Training in progress, step 5088
Browse files- model.safetensors +1 -1
- run-21/checkpoint-5088/config.json +331 -0
- run-21/checkpoint-5088/model.safetensors +3 -0
- run-21/checkpoint-5088/optimizer.pt +3 -0
- run-21/checkpoint-5088/rng_state.pth +3 -0
- run-21/checkpoint-5088/scheduler.pt +3 -0
- run-21/checkpoint-5088/special_tokens_map.json +7 -0
- run-21/checkpoint-5088/tokenizer.json +0 -0
- run-21/checkpoint-5088/tokenizer_config.json +55 -0
- run-21/checkpoint-5088/trainer_state.json +289 -0
- run-21/checkpoint-5088/training_args.bin +3 -0
- run-21/checkpoint-5088/vocab.txt +0 -0
- runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896107.01f54d6a8452.2040.17 +3 -0
- runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896139.01f54d6a8452.2040.18 +3 -0
- runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896172.01f54d6a8452.2040.19 +3 -0
- runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896203.01f54d6a8452.2040.20 +3 -0
- runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896235.01f54d6a8452.2040.21 +3 -0
- runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896268.01f54d6a8452.2040.22 +3 -0
- training_args.bin +1 -1
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 sha256:39ecc31dc5f9430ee6f498bdebf989ed5d65345605b61fcef7b129235f2f6de7
|
3 |
size 268290900
|
run-21/checkpoint-5088/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-21/checkpoint-5088/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:39ecc31dc5f9430ee6f498bdebf989ed5d65345605b61fcef7b129235f2f6de7
|
3 |
+
size 268290900
|
run-21/checkpoint-5088/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6ec43a00deeb8c1451b43355e5104cd1232f3cc0a93bab528bc66fe5556a50a7
|
3 |
+
size 536643898
|
run-21/checkpoint-5088/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3d40d17e5dd44e9069da6c48306436b7bd4a677df9d75342ab223ba35161a15b
|
3 |
+
size 14244
|
run-21/checkpoint-5088/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bab0e379373c55b14cac2ecf773c3afc3919c4cd89d531acf3dbf12b35e89bb4
|
3 |
+
size 1064
|
run-21/checkpoint-5088/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-21/checkpoint-5088/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
run-21/checkpoint-5088/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-21/checkpoint-5088/trainer_state.json
ADDED
@@ -0,0 +1,289 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 16.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 5088,
|
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": 0.3145744502544403,
|
14 |
+
"learning_rate": 0.00013773896055969116,
|
15 |
+
"loss": 0.2562,
|
16 |
+
"step": 317
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 1.0,
|
20 |
+
"eval_accuracy": 0.8893548387096775,
|
21 |
+
"eval_loss": 0.05541645735502243,
|
22 |
+
"eval_runtime": 2.6126,
|
23 |
+
"eval_samples_per_second": 1186.566,
|
24 |
+
"eval_steps_per_second": 24.88,
|
25 |
+
"step": 318
|
26 |
+
},
|
27 |
+
{
|
28 |
+
"epoch": 1.9937106918238994,
|
29 |
+
"grad_norm": 0.2050488293170929,
|
30 |
+
"learning_rate": 0.00012858715789831575,
|
31 |
+
"loss": 0.0488,
|
32 |
+
"step": 634
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 2.0,
|
36 |
+
"eval_accuracy": 0.9303225806451613,
