Hugging Face's logo

veekaybee
/
mrigankraman-4090

PEFT
Model card Files Community
veekaybee commited on
Commit
51d5a37
·
1 Parent(s): 5178253

Upload folder using huggingface_hub

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 +21 -0
  2. adapter_config.json +23 -0
  3. adapter_model.bin +3 -0
  4. checkpoint-1040/README.md +21 -0
  5. checkpoint-1040/adapter_config.json +23 -0
  6. checkpoint-1040/adapter_model.bin +3 -0
  7. checkpoint-1040/qwen.tiktoken +0 -0
  8. checkpoint-1040/rng_state.pth +3 -0
  9. checkpoint-1040/scheduler.pt +3 -0
  10. checkpoint-1040/special_tokens_map.json +3 -0
  11. checkpoint-1040/tokenizer_config.json +17 -0
  12. checkpoint-1040/trainer_state.json +643 -0
  13. checkpoint-1040/training_args.bin +3 -0
  14. checkpoint-1300/README.md +21 -0
  15. checkpoint-1300/adapter_config.json +23 -0
  16. checkpoint-1300/adapter_model.bin +3 -0
  17. checkpoint-1300/qwen.tiktoken +0 -0
  18. checkpoint-1300/rng_state.pth +3 -0
  19. checkpoint-1300/scheduler.pt +3 -0
  20. checkpoint-1300/special_tokens_map.json +3 -0
  21. checkpoint-1300/tokenizer_config.json +17 -0
  22. checkpoint-1300/trainer_state.json +799 -0
  23. checkpoint-1300/training_args.bin +3 -0
  24. checkpoint-1560/README.md +21 -0
  25. checkpoint-1560/adapter_config.json +23 -0
  26. checkpoint-1560/adapter_model.bin +3 -0
  27. checkpoint-1560/qwen.tiktoken +0 -0
  28. checkpoint-1560/rng_state.pth +3 -0
  29. checkpoint-1560/scheduler.pt +3 -0
  30. checkpoint-1560/special_tokens_map.json +3 -0
  31. checkpoint-1560/tokenizer_config.json +17 -0
  32. checkpoint-1560/trainer_state.json +955 -0
  33. checkpoint-1560/training_args.bin +3 -0
  34. checkpoint-1820/README.md +21 -0
  35. checkpoint-1820/adapter_config.json +23 -0
  36. checkpoint-1820/adapter_model.bin +3 -0
  37. checkpoint-1820/qwen.tiktoken +0 -0
  38. checkpoint-1820/rng_state.pth +3 -0
  39. checkpoint-1820/scheduler.pt +3 -0
  40. checkpoint-1820/special_tokens_map.json +3 -0
  41. checkpoint-1820/tokenizer_config.json +17 -0
  42. checkpoint-1820/trainer_state.json +1111 -0
  43. checkpoint-1820/training_args.bin +3 -0
  44. checkpoint-2080/README.md +21 -0
  45. checkpoint-2080/adapter_config.json +23 -0
  46. checkpoint-2080/adapter_model.bin +3 -0
  47. checkpoint-2080/qwen.tiktoken +0 -0
  48. checkpoint-2080/rng_state.pth +3 -0
  49. checkpoint-2080/scheduler.pt +3 -0
  50. checkpoint-2080/special_tokens_map.json +3 -0
README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0
adapter_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Qwen/Qwen-14B",
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,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 128,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "w2",
18
+ "c_attn",
19
+ "w1",
20
+ "c_proj"
21
+ ],
22
+ "task_type": "CAUSAL_LM"
23
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf86ce12124899d76768ff0e90a5a19ee21374331435c7600cb2f593a8be2c2f
3
+ size 892737101
checkpoint-1040/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0
checkpoint-1040/adapter_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Qwen/Qwen-14B",
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,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 128,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "w2",
18
+ "c_attn",
19
+ "w1",
20
+ "c_proj"
21
+ ],
22
+ "task_type": "CAUSAL_LM"
23
+ }
checkpoint-1040/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86de7148f6e10e6ea05500c55ace3715400467f4f33bc0feeffd6775398dd096
3
+ size 892742669
checkpoint-1040/qwen.tiktoken ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1040/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9740d07a509b3cc617ecd7a1aa791600189e7e8520f3dec6d612bb7f22183eda
3
+ size 14575
checkpoint-1040/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80c68bdec7235ab7f61454bb154d55f0fa43b8045e9e15c18da3bf6d917308da
3
+ size 627
checkpoint-1040/special_tokens_map.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "pad_token": "<|endoftext|>"
3
+ }
checkpoint-1040/tokenizer_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {},
3
+ "additional_special_tokens": [],
4
+ "auto_map": {
5
+ "AutoTokenizer": [
6
+ "Qwen/Qwen-14B--tokenization_qwen.QWenTokenizer",
7
+ null
8
+ ]
9
+ },
10
+ "clean_up_tokenization_spaces": true,
11
+ "legacy": false,
12
+ "model_max_length": 1600,
13
+ "pad_token": "<|endoftext|>",
14
+ "padding_side": "right",
15
+ "tokenizer_class": "QWenTokenizer",
16
+ "tokenizer_file": null
17
+ }
checkpoint-1040/trainer_state.json ADDED
@@ -0,0 +1,643 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.3367609254498714,
5
+ "eval_steps": 15000,
6
+ "global_step": 1040,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 3.1914893617021277e-06,
14
+ "loss": 0.8544,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.03,
19
+ "learning_rate": 6.3829787234042555e-06,
20
+ "loss": 0.8342,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.04,
25
+ "learning_rate": 9.574468085106385e-06,
26
+ "loss": 0.8016,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.05,
31
