Commit
•
53911bf
1
Parent(s):
835e889
Upload 10 files
Browse files- config.json +249 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +51 -0
- tokenizer.json +0 -0
- tokenizer_config.json +72 -0
- trainer_state.json +787 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,249 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "sentence-transformers/all-mpnet-base-v2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MPNetForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "control de bienes",
|
| 14 |
+
"1": "capacitaciones",
|
| 15 |
+
"2": "aval",
|
| 16 |
+
"3": "adquisiciones informe rt",
|
| 17 |
+
"4": "adquisiciones",
|
| 18 |
+
"5": "erp crm",
|
| 19 |
+
"6": "control de bienes prestamos",
|
| 20 |
+
"7": "control de cambios",
|
| 21 |
+
"8": "erp bienestar",
|
| 22 |
+
"9": "erp centros de produccion",
|
| 23 |
+
"10": "erp gestion academica evaluacion docente",
|
| 24 |
+
"11": "erp conceder acceso",
|
| 25 |
+
"12": "erp eliminar accesos",
|
| 26 |
+
"13": "erp financiero",
|
| 27 |
+
"14": "erp gestion academica",
|
| 28 |
+
"15": "erp no reportes",
|
| 29 |
+
"16": "erp gestion de biblioteca",
|
| 30 |
+
"17": "erp gestion logistica",
|
| 31 |
+
"18": "erp internacionalizacion",
|
| 32 |
+
"19": "erp investigacion cientifica",
|
| 33 |
+
"20": "erp talento humano",
|
| 34 |
+
"21": "erp salud ocupacional",
|
| 35 |
+
"22": "erp planificacion",
|
| 36 |
+
"23": "erp otros",
|
| 37 |
+
"24": "erp odoo monica otros",
|
| 38 |
+
"25": "evea conceder acceso",
|
| 39 |
+
"26": "erp vinculacion",
|
| 40 |
+
"27": "erp votaciones",
|
| 41 |
+
"28": "evea acceso al servicio evea",
|
| 42 |
+
"29": "evea cursos",
|
| 43 |
+
"30": "evea modificacion de la informacion",
|
| 44 |
+
"31": "evea eliminar accesos",
|
| 45 |
+
"32": "evea encuesta cuestionario tareas examenes",
|
| 46 |
+
"33": "evea interrupcion del servicio",
|
| 47 |
+
"34": "evea matriculacion en cursos",
|
| 48 |
+
"35": "evea reportes",
|
| 49 |
+
"36": "evea nuevo",
|
| 50 |
+
"37": "evea otro",
|
| 51 |
+
"38": "evea plataforma",
|
| 52 |
+
"39": "evea problema con la configuracion",
|
| 53 |
+
"40": "firma electronica quipux",
|
| 54 |
+
"41": "encuesta",
|
| 55 |
+
"42": "evea sincronizacion",
|
| 56 |
+
"43": "evea sincronizacion proveedor",
|
| 57 |
+
"44": "evea respaldos",
|
| 58 |
+
"45": "infraestructura respaldo y recuperacion",
|
| 59 |
+
"46": "gestion",
|
| 60 |
+
"47": "infraestructura conectividad",
|
| 61 |
+
"48": "infraestructura mantenimiento preventivo",
|
| 62 |
+
"49": "infraestructura uso de infraestructura fisica",
|
| 63 |
+
"50": "office 365 conceder acceso",
|
| 64 |
+
"51": "informe",
|
| 65 |
+
"52": "office 365",
|
| 66 |
+
"53": "office 365 acceso al servicio",
|
| 67 |
+
"54": "office 365 bloqueo de servicio",
|
| 68 |
+
"55": "office 365 problema con el servicio",
|
| 69 |
+
"56": "office 365 creacion de cuenta",
|
| 70 |
+
"57": "office 365 eliminar accesos",
|
| 71 |
+
"58": "office 365 interrupcion del servicio",
|
| 72 |
+
"59": "office 365 otro",
|
| 73 |
+
"60": "poa",
|
| 74 |
+
"61": "pagina web",
|
| 75 |
+
"62": "otros",
|
| 76 |
+
"63": "oficio",
|
| 77 |
+
"64": "office 365 roles y permisos",
|
| 78 |
+
"65": "reportes",
|
| 79 |
+
"66": "proveedor",
|
| 80 |
+
"67": "redes sociales",
|
| 81 |
+
"68": "proyectos",
|
| 82 |
+
"69": "repositorio base digital",
|
| 83 |
+
"70": "soporte equipos mantenimiento de hardware",
|
| 84 |
+
"71": "revistas",
|
| 85 |
+
"72": "servicios conceder accesos",
|
| 86 |
+
"73": "servicios elimnar accesos",
|
| 87 |
+
"74": "soporte equipos garantia",
|
| 88 |
+
"75": "soporte erp gestion logistica",
|
| 89 |
+
"76": "soporte equipos mantenimiento de software",
|
| 90 |
+
"77": "soporte erp bienestar",
|
| 91 |
+
"78": "soporte erp financiero",
|
| 92 |
+
"79": "soporte erp gestion academica",
|
| 93 |
+
"80": "soporte infraestructura",
|
| 94 |
+
"81": "soporte evea sincronizacion",
|
| 95 |
+
"82": "soporte erp talento humano",
|
| 96 |
+
"83": "soporte evea plataforma",
|
| 97 |
+
"84": "soporte erp otros",
|
| 98 |
+
"85": "soporte seguridad",
|
| 99 |
+
"86": "soporte otros",
|
| 100 |
+
"87": "soporte pagina web",
|
| 101 |
+
"88": "soporte repositorio base digital",
|
| 102 |
+
"89": "soporte revistas",
|
| 103 |
+
"90": "soporte odoo monica otros",
