Training in progress, epoch 6
Browse files- model.safetensors +1 -1
- run-0/checkpoint-5724/config.json +302 -302
- run-0/checkpoint-5724/model.safetensors +1 -1
- run-0/checkpoint-5724/optimizer.pt +1 -1
- run-0/checkpoint-5724/scaler.pt +1 -1
- run-0/checkpoint-5724/scheduler.pt +1 -1
- run-0/checkpoint-5724/trainer_state.json +69 -104
- run-0/checkpoint-5724/training_args.bin +2 -2
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 598898116
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81ac3f88bdb451e4316877bdf137e199de70b195697d9835e23b0052e2b2bace
|
| 3 |
size 598898116
|
run-0/checkpoint-5724/config.json
CHANGED
|
@@ -21,313 +21,313 @@
|
|
| 21 |
"hidden_activation": "gelu",
|
| 22 |
"hidden_size": 768,
|
| 23 |
"id2label": {
|
| 24 |
-
"0": "
|
| 25 |
-
"1": "
|
| 26 |
-
"2": "
|
| 27 |
-
"3": "
|
| 28 |
-
"4": "
|
| 29 |
-
"5": "
|
| 30 |
-
"6": "
|
| 31 |
-
"7": "
|
| 32 |
-
"8": "
|
| 33 |
-
"9": "
|
| 34 |
-
"10": "
|
| 35 |
-
"11": "
|
| 36 |
-
"12": "
|
| 37 |
-
"13": "
|
| 38 |
-
"14": "
|
| 39 |
-
"15": "
|
| 40 |
-
"16": "
|
| 41 |
-
"17": "
|
| 42 |
-
"18": "
|
| 43 |
-
"19": "
|
| 44 |
-
"20": "
|
| 45 |
-
"21": "
|
| 46 |
-
"22": "
|
| 47 |
-
"23": "
|
| 48 |
-
"24": "
|
| 49 |
-
"25": "
|
| 50 |
-
"26": "
|
| 51 |
-
"27": "
|
| 52 |
-
"28": "
|
| 53 |
-
"29": "
|
| 54 |
-
"30": "
|
| 55 |
-
"31": "
|
| 56 |
-
"32": "
|
| 57 |
-
"33": "
|
| 58 |
-
"34": "
|
| 59 |
-
"35": "
|
| 60 |
-
"36": "
|
| 61 |
-
"37": "
|
| 62 |
-
"38": "
|
| 63 |
-
"39": "
|
| 64 |
-
"40": "
|
| 65 |
-
"41": "
|
| 66 |
-
"42": "
|
| 67 |
-
"43": "
|
| 68 |
-
"44": "
|
| 69 |
-
"45": "
|
| 70 |
-
"46": "
|
| 71 |
-
"47": "
|
| 72 |
-
"48": "
|
| 73 |
-
"49": "
|
| 74 |
-
"50": "
|
| 75 |
-
"51": "
|
| 76 |
-
"52": "
|
| 77 |
-
"53": "
|
| 78 |
-
"54": "
|
| 79 |
-
"55": "
|
| 80 |
-
"56": "
|
| 81 |
-
"57": "
|
| 82 |
-
"58": "
|
| 83 |
-
"59": "
|
| 84 |
-
"60": "
|
| 85 |
-
"61": "
|
| 86 |
-
"62": "
|
| 87 |
-
"63": "
|
| 88 |
-
"64": "
|
| 89 |
-
"65": "
|
| 90 |
-
"66": "
|
| 91 |
-
"67": "
|
| 92 |
-
"68": "
|
| 93 |
-
"69": "
|
| 94 |
-
"70": "
|
| 95 |
-
"71": "
|
| 96 |
-
"72": "
|
| 97 |
-
"73": "
|
| 98 |
-
"74": "
|
| 99 |
-
"75": "
|
| 100 |
-
"76": "
|
| 101 |
-
"77": "
|
| 102 |
-
"78": "
|
| 103 |
-
"79": "
|
| 104 |
-
"80": "
|
| 105 |
-
"81": "
|
| 106 |
-
"82": "
|
| 107 |
-
"83": "
|
| 108 |
-
"84": "
|
| 109 |
-
"85": "
|
| 110 |
-
"86": "
|
| 111 |
-
"87": "
|
| 112 |
-
"88": "
|
| 113 |
-
"89": "
|
| 114 |
-
"90": "
|
| 115 |
-
"91": "
|
| 116 |
-
"92": "
|
| 117 |
-
"93": "
|
| 118 |
-
"94": "
|
| 119 |
-
"95": "
|
| 120 |
-
"96": "
|
| 121 |
-
"97": "
|
| 122 |
-
"98": "
|
| 123 |
-
"99": "
|
| 124 |
-
"100": "
|
| 125 |
-
"101": "
|
| 126 |
-
"102": "
|
| 127 |
-
"103": "
|
| 128 |
-
"104": "
|
| 129 |
-
"105": "
|
| 130 |
-
"106": "
|
| 131 |
-
"107": "
|
| 132 |
-
"108": "
|
| 133 |
-
"109": "
|
| 134 |
-
"110": "
|
| 135 |
-
"111": "
|
| 136 |
-
"112": "
|
| 137 |
-
"113": "
|
| 138 |
-
"114": "
|
| 139 |
-
"115": "
|
| 140 |
-
"116": "
|
| 141 |
-
"117": "
|
| 142 |
-
"118": "
|
| 143 |
-
"119": "
|
| 144 |
-
"120": "
|
| 145 |
-
"121": "
|
| 146 |
-
"122": "
|
| 147 |
-
