Training in progress, step 23532
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- model.safetensors +1 -1
- run-0/checkpoint-23532/config.json +331 -0
- run-0/checkpoint-23532/model.safetensors +3 -0
- run-0/checkpoint-23532/optimizer.pt +3 -0
- run-0/checkpoint-23532/rng_state.pth +3 -0
- run-0/checkpoint-23532/scheduler.pt +3 -0
- run-0/checkpoint-23532/special_tokens_map.json +7 -0
- run-0/checkpoint-23532/tokenizer.json +0 -0
- run-0/checkpoint-23532/tokenizer_config.json +55 -0
- run-0/checkpoint-23532/trainer_state.json +1212 -0
- run-0/checkpoint-23532/training_args.bin +3 -0
- run-0/checkpoint-23532/vocab.txt +0 -0
- run-1/checkpoint-24168/config.json +331 -0
- run-1/checkpoint-24168/model.safetensors +3 -0
- run-1/checkpoint-24168/optimizer.pt +3 -0
- run-1/checkpoint-24168/rng_state.pth +3 -0
- run-1/checkpoint-24168/scheduler.pt +3 -0
- run-1/checkpoint-24168/special_tokens_map.json +7 -0
- run-1/checkpoint-24168/tokenizer.json +0 -0
- run-1/checkpoint-24168/tokenizer_config.json +55 -0
- run-1/checkpoint-24168/trainer_state.json +1244 -0
- run-1/checkpoint-24168/training_args.bin +3 -0
- run-1/checkpoint-24168/vocab.txt +0 -0
- run-10/checkpoint-20670/config.json +331 -0
- run-10/checkpoint-20670/model.safetensors +3 -0
- run-10/checkpoint-20670/optimizer.pt +3 -0
- run-10/checkpoint-20670/rng_state.pth +3 -0
- run-10/checkpoint-20670/scheduler.pt +3 -0
- run-10/checkpoint-20670/special_tokens_map.json +7 -0
- run-10/checkpoint-20670/tokenizer.json +0 -0
- run-10/checkpoint-20670/tokenizer_config.json +55 -0
- run-10/checkpoint-20670/trainer_state.json +1068 -0
- run-10/checkpoint-20670/training_args.bin +3 -0
- run-10/checkpoint-20670/vocab.txt +0 -0
- run-12/checkpoint-20034/config.json +331 -0
- run-12/checkpoint-20034/model.safetensors +3 -0
- run-12/checkpoint-20034/optimizer.pt +3 -0
- run-12/checkpoint-20034/rng_state.pth +3 -0
- run-12/checkpoint-20034/scheduler.pt +3 -0
- run-12/checkpoint-20034/special_tokens_map.json +7 -0
- run-12/checkpoint-20034/tokenizer.json +0 -0
- run-12/checkpoint-20034/tokenizer_config.json +55 -0
- run-12/checkpoint-20034/trainer_state.json +1036 -0
- run-12/checkpoint-20034/training_args.bin +3 -0
- run-12/checkpoint-20034/vocab.txt +0 -0
- run-2/checkpoint-12402/config.json +331 -0
- run-2/checkpoint-12402/model.safetensors +3 -0
- run-2/checkpoint-12402/optimizer.pt +3 -0
- run-2/checkpoint-12402/rng_state.pth +3 -0
- run-2/checkpoint-12402/scheduler.pt +3 -0
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 268290900
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e22625dfcd91e832d3564d2a5452917d2b462351555ac58a11f0f3fb2cd840bc
|
| 3 |
size 268290900
|
run-0/checkpoint-23532/config.json
ADDED
|
@@ -0,0 +1,331 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilbert-base-uncased",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"hidden_dim": 3072,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "restaurant_reviews",
|
| 13 |
+
"1": "nutrition_info",
|
| 14 |
+
"2": "account_blocked",
|
| 15 |
+
"3": "oil_change_how",
|
| 16 |
+
"4": "time",
|
| 17 |
+
"5": "weather",
|
| 18 |
+
"6": "redeem_rewards",
|
| 19 |
+
"7": "interest_rate",
|
| 20 |
+
"8": "gas_type",
|
| 21 |
+
"9": "accept_reservations",
|
| 22 |
+
"10": "smart_home",
|
| 23 |
+
"11": "user_name",
|
| 24 |
+
"12": "report_lost_card",
|
| 25 |
+
"13": "repeat",
|
| 26 |
+
"14": "whisper_mode",
|
| 27 |
+
"15": "what_are_your_hobbies",
|
| 28 |
+
"16": "order",
|
| 29 |
+
"17": "jump_start",
|
| 30 |
+
"18": "schedule_meeting",
|
| 31 |
+
"19": "meeting_schedule",
|
| 32 |
+
"20": "freeze_account",
|
| 33 |
+
"21": "what_song",
|
| 34 |
+
"22": "meaning_of_life",
|
| 35 |
+
"23": "restaurant_reservation",
|
| 36 |
+
"24": "traffic",
|
| 37 |
+
"25": "make_call",
|
| 38 |
+
"26": "text",
|
| 39 |
+
"27": "bill_balance",
|
| 40 |
+
"28": "improve_credit_score",
|
| 41 |
+
"29": "change_language",
|
| 42 |
+
"30": "no",
|
| 43 |
+
"31": "measurement_conversion",
|
| 44 |
+
"32": "timer",
|
| 45 |
+
"33": "flip_coin",
|
| 46 |
+
"34": "do_you_have_pets",
|
| 47 |
+
"35": "balance",
|
| 48 |
+
"36": "tell_joke",
|
| 49 |
+
"37": "last_maintenance",
|
| 50 |
+
"38": "exchange_rate",
|
| 51 |
+
"39": "uber",
|
| 52 |
+
"40": "car_rental",
|
| 53 |
+
"41": "credit_limit",
|
| 54 |
+
"42": "oos",
|
| 55 |
+
"43": "shopping_list",
|
| 56 |
+
"44": "expiration_date",
|
| 57 |
+
"45": "routing",
|
| 58 |
+
"46": "meal_suggestion",
|
| 59 |
+
"47": "tire_change",
|
| 60 |
+
"48": "todo_list",
|
| 61 |
+
"49": "card_declined",
|
| 62 |
+
"50": "rewards_balance",
|
| 63 |
+
"51": "change_accent",
|
| 64 |
+
"52": "vaccines",
|
| 65 |
+
"53": "reminder_update",
|
| 66 |
+
"54": "food_last",
|
| 67 |
+
"55": "change_ai_name",
|
| 68 |
+
"56": "bill_due",
|
| 69 |
+
"57": "who_do_you_work_for",
|
| 70 |
+
"58": "share_location",
|
| 71 |
+
"59": "international_visa",
|
| 72 |
+
"60": "calendar",
|
| 73 |
+
"61": "translate",
|
| 74 |
+
"62": "carry_on",
|
| 75 |
+
"63": "book_flight",
|
| 76 |
+
"64": "insurance_change",
|
| 77 |
+
"65": "todo_list_update",
|
| 78 |
+
"66": "timezone",
|
| 79 |
+
"67": "cancel_reservation",
|
| 80 |
+
"68": "transactions",
|
| 81 |
+
"69": "credit_score",
|
| 82 |
+
"70": "report_fraud",
|
| 83 |
+
"71": "spending_history",
|
| 84 |
+
"72": "directions",
|
| 85 |
+
"73": "spelling",
|
| 86 |
+
"74": "insurance",
|
| 87 |
+
"75": "what_is_your_name",
|
| 88 |
+
"76": "reminder",
|
| 89 |
+
"77": "where_are_you_from",
|
| 90 |
+
"78": "distance",
|
| 91 |
+
"79": "payday",
|
| 92 |
+
"80": "flight_status",
|
| 93 |
+
"81": "find_phone",
|
| 94 |
+
"82": "greeting",
|
| 95 |
+
"83": "alarm",
|
| 96 |
+
"84": "order_status",
|
| 97 |
+
"85": "confirm_reservation",
|
| 98 |
+
"86": "cook_time",
|
| 99 |
+
"87": "damaged_card",
|
| 100 |
+
"88": "reset_settings",
|
| 101 |
+
"89": "pin_change",
|
| 102 |
+
"90": "replacement_card_duration",
|
| 103 |
+
"91": "new_card",
|
| 104 |
+
"92": "roll_dice",
|
| 105 |
+
"93": "income",
|
| 106 |
+
"94": "taxes",
|
| 107 |
+
"95": "date",
|
| 108 |
+
"96": "who_made_you",
|
| 109 |
+
"97": "pto_request",
|
| 110 |
+
"98": "tire_pressure",
|
| 111 |
+
"99": "how_old_are_you",
|
| 112 |
+
"100": "rollover_401k",
|
| 113 |
+
"101": "pto_request_status",
|
| 114 |
+
"102": "how_busy",
|
| 115 |
+
"103": "application_status",
|
| 116 |
+
"104": "recipe",
|
| 117 |
+
"105": "calendar_update",
|
| 118 |
+
"106": "play_music",
|
| 119 |
+
"107": "yes",
|
| 120 |
+
"108": "direct_deposit",
|
| 121 |
+
"109": "credit_limit_change",
|
| 122 |
+
"110": "gas",
|
| 123 |
+
"111": "pay_bill",
|
| 124 |
+
"112": "ingredients_list",
|
| 125 |
+
"113": "lost_luggage",
|
| 126 |
+
"114": "goodbye",
|
| 127 |
+
"115": "what_can_i_ask_you",
|
| 128 |
+
"116": "book_hotel",
|
| 129 |
+
"117": "are_you_a_bot",
|
| 130 |
+
"118": "next_song",
|
| 131 |
+
"119": "change_speed",
|
| 132 |
+
"120": "plug_type",
|
| 133 |
+
"121": "maybe",
|
| 134 |
+
"122": "w2",
|
| 135 |
+
"123": "oil_change_when",
|
| 136 |
+
"124": "thank_you",
|
| 137 |
+
"125": "shopping_list_update",
|
| 138 |
+
"126": "pto_balance",
|
| 139 |
+
"127": "order_checks",
|
| 140 |
+
"128": "travel_alert",
|
| 141 |
+
"129": "fun_fact",
|
| 142 |
+
"130": "sync_device",
|
| 143 |
+
"131": "schedule_maintenance",
|
| 144 |
+
"132": "apr",
|
| 145 |
+
"133": "transfer",
|
| 146 |
+
"134": "ingredient_substitution",
|
| 147 |
+
"135": "calories",
|
| 148 |
+
"136": "current_location",
|
| 149 |
+
"137": "international_fees",
|
| 150 |
+
"138": "calculator",
|
| 151 |
+
"139": "definition",
|
| 152 |
+
"140": "next_holiday",
|
| 153 |
+
"141": "update_playlist",
|
| 154 |
+
"142": "mpg",
|
| 155 |
+
"143": "min_payment",
|
| 156 |
+
"144": "change_user_name",
|
| 157 |
+
"145": "restaurant_suggestion",
|
| 158 |
+
"146": "travel_notification",
|
| 159 |
+
"147": "cancel",
|
| 160 |
+
"148": "pto_used",
|
| 161 |
+
"149": "travel_suggestion",
|
| 162 |
+
"150": "change_volume"
|
| 163 |
+
},
|
| 164 |
+
"initializer_range": 0.02,
|
| 165 |
+
"label2id": {
|
| 166 |
+
"accept_reservations": 9,
|
| 167 |
+
"account_blocked": 2,
|
| 168 |
+
"alarm": 83,
|
| 169 |
+
"application_status": 103,
|
| 170 |
+
"apr": 132,
|
| 171 |
+
"are_you_a_bot": 117,
|
| 172 |
+
"balance": 35,
|
| 173 |
+
"bill_balance": 27,
|
| 174 |
+
"bill_due": 56,
|
| 175 |
+
"book_flight": 63,
|
| 176 |
+
"book_hotel": 116,
|
| 177 |
+
"calculator": 138,
|
| 178 |
+
"calendar": 60,
|
| 179 |
+
"calendar_update": 105,
|
| 180 |
+
"calories": 135,
|
| 181 |
+
"cancel": 147,
|
| 182 |
+
"cancel_reservation": 67,
|
| 183 |
+
"car_rental": 40,
|
| 184 |
+
"card_declined": 49,
|
| 185 |
+
"carry_on": 62,
|
| 186 |
+
"change_accent": 51,
|
| 187 |
+
"change_ai_name": 55,
|
| 188 |
+
"change_language": 29,
|
| 189 |
+
"change_speed": 119,
|
| 190 |
+
"change_user_name": 144,
|
| 191 |
+
"change_volume": 150,
|
| 192 |
+
"confirm_reservation": 85,
|
| 193 |
+
"cook_time": 86,
|
| 194 |
+
"credit_limit": 41,
|
| 195 |
+
"credit_limit_change": 109,
|
| 196 |
+
"credit_score": 69,
|
| 197 |
+
"current_location": 136,
|
| 198 |
+
"damaged_card": 87,
|
| 199 |
+
"date": 95,
|
| 200 |
+
"definition": 139,
|
| 201 |
+
"direct_deposit": 108,
|
| 202 |
+
"directions": 72,
|
| 203 |
+
"distance": 78,
|
| 204 |
+
"do_you_have_pets": 34,
|
| 205 |
+
"exchange_rate": 38,
|
| 206 |
+
"expiration_date": 44,
|
| 207 |
+
"find_phone": 81,
|
| 208 |
+
"flight_status": 80,
|
| 209 |
+
"flip_coin": 33,
|
| 210 |
+
"food_last": 54,
|
| 211 |
+
"freeze_account": 20,
|
| 212 |
+
"fun_fact": 129,
|
| 213 |
+
"gas": 110,
|
| 214 |
+
"gas_type": 8,
|
| 215 |
+
"goodbye": 114,
|
| 216 |
+
"greeting": 82,
|
| 217 |
+
"how_busy": 102,
|
| 218 |
+
"how_old_are_you": 99,
|
| 219 |
+
"improve_credit_score": 28,
|
| 220 |
+
"income": 93,
|
| 221 |
+
"ingredient_substitution": 134,
|
| 222 |
+
"ingredients_list": 112,
|
| 223 |
+
"insurance": 74,
|
| 224 |
+
"insurance_change": 64,
|
| 225 |
+
"interest_rate": 7,
|
| 226 |
+
"international_fees": 137,
|
| 227 |
+
"international_visa": 59,
|
| 228 |
+
"jump_start": 17,
|
| 229 |
+
"last_maintenance": 37,
|
| 230 |
+
"lost_luggage": 113,
|
| 231 |
+
"make_call": 25,
|
| 232 |
+
"maybe": 121,
|
| 233 |
+
"meal_suggestion": 46,
|
| 234 |
+
"meaning_of_life": 22,
|
| 235 |
+
"measurement_conversion": 31,
|
| 236 |
+
"meeting_schedule": 19,
|
| 237 |
+
"min_payment": 143,
|
| 238 |
+
"mpg": 142,
|
| 239 |
+
"new_card": 91,
|
| 240 |
+
"next_holiday": 140,
|
| 241 |
+
"next_song": 118,
|
| 242 |
+
"no": 30,
|
| 243 |
+
"nutrition_info": 1,
|
| 244 |
+
"oil_change_how": 3,
|
| 245 |
+
"oil_change_when": 123,
|
| 246 |
+
"oos": 42,
|
| 247 |
+
"order": 16,
|
| 248 |
+
"order_checks": 127,
|
| 249 |
+
"order_status": 84,
|
| 250 |
+
"pay_bill": 111,
|
| 251 |
+
"payday": 79,
|
| 252 |
+
"pin_change": 89,
|
| 253 |
+
"play_music": 106,
|
| 254 |
+
"plug_type": 120,
|
| 255 |
+
"pto_balance": 126,
|
| 256 |
+
"pto_request": 97,
|
| 257 |
+
"pto_request_status": 101,
|
| 258 |
+
"pto_used": 148,
|
| 259 |
+
"recipe": 104,
|
| 260 |
+
"redeem_rewards": 6,
|
| 261 |
+
"reminder": 76,
|
| 262 |
+
"reminder_update": 53,
|
| 263 |
+
"repeat": 13,
|
| 264 |
+
"replacement_card_duration": 90,
|
| 265 |
+
"report_fraud": 70,
|
| 266 |
+
"report_lost_card": 12,
|
| 267 |
+
"reset_settings": 88,
|
| 268 |
+
"restaurant_reservation": 23,
|
| 269 |
+
"restaurant_reviews": 0,
|
| 270 |
+
"restaurant_suggestion": 145,
|
| 271 |
+
"rewards_balance": 50,
|
| 272 |
+
"roll_dice": 92,
|
| 273 |
+
"rollover_401k": 100,
|
| 274 |
+
"routing": 45,
|
| 275 |
+
"schedule_maintenance": 131,
|
| 276 |
+
"schedule_meeting": 18,
|
| 277 |
+
"share_location": 58,
|
| 278 |
+
"shopping_list": 43,
|
| 279 |
+
"shopping_list_update": 125,
|
| 280 |
+
"smart_home": 10,
|
| 281 |
+
"spelling": 73,
|
| 282 |
+
"spending_history": 71,
|
| 283 |
+
"sync_device": 130,
|
| 284 |
+
"taxes": 94,
|
| 285 |
+
"tell_joke": 36,
|
| 286 |
+
"text": 26,
|
| 287 |
+
"thank_you": 124,
|
| 288 |
+
"time": 4,
|
| 289 |
+
"timer": 32,
|
| 290 |
+
"timezone": 66,
|
| 291 |
+
"tire_change": 47,
|
| 292 |
+
"tire_pressure": 98,
|
| 293 |
+
"todo_list": 48,
|
| 294 |
+
"todo_list_update": 65,
|
| 295 |
+
"traffic": 24,
|
| 296 |
+
"transactions": 68,
|
| 297 |
+
"transfer": 133,
|
| 298 |
+
"translate": 61,
|
| 299 |
+
"travel_alert": 128,
|
| 300 |
+
"travel_notification": 146,
|
| 301 |
+
"travel_suggestion": 149,
|
| 302 |
+
"uber": 39,
|
| 303 |
+
"update_playlist": 141,
|
| 304 |
+
"user_name": 11,
|
| 305 |
+
"vaccines": 52,
|
| 306 |
+
"w2": 122,
|
| 307 |
+
"weather": 5,
|
| 308 |
+
"what_are_your_hobbies": 15,
|
| 309 |
+
"what_can_i_ask_you": 115,
|
| 310 |
+
"what_is_your_name": 75,
|
| 311 |
+
"what_song": 21,
|
| 312 |
+
"where_are_you_from": 77,
|
| 313 |
+
"whisper_mode": 14,
|
| 314 |
+
"who_do_you_work_for": 57,
|
| 315 |
+
"who_made_you": 96,
|
| 316 |
+
"yes": 107
|
| 317 |
+
},
|
| 318 |
+
"max_position_embeddings": 512,
|
| 319 |
+
"model_type": "distilbert",
|
| 320 |
+
"n_heads": 12,
|
| 321 |
+
"n_layers": 6,
|
| 322 |
+
"pad_token_id": 0,
|
| 323 |
+
"problem_type": "single_label_classification",
|
| 324 |
+
"qa_dropout": 0.1,
|
| 325 |
+
"seq_classif_dropout": 0.2,
|
| 326 |
+
"sinusoidal_pos_embds": false,
|
| 327 |
+
"tie_weights_": true,
|
| 328 |
+
"torch_dtype": "float32",
|
| 329 |
+
"transformers_version": "4.44.2",
|
| 330 |
+
"vocab_size": 30522
|
| 331 |
+
}
|
run-0/checkpoint-23532/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e22625dfcd91e832d3564d2a5452917d2b462351555ac58a11f0f3fb2cd840bc
|
| 3 |
+
size 268290900
|
run-0/checkpoint-23532/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0d3df8e206fc5d9ef3ed87e7474b38afc1702619f72f04e684aa35a57e12c29
|
| 3 |
+
size 536643898
|
run-0/checkpoint-23532/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b9c11fb4db653679cb4e1738c6da868a25b312984cb03fb9937dec61b2dd07f
|
| 3 |
+
size 14244
|
run-0/checkpoint-23532/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3abce43f3f6073de6ea15ea05747334d5bc2bdc0e94e2618f6a9c02e828fd3e1
|
| 3 |
+
size 1064
|
run-0/checkpoint-23532/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
run-0/checkpoint-23532/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
run-0/checkpoint-23532/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": true,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"mask_token": "[MASK]",
|
| 48 |
+
"model_max_length": 512,
|
| 49 |
+
"pad_token": "[PAD]",
|
| 50 |
+
"sep_token": "[SEP]",
|
| 51 |
+
"strip_accents": null,
|
| 52 |
+
"tokenize_chinese_chars": true,
|
| 53 |
+
"tokenizer_class": "DistilBertTokenizer",
|
| 54 |
+
"unk_token": "[UNK]"
|
| 55 |
+
}
|
run-0/checkpoint-23532/trainer_state.json
ADDED
|
@@ -0,0 +1,1212 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 74.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 23532,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.9968553459119497,
|
| 13 |
+
"grad_norm": 2.3475592136383057,
|
| 14 |
+
"learning_rate": 1.9730579636240016e-05,
|
| 15 |
+
"loss": 2.8866,
|
| 16 |
+
"step": 317
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_accuracy": 0.7316129032258064,
|
| 21 |
+
"eval_loss": 2.071580648422241,
|
| 22 |
+
"eval_runtime": 1.572,
|
| 23 |
+
"eval_samples_per_second": 1972.01,
|
| 24 |
+
"eval_steps_per_second": 41.349,
|
| 25 |
+
"step": 318
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.9937106918238994,
|
| 29 |
+
"grad_norm": 2.253408908843994,
|
| 30 |
+
"learning_rate": 1.9461159272480027e-05,
|
| 31 |
+
"loss": 1.5472,
|
| 32 |
+
"step": 634
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.0,
|
| 36 |
+
"eval_accuracy": 0.8774193548387097,
|
| 37 |
+
"eval_loss": 0.9601801633834839,
|
| 38 |
+
"eval_runtime": 1.5767,
|
| 39 |
+
"eval_samples_per_second": 1966.179,
|
| 40 |
+
"eval_steps_per_second": 41.226,
|
| 41 |
+
"step": 636
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 2.990566037735849,
|
| 45 |
+
"grad_norm": 2.595813274383545,
|
| 46 |
+
"learning_rate": 1.9191738908720045e-05,
|
| 47 |
+
"loss": 0.7257,
|
| 48 |
+
"step": 951
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 3.0,
|
| 52 |
+
"eval_accuracy": 0.9174193548387096,
|
| 53 |
+
"eval_loss": 0.514940083026886,
|
| 54 |
+
"eval_runtime": 1.5858,
|
| 55 |
+
"eval_samples_per_second": 1954.843,
|
| 56 |
+
"eval_steps_per_second": 40.989,
|
| 57 |
+
"step": 954
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 3.9874213836477987,
|
| 61 |
+
"grad_norm": 1.9288506507873535,
|
| 62 |
+
"learning_rate": 1.8922318544960056e-05,
|
| 63 |
+
"loss": 0.3992,
|
| 64 |
+
"step": 1268
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 4.0,
|
| 68 |
+
"eval_accuracy": 0.937741935483871,
|
| 69 |
+
"eval_loss": 0.3805593252182007,
|
| 70 |
+
"eval_runtime": 1.5729,
|
| 71 |
+
"eval_samples_per_second": 1970.858,
|
| 72 |
+
"eval_steps_per_second": 41.324,
|
| 73 |
+
"step": 1272
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.984276729559748,
|
| 77 |
+
"grad_norm": 0.8146032094955444,
|
| 78 |
+
"learning_rate": 1.865289818120007e-05,
|
| 79 |
+
"loss": 0.2851,
|
| 80 |
+
"step": 1585
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 5.0,
|
| 84 |
+
"eval_accuracy": 0.9429032258064516,
|
| 85 |
+
"eval_loss": 0.3391108512878418,
|
| 86 |
+
"eval_runtime": 1.572,
|
| 87 |
+
"eval_samples_per_second": 1972.021,
|
| 88 |
+
"eval_steps_per_second": 41.349,
|
| 89 |
+
"step": 1590
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 5.981132075471698,
|
| 93 |
+
"grad_norm": 0.6014134883880615,
|
| 94 |
+
"learning_rate": 1.8383477817440084e-05,
|
| 95 |
+
"loss": 0.2386,
|
| 96 |
+
"step": 1902
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 6.0,
|
| 100 |
+
"eval_accuracy": 0.9441935483870968,
|
| 101 |
+
"eval_loss": 0.32530343532562256,
|
| 102 |
+
"eval_runtime": 1.6034,
|
| 103 |
+
"eval_samples_per_second": 1933.35,
|
| 104 |
+
"eval_steps_per_second": 40.538,
|
| 105 |
+
"step": 1908
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 6.977987421383648,
|
| 109 |
+
"grad_norm": 1.808417797088623,
|
| 110 |
+
"learning_rate": 1.8114057453680095e-05,
|
| 111 |
+
"loss": 0.2158,
|
| 112 |
+
"step": 2219
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 7.0,
|
| 116 |
+
"eval_accuracy": 0.9454838709677419,
|
| 117 |
+
"eval_loss": 0.31467169523239136,
|
| 118 |
+
"eval_runtime": 1.5893,
|
| 119 |
+
"eval_samples_per_second": 1950.547,
|
| 120 |
+
"eval_steps_per_second": 40.899,
|
| 121 |
+
"step": 2226
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 7.9748427672955975,
|
| 125 |
+
"grad_norm": 0.3616284430027008,
|
| 126 |
+
"learning_rate": 1.784463708992011e-05,
|
| 127 |
+
"loss": 0.2045,
|
| 128 |
+
"step": 2536
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"eval_accuracy": 0.947741935483871,
|
| 133 |
+
"eval_loss": 0.30959999561309814,
|
| 134 |
+
"eval_runtime": 1.5809,
|
| 135 |
+
"eval_samples_per_second": 1960.963,
|
| 136 |
+
"eval_steps_per_second": 41.117,
|
| 137 |
+
"step": 2544
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 8.971698113207546,
|
| 141 |
+
"grad_norm": 0.4411262273788452,
|
| 142 |
+
"learning_rate": 1.7575216726160124e-05,
|
| 143 |
+
"loss": 0.1969,
|
| 144 |
+
"step": 2853
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 9.0,
|
| 148 |
+
"eval_accuracy": 0.9467741935483871,
|
| 149 |
+
"eval_loss": 0.30444082617759705,
|
| 150 |
+
"eval_runtime": 1.5824,
|
| 151 |
+
"eval_samples_per_second": 1959.054,
|
| 152 |
+
"eval_steps_per_second": 41.077,
|
| 153 |
+
"step": 2862
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 9.968553459119496,
|
| 157 |
+
"grad_norm": 0.32283931970596313,
|
| 158 |
+
"learning_rate": 1.7305796362400138e-05,
|
| 159 |
+
"loss": 0.1919,
|
| 160 |
+
"step": 3170
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 10.0,
|
| 164 |
+
"eval_accuracy": 0.9458064516129032,
|
| 165 |
+
"eval_loss": 0.30055466294288635,
|
| 166 |
+
"eval_runtime": 1.592,
|
| 167 |
+
"eval_samples_per_second": 1947.271,
|
| 168 |
+
"eval_steps_per_second": 40.83,
|
| 169 |
+
"step": 3180
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.965408805031446,
|
| 173 |
+
"grad_norm": 0.34415411949157715,
|
| 174 |
+
"learning_rate": 1.7036375998640152e-05,
|
| 175 |
+
"loss": 0.1887,
|
| 176 |
+
"step": 3487
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 11.0,
|
| 180 |
+
"eval_accuracy": 0.9461290322580646,
|
| 181 |
+
"eval_loss": 0.30284497141838074,
|
| 182 |
+
"eval_runtime": 1.5873,
|
| 183 |
+
"eval_samples_per_second": 1953.002,
|
| 184 |
+
"eval_steps_per_second": 40.95,
|
| 185 |
+
"step": 3498
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 11.962264150943396,
|
| 189 |
+
"grad_norm": 0.23879113793373108,
|
| 190 |
+
"learning_rate": 1.6766955634880163e-05,
|
| 191 |
+
"loss": 0.1864,
|
| 192 |
+
"step": 3804
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 12.0,
|
| 196 |
+
"eval_accuracy": 0.9493548387096774,
|
| 197 |
+
"eval_loss": 0.2960328757762909,
|
| 198 |
+
"eval_runtime": 1.5799,
|
| 199 |
+
"eval_samples_per_second": 1962.166,
|
| 200 |
+
"eval_steps_per_second": 41.142,
|
| 201 |
+
"step": 3816
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 12.959119496855346,
|
| 205 |
+
"grad_norm": 0.2762240171432495,
|
| 206 |
+
"learning_rate": 1.6497535271120178e-05,
|
| 207 |
+
"loss": 0.1836,
|
| 208 |
+
"step": 4121
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 13.0,
|
| 212 |
+
"eval_accuracy": 0.9454838709677419,
|
| 213 |
+
"eval_loss": 0.2947392165660858,
|
| 214 |
+
"eval_runtime": 1.5882,
|
| 215 |
+
"eval_samples_per_second": 1951.914,
|
| 216 |
+
"eval_steps_per_second": 40.927,
|
| 217 |
+
"step": 4134
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 13.955974842767295,
|
| 221 |
+
"grad_norm": 0.2626229524612427,
|
| 222 |
+
"learning_rate": 1.6228114907360192e-05,
|
| 223 |
+
"loss": 0.1818,
|
| 224 |
+
"step": 4438
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 14.0,
|
| 228 |
+
"eval_accuracy": 0.9483870967741935,
|
| 229 |
+
"eval_loss": 0.2958820164203644,
|
| 230 |
+
"eval_runtime": 1.5898,
|
| 231 |
+
"eval_samples_per_second": 1949.935,
|
| 232 |
+
"eval_steps_per_second": 40.886,
|
| 233 |
+
"step": 4452
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 14.952830188679245,
|
| 237 |
+
"grad_norm": 0.386436402797699,
|
| 238 |
+
"learning_rate": 1.5958694543600206e-05,
|
| 239 |
+
"loss": 0.1805,
|
| 240 |
+
"step": 4755
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 15.0,
|
| 244 |
+
"eval_accuracy": 0.9506451612903226,
|
| 245 |
+
"eval_loss": 0.29009997844696045,
|
| 246 |
+
"eval_runtime": 1.5859,
|
| 247 |
+
"eval_samples_per_second": 1954.683,
|
| 248 |
+
"eval_steps_per_second": 40.985,
|
| 249 |
+
"step": 4770
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 15.949685534591195,
|
| 253 |
+
"grad_norm": 0.22230026125907898,
|
| 254 |
+
"learning_rate": 1.568927417984022e-05,
|
| 255 |
+
"loss": 0.1798,
|
| 256 |
+
"step": 5072
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 16.0,
|
| 260 |
+
"eval_accuracy": 0.9461290322580646,
|
| 261 |
+
"eval_loss": 0.29720747470855713,
|
| 262 |
+
"eval_runtime": 1.5901,
|
| 263 |
+
"eval_samples_per_second": 1949.568,
|
| 264 |
+
"eval_steps_per_second": 40.878,
|
| 265 |
+
"step": 5088
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 16.946540880503143,
|
| 269 |
+
"grad_norm": 0.25611138343811035,
|
| 270 |
+
"learning_rate": 1.541985381608023e-05,
|
| 271 |
+
"loss": 0.1782,
|
| 272 |
+
"step": 5389
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 17.0,
|
| 276 |
+
"eval_accuracy": 0.9461290322580646,
|
| 277 |
+
"eval_loss": 0.29609400033950806,
|
| 278 |
+
"eval_runtime": 1.5952,
|
| 279 |
+
"eval_samples_per_second": 1943.373,
|
| 280 |
+
"eval_steps_per_second": 40.748,
|
| 281 |
+
"step": 5406
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 17.943396226415093,
|
| 285 |
+
"grad_norm": 0.21049971878528595,
|
| 286 |
+
"learning_rate": 1.5150433452320246e-05,
|
| 287 |
+
"loss": 0.1768,
|
| 288 |
+
"step": 5706
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 18.0,
|
| 292 |
+
"eval_accuracy": 0.95,
|
| 293 |
+
"eval_loss": 0.29309019446372986,
|
| 294 |
+
"eval_runtime": 1.5867,
|
| 295 |
+
"eval_samples_per_second": 1953.762,
|
| 296 |
+
"eval_steps_per_second": 40.966,
|
| 297 |
+
"step": 5724
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 18.940251572327043,
|
| 301 |
+
"grad_norm": 0.2441435158252716,
|
| 302 |
+
"learning_rate": 1.4881013088560258e-05,
|
| 303 |
+
"loss": 0.1761,
|
| 304 |
+
"step": 6023
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 19.0,
|
| 308 |
+
"eval_accuracy": 0.9474193548387096,
|
| 309 |
+
"eval_loss": 0.2915351390838623,
|
| 310 |
+
"eval_runtime": 1.5853,
|
| 311 |
+
"eval_samples_per_second": 1955.431,
|
| 312 |
+
"eval_steps_per_second": 41.001,
|
| 313 |
+
"step": 6042
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 19.937106918238992,
|
| 317 |
+
"grad_norm": 0.5714885592460632,
|
| 318 |
+
"learning_rate": 1.4611592724800274e-05,
|
| 319 |
+
"loss": 0.1749,
|
| 320 |
+
"step": 6340
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 20.0,
|
| 324 |
+
"eval_accuracy": 0.9480645161290323,
|
| 325 |
+
"eval_loss": 0.2894654870033264,
|
| 326 |
+
"eval_runtime": 1.5856,
|
| 327 |
+
"eval_samples_per_second": 1955.05,
|
| 328 |
+
"eval_steps_per_second": 40.993,
|
| 329 |
+
"step": 6360
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 20.933962264150942,
|
| 333 |
+
"grad_norm": 0.23249487578868866,
|
| 334 |
+
"learning_rate": 1.4342172361040287e-05,
|
| 335 |
+
"loss": 0.1743,
|
| 336 |
+
"step": 6657
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 21.0,
|
| 340 |
+
"eval_accuracy": 0.9483870967741935,
|
| 341 |
+
"eval_loss": 0.28891319036483765,
|
| 342 |
+
"eval_runtime": 1.5819,
|
| 343 |
+
"eval_samples_per_second": 1959.694,
|
| 344 |
+
"eval_steps_per_second": 41.09,
|
| 345 |
+
"step": 6678
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 21.930817610062892,
|
| 349 |
+
"grad_norm": 0.2100275754928589,
|
| 350 |
+
"learning_rate": 1.4072751997280301e-05,
|
| 351 |
+
"loss": 0.1735,
|
| 352 |
+
"step": 6974
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 22.0,
|
| 356 |
+
"eval_accuracy": 0.9487096774193549,
|
| 357 |
+
"eval_loss": 0.28882089257240295,
|
| 358 |
+
"eval_runtime": 1.5826,
|
| 359 |
+
"eval_samples_per_second": 1958.766,
|
| 360 |
+
"eval_steps_per_second": 41.071,
|
| 361 |
+
"step": 6996
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 22.927672955974842,
|
| 365 |
+