|
37 |
+
"eval_loss": 0.030376020818948746,
|
38 |
+
"eval_runtime": 2.6008,
|
39 |
+
"eval_samples_per_second": 1191.932,
|
40 |
+
"eval_steps_per_second": 24.992,
|
41 |
+
"step": 636
|
42 |
+
},
|
43 |
+
{
|
44 |
+
"epoch": 2.990566037735849,
|
45 |
+
"grad_norm": 0.20619478821754456,
|
46 |
+
"learning_rate": 0.00011943535523694032,
|
47 |
+
"loss": 0.0305,
|
48 |
+
"step": 951
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"epoch": 3.0,
|
52 |
+
"eval_accuracy": 0.9383870967741935,
|
53 |
+
"eval_loss": 0.023558245971798897,
|
54 |
+
"eval_runtime": 2.6497,
|
55 |
+
"eval_samples_per_second": 1169.948,
|
56 |
+
"eval_steps_per_second": 24.531,
|
57 |
+
"step": 954
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 3.9874213836477987,
|
61 |
+
"grad_norm": 0.13936841487884521,
|
62 |
+
"learning_rate": 0.00011028355257556492,
|
63 |
+
"loss": 0.0246,
|
64 |
+
"step": 1268
|
65 |
+
},
|
66 |
+
{
|
67 |
+
"epoch": 4.0,
|
68 |
+
"eval_accuracy": 0.9390322580645162,
|
69 |
+
"eval_loss": 0.02127992734313011,
|
70 |
+
"eval_runtime": 2.6543,
|
71 |
+
"eval_samples_per_second": 1167.921,
|
72 |
+
"eval_steps_per_second": 24.489,
|
73 |
+
"step": 1272
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 4.984276729559748,
|
77 |
+
"grad_norm": 0.09692877531051636,
|
78 |
+
"learning_rate": 0.0001011317499141895,
|
79 |
+
"loss": 0.0215,
|
80 |
+
"step": 1585
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 5.0,
|
84 |
+
"eval_accuracy": 0.9403225806451613,
|
85 |
+
"eval_loss": 0.01949879713356495,
|
86 |
+
"eval_runtime": 2.6309,
|
87 |
+
"eval_samples_per_second": 1178.314,
|
88 |
+
"eval_steps_per_second": 24.707,
|
89 |
+
"step": 1590
|
90 |
+
},
|
91 |
+
{
|
92 |
+
"epoch": 5.981132075471698,
|
93 |
+
"grad_norm": 0.09292752295732498,
|
94 |
+
"learning_rate": 9.197994725281409e-05,
|
95 |
+
"loss": 0.0197,
|
96 |
+
"step": 1902
|
97 |
+
},
|
98 |
+
{
|
99 |
+
"epoch": 6.0,
|
100 |
+
"eval_accuracy": 0.9435483870967742,
|
101 |
+
"eval_loss": 0.01897437684237957,
|
102 |
+
"eval_runtime": 2.6528,
|
103 |
+
"eval_samples_per_second": 1168.577,
|
104 |
+
"eval_steps_per_second": 24.502,
|
105 |
+
"step": 1908
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 6.977987421383648,
|
109 |
+
"grad_norm": 0.07617080211639404,
|
110 |
+
"learning_rate": 8.282814459143868e-05,
|
111 |
+
"loss": 0.0183,
|
112 |
+
"step": 2219
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"epoch": 7.0,
|
116 |
+
"eval_accuracy": 0.9441935483870968,
|
117 |
+
"eval_loss": 0.018397612497210503,
|
118 |
+
"eval_runtime": 2.6546,
|
119 |
+
"eval_samples_per_second": 1167.789,
|
120 |
+
"eval_steps_per_second": 24.486,
|
121 |
+
"step": 2226
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 7.9748427672955975,
|
125 |
+
"grad_norm": 0.07391500473022461,
|
126 |
+
"learning_rate": 7.367634193006326e-05,
|
127 |
+
"loss": 0.0171,
|
128 |
+
"step": 2536
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 8.0,
|
132 |
+
"eval_accuracy": 0.9461290322580646,
|
133 |
+
"eval_loss": 0.017443224787712097,
|
134 |
+
"eval_runtime": 2.6528,
|
135 |
+
"eval_samples_per_second": 1168.563,
|
136 |
+
"eval_steps_per_second": 24.502,
|
137 |
+
"step": 2544
|
138 |
+
