Training in progress, step 5000

last-checkpoint/added_tokens.json

@@ -0,0 +1,446 @@
+{
+  "ACV": 31381,
+  "Absolute": 31398,
+  "Acidic": 31331,
+  "Additive": 31301,
+  "Admittance": 31228,
+  "Affinity": 31407,
+  "Alkaline": 31410,
+  "Alternating": 31428,
+  "Aluminum": 31446,
+  "Amount": 31283,
+  "Amperostat": 31359,
+  "Analyzer": 31324,
+  "Angle": 31420,
+  "Angluar": 31356,
+  "Anglular": 31348,
+  "Angstrom": 31193,
+  "Angular": 31472,
+  "Anodic": 31334,
+  "Approximation": 31440,
+  "Aqueous": 31346,
+  "Areal": 31425,
+  "Areic": 31462,
+  "Artifact": 31456,
+  "Atom": 31352,
+  "Auxiliary": 31387,
+  "BAE": 31368,
+  "Bandgap": 31421,
+  "Battery": 31328,
+  "Binary": 31307,
+  "Binder": 31297,
+  "Bode": 31191,
+  "Breadth": 31338,
+  "Bulk": 31196,
+  "Butler": 31310,
+  "Button": 31467,
+  "CCCV": 31312,
+  "CMC": 31315,
+  "CPE": 31202,
+  "CSV": 31336,
+  "Capacitance": 31238,
+  "Capacity": 31295,
+  "Carnot": 31363,
+  "Cathodic": 31357,
+  "Characterisation": 31214,
+  "Charge": 31460,
+  "Charging": 31265,
+  "Closed": 31285,
+  "Cobalt": 31350,
+  "Coefficient": 31289,
+  "Coin": 31270,
+  "Collector": 31445,
+  "Conductance": 31263,
+  "Conductive": 31454,
+  "Conductivity": 31395,
+  "Conductor": 31284,
+  "Continuant": 31287,
+  "Continuum": 31430,
+  "Conversion": 31333,
+  "Coulombic": 31218,
+  "Coupling": 31192,
+  "Coverage": 31200,
+  "Crystalization": 31400,
+  "Curve": 31215,
+  "Cutoff": 31256,
+  "Cutting": 31216,
+  "Cycling": 31261,
+  "Cylindrical": 31325,
+  "DFN": 31293,
+  "DMC": 31302,
+  "DME": 31457,
+  "DPV": 31183,
+  "Debye": 31286,
+  "Dedomena": 31222,
+  "Discharge": 31274,
+  "Discharging": 31452,
+  "Disintegration": 31469,
+  "Displacement": 31451,
+  "Dissociation": 31354,
+  "Donnan": 31210,
+  "Doyle": 31366,
+  "Drag": 31306,
+  "EASA": 31320,
+  "ECSA": 31418,
+  "EIS": 31187,
+  "EMC": 31409,
+  "Elaboration": 31358,
+  "Electrochemical": 31384,
+  "Electrochemically": 31424,
+  "Electrode": 31322,
+  "Electrolyte": 31379,
+  "Electrolytic": 31185,
+  "Electromagnetic": 31203,
+  "Electronically": 31471,
+  "Electrons": 31271,
+  "Electrooxidation": 31374,
+  "Electroreduction": 31235,
+  "Elemental": 31288,
+  "Elementary": 31405,
+  "Embarassingly": 31329,
+  "Encoded": 31246,
+  "Endurant": 31355,
+  "Engineered": 31390,
+  "Entity": 31294,
+  "Equilibrium": 31209,
+  "Equipment": 31197,
+  "Erdey": 31237,
+  "Executable": 31205,
+  "Expansion": 31275,
+  "Extinction": 31199,
+  "FEC": 31389,
+  "FRA": 31221,
+  "Faraday": 31367,
+  "Fermionic": 31364,
+  "Float": 31231,
+  "Forming": 31220,
+  "Friction": 31386,
+  "Fuller": 31305,
+  "GDE": 31251,
+  "Galvanizing": 31442,
+  "Galvanostatic": 31415,
+  "Generative": 31253,
+  "Gravimetric": 31435,
+  "Gruz": 31280,
+  "Gyromagnetic": 31378,
+  "Helmholtz": 31239,
+  "Hold": 31250,
+  "Holistic": 31244,
+  "Icon": 31369,
+  "Impedance": 31273,
+  "Indifferent": 31184,
+  "Inductance": 31223,
+  "Inductive": 31383,
+  "Inert": 31186,
+  "Insertion": 31433,
+  "Insulance": 31402,
+  "Intangible": 31373,
+  "Integral": 31427,
+  "Intentional": 31317,
+  "Intercalation": 31225,
+  "Intermittent": 31245,
+  "Interruption": 31240,
+  "Invariant": 31278,