+ "learning_rate": 1.2765957446808511e-05,
32
+ "loss": 0.7802,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.06,
37
+ "learning_rate": 1.5957446808510637e-05,
38
+ "loss": 0.7127,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.08,
43
+ "learning_rate": 1.914893617021277e-05,
44
+ "loss": 0.6504,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.09,
49
+ "learning_rate": 2.2340425531914894e-05,
50
+ "loss": 0.6011,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.1,
55
+ "learning_rate": 2.5531914893617022e-05,
56
+ "loss": 0.6859,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.12,
61
+ "learning_rate": 2.872340425531915e-05,
62
+ "loss": 0.5766,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.13,
67
+ "learning_rate": 2.9999474091752356e-05,
68
+ "loss": 0.6617,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.14,
73
+ "learning_rate": 2.9996260341563078e-05,
74
+ "loss": 0.6804,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.15,
79
+ "learning_rate": 2.999012563764485e-05,
80
+ "loss": 0.5635,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.17,
85
+ "learning_rate": 2.9981071174907998e-05,
86
+ "loss": 0.5021,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.18,
91
+ "learning_rate": 2.9969098716969983e-05,
92
+ "loss": 0.575,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.19,
97
+ "learning_rate": 2.995421059581191e-05,
98
+ "loss": 0.6085,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.21,
103
+ "learning_rate": 2.993640971132428e-05,
104
+ "loss": 0.6241,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.22,
109
+ "learning_rate": 2.991569953074218e-05,
110
+ "loss": 0.6405,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.23,
115
+ "learning_rate": 2.9892084087969912e-05,
116
+ "loss": 0.6318,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.24,
121
+ "learning_rate": 2.98655679827953e-05,
122
+ "loss": 0.5592,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.26,
127
+ "learning_rate": 2.9836156379993726e-05,
128
+ "loss": 0.5625,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.27,
133
+ "learning_rate": 2.980385500832215e-05,
134
+ "loss": 0.5857,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.28,
139
+ "learning_rate": 2.9768670159403283e-05,
140
+ "loss": 0.6439,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.3,
145
+ "learning_rate": 2.9730608686500093e-05,
146
+ "loss": 0.4829,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.31,
151
+ "learning_rate": 2.9689678003180932e-05,
152
+ "loss": 0.5554,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.32,
157
+ "learning_rate": 2.964588608187556e-05,
158
+ "loss": 0.5177,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.33,
163
+ "learning_rate": 2.9599241452322243e-05,
164
+ "loss": 0.5375,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.35,
169
+ "learning_rate": 2.9549753199906383e-05,
170
+ "loss": 0.5527,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.36,
175
+ "learning_rate": 2.9497430963890838e-05,
176
+ "loss": 0.5166,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.37,
181
+ "learning_rate": 2.9442284935538424e-05,
182
+ "loss": 0.5353,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.39,
187
+ "learning_rate": 2.9384325856126867e-05,
188
+ "loss": 0.531,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.4,
193
+ "learning_rate": 2.932356501485663e-05,
194
+ "loss": 0.5583,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.41,
199
+ "learning_rate": 2.9260014246652e-05,
200
+ "loss": 0.5355,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.42,
205
+ "learning_rate": 2.919368592985593e-05,
206
+ "loss": 0.5351,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.44,
211
+ "learning_rate": 2.9124592983818963e-05,
212
+ "loss": 0.5349,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.45,
217
+ "learning_rate": 2.9052748866382848e-05,
218
+ "loss": 0.5094,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.46,
223
+ "learning_rate": 2.8978167571259202e-05,
224
+ "loss": 0.538,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.48,
229
+ "learning_rate": 2.8900863625303867e-05,
230
+ "loss": 0.5041,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.49,
235
+ "learning_rate": 2.8820852085687373e-05,
236
+ "loss": 0.4906,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.5,
241
+ "learning_rate": 2.873814853696211e-05,
242
+ "loss": 0.5246,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.51,
247
+ "learning_rate": 2.8652769088026798e-05,
248
+ "loss": 0.5484,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.53,
253
+ "learning_rate": 2.856473036898882e-05,
254
+ "loss": 0.4325,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.54,
259
+ "learning_rate": 2.847404952792501e-05,
260
+ "loss": 0.4965,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.55,
265
+ "learning_rate": 2.8380744227541592e-05,
266
+ "loss": 0.509,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.57,
271
+ "learning_rate": 2.8284832641733857e-05,
272
+ "loss": 0.5052,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.58,