},
|
139 |
+
{
|
140 |
+
"epoch": 8.971698113207546,
|
141 |
+
"grad_norm": 0.09221338480710983,
|
142 |
+
"learning_rate": 6.452453926868784e-05,
|
143 |
+
"loss": 0.0162,
|
144 |
+
"step": 2853
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 9.0,
|
148 |
+
"eval_accuracy": 0.9435483870967742,
|
149 |
+
"eval_loss": 0.017002489417791367,
|
150 |
+
"eval_runtime": 2.6461,
|
151 |
+
"eval_samples_per_second": 1171.522,
|
152 |
+
"eval_steps_per_second": 24.564,
|
153 |
+
"step": 2862
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 9.968553459119496,
|
157 |
+
"grad_norm": 0.07394809275865555,
|
158 |
+
"learning_rate": 5.537273660731243e-05,
|
159 |
+
"loss": 0.0155,
|
160 |
+
"step": 3170
|
161 |
+
},
|
162 |
+
{
|
163 |
+
"epoch": 10.0,
|
164 |
+
"eval_accuracy": 0.9429032258064516,
|
165 |
+
"eval_loss": 0.01677101105451584,
|
166 |
+
"eval_runtime": 2.635,
|
167 |
+
"eval_samples_per_second": 1176.456,
|
168 |
+
"eval_steps_per_second": 24.668,
|
169 |
+
"step": 3180
|
170 |
+
},
|
171 |
+
{
|
172 |
+
"epoch": 10.965408805031446,
|
173 |
+
"grad_norm": 0.0590691976249218,
|
174 |
+
"learning_rate": 4.622093394593702e-05,
|
175 |
+
"loss": 0.0149,
|
176 |
+
"step": 3487
|
177 |
+
},
|
178 |
+
{
|
179 |
+
"epoch": 11.0,
|
180 |
+
"eval_accuracy": 0.9464516129032258,
|
181 |
+
"eval_loss": 0.016378937289118767,
|
182 |
+
"eval_runtime": 2.6318,
|
183 |
+
"eval_samples_per_second": 1177.923,
|
184 |
+
"eval_steps_per_second": 24.698,
|
185 |
+
"step": 3498
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 11.962264150943396,
|
189 |
+
"grad_norm": 0.05711551755666733,
|
190 |
+
"learning_rate": 3.7069131284561605e-05,
|
191 |
+
"loss": 0.0143,
|
192 |
+
"step": 3804
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 12.0,
|
196 |
+
"eval_accuracy": 0.9461290322580646,
|
197 |
+
"eval_loss": 0.015853669494390488,
|
198 |
+
"eval_runtime": 2.6492,
|
199 |
+
"eval_samples_per_second": 1170.169,
|
200 |
+
"eval_steps_per_second": 24.536,
|
201 |
+
"step": 3816
|
202 |
+
},
|
203 |
+
{
|
204 |
+
"epoch": 12.959119496855346,
|
205 |
+
"grad_norm": 0.05840861052274704,
|
206 |
+
"learning_rate": 2.7917328623186196e-05,
|
207 |
+
"loss": 0.0137,
|
208 |
+
"step": 4121
|
209 |
+
},
|
210 |
+
{
|
211 |
+
"epoch": 13.0,
|
212 |
+
"eval_accuracy": 0.9461290322580646,
|
213 |
+
"eval_loss": 0.015562719665467739,
|
214 |
+
"eval_runtime": 2.6419,
|
215 |
+
"eval_samples_per_second": 1173.405,
|
216 |
+
"eval_steps_per_second": 24.604,
|
217 |
+
"step": 4134
|
218 |
+
},
|
219 |
+
{
|
220 |
+
"epoch": 13.955974842767295,
|
221 |
+
"grad_norm": 0.0555766336619854,
|
222 |
+
"learning_rate": 1.8765525961810784e-05,
|
223 |
+
"loss": 0.0133,
|
224 |
+
"step": 4438
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 14.0,
|
228 |
+
"eval_accuracy": 0.9467741935483871,
|
229 |
+
"eval_loss": 0.015005242079496384,
|
230 |
+
"eval_runtime": 2.6572,
|
231 |
+
"eval_samples_per_second": 1166.643,
|
232 |
+
"eval_steps_per_second": 24.462,
|
233 |
+
"step": 4452
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 14.952830188679245,
|
237 |
+
"grad_norm": 0.05531783401966095,
|
238 |
+
"learning_rate": 9.61372330043537e-06,
|
239 |
+
"loss": 0.0129,
|
240 |
+
