vedantjumle
commited on
Commit
·
7b64897
1
Parent(s):
4b975c9
Training in progress epoch 0
Browse files- README.md +55 -0
- config.json +343 -0
- merges.txt +0 -0
- special_tokens_map.json +6 -0
- tf_model.h5 +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +9 -0
- vocab.json +0 -0
README.md
ADDED
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
base_model: gpt2
|
4 |
+
tags:
|
5 |
+
- generated_from_keras_callback
|
6 |
+
model-index:
|
7 |
+
- name: vedantjumle/indo-ml-final-test-gpt2
|
8 |
+
results: []
|
9 |
+
---
|
10 |
+
|
11 |
+
<!-- This model card has been generated automatically according to the information Keras had access to. You should
|
12 |
+
probably proofread and complete it, then remove this comment. -->
|
13 |
+
|
14 |
+
# vedantjumle/indo-ml-final-test-gpt2
|
15 |
+
|
16 |
+
This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on an unknown dataset.
|
17 |
+
It achieves the following results on the evaluation set:
|
18 |
+
- Train Loss: 6.1590
|
19 |
+
- Validation Loss: 5.0940
|
20 |
+
- Train Accuracy: 0.01
|
21 |
+
- Epoch: 0
|
22 |
+
|
23 |
+
## Model description
|
24 |
+
|
25 |
+
More information needed
|
26 |
+
|
27 |
+
## Intended uses & limitations
|
28 |
+
|
29 |
+
More information needed
|
30 |
+
|
31 |
+
## Training and evaluation data
|
32 |
+
|
33 |
+
More information needed
|
34 |
+
|
35 |
+
## Training procedure
|
36 |
+
|
37 |
+
### Training hyperparameters
|
38 |
+
|
39 |
+
The following hyperparameters were used during training:
|
40 |
+
- optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 3000, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False}
|
41 |
+
- training_precision: float32
|
42 |
+
|
43 |
+
### Training results
|
44 |
+
|
45 |
+
| Train Loss | Validation Loss | Train Accuracy | Epoch |
|
46 |
+
|:----------:|:---------------:|:--------------:|:-----:|
|
47 |
+
| 6.1590 | 5.0940 | 0.01 | 0 |
|
48 |
+
|
49 |
+
|
50 |
+
### Framework versions
|
51 |
+
|
52 |
+
- Transformers 4.33.2
|
53 |
+
- TensorFlow 2.13.0
|
54 |
+
- Datasets 2.14.5
|
55 |
+
- Tokenizers 0.13.3
|
config.json
ADDED
@@ -0,0 +1,343 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "gpt2",
|
3 |
+
"activation_function": "gelu_new",
|
4 |
+
"architectures": [
|
5 |
+
"GPT2ForSequenceClassification"
|
6 |
+
],
|
7 |
+
"attn_pdrop": 0.1,
|
8 |
+
"bos_token_id": 50256,
|
9 |
+
"embd_pdrop": 0.1,
|
10 |
+
"eos_token_id": 50256,
|
11 |
+
"id2label": {
|
12 |
+
"accept reservations": 0,
|
13 |
+
"account blocked": 1,
|
14 |
+
"alarm": 2,
|
15 |
+
"application status": 3,
|
16 |
+
"apr": 4,
|
17 |
+
"are you a bot": 5,
|
18 |
+
"balance": 6,
|
19 |
+
"bill balance": 7,
|
20 |
+
"bill due": 8,
|
21 |
+
"book flight": 9,
|
22 |
+
"book hotel": 10,
|
23 |
+
"calculator": 11,
|
24 |
+
"calendar": 12,
|
25 |
+
"calendar update": 13,
|
26 |
+
"calories": 14,
|
27 |
+
"cancel": 15,
|
28 |
+
"cancel reservation": 16,
|
29 |
+
"car rental": 17,
|
30 |
+
"card declined": 18,
|
31 |
+
"carry on": 19,
|
32 |
+
"change accent": 20,
|
33 |
+
"change ai name": 21,
|
34 |
+
"change language": 22,
|
35 |
+
"change speed": 23,
|
36 |
+
"change user name": 24,
|
37 |
+
"change volume": 25,
|
38 |
+
"confirm reservation": 26,
|
39 |
+
"cook time": 27,
|
40 |
+
"credit limit": 28,
|
41 |
+
"credit limit change": 29,
|
42 |
+
"credit score": 30,
|
43 |
+
"current location": 31,
|
44 |
+
