Commit
•
ea72f6e
1
Parent(s):
1438451
Add nanoBERT model
Browse files- config.json +26 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +551 -0
- training_args.bin +3 -0
- vocab.json +27 -0
config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"RobertaForMaskedLM"
|
4 |
+
],
|
5 |
+
"attention_probs_dropout_prob": 0.1,
|
6 |
+
"bos_token_id": 0,
|
7 |
+
"classifier_dropout": null,
|
8 |
+
"eos_token_id": 2,
|
9 |
+
"hidden_act": "gelu",
|
10 |
+
"hidden_dropout_prob": 0.1,
|
11 |
+
"hidden_size": 320,
|
12 |
+
"initializer_range": 0.02,
|
13 |
+
"intermediate_size": 3072,
|
14 |
+
"layer_norm_eps": 1e-12,
|
15 |
+
"max_position_embeddings": 202,
|
16 |
+
"model_type": "roberta",
|
17 |
+
"num_attention_heads": 20,
|
18 |
+
"num_hidden_layers": 6,
|
19 |
+
"pad_token_id": 1,
|
20 |
+
"position_embedding_type": "absolute",
|
21 |
+
"torch_dtype": "float32",
|
22 |
+
"transformers_version": "4.34.0",
|
23 |
+
"type_vocab_size": 1,
|
24 |
+
"use_cache": true,
|
25 |
+
"vocab_size": 25
|
26 |
+
}
|
merges.txt
ADDED
File without changes
|
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:166ae04df4551eeabfffcf5a7d7bf1e445e2e9bf30575c14ab33b1b596f4cbac
|
3 |
+
size 115823034
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:abc474c6c6f23215d913d511eb771242317837f1e1880243e0076eae498e7fe2
|
3 |
+
size 57903730
|
rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a90152373dab951875e92c43446a0ac8948d44ed5965f9bb310bfa2b43bd3423
|
3 |
+
size 14180
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b914267d1a6232aa34982b741df9a8419974f9bd7ab32e4784e35ee04f2e5332
|
3 |
+
size 1064
|
trainer_state.json
ADDED
@@ -0,0 +1,551 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 3.722424669879707,
|
5 |
+
"eval_steps": 10000,
|
6 |
+
"global_step": 380000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.1,
|
13 |
+
"learning_rate": 9.83937751004016e-06,
|
14 |
+
"loss": 1.4799,
|
15 |
+
"step": 10000
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.1,
|
19 |
+
"eval_loss": 1.0304539203643799,
|
20 |
+
"eval_runtime": 69.6877,
|
21 |
+
"eval_samples_per_second": 1434.973,
|
22 |
+
"eval_steps_per_second": 14.952,
|
23 |
+
"step": 10000
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.2,
|
27 |
+
"learning_rate": 9.638654618473896e-06,
|
28 |
+
"loss": 0.9736,
|
29 |
+
"step": 20000
|
30 |
+
},
|
31 |
+
{
|
32 |
+
"epoch": 0.2,
|
33 |
+
"eval_loss": 0.9039866328239441,
|
34 |
+
"eval_runtime": 65.2654,
|
35 |
+
"eval_samples_per_second": 1532.205,
|
36 |
+
"eval_steps_per_second": 15.966,
|
37 |
+
"step": 20000
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.29,
|
41 |
+
"learning_rate": 9.437951807228917e-06,
|
42 |
+
"loss": 0.8922,
|
43 |
+
"step": 30000
|
44 |
+
},
|
45 |
+
{
|
46 |
+
"epoch": 0.29,
|
47 |
+
"eval_loss": 0.8560834527015686,
|
48 |
+
"eval_runtime": 64.4642,
|
49 |
+
"eval_samples_per_second": 1551.248,
|
50 |
+
"eval_steps_per_second": 16.164,
|
51 |
+
"step": 30000
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.39,