"123": "
|
| 148 |
-
"124": "
|
| 149 |
-
"125": "
|
| 150 |
-
"126": "
|
| 151 |
-
"127": "
|
| 152 |
-
"128": "
|
| 153 |
-
"129": "
|
| 154 |
-
"130": "
|
| 155 |
-
"131": "
|
| 156 |
-
"132": "
|
| 157 |
-
"133": "
|
| 158 |
-
"134": "
|
| 159 |
-
"135": "
|
| 160 |
-
"136": "
|
| 161 |
-
"137": "
|
| 162 |
-
"138": "
|
| 163 |
-
"139": "
|
| 164 |
-
"140": "
|
| 165 |
-
"141": "
|
| 166 |
-
"142": "
|
| 167 |
-
"143": "
|
| 168 |
-
"144": "
|
| 169 |
-
"145": "
|
| 170 |
-
"146": "
|
| 171 |
-
"147": "
|
| 172 |
-
"148": "
|
| 173 |
-
"149": "
|
| 174 |
-
"150": "
|
| 175 |
},
|
| 176 |
"initializer_cutoff_factor": 2.0,
|
| 177 |
"initializer_range": 0.02,
|
| 178 |
"intermediate_size": 1152,
|
| 179 |
"label2id": {
|
| 180 |
-
"
|
| 181 |
-
"
|
| 182 |
-
"
|
| 183 |
-
"
|
| 184 |
-
"
|
| 185 |
-
"
|
| 186 |
-
"
|
| 187 |
-
"
|
| 188 |
-
"
|
| 189 |
-
"
|
| 190 |
-
"
|
| 191 |
-
"
|
| 192 |
-
"
|
| 193 |
-
"
|
| 194 |
-
"
|
| 195 |
-
"
|
| 196 |
-
"
|
| 197 |
-
"
|
| 198 |
-
"
|
| 199 |
-
"
|
| 200 |
-
"
|
| 201 |
-
"
|
| 202 |
-
"
|
| 203 |
-
"
|
| 204 |
-
"
|
| 205 |
-
"
|
| 206 |
-
"
|
| 207 |
-
"
|
| 208 |
-
"
|
| 209 |
-
"
|
| 210 |
-
"
|
| 211 |
-
"
|
| 212 |
-
"
|
| 213 |
-
"
|
| 214 |
-
"
|
| 215 |
-
"
|
| 216 |
-
"
|
| 217 |
-
"
|
| 218 |
-
"
|
| 219 |
-
"
|
| 220 |
-
"
|
| 221 |
-
"
|
| 222 |
-
"
|
| 223 |
-
"
|
| 224 |
-
"
|
| 225 |
-
"
|
| 226 |
-
"
|
| 227 |
-
"
|
| 228 |
-
"
|
| 229 |
-
"
|
| 230 |
-
"
|
| 231 |
-
"
|
| 232 |
-
"
|
| 233 |
-
"
|
| 234 |
-
"
|
| 235 |
-
"
|
| 236 |
-
"
|
| 237 |
-
"
|
| 238 |
-
"
|
| 239 |
-
"
|
| 240 |
-
"
|
| 241 |
-
"
|
| 242 |
-
"
|
| 243 |
-
"
|
| 244 |
-
"
|
| 245 |
-
"
|
| 246 |
-
"
|
| 247 |
-
"
|
| 248 |
-
"
|
| 249 |
-
"
|
| 250 |
-
"
|
| 251 |
-
"
|
| 252 |
-
"
|
| 253 |
-
"
|
| 254 |
-
"
|
| 255 |
-
"
|
| 256 |
-
"
|
| 257 |
-
"
|
| 258 |
-
"
|
| 259 |
-
"
|
| 260 |
-
"
|
| 261 |
-
"
|
| 262 |
-
"
|
| 263 |
-
"
|
| 264 |
-
"
|
| 265 |
-
"
|
| 266 |
-
"
|
| 267 |
-
"
|
| 268 |
-
"
|
| 269 |
-
"
|
| 270 |
-
"
|
| 271 |
-
"
|
| 272 |
-
"
|
| 273 |
-
"
|
| 274 |
-
"
|
| 275 |
-
"
|
| 276 |
-
"
|
| 277 |
-
"
|
| 278 |
-
"
|
| 279 |
-
"
|
| 280 |
-
"
|
| 281 |
-
"
|
| 282 |
-
"
|
| 283 |
-
"
|
| 284 |
-
"
|
| 285 |
-
"
|
| 286 |
-
"
|
| 287 |
-
"
|
| 288 |
-
"
|
| 289 |
-
"
|
| 290 |
-
"
|
| 291 |
-
"
|
| 292 |
-
"
|
| 293 |
-
"
|
| 294 |
-
"
|
| 295 |
-
"
|
| 296 |
-
"
|
| 297 |
-
"
|
| 298 |
-
"
|
| 299 |
-
"
|
| 300 |
-
"
|
| 301 |
-
"
|
| 302 |
-
"
|
| 303 |
-
"
|
| 304 |
-
"
|
| 305 |
-
"
|
| 306 |
-
"
|
| 307 |
-
"
|
| 308 |
-
"
|
| 309 |
-
"
|
| 310 |
-
"
|
| 311 |
-
"
|
| 312 |
-
"
|
| 313 |
-
"
|
| 314 |
-
"
|
| 315 |
-
"
|
| 316 |
-
"
|
| 317 |
-
"
|
| 318 |
-
"
|
| 319 |
-
"
|
| 320 |
-
"
|
| 321 |
-
"
|
| 322 |
-
"
|
| 323 |
-
"
|
| 324 |
-
"
|
| 325 |
-
"
|
| 326 |
-
"
|
| 327 |
-
"
|
| 328 |
-
"
|
| 329 |
-
"
|
| 330 |
-
"
|
| 331 |
},
|
| 332 |
"layer_norm_eps": 1e-05,
|
| 333 |
"local_attention": 128,
|
|
|
|
| 21 |
"hidden_activation": "gelu",
|
| 22 |
"hidden_size": 768,
|
| 23 |
"id2label": {
|
| 24 |
+
"0": "restaurant_reviews",
|
| 25 |
+
"1": "nutrition_info",
|
| 26 |
+
"2": "account_blocked",
|
| 27 |
+
"3": "oil_change_how",
|
