gguichard commited on
Commit
7f33e2a
1 Parent(s): f0821c9

Training in progress, epoch 21, checkpoint

Browse files
checkpoint-4452/README.md ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+ ### Framework versions
7
+
8
+
9
+ - PEFT 0.5.0
checkpoint-4452/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "camembert/camembert-large",
4
+ "bias": "all",
5
+ "fan_in_fan_out": false,
6
+ "inference_mode": true,
7
+ "init_lora_weights": true,
8
+ "layers_pattern": null,
9
+ "layers_to_transform": null,
10
+ "lora_alpha": 16,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": [
13
+ "classifier.bias",
14
+ "classifier.weight"
15
+ ],
16
+ "peft_type": "LORA",
17
+ "r": 64,
18
+ "revision": null,
19
+ "target_modules": [
20
+ "query",
21
+ "key",
22
+ "value",
23
+ "dense"
24
+ ],
25
+ "task_type": "TOKEN_CLS"
26
+ }
checkpoint-4452/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3ebacd23b93e6b22e56777f61a48038693530918f950f3f65589b487cb2f676
3
+ size 117304221
checkpoint-4452/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d16cce7352308bb38420fd615dc1d9f3636e06e7246abd63dd49cc9f0f267ffe
3
+ size 234652550
checkpoint-4452/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94feff1666aaa0862922ded2c60ce6dc3ba16be8454b9fb078d671fdcaf614bd
3
+ size 14575
checkpoint-4452/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0bdd72ff2dc2702504b46387625646d2b92174d4bd62d6f7ffe7288b1cc944b4
3
+ size 627
checkpoint-4452/sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f98f266fdc548c94216aaadc13ffaaafacf0c8793303e2195322d954549ea261
3
+ size 808767
checkpoint-4452/special_tokens_map.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<s>NOTUSED",
4
+ "</s>NOTUSED"
5
+ ],
6
+ "bos_token": "<s>",
7
+ "cls_token": "<s>",
8
+ "eos_token": "</s>",
9
+ "mask_token": {
10
+ "content": "<mask>",
11
+ "lstrip": true,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "<pad>",
17
+ "sep_token": "</s>",
18
+ "unk_token": "<unk>"
19
+ }
checkpoint-4452/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-4452/tokenizer_config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<s>NOTUSED",
4
+ "</s>NOTUSED"
5
+ ],
6
+ "bos_token": "<s>",
7
+ "clean_up_tokenization_spaces": true,
8
+ "cls_token": "<s>",
9
+ "eos_token": "</s>",
10
+ "mask_token": {
11
+ "__type": "AddedToken",
12
+ "content": "<mask>",
13
+ "lstrip": true,
14
+ "normalized": true,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "model_max_length": 1000000000000000019884624838656,
19
+ "pad_token": "<pad>",
20
+ "sep_token": "</s>",
21
+ "sp_model_kwargs": {},
22
+ "tokenizer_class": "CamembertTokenizer",
23
+ "unk_token": "<unk>"
24
+ }
checkpoint-4452/trainer_state.json ADDED
@@ -0,0 +1,313 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 21.0,
5
+ "eval_steps": 500,
6
+ "global_step": 4452,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 1.0,
13
+ "learning_rate": 0.00019900471698113207,
14
+ "loss": 6.1329,
15
+ "step": 212
16
+ },
17
+ {
18
+ "epoch": 1.0,
19
+ "eval_loss": 5.14426851272583,
20
+ "eval_runtime": 3.8194,
21
+ "eval_samples_per_second": 176.991,
22
+ "eval_steps_per_second": 22.255,
23
+ "step": 212
24
+ },
25
+ {
26
+ "epoch": 2.0,
27
+ "learning_rate": 0.00019800471698113207,
28
+ "loss": 4.6548,
29
+ "step": 424
30
+ },
31
+ {
32
+ "epoch": 2.0,
33
+ "eval_loss": 3.559998035430908,
34
+ "eval_runtime": 3.8713,