|
55 |
+
"learning_rate": 9.237228915662652e-06,
|
56 |
+
"loss": 0.8531,
|
57 |
+
"step": 40000
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 0.39,
|
61 |
+
"eval_loss": 0.8226799368858337,
|
62 |
+
"eval_runtime": 64.3868,
|
63 |
+
"eval_samples_per_second": 1553.114,
|
64 |
+
"eval_steps_per_second": 16.183,
|
65 |
+
"step": 40000
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.49,
|
69 |
+
"learning_rate": 9.036526104417672e-06,
|
70 |
+
"loss": 0.829,
|
71 |
+
"step": 50000
|
72 |
+
},
|
73 |
+
{
|
74 |
+
"epoch": 0.49,
|
75 |
+
"eval_loss": 0.8042359948158264,
|
76 |
+
"eval_runtime": 64.3396,
|
77 |
+
"eval_samples_per_second": 1554.254,
|
78 |
+
"eval_steps_per_second": 16.195,
|
79 |
+
"step": 50000
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.59,
|
83 |
+
"learning_rate": 8.835823293172691e-06,
|
84 |
+
"loss": 0.811,
|
85 |
+
"step": 60000
|
86 |
+
},
|
87 |
+
{
|
88 |
+
"epoch": 0.59,
|
89 |
+
"eval_loss": 0.7950388789176941,
|
90 |
+
"eval_runtime": 64.5291,
|
91 |
+
"eval_samples_per_second": 1549.687,
|
92 |
+
"eval_steps_per_second": 16.148,
|
93 |
+
"step": 60000
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.69,
|
97 |
+
"learning_rate": 8.635120481927711e-06,
|
98 |
+
"loss": 0.7976,
|
99 |
+
"step": 70000
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 0.69,
|
103 |
+
"eval_loss": 0.7777369022369385,
|
104 |
+
"eval_runtime": 64.862,
|
105 |
+
"eval_samples_per_second": 1541.734,
|
106 |
+
"eval_steps_per_second": 16.065,
|
107 |
+
"step": 70000
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.78,
|
111 |
+
"learning_rate": 8.434417670682732e-06,
|
112 |
+
"loss": 0.7872,
|
113 |
+
"step": 80000
|
114 |
+
},
|
115 |
+
{
|
116 |
+
"epoch": 0.78,
|
117 |
+
"eval_loss": 0.7702258825302124,
|
118 |
+
"eval_runtime": 64.4619,
|
119 |
+
"eval_samples_per_second": 1551.303,
|
120 |
+
"eval_steps_per_second": 16.165,
|
121 |
+
"step": 80000
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.88,
|
125 |
+
"learning_rate": 8.233714859437752e-06,
|
126 |
+
"loss": 0.778,
|
127 |
+
"step": 90000
|
128 |
+
},
|
129 |
+
{
|
130 |
+
"epoch": 0.88,
|
131 |
+
"eval_loss": 0.760698139667511,
|
132 |
+
"eval_runtime": 64.4008,
|
133 |
+
"eval_samples_per_second": 1552.775,
|
134 |
+
"eval_steps_per_second": 16.18,
|
135 |
+
"step": 90000
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.98,
|
139 |
+
"learning_rate": 8.033012048192772e-06,
|
140 |
+
"loss": 0.7695,
|
141 |
+
"step": 100000
|
142 |
+
},
|
143 |
+
{
|
144 |
+
"epoch": 0.98,
|
145 |
+
"eval_loss": 0.7517786622047424,
|
146 |
+
"eval_runtime": 64.56,
|
147 |
+
"eval_samples_per_second": 1548.946,
|
148 |
+
"eval_steps_per_second": 16.14,
|
149 |
+
"step": 100000
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 1.08,
|
153 |
+
"learning_rate": 7.832309236947791e-06,
|
154 |
+
"loss": 0.7638,
|
155 |
+
"step": 110000
|
156 |
+
},
|
157 |
+
{
|
158 |
+
"epoch": 1.08,
|
159 |
+
"eval_loss": 0.7468145489692688,
|