277
+ "learning_rate": 2.8186333452046256e-05,
278
+ "loss": 0.5119,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.59,
283
+ "learning_rate": 2.808526584403364e-05,
284
+ "loss": 0.4206,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.6,
289
+ "learning_rate": 2.7981649503524323e-05,
290
+ "loss": 0.5834,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.62,
295
+ "learning_rate": 2.7875504612785717e-05,
296
+ "loss": 0.4789,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.63,
301
+ "learning_rate": 2.7766851846593215e-05,
302
+ "loss": 0.4196,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.64,
307
+ "learning_rate": 2.7655712368203213e-05,
308
+ "loss": 0.4959,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.66,
313
+ "learning_rate": 2.7542107825230944e-05,
314
+ "loss": 0.4784,
315
+ "step": 510
316
+ },
317
+ {
318
+ "epoch": 0.67,
319
+ "learning_rate": 2.7426060345433984e-05,
320
+ "loss": 0.52,
321
+ "step": 520
322
+ },
323
+ {
324
+ "epoch": 0.68,
325
+ "learning_rate": 2.7307592532402248e-05,
326
+ "loss": 0.517,
327
+ "step": 530
328
+ },
329
+ {
330
+ "epoch": 0.69,
331
+ "learning_rate": 2.7186727461155252e-05,
332
+ "loss": 0.4751,
333
+ "step": 540
334
+ },
335
+ {
336
+ "epoch": 0.71,
337
+ "learning_rate": 2.706348867364763e-05,
338
+ "loss": 0.4711,
339
+ "step": 550
340
+ },
341
+ {
342
+ "epoch": 0.72,
343
+ "learning_rate": 2.6937900174183637e-05,
344
+ "loss": 0.4961,
345
+ "step": 560
346
+ },
347
+ {
348
+ "epoch": 0.73,
349
+ "learning_rate": 2.6809986424741623e-05,
350
+ "loss": 0.5276,
351
+ "step": 570
352
+ },
353
+ {
354
+ "epoch": 0.75,
355
+ "learning_rate": 2.6679772340209362e-05,
356
+ "loss": 0.5341,
357
+ "step": 580
358
+ },
359
+ {
360
+ "epoch": 0.76,
361
+ "learning_rate": 2.6547283283531153e-05,
362
+ "loss": 0.5208,
363
+ "step": 590
364
+ },
365
+ {
366
+ "epoch": 0.77,
367
+ "learning_rate": 2.6412545060767663e-05,
368
+ "loss": 0.4345,
369
+ "step": 600
370
+ },
371
+ {
372
+ "epoch": 0.78,
373
+ "learning_rate": 2.6275583916069456e-05,
374
+ "loss": 0.4814,
375
+ "step": 610
376
+ },
377
+ {
378
+ "epoch": 0.8,
379
+ "learning_rate": 2.6136426526565168e-05,
380
+ "loss": 0.4222,
381
+ "step": 620
382
+ },
383
+ {
384
+ "epoch": 0.81,
385
+ "learning_rate": 2.5995099997165395e-05,
386
+ "loss": 0.5428,
387
+ "step": 630
388
+ },
389
+ {
390
+ "epoch": 0.82,
391
+ "learning_rate": 2.585163185528322e-05,
392
+ "loss": 0.4636,
393
+ "step": 640
394
+ },
395
+ {
396
+ "epoch": 0.84,
397
+ "learning_rate": 2.5706050045472446e-05,
398
+ "loss": 0.4903,
399
+ "step": 650
400
+ },
401
+ {
402
+ "epoch": 0.85,
403
+ "learning_rate": 2.55583829239846e-05,
404
+ "loss": 0.4702,
405
+ "step": 660
406
+ },
407
+ {
408
+ "epoch": 0.86,
409
+ "learning_rate": 2.540865925324574e-05,
410
+ "loss": 0.5231,
411
+ "step": 670
412
+ },
413
+ {
414
+ "epoch": 0.87,
415
+ "learning_rate": 2.5256908196254126e-05,
416
+ "loss": 0.5089,
417
+ "step": 680
418
+ },
419
+ {
420
+ "epoch": 0.89,
421
+ "learning_rate": 2.510315931089991e-05,
422
+ "loss": 0.4572,
423
+ "step": 690
424
+ },
425
+ {
426
+ "epoch": 0.9,
427
+ "learning_rate": 2.494744254420787e-05,
428
+ "loss": 0.471,
429
+ "step": 700
430
+ },
431
+ {
432
+ "epoch": 0.91,
433
+ "learning_rate": 2.478978822650436e-05,
434
+ "loss": 0.3929,
435
+ "step": 710
436
+ },
437
+ {
438
+ "epoch": 0.93,
439
+ "learning_rate": 2.4630227065509627e-05,
440
+ "loss": 0.476,
441
+ "step": 720
442
+ },
443
+ {
444
+ "epoch": 0.94,
445
+ "learning_rate": 2.446879014035656e-05,
446
+ "loss": 0.5348,
447
+ "step": 730
448
+ },
449
+ {
450
+ "epoch": 0.95,
451
+ "learning_rate": 2.4305508895537186e-05,
452
+ "loss": 0.4399,
453
+ "step": 740
454
+ },
455
+ {
456
+ "epoch": 0.96,
457
+ "learning_rate": 2.4140415134777895e-05,
458
+ "loss": 0.4443,
459
+ "step": 750
460
+ },
461
+ {
462
+ "epoch": 0.98,
463
+ "learning_rate": 2.3973541014844783e-05,
464
+ "loss": 0.5046,
465
+ "step": 760
466
+ },
467
+ {
468
+ "epoch": 0.99,
469
+ "learning_rate": 2.380491903928019e-05,
470
+ "loss": 0.4318,
471
+ "step": 770
472
+ },
473
+ {
474
+ "epoch": 1.0,
475
+ "learning_rate": 2.3634582052071682e-05,
476
+ "loss": 0.5244,
477
+ "step": 780
478
+ },
479
+ {
480
+ "epoch": 1.02,
481
+ "learning_rate": 2.3462563231254764e-05,
482
+ "loss": 0.5364,
483
+ "step": 790
484
+ },
485
+ {
486
+ "epoch": 1.03,
487
+ "learning_rate": 2.328889608245048e-05,
488
+ "loss": 0.5278,
489
+ "step": 800
490
+ },
491
+ {
492
+ "epoch": 1.04,
493
+ "learning_rate": 2.3113614432339247e-05,
494
+ "loss": 0.5021,
495
+ "step": 810
496
+ },
497
+ {
498
+ "epoch": 1.05,
499
+ "learning_rate": 2.2936752422072105e-05,
500
+ "loss": 0.4363,
501
+ "step": 820
502
+ },
503
+ {
504
+ "epoch": 1.07,
505
+ "learning_rate": 2.275834450062077e-05,
506
+ "loss": 0.4657,
507
+ "step": 830
508
+ },
509
+ {
510
+ "epoch": 1.08,
511
+ "learning_rate": 2.2578425418067713e-05,
512
+ "loss": 0.4355,
513
+ "step": 840
514
+ },
515
+ {
516
+ "epoch": 1.09,
517
+ "learning_rate": 2.239703021883754e-05,
518
+ "loss": 0.402,
519
+ "step": 850
520
+ },