"grad_norm": 0.18105138838291168,
|
| 366 |
+
"learning_rate": 1.3803331633520314e-05,
|
| 367 |
+
"loss": 0.173,
|
| 368 |
+
"step": 7291
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 23.0,
|
| 372 |
+
"eval_accuracy": 0.9503225806451613,
|
| 373 |
+
"eval_loss": 0.28733834624290466,
|
| 374 |
+
"eval_runtime": 1.5906,
|
| 375 |
+
"eval_samples_per_second": 1948.93,
|
| 376 |
+
"eval_steps_per_second": 40.865,
|
| 377 |
+
"step": 7314
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 23.92452830188679,
|
| 381 |
+
"grad_norm": 0.17866890132427216,
|
| 382 |
+
"learning_rate": 1.3533911269760327e-05,
|
| 383 |
+
"loss": 0.1727,
|
| 384 |
+
"step": 7608
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 24.0,
|
| 388 |
+
"eval_accuracy": 0.9509677419354838,
|
| 389 |
+
"eval_loss": 0.28442323207855225,
|
| 390 |
+
"eval_runtime": 1.5887,
|
| 391 |
+
"eval_samples_per_second": 1951.295,
|
| 392 |
+
"eval_steps_per_second": 40.914,
|
| 393 |
+
"step": 7632
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 24.92138364779874,
|
| 397 |
+
"grad_norm": 0.1832052618265152,
|
| 398 |
+
"learning_rate": 1.3264490906000341e-05,
|
| 399 |
+
"loss": 0.1723,
|
| 400 |
+
"step": 7925
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 25.0,
|
| 404 |
+
"eval_accuracy": 0.947741935483871,
|
| 405 |
+
"eval_loss": 0.28769010305404663,
|
| 406 |
+
"eval_runtime": 1.5884,
|
| 407 |
+
"eval_samples_per_second": 1951.652,
|
| 408 |
+
"eval_steps_per_second": 40.922,
|
| 409 |
+
"step": 7950
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 25.91823899371069,
|
| 413 |
+
"grad_norm": 0.14876896142959595,
|
| 414 |
+
"learning_rate": 1.2995070542240355e-05,
|
| 415 |
+
"loss": 0.1717,
|
| 416 |
+
"step": 8242
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 26.0,
|
| 420 |
+
"eval_accuracy": 0.9503225806451613,
|
| 421 |
+
"eval_loss": 0.2851002514362335,
|
| 422 |
+
"eval_runtime": 1.5925,
|
| 423 |
+
"eval_samples_per_second": 1946.621,
|
| 424 |
+
"eval_steps_per_second": 40.816,
|
| 425 |
+
"step": 8268
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 26.91509433962264,
|
| 429 |
+
"grad_norm": 0.18779420852661133,
|
| 430 |
+
"learning_rate": 1.2725650178480368e-05,
|
| 431 |
+
"loss": 0.1712,
|
| 432 |
+
"step": 8559
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 27.0,
|
| 436 |
+
"eval_accuracy": 0.9503225806451613,
|
| 437 |
+
"eval_loss": 0.28837448358535767,
|
| 438 |
+
"eval_runtime": 1.5825,
|
| 439 |
+
"eval_samples_per_second": 1958.963,
|
| 440 |
+
"eval_steps_per_second": 41.075,
|
| 441 |
+
"step": 8586
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 27.91194968553459,
|
| 445 |
+
"grad_norm": 0.17873169481754303,
|
| 446 |
+
"learning_rate": 1.2456229814720382e-05,
|
| 447 |
+
"loss": 0.1707,
|
| 448 |
+
"step": 8876
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 28.0,
|
| 452 |
+
"eval_accuracy": 0.9519354838709677,
|
| 453 |
+
"eval_loss": 0.28391382098197937,
|
| 454 |
+
"eval_runtime": 1.585,
|
| 455 |
+
"eval_samples_per_second": 1955.836,
|
| 456 |
+
"eval_steps_per_second": 41.009,
|
| 457 |
+
"step": 8904
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 28.90880503144654,
|
| 461 |
+
"grad_norm": 0.15820345282554626,
|
| 462 |
+
"learning_rate": 1.2186809450960395e-05,
|
| 463 |
+
"loss": 0.1704,
|
| 464 |
+
"step": 9193
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 29.0,
|
| 468 |
+
"eval_accuracy": 0.9490322580645161,
|
| 469 |
+
"eval_loss": 0.28534936904907227,
|
| 470 |
+
"eval_runtime": 1.5858,
|
| 471 |
+
"eval_samples_per_second": 1954.889,
|
| 472 |
+
"eval_steps_per_second": 40.99,
|
| 473 |
+
"step": 9222
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 29.90566037735849,
|
| 477 |
+
"grad_norm": 0.20750579237937927,
|
| 478 |
+
"learning_rate": 1.1917389087200408e-05,
|
| 479 |
+
"loss": 0.17,
|
| 480 |
+
"step": 9510
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 30.0,
|
| 484 |
+
"eval_accuracy": 0.9516129032258065,
|
| 485 |
+
"eval_loss": 0.2824024558067322,
|
| 486 |
+
"eval_runtime": 1.5833,
|
| 487 |
+
"eval_samples_per_second": 1957.892,
|
| 488 |
+
"eval_steps_per_second": 41.053,
|
| 489 |
+
"step": 9540
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 30.90251572327044,
|
| 493 |
+
"grad_norm": 0.2273702174425125,
|
| 494 |
+
"learning_rate": 1.1647968723440422e-05,
|
| 495 |
+
"loss": 0.1698,
|
| 496 |
+
"step": 9827
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 31.0,
|
| 500 |
+
"eval_accuracy": 0.9474193548387096,
|
| 501 |
+
"eval_loss": 0.28540265560150146,
|
| 502 |
+
"eval_runtime": 1.5831,
|
| 503 |
+
"eval_samples_per_second": 1958.199,
|
| 504 |
+
"eval_steps_per_second": 41.059,
|
| 505 |
+
"step": 9858
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 31.89937106918239,
|
| 509 |
+
"grad_norm": 0.15220269560813904,
|
| 510 |
+
"learning_rate": 1.1378548359680436e-05,
|
| 511 |
+
"loss": 0.1696,
|
| 512 |
+
"step": 10144
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 32.0,
|
| 516 |
+
"eval_accuracy": 0.9487096774193549,
|
| 517 |
+
"eval_loss": 0.2837918996810913,
|
| 518 |
+
"eval_runtime": 1.5913,
|
| 519 |
+
"eval_samples_per_second": 1948.087,
|
| 520 |
+
"eval_steps_per_second": 40.847,
|
| 521 |
+
"step": 10176
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 32.89622641509434,
|
| 525 |
+
"grad_norm": 0.12839631736278534,
|
| 526 |
+
"learning_rate": 1.110912799592045e-05,
|
| 527 |
+
"loss": 0.1692,
|
| 528 |
+
"step": 10461
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 33.0,
|
| 532 |
+
"eval_accuracy": 0.9506451612903226,
|
| 533 |
+
"eval_loss": 0.28212934732437134,
|
| 534 |
+
"eval_runtime": 1.5852,
|
| 535 |
+
"eval_samples_per_second": 1955.649,
|
| 536 |
+
"eval_steps_per_second": 41.006,
|
| 537 |
+
"step": 10494
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 33.893081761006286,
|
| 541 |
+
"grad_norm": 0.14741210639476776,
|
| 542 |
+
"learning_rate": 1.0839707632160463e-05,
|
| 543 |
+
"loss": 0.1691,
|
| 544 |
+
"step": 10778
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 34.0,
|
| 548 |
+
"eval_accuracy": 0.9516129032258065,
|
| 549 |
+
"eval_loss": 0.28252750635147095,
|
| 550 |
+
"eval_runtime": 1.5948,
|
| 551 |
+
"eval_samples_per_second": 1943.876,
|
| 552 |
+
"eval_steps_per_second": 40.759,
|
| 553 |
+
"step": 10812
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 34.88993710691824,
|
| 557 |
+
"grad_norm": 0.1750616878271103,
|
| 558 |
+
"learning_rate": 1.0570287268400476e-05,
|
| 559 |
+
"loss": 0.1688,
|
| 560 |
+
"step": 11095
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 35.0,
|
| 564 |
+
"eval_accuracy": 0.9506451612903226,
|
| 565 |
+
"eval_loss": 0.2832624018192291,
|
| 566 |
+
"eval_runtime": 1.5946,
|
| 567 |
+
"eval_samples_per_second": 1944.052,
|
| 568 |
+
"eval_steps_per_second": 40.762,
|
| 569 |
+
"step": 11130
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 35.886792452830186,
|
| 573 |
+
"grad_norm": 0.1566316783428192,
|
| 574 |
+
"learning_rate": 1.030086690464049e-05,
|
| 575 |
+
"loss": 0.1686,
|
| 576 |
+
"step": 11412
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 36.0,
|
| 580 |
+
"eval_accuracy": 0.9503225806451613,
|
| 581 |
+
"eval_loss": 0.28356868028640747,
|
| 582 |
+
"eval_runtime": 1.5928,
|
| 583 |
+
"eval_samples_per_second": 1946.198,
|
| 584 |
+
"eval_steps_per_second": 40.807,
|
| 585 |
+
"step": 11448
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 36.88364779874214,
|
| 589 |
+
"grad_norm": 0.17213183641433716,
|
| 590 |
+
"learning_rate": 1.0031446540880504e-05,
|
| 591 |
+
"loss": 0.1685,
|
| 592 |
+
"step": 11729
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 37.0,
|
| 596 |
+
"eval_accuracy": 0.9506451612903226,
|
| 597 |
+
"eval_loss": 0.28204575181007385,
|
| 598 |
+
"eval_runtime": 1.5933,
|
| 599 |
+
"eval_samples_per_second": 1945.652,
|
| 600 |
+
"eval_steps_per_second": 40.796,
|
| 601 |
+
"step": 11766
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 37.880503144654085,
|
| 605 |
+
"grad_norm": 0.1503811776638031,
|
| 606 |
+
"learning_rate": 9.762026177120517e-06,
|
| 607 |
+
"loss": 0.1685,
|
| 608 |
+
"step": 12046
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 38.0,
|
| 612 |
+
"eval_accuracy": 0.9506451612903226,
|
| 613 |
+
"eval_loss": 0.2826063930988312,
|
| 614 |
+
"eval_runtime": 1.5945,
|
| 615 |
+
"eval_samples_per_second": 1944.184,
|
| 616 |
+
"eval_steps_per_second": 40.765,
|
| 617 |
+
"step": 12084
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 38.87735849056604,
|
| 621 |
+
"grad_norm": 0.1548190414905548,
|
| 622 |
+
"learning_rate": 9.492605813360531e-06,
|
| 623 |
+
"loss": 0.1679,
|
| 624 |
+
"step": 12363
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 39.0,
|
| 628 |
+
"eval_accuracy": 0.9519354838709677,
|
| 629 |
+
"eval_loss": 0.2820790410041809,
|
| 630 |
+
"eval_runtime": 1.5867,
|
| 631 |
+
"eval_samples_per_second": 1953.739,
|
| 632 |
+
"eval_steps_per_second": 40.966,
|
| 633 |
+
"step": 12402
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 39.874213836477985,
|
| 637 |
+
"grad_norm": 0.15846258401870728,
|
| 638 |
+
"learning_rate": 9.223185449600544e-06,
|
| 639 |
+
"loss": 0.168,
|
| 640 |
+
"step": 12680
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 40.0,
|
| 644 |
+
"eval_accuracy": 0.9512903225806452,
|
| 645 |
+
"eval_loss": 0.27980658411979675,
|
| 646 |
+
"eval_runtime": 1.5972,
|
| 647 |
+
"eval_samples_per_second": 1940.923,
|
| 648 |
+
"eval_steps_per_second": 40.697,
|
| 649 |
+
"step": 12720
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 40.87106918238994,
|
| 653 |
+
"grad_norm": 0.1553773283958435,
|
| 654 |
+
"learning_rate": 8.953765085840558e-06,
|
| 655 |
+
"loss": 0.1677,
|
| 656 |
+
"step": 12997
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 41.0,
|
| 660 |
+
"eval_accuracy": 0.9519354838709677,
|
| 661 |
+
"eval_loss": 0.28106164932250977,
|
| 662 |
+
"eval_runtime": 1.5869,
|
| 663 |
+
"eval_samples_per_second": 1953.442,
|
| 664 |
+
"eval_steps_per_second": 40.959,
|
| 665 |
+
"step": 13038
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 41.867924528301884,
|
| 669 |
+
"grad_norm": 0.12552271783351898,
|
| 670 |
+
"learning_rate": 8.684344722080573e-06,
|
| 671 |
+
"loss": 0.1673,
|
| 672 |
+
"step": 13314
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 42.0,
|
| 676 |
+
"eval_accuracy": 0.9509677419354838,
|
| 677 |
+
"eval_loss": 0.28286346793174744,
|
| 678 |
+
"eval_runtime": 1.586,
|
| 679 |
+
"eval_samples_per_second": 1954.589,
|
| 680 |
+
"eval_steps_per_second": 40.983,
|
| 681 |
+
"step": 13356
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 42.86477987421384,
|
| 685 |
+
"grad_norm": 0.15184646844863892,
|
| 686 |
+
"learning_rate": 8.414924358320585e-06,
|
| 687 |
+
"loss": 0.1674,
|
| 688 |
+
"step": 13631
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 43.0,
|
| 692 |
+
"eval_accuracy": 0.9525806451612904,
|
| 693 |
+
"eval_loss": 0.2814062535762787,
|
| 694 |
+
"eval_runtime": 1.592,
|
| 695 |
+
"eval_samples_per_second": 1947.238,
|
| 696 |
+
"eval_steps_per_second": 40.829,
|
| 697 |
+
"step": 13674
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 43.861635220125784,
|
| 701 |
+
"grad_norm": 0.11857682466506958,
|
| 702 |
+
"learning_rate": 8.145503994560598e-06,
|
| 703 |
+
"loss": 0.1673,
|
| 704 |
+
"step": 13948
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 44.0,
|
| 708 |
+
"eval_accuracy": 0.952258064516129,
|
| 709 |
+
"eval_loss": 0.28064581751823425,
|
| 710 |
+
"eval_runtime": 1.589,
|
| 711 |
+
"eval_samples_per_second": 1950.913,
|
| 712 |
+
"eval_steps_per_second": 40.906,
|
| 713 |
+
"step": 13992
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 44.85849056603774,
|
| 717 |
+
"grad_norm": 0.13426660001277924,
|
| 718 |
+
"learning_rate": 7.876083630800612e-06,
|
| 719 |
+
"loss": 0.1671,
|
| 720 |
+
"step": 14265
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 45.0,
|
| 724 |
+
"eval_accuracy": 0.952258064516129,
|
| 725 |
+
"eval_loss": 0.2805019021034241,
|
| 726 |
+
"eval_runtime": 1.5857,
|
| 727 |
+
"eval_samples_per_second": 1954.917,
|
| 728 |
+
"eval_steps_per_second": 40.99,
|
| 729 |
+
"step": 14310
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 45.855345911949684,
|
| 733 |
+
"grad_norm": 0.1329268366098404,
|
| 734 |
+
"learning_rate": 7.606663267040626e-06,
|
| 735 |
+
"loss": 0.1666,
|
| 736 |
+
"step": 14582
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 46.0,
|
| 740 |
+
"eval_accuracy": 0.9512903225806452,
|
| 741 |
+
"eval_loss": 0.2800099849700928,
|
| 742 |
+
"eval_runtime": 1.5902,
|
| 743 |
+
"eval_samples_per_second": 1949.437,
|
| 744 |
+
"eval_steps_per_second": 40.875,
|
| 745 |
+
"step": 14628
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 46.85220125786164,
|
| 749 |
+
"grad_norm": 0.17084655165672302,
|
| 750 |
+
"learning_rate": 7.33724290328064e-06,
|
| 751 |
+
"loss": 0.1668,
|
| 752 |
+
"step": 14899
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 47.0,
|
| 756 |
+
"eval_accuracy": 0.952258064516129,
|
| 757 |
+
"eval_loss": 0.2803304195404053,
|
| 758 |
+
"eval_runtime": 1.5878,
|
| 759 |
+
"eval_samples_per_second": 1952.385,
|
| 760 |
+
"eval_steps_per_second": 40.937,
|
| 761 |
+
"step": 14946
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 47.84905660377358,
|
| 765 |
+
"grad_norm": 0.1700143814086914,
|
| 766 |
+
"learning_rate": 7.067822539520653e-06,
|
| 767 |
+
"loss": 0.1666,
|
| 768 |
+
"step": 15216
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 48.0,
|
| 772 |
+
"eval_accuracy": 0.9519354838709677,
|
| 773 |
+
"eval_loss": 0.27888280153274536,
|
| 774 |
+
"eval_runtime": 1.5824,
|
| 775 |
+
"eval_samples_per_second": 1959.015,
|
| 776 |
+
"eval_steps_per_second": 41.076,
|
| 777 |
+
"step": 15264
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 48.84591194968554,
|
| 781 |
+
"grad_norm": 0.1518237590789795,
|
| 782 |
+
"learning_rate": 6.798402175760667e-06,
|
| 783 |
+
"loss": 0.1668,
|
| 784 |
+
"step": 15533
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 49.0,
|
| 788 |
+
"eval_accuracy": 0.9509677419354838,
|
| 789 |
+
"eval_loss": 0.2804841995239258,
|
| 790 |
+
"eval_runtime": 1.5832,
|
| 791 |
+
"eval_samples_per_second": 1958.042,
|
| 792 |
+
"eval_steps_per_second": 41.056,
|
| 793 |
+
"step": 15582
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 49.84276729559748,
|
| 797 |
+
"grad_norm": 0.16697125136852264,
|
| 798 |
+
"learning_rate": 6.528981812000681e-06,
|
| 799 |
+
"loss": 0.1662,
|
| 800 |
+
"step": 15850
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 50.0,
|
| 804 |
+
"eval_accuracy": 0.9529032258064516,
|
| 805 |
+
"eval_loss": 0.2788369953632355,
|
| 806 |
+
"eval_runtime": 1.5946,
|
| 807 |
+
"eval_samples_per_second": 1944.025,
|
| 808 |
+
"eval_steps_per_second": 40.762,
|
| 809 |
+
"step": 15900
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 50.839622641509436,
|
| 813 |
+
"grad_norm": 0.14791205525398254,
|
| 814 |
+
"learning_rate": 6.259561448240694e-06,
|
| 815 |
+
"loss": 0.1663,
|
| 816 |
+
"step": 16167
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 51.0,
|
| 820 |
+
"eval_accuracy": 0.9516129032258065,
|
| 821 |
+
"eval_loss": 0.2813352644443512,
|
| 822 |
+
"eval_runtime": 1.5922,
|
| 823 |
+
"eval_samples_per_second": 1947.046,
|
| 824 |
+
"eval_steps_per_second": 40.825,
|
| 825 |
+
"step": 16218
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 51.83647798742138,
|
| 829 |
+
"grad_norm": 0.1317252218723297,
|
| 830 |
+
"learning_rate": 5.990141084480707e-06,
|
| 831 |
+
"loss": 0.1663,
|
| 832 |
+
"step": 16484
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 52.0,
|
| 836 |
+
"eval_accuracy": 0.9516129032258065,
|
| 837 |
+
"eval_loss": 0.278254896402359,
|
| 838 |
+
"eval_runtime": 1.5877,
|
| 839 |
+
"eval_samples_per_second": 1952.515,
|
| 840 |
+
"eval_steps_per_second": 40.94,
|
| 841 |
+
"step": 16536
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 52.833333333333336,
|
| 845 |
+
"grad_norm": 0.12177319079637527,
|
| 846 |
+
"learning_rate": 5.720720720720722e-06,
|
| 847 |
+
"loss": 0.1661,
|
| 848 |
+
"step": 16801
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 53.0,
|
| 852 |
+
"eval_accuracy": 0.9512903225806452,
|
| 853 |
+
"eval_loss": 0.2790258526802063,
|
| 854 |
+
"eval_runtime": 1.5943,
|
| 855 |
+
"eval_samples_per_second": 1944.473,
|
| 856 |
+
"eval_steps_per_second": 40.771,
|
| 857 |
+
"step": 16854
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 53.83018867924528,
|
| 861 |
+
"grad_norm": 0.1430531144142151,
|
| 862 |
+
"learning_rate": 5.451300356960735e-06,
|
| 863 |
+
"loss": 0.1658,
|
| 864 |
+
"step": 17118
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 54.0,
|
| 868 |
+
"eval_accuracy": 0.9525806451612904,
|
| 869 |
+
"eval_loss": 0.27935537695884705,
|
| 870 |
+
"eval_runtime": 1.5954,
|
| 871 |
+
"eval_samples_per_second": 1943.111,
|
| 872 |
+
"eval_steps_per_second": 40.743,
|
| 873 |
+
"step": 17172
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 54.827044025157235,
|
| 877 |
+
"grad_norm": 0.13812412321567535,
|
| 878 |
+
"learning_rate": 5.181879993200748e-06,
|
| 879 |
+
"loss": 0.1659,
|
| 880 |
+
"step": 17435
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 55.0,
|
| 884 |
+
"eval_accuracy": 0.9532258064516129,
|
| 885 |
+
"eval_loss": 0.28016936779022217,
|
| 886 |
+
"eval_runtime": 1.5893,
|
| 887 |
+
"eval_samples_per_second": 1950.573,
|
| 888 |
+
"eval_steps_per_second": 40.899,
|
| 889 |
+
"step": 17490
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 55.82389937106918,
|
| 893 |
+
"grad_norm": 0.19225658476352692,
|
| 894 |
+
"learning_rate": 4.912459629440762e-06,
|
| 895 |
+
"loss": 0.1661,
|
| 896 |
+
"step": 17752
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 56.0,
|
| 900 |
+
"eval_accuracy": 0.9525806451612904,
|
| 901 |
+
"eval_loss": 0.2795941233634949,
|
| 902 |
+
"eval_runtime": 1.5827,
|
| 903 |
+
"eval_samples_per_second": 1958.672,
|
| 904 |
+
"eval_steps_per_second": 41.069,
|
| 905 |
+
"step": 17808
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 56.820754716981135,
|
| 909 |
+
"grad_norm": 0.1685912162065506,
|
| 910 |
+
"learning_rate": 4.6430392656807755e-06,
|
| 911 |
+
"loss": 0.1654,
|
| 912 |
+
"step": 18069
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 57.0,
|
| 916 |
+
"eval_accuracy": 0.9516129032258065,
|
| 917 |
+
"eval_loss": 0.2792729437351227,
|
| 918 |
+
"eval_runtime": 1.5874,
|
| 919 |
+
"eval_samples_per_second": 1952.898,
|
| 920 |
+
"eval_steps_per_second": 40.948,
|
| 921 |
+
"step": 18126
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 57.81761006289308,
|
| 925 |
+
"grad_norm": 0.13640663027763367,
|
| 926 |
+
"learning_rate": 4.373618901920789e-06,
|
| 927 |
+
"loss": 0.1658,
|
| 928 |
+
"step": 18386
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 58.0,
|
| 932 |
+
"eval_accuracy": 0.952258064516129,
|
| 933 |
+
"eval_loss": 0.27936556935310364,
|
| 934 |
+
"eval_runtime": 1.5908,
|
| 935 |
+
"eval_samples_per_second": 1948.765,
|
| 936 |
+
"eval_steps_per_second": 40.861,
|
| 937 |
+
"step": 18444
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 58.814465408805034,
|
| 941 |
+
"grad_norm": 0.181793674826622,
|
| 942 |
+
"learning_rate": 4.1041985381608024e-06,
|
| 943 |
+
"loss": 0.1654,
|
| 944 |
+
"step": 18703
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 59.0,
|
| 948 |
+
"eval_accuracy": 0.9509677419354838,
|
| 949 |
+
"eval_loss": 0.27835920453071594,
|
| 950 |
+
"eval_runtime": 1.583,
|
| 951 |
+
"eval_samples_per_second": 1958.277,
|
| 952 |
+
"eval_steps_per_second": 41.061,
|
| 953 |
+
"step": 18762
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 59.81132075471698,
|
| 957 |
+
"grad_norm": 0.13843421638011932,
|
| 958 |
+
"learning_rate": 3.834778174400816e-06,
|
| 959 |
+
"loss": 0.1658,
|
| 960 |
+
"step": 19020
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 60.0,
|
| 964 |
+
"eval_accuracy": 0.9525806451612904,
|
| 965 |
+
"eval_loss": 0.28052493929862976,
|
| 966 |
+
"eval_runtime": 1.5893,
|
| 967 |
+
"eval_samples_per_second": 1950.524,
|
| 968 |
+
"eval_steps_per_second": 40.898,
|
| 969 |
+
"step": 19080
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 60.808176100628934,
|
| 973 |
+
"grad_norm": 0.14343231916427612,
|
| 974 |
+
"learning_rate": 3.56535781064083e-06,
|
| 975 |
+
"loss": 0.1653,
|
| 976 |
+
"step": 19337
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 61.0,
|
| 980 |
+
"eval_accuracy": 0.952258064516129,
|
| 981 |
+
"eval_loss": 0.27861472964286804,
|
| 982 |
+
"eval_runtime": 1.5821,
|
| 983 |
+
"eval_samples_per_second": 1959.483,
|
| 984 |
+
"eval_steps_per_second": 41.086,
|
| 985 |
+
"step": 19398
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 61.80503144654088,
|
| 989 |
+
"grad_norm": 0.14610852301120758,
|
| 990 |
+
"learning_rate": 3.2959374468808437e-06,
|
| 991 |
+
"loss": 0.1655,
|
| 992 |
+
"step": 19654
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 62.0,
|
| 996 |
+
"eval_accuracy": 0.9512903225806452,
|
| 997 |
+
"eval_loss": 0.27765244245529175,
|
| 998 |
+
"eval_runtime": 1.5812,
|
| 999 |
+
"eval_samples_per_second": 1960.523,
|
| 1000 |
+
"eval_steps_per_second": 41.108,
|
| 1001 |
+
"step": 19716
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 62.801886792452834,
|
| 1005 |
+
"grad_norm": 0.13857333362102509,
|
| 1006 |
+
"learning_rate": 3.0265170831208568e-06,
|
| 1007 |
+
"loss": 0.1653,
|
| 1008 |
+
"step": 19971
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 63.0,
|
| 1012 |
+
"eval_accuracy": 0.9532258064516129,
|
| 1013 |
+
"eval_loss": 0.27754127979278564,
|
| 1014 |
+
"eval_runtime": 1.5909,
|
| 1015 |
+
"eval_samples_per_second": 1948.602,
|
| 1016 |
+
"eval_steps_per_second": 40.858,
|
| 1017 |
+
"step": 20034
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 63.79874213836478,
|
| 1021 |
+
"grad_norm": 0.1268884837627411,
|
| 1022 |
+
"learning_rate": 2.7570967193608707e-06,
|
| 1023 |
+
"loss": 0.1652,
|
| 1024 |
+
"step": 20288
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 64.0,
|
| 1028 |
+
"eval_accuracy": 0.9519354838709677,
|
| 1029 |
+
"eval_loss": 0.2781769037246704,
|
| 1030 |
+
"eval_runtime": 1.5911,
|
| 1031 |
+
"eval_samples_per_second": 1948.283,
|
| 1032 |
+
"eval_steps_per_second": 40.851,
|
| 1033 |
+
"step": 20352
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 64.79559748427673,
|
| 1037 |
+
"grad_norm": 0.1306312382221222,
|
| 1038 |
+
"learning_rate": 2.487676355600884e-06,
|
| 1039 |
+
"loss": 0.1652,
|
| 1040 |
+
"step": 20605
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 65.0,
|
| 1044 |
+
"eval_accuracy": 0.9525806451612904,
|
| 1045 |
+
"eval_loss": 0.2785084843635559,
|
| 1046 |
+
"eval_runtime": 1.587,
|
| 1047 |
+
"eval_samples_per_second": 1953.372,
|
| 1048 |
+
"eval_steps_per_second": 40.958,
|
| 1049 |
+
"step": 20670
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 65.79245283018868,
|
| 1053 |
+
"grad_norm": 0.12314357608556747,
|
| 1054 |
+
"learning_rate": 2.2182559918408976e-06,
|
| 1055 |
+
"loss": 0.1654,
|
| 1056 |
+
"step": 20922
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 66.0,
|
| 1060 |
+
"eval_accuracy": 0.952258064516129,
|
| 1061 |
+
"eval_loss": 0.27857428789138794,
|
| 1062 |
+
"eval_runtime": 1.5854,
|
| 1063 |
+
"eval_samples_per_second": 1955.312,
|
| 1064 |
+
"eval_steps_per_second": 40.998,
|
| 1065 |
+
"step": 20988
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 66.78930817610063,
|
| 1069 |
+
"grad_norm": 0.15283074975013733,
|
| 1070 |
+
"learning_rate": 1.948835628080911e-06,
|
| 1071 |
+
"loss": 0.1649,
|
| 1072 |
+
"step": 21239
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 67.0,
|
| 1076 |
+
"eval_accuracy": 0.9525806451612904,
|
| 1077 |
+
"eval_loss": 0.2788286805152893,
|
| 1078 |
+
"eval_runtime": 1.5951,
|
| 1079 |
+
"eval_samples_per_second": 1943.445,
|
| 1080 |
+
"eval_steps_per_second": 40.75,
|
| 1081 |
+
"step": 21306
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 67.78616352201257,
|
| 1085 |
+
"grad_norm": 0.12186983227729797,
|
| 1086 |
+
"learning_rate": 1.679415264320925e-06,
|
| 1087 |
+
"loss": 0.165,
|
| 1088 |
+
"step": 21556
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 68.0,
|
| 1092 |
+
"eval_accuracy": 0.952258064516129,
|
| 1093 |
+
"eval_loss": 0.27811065316200256,
|
| 1094 |
+
"eval_runtime": 1.5914,
|
| 1095 |
+
"eval_samples_per_second": 1947.92,
|
| 1096 |
+
"eval_steps_per_second": 40.843,
|
| 1097 |
+
"step": 21624
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 68.78301886792453,
|
| 1101 |
+
"grad_norm": 0.1512852907180786,
|
| 1102 |
+
"learning_rate": 1.4099949005609384e-06,
|
| 1103 |
+
"loss": 0.165,
|
| 1104 |
+
"step": 21873
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 69.0,
|
| 1108 |
+
"eval_accuracy": 0.9519354838709677,
|
| 1109 |
+
"eval_loss": 0.277495801448822,
|
| 1110 |
+
"eval_runtime": 1.5848,
|
| 1111 |
+
"eval_samples_per_second": 1956.025,
|
| 1112 |
+
"eval_steps_per_second": 41.013,
|
| 1113 |
+
"step": 21942
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 69.77987421383648,
|
| 1117 |
+
"grad_norm": 0.14219249784946442,
|
| 1118 |
+
"learning_rate": 1.140574536800952e-06,
|
| 1119 |
+
"loss": 0.165,
|
| 1120 |
+
"step": 22190
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 70.0,
|
| 1124 |
+
"eval_accuracy": 0.952258064516129,
|
| 1125 |
+
"eval_loss": 0.2772570252418518,
|
| 1126 |
+
"eval_runtime": 1.5882,
|
| 1127 |
+
"eval_samples_per_second": 1951.939,
|
| 1128 |
+
"eval_steps_per_second": 40.928,
|
| 1129 |
+
"step": 22260
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 70.77672955974843,
|
| 1133 |
+
"grad_norm": 0.12878191471099854,
|
| 1134 |
+
"learning_rate": 8.711541730409656e-07,
|
| 1135 |
+
"loss": 0.165,
|
| 1136 |
+
"step": 22507
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 71.0,
|
| 1140 |
+
"eval_accuracy": 0.9525806451612904,
|
| 1141 |
+
"eval_loss": 0.27687469124794006,
|
| 1142 |
+
"eval_runtime": 1.5932,
|
| 1143 |
+
"eval_samples_per_second": 1945.737,
|
| 1144 |
+
"eval_steps_per_second": 40.798,
|
| 1145 |
+
"step": 22578
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 71.77358490566037,
|
| 1149 |
+
"grad_norm": 0.1408313363790512,
|
| 1150 |
+
"learning_rate": 6.017338092809792e-07,
|
| 1151 |
+
"loss": 0.1651,
|
| 1152 |
+
"step": 22824
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 72.0,
|
| 1156 |
+
"eval_accuracy": 0.9532258064516129,