160 |
+
"eval_runtime": 64.9838,
|
161 |
+
"eval_samples_per_second": 1538.846,
|
162 |
+
"eval_steps_per_second": 16.035,
|
163 |
+
"step": 110000
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 1.18,
|
167 |
+
"learning_rate": 7.631606425702813e-06,
|
168 |
+
"loss": 0.7573,
|
169 |
+
"step": 120000
|
170 |
+
},
|
171 |
+
{
|
172 |
+
"epoch": 1.18,
|
173 |
+
"eval_loss": 0.739628255367279,
|
174 |
+
"eval_runtime": 64.2595,
|
175 |
+
"eval_samples_per_second": 1556.191,
|
176 |
+
"eval_steps_per_second": 16.216,
|
177 |
+
"step": 120000
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.27,
|
181 |
+
"learning_rate": 7.4309036144578315e-06,
|
182 |
+
"loss": 0.7512,
|
183 |
+
"step": 130000
|
184 |
+
},
|
185 |
+
{
|
186 |
+
"epoch": 1.27,
|
187 |
+
"eval_loss": 0.7346537709236145,
|
188 |
+
"eval_runtime": 64.911,
|
189 |
+
"eval_samples_per_second": 1540.57,
|
190 |
+
"eval_steps_per_second": 16.053,
|
191 |
+
"step": 130000
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 1.37,
|
195 |
+
"learning_rate": 7.2301807228915665e-06,
|
196 |
+
"loss": 0.7468,
|
197 |
+
"step": 140000
|
198 |
+
},
|
199 |
+
{
|
200 |
+
"epoch": 1.37,
|
201 |
+
"eval_loss": 0.7300976514816284,
|
202 |
+
"eval_runtime": 64.8445,
|
203 |
+
"eval_samples_per_second": 1542.15,
|
204 |
+
"eval_steps_per_second": 16.069,
|
205 |
+
"step": 140000
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 1.47,
|
209 |
+
"learning_rate": 7.029477911646587e-06,
|
210 |
+
"loss": 0.7418,
|
211 |
+
"step": 150000
|
212 |
+
},
|
213 |
+
{
|
214 |
+
"epoch": 1.47,
|
215 |
+
"eval_loss": 0.7252103686332703,
|
216 |
+
"eval_runtime": 64.5503,
|
217 |
+
"eval_samples_per_second": 1549.179,
|
218 |
+
"eval_steps_per_second": 16.142,
|
219 |
+
"step": 150000
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.57,
|
223 |
+
"learning_rate": 6.828775100401607e-06,
|
224 |
+
"loss": 0.7379,
|
225 |
+
"step": 160000
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 1.57,
|
229 |
+
"eval_loss": 0.7203709483146667,
|
230 |
+
"eval_runtime": 64.4763,
|
231 |
+
"eval_samples_per_second": 1550.958,
|
232 |
+
"eval_steps_per_second": 16.161,
|
233 |
+
"step": 160000
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 1.67,
|
237 |
+
"learning_rate": 6.628072289156627e-06,
|
238 |
+
"loss": 0.7333,
|
239 |
+
"step": 170000
|
240 |
+
},
|
241 |
+
{
|
242 |
+
"epoch": 1.67,
|
243 |
+
"eval_loss": 0.7167079448699951,
|
244 |
+
"eval_runtime": 64.7852,
|
245 |
+
"eval_samples_per_second": 1543.562,
|
246 |
+
"eval_steps_per_second": 16.084,
|
247 |
+
"step": 170000
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 1.76,
|
251 |
+
"learning_rate": 6.427369477911647e-06,
|
252 |
+
"loss": 0.7298,
|
253 |
+
"step": 180000
|
254 |
+
},
|
255 |
+
{
|
256 |
+
"epoch": 1.76,
|
257 |
+
"eval_loss": 0.7129059433937073,
|
258 |
+
"eval_runtime": 64.609,
|
259 |
+
"eval_samples_per_second": 1547.772,
|
260 |
+
"eval_steps_per_second": 16.128,
|
261 |
+
"step": 180000