|
| 104 |
+
"91": "soporte infraestructura conectividad",
|
| 105 |
+
"92": "soporte infraestructura respaldo y recuperacion",
|
| 106 |
+
"93": "soporte mantenimiento correctivo",
|
| 107 |
+
"94": "soporte office 365",
|
| 108 |
+
"95": "streaming eventos",
|
| 109 |
+
"96": "soporte streaming eventos",
|
| 110 |
+
"97": "soporte turniting",
|
| 111 |
+
"98": "soporte zoom",
|
| 112 |
+
"99": "sin datos",
|
| 113 |
+
"100": "turnitin",
|
| 114 |
+
"101": "zoom credenciales zoom",
|
| 115 |
+
"102": "zoom interrupcion del servicio",
|
| 116 |
+
"103": "zoom licenciamiento",
|
| 117 |
+
"104": "zoom otros",
|
| 118 |
+
"105": "zoom webinar",
|
| 119 |
+
"106": "zoom conceder accesos",
|
| 120 |
+
"107": "zoom eliminar accesos",
|
| 121 |
+
"108": "zoom videoconferencia",
|
| 122 |
+
"109": "reporte informe"
|
| 123 |
+
},
|
| 124 |
+
"initializer_range": 0.02,
|
| 125 |
+
"intermediate_size": 3072,
|
| 126 |
+
"label2id": {
|
| 127 |
+
"adquisiciones": 4,
|
| 128 |
+
"adquisiciones informe rt": 3,
|
| 129 |
+
"aval": 2,
|
| 130 |
+
"capacitaciones": 1,
|
| 131 |
+
"control de bienes": 0,
|
| 132 |
+
"control de bienes prestamos": 6,
|
| 133 |
+
"control de cambios": 7,
|
| 134 |
+
"encuesta": 41,
|
| 135 |
+
"erp bienestar": 8,
|
| 136 |
+
"erp centros de produccion": 9,
|
| 137 |
+
"erp conceder acceso": 11,
|
| 138 |
+
"erp crm": 5,
|
| 139 |
+
"erp eliminar accesos": 12,
|
| 140 |
+
"erp financiero": 13,
|
| 141 |
+
"erp gestion academica": 14,
|
| 142 |
+
"erp gestion academica evaluacion docente": 10,
|
| 143 |
+
"erp gestion de biblioteca": 16,
|
| 144 |
+
"erp gestion logistica": 17,
|
| 145 |
+
"erp internacionalizacion": 18,
|
| 146 |
+
"erp investigacion cientifica": 19,
|
| 147 |
+
"erp no reportes": 15,
|
| 148 |
+
"erp odoo monica otros": 24,
|
| 149 |
+
"erp otros": 23,
|
| 150 |
+
"erp planificacion": 22,
|
| 151 |
+
"erp salud ocupacional": 21,
|
| 152 |
+
"erp talento humano": 20,
|
| 153 |
+
"erp vinculacion": 26,
|
| 154 |
+
"erp votaciones": 27,
|
| 155 |
+
"evea acceso al servicio evea": 28,
|
| 156 |
+
"evea conceder acceso": 25,
|
| 157 |
+
"evea cursos": 29,
|
| 158 |
+
"evea eliminar accesos": 31,
|
| 159 |
+
"evea encuesta cuestionario tareas examenes": 32,
|
| 160 |
+
"evea interrupcion del servicio": 33,
|
| 161 |
+
"evea matriculacion en cursos": 34,
|
| 162 |
+
"evea modificacion de la informacion": 30,
|
| 163 |
+
"evea nuevo": 36,
|
| 164 |
+
"evea otro": 37,
|
| 165 |
+
"evea plataforma": 38,
|
| 166 |
+
"evea problema con la configuracion": 39,
|
| 167 |
+
"evea reportes": 35,
|
| 168 |
+
"evea respaldos": 44,
|
| 169 |
+
"evea sincronizacion": 42,
|
| 170 |
+
"evea sincronizacion proveedor": 43,
|
| 171 |
+
"firma electronica quipux": 40,
|
| 172 |
+
"gestion": 46,
|
| 173 |
+
"informe": 51,
|
| 174 |
+
"infraestructura conectividad": 47,
|
| 175 |
+
"infraestructura mantenimiento preventivo": 48,
|
| 176 |
+
"infraestructura respaldo y recuperacion": 45,
|
| 177 |
+
"infraestructura uso de infraestructura fisica": 49,
|
| 178 |
+
"office 365": 52,
|
| 179 |
+
"office 365 acceso al servicio": 53,
|
| 180 |
+
"office 365 bloqueo de servicio": 54,
|
| 181 |
+
"office 365 conceder acceso": 50,
|
| 182 |
+
"office 365 creacion de cuenta": 56,
|
| 183 |
+
"office 365 eliminar accesos": 57,
|
| 184 |
+
"office 365 interrupcion del servicio": 58,
|
| 185 |
+
"office 365 otro": 59,
|
| 186 |
+
"office 365 problema con el servicio": 55,
|
| 187 |
+
"office 365 roles y permisos": 64,
|
| 188 |
+
"oficio": 63,
|
| 189 |
+
"otros": 62,
|
| 190 |
+
"pagina web": 61,
|
| 191 |
+
"poa": 60,
|
| 192 |
+
"proveedor": 66,
|
| 193 |
+
"proyectos": 68,
|
| 194 |
+
"redes sociales": 67,
|
| 195 |
+
"reporte informe": 109,
|
| 196 |
+
"reportes": 65,
|
| 197 |
+
"repositorio base digital": 69,
|
| 198 |
+
"revistas": 71,
|
| 199 |
+
"servicios conceder accesos": 72,
|
| 200 |
+
"servicios elimnar accesos": 73,
|
| 201 |
+
"sin datos": 99,
|
| 202 |
+
"soporte equipos garantia": 74,
|
| 203 |
+
"soporte equipos mantenimiento de hardware": 70,
|
| 204 |
+
"soporte equipos mantenimiento de software": 76,
|
| 205 |
+
"soporte erp bienestar": 77,
|
| 206 |
+
"soporte erp financiero": 78,
|
| 207 |
+
"soporte erp gestion academica": 79,
|
| 208 |
+
"soporte erp gestion logistica": 75,
|
