "\\Delta": 128068,
+  "\\Gamma": 128089,
+  "\\Im": 128291,
+  "\\Lambda": 128162,
+  "\\Leftarrow": 128283,
+  "\\Leftrightarrow": 128151,
+  "\\Longleftrightarrow": 128275,
+  "\\Longrightarrow": 128216,
+  "\\Omega": 128064,
+  "\\Phi": 128134,
+  "\\Pi": 128234,
+  "\\Pr": 128207,
+  "\\Psi": 128212,
+  "\\R": 128252,
+  "\\Re": 128243,
+  "\\Rightarrow": 128066,
+  "\\Sigma": 128136,
+  "\\Theta": 128208,
+  "\\Vert": 128141,
+  "\\\\": 128009,
+  "\\a": 128259,
+  "\\aleph": 128253,
+  "\\alpha": 128028,
+  "\\angle": 128165,
+  "\\approx": 128079,
+  "\\arcsin": 128284,
+  "\\arctan": 128191,
+  "\\arg": 128254,
+  "\\ast": 128139,
+  "\\b": 128286,
+  "\\backslash": 128156,
+  "\\bar": 128070,
+  "\\barwedge": 128257,
+  "\\begin": 128021,
+  "\\beta": 128046,
+  "\\big": 128107,
+  "\\bigcap": 128158,
+  "\\bigcup": 128121,
+  "\\bigg": 128127,
+  "\\biggl": 128299,
+  "\\biggr": 128293,
+  "\\bigl": 128215,
+  "\\bigoplus": 128242,
+  "\\bigr": 128203,
+  "\\bigtriangleup": 128277,
+  "\\bigwedge": 128300,
+  "\\binom": 128052,
+  "\\blacksquare": 128295,
+  "\\bmod": 128172,
+  "\\boldsymbol": 128160,
+  "\\bot": 128251,
+  "\\boxdot": 128231,
+  "\\boxed": 128260,
+  "\\bullet": 128187,
+  "\\cal": 128209,
+  "\\cancel": 128129,
+  "\\cap": 128038,
+  "\\cdot": 128001,
+  "\\cdots": 128023,
+  "\\cfrac": 128245,
+  "\\chi": 128106,
+  "\\choose": 128133,
+  "\\circ": 128056,
+  "\\colon": 128135,
+  "\\coloneqq": 128138,
+  "\\color": 128217,
+  "\\complement": 128120,
+  "\\cong": 128110,
+  "\\cos": 128024,
+  "\\cosh": 128188,
+  "\\cot": 128092,
+  "\\cr": 128204,
+  "\\csc": 128077,
+  "\\cup": 128050,
+  "\\d": 128241,
+  "\\dagger": 128199,
+  "\\ddot": 128287,
+  "\\ddots": 128213,
+  "\\deg": 128210,
+  "\\delta": 128048,
+  "\\det": 128157,
+  "\\dfrac": 128007,
+  "\\diamond": 128267,
+  "\\dim": 128150,
+  "\\displaystyle": 128080,
+  "\\div": 128124,
+  "\\dot": 128147,
+  "\\dots": 128078,
+  "\\dotsc": 128268,
+  "\\dotsm": 128279,
+  "\\downarrow": 128240,
+  "\\dt": 128296,
+  "\\dx": 128211,
+  "\\ell": 128118,
+  "\\emptyset": 128094,
+  "\\end": 128022,
+  "\\epsilon": 128033,
+  "\\equiv": 128054,
+  "\\eta": 128101,
+  "\\exists": 128096,
+  "\\exp": 128137,
+  "\\f": 128250,
+  "\\forall": 128049,
+  "\\frac": 128002,
+  "\\frak": 128273,
+  "\\gamma": 128062,
+  "\\gcd": 128194,
+  "\\ge": 128086,
+  "\\gep": 128228,
+  "\\geq": 128025,
+  "\\geqslant": 128224,
+  "\\ger": 128229,
+  "\\ges": 128225,
+  "\\get": 128230,
+  "\\gray": 128117,
+  "\\gt": 128155,
+  "\\hat": 128090,
+  "\\hline": 128183,
+  "\\hookrightarrow": 128256,
+  "\\hphantom": 128182,
+  "\\hspace": 128174,
+  "\\iff": 128144,
+  "\\iint": 128219,
+  "\\implies": 128087,
+  "\\in": 128006,
+  "\\inf": 128154,
+  "\\infty": 128011,
+  "\\int": 128014,
+  "\\iota": 128266,
+  "\\kappa": 128161,
+  "\\ker": 128167,
+  "\\lVert": 128186,
+  "\\lambda": 128035,
+  "\\land": 128140,
+  "\\langle": 128057,
+  "\\lbrace": 128197,
+  "\\lceil": 128226,
+  "\\lcm": 128180,
+  "\\ldots": 128072,
+  "\\le": 128051,
+  "\\left": 128003,
+  "\\leftarrow": 128274,
+  "\\leftrightarrow": 128255,
+  "\\lep": 128232,
+  "\\leq": 128016,
+  "\\leqslant": 128177,
+  "\\ler": 128223,
+  "\\les": 128222,
+  "\\let": 128227,
+  "\\lfloor": 128126,
+  "\\lim": 128030,
+  "\\liminf": 128221,
+  "\\limits": 128032,
+  "\\limsup": 128185,
+  "\\ln": 128042,
+  "\\lnot": 128173,
+  "\\log": 128037,
+  "\\longrightarrow": 128159,
+  "\\lor": 128163,
+  "\\lt": 128104,
+  "\\lvert": 128146,
+  "\\mapsto": 128105,
+  "\\mathbb": 128005,
+  "\\mathbf": 128036,
+  "\\mathcal": 128027,
+  "\\mathfrak": 128071,
+  "\\mathop": 128244,
+  "\\mathrm": 128029,
+  "\\mathscr": 128112,
+  "\\mathsf": 128169,
+  "\\max": 128122,
+  "\\mid": 128058,
+  "\\middle": 128184,
+  "\\min": 128143,
+  "\\mod": 128145,
+  "\\models": 128281,
+  "\\mu": 128044,
+  "\\nabla": 128111,
+  "\\ne": 128113,
+  "\\neg": 128148,
+  "\\neq": 128013,
+  "\\newcommand": 128214,
+  "\\ni": 128276,
+  "\\nlet": 128278,
+  "\\nmid": 128249,
+  "\\not": 128130,
+  "\\notin": 128152,
+  "\\nu": 128091,
+  "\\odot": 128246,
+  "\\oint": 128261,
+  "\\omega": 128059,
+  "\\operatorname": 128039,
+  "\\oplus": 128114,
+  "\\oslash": 128258,
+  "\\otimes": 128088,
+  "\\over": 128103,
+  "\\overline": 128063,
+  "\\overset": 128196,
+  "\\p": 128288,
+  "\\parallel": 128292,
+  "\\pars": 128270,
+  "\\partial": 128026,
+  "\\perp": 128170,
+  "\\phantom": 128192,
+  "\\phi": 128041,
+  "\\pi": 128010,
+  "\\pm": 128047,
+  "\\pmatrix": 128247,
+  "\\pmod": 128102,
+  "\\pmxd": 128200,
+  "\\prec": 128294,
+  "\\prime": 128100,
+  "\\prod": 128098,
+  "\\psi": 128081,
+  "\\qquad": 128116,
+  "\\quad": 128060,
+  "\\rVert": 128189,
+  "\\rangle": 128055,
+  "\\rbrace": 128201,
+  "\\rceil": 128220,
+  "\\rfloor": 128125,
+  "\\rho": 128074,
+  "\\right": 128004,
+  "\\rightarrow": 128034,
+  "\\root": 128263,
+  "\\rvert": 128149,
+  "\\scriptsize": 128239,
+  "\\sec": 128076,
+  "\\setminus": 128084,
+  "\\sf": 128297,
+  "\\sigma": 128045,
+  "\\sim": 128132,
+  "\\simeq": 128179,
+  "\\sin": 128019,
+  "\\sinh": 128193,
+  "\\space": 128097,
+  "\\sqrt": 128012,
+  "\\square": 128235,
+  "\\stackrel": 128198,
+  "\\star": 128166,
+  "\\subset": 128082,
+  "\\subseteq": 128093,
+  "\\subsetneq": 128269,
+  "\\substack": 128181,
+  "\\sum": 128015,
+  "\\sup": 128099,
+  "\\supset": 128265,
+  "\\supseteq": 128298,
+  "\\tag": 128083,
+  "\\tan": 128061,
+  "\\tanh": 128289,
+  "\\tau": 128069,
+  "\\text": 128018,
+  "\\textbf": 128119,
+  "\\textit": 128282,
+  "\\textrm": 128164,
+  "\\tfrac": 128040,
+  "\\therefore": 128248,
+  "\\theta": 128031,
+  "\\thinspace": 128043,
+  "\\tilde": 128115,
+  "\\times": 128008,
+  "\\to": 128017,
+  "\\top": 128218,
+  "\\triangle": 128178,
+  "\\triangleleft": 128290,
+  "\\underbrace": 128175,
+  "\\underline": 128176,
+  "\\uparrow": 128236,
+  "\\varepsilon": 128067,
+  "\\varnothing": 128171,
+  "\\varphi": 128065,
+  "\\vartheta": 128128,
+  "\\vdash": 128237,
+  "\\vdots": 128142,
+  "\\vec": 128131,
+  "\\vee": 128168,
+  "\\vert": 128109,
+  "\\vphantom": 128190,
+  "\\wedge": 128108,
+  "\\widehat": 128202,
+  "\\widetilde": 128206,
+  "\\x": 128205,
+  "\\xi": 128085,
+  "\\xrightarrow": 128195,
+  "\\zeta": 128073,
+  "\\|": 128020,
+  "align*": 128123,
+  "bmatrix": 128095,
+  "eqnarray": 128233,
+  "eqnarray*": 128271,
+  "equation": 128153,
+  "equation*": 128238,
+  "pmatrix": 128075,
+  "vmatrix": 128272
+}