| 28 |
+
"4": "time",
|
| 29 |
+
"5": "weather",
|
| 30 |
+
"6": "redeem_rewards",
|
| 31 |
+
"7": "interest_rate",
|
| 32 |
+
"8": "gas_type",
|
| 33 |
+
"9": "accept_reservations",
|
| 34 |
+
"10": "smart_home",
|
| 35 |
+
"11": "user_name",
|
| 36 |
+
"12": "report_lost_card",
|
| 37 |
+
"13": "repeat",
|
| 38 |
+
"14": "whisper_mode",
|
| 39 |
+
"15": "what_are_your_hobbies",
|
| 40 |
+
"16": "order",
|
| 41 |
+
"17": "jump_start",
|
| 42 |
+
"18": "schedule_meeting",
|
| 43 |
+
"19": "meeting_schedule",
|
| 44 |
+
"20": "freeze_account",
|
| 45 |
+
"21": "what_song",
|
| 46 |
+
"22": "meaning_of_life",
|
| 47 |
+
"23": "restaurant_reservation",
|
| 48 |
+
"24": "traffic",
|
| 49 |
+
"25": "make_call",
|
| 50 |
+
"26": "text",
|
| 51 |
+
"27": "bill_balance",
|
| 52 |
+
"28": "improve_credit_score",
|
| 53 |
+
"29": "change_language",
|
| 54 |
+
"30": "no",
|
| 55 |
+
"31": "measurement_conversion",
|
| 56 |
+
"32": "timer",
|
| 57 |
+
"33": "flip_coin",
|
| 58 |
+
"34": "do_you_have_pets",
|
| 59 |
+
"35": "balance",
|
| 60 |
+
"36": "tell_joke",
|
| 61 |
+
"37": "last_maintenance",
|
| 62 |
+
"38": "exchange_rate",
|
| 63 |
+
"39": "uber",
|
| 64 |
+
"40": "car_rental",
|
| 65 |
+
"41": "credit_limit",
|
| 66 |
+
"42": "oos",
|
| 67 |
+
"43": "shopping_list",
|
| 68 |
+
"44": "expiration_date",
|
| 69 |
+
"45": "routing",
|
| 70 |
+
"46": "meal_suggestion",
|
| 71 |
+
"47": "tire_change",
|
| 72 |
+
"48": "todo_list",
|
| 73 |
+
"49": "card_declined",
|
| 74 |
+
"50": "rewards_balance",
|
| 75 |
+
"51": "change_accent",
|
| 76 |
+
"52": "vaccines",
|
| 77 |
+
"53": "reminder_update",
|
| 78 |
+
"54": "food_last",
|
| 79 |
+
"55": "change_ai_name",
|
| 80 |
+
"56": "bill_due",
|
| 81 |
+
"57": "who_do_you_work_for",
|
| 82 |
+
"58": "share_location",
|
| 83 |
+
"59": "international_visa",
|
| 84 |
+
"60": "calendar",
|
| 85 |
+
"61": "translate",
|
| 86 |
+
"62": "carry_on",
|
| 87 |
+
"63": "book_flight",
|
| 88 |
+
"64": "insurance_change",
|
| 89 |
+
"65": "todo_list_update",
|
| 90 |
+
"66": "timezone",
|
| 91 |
+
"67": "cancel_reservation",
|
| 92 |
+
"68": "transactions",
|
| 93 |
+
"69": "credit_score",
|
| 94 |
+
"70": "report_fraud",
|
| 95 |
+
"71": "spending_history",
|
| 96 |
+
"72": "directions",
|
| 97 |
+
"73": "spelling",
|
| 98 |
+
"74": "insurance",
|
| 99 |
+
"75": "what_is_your_name",
|
| 100 |
+
"76": "reminder",
|
| 101 |
+
"77": "where_are_you_from",
|
| 102 |
+
"78": "distance",
|
| 103 |
+
"79": "payday",
|
| 104 |
+
"80": "flight_status",
|
| 105 |
+
"81": "find_phone",
|
| 106 |
+
"82": "greeting",
|
| 107 |
+
"83": "alarm",
|
| 108 |
+
"84": "order_status",
|
| 109 |
+
"85": "confirm_reservation",
|
| 110 |
+
"86": "cook_time",
|
| 111 |
+
"87": "damaged_card",
|
| 112 |
+
"88": "reset_settings",
|
| 113 |
+
"89": "pin_change",
|
| 114 |
+
"90": "replacement_card_duration",
|
| 115 |
+
"91": "new_card",
|
| 116 |
+
"92": "roll_dice",
|
| 117 |
+
"93": "income",
|
| 118 |
+
"94": "taxes",
|
| 119 |
+
"95": "date",
|
| 120 |
+
"96": "who_made_you",
|
| 121 |
+
"97": "pto_request",
|
| 122 |
+
"98": "tire_pressure",
|
| 123 |
+
"99": "how_old_are_you",
|
| 124 |
+
"100": "rollover_401k",