| 209 |
+
"soporte erp otros": 84,
|
| 210 |
+
"soporte erp talento humano": 82,
|
| 211 |
+
"soporte evea plataforma": 83,
|
| 212 |
+
"soporte evea sincronizacion": 81,
|
| 213 |
+
"soporte infraestructura": 80,
|
| 214 |
+
"soporte infraestructura conectividad": 91,
|
| 215 |
+
"soporte infraestructura respaldo y recuperacion": 92,
|
| 216 |
+
"soporte mantenimiento correctivo": 93,
|
| 217 |
+
"soporte odoo monica otros": 90,
|
| 218 |
+
"soporte office 365": 94,
|
| 219 |
+
"soporte otros": 86,
|
| 220 |
+
"soporte pagina web": 87,
|
| 221 |
+
"soporte repositorio base digital": 88,
|
| 222 |
+
"soporte revistas": 89,
|
| 223 |
+
"soporte seguridad": 85,
|
| 224 |
+
"soporte streaming eventos": 96,
|
| 225 |
+
"soporte turniting": 97,
|
| 226 |
+
"soporte zoom": 98,
|
| 227 |
+
"streaming eventos": 95,
|
| 228 |
+
"turnitin": 100,
|
| 229 |
+
"zoom conceder accesos": 106,
|
| 230 |
+
"zoom credenciales zoom": 101,
|
| 231 |
+
"zoom eliminar accesos": 107,
|
| 232 |
+
"zoom interrupcion del servicio": 102,
|
| 233 |
+
"zoom licenciamiento": 103,
|
| 234 |
+
"zoom otros": 104,
|
| 235 |
+
"zoom videoconferencia": 108,
|
| 236 |
+
"zoom webinar": 105
|
| 237 |
+
},
|
| 238 |
+
"layer_norm_eps": 1e-05,
|
| 239 |
+
"max_position_embeddings": 514,
|
| 240 |
+
"model_type": "mpnet",
|
| 241 |
+
"num_attention_heads": 12,
|
| 242 |
+
"num_hidden_layers": 12,
|
| 243 |
+
"pad_token_id": 1,
|
| 244 |
+
"problem_type": "single_label_classification",
|
| 245 |
+
"relative_attention_num_buckets": 32,
|
| 246 |
+
"torch_dtype": "float32",
|
| 247 |
+
"transformers_version": "4.39.3",
|
| 248 |
+
"vocab_size": 30527
|
| 249 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3c34868aeab3207ec7c2c4e275ceadaefae544de0922c88fc3579f2be57654d
|
| 3 |
+
size 876736442
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd1fd254665e11824f0cb7c37c036de4909ac73fafa6d0dc380b4c0408a4f806
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cf06e273b5aeea9357f7cc59fe206491bb60037d5e532e3fcf971a7cf99a8c15
|
| 3 |
+
size 1064
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": true,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "[UNK]",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": true,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"104": {
|
| 36 |
+
"content": "[UNK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
},
|
| 43 |
+
"30526": {
|
| 44 |
+
"content": "<mask>",
|
| 45 |
+
"lstrip": true,
|
| 46 |
+
"normalized": false,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": true
|
| 50 |
+
}
|
| 51 |
+
},
|
| 52 |
+
"bos_token": "<s>",
|
| 53 |
+
"clean_up_tokenization_spaces": true,
|
| 54 |
+
"cls_token": "<s>",
|
| 55 |
+
"do_lower_case": true,
|
| 56 |
+
"eos_token": "</s>",
|
| 57 |
+
"mask_token": "<mask>",
|
| 58 |
+
"max_length": 128,
|
| 59 |
+
"model_max_length": 512,
|
| 60 |
+
"pad_to_multiple_of": null,
|
| 61 |
+
"pad_token": "<pad>",
|
| 62 |
+
"pad_token_type_id": 0,
|
| 63 |
+
"padding_side": "right",
|
| 64 |
+
"sep_token": "</s>",
|
| 65 |
+
"stride": 0,
|
| 66 |
+
"strip_accents": null,
|
| 67 |
+
"tokenize_chinese_chars": true,
|
| 68 |
+
"tokenizer_class": "MPNetTokenizer",
|
| 69 |
+
"truncation_side": "right",
|
| 70 |
+
"truncation_strategy": "longest_first",
|
| 71 |
+
"unk_token": "[UNK]"
|
| 72 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,787 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.022735346358792183,
|
| 3 |
+
"best_model_checkpoint": "./results/checkpoint-10634",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 10634,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"grad_norm": 0.7369279265403748,
|
| 14 |
+
"learning_rate": 4.952981004325748e-05,
|
| 15 |
+
"loss": 4.7032,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.04,
|
| 20 |
+
"grad_norm": 0.5547892451286316,
|
| 21 |
+
"learning_rate": 4.905962008651496e-05,
|
| 22 |
+
"loss": 4.7033,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.06,
|
| 27 |
+
"grad_norm": 0.5557618737220764,
|
| 28 |
+
"learning_rate": 4.858943012977243e-05,
|
| 29 |
+
"loss": 4.7015,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.08,
|
| 34 |
+
"grad_norm": 0.5398121476173401,
|
| 35 |
+
"learning_rate": 4.8119240173029906e-05,
|
| 36 |
+
"loss": 4.7038,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09,
