Hugging Face's logo

kinleyrabgay
/
nllb-200-600M-dzo-eng-30k

Text Generation
Transformers
Safetensors
m2m_100
text2text-generation
Generated from Trainer
Model card Files Community
kinleyrabgay commited on
Commit
2cc22c2
·
verified ·
1 Parent(s): 143553a

Initial release: Fine-tuned NLLB for Dzongkha-English

Browse files
Files changed (3) hide show
  1. all_results.json +8 -0
  2. train_results.json +8 -0
  3. trainer_state.json +263 -0
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "total_flos": 2.70888075264e+16,
4
+ "train_loss": 0.04798975685119629,
5
+ "train_runtime": 11153.1781,
6
+ "train_samples_per_second": 8.966,
7
+ "train_steps_per_second": 1.121
8
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "total_flos": 2.70888075264e+16,
4
+ "train_loss": 0.04798975685119629,
5
+ "train_runtime": 11153.1781,
6
+ "train_samples_per_second": 8.966,
7
+ "train_steps_per_second": 1.121
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,263 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 10000,
3
+ "best_metric": 63.44155467990254,
4
+ "best_model_checkpoint": "nllb-200-600M-dzo-eng-30k-checkpoints/checkpoint-10000",
5
+ "epoch": 5.0,
6
+ "eval_steps": 500,
7
+ "global_step": 12500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.2,
14
+ "grad_norm": 0.2576698958873749,
15
+ "learning_rate": 2.88024e-05,
16
+ "loss": 0.0753,
17
+ "step": 500
18
+ },
19
+ {
20
+ "epoch": 0.4,
21
+ "grad_norm": 0.19992510974407196,
22
+ "learning_rate": 2.7602400000000002e-05,
23
+ "loss": 0.0735,
24
+ "step": 1000
25
+ },
26
+ {
27
+ "epoch": 0.6,
28
+ "grad_norm": 0.2977311909198761,
29
+ "learning_rate": 2.64024e-05,
30
+ "loss": 0.0713,
31
+ "step": 1500
32
+ },
33
+ {
34
+ "epoch": 0.8,
35
+ "grad_norm": 0.21389248967170715,
36
+ "learning_rate": 2.52024e-05,
37
+ "loss": 0.071,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 1.0,
42
+ "grad_norm": 0.2741245627403259,
43
+ "learning_rate": 2.4002400000000002e-05,
44
+ "loss": 0.0688,
45
+ "step": 2500
46
+ },
47
+ {
48
+ "epoch": 1.0,
49
+ "eval_bleu": 62.03431583845565,
50
+ "eval_loss": 0.06440743803977966,
51
+ "eval_runtime": 243.3832,
52
+ "eval_samples_per_second": 6.163,
53
+ "eval_steps_per_second": 1.541,
54
+ "step": 2500
55
+ },
56
+ {
57
+ "epoch": 1.2,
58
+ "grad_norm": 0.24694935977458954,
59
+ "learning_rate": 2.28024e-05,
60
+ "loss": 0.0553,
61
+ "step": 3000
62
+ },
63
+ {
64
+ "epoch": 1.4,
65
+ "grad_norm": 0.26646921038627625,
66
+ "learning_rate": 2.1602400000000002e-05,
67
+ "loss": 0.0563,
68
+ "step": 3500
69
+ },
70
+ {
71
+ "epoch": 1.6,
72
+ "grad_norm": 0.25563061237335205,
73
+ "learning_rate": 2.04024e-05,
74
+ "loss": 0.0538,
75
+ "step": 4000
76
+ },
77
+ {
78
+ "epoch": 1.8,
79
+ "grad_norm": 0.25169041752815247,
80
+ "learning_rate": 1.92024e-05,
81
+ "loss": 0.0546,
82
+ "step": 4500
83
+ },
84
+ {
85
+ "epoch": 2.0,
86
+ "grad_norm": 0.2201855629682541,
87
+ "learning_rate": 1.80024e-05,
88
+ "loss": 0.0543,
89
+ "step": 5000
90
+ },
91
+ {
92
+ "epoch": 2.0,
93
+ "eval_bleu": 62.84538472555695,
94
+ "eval_loss": 0.06339649111032486,
95
+ "eval_runtime": 245.6581,
96
+ "eval_samples_per_second": 6.106,
97
+ "eval_steps_per_second": 1.527,
98
+ "step": 5000
99
+ },
100
+ {
101
+ "epoch": 2.2,
102
+ "grad_norm": 0.20503699779510498,
103
+ "learning_rate": 1.68024e-05,
104
+ "loss": 0.0436,
105
+ "step": 5500
106
+ },
107
+ {
108
+ "epoch": 2.4,
109
+ "grad_norm": 0.18902115523815155,
110
+ "learning_rate": 1.56024e-05,
111
+ "loss": 0.0444,
112
+ "step": 6000
113
+ },
114
+ {
115
+ "epoch": 2.6,