|
| 1157 |
+
"eval_loss": 0.2769748866558075,
|
| 1158 |
+
"eval_runtime": 1.5892,
|
| 1159 |
+
"eval_samples_per_second": 1950.609,
|
| 1160 |
+
"eval_steps_per_second": 40.9,
|
| 1161 |
+
"step": 22896
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 72.77044025157232,
|
| 1165 |
+
"grad_norm": 0.13329148292541504,
|
| 1166 |
+
"learning_rate": 3.3231344552099275e-07,
|
| 1167 |
+
"loss": 0.1648,
|
| 1168 |
+
"step": 23141
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 73.0,
|
| 1172 |
+
"eval_accuracy": 0.9535483870967741,
|
| 1173 |
+
"eval_loss": 0.2772596478462219,
|
| 1174 |
+
"eval_runtime": 1.5912,
|
| 1175 |
+
"eval_samples_per_second": 1948.155,
|
| 1176 |
+
"eval_steps_per_second": 40.848,
|
| 1177 |
+
"step": 23214
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 73.76729559748428,
|
| 1181 |
+
"grad_norm": 0.11299853026866913,
|
| 1182 |
+
"learning_rate": 6.289308176100629e-08,
|
| 1183 |
+
"loss": 0.1652,
|
| 1184 |
+
"step": 23458
|
| 1185 |
+
}
|
| 1186 |
+
],
|
| 1187 |
+
"logging_steps": 317,
|
| 1188 |
+
"max_steps": 23532,
|
| 1189 |
+
"num_input_tokens_seen": 0,
|
| 1190 |
+
"num_train_epochs": 74,
|
| 1191 |
+
"save_steps": 1000000000.0,
|
| 1192 |
+
"stateful_callbacks": {
|
| 1193 |
+
"TrainerControl": {
|
| 1194 |
+
"args": {
|
| 1195 |
+
"should_epoch_stop": false,
|
| 1196 |
+
"should_evaluate": false,
|
| 1197 |
+
"should_log": false,
|
| 1198 |
+
"should_save": true,
|
| 1199 |
+
"should_training_stop": true
|
| 1200 |
+
},
|
| 1201 |
+
"attributes": {}
|
| 1202 |
+
}
|
| 1203 |
+
},
|
| 1204 |
+
"total_flos": 6098875529253696.0,
|
| 1205 |
+
"train_batch_size": 48,
|
| 1206 |
+
"trial_name": null,
|
| 1207 |
+
"trial_params": {
|
| 1208 |
+
"alpha": 0.5969475255349722,
|
| 1209 |
+
"num_train_epochs": 74,
|
| 1210 |
+
"temperature": 2
|
| 1211 |
+
}
|
| 1212 |
+
}
|
run-0/checkpoint-23532/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d11aee97ba5d0d9bc630562bed55f9ef81cbe28c4bdd939c7abdefc104c065c
|
| 3 |
+
size 5304
|
run-0/checkpoint-23532/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
run-1/checkpoint-24168/config.json
ADDED
|
@@ -0,0 +1,331 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilbert-base-uncased",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"hidden_dim": 3072,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "restaurant_reviews",
|
| 13 |
+
"1": "nutrition_info",
|
| 14 |
+
"2": "account_blocked",
|
| 15 |
+
"3": "oil_change_how",
|
| 16 |
+
"4": "time",
|
| 17 |
+
"5": "weather",
|
| 18 |
+
"6": "redeem_rewards",
|
| 19 |
+
"7": "interest_rate",
|
| 20 |
+
"8": "gas_type",
|
| 21 |
+
"9": "accept_reservations",
|
| 22 |
+
"10": "smart_home",
|
| 23 |
+
"11": "user_name",
|
| 24 |
+
"12": "report_lost_card",
|
| 25 |
+
"13": "repeat",
|
| 26 |
+
"14": "whisper_mode",
|
| 27 |
+
"15": "what_are_your_hobbies",
|
| 28 |
+
"16": "order",
|
| 29 |
+
"17": "jump_start",
|
| 30 |
+
"18": "schedule_meeting",
|
| 31 |
+
"19": "meeting_schedule",
|
| 32 |
+
"20": "freeze_account",
|
| 33 |
+
"21": "what_song",
|
| 34 |
+
"22": "meaning_of_life",
|
| 35 |
+
"23": "restaurant_reservation",
|
| 36 |
+
"24": "traffic",
|
| 37 |
+
"25": "make_call",
|
| 38 |
+
"26": "text",
|
| 39 |
+
"27": "bill_balance",
|
| 40 |
+
"28": "improve_credit_score",
|
| 41 |
+
"29": "change_language",
|
| 42 |
+
"30": "no",
|
| 43 |
+
"31": "measurement_conversion",
|
| 44 |
+
"32": "timer",
|
| 45 |
+
"33": "flip_coin",
|
| 46 |
+
"34": "do_you_have_pets",
|
| 47 |
+
"35": "balance",
|
| 48 |
+
"36": "tell_joke",
|
| 49 |
+
"37": "last_maintenance",
|
| 50 |
+
"38": "exchange_rate",
|
| 51 |
+
"39": "uber",
|
| 52 |
+
"40": "car_rental",
|
| 53 |
+
"41": "credit_limit",
|
| 54 |
+
"42": "oos",
|
| 55 |
+
"43": "shopping_list",
|
| 56 |
+
"44": "expiration_date",
|
| 57 |
+
"45": "routing",
|
| 58 |
+
"46": "meal_suggestion",
|
| 59 |
+
"47": "tire_change",
|
| 60 |
+
"48": "todo_list",
|
| 61 |
+
"49": "card_declined",
|
| 62 |
+
"50": "rewards_balance",
|
| 63 |
+
"51": "change_accent",
|
| 64 |
+
"52": "vaccines",
|
| 65 |
+
"53": "reminder_update",
|
| 66 |
+
"54": "food_last",
|
| 67 |
+
"55": "change_ai_name",
|
| 68 |
+
"56": "bill_due",
|
| 69 |
+
"57": "who_do_you_work_for",
|
| 70 |
+
"58": "share_location",
|
| 71 |
+
"59": "international_visa",
|
| 72 |
+
"60": "calendar",
|
| 73 |
+
"61": "translate",
|
| 74 |
+
"62": "carry_on",
|
| 75 |
+
"63": "book_flight",
|
| 76 |
+
"64": "insurance_change",
|
| 77 |
+
"65": "todo_list_update",
|
| 78 |
+
"66": "timezone",
|
| 79 |
+
"67": "cancel_reservation",
|
| 80 |
+
"68": "transactions",
|
| 81 |
+
"69": "credit_score",
|
| 82 |
+
"70": "report_fraud",
|
| 83 |
+
"71": "spending_history",
|
| 84 |
+
"72": "directions",
|
| 85 |
+
"73": "spelling",
|
| 86 |
+
"74": "insurance",
|
| 87 |
+
"75": "what_is_your_name",
|
| 88 |
+
"76": "reminder",
|
| 89 |
+
"77": "where_are_you_from",
|
| 90 |
+
"78": "distance",
|
| 91 |
+
"79": "payday",
|
| 92 |
+
"80": "flight_status",
|
| 93 |
+
"81": "find_phone",
|
| 94 |
+
"82": "greeting",
|
| 95 |
+
"83": "alarm",
|
| 96 |
+
"84": "order_status",
|
| 97 |
+
"85": "confirm_reservation",
|
| 98 |
+
"86": "cook_time",
|
| 99 |
+
"87": "damaged_card",
|
| 100 |
+
"88": "reset_settings",
|
| 101 |
+
"89": "pin_change",
|
| 102 |
+
"90": "replacement_card_duration",
|
| 103 |
+
"91": "new_card",
|
| 104 |
+
"92": "roll_dice",
|
| 105 |
+
"93": "income",
|
| 106 |
+
"94": "taxes",
|
| 107 |
+
"95": "date",
|
| 108 |
+
"96": "who_made_you",
|
| 109 |
+
"97": "pto_request",
|
| 110 |
+
"98": "tire_pressure",
|
| 111 |
+
"99": "how_old_are_you",
|
| 112 |
+
"100": "rollover_401k",
|
| 113 |
+
"101": "pto_request_status",
|
| 114 |
+
"102": "how_busy",
|
| 115 |
+
"103": "application_status",
|
| 116 |
+
"104": "recipe",
|
| 117 |
+
"105": "calendar_update",
|
| 118 |
+
"106": "play_music",
|
| 119 |
+
"107": "yes",
|
| 120 |
+
"108": "direct_deposit",
|
| 121 |
+
"109": "credit_limit_change",
|
| 122 |
+
"110": "gas",
|
| 123 |
+
"111": "pay_bill",
|
| 124 |
+
"112": "ingredients_list",
|
| 125 |
+
"113": "lost_luggage",
|
| 126 |
+
"114": "goodbye",
|
| 127 |
+
"115": "what_can_i_ask_you",
|
| 128 |
+
"116": "book_hotel",
|
| 129 |
+
"117": "are_you_a_bot",
|
| 130 |
+
"118": "next_song",
|
| 131 |
+
"119": "change_speed",
|
| 132 |
+
"120": "plug_type",
|
| 133 |
+
"121": "maybe",
|
| 134 |
+
"122": "w2",
|
| 135 |
+
"123": "oil_change_when",
|
| 136 |
+
"124": "thank_you",
|
| 137 |
+
"125": "shopping_list_update",
|
| 138 |
+
"126": "pto_balance",
|
| 139 |
+
"127": "order_checks",
|
| 140 |
+
"128": "travel_alert",
|
| 141 |
+
"129": "fun_fact",
|
| 142 |
+
"130": "sync_device",
|
| 143 |
+
"131": "schedule_maintenance",
|
| 144 |
+
"132": "apr",
|
| 145 |
+
"133": "transfer",
|
| 146 |
+
"134": "ingredient_substitution",
|
| 147 |
+
"135": "calories",
|
| 148 |
+
"136": "current_location",
|
| 149 |
+
"137": "international_fees",
|
| 150 |
+
"138": "calculator",
|
| 151 |
+
"139": "definition",
|
| 152 |
+
"140": "next_holiday",
|
| 153 |
+
"141": "update_playlist",
|
| 154 |
+
"142": "mpg",
|
| 155 |
+
"143": "min_payment",
|
| 156 |
+
"144": "change_user_name",
|
| 157 |
+
"145": "restaurant_suggestion",
|
| 158 |
+
"146": "travel_notification",
|
| 159 |
+
"147": "cancel",
|
| 160 |
+
"148": "pto_used",
|
| 161 |
+
"149": "travel_suggestion",
|
| 162 |
+
"150": "change_volume"
|
| 163 |
+
},
|
| 164 |
+
"initializer_range": 0.02,
|
| 165 |
+
"label2id": {
|
| 166 |
+
"accept_reservations": 9,
|
| 167 |
+
"account_blocked": 2,
|
| 168 |
+
"alarm": 83,
|
| 169 |
+
"application_status": 103,
|
| 170 |
+
"apr": 132,
|
| 171 |
+
"are_you_a_bot": 117,
|
| 172 |
+
"balance": 35,
|
| 173 |
+
"bill_balance": 27,
|
| 174 |
+
"bill_due": 56,
|
| 175 |
+
"book_flight": 63,
|
| 176 |
+
"book_hotel": 116,
|
| 177 |
+
"calculator": 138,
|
| 178 |
+
"calendar": 60,
|
| 179 |
+
"calendar_update": 105,
|
| 180 |
+
"calories": 135,
|
| 181 |
+
"cancel": 147,
|
| 182 |
+
"cancel_reservation": 67,
|
| 183 |
+
"car_rental": 40,
|
| 184 |
+
"card_declined": 49,
|
| 185 |
+
"carry_on": 62,
|
| 186 |
+
"change_accent": 51,
|
| 187 |
+
"change_ai_name": 55,
|
| 188 |
+
"change_language": 29,
|
| 189 |
+
"change_speed": 119,
|
| 190 |
+
"change_user_name": 144,
|
| 191 |
+
"change_volume": 150,
|
| 192 |
+
"confirm_reservation": 85,
|
| 193 |
+
"cook_time": 86,
|
| 194 |
+
"credit_limit": 41,
|
| 195 |
+
"credit_limit_change": 109,
|
| 196 |
+
"credit_score": 69,
|
| 197 |
+
"current_location": 136,
|
| 198 |
+
"damaged_card": 87,
|
| 199 |
+
"date": 95,
|
| 200 |
+
"definition": 139,
|
| 201 |
+
"direct_deposit": 108,
|
| 202 |
+
"directions": 72,
|
| 203 |
+
"distance": 78,
|
| 204 |
+
"do_you_have_pets": 34,
|
| 205 |
+
"exchange_rate": 38,
|
| 206 |
+
"expiration_date": 44,
|
| 207 |
+
"find_phone": 81,
|
| 208 |
+
"flight_status": 80,
|
| 209 |
+
"flip_coin": 33,
|
| 210 |
+
"food_last": 54,
|
| 211 |
+
"freeze_account": 20,
|
| 212 |
+
"fun_fact": 129,
|
| 213 |
+
"gas": 110,
|
| 214 |
+
"gas_type": 8,
|
| 215 |
+
"goodbye": 114,
|
| 216 |
+
"greeting": 82,
|
| 217 |
+
"how_busy": 102,
|
| 218 |
+
"how_old_are_you": 99,
|
| 219 |
+
"improve_credit_score": 28,
|
| 220 |
+
"income": 93,
|
| 221 |
+
"ingredient_substitution": 134,
|
| 222 |
+
"ingredients_list": 112,
|
| 223 |
+
"insurance": 74,
|
| 224 |
+
"insurance_change": 64,
|
| 225 |
+
"interest_rate": 7,
|
| 226 |
+
"international_fees": 137,
|
| 227 |
+
"international_visa": 59,
|
| 228 |
+
"jump_start": 17,
|
| 229 |
+
"last_maintenance": 37,
|
| 230 |
+
"lost_luggage": 113,
|
| 231 |
+
"make_call": 25,
|
| 232 |
+
"maybe": 121,
|
| 233 |
+
"meal_suggestion": 46,
|
| 234 |
+
"meaning_of_life": 22,
|
| 235 |
+
"measurement_conversion": 31,
|
| 236 |
+
"meeting_schedule": 19,
|
| 237 |
+
"min_payment": 143,
|
| 238 |
+
"mpg": 142,
|
| 239 |
+
"new_card": 91,
|
| 240 |
+
"next_holiday": 140,
|
| 241 |
+
"next_song": 118,
|
| 242 |
+
"no": 30,
|
| 243 |
+
"nutrition_info": 1,
|
| 244 |
+
"oil_change_how": 3,
|
| 245 |
+
"oil_change_when": 123,
|
| 246 |
+
"oos": 42,
|
| 247 |
+
"order": 16,
|
| 248 |
+
"order_checks": 127,
|
| 249 |
+
"order_status": 84,
|
| 250 |
+
"pay_bill": 111,
|
| 251 |
+
"payday": 79,
|
| 252 |
+
"pin_change": 89,
|
| 253 |
+
"play_music": 106,
|
| 254 |
+
"plug_type": 120,
|
| 255 |
+
"pto_balance": 126,
|
| 256 |
+
"pto_request": 97,
|
| 257 |
+
"pto_request_status": 101,
|
| 258 |
+
"pto_used": 148,
|
| 259 |
+
"recipe": 104,
|
| 260 |
+
"redeem_rewards": 6,
|
| 261 |
+
"reminder": 76,
|
| 262 |
+
"reminder_update": 53,
|
| 263 |
+
"repeat": 13,
|
| 264 |
+
"replacement_card_duration": 90,
|
| 265 |
+
"report_fraud": 70,
|
| 266 |
+
"report_lost_card": 12,
|
| 267 |
+
"reset_settings": 88,
|
| 268 |
+
"restaurant_reservation": 23,
|
| 269 |
+
"restaurant_reviews": 0,
|
| 270 |
+
"restaurant_suggestion": 145,
|
| 271 |
+
"rewards_balance": 50,
|
| 272 |
+
"roll_dice": 92,
|
| 273 |
+
"rollover_401k": 100,
|
| 274 |
+
"routing": 45,
|
| 275 |
+
"schedule_maintenance": 131,
|
| 276 |
+
"schedule_meeting": 18,
|
| 277 |
+
"share_location": 58,
|
| 278 |
+
"shopping_list": 43,
|
| 279 |
+
"shopping_list_update": 125,
|
| 280 |
+
"smart_home": 10,
|
| 281 |
+
"spelling": 73,
|
| 282 |
+
"spending_history": 71,
|
| 283 |
+
"sync_device": 130,
|
| 284 |
+
"taxes": 94,
|
| 285 |
+
"tell_joke": 36,
|
| 286 |
+
"text": 26,
|
| 287 |
+
"thank_you": 124,
|
| 288 |
+
"time": 4,
|
| 289 |
+
"timer": 32,
|
| 290 |
+
"timezone": 66,
|
| 291 |
+
"tire_change": 47,
|
| 292 |
+
"tire_pressure": 98,
|
| 293 |
+
"todo_list": 48,
|
| 294 |
+
"todo_list_update": 65,
|
| 295 |
+
"traffic": 24,
|
| 296 |
+
"transactions": 68,
|
| 297 |
+
"transfer": 133,
|
| 298 |
+
"translate": 61,
|
| 299 |
+
"travel_alert": 128,
|
| 300 |
+
"travel_notification": 146,
|
| 301 |
+
"travel_suggestion": 149,
|
| 302 |
+
"uber": 39,
|
| 303 |
+
"update_playlist": 141,
|
| 304 |
+
"user_name": 11,
|
| 305 |
+
"vaccines": 52,
|
| 306 |
+
"w2": 122,
|
| 307 |
+
"weather": 5,
|
| 308 |
+
"what_are_your_hobbies": 15,
|
| 309 |
+
"what_can_i_ask_you": 115,
|
| 310 |
+
"what_is_your_name": 75,
|
| 311 |
+
"what_song": 21,
|
| 312 |
+
"where_are_you_from": 77,
|
| 313 |
+
"whisper_mode": 14,
|
| 314 |
+
"who_do_you_work_for": 57,
|
| 315 |
+
"who_made_you": 96,
|
| 316 |
+
"yes": 107
|
| 317 |
+
},
|
| 318 |
+
"max_position_embeddings": 512,
|
| 319 |
+
"model_type": "distilbert",
|
| 320 |
+
"n_heads": 12,
|
| 321 |
+
"n_layers": 6,
|
| 322 |
+
"pad_token_id": 0,
|
| 323 |
+
"problem_type": "single_label_classification",
|
| 324 |
+
"qa_dropout": 0.1,
|
| 325 |
+
"seq_classif_dropout": 0.2,
|
| 326 |
+
"sinusoidal_pos_embds": false,
|
| 327 |
+
"tie_weights_": true,
|
| 328 |
+
"torch_dtype": "float32",
|
| 329 |
+
"transformers_version": "4.44.2",
|
| 330 |
+
"vocab_size": 30522
|
| 331 |
+
}
|
run-1/checkpoint-24168/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9bceb18af4151eb56df29f1ff07a865d4aa73d09bd377d20bbcdc87bb0a06d57
|
| 3 |
+
size 268290900
|
run-1/checkpoint-24168/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd782aa683c16dd7427c1ccc06a9c61c48bbe1d6da106d0031b1b99a661148e5
|
| 3 |
+
size 536643898
|
run-1/checkpoint-24168/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9a95fbf30085408132a4533fead34661bffd7b80a7155677e5c7e3cb70ff754f
|
| 3 |
+
size 14244
|
run-1/checkpoint-24168/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eb6dcdb206371c49a63809b060a79f1a9f3c56fe50c7c2a2b5582b435d9a234b
|
| 3 |
+
size 1064
|
run-1/checkpoint-24168/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
run-1/checkpoint-24168/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
run-1/checkpoint-24168/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": true,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"mask_token": "[MASK]",
|
| 48 |
+
"model_max_length": 512,
|
| 49 |
+
"pad_token": "[PAD]",
|
| 50 |
+
"sep_token": "[SEP]",
|
| 51 |
+
"strip_accents": null,
|
| 52 |
+
"tokenize_chinese_chars": true,
|
| 53 |
+
"tokenizer_class": "DistilBertTokenizer",
|
| 54 |
+
"unk_token": "[UNK]"
|
| 55 |
+
}
|
run-1/checkpoint-24168/trainer_state.json
ADDED
|
@@ -0,0 +1,1244 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 76.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 24168,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.9968553459119497,
|
| 13 |
+
"grad_norm": NaN,
|
| 14 |
+
"learning_rate": 1.973766964581265e-05,
|
| 15 |
+
"loss": 0.0,
|
| 16 |
+
"step": 317
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 21 |
+
"eval_loss": NaN,
|
| 22 |
+
"eval_runtime": 1.5444,
|
| 23 |
+
"eval_samples_per_second": 2007.293,
|
| 24 |
+
"eval_steps_per_second": 42.088,
|
| 25 |
+
"step": 318
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.9937106918238994,
|
| 29 |
+
"grad_norm": NaN,
|
| 30 |
+
"learning_rate": 1.947533929162529e-05,
|
| 31 |
+
"loss": 0.0,
|
| 32 |
+
"step": 634
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.0,
|
| 36 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 37 |
+
"eval_loss": NaN,
|
| 38 |
+
"eval_runtime": 1.5402,
|
| 39 |
+
"eval_samples_per_second": 2012.771,
|
| 40 |
+
"eval_steps_per_second": 42.203,
|
| 41 |
+
"step": 636
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 2.990566037735849,
|
| 45 |
+
"grad_norm": NaN,
|
| 46 |
+
"learning_rate": 1.9213008937437935e-05,
|
| 47 |
+
"loss": 0.0,
|
| 48 |
+
"step": 951
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 3.0,
|
| 52 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 53 |
+
"eval_loss": NaN,
|
| 54 |
+
"eval_runtime": 1.5431,
|
| 55 |
+
"eval_samples_per_second": 2008.973,
|
| 56 |
+
"eval_steps_per_second": 42.124,
|
| 57 |
+
"step": 954
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 3.9874213836477987,
|
| 61 |
+
"grad_norm": NaN,
|
| 62 |
+
"learning_rate": 1.895067858325058e-05,
|
| 63 |
+
"loss": 0.0,
|
| 64 |
+
"step": 1268
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 4.0,
|
| 68 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 69 |
+
"eval_loss": NaN,
|
| 70 |
+
"eval_runtime": 1.5447,
|
| 71 |
+
"eval_samples_per_second": 2006.804,
|
| 72 |
+
"eval_steps_per_second": 42.078,
|
| 73 |
+
"step": 1272
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.984276729559748,
|
| 77 |
+
"grad_norm": NaN,
|
| 78 |
+
"learning_rate": 1.8688348229063225e-05,
|
| 79 |
+
"loss": 0.0,
|
| 80 |
+
"step": 1585
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 5.0,
|
| 84 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 85 |
+
"eval_loss": NaN,
|
| 86 |
+
"eval_runtime": 1.5441,
|
| 87 |
+
"eval_samples_per_second": 2007.665,
|
| 88 |
+
"eval_steps_per_second": 42.096,
|
| 89 |
+
"step": 1590
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 5.981132075471698,
|
| 93 |
+
"grad_norm": NaN,
|
| 94 |
+
"learning_rate": 1.842601787487587e-05,
|
| 95 |
+
"loss": 0.0,
|
| 96 |
+
"step": 1902
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 6.0,
|
| 100 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 101 |
+
"eval_loss": NaN,
|
| 102 |
+
"eval_runtime": 1.5537,
|
| 103 |
+
"eval_samples_per_second": 1995.266,
|
| 104 |
+
"eval_steps_per_second": 41.836,
|
| 105 |
+
"step": 1908
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 6.977987421383648,
|
| 109 |
+
"grad_norm": NaN,
|
| 110 |
+
"learning_rate": 1.8163687520688515e-05,
|
| 111 |
+
"loss": 0.0,
|
| 112 |
+
"step": 2219
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 7.0,
|
| 116 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 117 |
+
"eval_loss": NaN,
|
| 118 |
+
"eval_runtime": 1.5462,
|
| 119 |
+
"eval_samples_per_second": 2004.977,
|
| 120 |
+
"eval_steps_per_second": 42.04,
|
| 121 |
+
"step": 2226
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 7.9748427672955975,
|
| 125 |
+
"grad_norm": NaN,
|
| 126 |
+
"learning_rate": 1.790135716650116e-05,
|
| 127 |
+
"loss": 0.0,
|
| 128 |
+
"step": 2536
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 133 |
+
"eval_loss": NaN,
|
| 134 |
+
"eval_runtime": 1.5503,
|
| 135 |
+
"eval_samples_per_second": 1999.613,
|
| 136 |
+
"eval_steps_per_second": 41.927,
|
| 137 |
+
"step": 2544
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 8.971698113207546,
|
| 141 |
+
"grad_norm": NaN,
|
| 142 |
+
"learning_rate": 1.7639026812313805e-05,
|
| 143 |
+
"loss": 0.0,
|
| 144 |
+
"step": 2853
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 9.0,
|
| 148 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 149 |
+
"eval_loss": NaN,
|
| 150 |
+
"eval_runtime": 1.5538,
|
| 151 |
+
"eval_samples_per_second": 1995.075,
|
| 152 |
+
"eval_steps_per_second": 41.832,
|
| 153 |
+
"step": 2862
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 9.968553459119496,
|
| 157 |
+
"grad_norm": NaN,
|
| 158 |
+
"learning_rate": 1.7376696458126448e-05,
|
| 159 |
+
"loss": 0.0,
|
| 160 |
+
"step": 3170
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 10.0,
|
| 164 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 165 |
+
"eval_loss": NaN,
|
| 166 |
+
"eval_runtime": 1.5525,
|
| 167 |
+
"eval_samples_per_second": 1996.753,
|
| 168 |
+
"eval_steps_per_second": 41.867,
|
| 169 |
+
"step": 3180
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.965408805031446,
|
| 173 |
+
"grad_norm": NaN,
|
| 174 |
+
"learning_rate": 1.7114366103939095e-05,
|
| 175 |
+
"loss": 0.0,
|
| 176 |
+
"step": 3487
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 11.0,
|
| 180 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 181 |
+
"eval_loss": NaN,
|
| 182 |
+
"eval_runtime": 1.5476,
|
| 183 |
+
"eval_samples_per_second": 2003.139,
|
| 184 |
+
"eval_steps_per_second": 42.001,
|
| 185 |
+
"step": 3498
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 11.962264150943396,
|
| 189 |
+
"grad_norm": NaN,
|
| 190 |
+
"learning_rate": 1.6852035749751738e-05,
|
| 191 |
+
"loss": 0.0,
|
| 192 |
+
"step": 3804
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 12.0,
|
| 196 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 197 |
+
"eval_loss": NaN,
|
| 198 |
+
"eval_runtime": 1.5495,
|
| 199 |
+
"eval_samples_per_second": 2000.683,
|
| 200 |
+
"eval_steps_per_second": 41.95,
|
| 201 |
+
"step": 3816
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 12.959119496855346,
|
| 205 |
+
"grad_norm": NaN,
|
| 206 |
+
"learning_rate": 1.6589705395564385e-05,
|
| 207 |
+
"loss": 0.0,
|
| 208 |
+
"step": 4121
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 13.0,
|
| 212 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 213 |
+
"eval_loss": NaN,
|
| 214 |
+
"eval_runtime": 1.552,
|
| 215 |
+
"eval_samples_per_second": 1997.38,
|
| 216 |
+
"eval_steps_per_second": 41.881,
|
| 217 |
+
"step": 4134
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 13.955974842767295,
|
| 221 |
+
"grad_norm": NaN,
|
| 222 |
+
"learning_rate": 1.6327375041377028e-05,
|
| 223 |
+
"loss": 0.0,
|
| 224 |
+
"step": 4438
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 14.0,
|
| 228 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 229 |
+
"eval_loss": NaN,
|
| 230 |
+
"eval_runtime": 1.5458,
|
| 231 |
+
"eval_samples_per_second": 2005.461,
|
| 232 |
+
"eval_steps_per_second": 42.05,
|
| 233 |
+
"step": 4452
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 14.952830188679245,
|
| 237 |
+
"grad_norm": NaN,
|
| 238 |
+
"learning_rate": 1.606504468718967e-05,
|
| 239 |
+
"loss": 0.0,
|
| 240 |
+
"step": 4755
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 15.0,
|
| 244 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 245 |
+
"eval_loss": NaN,
|
| 246 |
+
"eval_runtime": 1.5467,
|
| 247 |
+
"eval_samples_per_second": 2004.26,
|
| 248 |
+
"eval_steps_per_second": 42.025,
|
| 249 |
+
"step": 4770
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 15.949685534591195,
|
| 253 |
+
"grad_norm": NaN,
|
| 254 |
+
"learning_rate": 1.5802714333002318e-05,
|
| 255 |
+
"loss": 0.0,
|
| 256 |
+
"step": 5072
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 16.0,
|
| 260 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 261 |
+
"eval_loss": NaN,
|
| 262 |
+
"eval_runtime": 1.5421,
|
| 263 |
+
"eval_samples_per_second": 2010.27,
|
| 264 |
+
"eval_steps_per_second": 42.151,
|
| 265 |
+
"step": 5088
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 16.946540880503143,
|
| 269 |
+
"grad_norm": NaN,
|
| 270 |
+
"learning_rate": 1.5540383978814965e-05,
|
| 271 |
+
"loss": 0.0,
|
| 272 |
+
"step": 5389
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 17.0,
|
| 276 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 277 |
+
"eval_loss": NaN,
|
| 278 |
+
"eval_runtime": 1.5512,
|
| 279 |
+
"eval_samples_per_second": 1998.496,
|
| 280 |
+
"eval_steps_per_second": 41.904,
|
| 281 |
+
"step": 5406
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 17.943396226415093,
|
| 285 |
+
"grad_norm": NaN,
|
| 286 |
+
"learning_rate": 1.5278053624627608e-05,
|
| 287 |
+
"loss": 0.0,
|
| 288 |
+
"step": 5706
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 18.0,
|
| 292 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 293 |
+
"eval_loss": NaN,
|
| 294 |
+
"eval_runtime": 1.5483,
|
| 295 |
+
"eval_samples_per_second": 2002.15,
|
| 296 |
+
"eval_steps_per_second": 41.981,
|
| 297 |
+
"step": 5724
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 18.940251572327043,
|
| 301 |
+
"grad_norm": NaN,
|
| 302 |
+
"learning_rate": 1.5015723270440253e-05,
|
| 303 |
+
"loss": 0.0,
|
| 304 |
+
"step": 6023
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 19.0,
|
| 308 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 309 |
+
"eval_loss": NaN,
|
| 310 |
+
"eval_runtime": 1.5501,
|
| 311 |
+
"eval_samples_per_second": 1999.815,
|
| 312 |
+
"eval_steps_per_second": 41.932,
|
| 313 |
+
"step": 6042
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 19.937106918238992,
|
| 317 |
+
"grad_norm": NaN,
|
| 318 |
+
"learning_rate": 1.4753392916252898e-05,
|
| 319 |
+
"loss": 0.0,
|
| 320 |
+
"step": 6340
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 20.0,
|
| 324 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 325 |
+
"eval_loss": NaN,
|
| 326 |
+
"eval_runtime": 1.563,
|
| 327 |
+
"eval_samples_per_second": 1983.404,
|
| 328 |
+
"eval_steps_per_second": 41.588,
|
| 329 |
+
"step": 6360
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 20.933962264150942,
|
| 333 |
+
"grad_norm": NaN,
|
| 334 |
+
"learning_rate": 1.4491062562065543e-05,
|
| 335 |
+
"loss": 0.0,
|
| 336 |
+
"step": 6657
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 21.0,
|
| 340 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 341 |
+
"eval_loss": NaN,
|
| 342 |
+
"eval_runtime": 1.5543,
|
| 343 |
+
"eval_samples_per_second": 1994.438,
|
| 344 |
+
"eval_steps_per_second": 41.819,
|
| 345 |
+
"step": 6678
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 21.930817610062892,
|
| 349 |
+
"grad_norm": NaN,
|
| 350 |
+
"learning_rate": 1.4228732207878186e-05,
|
| 351 |
+
"loss": 0.0,
|
| 352 |
+
"step": 6974
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 22.0,
|
| 356 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 357 |
+
"eval_loss": NaN,
|
| 358 |
+
"eval_runtime": 1.5533,
|
| 359 |
+
"eval_samples_per_second": 1995.714,
|
| 360 |
+
"eval_steps_per_second": 41.846,
|
| 361 |
+
"step": 6996
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 22.927672955974842,
|
| 365 |
+
"grad_norm": NaN,
|
| 366 |
+
"learning_rate": 1.3966401853690831e-05,
|
| 367 |
+
"loss": 0.0,
|
| 368 |
+
"step": 7291
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 23.0,
|
| 372 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 373 |
+
"eval_loss": NaN,
|
| 374 |
+
"eval_runtime": 1.5481,
|
| 375 |
+
"eval_samples_per_second": 2002.435,
|
| 376 |
+
"eval_steps_per_second": 41.987,
|
| 377 |
+
"step": 7314
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 23.92452830188679,
|
| 381 |
+
"grad_norm": NaN,
|
| 382 |
+
"learning_rate": 1.3704071499503478e-05,
|
| 383 |
+
"loss": 0.0,
|
| 384 |
+
"step": 7608
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 24.0,
|
| 388 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 389 |
+
"eval_loss": NaN,
|
| 390 |
+
"eval_runtime": 1.539,
|
| 391 |
+
"eval_samples_per_second": 2014.299,
|
| 392 |
+
"eval_steps_per_second": 42.235,
|
| 393 |
+
"step": 7632
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 24.92138364779874,
|
| 397 |
+
"grad_norm": NaN,
|
| 398 |
+
"learning_rate": 1.3441741145316121e-05,
|
| 399 |
+
"loss": 0.0,
|
| 400 |
+
"step": 7925
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 25.0,
|