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 1.86,
|
265 |
+
"learning_rate": 6.2266666666666675e-06,
|
266 |
+
"loss": 0.7258,
|
267 |
+
"step": 190000
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 1.86,
|
271 |
+
"eval_loss": 0.7089965343475342,
|
272 |
+
"eval_runtime": 64.6825,
|
273 |
+
"eval_samples_per_second": 1546.013,
|
274 |
+
"eval_steps_per_second": 16.109,
|
275 |
+
"step": 190000
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 1.96,
|
279 |
+
"learning_rate": 6.025963855421687e-06,
|
280 |
+
"loss": 0.7229,
|
281 |
+
"step": 200000
|
282 |
+
},
|
283 |
+
{
|
284 |
+
"epoch": 1.96,
|
285 |
+
"eval_loss": 0.7047411799430847,
|
286 |
+
"eval_runtime": 64.3501,
|
287 |
+
"eval_samples_per_second": 1553.999,
|
288 |
+
"eval_steps_per_second": 16.193,
|
289 |
+
"step": 200000
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 2.06,
|
293 |
+
"learning_rate": 5.825261044176708e-06,
|
294 |
+
"loss": 0.7194,
|
295 |
+
"step": 210000
|
296 |
+
},
|
297 |
+
{
|
298 |
+
"epoch": 2.06,
|
299 |
+
"eval_loss": 0.7026636004447937,
|
300 |
+
"eval_runtime": 64.4785,
|
301 |
+
"eval_samples_per_second": 1550.905,
|
302 |
+
"eval_steps_per_second": 16.16,
|
303 |
+
"step": 210000
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 2.16,
|
307 |
+
"learning_rate": 5.624558232931727e-06,
|
308 |
+
"loss": 0.7171,
|
309 |
+
"step": 220000
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 2.16,
|
313 |
+
"eval_loss": 0.702177107334137,
|
314 |
+
"eval_runtime": 64.5209,
|
315 |
+
"eval_samples_per_second": 1549.886,
|
316 |
+
"eval_steps_per_second": 16.15,
|
317 |
+
"step": 220000
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 2.25,
|
321 |
+
"learning_rate": 5.423855421686748e-06,
|
322 |
+
"loss": 0.7138,
|
323 |
+
"step": 230000
|
324 |
+
},
|
325 |
+
{
|
326 |
+
"epoch": 2.25,
|
327 |
+
"eval_loss": 0.6978575587272644,
|
328 |
+
"eval_runtime": 64.573,
|
329 |
+
"eval_samples_per_second": 1548.634,
|
330 |
+
"eval_steps_per_second": 16.137,
|
331 |
+
"step": 230000
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 2.35,
|
335 |
+
"learning_rate": 5.2231526104417676e-06,
|
336 |
+
"loss": 0.7113,
|
337 |
+
"step": 240000
|
338 |
+
},
|
339 |
+
{
|
340 |
+
"epoch": 2.35,
|
341 |
+
"eval_loss": 0.6939824819564819,
|
342 |
+
"eval_runtime": 64.7414,
|
343 |
+
"eval_samples_per_second": 1544.606,
|
344 |
+
"eval_steps_per_second": 16.095,
|
345 |
+
"step": 240000
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 2.45,
|
349 |
+
"learning_rate": 5.022449799196788e-06,
|
350 |
+
"loss": 0.7087,
|
351 |
+
"step": 250000
|
352 |
+
},
|
353 |
+
{
|
354 |
+
"epoch": 2.45,
|
355 |
+
"eval_loss": 0.6922757029533386,
|
356 |
+
"eval_runtime": 64.5094,
|
357 |
+
"eval_samples_per_second": 1550.162,
|
358 |
+
"eval_steps_per_second": 16.153,
|
359 |
+
"step": 250000
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 2.55,
|
363 |
+
"learning_rate": 4.821746987951808e-06,
|
364 |
+
"loss": 0.7071,
|
365 |
+
"step": 260000
|
366 |
+
},
|
367 |