"step": 4755
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 15.0,
|
244 |
+
"eval_accuracy": 0.9461290322580646,
|
245 |
+
"eval_loss": 0.01490978617221117,
|
246 |
+
"eval_runtime": 2.6507,
|
247 |
+
"eval_samples_per_second": 1169.499,
|
248 |
+
"eval_steps_per_second": 24.522,
|
249 |
+
"step": 4770
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 15.949685534591195,
|
253 |
+
"grad_norm": 0.05344949662685394,
|
254 |
+
"learning_rate": 4.6192063905995774e-07,
|
255 |
+
"loss": 0.0125,
|
256 |
+
"step": 5072
|
257 |
+
}
|
258 |
+
],
|
259 |
+
"logging_steps": 317,
|
260 |
+
"max_steps": 5088,
|
261 |
+
"num_input_tokens_seen": 0,
|
262 |
+
"num_train_epochs": 16,
|
263 |
+
"save_steps": 1000000000.0,
|
264 |
+
"stateful_callbacks": {
|
265 |
+
"TrainerControl": {
|
266 |
+
"args": {
|
267 |
+
"should_epoch_stop": false,
|
268 |
+
"should_evaluate": false,
|
269 |
+
"should_log": false,
|
270 |
+
"should_save": true,
|
271 |
+
"should_training_stop": true
|
272 |
+
},
|
273 |
+
"attributes": {}
|
274 |
+
}
|
275 |
+
},
|
276 |
+
"total_flos": 1319419533829824.0,
|
277 |
+
"train_batch_size": 48,
|
278 |
+
"trial_name": null,
|
279 |
+
"trial_params": {
|
280 |
+
"alpha": 0.8496454567338295,
|
281 |
+
"fp16": false,
|
282 |
+
"learning_rate": 0.0001430799179488219,
|
283 |
+
"lr_scheduler": "cosine",
|
284 |
+
"num_train_epochs": 16,
|
285 |
+
"temperature": 12,
|
286 |
+
"warmup_steps": 132,
|
287 |
+
"weight_decay": 0.20337985521529534
|
288 |
+
}
|
289 |
+
}
|
run-21/checkpoint-5088/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:13a23c6b5c1e288fd01c7592b7c0967906736786d63802f1047b84af4a3822d5
|
3 |
+
size 5240
|
run-21/checkpoint-5088/vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896107.01f54d6a8452.2040.17
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6e1f1904d0dc4b29e1ae8973bb9915ea66d60a10ffe505d3df136c34cf4749dc
|
3 |
+
size 13022
|
runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896139.01f54d6a8452.2040.18
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cb21712f0fbce11a8b246856dcd517d11ae2623eed19d8fc18a617f519ca3738
|
3 |
+
size 13021
|
runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896172.01f54d6a8452.2040.19
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:45b0677868910ce767464be513df45e663d2377be8ef534731360c377a4cbc66
|
3 |
+
size 13020
|
runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896203.01f54d6a8452.2040.20
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d5fc61841adf42b7d500b29cc64d7c731fe3b5b40f6a3fa6e13d859a5a088abd
|
3 |
+
size 13021
|
runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896235.01f54d6a8452.2040.21
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ed5a9e10f7efeb851a18035168f2d2923c2ab750f083baa8546eb498916be286
|
3 |
+
size 13021
|
runs/Oct02_17-16-46_01f54d6a8452/events.out.tfevents.1727896268.01f54d6a8452.2040.22
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:03e5deb8d843038dfb3f5252827973279f9b4e425efec7a339631d23282046d2
|
3 |
+
size 21031
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5240
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:13a23c6b5c1e288fd01c7592b7c0967906736786d63802f1047b84af4a3822d5
|
3 |
size 5240
|