|
| 41 |
+
"grad_norm": 0.5128409266471863,
|
| 42 |
+
"learning_rate": 4.7649050216287384e-05,
|
| 43 |
+
"loss": 4.7003,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.11,
|
| 48 |
+
"grad_norm": 0.57951420545578,
|
| 49 |
+
"learning_rate": 4.717886025954486e-05,
|
| 50 |
+
"loss": 4.7018,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.13,
|
| 55 |
+
"grad_norm": 0.48418277502059937,
|
| 56 |
+
"learning_rate": 4.670867030280233e-05,
|
| 57 |
+
"loss": 4.7005,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.15,
|
| 62 |
+
"grad_norm": 0.5264779329299927,
|
| 63 |
+
"learning_rate": 4.623848034605981e-05,
|
| 64 |
+
"loss": 4.7034,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.17,
|
| 69 |
+
"grad_norm": 0.4631044268608093,
|
| 70 |
+
"learning_rate": 4.576829038931729e-05,
|
| 71 |
+
"loss": 4.7018,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.19,
|
| 76 |
+
"grad_norm": 0.49229735136032104,
|
| 77 |
+
"learning_rate": 4.5298100432574765e-05,
|
| 78 |
+
"loss": 4.6994,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.21,
|
| 83 |
+
"grad_norm": 0.5025375485420227,
|
| 84 |
+
"learning_rate": 4.482791047583224e-05,
|
| 85 |
+
"loss": 4.7019,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.23,
|
| 90 |
+
"grad_norm": 0.5106560587882996,
|
| 91 |
+
"learning_rate": 4.4357720519089714e-05,
|
| 92 |
+
"loss": 4.7017,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.24,
|
| 97 |
+
"grad_norm": 0.4811244308948517,
|
| 98 |
+
"learning_rate": 4.388753056234719e-05,
|
| 99 |
+
"loss": 4.7015,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.26,
|
| 104 |
+
"grad_norm": 0.5047605037689209,
|
| 105 |
+
"learning_rate": 4.341734060560467e-05,
|
| 106 |
+
"loss": 4.7011,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.28,
|
| 111 |
+
"grad_norm": 0.4720272123813629,
|
| 112 |
+
"learning_rate": 4.2947150648862146e-05,
|
| 113 |
+
"loss": 4.7004,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3,
|
| 118 |
+
"grad_norm": 0.3932195007801056,
|
| 119 |
+
"learning_rate": 4.247696069211962e-05,
|
| 120 |
+
"loss": 4.7028,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.32,
|
| 125 |
+
"grad_norm": 0.39717328548431396,
|
| 126 |
+
"learning_rate": 4.2006770735377095e-05,
|
| 127 |
+
"loss": 4.702,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.34,
|
| 132 |
+
"grad_norm": 0.46395114064216614,
|
| 133 |
+
"learning_rate": 4.153658077863457e-05,
|
| 134 |
+
"loss": 4.7005,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.36,
|
| 139 |
+
"grad_norm": 0.4270722270011902,
|
| 140 |
+
"learning_rate": 4.106639082189205e-05,
|
| 141 |
+
"loss": 4.7,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.38,
|
| 146 |
+
"grad_norm": 0.4068189561367035,
|
| 147 |
+
"learning_rate": 4.059620086514952e-05,
|
| 148 |
+
"loss": 4.7016,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.39,
|
| 153 |
+
"grad_norm": 0.4495134651660919,
|
| 154 |
+
"learning_rate": 4.0126010908407e-05,
|
| 155 |
+
"loss": 4.7,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.41,
|
| 160 |
+
"grad_norm": 0.4774065613746643,
|
| 161 |
+
"learning_rate": 3.9655820951664476e-05,
|
| 162 |
+
"loss": 4.7022,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.43,
|
| 167 |
+
"grad_norm": 0.4217697083950043,
|
| 168 |
+
"learning_rate": 3.9185630994921954e-05,
|
| 169 |
+
"loss": 4.7016,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.45,
|
| 174 |
+
"grad_norm": 0.4176025092601776,
|
| 175 |
+
"learning_rate": 3.871544103817943e-05,
|
| 176 |
+
"loss": 4.7005,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.47,
|
| 181 |
+
"grad_norm": 0.43538960814476013,
|
| 182 |
+
"learning_rate": 3.82452510814369e-05,
|
| 183 |
+
"loss": 4.7002,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.49,
|
| 188 |
+
"grad_norm": 0.3962569236755371,
|
| 189 |
+
"learning_rate": 3.777506112469438e-05,
|
| 190 |
+
"loss": 4.7015,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.51,
|
| 195 |
+
"grad_norm": 0.41195255517959595,
|
| 196 |
+
"learning_rate": 3.730487116795186e-05,
|
| 197 |
+
"loss": 4.7014,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.53,
|
| 202 |
+
"grad_norm": 0.3937325179576874,
|
| 203 |
+
"learning_rate": 3.6834681211209335e-05,