| 404 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 405 |
+
"eval_loss": NaN,
|
| 406 |
+
"eval_runtime": 1.5516,
|
| 407 |
+
"eval_samples_per_second": 1997.904,
|
| 408 |
+
"eval_steps_per_second": 41.892,
|
| 409 |
+
"step": 7950
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 25.91823899371069,
|
| 413 |
+
"grad_norm": NaN,
|
| 414 |
+
"learning_rate": 1.3179410791128766e-05,
|
| 415 |
+
"loss": 0.0,
|
| 416 |
+
"step": 8242
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 26.0,
|
| 420 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 421 |
+
"eval_loss": NaN,
|
| 422 |
+
"eval_runtime": 1.5404,
|
| 423 |
+
"eval_samples_per_second": 2012.497,
|
| 424 |
+
"eval_steps_per_second": 42.198,
|
| 425 |
+
"step": 8268
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 26.91509433962264,
|
| 429 |
+
"grad_norm": NaN,
|
| 430 |
+
"learning_rate": 1.2917080436941411e-05,
|
| 431 |
+
"loss": 0.0,
|
| 432 |
+
"step": 8559
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 27.0,
|
| 436 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 437 |
+
"eval_loss": NaN,
|
| 438 |
+
"eval_runtime": 1.5468,
|
| 439 |
+
"eval_samples_per_second": 2004.126,
|
| 440 |
+
"eval_steps_per_second": 42.022,
|
| 441 |
+
"step": 8586
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 27.91194968553459,
|
| 445 |
+
"grad_norm": NaN,
|
| 446 |
+
"learning_rate": 1.2654750082754058e-05,
|
| 447 |
+
"loss": 0.0,
|
| 448 |
+
"step": 8876
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 28.0,
|
| 452 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 453 |
+
"eval_loss": NaN,
|
| 454 |
+
"eval_runtime": 1.5481,
|
| 455 |
+
"eval_samples_per_second": 2002.49,
|
| 456 |
+
"eval_steps_per_second": 41.988,
|
| 457 |
+
"step": 8904
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 28.90880503144654,
|
| 461 |
+
"grad_norm": NaN,
|
| 462 |
+
"learning_rate": 1.23924197285667e-05,
|
| 463 |
+
"loss": 0.0,
|
| 464 |
+
"step": 9193
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 29.0,
|
| 468 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 469 |
+
"eval_loss": NaN,
|
| 470 |
+
"eval_runtime": 1.5508,
|
| 471 |
+
"eval_samples_per_second": 1999.025,
|
| 472 |
+
"eval_steps_per_second": 41.915,
|
| 473 |
+
"step": 9222
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 29.90566037735849,
|
| 477 |
+
"grad_norm": NaN,
|
| 478 |
+
"learning_rate": 1.2130089374379346e-05,
|
| 479 |
+
"loss": 0.0,
|
| 480 |
+
"step": 9510
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 30.0,
|
| 484 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 485 |
+
"eval_loss": NaN,
|
| 486 |
+
"eval_runtime": 1.5591,
|
| 487 |
+
"eval_samples_per_second": 1988.293,
|
| 488 |
+
"eval_steps_per_second": 41.69,
|
| 489 |
+
"step": 9540
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 30.90251572327044,
|
| 493 |
+
"grad_norm": NaN,
|
| 494 |
+
"learning_rate": 1.1867759020191991e-05,
|
| 495 |
+
"loss": 0.0,
|
| 496 |
+
"step": 9827
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 31.0,
|
| 500 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 501 |
+
"eval_loss": NaN,
|
| 502 |
+
"eval_runtime": 1.5452,
|
| 503 |
+
"eval_samples_per_second": 2006.189,
|
| 504 |
+
"eval_steps_per_second": 42.065,
|
| 505 |
+
"step": 9858
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 31.89937106918239,
|
| 509 |
+
"grad_norm": NaN,
|
| 510 |
+
"learning_rate": 1.1605428666004634e-05,
|
| 511 |
+
"loss": 0.0,
|
| 512 |
+
"step": 10144
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 32.0,
|
| 516 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 517 |
+
"eval_loss": NaN,
|
| 518 |
+
"eval_runtime": 1.549,
|
| 519 |
+
"eval_samples_per_second": 2001.291,
|
| 520 |
+
"eval_steps_per_second": 41.963,
|
| 521 |
+
"step": 10176
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 32.89622641509434,
|
| 525 |
+
"grad_norm": NaN,
|
| 526 |
+
"learning_rate": 1.134309831181728e-05,
|
| 527 |
+
"loss": 0.0,
|
| 528 |
+
"step": 10461
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 33.0,
|
| 532 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 533 |
+
"eval_loss": NaN,
|
| 534 |
+
"eval_runtime": 1.5482,
|
| 535 |
+
"eval_samples_per_second": 2002.327,
|
| 536 |
+
"eval_steps_per_second": 41.984,
|
| 537 |
+
"step": 10494
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 33.893081761006286,
|
| 541 |
+
"grad_norm": NaN,
|
| 542 |
+
"learning_rate": 1.1080767957629926e-05,
|
| 543 |
+
"loss": 0.0,
|
| 544 |
+
"step": 10778
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 34.0,
|
| 548 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 549 |
+
"eval_loss": NaN,
|
| 550 |
+
"eval_runtime": 1.551,
|
| 551 |
+
"eval_samples_per_second": 1998.704,
|
| 552 |
+
"eval_steps_per_second": 41.908,
|
| 553 |
+
"step": 10812
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 34.88993710691824,
|
| 557 |
+
"grad_norm": NaN,
|
| 558 |
+
"learning_rate": 1.081843760344257e-05,
|
| 559 |
+
"loss": 0.0,
|
| 560 |
+
"step": 11095
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 35.0,
|
| 564 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 565 |
+
"eval_loss": NaN,
|
| 566 |
+
"eval_runtime": 1.5487,
|
| 567 |
+
"eval_samples_per_second": 2001.682,
|
| 568 |
+
"eval_steps_per_second": 41.971,
|
| 569 |
+
"step": 11130
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 35.886792452830186,
|
| 573 |
+
"grad_norm": NaN,
|
| 574 |
+
"learning_rate": 1.0556107249255214e-05,
|
| 575 |
+
"loss": 0.0,
|
| 576 |
+
"step": 11412
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 36.0,
|
| 580 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 581 |
+
"eval_loss": NaN,
|
| 582 |
+
"eval_runtime": 1.542,
|
| 583 |
+
"eval_samples_per_second": 2010.376,
|
| 584 |
+
"eval_steps_per_second": 42.153,
|
| 585 |
+
"step": 11448
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 36.88364779874214,
|
| 589 |
+
"grad_norm": NaN,
|
| 590 |
+
"learning_rate": 1.029377689506786e-05,
|
| 591 |
+
"loss": 0.0,
|
| 592 |
+
"step": 11729
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 37.0,
|
| 596 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 597 |
+
"eval_loss": NaN,
|
| 598 |
+
"eval_runtime": 1.5462,
|
| 599 |
+
"eval_samples_per_second": 2004.941,
|
| 600 |
+
"eval_steps_per_second": 42.039,
|
| 601 |
+
"step": 11766
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 37.880503144654085,
|
| 605 |
+
"grad_norm": NaN,
|
| 606 |
+
"learning_rate": 1.0031446540880504e-05,
|
| 607 |
+
"loss": 0.0,
|
| 608 |
+
"step": 12046
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 38.0,
|
| 612 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 613 |
+
"eval_loss": NaN,
|
| 614 |
+
"eval_runtime": 1.5538,
|
| 615 |
+
"eval_samples_per_second": 1995.139,
|
| 616 |
+
"eval_steps_per_second": 41.834,
|
| 617 |
+
"step": 12084
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 38.87735849056604,
|
| 621 |
+
"grad_norm": NaN,
|
| 622 |
+
"learning_rate": 9.769116186693148e-06,
|
| 623 |
+
"loss": 0.0,
|
| 624 |
+
"step": 12363
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 39.0,
|
| 628 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 629 |
+
"eval_loss": NaN,
|
| 630 |
+
"eval_runtime": 1.544,
|
| 631 |
+
"eval_samples_per_second": 2007.773,
|
| 632 |
+
"eval_steps_per_second": 42.098,
|
| 633 |
+
"step": 12402
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 39.874213836477985,
|
| 637 |
+
"grad_norm": NaN,
|
| 638 |
+
"learning_rate": 9.506785832505794e-06,
|
| 639 |
+
"loss": 0.0,
|
| 640 |
+
"step": 12680
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 40.0,
|
| 644 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 645 |
+
"eval_loss": NaN,
|
| 646 |
+
"eval_runtime": 1.5472,
|
| 647 |
+
"eval_samples_per_second": 2003.631,
|
| 648 |
+
"eval_steps_per_second": 42.012,
|
| 649 |
+
"step": 12720
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 40.87106918238994,
|
| 653 |
+
"grad_norm": NaN,
|
| 654 |
+
"learning_rate": 9.244455478318438e-06,
|
| 655 |
+
"loss": 0.0,
|
| 656 |
+
"step": 12997
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 41.0,
|
| 660 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 661 |
+
"eval_loss": NaN,
|
| 662 |
+
"eval_runtime": 1.5472,
|
| 663 |
+
"eval_samples_per_second": 2003.584,
|
| 664 |
+
"eval_steps_per_second": 42.011,
|
| 665 |
+
"step": 13038
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 41.867924528301884,
|
| 669 |
+
"grad_norm": NaN,
|
| 670 |
+
"learning_rate": 8.982125124131084e-06,
|
| 671 |
+
"loss": 0.0,
|
| 672 |
+
"step": 13314
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 42.0,
|
| 676 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 677 |
+
"eval_loss": NaN,
|
| 678 |
+
"eval_runtime": 1.5517,
|
| 679 |
+
"eval_samples_per_second": 1997.766,
|
| 680 |
+
"eval_steps_per_second": 41.889,
|
| 681 |
+
"step": 13356
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 42.86477987421384,
|
| 685 |
+
"grad_norm": NaN,
|
| 686 |
+
"learning_rate": 8.719794769943728e-06,
|
| 687 |
+
"loss": 0.0,
|
| 688 |
+
"step": 13631
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 43.0,
|
| 692 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 693 |
+
"eval_loss": NaN,
|
| 694 |
+
"eval_runtime": 1.5535,
|
| 695 |
+
"eval_samples_per_second": 1995.529,
|
| 696 |
+
"eval_steps_per_second": 41.842,
|
| 697 |
+
"step": 13674
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 43.861635220125784,
|
| 701 |
+
"grad_norm": NaN,
|
| 702 |
+
"learning_rate": 8.457464415756373e-06,
|
| 703 |
+
"loss": 0.0,
|
| 704 |
+
"step": 13948
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 44.0,
|
| 708 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 709 |
+
"eval_loss": NaN,
|
| 710 |
+
"eval_runtime": 1.5578,
|
| 711 |
+
"eval_samples_per_second": 1989.984,
|
| 712 |
+
"eval_steps_per_second": 41.725,
|
| 713 |
+
"step": 13992
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 44.85849056603774,
|
| 717 |
+
"grad_norm": NaN,
|
| 718 |
+
"learning_rate": 8.195134061569018e-06,
|
| 719 |
+
"loss": 0.0,
|
| 720 |
+
"step": 14265
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 45.0,
|
| 724 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 725 |
+
"eval_loss": NaN,
|
| 726 |
+
"eval_runtime": 1.548,
|
| 727 |
+
"eval_samples_per_second": 2002.591,
|
| 728 |
+
"eval_steps_per_second": 41.99,
|
| 729 |
+
"step": 14310
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 45.855345911949684,
|
| 733 |
+
"grad_norm": NaN,
|
| 734 |
+
"learning_rate": 7.932803707381663e-06,
|
| 735 |
+
"loss": 0.0,
|
| 736 |
+
"step": 14582
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 46.0,
|
| 740 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 741 |
+
"eval_loss": NaN,
|
| 742 |
+
"eval_runtime": 1.5507,
|
| 743 |
+
"eval_samples_per_second": 1999.086,
|
| 744 |
+
"eval_steps_per_second": 41.916,
|
| 745 |
+
"step": 14628
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 46.85220125786164,
|
| 749 |
+
"grad_norm": NaN,
|
| 750 |
+
"learning_rate": 7.670473353194308e-06,
|
| 751 |
+
"loss": 0.0,
|
| 752 |
+
"step": 14899
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 47.0,
|
| 756 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 757 |
+
"eval_loss": NaN,
|
| 758 |
+
"eval_runtime": 1.5507,
|
| 759 |
+
"eval_samples_per_second": 1999.054,
|
| 760 |
+
"eval_steps_per_second": 41.916,
|
| 761 |
+
"step": 14946
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 47.84905660377358,
|
| 765 |
+
"grad_norm": NaN,
|
| 766 |
+
"learning_rate": 7.408142999006952e-06,
|
| 767 |
+
"loss": 0.0,
|
| 768 |
+
"step": 15216
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 48.0,
|
| 772 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 773 |
+
"eval_loss": NaN,
|
| 774 |
+
"eval_runtime": 1.5435,
|
| 775 |
+
"eval_samples_per_second": 2008.361,
|
| 776 |
+
"eval_steps_per_second": 42.111,
|
| 777 |
+
"step": 15264
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 48.84591194968554,
|
| 781 |
+
"grad_norm": NaN,
|
| 782 |
+
"learning_rate": 7.145812644819597e-06,
|
| 783 |
+
"loss": 0.0,
|
| 784 |
+
"step": 15533
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 49.0,
|
| 788 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 789 |
+
"eval_loss": NaN,
|
| 790 |
+
"eval_runtime": 1.5454,
|
| 791 |
+
"eval_samples_per_second": 2005.987,
|
| 792 |
+
"eval_steps_per_second": 42.061,
|
| 793 |
+
"step": 15582
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 49.84276729559748,
|
| 797 |
+
"grad_norm": NaN,
|
| 798 |
+
"learning_rate": 6.883482290632242e-06,
|
| 799 |
+
"loss": 0.0,
|
| 800 |
+
"step": 15850
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 50.0,
|
| 804 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 805 |
+
"eval_loss": NaN,
|
| 806 |
+
"eval_runtime": 1.555,
|
| 807 |
+
"eval_samples_per_second": 1993.568,
|
| 808 |
+
"eval_steps_per_second": 41.801,
|
| 809 |
+
"step": 15900
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 50.839622641509436,
|
| 813 |
+
"grad_norm": NaN,
|
| 814 |
+
"learning_rate": 6.621151936444886e-06,
|
| 815 |
+
"loss": 0.0,
|
| 816 |
+
"step": 16167
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 51.0,
|
| 820 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 821 |
+
"eval_loss": NaN,
|
| 822 |
+
"eval_runtime": 1.5501,
|
| 823 |
+
"eval_samples_per_second": 1999.83,
|
| 824 |
+
"eval_steps_per_second": 41.932,
|
| 825 |
+
"step": 16218
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 51.83647798742138,
|
| 829 |
+
"grad_norm": NaN,
|
| 830 |
+
"learning_rate": 6.358821582257532e-06,
|
| 831 |
+
"loss": 0.0,
|
| 832 |
+
"step": 16484
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 52.0,
|
| 836 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 837 |
+
"eval_loss": NaN,
|
| 838 |
+
"eval_runtime": 1.5506,
|
| 839 |
+
"eval_samples_per_second": 1999.181,
|
| 840 |
+
"eval_steps_per_second": 41.918,
|
| 841 |
+
"step": 16536
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 52.833333333333336,
|
| 845 |
+
"grad_norm": NaN,
|
| 846 |
+
"learning_rate": 6.096491228070176e-06,
|
| 847 |
+
"loss": 0.0,
|
| 848 |
+
"step": 16801
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 53.0,
|
| 852 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 853 |
+
"eval_loss": NaN,
|
| 854 |
+
"eval_runtime": 1.5631,
|
| 855 |
+
"eval_samples_per_second": 1983.281,
|
| 856 |
+
"eval_steps_per_second": 41.585,
|
| 857 |
+
"step": 16854
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 53.83018867924528,
|
| 861 |
+
"grad_norm": NaN,
|
| 862 |
+
"learning_rate": 5.834160873882821e-06,
|
| 863 |
+
"loss": 0.0,
|
| 864 |
+
"step": 17118
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 54.0,
|
| 868 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 869 |
+
"eval_loss": NaN,
|
| 870 |
+
"eval_runtime": 1.5457,
|
| 871 |
+
"eval_samples_per_second": 2005.561,
|
| 872 |
+
"eval_steps_per_second": 42.052,
|
| 873 |
+
"step": 17172
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 54.827044025157235,
|
| 877 |
+
"grad_norm": NaN,
|
| 878 |
+
"learning_rate": 5.571830519695466e-06,
|
| 879 |
+
"loss": 0.0,
|
| 880 |
+
"step": 17435
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 55.0,
|
| 884 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 885 |
+
"eval_loss": NaN,
|
| 886 |
+
"eval_runtime": 1.5494,
|
| 887 |
+
"eval_samples_per_second": 2000.789,
|
| 888 |
+
"eval_steps_per_second": 41.952,
|
| 889 |
+
"step": 17490
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 55.82389937106918,
|
| 893 |
+
"grad_norm": NaN,
|
| 894 |
+
"learning_rate": 5.30950016550811e-06,
|
| 895 |
+
"loss": 0.0,
|
| 896 |
+
"step": 17752
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 56.0,
|
| 900 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 901 |
+
"eval_loss": NaN,
|
| 902 |
+
"eval_runtime": 1.5572,
|
| 903 |
+
"eval_samples_per_second": 1990.807,
|
| 904 |
+
"eval_steps_per_second": 41.743,
|
| 905 |
+
"step": 17808
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 56.820754716981135,
|
| 909 |
+
"grad_norm": NaN,
|
| 910 |
+
"learning_rate": 5.047169811320756e-06,
|
| 911 |
+
"loss": 0.0,
|
| 912 |
+
"step": 18069
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 57.0,
|
| 916 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 917 |
+
"eval_loss": NaN,
|
| 918 |
+
"eval_runtime": 1.5473,
|
| 919 |
+
"eval_samples_per_second": 2003.467,
|
| 920 |
+
"eval_steps_per_second": 42.008,
|
| 921 |
+
"step": 18126
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 57.81761006289308,
|
| 925 |
+
"grad_norm": NaN,
|
| 926 |
+
"learning_rate": 4.7848394571334e-06,
|
| 927 |
+
"loss": 0.0,
|
| 928 |
+
"step": 18386
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 58.0,
|
| 932 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 933 |
+
"eval_loss": NaN,
|
| 934 |
+
"eval_runtime": 1.5472,
|
| 935 |
+
"eval_samples_per_second": 2003.565,
|
| 936 |
+
"eval_steps_per_second": 42.01,
|
| 937 |
+
"step": 18444
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 58.814465408805034,
|
| 941 |
+
"grad_norm": NaN,
|
| 942 |
+
"learning_rate": 4.522509102946045e-06,
|
| 943 |
+
"loss": 0.0,
|
| 944 |
+
"step": 18703
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 59.0,
|
| 948 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 949 |
+
"eval_loss": NaN,
|
| 950 |
+
"eval_runtime": 1.5591,
|
| 951 |
+
"eval_samples_per_second": 1988.309,
|
| 952 |
+
"eval_steps_per_second": 41.69,
|
| 953 |
+
"step": 18762
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 59.81132075471698,
|
| 957 |
+
"grad_norm": NaN,
|
| 958 |
+
"learning_rate": 4.260178748758689e-06,
|
| 959 |
+
"loss": 0.0,
|
| 960 |
+
"step": 19020
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 60.0,
|
| 964 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 965 |
+
"eval_loss": NaN,
|
| 966 |
+
"eval_runtime": 1.5461,
|
| 967 |
+
"eval_samples_per_second": 2005.092,
|
| 968 |
+
"eval_steps_per_second": 42.042,
|
| 969 |
+
"step": 19080
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 60.808176100628934,
|
| 973 |
+
"grad_norm": NaN,
|
| 974 |
+
"learning_rate": 3.997848394571334e-06,
|
| 975 |
+
"loss": 0.0,
|
| 976 |
+
"step": 19337
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 61.0,
|
| 980 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 981 |
+
"eval_loss": NaN,
|
| 982 |
+
"eval_runtime": 1.5464,
|
| 983 |
+
"eval_samples_per_second": 2004.597,
|
| 984 |
+
"eval_steps_per_second": 42.032,
|
| 985 |
+
"step": 19398
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 61.80503144654088,
|
| 989 |
+
"grad_norm": NaN,
|
| 990 |
+
"learning_rate": 3.735518040383979e-06,
|
| 991 |
+
"loss": 0.0,
|
| 992 |
+
"step": 19654
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 62.0,
|
| 996 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 997 |
+
"eval_loss": NaN,
|
| 998 |
+
"eval_runtime": 1.5593,
|
| 999 |
+
"eval_samples_per_second": 1988.116,
|
| 1000 |
+
"eval_steps_per_second": 41.686,
|
| 1001 |
+
"step": 19716
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 62.801886792452834,
|
| 1005 |
+
"grad_norm": NaN,
|
| 1006 |
+
"learning_rate": 3.473187686196624e-06,
|
| 1007 |
+
"loss": 0.0,
|
| 1008 |
+
"step": 19971
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 63.0,
|
| 1012 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1013 |
+
"eval_loss": NaN,
|
| 1014 |
+
"eval_runtime": 1.5574,
|
| 1015 |
+
"eval_samples_per_second": 1990.46,
|
| 1016 |
+
"eval_steps_per_second": 41.735,
|
| 1017 |
+
"step": 20034
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 63.79874213836478,
|
| 1021 |
+
"grad_norm": NaN,
|
| 1022 |
+
"learning_rate": 3.210857332009269e-06,
|
| 1023 |
+
"loss": 0.0,
|
| 1024 |
+
"step": 20288
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 64.0,
|
| 1028 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1029 |
+
"eval_loss": NaN,
|
| 1030 |
+
"eval_runtime": 1.5504,
|
| 1031 |
+
"eval_samples_per_second": 1999.508,
|
| 1032 |
+
"eval_steps_per_second": 41.925,
|
| 1033 |
+
"step": 20352
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 64.79559748427673,
|
| 1037 |
+
"grad_norm": NaN,
|
| 1038 |
+
"learning_rate": 2.9485269778219135e-06,
|
| 1039 |
+
"loss": 0.0,
|
| 1040 |
+
"step": 20605
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 65.0,
|
| 1044 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1045 |
+
"eval_loss": NaN,
|
| 1046 |
+
"eval_runtime": 1.5516,
|
| 1047 |
+
"eval_samples_per_second": 1997.884,
|
| 1048 |
+
"eval_steps_per_second": 41.891,
|
| 1049 |
+
"step": 20670
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 65.79245283018868,
|
| 1053 |
+
"grad_norm": NaN,
|
| 1054 |
+
"learning_rate": 2.686196623634558e-06,
|
| 1055 |
+
"loss": 0.0,
|
| 1056 |
+
"step": 20922
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 66.0,
|
| 1060 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1061 |
+
"eval_loss": NaN,
|
| 1062 |
+
"eval_runtime": 1.5621,
|
| 1063 |
+
"eval_samples_per_second": 1984.56,
|
| 1064 |
+
"eval_steps_per_second": 41.612,
|
| 1065 |
+
"step": 20988
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 66.78930817610063,
|
| 1069 |
+
"grad_norm": NaN,
|
| 1070 |
+
"learning_rate": 2.423866269447203e-06,
|
| 1071 |
+
"loss": 0.0,
|
| 1072 |
+
"step": 21239
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 67.0,
|
| 1076 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1077 |
+
"eval_loss": NaN,
|
| 1078 |
+
"eval_runtime": 1.5469,
|
| 1079 |
+
"eval_samples_per_second": 2004.011,
|
| 1080 |
+
"eval_steps_per_second": 42.02,
|
| 1081 |
+
"step": 21306
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 67.78616352201257,
|
| 1085 |
+
"grad_norm": NaN,
|
| 1086 |
+
"learning_rate": 2.1615359152598476e-06,
|
| 1087 |
+
"loss": 0.0,
|
| 1088 |
+
"step": 21556
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 68.0,
|
| 1092 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1093 |
+
"eval_loss": NaN,
|
| 1094 |
+
"eval_runtime": 1.5535,
|
| 1095 |
+
"eval_samples_per_second": 1995.435,
|
| 1096 |
+
"eval_steps_per_second": 41.84,
|
| 1097 |
+
"step": 21624
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 68.78301886792453,
|
| 1101 |
+
"grad_norm": NaN,
|
| 1102 |
+
"learning_rate": 1.8992055610724928e-06,
|
| 1103 |
+
"loss": 0.0,
|
| 1104 |
+
"step": 21873
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 69.0,
|
| 1108 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1109 |
+
"eval_loss": NaN,
|
| 1110 |
+
"eval_runtime": 1.5443,
|
| 1111 |
+
"eval_samples_per_second": 2007.325,
|
| 1112 |
+
"eval_steps_per_second": 42.089,
|
| 1113 |
+
"step": 21942
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 69.77987421383648,
|
| 1117 |
+
"grad_norm": NaN,
|
| 1118 |
+
"learning_rate": 1.6368752068851376e-06,
|
| 1119 |
+
"loss": 0.0,
|
| 1120 |
+
"step": 22190
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 70.0,
|
| 1124 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1125 |
+
"eval_loss": NaN,
|
| 1126 |
+
"eval_runtime": 1.5594,
|
| 1127 |
+
"eval_samples_per_second": 1987.916,
|
| 1128 |
+
"eval_steps_per_second": 41.682,
|
| 1129 |
+
"step": 22260
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 70.77672955974843,
|
| 1133 |
+
"grad_norm": NaN,
|
| 1134 |
+
"learning_rate": 1.3745448526977824e-06,
|
| 1135 |
+
"loss": 0.0,
|
| 1136 |
+
"step": 22507
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 71.0,
|
| 1140 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1141 |
+
"eval_loss": NaN,
|
| 1142 |
+
"eval_runtime": 1.5504,
|
| 1143 |
+
"eval_samples_per_second": 1999.547,
|
| 1144 |
+
"eval_steps_per_second": 41.926,
|
| 1145 |
+
"step": 22578
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 71.77358490566037,
|
| 1149 |
+
"grad_norm": NaN,
|
| 1150 |
+
"learning_rate": 1.1122144985104271e-06,
|
| 1151 |
+
"loss": 0.0,
|
| 1152 |
+
"step": 22824
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 72.0,
|
| 1156 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1157 |
+
"eval_loss": NaN,
|
| 1158 |
+
"eval_runtime": 1.5487,
|
| 1159 |
+
"eval_samples_per_second": 2001.694,
|
| 1160 |
+
"eval_steps_per_second": 41.971,
|
| 1161 |
+
"step": 22896
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 72.77044025157232,
|
| 1165 |
+
"grad_norm": NaN,
|
| 1166 |
+
"learning_rate": 8.498841443230719e-07,
|
| 1167 |
+
"loss": 0.0,
|
| 1168 |
+
"step": 23141
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 73.0,
|
| 1172 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1173 |
+
"eval_loss": NaN,
|
| 1174 |
+
"eval_runtime": 1.5548,
|
| 1175 |
+
"eval_samples_per_second": 1993.874,
|
| 1176 |
+
"eval_steps_per_second": 41.807,
|
| 1177 |
+
"step": 23214
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 73.76729559748428,
|
| 1181 |
+
"grad_norm": NaN,
|
| 1182 |
+
"learning_rate": 5.875537901357168e-07,
|
| 1183 |
+
"loss": 0.0,
|
| 1184 |
+
"step": 23458
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 74.0,
|
| 1188 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1189 |
+
"eval_loss": NaN,
|
| 1190 |
+
"eval_runtime": 1.5427,
|
| 1191 |
+
"eval_samples_per_second": 2009.409,
|
| 1192 |
+
"eval_steps_per_second": 42.133,
|
| 1193 |
+
"step": 23532
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 74.76415094339623,
|
| 1197 |
+
"grad_norm": NaN,
|
| 1198 |
+
"learning_rate": 3.252234359483615e-07,
|
| 1199 |
+
"loss": 0.0,
|
| 1200 |
+
"step": 23775
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 75.0,
|
| 1204 |
+
"eval_accuracy": 0.0064516129032258064,
|
| 1205 |
+
"eval_loss": NaN,
|
| 1206 |
+
"eval_runtime": 1.5505,
|
| 1207 |
+
"eval_samples_per_second": 1999.347,
|
| 1208 |
+
"eval_steps_per_second": 41.922,
|
| 1209 |
+
"step": 23850
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 75.76100628930817,
|
| 1213 |
+
"grad_norm": NaN,
|
| 1214 |
+
"learning_rate": 6.289308176100629e-08,
|
| 1215 |
+
"loss": 0.0,
|
| 1216 |
+
"step": 24092
|
| 1217 |
+
}
|
| 1218 |
+
],
|
| 1219 |
+
"logging_steps": 317,
|
| 1220 |
+
"max_steps": 24168,
|
| 1221 |
+
"num_input_tokens_seen": 0,
|
| 1222 |
+
"num_train_epochs": 76,
|
| 1223 |
+
"save_steps": 1000000000.0,
|
| 1224 |
+
"stateful_callbacks": {
|
| 1225 |
+
"TrainerControl": {
|
| 1226 |
+
"args": {
|
| 1227 |
+
"should_epoch_stop": false,
|
| 1228 |
+
"should_evaluate": false,
|
| 1229 |
+
"should_log": false,
|
| 1230 |
+
"should_save": true,
|
| 1231 |
+