+  "Inverse": 31453,
+  "Ionic": 31434,
+  "Ionization": 31319,
+  "Jar": 31464,
+  "Kynar": 31343,
+  "LCO": 31401,
+  "LFP": 31316,
+  "LMO": 31303,
+  "LNO": 31224,
+  "LSV": 31365,
+  "LTO": 31188,
+  "Larmor": 31392,
+  "Lifetime": 31236,
+  "Limit": 31298,
+  "Limiting": 31432,
+  "Lineic": 31466,
+  "Liquid": 31441,
+  "Lithium": 31423,
+  "Loading": 31266,
+  "Loose": 31397,
+  "Lorenz": 31309,
+  "Macroscopic": 31361,
+  "Magnetizing": 31399,
+  "Magnetogyric": 31388,
+  "Manganese": 31443,
+  "Massic": 31272,
+  "Maximal": 31360,
+  "Measurand": 31347,
+  "Measured": 31217,
+  "Mereological": 31242,
+  "Micron": 31438,
+  "Minute": 31465,
+  "Mixture": 31431,
+  "Modulus": 31308,
+  "Molar": 31201,
+  "Molarity": 31257,
+  "Mole": 31351,
+  "Molecule": 31447,
+  "Moment": 31254,
+  "Multiplicity": 31323,
+  "NCA": 31439,
+  "NCM": 31448,
+  "NHE": 31342,
+  "NMC": 31414,
+  "NPV": 31195,
+  "Nernst": 31417,
+  "Neutral": 31335,
+  "Nickel": 31436,
+  "Nucleon": 31264,
+  "Numeral": 31458,
+  "OCV": 31416,
+  "OSWV": 31207,
+  "Occurrent": 31337,
+  "Ohmic": 31403,
+  "Ordinary": 31455,
+  "Osmotic": 31313,
+  "Osteryoung": 31194,
+  "Overpotential": 31429,
+  "Overvoltage": 31259,
+  "Oxidation": 31226,
+  "Oxide": 31394,
+  "PTFE": 31376,
+  "PVDF": 31468,
+  "Partion": 31426,
+  "Perdurant": 31233,
+  "Permeability": 31304,
+  "Permittivity": 31345,
+  "Phosphate": 31385,
+  "Plane": 31437,
+  "Plot": 31375,
+  "Polarization": 31396,
+  "Polarography": 31372,
+  "Potentiostatic": 31269,
+  "Pouch": 31268,
+  "Pressureless": 31327,
+  "Primitive": 31419,
+  "Processed": 31299,
+  "Qt": 31249,
+  "Quantity": 31291,
+  "Radioactive": 31408,
+  "Radius": 31241,
+  "Reachable": 31211,
+  "Reactance": 31296,
+  "Reactive": 31281,
+  "Reciprocal": 31230,
+  "Reluctance": 31413,
+  "Repetency": 31255,
+  "Resistivity": 31341,
+  "Retention": 31382,
+  "Reversal": 31258,
+  "Reversible": 31473,
+  "Rolling": 31461,
+  "SAE": 31189,
+  "SBR": 31349,
+  "SEI": 31318,
+  "SHE": 31326,
+  "SWV": 31277,
+  "Scan": 31267,
+  "Selectivity": 31362,
+  "Semiosis": 31292,
+  "Semiotic": 31377,
+  "Separator": 31260,
+  "Shear": 31262,
+  "Shelf": 31444,
+  "Shuttlecock": 31321,
+  "Simulacrum": 31449,
+  "Sintering": 31463,
+  "Solubility": 31290,
+  "Solute": 31332,
+  "Solvent": 31243,
+  "Spatially": 31353,
+  "Spatiialnteraction": 31314,
+  "Spatio": 31340,
+  "Stopping": 31370,
+  "Strength": 31198,
+  "Substance": 31190,
+  "Supercapacitor": 31412,
+  "Swagelok": 31248,
+  "Sweep": 31344,
+  "Swing": 31232,
+  "TAST": 31404,
+  "TFSI": 31252,
+  "Tangible": 31422,
+  "Teflon": 31219,
+  "Tension": 31276,
+  "Theorization": 31406,
+  "Thermodynamic": 31208,
+  "Thermogalvanic": 31247,
+  "Tiling": 31470,
+  "Titanium": 31380,
+  "Titration": 31229,
+  "Transferred": 31300,
+  "Transferrence": 31339,
+  "Transmittance": 31459,
+  "Vacuum": 31282,
+  "Vapour": 31212,
+  "Viscosity": 31311,
+  "Volmer": 31393,
+  "Voltameter": 31279,
+  "Voltammetry": 31213,
+  "Volumetric": 31450,
+  "Volumic": 31411,
+  "Wholistic": 31227,
+  "Window": 31206,
+  "Workflow": 31391,
+  "Youngs": 31204,
+  "Zinc": 31234,
+  "and_logical": 31534,
+  "attr_close": 31519,