521
+ {
522
+ "epoch": 1.11,
523
+ "learning_rate": 2.221419423487113e-05,
524
+ "loss": 0.5007,
525
+ "step": 860
526
+ },
527
+ {
528
+ "epoch": 1.12,
529
+ "learning_rate": 2.202995307874367e-05,
530
+ "loss": 0.4388,
531
+ "step": 870
532
+ },
533
+ {
534
+ "epoch": 1.13,
535
+ "learning_rate": 2.1844342636728133e-05,
536
+ "loss": 0.5182,
537
+ "step": 880
538
+ },
539
+ {
540
+ "epoch": 1.14,
541
+ "learning_rate": 2.165739906180535e-05,
542
+ "loss": 0.5118,
543
+ "step": 890
544
+ },
545
+ {
546
+ "epoch": 1.16,
547
+ "learning_rate": 2.1469158766622195e-05,
548
+ "loss": 0.3888,
549
+ "step": 900
550
+ },
551
+ {
552
+ "epoch": 1.17,
553
+ "learning_rate": 2.127965841639918e-05,
554
+ "loss": 0.3978,
555
+ "step": 910
556
+ },
557
+ {
558
+ "epoch": 1.18,
559
+ "learning_rate": 2.1088934921788813e-05,
560
+ "loss": 0.5071,
561
+ "step": 920
562
+ },
563
+ {
564
+ "epoch": 1.2,
565
+ "learning_rate": 2.0897025431686234e-05,
566
+ "loss": 0.4999,
567
+ "step": 930
568
+ },
569
+ {
570
+ "epoch": 1.21,
571
+ "learning_rate": 2.0703967325993373e-05,
572
+ "loss": 0.4734,
573
+ "step": 940
574
+ },
575
+ {
576
+ "epoch": 1.22,
577
+ "learning_rate": 2.0509798208338107e-05,
578
+ "loss": 0.5421,
579
+ "step": 950
580
+ },
581
+ {
582
+ "epoch": 1.23,
583
+ "learning_rate": 2.0314555898749934e-05,
584
+ "loss": 0.4867,
585
+ "step": 960
586
+ },
587
+ {
588
+ "epoch": 1.25,
589
+ "learning_rate": 2.0118278426293376e-05,
590
+ "loss": 0.4662,
591
+ "step": 970
592
+ },
593
+ {
594
+ "epoch": 1.26,
595
+ "learning_rate": 1.9921004021660762e-05,
596
+ "loss": 0.4709,
597
+ "step": 980
598
+ },
599
+ {
600
+ "epoch": 1.27,
601
+ "learning_rate": 1.972277110972567e-05,
602
+ "loss": 0.5178,
603
+ "step": 990
604
+ },
605
+ {
606
+ "epoch": 1.29,
607
+ "learning_rate": 1.952361830205859e-05,
608
+ "loss": 0.5188,
609
+ "step": 1000
610
+ },
611
+ {
612
+ "epoch": 1.3,
613
+ "learning_rate": 1.932358438940622e-05,
614
+ "loss": 0.412,
615
+ "step": 1010
616
+ },
617
+ {
618
+ "epoch": 1.31,
619
+ "learning_rate": 1.9122708334135816e-05,
620
+ "loss": 0.4236,
621
+ "step": 1020
622
+ },
623
+ {
624
+ "epoch": 1.32,
625
+ "learning_rate": 1.8921029262646187e-05,
626
+ "loss": 0.4594,
627
+ "step": 1030
628
+ },
629
+ {
630
+ "epoch": 1.34,
631
+ "learning_rate": 1.8718586457746694e-05,
632
+ "loss": 0.4521,
633
+ "step": 1040
634
+ }
635
+ ],
636
+ "logging_steps": 10,
637
+ "max_steps": 2345,
638
+ "num_train_epochs": 4,
639
+ "save_steps": 260,
640
+ "total_flos": 5.928126824354918e+17,
641
+ "trial_name": null,
642
+ "trial_params": null
643
+ }
checkpoint-1040/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c490f195e94c140704efa846a450585480b78d2abae7738c4ecc5de88d1d30a
3
+ size 4091
checkpoint-1300/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0
checkpoint-1300/adapter_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Qwen/Qwen-14B",
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,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 128,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "w2",
18
+ "c_attn",
19
+ "w1",
20
+ "c_proj"
21
+ ],
22
+ "task_type": "CAUSAL_LM"
23
+ }
checkpoint-1300/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:595947e9038ec4a24de401e1c4bd84124a88898ce30ac4b8314163faf0031480
3
+ size 892742669
checkpoint-1300/qwen.tiktoken ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1300/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:021cef3f1377fe7c397d106b37d7c223ce60d61411e79e9b8e8a5396781adc73
3
+ size 14575
checkpoint-1300/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5406635812e0647e144c494aaaf8c571bd2bf27e7c9daed245a0cc176902b942
3
+ size 627
checkpoint-1300/special_tokens_map.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "pad_token": "<|endoftext|>"
3
+ }
checkpoint-1300/tokenizer_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {},
3
+ "additional_special_tokens": [],
4
+ "auto_map": {
5
+ "AutoTokenizer": [
6
+ "Qwen/Qwen-14B--tokenization_qwen.QWenTokenizer",
7
+ null
8
+ ]
9
+ },
10
+ "clean_up_tokenization_spaces": true,
11
+ "legacy": false,
12
+ "model_max_length": 1600,
13
+ "pad_token": "<|endoftext|>",
14
+ "padding_side": "right",
15
+ "tokenizer_class": "QWenTokenizer",
16
+ "tokenizer_file": null
17
+ }
checkpoint-1300/trainer_state.json ADDED
@@ -0,0 +1,799 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.6709511568123392,
5
+ "eval_steps": 15000,
6
+ "global_step": 1300,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 3.1914893617021277e-06,
14
+ "loss": 0.8544,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.03,
19
+ "learning_rate": 6.3829787234042555e-06,
20
+ "loss": 0.8342,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.04,
25
+ "learning_rate": 9.574468085106385e-06,
26
+ "loss": 0.8016,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.05,
31
+ "learning_rate": 1.2765957446808511e-05,
32
+ "loss": 0.7802,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.06,
37
+ "learning_rate": 1.5957446808510637e-05,
38
+ "loss": 0.7127,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.08,