|
| 204 |
+
"loss": 4.7015,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.55,
|
| 209 |
+
"grad_norm": 0.39071130752563477,
|
| 210 |
+
"learning_rate": 3.6364491254466806e-05,
|
| 211 |
+
"loss": 4.6994,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.56,
|
| 216 |
+
"grad_norm": 0.40909039974212646,
|
| 217 |
+
"learning_rate": 3.5894301297724284e-05,
|
| 218 |
+
"loss": 4.7019,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.58,
|
| 223 |
+
"grad_norm": 0.3900412321090698,
|
| 224 |
+
"learning_rate": 3.542411134098176e-05,
|
| 225 |
+
"loss": 4.6996,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.6,
|
| 230 |
+
"grad_norm": 0.5122581720352173,
|
| 231 |
+
"learning_rate": 3.495392138423923e-05,
|
| 232 |
+
"loss": 4.6992,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.62,
|
| 237 |
+
"grad_norm": 0.38495704531669617,
|
| 238 |
+
"learning_rate": 3.448373142749671e-05,
|
| 239 |
+
"loss": 4.7024,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.64,
|
| 244 |
+
"grad_norm": 0.40592727065086365,
|
| 245 |
+
"learning_rate": 3.401354147075419e-05,
|
| 246 |
+
"loss": 4.7002,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.66,
|
| 251 |
+
"grad_norm": 0.4175421893596649,
|
| 252 |
+
"learning_rate": 3.354335151401166e-05,
|
| 253 |
+
"loss": 4.6993,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.68,
|
| 258 |
+
"grad_norm": 0.43954411149024963,
|
| 259 |
+
"learning_rate": 3.3073161557269136e-05,
|
| 260 |
+
"loss": 4.7016,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.7,
|
| 265 |
+
"grad_norm": 0.501674473285675,
|
| 266 |
+
"learning_rate": 3.2602971600526614e-05,
|
| 267 |
+
"loss": 4.6995,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.71,
|
| 272 |
+
"grad_norm": 0.3901698589324951,
|
| 273 |
+
"learning_rate": 3.2132781643784084e-05,
|
| 274 |
+
"loss": 4.7006,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.73,
|
| 279 |
+
"grad_norm": 0.3993096649646759,
|
| 280 |
+
"learning_rate": 3.166259168704156e-05,
|
| 281 |
+
"loss": 4.699,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.75,
|
| 286 |
+
"grad_norm": 0.43237388134002686,
|
| 287 |
+
"learning_rate": 3.119240173029904e-05,
|
| 288 |
+
"loss": 4.7001,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.77,
|
| 293 |
+
"grad_norm": 2.5795607566833496,
|
| 294 |
+
"learning_rate": 3.072221177355652e-05,
|
| 295 |
+
"loss": 4.6928,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.79,
|
| 300 |
+
"grad_norm": 0.9452407956123352,
|
| 301 |
+
"learning_rate": 3.0252021816814e-05,
|
| 302 |
+
"loss": 4.7014,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.81,
|
| 307 |
+
"grad_norm": 0.7060949802398682,
|
| 308 |
+
"learning_rate": 2.978183186007147e-05,
|
| 309 |
+
"loss": 4.707,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.83,
|
| 314 |
+
"grad_norm": 1.052825689315796,
|
| 315 |
+
"learning_rate": 2.9311641903328947e-05,
|
| 316 |
+
"loss": 4.7014,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.85,
|
| 321 |
+
"grad_norm": 0.7108451128005981,
|
| 322 |
+
"learning_rate": 2.8841451946586424e-05,
|
| 323 |
+
"loss": 4.7027,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.87,
|
| 328 |
+
"grad_norm": 0.6421079635620117,
|
| 329 |
+
"learning_rate": 2.83712619898439e-05,
|
| 330 |
+
"loss": 4.7033,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.88,
|
| 335 |
+
"grad_norm": 0.5927799940109253,
|
| 336 |
+
"learning_rate": 2.7901072033101373e-05,
|
| 337 |
+
"loss": 4.7044,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.9,
|
| 342 |
+
"grad_norm": 0.5713520050048828,
|
| 343 |
+
"learning_rate": 2.743088207635885e-05,
|
| 344 |
+
"loss": 4.7022,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.92,
|
| 349 |
+
"grad_norm": 0.5557407140731812,
|
| 350 |
+
"learning_rate": 2.6960692119616325e-05,
|
| 351 |
+
"loss": 4.7036,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.94,
|
| 356 |
+
"grad_norm": 0.37561553716659546,
|
| 357 |
+
"learning_rate": 2.6490502162873802e-05,
|
| 358 |
+
"loss": 4.7011,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.96,
|
| 363 |
+
"grad_norm": 0.404910683631897,
|
| 364 |
+
"learning_rate": 2.602031220613128e-05,
|
| 365 |
+
"loss": 4.7007,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.98,
|
