Hugging Face's logo

wangrongsheng
/
CareLlama2-7b-chat-sft-multi

PEFT
Model card Files Community
wangrongsheng commited on
Commit
a97d6fd
•
1 Parent(s): b079eb3

commit from root

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. README.md +20 -0
  2. adapter_config.json +21 -0
  3. adapter_model.bin +3 -0
  4. all_results.json +7 -0
  5. checkpoint-1000/README.md +20 -0
  6. checkpoint-1000/adapter_config.json +21 -0
  7. checkpoint-1000/adapter_model.bin +3 -0
  8. checkpoint-1000/finetuning_args.json +16 -0
  9. checkpoint-1000/optimizer.pt +3 -0
  10. checkpoint-1000/rng_state.pth +3 -0
  11. checkpoint-1000/scheduler.pt +3 -0
  12. checkpoint-1000/trainer_state.json +616 -0
  13. checkpoint-1000/training_args.bin +3 -0
  14. checkpoint-2000/README.md +20 -0
  15. checkpoint-2000/adapter_config.json +21 -0
  16. checkpoint-2000/adapter_model.bin +3 -0
  17. checkpoint-2000/finetuning_args.json +16 -0
  18. checkpoint-2000/optimizer.pt +3 -0
  19. checkpoint-2000/rng_state.pth +3 -0
  20. checkpoint-2000/scheduler.pt +3 -0
  21. checkpoint-2000/trainer_state.json +1216 -0
  22. checkpoint-2000/training_args.bin +3 -0
  23. checkpoint-3000/README.md +20 -0
  24. checkpoint-3000/adapter_config.json +21 -0
  25. checkpoint-3000/adapter_model.bin +3 -0
  26. checkpoint-3000/finetuning_args.json +16 -0
  27. checkpoint-3000/optimizer.pt +3 -0
  28. checkpoint-3000/rng_state.pth +3 -0
  29. checkpoint-3000/scheduler.pt +3 -0
  30. checkpoint-3000/trainer_state.json +1816 -0
  31. checkpoint-3000/training_args.bin +3 -0
  32. checkpoint-4000/README.md +20 -0
  33. checkpoint-4000/adapter_config.json +21 -0
  34. checkpoint-4000/adapter_model.bin +3 -0
  35. checkpoint-4000/finetuning_args.json +16 -0
  36. checkpoint-4000/optimizer.pt +3 -0
  37. checkpoint-4000/rng_state.pth +3 -0
  38. checkpoint-4000/scheduler.pt +3 -0
  39. checkpoint-4000/trainer_state.json +2416 -0
  40. checkpoint-4000/training_args.bin +3 -0
  41. checkpoint-5000/README.md +20 -0
  42. checkpoint-5000/adapter_config.json +21 -0
  43. checkpoint-5000/adapter_model.bin +3 -0
  44. checkpoint-5000/finetuning_args.json +16 -0
  45. checkpoint-5000/optimizer.pt +3 -0
  46. checkpoint-5000/rng_state.pth +3 -0
  47. checkpoint-5000/scheduler.pt +3 -0
  48. checkpoint-5000/trainer_state.json +3016 -0
  49. checkpoint-5000/training_args.bin +3 -0
  50. checkpoint-6000/README.md +20 -0
README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "./Llama-2-7b-chat-hf",
4
+ "bias": "none",
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": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4385504f53e2af7c1a74d5219916e89bff507fe42d1f92fed316de91d257f9d
3
+ size 16821197
all_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.0,
3
+ "train_loss": 0.7045552698505881,
4
+ "train_runtime": 69586.6665,
5
+ "train_samples_per_second": 1.984,
6
+ "train_steps_per_second": 0.124
7
+ }
checkpoint-1000/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-1000/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "./Llama-2-7b-chat-hf",
4
+ "bias": "none",
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": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-1000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1971bd4b1027249696ed44f8bd7423a5c676f0e00c1518dd8ff56189b5baa0f9
3
+ size 16821197
checkpoint-1000/finetuning_args.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dpo_beta": 0.1,
3
+ "finetuning_type": "lora",
4
+ "lora_alpha": 32.0,
5
+ "lora_dropout": 0.1,
6
+ "lora_rank": 8,
7
+ "lora_target": [
8
+ "q_proj",
9
+ "v_proj"
10
+ ],
11
+ "name_module_trainable": "mlp",
12
+ "num_hidden_layers": 32,
13
+ "num_layer_trainable": 3,
14
+ "ppo_score_norm": false,
15
+ "resume_lora_training": true
16
+ }
checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:54704d419d264b492ca7ea46e8c032568229572e417454d6a9ac17635f8b0c5b
3
+ size 33661637
checkpoint-1000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c538fd6f76b188ed4f326f25da47706880a9892e5472c9849503968db129379
3
+ size 18663
checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6b3bd6984e103bebadae8a43822b167594bfea1e481bce2610d82fd4b4e0756
3
+ size 627
checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,616 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.23180343069077422,
5
+ "global_step": 1000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.999983427462679e-05,
13
+ "loss": 1.3476,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 4.999933710070436e-05,
19
+ "loss": 1.1454,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 4.999850848482424e-05,
25
+ "loss": 1.0263,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 4.999734843797226e-05,
31
+ "loss": 1.0283,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 4.9995856975528347e-05,
37
+ "loss": 0.9733,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 4.999403411726636e-05,
43
+ "loss": 0.9591,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 4.999187988735381e-05,
49
+ "loss": 0.9617,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 4.998939431435153e-05,
55
+ "loss": 0.9414,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 4.998657743121333e-05,
61
+ "loss": 0.9233,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 4.998342927528553e-05,
67
+ "loss": 0.9236,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 4.997994988830648e-05,
73
+ "loss": 0.8743,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 4.997613931640598e-05,
79
+ "loss": 0.895,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 4.997199761010472e-05,
85
+ "loss": 0.8831,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 4.996752482431356e-05,
91
+ "loss": 0.8799,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 4.996272101833284e-05,
97
+ "loss": 0.877,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 4.995758625585155e-05,
103
+ "loss": 0.8745,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 4.995212060494653e-05,
109
+ "loss": 0.8716,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 4.9946324138081534e-05,
115
+ "loss": 0.8534,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 4.99401969321063e-05,
121
+ "loss": 0.8449,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.05,
126
+ "learning_rate": 4.9933739068255513e-05,
127
+ "loss": 0.8588,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 4.992695063214772e-05,
133
+ "loss": 0.8409,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 4.99198317137842e-05,
139
+ "loss": 0.8517,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 4.99123824075478e-05,
145
+ "loss": 0.8698,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.06,
150
+ "learning_rate": 4.990460281220164e-05,
151
+ "loss": 0.848,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.06,
156
+ "learning_rate": 4.9896493030887815e-05,
157
+ "loss": 0.8322,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 4.988805317112607e-05,
163
+ "loss": 0.8376,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 4.98792833448123e-05,
169
+ "loss": 0.8309,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 4.987018366821713e-05,
175
+ "loss": 0.8195,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.07,
180
+ "learning_rate": 4.9860754261984346e-05,
181
+ "loss": 0.8313,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 4.98509952511293e-05,
187
+ "loss": 0.8115,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 4.984090676503723e-05,
193
+ "loss": 0.8224,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 4.983048893746162e-05,
199
+ "loss": 0.8157,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.08,
204
+ "learning_rate": 4.9819741906522306e-05,
205
+ "loss": 0.8165,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.08,
210
+ "learning_rate": 4.980866581470376e-05,
211
+ "loss": 0.8144,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 4.979726080885314e-05,
217
+ "loss": 0.8147,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 4.978552704017835e-05,
223
+ "loss": 0.7894,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.09,
228
+ "learning_rate": 4.977346466424605e-05,
229
+ "loss": 0.8144,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.09,
234
+ "learning_rate": 4.976107384097957e-05,
235
+ "loss": 0.7984,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 4.974835473465682e-05,
241
+ "loss": 0.8188,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 4.97353075139081e-05,
247
+ "loss": 0.8049,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.1,
252
+ "learning_rate": 4.9721932351713846e-05,
253
+ "loss": 0.8154,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.1,
258
+ "learning_rate": 4.9708229425402354e-05,
259
+ "loss": 0.7953,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.1,
264
+ "learning_rate": 4.9694198916647425e-05,
265
+ "loss": 0.7996,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 4.967984101146598e-05,
271
+ "loss": 0.7979,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 4.966515590021554e-05,
277
+ "loss": 0.8106,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.11,
282
+ "learning_rate": 4.965014377759176e-05,
283
+ "loss": 0.8008,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.11,
288
+ "learning_rate": 4.9634804842625807e-05,
289
+ "loss": 0.8045,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 4.961913929868172e-05,
295
+ "loss": 0.786,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 4.960314735345377e-05,
301
+ "loss": 0.7897,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.12,
306
+ "learning_rate": 4.958682921896365e-05,
307
+ "loss": 0.7855,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.12,
312
+ "learning_rate": 4.957018511155765e-05,
313
+ "loss": 0.7657,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.12,
318
+ "learning_rate": 4.955321525190386e-05,
319
+ "loss": 0.779,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 4.953591986498918e-05,
325
+ "loss": 0.7923,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.13,
330
+ "learning_rate": 4.9518299180116366e-05,
331
+ "loss": 0.7732,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.13,
336
+ "learning_rate": 4.9500353430900984e-05,
337
+ "loss": 0.7898,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.13,
342
+ "learning_rate": 4.948208285526832e-05,
343
+ "loss": 0.7796,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.13,
348
+ "learning_rate": 4.94634876954502e-05,
349
+ "loss": 0.7819,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 4.944456819798182e-05,
355
+ "loss": 0.7777,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.14,
360
+ "learning_rate": 4.942532461369843e-05,
361
+ "loss": 0.7815,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.14,
366
+ "learning_rate": 4.940575719773205e-05,
367
+ "loss": 0.7807,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.14,
372
+ "learning_rate": 4.938586620950807e-05,
373
+ "loss": 0.7836,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 4.9365651912741806e-05,
379
+ "loss": 0.7795,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.15,
384
+ "learning_rate": 4.934511457543499e-05,
385
+ "loss": 0.7844,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.15,
390
+ "learning_rate": 4.932425446987228e-05,
391
+ "loss": 0.7865,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.15,
396
+ "learning_rate": 4.930307187261757e-05,
397
+ "loss": 0.7751,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.15,
402
+ "learning_rate": 4.928156706451035e-05,
403
+ "loss": 0.7705,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.16,
408
+ "learning_rate": 4.925974033066202e-05,
409
+ "loss": 0.773,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.16,
414
+ "learning_rate": 4.923759196045208e-05,
415
+ "loss": 0.7714,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.16,
420
+ "learning_rate": 4.921512224752426e-05,
421
+ "loss": 0.7788,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.16,
426
+ "learning_rate": 4.9192331489782705e-05,
427
+ "loss": 0.7755,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 4.9169219989387956e-05,
433
+ "loss": 0.7682,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.17,
438
+ "learning_rate": 4.9145788052752975e-05,
439
+ "loss": 0.788,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.17,
444
+ "learning_rate": 4.912203599053907e-05,
445
+ "loss": 0.7752,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.17,
450
+ "learning_rate": 4.90979641176518e-05,
451
+ "loss": 0.7786,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.17,
456
+ "learning_rate": 4.907357275323677e-05,
457
+ "loss": 0.7848,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.18,
462
+ "learning_rate": 4.904886222067542e-05,
463
+ "loss": 0.7747,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.18,
468
+ "learning_rate": 4.9023832847580724e-05,
469
+ "loss": 0.7784,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.18,
474
+ "learning_rate": 4.899848496579287e-05,
475
+ "loss": 0.768,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.18,
480
+ "learning_rate": 4.897281891137481e-05,
481
+ "loss": 0.7684,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.19,
486
+ "learning_rate": 4.8946835024607885e-05,
487
+ "loss": 0.7696,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.19,
492
+ "learning_rate": 4.892053364998722e-05,
493
+ "loss": 0.7552,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.19,
498
+ "learning_rate": 4.889391513621725e-05,
499
+ "loss": 0.7584,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.19,
504
+ "learning_rate": 4.886697983620699e-05,
505
+ "loss": 0.7682,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.19,
510
+ "learning_rate": 4.883972810706548e-05,
511
+ "loss": 0.7701,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.2,
516
+ "learning_rate": 4.881216031009696e-05,
517
+ "loss": 0.7637,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.2,
522
+ "learning_rate": 4.878427681079608e-05,
523
+ "loss": 0.7673,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.2,
528
+ "learning_rate": 4.875607797884312e-05,
529
+ "loss": 0.7579,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.2,
534
+ "learning_rate": 4.8727564188099054e-05,
535
+ "loss": 0.7629,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.21,
540
+ "learning_rate": 4.8698735816600535e-05,
541
+ "loss": 0.7766,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.21,
546
+ "learning_rate": 4.866959324655501e-05,
547
+ "loss": 0.7604,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.21,
552
+ "learning_rate": 4.864013686433552e-05,
553
+ "loss": 0.7515,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.21,
558
+ "learning_rate": 4.861036706047566e-05,
559
+ "loss": 0.7508,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.22,
564
+ "learning_rate": 4.858028422966439e-05,
565
+ "loss": 0.7559,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.22,
570
+ "learning_rate": 4.854988877074076e-05,
571
+ "loss": 0.7684,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.22,
576
+ "learning_rate": 4.8519181086688695e-05,
577
+ "loss": 0.7514,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.22,
582
+ "learning_rate": 4.8488161584631574e-05,
583
+ "loss": 0.7543,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.22,
588
+ "learning_rate": 4.845683067582689e-05,
589
+ "loss": 0.7651,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.23,
594
+ "learning_rate": 4.842518877566076e-05,
595
+ "loss": 0.7634,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.23,
600
+ "learning_rate": 4.839323630364243e-05,
601
+ "loss": 0.7477,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.23,
606
+ "learning_rate": 4.836097368339875e-05,
607
+ "loss": 0.7559,
608
+ "step": 1000
609
+ }
610
+ ],
611
+ "max_steps": 8628,
612
+ "num_train_epochs": 2,
613
+ "total_flos": 3.1584515936537805e+17,
614
+ "trial_name": null,
615
+ "trial_params": null
616
+ }
checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ce444f5cfe8a7851710b073da2f9ede89d9fad5a6d091ebf47aee5e6b0d0797
3
+ size 3259
checkpoint-2000/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-2000/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "./Llama-2-7b-chat-hf",
4
+ "bias": "none",
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": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-2000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59027433e7e72159b578e4981569825719b027c6ff576cc7006c861b71506f7f
3
+ size 16821197
checkpoint-2000/finetuning_args.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dpo_beta": 0.1,
3
+ "finetuning_type": "lora",
4
+ "lora_alpha": 32.0,
5
+ "lora_dropout": 0.1,
6
+ "lora_rank": 8,
7
+ "lora_target": [
8
+ "q_proj",
9
+ "v_proj"
10
+ ],
11
+ "name_module_trainable": "mlp",
12
+ "num_hidden_layers": 32,
13
+ "num_layer_trainable": 3,
14
+ "ppo_score_norm": false,
15
+ "resume_lora_training": true
16
+ }
checkpoint-2000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44fd8d3837d11c723cb03ee6d40a0fa9641bc4fed83134ac9b0d913bce1acbcd
3
+ size 33661637
checkpoint-2000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d80ae3653f3a6d6e870eb73d506a8b7b69fc19d5b22dfdb97c2ce76cec263a4e
3
+ size 18663
checkpoint-2000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cad3bc48d491186d577a52b69ed0c392a35794bf8c9c59290f6422da713a090f
3
+ size 627
checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,1216 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.46360686138154844,
5
+ "global_step": 2000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.999983427462679e-05,
13
+ "loss": 1.3476,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 4.999933710070436e-05,
19
+ "loss": 1.1454,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 4.999850848482424e-05,
25
+ "loss": 1.0263,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 4.999734843797226e-05,
31
+ "loss": 1.0283,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 4.9995856975528347e-05,
37
+ "loss": 0.9733,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 4.999403411726636e-05,
43
+ "loss": 0.9591,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 4.999187988735381e-05,
49
+ "loss": 0.9617,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 4.998939431435153e-05,
55
+ "loss": 0.9414,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 4.998657743121333e-05,
61
+ "loss": 0.9233,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 4.998342927528553e-05,
67
+ "loss": 0.9236,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 4.997994988830648e-05,
73
+ "loss": 0.8743,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 4.997613931640598e-05,
79
+ "loss": 0.895,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 4.997199761010472e-05,
85
+ "loss": 0.8831,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 4.996752482431356e-05,
91
+ "loss": 0.8799,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 4.996272101833284e-05,
97
+ "loss": 0.877,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 4.995758625585155e-05,
103
+ "loss": 0.8745,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 4.995212060494653e-05,
109
+ "loss": 0.8716,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 4.9946324138081534e-05,
115
+ "loss": 0.8534,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 4.99401969321063e-05,
121
+ "loss": 0.8449,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.05,
126
+ "learning_rate": 4.9933739068255513e-05,
127
+ "loss": 0.8588,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 4.992695063214772e-05,
133
+ "loss": 0.8409,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 4.99198317137842e-05,
139
+ "loss": 0.8517,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 4.99123824075478e-05,
145
+ "loss": 0.8698,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.06,
150
+ "learning_rate": 4.990460281220164e-05,
151
+ "loss": 0.848,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.06,
156
+ "learning_rate": 4.9896493030887815e-05,
157
+ "loss": 0.8322,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 4.988805317112607e-05,
163
+ "loss": 0.8376,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 4.98792833448123e-05,
169
+ "loss": 0.8309,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 4.987018366821713e-05,
175
+ "loss": 0.8195,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.07,
180
+ "learning_rate": 4.9860754261984346e-05,
181
+ "loss": 0.8313,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 4.98509952511293e-05,
187
+ "loss": 0.8115,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 4.984090676503723e-05,
193
+ "loss": 0.8224,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 4.983048893746162e-05,
199
+ "loss": 0.8157,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.08,
204
+ "learning_rate": 4.9819741906522306e-05,
205
+ "loss": 0.8165,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.08,
210
+ "learning_rate": 4.980866581470376e-05,
211
+ "loss": 0.8144,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 4.979726080885314e-05,
217
+ "loss": 0.8147,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 4.978552704017835e-05,
223
+ "loss": 0.7894,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.09,
228
+ "learning_rate": 4.977346466424605e-05,
229
+ "loss": 0.8144,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.09,
234
+ "learning_rate": 4.976107384097957e-05,
235
+ "loss": 0.7984,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 4.974835473465682e-05,
241
+ "loss": 0.8188,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 4.97353075139081e-05,
247
+ "loss": 0.8049,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.1,
252
+ "learning_rate": 4.9721932351713846e-05,
253
+ "loss": 0.8154,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.1,
258
+ "learning_rate": 4.9708229425402354e-05,
259
+ "loss": 0.7953,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.1,
264
+ "learning_rate": 4.9694198916647425e-05,
265
+ "loss": 0.7996,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 4.967984101146598e-05,
271
+ "loss": 0.7979,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 4.966515590021554e-05,
277
+ "loss": 0.8106,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.11,
282
+ "learning_rate": 4.965014377759176e-05,
283
+ "loss": 0.8008,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.11,
288
+ "learning_rate": 4.9634804842625807e-05,
289
+ "loss": 0.8045,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 4.961913929868172e-05,
295
+ "loss": 0.786,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 4.960314735345377e-05,
301
+ "loss": 0.7897,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.12,
306
+ "learning_rate": 4.958682921896365e-05,
307
+ "loss": 0.7855,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.12,
312
+ "learning_rate": 4.957018511155765e-05,
313
+ "loss": 0.7657,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.12,
318
+ "learning_rate": 4.955321525190386e-05,
319
+ "loss": 0.779,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 4.953591986498918e-05,
325
+ "loss": 0.7923,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.13,
330
+ "learning_rate": 4.9518299180116366e-05,
331
+ "loss": 0.7732,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.13,
336
+ "learning_rate": 4.9500353430900984e-05,
337
+ "loss": 0.7898,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.13,
342
+ "learning_rate": 4.948208285526832e-05,
343
+ "loss": 0.7796,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.13,
348
+ "learning_rate": 4.94634876954502e-05,
349
+ "loss": 0.7819,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 4.944456819798182e-05,
355
+ "loss": 0.7777,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.14,
360
+ "learning_rate": 4.942532461369843e-05,
361
+ "loss": 0.7815,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.14,
366
+ "learning_rate": 4.940575719773205e-05,
367
+ "loss": 0.7807,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.14,
372
+ "learning_rate": 4.938586620950807e-05,
373
+ "loss": 0.7836,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 4.9365651912741806e-05,
379
+ "loss": 0.7795,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.15,
384
+ "learning_rate": 4.934511457543499e-05,
385
+ "loss": 0.7844,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.15,
390
+ "learning_rate": 4.932425446987228e-05,
391
+ "loss": 0.7865,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.15,
396
+ "learning_rate": 4.930307187261757e-05,
397
+ "loss": 0.7751,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.15,
402
+ "learning_rate": 4.928156706451035e-05,
403
+ "loss": 0.7705,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.16,
408
+ "learning_rate": 4.925974033066202e-05,
409
+ "loss": 0.773,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.16,
414