"should_training_stop": true
|
| 1232 |
+
},
|
| 1233 |
+
"attributes": {}
|
| 1234 |
+
}
|
| 1235 |
+
},
|
| 1236 |
+
"total_flos": 6263008648762236.0,
|
| 1237 |
+
"train_batch_size": 48,
|
| 1238 |
+
"trial_name": null,
|
| 1239 |
+
"trial_params": {
|
| 1240 |
+
"alpha": 0.12159569042558904,
|
| 1241 |
+
"num_train_epochs": 76,
|
| 1242 |
+
"temperature": 0
|
| 1243 |
+
}
|
| 1244 |
+
}
|
run-1/checkpoint-24168/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e3a41cece78d4d2e0dc66507c655f7b6250bdfa5a0cd8bd67ff5a19d5e73b5e9
|
| 3 |
+
size 5304
|
run-1/checkpoint-24168/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
run-10/checkpoint-20670/config.json
ADDED
|
@@ -0,0 +1,331 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilbert-base-uncased",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"hidden_dim": 3072,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "restaurant_reviews",
|
| 13 |
+
"1": "nutrition_info",
|
| 14 |
+
"2": "account_blocked",
|
| 15 |
+
"3": "oil_change_how",
|
| 16 |
+
"4": "time",
|
| 17 |
+
"5": "weather",
|
| 18 |
+
"6": "redeem_rewards",
|
| 19 |
+
"7": "interest_rate",
|
| 20 |
+
"8": "gas_type",
|
| 21 |
+
"9": "accept_reservations",
|
| 22 |
+
"10": "smart_home",
|
| 23 |
+
"11": "user_name",
|
| 24 |
+
"12": "report_lost_card",
|
| 25 |
+
"13": "repeat",
|
| 26 |
+
"14": "whisper_mode",
|
| 27 |
+
"15": "what_are_your_hobbies",
|
| 28 |
+
"16": "order",
|
| 29 |
+
"17": "jump_start",
|
| 30 |
+
"18": "schedule_meeting",
|
| 31 |
+
"19": "meeting_schedule",
|
| 32 |
+
"20": "freeze_account",
|
| 33 |
+
"21": "what_song",
|
| 34 |
+
"22": "meaning_of_life",
|
| 35 |
+
"23": "restaurant_reservation",
|
| 36 |
+
"24": "traffic",
|
| 37 |
+
"25": "make_call",
|
| 38 |
+
"26": "text",
|
| 39 |
+
"27": "bill_balance",
|
| 40 |
+
"28": "improve_credit_score",
|
| 41 |
+
"29": "change_language",
|
| 42 |
+
"30": "no",
|
| 43 |
+
"31": "measurement_conversion",
|
| 44 |
+
"32": "timer",
|
| 45 |
+
"33": "flip_coin",
|
| 46 |
+
"34": "do_you_have_pets",
|
| 47 |
+
"35": "balance",
|
| 48 |
+
"36": "tell_joke",
|
| 49 |
+
"37": "last_maintenance",
|
| 50 |
+
"38": "exchange_rate",
|
| 51 |
+
"39": "uber",
|
| 52 |
+
"40": "car_rental",
|
| 53 |
+
"41": "credit_limit",
|
| 54 |
+
"42": "oos",
|
| 55 |
+
"43": "shopping_list",
|
| 56 |
+
"44": "expiration_date",
|
| 57 |
+
"45": "routing",
|
| 58 |
+
"46": "meal_suggestion",
|
| 59 |
+
"47": "tire_change",
|
| 60 |
+
"48": "todo_list",
|
| 61 |
+
"49": "card_declined",
|
| 62 |
+
"50": "rewards_balance",
|
| 63 |
+
"51": "change_accent",
|
| 64 |
+
"52": "vaccines",
|
| 65 |
+
"53": "reminder_update",
|
| 66 |
+
"54": "food_last",
|
| 67 |
+
"55": "change_ai_name",
|
| 68 |
+
"56": "bill_due",
|
| 69 |
+
"57": "who_do_you_work_for",
|
| 70 |
+
"58": "share_location",
|
| 71 |
+
"59": "international_visa",
|
| 72 |
+
"60": "calendar",
|
| 73 |
+
"61": "translate",
|
| 74 |
+
"62": "carry_on",
|
| 75 |
+
"63": "book_flight",
|
| 76 |
+
"64": "insurance_change",
|
| 77 |
+
"65": "todo_list_update",
|
| 78 |
+
"66": "timezone",
|
| 79 |
+
"67": "cancel_reservation",
|
| 80 |
+
"68": "transactions",
|
| 81 |
+
"69": "credit_score",
|
| 82 |
+
"70": "report_fraud",
|
| 83 |
+
"71": "spending_history",
|
| 84 |
+
"72": "directions",
|
| 85 |
+
"73": "spelling",
|
| 86 |
+
"74": "insurance",
|
| 87 |
+
"75": "what_is_your_name",
|
| 88 |
+
"76": "reminder",
|
| 89 |
+
"77": "where_are_you_from",
|
| 90 |
+
"78": "distance",
|
| 91 |
+
"79": "payday",
|
| 92 |
+
"80": "flight_status",
|
| 93 |
+
"81": "find_phone",
|
| 94 |
+
"82": "greeting",
|
| 95 |
+
"83": "alarm",
|
| 96 |
+
"84": "order_status",
|
| 97 |
+
"85": "confirm_reservation",
|
| 98 |
+
"86": "cook_time",
|
| 99 |
+
"87": "damaged_card",
|
| 100 |
+
"88": "reset_settings",
|
| 101 |
+
"89": "pin_change",
|
| 102 |
+
"90": "replacement_card_duration",
|
| 103 |
+
"91": "new_card",
|
| 104 |
+
"92": "roll_dice",
|
| 105 |
+
"93": "income",
|
| 106 |
+
"94": "taxes",
|
| 107 |
+
"95": "date",
|
| 108 |
+
"96": "who_made_you",
|
| 109 |
+
"97": "pto_request",
|
| 110 |
+
"98": "tire_pressure",
|
| 111 |
+
"99": "how_old_are_you",
|
| 112 |
+
"100": "rollover_401k",
|
| 113 |
+
"101": "pto_request_status",
|
| 114 |
+
"102": "how_busy",
|
| 115 |
+
"103": "application_status",
|
| 116 |
+
"104": "recipe",
|
| 117 |
+
"105": "calendar_update",
|
| 118 |
+
"106": "play_music",
|
| 119 |
+
"107": "yes",
|
| 120 |
+
"108": "direct_deposit",
|
| 121 |
+
"109": "credit_limit_change",
|
| 122 |
+
"110": "gas",
|
| 123 |
+
"111": "pay_bill",
|
| 124 |
+
"112": "ingredients_list",
|
| 125 |
+
"113": "lost_luggage",
|
| 126 |
+
"114": "goodbye",
|
| 127 |
+
"115": "what_can_i_ask_you",
|
| 128 |
+
"116": "book_hotel",
|
| 129 |
+
"117": "are_you_a_bot",
|
| 130 |
+
"118": "next_song",
|
| 131 |
+
"119": "change_speed",
|
| 132 |
+
"120": "plug_type",
|
| 133 |
+
"121": "maybe",
|
| 134 |
+
"122": "w2",
|
| 135 |
+
"123": "oil_change_when",
|
| 136 |
+
"124": "thank_you",
|
| 137 |
+
"125": "shopping_list_update",
|
| 138 |
+
"126": "pto_balance",
|
| 139 |
+
"127": "order_checks",
|
| 140 |
+
"128": "travel_alert",
|
| 141 |
+
"129": "fun_fact",
|
| 142 |
+
"130": "sync_device",
|
| 143 |
+
"131": "schedule_maintenance",
|
| 144 |
+
"132": "apr",
|
| 145 |
+
"133": "transfer",
|
| 146 |
+
"134": "ingredient_substitution",
|
| 147 |
+
"135": "calories",
|
| 148 |
+
"136": "current_location",
|
| 149 |
+
"137": "international_fees",
|
| 150 |
+
"138": "calculator",
|
| 151 |
+
"139": "definition",
|
| 152 |
+
"140": "next_holiday",
|
| 153 |
+
"141": "update_playlist",
|
| 154 |
+
"142": "mpg",
|
| 155 |
+
"143": "min_payment",
|
| 156 |
+
"144": "change_user_name",
|
| 157 |
+
"145": "restaurant_suggestion",
|
| 158 |
+
"146": "travel_notification",
|
| 159 |
+
"147": "cancel",
|
| 160 |
+
"148": "pto_used",
|
| 161 |
+
"149": "travel_suggestion",
|
| 162 |
+
"150": "change_volume"
|
| 163 |
+
},
|
| 164 |
+
"initializer_range": 0.02,
|
| 165 |
+
"label2id": {
|
| 166 |
+
"accept_reservations": 9,
|
| 167 |
+
"account_blocked": 2,
|
| 168 |
+
"alarm": 83,
|
| 169 |
+
"application_status": 103,
|
| 170 |
+
"apr": 132,
|
| 171 |
+
"are_you_a_bot": 117,
|
| 172 |
+
"balance": 35,
|
| 173 |
+
"bill_balance": 27,
|
| 174 |
+
"bill_due": 56,
|
| 175 |
+
"book_flight": 63,
|
| 176 |
+
"book_hotel": 116,
|
| 177 |
+
"calculator": 138,
|
| 178 |
+
"calendar": 60,
|
| 179 |
+
"calendar_update": 105,
|
| 180 |
+
"calories": 135,
|
| 181 |
+
"cancel": 147,
|
| 182 |
+
"cancel_reservation": 67,
|
| 183 |
+
"car_rental": 40,
|
| 184 |
+
"card_declined": 49,
|
| 185 |
+
"carry_on": 62,
|
| 186 |
+
"change_accent": 51,
|
| 187 |
+
"change_ai_name": 55,
|
| 188 |
+
"change_language": 29,
|
| 189 |
+
"change_speed": 119,
|
| 190 |
+
"change_user_name": 144,
|
| 191 |
+
"change_volume": 150,
|
| 192 |
+
"confirm_reservation": 85,
|
| 193 |
+
"cook_time": 86,
|
| 194 |
+
"credit_limit": 41,
|
| 195 |
+
"credit_limit_change": 109,
|
| 196 |
+
"credit_score": 69,
|
| 197 |
+
"current_location": 136,
|
| 198 |
+
"damaged_card": 87,
|
| 199 |
+
"date": 95,
|
| 200 |
+
"definition": 139,
|
| 201 |
+
"direct_deposit": 108,
|
| 202 |
+
"directions": 72,
|
| 203 |
+
"distance": 78,
|
| 204 |
+
"do_you_have_pets": 34,
|
| 205 |
+
"exchange_rate": 38,
|
| 206 |
+
"expiration_date": 44,
|
| 207 |
+
"find_phone": 81,
|
| 208 |
+
"flight_status": 80,
|
| 209 |
+
"flip_coin": 33,
|
| 210 |
+
"food_last": 54,
|
| 211 |
+
"freeze_account": 20,
|
| 212 |
+
"fun_fact": 129,
|
| 213 |
+
"gas": 110,
|
| 214 |
+
"gas_type": 8,
|
| 215 |
+
"goodbye": 114,
|
| 216 |
+
"greeting": 82,
|
| 217 |
+
"how_busy": 102,
|
| 218 |
+
"how_old_are_you": 99,
|
| 219 |
+
"improve_credit_score": 28,
|
| 220 |
+
"income": 93,
|
| 221 |
+
"ingredient_substitution": 134,
|
| 222 |
+
"ingredients_list": 112,
|
| 223 |
+
"insurance": 74,
|
| 224 |
+
"insurance_change": 64,
|
| 225 |
+
"interest_rate": 7,
|
| 226 |
+
"international_fees": 137,
|
| 227 |
+
"international_visa": 59,
|
| 228 |
+
"jump_start": 17,
|
| 229 |
+
"last_maintenance": 37,
|
| 230 |
+
"lost_luggage": 113,
|
| 231 |
+
"make_call": 25,
|
| 232 |
+
"maybe": 121,
|
| 233 |
+
"meal_suggestion": 46,
|
| 234 |
+
"meaning_of_life": 22,
|
| 235 |
+
"measurement_conversion": 31,
|
| 236 |
+
"meeting_schedule": 19,
|
| 237 |
+
"min_payment": 143,
|
| 238 |
+
"mpg": 142,
|
| 239 |
+
"new_card": 91,
|
| 240 |
+
"next_holiday": 140,
|
| 241 |
+
"next_song": 118,
|
| 242 |
+
"no": 30,
|
| 243 |
+
"nutrition_info": 1,
|
| 244 |
+
"oil_change_how": 3,
|
| 245 |
+
"oil_change_when": 123,
|
| 246 |
+
"oos": 42,
|
| 247 |
+
"order": 16,
|
| 248 |
+
"order_checks": 127,
|
| 249 |
+
"order_status": 84,
|
| 250 |
+
"pay_bill": 111,
|
| 251 |
+
"payday": 79,
|
| 252 |
+
"pin_change": 89,
|
| 253 |
+
"play_music": 106,
|
| 254 |
+
"plug_type": 120,
|
| 255 |
+
"pto_balance": 126,
|
| 256 |
+
"pto_request": 97,
|
| 257 |
+
"pto_request_status": 101,
|
| 258 |
+
"pto_used": 148,
|
| 259 |
+
"recipe": 104,
|
| 260 |
+
"redeem_rewards": 6,
|
| 261 |
+
"reminder": 76,
|
| 262 |
+
"reminder_update": 53,
|
| 263 |
+
"repeat": 13,
|
| 264 |
+
"replacement_card_duration": 90,
|
| 265 |
+
"report_fraud": 70,
|
| 266 |
+
"report_lost_card": 12,
|
| 267 |
+
"reset_settings": 88,
|
| 268 |
+
"restaurant_reservation": 23,
|
| 269 |
+
"restaurant_reviews": 0,
|
| 270 |
+
"restaurant_suggestion": 145,
|
| 271 |
+
"rewards_balance": 50,
|
| 272 |
+
"roll_dice": 92,
|
| 273 |
+
"rollover_401k": 100,
|
| 274 |
+
"routing": 45,
|
| 275 |
+
"schedule_maintenance": 131,
|
| 276 |
+
"schedule_meeting": 18,
|
| 277 |
+
"share_location": 58,
|
| 278 |
+
"shopping_list": 43,
|
| 279 |
+
"shopping_list_update": 125,
|
| 280 |
+
"smart_home": 10,
|
| 281 |
+
"spelling": 73,
|
| 282 |
+
"spending_history": 71,
|
| 283 |
+
"sync_device": 130,
|
| 284 |
+
"taxes": 94,
|
| 285 |
+
"tell_joke": 36,
|
| 286 |
+
"text": 26,
|
| 287 |
+
"thank_you": 124,
|
| 288 |
+
"time": 4,
|
| 289 |
+
"timer": 32,
|
| 290 |
+
"timezone": 66,
|
| 291 |
+
"tire_change": 47,
|
| 292 |
+
"tire_pressure": 98,
|
| 293 |
+
"todo_list": 48,
|
| 294 |
+
"todo_list_update": 65,
|
| 295 |
+
"traffic": 24,
|
| 296 |
+
"transactions": 68,
|
| 297 |
+
"transfer": 133,
|
| 298 |
+
"translate": 61,
|
| 299 |
+
"travel_alert": 128,
|
| 300 |
+
"travel_notification": 146,
|
| 301 |
+
"travel_suggestion": 149,
|
| 302 |
+
"uber": 39,
|
| 303 |
+
"update_playlist": 141,
|
| 304 |
+
"user_name": 11,
|
| 305 |
+
"vaccines": 52,
|
| 306 |
+
"w2": 122,
|
| 307 |
+
"weather": 5,
|
| 308 |
+
"what_are_your_hobbies": 15,
|
| 309 |
+
"what_can_i_ask_you": 115,
|
| 310 |
+
"what_is_your_name": 75,
|
| 311 |
+
"what_song": 21,
|
| 312 |
+
"where_are_you_from": 77,
|
| 313 |
+
"whisper_mode": 14,
|
| 314 |
+
"who_do_you_work_for": 57,
|
| 315 |
+
"who_made_you": 96,
|
| 316 |
+
"yes": 107
|
| 317 |
+
},
|
| 318 |
+
"max_position_embeddings": 512,
|
| 319 |
+
"model_type": "distilbert",
|
| 320 |
+
"n_heads": 12,
|
| 321 |
+
"n_layers": 6,
|
| 322 |
+
"pad_token_id": 0,
|
| 323 |
+
"problem_type": "single_label_classification",
|
| 324 |
+
"qa_dropout": 0.1,
|
| 325 |
+
"seq_classif_dropout": 0.2,
|
| 326 |
+
"sinusoidal_pos_embds": false,
|
| 327 |
+
"tie_weights_": true,
|
| 328 |
+
"torch_dtype": "float32",
|
| 329 |
+
"transformers_version": "4.44.2",
|
| 330 |
+
"vocab_size": 30522
|
| 331 |
+
}
|
run-10/checkpoint-20670/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:744b9003a01b5dbac49a192b07d7ea152ac1018e4348b753960a2ac74b23f326
|
| 3 |
+
size 268290900
|
run-10/checkpoint-20670/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8b0e1177b42431136fa09671a270f02c586044ba3c2dccb7b92015591975e6e
|
| 3 |
+
size 536643898
|
run-10/checkpoint-20670/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8778887608feaac66ef895d461dee9594940d0493c8c35fe7020e00768acfac
|
| 3 |
+
size 14244
|
run-10/checkpoint-20670/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:164e003ec7442091f601e3f5f9485a117ce41fe8f9140e192a7e7673dc00cb53
|
| 3 |
+
size 1064
|
run-10/checkpoint-20670/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
run-10/checkpoint-20670/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
run-10/checkpoint-20670/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": true,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"mask_token": "[MASK]",
|
| 48 |
+
"model_max_length": 512,
|
| 49 |
+
"pad_token": "[PAD]",
|
| 50 |
+
"sep_token": "[SEP]",
|
| 51 |
+
"strip_accents": null,
|
| 52 |
+
"tokenize_chinese_chars": true,
|
| 53 |
+
"tokenizer_class": "DistilBertTokenizer",
|
| 54 |
+
"unk_token": "[UNK]"
|
| 55 |
+
}
|
run-10/checkpoint-20670/trainer_state.json
ADDED
|
@@ -0,0 +1,1068 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 65.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 20670,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.9968553459119497,
|
| 13 |
+
"grad_norm": 1.7892354726791382,
|
| 14 |
+
"learning_rate": 1.9693275278180943e-05,
|
| 15 |
+
"loss": 2.2558,
|
| 16 |
+
"step": 317
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_accuracy": 0.7374193548387097,
|
| 21 |
+
"eval_loss": 1.617253303527832,
|
| 22 |
+
"eval_runtime": 1.5797,
|
| 23 |
+
"eval_samples_per_second": 1962.428,
|
| 24 |
+
"eval_steps_per_second": 41.148,
|
| 25 |
+
"step": 318
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.9937106918238994,
|
| 29 |
+
"grad_norm": 1.7214646339416504,
|
| 30 |
+
"learning_rate": 1.9386550556361877e-05,
|
| 31 |
+
"loss": 1.2223,
|
| 32 |
+
"step": 634
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.0,
|
| 36 |
+
"eval_accuracy": 0.8806451612903226,
|
| 37 |
+
"eval_loss": 0.7785580158233643,
|
| 38 |
+
"eval_runtime": 1.5855,
|
| 39 |
+
"eval_samples_per_second": 1955.164,
|
| 40 |
+
"eval_steps_per_second": 40.995,
|
| 41 |
+
"step": 636
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 2.990566037735849,
|
| 45 |
+
"grad_norm": 2.026827812194824,
|
| 46 |
+
"learning_rate": 1.9079825834542818e-05,
|
| 47 |
+
"loss": 0.5979,
|
| 48 |
+
"step": 951
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 3.0,
|
| 52 |
+
"eval_accuracy": 0.9235483870967742,
|
| 53 |
+
"eval_loss": 0.4162185788154602,
|
| 54 |
+
"eval_runtime": 1.596,
|
| 55 |
+
"eval_samples_per_second": 1942.32,
|
| 56 |
+
"eval_steps_per_second": 40.726,
|
| 57 |
+
"step": 954
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 3.9874213836477987,
|
| 61 |
+
"grad_norm": 1.5834722518920898,
|
| 62 |
+
"learning_rate": 1.8773101112723755e-05,
|
| 63 |
+
"loss": 0.3243,
|
| 64 |
+
"step": 1268
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 4.0,
|
| 68 |
+
"eval_accuracy": 0.9380645161290323,
|
| 69 |
+
"eval_loss": 0.2845900356769562,
|
| 70 |
+
"eval_runtime": 1.584,
|
| 71 |
+
"eval_samples_per_second": 1957.074,
|
| 72 |
+
"eval_steps_per_second": 41.035,
|
| 73 |
+
"step": 1272
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.984276729559748,
|
| 77 |
+
"grad_norm": 0.5995246767997742,
|
| 78 |
+
"learning_rate": 1.8466376390904693e-05,
|
| 79 |
+
"loss": 0.2172,
|
| 80 |
+
"step": 1585
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 5.0,
|
| 84 |
+
"eval_accuracy": 0.9432258064516129,
|
| 85 |
+
"eval_loss": 0.24240821599960327,
|
| 86 |
+
"eval_runtime": 1.5902,
|
| 87 |
+
"eval_samples_per_second": 1949.469,
|
| 88 |
+
"eval_steps_per_second": 40.876,
|
| 89 |
+
"step": 1590
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 5.981132075471698,
|
| 93 |
+
"grad_norm": 0.5339710712432861,
|
| 94 |
+
"learning_rate": 1.8159651669085634e-05,
|
| 95 |
+
"loss": 0.1736,
|
| 96 |
+
"step": 1902
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 6.0,
|
| 100 |
+
"eval_accuracy": 0.9438709677419355,
|
| 101 |
+
"eval_loss": 0.22936078906059265,
|
| 102 |
+
"eval_runtime": 1.597,
|
| 103 |
+
"eval_samples_per_second": 1941.162,
|
| 104 |
+
"eval_steps_per_second": 40.702,
|
| 105 |
+
"step": 1908
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 6.977987421383648,
|
| 109 |
+
"grad_norm": 0.650678813457489,
|
| 110 |
+
"learning_rate": 1.785292694726657e-05,
|
| 111 |
+
"loss": 0.1522,
|
| 112 |
+
"step": 2219
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 7.0,
|
| 116 |
+
"eval_accuracy": 0.9464516129032258,
|
| 117 |
+
"eval_loss": 0.2193584442138672,
|
| 118 |
+
"eval_runtime": 1.583,
|
| 119 |
+
"eval_samples_per_second": 1958.299,
|
| 120 |
+
"eval_steps_per_second": 41.061,
|
| 121 |
+
"step": 2226
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 7.9748427672955975,
|
| 125 |
+
"grad_norm": 0.3262224793434143,
|
| 126 |
+
"learning_rate": 1.754620222544751e-05,
|
| 127 |
+
"loss": 0.141,
|
| 128 |
+
"step": 2536
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"eval_accuracy": 0.9467741935483871,
|
| 133 |
+
"eval_loss": 0.21464666724205017,
|
| 134 |
+
"eval_runtime": 1.6039,
|
| 135 |
+
"eval_samples_per_second": 1932.766,
|
| 136 |
+
"eval_steps_per_second": 40.526,
|
| 137 |
+
"step": 2544
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 8.971698113207546,
|
| 141 |
+
"grad_norm": 0.31927967071533203,
|
| 142 |
+
"learning_rate": 1.7239477503628447e-05,
|
| 143 |
+
"loss": 0.1338,
|
| 144 |
+
"step": 2853
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 9.0,
|
| 148 |
+
"eval_accuracy": 0.9458064516129032,
|
| 149 |
+
"eval_loss": 0.21053576469421387,
|
| 150 |
+
"eval_runtime": 1.5796,
|
| 151 |
+
"eval_samples_per_second": 1962.498,
|
| 152 |
+
"eval_steps_per_second": 41.149,
|
| 153 |
+
"step": 2862
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 9.968553459119496,
|
| 157 |
+
"grad_norm": 0.2677464783191681,
|
| 158 |
+
"learning_rate": 1.6932752781809388e-05,
|
| 159 |
+
"loss": 0.1285,
|
| 160 |
+
"step": 3170
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 10.0,
|
| 164 |
+
"eval_accuracy": 0.9458064516129032,
|
| 165 |
+
"eval_loss": 0.20650111138820648,
|
| 166 |
+
"eval_runtime": 1.5867,
|
| 167 |
+
"eval_samples_per_second": 1953.698,
|
| 168 |
+
"eval_steps_per_second": 40.965,
|
| 169 |
+
"step": 3180
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.965408805031446,
|
| 173 |
+
"grad_norm": 0.2738747000694275,
|
| 174 |
+
"learning_rate": 1.6626028059990325e-05,
|
| 175 |
+
"loss": 0.1249,
|
| 176 |
+
"step": 3487
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 11.0,
|
| 180 |
+
"eval_accuracy": 0.9451612903225807,
|
| 181 |
+
"eval_loss": 0.2075880914926529,
|
| 182 |
+
"eval_runtime": 1.5904,
|
| 183 |
+
"eval_samples_per_second": 1949.165,
|
| 184 |
+
"eval_steps_per_second": 40.87,
|
| 185 |
+
"step": 3498
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 11.962264150943396,
|
| 189 |
+
"grad_norm": 0.1720859259366989,
|
| 190 |
+
"learning_rate": 1.6319303338171263e-05,
|
| 191 |
+
"loss": 0.1224,
|
| 192 |
+
"step": 3804
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 12.0,
|
| 196 |
+
"eval_accuracy": 0.9487096774193549,
|
| 197 |
+
"eval_loss": 0.20247511565685272,
|
| 198 |
+
"eval_runtime": 1.5789,
|
| 199 |
+
"eval_samples_per_second": 1963.422,
|
| 200 |
+
"eval_steps_per_second": 41.169,
|
| 201 |
+
"step": 3816
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 12.959119496855346,
|
| 205 |
+
"grad_norm": 0.18272709846496582,
|
| 206 |
+
"learning_rate": 1.6012578616352204e-05,
|
| 207 |
+
"loss": 0.1195,
|
| 208 |
+
"step": 4121
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 13.0,
|
| 212 |
+
"eval_accuracy": 0.9461290322580646,
|
| 213 |
+
"eval_loss": 0.20313145220279694,
|
| 214 |
+
"eval_runtime": 1.586,
|
| 215 |
+
"eval_samples_per_second": 1954.586,
|
| 216 |
+
"eval_steps_per_second": 40.983,
|
| 217 |
+
"step": 4134
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 13.955974842767295,
|
| 221 |
+
"grad_norm": 0.169128879904747,
|
| 222 |
+
"learning_rate": 1.570585389453314e-05,
|
| 223 |
+
"loss": 0.1178,
|
| 224 |
+
"step": 4438
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 14.0,
|
| 228 |
+
"eval_accuracy": 0.9470967741935484,
|
| 229 |
+
"eval_loss": 0.2029477208852768,
|
| 230 |
+
"eval_runtime": 1.5852,
|
| 231 |
+
"eval_samples_per_second": 1955.55,
|
| 232 |
+
"eval_steps_per_second": 41.003,
|
| 233 |
+
"step": 4452
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 14.952830188679245,
|
| 237 |
+
"grad_norm": 0.23086608946323395,
|
| 238 |
+
"learning_rate": 1.539912917271408e-05,
|
| 239 |
+
"loss": 0.1164,
|
| 240 |
+
"step": 4755
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 15.0,
|
| 244 |
+
"eval_accuracy": 0.9474193548387096,
|
| 245 |
+
"eval_loss": 0.19893775880336761,
|
| 246 |
+
"eval_runtime": 1.5901,
|
| 247 |
+
"eval_samples_per_second": 1949.554,
|
| 248 |
+
"eval_steps_per_second": 40.878,
|
| 249 |
+
"step": 4770
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 15.949685534591195,
|
| 253 |
+
"grad_norm": 0.162359818816185,
|
| 254 |
+
"learning_rate": 1.5092404450895017e-05,
|
| 255 |
+
"loss": 0.1148,
|
| 256 |
+
"step": 5072
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 16.0,
|
| 260 |
+
"eval_accuracy": 0.9448387096774193,
|
| 261 |
+
"eval_loss": 0.2043694704771042,
|
| 262 |
+
"eval_runtime": 1.5835,
|
| 263 |
+
"eval_samples_per_second": 1957.683,
|
| 264 |
+
"eval_steps_per_second": 41.048,
|
| 265 |
+
"step": 5088
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 16.946540880503143,
|
| 269 |
+
"grad_norm": 0.19157171249389648,
|
| 270 |
+
"learning_rate": 1.4785679729075956e-05,
|
| 271 |
+
"loss": 0.1136,
|
| 272 |
+
"step": 5389
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 17.0,
|
| 276 |
+
"eval_accuracy": 0.9464516129032258,
|
| 277 |
+
"eval_loss": 0.20171256363391876,
|
| 278 |
+
"eval_runtime": 1.591,
|
| 279 |
+
"eval_samples_per_second": 1948.448,
|
| 280 |
+
"eval_steps_per_second": 40.855,
|
| 281 |
+
"step": 5406
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 17.943396226415093,
|
| 285 |
+
"grad_norm": 0.1393335908651352,
|
| 286 |
+
"learning_rate": 1.4478955007256896e-05,
|
| 287 |
+
"loss": 0.1125,
|
| 288 |
+
"step": 5706
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 18.0,
|
| 292 |
+
"eval_accuracy": 0.9480645161290323,
|
| 293 |
+
"eval_loss": 0.19980596005916595,
|
| 294 |
+
"eval_runtime": 1.5855,
|
| 295 |
+
"eval_samples_per_second": 1955.186,
|
| 296 |
+
"eval_steps_per_second": 40.996,
|
| 297 |
+
"step": 5724
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 18.940251572327043,
|
| 301 |
+
"grad_norm": 0.17489531636238098,
|
| 302 |
+
"learning_rate": 1.4172230285437833e-05,
|
| 303 |
+
"loss": 0.1115,
|
| 304 |
+
"step": 6023
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 19.0,
|
| 308 |
+
"eval_accuracy": 0.9480645161290323,
|
| 309 |
+
"eval_loss": 0.19913578033447266,
|
| 310 |
+
"eval_runtime": 1.5874,
|
| 311 |
+
"eval_samples_per_second": 1952.884,
|
| 312 |
+
"eval_steps_per_second": 40.948,
|
| 313 |
+
"step": 6042
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 19.937106918238992,
|
| 317 |
+
"grad_norm": 0.14055393636226654,
|
| 318 |
+
"learning_rate": 1.3865505563618772e-05,
|
| 319 |
+
"loss": 0.1104,
|
| 320 |
+
"step": 6340
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 20.0,
|
| 324 |
+
"eval_accuracy": 0.9490322580645161,
|
| 325 |
+
"eval_loss": 0.1966058611869812,
|
| 326 |
+
"eval_runtime": 1.5873,
|
| 327 |
+
"eval_samples_per_second": 1953.007,
|
| 328 |
+
"eval_steps_per_second": 40.95,
|
| 329 |
+
"step": 6360
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 20.933962264150942,
|
| 333 |
+
"grad_norm": 0.15471237897872925,
|
| 334 |
+
"learning_rate": 1.355878084179971e-05,
|
| 335 |
+
"loss": 0.11,
|
| 336 |
+
"step": 6657
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 21.0,
|
| 340 |
+
"eval_accuracy": 0.947741935483871,
|
| 341 |
+
"eval_loss": 0.1969558149576187,
|
| 342 |
+
"eval_runtime": 1.5977,
|
| 343 |
+
"eval_samples_per_second": 1940.313,
|
| 344 |
+
"eval_steps_per_second": 40.684,
|
| 345 |
+
"step": 6678
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 21.930817610062892,
|
| 349 |
+
"grad_norm": 0.13617640733718872,
|
| 350 |
+
"learning_rate": 1.325205611998065e-05,
|
| 351 |
+
"loss": 0.109,
|
| 352 |
+
"step": 6974
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 22.0,
|
| 356 |
+
"eval_accuracy": 0.9483870967741935,
|
| 357 |
+
"eval_loss": 0.19681183993816376,
|
| 358 |
+
"eval_runtime": 1.5836,
|
| 359 |
+
"eval_samples_per_second": 1957.549,
|
| 360 |
+
"eval_steps_per_second": 41.045,
|
| 361 |
+
"step": 6996
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 22.927672955974842,
|
| 365 |
+
"grad_norm": 0.12482684850692749,
|
| 366 |
+
"learning_rate": 1.2945331398161587e-05,
|
| 367 |
+
"loss": 0.1086,
|
| 368 |
+
"step": 7291
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 23.0,
|
| 372 |
+
"eval_accuracy": 0.9493548387096774,
|
| 373 |
+
"eval_loss": 0.19712504744529724,
|
| 374 |
+
"eval_runtime": 1.5908,
|
| 375 |
+
"eval_samples_per_second": 1948.729,
|
| 376 |
+
"eval_steps_per_second": 40.86,
|
| 377 |
+
"step": 7314
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 23.92452830188679,
|
| 381 |
+
"grad_norm": 0.1342603862285614,
|
| 382 |
+
"learning_rate": 1.2638606676342526e-05,
|
| 383 |
+
"loss": 0.1081,
|
| 384 |
+
"step": 7608
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 24.0,
|
| 388 |
+
"eval_accuracy": 0.9496774193548387,
|
| 389 |
+
"eval_loss": 0.19047369062900543,
|
| 390 |
+
"eval_runtime": 1.583,
|
| 391 |
+
"eval_samples_per_second": 1958.273,
|
| 392 |
+
"eval_steps_per_second": 41.061,
|
| 393 |
+
"step": 7632
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 24.92138364779874,
|
| 397 |
+
"grad_norm": 0.12532541155815125,
|
| 398 |
+
"learning_rate": 1.2331881954523466e-05,
|
| 399 |
+
"loss": 0.1076,
|
| 400 |
+
"step": 7925
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 25.0,
|
| 404 |
+
"eval_accuracy": 0.9490322580645161,
|
| 405 |
+
"eval_loss": 0.19533421099185944,
|
| 406 |
+
"eval_runtime": 1.5923,
|
| 407 |
+
"eval_samples_per_second": 1946.876,
|
| 408 |
+
"eval_steps_per_second": 40.822,
|
| 409 |
+
"step": 7950
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 25.91823899371069,
|
| 413 |
+
"grad_norm": 0.10873907059431076,
|
| 414 |
+
"learning_rate": 1.2025157232704403e-05,
|
| 415 |
+
"loss": 0.1071,
|
| 416 |
+
"step": 8242
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 26.0,
|
| 420 |
+
"eval_accuracy": 0.9490322580645161,
|
| 421 |
+
"eval_loss": 0.19298453629016876,
|
| 422 |
+
"eval_runtime": 1.5821,
|
| 423 |
+
"eval_samples_per_second": 1959.435,
|
| 424 |
+
"eval_steps_per_second": 41.085,