"damaged card": 32,
|
45 |
+
"date": 33,
|
46 |
+
"definition": 34,
|
47 |
+
"direct deposit": 35,
|
48 |
+
"directions": 36,
|
49 |
+
"distance": 37,
|
50 |
+
"do you have pets": 38,
|
51 |
+
"exchange rate": 39,
|
52 |
+
"expiration date": 40,
|
53 |
+
"find phone": 41,
|
54 |
+
"flight status": 42,
|
55 |
+
"flip coin": 43,
|
56 |
+
"food last": 44,
|
57 |
+
"freeze account": 45,
|
58 |
+
"fun fact": 46,
|
59 |
+
"gas": 47,
|
60 |
+
"gas type": 48,
|
61 |
+
"goodbye": 49,
|
62 |
+
"greeting": 50,
|
63 |
+
"how busy": 51,
|
64 |
+
"how old are you": 52,
|
65 |
+
"improve credit score": 53,
|
66 |
+
"income": 54,
|
67 |
+
"ingredient substitution": 55,
|
68 |
+
"ingredients list": 56,
|
69 |
+
"insurance": 57,
|
70 |
+
"insurance change": 58,
|
71 |
+
"interest rate": 59,
|
72 |
+
"international fees": 60,
|
73 |
+
"international visa": 61,
|
74 |
+
"jump start": 62,
|
75 |
+
"last maintenance": 63,
|
76 |
+
"lost luggage": 64,
|
77 |
+
"make call": 65,
|
78 |
+
"maybe": 66,
|
79 |
+
"meal suggestion": 67,
|
80 |
+
"meaning of life": 68,
|
81 |
+
"measurement conversion": 69,
|
82 |
+
"meeting schedule": 70,
|
83 |
+
"min payment": 71,
|
84 |
+
"mpg": 72,
|
85 |
+
"new card": 73,
|
86 |
+
"next holiday": 74,
|
87 |
+
"next song": 75,
|
88 |
+
"no": 76,
|
89 |
+
"nutrition info": 77,
|
90 |
+
"oil change how": 78,
|
91 |
+
"oil change when": 79,
|
92 |
+
"order": 80,
|
93 |
+
"order checks": 81,
|
94 |
+
"order status": 82,
|
95 |
+
"pay bill": 83,
|
96 |
+
"payday": 84,
|
97 |
+
"pin change": 85,
|
98 |
+
"play music": 86,
|
99 |
+
"plug type": 87,
|
100 |
+
"pto balance": 88,
|
101 |
+
"pto request": 89,
|
102 |
+
"pto request status": 90,
|
103 |
+
"pto used": 91,
|
104 |
+
"recipe": 92,
|
105 |
+
"redeem rewards": 93,
|
106 |
+
"reminder": 94,
|
107 |
+
"reminder update": 95,
|
108 |
+
"repeat": 96,
|
109 |
+
"replacement card duration": 97,
|
110 |
+
"report fraud": 98,
|
111 |
+
"report lost card": 99,
|
112 |
+
"reset settings": 100,
|
113 |
+
"restaurant reservation": 101,
|
114 |
+
"restaurant reviews": 102,
|
115 |
+
"restaurant suggestion": 103,
|
116 |
+
"rewards balance": 104,
|
117 |
+
"roll dice": 105,
|
118 |
+
"rollover 401k": 106,
|
119 |
+
"routing": 107,
|
120 |
+
"schedule maintenance": 108,
|
121 |
+
"schedule meeting": 109,
|
122 |
+
"share location": 110,
|
123 |
+
"shopping list": 111,
|
124 |
+
"shopping list update": 112,
|
125 |
+
"smart home": 113,
|
126 |
+
"spelling": 114,
|
127 |
+
"spending history": 115,
|
128 |
+
"sync device": 116,
|
129 |
+
"taxes": 117,
|
130 |
+
"tell joke": 118,
|
131 |
+
"text": 119,
|
132 |
+
"thank you": 120,
|
133 |
+
"time": 121,
|
134 |
+
"timer": 122,
|
135 |
+
"timezone": 123,
|
136 |
+
"tire change": 124,
|
137 |
+
"tire pressure": 125,
|
138 |
+
"todo list": 126,
|
139 |
+
"todo list update": 127,
|
140 |
+
"traffic": 128,
|
141 |
+
"transactions": 129,
|
142 |
+
"transfer": 130,
|
143 |
+
"translate": 131,
|
144 |
+
"travel alert": 132,
|
145 |
+
"travel notification": 133,
|
146 |
+
"travel suggestion": 134,
|
147 |
+
"uber": 135,
|
148 |
+
"update playlist": 136,
|
149 |
+
"user name": 137,
|
150 |
+
"vaccines": 138,
|
151 |
+
"w2": 139,
|
152 |
+
"weather": 140,
|
153 |
+
"what are your hobbies": 141,
|
154 |
+
"what can i ask you": 142,
|
155 |
+
"what is your name": 143,
|
156 |
+
"what song": 144,
|
157 |
+
"where are you from": 145,