+ "learning_rate": 4.923759196045208e-05,
415
+ "loss": 0.7714,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.16,
420
+ "learning_rate": 4.921512224752426e-05,
421
+ "loss": 0.7788,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.16,
426
+ "learning_rate": 4.9192331489782705e-05,
427
+ "loss": 0.7755,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 4.9169219989387956e-05,
433
+ "loss": 0.7682,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.17,
438
+ "learning_rate": 4.9145788052752975e-05,
439
+ "loss": 0.788,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.17,
444
+ "learning_rate": 4.912203599053907e-05,
445
+ "loss": 0.7752,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.17,
450
+ "learning_rate": 4.90979641176518e-05,
451
+ "loss": 0.7786,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.17,
456
+ "learning_rate": 4.907357275323677e-05,
457
+ "loss": 0.7848,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.18,
462
+ "learning_rate": 4.904886222067542e-05,
463
+ "loss": 0.7747,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.18,
468
+ "learning_rate": 4.9023832847580724e-05,
469
+ "loss": 0.7784,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.18,
474
+ "learning_rate": 4.899848496579287e-05,
475
+ "loss": 0.768,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.18,
480
+ "learning_rate": 4.897281891137481e-05,
481
+ "loss": 0.7684,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.19,
486
+ "learning_rate": 4.8946835024607885e-05,
487
+ "loss": 0.7696,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.19,
492
+ "learning_rate": 4.892053364998722e-05,
493
+ "loss": 0.7552,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.19,
498
+ "learning_rate": 4.889391513621725e-05,
499
+ "loss": 0.7584,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.19,
504
+ "learning_rate": 4.886697983620699e-05,
505
+ "loss": 0.7682,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.19,
510
+ "learning_rate": 4.883972810706548e-05,
511
+ "loss": 0.7701,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.2,
516
+ "learning_rate": 4.881216031009696e-05,
517
+ "loss": 0.7637,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.2,
522
+ "learning_rate": 4.878427681079608e-05,
523
+ "loss": 0.7673,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.2,
528
+ "learning_rate": 4.875607797884312e-05,
529
+ "loss": 0.7579,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.2,
534
+ "learning_rate": 4.8727564188099054e-05,
535
+ "loss": 0.7629,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.21,
540
+ "learning_rate": 4.8698735816600535e-05,
541
+ "loss": 0.7766,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.21,
546
+ "learning_rate": 4.866959324655501e-05,
547
+ "loss": 0.7604,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.21,
552
+ "learning_rate": 4.864013686433552e-05,
553
+ "loss": 0.7515,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.21,
558
+ "learning_rate": 4.861036706047566e-05,
559
+ "loss": 0.7508,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.22,
564
+ "learning_rate": 4.858028422966439e-05,
565
+ "loss": 0.7559,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.22,
570
+ "learning_rate": 4.854988877074076e-05,
571
+ "loss": 0.7684,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.22,
576
+ "learning_rate": 4.8519181086688695e-05,
577
+ "loss": 0.7514,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.22,
582
+ "learning_rate": 4.8488161584631574e-05,
583
+ "loss": 0.7543,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.22,
588
+ "learning_rate": 4.845683067582689e-05,
589
+ "loss": 0.7651,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.23,
594
+ "learning_rate": 4.842518877566076e-05,
595
+ "loss": 0.7634,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.23,
600
+ "learning_rate": 4.839323630364243e-05,
601
+ "loss": 0.7477,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.23,
606
+ "learning_rate": 4.836097368339875e-05,
607
+ "loss": 0.7559,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.23,
612
+ "learning_rate": 4.832840134266849e-05,
613
+ "loss": 0.7729,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.24,
618
+ "learning_rate": 4.829551971329672e-05,
619
+ "loss": 0.7518,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.24,
624
+ "learning_rate": 4.826232923122907e-05,
625
+ "loss": 0.7551,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.24,
630
+ "learning_rate": 4.8228830336505926e-05,
631
+ "loss": 0.7613,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.24,
636
+ "learning_rate": 4.819502347325665e-05,
637
+ "loss": 0.7529,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.25,
642
+ "learning_rate": 4.816090908969363e-05,
643
+ "loss": 0.7476,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.25,
648
+ "learning_rate": 4.81264876381064e-05,
649
+ "loss": 0.765,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.25,
654
+ "learning_rate": 4.809175957485558e-05,
655
+ "loss": 0.7417,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.25,
660
+ "learning_rate": 4.805672536036686e-05,
661
+ "loss": 0.7398,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.25,
666
+ "learning_rate": 4.8021385459124925e-05,
667
+ "loss": 0.7489,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.26,
672
+ "learning_rate": 4.798574033966723e-05,
673
+ "loss": 0.7546,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.26,
678
+ "learning_rate": 4.794979047457783e-05,
679
+ "loss": 0.738,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.26,
684
+ "learning_rate": 4.79135363404811e-05,
685
+ "loss": 0.7495,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.26,
690
+ "learning_rate": 4.7876978418035455e-05,
691
+ "loss": 0.7622,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.27,
696
+ "learning_rate": 4.7840117191926915e-05,
697
+ "loss": 0.7465,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.27,
702
+ "learning_rate": 4.780295315086271e-05,
703
+ "loss": 0.7446,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.27,
708
+ "learning_rate": 4.7765486787564814e-05,
709
+ "loss": 0.7425,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.27,
714
+ "learning_rate": 4.772771859876338e-05,
715
+ "loss": 0.7583,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.28,
720
+ "learning_rate": 4.768964908519019e-05,
721
+ "loss": 0.7579,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.28,
726
+ "learning_rate": 4.7651278751571984e-05,
727
+ "loss": 0.7509,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.28,
732
+ "learning_rate": 4.76126081066238e-05,
733
+ "loss": 0.7358,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.28,
738
+ "learning_rate": 4.7573637663042195e-05,
739
+ "loss": 0.7409,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.29,
744
+ "learning_rate": 4.753436793749848e-05,
745
+ "loss": 0.7289,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.29,
750
+ "learning_rate": 4.749479945063186e-05,
751
+ "loss": 0.7488,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.29,
756
+ "learning_rate": 4.745493272704249e-05,
757
+ "loss": 0.7514,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.29,
762
+ "learning_rate": 4.7414768295284594e-05,
763
+ "loss": 0.7384,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.29,
768
+ "learning_rate": 4.7374306687859405e-05,
769
+ "loss": 0.739,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.3,
774
+ "learning_rate": 4.7333548441208134e-05,
775
+ "loss": 0.7494,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.3,
780
+ "learning_rate": 4.729249409570481e-05,
781
+ "loss": 0.7503,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.3,
786
+ "learning_rate": 4.725114419564919e-05,
787
+ "loss": 0.7388,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.3,
792
+ "learning_rate": 4.720949928925948e-05,
793
+ "loss": 0.7601,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.31,
798
+ "learning_rate": 4.716755992866509e-05,
799
+ "loss": 0.7416,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.31,
804
+ "learning_rate": 4.712532666989931e-05,
805
+ "loss": 0.7365,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.31,
810
+ "learning_rate": 4.708280007289195e-05,
811
+ "loss": 0.7455,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.31,
816
+ "learning_rate": 4.703998070146191e-05,
817
+ "loss": 0.7342,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.32,
822
+ "learning_rate": 4.699686912330969e-05,
823
+ "loss": 0.7374,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.32,
828
+ "learning_rate": 4.695346591000986e-05,
829
+ "loss": 0.7312,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.32,
834
+ "learning_rate": 4.6909771637003556e-05,
835
+ "loss": 0.7446,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.32,
840
+ "learning_rate": 4.686578688359072e-05,
841
+ "loss": 0.7387,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.32,
846
+ "learning_rate": 4.6821512232922546e-05,
847
+ "loss": 0.73,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.33,
852
+ "learning_rate": 4.677694827199367e-05,
853
+ "loss": 0.7357,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.33,
858
+ "learning_rate": 4.673209559163442e-05,
859
+ "loss": 0.7441,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.33,
864
+ "learning_rate": 4.668695478650296e-05,
865
+ "loss": 0.7359,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.33,
870
+ "learning_rate": 4.664152645507744e-05,
871
+ "loss": 0.7467,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.34,
876
+ "learning_rate": 4.659581119964803e-05,
877
+ "loss": 0.7221,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.34,
882
+ "learning_rate": 4.654980962630896e-05,
883
+ "loss": 0.7334,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.34,
888
+ "learning_rate": 4.6503522344950454e-05,
889
+ "loss": 0.7405,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.34,
894
+ "learning_rate": 4.6456949969250674e-05,
895
+ "loss": 0.7424,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.35,
900
+ "learning_rate": 4.6410093116667566e-05,
901
+ "loss": 0.7378,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.35,
906
+ "learning_rate": 4.636295240843068e-05,
907
+ "loss": 0.7479,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.35,
912
+ "learning_rate": 4.6315528469532945e-05,
913
+ "loss": 0.7479,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.35,
918
+ "learning_rate": 4.6267821928722345e-05,
919
+ "loss": 0.7337,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.35,
924
+ "learning_rate": 4.621983341849362e-05,
925
+ "loss": 0.7403,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.36,
930
+ "learning_rate": 4.6171563575079876e-05,
931
+ "loss": 0.7353,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.36,
936
+ "learning_rate": 4.612301303844415e-05,
937
+ "loss": 0.7237,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.36,
942
+ "learning_rate": 4.607418245227089e-05,
943
+ "loss": 0.739,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.36,
948
+ "learning_rate": 4.602507246395747e-05,
949
+ "loss": 0.7292,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.37,
954
+ "learning_rate": 4.5975683724605584e-05,
955
+ "loss": 0.745,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.37,
960
+ "learning_rate": 4.5926016889012616e-05,
961
+ "loss": 0.7276,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.37,
966
+ "learning_rate": 4.587607261566294e-05,
967
+ "loss": 0.7481,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.37,
972
+ "learning_rate": 4.582585156671925e-05,
973
+ "loss": 0.7362,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.38,
978
+ "learning_rate": 4.577535440801368e-05,
979
+ "loss": 0.734,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.38,
984
+ "learning_rate": 4.572458180903909e-05,
985
+ "loss": 0.7428,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.38,
990
+ "learning_rate": 4.5673534442940113e-05,
991
+ "loss": 0.732,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.38,
996
+ "learning_rate": 4.5622212986504245e-05,
997
+ "loss": 0.7191,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.38,
1002
+ "learning_rate": 4.557061812015289e-05,
1003
+ "loss": 0.727,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.39,
1008
+ "learning_rate": 4.551875052793232e-05,
1009
+ "loss": 0.728,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.39,
1014
+ "learning_rate": 4.546661089750462e-05,
1015
+ "loss": 0.7366,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.39,
1020
+ "learning_rate": 4.541419992013859e-05,
1021
+ "loss": 0.7274,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.39,
1026
+ "learning_rate": 4.536151829070052e-05,
1027
+ "loss": 0.7185,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.4,
1032
+ "learning_rate": 4.530856670764502e-05,
1033
+ "loss": 0.7463,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.4,
1038
+ "learning_rate": 4.525534587300575e-05,
1039
+ "loss": 0.734,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.4,
1044
+ "learning_rate": 4.520185649238615e-05,
1045
+ "loss": 0.7325,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.4,
1050
+ "learning_rate": 4.5148099274950005e-05,
1051
+ "loss": 0.7429,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.41,
1056
+ "learning_rate": 4.509407493341212e-05,
1057
+ "loss": 0.7333,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.41,
1062
+ "learning_rate": 4.503978418402883e-05,
1063
+ "loss": 0.7227,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.41,
1068
+ "learning_rate": 4.49852277465885e-05,
1069
+ "loss": 0.7139,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.41,
1074
+ "learning_rate": 4.4930406344402006e-05,
1075
+ "loss": 0.7387,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.41,
1080
+ "learning_rate": 4.487532070429315e-05,
1081
+ "loss": 0.7276,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.42,
1086
+ "learning_rate": 4.481997155658898e-05,
1087
+ "loss": 0.7251,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.42,
1092
+ "learning_rate": 4.476435963511015e-05,
1093
+ "loss": 0.7287,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.42,
1098
+ "learning_rate": 4.470848567716119e-05,
1099
+ "loss": 0.7169,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.42,
1104
+ "learning_rate": 4.4652350423520685e-05,
1105
+ "loss": 0.725,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.43,
1110
+ "learning_rate": 4.45959546184315e-05,
1111
+ "loss": 0.7089,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.43,
1116
+ "learning_rate": 4.4539299009590926e-05,
1117
+ "loss": 0.722,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.43,
1122
+ "learning_rate": 4.4482384348140695e-05,
1123
+ "loss": 0.7197,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.43,
1128
+ "learning_rate": 4.44252113886571e-05,
1129
+ "loss": 0.7135,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.44,
1134
+ "learning_rate": 4.4367780889140945e-05,
1135
+ "loss": 0.7168,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.44,
1140
+ "learning_rate": 4.431009361100748e-05,
1141
+ "loss": 0.7267,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.44,
1146
+ "learning_rate": 4.4252150319076416e-05,
1147
+ "loss": 0.7241,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.44,
1152
+ "learning_rate": 4.419395178156159e-05,
1153
+ "loss": 0.7294,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.45,
1158
+ "learning_rate": 4.413549877006099e-05,
1159
+ "loss": 0.7392,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.45,
1164
+ "learning_rate": 4.4076792059546365e-05,
1165
+ "loss": 0.7058,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.45,
1170
+ "learning_rate": 4.401783242835306e-05,
1171
+ "loss": 0.721,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.45,
1176
+ "learning_rate": 4.395862065816959e-05,
1177
+ "loss": 0.7284,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.45,
1182
+ "learning_rate": 4.3899157534027396e-05,
1183
+ "loss": 0.7259,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.46,
1188
+ "learning_rate": 4.383944384429035e-05,
1189
+ "loss": 0.7378,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.46,
1194
+ "learning_rate": 4.3779480380644337e-05,
1195
+ "loss": 0.7118,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.46,
1200
+ "learning_rate": 4.3719267938086736e-05,
1201
+ "loss": 0.7318,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.46,
1206
+ "learning_rate": 4.3658807314915915e-05,
1207
+ "loss": 0.7335,
1208
+ "step": 2000
1209
+ }
1210
+ ],
1211
+ "max_steps": 8628,
1212
+ "num_train_epochs": 2,
1213
+ "total_flos": 6.316517778417254e+17,
1214
+ "trial_name": null,
1215
+ "trial_params": null
1216
+ }
checkpoint-2000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ce444f5cfe8a7851710b073da2f9ede89d9fad5a6d091ebf47aee5e6b0d0797
3
+ size 3259
checkpoint-3000/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-3000/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "./Llama-2-7b-chat-hf",
4
+ "bias": "none",
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": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-3000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7104b0c2e20a86f70d274d994525e3a78d04c0460fcff0cd49912ee2f037b17a
3
+ size 16821197
checkpoint-3000/finetuning_args.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dpo_beta": 0.1,
3
+ "finetuning_type": "lora",
4
+ "lora_alpha": 32.0,
5
+ "lora_dropout": 0.1,
6
+ "lora_rank": 8,
7
+ "lora_target": [
8
+ "q_proj",
9
+ "v_proj"
10
+ ],
11
+ "name_module_trainable": "mlp",
12
+ "num_hidden_layers": 32,
13
+ "num_layer_trainable": 3,
14
+ "ppo_score_norm": false,
15
+ "resume_lora_training": true
16
+ }
checkpoint-3000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6597eb71b48e043e51cf6e13d5a556f5a083db56f4d1166f344bc26731ff95d3
3
+ size 33661637
checkpoint-3000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e6a2a9803b60640592b261dfbc783aaaf10a005e3d15752bcf0f10866c127de
3
+ size 18663
checkpoint-3000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17207ad0374cc469cb6a6cdc56c80cf6a98deb305746fb2b5c4267d65d7c273f
3
+ size 627
checkpoint-3000/trainer_state.json ADDED
@@ -0,0 +1,1816 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.6954102920723226,
5
+ "global_step": 3000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.999983427462679e-05,
13
+ "loss": 1.3476,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 4.999933710070436e-05,
19
+ "loss": 1.1454,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 4.999850848482424e-05,
25
+ "loss": 1.0263,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 4.999734843797226e-05,
31
+ "loss": 1.0283,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 4.9995856975528347e-05,
37
+ "loss": 0.9733,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 4.999403411726636e-05,
43
+ "loss": 0.9591,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 4.999187988735381e-05,
49
+ "loss": 0.9617,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 4.998939431435153e-05,
55
+ "loss": 0.9414,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 4.998657743121333e-05,
61
+ "loss": 0.9233,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 4.998342927528553e-05,
67
+ "loss": 0.9236,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 4.997994988830648e-05,
73
+ "loss": 0.8743,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 4.997613931640598e-05,
79
+ "loss": 0.895,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 4.997199761010472e-05,
85
+ "loss": 0.8831,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 4.996752482431356e-05,
91
+ "loss": 0.8799,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 4.996272101833284e-05,
97
+ "loss": 0.877,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 4.995758625585155e-05,
103
+ "loss": 0.8745,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 4.995212060494653e-05,
109
+ "loss": 0.8716,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 4.9946324138081534e-05,
115
+ "loss": 0.8534,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 4.99401969321063e-05,
121
+ "loss": 0.8449,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.05,
126
+ "learning_rate": 4.9933739068255513e-05,
127
+ "loss": 0.8588,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 4.992695063214772e-05,
133
+ "loss": 0.8409,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 4.99198317137842e-05,
139
+ "loss": 0.8517,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 4.99123824075478e-05,
145
+ "loss": 0.8698,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.06,
150
+ "learning_rate": 4.990460281220164e-05,
151
+ "loss": 0.848,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.06,
156
+ "learning_rate": 4.9896493030887815e-05,
157
+ "loss": 0.8322,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 4.988805317112607e-05,
163
+ "loss": 0.8376,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 4.98792833448123e-05,
169
+ "loss": 0.8309,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 4.987018366821713e-05,
175
+ "loss": 0.8195,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.07,
180
+ "learning_rate": 4.9860754261984346e-05,
181
+ "loss": 0.8313,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 4.98509952511293e-05,
187
+ "loss": 0.8115,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 4.984090676503723e-05,
193
+ "loss": 0.8224,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 4.983048893746162e-05,
199
+ "loss": 0.8157,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.08,
204
+ "learning_rate": 4.9819741906522306e-05,
205
+ "loss": 0.8165,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.08,
210
+ "learning_rate": 4.980866581470376e-05,
211
+ "loss": 0.8144,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 4.979726080885314e-05,
217
+ "loss": 0.8147,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 4.978552704017835e-05,
223
+ "loss": 0.7894,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.09,
228
+ "learning_rate": 4.977346466424605e-05,
229
+ "loss": 0.8144,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.09,
234
+ "learning_rate": 4.976107384097957e-05,
235
+ "loss": 0.7984,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 4.974835473465682e-05,
241
+ "loss": 0.8188,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 4.97353075139081e-05,
247
+ "loss": 0.8049,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.1,
252
+ "learning_rate": 4.9721932351713846e-05,
253
+ "loss": 0.8154,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.1,
258
+ "learning_rate": 4.9708229425402354e-05,
259
+ "loss": 0.7953,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.1,
264
+ "learning_rate": 4.9694198916647425e-05,
265
+ "loss": 0.7996,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 4.967984101146598e-05,
271
+ "loss": 0.7979,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 4.966515590021554e-05,
277
+ "loss": 0.8106,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.11,
282
+ "learning_rate": 4.965014377759176e-05,
283
+ "loss": 0.8008,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.11,
288
+ "learning_rate": 4.9634804842625807e-05,
289
+ "loss": 0.8045,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 4.961913929868172e-05,
295
+ "loss": 0.786,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 4.960314735345377e-05,
301
+ "loss": 0.7897,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.12,
306
+ "learning_rate": 4.958682921896365e-05,
307
+ "loss": 0.7855,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.12,
312
+ "learning_rate": 4.957018511155765e-05,
313
+ "loss": 0.7657,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.12,
318
+ "learning_rate": 4.955321525190386e-05,
319
+ "loss": 0.779,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 4.953591986498918e-05,
325
+ "loss": 0.7923,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.13,
330
+ "learning_rate": 4.9518299180116366e-05,
331
+ "loss": 0.7732,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.13,
336
+ "learning_rate": 4.9500353430900984e-05,
337
+ "loss": 0.7898,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.13,
342
+ "learning_rate": 4.948208285526832e-05,
343
+ "loss": 0.7796,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.13,
348
+ "learning_rate": 4.94634876954502e-05,
349
+ "loss": 0.7819,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 4.944456819798182e-05,
355
+ "loss": 0.7777,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.14,
360
+ "learning_rate": 4.942532461369843e-05,
361
+ "loss": 0.7815,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.14,
366
+ "learning_rate": 4.940575719773205e-05,
367
+ "loss": 0.7807,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.14,