config.json

@@ -0,0 +1,35 @@
+{
+  "_name_or_path": "../../final/models/pretraining/DML-MFM-NFIR-FFIR-MTM",
+  "architectures": [
+    "DebertaV2ForMaskedLM"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-07,
+  "max_position_embeddings": 512,
+  "max_relative_positions": -1,
+  "model_type": "deberta-v2",
+  "norm_rel_ebd": "layer_norm",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "pooler_dropout": 0,
+  "pooler_hidden_act": "gelu",
+  "pooler_hidden_size": 768,
+  "pos_att_type": [
+    "p2c",
+    "c2p"
+  ],
+  "position_biased_input": false,
+  "position_buckets": 256,
+  "relative_attention": true,
+  "share_att_key": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.30.2",
+  "type_vocab_size": 0,
+  "vocab_size": 128301
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e303241bfaca8dcb6b3b07b6f97fd88b9e05c91984b847818f964991c05d4d5
+size 738388721

special_tokens_map.json

@@ -0,0 +1,9 @@
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
+size 2464616

tokenizer_config.json

@@ -0,0 +1,16 @@
+{
+  "bos_token": "[CLS]",
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "sp_model_kwargs": {},
+  "split_by_punct": false,
+  "tokenizer_class": "DebertaV2Tokenizer",
+  "unk_token": "[UNK]",
+  "vocab_type": "spm"
+}