35
+ "eval_samples_per_second": 174.616,
36
+ "eval_steps_per_second": 21.956,
37
+ "step": 424
38
+ },
39
+ {
40
+ "epoch": 3.0,
41
+ "learning_rate": 0.00019700471698113207,
42
+ "loss": 3.2979,
43
+ "step": 636
44
+ },
45
+ {
46
+ "epoch": 3.0,
47
+ "eval_loss": 2.3220388889312744,
48
+ "eval_runtime": 3.7585,
49
+ "eval_samples_per_second": 179.86,
50
+ "eval_steps_per_second": 22.616,
51
+ "step": 636
52
+ },
53
+ {
54
+ "epoch": 4.0,
55
+ "learning_rate": 0.00019600471698113207,
56
+ "loss": 2.2394,
57
+ "step": 848
58
+ },
59
+ {
60
+ "epoch": 4.0,
61
+ "eval_loss": 1.5198729038238525,
62
+ "eval_runtime": 3.8733,
63
+ "eval_samples_per_second": 174.529,
64
+ "eval_steps_per_second": 21.945,
65
+ "step": 848
66
+ },
67
+ {
68
+ "epoch": 5.0,
69
+ "learning_rate": 0.00019500471698113208,
70
+ "loss": 1.5436,
71
+ "step": 1060
72
+ },
73
+ {
74
+ "epoch": 5.0,
75
+ "eval_loss": 1.0844509601593018,
76
+ "eval_runtime": 3.7882,
77
+ "eval_samples_per_second": 178.45,
78
+ "eval_steps_per_second": 22.438,
79
+ "step": 1060
80
+ },
81
+ {
82
+ "epoch": 6.0,
83
+ "learning_rate": 0.0001940047169811321,
84
+ "loss": 1.1087,
85
+ "step": 1272
86
+ },
87
+ {
88
+ "epoch": 6.0,
89
+ "eval_loss": 0.8275050520896912,
90
+ "eval_runtime": 3.8596,
91
+ "eval_samples_per_second": 175.146,
92
+ "eval_steps_per_second": 22.023,
93
+ "step": 1272
94
+ },
95
+ {
96
+ "epoch": 7.0,
97
+ "learning_rate": 0.00019300943396226416,
98
+ "loss": 0.8094,
99
+ "step": 1484
100
+ },
101
+ {
102
+ "epoch": 7.0,
103
+ "eval_loss": 0.6410268545150757,
104
+ "eval_runtime": 3.826,
105
+ "eval_samples_per_second": 176.686,
106
+ "eval_steps_per_second": 22.216,
107
+ "step": 1484
108
+ },
109
+ {
110
+ "epoch": 8.0,
111
+ "learning_rate": 0.00019200943396226417,
112
+ "loss": 0.5999,
113
+ "step": 1696
114
+ },
115
+ {
116
+ "epoch": 8.0,
117
+ "eval_loss": 0.5242534279823303,
118
+ "eval_runtime": 3.7888,
119
+ "eval_samples_per_second": 178.42,
120
+ "eval_steps_per_second": 22.434,
121
+ "step": 1696
122
+ },
123
+ {
124
+ "epoch": 9.0,
125
+ "learning_rate": 0.00019100943396226417,
126
+ "loss": 0.4333,
127
+ "step": 1908
128
+ },
129
+ {
130
+ "epoch": 9.0,
131
+ "eval_loss": 0.4525991678237915,
132
+ "eval_runtime": 3.7762,
133
+ "eval_samples_per_second": 179.014,
134
+ "eval_steps_per_second": 22.509,
135
+ "step": 1908
136
+ },
137
+ {
138
+ "epoch": 10.0,
139
+ "learning_rate": 0.00019000943396226417,
140
+ "loss": 0.3096,
141
+ "step": 2120
142
+ },
143
+ {
144
+ "epoch": 10.0,
145
+ "eval_loss": 0.3988202214241028,
146
+ "eval_runtime": 3.3771,
147
+ "eval_samples_per_second": 200.169,
148
+ "eval_steps_per_second": 25.169,
149
+ "step": 2120
150
+ },
151
+ {
152
+ "epoch": 11.0,
153
+ "learning_rate": 0.00018900943396226415,
154
+ "loss": 0.225,
155
+ "step": 2332
156
+ },
157
+ {
158
+ "epoch": 11.0,
159
+ "eval_loss": 0.37283995747566223,
160
+ "eval_runtime": 3.3836,
161
+ "eval_samples_per_second": 199.788,
162
+ "eval_steps_per_second": 25.121,
163
+ "step": 2332
164
+ },
165
+ {
166
+ "epoch": 12.0,
167
+ "learning_rate": 0.00018800943396226415,
168
+ "loss": 0.1484,
169
+ "step": 2544
170
+ },
171
+ {
172
+ "epoch": 12.0,
173
+ "eval_loss": 0.36433306336402893,
174
+ "eval_runtime": 3.9173,
175
+ "eval_samples_per_second": 172.567,
176
+ "eval_steps_per_second": 21.698,
177
+ "step": 2544
178
+ },