372
+ "learning_rate": 4.938586620950807e-05,
373
+ "loss": 0.7836,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 4.9365651912741806e-05,
379
+ "loss": 0.7795,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.15,
384
+ "learning_rate": 4.934511457543499e-05,
385
+ "loss": 0.7844,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.15,
390
+ "learning_rate": 4.932425446987228e-05,
391
+ "loss": 0.7865,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.15,
396
+ "learning_rate": 4.930307187261757e-05,
397
+ "loss": 0.7751,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.15,
402
+ "learning_rate": 4.928156706451035e-05,
403
+ "loss": 0.7705,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.16,
408
+ "learning_rate": 4.925974033066202e-05,
409
+ "loss": 0.773,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.16,
414
+ "learning_rate": 4.923759196045208e-05,
415
+ "loss": 0.7714,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.16,
420
+ "learning_rate": 4.921512224752426e-05,
421
+ "loss": 0.7788,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.16,
426
+ "learning_rate": 4.9192331489782705e-05,
427
+ "loss": 0.7755,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 4.9169219989387956e-05,
433
+ "loss": 0.7682,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.17,
438
+ "learning_rate": 4.9145788052752975e-05,
439
+ "loss": 0.788,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.17,
444
+ "learning_rate": 4.912203599053907e-05,
445
+ "loss": 0.7752,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.17,
450
+ "learning_rate": 4.90979641176518e-05,
451
+ "loss": 0.7786,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.17,
456
+ "learning_rate": 4.907357275323677e-05,
457
+ "loss": 0.7848,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.18,
462
+ "learning_rate": 4.904886222067542e-05,
463
+ "loss": 0.7747,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.18,
468
+ "learning_rate": 4.9023832847580724e-05,
469
+ "loss": 0.7784,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.18,
474
+ "learning_rate": 4.899848496579287e-05,
475
+ "loss": 0.768,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.18,
480
+ "learning_rate": 4.897281891137481e-05,
481
+ "loss": 0.7684,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.19,
486
+ "learning_rate": 4.8946835024607885e-05,
487
+ "loss": 0.7696,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.19,
492
+ "learning_rate": 4.892053364998722e-05,
493
+ "loss": 0.7552,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.19,
498
+ "learning_rate": 4.889391513621725e-05,
499
+ "loss": 0.7584,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.19,
504
+ "learning_rate": 4.886697983620699e-05,
505
+ "loss": 0.7682,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.19,
510
+ "learning_rate": 4.883972810706548e-05,
511
+ "loss": 0.7701,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.2,
516
+ "learning_rate": 4.881216031009696e-05,
517
+ "loss": 0.7637,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.2,
522
+ "learning_rate": 4.878427681079608e-05,
523
+ "loss": 0.7673,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.2,
528
+ "learning_rate": 4.875607797884312e-05,
529
+ "loss": 0.7579,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.2,
534
+ "learning_rate": 4.8727564188099054e-05,
535
+ "loss": 0.7629,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.21,
540
+ "learning_rate": 4.8698735816600535e-05,
541
+ "loss": 0.7766,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.21,
546
+ "learning_rate": 4.866959324655501e-05,
547
+ "loss": 0.7604,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.21,
552
+ "learning_rate": 4.864013686433552e-05,
553
+ "loss": 0.7515,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.21,
558
+ "learning_rate": 4.861036706047566e-05,
559
+ "loss": 0.7508,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.22,
564
+ "learning_rate": 4.858028422966439e-05,
565
+ "loss": 0.7559,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.22,
570
+ "learning_rate": 4.854988877074076e-05,
571
+ "loss": 0.7684,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.22,
576
+ "learning_rate": 4.8519181086688695e-05,
577
+ "loss": 0.7514,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.22,
582
+ "learning_rate": 4.8488161584631574e-05,
583
+ "loss": 0.7543,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.22,
588
+ "learning_rate": 4.845683067582689e-05,
589
+ "loss": 0.7651,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.23,
594
+ "learning_rate": 4.842518877566076e-05,
595
+ "loss": 0.7634,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.23,
600
+ "learning_rate": 4.839323630364243e-05,
601
+ "loss": 0.7477,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.23,
606
+ "learning_rate": 4.836097368339875e-05,
607
+ "loss": 0.7559,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.23,
612
+ "learning_rate": 4.832840134266849e-05,
613
+ "loss": 0.7729,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.24,
618
+ "learning_rate": 4.829551971329672e-05,
619
+ "loss": 0.7518,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.24,
624
+ "learning_rate": 4.826232923122907e-05,
625
+ "loss": 0.7551,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.24,
630
+ "learning_rate": 4.8228830336505926e-05,
631
+ "loss": 0.7613,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.24,
636
+ "learning_rate": 4.819502347325665e-05,
637
+ "loss": 0.7529,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.25,
642
+ "learning_rate": 4.816090908969363e-05,
643
+ "loss": 0.7476,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.25,
648
+ "learning_rate": 4.81264876381064e-05,
649
+ "loss": 0.765,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.25,
654
+ "learning_rate": 4.809175957485558e-05,
655
+ "loss": 0.7417,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.25,
660
+ "learning_rate": 4.805672536036686e-05,
661
+ "loss": 0.7398,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.25,
666
+ "learning_rate": 4.8021385459124925e-05,
667
+ "loss": 0.7489,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.26,
672
+ "learning_rate": 4.798574033966723e-05,
673
+ "loss": 0.7546,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.26,
678
+ "learning_rate": 4.794979047457783e-05,
679
+ "loss": 0.738,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.26,
684
+ "learning_rate": 4.79135363404811e-05,
685
+ "loss": 0.7495,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.26,
690
+ "learning_rate": 4.7876978418035455e-05,
691
+ "loss": 0.7622,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.27,
696
+ "learning_rate": 4.7840117191926915e-05,
697
+ "loss": 0.7465,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.27,
702
+ "learning_rate": 4.780295315086271e-05,
703
+ "loss": 0.7446,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.27,
708
+ "learning_rate": 4.7765486787564814e-05,
709
+ "loss": 0.7425,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.27,
714
+ "learning_rate": 4.772771859876338e-05,
715
+ "loss": 0.7583,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.28,
720
+ "learning_rate": 4.768964908519019e-05,
721
+ "loss": 0.7579,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.28,
726
+ "learning_rate": 4.7651278751571984e-05,
727
+ "loss": 0.7509,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.28,
732
+ "learning_rate": 4.76126081066238e-05,
733
+ "loss": 0.7358,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.28,
738
+ "learning_rate": 4.7573637663042195e-05,
739
+ "loss": 0.7409,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.29,
744
+ "learning_rate": 4.753436793749848e-05,
745
+ "loss": 0.7289,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.29,
750
+ "learning_rate": 4.749479945063186e-05,
751
+ "loss": 0.7488,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.29,
756
+ "learning_rate": 4.745493272704249e-05,
757
+ "loss": 0.7514,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.29,
762
+ "learning_rate": 4.7414768295284594e-05,
763
+ "loss": 0.7384,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.29,
768
+ "learning_rate": 4.7374306687859405e-05,
769
+ "loss": 0.739,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.3,
774
+ "learning_rate": 4.7333548441208134e-05,
775
+ "loss": 0.7494,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.3,
780
+ "learning_rate": 4.729249409570481e-05,
781
+ "loss": 0.7503,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.3,
786
+ "learning_rate": 4.725114419564919e-05,
787
+ "loss": 0.7388,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.3,
792
+ "learning_rate": 4.720949928925948e-05,
793
+ "loss": 0.7601,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.31,
798
+ "learning_rate": 4.716755992866509e-05,
799
+ "loss": 0.7416,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.31,
804
+ "learning_rate": 4.712532666989931e-05,
805
+ "loss": 0.7365,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.31,
810
+ "learning_rate": 4.708280007289195e-05,
811
+ "loss": 0.7455,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.31,
816
+ "learning_rate": 4.703998070146191e-05,
817
+ "loss": 0.7342,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.32,
822
+ "learning_rate": 4.699686912330969e-05,
823
+ "loss": 0.7374,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.32,
828
+ "learning_rate": 4.695346591000986e-05,
829
+ "loss": 0.7312,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.32,
834
+ "learning_rate": 4.6909771637003556e-05,
835
+ "loss": 0.7446,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.32,
840
+ "learning_rate": 4.686578688359072e-05,
841
+ "loss": 0.7387,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.32,
846
+ "learning_rate": 4.6821512232922546e-05,
847
+ "loss": 0.73,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.33,
852
+ "learning_rate": 4.677694827199367e-05,
853
+ "loss": 0.7357,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.33,
858
+ "learning_rate": 4.673209559163442e-05,
859
+ "loss": 0.7441,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.33,
864
+ "learning_rate": 4.668695478650296e-05,
865
+ "loss": 0.7359,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.33,
870
+ "learning_rate": 4.664152645507744e-05,
871
+ "loss": 0.7467,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.34,
876
+ "learning_rate": 4.659581119964803e-05,
877
+ "loss": 0.7221,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.34,
882
+ "learning_rate": 4.654980962630896e-05,
883
+ "loss": 0.7334,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.34,
888
+ "learning_rate": 4.6503522344950454e-05,
889
+ "loss": 0.7405,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.34,
894
+ "learning_rate": 4.6456949969250674e-05,
895
+ "loss": 0.7424,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.35,
900
+ "learning_rate": 4.6410093116667566e-05,
901
+ "loss": 0.7378,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.35,
906
+ "learning_rate": 4.636295240843068e-05,
907
+ "loss": 0.7479,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.35,
912
+ "learning_rate": 4.6315528469532945e-05,
913
+ "loss": 0.7479,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.35,
918
+ "learning_rate": 4.6267821928722345e-05,
919
+ "loss": 0.7337,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.35,
924
+ "learning_rate": 4.621983341849362e-05,
925
+ "loss": 0.7403,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.36,
930
+ "learning_rate": 4.6171563575079876e-05,
931
+ "loss": 0.7353,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.36,
936
+ "learning_rate": 4.612301303844415e-05,
937
+ "loss": 0.7237,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.36,
942
+ "learning_rate": 4.607418245227089e-05,
943
+ "loss": 0.739,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.36,
948
+ "learning_rate": 4.602507246395747e-05,
949
+ "loss": 0.7292,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.37,
954
+ "learning_rate": 4.5975683724605584e-05,
955
+ "loss": 0.745,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.37,
960
+ "learning_rate": 4.5926016889012616e-05,
961
+ "loss": 0.7276,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.37,
966
+ "learning_rate": 4.587607261566294e-05,
967
+ "loss": 0.7481,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.37,
972
+ "learning_rate": 4.582585156671925e-05,
973
+ "loss": 0.7362,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.38,
978
+ "learning_rate": 4.577535440801368e-05,
979
+ "loss": 0.734,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.38,
984
+ "learning_rate": 4.572458180903909e-05,
985
+ "loss": 0.7428,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.38,
990
+ "learning_rate": 4.5673534442940113e-05,
991
+ "loss": 0.732,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.38,
996
+ "learning_rate": 4.5622212986504245e-05,
997
+ "loss": 0.7191,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.38,
1002
+ "learning_rate": 4.557061812015289e-05,
1003
+ "loss": 0.727,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.39,
1008
+ "learning_rate": 4.551875052793232e-05,
1009
+ "loss": 0.728,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.39,
1014
+ "learning_rate": 4.546661089750462e-05,
1015
+ "loss": 0.7366,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.39,
1020
+ "learning_rate": 4.541419992013859e-05,
1021
+ "loss": 0.7274,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.39,
1026
+ "learning_rate": 4.536151829070052e-05,
1027
+ "loss": 0.7185,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.4,
1032
+ "learning_rate": 4.530856670764502e-05,
1033
+ "loss": 0.7463,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.4,
1038
+ "learning_rate": 4.525534587300575e-05,
1039
+ "loss": 0.734,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.4,
1044
+ "learning_rate": 4.520185649238615e-05,
1045
+ "loss": 0.7325,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.4,
1050
+ "learning_rate": 4.5148099274950005e-05,
1051
+ "loss": 0.7429,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.41,
1056
+ "learning_rate": 4.509407493341212e-05,
1057
+ "loss": 0.7333,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.41,
1062
+ "learning_rate": 4.503978418402883e-05,
1063
+ "loss": 0.7227,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.41,
1068
+ "learning_rate": 4.49852277465885e-05,
1069
+ "loss": 0.7139,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.41,
1074
+ "learning_rate": 4.4930406344402006e-05,
1075
+ "loss": 0.7387,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.41,
1080
+ "learning_rate": 4.487532070429315e-05,
1081
+ "loss": 0.7276,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.42,
1086
+ "learning_rate": 4.481997155658898e-05,
1087
+ "loss": 0.7251,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.42,
1092
+ "learning_rate": 4.476435963511015e-05,
1093
+ "loss": 0.7287,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.42,
1098
+ "learning_rate": 4.470848567716119e-05,
1099
+ "loss": 0.7169,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.42,
1104
+ "learning_rate": 4.4652350423520685e-05,
1105
+ "loss": 0.725,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.43,
1110
+ "learning_rate": 4.45959546184315e-05,
1111
+ "loss": 0.7089,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.43,
1116
+ "learning_rate": 4.4539299009590926e-05,
1117
+ "loss": 0.722,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.43,
1122
+ "learning_rate": 4.4482384348140695e-05,
1123
+ "loss": 0.7197,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.43,
1128
+ "learning_rate": 4.44252113886571e-05,
1129
+ "loss": 0.7135,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.44,
1134
+ "learning_rate": 4.4367780889140945e-05,
1135
+ "loss": 0.7168,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.44,
1140
+ "learning_rate": 4.431009361100748e-05,
1141
+ "loss": 0.7267,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.44,
1146
+ "learning_rate": 4.4252150319076416e-05,
1147
+ "loss": 0.7241,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.44,
1152
+ "learning_rate": 4.419395178156159e-05,
1153
+ "loss": 0.7294,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.45,
1158
+ "learning_rate": 4.413549877006099e-05,
1159
+ "loss": 0.7392,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.45,
1164
+ "learning_rate": 4.4076792059546365e-05,
1165
+ "loss": 0.7058,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.45,
1170
+ "learning_rate": 4.401783242835306e-05,
1171
+ "loss": 0.721,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.45,
1176
+ "learning_rate": 4.395862065816959e-05,
1177
+ "loss": 0.7284,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.45,
1182
+ "learning_rate": 4.3899157534027396e-05,
1183
+ "loss": 0.7259,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.46,
1188
+ "learning_rate": 4.383944384429035e-05,
1189
+ "loss": 0.7378,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.46,
1194
+ "learning_rate": 4.3779480380644337e-05,
1195
+ "loss": 0.7118,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.46,
1200
+ "learning_rate": 4.3719267938086736e-05,
1201
+ "loss": 0.7318,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.46,
1206
+ "learning_rate": 4.3658807314915915e-05,
1207
+ "loss": 0.7335,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 0.47,
1212
+ "learning_rate": 4.359809931272062e-05,
1213
+ "loss": 0.7162,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 0.47,
1218
+ "learning_rate": 4.353714473636937e-05,
1219
+ "loss": 0.7279,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 0.47,
1224
+ "learning_rate": 4.3475944393999736e-05,
1225
+ "loss": 0.7344,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 0.47,
1230
+ "learning_rate": 4.34144990970077e-05,
1231
+ "loss": 0.7258,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 0.48,
1236
+ "learning_rate": 4.335280966003684e-05,
1237
+ "loss": 0.7202,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 0.48,
1242
+ "learning_rate": 4.329087690096755e-05,
1243
+ "loss": 0.7225,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 0.48,
1248
+ "learning_rate": 4.3228701640906197e-05,
1249
+ "loss": 0.7073,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 0.48,
1254
+ "learning_rate": 4.3166284704174244e-05,
1255
+ "loss": 0.7265,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 0.48,
1260
+ "learning_rate": 4.310362691829729e-05,
1261
+ "loss": 0.7229,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 0.49,
1266
+ "learning_rate": 4.304072911399415e-05,
1267
+ "loss": 0.6985,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 0.49,
1272
+ "learning_rate": 4.297759212516577e-05,
1273
+ "loss": 0.7316,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 0.49,
1278
+ "learning_rate": 4.291421678888425e-05,
1279
+ "loss": 0.7332,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 0.49,
1284
+ "learning_rate": 4.285060394538167e-05,
1285
+ "loss": 0.7162,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 0.5,
1290
+ "learning_rate": 4.278675443803903e-05,
1291
+ "loss": 0.7064,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 0.5,
1296
+ "learning_rate": 4.2722669113375e-05,
1297
+ "loss": 0.7234,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 0.5,
1302
+ "learning_rate": 4.2658348821034724e-05,
1303
+ "loss": 0.7086,
1304
+ "step": 2160
1305
+ },
1306
+ {
1307
+ "epoch": 0.5,
1308
+ "learning_rate": 4.2593794413778545e-05,
1309
+ "loss": 0.7277,
1310
+ "step": 2170
1311
+ },
1312
+ {
1313
+ "epoch": 0.51,
1314
+ "learning_rate": 4.2529006747470745e-05,
1315
+ "loss": 0.6973,
1316
+ "step": 2180
1317
+ },
1318
+ {
1319
+ "epoch": 0.51,
1320
+ "learning_rate": 4.2463986681068124e-05,
1321
+ "loss": 0.7221,
1322
+ "step": 2190
1323
+ },
1324
+ {
1325
+ "epoch": 0.51,
1326
+ "learning_rate": 4.2398735076608675e-05,
1327
+ "loss": 0.7184,
1328
+ "step": 2200
1329
+ },
1330
+ {
1331
+ "epoch": 0.51,
1332
+ "learning_rate": 4.233325279920011e-05,
1333
+ "loss": 0.7132,
1334
+ "step": 2210
1335
+ },
1336
+ {
1337
+ "epoch": 0.51,
1338
+ "learning_rate": 4.226754071700841e-05,
1339
+ "loss": 0.7208,
1340
+ "step": 2220
1341
+ },
1342
+ {
1343
+ "epoch": 0.52,
1344
+ "learning_rate": 4.220159970124633e-05,
1345
+ "loss": 0.7078,
1346
+ "step": 2230
1347
+ },
1348
+ {
1349
+ "epoch": 0.52,
1350
+ "learning_rate": 4.2135430626161834e-05,
1351
+ "loss": 0.7196,
1352
+ "step": 2240
1353
+ },
1354
+ {
1355
+ "epoch": 0.52,
1356
+ "learning_rate": 4.206903436902649e-05,
1357
+ "loss": 0.7239,
1358
+ "step": 2250
1359
+ },
1360
+ {
1361
+ "epoch": 0.52,
1362
+ "learning_rate": 4.200241181012385e-05,
1363
+ "loss": 0.7172,
1364
+ "step": 2260
1365
+ },
1366
+ {
1367
+ "epoch": 0.53,
1368
+ "learning_rate": 4.19355638327378e-05,
1369
+ "loss": 0.7256,
1370
+ "step": 2270
1371
+ },
1372
+ {
1373
+ "epoch": 0.53,
1374
+ "learning_rate": 4.1868491323140816e-05,
1375
+ "loss": 0.7117,
1376
+ "step": 2280
1377
+ },
1378
+ {
1379
+ "epoch": 0.53,
1380
+ "learning_rate": 4.180119517058222e-05,
1381
+ "loss": 0.7199,
1382
+ "step": 2290
1383
+ },
1384
+ {
1385
+ "epoch": 0.53,
1386
+ "learning_rate": 4.173367626727643e-05,
1387
+ "loss": 0.7023,
1388
+ "step": 2300
1389
+ },
1390
+ {
1391
+ "epoch": 0.54,
1392
+ "learning_rate": 4.166593550839108e-05,
1393
+ "loss": 0.7185,
1394
+ "step": 2310
1395
+ },
1396
+ {
1397
+ "epoch": 0.54,
1398
+ "learning_rate": 4.159797379203516e-05,
1399
+ "loss": 0.7167,
1400
+ "step": 2320
1401
+ },
1402
+ {
1403
+ "epoch": 0.54,
1404
+ "learning_rate": 4.152979201924715e-05,
1405
+ "loss": 0.7194,
1406
+ "step": 2330
1407
+ },
1408
+ {
1409
+ "epoch": 0.54,
1410
+ "learning_rate": 4.146139109398303e-05,
1411
+ "loss": 0.717,
1412
+ "step": 2340
1413
+ },
1414
+ {
1415
+ "epoch": 0.54,
1416
+ "learning_rate": 4.139277192310429e-05,
1417
+ "loss": 0.724,
1418
+ "step": 2350
1419
+ },
1420
+ {
1421
+ "epoch": 0.55,
1422
+ "learning_rate": 4.1323935416365965e-05,
1423
+ "loss": 0.6933,
1424
+ "step": 2360
1425
+ },
1426
+ {
1427
+ "epoch": 0.55,
1428
+ "learning_rate": 4.125488248640451e-05,
1429
+ "loss": 0.7074,
1430
+ "step": 2370
1431
+ },
1432
+ {
1433
+ "epoch": 0.55,
1434
+ "learning_rate": 4.118561404872574e-05,
1435
+ "loss": 0.7028,
1436
+ "step": 2380
1437
+ },
1438
+ {
1439
+ "epoch": 0.55,
1440
+ "learning_rate": 4.111613102169265e-05,
1441
+ "loss": 0.7182,
1442
+ "step": 2390
1443
+ },
1444
+ {
1445
+ "epoch": 0.56,
1446
+ "learning_rate": 4.1046434326513305e-05,
1447
+ "loss": 0.7224,
1448
+ "step": 2400
1449
+ },
1450
+ {
1451
+ "epoch": 0.56,
1452
+ "learning_rate": 4.097652488722856e-05,
1453
+ "loss": 0.7065,
1454
+ "step": 2410
1455
+ },
1456
+ {
1457
+ "epoch": 0.56,
1458
+ "learning_rate": 4.090640363069986e-05,
1459
+ "loss": 0.7074,
1460
+ "step": 2420
1461
+ },
1462
+ {
1463
+ "epoch": 0.56,
1464
+ "learning_rate": 4.08360714865969e-05,
1465
+ "loss": 0.7246,
1466
+ "step": 2430
1467
+ },
1468
+ {
1469
+ "epoch": 0.57,
1470
+ "learning_rate": 4.076552938738537e-05,
1471
+ "loss": 0.7071,
1472
+ "step": 2440
1473
+ },
1474
+ {
1475
+ "epoch": 0.57,
1476
+ "learning_rate": 4.069477826831451e-05,
1477
+ "loss": 0.7082,
1478
+ "step": 2450
1479
+ },
1480
+ {
1481
+ "epoch": 0.57,
1482
+ "learning_rate": 4.062381906740477e-05,
1483
+ "loss": 0.7042,
1484
+ "step": 2460
1485
+ },
1486
+ {
1487
+ "epoch": 0.57,
1488
+ "learning_rate": 4.055265272543536e-05,
1489
+ "loss": 0.7011,
1490
+ "step": 2470
1491
+ },
1492
+ {
1493
+ "epoch": 0.57,
1494
+ "learning_rate": 4.048128018593177e-05,
1495
+ "loss": 0.7155,
1496
+ "step": 2480
1497
+ },
1498
+ {
1499
+ "epoch": 0.58,
1500
+ "learning_rate": 4.040970239515325e-05,
1501
+ "loss": 0.7149,
1502
+ "step": 2490
1503
+ },
1504
+ {
1505
+ "epoch": 0.58,
1506
+ "learning_rate": 4.033792030208029e-05,
1507
+ "loss": 0.717,
1508
+ "step": 2500
1509
+ },
1510
+ {
1511
+ "epoch": 0.58,
1512
+ "learning_rate": 4.026593485840203e-05,
1513
+ "loss": 0.7237,
1514
+ "step": 2510
1515
+ },
1516
+ {
1517
+ "epoch": 0.58,
1518
+ "learning_rate": 4.019374701850362e-05,
1519
+ "loss": 0.7251,
1520
+ "step": 2520
1521
+ },
1522
+ {
1523
+ "epoch": 0.59,
1524
+ "learning_rate": 4.0121357739453604e-05,
1525
+ "loss": 0.7179,
1526
+ "step": 2530
1527
+ },
1528
+ {
1529
+ "epoch": 0.59,
1530
+ "learning_rate": 4.0048767980991206e-05,
1531
+ "loss": 0.7171,
1532
+ "step": 2540
1533
+ },
1534
+ {
1535
+ "epoch": 0.59,
1536
+ "learning_rate": 3.997597870551361e-05,
1537
+ "loss": 0.7124,
1538
+ "step": 2550
1539
+ },
1540
+ {
1541
+ "epoch": 0.59,
1542
+ "learning_rate": 3.990299087806319e-05,
1543
+ "loss": 0.7206,
1544
+ "step": 2560
1545
+ },
1546
+ {
1547
+ "epoch": 0.6,
1548