|
| 125 |
+
"101": "pto_request_status",
|
| 126 |
+
"102": "how_busy",
|
| 127 |
+
"103": "application_status",
|
| 128 |
+
"104": "recipe",
|
| 129 |
+
"105": "calendar_update",
|
| 130 |
+
"106": "play_music",
|
| 131 |
+
"107": "yes",
|
| 132 |
+
"108": "direct_deposit",
|
| 133 |
+
"109": "credit_limit_change",
|
| 134 |
+
"110": "gas",
|
| 135 |
+
"111": "pay_bill",
|
| 136 |
+
"112": "ingredients_list",
|
| 137 |
+
"113": "lost_luggage",
|
| 138 |
+
"114": "goodbye",
|
| 139 |
+
"115": "what_can_i_ask_you",
|
| 140 |
+
"116": "book_hotel",
|
| 141 |
+
"117": "are_you_a_bot",
|
| 142 |
+
"118": "next_song",
|
| 143 |
+
"119": "change_speed",
|
| 144 |
+
"120": "plug_type",
|
| 145 |
+
"121": "maybe",
|
| 146 |
+
"122": "w2",
|
| 147 |
+
"123": "oil_change_when",
|
| 148 |
+
"124": "thank_you",
|
| 149 |
+
"125": "shopping_list_update",
|
| 150 |
+
"126": "pto_balance",
|
| 151 |
+
"127": "order_checks",
|
| 152 |
+
"128": "travel_alert",
|
| 153 |
+
"129": "fun_fact",
|
| 154 |
+
"130": "sync_device",
|
| 155 |
+
"131": "schedule_maintenance",
|
| 156 |
+
"132": "apr",
|
| 157 |
+
"133": "transfer",
|
| 158 |
+
"134": "ingredient_substitution",
|
| 159 |
+
"135": "calories",
|
| 160 |
+
"136": "current_location",
|
| 161 |
+
"137": "international_fees",
|
| 162 |
+
"138": "calculator",
|
| 163 |
+
"139": "definition",
|
| 164 |
+
"140": "next_holiday",
|
| 165 |
+
"141": "update_playlist",
|
| 166 |
+
"142": "mpg",
|
| 167 |
+
"143": "min_payment",
|
| 168 |
+
"144": "change_user_name",
|
| 169 |
+
"145": "restaurant_suggestion",
|
| 170 |
+
"146": "travel_notification",
|
| 171 |
+
"147": "cancel",
|
| 172 |
+
"148": "pto_used",
|
| 173 |
+
"149": "travel_suggestion",
|
| 174 |
+
"150": "change_volume"
|
| 175 |
},
|
| 176 |
"initializer_cutoff_factor": 2.0,
|
| 177 |
"initializer_range": 0.02,
|
| 178 |
"intermediate_size": 1152,
|
| 179 |
"label2id": {
|
| 180 |
+
"accept_reservations": 9,
|
| 181 |
+
"account_blocked": 2,
|
| 182 |
+
"alarm": 83,
|
| 183 |
+
"application_status": 103,
|
| 184 |
+
"apr": 132,
|
| 185 |
+
"are_you_a_bot": 117,
|
| 186 |
+
"balance": 35,
|
| 187 |
+
"bill_balance": 27,
|
| 188 |
+
"bill_due": 56,
|
| 189 |
+
"book_flight": 63,
|
| 190 |
+
"book_hotel": 116,
|
| 191 |
+
"calculator": 138,
|
| 192 |
+
"calendar": 60,
|
| 193 |
+
"calendar_update": 105,
|
| 194 |
+
"calories": 135,
|
| 195 |
+
"cancel": 147,
|
| 196 |
+
"cancel_reservation": 67,
|
| 197 |
+
"car_rental": 40,
|
| 198 |
+
"card_declined": 49,
|
| 199 |
+
"carry_on": 62,
|
| 200 |
+
"change_accent": 51,
|
| 201 |
+
"change_ai_name": 55,
|
| 202 |
+
"change_language": 29,
|
| 203 |
+
"change_speed": 119,
|
| 204 |
+
"change_user_name": 144,
|
| 205 |
+
"change_volume": 150,
|
| 206 |
+
"confirm_reservation": 85,
|
| 207 |
+
"cook_time": 86,
|
| 208 |
+
"credit_limit": 41,
|
| 209 |
+
"credit_limit_change": 109,
|
| 210 |
+
"credit_score": 69,
|
| 211 |
+
"current_location": 136,
|
| 212 |
+
"damaged_card": 87,
|
| 213 |
+
"date": 95,
|
| 214 |
+
"definition": 139,
|
| 215 |
+
"direct_deposit": 108,
|
| 216 |
+
"directions": 72,
|
| 217 |
+
"distance": 78,
|
| 218 |
+
"do_you_have_pets": 34,
|