43
+ "learning_rate": 1.914893617021277e-05,
44
+ "loss": 0.6504,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.09,
49
+ "learning_rate": 2.2340425531914894e-05,
50
+ "loss": 0.6011,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.1,
55
+ "learning_rate": 2.5531914893617022e-05,
56
+ "loss": 0.6859,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.12,
61
+ "learning_rate": 2.872340425531915e-05,
62
+ "loss": 0.5766,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.13,
67
+ "learning_rate": 2.9999474091752356e-05,
68
+ "loss": 0.6617,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.14,
73
+ "learning_rate": 2.9996260341563078e-05,
74
+ "loss": 0.6804,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.15,
79
+ "learning_rate": 2.999012563764485e-05,
80
+ "loss": 0.5635,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.17,
85
+ "learning_rate": 2.9981071174907998e-05,
86
+ "loss": 0.5021,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.18,
91
+ "learning_rate": 2.9969098716969983e-05,
92
+ "loss": 0.575,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.19,
97
+ "learning_rate": 2.995421059581191e-05,
98
+ "loss": 0.6085,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.21,
103
+ "learning_rate": 2.993640971132428e-05,
104
+ "loss": 0.6241,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.22,
109
+ "learning_rate": 2.991569953074218e-05,
110
+ "loss": 0.6405,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.23,
115
+ "learning_rate": 2.9892084087969912e-05,
116
+ "loss": 0.6318,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.24,
121
+ "learning_rate": 2.98655679827953e-05,
122
+ "loss": 0.5592,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.26,
127
+ "learning_rate": 2.9836156379993726e-05,
128
+ "loss": 0.5625,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.27,
133
+ "learning_rate": 2.980385500832215e-05,
134
+ "loss": 0.5857,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.28,
139
+ "learning_rate": 2.9768670159403283e-05,
140
+ "loss": 0.6439,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.3,
145
+ "learning_rate": 2.9730608686500093e-05,
146
+ "loss": 0.4829,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.31,
151
+ "learning_rate": 2.9689678003180932e-05,
152
+ "loss": 0.5554,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.32,
157
+ "learning_rate": 2.964588608187556e-05,
158
+ "loss": 0.5177,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.33,
163
+ "learning_rate": 2.9599241452322243e-05,
164
+ "loss": 0.5375,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.35,
169
+ "learning_rate": 2.9549753199906383e-05,
170
+ "loss": 0.5527,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.36,
175
+ "learning_rate": 2.9497430963890838e-05,
176
+ "loss": 0.5166,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.37,
181
+ "learning_rate": 2.9442284935538424e-05,
182
+ "loss": 0.5353,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.39,
187
+ "learning_rate": 2.9384325856126867e-05,
188
+ "loss": 0.531,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.4,
193
+ "learning_rate": 2.932356501485663e-05,
194
+ "loss": 0.5583,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.41,
199
+ "learning_rate": 2.9260014246652e-05,
200
+ "loss": 0.5355,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.42,
205
+ "learning_rate": 2.919368592985593e-05,
206
+ "loss": 0.5351,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.44,
211
+ "learning_rate": 2.9124592983818963e-05,
212
+ "loss": 0.5349,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.45,
217
+ "learning_rate": 2.9052748866382848e-05,
218
+ "loss": 0.5094,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.46,
223
+ "learning_rate": 2.8978167571259202e-05,
224
+ "loss": 0.538,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.48,
229
+ "learning_rate": 2.8900863625303867e-05,
230
+ "loss": 0.5041,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.49,
235
+ "learning_rate": 2.8820852085687373e-05,
236
+ "loss": 0.4906,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.5,
241
+ "learning_rate": 2.873814853696211e-05,
242
+ "loss": 0.5246,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.51,
247
+ "learning_rate": 2.8652769088026798e-05,
248
+ "loss": 0.5484,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.53,
253
+ "learning_rate": 2.856473036898882e-05,
254
+ "loss": 0.4325,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.54,
259
+ "learning_rate": 2.847404952792501e-05,
260
+ "loss": 0.4965,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.55,
265
+ "learning_rate": 2.8380744227541592e-05,
266
+ "loss": 0.509,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.57,
271
+ "learning_rate": 2.8284832641733857e-05,
272
+ "loss": 0.5052,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.58,
277
+ "learning_rate": 2.8186333452046256e-05,
278
+ "loss": 0.5119,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.59,
283
+ "learning_rate": 2.808526584403364e-05,
284