+ "learning_rate": 3.9829805466314775e-05,
1549
+ "loss": 0.72,
1550
+ "step": 2570
1551
+ },
1552
+ {
1553
+ "epoch": 0.6,
1554
+ "learning_rate": 3.975642344056271e-05,
1555
+ "loss": 0.6972,
1556
+ "step": 2580
1557
+ },
1558
+ {
1559
+ "epoch": 0.6,
1560
+ "learning_rate": 3.9682845773708085e-05,
1561
+ "loss": 0.7158,
1562
+ "step": 2590
1563
+ },
1564
+ {
1565
+ "epoch": 0.6,
1566
+ "learning_rate": 3.960907344124581e-05,
1567
+ "loss": 0.7083,
1568
+ "step": 2600
1569
+ },
1570
+ {
1571
+ "epoch": 0.61,
1572
+ "learning_rate": 3.9535107421251676e-05,
1573
+ "loss": 0.7069,
1574
+ "step": 2610
1575
+ },
1576
+ {
1577
+ "epoch": 0.61,
1578
+ "learning_rate": 3.946094869436938e-05,
1579
+ "loss": 0.7018,
1580
+ "step": 2620
1581
+ },
1582
+ {
1583
+ "epoch": 0.61,
1584
+ "learning_rate": 3.938659824379754e-05,
1585
+ "loss": 0.6973,
1586
+ "step": 2630
1587
+ },
1588
+ {
1589
+ "epoch": 0.61,
1590
+ "learning_rate": 3.931205705527664e-05,
1591
+ "loss": 0.7064,
1592
+ "step": 2640
1593
+ },
1594
+ {
1595
+ "epoch": 0.61,
1596
+ "learning_rate": 3.9237326117075995e-05,
1597
+ "loss": 0.702,
1598
+ "step": 2650
1599
+ },
1600
+ {
1601
+ "epoch": 0.62,
1602
+ "learning_rate": 3.9162406419980605e-05,
1603
+ "loss": 0.713,
1604
+ "step": 2660
1605
+ },
1606
+ {
1607
+ "epoch": 0.62,
1608
+ "learning_rate": 3.908729895727806e-05,
1609
+ "loss": 0.7054,
1610
+ "step": 2670
1611
+ },
1612
+ {
1613
+ "epoch": 0.62,
1614
+ "learning_rate": 3.9012004724745347e-05,
1615
+ "loss": 0.7056,
1616
+ "step": 2680
1617
+ },
1618
+ {
1619
+ "epoch": 0.62,
1620
+ "learning_rate": 3.893652472063562e-05,
1621
+ "loss": 0.7047,
1622
+ "step": 2690
1623
+ },
1624
+ {
1625
+ "epoch": 0.63,
1626
+ "learning_rate": 3.886085994566506e-05,
1627
+ "loss": 0.705,
1628
+ "step": 2700
1629
+ },
1630
+ {
1631
+ "epoch": 0.63,
1632
+ "learning_rate": 3.87850114029995e-05,
1633
+ "loss": 0.6996,
1634
+ "step": 2710
1635
+ },
1636
+ {
1637
+ "epoch": 0.63,
1638
+ "learning_rate": 3.870898009824118e-05,
1639
+ "loss": 0.6986,
1640
+ "step": 2720
1641
+ },
1642
+ {
1643
+ "epoch": 0.63,
1644
+ "learning_rate": 3.863276703941542e-05,
1645
+ "loss": 0.7054,
1646
+ "step": 2730
1647
+ },
1648
+ {
1649
+ "epoch": 0.64,
1650
+ "learning_rate": 3.855637323695721e-05,
1651
+ "loss": 0.7094,
1652
+ "step": 2740
1653
+ },
1654
+ {
1655
+ "epoch": 0.64,
1656
+ "learning_rate": 3.8479799703697884e-05,
1657
+ "loss": 0.7103,
1658
+ "step": 2750
1659
+ },
1660
+ {
1661
+ "epoch": 0.64,
1662
+ "learning_rate": 3.840304745485162e-05,
1663
+ "loss": 0.6921,
1664
+ "step": 2760
1665
+ },
1666
+ {
1667
+ "epoch": 0.64,
1668
+ "learning_rate": 3.8326117508002025e-05,
1669
+ "loss": 0.7177,
1670
+ "step": 2770
1671
+ },
1672
+ {
1673
+ "epoch": 0.64,
1674
+ "learning_rate": 3.8249010883088646e-05,
1675
+ "loss": 0.7163,
1676
+ "step": 2780
1677
+ },
1678
+ {
1679
+ "epoch": 0.65,
1680
+ "learning_rate": 3.817172860239339e-05,
1681
+ "loss": 0.7135,
1682
+ "step": 2790
1683
+ },
1684
+ {
1685
+ "epoch": 0.65,
1686
+ "learning_rate": 3.8094271690527076e-05,
1687
+ "loss": 0.7141,
1688
+ "step": 2800
1689
+ },
1690
+ {
1691
+ "epoch": 0.65,
1692
+ "learning_rate": 3.801664117441572e-05,
1693
+ "loss": 0.704,
1694
+ "step": 2810
1695
+ },
1696
+ {
1697
+ "epoch": 0.65,
1698
+ "learning_rate": 3.7938838083287056e-05,
1699
+ "loss": 0.7092,
1700
+ "step": 2820
1701
+ },
1702
+ {
1703
+ "epoch": 0.66,
1704
+ "learning_rate": 3.7860863448656765e-05,
1705
+ "loss": 0.6986,
1706
+ "step": 2830
1707
+ },
1708
+ {
1709
+ "epoch": 0.66,
1710
+ "learning_rate": 3.778271830431488e-05,
1711
+ "loss": 0.7153,
1712
+ "step": 2840
1713
+ },
1714
+ {
1715
+ "epoch": 0.66,
1716
+ "learning_rate": 3.770440368631207e-05,
1717
+ "loss": 0.7048,
1718
+ "step": 2850
1719
+ },
1720
+ {
1721
+ "epoch": 0.66,
1722
+ "learning_rate": 3.762592063294588e-05,
1723
+ "loss": 0.7103,
1724
+ "step": 2860
1725
+ },
1726
+ {
1727
+ "epoch": 0.67,
1728
+ "learning_rate": 3.7547270184746976e-05,
1729
+ "loss": 0.7048,
1730
+ "step": 2870
1731
+ },
1732
+ {
1733
+ "epoch": 0.67,
1734
+ "learning_rate": 3.746845338446533e-05,
1735
+ "loss": 0.7083,
1736
+ "step": 2880
1737
+ },
1738
+ {
1739
+ "epoch": 0.67,
1740
+ "learning_rate": 3.738947127705644e-05,
1741
+ "loss": 0.7011,
1742
+ "step": 2890
1743
+ },
1744
+ {
1745
+ "epoch": 0.67,
1746
+ "learning_rate": 3.731032490966746e-05,
1747
+ "loss": 0.6959,
1748
+ "step": 2900
1749
+ },
1750
+ {
1751
+ "epoch": 0.67,
1752
+ "learning_rate": 3.723101533162329e-05,
1753
+ "loss": 0.7043,
1754
+ "step": 2910
1755
+ },
1756
+ {
1757
+ "epoch": 0.68,
1758
+ "learning_rate": 3.715154359441265e-05,
1759
+ "loss": 0.7101,
1760
+ "step": 2920
1761
+ },
1762
+ {
1763
+ "epoch": 0.68,
1764
+ "learning_rate": 3.7071910751674256e-05,
1765
+ "loss": 0.7071,
1766
+ "step": 2930
1767
+ },
1768
+ {
1769
+ "epoch": 0.68,
1770
+ "learning_rate": 3.6992117859182674e-05,
1771
+ "loss": 0.7158,
1772
+ "step": 2940
1773
+ },
1774
+ {
1775
+ "epoch": 0.68,
1776
+ "learning_rate": 3.691216597483447e-05,
1777
+ "loss": 0.7046,
1778
+ "step": 2950
1779
+ },
1780
+ {
1781
+ "epoch": 0.69,
1782
+ "learning_rate": 3.683205615863412e-05,
1783
+ "loss": 0.7003,
1784
+ "step": 2960
1785
+ },
1786
+ {
1787
+ "epoch": 0.69,
1788
+ "learning_rate": 3.6751789472679946e-05,
1789
+ "loss": 0.7067,
1790
+ "step": 2970
1791
+ },
1792
+ {
1793
+ "epoch": 0.69,
1794
+ "learning_rate": 3.667136698115008e-05,
1795
+ "loss": 0.7002,
1796
+ "step": 2980
1797
+ },
1798
+ {
1799
+ "epoch": 0.69,
1800
+ "learning_rate": 3.65907897502883e-05,
1801
+ "loss": 0.707,
1802
+ "step": 2990
1803
+ },
1804
+ {
1805
+ "epoch": 0.7,
1806
+ "learning_rate": 3.651005884838995e-05,
1807
+ "loss": 0.7069,
1808
+ "step": 3000
1809
+ }
1810
+ ],
1811
+ "max_steps": 8628,
1812
+ "num_train_epochs": 2,
1813
+ "total_flos": 9.47445766429868e+17,
1814
+ "trial_name": null,
1815
+ "trial_params": null
1816
+ }
checkpoint-3000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ce444f5cfe8a7851710b073da2f9ede89d9fad5a6d091ebf47aee5e6b0d0797
3
+ size 3259
checkpoint-4000/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-4000/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "./Llama-2-7b-chat-hf",
4
+ "bias": "none",
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": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-4000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69e2ba2e4caefd750288302544496e94b08c73a9bd2c4108c4791a48b06440fd
3
+ size 16821197
checkpoint-4000/finetuning_args.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dpo_beta": 0.1,
3
+ "finetuning_type": "lora",
4
+ "lora_alpha": 32.0,
5
+ "lora_dropout": 0.1,
6
+ "lora_rank": 8,
7
+ "lora_target": [
8
+ "q_proj",
9
+ "v_proj"
10
+ ],
11
+ "name_module_trainable": "mlp",
12
+ "num_hidden_layers": 32,
13
+ "num_layer_trainable": 3,
14
+ "ppo_score_norm": false,
15
+ "resume_lora_training": true
16
+ }
checkpoint-4000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9e6211a8b89462696a9a3484d74bd36cbd0936c16884e9c626a859b43ec4d54
3
+ size 33661637
checkpoint-4000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dfb2de8832d37b94e07204ea24a1338ba8b7f5fa0a83cdb814759bf3f783150a
3
+ size 18663
checkpoint-4000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b96c949af3c1e9df7efe1c68b03944122f4f571399fb8d9ce5b5787932c8339
3
+ size 627
checkpoint-4000/trainer_state.json ADDED
@@ -0,0 +1,2416 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9272137227630969,
5
+ "global_step": 4000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.999983427462679e-05,
13
+ "loss": 1.3476,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 4.999933710070436e-05,
19
+ "loss": 1.1454,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 4.999850848482424e-05,
25
+ "loss": 1.0263,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 4.999734843797226e-05,
31
+ "loss": 1.0283,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 4.9995856975528347e-05,
37
+ "loss": 0.9733,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 4.999403411726636e-05,
43
+ "loss": 0.9591,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 4.999187988735381e-05,
49
+ "loss": 0.9617,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 4.998939431435153e-05,
55
+ "loss": 0.9414,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 4.998657743121333e-05,
61
+ "loss": 0.9233,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 4.998342927528553e-05,
67
+ "loss": 0.9236,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 4.997994988830648e-05,
73
+ "loss": 0.8743,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 4.997613931640598e-05,
79
+ "loss": 0.895,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 4.997199761010472e-05,
85
+ "loss": 0.8831,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 4.996752482431356e-05,
91
+ "loss": 0.8799,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 4.996272101833284e-05,
97
+ "loss": 0.877,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 4.995758625585155e-05,
103
+ "loss": 0.8745,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 4.995212060494653e-05,
109
+ "loss": 0.8716,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 4.9946324138081534e-05,
115
+ "loss": 0.8534,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 4.99401969321063e-05,
121
+ "loss": 0.8449,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.05,
126
+ "learning_rate": 4.9933739068255513e-05,
127
+ "loss": 0.8588,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 4.992695063214772e-05,
133
+ "loss": 0.8409,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 4.99198317137842e-05,
139
+ "loss": 0.8517,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 4.99123824075478e-05,
145
+ "loss": 0.8698,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.06,
150
+ "learning_rate": 4.990460281220164e-05,
151
+ "loss": 0.848,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.06,
156
+ "learning_rate": 4.9896493030887815e-05,
157
+ "loss": 0.8322,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 4.988805317112607e-05,
163
+ "loss": 0.8376,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 4.98792833448123e-05,
169
+ "loss": 0.8309,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 4.987018366821713e-05,
175
+ "loss": 0.8195,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.07,
180
+ "learning_rate": 4.9860754261984346e-05,
181
+ "loss": 0.8313,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 4.98509952511293e-05,
187
+ "loss": 0.8115,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 4.984090676503723e-05,
193
+ "loss": 0.8224,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 4.983048893746162e-05,
199
+ "loss": 0.8157,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.08,
204
+ "learning_rate": 4.9819741906522306e-05,
205
+ "loss": 0.8165,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.08,
210
+ "learning_rate": 4.980866581470376e-05,
211
+ "loss": 0.8144,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 4.979726080885314e-05,
217
+ "loss": 0.8147,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 4.978552704017835e-05,
223
+ "loss": 0.7894,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.09,
228
+ "learning_rate": 4.977346466424605e-05,
229
+ "loss": 0.8144,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.09,
234
+ "learning_rate": 4.976107384097957e-05,
235
+ "loss": 0.7984,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 4.974835473465682e-05,
241
+ "loss": 0.8188,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 4.97353075139081e-05,
247
+ "loss": 0.8049,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.1,
252
+ "learning_rate": 4.9721932351713846e-05,
253
+ "loss": 0.8154,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.1,
258
+ "learning_rate": 4.9708229425402354e-05,
259
+ "loss": 0.7953,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.1,
264
+ "learning_rate": 4.9694198916647425e-05,
265
+ "loss": 0.7996,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 4.967984101146598e-05,
271
+ "loss": 0.7979,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 4.966515590021554e-05,
277
+ "loss": 0.8106,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.11,
282
+ "learning_rate": 4.965014377759176e-05,
283
+ "loss": 0.8008,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.11,
288
+ "learning_rate": 4.9634804842625807e-05,
289
+ "loss": 0.8045,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 4.961913929868172e-05,
295
+ "loss": 0.786,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 4.960314735345377e-05,
301
+ "loss": 0.7897,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.12,
306
+ "learning_rate": 4.958682921896365e-05,
307
+ "loss": 0.7855,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.12,
312
+ "learning_rate": 4.957018511155765e-05,
313
+ "loss": 0.7657,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.12,
318
+ "learning_rate": 4.955321525190386e-05,
319
+ "loss": 0.779,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 4.953591986498918e-05,
325
+ "loss": 0.7923,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.13,
330
+ "learning_rate": 4.9518299180116366e-05,
331
+ "loss": 0.7732,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.13,
336
+ "learning_rate": 4.9500353430900984e-05,
337
+ "loss": 0.7898,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.13,
342
+ "learning_rate": 4.948208285526832e-05,
343
+ "loss": 0.7796,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.13,
348
+ "learning_rate": 4.94634876954502e-05,
349
+ "loss": 0.7819,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 4.944456819798182e-05,
355
+ "loss": 0.7777,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.14,
360
+ "learning_rate": 4.942532461369843e-05,
361
+ "loss": 0.7815,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.14,
366
+ "learning_rate": 4.940575719773205e-05,
367
+ "loss": 0.7807,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.14,
372
+ "learning_rate": 4.938586620950807e-05,
373
+ "loss": 0.7836,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 4.9365651912741806e-05,
379
+ "loss": 0.7795,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.15,
384
+ "learning_rate": 4.934511457543499e-05,
385
+ "loss": 0.7844,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.15,
390
+ "learning_rate": 4.932425446987228e-05,
391
+ "loss": 0.7865,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.15,
396
+ "learning_rate": 4.930307187261757e-05,
397
+ "loss": 0.7751,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.15,
402
+ "learning_rate": 4.928156706451035e-05,
403
+ "loss": 0.7705,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.16,
408
+ "learning_rate": 4.925974033066202e-05,
409
+ "loss": 0.773,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.16,
414
+ "learning_rate": 4.923759196045208e-05,
415
+ "loss": 0.7714,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.16,
420
+ "learning_rate": 4.921512224752426e-05,
421
+ "loss": 0.7788,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.16,
426
+ "learning_rate": 4.9192331489782705e-05,
427
+ "loss": 0.7755,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 4.9169219989387956e-05,
433
+ "loss": 0.7682,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.17,
438
+ "learning_rate": 4.9145788052752975e-05,
439
+ "loss": 0.788,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.17,
444
+ "learning_rate": 4.912203599053907e-05,
445
+ "loss": 0.7752,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.17,
450
+ "learning_rate": 4.90979641176518e-05,
451
+ "loss": 0.7786,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.17,
456
+ "learning_rate": 4.907357275323677e-05,
457
+ "loss": 0.7848,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.18,
462
+ "learning_rate": 4.904886222067542e-05,
463
+ "loss": 0.7747,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.18,
468
+ "learning_rate": 4.9023832847580724e-05,
469
+ "loss": 0.7784,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.18,
474
+ "learning_rate": 4.899848496579287e-05,
475
+ "loss": 0.768,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.18,
480
+ "learning_rate": 4.897281891137481e-05,
481
+ "loss": 0.7684,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.19,
486
+ "learning_rate": 4.8946835024607885e-05,
487
+ "loss": 0.7696,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.19,
492
+ "learning_rate": 4.892053364998722e-05,
493
+ "loss": 0.7552,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.19,
498
+ "learning_rate": 4.889391513621725e-05,
499
+ "loss": 0.7584,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.19,
504
+ "learning_rate": 4.886697983620699e-05,
505
+ "loss": 0.7682,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.19,
510
+ "learning_rate": 4.883972810706548e-05,
511
+ "loss": 0.7701,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.2,
516
+ "learning_rate": 4.881216031009696e-05,
517
+ "loss": 0.7637,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.2,
522
+ "learning_rate": 4.878427681079608e-05,
523
+ "loss": 0.7673,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.2,
528
+ "learning_rate": 4.875607797884312e-05,
529
+ "loss": 0.7579,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.2,
534
+ "learning_rate": 4.8727564188099054e-05,
535
+ "loss": 0.7629,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.21,
540
+ "learning_rate": 4.8698735816600535e-05,
541
+ "loss": 0.7766,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.21,
546
+ "learning_rate": 4.866959324655501e-05,
547
+ "loss": 0.7604,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.21,
552
+ "learning_rate": 4.864013686433552e-05,
553
+ "loss": 0.7515,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.21,
558
+ "learning_rate": 4.861036706047566e-05,
559
+ "loss": 0.7508,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.22,
564
+ "learning_rate": 4.858028422966439e-05,
565
+ "loss": 0.7559,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.22,
570
+ "learning_rate": 4.854988877074076e-05,
571
+ "loss": 0.7684,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.22,
576
+ "learning_rate": 4.8519181086688695e-05,
577
+ "loss": 0.7514,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.22,
582
+ "learning_rate": 4.8488161584631574e-05,
583
+ "loss": 0.7543,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.22,
588
+ "learning_rate": 4.845683067582689e-05,
589
+ "loss": 0.7651,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.23,
594
+ "learning_rate": 4.842518877566076e-05,
595
+ "loss": 0.7634,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.23,
600
+ "learning_rate": 4.839323630364243e-05,
601
+ "loss": 0.7477,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.23,
606
+ "learning_rate": 4.836097368339875e-05,
607
+ "loss": 0.7559,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.23,
612
+ "learning_rate": 4.832840134266849e-05,
613
+ "loss": 0.7729,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.24,
618
+ "learning_rate": 4.829551971329672e-05,
619
+ "loss": 0.7518,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.24,
624
+ "learning_rate": 4.826232923122907e-05,
625
+ "loss": 0.7551,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.24,
630
+ "learning_rate": 4.8228830336505926e-05,
631
+ "loss": 0.7613,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.24,
636
+ "learning_rate": 4.819502347325665e-05,
637
+ "loss": 0.7529,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.25,
642
+ "learning_rate": 4.816090908969363e-05,
643
+ "loss": 0.7476,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.25,
648
+ "learning_rate": 4.81264876381064e-05,
649
+ "loss": 0.765,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.25,
654
+ "learning_rate": 4.809175957485558e-05,
655
+ "loss": 0.7417,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.25,
660
+ "learning_rate": 4.805672536036686e-05,
661
+ "loss": 0.7398,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.25,
666
+ "learning_rate": 4.8021385459124925e-05,
667
+ "loss": 0.7489,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.26,
672
+ "learning_rate": 4.798574033966723e-05,
673
+ "loss": 0.7546,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.26,
678
+ "learning_rate": 4.794979047457783e-05,
679
+ "loss": 0.738,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.26,
684
+ "learning_rate": 4.79135363404811e-05,
685
+ "loss": 0.7495,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.26,
690
+ "learning_rate": 4.7876978418035455e-05,
691
+ "loss": 0.7622,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.27,
696
+ "learning_rate": 4.7840117191926915e-05,
697
+ "loss": 0.7465,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.27,
702
+ "learning_rate": 4.780295315086271e-05,
703
+ "loss": 0.7446,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.27,
708
+ "learning_rate": 4.7765486787564814e-05,
709
+ "loss": 0.7425,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.27,
714
+ "learning_rate": 4.772771859876338e-05,
715
+ "loss": 0.7583,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.28,
720
+ "learning_rate": 4.768964908519019e-05,
721
+ "loss": 0.7579,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.28,
726
+ "learning_rate": 4.7651278751571984e-05,
727
+ "loss": 0.7509,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.28,
732
+ "learning_rate": 4.76126081066238e-05,
733
+ "loss": 0.7358,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.28,
738
+ "learning_rate": 4.7573637663042195e-05,
739
+ "loss": 0.7409,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.29,
744
+ "learning_rate": 4.753436793749848e-05,
745
+ "loss": 0.7289,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.29,
750
+ "learning_rate": 4.749479945063186e-05,
751
+ "loss": 0.7488,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.29,
756
+ "learning_rate": 4.745493272704249e-05,
757
+ "loss": 0.7514,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.29,
762
+ "learning_rate": 4.7414768295284594e-05,
763
+ "loss": 0.7384,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.29,
768
+ "learning_rate": 4.7374306687859405e-05,
769
+ "loss": 0.739,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.3,
774
+ "learning_rate": 4.7333548441208134e-05,
775
+ "loss": 0.7494,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.3,
780
+ "learning_rate": 4.729249409570481e-05,
781
+ "loss": 0.7503,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.3,
786
+ "learning_rate": 4.725114419564919e-05,
787
+ "loss": 0.7388,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.3,
792
+ "learning_rate": 4.720949928925948e-05,
793
+ "loss": 0.7601,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.31,
798
+ "learning_rate": 4.716755992866509e-05,
799
+ "loss": 0.7416,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.31,
804
+ "learning_rate": 4.712532666989931e-05,
805
+ "loss": 0.7365,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.31,
810
+ "learning_rate": 4.708280007289195e-05,
811
+ "loss": 0.7455,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.31,
816
+ "learning_rate": 4.703998070146191e-05,
817
+ "loss": 0.7342,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.32,
822
+ "learning_rate": 4.699686912330969e-05,
823
+ "loss": 0.7374,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.32,
828
+ "learning_rate": 4.695346591000986e-05,
829
+ "loss": 0.7312,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.32,
834
+ "learning_rate": 4.6909771637003556e-05,
835
+ "loss": 0.7446,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.32,
840
+ "learning_rate": 4.686578688359072e-05,
841
+ "loss": 0.7387,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.32,
846
+ "learning_rate": 4.6821512232922546e-05,
847
+ "loss": 0.73,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.33,
852
+ "learning_rate": 4.677694827199367e-05,
853
+ "loss": 0.7357,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.33,
858
+ "learning_rate": 4.673209559163442e-05,
859
+ "loss": 0.7441,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.33,
864
+ "learning_rate": 4.668695478650296e-05,
865
+ "loss": 0.7359,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.33,
870
+ "learning_rate": 4.664152645507744e-05,
871
+ "loss": 0.7467,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.34,
876
+ "learning_rate": 4.659581119964803e-05,
877
+ "loss": 0.7221,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.34,
882
+ "learning_rate": 4.654980962630896e-05,
883
+ "loss": 0.7334,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.34,
888
+ "learning_rate": 4.6503522344950454e-05,
889
+ "loss": 0.7405,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.34,
894
+ "learning_rate": 4.6456949969250674e-05,
895