|
158 |
+
"whisper mode": 146,
|
159 |
+
"who do you work for": 147,
|
160 |
+
"who made you": 148,
|
161 |
+
"yes": 149
|
162 |
+
},
|
163 |
+
"initializer_range": 0.02,
|
164 |
+
"label2id": {
|
165 |
+
"0": "accept reservations",
|
166 |
+
"1": "account blocked",
|
167 |
+
"2": "alarm",
|
168 |
+
"3": "application status",
|
169 |
+
"4": "apr",
|
170 |
+
"5": "are you a bot",
|
171 |
+
"6": "balance",
|
172 |
+
"7": "bill balance",
|
173 |
+
"8": "bill due",
|
174 |
+
"9": "book flight",
|
175 |
+
"10": "book hotel",
|
176 |
+
"11": "calculator",
|
177 |
+
"12": "calendar",
|
178 |
+
"13": "calendar update",
|
179 |
+
"14": "calories",
|
180 |
+
"15": "cancel",
|
181 |
+
"16": "cancel reservation",
|
182 |
+
"17": "car rental",
|
183 |
+
"18": "card declined",
|
184 |
+
"19": "carry on",
|
185 |
+
"20": "change accent",
|
186 |
+
"21": "change ai name",
|
187 |
+
"22": "change language",
|
188 |
+
"23": "change speed",
|
189 |
+
"24": "change user name",
|
190 |
+
"25": "change volume",
|
191 |
+
"26": "confirm reservation",
|
192 |
+
"27": "cook time",
|
193 |
+
"28": "credit limit",
|
194 |
+
"29": "credit limit change",
|
195 |
+
"30": "credit score",
|
196 |
+
"31": "current location",
|
197 |
+
"32": "damaged card",
|
198 |
+
"33": "date",
|
199 |
+
"34": "definition",
|
200 |
+
"35": "direct deposit",
|
201 |
+
"36": "directions",
|
202 |
+
"37": "distance",
|
203 |
+
"38": "do you have pets",
|
204 |
+
"39": "exchange rate",
|
205 |
+
"40": "expiration date",
|
206 |
+
"41": "find phone",
|
207 |
+
"42": "flight status",
|
208 |
+
"43": "flip coin",
|
209 |
+
"44": "food last",
|
210 |
+
"45": "freeze account",
|
211 |
+
"46": "fun fact",
|
212 |
+
"47": "gas",
|
213 |
+
"48": "gas type",
|
214 |
+
"49": "goodbye",
|
215 |
+
"50": "greeting",
|
216 |
+
"51": "how busy",
|
217 |
+
"52": "how old are you",
|
218 |
+
"53": "improve credit score",
|
219 |
+
"54": "income",
|
220 |
+
"55": "ingredient substitution",
|
221 |
+
"56": "ingredients list",
|
222 |
+
"57": "insurance",
|
223 |
+
"58": "insurance change",
|
224 |
+
"59": "interest rate",
|
225 |
+
"60": "international fees",
|
226 |
+
"61": "international visa",
|
227 |
+
"62": "jump start",
|
228 |
+
"63": "last maintenance",
|
229 |
+
"64": "lost luggage",
|
230 |
+
"65": "make call",
|
231 |
+
"66": "maybe",
|
232 |
+
"67": "meal suggestion",
|
233 |
+
"68": "meaning of life",
|
234 |
+
"69": "measurement conversion",
|
235 |
+
"70": "meeting schedule",
|
236 |
+
"71": "min payment",
|
237 |
+
"72": "mpg",
|
238 |
+
"73": "new card",
|
239 |
+
"74": "next holiday",
|
240 |
+
"75": "next song",
|
241 |
+
"76": "no",
|
242 |
+
"77": "nutrition info",
|
243 |
+
"78": "oil change how",
|
244 |
+
"79": "oil change when",
|
245 |
+
"80": "order",
|
246 |
+
"81": "order checks",
|
247 |
+
"82": "order status",
|
248 |
+
"83": "pay bill",
|
249 |
+
"84": "payday",
|
250 |
+
"85": "pin change",
|
251 |
+
"86": "play music",
|
252 |
+
"87": "plug type",
|
253 |
+
"88": "pto balance",
|
254 |
+
"89": "pto request",
|
255 |
+
"90": "pto request status",
|
256 |
+
"91": "pto used",
|
257 |
+
"92": "recipe",
|
258 |
+
"93": "redeem rewards",
|
259 |
+
"94": "reminder",
|
260 |
+
"95": "reminder update",
|
261 |
+
"96": "repeat",
|
262 |
+
"97": "replacement card duration",
|
263 |
+
"98": "report fraud",
|
264 |
+
"99": "report lost card",