+  "attr_open": 31531,
+  "avg": 31485,
+  "battery__": 31182,
+  "bool_not": 31518,
+  "brack_close": 31530,
+  "brack_open": 31538,
+  "brick__": 31163,
+  "ceil": 31516,
+  "choose_by_max": 31477,
+  "choose_by_min": 31515,
+  "coalesce": 31491,
+  "concat": 31497,
+  "constant__": 31171,
+  "csvw__": 31127,
+  "datatype": 31480,
+  "dbo__": 31156,
+  "dbp__": 31167,
+  "dbr__": 31165,
+  "dc__": 31151,
+  "dcam__": 31132,
+  "dcat__": 31120,
+  "dcmitype__": 31142,
+  "dcterms__": 31168,
+  "doap__": 31118,
+  "dur_to_usecs": 31508,
+  "electrochemistry__": 31181,
+  "electrochemistrys__": 31148,
+  "electromotive": 31371,
+  "emmo__": 31116,
+  "encode_for_uri": 31489,
+  "foaf__": 31162,
+  "geo__": 31158,
+  "group_concat": 31484,
+  "hamming_dist": 31475,
+  "haversine_dist": 31503,
+  "isblank": 31482,
+  "isiri": 31513,
+  "isliteral": 31494,
+  "isnumeric": 31496,
+  "isuri": 31481,
+  "lang_at": 31528,
+  "langmatches": 31486,
+  "lcase": 31509,
+  "levenshtein_dist": 31488,
+  "log2": 31474,
+  "manufacturing__": 31125,
+  "map__": 31170,
+  "math_eql": 31539,
+  "math_geq": 31535,
+  "math_gt": 31536,
+  "math_leq": 31521,
+  "math_lt": 31537,
+  "math_neq": 31523,
+  "md5": 31504,
+  "odrl__": 31136,
+  "or_logical": 31524,
+  "org__": 31175,
+  "owl__": 31131,
+  "p__": 31176,
+  "par_close": 31526,
+  "par_open": 31520,
+  "periodictable__": 31145,
+  "pmanufacturing__": 31149,
+  "pq__": 31122,
+  "pqn__": 31174,
+  "pqv__": 31128,
+  "pr__": 31166,
+  "prefix__": 31119,
+  "prn__": 31146,
+  "prof__": 31124,
+  "prov__": 31159,
+  "prv__": 31173,
+  "ps__": 31141,
+  "psn__": 31140,
+  "psv__": 31129,
+  "qb__": 31137,
+  "qkdv__": 31157,
+  "quantitykind__": 31154,
+  "qudt__": 31178,
+  "quote_str": 31517,
+  "radians": 31512,
+  "rand": 31495,
+  "rdf__": 31134,
+  "rdfs__": 31138,
+  "regex": 31476,
+  "rounddown": 31511,
+  "roundup": 31499,
+  "sameterm": 31493,
+  "schema__": 31160,
+  "seconds_dbl": 31483,
+  "sep_com": 31525,
+  "sep_dot": 31529,
+  "sep_semi": 31532,
+  "sh__": 31153,
+  "sha1": 31506,
+  "sha256": 31487,
+  "sha384": 31507,
+  "sha512": 31502,
+  "skos__": 31130,
+  "soqk__": 31164,
+  "sosa__": 31150,
+  "sou__": 31155,
+  "ssn__": 31152,
+  "str_type": 31533,
+  "strafter": 31500,
+  "strbefore": 31498,
+  "strends": 31479,
+  "strlen": 31510,
+  "strstarts": 31514,
+  "swrl__": 31144,
+  "time__": 31161,
+  "timezone": 31478,
+  "tz": 31490,
+  "ucase": 31505,
+  "unit__": 31177,
+  "usecs_to_dur": 31501,
+  "value0": 31550,
+  "value1": 31551,
+  "value2": 31552,
+  "value3": 31553,
+  "value4": 31554,
+  "value5": 31555,
+  "value6": 31556,
+  "value7": 31557,
+  "value8": 31558,
+  "value9": 31559,
+  "vann__": 31143,
+  "var0": 31540,
+  "var1": 31541,
+  "var2": 31542,
+  "var3": 31543,
+  "var4": 31544,
+  "var5": 31545,
+  "var6": 31546,
+  "var7": 31547,
+  "var8": 31548,
+  "var9": 31549,
+  "var_": 31522,
+  "varp": 31492,
+  "void__": 31123,
+  "voltammetry": 31330,
+  "wd__": 31139,
+  "wdata__": 31135,
+  "wdno__": 31169,
+  "wdref__": 31179,
+  "wds__": 31117,
+  "wdt__": 31147,
+  "wdtn__": 31126,
+  "wdv__": 31133,
+  "wgs__": 31180,
+  "wildcard": 31527,
+  "xml__": 31121,
+  "xsd__": 31172
+}