+ "loss": 0.7424,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.35,
900
+ "learning_rate": 4.6410093116667566e-05,
901
+ "loss": 0.7378,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.35,
906
+ "learning_rate": 4.636295240843068e-05,
907
+ "loss": 0.7479,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.35,
912
+ "learning_rate": 4.6315528469532945e-05,
913
+ "loss": 0.7479,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.35,
918
+ "learning_rate": 4.6267821928722345e-05,
919
+ "loss": 0.7337,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.35,
924
+ "learning_rate": 4.621983341849362e-05,
925
+ "loss": 0.7403,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.36,
930
+ "learning_rate": 4.6171563575079876e-05,
931
+ "loss": 0.7353,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.36,
936
+ "learning_rate": 4.612301303844415e-05,
937
+ "loss": 0.7237,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.36,
942
+ "learning_rate": 4.607418245227089e-05,
943
+ "loss": 0.739,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.36,
948
+ "learning_rate": 4.602507246395747e-05,
949
+ "loss": 0.7292,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.37,
954
+ "learning_rate": 4.5975683724605584e-05,
955
+ "loss": 0.745,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.37,
960
+ "learning_rate": 4.5926016889012616e-05,
961
+ "loss": 0.7276,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.37,
966
+ "learning_rate": 4.587607261566294e-05,
967
+ "loss": 0.7481,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.37,
972
+ "learning_rate": 4.582585156671925e-05,
973
+ "loss": 0.7362,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.38,
978
+ "learning_rate": 4.577535440801368e-05,
979
+ "loss": 0.734,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.38,
984
+ "learning_rate": 4.572458180903909e-05,
985
+ "loss": 0.7428,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.38,
990
+ "learning_rate": 4.5673534442940113e-05,
991
+ "loss": 0.732,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.38,
996
+ "learning_rate": 4.5622212986504245e-05,
997
+ "loss": 0.7191,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.38,
1002
+ "learning_rate": 4.557061812015289e-05,
1003
+ "loss": 0.727,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.39,
1008
+ "learning_rate": 4.551875052793232e-05,
1009
+ "loss": 0.728,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.39,
1014
+ "learning_rate": 4.546661089750462e-05,
1015
+ "loss": 0.7366,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.39,
1020
+ "learning_rate": 4.541419992013859e-05,
1021
+ "loss": 0.7274,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.39,
1026
+ "learning_rate": 4.536151829070052e-05,
1027
+ "loss": 0.7185,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.4,
1032
+ "learning_rate": 4.530856670764502e-05,
1033
+ "loss": 0.7463,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.4,
1038
+ "learning_rate": 4.525534587300575e-05,
1039
+ "loss": 0.734,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.4,
1044
+ "learning_rate": 4.520185649238615e-05,
1045
+ "loss": 0.7325,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.4,
1050
+ "learning_rate": 4.5148099274950005e-05,
1051
+ "loss": 0.7429,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.41,
1056
+ "learning_rate": 4.509407493341212e-05,
1057
+ "loss": 0.7333,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.41,
1062
+ "learning_rate": 4.503978418402883e-05,
1063
+ "loss": 0.7227,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.41,
1068
+ "learning_rate": 4.49852277465885e-05,
1069
+ "loss": 0.7139,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.41,
1074
+ "learning_rate": 4.4930406344402006e-05,
1075
+ "loss": 0.7387,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.41,
1080
+ "learning_rate": 4.487532070429315e-05,
1081
+ "loss": 0.7276,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.42,
1086
+ "learning_rate": 4.481997155658898e-05,
1087
+ "loss": 0.7251,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.42,
1092
+ "learning_rate": 4.476435963511015e-05,
1093
+ "loss": 0.7287,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.42,
1098
+ "learning_rate": 4.470848567716119e-05,
1099
+ "loss": 0.7169,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.42,
1104
+ "learning_rate": 4.4652350423520685e-05,
1105
+ "loss": 0.725,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.43,
1110
+ "learning_rate": 4.45959546184315e-05,
1111
+ "loss": 0.7089,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.43,
1116
+ "learning_rate": 4.4539299009590926e-05,
1117
+ "loss": 0.722,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.43,
1122
+ "learning_rate": 4.4482384348140695e-05,
1123
+ "loss": 0.7197,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.43,
1128
+ "learning_rate": 4.44252113886571e-05,
1129
+ "loss": 0.7135,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.44,
1134
+ "learning_rate": 4.4367780889140945e-05,
1135
+ "loss": 0.7168,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.44,
1140
+ "learning_rate": 4.431009361100748e-05,
1141
+ "loss": 0.7267,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.44,
1146
+ "learning_rate": 4.4252150319076416e-05,
1147
+ "loss": 0.7241,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.44,
1152
+ "learning_rate": 4.419395178156159e-05,
1153
+ "loss": 0.7294,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.45,
1158
+ "learning_rate": 4.413549877006099e-05,
1159
+ "loss": 0.7392,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.45,
1164
+ "learning_rate": 4.4076792059546365e-05,
1165
+ "loss": 0.7058,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.45,
1170
+ "learning_rate": 4.401783242835306e-05,
1171
+ "loss": 0.721,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.45,
1176
+ "learning_rate": 4.395862065816959e-05,
1177
+ "loss": 0.7284,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.45,
1182
+ "learning_rate": 4.3899157534027396e-05,
1183
+ "loss": 0.7259,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.46,
1188
+ "learning_rate": 4.383944384429035e-05,
1189
+ "loss": 0.7378,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.46,
1194
+ "learning_rate": 4.3779480380644337e-05,
1195
+ "loss": 0.7118,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.46,
1200
+ "learning_rate": 4.3719267938086736e-05,
1201
+ "loss": 0.7318,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.46,
1206
+ "learning_rate": 4.3658807314915915e-05,
1207
+ "loss": 0.7335,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 0.47,
1212
+ "learning_rate": 4.359809931272062e-05,
1213
+ "loss": 0.7162,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 0.47,
1218
+ "learning_rate": 4.353714473636937e-05,
1219
+ "loss": 0.7279,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 0.47,
1224
+ "learning_rate": 4.3475944393999736e-05,
1225
+ "loss": 0.7344,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 0.47,
1230
+ "learning_rate": 4.34144990970077e-05,
1231
+ "loss": 0.7258,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 0.48,
1236
+ "learning_rate": 4.335280966003684e-05,
1237
+ "loss": 0.7202,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 0.48,
1242
+ "learning_rate": 4.329087690096755e-05,
1243
+ "loss": 0.7225,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 0.48,
1248
+ "learning_rate": 4.3228701640906197e-05,
1249
+ "loss": 0.7073,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 0.48,
1254
+ "learning_rate": 4.3166284704174244e-05,
1255
+ "loss": 0.7265,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 0.48,
1260
+ "learning_rate": 4.310362691829729e-05,
1261
+ "loss": 0.7229,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 0.49,
1266
+ "learning_rate": 4.304072911399415e-05,
1267
+ "loss": 0.6985,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 0.49,
1272
+ "learning_rate": 4.297759212516577e-05,
1273
+ "loss": 0.7316,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 0.49,
1278
+ "learning_rate": 4.291421678888425e-05,
1279
+ "loss": 0.7332,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 0.49,
1284
+ "learning_rate": 4.285060394538167e-05,
1285
+ "loss": 0.7162,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 0.5,
1290
+ "learning_rate": 4.278675443803903e-05,
1291
+ "loss": 0.7064,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 0.5,
1296
+ "learning_rate": 4.2722669113375e-05,
1297
+ "loss": 0.7234,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 0.5,
1302
+ "learning_rate": 4.2658348821034724e-05,
1303
+ "loss": 0.7086,
1304
+ "step": 2160
1305
+ },
1306
+ {
1307
+ "epoch": 0.5,
1308
+ "learning_rate": 4.2593794413778545e-05,
1309
+ "loss": 0.7277,
1310
+ "step": 2170
1311
+ },
1312
+ {
1313
+ "epoch": 0.51,
1314
+ "learning_rate": 4.2529006747470745e-05,
1315
+ "loss": 0.6973,
1316
+ "step": 2180
1317
+ },
1318
+ {
1319
+ "epoch": 0.51,
1320
+ "learning_rate": 4.2463986681068124e-05,
1321
+ "loss": 0.7221,
1322
+ "step": 2190
1323
+ },
1324
+ {
1325
+ "epoch": 0.51,
1326
+ "learning_rate": 4.2398735076608675e-05,
1327
+ "loss": 0.7184,
1328
+ "step": 2200
1329
+ },
1330
+ {
1331
+ "epoch": 0.51,
1332
+ "learning_rate": 4.233325279920011e-05,
1333
+ "loss": 0.7132,
1334
+ "step": 2210
1335
+ },
1336
+ {
1337
+ "epoch": 0.51,
1338
+ "learning_rate": 4.226754071700841e-05,
1339
+ "loss": 0.7208,
1340
+ "step": 2220
1341
+ },
1342
+ {
1343
+ "epoch": 0.52,
1344
+ "learning_rate": 4.220159970124633e-05,
1345
+ "loss": 0.7078,
1346
+ "step": 2230
1347
+ },
1348
+ {
1349
+ "epoch": 0.52,
1350
+ "learning_rate": 4.2135430626161834e-05,
1351
+ "loss": 0.7196,
1352
+ "step": 2240
1353
+ },
1354
+ {
1355
+ "epoch": 0.52,
1356
+ "learning_rate": 4.206903436902649e-05,
1357
+ "loss": 0.7239,
1358
+ "step": 2250
1359
+ },
1360
+ {
1361
+ "epoch": 0.52,
1362
+ "learning_rate": 4.200241181012385e-05,
1363
+ "loss": 0.7172,
1364
+ "step": 2260
1365
+ },
1366
+ {
1367
+ "epoch": 0.53,
1368
+ "learning_rate": 4.19355638327378e-05,
1369
+ "loss": 0.7256,
1370
+ "step": 2270
1371
+ },
1372
+ {
1373
+ "epoch": 0.53,
1374
+ "learning_rate": 4.1868491323140816e-05,
1375
+ "loss": 0.7117,
1376
+ "step": 2280
1377
+ },
1378
+ {
1379
+ "epoch": 0.53,
1380
+ "learning_rate": 4.180119517058222e-05,
1381
+ "loss": 0.7199,
1382
+ "step": 2290
1383
+ },
1384
+ {
1385
+ "epoch": 0.53,
1386
+ "learning_rate": 4.173367626727643e-05,
1387
+ "loss": 0.7023,
1388
+ "step": 2300
1389
+ },
1390
+ {
1391
+ "epoch": 0.54,
1392
+ "learning_rate": 4.166593550839108e-05,
1393
+ "loss": 0.7185,
1394
+ "step": 2310
1395
+ },
1396
+ {
1397
+ "epoch": 0.54,
1398
+ "learning_rate": 4.159797379203516e-05,
1399
+ "loss": 0.7167,
1400
+ "step": 2320
1401
+ },
1402
+ {
1403
+ "epoch": 0.54,
1404
+ "learning_rate": 4.152979201924715e-05,
1405
+ "loss": 0.7194,
1406
+ "step": 2330
1407
+ },
1408
+ {
1409
+ "epoch": 0.54,
1410
+ "learning_rate": 4.146139109398303e-05,
1411
+ "loss": 0.717,
1412
+ "step": 2340
1413
+ },
1414
+ {
1415
+ "epoch": 0.54,
1416
+ "learning_rate": 4.139277192310429e-05,
1417
+ "loss": 0.724,
1418
+ "step": 2350
1419
+ },
1420
+ {
1421
+ "epoch": 0.55,
1422
+ "learning_rate": 4.1323935416365965e-05,
1423
+ "loss": 0.6933,
1424
+ "step": 2360
1425
+ },
1426
+ {
1427
+ "epoch": 0.55,
1428
+ "learning_rate": 4.125488248640451e-05,
1429
+ "loss": 0.7074,
1430
+ "step": 2370
1431
+ },
1432
+ {
1433
+ "epoch": 0.55,
1434
+ "learning_rate": 4.118561404872574e-05,
1435
+ "loss": 0.7028,
1436
+ "step": 2380
1437
+ },
1438
+ {
1439
+ "epoch": 0.55,
1440
+ "learning_rate": 4.111613102169265e-05,
1441
+ "loss": 0.7182,
1442
+ "step": 2390
1443
+ },
1444
+ {
1445
+ "epoch": 0.56,
1446
+ "learning_rate": 4.1046434326513305e-05,
1447
+ "loss": 0.7224,
1448
+ "step": 2400
1449
+ },
1450
+ {
1451
+ "epoch": 0.56,
1452
+ "learning_rate": 4.097652488722856e-05,
1453
+ "loss": 0.7065,
1454
+ "step": 2410
1455
+ },
1456
+ {
1457
+ "epoch": 0.56,
1458
+ "learning_rate": 4.090640363069986e-05,
1459
+ "loss": 0.7074,
1460
+ "step": 2420
1461
+ },
1462
+ {
1463
+ "epoch": 0.56,
1464
+ "learning_rate": 4.08360714865969e-05,
1465
+ "loss": 0.7246,
1466
+ "step": 2430
1467
+ },
1468
+ {
1469
+ "epoch": 0.57,
1470
+ "learning_rate": 4.076552938738537e-05,
1471
+ "loss": 0.7071,
1472
+ "step": 2440
1473
+ },
1474
+ {
1475
+ "epoch": 0.57,
1476
+ "learning_rate": 4.069477826831451e-05,
1477
+ "loss": 0.7082,
1478
+ "step": 2450
1479
+ },
1480
+ {
1481
+ "epoch": 0.57,
1482
+ "learning_rate": 4.062381906740477e-05,
1483
+ "loss": 0.7042,
1484
+ "step": 2460
1485
+ },
1486
+ {
1487
+ "epoch": 0.57,
1488
+ "learning_rate": 4.055265272543536e-05,
1489
+ "loss": 0.7011,
1490
+ "step": 2470
1491
+ },
1492
+ {
1493
+ "epoch": 0.57,
1494
+ "learning_rate": 4.048128018593177e-05,
1495
+ "loss": 0.7155,
1496
+ "step": 2480
1497
+ },
1498
+ {
1499
+ "epoch": 0.58,
1500
+ "learning_rate": 4.040970239515325e-05,
1501
+ "loss": 0.7149,
1502
+ "step": 2490
1503
+ },
1504
+ {
1505
+ "epoch": 0.58,
1506
+ "learning_rate": 4.033792030208029e-05,
1507
+ "loss": 0.717,
1508
+ "step": 2500
1509
+ },
1510
+ {
1511
+ "epoch": 0.58,
1512
+ "learning_rate": 4.026593485840203e-05,
1513
+ "loss": 0.7237,
1514
+ "step": 2510
1515
+ },
1516
+ {
1517
+ "epoch": 0.58,
1518
+ "learning_rate": 4.019374701850362e-05,
1519
+ "loss": 0.7251,
1520
+ "step": 2520
1521
+ },
1522
+ {
1523
+ "epoch": 0.59,
1524
+ "learning_rate": 4.0121357739453604e-05,
1525
+ "loss": 0.7179,
1526
+ "step": 2530
1527
+ },
1528
+ {
1529
+ "epoch": 0.59,
1530
+ "learning_rate": 4.0048767980991206e-05,
1531
+ "loss": 0.7171,
1532
+ "step": 2540
1533
+ },
1534
+ {
1535
+ "epoch": 0.59,
1536
+ "learning_rate": 3.997597870551361e-05,
1537
+ "loss": 0.7124,
1538
+ "step": 2550
1539
+ },
1540
+ {
1541
+ "epoch": 0.59,
1542
+ "learning_rate": 3.990299087806319e-05,
1543
+ "loss": 0.7206,
1544
+ "step": 2560
1545
+ },
1546
+ {
1547
+ "epoch": 0.6,
1548
+ "learning_rate": 3.9829805466314775e-05,
1549
+ "loss": 0.72,
1550
+ "step": 2570
1551
+ },
1552
+ {
1553
+ "epoch": 0.6,
1554
+ "learning_rate": 3.975642344056271e-05,
1555
+ "loss": 0.6972,
1556
+ "step": 2580
1557
+ },
1558
+ {
1559
+ "epoch": 0.6,
1560
+ "learning_rate": 3.9682845773708085e-05,
1561
+ "loss": 0.7158,
1562
+ "step": 2590
1563
+ },
1564
+ {
1565
+ "epoch": 0.6,
1566
+ "learning_rate": 3.960907344124581e-05,
1567
+ "loss": 0.7083,
1568
+ "step": 2600
1569
+ },
1570
+ {
1571
+ "epoch": 0.61,
1572
+ "learning_rate": 3.9535107421251676e-05,
1573
+ "loss": 0.7069,
1574
+ "step": 2610
1575
+ },
1576
+ {
1577
+ "epoch": 0.61,
1578
+ "learning_rate": 3.946094869436938e-05,
1579
+ "loss": 0.7018,
1580
+ "step": 2620
1581
+ },
1582
+ {
1583
+ "epoch": 0.61,
1584
+ "learning_rate": 3.938659824379754e-05,
1585
+ "loss": 0.6973,
1586
+ "step": 2630
1587
+ },
1588
+ {
1589
+ "epoch": 0.61,
1590
+ "learning_rate": 3.931205705527664e-05,
1591
+ "loss": 0.7064,
1592
+ "step": 2640
1593
+ },
1594
+ {
1595
+ "epoch": 0.61,
1596
+ "learning_rate": 3.9237326117075995e-05,
1597
+ "loss": 0.702,
1598
+ "step": 2650
1599
+ },
1600
+ {
1601
+ "epoch": 0.62,
1602
+ "learning_rate": 3.9162406419980605e-05,
1603
+ "loss": 0.713,
1604
+ "step": 2660
1605
+ },
1606
+ {
1607
+ "epoch": 0.62,
1608
+ "learning_rate": 3.908729895727806e-05,
1609
+ "loss": 0.7054,
1610
+ "step": 2670
1611
+ },
1612
+ {
1613
+ "epoch": 0.62,
1614
+ "learning_rate": 3.9012004724745347e-05,
1615
+ "loss": 0.7056,
1616
+ "step": 2680
1617
+ },
1618
+ {
1619
+ "epoch": 0.62,
1620
+ "learning_rate": 3.893652472063562e-05,
1621
+ "loss": 0.7047,
1622
+ "step": 2690
1623
+ },
1624
+ {
1625
+ "epoch": 0.63,
1626
+ "learning_rate": 3.886085994566506e-05,
1627
+ "loss": 0.705,
1628
+ "step": 2700
1629
+ },
1630
+ {
1631
+ "epoch": 0.63,
1632
+ "learning_rate": 3.87850114029995e-05,
1633
+ "loss": 0.6996,
1634
+ "step": 2710
1635
+ },
1636
+ {
1637
+ "epoch": 0.63,
1638
+ "learning_rate": 3.870898009824118e-05,
1639
+ "loss": 0.6986,
1640
+ "step": 2720
1641
+ },
1642
+ {
1643
+ "epoch": 0.63,
1644
+ "learning_rate": 3.863276703941542e-05,
1645
+ "loss": 0.7054,
1646
+ "step": 2730
1647
+ },
1648
+ {
1649
+ "epoch": 0.64,
1650
+ "learning_rate": 3.855637323695721e-05,
1651
+ "loss": 0.7094,
1652
+ "step": 2740
1653
+ },
1654
+ {
1655
+ "epoch": 0.64,
1656
+ "learning_rate": 3.8479799703697884e-05,
1657
+ "loss": 0.7103,
1658
+ "step": 2750
1659
+ },
1660
+ {
1661
+ "epoch": 0.64,
1662
+ "learning_rate": 3.840304745485162e-05,
1663
+ "loss": 0.6921,
1664
+ "step": 2760
1665
+ },
1666
+ {
1667
+ "epoch": 0.64,
1668
+ "learning_rate": 3.8326117508002025e-05,
1669
+ "loss": 0.7177,
1670
+ "step": 2770
1671
+ },
1672
+ {
1673
+ "epoch": 0.64,
1674
+ "learning_rate": 3.8249010883088646e-05,
1675
+ "loss": 0.7163,
1676
+ "step": 2780
1677
+ },
1678
+ {
1679
+ "epoch": 0.65,
1680
+ "learning_rate": 3.817172860239339e-05,
1681
+ "loss": 0.7135,
1682
+ "step": 2790
1683
+ },
1684
+ {
1685
+ "epoch": 0.65,
1686
+ "learning_rate": 3.8094271690527076e-05,
1687
+ "loss": 0.7141,
1688
+ "step": 2800
1689
+ },
1690
+ {
1691
+ "epoch": 0.65,
1692
+ "learning_rate": 3.801664117441572e-05,
1693
+ "loss": 0.704,
1694
+ "step": 2810
1695
+ },
1696
+ {
1697
+ "epoch": 0.65,
1698
+ "learning_rate": 3.7938838083287056e-05,
1699
+ "loss": 0.7092,
1700
+ "step": 2820
1701
+ },
1702
+ {
1703
+ "epoch": 0.66,
1704
+ "learning_rate": 3.7860863448656765e-05,
1705
+ "loss": 0.6986,
1706
+ "step": 2830
1707
+ },
1708
+ {
1709
+ "epoch": 0.66,
1710
+ "learning_rate": 3.778271830431488e-05,
1711
+ "loss": 0.7153,
1712
+ "step": 2840
1713
+ },
1714
+ {
1715
+ "epoch": 0.66,
1716
+ "learning_rate": 3.770440368631207e-05,
1717
+ "loss": 0.7048,
1718
+ "step": 2850
1719
+ },
1720
+ {
1721
+ "epoch": 0.66,
1722
+ "learning_rate": 3.762592063294588e-05,
1723
+ "loss": 0.7103,
1724
+ "step": 2860
1725
+ },
1726
+ {
1727
+ "epoch": 0.67,
1728
+ "learning_rate": 3.7547270184746976e-05,
1729
+ "loss": 0.7048,
1730
+ "step": 2870
1731
+ },
1732
+ {
1733
+ "epoch": 0.67,
1734
+ "learning_rate": 3.746845338446533e-05,
1735
+ "loss": 0.7083,
1736
+ "step": 2880
1737
+ },
1738
+ {
1739
+ "epoch": 0.67,
1740
+ "learning_rate": 3.738947127705644e-05,
1741
+ "loss": 0.7011,
1742
+ "step": 2890
1743
+ },
1744
+ {
1745
+ "epoch": 0.67,
1746
+ "learning_rate": 3.731032490966746e-05,
1747
+ "loss": 0.6959,
1748
+ "step": 2900
1749
+ },
1750
+ {
1751
+ "epoch": 0.67,
1752
+ "learning_rate": 3.723101533162329e-05,
1753
+ "loss": 0.7043,
1754
+ "step": 2910
1755
+ },
1756
+ {
1757
+ "epoch": 0.68,
1758
+ "learning_rate": 3.715154359441265e-05,
1759
+ "loss": 0.7101,
1760
+ "step": 2920
1761
+ },
1762
+ {
1763
+ "epoch": 0.68,
1764
+ "learning_rate": 3.7071910751674256e-05,
1765
+ "loss": 0.7071,
1766
+ "step": 2930
1767
+ },
1768
+ {
1769
+ "epoch": 0.68,
1770
+ "learning_rate": 3.6992117859182674e-05,
1771
+ "loss": 0.7158,
1772
+ "step": 2940
1773
+ },
1774
+ {
1775
+ "epoch": 0.68,
1776
+ "learning_rate": 3.691216597483447e-05,
1777
+ "loss": 0.7046,
1778
+ "step": 2950
1779
+ },
1780
+ {
1781
+ "epoch": 0.69,
1782
+ "learning_rate": 3.683205615863412e-05,
1783
+ "loss": 0.7003,
1784
+ "step": 2960
1785
+ },
1786
+ {
1787
+ "epoch": 0.69,
1788
+ "learning_rate": 3.6751789472679946e-05,
1789
+ "loss": 0.7067,
1790
+ "step": 2970
1791
+ },
1792
+ {
1793
+ "epoch": 0.69,
1794
+ "learning_rate": 3.667136698115008e-05,
1795
+ "loss": 0.7002,
1796
+ "step": 2980
1797
+ },
1798
+ {
1799
+ "epoch": 0.69,
1800
+ "learning_rate": 3.65907897502883e-05,
1801
+ "loss": 0.707,
1802
+ "step": 2990
1803
+ },
1804
+ {
1805
+ "epoch": 0.7,
1806
+ "learning_rate": 3.651005884838995e-05,
1807
+ "loss": 0.7069,
1808
+ "step": 3000
1809
+ },
1810
+ {
1811
+ "epoch": 0.7,
1812
+ "learning_rate": 3.6429175345787735e-05,
1813
+ "loss": 0.7023,
1814
+ "step": 3010
1815
+ },
1816
+ {
1817
+ "epoch": 0.7,
1818
+ "learning_rate": 3.6348140314837553e-05,
1819
+ "loss": 0.6921,
1820
+ "step": 3020
1821
+ },
1822
+ {
1823
+ "epoch": 0.7,
1824
+ "learning_rate": 3.626695482990425e-05,
1825
+ "loss": 0.6896,
1826
+ "step": 3030
1827
+ },
1828
+ {
1829
+ "epoch": 0.7,
1830
+ "learning_rate": 3.618561996734743e-05,
1831
+ "loss": 0.7025,
1832
+ "step": 3040
1833
+ },
1834
+ {
1835
+ "epoch": 0.71,
1836
+ "learning_rate": 3.610413680550711e-05,
1837
+ "loss": 0.7057,
1838
+ "step": 3050
1839
+ },
1840
+ {
1841
+ "epoch": 0.71,
1842
+ "learning_rate": 3.6022506424689485e-05,
1843
+ "loss": 0.6988,
1844
+ "step": 3060
1845
+ },
1846
+ {
1847
+ "epoch": 0.71,
1848
+ "learning_rate": 3.594072990715258e-05,
1849
+ "loss": 0.722,
1850
+ "step": 3070
1851
+ },
1852
+ {
1853
+ "epoch": 0.71,
1854
+ "learning_rate": 3.5858808337091925e-05,
1855
+ "loss": 0.7108,
1856
+ "step": 3080
1857
+ },
1858
+ {
1859
+ "epoch": 0.72,
1860
+ "learning_rate": 3.577674280062612e-05,
1861
+ "loss": 0.7075,
1862
+ "step": 3090
1863
+ },
1864
+ {
1865
+ "epoch": 0.72,
1866
+ "learning_rate": 3.569453438578251e-05,
1867
+ "loss": 0.7048,
1868
+ "step": 3100
1869
+ },
1870
+ {
1871
+ "epoch": 0.72,
1872
+ "learning_rate": 3.561218418248271e-05,
1873
+ "loss": 0.6934,
1874
+ "step": 3110
1875
+ },
1876
+ {
1877
+ "epoch": 0.72,
1878
+ "learning_rate": 3.552969328252817e-05,
1879
+ "loss": 0.7007,
1880
+ "step": 3120
1881
+ },
1882
+ {
1883
+ "epoch": 0.73,
1884
+ "learning_rate": 3.544706277958572e-05,
1885
+ "loss": 0.7221,
1886
+ "step": 3130
1887
+ },
1888
+ {
1889
+ "epoch": 0.73,
1890
+ "learning_rate": 3.5364293769173e-05,
1891
+ "loss": 0.7069,
1892
+ "step": 3140
1893
+ },
1894
+ {
1895
+ "epoch": 0.73,
1896
+ "learning_rate": 3.528138734864406e-05,
1897
+ "loss": 0.7013,
1898
+ "step": 3150
1899
+ },
1900
+ {
1901
+ "epoch": 0.73,
1902
+ "learning_rate": 3.5198344617174685e-05,
1903
+ "loss": 0.7161,
1904
+ "step": 3160
1905
+ },
1906
+ {
1907
+ "epoch": 0.73,
1908
+ "learning_rate": 3.511516667574788e-05,
1909
+ "loss": 0.6995,
1910
+ "step": 3170
1911
+ },
1912
+ {
1913
+ "epoch": 0.74,
1914
+ "learning_rate": 3.5031854627139284e-05,
1915
+ "loss": 0.7157,
1916
+ "step": 3180
1917
+ },
1918
+ {
1919
+ "epoch": 0.74,
1920
+ "learning_rate": 3.4948409575902516e-05,
1921
+ "loss": 0.6929,
1922
+ "step": 3190
1923
+ },
1924
+ {
1925
+ "epoch": 0.74,
1926
+ "learning_rate": 3.486483262835457e-05,
1927
+ "loss": 0.6973,
1928
+ "step": 3200
1929
+ },
1930
+ {
1931
+ "epoch": 0.74,
1932
+ "learning_rate": 3.478112489256111e-05,
1933
+ "loss": 0.6941,
1934
+ "step": 3210
1935
+ },
1936
+ {
1937
+ "epoch": 0.75,
1938
+ "learning_rate": 3.4697287478321784e-05,
1939
+ "loss": 0.7059,
1940
+ "step": 3220
1941
+ },
1942
+ {
1943
+ "epoch": 0.75,
1944
+ "learning_rate": 3.4613321497155546e-05,
1945
+ "loss": 0.6833,
1946
+ "step": 3230
1947
+ },
1948
+ {
1949
+ "epoch": 0.75,
1950
+ "learning_rate": 3.452922806228587e-05,
1951
+ "loss": 0.6766,
1952
+ "step": 3240
1953
+ },
1954
+ {
1955
+ "epoch": 0.75,
1956
+ "learning_rate": 3.444500828862604e-05,
1957
+ "loss": 0.7001,
1958
+ "step": 3250
1959
+ },
1960
+ {
1961
+ "epoch": 0.76,
1962
+ "learning_rate": 3.436066329276432e-05,
1963
+ "loss": 0.7082,
1964
+ "step": 3260
1965
+ },
1966
+ {
1967
+ "epoch": 0.76,
1968
+ "learning_rate": 3.4276194192949185e-05,
1969
+ "loss": 0.7035,
1970
+ "step": 3270
1971
+ },
1972
+ {
1973
+ "epoch": 0.76,
1974
+ "learning_rate": 3.4191602109074484e-05,
1975
+ "loss": 0.6948,
1976
+ "step": 3280
1977
+ },
1978
+ {
1979
+ "epoch": 0.76,
1980
+ "learning_rate": 3.4106888162664584e-05,
1981
+ "loss": 0.7041,
1982
+ "step": 3290
1983
+ },
1984
+ {
1985
+ "epoch": 0.76,
1986
+ "learning_rate": 3.402205347685953e-05,
1987
+ "loss": 0.7004,
1988
+ "step": 3300
1989
+ },
1990
+ {
1991
+ "epoch": 0.77,
1992
+ "learning_rate": 3.39370991764001e-05,
1993
+ "loss": 0.7003,
1994
+ "step": 3310
1995
+ },
1996
+ {
1997
+ "epoch": 0.77,
1998
+ "learning_rate": 3.385202638761297e-05,
1999
+ "loss": 0.6908,
2000
+ "step": 3320
2001
+ },
2002
+ {
2003
+ "epoch": 0.77,
2004
+ "learning_rate": 3.376683623839569e-05,
2005
+ "loss": 0.7059,
2006
+ "step": 3330
2007
+ },
2008
+ {
2009
+ "epoch": 0.77,
2010
+ "learning_rate": 3.3681529858201814e-05,
2011
+ "loss": 0.7027,
2012
+ "step": 3340
2013
+ },
2014
+ {
2015
+ "epoch": 0.78,
2016
+ "learning_rate": 3.3596108378025875e-05,
2017
+ "loss": 0.6984,
2018
+ "step": 3350
2019
+ },
2020
+ {
2021
+ "epoch": 0.78,
2022
+ "learning_rate": 3.351057293038841e-05,
2023
+ "loss": 0.7054,
2024
+ "step": 3360
2025
+ },
2026
+ {
2027
+ "epoch": 0.78,
2028
+ "learning_rate": 3.342492464932094e-05,
2029
+ "loss": 0.6803,
2030
+ "step": 3370
2031
+ },
2032
+ {
2033
+ "epoch": 0.78,
2034
+ "learning_rate": 3.333916467035092e-05,
2035
+ "loss": 0.6924,
2036
+ "step": 3380
2037
+ },
2038
+ {
2039
+ "epoch": 0.79,
2040
+ "learning_rate": 3.325329413048673e-05,
2041
+ "loss": 0.7143,
2042
+ "step": 3390
2043
+ },
2044
+ {
2045
+ "epoch": 0.79,
2046
+ "learning_rate": 3.316731416820254e-05,
2047
+ "loss": 0.6986,