|
| 425 |
+
"step": 8268
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 26.91509433962264,
|
| 429 |
+
"grad_norm": 0.13738222420215607,
|
| 430 |
+
"learning_rate": 1.1718432510885342e-05,
|
| 431 |
+
"loss": 0.1065,
|
| 432 |
+
"step": 8559
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 27.0,
|
| 436 |
+
"eval_accuracy": 0.9474193548387096,
|
| 437 |
+
"eval_loss": 0.19757185876369476,
|
| 438 |
+
"eval_runtime": 1.5903,
|
| 439 |
+
"eval_samples_per_second": 1949.304,
|
| 440 |
+
"eval_steps_per_second": 40.873,
|
| 441 |
+
"step": 8586
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 27.91194968553459,
|
| 445 |
+
"grad_norm": 0.11282658576965332,
|
| 446 |
+
"learning_rate": 1.141170778906628e-05,
|
| 447 |
+
"loss": 0.1063,
|
| 448 |
+
"step": 8876
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 28.0,
|
| 452 |
+
"eval_accuracy": 0.9496774193548387,
|
| 453 |
+
"eval_loss": 0.1946183145046234,
|
| 454 |
+
"eval_runtime": 1.5849,
|
| 455 |
+
"eval_samples_per_second": 1955.904,
|
| 456 |
+
"eval_steps_per_second": 41.011,
|
| 457 |
+
"step": 8904
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 28.90880503144654,
|
| 461 |
+
"grad_norm": 0.11028722673654556,
|
| 462 |
+
"learning_rate": 1.110498306724722e-05,
|
| 463 |
+
"loss": 0.1057,
|
| 464 |
+
"step": 9193
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 29.0,
|
| 468 |
+
"eval_accuracy": 0.9493548387096774,
|
| 469 |
+
"eval_loss": 0.1943337470293045,
|
| 470 |
+
"eval_runtime": 1.5802,
|
| 471 |
+
"eval_samples_per_second": 1961.812,
|
| 472 |
+
"eval_steps_per_second": 41.135,
|
| 473 |
+
"step": 9222
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 29.90566037735849,
|
| 477 |
+
"grad_norm": 0.14671742916107178,
|
| 478 |
+
"learning_rate": 1.0798258345428159e-05,
|
| 479 |
+
"loss": 0.1054,
|
| 480 |
+
"step": 9510
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 30.0,
|
| 484 |
+
"eval_accuracy": 0.9493548387096774,
|
| 485 |
+
"eval_loss": 0.1915769875049591,
|
| 486 |
+
"eval_runtime": 1.5847,
|
| 487 |
+
"eval_samples_per_second": 1956.259,
|
| 488 |
+
"eval_steps_per_second": 41.018,
|
| 489 |
+
"step": 9540
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 30.90251572327044,
|
| 493 |
+
"grad_norm": 0.14922679960727692,
|
| 494 |
+
"learning_rate": 1.0491533623609096e-05,
|
| 495 |
+
"loss": 0.1051,
|
| 496 |
+
"step": 9827
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 31.0,
|
| 500 |
+
"eval_accuracy": 0.9467741935483871,
|
| 501 |
+
"eval_loss": 0.19569595158100128,
|
| 502 |
+
"eval_runtime": 1.5802,
|
| 503 |
+
"eval_samples_per_second": 1961.721,
|
| 504 |
+
"eval_steps_per_second": 41.133,
|
| 505 |
+
"step": 9858
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 31.89937106918239,
|
| 509 |
+
"grad_norm": 0.10370077192783356,
|
| 510 |
+
"learning_rate": 1.0184808901790036e-05,
|
| 511 |
+
"loss": 0.105,
|
| 512 |
+
"step": 10144
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 32.0,
|
| 516 |
+
"eval_accuracy": 0.95,
|
| 517 |
+
"eval_loss": 0.19266918301582336,
|
| 518 |
+
"eval_runtime": 1.5815,
|
| 519 |
+
"eval_samples_per_second": 1960.158,
|
| 520 |
+
"eval_steps_per_second": 41.1,
|
| 521 |
+
"step": 10176
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 32.89622641509434,
|
| 525 |
+
"grad_norm": 0.0962880551815033,
|
| 526 |
+
"learning_rate": 9.878084179970973e-06,
|
| 527 |
+
"loss": 0.1047,
|
| 528 |
+
"step": 10461
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 33.0,
|
| 532 |
+
"eval_accuracy": 0.9496774193548387,
|
| 533 |
+
"eval_loss": 0.19326689839363098,
|
| 534 |
+
"eval_runtime": 1.579,
|
| 535 |
+
"eval_samples_per_second": 1963.275,
|
| 536 |
+
"eval_steps_per_second": 41.165,
|
| 537 |
+
"step": 10494
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 33.893081761006286,
|
| 541 |
+
"grad_norm": 0.10902892053127289,
|
| 542 |
+
"learning_rate": 9.571359458151912e-06,
|
| 543 |
+
"loss": 0.1043,
|
| 544 |
+
"step": 10778
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 34.0,
|
| 548 |
+
"eval_accuracy": 0.9496774193548387,
|
| 549 |
+
"eval_loss": 0.1923554390668869,
|
| 550 |
+
"eval_runtime": 1.5857,
|
| 551 |
+
"eval_samples_per_second": 1954.952,
|
| 552 |
+
"eval_steps_per_second": 40.991,
|
| 553 |
+
"step": 10812
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 34.88993710691824,
|
| 557 |
+
"grad_norm": 0.12455958127975464,
|
| 558 |
+
"learning_rate": 9.26463473633285e-06,
|
| 559 |
+
"loss": 0.1043,
|
| 560 |
+
"step": 11095
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 35.0,
|
| 564 |
+
"eval_accuracy": 0.9493548387096774,
|
| 565 |
+
"eval_loss": 0.19354650378227234,
|
| 566 |
+
"eval_runtime": 1.5891,
|
| 567 |
+
"eval_samples_per_second": 1950.802,
|
| 568 |
+
"eval_steps_per_second": 40.904,
|
| 569 |
+
"step": 11130
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 35.886792452830186,
|
| 573 |
+
"grad_norm": 0.1168413907289505,
|
| 574 |
+
"learning_rate": 8.95791001451379e-06,
|
| 575 |
+
"loss": 0.1039,
|
| 576 |
+
"step": 11412
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 36.0,
|
| 580 |
+
"eval_accuracy": 0.9496774193548387,
|
| 581 |
+
"eval_loss": 0.19258633255958557,
|
| 582 |
+
"eval_runtime": 1.582,
|
| 583 |
+
"eval_samples_per_second": 1959.532,
|
| 584 |
+
"eval_steps_per_second": 41.087,
|
| 585 |
+
"step": 11448
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 36.88364779874214,
|
| 589 |
+
"grad_norm": 0.11007039248943329,
|
| 590 |
+
"learning_rate": 8.651185292694727e-06,
|
| 591 |
+
"loss": 0.104,
|
| 592 |
+
"step": 11729
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 37.0,
|
| 596 |
+
"eval_accuracy": 0.9506451612903226,
|
| 597 |
+
"eval_loss": 0.19169464707374573,
|
| 598 |
+
"eval_runtime": 1.5851,
|
| 599 |
+
"eval_samples_per_second": 1955.757,
|
| 600 |
+
"eval_steps_per_second": 41.008,
|
| 601 |
+
"step": 11766
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 37.880503144654085,
|
| 605 |
+
"grad_norm": 0.10598143190145493,
|
| 606 |
+
"learning_rate": 8.344460570875665e-06,
|
| 607 |
+
"loss": 0.1037,
|
| 608 |
+
"step": 12046
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 38.0,
|
| 612 |
+
"eval_accuracy": 0.9503225806451613,
|
| 613 |
+
"eval_loss": 0.19138343632221222,
|
| 614 |
+
"eval_runtime": 1.5756,
|
| 615 |
+
"eval_samples_per_second": 1967.531,
|
| 616 |
+
"eval_steps_per_second": 41.255,
|
| 617 |
+
"step": 12084
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 38.87735849056604,
|
| 621 |
+
"grad_norm": 0.12342941761016846,
|
| 622 |
+
"learning_rate": 8.037735849056606e-06,
|
| 623 |
+
"loss": 0.1031,
|
| 624 |
+
"step": 12363
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 39.0,
|
| 628 |
+
"eval_accuracy": 0.9490322580645161,
|
| 629 |
+
"eval_loss": 0.1933196783065796,
|
| 630 |
+
"eval_runtime": 1.5808,
|
| 631 |
+
"eval_samples_per_second": 1961.015,
|
| 632 |
+
"eval_steps_per_second": 41.118,
|
| 633 |
+
"step": 12402
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 39.874213836477985,
|
| 637 |
+
"grad_norm": 0.10500668734312057,
|
| 638 |
+
"learning_rate": 7.731011127237543e-06,
|
| 639 |
+
"loss": 0.1032,
|
| 640 |
+
"step": 12680
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 40.0,
|
| 644 |
+
"eval_accuracy": 0.9512903225806452,
|
| 645 |
+
"eval_loss": 0.18965306878089905,
|
| 646 |
+
"eval_runtime": 1.5814,
|
| 647 |
+
"eval_samples_per_second": 1960.237,
|
| 648 |
+
"eval_steps_per_second": 41.102,
|
| 649 |
+
"step": 12720
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 40.87106918238994,
|
| 653 |
+
"grad_norm": 0.11004229635000229,
|
| 654 |
+
"learning_rate": 7.424286405418482e-06,
|
| 655 |
+
"loss": 0.103,
|
| 656 |
+
"step": 12997
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 41.0,
|
| 660 |
+
"eval_accuracy": 0.95,
|
| 661 |
+
"eval_loss": 0.19283445179462433,
|
| 662 |
+
"eval_runtime": 1.5854,
|
| 663 |
+
"eval_samples_per_second": 1955.382,
|
| 664 |
+
"eval_steps_per_second": 41.0,
|
| 665 |
+
"step": 13038
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 41.867924528301884,
|
| 669 |
+
"grad_norm": 0.08491053432226181,
|
| 670 |
+
"learning_rate": 7.11756168359942e-06,
|
| 671 |
+
"loss": 0.1027,
|
| 672 |
+
"step": 13314
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 42.0,
|
| 676 |
+
"eval_accuracy": 0.9496774193548387,
|
| 677 |
+
"eval_loss": 0.19381901621818542,
|
| 678 |
+
"eval_runtime": 1.5827,
|
| 679 |
+
"eval_samples_per_second": 1958.66,
|
| 680 |
+
"eval_steps_per_second": 41.069,
|
| 681 |
+
"step": 13356
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 42.86477987421384,
|
| 685 |
+
"grad_norm": 0.10829228907823563,
|
| 686 |
+
"learning_rate": 6.8108369617803585e-06,
|
| 687 |
+
"loss": 0.1026,
|
| 688 |
+
"step": 13631
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 43.0,
|
| 692 |
+
"eval_accuracy": 0.9496774193548387,
|
| 693 |
+
"eval_loss": 0.19210335612297058,
|
| 694 |
+
"eval_runtime": 1.5793,
|
| 695 |
+
"eval_samples_per_second": 1962.945,
|
| 696 |
+
"eval_steps_per_second": 41.159,
|
| 697 |
+
"step": 13674
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 43.861635220125784,
|
| 701 |
+
"grad_norm": 0.08409814536571503,
|
| 702 |
+
"learning_rate": 6.504112239961297e-06,
|
| 703 |
+
"loss": 0.1026,
|
| 704 |
+
"step": 13948
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 44.0,
|
| 708 |
+
"eval_accuracy": 0.9490322580645161,
|
| 709 |
+
"eval_loss": 0.19112293422222137,
|
| 710 |
+
"eval_runtime": 1.5829,
|
| 711 |
+
"eval_samples_per_second": 1958.426,
|
| 712 |
+
"eval_steps_per_second": 41.064,
|
| 713 |
+
"step": 13992
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 44.85849056603774,
|
| 717 |
+
"grad_norm": 0.09595589339733124,
|
| 718 |
+
"learning_rate": 6.197387518142236e-06,
|
| 719 |
+
"loss": 0.1025,
|
| 720 |
+
"step": 14265
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 45.0,
|
| 724 |
+
"eval_accuracy": 0.9509677419354838,
|
| 725 |
+
"eval_loss": 0.19056861102581024,
|
| 726 |
+
"eval_runtime": 1.592,
|
| 727 |
+
"eval_samples_per_second": 1947.236,
|
| 728 |
+
"eval_steps_per_second": 40.829,
|
| 729 |
+
"step": 14310
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 45.855345911949684,
|
| 733 |
+
"grad_norm": 0.09873504191637039,
|
| 734 |
+
"learning_rate": 5.890662796323175e-06,
|
| 735 |
+
"loss": 0.102,
|
| 736 |
+
"step": 14582
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 46.0,
|
| 740 |
+
"eval_accuracy": 0.9516129032258065,
|
| 741 |
+
"eval_loss": 0.1900680810213089,
|
| 742 |
+
"eval_runtime": 1.5864,
|
| 743 |
+
"eval_samples_per_second": 1954.141,
|
| 744 |
+
"eval_steps_per_second": 40.974,
|
| 745 |
+
"step": 14628
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 46.85220125786164,
|
| 749 |
+
"grad_norm": 0.11340785026550293,
|
| 750 |
+
"learning_rate": 5.583938074504112e-06,
|
| 751 |
+
"loss": 0.102,
|
| 752 |
+
"step": 14899
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 47.0,
|
| 756 |
+
"eval_accuracy": 0.9516129032258065,
|
| 757 |
+
"eval_loss": 0.19092942774295807,
|
| 758 |
+
"eval_runtime": 1.5847,
|
| 759 |
+
"eval_samples_per_second": 1956.237,
|
| 760 |
+
"eval_steps_per_second": 41.018,
|
| 761 |
+
"step": 14946
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 47.84905660377358,
|
| 765 |
+
"grad_norm": 0.10237148404121399,
|
| 766 |
+
"learning_rate": 5.277213352685051e-06,
|
| 767 |
+
"loss": 0.1019,
|
| 768 |
+
"step": 15216
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 48.0,
|
| 772 |
+
"eval_accuracy": 0.9516129032258065,
|
| 773 |
+
"eval_loss": 0.1910076141357422,
|
| 774 |
+
"eval_runtime": 1.5896,
|
| 775 |
+
"eval_samples_per_second": 1950.15,
|
| 776 |
+
"eval_steps_per_second": 40.89,
|
| 777 |
+
"step": 15264
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 48.84591194968554,
|
| 781 |
+
"grad_norm": 0.1059408113360405,
|
| 782 |
+
"learning_rate": 4.97048863086599e-06,
|
| 783 |
+
"loss": 0.102,
|
| 784 |
+
"step": 15533
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 49.0,
|
| 788 |
+
"eval_accuracy": 0.9509677419354838,
|
| 789 |
+
"eval_loss": 0.19144326448440552,
|
| 790 |
+
"eval_runtime": 1.5991,
|
| 791 |
+
"eval_samples_per_second": 1938.605,
|
| 792 |
+
"eval_steps_per_second": 40.648,
|
| 793 |
+
"step": 15582
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 49.84276729559748,
|
| 797 |
+
"grad_norm": 0.12535175681114197,
|
| 798 |
+
"learning_rate": 4.6637639090469286e-06,
|
| 799 |
+
"loss": 0.1016,
|
| 800 |
+
"step": 15850
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 50.0,
|
| 804 |
+
"eval_accuracy": 0.9506451612903226,
|
| 805 |
+
"eval_loss": 0.19012872874736786,
|
| 806 |
+
"eval_runtime": 1.579,
|
| 807 |
+
"eval_samples_per_second": 1963.268,
|
| 808 |
+
"eval_steps_per_second": 41.165,
|
| 809 |
+
"step": 15900
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 50.839622641509436,
|
| 813 |
+
"grad_norm": 0.1147344708442688,
|
| 814 |
+
"learning_rate": 4.357039187227867e-06,
|
| 815 |
+
"loss": 0.1016,
|
| 816 |
+
"step": 16167
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 51.0,
|
| 820 |
+
"eval_accuracy": 0.95,
|
| 821 |
+
"eval_loss": 0.19135929644107819,
|
| 822 |
+
"eval_runtime": 1.5886,
|
| 823 |
+
"eval_samples_per_second": 1951.429,
|
| 824 |
+
"eval_steps_per_second": 40.917,
|
| 825 |
+
"step": 16218
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 51.83647798742138,
|
| 829 |
+
"grad_norm": 0.09682757407426834,
|
| 830 |
+
"learning_rate": 4.0503144654088055e-06,
|
| 831 |
+
"loss": 0.1017,
|
| 832 |
+
"step": 16484
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 52.0,
|
| 836 |
+
"eval_accuracy": 0.9509677419354838,
|
| 837 |
+
"eval_loss": 0.19057147204875946,
|
| 838 |
+
"eval_runtime": 1.5852,
|
| 839 |
+
"eval_samples_per_second": 1955.562,
|
| 840 |
+
"eval_steps_per_second": 41.004,
|
| 841 |
+
"step": 16536
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 52.833333333333336,
|
| 845 |
+
"grad_norm": 0.08178399503231049,
|
| 846 |
+
"learning_rate": 3.743589743589744e-06,
|
| 847 |
+
"loss": 0.1015,
|
| 848 |
+
"step": 16801
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 53.0,
|
| 852 |
+
"eval_accuracy": 0.9493548387096774,
|
| 853 |
+
"eval_loss": 0.1907467395067215,
|
| 854 |
+
"eval_runtime": 1.5863,
|
| 855 |
+
"eval_samples_per_second": 1954.248,
|
| 856 |
+
"eval_steps_per_second": 40.976,
|
| 857 |
+
"step": 16854
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 53.83018867924528,
|
| 861 |
+
"grad_norm": 0.11521653831005096,
|
| 862 |
+
"learning_rate": 3.4368650217706824e-06,
|
| 863 |
+
"loss": 0.1014,
|
| 864 |
+
"step": 17118
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 54.0,
|
| 868 |
+
"eval_accuracy": 0.9509677419354838,
|
| 869 |
+
"eval_loss": 0.1901290863752365,
|
| 870 |
+
"eval_runtime": 1.5866,
|
| 871 |
+
"eval_samples_per_second": 1953.871,
|
| 872 |
+
"eval_steps_per_second": 40.968,
|
| 873 |
+
"step": 17172
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 54.827044025157235,
|
| 877 |
+
"grad_norm": 0.09569505602121353,
|
| 878 |
+
"learning_rate": 3.1301402999516213e-06,
|
| 879 |
+
"loss": 0.1014,
|
| 880 |
+
"step": 17435
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 55.0,
|
| 884 |
+
"eval_accuracy": 0.9506451612903226,
|
| 885 |
+
"eval_loss": 0.19127020239830017,
|
| 886 |
+
"eval_runtime": 1.5853,
|
| 887 |
+
"eval_samples_per_second": 1955.414,
|
| 888 |
+
"eval_steps_per_second": 41.001,
|
| 889 |
+
"step": 17490
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 55.82389937106918,
|
| 893 |
+
"grad_norm": 0.1354704350233078,
|
| 894 |
+
"learning_rate": 2.8234155781325597e-06,
|
| 895 |
+
"loss": 0.1014,
|
| 896 |
+
"step": 17752
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 56.0,
|
| 900 |
+
"eval_accuracy": 0.9503225806451613,
|
| 901 |
+
"eval_loss": 0.19035856425762177,
|
| 902 |
+
"eval_runtime": 1.5818,
|
| 903 |
+
"eval_samples_per_second": 1959.757,
|
| 904 |
+
"eval_steps_per_second": 41.092,
|
| 905 |
+
"step": 17808
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 56.820754716981135,
|
| 909 |
+
"grad_norm": 0.10605770349502563,
|
| 910 |
+
"learning_rate": 2.5166908563134977e-06,
|
| 911 |
+
"loss": 0.101,
|
| 912 |
+
"step": 18069
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 57.0,
|
| 916 |
+
"eval_accuracy": 0.9503225806451613,
|
| 917 |
+
"eval_loss": 0.19088035821914673,
|
| 918 |
+
"eval_runtime": 1.5926,
|
| 919 |
+
"eval_samples_per_second": 1946.458,
|
| 920 |
+
"eval_steps_per_second": 40.813,
|
| 921 |
+
"step": 18126
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 57.81761006289308,
|
| 925 |
+
"grad_norm": 0.09835700690746307,
|
| 926 |
+
"learning_rate": 2.2099661344944366e-06,
|
| 927 |
+
"loss": 0.1012,
|
| 928 |
+
"step": 18386
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 58.0,
|
| 932 |
+
"eval_accuracy": 0.9509677419354838,
|
| 933 |
+
"eval_loss": 0.18963685631752014,
|
| 934 |
+
"eval_runtime": 1.5851,
|
| 935 |
+
"eval_samples_per_second": 1955.697,
|
| 936 |
+
"eval_steps_per_second": 41.007,
|
| 937 |
+
"step": 18444
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 58.814465408805034,
|
| 941 |
+
"grad_norm": 0.1112636849284172,
|
| 942 |
+
"learning_rate": 1.903241412675375e-06,
|
| 943 |
+
"loss": 0.101,
|
| 944 |
+
"step": 18703
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 59.0,
|
| 948 |
+
"eval_accuracy": 0.9503225806451613,
|
| 949 |
+
"eval_loss": 0.18980875611305237,
|
| 950 |
+
"eval_runtime": 1.5835,
|
| 951 |
+
"eval_samples_per_second": 1957.658,
|
| 952 |
+
"eval_steps_per_second": 41.048,
|
| 953 |
+
"step": 18762
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 59.81132075471698,
|
| 957 |
+
"grad_norm": 0.0956527441740036,
|
| 958 |
+
"learning_rate": 1.5965166908563137e-06,
|
| 959 |
+
"loss": 0.1012,
|
| 960 |
+
"step": 19020
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 60.0,
|
| 964 |
+
"eval_accuracy": 0.9503225806451613,
|
| 965 |
+
"eval_loss": 0.19082389771938324,
|
| 966 |
+
"eval_runtime": 1.583,
|
| 967 |
+
"eval_samples_per_second": 1958.298,
|
| 968 |
+
"eval_steps_per_second": 41.061,
|
| 969 |
+
"step": 19080
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 60.808176100628934,
|
| 973 |
+
"grad_norm": 0.11568117886781693,
|
| 974 |
+
"learning_rate": 1.289791969037252e-06,
|
| 975 |
+
"loss": 0.101,
|
| 976 |
+
"step": 19337
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 61.0,
|
| 980 |
+
"eval_accuracy": 0.9509677419354838,
|
| 981 |
+
"eval_loss": 0.19020406901836395,
|
| 982 |
+
"eval_runtime": 1.5863,
|
| 983 |
+
"eval_samples_per_second": 1954.251,
|
| 984 |
+
"eval_steps_per_second": 40.976,
|
| 985 |
+
"step": 19398
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 61.80503144654088,
|
| 989 |
+
"grad_norm": 0.10544057190418243,
|
| 990 |
+
"learning_rate": 9.830672472181907e-07,
|
| 991 |
+
"loss": 0.101,
|
| 992 |
+
"step": 19654
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 62.0,
|
| 996 |
+
"eval_accuracy": 0.9506451612903226,
|
| 997 |
+
"eval_loss": 0.18987761437892914,
|
| 998 |
+
"eval_runtime": 1.5906,
|
| 999 |
+
"eval_samples_per_second": 1948.959,
|
| 1000 |
+
"eval_steps_per_second": 40.865,
|
| 1001 |
+
"step": 19716
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 62.801886792452834,
|
| 1005 |
+
"grad_norm": 0.10238124430179596,
|
| 1006 |
+
"learning_rate": 6.763425253991292e-07,
|
| 1007 |
+
"loss": 0.101,
|
| 1008 |
+
"step": 19971
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 63.0,
|
| 1012 |
+
"eval_accuracy": 0.9512903225806452,
|
| 1013 |
+
"eval_loss": 0.18937255442142487,
|
| 1014 |
+
"eval_runtime": 1.5952,
|
| 1015 |
+
"eval_samples_per_second": 1943.308,
|
| 1016 |
+
"eval_steps_per_second": 40.747,
|
| 1017 |
+
"step": 20034
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 63.79874213836478,
|
| 1021 |
+
"grad_norm": 0.08944712579250336,
|
| 1022 |
+
"learning_rate": 3.696178035800677e-07,
|
| 1023 |
+
"loss": 0.1009,
|
| 1024 |
+
"step": 20288
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 64.0,
|
| 1028 |
+
"eval_accuracy": 0.9512903225806452,
|
| 1029 |
+
"eval_loss": 0.18973702192306519,
|
| 1030 |
+
"eval_runtime": 1.591,
|
| 1031 |
+
"eval_samples_per_second": 1948.431,
|
| 1032 |
+
"eval_steps_per_second": 40.854,
|
| 1033 |
+
"step": 20352
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 64.79559748427673,
|
| 1037 |
+
"grad_norm": 0.09156543016433716,
|
| 1038 |
+
"learning_rate": 6.289308176100629e-08,
|
| 1039 |
+
"loss": 0.1009,
|
| 1040 |
+
"step": 20605
|
| 1041 |
+
}
|
| 1042 |
+
],
|
| 1043 |
+
"logging_steps": 317,
|
| 1044 |
+
"max_steps": 20670,
|
| 1045 |
+
"num_input_tokens_seen": 0,
|
| 1046 |
+
"num_train_epochs": 65,
|
| 1047 |
+
"save_steps": 1000000000.0,
|
| 1048 |
+
"stateful_callbacks": {
|
| 1049 |
+
"TrainerControl": {
|
| 1050 |
+
"args": {
|
| 1051 |
+
"should_epoch_stop": false,
|
| 1052 |
+
"should_evaluate": false,
|
| 1053 |
+
"should_log": false,
|
| 1054 |
+
"should_save": true,
|
| 1055 |
+
"should_training_stop": true
|
| 1056 |
+
},
|
| 1057 |
+
"attributes": {}
|
| 1058 |
+
}
|
| 1059 |
+
},
|
| 1060 |
+
"total_flos": 5356831747911420.0,
|
| 1061 |
+
"train_batch_size": 48,
|
| 1062 |
+
"trial_name": null,
|
| 1063 |
+
"trial_params": {
|
| 1064 |
+
"alpha": 0.4628195497376964,
|
| 1065 |
+
"num_train_epochs": 65,
|
| 1066 |
+
"temperature": 4
|
| 1067 |
+
}
|
| 1068 |
+
}
|
run-10/checkpoint-20670/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5d0e048ad3243f3df054d98952bd88da3e255967a5b0c48dd11bcc42248d7f0
|
| 3 |
+
size 5304
|
run-10/checkpoint-20670/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
run-12/checkpoint-20034/config.json
ADDED
|
@@ -0,0 +1,331 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilbert-base-uncased",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"hidden_dim": 3072,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "restaurant_reviews",
|
| 13 |
+
"1": "nutrition_info",
|
| 14 |
+
"2": "account_blocked",
|
| 15 |
+
"3": "oil_change_how",
|
| 16 |
+
"4": "time",
|
| 17 |
+
"5": "weather",
|
| 18 |
+
"6": "redeem_rewards",
|
| 19 |
+
"7": "interest_rate",
|
| 20 |
+
"8": "gas_type",
|
| 21 |
+
"9": "accept_reservations",
|
| 22 |
+
"10": "smart_home",
|
| 23 |
+
"11": "user_name",
|
| 24 |
+
"12": "report_lost_card",
|
| 25 |
+
"13": "repeat",
|
| 26 |
+
"14": "whisper_mode",
|
| 27 |
+
"15": "what_are_your_hobbies",
|
| 28 |
+
"16": "order",
|
| 29 |
+
"17": "jump_start",
|
| 30 |
+
"18": "schedule_meeting",
|
| 31 |
+
"19": "meeting_schedule",
|
| 32 |
+
"20": "freeze_account",
|
| 33 |
+
"21": "what_song",
|
| 34 |
+
"22": "meaning_of_life",
|
| 35 |
+
"23": "restaurant_reservation",
|
| 36 |
+
"24": "traffic",
|
| 37 |
+
"25": "make_call",
|
| 38 |
+
"26": "text",
|
| 39 |
+
"27": "bill_balance",
|
| 40 |
+
"28": "improve_credit_score",
|
| 41 |
+
"29": "change_language",
|
| 42 |
+
"30": "no",
|
| 43 |
+
"31": "measurement_conversion",
|
| 44 |
+
"32": "timer",
|
| 45 |
+
"33": "flip_coin",
|
| 46 |
+
"34": "do_you_have_pets",
|
| 47 |
+
"35": "balance",
|
| 48 |
+
"36": "tell_joke",
|
| 49 |
+
"37": "last_maintenance",
|
| 50 |
+
"38": "exchange_rate",
|
| 51 |
+
"39": "uber",
|
| 52 |
+
"40": "car_rental",
|
| 53 |
+
"41": "credit_limit",
|
| 54 |
+
"42": "oos",
|
| 55 |
+
"43": "shopping_list",
|
| 56 |
+
"44": "expiration_date",
|
| 57 |
+
"45": "routing",
|
| 58 |
+
"46": "meal_suggestion",
|
| 59 |
+
"47": "tire_change",
|
| 60 |
+
"48": "todo_list",
|
| 61 |
+
"49": "card_declined",
|
| 62 |
+
"50": "rewards_balance",
|
| 63 |
+
"51": "change_accent",
|
| 64 |
+
"52": "vaccines",
|
| 65 |
+
"53": "reminder_update",
|
| 66 |
+
"54": "food_last",
|
| 67 |
+
"55": "change_ai_name",
|
| 68 |
+
"56": "bill_due",
|
| 69 |
+
"57": "who_do_you_work_for",
|
| 70 |
+
"58": "share_location",
|
| 71 |
+
"59": "international_visa",
|
| 72 |
+
"60": "calendar",
|
| 73 |
+
"61": "translate",
|
| 74 |
+
"62": "carry_on",
|
| 75 |
+
"63": "book_flight",
|
| 76 |
+
"64": "insurance_change",
|
| 77 |
+
"65": "todo_list_update",
|
| 78 |
+
"66": "timezone",
|
| 79 |
+
"67": "cancel_reservation",
|
| 80 |
+
"68": "transactions",
|
| 81 |
+
"69": "credit_score",
|
| 82 |
+
"70": "report_fraud",
|
| 83 |
+
"71": "spending_history",
|
| 84 |
+
"72": "directions",
|
| 85 |
+
"73": "spelling",
|
| 86 |
+
"74": "insurance",
|
| 87 |
+
"75": "what_is_your_name",
|
| 88 |
+
"76": "reminder",
|
| 89 |
+
"77": "where_are_you_from",
|
| 90 |
+
"78": "distance",
|
| 91 |
+
"79": "payday",
|
| 92 |
+
"80": "flight_status",
|
| 93 |
+
"81": "find_phone",
|
| 94 |
+
"82": "greeting",
|
| 95 |
+
"83": "alarm",
|
| 96 |
+
"84": "order_status",
|
| 97 |
+
"85": "confirm_reservation",
|
| 98 |
+
"86": "cook_time",
|
| 99 |
+
"87": "damaged_card",
|
| 100 |
+
"88": "reset_settings",
|
| 101 |
+
"89": "pin_change",
|
| 102 |
+
"90": "replacement_card_duration",
|
| 103 |
+
"91": "new_card",
|
| 104 |
+
"92": "roll_dice",
|
| 105 |
+
"93": "income",
|
| 106 |
+
"94": "taxes",
|
| 107 |
+
"95": "date",
|
| 108 |
+
"96": "who_made_you",
|
| 109 |
+
"97": "pto_request",
|
| 110 |
+
"98": "tire_pressure",
|
| 111 |
+
"99": "how_old_are_you",
|
| 112 |
+
"100": "rollover_401k",
|
| 113 |
+
"101": "pto_request_status",
|
| 114 |
+
"102": "how_busy",
|
| 115 |
+
"103": "application_status",
|
| 116 |
+
"104": "recipe",
|
| 117 |
+
"105": "calendar_update",
|
| 118 |
+
"106": "play_music",
|
| 119 |
+
"107": "yes",
|
| 120 |
+
"108": "direct_deposit",
|
| 121 |
+
"109": "credit_limit_change",
|
| 122 |
+
"110": "gas",
|
| 123 |
+
"111": "pay_bill",
|
| 124 |
+
"112": "ingredients_list",
|
| 125 |
+
"113": "lost_luggage",
|
| 126 |
+
"114": "goodbye",
|
| 127 |
+
"115": "what_can_i_ask_you",
|
| 128 |
+
"116": "book_hotel",
|
| 129 |
+
"117": "are_you_a_bot",
|
| 130 |
+
"118": "next_song",
|
| 131 |
+
"119": "change_speed",
|
| 132 |
+
"120": "plug_type",
|
| 133 |
+
"121": "maybe",
|
| 134 |
+
"122": "w2",
|
| 135 |
+
"123": "oil_change_when",
|
| 136 |
+
"124": "thank_you",
|
| 137 |
+
"125": "shopping_list_update",
|
| 138 |
+
"126": "pto_balance",
|
| 139 |
+
"127": "order_checks",
|
| 140 |
+
"128": "travel_alert",
|
| 141 |
+
"129": "fun_fact",
|
| 142 |
+
"130": "sync_device",
|
| 143 |
+
"131": "schedule_maintenance",
|
| 144 |
+
"132": "apr",
|
| 145 |
+
"133": "transfer",
|
| 146 |
+
"134": "ingredient_substitution",
|
| 147 |
+
"135": "calories",
|
| 148 |
+
"136": "current_location",
|
| 149 |
+
"137": "international_fees",
|
| 150 |
+