+
{
|
368 |
+
"epoch": 2.55,
|
369 |
+
"eval_loss": 0.6915081739425659,
|
370 |
+
"eval_runtime": 65.0605,
|
371 |
+
"eval_samples_per_second": 1537.031,
|
372 |
+
"eval_steps_per_second": 16.016,
|
373 |
+
"step": 260000
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 2.64,
|
377 |
+
"learning_rate": 4.6210441767068274e-06,
|
378 |
+
"loss": 0.7047,
|
379 |
+
"step": 270000
|
380 |
+
},
|
381 |
+
{
|
382 |
+
"epoch": 2.64,
|
383 |
+
"eval_loss": 0.6870027184486389,
|
384 |
+
"eval_runtime": 64.7199,
|
385 |
+
"eval_samples_per_second": 1545.119,
|
386 |
+
"eval_steps_per_second": 16.1,
|
387 |
+
"step": 270000
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 2.74,
|
391 |
+
"learning_rate": 4.420341365461848e-06,
|
392 |
+
"loss": 0.7027,
|
393 |
+
"step": 280000
|
394 |
+
},
|
395 |
+
{
|
396 |
+
"epoch": 2.74,
|
397 |
+
"eval_loss": 0.6844470500946045,
|
398 |
+
"eval_runtime": 64.906,
|
399 |
+
"eval_samples_per_second": 1540.69,
|
400 |
+
"eval_steps_per_second": 16.054,
|
401 |
+
"step": 280000
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 2.84,
|
405 |
+
"learning_rate": 4.219638554216868e-06,
|
406 |
+
"loss": 0.6999,
|
407 |
+
"step": 290000
|
408 |
+
},
|
409 |
+
{
|
410 |
+
"epoch": 2.84,
|
411 |
+
"eval_loss": 0.6833365559577942,
|
412 |
+
"eval_runtime": 64.5739,
|
413 |
+
"eval_samples_per_second": 1548.614,
|
414 |
+
"eval_steps_per_second": 16.137,
|
415 |
+
"step": 290000
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 2.94,
|
419 |
+
"learning_rate": 4.018935742971888e-06,
|
420 |
+
"loss": 0.6992,
|
421 |
+
"step": 300000
|
422 |
+
},
|
423 |
+
{
|
424 |
+
"epoch": 2.94,
|
425 |
+
"eval_loss": 0.6834575533866882,
|
426 |
+
"eval_runtime": 64.6706,
|
427 |
+
"eval_samples_per_second": 1546.298,
|
428 |
+
"eval_steps_per_second": 16.112,
|
429 |
+
"step": 300000
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 3.04,
|
433 |
+
"learning_rate": 3.818253012048193e-06,
|
434 |
+
"loss": 0.6975,
|
435 |
+
"step": 310000
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 3.04,
|
439 |
+
"eval_loss": 0.6793897151947021,
|
440 |
+
"eval_runtime": 64.8495,
|
441 |
+
"eval_samples_per_second": 1542.031,
|
442 |
+
"eval_steps_per_second": 16.068,
|
443 |
+
"step": 310000
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 3.13,
|
447 |
+
"learning_rate": 3.617530120481928e-06,
|
448 |
+
"loss": 0.6947,
|
449 |
+
"step": 320000
|
450 |
+
},
|
451 |
+
{
|
452 |
+
"epoch": 3.13,
|
453 |
+
"eval_loss": 0.6794592142105103,
|
454 |
+
"eval_runtime": 64.6554,
|
455 |
+
"eval_samples_per_second": 1546.661,
|
456 |
+
"eval_steps_per_second": 16.116,
|
457 |
+
"step": 320000
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 3.23,
|
461 |
+
"learning_rate": 3.416827309236948e-06,
|
462 |
+
"loss": 0.6947,
|
463 |
+
"step": 330000
|
464 |
+
},
|
465 |
+
{
|
466 |
+
"epoch": 3.23,
|
467 |
+
"eval_loss": 0.6781629920005798,
|
468 |
+
"eval_runtime": 65.93,
|
469 |
+
"eval_samples_per_second": 1516.76,
|
470 |
+