2048
+ "step": 3400
2049
+ },
2050
+ {
2051
+ "epoch": 0.79,
2052
+ "learning_rate": 3.308122592342326e-05,
2053
+ "loss": 0.6875,
2054
+ "step": 3410
2055
+ },
2056
+ {
2057
+ "epoch": 0.79,
2058
+ "learning_rate": 3.2995030537509396e-05,
2059
+ "loss": 0.6934,
2060
+ "step": 3420
2061
+ },
2062
+ {
2063
+ "epoch": 0.8,
2064
+ "learning_rate": 3.290872915324197e-05,
2065
+ "loss": 0.698,
2066
+ "step": 3430
2067
+ },
2068
+ {
2069
+ "epoch": 0.8,
2070
+ "learning_rate": 3.2822322914807305e-05,
2071
+ "loss": 0.6976,
2072
+ "step": 3440
2073
+ },
2074
+ {
2075
+ "epoch": 0.8,
2076
+ "learning_rate": 3.273581296778188e-05,
2077
+ "loss": 0.6914,
2078
+ "step": 3450
2079
+ },
2080
+ {
2081
+ "epoch": 0.8,
2082
+ "learning_rate": 3.264920045911716e-05,
2083
+ "loss": 0.6902,
2084
+ "step": 3460
2085
+ },
2086
+ {
2087
+ "epoch": 0.8,
2088
+ "learning_rate": 3.256248653712437e-05,
2089
+ "loss": 0.6993,
2090
+ "step": 3470
2091
+ },
2092
+ {
2093
+ "epoch": 0.81,
2094
+ "learning_rate": 3.247567235145929e-05,
2095
+ "loss": 0.679,
2096
+ "step": 3480
2097
+ },
2098
+ {
2099
+ "epoch": 0.81,
2100
+ "learning_rate": 3.2388759053106965e-05,
2101
+ "loss": 0.6851,
2102
+ "step": 3490
2103
+ },
2104
+ {
2105
+ "epoch": 0.81,
2106
+ "learning_rate": 3.230174779436651e-05,
2107
+ "loss": 0.6802,
2108
+ "step": 3500
2109
+ },
2110
+ {
2111
+ "epoch": 0.81,
2112
+ "learning_rate": 3.22146397288358e-05,
2113
+ "loss": 0.7058,
2114
+ "step": 3510
2115
+ },
2116
+ {
2117
+ "epoch": 0.82,
2118
+ "learning_rate": 3.2127436011396156e-05,
2119
+ "loss": 0.6926,
2120
+ "step": 3520
2121
+ },
2122
+ {
2123
+ "epoch": 0.82,
2124
+ "learning_rate": 3.2040137798197066e-05,
2125
+ "loss": 0.7031,
2126
+ "step": 3530
2127
+ },
2128
+ {
2129
+ "epoch": 0.82,
2130
+ "learning_rate": 3.195274624664086e-05,
2131
+ "loss": 0.7132,
2132
+ "step": 3540
2133
+ },
2134
+ {
2135
+ "epoch": 0.82,
2136
+ "learning_rate": 3.186526251536733e-05,
2137
+ "loss": 0.6929,
2138
+ "step": 3550
2139
+ },
2140
+ {
2141
+ "epoch": 0.83,
2142
+ "learning_rate": 3.177768776423839e-05,
2143
+ "loss": 0.6959,
2144
+ "step": 3560
2145
+ },
2146
+ {
2147
+ "epoch": 0.83,
2148
+ "learning_rate": 3.1690023154322714e-05,
2149
+ "loss": 0.7035,
2150
+ "step": 3570
2151
+ },
2152
+ {
2153
+ "epoch": 0.83,
2154
+ "learning_rate": 3.1602269847880315e-05,
2155
+ "loss": 0.6965,
2156
+ "step": 3580
2157
+ },
2158
+ {
2159
+ "epoch": 0.83,
2160
+ "learning_rate": 3.1514429008347154e-05,
2161
+ "loss": 0.679,
2162
+ "step": 3590
2163
+ },
2164
+ {
2165
+ "epoch": 0.83,
2166
+ "learning_rate": 3.14265018003197e-05,
2167
+ "loss": 0.6861,
2168
+ "step": 3600
2169
+ },
2170
+ {
2171
+ "epoch": 0.84,
2172
+ "learning_rate": 3.13384893895395e-05,
2173
+ "loss": 0.6964,
2174
+ "step": 3610
2175
+ },
2176
+ {
2177
+ "epoch": 0.84,
2178
+ "learning_rate": 3.125039294287773e-05,
2179
+ "loss": 0.7003,
2180
+ "step": 3620
2181
+ },
2182
+ {
2183
+ "epoch": 0.84,
2184
+ "learning_rate": 3.11622136283197e-05,
2185
+ "loss": 0.6845,
2186
+ "step": 3630
2187
+ },
2188
+ {
2189
+ "epoch": 0.84,
2190
+ "learning_rate": 3.1073952614949406e-05,
2191
+ "loss": 0.6926,
2192
+ "step": 3640
2193
+ },
2194
+ {
2195
+ "epoch": 0.85,
2196
+ "learning_rate": 3.0985611072934e-05,
2197
+ "loss": 0.6953,
2198
+ "step": 3650
2199
+ },
2200
+ {
2201
+ "epoch": 0.85,
2202
+ "learning_rate": 3.089719017350828e-05,
2203
+ "loss": 0.6821,
2204
+ "step": 3660
2205
+ },
2206
+ {
2207
+ "epoch": 0.85,
2208
+ "learning_rate": 3.080869108895915e-05,
2209
+ "loss": 0.6903,
2210
+ "step": 3670
2211
+ },
2212
+ {
2213
+ "epoch": 0.85,
2214
+ "learning_rate": 3.072011499261016e-05,
2215
+ "loss": 0.7159,
2216
+ "step": 3680
2217
+ },
2218
+ {
2219
+ "epoch": 0.86,
2220
+ "learning_rate": 3.0631463058805796e-05,
2221
+ "loss": 0.6961,
2222
+ "step": 3690
2223
+ },
2224
+ {
2225
+ "epoch": 0.86,
2226
+ "learning_rate": 3.0542736462896065e-05,
2227
+ "loss": 0.7061,
2228
+ "step": 3700
2229
+ },
2230
+ {
2231
+ "epoch": 0.86,
2232
+ "learning_rate": 3.045393638122082e-05,
2233
+ "loss": 0.7089,
2234
+ "step": 3710
2235
+ },
2236
+ {
2237
+ "epoch": 0.86,
2238
+ "learning_rate": 3.03650639910942e-05,
2239
+ "loss": 0.672,
2240
+ "step": 3720
2241
+ },
2242
+ {
2243
+ "epoch": 0.86,
2244
+ "learning_rate": 3.0276120470789004e-05,
2245
+ "loss": 0.7006,
2246
+ "step": 3730
2247
+ },
2248
+ {
2249
+ "epoch": 0.87,
2250
+ "learning_rate": 3.0187106999521074e-05,
2251
+ "loss": 0.6945,
2252
+ "step": 3740
2253
+ },
2254
+ {
2255
+ "epoch": 0.87,
2256
+ "learning_rate": 3.0098024757433675e-05,
2257
+ "loss": 0.6957,
2258
+ "step": 3750
2259
+ },
2260
+ {
2261
+ "epoch": 0.87,
2262
+ "learning_rate": 3.000887492558183e-05,
2263
+ "loss": 0.6819,
2264
+ "step": 3760
2265
+ },
2266
+ {
2267
+ "epoch": 0.87,
2268
+ "learning_rate": 2.991965868591668e-05,
2269
+ "loss": 0.6861,
2270
+ "step": 3770
2271
+ },
2272
+ {
2273
+ "epoch": 0.88,
2274
+ "learning_rate": 2.983037722126978e-05,
2275
+ "loss": 0.7015,
2276
+ "step": 3780
2277
+ },
2278
+ {
2279
+ "epoch": 0.88,
2280
+ "learning_rate": 2.9741031715337464e-05,
2281
+ "loss": 0.6852,
2282
+ "step": 3790
2283
+ },
2284
+ {
2285
+ "epoch": 0.88,
2286
+ "learning_rate": 2.9651623352665114e-05,
2287
+ "loss": 0.7009,
2288
+ "step": 3800
2289
+ },
2290
+ {
2291
+ "epoch": 0.88,
2292
+ "learning_rate": 2.9562153318631468e-05,
2293
+ "loss": 0.6885,
2294
+ "step": 3810
2295
+ },
2296
+ {
2297
+ "epoch": 0.89,
2298
+ "learning_rate": 2.947262279943292e-05,
2299
+ "loss": 0.6701,
2300
+ "step": 3820
2301
+ },
2302
+ {
2303
+ "epoch": 0.89,
2304
+ "learning_rate": 2.9383032982067753e-05,
2305
+ "loss": 0.6847,
2306
+ "step": 3830
2307
+ },
2308
+ {
2309
+ "epoch": 0.89,
2310
+ "learning_rate": 2.929338505432045e-05,
2311
+ "loss": 0.6903,
2312
+ "step": 3840
2313
+ },
2314
+ {
2315
+ "epoch": 0.89,
2316
+ "learning_rate": 2.9203680204745914e-05,
2317
+ "loss": 0.6992,
2318
+ "step": 3850
2319
+ },
2320
+ {
2321
+ "epoch": 0.89,
2322
+ "learning_rate": 2.9113919622653707e-05,
2323
+ "loss": 0.7075,
2324
+ "step": 3860
2325
+ },
2326
+ {
2327
+ "epoch": 0.9,
2328
+ "learning_rate": 2.902410449809232e-05,
2329
+ "loss": 0.7072,
2330
+ "step": 3870
2331
+ },
2332
+ {
2333
+ "epoch": 0.9,
2334
+ "learning_rate": 2.8934236021833348e-05,
2335
+ "loss": 0.6897,
2336
+ "step": 3880
2337
+ },
2338
+ {
2339
+ "epoch": 0.9,
2340
+ "learning_rate": 2.8844315385355737e-05,
2341
+ "loss": 0.7081,
2342
+ "step": 3890
2343
+ },
2344
+ {
2345
+ "epoch": 0.9,
2346
+ "learning_rate": 2.8754343780829968e-05,
2347
+ "loss": 0.7032,
2348
+ "step": 3900
2349
+ },
2350
+ {
2351
+ "epoch": 0.91,
2352
+ "learning_rate": 2.866432240110226e-05,
2353
+ "loss": 0.6811,
2354
+ "step": 3910
2355
+ },
2356
+ {
2357
+ "epoch": 0.91,
2358
+ "learning_rate": 2.857425243967875e-05,
2359
+ "loss": 0.6911,
2360
+ "step": 3920
2361
+ },
2362
+ {
2363
+ "epoch": 0.91,
2364
+ "learning_rate": 2.8484135090709686e-05,
2365
+ "loss": 0.6862,
2366
+ "step": 3930
2367
+ },
2368
+ {
2369
+ "epoch": 0.91,
2370
+ "learning_rate": 2.839397154897356e-05,
2371
+ "loss": 0.6785,
2372
+ "step": 3940
2373
+ },
2374
+ {
2375
+ "epoch": 0.92,
2376
+ "learning_rate": 2.83037630098613e-05,
2377
+ "loss": 0.7007,
2378
+ "step": 3950
2379
+ },
2380
+ {
2381
+ "epoch": 0.92,
2382
+ "learning_rate": 2.821351066936043e-05,
2383
+ "loss": 0.675,
2384
+ "step": 3960
2385
+ },
2386
+ {
2387
+ "epoch": 0.92,
2388
+ "learning_rate": 2.812321572403915e-05,
2389
+ "loss": 0.6969,
2390
+ "step": 3970
2391
+ },
2392
+ {
2393
+ "epoch": 0.92,
2394
+ "learning_rate": 2.803287937103055e-05,
2395
+ "loss": 0.694,
2396
+ "step": 3980
2397
+ },
2398
+ {
2399
+ "epoch": 0.92,
2400
+ "learning_rate": 2.7942502808016703e-05,
2401
+ "loss": 0.6951,
2402
+ "step": 3990
2403
+ },
2404
+ {
2405
+ "epoch": 0.93,
2406
+ "learning_rate": 2.7852087233212763e-05,
2407
+ "loss": 0.679,
2408
+ "step": 4000
2409
+ }
2410
+ ],
2411
+ "max_steps": 8628,
2412
+ "num_train_epochs": 2,
2413
+ "total_flos": 1.2632560281122243e+18,
2414
+ "trial_name": null,
2415
+ "trial_params": null
2416
+ }
checkpoint-4000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ce444f5cfe8a7851710b073da2f9ede89d9fad5a6d091ebf47aee5e6b0d0797
3
+ size 3259
checkpoint-5000/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
checkpoint-5000/adapter_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "./Llama-2-7b-chat-hf",
4
+ "bias": "none",
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": 32.0,
11
+ "lora_dropout": 0.1,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 8,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "q_proj",
18
+ "v_proj"
19
+ ],
20
+ "task_type": "CAUSAL_LM"
21
+ }
checkpoint-5000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d427f68435cea8f41617a7481c34e89d590351f9dc4b2189a03cec353683da7
3
+ size 16821197
checkpoint-5000/finetuning_args.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dpo_beta": 0.1,
3
+ "finetuning_type": "lora",
4
+ "lora_alpha": 32.0,
5
+ "lora_dropout": 0.1,
6
+ "lora_rank": 8,
7
+ "lora_target": [
8
+ "q_proj",
9
+ "v_proj"
10
+ ],
11
+ "name_module_trainable": "mlp",
12
+ "num_hidden_layers": 32,
13
+ "num_layer_trainable": 3,
14
+ "ppo_score_norm": false,
15
+ "resume_lora_training": true
16
+ }
checkpoint-5000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a005dee1942e29d8cced0207d73d078856515733526d8cdeeeaa9e4505043845
3
+ size 33661637
checkpoint-5000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e3e5a574a8619e4eed9e85b6d91988d30897ee004d5f8448ee000d274a1f212
3
+ size 18663
checkpoint-5000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ebb26972b6123215ec703a7cfc0ae0226c90a7ece86ba4cb59c9ecf2ccb5ef1
3
+ size 627
checkpoint-5000/trainer_state.json ADDED
@@ -0,0 +1,3016 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.1590171534538711,
5
+ "global_step": 5000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.999983427462679e-05,
13
+ "loss": 1.3476,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 4.999933710070436e-05,
19
+ "loss": 1.1454,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 4.999850848482424e-05,
25
+ "loss": 1.0263,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 4.999734843797226e-05,
31
+ "loss": 1.0283,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 4.9995856975528347e-05,
37
+ "loss": 0.9733,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 4.999403411726636e-05,
43
+ "loss": 0.9591,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 4.999187988735381e-05,
49
+ "loss": 0.9617,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 4.998939431435153e-05,
55
+ "loss": 0.9414,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 4.998657743121333e-05,
61
+ "loss": 0.9233,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 4.998342927528553e-05,
67
+ "loss": 0.9236,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 4.997994988830648e-05,
73
+ "loss": 0.8743,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.03,
78
+ "learning_rate": 4.997613931640598e-05,
79
+ "loss": 0.895,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 4.997199761010472e-05,
85
+ "loss": 0.8831,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 4.996752482431356e-05,
91
+ "loss": 0.8799,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 4.996272101833284e-05,
97
+ "loss": 0.877,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.04,
102
+ "learning_rate": 4.995758625585155e-05,
103
+ "loss": 0.8745,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.04,
108
+ "learning_rate": 4.995212060494653e-05,
109
+ "loss": 0.8716,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 4.9946324138081534e-05,
115
+ "loss": 0.8534,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 4.99401969321063e-05,
121
+ "loss": 0.8449,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.05,
126
+ "learning_rate": 4.9933739068255513e-05,
127
+ "loss": 0.8588,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.05,
132
+ "learning_rate": 4.992695063214772e-05,
133
+ "loss": 0.8409,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.05,
138
+ "learning_rate": 4.99198317137842e-05,
139
+ "loss": 0.8517,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 4.99123824075478e-05,
145
+ "loss": 0.8698,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.06,
150
+ "learning_rate": 4.990460281220164e-05,
151
+ "loss": 0.848,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.06,
156
+ "learning_rate": 4.9896493030887815e-05,
157
+ "loss": 0.8322,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.06,
162
+ "learning_rate": 4.988805317112607e-05,
163
+ "loss": 0.8376,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.06,
168
+ "learning_rate": 4.98792833448123e-05,
169
+ "loss": 0.8309,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 4.987018366821713e-05,
175
+ "loss": 0.8195,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.07,
180
+ "learning_rate": 4.9860754261984346e-05,
181
+ "loss": 0.8313,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.07,
186
+ "learning_rate": 4.98509952511293e-05,
187
+ "loss": 0.8115,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.07,
192
+ "learning_rate": 4.984090676503723e-05,
193
+ "loss": 0.8224,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.07,
198
+ "learning_rate": 4.983048893746162e-05,
199
+ "loss": 0.8157,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.08,
204
+ "learning_rate": 4.9819741906522306e-05,
205
+ "loss": 0.8165,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.08,
210
+ "learning_rate": 4.980866581470376e-05,
211
+ "loss": 0.8144,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.08,
216
+ "learning_rate": 4.979726080885314e-05,
217
+ "loss": 0.8147,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.08,
222
+ "learning_rate": 4.978552704017835e-05,
223
+ "loss": 0.7894,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.09,
228
+ "learning_rate": 4.977346466424605e-05,
229
+ "loss": 0.8144,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.09,
234
+ "learning_rate": 4.976107384097957e-05,
235
+ "loss": 0.7984,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.09,
240
+ "learning_rate": 4.974835473465682e-05,
241
+ "loss": 0.8188,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.09,
246
+ "learning_rate": 4.97353075139081e-05,
247
+ "loss": 0.8049,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.1,
252
+ "learning_rate": 4.9721932351713846e-05,
253
+ "loss": 0.8154,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.1,
258
+ "learning_rate": 4.9708229425402354e-05,
259
+ "loss": 0.7953,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.1,
264
+ "learning_rate": 4.9694198916647425e-05,
265
+ "loss": 0.7996,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.1,
270
+ "learning_rate": 4.967984101146598e-05,
271
+ "loss": 0.7979,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.1,
276
+ "learning_rate": 4.966515590021554e-05,
277
+ "loss": 0.8106,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.11,
282
+ "learning_rate": 4.965014377759176e-05,
283
+ "loss": 0.8008,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.11,
288
+ "learning_rate": 4.9634804842625807e-05,
289
+ "loss": 0.8045,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.11,
294
+ "learning_rate": 4.961913929868172e-05,
295
+ "loss": 0.786,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.11,
300
+ "learning_rate": 4.960314735345377e-05,
301
+ "loss": 0.7897,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.12,
306
+ "learning_rate": 4.958682921896365e-05,
307
+ "loss": 0.7855,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.12,
312
+ "learning_rate": 4.957018511155765e-05,
313
+ "loss": 0.7657,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.12,
318
+ "learning_rate": 4.955321525190386e-05,
319
+ "loss": 0.779,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.12,
324
+ "learning_rate": 4.953591986498918e-05,
325
+ "loss": 0.7923,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.13,
330
+ "learning_rate": 4.9518299180116366e-05,
331
+ "loss": 0.7732,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.13,
336
+ "learning_rate": 4.9500353430900984e-05,
337
+ "loss": 0.7898,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.13,
342
+ "learning_rate": 4.948208285526832e-05,
343
+ "loss": 0.7796,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.13,
348
+ "learning_rate": 4.94634876954502e-05,
349
+ "loss": 0.7819,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.13,
354
+ "learning_rate": 4.944456819798182e-05,
355
+ "loss": 0.7777,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.14,
360
+ "learning_rate": 4.942532461369843e-05,
361
+ "loss": 0.7815,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.14,
366
+ "learning_rate": 4.940575719773205e-05,
367
+ "loss": 0.7807,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.14,
372
+ "learning_rate": 4.938586620950807e-05,
373
+ "loss": 0.7836,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 0.14,
378
+ "learning_rate": 4.9365651912741806e-05,
379
+ "loss": 0.7795,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 0.15,
384
+ "learning_rate": 4.934511457543499e-05,
385
+ "loss": 0.7844,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 0.15,
390
+ "learning_rate": 4.932425446987228e-05,
391
+ "loss": 0.7865,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 0.15,
396
+ "learning_rate": 4.930307187261757e-05,
397
+ "loss": 0.7751,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 0.15,
402
+ "learning_rate": 4.928156706451035e-05,
403
+ "loss": 0.7705,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 0.16,
408
+ "learning_rate": 4.925974033066202e-05,
409
+ "loss": 0.773,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 0.16,
414
+ "learning_rate": 4.923759196045208e-05,
415
+ "loss": 0.7714,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 0.16,
420
+ "learning_rate": 4.921512224752426e-05,
421
+ "loss": 0.7788,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 0.16,
426
+ "learning_rate": 4.9192331489782705e-05,
427
+ "loss": 0.7755,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 0.16,
432
+ "learning_rate": 4.9169219989387956e-05,
433
+ "loss": 0.7682,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 0.17,
438
+ "learning_rate": 4.9145788052752975e-05,
439
+ "loss": 0.788,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 0.17,
444
+ "learning_rate": 4.912203599053907e-05,
445
+ "loss": 0.7752,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 0.17,
450
+ "learning_rate": 4.90979641176518e-05,
451
+ "loss": 0.7786,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 0.17,
456
+ "learning_rate": 4.907357275323677e-05,
457
+ "loss": 0.7848,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 0.18,
462
+ "learning_rate": 4.904886222067542e-05,
463
+ "loss": 0.7747,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 0.18,
468
+ "learning_rate": 4.9023832847580724e-05,
469
+ "loss": 0.7784,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 0.18,
474
+ "learning_rate": 4.899848496579287e-05,
475
+ "loss": 0.768,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 0.18,
480
+ "learning_rate": 4.897281891137481e-05,
481
+ "loss": 0.7684,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 0.19,
486
+ "learning_rate": 4.8946835024607885e-05,
487
+ "loss": 0.7696,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 0.19,
492
+ "learning_rate": 4.892053364998722e-05,
493
+ "loss": 0.7552,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 0.19,
498
+ "learning_rate": 4.889391513621725e-05,
499
+ "loss": 0.7584,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 0.19,
504
+ "learning_rate": 4.886697983620699e-05,
505
+ "loss": 0.7682,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 0.19,
510
+ "learning_rate": 4.883972810706548e-05,
511
+ "loss": 0.7701,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 0.2,
516
+ "learning_rate": 4.881216031009696e-05,
517
+ "loss": 0.7637,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 0.2,
522
+ "learning_rate": 4.878427681079608e-05,
523
+ "loss": 0.7673,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 0.2,
528
+ "learning_rate": 4.875607797884312e-05,
529
+ "loss": 0.7579,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 0.2,
534
+ "learning_rate": 4.8727564188099054e-05,
535
+ "loss": 0.7629,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 0.21,
540
+ "learning_rate": 4.8698735816600535e-05,
541
+ "loss": 0.7766,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 0.21,
546
+ "learning_rate": 4.866959324655501e-05,
547
+ "loss": 0.7604,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 0.21,
552
+ "learning_rate": 4.864013686433552e-05,
553
+ "loss": 0.7515,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 0.21,
558
+ "learning_rate": 4.861036706047566e-05,
559
+ "loss": 0.7508,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 0.22,
564
+ "learning_rate": 4.858028422966439e-05,
565
+ "loss": 0.7559,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 0.22,
570
+ "learning_rate": 4.854988877074076e-05,
571
+ "loss": 0.7684,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 0.22,
576
+ "learning_rate": 4.8519181086688695e-05,
577
+ "loss": 0.7514,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 0.22,
582
+ "learning_rate": 4.8488161584631574e-05,
583
+ "loss": 0.7543,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 0.22,
588
+ "learning_rate": 4.845683067582689e-05,
589
+ "loss": 0.7651,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 0.23,
594
+ "learning_rate": 4.842518877566076e-05,
595
+ "loss": 0.7634,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 0.23,
600
+ "learning_rate": 4.839323630364243e-05,
601
+ "loss": 0.7477,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 0.23,
606
+ "learning_rate": 4.836097368339875e-05,
607
+ "loss": 0.7559,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 0.23,
612
+ "learning_rate": 4.832840134266849e-05,
613
+ "loss": 0.7729,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 0.24,
618
+ "learning_rate": 4.829551971329672e-05,
619
+ "loss": 0.7518,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 0.24,
624
+ "learning_rate": 4.826232923122907e-05,
625
+ "loss": 0.7551,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 0.24,
630
+ "learning_rate": 4.8228830336505926e-05,
631
+ "loss": 0.7613,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 0.24,
636
+ "learning_rate": 4.819502347325665e-05,
637
+ "loss": 0.7529,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 0.25,
642
+ "learning_rate": 4.816090908969363e-05,
643
+ "loss": 0.7476,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 0.25,
648
+ "learning_rate": 4.81264876381064e-05,
649
+ "loss": 0.765,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 0.25,
654
+ "learning_rate": 4.809175957485558e-05,
655
+ "loss": 0.7417,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 0.25,
660
+ "learning_rate": 4.805672536036686e-05,
661
+ "loss": 0.7398,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 0.25,
666
+ "learning_rate": 4.8021385459124925e-05,
667
+ "loss": 0.7489,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 0.26,
672
+ "learning_rate": 4.798574033966723e-05,
673
+ "loss": 0.7546,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 0.26,
678
+ "learning_rate": 4.794979047457783e-05,
679
+ "loss": 0.738,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 0.26,
684
+ "learning_rate": 4.79135363404811e-05,
685
+ "loss": 0.7495,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 0.26,
690
+ "learning_rate": 4.7876978418035455e-05,
691
+ "loss": 0.7622,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 0.27,
696
+ "learning_rate": 4.7840117191926915e-05,
697
+ "loss": 0.7465,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 0.27,
702
+ "learning_rate": 4.780295315086271e-05,
703
+ "loss": 0.7446,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 0.27,
708
+ "learning_rate": 4.7765486787564814e-05,
709
+ "loss": 0.7425,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 0.27,
714
+ "learning_rate": 4.772771859876338e-05,
715
+ "loss": 0.7583,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 0.28,
720
+ "learning_rate": 4.768964908519019e-05,
721
+ "loss": 0.7579,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 0.28,
726
+ "learning_rate": 4.7651278751571984e-05,
727
+ "loss": 0.7509,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 0.28,
732
+ "learning_rate": 4.76126081066238e-05,
733
+ "loss": 0.7358,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 0.28,
738
+ "learning_rate": 4.7573637663042195e-05,
739
+ "loss": 0.7409,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 0.29,
744
+ "learning_rate": 4.753436793749848e-05,
745
+ "loss": 0.7289,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 0.29,