last-checkpoint/config.json

@@ -0,0 +1,182 @@
+{
+  "_commit_hash": null,
+  "_name_or_path": "bert2bert_added_vocab",
+  "architectures": [
+    "EncoderDecoderModel"
+  ],
+  "decoder": {
+    "_name_or_path": "decoder_added_vocab",
+    "add_cross_attention": true,
+    "architectures": [
+      "BertModel"
+    ],
+    "attention_probs_dropout_prob": 0.1,
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": null,
+    "chunk_size_feed_forward": 0,
+    "classifier_dropout": null,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "gradient_checkpointing": false,
+    "hidden_act": "gelu",
+    "hidden_dropout_prob": 0.1,
+    "hidden_size": 768,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 3072,
+    "is_decoder": true,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "layer_norm_eps": 1e-12,
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "max_position_embeddings": 512,
+    "min_length": 0,
+    "model_type": "bert",
+    "no_repeat_ngram_size": 0,
+    "num_attention_heads": 12,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_hidden_layers": 12,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": 0,
+    "position_embedding_type": "absolute",
+    "prefix": null,
+    "problem_type": null,
+    "pruned_heads": {},
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": "float32",
+    "torchscript": false,
+    "transformers_version": "4.30.0.dev0",
+    "type_vocab_size": 2,
+    "typical_p": 1.0,
+    "use_bfloat16": false,
+    "use_cache": true,
+    "vocab_size": 31560
+  },
+  "decoder_start_token_id": 101,
+  "early_stopping": true,
+  "encoder": {
+    "_name_or_path": "encoder_added_vocab",
+    "add_cross_attention": false,
+    "architectures": [
+      "BertModel"
+    ],
+    "attention_probs_dropout_prob": 0.1,
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": null,
+    "chunk_size_feed_forward": 0,
+    "classifier_dropout": null,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "hidden_act": "gelu",
+    "hidden_dropout_prob": 0.1,
+    "hidden_size": 768,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 3072,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "layer_norm_eps": 1e-12,
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "max_position_embeddings": 512,
+    "min_length": 0,
+    "model_type": "bert",
+    "no_repeat_ngram_size": 0,
+    "num_attention_heads": 12,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_hidden_layers": 12,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": 0,
+    "position_embedding_type": "absolute",
+    "prefix": null,
+    "problem_type": null,
+    "pruned_heads": {},
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": "float32",
+    "torchscript": false,
+    "transformers_version": "4.30.0.dev0",
+    "type_vocab_size": 2,
+    "typical_p": 1.0,
+    "use_bfloat16": false,
+    "use_cache": true,
+    "vocab_size": 31560
+  },
+  "eos_token_id": 102,
+  "is_encoder_decoder": true,
+  "max_length": 256,
+  "min_length": 16,
+  "model_type": "encoder-decoder",
+  "num_beams": 10,
+  "pad_token_id": 0,
+  "torch_dtype": "float32",
+  "transformers_version": null,
+  "vocab_size": 31560
+}

