Training in progress, epoch 1

config.json

@@ -0,0 +1,393 @@
+{
+  "architectures": [
+    "ModernBertForSequenceClassification"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "dtype": "float32",
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "id2label": {
+    "0": "AGS Script",
+    "1": "Ada",
+    "10": "C#",
+    "100": "Perl",
+    "101": "Pickle",
+    "102": "PicoLisp",
+    "103": "PlantUML",
+    "104": "PostScript",
+    "105": "PowerShell",
+    "106": "Prolog",
+    "107": "Protocol Buffer Text Format",
+    "108": "PureScript",
+    "109": "Python",
+    "11": "C++",
+    "110": "R",
+    "111": "RDoc",
+    "112": "REALbasic",
+    "113": "Ragel in Ruby Host",
+    "114": "Rascal",
+    "115": "Raw token data",
+    "116": "Redcode",
+    "117": "Ruby",
+    "118": "Rust",
+    "119": "SCSS",
+    "12": "CMake",
+    "120": "SQF",
+    "121": "SQL",
+    "122": "Sass",
+    "123": "Scala",
+    "124": "Scheme",
+    "125": "Scilab",
+    "126": "Shell",
+    "127": "Smali",
+    "128": "Solidity",
+    "129": "Squirrel",
+    "13": "CODEOWNERS",
+    "130": "Starlark",
+    "131": "Stylus",
+    "132": "Swift",
+    "133": "TSV",
+    "134": "TSX",
+    "135": "Thrift",
+    "136": "Turtle",
+    "137": "Twig",
+    "138": "TypeScript",
+    "139": "Unity3D Asset",
+    "14": "Chapel",
+    "140": "Unix Assembly",
+    "141": "Unknown",
+    "142": "VBScript",
+    "143": "VCL",
+    "144": "VHDL",
+    "145": "Verilog",
+    "146": "Vim Script",
+    "147": "Vim Snippet",
+    "148": "Visual Basic",
+    "149": "Visual Basic .NET",
+    "15": "CoffeeScript",
+    "150": "Vue",
+    "151": "Wavefront Object",
+    "152": "XML Property List",
+    "153": "XS",
+    "154": "YANG",
+    "16": "ColdFusion",
+    "17": "Common Lisp",
+    "18": "Crystal",
+    "19": "Csound",
+    "2": "Adobe Font Metrics",
+    "20": "Csound Document",
+    "21": "D",
+    "22": "DIGITAL Command Language",
+    "23": "Dart",
+    "24": "Dockerfile",
+    "25": "ECL",
+    "26": "EJS",
+    "27": "Eagle",
+    "28": "Edoid",
+    "29": "Erlang",
+    "3": "ApacheConf",
+    "30": "Fish",
+    "31": "Fluent",
+    "32": "Fortran Free Form",
+    "33": "FreeBasic",
+    "34": "FreeMarker",
+    "35": "G-code",
+    "36": "GAP",
+    "37": "GAS",
+    "38": "GDB",
+    "39": "GDScript",
+    "4": "Apex",
+    "40": "GLSL",
+    "41": "Gerber Image",
+    "42": "Gettext Catalog",
+    "43": "Gherkin",
+    "44": "Git Config",
+    "45": "Gnuplot",
+    "46": "Go",
+    "47": "Go Module",
+    "48": "Gradle",
+    "49": "Graphviz (DOT)",
+    "5": "AsciiDoc",
+    "50": "Groovy",
+    "51": "HCL",
+    "52": "HTML+Razor",
+    "53": "Hack",
+    "54": "Haskell",
+    "55": "Haxe",
+    "56": "Hoon",
+    "57": "Ignore List",
+    "58": "ImageJ Macro",
+    "59": "Inform 7",
+    "6": "Assembly",
+    "60": "Inno Setup",
+    "61": "Io",
+    "62": "Ioke",
+    "63": "Isabelle",
+    "64": "J",
+    "65": "JAR Manifest",
+    "66": "JQ",
+    "67": "Java",
+    "68": "JavaScript",
+    "69": "Julia",
+    "7": "Blade",
+    "70": "Jupyter Notebook",
+    "71": "KiCad Layout",
+    "72": "Kit",
+    "73": "Kotlin",
+    "74": "Kvlang",
+    "75": "LLVM",
+    "76": "Lean",
+    "77": "Less",
+    "78": "Lex",
+    "79": "Logtalk",
+    "8": "Bluespec",
+    "80": "LookML",
+    "81": "Lua",
+    "82": "MATLAB",
+    "83": "Makefile",
+    "84": "Mathematica",
+    "85": "Metal",
+    "86": "Modelica",
+    "87": "Nim",
+    "88": "OCaml",
+    "89": "ObjDump",
+    "9": "C",
+    "90": "Objective-C",
+    "91": "Objective-C++",
+    "92": "Open Policy Agent",
+    "93": "OpenEdge ABL",
+    "94": "OpenStep Property List",
+    "95": "OpenType Feature File",
+    "96": "PHP",
+    "97": "PLpgSQL",
+    "98": "POV-Ray SDL",
+    "99": "Pascal"
+  },
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "label2id": {
+    "AGS Script": "0",
+    "Ada": "1",
+    "Adobe Font Metrics": "2",
+    "ApacheConf": "3",
+    "Apex": "4",
+    "AsciiDoc": "5",
+    "Assembly": "6",
+    "Blade": "7",
+    "Bluespec": "8",
+    "C": "9",