| 219 |
+
"exchange_rate": 38,
|
| 220 |
+
"expiration_date": 44,
|
| 221 |
+
"find_phone": 81,
|
| 222 |
+
"flight_status": 80,
|
| 223 |
+
"flip_coin": 33,
|
| 224 |
+
"food_last": 54,
|
| 225 |
+
"freeze_account": 20,
|
| 226 |
+
"fun_fact": 129,
|
| 227 |
+
"gas": 110,
|
| 228 |
+
"gas_type": 8,
|
| 229 |
+
"goodbye": 114,
|
| 230 |
+
"greeting": 82,
|
| 231 |
+
"how_busy": 102,
|
| 232 |
+
"how_old_are_you": 99,
|
| 233 |
+
"improve_credit_score": 28,
|
| 234 |
+
"income": 93,
|
| 235 |
+
"ingredient_substitution": 134,
|
| 236 |
+
"ingredients_list": 112,
|
| 237 |
+
"insurance": 74,
|
| 238 |
+
"insurance_change": 64,
|
| 239 |
+
"interest_rate": 7,
|
| 240 |
+
"international_fees": 137,
|
| 241 |
+
"international_visa": 59,
|
| 242 |
+
"jump_start": 17,
|
| 243 |
+
"last_maintenance": 37,
|
| 244 |
+
"lost_luggage": 113,
|
| 245 |
+
"make_call": 25,
|
| 246 |
+
"maybe": 121,
|
| 247 |
+
"meal_suggestion": 46,
|
| 248 |
+
"meaning_of_life": 22,
|
| 249 |
+
"measurement_conversion": 31,
|
| 250 |
+
"meeting_schedule": 19,
|
| 251 |
+
"min_payment": 143,
|
| 252 |
+
"mpg": 142,
|
| 253 |
+
"new_card": 91,
|
| 254 |
+
"next_holiday": 140,
|
| 255 |
+
"next_song": 118,
|
| 256 |
+
"no": 30,
|
| 257 |
+
"nutrition_info": 1,
|
| 258 |
+
"oil_change_how": 3,
|
| 259 |
+
"oil_change_when": 123,
|
| 260 |
+
"oos": 42,
|
| 261 |
+
"order": 16,
|
| 262 |
+
"order_checks": 127,
|
| 263 |
+
"order_status": 84,
|
| 264 |
+
"pay_bill": 111,
|
| 265 |
+
"payday": 79,
|
| 266 |
+
"pin_change": 89,
|
| 267 |
+
"play_music": 106,
|
| 268 |
+
"plug_type": 120,
|
| 269 |
+
"pto_balance": 126,
|
| 270 |
+
"pto_request": 97,
|
| 271 |
+
"pto_request_status": 101,
|
| 272 |
+
"pto_used": 148,
|
| 273 |
+
"recipe": 104,
|
| 274 |
+
"redeem_rewards": 6,
|
| 275 |
+
"reminder": 76,
|
| 276 |
+
"reminder_update": 53,
|
| 277 |
+
"repeat": 13,
|
| 278 |
+
"replacement_card_duration": 90,
|
| 279 |
+
"report_fraud": 70,
|
| 280 |
+
"report_lost_card": 12,
|
| 281 |
+
"reset_settings": 88,
|
| 282 |
+
"restaurant_reservation": 23,
|
| 283 |
+
"restaurant_reviews": 0,
|
| 284 |
+
"restaurant_suggestion": 145,
|
| 285 |
+
"rewards_balance": 50,
|
| 286 |
+
"roll_dice": 92,
|
| 287 |
+
"rollover_401k": 100,
|
| 288 |
+
"routing": 45,
|
| 289 |
+
"schedule_maintenance": 131,
|
| 290 |
+
"schedule_meeting": 18,
|
| 291 |
+
"share_location": 58,
|
| 292 |
+
"shopping_list": 43,
|
| 293 |
+
"shopping_list_update": 125,
|
| 294 |
+
"smart_home": 10,
|
| 295 |
+
"spelling": 73,
|
| 296 |
+
"spending_history": 71,
|
| 297 |
+
"sync_device": 130,
|
| 298 |
+
"taxes": 94,
|
| 299 |
+
"tell_joke": 36,
|
| 300 |
+
"text": 26,
|
| 301 |
+
"thank_you": 124,
|
| 302 |
+
"time": 4,
|
| 303 |
+
"timer": 32,
|
| 304 |
+
"timezone": 66,
|
| 305 |
+
"tire_change": 47,
|
| 306 |
+
"tire_pressure": 98,
|
| 307 |
+
"todo_list": 48,
|
| 308 |
+
"todo_list_update": 65,
|
| 309 |
+
"traffic": 24,
|
| 310 |
+
"transactions": 68,
|
| 311 |
+
"transfer": 133,
|
| 312 |
+
"translate": 61,
|
| 313 |
+
"travel_alert": 128,
|
| 314 |
+
"travel_notification": 146,
|
| 315 |
+
"travel_suggestion": 149,
|
| 316 |
+
"uber": 39,
|
| 317 |
+
"update_playlist": 141,
|
| 318 |