| 370 |
+
"grad_norm": 0.39087265729904175,
|
| 371 |
+
"learning_rate": 2.5550122249388754e-05,
|
| 372 |
+
"loss": 4.7021,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0,
|
| 377 |
+
"grad_norm": 0.3689398169517517,
|
| 378 |
+
"learning_rate": 2.507993229264623e-05,
|
| 379 |
+
"loss": 4.7004,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.0,
|
| 384 |
+
"eval_accuracy": 0.008644168146832444,
|
| 385 |
+
"eval_f1": 0.00014816254395838658,
|
| 386 |
+
"eval_loss": 4.700786113739014,
|
| 387 |
+
"eval_precision": 7.472164295071267e-05,
|
| 388 |
+
"eval_recall": 0.008644168146832444,
|
| 389 |
+
"eval_runtime": 96.2388,
|
| 390 |
+
"eval_samples_per_second": 87.75,
|
| 391 |
+
"eval_steps_per_second": 4.884,
|
| 392 |
+
"step": 5317
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.02,
|
| 396 |
+
"grad_norm": 0.3822825849056244,
|
| 397 |
+
"learning_rate": 2.4609742335903706e-05,
|
| 398 |
+
"loss": 4.7002,
|
| 399 |
+
"step": 5400
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.03,
|
| 403 |
+
"grad_norm": 0.421373575925827,
|
| 404 |
+
"learning_rate": 2.413955237916118e-05,
|
| 405 |
+
"loss": 4.6991,
|
| 406 |
+
"step": 5500
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.05,
|
| 410 |
+
"grad_norm": 0.4438491463661194,
|
| 411 |
+
"learning_rate": 2.3669362422418658e-05,
|
| 412 |
+
"loss": 4.7001,
|
| 413 |
+
"step": 5600
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.07,
|
| 417 |
+
"grad_norm": 0.812016487121582,
|
| 418 |
+
"learning_rate": 2.3199172465676132e-05,
|
| 419 |
+
"loss": 4.6941,
|
| 420 |
+
"step": 5700
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.09,
|
| 424 |
+
"grad_norm": 1.119163155555725,
|
| 425 |
+
"learning_rate": 2.272898250893361e-05,
|
| 426 |
+
"loss": 4.6807,
|
| 427 |
+
"step": 5800
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.11,
|
| 431 |
+
"grad_norm": 1.1431705951690674,
|
| 432 |
+
"learning_rate": 2.2258792552191084e-05,
|
| 433 |
+
"loss": 4.6842,
|
| 434 |
+
"step": 5900
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.13,
|
| 438 |
+
"grad_norm": 1.4402434825897217,
|
| 439 |
+
"learning_rate": 2.1788602595448562e-05,
|
| 440 |
+
"loss": 4.6653,
|
| 441 |
+
"step": 6000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.15,
|
| 445 |
+
"grad_norm": 2.073824644088745,
|
| 446 |
+
"learning_rate": 2.131841263870604e-05,
|
| 447 |
+
"loss": 4.6543,
|
| 448 |
+
"step": 6100
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.17,
|
| 452 |
+
"grad_norm": 1.6333357095718384,
|
| 453 |
+
"learning_rate": 2.0848222681963514e-05,
|
| 454 |
+
"loss": 4.6394,
|
| 455 |
+
"step": 6200
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.18,
|
| 459 |
+
"grad_norm": 1.206061840057373,
|
| 460 |
+
"learning_rate": 2.037803272522099e-05,
|
| 461 |
+
"loss": 4.6213,
|
| 462 |
+
"step": 6300
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 1.2,
|
| 466 |
+
"grad_norm": 0.5230748653411865,
|
| 467 |
+
"learning_rate": 1.9907842768478466e-05,
|
| 468 |
+
"loss": 4.7044,
|
| 469 |
+
"step": 6400
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 1.22,
|
| 473 |
+
"grad_norm": 0.4634922444820404,
|
| 474 |
+
"learning_rate": 1.9437652811735943e-05,
|
| 475 |
+
"loss": 4.7031,
|
| 476 |
+
"step": 6500
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.24,
|
| 480 |
+
"grad_norm": 0.47288939356803894,
|
| 481 |
+
"learning_rate": 1.8967462854993417e-05,
|
| 482 |
+
"loss": 4.7027,
|
| 483 |
+
"step": 6600
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.26,
|
| 487 |
+
"grad_norm": 0.4731047749519348,
|
| 488 |
+
"learning_rate": 1.8497272898250895e-05,
|
| 489 |
+
"loss": 4.7014,
|
| 490 |
+
"step": 6700
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 1.28,
|
| 494 |
+
"grad_norm": 0.5110020637512207,
|
| 495 |
+
"learning_rate": 1.802708294150837e-05,
|
| 496 |
+
"loss": 4.6995,
|
| 497 |
+
"step": 6800
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.3,
|
| 501 |
+
"grad_norm": 0.5348946452140808,
|
| 502 |
+
"learning_rate": 1.7556892984765847e-05,
|
| 503 |
+
"loss": 4.7013,
|
| 504 |
+
"step": 6900
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.32,
|
| 508 |
+
"grad_norm": 0.5182259678840637,
|
| 509 |
+
"learning_rate": 1.708670302802332e-05,
|
| 510 |
+
"loss": 4.6991,
|
| 511 |