179
+ {
180
+ "epoch": 13.0,
181
+ "learning_rate": 0.00018700943396226415,
182
+ "loss": 0.1086,
183
+ "step": 2756
184
+ },
185
+ {
186
+ "epoch": 13.0,
187
+ "eval_loss": 0.3588293790817261,
188
+ "eval_runtime": 3.3573,
189
+ "eval_samples_per_second": 201.351,
190
+ "eval_steps_per_second": 25.318,
191
+ "step": 2756
192
+ },
193
+ {
194
+ "epoch": 14.0,
195
+ "learning_rate": 0.00018600943396226415,
196
+ "loss": 0.0794,
197
+ "step": 2968
198
+ },
199
+ {
200
+ "epoch": 14.0,
201
+ "eval_loss": 0.34358447790145874,
202
+ "eval_runtime": 3.5044,
203
+ "eval_samples_per_second": 192.9,
204
+ "eval_steps_per_second": 24.255,
205
+ "step": 2968
206
+ },
207
+ {
208
+ "epoch": 15.0,
209
+ "learning_rate": 0.00018500943396226416,
210
+ "loss": 0.0629,
211
+ "step": 3180
212
+ },
213
+ {
214
+ "epoch": 15.0,
215
+ "eval_loss": 0.3459497392177582,
216
+ "eval_runtime": 3.9331,
217
+ "eval_samples_per_second": 171.873,
218
+ "eval_steps_per_second": 21.611,
219
+ "step": 3180
220
+ },
221
+ {
222
+ "epoch": 16.0,
223
+ "learning_rate": 0.00018400943396226416,
224
+ "loss": 0.0499,
225
+ "step": 3392
226
+ },
227
+ {
228
+ "epoch": 16.0,
229
+ "eval_loss": 0.35890209674835205,
230
+ "eval_runtime": 3.9531,
231
+ "eval_samples_per_second": 171.003,
232
+ "eval_steps_per_second": 21.502,
233
+ "step": 3392
234
+ },
235
+ {
236
+ "epoch": 17.0,
237
+ "learning_rate": 0.00018300943396226416,
238
+ "loss": 0.0381,
239
+ "step": 3604
240
+ },
241
+ {
242
+ "epoch": 17.0,
243
+ "eval_loss": 0.36232635378837585,
244
+ "eval_runtime": 3.436,
245
+ "eval_samples_per_second": 196.742,
246
+ "eval_steps_per_second": 24.738,
247
+ "step": 3604
248
+ },
249
+ {
250
+ "epoch": 18.0,
251
+ "learning_rate": 0.00018200943396226417,
252
+ "loss": 0.0328,
253
+ "step": 3816
254
+ },
255
+ {
256
+ "epoch": 18.0,
257
+ "eval_loss": 0.33994847536087036,
258
+ "eval_runtime": 3.6854,
259
+ "eval_samples_per_second": 183.428,
260
+ "eval_steps_per_second": 23.064,
261
+ "step": 3816
262
+ },
263
+ {
264
+ "epoch": 19.0,
265
+ "learning_rate": 0.00018100943396226414,
266
+ "loss": 0.0271,
267
+ "step": 4028
268
+ },
269
+ {
270
+ "epoch": 19.0,
271
+ "eval_loss": 0.3786211609840393,
272
+ "eval_runtime": 3.5861,
273
+ "eval_samples_per_second": 188.507,
274
+ "eval_steps_per_second": 23.703,
275
+ "step": 4028
276
+ },
277
+ {
278
+ "epoch": 20.0,
279
+ "learning_rate": 0.00018000943396226414,
280
+ "loss": 0.0237,
281
+ "step": 4240
282
+ },
283
+ {
284
+ "epoch": 20.0,
285
+ "eval_loss": 0.3512681722640991,
286
+ "eval_runtime": 3.6632,
287
+ "eval_samples_per_second": 184.54,
288
+ "eval_steps_per_second": 23.204,
289
+ "step": 4240
290
+ },
291
+ {
292
+ "epoch": 21.0,
293
+ "learning_rate": 0.00017900943396226415,
294
+ "loss": 0.0186,
295
+ "step": 4452
296
+ },
297
+ {
298
+ "epoch": 21.0,
299
+ "eval_loss": 0.374735951423645,
300
+ "eval_runtime": 3.9023,
301
+ "eval_samples_per_second": 173.23,
302
+ "eval_steps_per_second": 21.782,
303
+ "step": 4452
304
+ }
305
+ ],
306
+ "logging_steps": 500,
307
+ "max_steps": 42400,
308
+ "num_train_epochs": 200,
309
+ "save_steps": 500,
310
+ "total_flos": 5991938729727324.0,
311
+ "trial_name": null,
312
+ "trial_params": null
313
+ }
checkpoint-4452/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11396e77b2185e0d94bce5383c2512c5afd7f340883723b782abc68f5d11c5a6
3
+ size 4155