+
"user_name": 11,
|
| 319 |
+
"vaccines": 52,
|
| 320 |
+
"w2": 122,
|
| 321 |
+
"weather": 5,
|
| 322 |
+
"what_are_your_hobbies": 15,
|
| 323 |
+
"what_can_i_ask_you": 115,
|
| 324 |
+
"what_is_your_name": 75,
|
| 325 |
+
"what_song": 21,
|
| 326 |
+
"where_are_you_from": 77,
|
| 327 |
+
"whisper_mode": 14,
|
| 328 |
+
"who_do_you_work_for": 57,
|
| 329 |
+
"who_made_you": 96,
|
| 330 |
+
"yes": 107
|
| 331 |
},
|
| 332 |
"layer_norm_eps": 1e-05,
|
| 333 |
"local_attention": 128,
|
run-0/checkpoint-5724/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 598898116
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81ac3f88bdb451e4316877bdf137e199de70b195697d9835e23b0052e2b2bace
|
| 3 |
size 598898116
|
run-0/checkpoint-5724/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1197886411
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:022d1a6aa461991f7bc9e38c5ac0505359b0feab028c560565db26ffc81ef9f4
|
| 3 |
size 1197886411
|
run-0/checkpoint-5724/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:507b0db25825069fb54cb54f098269305d212c34ae4cb16af06f08bb514410fc
|
| 3 |
size 1383
|
run-0/checkpoint-5724/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:842c477c03985d57c436f042185e94f8b68859521c9e0660a688efc82daf99b1
|
| 3 |
size 1465
|
run-0/checkpoint-5724/trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 3816,
|
| 3 |
-
"best_metric": 0.
|
| 4 |
"best_model_checkpoint": "ModernBERT-base-finetuned-distilled-clinc/run-0/checkpoint-3816",
|
| 5 |
"epoch": 6.0,
|
| 6 |
"eval_steps": 500,
|
|
@@ -10,142 +10,107 @@
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate": 1.
|
| 16 |
-
"loss":
|
| 17 |
-
"step":
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 1.0,
|
| 21 |
-
"eval_accuracy": 0.
|
| 22 |
-
"eval_loss": 1.
|
| 23 |
-
"eval_runtime":
|
| 24 |
-
"eval_samples_per_second":
|
| 25 |
-
"eval_steps_per_second":
|
| 26 |
"step": 954
|
| 27 |
},
|
| 28 |
{
|
| 29 |
-
"epoch": 1.
|
| 30 |
-
"grad_norm":
|
| 31 |
-
"learning_rate": 1.
|
| 32 |
-
"loss":
|
| 33 |
-
"step":
|
| 34 |
-
},
|
| 35 |
-
{
|
| 36 |
-
"epoch": 1.5723270440251573,
|
| 37 |
-
"grad_norm": 24.63665199279785,
|
| 38 |
-
"learning_rate": 1.6508269275564874e-05,
|
| 39 |
-
"loss": 1.3397,
|
| 40 |
-
"step": 1500
|
| 41 |
},
|
| 42 |
{
|
| 43 |
"epoch": 2.0,
|
| 44 |
-
"eval_accuracy": 0.
|
| 45 |
-
"eval_loss":
|
| 46 |
-
"eval_runtime":
|
| 47 |
-
"eval_samples_per_second":
|
| 48 |
-
"eval_steps_per_second":
|
| 49 |
"step": 1908
|
| 50 |
},
|
| 51 |
{
|
| 52 |
-
"epoch": 2.
|
| 53 |
-
"grad_norm": 5.
|
| 54 |
-
"learning_rate": 1.
|
| 55 |
-
"loss": 0.
|
| 56 |
-
"step":
|
| 57 |
-
},
|
| 58 |
-
{
|
| 59 |
-
"epoch": 2.620545073375262,
|
| 60 |
-
"grad_norm": 3.776374101638794,
|
| 61 |
-
"learning_rate": 1.4178895877009086e-05,
|
| 62 |
-
"loss": 0.7147,
|
| 63 |
-
"step": 2500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 3.0,
|
| 67 |
-
"eval_accuracy": 0.
|
| 68 |
-
"eval_loss": 0.
|
| 69 |
-
"eval_runtime":
|
| 70 |
-
"eval_samples_per_second":
|
| 71 |
-
"eval_steps_per_second":
|
| 72 |
"step": 2862
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"epoch": 3.
|
| 76 |
-
"grad_norm": 3.