116
+ "grad_norm": 0.39264440536499023,
117
+ "learning_rate": 1.4402400000000001e-05,
118
+ "loss": 0.0434,
119
+ "step": 6500
120
+ },
121
+ {
122
+ "epoch": 2.8,
123
+ "grad_norm": 0.22416116297245026,
124
+ "learning_rate": 1.3202400000000001e-05,
125
+ "loss": 0.0443,
126
+ "step": 7000
127
+ },
128
+ {
129
+ "epoch": 3.0,
130
+ "grad_norm": 0.2807520925998688,
131
+ "learning_rate": 1.20024e-05,
132
+ "loss": 0.0444,
133
+ "step": 7500
134
+ },
135
+ {
136
+ "epoch": 3.0,
137
+ "eval_bleu": 63.39396750181061,
138
+ "eval_loss": 0.06433024257421494,
139
+ "eval_runtime": 245.8516,
140
+ "eval_samples_per_second": 6.101,
141
+ "eval_steps_per_second": 1.525,
142
+ "step": 7500
143
+ },
144
+ {
145
+ "epoch": 3.2,
146
+ "grad_norm": 0.2532879710197449,
147
+ "learning_rate": 1.08024e-05,
148
+ "loss": 0.0371,
149
+ "step": 8000
150
+ },
151
+ {
152
+ "epoch": 3.4,
153
+ "grad_norm": 0.20367200672626495,
154
+ "learning_rate": 9.6024e-06,
155
+ "loss": 0.036,
156
+ "step": 8500
157
+ },
158
+ {
159
+ "epoch": 3.6,
160
+ "grad_norm": 0.24893485009670258,
161
+ "learning_rate": 8.4024e-06,
162
+ "loss": 0.0364,
163
+ "step": 9000
164
+ },
165
+ {
166
+ "epoch": 3.8,
167
+ "grad_norm": 0.27603277564048767,
168
+ "learning_rate": 7.2024e-06,
169
+ "loss": 0.0366,
170
+ "step": 9500
171
+ },
172
+ {
173
+ "epoch": 4.0,
174
+ "grad_norm": 0.21856920421123505,
175
+ "learning_rate": 6.0024e-06,
176
+ "loss": 0.0377,
177
+ "step": 10000
178
+ },
179
+ {
180
+ "epoch": 4.0,
181
+ "eval_bleu": 63.44155467990254,
182
+ "eval_loss": 0.06584766507148743,
183
+ "eval_runtime": 246.2205,
184
+ "eval_samples_per_second": 6.092,
185
+ "eval_steps_per_second": 1.523,
186
+ "step": 10000
187
+ },
188
+ {
189
+ "epoch": 4.2,
190
+ "grad_norm": 0.21814288198947906,
191
+ "learning_rate": 4.8024e-06,
192
+ "loss": 0.0328,
193
+ "step": 10500
194
+ },
195
+ {
196
+ "epoch": 4.4,
197
+ "grad_norm": 0.20107074081897736,
198
+ "learning_rate": 3.6024000000000004e-06,
199
+ "loss": 0.0321,
200
+ "step": 11000
201
+ },
202
+ {
203
+ "epoch": 4.6,
204
+ "grad_norm": 0.2721525728702545,
205
+ "learning_rate": 2.4024e-06,
206
+ "loss": 0.0323,
207
+ "step": 11500
208
+ },
209
+ {
210
+ "epoch": 4.8,
211
+ "grad_norm": 0.33091649413108826,
212
+ "learning_rate": 1.2024e-06,
213
+ "loss": 0.0329,
214
+ "step": 12000
215
+ },
216
+ {
217
+ "epoch": 5.0,
218
+ "grad_norm": 0.2466353476047516,
219
+ "learning_rate": 2.4000000000000004e-09,
220
+ "loss": 0.0315,
221
+ "step": 12500
222
+ },
223
+ {
224
+ "epoch": 5.0,
225
+ "eval_bleu": 63.062347513148154,
226
+ "eval_loss": 0.06781521439552307,
227
+ "eval_runtime": 251.1652,
228
+ "eval_samples_per_second": 5.972,
229
+ "eval_steps_per_second": 1.493,
230
+ "step": 12500
231
+ },
232
+ {
233
+ "epoch": 5.0,
234
+ "step": 12500,
235
+ "total_flos": 2.70888075264e+16,
236
+ "train_loss": 0.04798975685119629,
237
+ "train_runtime": 11153.1781,
238
+ "train_samples_per_second": 8.966,
239
+ "train_steps_per_second": 1.121
240
+ }
241
+ ],
242
+ "logging_steps": 500,
243
+ "max_steps": 12500,
244
+ "num_input_tokens_seen": 0,
245
+ "num_train_epochs": 5,
246
+ "save_steps": 500,
247
+ "stateful_callbacks": {
248
+ "TrainerControl": {
249
+ "args": {
250
+ "should_epoch_stop": false,
251
+ "should_evaluate": false,
252
+ "should_log": false,
253
+ "should_save": true,
254
+ "should_training_stop": true
255
+ },
256
+ "attributes": {}
257
+ }
258
+ },
259
+ "total_flos": 2.70888075264e+16,
260
+ "train_batch_size": 4,
261
+ "trial_name": null,
262
+ "trial_params": null
263
+ }