+ "loss": 0.4206,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.6,
289
+ "learning_rate": 2.7981649503524323e-05,
290
+ "loss": 0.5834,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.62,
295
+ "learning_rate": 2.7875504612785717e-05,
296
+ "loss": 0.4789,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.63,
301
+ "learning_rate": 2.7766851846593215e-05,
302
+ "loss": 0.4196,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.64,
307
+ "learning_rate": 2.7655712368203213e-05,
308
+ "loss": 0.4959,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.66,
313
+ "learning_rate": 2.7542107825230944e-05,
314
+ "loss": 0.4784,
315
+ "step": 510
316
+ },
317
+ {
318
+ "epoch": 0.67,
319
+ "learning_rate": 2.7426060345433984e-05,
320
+ "loss": 0.52,
321
+ "step": 520
322
+ },
323
+ {
324
+ "epoch": 0.68,
325
+ "learning_rate": 2.7307592532402248e-05,
326
+ "loss": 0.517,
327
+ "step": 530
328
+ },
329
+ {
330
+ "epoch": 0.69,
331
+ "learning_rate": 2.7186727461155252e-05,
332
+ "loss": 0.4751,
333
+ "step": 540
334
+ },
335
+ {
336
+ "epoch": 0.71,
337
+ "learning_rate": 2.706348867364763e-05,
338
+ "loss": 0.4711,
339
+ "step": 550
340
+ },
341
+ {
342
+ "epoch": 0.72,
343
+ "learning_rate": 2.6937900174183637e-05,
344
+ "loss": 0.4961,
345
+ "step": 560
346
+ },
347
+ {
348
+ "epoch": 0.73,
349
+ "learning_rate": 2.6809986424741623e-05,
350
+ "loss": 0.5276,
351
+ "step": 570
352
+ },
353
+ {
354
+ "epoch": 0.75,
355
+ "learning_rate": 2.6679772340209362e-05,
356
+ "loss": 0.5341,
357
+ "step": 580
358
+ },
359
+ {
360
+ "epoch": 0.76,
361
+ "learning_rate": 2.6547283283531153e-05,
362
+ "loss": 0.5208,
363
+ "step": 590
364
+ },
365
+ {
366
+ "epoch": 0.77,
367
+ "learning_rate": 2.6412545060767663e-05,
368
+ "loss": 0.4345,
369
+ "step": 600
370
+ },
371
+ {
372
+ "epoch": 0.78,
373
+ "learning_rate": 2.6275583916069456e-05,
374
+ "loss": 0.4814,
375
+ "step": 610
376
+ },
377
+ {
378
+ "epoch": 0.8,
379
+ "learning_rate": 2.6136426526565168e-05,
380
+ "loss": 0.4222,
381
+ "step": 620
382
+ },
383
+ {
384
+ "epoch": 0.81,
385
+ "learning_rate": 2.5995099997165395e-05,
386
+ "loss": 0.5428,
387
+ "step": 630
388
+ },
389
+ {
390
+ "epoch": 0.82,
391
+ "learning_rate": 2.585163185528322e-05,
392
+ "loss": 0.4636,
393
+ "step": 640
394
+ },
395
+ {
396
+ "epoch": 0.84,
397
+ "learning_rate": 2.5706050045472446e-05,
398
+ "loss": 0.4903,
399
+ "step": 650
400
+ },
401
+ {
402
+ "epoch": 0.85,
403
+ "learning_rate": 2.55583829239846e-05,
404
+ "loss": 0.4702,
405
+ "step": 660
406
+ },
407
+ {
408
+ "epoch": 0.86,
409
+ "learning_rate": 2.540865925324574e-05,
410
+ "loss": 0.5231,
411
+ "step": 670
412
+ },
413
+ {
414
+ "epoch": 0.87,
415
+ "learning_rate": 2.5256908196254126e-05,
416
+ "loss": 0.5089,
417
+ "step": 680
418
+ },
419
+ {
420
+ "epoch": 0.89,
421
+ "learning_rate": 2.510315931089991e-05,
422
+ "loss": 0.4572,
423
+ "step": 690
424
+ },
425
+ {
426
+ "epoch": 0.9,
427
+ "learning_rate": 2.494744254420787e-05,
428
+ "loss": 0.471,
429
+ "step": 700
430
+ },
431
+ {
432
+ "epoch": 0.91,
433
+ "learning_rate": 2.478978822650436e-05,
434
+ "loss": 0.3929,
435
+ "step": 710
436
+ },
437
+ {
438
+ "epoch": 0.93,
439
+ "learning_rate": 2.4630227065509627e-05,
440
+ "loss": 0.476,
441
+ "step": 720
442
+ },
443
+ {
444
+ "epoch": 0.94,
445
+ "learning_rate": 2.446879014035656e-05,
446
+ "loss": 0.5348,
447
+ "step": 730
448
+ },
449
+ {
450
+ "epoch": 0.95,
451
+ "learning_rate": 2.4305508895537186e-05,
452
+ "loss": 0.4399,
453
+ "step": 740
454
+ },
455
+ {
456
+ "epoch": 0.96,
457
+ "learning_rate": 2.4140415134777895e-05,
458
+ "loss": 0.4443,
459
+ "step": 750
460
+ },
461
+ {
462
+ "epoch": 0.98,
463
+ "learning_rate": 2.3973541014844783e-05,
464
+ "loss": 0.5046,
465
+ "step": 760
466
+ },
467
+ {
468
+ "epoch": 0.99,
469
+ "learning_rate": 2.380491903928019e-05,
470
+ "loss": 0.4318,
471
+ "step": 770
472
+ },
473
+ {
474
+ "epoch": 1.0,
475
+ "learning_rate": 2.3634582052071682e-05,
476
+ "loss": 0.5244,
477
+ "step": 780
478
+ },
479
+ {
480
+ "epoch": 1.02,
481
+ "learning_rate": 2.3462563231254764e-05,
482
+ "loss": 0.5364,
483
+ "step": 790
484
+ },
485
+ {
486
+ "epoch": 1.03,
487
+ "learning_rate": 2.328889608245048e-05,
488
+ "loss": 0.5278,
489
+ "step": 800
490
+ },
491
+ {
492
+ "epoch": 1.04,
493
+ "learning_rate": 2.3113614432339247e-05,
494
+ "loss": 0.5021,
495
+ "step": 810
496
+ },
497
+ {
498
+ "epoch": 1.05,
499
+ "learning_rate": 2.2936752422072105e-05,
500
+ "loss": 0.4363,
501
+ "step": 820
502
+ },
503
+ {
504
+ "epoch": 1.07,
505
+ "learning_rate": 2.275834450062077e-05,
506
+ "loss": 0.4657,
507
+ "step": 830
508
+ },
509
+ {
510
+ "epoch": 1.08,
511
+ "learning_rate": 2.2578425418067713e-05,
512
+ "loss": 0.4355,
513
+ "step": 840
514
+ },
515
+ {
516
+ "epoch": 1.09,
517
+ "learning_rate": 2.239703021883754e-05,
518
+ "loss": 0.402,
519
+ "step": 850
520
+ },
521
+ {
522
+ "epoch": 1.11,
523
+ "learning_rate": 2.221419423487113e-05,
524
+ "loss": 0.5007,
525
+ "step": 860
526
+ },
527
+ {
528
+ "epoch": 1.12,
529
+ "learning_rate": 2.202995307874367e-05,
530
+ "loss": 0.4388,
531