750
+ "learning_rate": 4.749479945063186e-05,
751
+ "loss": 0.7488,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 0.29,
756
+ "learning_rate": 4.745493272704249e-05,
757
+ "loss": 0.7514,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 0.29,
762
+ "learning_rate": 4.7414768295284594e-05,
763
+ "loss": 0.7384,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 0.29,
768
+ "learning_rate": 4.7374306687859405e-05,
769
+ "loss": 0.739,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 0.3,
774
+ "learning_rate": 4.7333548441208134e-05,
775
+ "loss": 0.7494,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 0.3,
780
+ "learning_rate": 4.729249409570481e-05,
781
+ "loss": 0.7503,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 0.3,
786
+ "learning_rate": 4.725114419564919e-05,
787
+ "loss": 0.7388,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 0.3,
792
+ "learning_rate": 4.720949928925948e-05,
793
+ "loss": 0.7601,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 0.31,
798
+ "learning_rate": 4.716755992866509e-05,
799
+ "loss": 0.7416,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 0.31,
804
+ "learning_rate": 4.712532666989931e-05,
805
+ "loss": 0.7365,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 0.31,
810
+ "learning_rate": 4.708280007289195e-05,
811
+ "loss": 0.7455,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 0.31,
816
+ "learning_rate": 4.703998070146191e-05,
817
+ "loss": 0.7342,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 0.32,
822
+ "learning_rate": 4.699686912330969e-05,
823
+ "loss": 0.7374,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 0.32,
828
+ "learning_rate": 4.695346591000986e-05,
829
+ "loss": 0.7312,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 0.32,
834
+ "learning_rate": 4.6909771637003556e-05,
835
+ "loss": 0.7446,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 0.32,
840
+ "learning_rate": 4.686578688359072e-05,
841
+ "loss": 0.7387,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 0.32,
846
+ "learning_rate": 4.6821512232922546e-05,
847
+ "loss": 0.73,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 0.33,
852
+ "learning_rate": 4.677694827199367e-05,
853
+ "loss": 0.7357,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 0.33,
858
+ "learning_rate": 4.673209559163442e-05,
859
+ "loss": 0.7441,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 0.33,
864
+ "learning_rate": 4.668695478650296e-05,
865
+ "loss": 0.7359,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 0.33,
870
+ "learning_rate": 4.664152645507744e-05,
871
+ "loss": 0.7467,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 0.34,
876
+ "learning_rate": 4.659581119964803e-05,
877
+ "loss": 0.7221,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 0.34,
882
+ "learning_rate": 4.654980962630896e-05,
883
+ "loss": 0.7334,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 0.34,
888
+ "learning_rate": 4.6503522344950454e-05,
889
+ "loss": 0.7405,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 0.34,
894
+ "learning_rate": 4.6456949969250674e-05,
895
+ "loss": 0.7424,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 0.35,
900
+ "learning_rate": 4.6410093116667566e-05,
901
+ "loss": 0.7378,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 0.35,
906
+ "learning_rate": 4.636295240843068e-05,
907
+ "loss": 0.7479,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 0.35,
912
+ "learning_rate": 4.6315528469532945e-05,
913
+ "loss": 0.7479,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 0.35,
918
+ "learning_rate": 4.6267821928722345e-05,
919
+ "loss": 0.7337,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 0.35,
924
+ "learning_rate": 4.621983341849362e-05,
925
+ "loss": 0.7403,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 0.36,
930
+ "learning_rate": 4.6171563575079876e-05,
931
+ "loss": 0.7353,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 0.36,
936
+ "learning_rate": 4.612301303844415e-05,
937
+ "loss": 0.7237,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 0.36,
942
+ "learning_rate": 4.607418245227089e-05,
943
+ "loss": 0.739,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 0.36,
948
+ "learning_rate": 4.602507246395747e-05,
949
+ "loss": 0.7292,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 0.37,
954
+ "learning_rate": 4.5975683724605584e-05,
955
+ "loss": 0.745,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 0.37,
960
+ "learning_rate": 4.5926016889012616e-05,
961
+ "loss": 0.7276,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 0.37,
966
+ "learning_rate": 4.587607261566294e-05,
967
+ "loss": 0.7481,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 0.37,
972
+ "learning_rate": 4.582585156671925e-05,
973
+ "loss": 0.7362,
974
+ "step": 1610
975
+ },
976
+ {
977
+ "epoch": 0.38,
978
+ "learning_rate": 4.577535440801368e-05,
979
+ "loss": 0.734,
980
+ "step": 1620
981
+ },
982
+ {
983
+ "epoch": 0.38,
984
+ "learning_rate": 4.572458180903909e-05,
985
+ "loss": 0.7428,
986
+ "step": 1630
987
+ },
988
+ {
989
+ "epoch": 0.38,
990
+ "learning_rate": 4.5673534442940113e-05,
991
+ "loss": 0.732,
992
+ "step": 1640
993
+ },
994
+ {
995
+ "epoch": 0.38,
996
+ "learning_rate": 4.5622212986504245e-05,
997
+ "loss": 0.7191,
998
+ "step": 1650
999
+ },
1000
+ {
1001
+ "epoch": 0.38,
1002
+ "learning_rate": 4.557061812015289e-05,
1003
+ "loss": 0.727,
1004
+ "step": 1660
1005
+ },
1006
+ {
1007
+ "epoch": 0.39,
1008
+ "learning_rate": 4.551875052793232e-05,
1009
+ "loss": 0.728,
1010
+ "step": 1670
1011
+ },
1012
+ {
1013
+ "epoch": 0.39,
1014
+ "learning_rate": 4.546661089750462e-05,
1015
+ "loss": 0.7366,
1016
+ "step": 1680
1017
+ },
1018
+ {
1019
+ "epoch": 0.39,
1020
+ "learning_rate": 4.541419992013859e-05,
1021
+ "loss": 0.7274,
1022
+ "step": 1690
1023
+ },
1024
+ {
1025
+ "epoch": 0.39,
1026
+ "learning_rate": 4.536151829070052e-05,
1027
+ "loss": 0.7185,
1028
+ "step": 1700
1029
+ },
1030
+ {
1031
+ "epoch": 0.4,
1032
+ "learning_rate": 4.530856670764502e-05,
1033
+ "loss": 0.7463,
1034
+ "step": 1710
1035
+ },
1036
+ {
1037
+ "epoch": 0.4,
1038
+ "learning_rate": 4.525534587300575e-05,
1039
+ "loss": 0.734,
1040
+ "step": 1720
1041
+ },
1042
+ {
1043
+ "epoch": 0.4,
1044
+ "learning_rate": 4.520185649238615e-05,
1045
+ "loss": 0.7325,
1046
+ "step": 1730
1047
+ },
1048
+ {
1049
+ "epoch": 0.4,
1050
+ "learning_rate": 4.5148099274950005e-05,
1051
+ "loss": 0.7429,
1052
+ "step": 1740
1053
+ },
1054
+ {
1055
+ "epoch": 0.41,
1056
+ "learning_rate": 4.509407493341212e-05,
1057
+ "loss": 0.7333,
1058
+ "step": 1750
1059
+ },
1060
+ {
1061
+ "epoch": 0.41,
1062
+ "learning_rate": 4.503978418402883e-05,
1063
+ "loss": 0.7227,
1064
+ "step": 1760
1065
+ },
1066
+ {
1067
+ "epoch": 0.41,
1068
+ "learning_rate": 4.49852277465885e-05,
1069
+ "loss": 0.7139,
1070
+ "step": 1770
1071
+ },
1072
+ {
1073
+ "epoch": 0.41,
1074
+ "learning_rate": 4.4930406344402006e-05,
1075
+ "loss": 0.7387,
1076
+ "step": 1780
1077
+ },
1078
+ {
1079
+ "epoch": 0.41,
1080
+ "learning_rate": 4.487532070429315e-05,
1081
+ "loss": 0.7276,
1082
+ "step": 1790
1083
+ },
1084
+ {
1085
+ "epoch": 0.42,
1086
+ "learning_rate": 4.481997155658898e-05,
1087
+ "loss": 0.7251,
1088
+ "step": 1800
1089
+ },
1090
+ {
1091
+ "epoch": 0.42,
1092
+ "learning_rate": 4.476435963511015e-05,
1093
+ "loss": 0.7287,
1094
+ "step": 1810
1095
+ },
1096
+ {
1097
+ "epoch": 0.42,
1098
+ "learning_rate": 4.470848567716119e-05,
1099
+ "loss": 0.7169,
1100
+ "step": 1820
1101
+ },
1102
+ {
1103
+ "epoch": 0.42,
1104
+ "learning_rate": 4.4652350423520685e-05,
1105
+ "loss": 0.725,
1106
+ "step": 1830
1107
+ },
1108
+ {
1109
+ "epoch": 0.43,
1110
+ "learning_rate": 4.45959546184315e-05,
1111
+ "loss": 0.7089,
1112
+ "step": 1840
1113
+ },
1114
+ {
1115
+ "epoch": 0.43,
1116
+ "learning_rate": 4.4539299009590926e-05,
1117
+ "loss": 0.722,
1118
+ "step": 1850
1119
+ },
1120
+ {
1121
+ "epoch": 0.43,
1122
+ "learning_rate": 4.4482384348140695e-05,
1123
+ "loss": 0.7197,
1124
+ "step": 1860
1125
+ },
1126
+ {
1127
+ "epoch": 0.43,
1128
+ "learning_rate": 4.44252113886571e-05,
1129
+ "loss": 0.7135,
1130
+ "step": 1870
1131
+ },
1132
+ {
1133
+ "epoch": 0.44,
1134
+ "learning_rate": 4.4367780889140945e-05,
1135
+ "loss": 0.7168,
1136
+ "step": 1880
1137
+ },
1138
+ {
1139
+ "epoch": 0.44,
1140
+ "learning_rate": 4.431009361100748e-05,
1141
+ "loss": 0.7267,
1142
+ "step": 1890
1143
+ },
1144
+ {
1145
+ "epoch": 0.44,
1146
+ "learning_rate": 4.4252150319076416e-05,
1147
+ "loss": 0.7241,
1148
+ "step": 1900
1149
+ },
1150
+ {
1151
+ "epoch": 0.44,
1152
+ "learning_rate": 4.419395178156159e-05,
1153
+ "loss": 0.7294,
1154
+ "step": 1910
1155
+ },
1156
+ {
1157
+ "epoch": 0.45,
1158
+ "learning_rate": 4.413549877006099e-05,
1159
+ "loss": 0.7392,
1160
+ "step": 1920
1161
+ },
1162
+ {
1163
+ "epoch": 0.45,
1164
+ "learning_rate": 4.4076792059546365e-05,
1165
+ "loss": 0.7058,
1166
+ "step": 1930
1167
+ },
1168
+ {
1169
+ "epoch": 0.45,
1170
+ "learning_rate": 4.401783242835306e-05,
1171
+ "loss": 0.721,
1172
+ "step": 1940
1173
+ },
1174
+ {
1175
+ "epoch": 0.45,
1176
+ "learning_rate": 4.395862065816959e-05,
1177
+ "loss": 0.7284,
1178
+ "step": 1950
1179
+ },
1180
+ {
1181
+ "epoch": 0.45,
1182
+ "learning_rate": 4.3899157534027396e-05,
1183
+ "loss": 0.7259,
1184
+ "step": 1960
1185
+ },
1186
+ {
1187
+ "epoch": 0.46,
1188
+ "learning_rate": 4.383944384429035e-05,
1189
+ "loss": 0.7378,
1190
+ "step": 1970
1191
+ },
1192
+ {
1193
+ "epoch": 0.46,
1194
+ "learning_rate": 4.3779480380644337e-05,
1195
+ "loss": 0.7118,
1196
+ "step": 1980
1197
+ },
1198
+ {
1199
+ "epoch": 0.46,
1200
+ "learning_rate": 4.3719267938086736e-05,
1201
+ "loss": 0.7318,
1202
+ "step": 1990
1203
+ },
1204
+ {
1205
+ "epoch": 0.46,
1206
+ "learning_rate": 4.3658807314915915e-05,
1207
+ "loss": 0.7335,
1208
+ "step": 2000
1209
+ },
1210
+ {
1211
+ "epoch": 0.47,
1212
+ "learning_rate": 4.359809931272062e-05,
1213
+ "loss": 0.7162,
1214
+ "step": 2010
1215
+ },
1216
+ {
1217
+ "epoch": 0.47,
1218
+ "learning_rate": 4.353714473636937e-05,
1219
+ "loss": 0.7279,
1220
+ "step": 2020
1221
+ },
1222
+ {
1223
+ "epoch": 0.47,
1224
+ "learning_rate": 4.3475944393999736e-05,
1225
+ "loss": 0.7344,
1226
+ "step": 2030
1227
+ },
1228
+ {
1229
+ "epoch": 0.47,
1230
+ "learning_rate": 4.34144990970077e-05,
1231
+ "loss": 0.7258,
1232
+ "step": 2040
1233
+ },
1234
+ {
1235
+ "epoch": 0.48,
1236
+ "learning_rate": 4.335280966003684e-05,
1237
+ "loss": 0.7202,
1238
+ "step": 2050
1239
+ },
1240
+ {
1241
+ "epoch": 0.48,
1242
+ "learning_rate": 4.329087690096755e-05,
1243
+ "loss": 0.7225,
1244
+ "step": 2060
1245
+ },
1246
+ {
1247
+ "epoch": 0.48,
1248
+ "learning_rate": 4.3228701640906197e-05,
1249
+ "loss": 0.7073,
1250
+ "step": 2070
1251
+ },
1252
+ {
1253
+ "epoch": 0.48,
1254
+ "learning_rate": 4.3166284704174244e-05,
1255
+ "loss": 0.7265,
1256
+ "step": 2080
1257
+ },
1258
+ {
1259
+ "epoch": 0.48,
1260
+ "learning_rate": 4.310362691829729e-05,
1261
+ "loss": 0.7229,
1262
+ "step": 2090
1263
+ },
1264
+ {
1265
+ "epoch": 0.49,
1266
+ "learning_rate": 4.304072911399415e-05,
1267
+ "loss": 0.6985,
1268
+ "step": 2100
1269
+ },
1270
+ {
1271
+ "epoch": 0.49,
1272
+ "learning_rate": 4.297759212516577e-05,
1273
+ "loss": 0.7316,
1274
+ "step": 2110
1275
+ },
1276
+ {
1277
+ "epoch": 0.49,
1278
+ "learning_rate": 4.291421678888425e-05,
1279
+ "loss": 0.7332,
1280
+ "step": 2120
1281
+ },
1282
+ {
1283
+ "epoch": 0.49,
1284
+ "learning_rate": 4.285060394538167e-05,
1285
+ "loss": 0.7162,
1286
+ "step": 2130
1287
+ },
1288
+ {
1289
+ "epoch": 0.5,
1290
+ "learning_rate": 4.278675443803903e-05,
1291
+ "loss": 0.7064,
1292
+ "step": 2140
1293
+ },
1294
+ {
1295
+ "epoch": 0.5,
1296
+ "learning_rate": 4.2722669113375e-05,
1297
+ "loss": 0.7234,
1298
+ "step": 2150
1299
+ },
1300
+ {
1301
+ "epoch": 0.5,
1302
+ "learning_rate": 4.2658348821034724e-05,
1303
+ "loss": 0.7086,
1304
+ "step": 2160
1305
+ },
1306
+ {
1307
+ "epoch": 0.5,
1308
+ "learning_rate": 4.2593794413778545e-05,
1309
+ "loss": 0.7277,
1310
+ "step": 2170
1311
+ },
1312
+ {
1313
+ "epoch": 0.51,
1314
+ "learning_rate": 4.2529006747470745e-05,
1315
+ "loss": 0.6973,
1316
+ "step": 2180
1317
+ },
1318
+ {
1319
+ "epoch": 0.51,
1320
+ "learning_rate": 4.2463986681068124e-05,
1321
+ "loss": 0.7221,
1322
+ "step": 2190
1323
+ },
1324
+ {
1325
+ "epoch": 0.51,
1326
+ "learning_rate": 4.2398735076608675e-05,
1327
+ "loss": 0.7184,
1328
+ "step": 2200
1329
+ },
1330
+ {
1331
+ "epoch": 0.51,
1332
+ "learning_rate": 4.233325279920011e-05,
1333
+ "loss": 0.7132,
1334
+ "step": 2210
1335
+ },
1336
+ {
1337
+ "epoch": 0.51,
1338
+ "learning_rate": 4.226754071700841e-05,
1339
+ "loss": 0.7208,
1340
+ "step": 2220
1341
+ },
1342
+ {
1343
+ "epoch": 0.52,
1344
+ "learning_rate": 4.220159970124633e-05,
1345
+ "loss": 0.7078,
1346
+ "step": 2230
1347
+ },
1348
+ {
1349
+ "epoch": 0.52,
1350
+ "learning_rate": 4.2135430626161834e-05,
1351
+ "loss": 0.7196,
1352
+ "step": 2240
1353
+ },
1354
+ {
1355
+ "epoch": 0.52,
1356
+ "learning_rate": 4.206903436902649e-05,
1357
+ "loss": 0.7239,
1358
+ "step": 2250
1359
+ },
1360
+ {
1361
+ "epoch": 0.52,
1362
+ "learning_rate": 4.200241181012385e-05,
1363
+ "loss": 0.7172,
1364
+ "step": 2260
1365
+ },
1366
+ {
1367
+ "epoch": 0.53,
1368
+ "learning_rate": 4.19355638327378e-05,
1369
+ "loss": 0.7256,
1370
+ "step": 2270
1371
+ },
1372
+ {
1373
+ "epoch": 0.53,
1374
+ "learning_rate": 4.1868491323140816e-05,
1375
+ "loss": 0.7117,
1376
+ "step": 2280
1377
+ },
1378
+ {
1379
+ "epoch": 0.53,
1380
+ "learning_rate": 4.180119517058222e-05,
1381
+ "loss": 0.7199,
1382
+ "step": 2290
1383
+ },
1384
+ {
1385
+ "epoch": 0.53,
1386
+ "learning_rate": 4.173367626727643e-05,
1387
+ "loss": 0.7023,
1388
+ "step": 2300
1389
+ },
1390
+ {
1391
+ "epoch": 0.54,
1392
+ "learning_rate": 4.166593550839108e-05,
1393
+ "loss": 0.7185,
1394
+ "step": 2310
1395
+ },
1396
+ {
1397
+ "epoch": 0.54,
1398
+ "learning_rate": 4.159797379203516e-05,
1399
+ "loss": 0.7167,
1400
+ "step": 2320
1401
+ },
1402
+ {
1403
+ "epoch": 0.54,
1404
+ "learning_rate": 4.152979201924715e-05,
1405
+ "loss": 0.7194,
1406
+ "step": 2330
1407
+ },
1408
+ {
1409
+ "epoch": 0.54,
1410
+ "learning_rate": 4.146139109398303e-05,
1411
+ "loss": 0.717,
1412
+ "step": 2340
1413
+ },
1414
+ {
1415
+ "epoch": 0.54,
1416
+ "learning_rate": 4.139277192310429e-05,
1417
+ "loss": 0.724,
1418
+ "step": 2350
1419
+ },
1420
+ {
1421
+ "epoch": 0.55,
1422
+ "learning_rate": 4.1323935416365965e-05,
1423
+ "loss": 0.6933,
1424
+ "step": 2360
1425
+ },
1426
+ {
1427
+ "epoch": 0.55,
1428
+ "learning_rate": 4.125488248640451e-05,
1429
+ "loss": 0.7074,
1430
+ "step": 2370
1431
+ },
1432
+ {
1433
+ "epoch": 0.55,
1434
+ "learning_rate": 4.118561404872574e-05,
1435
+ "loss": 0.7028,
1436
+ "step": 2380
1437
+ },
1438
+ {
1439
+ "epoch": 0.55,
1440
+ "learning_rate": 4.111613102169265e-05,
1441
+ "loss": 0.7182,
1442
+ "step": 2390
1443
+ },
1444
+ {
1445
+ "epoch": 0.56,
1446
+ "learning_rate": 4.1046434326513305e-05,
1447
+ "loss": 0.7224,
1448
+ "step": 2400
1449
+ },
1450
+ {
1451
+ "epoch": 0.56,
1452
+ "learning_rate": 4.097652488722856e-05,
1453
+ "loss": 0.7065,
1454
+ "step": 2410
1455
+ },
1456
+ {
1457
+ "epoch": 0.56,
1458
+ "learning_rate": 4.090640363069986e-05,
1459
+ "loss": 0.7074,
1460
+ "step": 2420
1461
+ },
1462
+ {
1463
+ "epoch": 0.56,
1464
+ "learning_rate": 4.08360714865969e-05,
1465
+ "loss": 0.7246,
1466
+ "step": 2430
1467
+ },
1468
+ {
1469
+ "epoch": 0.57,
1470
+ "learning_rate": 4.076552938738537e-05,
1471
+ "loss": 0.7071,
1472
+ "step": 2440
1473
+ },
1474
+ {
1475
+ "epoch": 0.57,
1476
+ "learning_rate": 4.069477826831451e-05,
1477
+ "loss": 0.7082,
1478
+ "step": 2450
1479
+ },
1480
+ {
1481
+ "epoch": 0.57,
1482
+ "learning_rate": 4.062381906740477e-05,
1483
+ "loss": 0.7042,
1484
+ "step": 2460
1485
+ },
1486
+ {
1487
+ "epoch": 0.57,
1488
+ "learning_rate": 4.055265272543536e-05,
1489
+ "loss": 0.7011,
1490
+ "step": 2470
1491
+ },
1492
+ {
1493
+ "epoch": 0.57,
1494
+ "learning_rate": 4.048128018593177e-05,
1495
+ "loss": 0.7155,
1496
+ "step": 2480
1497
+ },
1498
+ {
1499
+ "epoch": 0.58,
1500
+ "learning_rate": 4.040970239515325e-05,
1501
+ "loss": 0.7149,
1502
+ "step": 2490
1503
+ },
1504
+ {
1505
+ "epoch": 0.58,
1506
+ "learning_rate": 4.033792030208029e-05,
1507
+ "loss": 0.717,
1508
+ "step": 2500
1509
+ },
1510
+ {
1511
+ "epoch": 0.58,
1512
+ "learning_rate": 4.026593485840203e-05,
1513
+ "loss": 0.7237,
1514
+ "step": 2510
1515
+ },
1516
+ {
1517
+ "epoch": 0.58,
1518
+ "learning_rate": 4.019374701850362e-05,
1519
+ "loss": 0.7251,
1520
+ "step": 2520
1521
+ },
1522
+ {
1523
+ "epoch": 0.59,
1524
+ "learning_rate": 4.0121357739453604e-05,
1525
+ "loss": 0.7179,
1526
+ "step": 2530
1527
+ },
1528
+ {
1529
+ "epoch": 0.59,
1530
+ "learning_rate": 4.0048767980991206e-05,
1531
+ "loss": 0.7171,
1532
+ "step": 2540
1533
+ },
1534
+ {
1535
+ "epoch": 0.59,
1536
+ "learning_rate": 3.997597870551361e-05,
1537
+ "loss": 0.7124,
1538
+ "step": 2550
1539
+ },
1540
+ {
1541
+ "epoch": 0.59,
1542
+ "learning_rate": 3.990299087806319e-05,
1543
+ "loss": 0.7206,
1544
+ "step": 2560
1545
+ },
1546
+ {
1547
+ "epoch": 0.6,
1548
+ "learning_rate": 3.9829805466314775e-05,
1549
+ "loss": 0.72,
1550
+ "step": 2570
1551
+ },
1552
+ {
1553
+ "epoch": 0.6,
1554
+ "learning_rate": 3.975642344056271e-05,
1555
+ "loss": 0.6972,
1556
+ "step": 2580
1557
+ },
1558
+ {
1559
+ "epoch": 0.6,
1560
+ "learning_rate": 3.9682845773708085e-05,
1561
+ "loss": 0.7158,
1562
+ "step": 2590
1563
+ },
1564
+ {
1565
+ "epoch": 0.6,
1566
+ "learning_rate": 3.960907344124581e-05,
1567
+ "loss": 0.7083,
1568
+ "step": 2600
1569
+ },
1570
+ {
1571
+ "epoch": 0.61,
1572
+ "learning_rate": 3.9535107421251676e-05,
1573
+ "loss": 0.7069,
1574
+ "step": 2610
1575
+ },
1576
+ {
1577
+ "epoch": 0.61,
1578
+ "learning_rate": 3.946094869436938e-05,
1579
+ "loss": 0.7018,
1580
+ "step": 2620
1581
+ },
1582
+ {
1583
+ "epoch": 0.61,
1584
+ "learning_rate": 3.938659824379754e-05,
1585
+ "loss": 0.6973,
1586
+ "step": 2630
1587
+ },
1588
+ {
1589
+ "epoch": 0.61,
1590
+ "learning_rate": 3.931205705527664e-05,
1591
+ "loss": 0.7064,
1592
+ "step": 2640
1593
+ },
1594
+ {
1595
+ "epoch": 0.61,
1596
+ "learning_rate": 3.9237326117075995e-05,
1597
+ "loss": 0.702,
1598
+ "step": 2650
1599
+ },
1600
+ {
1601
+ "epoch": 0.62,
1602
+ "learning_rate": 3.9162406419980605e-05,
1603
+ "loss": 0.713,
1604
+ "step": 2660
1605
+ },
1606
+ {
1607
+ "epoch": 0.62,
1608
+ "learning_rate": 3.908729895727806e-05,
1609
+ "loss": 0.7054,
1610
+ "step": 2670
1611
+ },
1612
+ {
1613
+ "epoch": 0.62,
1614
+ "learning_rate": 3.9012004724745347e-05,
1615
+ "loss": 0.7056,
1616
+ "step": 2680
1617
+ },
1618
+ {
1619
+ "epoch": 0.62,
1620
+ "learning_rate": 3.893652472063562e-05,
1621
+ "loss": 0.7047,
1622
+ "step": 2690
1623
+ },
1624
+ {
1625
+ "epoch": 0.63,
1626
+ "learning_rate": 3.886085994566506e-05,
1627
+ "loss": 0.705,
1628
+ "step": 2700
1629
+ },
1630
+ {
1631
+ "epoch": 0.63,
1632
+ "learning_rate": 3.87850114029995e-05,
1633
+ "loss": 0.6996,
1634
+ "step": 2710
1635
+ },
1636
+ {
1637
+ "epoch": 0.63,
1638
+ "learning_rate": 3.870898009824118e-05,
1639
+ "loss": 0.6986,
1640
+ "step": 2720
1641
+ },
1642
+ {
1643
+ "epoch": 0.63,
1644
+ "learning_rate": 3.863276703941542e-05,
1645
+ "loss": 0.7054,
1646
+ "step": 2730
1647
+ },
1648
+ {
1649
+ "epoch": 0.64,
1650
+ "learning_rate": 3.855637323695721e-05,
1651
+ "loss": 0.7094,
1652
+ "step": 2740
1653
+ },
1654
+ {
1655
+ "epoch": 0.64,
1656
+ "learning_rate": 3.8479799703697884e-05,
1657
+ "loss": 0.7103,
1658
+ "step": 2750
1659
+ },
1660
+ {
1661
+ "epoch": 0.64,
1662
+ "learning_rate": 3.840304745485162e-05,
1663
+ "loss": 0.6921,
1664
+ "step": 2760
1665
+ },
1666
+ {
1667
+ "epoch": 0.64,
1668
+ "learning_rate": 3.8326117508002025e-05,
1669
+ "loss": 0.7177,
1670
+ "step": 2770
1671
+ },
1672
+ {
1673
+ "epoch": 0.64,
1674
+ "learning_rate": 3.8249010883088646e-05,
1675
+ "loss": 0.7163,
1676
+ "step": 2780
1677
+ },
1678
+ {
1679
+ "epoch": 0.65,
1680
+ "learning_rate": 3.817172860239339e-05,
1681
+ "loss": 0.7135,
1682
+ "step": 2790
1683
+ },
1684
+ {
1685
+ "epoch": 0.65,
1686
+ "learning_rate": 3.8094271690527076e-05,
1687
+ "loss": 0.7141,
1688
+ "step": 2800
1689
+ },
1690
+ {
1691
+ "epoch": 0.65,
1692
+ "learning_rate": 3.801664117441572e-05,
1693
+ "loss": 0.704,
1694
+ "step": 2810
1695
+ },
1696
+ {
1697
+ "epoch": 0.65,
1698
+ "learning_rate": 3.7938838083287056e-05,
1699
+ "loss": 0.7092,
1700
+ "step": 2820
1701
+ },
1702
+ {
1703
+ "epoch": 0.66,
1704
+ "learning_rate": 3.7860863448656765e-05,
1705
+ "loss": 0.6986,
1706
+ "step": 2830
1707
+ },
1708
+ {
1709
+ "epoch": 0.66,
1710
+ "learning_rate": 3.778271830431488e-05,
1711
+ "loss": 0.7153,
1712
+ "step": 2840
1713
+ },
1714
+ {
1715
+ "epoch": 0.66,
1716
+ "learning_rate": 3.770440368631207e-05,
1717
+ "loss": 0.7048,
1718
+ "step": 2850
1719
+ },
1720
+ {
1721
+ "epoch": 0.66,
1722
+ "learning_rate": 3.762592063294588e-05,
1723
+ "loss": 0.7103,
1724
+ "step": 2860
1725
+ },
1726
+ {
1727
+ "epoch": 0.67,
1728
+ "learning_rate": 3.7547270184746976e-05,
1729
+ "loss": 0.7048,
1730
+ "step": 2870
1731
+ },
1732
+ {
1733
+ "epoch": 0.67,
1734
+ "learning_rate": 3.746845338446533e-05,
1735
+ "loss": 0.7083,
1736
+ "step": 2880
1737
+ },
1738
+ {
1739
+ "epoch": 0.67,
1740
+ "learning_rate": 3.738947127705644e-05,
1741
+ "loss": 0.7011,
1742
+ "step": 2890
1743
+ },
1744
+ {
1745
+ "epoch": 0.67,
1746
+ "learning_rate": 3.731032490966746e-05,
1747
+ "loss": 0.6959,
1748
+ "step": 2900
1749
+ },
1750
+ {
1751
+ "epoch": 0.67,
1752
+ "learning_rate": 3.723101533162329e-05,
1753
+ "loss": 0.7043,
1754
+ "step": 2910
1755
+ },
1756
+ {
1757
+ "epoch": 0.68,
1758
+ "learning_rate": 3.715154359441265e-05,
1759
+ "loss": 0.7101,
1760
+ "step": 2920
1761
+ },
1762
+ {
1763
+ "epoch": 0.68,
1764
+ "learning_rate": 3.7071910751674256e-05,
1765
+ "loss": 0.7071,
1766
+ "step": 2930
1767
+ },
1768
+ {
1769
+ "epoch": 0.68,
1770
+ "learning_rate": 3.6992117859182674e-05,
1771
+ "loss": 0.7158,
1772
+ "step": 2940
1773
+ },
1774
+ {
1775
+ "epoch": 0.68,
1776
+ "learning_rate": 3.691216597483447e-05,
1777
+ "loss": 0.7046,
1778
+ "step": 2950
1779
+ },
1780
+ {
1781
+ "epoch": 0.69,
1782
+ "learning_rate": 3.683205615863412e-05,
1783
+ "loss": 0.7003,
1784
+ "step": 2960
1785
+ },
1786
+ {
1787
+ "epoch": 0.69,
1788
+ "learning_rate": 3.6751789472679946e-05,
1789
+ "loss": 0.7067,
1790
+ "step": 2970
1791
+ },
1792
+ {
1793
+ "epoch": 0.69,
1794
+ "learning_rate": 3.667136698115008e-05,
1795
+ "loss": 0.7002,
1796
+ "step": 2980
1797
+ },
1798
+ {
1799
+ "epoch": 0.69,
1800
+ "learning_rate": 3.65907897502883e-05,
1801
+ "loss": 0.707,
1802
+ "step": 2990
1803
+ },
1804
+ {
1805
+ "epoch": 0.7,
1806
+ "learning_rate": 3.651005884838995e-05,
1807
+ "loss": 0.7069,
1808
+ "step": 3000
1809
+ },
1810
+ {
1811
+ "epoch": 0.7,
1812
+ "learning_rate": 3.6429175345787735e-05,
1813
+ "loss": 0.7023,
1814
+ "step": 3010
1815
+ },
1816
+ {
1817
+ "epoch": 0.7,
1818
+ "learning_rate": 3.6348140314837553e-05,
1819
+ "loss": 0.6921,
1820
+ "step": 3020
1821
+ },
1822
+ {
1823
+ "epoch": 0.7,
1824
+ "learning_rate": 3.626695482990425e-05,
1825
+ "loss": 0.6896,
1826
+ "step": 3030
1827
+ },
1828
+ {
1829
+ "epoch": 0.7,
1830
+ "learning_rate": 3.618561996734743e-05,
1831
+ "loss": 0.7025,
1832
+ "step": 3040
1833
+ },
1834
+ {
1835
+ "epoch": 0.71,
1836
+ "learning_rate": 3.610413680550711e-05,
1837
+ "loss": 0.7057,
1838
+ "step": 3050
1839
+ },
1840
+ {
1841
+ "epoch": 0.71,
1842
+ "learning_rate": 3.6022506424689485e-05,
1843
+ "loss": 0.6988,
1844
+ "step": 3060
1845
+ },
1846
+ {
1847
+ "epoch": 0.71,
1848
+ "learning_rate": 3.594072990715258e-05,
1849
+ "loss": 0.722,
1850
+ "step": 3070
1851
+ },
1852
+ {
1853
+ "epoch": 0.71,
1854
+ "learning_rate": 3.5858808337091925e-05,
1855
+ "loss": 0.7108,
1856
+ "step": 3080
1857
+ },
1858
+ {
1859
+ "epoch": 0.72,
1860
+ "learning_rate": 3.577674280062612e-05,
1861
+ "loss": 0.7075,
1862
+ "step": 3090
1863
+ },
1864
+ {
1865
+ "epoch": 0.72,
1866
+ "learning_rate": 3.569453438578251e-05,
1867
+ "loss": 0.7048,
1868
+ "step": 3100
1869
+ },
1870
+ {
1871
+ "epoch": 0.72,
1872
+ "learning_rate": 3.561218418248271e-05,
1873
+ "loss": 0.6934,
1874
+ "step": 3110
1875
+ },
1876
+ {
1877
+ "epoch": 0.72,
1878
+ "learning_rate": 3.552969328252817e-05,
1879
+ "loss": 0.7007,
1880
+ "step": 3120
1881
+ },
1882
+ {
1883
+ "epoch": 0.73,
1884
+ "learning_rate": 3.544706277958572e-05,
1885
+ "loss": 0.7221,
1886
+ "step": 3130
1887
+ },
1888
+ {
1889
+ "epoch": 0.73,
1890
+ "learning_rate": 3.5364293769173e-05,
1891
+ "loss": 0.7069,
1892
+ "step": 3140
1893
+ },
1894
+ {
1895
+ "epoch": 0.73,
1896