"eval_steps_per_second": 15.805,
|
471 |
+
"step": 330000
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 3.33,
|
475 |
+
"learning_rate": 3.216124497991968e-06,
|
476 |
+
"loss": 0.6932,
|
477 |
+
"step": 340000
|
478 |
+
},
|
479 |
+
{
|
480 |
+
"epoch": 3.33,
|
481 |
+
"eval_loss": 0.6771513819694519,
|
482 |
+
"eval_runtime": 64.7971,
|
483 |
+
"eval_samples_per_second": 1543.279,
|
484 |
+
"eval_steps_per_second": 16.081,
|
485 |
+
"step": 340000
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 3.43,
|
489 |
+
"learning_rate": 3.015401606425703e-06,
|
490 |
+
"loss": 0.6914,
|
491 |
+
"step": 350000
|
492 |
+
},
|
493 |
+
{
|
494 |
+
"epoch": 3.43,
|
495 |
+
"eval_loss": 0.6757428050041199,
|
496 |
+
"eval_runtime": 64.8856,
|
497 |
+
"eval_samples_per_second": 1541.173,
|
498 |
+
"eval_steps_per_second": 16.059,
|
499 |
+
"step": 350000
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 3.53,
|
503 |
+
"learning_rate": 2.8146987951807233e-06,
|
504 |
+
"loss": 0.6907,
|
505 |
+
"step": 360000
|
506 |
+
},
|
507 |
+
{
|
508 |
+
"epoch": 3.53,
|
509 |
+
"eval_loss": 0.6730121374130249,
|
510 |
+
"eval_runtime": 65.1098,
|
511 |
+
"eval_samples_per_second": 1535.867,
|
512 |
+
"eval_steps_per_second": 16.004,
|
513 |
+
"step": 360000
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 3.62,
|
517 |
+
"learning_rate": 2.6139959839357434e-06,
|
518 |
+
"loss": 0.689,
|
519 |
+
"step": 370000
|
520 |
+
},
|
521 |
+
{
|
522 |
+
"epoch": 3.62,
|
523 |
+
"eval_loss": 0.6724720597267151,
|
524 |
+
"eval_runtime": 65.1981,
|
525 |
+
"eval_samples_per_second": 1533.787,
|
526 |
+
"eval_steps_per_second": 15.982,
|
527 |
+
"step": 370000
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 3.72,
|
531 |
+
"learning_rate": 2.413293172690763e-06,
|
532 |
+
"loss": 0.6885,
|
533 |
+
"step": 380000
|
534 |
+
},
|
535 |
+
{
|
536 |
+
"epoch": 3.72,
|
537 |
+
"eval_loss": 0.6744215488433838,
|
538 |
+
"eval_runtime": 64.9859,
|
539 |
+
"eval_samples_per_second": 1538.796,
|
540 |
+
"eval_steps_per_second": 16.034,
|
541 |
+
"step": 380000
|
542 |
+
}
|
543 |
+
],
|
544 |
+
"logging_steps": 10000,
|
545 |
+
"max_steps": 500000,
|
546 |
+
"num_train_epochs": 5,
|
547 |
+
"save_steps": 10000,
|
548 |
+
"total_flos": 6.301004166565632e+17,
|
549 |
+
"trial_name": null,
|
550 |
+
"trial_params": null
|
551 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:52cfee9909bae69ef99be6ff3749cceaf5f5a35741ce1b069ebf8f2caea16cef
|
3 |
+
size 4600
|
vocab.json
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"<s>": 0,
|
3 |
+
"</s>": 2,
|
4 |
+
"<pad>": 1,
|
5 |
+
"<unk>": 3,
|
6 |
+
"<mask>": 4,
|
7 |
+
"A": 5,
|
8 |
+
"C": 6,
|
9 |
+
"D": 7,
|
10 |
+
"E": 8,
|
11 |
+
"F": 9,
|
12 |
+
"G": 10,
|
13 |
+
"H": 11,
|
14 |
+
"I": 12,
|
15 |
+
"K": 13,
|
16 |
+
"L": 14,
|
17 |
+
"M": 15,
|
18 |
+
"N": 16,
|
19 |
+
"P": 17,
|
20 |
+
"Q": 18,
|
21 |
+
"R": 19,
|
22 |
+
"S": 20,
|
23 |
+
"T": 21,
|
24 |
+
"V": 22,
|
25 |
+
"W": 23,
|
26 |
+
"Y": 24
|
27 |
+
}
|