+ "step": 870
532
+ },
533
+ {
534
+ "epoch": 1.13,
535
+ "learning_rate": 2.1844342636728133e-05,
536
+ "loss": 0.5182,
537
+ "step": 880
538
+ },
539
+ {
540
+ "epoch": 1.14,
541
+ "learning_rate": 2.165739906180535e-05,
542
+ "loss": 0.5118,
543
+ "step": 890
544
+ },
545
+ {
546
+ "epoch": 1.16,
547
+ "learning_rate": 2.1469158766622195e-05,
548
+ "loss": 0.3888,
549
+ "step": 900
550
+ },
551
+ {
552
+ "epoch": 1.17,
553
+ "learning_rate": 2.127965841639918e-05,
554
+ "loss": 0.3978,
555
+ "step": 910
556
+ },
557
+ {
558
+ "epoch": 1.18,
559
+ "learning_rate": 2.1088934921788813e-05,
560
+ "loss": 0.5071,
561
+ "step": 920
562
+ },
563
+ {
564
+ "epoch": 1.2,
565
+ "learning_rate": 2.0897025431686234e-05,
566
+ "loss": 0.4999,
567
+ "step": 930
568
+ },
569
+ {
570
+ "epoch": 1.21,
571
+ "learning_rate": 2.0703967325993373e-05,
572
+ "loss": 0.4734,
573
+ "step": 940
574
+ },
575
+ {
576
+ "epoch": 1.22,
577
+ "learning_rate": 2.0509798208338107e-05,
578
+ "loss": 0.5421,
579
+ "step": 950
580
+ },
581
+ {
582
+ "epoch": 1.23,
583
+ "learning_rate": 2.0314555898749934e-05,
584
+ "loss": 0.4867,
585
+ "step": 960
586
+ },
587
+ {
588
+ "epoch": 1.25,
589
+ "learning_rate": 2.0118278426293376e-05,
590
+ "loss": 0.4662,
591
+ "step": 970
592
+ },
593
+ {
594
+ "epoch": 1.26,
595
+ "learning_rate": 1.9921004021660762e-05,
596
+ "loss": 0.4709,
597
+ "step": 980
598
+ },
599
+ {
600
+ "epoch": 1.27,
601
+ "learning_rate": 1.972277110972567e-05,
602
+ "loss": 0.5178,
603
+ "step": 990
604
+ },
605
+ {
606
+ "epoch": 1.29,
607
+ "learning_rate": 1.952361830205859e-05,
608
+ "loss": 0.5188,
609
+ "step": 1000
610
+ },
611
+ {
612
+ "epoch": 1.3,
613
+ "learning_rate": 1.932358438940622e-05,
614
+ "loss": 0.412,
615
+ "step": 1010
616
+ },
617
+ {
618
+ "epoch": 1.31,
619
+ "learning_rate": 1.9122708334135816e-05,
620
+ "loss": 0.4236,
621
+ "step": 1020
622
+ },
623
+ {
624
+ "epoch": 1.32,
625
+ "learning_rate": 1.8921029262646187e-05,
626
+ "loss": 0.4594,
627
+ "step": 1030
628
+ },
629
+ {
630
+ "epoch": 1.34,
631
+ "learning_rate": 1.8718586457746694e-05,
632
+ "loss": 0.4521,
633
+ "step": 1040
634
+ },
635
+ {
636
+ "epoch": 1.35,
637
+ "learning_rate": 1.8515419351005745e-05,
638
+ "loss": 0.4427,
639
+ "step": 1050
640
+ },
641
+ {
642
+ "epoch": 1.36,
643
+ "learning_rate": 1.8311567515070438e-05,
644
+ "loss": 0.4509,
645
+ "step": 1060
646
+ },
647
+ {
648
+ "epoch": 1.38,
649
+ "learning_rate": 1.8107070655958584e-05,
650
+ "loss": 0.4548,
651
+ "step": 1070
652
+ },
653
+ {
654
+ "epoch": 1.39,
655
+ "learning_rate": 1.7901968605324847e-05,
656
+ "loss": 0.479,
657
+ "step": 1080
658
+ },
659
+ {
660
+ "epoch": 1.4,
661
+ "learning_rate": 1.7696301312702383e-05,
662
+ "loss": 0.4867,
663
+ "step": 1090
664
+ },
665
+ {
666
+ "epoch": 1.41,
667
+ "learning_rate": 1.7490108837721493e-05,
668
+ "loss": 0.4894,
669
+ "step": 1100
670
+ },
671
+ {
672
+ "epoch": 1.43,
673
+ "learning_rate": 1.7283431342306884e-05,
674
+ "loss": 0.4646,
675
+ "step": 1110
676
+ },
677
+ {
678
+ "epoch": 1.44,
679
+ "learning_rate": 1.7076309082854963e-05,
680
+ "loss": 0.4802,
681
+ "step": 1120
682
+ },
683
+ {
684
+ "epoch": 1.45,
685
+ "learning_rate": 1.686878240239274e-05,
686
+ "loss": 0.4287,
687
+ "step": 1130
688
+ },
689
+ {
690
+ "epoch": 1.47,
691
+ "learning_rate": 1.666089172271988e-05,
692
+ "loss": 0.4808,
693
+ "step": 1140
694
+ },
695
+ {
696
+ "epoch": 1.48,
697
+ "learning_rate": 1.6452677536535375e-05,
698
+ "loss": 0.4639,
699
+ "step": 1150
700
+ },
701
+ {
702
+ "epoch": 1.49,
703
+ "learning_rate": 1.624418039955042e-05,
704
+ "loss": 0.4263,
705
+ "step": 1160
706
+ },
707
+ {
708
+ "epoch": 1.5,
709
+ "learning_rate": 1.6035440922589054e-05,
710
+ "loss": 0.492,
711
+ "step": 1170
712
+ },
713
+ {
714
+ "epoch": 1.52,
715
+ "learning_rate": 1.5826499763677982e-05,
716
+ "loss": 0.487,
717
+ "step": 1180
718
+ },
719
+ {
720
+ "epoch": 1.53,
721
+ "learning_rate": 1.561739762012729e-05,
722
+ "loss": 0.3992,
723
+ "step": 1190
724
+ },
725
+ {
726
+ "epoch": 1.54,
727
+ "learning_rate": 1.5408175220603464e-05,
728
+ "loss": 0.4221,
729
+ "step": 1200
730
+ },
731
+ {
732
+ "epoch": 1.56,
733
+ "learning_rate": 1.5198873317196304e-05,
734
+ "loss": 0.4954,
735
+ "step": 1210
736
+ },
737
+ {
738
+ "epoch": 1.57,
739
+ "learning_rate": 1.4989532677481273e-05,
740
+ "loss": 0.4486,
741
+ "step": 1220
742
+ },
743
+ {
744
+ "epoch": 1.58,
745
+ "learning_rate": 1.4780194076578861e-05,
746
+ "loss": 0.4657,
747
+ "step": 1230
748
+ },
749
+ {
750
+ "epoch": 1.59,
751
+ "learning_rate": 1.4570898289212423e-05,
752
+ "loss": 0.3984,
753
+ "step": 1240
754
+ },
755
+ {
756
+ "epoch": 1.61,
757
+ "learning_rate": 1.4361686081766154e-05,
758
+ "loss": 0.512,
759
+ "step": 1250
760
+ },
761
+ {
762
+ "epoch": 1.62,
763
+ "learning_rate": 1.4152598204344656e-05,
764
+ "loss": 0.4178,
765
+ "step": 1260
766
+ },
767
+ {
768
+ "epoch": 1.63,
769
+ "learning_rate": 1.3943675382835672e-05,
770
+ "loss": 0.3971,
771
+ "step": 1270
772
+ },
773
+ {
774
+ "epoch": 1.65,
775
+ "learning_rate": 1.3734958310977542e-05,