+
"step": 7000
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 1.34,
|
| 515 |
+
"grad_norm": 0.6097771525382996,
|
| 516 |
+
"learning_rate": 1.66165130712808e-05,
|
| 517 |
+
"loss": 4.7022,
|
| 518 |
+
"step": 7100
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.35,
|
| 522 |
+
"grad_norm": 0.49007490277290344,
|
| 523 |
+
"learning_rate": 1.6146323114538276e-05,
|
| 524 |
+
"loss": 4.7012,
|
| 525 |
+
"step": 7200
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 1.37,
|
| 529 |
+
"grad_norm": 0.5180237293243408,
|
| 530 |
+
"learning_rate": 1.567613315779575e-05,
|
| 531 |
+
"loss": 4.6998,
|
| 532 |
+
"step": 7300
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 1.39,
|
| 536 |
+
"grad_norm": 0.4303135275840759,
|
| 537 |
+
"learning_rate": 1.5205943201053227e-05,
|
| 538 |
+
"loss": 4.7025,
|
| 539 |
+
"step": 7400
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.41,
|
| 543 |
+
"grad_norm": 0.6103301644325256,
|
| 544 |
+
"learning_rate": 1.47357532443107e-05,
|
| 545 |
+
"loss": 4.7027,
|
| 546 |
+
"step": 7500
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 1.43,
|
| 550 |
+
"grad_norm": 0.5991978049278259,
|
| 551 |
+
"learning_rate": 1.4265563287568178e-05,
|
| 552 |
+
"loss": 4.7016,
|
| 553 |
+
"step": 7600
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 1.45,
|
| 557 |
+
"grad_norm": 0.590713381767273,
|
| 558 |
+
"learning_rate": 1.3795373330825653e-05,
|
| 559 |
+
"loss": 4.6992,
|
| 560 |
+
"step": 7700
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.47,
|
| 564 |
+
"grad_norm": 0.6496405601501465,
|
| 565 |
+
"learning_rate": 1.332518337408313e-05,
|
| 566 |
+
"loss": 4.7005,
|
| 567 |
+
"step": 7800
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 1.49,
|
| 571 |
+
"grad_norm": 0.5405462980270386,
|
| 572 |
+
"learning_rate": 1.2854993417340605e-05,
|
| 573 |
+
"loss": 4.704,
|
| 574 |
+
"step": 7900
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.5,
|
| 578 |
+
"grad_norm": 0.5655048489570618,
|
| 579 |
+
"learning_rate": 1.2384803460598082e-05,
|
| 580 |
+
"loss": 4.701,
|
| 581 |
+
"step": 8000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 1.52,
|
| 585 |
+
"grad_norm": 1.1468925476074219,
|
| 586 |
+
"learning_rate": 1.1914613503855558e-05,
|
| 587 |
+
"loss": 4.6547,
|
| 588 |
+
"step": 8100
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 1.54,
|
| 592 |
+
"grad_norm": 3.2609410285949707,
|
| 593 |
+
"learning_rate": 1.1444423547113034e-05,
|
| 594 |
+
"loss": 4.6054,
|
| 595 |
+
"step": 8200
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 1.56,
|
| 599 |
+
"grad_norm": 3.0458478927612305,
|
| 600 |
+
"learning_rate": 1.097423359037051e-05,
|
| 601 |
+
"loss": 4.5575,
|
| 602 |
+
"step": 8300
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.58,
|
| 606 |
+
"grad_norm": 4.3110527992248535,
|
| 607 |
+
"learning_rate": 1.0504043633627986e-05,
|
| 608 |
+
"loss": 4.5948,
|
| 609 |
+
"step": 8400
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 1.6,
|
| 613 |
+
"grad_norm": 1.7174781560897827,
|
| 614 |
+
"learning_rate": 1.0033853676885462e-05,
|
| 615 |
+
"loss": 4.5585,
|
| 616 |
+
"step": 8500
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 1.62,
|
| 620 |
+
"grad_norm": 3.2902700901031494,
|
| 621 |
+
"learning_rate": 9.56366372014294e-06,
|
| 622 |
+
"loss": 4.5886,
|
| 623 |
+
"step": 8600
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 1.64,
|
| 627 |
+
"grad_norm": 1.1743800640106201,
|
| 628 |
+
"learning_rate": 9.093473763400415e-06,
|
| 629 |
+
"loss": 4.5534,
|
| 630 |
+
"step": 8700
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 1.66,
|
| 634 |
+
"grad_norm": 1.5308388471603394,
|
| 635 |
+
"learning_rate": 8.623283806657891e-06,
|
| 636 |
+
"loss": 4.547,
|
| 637 |
+
"step": 8800
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 1.67,
|
| 641 |
+
"grad_norm": 2.839127540588379,
|
| 642 |
+
"learning_rate": 8.153093849915366e-06,
|
| 643 |
+
"loss": 4.5339,
|
| 644 |
+
"step": 8900
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.69,
|
| 648 |
+
"grad_norm": 1.9120197296142578,
|
| 649 |
+
"learning_rate": 7.682903893172842e-06,
|
| 650 |
+
"loss": 4.5439,
|
| 651 |
+
"step": 9000
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 1.71,
|
| 655 |
+