+    "C#": "10",
+    "C++": "11",
+    "CMake": "12",
+    "CODEOWNERS": "13",
+    "Chapel": "14",
+    "CoffeeScript": "15",
+    "ColdFusion": "16",
+    "Common Lisp": "17",
+    "Crystal": "18",
+    "Csound": "19",
+    "Csound Document": "20",
+    "D": "21",
+    "DIGITAL Command Language": "22",
+    "Dart": "23",
+    "Dockerfile": "24",
+    "ECL": "25",
+    "EJS": "26",
+    "Eagle": "27",
+    "Edoid": "28",
+    "Erlang": "29",
+    "Fish": "30",
+    "Fluent": "31",
+    "Fortran Free Form": "32",
+    "FreeBasic": "33",
+    "FreeMarker": "34",
+    "G-code": "35",
+    "GAP": "36",
+    "GAS": "37",
+    "GDB": "38",
+    "GDScript": "39",
+    "GLSL": "40",
+    "Gerber Image": "41",
+    "Gettext Catalog": "42",
+    "Gherkin": "43",
+    "Git Config": "44",
+    "Gnuplot": "45",
+    "Go": "46",
+    "Go Module": "47",
+    "Gradle": "48",
+    "Graphviz (DOT)": "49",
+    "Groovy": "50",
+    "HCL": "51",
+    "HTML+Razor": "52",
+    "Hack": "53",
+    "Haskell": "54",
+    "Haxe": "55",
+    "Hoon": "56",
+    "Ignore List": "57",
+    "ImageJ Macro": "58",
+    "Inform 7": "59",
+    "Inno Setup": "60",
+    "Io": "61",
+    "Ioke": "62",
+    "Isabelle": "63",
+    "J": "64",
+    "JAR Manifest": "65",
+    "JQ": "66",
+    "Java": "67",
+    "JavaScript": "68",
+    "Julia": "69",
+    "Jupyter Notebook": "70",
+    "KiCad Layout": "71",
+    "Kit": "72",
+    "Kotlin": "73",
+    "Kvlang": "74",
+    "LLVM": "75",
+    "Lean": "76",
+    "Less": "77",
+    "Lex": "78",
+    "Logtalk": "79",
+    "LookML": "80",
+    "Lua": "81",
+    "MATLAB": "82",
+    "Makefile": "83",
+    "Mathematica": "84",
+    "Metal": "85",
+    "Modelica": "86",
+    "Nim": "87",
+    "OCaml": "88",
+    "ObjDump": "89",
+    "Objective-C": "90",
+    "Objective-C++": "91",
+    "Open Policy Agent": "92",
+    "OpenEdge ABL": "93",
+    "OpenStep Property List": "94",
+    "OpenType Feature File": "95",
+    "PHP": "96",
+    "PLpgSQL": "97",
+    "POV-Ray SDL": "98",
+    "Pascal": "99",
+    "Perl": "100",
+    "Pickle": "101",
+    "PicoLisp": "102",
+    "PlantUML": "103",
+    "PostScript": "104",
+    "PowerShell": "105",
+    "Prolog": "106",
+    "Protocol Buffer Text Format": "107",
+    "PureScript": "108",
+    "Python": "109",
+    "R": "110",
+    "RDoc": "111",
+    "REALbasic": "112",
+    "Ragel in Ruby Host": "113",
+    "Rascal": "114",
+    "Raw token data": "115",
+    "Redcode": "116",
+    "Ruby": "117",
+    "Rust": "118",
+    "SCSS": "119",
+    "SQF": "120",
+    "SQL": "121",
+    "Sass": "122",
+    "Scala": "123",
+    "Scheme": "124",
+    "Scilab": "125",
+    "Shell": "126",
+    "Smali": "127",
+    "Solidity": "128",
+    "Squirrel": "129",
+    "Starlark": "130",
+    "Stylus": "131",
+    "Swift": "132",
+    "TSV": "133",
+    "TSX": "134",
+    "Thrift": "135",
+    "Turtle": "136",
+    "Twig": "137",
+    "TypeScript": "138",
+    "Unity3D Asset": "139",
+    "Unix Assembly": "140",
+    "Unknown": "141",
+    "VBScript": "142",
+    "VCL": "143",
+    "VHDL": "144",
+    "Verilog": "145",
+    "Vim Script": "146",
+    "Vim Snippet": "147",
+    "Visual Basic": "148",
+    "Visual Basic .NET": "149",
+    "Vue": "150",
+    "Wavefront Object": "151",
+    "XML Property List": "152",
+    "XS": "153",
+    "YANG": "154"
+  },
+  "layer_norm_eps": 1e-05,
+  "layer_types": [
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "local_attention": 128,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "rope_parameters": {
+    "full_attention": {
+      "rope_theta": 160000.0,
+      "rope_type": "default"
+    },
+    "sliding_attention": {
+      "rope_theta": 10000.0,
+      "rope_type": "default"
+    }
+  },
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.15.0",
+  "use_cache": false,
+  "vocab_size": 50368
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4308e42071e0edfe38e1002c537e85b71e8a2600637953b342c6c294eac267d0
+size 598910420

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce057b097ec08c103f446eb13f15363b2620f69fb45b983a9bb15edf72c5ace6
+size 6033