+ "learning_rate": 3.528138734864406e-05,
1897
+ "loss": 0.7013,
1898
+ "step": 3150
1899
+ },
1900
+ {
1901
+ "epoch": 0.73,
1902
+ "learning_rate": 3.5198344617174685e-05,
1903
+ "loss": 0.7161,
1904
+ "step": 3160
1905
+ },
1906
+ {
1907
+ "epoch": 0.73,
1908
+ "learning_rate": 3.511516667574788e-05,
1909
+ "loss": 0.6995,
1910
+ "step": 3170
1911
+ },
1912
+ {
1913
+ "epoch": 0.74,
1914
+ "learning_rate": 3.5031854627139284e-05,
1915
+ "loss": 0.7157,
1916
+ "step": 3180
1917
+ },
1918
+ {
1919
+ "epoch": 0.74,
1920
+ "learning_rate": 3.4948409575902516e-05,
1921
+ "loss": 0.6929,
1922
+ "step": 3190
1923
+ },
1924
+ {
1925
+ "epoch": 0.74,
1926
+ "learning_rate": 3.486483262835457e-05,
1927
+ "loss": 0.6973,
1928
+ "step": 3200
1929
+ },
1930
+ {
1931
+ "epoch": 0.74,
1932
+ "learning_rate": 3.478112489256111e-05,
1933
+ "loss": 0.6941,
1934
+ "step": 3210
1935
+ },
1936
+ {
1937
+ "epoch": 0.75,
1938
+ "learning_rate": 3.4697287478321784e-05,
1939
+ "loss": 0.7059,
1940
+ "step": 3220
1941
+ },
1942
+ {
1943
+ "epoch": 0.75,
1944
+ "learning_rate": 3.4613321497155546e-05,
1945
+ "loss": 0.6833,
1946
+ "step": 3230
1947
+ },
1948
+ {
1949
+ "epoch": 0.75,
1950
+ "learning_rate": 3.452922806228587e-05,
1951
+ "loss": 0.6766,
1952
+ "step": 3240
1953
+ },
1954
+ {
1955
+ "epoch": 0.75,
1956
+ "learning_rate": 3.444500828862604e-05,
1957
+ "loss": 0.7001,
1958
+ "step": 3250
1959
+ },
1960
+ {
1961
+ "epoch": 0.76,
1962
+ "learning_rate": 3.436066329276432e-05,
1963
+ "loss": 0.7082,
1964
+ "step": 3260
1965
+ },
1966
+ {
1967
+ "epoch": 0.76,
1968
+ "learning_rate": 3.4276194192949185e-05,
1969
+ "loss": 0.7035,
1970
+ "step": 3270
1971
+ },
1972
+ {
1973
+ "epoch": 0.76,
1974
+ "learning_rate": 3.4191602109074484e-05,
1975
+ "loss": 0.6948,
1976
+ "step": 3280
1977
+ },
1978
+ {
1979
+ "epoch": 0.76,
1980
+ "learning_rate": 3.4106888162664584e-05,
1981
+ "loss": 0.7041,
1982
+ "step": 3290
1983
+ },
1984
+ {
1985
+ "epoch": 0.76,
1986
+ "learning_rate": 3.402205347685953e-05,
1987
+ "loss": 0.7004,
1988
+ "step": 3300
1989
+ },
1990
+ {
1991
+ "epoch": 0.77,
1992
+ "learning_rate": 3.39370991764001e-05,
1993
+ "loss": 0.7003,
1994
+ "step": 3310
1995
+ },
1996
+ {
1997
+ "epoch": 0.77,
1998
+ "learning_rate": 3.385202638761297e-05,
1999
+ "loss": 0.6908,
2000
+ "step": 3320
2001
+ },
2002
+ {
2003
+ "epoch": 0.77,
2004
+ "learning_rate": 3.376683623839569e-05,
2005
+ "loss": 0.7059,
2006
+ "step": 3330
2007
+ },
2008
+ {
2009
+ "epoch": 0.77,
2010
+ "learning_rate": 3.3681529858201814e-05,
2011
+ "loss": 0.7027,
2012
+ "step": 3340
2013
+ },
2014
+ {
2015
+ "epoch": 0.78,
2016
+ "learning_rate": 3.3596108378025875e-05,
2017
+ "loss": 0.6984,
2018
+ "step": 3350
2019
+ },
2020
+ {
2021
+ "epoch": 0.78,
2022
+ "learning_rate": 3.351057293038841e-05,
2023
+ "loss": 0.7054,
2024
+ "step": 3360
2025
+ },
2026
+ {
2027
+ "epoch": 0.78,
2028
+ "learning_rate": 3.342492464932094e-05,
2029
+ "loss": 0.6803,
2030
+ "step": 3370
2031
+ },
2032
+ {
2033
+ "epoch": 0.78,
2034
+ "learning_rate": 3.333916467035092e-05,
2035
+ "loss": 0.6924,
2036
+ "step": 3380
2037
+ },
2038
+ {
2039
+ "epoch": 0.79,
2040
+ "learning_rate": 3.325329413048673e-05,
2041
+ "loss": 0.7143,
2042
+ "step": 3390
2043
+ },
2044
+ {
2045
+ "epoch": 0.79,
2046
+ "learning_rate": 3.316731416820254e-05,
2047
+ "loss": 0.6986,
2048
+ "step": 3400
2049
+ },
2050
+ {
2051
+ "epoch": 0.79,
2052
+ "learning_rate": 3.308122592342326e-05,
2053
+ "loss": 0.6875,
2054
+ "step": 3410
2055
+ },
2056
+ {
2057
+ "epoch": 0.79,
2058
+ "learning_rate": 3.2995030537509396e-05,
2059
+ "loss": 0.6934,
2060
+ "step": 3420
2061
+ },
2062
+ {
2063
+ "epoch": 0.8,
2064
+ "learning_rate": 3.290872915324197e-05,
2065
+ "loss": 0.698,
2066
+ "step": 3430
2067
+ },
2068
+ {
2069
+ "epoch": 0.8,
2070
+ "learning_rate": 3.2822322914807305e-05,
2071
+ "loss": 0.6976,
2072
+ "step": 3440
2073
+ },
2074
+ {
2075
+ "epoch": 0.8,
2076
+ "learning_rate": 3.273581296778188e-05,
2077
+ "loss": 0.6914,
2078
+ "step": 3450
2079
+ },
2080
+ {
2081
+ "epoch": 0.8,
2082
+ "learning_rate": 3.264920045911716e-05,
2083
+ "loss": 0.6902,
2084
+ "step": 3460
2085
+ },
2086
+ {
2087
+ "epoch": 0.8,
2088
+ "learning_rate": 3.256248653712437e-05,
2089
+ "loss": 0.6993,
2090
+ "step": 3470
2091
+ },
2092
+ {
2093
+ "epoch": 0.81,
2094
+ "learning_rate": 3.247567235145929e-05,
2095
+ "loss": 0.679,
2096
+ "step": 3480
2097
+ },
2098
+ {
2099
+ "epoch": 0.81,
2100
+ "learning_rate": 3.2388759053106965e-05,
2101
+ "loss": 0.6851,
2102
+ "step": 3490
2103
+ },
2104
+ {
2105
+ "epoch": 0.81,
2106
+ "learning_rate": 3.230174779436651e-05,
2107
+ "loss": 0.6802,
2108
+ "step": 3500
2109
+ },
2110
+ {
2111
+ "epoch": 0.81,
2112
+ "learning_rate": 3.22146397288358e-05,
2113
+ "loss": 0.7058,
2114
+ "step": 3510
2115
+ },
2116
+ {
2117
+ "epoch": 0.82,
2118
+ "learning_rate": 3.2127436011396156e-05,
2119
+ "loss": 0.6926,
2120
+ "step": 3520
2121
+ },
2122
+ {
2123
+ "epoch": 0.82,
2124
+ "learning_rate": 3.2040137798197066e-05,
2125
+ "loss": 0.7031,
2126
+ "step": 3530
2127
+ },
2128
+ {
2129
+ "epoch": 0.82,
2130
+ "learning_rate": 3.195274624664086e-05,
2131
+ "loss": 0.7132,
2132
+ "step": 3540
2133
+ },
2134
+ {
2135
+ "epoch": 0.82,
2136
+ "learning_rate": 3.186526251536733e-05,
2137
+ "loss": 0.6929,
2138
+ "step": 3550
2139
+ },
2140
+ {
2141
+ "epoch": 0.83,
2142
+ "learning_rate": 3.177768776423839e-05,
2143
+ "loss": 0.6959,
2144
+ "step": 3560
2145
+ },
2146
+ {
2147
+ "epoch": 0.83,
2148
+ "learning_rate": 3.1690023154322714e-05,
2149
+ "loss": 0.7035,
2150
+ "step": 3570
2151
+ },
2152
+ {
2153
+ "epoch": 0.83,
2154
+ "learning_rate": 3.1602269847880315e-05,
2155
+ "loss": 0.6965,
2156
+ "step": 3580
2157
+ },
2158
+ {
2159
+ "epoch": 0.83,
2160
+ "learning_rate": 3.1514429008347154e-05,
2161
+ "loss": 0.679,
2162
+ "step": 3590
2163
+ },
2164
+ {
2165
+ "epoch": 0.83,
2166
+ "learning_rate": 3.14265018003197e-05,
2167
+ "loss": 0.6861,
2168
+ "step": 3600
2169
+ },
2170
+ {
2171
+ "epoch": 0.84,
2172
+ "learning_rate": 3.13384893895395e-05,
2173
+ "loss": 0.6964,
2174
+ "step": 3610
2175
+ },
2176
+ {
2177
+ "epoch": 0.84,
2178
+ "learning_rate": 3.125039294287773e-05,
2179
+ "loss": 0.7003,
2180
+ "step": 3620
2181
+ },
2182
+ {
2183
+ "epoch": 0.84,
2184
+ "learning_rate": 3.11622136283197e-05,
2185
+ "loss": 0.6845,
2186
+ "step": 3630
2187
+ },
2188
+ {
2189
+ "epoch": 0.84,
2190
+ "learning_rate": 3.1073952614949406e-05,
2191
+ "loss": 0.6926,
2192
+ "step": 3640
2193
+ },
2194
+ {
2195
+ "epoch": 0.85,
2196
+ "learning_rate": 3.0985611072934e-05,
2197
+ "loss": 0.6953,
2198
+ "step": 3650
2199
+ },
2200
+ {
2201
+ "epoch": 0.85,
2202
+ "learning_rate": 3.089719017350828e-05,
2203
+ "loss": 0.6821,
2204
+ "step": 3660
2205
+ },
2206
+ {
2207
+ "epoch": 0.85,
2208
+ "learning_rate": 3.080869108895915e-05,
2209
+ "loss": 0.6903,
2210
+ "step": 3670
2211
+ },
2212
+ {
2213
+ "epoch": 0.85,
2214
+ "learning_rate": 3.072011499261016e-05,
2215
+ "loss": 0.7159,
2216
+ "step": 3680
2217
+ },
2218
+ {
2219
+ "epoch": 0.86,
2220
+ "learning_rate": 3.0631463058805796e-05,
2221
+ "loss": 0.6961,
2222
+ "step": 3690
2223
+ },
2224
+ {
2225
+ "epoch": 0.86,
2226
+ "learning_rate": 3.0542736462896065e-05,
2227
+ "loss": 0.7061,
2228
+ "step": 3700
2229
+ },
2230
+ {
2231
+ "epoch": 0.86,
2232
+ "learning_rate": 3.045393638122082e-05,
2233
+ "loss": 0.7089,
2234
+ "step": 3710
2235
+ },
2236
+ {
2237
+ "epoch": 0.86,
2238
+ "learning_rate": 3.03650639910942e-05,
2239
+ "loss": 0.672,
2240
+ "step": 3720
2241
+ },
2242
+ {
2243
+ "epoch": 0.86,
2244
+ "learning_rate": 3.0276120470789004e-05,
2245
+ "loss": 0.7006,
2246
+ "step": 3730
2247
+ },
2248
+ {
2249
+ "epoch": 0.87,
2250
+ "learning_rate": 3.0187106999521074e-05,
2251
+ "loss": 0.6945,
2252
+ "step": 3740
2253
+ },
2254
+ {
2255
+ "epoch": 0.87,
2256
+ "learning_rate": 3.0098024757433675e-05,
2257
+ "loss": 0.6957,
2258
+ "step": 3750
2259
+ },
2260
+ {
2261
+ "epoch": 0.87,
2262
+ "learning_rate": 3.000887492558183e-05,
2263
+ "loss": 0.6819,
2264
+ "step": 3760
2265
+ },
2266
+ {
2267
+ "epoch": 0.87,
2268
+ "learning_rate": 2.991965868591668e-05,
2269
+ "loss": 0.6861,
2270
+ "step": 3770
2271
+ },
2272
+ {
2273
+ "epoch": 0.88,
2274
+ "learning_rate": 2.983037722126978e-05,
2275
+ "loss": 0.7015,
2276
+ "step": 3780
2277
+ },
2278
+ {
2279
+ "epoch": 0.88,
2280
+ "learning_rate": 2.9741031715337464e-05,
2281
+ "loss": 0.6852,
2282
+ "step": 3790
2283
+ },
2284
+ {
2285
+ "epoch": 0.88,
2286
+ "learning_rate": 2.9651623352665114e-05,
2287
+ "loss": 0.7009,
2288
+ "step": 3800
2289
+ },
2290
+ {
2291
+ "epoch": 0.88,
2292
+ "learning_rate": 2.9562153318631468e-05,
2293
+ "loss": 0.6885,
2294
+ "step": 3810
2295
+ },
2296
+ {
2297
+ "epoch": 0.89,
2298
+ "learning_rate": 2.947262279943292e-05,
2299
+ "loss": 0.6701,
2300
+ "step": 3820
2301
+ },
2302
+ {
2303
+ "epoch": 0.89,
2304
+ "learning_rate": 2.9383032982067753e-05,
2305
+ "loss": 0.6847,
2306
+ "step": 3830
2307
+ },
2308
+ {
2309
+ "epoch": 0.89,
2310
+ "learning_rate": 2.929338505432045e-05,
2311
+ "loss": 0.6903,
2312
+ "step": 3840
2313
+ },
2314
+ {
2315
+ "epoch": 0.89,
2316
+ "learning_rate": 2.9203680204745914e-05,
2317
+ "loss": 0.6992,
2318
+ "step": 3850
2319
+ },
2320
+ {
2321
+ "epoch": 0.89,
2322
+ "learning_rate": 2.9113919622653707e-05,
2323
+ "loss": 0.7075,
2324
+ "step": 3860
2325
+ },
2326
+ {
2327
+ "epoch": 0.9,
2328
+ "learning_rate": 2.902410449809232e-05,
2329
+ "loss": 0.7072,
2330
+ "step": 3870
2331
+ },
2332
+ {
2333
+ "epoch": 0.9,
2334
+ "learning_rate": 2.8934236021833348e-05,
2335
+ "loss": 0.6897,
2336
+ "step": 3880
2337
+ },
2338
+ {
2339
+ "epoch": 0.9,
2340
+ "learning_rate": 2.8844315385355737e-05,
2341
+ "loss": 0.7081,
2342
+ "step": 3890
2343
+ },
2344
+ {
2345
+ "epoch": 0.9,
2346
+ "learning_rate": 2.8754343780829968e-05,
2347
+ "loss": 0.7032,
2348
+ "step": 3900
2349
+ },
2350
+ {
2351
+ "epoch": 0.91,
2352
+ "learning_rate": 2.866432240110226e-05,
2353
+ "loss": 0.6811,
2354
+ "step": 3910
2355
+ },
2356
+ {
2357
+ "epoch": 0.91,
2358
+ "learning_rate": 2.857425243967875e-05,
2359
+ "loss": 0.6911,
2360
+ "step": 3920
2361
+ },
2362
+ {
2363
+ "epoch": 0.91,
2364
+ "learning_rate": 2.8484135090709686e-05,
2365
+ "loss": 0.6862,
2366
+ "step": 3930
2367
+ },
2368
+ {
2369
+ "epoch": 0.91,
2370
+ "learning_rate": 2.839397154897356e-05,
2371
+ "loss": 0.6785,
2372
+ "step": 3940
2373
+ },
2374
+ {
2375
+ "epoch": 0.92,
2376
+ "learning_rate": 2.83037630098613e-05,
2377
+ "loss": 0.7007,
2378
+ "step": 3950
2379
+ },
2380
+ {
2381
+ "epoch": 0.92,
2382
+ "learning_rate": 2.821351066936043e-05,
2383
+ "loss": 0.675,
2384
+ "step": 3960
2385
+ },
2386
+ {
2387
+ "epoch": 0.92,
2388
+ "learning_rate": 2.812321572403915e-05,
2389
+ "loss": 0.6969,
2390
+ "step": 3970
2391
+ },
2392
+ {
2393
+ "epoch": 0.92,
2394
+ "learning_rate": 2.803287937103055e-05,
2395
+ "loss": 0.694,
2396
+ "step": 3980
2397
+ },
2398
+ {
2399
+ "epoch": 0.92,
2400
+ "learning_rate": 2.7942502808016703e-05,
2401
+ "loss": 0.6951,
2402
+ "step": 3990
2403
+ },
2404
+ {
2405
+ "epoch": 0.93,
2406
+ "learning_rate": 2.7852087233212763e-05,
2407
+ "loss": 0.679,
2408
+ "step": 4000
2409
+ },
2410
+ {
2411
+ "epoch": 0.93,
2412
+ "learning_rate": 2.7761633845351142e-05,
2413
+ "loss": 0.7046,
2414
+ "step": 4010
2415
+ },
2416
+ {
2417
+ "epoch": 0.93,
2418
+ "learning_rate": 2.767114384366555e-05,
2419
+ "loss": 0.6803,
2420
+ "step": 4020
2421
+ },
2422
+ {
2423
+ "epoch": 0.93,
2424
+ "learning_rate": 2.758061842787511e-05,
2425
+ "loss": 0.6927,
2426
+ "step": 4030
2427
+ },
2428
+ {
2429
+ "epoch": 0.94,
2430
+ "learning_rate": 2.7490058798168523e-05,
2431
+ "loss": 0.7024,
2432
+ "step": 4040
2433
+ },
2434
+ {
2435
+ "epoch": 0.94,
2436
+ "learning_rate": 2.7399466155188035e-05,
2437
+ "loss": 0.6839,
2438
+ "step": 4050
2439
+ },
2440
+ {
2441
+ "epoch": 0.94,
2442
+ "learning_rate": 2.7308841700013622e-05,
2443
+ "loss": 0.6846,
2444
+ "step": 4060
2445
+ },
2446
+ {
2447
+ "epoch": 0.94,
2448
+ "learning_rate": 2.721818663414702e-05,
2449
+ "loss": 0.6835,
2450
+ "step": 4070
2451
+ },
2452
+ {
2453
+ "epoch": 0.95,
2454
+ "learning_rate": 2.7127502159495782e-05,
2455
+ "loss": 0.6879,
2456
+ "step": 4080
2457
+ },
2458
+ {
2459
+ "epoch": 0.95,
2460
+ "learning_rate": 2.703678947835741e-05,
2461
+ "loss": 0.6936,
2462
+ "step": 4090
2463
+ },
2464
+ {
2465
+ "epoch": 0.95,
2466
+ "learning_rate": 2.694604979340331e-05,
2467
+ "loss": 0.7074,
2468
+ "step": 4100
2469
+ },
2470
+ {
2471
+ "epoch": 0.95,
2472
+ "learning_rate": 2.6855284307662952e-05,
2473
+ "loss": 0.7018,
2474
+ "step": 4110
2475
+ },
2476
+ {
2477
+ "epoch": 0.96,
2478
+ "learning_rate": 2.6764494224507846e-05,
2479
+ "loss": 0.6998,
2480
+ "step": 4120
2481
+ },
2482
+ {
2483
+ "epoch": 0.96,
2484
+ "learning_rate": 2.6673680747635637e-05,
2485
+ "loss": 0.6834,
2486
+ "step": 4130
2487
+ },
2488
+ {
2489
+ "epoch": 0.96,
2490
+ "learning_rate": 2.6582845081054096e-05,
2491
+ "loss": 0.6906,
2492
+ "step": 4140
2493
+ },
2494
+ {
2495
+ "epoch": 0.96,
2496
+ "learning_rate": 2.6491988429065224e-05,
2497
+ "loss": 0.6902,
2498
+ "step": 4150
2499
+ },
2500
+ {
2501
+ "epoch": 0.96,
2502
+ "learning_rate": 2.64011119962492e-05,
2503
+ "loss": 0.6787,
2504
+ "step": 4160
2505
+ },
2506
+ {
2507
+ "epoch": 0.97,
2508
+ "learning_rate": 2.6310216987448504e-05,
2509
+ "loss": 0.696,
2510
+ "step": 4170
2511
+ },
2512
+ {
2513
+ "epoch": 0.97,
2514
+ "learning_rate": 2.6219304607751865e-05,
2515
+ "loss": 0.6971,
2516
+ "step": 4180
2517
+ },
2518
+ {
2519
+ "epoch": 0.97,
2520
+ "learning_rate": 2.612837606247834e-05,
2521
+ "loss": 0.7099,
2522
+ "step": 4190
2523
+ },
2524
+ {
2525
+ "epoch": 0.97,
2526
+ "learning_rate": 2.6037432557161285e-05,
2527
+ "loss": 0.6855,
2528
+ "step": 4200
2529
+ },
2530
+ {
2531
+ "epoch": 0.98,
2532
+ "learning_rate": 2.5946475297532413e-05,
2533
+ "loss": 0.679,
2534
+ "step": 4210
2535
+ },
2536
+ {
2537
+ "epoch": 0.98,
2538
+ "learning_rate": 2.585550548950579e-05,
2539
+ "loss": 0.6911,
2540
+ "step": 4220
2541
+ },
2542
+ {
2543
+ "epoch": 0.98,
2544
+ "learning_rate": 2.5764524339161837e-05,
2545
+ "loss": 0.6831,
2546
+ "step": 4230
2547
+ },
2548
+ {
2549
+ "epoch": 0.98,
2550
+ "learning_rate": 2.567353305273138e-05,
2551
+ "loss": 0.6788,
2552
+ "step": 4240
2553
+ },
2554
+ {
2555
+ "epoch": 0.99,
2556
+ "learning_rate": 2.5582532836579587e-05,
2557
+ "loss": 0.6832,
2558
+ "step": 4250
2559
+ },
2560
+ {
2561
+ "epoch": 0.99,
2562
+ "learning_rate": 2.5491524897190068e-05,
2563
+ "loss": 0.6939,
2564
+ "step": 4260
2565
+ },
2566
+ {
2567
+ "epoch": 0.99,
2568
+ "learning_rate": 2.540051044114877e-05,
2569
+ "loss": 0.6848,
2570
+ "step": 4270
2571
+ },
2572
+ {
2573
+ "epoch": 0.99,
2574
+ "learning_rate": 2.53094906751281e-05,
2575
+ "loss": 0.6914,
2576
+ "step": 4280
2577
+ },
2578
+ {
2579
+ "epoch": 0.99,
2580
+ "learning_rate": 2.521846680587081e-05,
2581
+ "loss": 0.6926,
2582
+ "step": 4290
2583
+ },
2584
+ {
2585
+ "epoch": 1.0,
2586
+ "learning_rate": 2.5127440040174087e-05,
2587
+ "loss": 0.686,
2588
+ "step": 4300
2589
+ },
2590
+ {
2591
+ "epoch": 1.0,
2592
+ "learning_rate": 2.503641158487351e-05,
2593
+ "loss": 0.6789,
2594
+ "step": 4310
2595
+ },
2596
+ {
2597
+ "epoch": 1.0,
2598
+ "learning_rate": 2.4945382646827058e-05,
2599
+ "loss": 0.6505,
2600
+ "step": 4320
2601
+ },
2602
+ {
2603
+ "epoch": 1.0,
2604
+ "learning_rate": 2.48543544328991e-05,
2605
+ "loss": 0.6762,
2606
+ "step": 4330
2607
+ },
2608
+ {
2609
+ "epoch": 1.01,
2610
+ "learning_rate": 2.476332814994441e-05,
2611
+ "loss": 0.6765,
2612
+ "step": 4340
2613
+ },
2614
+ {
2615
+ "epoch": 1.01,
2616
+ "learning_rate": 2.4672305004792192e-05,
2617
+ "loss": 0.6742,
2618
+ "step": 4350
2619
+ },
2620
+ {
2621
+ "epoch": 1.01,
2622
+ "learning_rate": 2.458128620422999e-05,
2623
+ "loss": 0.6881,
2624
+ "step": 4360
2625
+ },
2626
+ {
2627
+ "epoch": 1.01,
2628
+ "learning_rate": 2.4490272954987796e-05,
2629
+ "loss": 0.6686,
2630
+ "step": 4370
2631
+ },
2632
+ {
2633
+ "epoch": 1.02,
2634
+ "learning_rate": 2.4399266463721993e-05,
2635
+ "loss": 0.6996,
2636
+ "step": 4380
2637
+ },
2638
+ {
2639
+ "epoch": 1.02,
2640
+ "learning_rate": 2.430826793699934e-05,
2641
+ "loss": 0.6819,
2642
+ "step": 4390
2643
+ },
2644
+ {
2645
+ "epoch": 1.02,
2646
+ "learning_rate": 2.4217278581281035e-05,
2647
+ "loss": 0.685,
2648
+ "step": 4400
2649
+ },
2650
+ {
2651
+ "epoch": 1.02,
2652
+ "learning_rate": 2.4126299602906665e-05,
2653
+ "loss": 0.685,
2654
+ "step": 4410
2655
+ },
2656
+ {
2657
+ "epoch": 1.02,
2658
+ "learning_rate": 2.4035332208078252e-05,
2659
+ "loss": 0.7007,
2660
+ "step": 4420
2661
+ },
2662
+ {
2663
+ "epoch": 1.03,
2664
+ "learning_rate": 2.3944377602844226e-05,
2665
+ "loss": 0.6671,
2666
+ "step": 4430
2667
+ },
2668
+ {
2669
+ "epoch": 1.03,
2670
+ "learning_rate": 2.3853436993083453e-05,
2671
+ "loss": 0.6851,
2672
+ "step": 4440
2673
+ },
2674
+ {
2675
+ "epoch": 1.03,
2676
+ "learning_rate": 2.3762511584489256e-05,
2677
+ "loss": 0.6677,
2678
+ "step": 4450
2679
+ },
2680
+ {
2681
+ "epoch": 1.03,
2682
+ "learning_rate": 2.3671602582553436e-05,
2683
+ "loss": 0.6871,
2684
+ "step": 4460
2685
+ },
2686
+ {
2687
+ "epoch": 1.04,
2688
+ "learning_rate": 2.3580711192550216e-05,
2689
+ "loss": 0.6797,
2690
+ "step": 4470
2691
+ },
2692
+ {
2693
+ "epoch": 1.04,
2694
+ "learning_rate": 2.348983861952039e-05,
2695
+ "loss": 0.6835,
2696
+ "step": 4480
2697
+ },
2698
+ {
2699
+ "epoch": 1.04,
2700
+ "learning_rate": 2.3398986068255235e-05,
2701
+ "loss": 0.6762,
2702
+ "step": 4490
2703
+ },
2704
+ {
2705
+ "epoch": 1.04,
2706
+ "learning_rate": 2.3308154743280595e-05,
2707
+ "loss": 0.6751,
2708
+ "step": 4500
2709
+ },
2710
+ {
2711
+ "epoch": 1.05,
2712
+ "learning_rate": 2.3217345848840873e-05,
2713
+ "loss": 0.6891,
2714
+ "step": 4510
2715
+ },
2716
+ {
2717
+ "epoch": 1.05,
2718
+ "learning_rate": 2.3126560588883113e-05,
2719
+ "loss": 0.6641,
2720
+ "step": 4520
2721
+ },
2722
+ {
2723
+ "epoch": 1.05,
2724
+ "learning_rate": 2.3035800167041007e-05,
2725
+ "loss": 0.6741,
2726
+ "step": 4530
2727
+ },
2728
+ {
2729
+ "epoch": 1.05,
2730
+ "learning_rate": 2.2945065786618925e-05,
2731
+ "loss": 0.6742,
2732
+ "step": 4540
2733
+ },
2734
+ {
2735
+ "epoch": 1.05,
2736
+ "learning_rate": 2.2854358650575996e-05,
2737
+ "loss": 0.6875,
2738
+ "step": 4550
2739
+ },
2740
+ {
2741
+ "epoch": 1.06,
2742
+ "learning_rate": 2.2763679961510137e-05,
2743
+ "loss": 0.6836,
2744
+ "step": 4560
2745
+ },
2746
+ {
2747
+ "epoch": 1.06,
2748
+ "learning_rate": 2.267303092164213e-05,
2749
+ "loss": 0.6688,
2750
+ "step": 4570
2751
+ },
2752
+ {
2753
+ "epoch": 1.06,
2754
+ "learning_rate": 2.2582412732799626e-05,
2755
+ "loss": 0.6831,
2756
+ "step": 4580
2757
+ },
2758
+ {
2759
+ "epoch": 1.06,
2760
+ "learning_rate": 2.2491826596401297e-05,
2761
+ "loss": 0.682,
2762
+ "step": 4590
2763
+ },
2764
+ {
2765
+ "epoch": 1.07,
2766
+ "learning_rate": 2.2401273713440837e-05,
2767
+ "loss": 0.6814,
2768
+ "step": 4600
2769
+ },
2770
+ {
2771
+ "epoch": 1.07,
2772
+ "learning_rate": 2.231075528447108e-05,
2773
+ "loss": 0.6753,
2774
+ "step": 4610
2775
+ },
2776
+ {
2777
+ "epoch": 1.07,
2778
+ "learning_rate": 2.222027250958805e-05,
2779
+ "loss": 0.6776,
2780
+ "step": 4620
2781
+ },
2782
+ {
2783
+ "epoch": 1.07,
2784
+ "learning_rate": 2.212982658841508e-05,
2785
+ "loss": 0.6726,
2786
+ "step": 4630
2787
+ },
2788
+ {
2789
+ "epoch": 1.08,
2790
+ "learning_rate": 2.2039418720086887e-05,
2791
+ "loss": 0.6769,
2792
+ "step": 4640
2793
+ },
2794
+ {
2795
+ "epoch": 1.08,
2796
+ "learning_rate": 2.1949050103233715e-05,
2797
+ "loss": 0.6828,
2798
+ "step": 4650
2799
+ },
2800
+ {
2801
+ "epoch": 1.08,
2802
+ "learning_rate": 2.185872193596535e-05,
2803
+ "loss": 0.6581,
2804
+ "step": 4660
2805
+ },
2806
+ {
2807
+ "epoch": 1.08,
2808
+ "learning_rate": 2.176843541585535e-05,
2809
+ "loss": 0.6853,
2810
+ "step": 4670
2811
+ },
2812
+ {
2813
+ "epoch": 1.08,
2814
+ "learning_rate": 2.1678191739925107e-05,
2815
+ "loss": 0.6925,
2816
+ "step": 4680
2817
+ },
2818
+ {
2819
+ "epoch": 1.09,
2820
+ "learning_rate": 2.158799210462793e-05,
2821
+ "loss": 0.6778,
2822
+ "step": 4690
2823
+ },
2824
+ {
2825
+ "epoch": 1.09,
2826
+ "learning_rate": 2.149783770583331e-05,
2827
+ "loss": 0.6912,
2828
+ "step": 4700
2829
+ },
2830
+ {
2831
+ "epoch": 1.09,
2832
+ "learning_rate": 2.140772973881095e-05,
2833
+ "loss": 0.6981,
2834
+ "step": 4710
2835
+ },
2836
+ {
2837
+ "epoch": 1.09,
2838
+ "learning_rate": 2.1317669398214972e-05,
2839
+ "loss": 0.6911,
2840
+ "step": 4720
2841
+ },
2842
+ {
2843
+ "epoch": 1.1,
2844
+ "learning_rate": 2.122765787806805e-05,
2845
+ "loss": 0.6765,
2846
+ "step": 4730
2847
+ },
2848
+ {
2849
+ "epoch": 1.1,
2850
+ "learning_rate": 2.1137696371745608e-05,
2851
+ "loss": 0.6822,
2852
+ "step": 4740
2853
+ },
2854
+ {
2855
+ "epoch": 1.1,
2856
+ "learning_rate": 2.1047786071959983e-05,
2857
+ "loss": 0.6796,
2858
+ "step": 4750
2859
+ },
2860
+ {
2861
+ "epoch": 1.1,
2862
+ "learning_rate": 2.095792817074463e-05,
2863
+ "loss": 0.6671,
2864
+ "step": 4760
2865
+ },
2866
+ {
2867
+ "epoch": 1.11,
2868
+ "learning_rate": 2.0868123859438257e-05,
2869
+ "loss": 0.6893,
2870
+ "step": 4770
2871
+ },
2872
+ {
2873
+ "epoch": 1.11,
2874
+ "learning_rate": 2.0778374328669126e-05,
2875
+ "loss": 0.6746,
2876
+ "step": 4780
2877
+ },
2878
+ {
2879
+ "epoch": 1.11,
2880
+ "learning_rate": 2.0688680768339185e-05,
2881
+ "loss": 0.6851,
2882
+ "step": 4790
2883
+ },
2884
+ {
2885
+ "epoch": 1.11,
2886
+ "learning_rate": 2.059904436760835e-05,
2887
+ "loss": 0.6854,
2888
+ "step": 4800
2889
+ },
2890
+ {
2891
+ "epoch": 1.11,
2892
+ "learning_rate": 2.0509466314878682e-05,
2893
+ "loss": 0.6623,
2894
+ "step": 4810
2895
+ },
2896
+ {
2897
+ "epoch": 1.12,
2898
+ "learning_rate": 2.041994779777868e-05,
2899
+ "loss": 0.6776,
2900
+ "step": 4820
2901
+ },
2902
+ {
2903
+ "epoch": 1.12,
2904
+ "learning_rate": 2.0330490003147528e-05,
2905
+ "loss": 0.6803,
2906
+ "step": 4830
2907
+ },
2908
+ {
2909
+ "epoch": 1.12,
2910
+ "learning_rate": 2.0241094117019328e-05,
2911
+ "loss": 0.6808,
2912
+ "step": 4840
2913
+ },
2914
+ {
2915
+ "epoch": 1.12,
2916
+ "learning_rate": 2.0151761324607405e-05,
2917
+ "loss": 0.6666,
2918
+ "step": 4850
2919
+ },
2920
+ {
2921
+ "epoch": 1.13,
2922
+ "learning_rate": 2.006249281028859e-05,
2923
+ "loss": 0.6707,
2924
+ "step": 4860
2925
+ },
2926
+ {
2927
+ "epoch": 1.13,
2928
+ "learning_rate": 1.997328975758753e-05,
2929
+ "loss": 0.6726,
2930
+ "step": 4870
2931
+ },
2932
+ {
2933
+ "epoch": 1.13,
2934
+ "learning_rate": 1.9884153349160926e-05,
2935
+ "loss": 0.672,
2936
+ "step": 4880
2937
+ },
2938
+ {
2939
+ "epoch": 1.13,
2940
+ "learning_rate": 1.9795084766781975e-05,
2941
+ "loss": 0.7034,
2942
+ "step": 4890
2943
+ },
2944
+ {
2945
+ "epoch": 1.14,
2946
+ "learning_rate": 1.970608519132458e-05,
2947
+ "loss": 0.6758,
2948
+ "step": 4900
2949
+ },
2950
+ {
2951
+ "epoch": 1.14,
2952
+ "learning_rate": 1.9617155802747787e-05,
2953
+ "loss": 0.6741,
2954
+ "step": 4910
2955
+ },
2956
+ {
2957
+ "epoch": 1.14,
2958
+ "learning_rate": 1.9528297780080067e-05,
2959
+ "loss": 0.6684,
2960
+ "step": 4920
2961
+ },
2962
+ {
2963
+ "epoch": 1.14,
2964
+ "learning_rate": 1.9439512301403747e-05,
2965
+ "loss": 0.6743,
2966
+ "step": 4930
2967
+ },
2968
+ {
2969
+ "epoch": 1.15,
2970
+ "learning_rate": 1.9350800543839355e-05,
2971
+ "loss": 0.6759,
2972
+ "step": 4940
2973
+ },
2974
+ {
2975
+ "epoch": 1.15,
2976
+ "learning_rate": 1.926216368353002e-05,
2977
+ "loss": 0.6745,
2978
+ "step": 4950
2979
+ },
2980
+ {
2981
+ "epoch": 1.15,
2982
+ "learning_rate": 1.917360289562588e-05,
2983
+ "loss": 0.6712,
2984
+ "step": 4960
2985
+ },
2986
+ {
2987
+ "epoch": 1.15,
2988
+ "learning_rate": 1.90851193542685e-05,
2989
+ "loss": 0.6762,
2990
+ "step": 4970
2991
+ },
2992
+ {
2993
+ "epoch": 1.15,
2994
+ "learning_rate": 1.899671423257534e-05,
2995
+ "loss": 0.6862,
2996
+ "step": 4980
2997
+ },
2998
+ {
2999
+ "epoch": 1.16,
3000
+ "learning_rate": 1.8908388702624107e-05,
3001
+ "loss": 0.6784,
3002
+ "step": 4990
3003
+ },
3004
+ {
3005
+ "epoch": 1.16,
3006
+ "learning_rate": 1.8820143935437333e-05,
3007
+ "loss": 0.684,
3008
+ "step": 5000
3009
+ }
3010
+ ],
3011
+ "max_steps": 8628,
3012
+ "num_train_epochs": 2,
3013
+ "total_flos": 1.5789628983890412e+18,
3014
+ "trial_name": null,
3015
+ "trial_params": null
3016
+ }
checkpoint-5000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ce444f5cfe8a7851710b073da2f9ede89d9fad5a6d091ebf47aee5e6b0d0797
3
+ size 3259
checkpoint-6000/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: None
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: float16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0