"grad_norm": 1.549185872077942,
|
| 656 |
+
"learning_rate": 7.212713936430318e-06,
|
| 657 |
+
"loss": 4.5276,
|
| 658 |
+
"step": 9100
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 1.73,
|
| 662 |
+
"grad_norm": 2.3087658882141113,
|
| 663 |
+
"learning_rate": 6.742523979687794e-06,
|
| 664 |
+
"loss": 4.5081,
|
| 665 |
+
"step": 9200
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 1.75,
|
| 669 |
+
"grad_norm": 1.4830306768417358,
|
| 670 |
+
"learning_rate": 6.27233402294527e-06,
|
| 671 |
+
"loss": 4.494,
|
| 672 |
+
"step": 9300
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.77,
|
| 676 |
+
"grad_norm": 2.1838178634643555,
|
| 677 |
+
"learning_rate": 5.802144066202746e-06,
|
| 678 |
+
"loss": 4.4904,
|
| 679 |
+
"step": 9400
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 1.79,
|
| 683 |
+
"grad_norm": 3.3637304306030273,
|
| 684 |
+
"learning_rate": 5.331954109460222e-06,
|
| 685 |
+
"loss": 4.4807,
|
| 686 |
+
"step": 9500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 1.81,
|
| 690 |
+
"grad_norm": 1.637593388557434,
|
| 691 |
+
"learning_rate": 4.861764152717698e-06,
|
| 692 |
+
"loss": 4.474,
|
| 693 |
+
"step": 9600
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 1.82,
|
| 697 |
+
"grad_norm": 1.833465576171875,
|
| 698 |
+
"learning_rate": 4.391574195975174e-06,
|
| 699 |
+
"loss": 4.4719,
|
| 700 |
+
"step": 9700
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 1.84,
|
| 704 |
+
"grad_norm": 2.6578447818756104,
|
| 705 |
+
"learning_rate": 3.921384239232651e-06,
|
| 706 |
+
"loss": 4.4593,
|
| 707 |
+
"step": 9800
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 1.86,
|
| 711 |
+
"grad_norm": 2.121244192123413,
|
| 712 |
+
"learning_rate": 3.451194282490126e-06,
|
| 713 |
+
"loss": 4.4514,
|
| 714 |
+
"step": 9900
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 1.88,
|
| 718 |
+
"grad_norm": 2.080278158187866,
|
| 719 |
+
"learning_rate": 2.981004325747602e-06,
|
| 720 |
+
"loss": 4.4407,
|
| 721 |
+
"step": 10000
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 1.9,
|
| 725 |
+
"grad_norm": 2.2840731143951416,
|
| 726 |
+
"learning_rate": 2.510814369005078e-06,
|
| 727 |
+
"loss": 4.4289,
|
| 728 |
+
"step": 10100
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.92,
|
| 732 |
+
"grad_norm": 2.203077554702759,
|
| 733 |
+
"learning_rate": 2.0406244122625544e-06,
|
| 734 |
+
"loss": 4.4071,
|
| 735 |
+
"step": 10200
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 1.94,
|
| 739 |
+
"grad_norm": 2.4338624477386475,
|
| 740 |
+
"learning_rate": 1.5704344555200301e-06,
|
| 741 |
+
"loss": 4.4166,
|
| 742 |
+
"step": 10300
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 1.96,
|
| 746 |
+
"grad_norm": 1.855125069618225,
|
| 747 |
+
"learning_rate": 1.1002444987775063e-06,
|
| 748 |
+
"loss": 4.4079,
|
| 749 |
+
"step": 10400
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 1.97,
|
| 753 |
+
"grad_norm": 2.9419732093811035,
|
| 754 |
+
"learning_rate": 6.300545420349821e-07,
|
| 755 |
+
"loss": 4.4227,
|
| 756 |
+
"step": 10500
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 1.99,
|
| 760 |
+
"grad_norm": 3.1695730686187744,
|
| 761 |
+
"learning_rate": 1.5986458529245815e-07,
|
| 762 |
+
"loss": 4.4244,
|
| 763 |
+
"step": 10600
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 2.0,
|
| 767 |
+
"eval_accuracy": 0.022735346358792183,
|
| 768 |
+
"eval_f1": 0.00212270922930165,
|
| 769 |
+
"eval_loss": 4.410290241241455,
|
| 770 |
+
"eval_precision": 0.0011595953132254675,
|
| 771 |
+
"eval_recall": 0.022735346358792183,
|
| 772 |
+
"eval_runtime": 96.0024,
|
| 773 |
+
"eval_samples_per_second": 87.967,
|
| 774 |
+
"eval_steps_per_second": 4.896,
|
| 775 |
+
"step": 10634
|
| 776 |
+
}
|
| 777 |
+
],
|
| 778 |
+
"logging_steps": 100,
|
| 779 |
+
"max_steps": 10634,
|
| 780 |
+
"num_input_tokens_seen": 0,
|
| 781 |
+
"num_train_epochs": 2,
|
| 782 |
+
"save_steps": 500,
|
| 783 |
+
"total_flos": 5.041092263921664e+16,
|
| 784 |
+
"train_batch_size": 18,
|
| 785 |
+
"trial_name": null,
|
| 786 |
+
"trial_params": null
|
| 787 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7ac24511dda067ae8ccff9fc3cfe145f435c4e027dd443fec4e9340547a2236d
|
| 3 |
+
size 4856
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|