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 0.
|
| 79 |
-
"step":
|
| 80 |
-
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 3.668763102725367,
|
| 83 |
-
"grad_norm": 4.3173909187316895,
|
| 84 |
-
"learning_rate": 1.1849522478453297e-05,
|
| 85 |
-
"loss": 0.5303,
|
| 86 |
-
"step": 3500
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 4.0,
|
| 90 |
-
"eval_accuracy": 0.
|
| 91 |
-
"eval_loss": 0.
|
| 92 |
-
"eval_runtime":
|
| 93 |
-
"eval_samples_per_second":
|
| 94 |
-
"eval_steps_per_second":
|
| 95 |
"step": 3816
|
| 96 |
},
|
| 97 |
{
|
| 98 |
-
"epoch": 4.
|
| 99 |
-
"grad_norm":
|
| 100 |
-
"learning_rate":
|
| 101 |
-
"loss": 0.
|
| 102 |
-
"step":
|
| 103 |
-
},
|
| 104 |
-
{
|
| 105 |
-
"epoch": 4.716981132075472,
|
| 106 |
-
"grad_norm": 3.2583296298980713,
|
| 107 |
-
"learning_rate": 9.520149079897509e-06,
|
| 108 |
-
"loss": 0.4425,
|
| 109 |
-
"step": 4500
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 5.0,
|
| 113 |
-
"eval_accuracy": 0.
|
| 114 |
-
"eval_loss": 0.
|
| 115 |
-
"eval_runtime":
|
| 116 |
-
"eval_samples_per_second":
|
| 117 |
-
"eval_steps_per_second":
|
| 118 |
"step": 4770
|
| 119 |
},
|
| 120 |
{
|
| 121 |
-
"epoch": 5.
|
| 122 |
-
"grad_norm":
|
| 123 |
-
"learning_rate":
|
| 124 |
-
"loss": 0.
|
| 125 |
-
"step":
|
| 126 |
-
},
|
| 127 |
-
{
|
| 128 |
-
"epoch": 5.765199161425577,
|
| 129 |
-
"grad_norm": 3.1272575855255127,
|
| 130 |
-
"learning_rate": 7.190775681341719e-06,
|
| 131 |
-
"loss": 0.382,
|
| 132 |
-
"step": 5500
|
| 133 |
},
|
| 134 |
{
|
| 135 |
"epoch": 6.0,
|
| 136 |
-
"eval_accuracy": 0.
|
| 137 |
-
"eval_loss": 0.
|
| 138 |
-
"eval_runtime":
|
| 139 |
-
"eval_samples_per_second":
|
| 140 |
-
"eval_steps_per_second":
|
| 141 |
"step": 5724
|
| 142 |
}
|
| 143 |
],
|
| 144 |
-
"logging_steps":
|
| 145 |
-
"max_steps":
|
| 146 |
"num_input_tokens_seen": 0,
|
| 147 |
-
"num_train_epochs":
|
| 148 |
-
"save_steps":
|
| 149 |
"stateful_callbacks": {
|
| 150 |
"EarlyStoppingCallback": {
|
| 151 |
"args": {
|
|
@@ -167,12 +132,12 @@
|
|
| 167 |
"attributes": {}
|
| 168 |
}
|
| 169 |
},
|
| 170 |
-
"total_flos":
|
| 171 |
"train_batch_size": 16,
|
| 172 |
"trial_name": null,
|
| 173 |
"trial_params": {
|
| 174 |
-
"alpha": 0.