|
265 |
+
"100": "reset settings",
|
266 |
+
"101": "restaurant reservation",
|
267 |
+
"102": "restaurant reviews",
|
268 |
+
"103": "restaurant suggestion",
|
269 |
+
"104": "rewards balance",
|
270 |
+
"105": "roll dice",
|
271 |
+
"106": "rollover 401k",
|
272 |
+
"107": "routing",
|
273 |
+
"108": "schedule maintenance",
|
274 |
+
"109": "schedule meeting",
|
275 |
+
"110": "share location",
|
276 |
+
"111": "shopping list",
|
277 |
+
"112": "shopping list update",
|
278 |
+
"113": "smart home",
|
279 |
+
"114": "spelling",
|
280 |
+
"115": "spending history",
|
281 |
+
"116": "sync device",
|
282 |
+
"117": "taxes",
|
283 |
+
"118": "tell joke",
|
284 |
+
"119": "text",
|
285 |
+
"120": "thank you",
|
286 |
+
"121": "time",
|
287 |
+
"122": "timer",
|
288 |
+
"123": "timezone",
|
289 |
+
"124": "tire change",
|
290 |
+
"125": "tire pressure",
|
291 |
+
"126": "todo list",
|
292 |
+
"127": "todo list update",
|
293 |
+
"128": "traffic",
|
294 |
+
"129": "transactions",
|
295 |
+
"130": "transfer",
|
296 |
+
"131": "translate",
|
297 |
+
"132": "travel alert",
|
298 |
+
"133": "travel notification",
|
299 |
+
"134": "travel suggestion",
|
300 |
+
"135": "uber",
|
301 |
+
"136": "update playlist",
|
302 |
+
"137": "user name",
|
303 |
+
"138": "vaccines",
|
304 |
+
"139": "w2",
|
305 |
+
"140": "weather",
|
306 |
+
"141": "what are your hobbies",
|
307 |
+
"142": "what can i ask you",
|
308 |
+
"143": "what is your name",
|
309 |
+
"144": "what song",
|
310 |
+
"145": "where are you from",
|
311 |
+
"146": "whisper mode",
|
312 |
+
"147": "who do you work for",
|
313 |
+
"148": "who made you",
|
314 |
+
"149": "yes"
|
315 |
+
},
|
316 |
+
"layer_norm_epsilon": 1e-05,
|
317 |
+
"model_type": "gpt2",
|
318 |
+
"n_ctx": 1024,
|
319 |
+
"n_embd": 768,
|
320 |
+
"n_head": 12,
|
321 |
+
"n_inner": null,
|
322 |
+
"n_layer": 12,
|
323 |
+
"n_positions": 1024,
|
324 |
+
"pad_token_id": 50256,
|
325 |
+
"reorder_and_upcast_attn": false,
|
326 |
+
"resid_pdrop": 0.1,
|
327 |
+
"scale_attn_by_inverse_layer_idx": false,
|
328 |
+
"scale_attn_weights": true,
|
329 |
+
"summary_activation": null,
|
330 |
+
"summary_first_dropout": 0.1,
|
331 |
+
"summary_proj_to_labels": true,
|
332 |
+
"summary_type": "cls_index",
|
333 |
+
"summary_use_proj": true,
|
334 |
+
"task_specific_params": {
|
335 |
+
"text-generation": {
|
336 |
+
"do_sample": true,
|
337 |
+
"max_length": 50
|
338 |
+
}
|
339 |
+
},
|
340 |
+
"transformers_version": "4.33.2",
|
341 |
+
"use_cache": true,
|
342 |
+
"vocab_size": 50257
|
343 |
+
}
|
merges.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
special_tokens_map.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<|endoftext|>",
|
3 |
+
"eos_token": "<|endoftext|>",
|
4 |
+
"pad_token": "<|endoftext|>",
|
5 |
+
"unk_token": "<|endoftext|>"
|
6 |
+
}
|
tf_model.h5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0acd6f2184c5252895a19ee3ec466089b24a98aaafe29bb230c95454cbcbd3c7
|
3 |
+
size 498407920
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"add_prefix_space": false,
|
3 |
+
"bos_token": "<|endoftext|>",
|
4 |
+
"clean_up_tokenization_spaces": true,
|
5 |
+
"eos_token": "<|endoftext|>",
|
6 |
+
"model_max_length": 1024,
|
7 |
+
"tokenizer_class": "GPT2Tokenizer",
|
8 |
+
"unk_token": "<|endoftext|>"
|
9 |
+
}
|
vocab.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|