"138": "calculator",
|
| 151 |
+
"139": "definition",
|
| 152 |
+
"140": "next_holiday",
|
| 153 |
+
"141": "update_playlist",
|
| 154 |
+
"142": "mpg",
|
| 155 |
+
"143": "min_payment",
|
| 156 |
+
"144": "change_user_name",
|
| 157 |
+
"145": "restaurant_suggestion",
|
| 158 |
+
"146": "travel_notification",
|
| 159 |
+
"147": "cancel",
|
| 160 |
+
"148": "pto_used",
|
| 161 |
+
"149": "travel_suggestion",
|
| 162 |
+
"150": "change_volume"
|
| 163 |
+
},
|
| 164 |
+
"initializer_range": 0.02,
|
| 165 |
+
"label2id": {
|
| 166 |
+
"accept_reservations": 9,
|
| 167 |
+
"account_blocked": 2,
|
| 168 |
+
"alarm": 83,
|
| 169 |
+
"application_status": 103,
|
| 170 |
+
"apr": 132,
|
| 171 |
+
"are_you_a_bot": 117,
|
| 172 |
+
"balance": 35,
|
| 173 |
+
"bill_balance": 27,
|
| 174 |
+
"bill_due": 56,
|
| 175 |
+
"book_flight": 63,
|
| 176 |
+
"book_hotel": 116,
|
| 177 |
+
"calculator": 138,
|
| 178 |
+
"calendar": 60,
|
| 179 |
+
"calendar_update": 105,
|
| 180 |
+
"calories": 135,
|
| 181 |
+
"cancel": 147,
|
| 182 |
+
"cancel_reservation": 67,
|
| 183 |
+
"car_rental": 40,
|
| 184 |
+
"card_declined": 49,
|
| 185 |
+
"carry_on": 62,
|
| 186 |
+
"change_accent": 51,
|
| 187 |
+
"change_ai_name": 55,
|
| 188 |
+
"change_language": 29,
|
| 189 |
+
"change_speed": 119,
|
| 190 |
+
"change_user_name": 144,
|
| 191 |
+
"change_volume": 150,
|
| 192 |
+
"confirm_reservation": 85,
|
| 193 |
+
"cook_time": 86,
|
| 194 |
+
"credit_limit": 41,
|
| 195 |
+
"credit_limit_change": 109,
|
| 196 |
+
"credit_score": 69,
|
| 197 |
+
"current_location": 136,
|
| 198 |
+
"damaged_card": 87,
|
| 199 |
+
"date": 95,
|
| 200 |
+
"definition": 139,
|
| 201 |
+
"direct_deposit": 108,
|
| 202 |
+
"directions": 72,
|
| 203 |
+
"distance": 78,
|
| 204 |
+
"do_you_have_pets": 34,
|
| 205 |
+
"exchange_rate": 38,
|
| 206 |
+
"expiration_date": 44,
|
| 207 |
+
"find_phone": 81,
|
| 208 |
+
"flight_status": 80,
|
| 209 |
+
"flip_coin": 33,
|
| 210 |
+
"food_last": 54,
|
| 211 |
+
"freeze_account": 20,
|
| 212 |
+
"fun_fact": 129,
|
| 213 |
+
"gas": 110,
|
| 214 |
+
"gas_type": 8,
|
| 215 |
+
"goodbye": 114,
|
| 216 |
+
"greeting": 82,
|
| 217 |
+
"how_busy": 102,
|
| 218 |
+
"how_old_are_you": 99,
|
| 219 |
+
"improve_credit_score": 28,
|
| 220 |
+
"income": 93,
|
| 221 |
+
"ingredient_substitution": 134,
|
| 222 |
+
"ingredients_list": 112,
|
| 223 |
+
"insurance": 74,
|
| 224 |
+
"insurance_change": 64,
|
| 225 |
+
"interest_rate": 7,
|
| 226 |
+
"international_fees": 137,
|
| 227 |
+
"international_visa": 59,
|
| 228 |
+
"jump_start": 17,
|
| 229 |
+
"last_maintenance": 37,
|
| 230 |
+
"lost_luggage": 113,
|
| 231 |
+
"make_call": 25,
|
| 232 |
+
"maybe": 121,
|
| 233 |
+
"meal_suggestion": 46,
|
| 234 |
+
"meaning_of_life": 22,
|
| 235 |
+
"measurement_conversion": 31,
|
| 236 |
+
"meeting_schedule": 19,
|
| 237 |
+
"min_payment": 143,
|
| 238 |
+
"mpg": 142,
|
| 239 |
+
"new_card": 91,
|
| 240 |
+
"next_holiday": 140,
|
| 241 |
+
"next_song": 118,
|
| 242 |
+
"no": 30,
|
| 243 |
+
"nutrition_info": 1,
|
| 244 |
+
"oil_change_how": 3,
|
| 245 |
+
"oil_change_when": 123,
|
| 246 |
+
"oos": 42,
|
| 247 |
+
"order": 16,
|
| 248 |
+
"order_checks": 127,
|
| 249 |
+
"order_status": 84,
|
| 250 |
+
"pay_bill": 111,
|
| 251 |
+
"payday": 79,
|
| 252 |
+
"pin_change": 89,
|
| 253 |
+
"play_music": 106,
|
| 254 |
+
"plug_type": 120,
|
| 255 |
+
"pto_balance": 126,
|
| 256 |
+
"pto_request": 97,
|
| 257 |
+
"pto_request_status": 101,
|
| 258 |
+
"pto_used": 148,
|
| 259 |
+
"recipe": 104,
|
| 260 |
+
"redeem_rewards": 6,
|
| 261 |
+
"reminder": 76,
|
| 262 |
+
"reminder_update": 53,
|
| 263 |
+
"repeat": 13,
|
| 264 |
+
"replacement_card_duration": 90,
|
| 265 |
+
"report_fraud": 70,
|
| 266 |
+
"report_lost_card": 12,
|
| 267 |
+
"reset_settings": 88,
|
| 268 |
+
"restaurant_reservation": 23,
|
| 269 |
+
"restaurant_reviews": 0,
|
| 270 |
+
"restaurant_suggestion": 145,
|
| 271 |
+
"rewards_balance": 50,
|
| 272 |
+
"roll_dice": 92,
|
| 273 |
+
"rollover_401k": 100,
|
| 274 |
+
"routing": 45,
|
| 275 |
+
"schedule_maintenance": 131,
|
| 276 |
+
"schedule_meeting": 18,
|
| 277 |
+
"share_location": 58,
|
| 278 |
+
"shopping_list": 43,
|
| 279 |
+
"shopping_list_update": 125,
|
| 280 |
+
"smart_home": 10,
|
| 281 |
+
"spelling": 73,
|
| 282 |
+
"spending_history": 71,
|
| 283 |
+
"sync_device": 130,
|
| 284 |
+
"taxes": 94,
|
| 285 |
+
"tell_joke": 36,
|
| 286 |
+
"text": 26,
|
| 287 |
+
"thank_you": 124,
|
| 288 |
+
"time": 4,
|
| 289 |
+
"timer": 32,
|
| 290 |
+
"timezone": 66,
|
| 291 |
+
"tire_change": 47,
|
| 292 |
+
"tire_pressure": 98,
|
| 293 |
+
"todo_list": 48,
|
| 294 |
+
"todo_list_update": 65,
|
| 295 |
+
"traffic": 24,
|
| 296 |
+
"transactions": 68,
|
| 297 |
+
"transfer": 133,
|
| 298 |
+
"translate": 61,
|
| 299 |
+
"travel_alert": 128,
|
| 300 |
+
"travel_notification": 146,
|
| 301 |
+
"travel_suggestion": 149,
|
| 302 |
+
"uber": 39,
|
| 303 |
+
"update_playlist": 141,
|
| 304 |
+
"user_name": 11,
|
| 305 |
+
"vaccines": 52,
|
| 306 |
+
"w2": 122,
|
| 307 |
+
"weather": 5,
|
| 308 |
+
"what_are_your_hobbies": 15,
|
| 309 |
+
"what_can_i_ask_you": 115,
|
| 310 |
+
"what_is_your_name": 75,
|
| 311 |
+
"what_song": 21,
|
| 312 |
+
"where_are_you_from": 77,
|
| 313 |
+
"whisper_mode": 14,
|
| 314 |
+
"who_do_you_work_for": 57,
|
| 315 |
+
"who_made_you": 96,
|
| 316 |
+
"yes": 107
|
| 317 |
+
},
|
| 318 |
+
"max_position_embeddings": 512,
|
| 319 |
+
"model_type": "distilbert",
|
| 320 |
+
"n_heads": 12,
|
| 321 |
+
"n_layers": 6,
|
| 322 |
+
"pad_token_id": 0,
|
| 323 |
+
"problem_type": "single_label_classification",
|
| 324 |
+
"qa_dropout": 0.1,
|
| 325 |
+
"seq_classif_dropout": 0.2,
|
| 326 |
+
"sinusoidal_pos_embds": false,
|
| 327 |
+
"tie_weights_": true,
|
| 328 |
+
"torch_dtype": "float32",
|
| 329 |
+
"transformers_version": "4.44.2",
|
| 330 |
+
"vocab_size": 30522
|
| 331 |
+
}
|
run-12/checkpoint-20034/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ef3ae8b006a1eb27181fa9b110b28531f4878b49b9b39450027f12306c979f54
|
| 3 |
+
size 268290900
|
run-12/checkpoint-20034/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8081143cd35e2983c46d504e342dcac99b07aeac8f9356ef8bfcb386630f7445
|
| 3 |
+
size 536643898
|
run-12/checkpoint-20034/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f4832671dc909dae4ef67b1908ecdd2d91a8035ca871f57971390aa62b8d5ea
|
| 3 |
+
size 14244
|
run-12/checkpoint-20034/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d63d59f0a622b88321e4168ab1d9091f9648ac72f53082b2ec881163e975be34
|
| 3 |
+
size 1064
|
run-12/checkpoint-20034/special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
run-12/checkpoint-20034/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
run-12/checkpoint-20034/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": true,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"mask_token": "[MASK]",
|
| 48 |
+
"model_max_length": 512,
|
| 49 |
+
"pad_token": "[PAD]",
|
| 50 |
+
"sep_token": "[SEP]",
|
| 51 |
+
"strip_accents": null,
|
| 52 |
+
"tokenize_chinese_chars": true,
|
| 53 |
+
"tokenizer_class": "DistilBertTokenizer",
|
| 54 |
+
"unk_token": "[UNK]"
|
| 55 |
+
}
|
run-12/checkpoint-20034/trainer_state.json
ADDED
|
@@ -0,0 +1,1036 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 63.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 20034,
|
| 7 |
+
"is_hyper_param_search": true,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.9968553459119497,
|
| 13 |
+
"grad_norm": 1.8778293132781982,
|
| 14 |
+
"learning_rate": 1.9683537985424778e-05,
|
| 15 |
+
"loss": 2.3675,
|
| 16 |
+
"step": 317
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_accuracy": 0.7393548387096774,
|
| 21 |
+
"eval_loss": 1.7057816982269287,
|
| 22 |
+
"eval_runtime": 1.5983,
|
| 23 |
+
"eval_samples_per_second": 1939.591,
|
| 24 |
+
"eval_steps_per_second": 40.669,
|
| 25 |
+
"step": 318
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.9937106918238994,
|
| 29 |
+
"grad_norm": 1.8159964084625244,
|
| 30 |
+
"learning_rate": 1.9367075970849558e-05,
|
| 31 |
+
"loss": 1.2883,
|
| 32 |
+
"step": 634
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.0,
|
| 36 |
+
"eval_accuracy": 0.88,
|
| 37 |
+
"eval_loss": 0.8210179805755615,
|
| 38 |
+
"eval_runtime": 1.5815,
|
| 39 |
+
"eval_samples_per_second": 1960.208,
|
| 40 |
+
"eval_steps_per_second": 41.101,
|
| 41 |
+
"step": 636
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 2.990566037735849,
|
| 45 |
+
"grad_norm": 2.137014865875244,
|
| 46 |
+
"learning_rate": 1.9050613956274335e-05,
|
| 47 |
+
"loss": 0.6277,
|
| 48 |
+
"step": 951
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 3.0,
|
| 52 |
+
"eval_accuracy": 0.9216129032258065,
|
| 53 |
+
"eval_loss": 0.4372541308403015,
|
| 54 |
+
"eval_runtime": 1.5878,
|
| 55 |
+
"eval_samples_per_second": 1952.368,
|
| 56 |
+
"eval_steps_per_second": 40.937,
|
| 57 |
+
"step": 954
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 3.9874213836477987,
|
| 61 |
+
"grad_norm": 1.6517659425735474,
|
| 62 |
+
"learning_rate": 1.873415194169911e-05,
|
| 63 |
+
"loss": 0.3377,
|
| 64 |
+
"step": 1268
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 4.0,
|
| 68 |
+
"eval_accuracy": 0.9374193548387096,
|
| 69 |
+
"eval_loss": 0.2971399426460266,
|
| 70 |
+
"eval_runtime": 1.5915,
|
| 71 |
+
"eval_samples_per_second": 1947.808,
|
| 72 |
+
"eval_steps_per_second": 40.841,
|
| 73 |
+
"step": 1272
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.984276729559748,
|
| 77 |
+
"grad_norm": 0.6481015086174011,
|
| 78 |
+
"learning_rate": 1.841768992712389e-05,
|
| 79 |
+
"loss": 0.223,
|
| 80 |
+
"step": 1585
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 5.0,
|
| 84 |
+
"eval_accuracy": 0.9425806451612904,
|
| 85 |
+
"eval_loss": 0.2514994442462921,
|
| 86 |
+
"eval_runtime": 1.5856,
|
| 87 |
+
"eval_samples_per_second": 1955.111,
|
| 88 |
+
"eval_steps_per_second": 40.994,
|
| 89 |
+
"step": 1590
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 5.981132075471698,
|
| 93 |
+
"grad_norm": 0.5463038682937622,
|
| 94 |
+
"learning_rate": 1.8101227912548668e-05,
|
| 95 |
+
"loss": 0.1761,
|
| 96 |
+
"step": 1902
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 6.0,
|
| 100 |
+
"eval_accuracy": 0.944516129032258,
|
| 101 |
+
"eval_loss": 0.23758269846439362,
|
| 102 |
+
"eval_runtime": 1.5787,
|
| 103 |
+
"eval_samples_per_second": 1963.585,
|
| 104 |
+
"eval_steps_per_second": 41.172,
|
| 105 |
+
"step": 1908
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 6.977987421383648,
|
| 109 |
+
"grad_norm": 0.6665645837783813,
|
| 110 |
+
"learning_rate": 1.7784765897973445e-05,
|
| 111 |
+
"loss": 0.1533,
|
| 112 |
+
"step": 2219
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 7.0,
|
| 116 |
+
"eval_accuracy": 0.9467741935483871,
|
| 117 |
+
"eval_loss": 0.22703349590301514,
|
| 118 |
+
"eval_runtime": 1.5867,
|
| 119 |
+
"eval_samples_per_second": 1953.7,
|
| 120 |
+
"eval_steps_per_second": 40.965,
|
| 121 |
+
"step": 2226
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 7.9748427672955975,
|
| 125 |
+
"grad_norm": 0.3181888461112976,
|
| 126 |
+
"learning_rate": 1.7468303883398225e-05,
|
| 127 |
+
"loss": 0.1416,
|
| 128 |
+
"step": 2536
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"eval_accuracy": 0.9467741935483871,
|
| 133 |
+
"eval_loss": 0.22196610271930695,
|
| 134 |
+
"eval_runtime": 1.584,
|
| 135 |
+
"eval_samples_per_second": 1957.081,
|
| 136 |
+
"eval_steps_per_second": 41.036,
|
| 137 |
+
"step": 2544
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 8.971698113207546,
|
| 141 |
+
"grad_norm": 0.31845995783805847,
|
| 142 |
+
"learning_rate": 1.7151841868823e-05,
|
| 143 |
+
"loss": 0.1339,
|
| 144 |
+
"step": 2853
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 9.0,
|
| 148 |
+
"eval_accuracy": 0.9467741935483871,
|
| 149 |
+
"eval_loss": 0.21790818870067596,
|
| 150 |
+
"eval_runtime": 1.603,
|
| 151 |
+
"eval_samples_per_second": 1933.897,
|
| 152 |
+
"eval_steps_per_second": 40.549,
|
| 153 |
+
"step": 2862
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 9.968553459119496,
|
| 157 |
+
"grad_norm": 0.2679583430290222,
|
| 158 |
+
"learning_rate": 1.6835379854247778e-05,
|
| 159 |
+
"loss": 0.1284,
|
| 160 |
+
"step": 3170
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 10.0,
|
| 164 |
+
"eval_accuracy": 0.9441935483870968,
|
| 165 |
+
"eval_loss": 0.21365021169185638,
|
| 166 |
+
"eval_runtime": 1.5889,
|
| 167 |
+
"eval_samples_per_second": 1951.025,
|
| 168 |
+
"eval_steps_per_second": 40.909,
|
| 169 |
+
"step": 3180
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.965408805031446,
|
| 173 |
+
"grad_norm": 0.2746322751045227,
|
| 174 |
+
"learning_rate": 1.6518917839672558e-05,
|
| 175 |
+
"loss": 0.1247,
|
| 176 |
+
"step": 3487
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 11.0,
|
| 180 |
+
"eval_accuracy": 0.9454838709677419,
|
| 181 |
+
"eval_loss": 0.21490098536014557,
|
| 182 |
+
"eval_runtime": 1.5913,
|
| 183 |
+
"eval_samples_per_second": 1948.133,
|
| 184 |
+
"eval_steps_per_second": 40.848,
|
| 185 |
+
"step": 3498
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 11.962264150943396,
|
| 189 |
+
"grad_norm": 0.17256155610084534,
|
| 190 |
+
"learning_rate": 1.6202455825097335e-05,
|
| 191 |
+
"loss": 0.1222,
|
| 192 |
+
"step": 3804
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 12.0,
|
| 196 |
+
"eval_accuracy": 0.9490322580645161,
|
| 197 |
+
"eval_loss": 0.20987379550933838,
|
| 198 |
+
"eval_runtime": 1.5991,
|
| 199 |
+
"eval_samples_per_second": 1938.633,
|
| 200 |
+
"eval_steps_per_second": 40.649,
|
| 201 |
+
"step": 3816
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 12.959119496855346,
|
| 205 |
+
"grad_norm": 0.18696637451648712,
|
| 206 |
+
"learning_rate": 1.5885993810522115e-05,
|
| 207 |
+
"loss": 0.1192,
|
| 208 |
+
"step": 4121
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 13.0,
|
| 212 |
+
"eval_accuracy": 0.9454838709677419,
|
| 213 |
+
"eval_loss": 0.20995916426181793,
|
| 214 |
+
"eval_runtime": 1.591,
|
| 215 |
+
"eval_samples_per_second": 1948.425,
|
| 216 |
+
"eval_steps_per_second": 40.854,
|
| 217 |
+
"step": 4134
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 13.955974842767295,
|
| 221 |
+
"grad_norm": 0.16687585413455963,
|
| 222 |
+
"learning_rate": 1.556953179594689e-05,
|
| 223 |
+
"loss": 0.1175,
|
| 224 |
+
"step": 4438
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 14.0,
|
| 228 |
+
"eval_accuracy": 0.9470967741935484,
|
| 229 |
+
"eval_loss": 0.20942126214504242,
|
| 230 |
+
"eval_runtime": 1.5854,
|
| 231 |
+
"eval_samples_per_second": 1955.322,
|
| 232 |
+
"eval_steps_per_second": 40.999,
|
| 233 |
+
"step": 4452
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 14.952830188679245,
|
| 237 |
+
"grad_norm": 0.2222217470407486,
|
| 238 |
+
"learning_rate": 1.525306978137167e-05,
|
| 239 |
+
"loss": 0.116,
|
| 240 |
+
"step": 4755
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 15.0,
|
| 244 |
+
"eval_accuracy": 0.9483870967741935,
|
| 245 |
+
"eval_loss": 0.20641160011291504,
|
| 246 |
+
"eval_runtime": 1.5853,
|
| 247 |
+
"eval_samples_per_second": 1955.458,
|
| 248 |
+
"eval_steps_per_second": 41.002,
|
| 249 |
+
"step": 4770
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 15.949685534591195,
|
| 253 |
+
"grad_norm": 0.15665258467197418,
|
| 254 |
+
"learning_rate": 1.4936607766796446e-05,
|
| 255 |
+
"loss": 0.1147,
|
| 256 |
+
"step": 5072
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 16.0,
|
| 260 |
+
"eval_accuracy": 0.944516129032258,
|
| 261 |
+
"eval_loss": 0.21185828745365143,
|
| 262 |
+
"eval_runtime": 1.5843,
|
| 263 |
+
"eval_samples_per_second": 1956.641,
|
| 264 |
+
"eval_steps_per_second": 41.026,
|
| 265 |
+
"step": 5088
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 16.946540880503143,
|
| 269 |
+
"grad_norm": 0.21620287001132965,
|
| 270 |
+
"learning_rate": 1.4620145752221225e-05,
|
| 271 |
+
"loss": 0.1132,
|
| 272 |
+
"step": 5389
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 17.0,
|
| 276 |
+
"eval_accuracy": 0.947741935483871,
|
| 277 |
+
"eval_loss": 0.20680664479732513,
|
| 278 |
+
"eval_runtime": 1.579,
|
| 279 |
+
"eval_samples_per_second": 1963.269,
|
| 280 |
+
"eval_steps_per_second": 41.165,
|
| 281 |
+
"step": 5406
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 17.943396226415093,
|
| 285 |
+
"grad_norm": 0.13292711973190308,
|
| 286 |
+
"learning_rate": 1.4303683737646003e-05,
|
| 287 |
+
"loss": 0.1121,
|
| 288 |
+
"step": 5706
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 18.0,
|
| 292 |
+
"eval_accuracy": 0.9493548387096774,
|
| 293 |
+
"eval_loss": 0.20597019791603088,
|
| 294 |
+
"eval_runtime": 1.5863,
|
| 295 |
+
"eval_samples_per_second": 1954.265,
|
| 296 |
+
"eval_steps_per_second": 40.977,
|
| 297 |
+
"step": 5724
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 18.940251572327043,
|
| 301 |
+
"grad_norm": 0.17331534624099731,
|
| 302 |
+
"learning_rate": 1.3987221723070781e-05,
|
| 303 |
+
"loss": 0.1112,
|
| 304 |
+
"step": 6023
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 19.0,
|
| 308 |
+
"eval_accuracy": 0.9490322580645161,
|
| 309 |
+
"eval_loss": 0.20577199757099152,
|
| 310 |
+
"eval_runtime": 1.593,
|
| 311 |
+
"eval_samples_per_second": 1946.01,
|
| 312 |
+
"eval_steps_per_second": 40.803,
|
| 313 |
+
"step": 6042
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 19.937106918238992,
|
| 317 |
+
"grad_norm": 0.179490864276886,
|
| 318 |
+
"learning_rate": 1.3670759708495558e-05,
|
| 319 |
+
"loss": 0.1098,
|
| 320 |
+
"step": 6340
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 20.0,
|
| 324 |
+
"eval_accuracy": 0.947741935483871,
|
| 325 |
+
"eval_loss": 0.20335346460342407,
|
| 326 |
+
"eval_runtime": 1.5965,
|
| 327 |
+
"eval_samples_per_second": 1941.713,
|
| 328 |
+
"eval_steps_per_second": 40.713,
|
| 329 |
+
"step": 6360
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 20.933962264150942,
|
| 333 |
+
"grad_norm": 0.1514139026403427,
|
| 334 |
+
"learning_rate": 1.3354297693920338e-05,
|
| 335 |
+
"loss": 0.1095,
|
| 336 |
+
"step": 6657
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 21.0,
|
| 340 |
+
"eval_accuracy": 0.9467741935483871,
|
| 341 |
+
"eval_loss": 0.2035306692123413,
|
| 342 |
+
"eval_runtime": 1.5835,
|
| 343 |
+
"eval_samples_per_second": 1957.701,
|
| 344 |
+
"eval_steps_per_second": 41.049,
|
| 345 |
+
"step": 6678
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 21.930817610062892,
|
| 349 |
+
"grad_norm": 0.13116957247257233,
|
| 350 |
+
"learning_rate": 1.3037835679345115e-05,
|
| 351 |
+
"loss": 0.1086,
|
| 352 |
+
"step": 6974
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 22.0,
|
| 356 |
+
"eval_accuracy": 0.9480645161290323,
|
| 357 |
+
"eval_loss": 0.20538054406642914,
|
| 358 |
+
"eval_runtime": 1.5892,
|
| 359 |
+
"eval_samples_per_second": 1950.626,
|
| 360 |
+
"eval_steps_per_second": 40.9,
|
| 361 |
+
"step": 6996
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 22.927672955974842,
|
| 365 |
+
"grad_norm": 0.13380903005599976,
|
| 366 |
+
"learning_rate": 1.2721373664769891e-05,
|
| 367 |
+
"loss": 0.1083,
|
| 368 |
+
"step": 7291
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 23.0,
|
| 372 |
+
"eval_accuracy": 0.9480645161290323,
|
| 373 |
+
"eval_loss": 0.2019556611776352,
|
| 374 |
+
"eval_runtime": 1.5887,
|
| 375 |
+
"eval_samples_per_second": 1951.225,
|
| 376 |
+
"eval_steps_per_second": 40.913,
|
| 377 |
+
"step": 7314
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 23.92452830188679,
|
| 381 |
+
"grad_norm": 0.1300133466720581,
|
| 382 |
+
"learning_rate": 1.2404911650194671e-05,
|
| 383 |
+
"loss": 0.1077,
|
| 384 |
+
"step": 7608
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 24.0,
|
| 388 |
+
"eval_accuracy": 0.9474193548387096,
|
| 389 |
+
"eval_loss": 0.2024136185646057,
|
| 390 |
+
"eval_runtime": 1.5799,
|
| 391 |
+
"eval_samples_per_second": 1962.185,
|
| 392 |
+
"eval_steps_per_second": 41.143,
|
| 393 |
+
"step": 7632
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 24.92138364779874,
|
| 397 |
+
"grad_norm": 0.14925934374332428,
|
| 398 |
+
"learning_rate": 1.2088449635619448e-05,
|
| 399 |
+
"loss": 0.107,
|
| 400 |
+
"step": 7925
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 25.0,
|
| 404 |
+
"eval_accuracy": 0.9480645161290323,
|
| 405 |
+
"eval_loss": 0.2033989429473877,
|
| 406 |
+
"eval_runtime": 1.5892,
|
| 407 |
+
"eval_samples_per_second": 1950.644,
|
| 408 |
+
"eval_steps_per_second": 40.901,
|
| 409 |
+
"step": 7950
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 25.91823899371069,
|
| 413 |
+
"grad_norm": 0.11126036942005157,
|
| 414 |
+
"learning_rate": 1.1771987621044224e-05,
|
| 415 |
+
"loss": 0.1068,
|
| 416 |
+
"step": 8242
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 26.0,
|
| 420 |
+
"eval_accuracy": 0.95,
|
| 421 |
+
"eval_loss": 0.20260857045650482,
|
| 422 |
+
"eval_runtime": 1.5864,
|
| 423 |
+
"eval_samples_per_second": 1954.067,
|
| 424 |
+
"eval_steps_per_second": 40.972,
|
| 425 |
+
"step": 8268
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 26.91509433962264,
|
| 429 |
+
"grad_norm": 0.12588472664356232,
|
| 430 |
+
"learning_rate": 1.1455525606469004e-05,
|
| 431 |
+
"loss": 0.1061,
|
| 432 |
+
"step": 8559
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 27.0,
|
| 436 |
+
"eval_accuracy": 0.9474193548387096,
|
| 437 |
+
"eval_loss": 0.2059621959924698,
|
| 438 |
+
"eval_runtime": 1.5915,
|
| 439 |
+
"eval_samples_per_second": 1947.843,
|
| 440 |
+
"eval_steps_per_second": 40.842,
|
| 441 |
+
"step": 8586
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 27.91194968553459,
|
| 445 |
+
"grad_norm": 0.11613517254590988,
|
| 446 |
+
"learning_rate": 1.1139063591893781e-05,
|
| 447 |
+
"loss": 0.1058,
|
| 448 |
+
"step": 8876
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 28.0,
|
| 452 |
+
"eval_accuracy": 0.9496774193548387,
|
| 453 |
+
"eval_loss": 0.20328474044799805,
|
| 454 |
+
"eval_runtime": 1.5805,
|
| 455 |
+
"eval_samples_per_second": 1961.399,
|
| 456 |
+
"eval_steps_per_second": 41.126,
|
| 457 |
+
"step": 8904
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 28.90880503144654,
|
| 461 |
+
"grad_norm": 0.10675258934497833,
|
| 462 |
+
"learning_rate": 1.082260157731856e-05,
|
| 463 |
+
"loss": 0.1054,
|
| 464 |
+
"step": 9193
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 29.0,
|
| 468 |
+
"eval_accuracy": 0.9493548387096774,
|
| 469 |
+
"eval_loss": 0.20192357897758484,
|
| 470 |
+
"eval_runtime": 1.5798,
|
| 471 |
+
"eval_samples_per_second": 1962.226,
|
| 472 |
+
"eval_steps_per_second": 41.143,
|
| 473 |
+
"step": 9222
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 29.90566037735849,
|
| 477 |
+
"grad_norm": 0.16315770149230957,
|
| 478 |
+
"learning_rate": 1.0506139562743338e-05,
|
| 479 |
+
"loss": 0.105,
|
| 480 |
+
"step": 9510
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 30.0,
|
| 484 |
+
"eval_accuracy": 0.9493548387096774,
|
| 485 |
+
"eval_loss": 0.19851729273796082,
|
| 486 |
+
"eval_runtime": 1.5919,
|
| 487 |
+
"eval_samples_per_second": 1947.418,
|
| 488 |
+
"eval_steps_per_second": 40.833,
|
| 489 |
+
"step": 9540
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 30.90251572327044,
|
| 493 |
+
"grad_norm": 0.1677229255437851,
|
| 494 |
+
"learning_rate": 1.0189677548168116e-05,
|
| 495 |
+
"loss": 0.1049,
|
| 496 |
+
"step": 9827
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 31.0,
|
| 500 |
+
"eval_accuracy": 0.947741935483871,
|
| 501 |
+
"eval_loss": 0.20081491768360138,
|
| 502 |
+
"eval_runtime": 1.5792,
|
| 503 |
+
"eval_samples_per_second": 1963.049,
|
| 504 |
+
"eval_steps_per_second": 41.161,
|
| 505 |
+
"step": 9858
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 31.89937106918239,
|
| 509 |
+
"grad_norm": 0.10275265574455261,
|
| 510 |
+
"learning_rate": 9.873215533592893e-06,
|
| 511 |
+
"loss": 0.1045,
|
| 512 |
+
"step": 10144
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 32.0,
|
| 516 |
+
"eval_accuracy": 0.947741935483871,
|
| 517 |
+
"eval_loss": 0.19913455843925476,
|
| 518 |
+
"eval_runtime": 1.5774,
|
| 519 |
+
"eval_samples_per_second": 1965.226,
|
| 520 |
+
"eval_steps_per_second": 41.206,
|
| 521 |
+
"step": 10176
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 32.89622641509434,
|
| 525 |
+
"grad_norm": 0.09303146600723267,
|
| 526 |
+
"learning_rate": 9.556753519017671e-06,
|
| 527 |
+
"loss": 0.1041,
|
| 528 |
+
"step": 10461
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 33.0,
|
| 532 |
+
"eval_accuracy": 0.9480645161290323,
|
| 533 |
+
"eval_loss": 0.2002706229686737,
|
| 534 |
+
"eval_runtime": 1.5858,
|
| 535 |
+
"eval_samples_per_second": 1954.895,
|
| 536 |
+
"eval_steps_per_second": 40.99,
|
| 537 |
+
"step": 10494
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 33.893081761006286,
|
| 541 |
+
"grad_norm": 0.10777819156646729,
|
| 542 |
+
"learning_rate": 9.24029150444245e-06,
|
| 543 |
+
"loss": 0.104,
|
| 544 |
+
"step": 10778
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 34.0,
|
| 548 |
+
"eval_accuracy": 0.9474193548387096,
|
| 549 |
+
"eval_loss": 0.20254099369049072,
|
| 550 |
+
"eval_runtime": 1.5867,
|
| 551 |
+
"eval_samples_per_second": 1953.68,
|
| 552 |
+
"eval_steps_per_second": 40.964,
|
| 553 |
+
"step": 10812
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 34.88993710691824,
|
| 557 |
+
"grad_norm": 0.12207500636577606,
|
| 558 |
+
"learning_rate": 8.923829489867226e-06,
|
| 559 |
+
"loss": 0.1039,
|
| 560 |
+
"step": 11095
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 35.0,
|
| 564 |
+
"eval_accuracy": 0.9503225806451613,
|
| 565 |
+
"eval_loss": 0.1988000124692917,
|
| 566 |
+
"eval_runtime": 1.5822,
|
| 567 |
+
"eval_samples_per_second": 1959.306,
|
| 568 |
+
"eval_steps_per_second": 41.082,
|
| 569 |
+
"step": 11130
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 35.886792452830186,
|
| 573 |
+
"grad_norm": 0.11546960473060608,
|
| 574 |
+
"learning_rate": 8.607367475292004e-06,
|
| 575 |
+
"loss": 0.1036,
|
| 576 |
+
"step": 11412
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 36.0,
|
| 580 |
+
"eval_accuracy": 0.9480645161290323,
|
| 581 |
+
"eval_loss": 0.2011515349149704,
|
| 582 |
+
"eval_runtime": 1.5802,
|
| 583 |
+
"eval_samples_per_second": 1961.837,
|
| 584 |
+
"eval_steps_per_second": 41.135,
|
| 585 |
+
"step": 11448
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 36.88364779874214,
|
| 589 |
+