last-checkpoint/generation_config.json

@@ -0,0 +1,10 @@
+{
+  "decoder_start_token_id": 101,
+  "early_stopping": true,
+  "eos_token_id": 102,
+  "max_length": 256,
+  "min_length": 16,
+  "num_beams": 10,
+  "pad_token_id": 0,
+  "transformers_version": "4.30.0.dev0"
+}

last-checkpoint/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8af2ce795803cccd70df17a5741d05c47630956ee6b41ca0e69110e78e3b67be
+size 1987250795

last-checkpoint/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9bb592d3f4e42471c4963cdf02b330d1cc67d6f2de554bbd0af9a4008052e45e
+size 996026489

last-checkpoint/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44df090ab45db81c35c42b16815291354adb530b8e2fdb8c60dae8dcd0293442
+size 14575

last-checkpoint/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dd12b75c68171d6b843ea8287729b2926e26ccf6f0ecf6e8495fe66f0329742
+size 627

last-checkpoint/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

last-checkpoint/tokenizer_config.json

@@ -0,0 +1,15 @@
+{
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 256,
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}

last-checkpoint/trainer_state.json

@@ -0,0 +1,340 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.074688796680498,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 87.4023,
+      "step": 100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 52.564,
+      "step": 200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 6e-06,
+      "loss": 41.0172,
+      "step": 300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 36.522,
+      "step": 400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 33.7109,
+      "step": 500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.2e-05,
+      "loss": 31.384,
+      "step": 600
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 29.2667,
+      "step": 700
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 27.6322,
+      "step": 800
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.8e-05,
+      "loss": 25.95,
+      "step": 900
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 24.3416,
+      "step": 1000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 22.5298,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.4e-05,
+      "loss": 20.7681,
+      "step": 1200
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 19.4657,
+      "step": 1300
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 17.7138,
+      "step": 1400
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3e-05,
+      "loss": 16.3022,
+      "step": 1500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 14.6267,
+      "step": 1600
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 13.0298,
+      "step": 1700
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.6e-05,
+      "loss": 11.5227,
+      "step": 1800
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.8e-05,
+      "loss": 9.8717,
+      "step": 1900
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4e-05,
+      "loss": 8.0718,
+      "step": 2000
+    },
+    {
+      "epoch": 0.83,
+      "eval_bleu": 5.8841,
+      "eval_em": 0.0,
+      "eval_gen_len": 126.98,
+      "eval_loss": 6.968929290771484,
+      "eval_rm": 0.0,
+      "eval_runtime": 84.2898,
+      "eval_samples_per_second": 0.593,
+      "eval_steps_per_second": 0.083,
+      "step": 2000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.2e-05,
+      "loss": 7.3621,
+      "step": 2100
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 6.6663,
+      "step": 2200
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 6.4464,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8e-05,
+      "loss": 6.0977,
+      "step": 2400
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 5.8656,
+      "step": 2500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.976851851851852e-05,
+      "loss": 5.6319,
+      "step": 2600
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.9537037037037035e-05,
+      "loss": 5.2821,
+      "step": 2700
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.930555555555556e-05,
+      "loss": 4.9816,
+      "step": 2800
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.9074074074074075e-05,
+      "loss": 5.0416,
+      "step": 2900