|
| 175 |
-
"num_train_epochs":
|
| 176 |
-
"temperature":
|
| 177 |
}
|
| 178 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 3816,
|
| 3 |
+
"best_metric": 0.9638709677419355,
|
| 4 |
"best_model_checkpoint": "ModernBERT-base-finetuned-distilled-clinc/run-0/checkpoint-3816",
|
| 5 |
"epoch": 6.0,
|
| 6 |
"eval_steps": 500,
|
|
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.9989517819706499,
|
| 14 |
+
"grad_norm": 7.180769920349121,
|
| 15 |
+
"learning_rate": 1.7148846960167716e-05,
|
| 16 |
+
"loss": 4.5857,
|
| 17 |
+
"step": 953
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 1.0,
|
| 21 |
+
"eval_accuracy": 0.9364516129032258,
|
| 22 |
+
"eval_loss": 1.5901516675949097,
|
| 23 |
+
"eval_runtime": 13.3547,
|
| 24 |
+
"eval_samples_per_second": 232.128,
|
| 25 |
+
"eval_steps_per_second": 14.527,
|
| 26 |
"step": 954
|
| 27 |
},
|
| 28 |
{
|
| 29 |
+
"epoch": 1.9979035639412999,
|
| 30 |
+
"grad_norm": 9.229886054992676,
|
| 31 |
+
"learning_rate": 1.4294699011680145e-05,
|
| 32 |
+
"loss": 1.0911,
|
| 33 |
+
"step": 1906
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 2.0,
|
| 37 |
+
"eval_accuracy": 0.9570967741935484,
|
| 38 |
+
"eval_loss": 0.8912312388420105,
|
| 39 |
+
"eval_runtime": 13.3096,
|
| 40 |
+
"eval_samples_per_second": 232.915,
|
| 41 |
+
"eval_steps_per_second": 14.576,
|
| 42 |
"step": 1908
|
| 43 |
},
|
| 44 |
{
|
| 45 |
+
"epoch": 2.99685534591195,
|
| 46 |
+
"grad_norm": 5.046926975250244,
|
| 47 |
+
"learning_rate": 1.1440551063192572e-05,
|
| 48 |
+
"loss": 0.6315,
|
| 49 |
+
"step": 2859
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
},
|
| 51 |
{
|
| 52 |
"epoch": 3.0,
|
| 53 |
+
"eval_accuracy": 0.9609677419354838,
|
| 54 |
+
"eval_loss": 0.7318851351737976,
|
| 55 |
+
"eval_runtime": 13.4127,
|
| 56 |
+
"eval_samples_per_second": 231.124,
|
| 57 |
+
"eval_steps_per_second": 14.464,
|
| 58 |
"step": 2862
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 3.9958071278825997,
|
| 62 |
+
"grad_norm": 3.7578282356262207,
|
| 63 |
+
"learning_rate": 8.586403114705001e-06,
|
| 64 |
+
"loss": 0.481,
|
| 65 |
+
"step": 3812
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 4.0,
|
| 69 |
+
"eval_accuracy": 0.9638709677419355,
|
| 70 |
+
"eval_loss": 0.6467007994651794,
|
| 71 |
+
"eval_runtime": 13.39,
|
| 72 |
+
"eval_samples_per_second": 231.516,
|
| 73 |
+
"eval_steps_per_second": 14.488,
|
| 74 |
"step": 3816
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 4.99475890985325,
|
| 78 |
+
"grad_norm": 2.5420217514038086,
|
| 79 |
+
"learning_rate": 5.732255166217431e-06,
|
| 80 |
+
"loss": 0.4072,
|
| 81 |
+
"step": 4765
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 5.0,
|
| 85 |
+
"eval_accuracy": 0.9632258064516129,
|
| 86 |
+
"eval_loss": 0.607825517654419,
|
| 87 |
+
"eval_runtime": 13.2517,
|
| 88 |
+
"eval_samples_per_second": 233.933,
|
| 89 |
+
"eval_steps_per_second": 14.64,
|
| 90 |
"step": 4770
|
| 91 |
},
|
| 92 |
{
|
| 93 |
+
"epoch": 5.9937106918239,
|
| 94 |
+
"grad_norm": 6.932933330535889,
|
| 95 |
+
"learning_rate": 2.8781072177298598e-06,
|
| 96 |
+
"loss": 0.3637,
|
| 97 |
+
"step": 5718
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 98 |
},
|
| 99 |
{
|
| 100 |
"epoch": 6.0,
|
| 101 |
+
"eval_accuracy": 0.9632258064516129,
|
| 102 |
+
"eval_loss": 0.5857376456260681,
|
| 103 |
+
"eval_runtime": 13.3136,
|
| 104 |
+
"eval_samples_per_second": 232.845,
|
| 105 |
+
"eval_steps_per_second": 14.572,
|
| 106 |
"step": 5724
|
| 107 |
}
|
| 108 |
],
|
| 109 |
+
"logging_steps": 953,
|
| 110 |
+
"max_steps": 6678,
|
| 111 |
"num_input_tokens_seen": 0,
|
| 112 |
+
"num_train_epochs": 7,
|
| 113 |
+
"save_steps": 1000000000.0,
|
| 114 |
"stateful_callbacks": {
|
| 115 |
"EarlyStoppingCallback": {
|
| 116 |
"args": {
|
|
|
|
| 132 |
"attributes": {}
|
| 133 |
}
|
| 134 |
},
|
| 135 |
+
"total_flos": 1255667955529344.0,
|
| 136 |
"train_batch_size": 16,
|
| 137 |
"trial_name": null,
|
| 138 |
"trial_params": {
|
| 139 |
+
"alpha": 0.5630385610905387,
|
| 140 |
+
"num_train_epochs": 7,
|
| 141 |
+
"temperature": 9
|
| 142 |
}
|
| 143 |
}
|
run-0/checkpoint-5724/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7d2c126b5b10d3fa5a3cf6095b18e906cad6909d2dc20a3626092a9a968fbe44
|
| 3 |
+
size 5905
|