"grad_norm": 0.11202769726514816,
|
| 590 |
+
"learning_rate": 8.290905460716781e-06,
|
| 591 |
+
"loss": 0.1037,
|
| 592 |
+
"step": 11729
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 37.0,
|
| 596 |
+
"eval_accuracy": 0.9483870967741935,
|
| 597 |
+
"eval_loss": 0.19893702864646912,
|
| 598 |
+
"eval_runtime": 1.5862,
|
| 599 |
+
"eval_samples_per_second": 1954.403,
|
| 600 |
+
"eval_steps_per_second": 40.979,
|
| 601 |
+
"step": 11766
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 37.880503144654085,
|
| 605 |
+
"grad_norm": 0.1050727367401123,
|
| 606 |
+
"learning_rate": 7.97444344614156e-06,
|
| 607 |
+
"loss": 0.1033,
|
| 608 |
+
"step": 12046
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 38.0,
|
| 612 |
+
"eval_accuracy": 0.9490322580645161,
|
| 613 |
+
"eval_loss": 0.1999415010213852,
|
| 614 |
+
"eval_runtime": 1.5888,
|
| 615 |
+
"eval_samples_per_second": 1951.155,
|
| 616 |
+
"eval_steps_per_second": 40.911,
|
| 617 |
+
"step": 12084
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 38.87735849056604,
|
| 621 |
+
"grad_norm": 0.11297059059143066,
|
| 622 |
+
"learning_rate": 7.657981431566338e-06,
|
| 623 |
+
"loss": 0.1028,
|
| 624 |
+
"step": 12363
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 39.0,
|
| 628 |
+
"eval_accuracy": 0.9487096774193549,
|
| 629 |
+
"eval_loss": 0.20117522776126862,
|
| 630 |
+
"eval_runtime": 1.5892,
|
| 631 |
+
"eval_samples_per_second": 1950.712,
|
| 632 |
+
"eval_steps_per_second": 40.902,
|
| 633 |
+
"step": 12402
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 39.874213836477985,
|
| 637 |
+
"grad_norm": 0.10458573698997498,
|
| 638 |
+
"learning_rate": 7.341519416991115e-06,
|
| 639 |
+
"loss": 0.1029,
|
| 640 |
+
"step": 12680
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 40.0,
|
| 644 |
+
"eval_accuracy": 0.9490322580645161,
|
| 645 |
+
"eval_loss": 0.1978650689125061,
|
| 646 |
+
"eval_runtime": 1.5862,
|
| 647 |
+
"eval_samples_per_second": 1954.371,
|
| 648 |
+
"eval_steps_per_second": 40.979,
|
| 649 |
+
"step": 12720
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 40.87106918238994,
|
| 653 |
+
"grad_norm": 0.1049540638923645,
|
| 654 |
+
"learning_rate": 7.025057402415893e-06,
|
| 655 |
+
"loss": 0.1027,
|
| 656 |
+
"step": 12997
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 41.0,
|
| 660 |
+
"eval_accuracy": 0.9496774193548387,
|
| 661 |
+
"eval_loss": 0.19904379546642303,
|
| 662 |
+
"eval_runtime": 1.5886,
|
| 663 |
+
"eval_samples_per_second": 1951.406,
|
| 664 |
+
"eval_steps_per_second": 40.917,
|
| 665 |
+
"step": 13038
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 41.867924528301884,
|
| 669 |
+
"grad_norm": 0.08443975448608398,
|
| 670 |
+
"learning_rate": 6.708595387840672e-06,
|
| 671 |
+
"loss": 0.1024,
|
| 672 |
+
"step": 13314
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 42.0,
|
| 676 |
+
"eval_accuracy": 0.947741935483871,
|
| 677 |
+
"eval_loss": 0.1994916945695877,
|
| 678 |
+
"eval_runtime": 1.5833,
|
| 679 |
+
"eval_samples_per_second": 1957.947,
|
| 680 |
+
"eval_steps_per_second": 41.054,
|
| 681 |
+
"step": 13356
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 42.86477987421384,
|
| 685 |
+
"grad_norm": 0.10657757520675659,
|
| 686 |
+
"learning_rate": 6.392133373265449e-06,
|
| 687 |
+
"loss": 0.1024,
|
| 688 |
+
"step": 13631
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 43.0,
|
| 692 |
+
"eval_accuracy": 0.9509677419354838,
|
| 693 |
+
"eval_loss": 0.19779588282108307,
|
| 694 |
+
"eval_runtime": 1.6024,
|
| 695 |
+
"eval_samples_per_second": 1934.587,
|
| 696 |
+
"eval_steps_per_second": 40.564,
|
| 697 |
+
"step": 13674
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 43.861635220125784,
|
| 701 |
+
"grad_norm": 0.09864117950201035,
|
| 702 |
+
"learning_rate": 6.0756713586902274e-06,
|
| 703 |
+
"loss": 0.1023,
|
| 704 |
+
"step": 13948
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 44.0,
|
| 708 |
+
"eval_accuracy": 0.9496774193548387,
|
| 709 |
+
"eval_loss": 0.1969560831785202,
|
| 710 |
+
"eval_runtime": 1.5862,
|
| 711 |
+
"eval_samples_per_second": 1954.296,
|
| 712 |
+
"eval_steps_per_second": 40.977,
|
| 713 |
+
"step": 13992
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 44.85849056603774,
|
| 717 |
+
"grad_norm": 0.09396997094154358,
|
| 718 |
+
"learning_rate": 5.759209344115006e-06,
|
| 719 |
+
"loss": 0.1022,
|
| 720 |
+
"step": 14265
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 45.0,
|
| 724 |
+
"eval_accuracy": 0.9487096774193549,
|
| 725 |
+
"eval_loss": 0.19702459871768951,
|
| 726 |
+
"eval_runtime": 1.6077,
|
| 727 |
+
"eval_samples_per_second": 1928.217,
|
| 728 |
+
"eval_steps_per_second": 40.43,
|
| 729 |
+
"step": 14310
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 45.855345911949684,
|
| 733 |
+
"grad_norm": 0.09406633675098419,
|
| 734 |
+
"learning_rate": 5.442747329539782e-06,
|
| 735 |
+
"loss": 0.1018,
|
| 736 |
+
"step": 14582
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 46.0,
|
| 740 |
+
"eval_accuracy": 0.95,
|
| 741 |
+
"eval_loss": 0.1965072900056839,
|
| 742 |
+
"eval_runtime": 1.5917,
|
| 743 |
+
"eval_samples_per_second": 1947.581,
|
| 744 |
+
"eval_steps_per_second": 40.836,
|
| 745 |
+
"step": 14628
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 46.85220125786164,
|
| 749 |
+
"grad_norm": 0.11108892410993576,
|
| 750 |
+
"learning_rate": 5.126285314964561e-06,
|
| 751 |
+
"loss": 0.1018,
|
| 752 |
+
"step": 14899
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 47.0,
|
| 756 |
+
"eval_accuracy": 0.9512903225806452,
|
| 757 |
+
"eval_loss": 0.1974688321352005,
|
| 758 |
+
"eval_runtime": 1.5999,
|
| 759 |
+
"eval_samples_per_second": 1937.681,
|
| 760 |
+
"eval_steps_per_second": 40.629,
|
| 761 |
+
"step": 14946
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 47.84905660377358,
|
| 765 |
+
"grad_norm": 0.09912719577550888,
|
| 766 |
+
"learning_rate": 4.809823300389338e-06,
|
| 767 |
+
"loss": 0.1017,
|
| 768 |
+
"step": 15216
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 48.0,
|
| 772 |
+
"eval_accuracy": 0.95,
|
| 773 |
+
"eval_loss": 0.19748900830745697,
|
| 774 |
+
"eval_runtime": 1.5807,
|
| 775 |
+
"eval_samples_per_second": 1961.164,
|
| 776 |
+
"eval_steps_per_second": 41.121,
|
| 777 |
+
"step": 15264
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 48.84591194968554,
|
| 781 |
+
"grad_norm": 0.10613682121038437,
|
| 782 |
+
"learning_rate": 4.4933612858141165e-06,
|
| 783 |
+
"loss": 0.1018,
|
| 784 |
+
"step": 15533
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 49.0,
|
| 788 |
+
"eval_accuracy": 0.9516129032258065,
|
| 789 |
+
"eval_loss": 0.19778111577033997,
|
| 790 |
+
"eval_runtime": 1.5919,
|
| 791 |
+
"eval_samples_per_second": 1947.314,
|
| 792 |
+
"eval_steps_per_second": 40.831,
|
| 793 |
+
"step": 15582
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 49.84276729559748,
|
| 797 |
+
"grad_norm": 0.12485523521900177,
|
| 798 |
+
"learning_rate": 4.176899271238895e-06,
|
| 799 |
+
"loss": 0.1014,
|
| 800 |
+
"step": 15850
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 50.0,
|
| 804 |
+
"eval_accuracy": 0.9509677419354838,
|
| 805 |
+
"eval_loss": 0.19711264967918396,
|
| 806 |
+
"eval_runtime": 1.5852,
|
| 807 |
+
"eval_samples_per_second": 1955.593,
|
| 808 |
+
"eval_steps_per_second": 41.004,
|
| 809 |
+
"step": 15900
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 50.839622641509436,
|
| 813 |
+
"grad_norm": 0.10598266869783401,
|
| 814 |
+
"learning_rate": 3.860437256663672e-06,
|
| 815 |
+
"loss": 0.1013,
|
| 816 |
+
"step": 16167
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 51.0,
|
| 820 |
+
"eval_accuracy": 0.9490322580645161,
|
| 821 |
+
"eval_loss": 0.19864656031131744,
|
| 822 |
+
"eval_runtime": 1.5959,
|
| 823 |
+
"eval_samples_per_second": 1942.538,
|
| 824 |
+
"eval_steps_per_second": 40.731,
|
| 825 |
+
"step": 16218
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 51.83647798742138,
|
| 829 |
+
"grad_norm": 0.09576704353094101,
|
| 830 |
+
"learning_rate": 3.5439752420884498e-06,
|
| 831 |
+
"loss": 0.1014,
|
| 832 |
+
"step": 16484
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 52.0,
|
| 836 |
+
"eval_accuracy": 0.9496774193548387,
|
| 837 |
+
"eval_loss": 0.1968703716993332,
|
| 838 |
+
"eval_runtime": 1.5849,
|
| 839 |
+
"eval_samples_per_second": 1955.923,
|
| 840 |
+
"eval_steps_per_second": 41.011,
|
| 841 |
+
"step": 16536
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 52.833333333333336,
|
| 845 |
+
"grad_norm": 0.08170927315950394,
|
| 846 |
+
"learning_rate": 3.2275132275132277e-06,
|
| 847 |
+
"loss": 0.1013,
|
| 848 |
+
"step": 16801
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 53.0,
|
| 852 |
+
"eval_accuracy": 0.9493548387096774,
|
| 853 |
+
"eval_loss": 0.19717678427696228,
|
| 854 |
+
"eval_runtime": 1.5789,
|
| 855 |
+
"eval_samples_per_second": 1963.405,
|
| 856 |
+
"eval_steps_per_second": 41.168,
|
| 857 |
+
"step": 16854
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 53.83018867924528,
|
| 861 |
+
"grad_norm": 0.10985760390758514,
|
| 862 |
+
"learning_rate": 2.911051212938006e-06,
|
| 863 |
+
"loss": 0.1012,
|
| 864 |
+
"step": 17118
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 54.0,
|
| 868 |
+
"eval_accuracy": 0.9496774193548387,
|
| 869 |
+
"eval_loss": 0.19691714644432068,
|
| 870 |
+
"eval_runtime": 1.5823,
|
| 871 |
+
"eval_samples_per_second": 1959.168,
|
| 872 |
+
"eval_steps_per_second": 41.079,
|
| 873 |
+
"step": 17172
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 54.827044025157235,
|
| 877 |
+
"grad_norm": 0.09639162570238113,
|
| 878 |
+
"learning_rate": 2.5945891983627835e-06,
|
| 879 |
+
"loss": 0.1012,
|
| 880 |
+
"step": 17435
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 55.0,
|
| 884 |
+
"eval_accuracy": 0.9503225806451613,
|
| 885 |
+
"eval_loss": 0.19801288843154907,
|
| 886 |
+
"eval_runtime": 1.5937,
|
| 887 |
+
"eval_samples_per_second": 1945.19,
|
| 888 |
+
"eval_steps_per_second": 40.786,
|
| 889 |
+
"step": 17490
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 55.82389937106918,
|
| 893 |
+
"grad_norm": 0.14052703976631165,
|
| 894 |
+
"learning_rate": 2.2781271837875614e-06,
|
| 895 |
+
"loss": 0.1012,
|
| 896 |
+
"step": 17752
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 56.0,
|
| 900 |
+
"eval_accuracy": 0.9493548387096774,
|
| 901 |
+
"eval_loss": 0.19718502461910248,
|
| 902 |
+
"eval_runtime": 1.5839,
|
| 903 |
+
"eval_samples_per_second": 1957.213,
|
| 904 |
+
"eval_steps_per_second": 41.038,
|
| 905 |
+
"step": 17808
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 56.820754716981135,
|
| 909 |
+
"grad_norm": 0.1034596860408783,
|
| 910 |
+
"learning_rate": 1.9616651692123393e-06,
|
| 911 |
+
"loss": 0.1008,
|
| 912 |
+
"step": 18069
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 57.0,
|
| 916 |
+
"eval_accuracy": 0.9503225806451613,
|
| 917 |
+
"eval_loss": 0.19776946306228638,
|
| 918 |
+
"eval_runtime": 1.5849,
|
| 919 |
+
"eval_samples_per_second": 1955.947,
|
| 920 |
+
"eval_steps_per_second": 41.012,
|
| 921 |
+
"step": 18126
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 57.81761006289308,
|
| 925 |
+
"grad_norm": 0.09913316369056702,
|
| 926 |
+
"learning_rate": 1.645203154637117e-06,
|
| 927 |
+
"loss": 0.101,
|
| 928 |
+
"step": 18386
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 58.0,
|
| 932 |
+
"eval_accuracy": 0.9506451612903226,
|
| 933 |
+
"eval_loss": 0.19688402116298676,
|
| 934 |
+
"eval_runtime": 1.5713,
|
| 935 |
+
"eval_samples_per_second": 1972.827,
|
| 936 |
+
"eval_steps_per_second": 41.366,
|
| 937 |
+
"step": 18444
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 58.814465408805034,
|
| 941 |
+
"grad_norm": 0.10723335295915604,
|
| 942 |
+
"learning_rate": 1.3287411400618948e-06,
|
| 943 |
+
"loss": 0.1008,
|
| 944 |
+
"step": 18703
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 59.0,
|
| 948 |
+
"eval_accuracy": 0.95,
|
| 949 |
+
"eval_loss": 0.1969434767961502,
|
| 950 |
+
"eval_runtime": 1.6034,
|
| 951 |
+
"eval_samples_per_second": 1933.451,
|
| 952 |
+
"eval_steps_per_second": 40.54,
|
| 953 |
+
"step": 18762
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 59.81132075471698,
|
| 957 |
+
"grad_norm": 0.09432651847600937,
|
| 958 |
+
"learning_rate": 1.0122791254866727e-06,
|
| 959 |
+
"loss": 0.101,
|
| 960 |
+
"step": 19020
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 60.0,
|
| 964 |
+
"eval_accuracy": 0.9503225806451613,
|
| 965 |
+
"eval_loss": 0.19751064479351044,
|
| 966 |
+
"eval_runtime": 1.5841,
|
| 967 |
+
"eval_samples_per_second": 1956.971,
|
| 968 |
+
"eval_steps_per_second": 41.033,
|
| 969 |
+
"step": 19080
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 60.808176100628934,
|
| 973 |
+
"grad_norm": 0.11219661682844162,
|
| 974 |
+
"learning_rate": 6.958171109114506e-07,
|
| 975 |
+
"loss": 0.1008,
|
| 976 |
+
"step": 19337
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 61.0,
|
| 980 |
+
"eval_accuracy": 0.9503225806451613,
|
| 981 |
+
"eval_loss": 0.1971895694732666,
|
| 982 |
+
"eval_runtime": 1.582,
|
| 983 |
+
"eval_samples_per_second": 1959.513,
|
| 984 |
+
"eval_steps_per_second": 41.087,
|
| 985 |
+
"step": 19398
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 61.80503144654088,
|
| 989 |
+
"grad_norm": 0.10517913103103638,
|
| 990 |
+
"learning_rate": 3.7935509633622843e-07,
|
| 991 |
+
"loss": 0.1009,
|
| 992 |
+
"step": 19654
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 62.0,
|
| 996 |
+
"eval_accuracy": 0.9506451612903226,
|
| 997 |
+
"eval_loss": 0.19690674543380737,
|
| 998 |
+
"eval_runtime": 1.5839,
|
| 999 |
+
"eval_samples_per_second": 1957.163,
|
| 1000 |
+
"eval_steps_per_second": 41.037,
|
| 1001 |
+
"step": 19716
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 62.801886792452834,
|
| 1005 |
+
"grad_norm": 0.10251986980438232,
|
| 1006 |
+
"learning_rate": 6.289308176100629e-08,
|
| 1007 |
+
"loss": 0.1008,
|
| 1008 |
+
"step": 19971
|
| 1009 |
+
}
|
| 1010 |
+
],
|
| 1011 |
+
"logging_steps": 317,
|
| 1012 |
+
"max_steps": 20034,
|
| 1013 |
+
"num_input_tokens_seen": 0,
|
| 1014 |
+
"num_train_epochs": 63,
|
| 1015 |
+
"save_steps": 1000000000.0,
|
| 1016 |
+
"stateful_callbacks": {
|
| 1017 |
+
"TrainerControl": {
|
| 1018 |
+
"args": {
|
| 1019 |
+
"should_epoch_stop": false,
|
| 1020 |
+
"should_evaluate": false,
|
| 1021 |
+
"should_log": false,
|
| 1022 |
+
"should_save": true,
|
| 1023 |
+
"should_training_stop": true
|
| 1024 |
+
},
|
| 1025 |
+
"attributes": {}
|
| 1026 |
+
}
|
| 1027 |
+
},
|
| 1028 |
+
"total_flos": 5194179875913420.0,
|
| 1029 |
+
"train_batch_size": 48,
|
| 1030 |
+
"trial_name": null,
|
| 1031 |
+
"trial_params": {
|
| 1032 |
+
"alpha": 0.4920075736426351,
|
| 1033 |
+
"num_train_epochs": 63,
|
| 1034 |
+
"temperature": 4
|
| 1035 |
+
}
|
| 1036 |
+
}
|
run-12/checkpoint-20034/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:27e5c70a607e87b053e47e411669b95ad09cdf8b3876ae9671e3e06c72f4f7f4
|
| 3 |
+
size 5304
|
run-12/checkpoint-20034/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
run-2/checkpoint-12402/config.json
ADDED
|
@@ -0,0 +1,331 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilbert-base-uncased",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"hidden_dim": 3072,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "restaurant_reviews",
|
| 13 |
+
"1": "nutrition_info",
|
| 14 |
+
"2": "account_blocked",
|
| 15 |
+
"3": "oil_change_how",
|
| 16 |
+
"4": "time",
|
| 17 |
+
"5": "weather",
|
| 18 |
+
"6": "redeem_rewards",
|
| 19 |
+
"7": "interest_rate",
|
| 20 |
+
"8": "gas_type",
|
| 21 |
+
"9": "accept_reservations",
|
| 22 |
+
"10": "smart_home",
|
| 23 |
+
"11": "user_name",
|
| 24 |
+
"12": "report_lost_card",
|
| 25 |
+
"13": "repeat",
|
| 26 |
+
"14": "whisper_mode",
|
| 27 |
+
"15": "what_are_your_hobbies",
|
| 28 |
+
"16": "order",
|
| 29 |
+
"17": "jump_start",
|
| 30 |
+
"18": "schedule_meeting",
|
| 31 |
+
"19": "meeting_schedule",
|
| 32 |
+
"20": "freeze_account",
|
| 33 |
+
"21": "what_song",
|
| 34 |
+
"22": "meaning_of_life",
|
| 35 |
+
"23": "restaurant_reservation",
|
| 36 |
+
"24": "traffic",
|
| 37 |
+
"25": "make_call",
|
| 38 |
+
"26": "text",
|
| 39 |
+
"27": "bill_balance",
|
| 40 |
+
"28": "improve_credit_score",
|
| 41 |
+
"29": "change_language",
|
| 42 |
+
"30": "no",
|
| 43 |
+
"31": "measurement_conversion",
|
| 44 |
+
"32": "timer",
|
| 45 |
+
"33": "flip_coin",
|
| 46 |
+
"34": "do_you_have_pets",
|
| 47 |
+
"35": "balance",
|
| 48 |
+
"36": "tell_joke",
|
| 49 |
+
"37": "last_maintenance",
|
| 50 |
+
"38": "exchange_rate",
|
| 51 |
+
"39": "uber",
|
| 52 |
+
"40": "car_rental",
|
| 53 |
+
"41": "credit_limit",
|
| 54 |
+
"42": "oos",
|
| 55 |
+
"43": "shopping_list",
|
| 56 |
+
"44": "expiration_date",
|
| 57 |
+
"45": "routing",
|
| 58 |
+
"46": "meal_suggestion",
|
| 59 |
+
"47": "tire_change",
|
| 60 |
+
"48": "todo_list",
|
| 61 |
+
"49": "card_declined",
|
| 62 |
+
"50": "rewards_balance",
|
| 63 |
+
"51": "change_accent",
|
| 64 |
+
"52": "vaccines",
|
| 65 |
+
"53": "reminder_update",
|
| 66 |
+
"54": "food_last",
|
| 67 |
+
"55": "change_ai_name",
|
| 68 |
+
"56": "bill_due",
|
| 69 |
+
"57": "who_do_you_work_for",
|
| 70 |
+
"58": "share_location",
|
| 71 |
+
"59": "international_visa",
|
| 72 |
+
"60": "calendar",
|
| 73 |
+
"61": "translate",
|
| 74 |
+
"62": "carry_on",
|
| 75 |
+
"63": "book_flight",
|
| 76 |
+
"64": "insurance_change",
|
| 77 |
+
"65": "todo_list_update",
|
| 78 |
+
"66": "timezone",
|
| 79 |
+
"67": "cancel_reservation",
|
| 80 |
+
"68": "transactions",
|
| 81 |
+
"69": "credit_score",
|
| 82 |
+
"70": "report_fraud",
|
| 83 |
+
"71": "spending_history",
|
| 84 |
+
"72": "directions",
|
| 85 |
+
"73": "spelling",
|
| 86 |
+
"74": "insurance",
|
| 87 |
+
"75": "what_is_your_name",
|
| 88 |
+
"76": "reminder",
|
| 89 |
+
"77": "where_are_you_from",
|
| 90 |
+
"78": "distance",
|
| 91 |
+
"79": "payday",
|
| 92 |
+
"80": "flight_status",
|
| 93 |
+
"81": "find_phone",
|
| 94 |
+
"82": "greeting",
|
| 95 |
+
"83": "alarm",
|
| 96 |
+
"84": "order_status",
|
| 97 |
+
"85": "confirm_reservation",
|
| 98 |
+
"86": "cook_time",
|
| 99 |
+
"87": "damaged_card",
|
| 100 |
+
"88": "reset_settings",
|
| 101 |
+
"89": "pin_change",
|
| 102 |
+
"90": "replacement_card_duration",
|
| 103 |
+
"91": "new_card",
|
| 104 |
+
"92": "roll_dice",
|
| 105 |
+
"93": "income",
|
| 106 |
+
"94": "taxes",
|
| 107 |
+
"95": "date",
|
| 108 |
+
"96": "who_made_you",
|
| 109 |
+
"97": "pto_request",
|
| 110 |
+
"98": "tire_pressure",
|
| 111 |
+
"99": "how_old_are_you",
|
| 112 |
+
"100": "rollover_401k",
|
| 113 |
+
"101": "pto_request_status",
|
| 114 |
+
"102": "how_busy",
|
| 115 |
+
"103": "application_status",
|
| 116 |
+
"104": "recipe",
|
| 117 |
+
"105": "calendar_update",
|
| 118 |
+
"106": "play_music",
|
| 119 |
+
"107": "yes",
|
| 120 |
+
"108": "direct_deposit",
|
| 121 |
+
"109": "credit_limit_change",
|
| 122 |
+
"110": "gas",
|
| 123 |
+
"111": "pay_bill",
|
| 124 |
+
"112": "ingredients_list",
|
| 125 |
+
"113": "lost_luggage",
|
| 126 |
+
"114": "goodbye",
|
| 127 |
+
"115": "what_can_i_ask_you",
|
| 128 |
+
"116": "book_hotel",
|
| 129 |
+
"117": "are_you_a_bot",
|
| 130 |
+
"118": "next_song",
|
| 131 |
+
"119": "change_speed",
|
| 132 |
+
"120": "plug_type",
|
| 133 |
+
"121": "maybe",
|
| 134 |
+
"122": "w2",
|
| 135 |
+
"123": "oil_change_when",
|
| 136 |
+
"124": "thank_you",
|
| 137 |
+
"125": "shopping_list_update",
|
| 138 |
+
"126": "pto_balance",
|
| 139 |
+
"127": "order_checks",
|
| 140 |
+
"128": "travel_alert",
|
| 141 |
+
"129": "fun_fact",
|
| 142 |
+
"130": "sync_device",
|
| 143 |
+
"131": "schedule_maintenance",
|
| 144 |
+
"132": "apr",
|
| 145 |
+
"133": "transfer",
|
| 146 |
+
"134": "ingredient_substitution",
|
| 147 |
+
"135": "calories",
|
| 148 |
+
"136": "current_location",
|
| 149 |
+
"137": "international_fees",
|
| 150 |
+
"138": "calculator",
|
| 151 |
+
"139": "definition",
|
| 152 |
+
"140": "next_holiday",
|
| 153 |
+
"141": "update_playlist",
|
| 154 |
+
"142": "mpg",
|
| 155 |
+
"143": "min_payment",
|
| 156 |
+
"144": "change_user_name",
|
| 157 |
+
"145": "restaurant_suggestion",
|
| 158 |
+
"146": "travel_notification",
|
| 159 |
+
"147": "cancel",
|
| 160 |
+
"148": "pto_used",
|
| 161 |
+
"149": "travel_suggestion",
|
| 162 |
+
"150": "change_volume"
|
| 163 |
+
},
|
| 164 |
+
"initializer_range": 0.02,
|
| 165 |
+
"label2id": {
|
| 166 |
+
"accept_reservations": 9,
|
| 167 |
+
"account_blocked": 2,
|
| 168 |
+
"alarm": 83,
|
| 169 |
+
"application_status": 103,
|
| 170 |
+
"apr": 132,
|
| 171 |
+
"are_you_a_bot": 117,
|
| 172 |
+
"balance": 35,
|
| 173 |
+
"bill_balance": 27,
|
| 174 |
+
"bill_due": 56,
|
| 175 |
+
"book_flight": 63,
|
| 176 |
+
"book_hotel": 116,
|
| 177 |
+
"calculator": 138,
|
| 178 |
+
"calendar": 60,
|
| 179 |
+
"calendar_update": 105,
|
| 180 |
+
"calories": 135,
|
| 181 |
+
"cancel": 147,
|
| 182 |
+
"cancel_reservation": 67,
|
| 183 |
+
"car_rental": 40,
|
| 184 |
+
"card_declined": 49,
|
| 185 |
+
"carry_on": 62,
|
| 186 |
+
"change_accent": 51,
|
| 187 |
+
"change_ai_name": 55,
|
| 188 |
+
"change_language": 29,
|
| 189 |
+
"change_speed": 119,
|
| 190 |
+
"change_user_name": 144,
|
| 191 |
+
"change_volume": 150,
|
| 192 |
+
"confirm_reservation": 85,
|
| 193 |
+
"cook_time": 86,
|
| 194 |
+
"credit_limit": 41,
|
| 195 |
+
"credit_limit_change": 109,
|
| 196 |
+
"credit_score": 69,
|
| 197 |
+
"current_location": 136,
|
| 198 |
+
"damaged_card": 87,
|
| 199 |
+
"date": 95,
|
| 200 |
+
"definition": 139,
|
| 201 |
+
"direct_deposit": 108,
|
| 202 |
+
"directions": 72,
|
| 203 |
+
"distance": 78,
|
| 204 |
+
"do_you_have_pets": 34,
|
| 205 |
+
"exchange_rate": 38,
|
| 206 |
+
"expiration_date": 44,
|
| 207 |
+
"find_phone": 81,
|
| 208 |
+
"flight_status": 80,
|
| 209 |
+
"flip_coin": 33,
|
| 210 |
+
"food_last": 54,
|
| 211 |
+
"freeze_account": 20,
|
| 212 |
+
"fun_fact": 129,
|
| 213 |
+
"gas": 110,
|
| 214 |
+
"gas_type": 8,
|
| 215 |
+
"goodbye": 114,
|
| 216 |
+
"greeting": 82,
|
| 217 |
+
"how_busy": 102,
|
| 218 |
+
"how_old_are_you": 99,
|
| 219 |
+
"improve_credit_score": 28,
|
| 220 |
+
"income": 93,
|
| 221 |
+
"ingredient_substitution": 134,
|
| 222 |
+
"ingredients_list": 112,
|
| 223 |
+
"insurance": 74,
|
| 224 |
+
"insurance_change": 64,
|
| 225 |
+
"interest_rate": 7,
|
| 226 |
+
"international_fees": 137,
|
| 227 |
+
"international_visa": 59,
|
| 228 |
+
"jump_start": 17,
|
| 229 |
+
"last_maintenance": 37,
|
| 230 |
+
"lost_luggage": 113,
|
| 231 |
+
"make_call": 25,
|
| 232 |
+
"maybe": 121,
|
| 233 |
+
"meal_suggestion": 46,
|
| 234 |
+
"meaning_of_life": 22,
|
| 235 |
+
"measurement_conversion": 31,
|
| 236 |
+
"meeting_schedule": 19,
|
| 237 |
+
"min_payment": 143,
|
| 238 |
+
"mpg": 142,
|
| 239 |
+
"new_card": 91,
|
| 240 |
+
"next_holiday": 140,
|
| 241 |
+
"next_song": 118,
|
| 242 |
+
"no": 30,
|
| 243 |
+
"nutrition_info": 1,
|
| 244 |
+
"oil_change_how": 3,
|
| 245 |
+
"oil_change_when": 123,
|
| 246 |
+
"oos": 42,
|
| 247 |
+
"order": 16,
|
| 248 |
+
"order_checks": 127,
|
| 249 |
+
"order_status": 84,
|
| 250 |
+
"pay_bill": 111,
|
| 251 |
+
"payday": 79,
|
| 252 |
+
"pin_change": 89,
|
| 253 |
+
"play_music": 106,
|
| 254 |
+
"plug_type": 120,
|
| 255 |
+
"pto_balance": 126,
|
| 256 |
+
"pto_request": 97,
|
| 257 |
+
"pto_request_status": 101,
|
| 258 |
+
"pto_used": 148,
|
| 259 |
+
"recipe": 104,
|
| 260 |
+
"redeem_rewards": 6,
|
| 261 |
+
"reminder": 76,
|
| 262 |
+
"reminder_update": 53,
|
| 263 |
+
"repeat": 13,
|
| 264 |
+
"replacement_card_duration": 90,
|
| 265 |
+
"report_fraud": 70,
|
| 266 |
+
"report_lost_card": 12,
|
| 267 |
+
"reset_settings": 88,
|
| 268 |
+
"restaurant_reservation": 23,
|
| 269 |
+
"restaurant_reviews": 0,
|
| 270 |
+
"restaurant_suggestion": 145,
|
| 271 |
+
"rewards_balance": 50,
|
| 272 |
+
"roll_dice": 92,
|
| 273 |
+
"rollover_401k": 100,
|
| 274 |
+
"routing": 45,
|
| 275 |
+
"schedule_maintenance": 131,
|
| 276 |
+
"schedule_meeting": 18,
|
| 277 |
+
"share_location": 58,
|
| 278 |
+
"shopping_list": 43,
|
| 279 |
+
"shopping_list_update": 125,
|
| 280 |
+
"smart_home": 10,
|
| 281 |
+
"spelling": 73,
|
| 282 |
+
"spending_history": 71,
|
| 283 |
+
"sync_device": 130,
|
| 284 |
+
"taxes": 94,
|
| 285 |
+
"tell_joke": 36,
|
| 286 |
+
"text": 26,
|
| 287 |
+
"thank_you": 124,
|
| 288 |
+
"time": 4,
|
| 289 |
+
"timer": 32,
|
| 290 |
+
"timezone": 66,
|
| 291 |
+
"tire_change": 47,
|
| 292 |
+
"tire_pressure": 98,
|
| 293 |
+
"todo_list": 48,
|
| 294 |
+
"todo_list_update": 65,
|
| 295 |
+
"traffic": 24,
|
| 296 |
+
"transactions": 68,
|
| 297 |
+
"transfer": 133,
|
| 298 |
+
"translate": 61,
|
| 299 |
+
"travel_alert": 128,
|
| 300 |
+
"travel_notification": 146,
|
| 301 |
+
"travel_suggestion": 149,
|
| 302 |
+
"uber": 39,
|
| 303 |
+
"update_playlist": 141,
|
| 304 |
+
"user_name": 11,
|
| 305 |
+
"vaccines": 52,
|
| 306 |
+
"w2": 122,
|
| 307 |
+
"weather": 5,
|
| 308 |
+
"what_are_your_hobbies": 15,
|
| 309 |
+
"what_can_i_ask_you": 115,
|
| 310 |
+
"what_is_your_name": 75,
|
| 311 |
+
"what_song": 21,
|
| 312 |
+
"where_are_you_from": 77,
|
| 313 |
+
"whisper_mode": 14,
|
| 314 |
+
"who_do_you_work_for": 57,
|
| 315 |
+
"who_made_you": 96,
|
| 316 |
+
"yes": 107
|
| 317 |
+
},
|
| 318 |
+
"max_position_embeddings": 512,
|
| 319 |
+
"model_type": "distilbert",
|
| 320 |
+
"n_heads": 12,
|
| 321 |
+
"n_layers": 6,
|
| 322 |
+
"pad_token_id": 0,
|
| 323 |
+
"problem_type": "single_label_classification",
|
| 324 |
+
"qa_dropout": 0.1,
|
| 325 |
+
"seq_classif_dropout": 0.2,
|
| 326 |
+
"sinusoidal_pos_embds": false,
|
| 327 |
+
"tie_weights_": true,
|
| 328 |
+
"torch_dtype": "float32",
|
| 329 |
+
"transformers_version": "4.44.2",
|
| 330 |
+
"vocab_size": 30522
|
| 331 |
+
}
|
run-2/checkpoint-12402/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad3a7f5ea7ebc74dc8716d8f00c605df72236c5c013949d1aea504f2db8ae81f
|
| 3 |
+
size 268290900
|
run-2/checkpoint-12402/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:241ceac887a0e24c2eeb83c59a904fdaf4ea2410a1dd2a9359dadb728a4a0e12
|
| 3 |
+
size 536643898
|
run-2/checkpoint-12402/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5ea40b3ef1c8cbccb37bb4ed27c88154d280adaa70c2f4b06de9a9a00f45c4c
|
| 3 |
+
size 14244
|
run-2/checkpoint-12402/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c2cc569715520ce44b4e7b901367f7e4bb95b8f9aabf6d6fb54daa7769f31e30
|
| 3 |
+
size 1064
|