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.8842592592592595e-05,
+      "loss": 4.9786,
+      "step": 3000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.8611111111111115e-05,
+      "loss": 4.7885,
+      "step": 3100
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.837962962962963e-05,
+      "loss": 4.6081,
+      "step": 3200
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 4.4985,
+      "step": 3300
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.791666666666667e-05,
+      "loss": 4.3774,
+      "step": 3400
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.768518518518519e-05,
+      "loss": 4.2805,
+      "step": 3500
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.745370370370371e-05,
+      "loss": 3.9643,
+      "step": 3600
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 3.9951,
+      "step": 3700
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.699074074074074e-05,
+      "loss": 3.9219,
+      "step": 3800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.675925925925926e-05,
+      "loss": 3.7154,
+      "step": 3900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.652777777777778e-05,
+      "loss": 3.6551,
+      "step": 4000
+    },
+    {
+      "epoch": 1.66,
+      "eval_bleu": 20.8847,
+      "eval_em": 0.0,
+      "eval_gen_len": 30.46,
+      "eval_loss": 3.524771213531494,
+      "eval_rm": 0.0,
+      "eval_runtime": 30.3441,
+      "eval_samples_per_second": 1.648,
+      "eval_steps_per_second": 0.231,
+      "step": 4000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 3.6855,
+      "step": 4100
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.6064814814814814e-05,
+      "loss": 3.5543,
+      "step": 4200
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 3.4561,
+      "step": 4300
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.5601851851851854e-05,
+      "loss": 3.3651,
+      "step": 4400
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 4.5370370370370374e-05,
+      "loss": 3.1537,
+      "step": 4500
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 4.5138888888888894e-05,
+      "loss": 3.3103,
+      "step": 4600
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 4.490740740740741e-05,
+      "loss": 3.1797,
+      "step": 4700
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 4.467592592592593e-05,
+      "loss": 3.1758,
+      "step": 4800
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 2.9908,
+      "step": 4900
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.4212962962962966e-05,
+      "loss": 2.7989,
+      "step": 5000
+    }
+  ],
+  "max_steps": 24100,
+  "num_train_epochs": 10,
+  "total_flos": 1401098107488768.0,
+  "trial_name": null,
+  "trial_params": null
+}

last-checkpoint/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7b57b26a1bbcf325616c7b7ce00f0d02fdbc04790c3b9ca67540017bb980b08
+size 4155

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4b2f7ee2080980e630ae668e1dc510bdf00cfdff869a925b78d26ae1342fb7fb
+oid sha256:9bb592d3f4e42471c4963cdf02b330d1cc67d6f2de554bbd0af9a4008052e45e
 size 996026489

runs/May23_15-22-34_9fa8c32f042d/1684855595.7281685/events.out.tfevents.1684855595.9fa8c32f042d.2702.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6235636b961b455a4b45e8c18481c9e41fe864fea5831bd619b7c41ca4db13d5
+size 6302

runs/May23_15-22-34_9fa8c32f042d/events.out.tfevents.1684855595.9fa8c32f042d.2702.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c42d2cb59c5db701b7496602dce695c68736aa5453b1c229f3981f15281e04fb
+size 17316

tokenizer.json

@@ -6746,7 +6746,7 @@
     "clean_text": true,
     "handle_chinese_chars": true,
     "strip_accents": null,
-    "lowercase": true
+    "lowercase": false
   },
   "pre_tokenizer": {
     "type": "BertPreTokenizer"

tokenizer_config.json

@@ -2,7 +2,7 @@
   "clean_up_tokenization_spaces": true,
   "cls_token": "[CLS]",
   "do_basic_tokenize": true,
-  "do_lower_case": true,
+  "do_lower_case": false,
   "mask_token": "[MASK]",
   "model_max_length": 256,
   "never_split": null,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f39883603e0f5d5f423c83fa5b774cffc6fd43f8970ddb1f1665c98b8cf652f9
-size 4219
+oid sha256:e7b57b26a1bbcf325616c7b7ce00f0d02fdbc04790c3b9ca67540017bb980b08
+size 4155