776
+ "loss": 0.4343,
777
+ "step": 1280
778
+ },
779
+ {
780
+ "epoch": 1.66,
781
+ "learning_rate": 1.3526487642432953e-05,
782
+ "loss": 0.4675,
783
+ "step": 1290
784
+ },
785
+ {
786
+ "epoch": 1.67,
787
+ "learning_rate": 1.331830398287044e-05,
788
+ "loss": 0.469,
789
+ "step": 1300
790
+ }
791
+ ],
792
+ "logging_steps": 10,
793
+ "max_steps": 2345,
794
+ "num_train_epochs": 4,
795
+ "save_steps": 260,
796
+ "total_flos": 7.430587118411674e+17,
797
+ "trial_name": null,
798
+ "trial_params": null
799
+ }
checkpoint-1300/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c490f195e94c140704efa846a450585480b78d2abae7738c4ecc5de88d1d30a
3
+ size 4091
checkpoint-1560/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0
checkpoint-1560/adapter_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Qwen/Qwen-14B",
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,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 128,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "w2",
18
+ "c_attn",
19
+ "w1",
20
+ "c_proj"
21
+ ],
22
+ "task_type": "CAUSAL_LM"
23
+ }
checkpoint-1560/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9643cbd07a0fba8c579f70a75346cc208a479bb22a48a78a4cb2b32f83c4ce8d
3
+ size 892742669
checkpoint-1560/qwen.tiktoken ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1560/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d940465d95f218f9242372a21f394b18864a580055e91878ae4605f5f77b9ce2
3
+ size 14575
checkpoint-1560/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:606afc16717a1d21ca82fdab9d1899f58de2fc94fde1c87491e6e52d3a2eb873
3
+ size 627
checkpoint-1560/special_tokens_map.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "pad_token": "<|endoftext|>"
3
+ }
checkpoint-1560/tokenizer_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {},
3
+ "additional_special_tokens": [],
4
+ "auto_map": {
5
+ "AutoTokenizer": [
6
+ "Qwen/Qwen-14B--tokenization_qwen.QWenTokenizer",
7
+ null
8
+ ]
9
+ },
10
+ "clean_up_tokenization_spaces": true,
11
+ "legacy": false,
12
+ "model_max_length": 1600,
13
+ "pad_token": "<|endoftext|>",
14
+ "padding_side": "right",
15
+ "tokenizer_class": "QWenTokenizer",
16
+ "tokenizer_file": null
17
+ }
checkpoint-1560/trainer_state.json ADDED
@@ -0,0 +1,955 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.005141388174807,
5
+ "eval_steps": 15000,
6
+ "global_step": 1560,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 3.1914893617021277e-06,
14
+ "loss": 0.8544,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.03,
19
+ "learning_rate": 6.3829787234042555e-06,
20
+ "loss": 0.8342,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.04,
25
+ "learning_rate": 9.574468085106385e-06,
26
+ "loss": 0.8016,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.05,
31
+ "learning_rate": 1.2765957446808511e-05,
32
+ "loss": 0.7802,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.06,
37
+ "learning_rate": 1.5957446808510637e-05,
38
+ "loss": 0.7127,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.08,
43
+ "learning_rate": 1.914893617021277e-05,
44
+ "loss": 0.6504,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.09,
49
+ "learning_rate": 2.2340425531914894e-05,
50
+ "loss": 0.6011,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.1,
55
+ "learning_rate": 2.5531914893617022e-05,
56
+ "loss": 0.6859,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.12,
61
+ "learning_rate": 2.872340425531915e-05,
62
+ "loss": 0.5766,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.13,
67
+ "learning_rate": 2.9999474091752356e-05,
68
+ "loss": 0.6617,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.14,
73
+ "learning_rate": 2.9996260341563078e-05,
74
+ "loss": 0.6804,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.15,
79
+ "learning_rate": 2.999012563764485e-05,
80
+ "loss": 0.5635,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.17,
85
+ "learning_rate": 2.9981071174907998e-05,
86
+ "loss": 0.5021,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.18,
91
+ "learning_rate": 2.9969098716969983e-05,
92
+ "loss": 0.575,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.19,
97
+ "learning_rate": 2.995421059581191e-05,
98
+ "loss": 0.6085,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.21,
103
+ "learning_rate": 2.993640971132428e-05,
104
+ "loss": 0.6241,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.22,
109
+ "learning_rate": 2.991569953074218e-05,
110
+ "loss": 0.6405,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.23,
115
+ "learning_rate": 2.9892084087969912e-05,
116
+ "loss": 0.6318,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.24,
121
+ "learning_rate": 2.98655679827953e-05,
122
+ "loss": 0.5592,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.26,
127
+ "learning_rate": 2.9836156379993726e-05,
128
+ "loss": 0.5625,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.27,
133
+ "learning_rate": 2.980385500832215e-05,
134
+ "loss": 0.5857,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.28,
139
+ "learning_rate": 2.9768670159403283e-05,
140
+ "loss": 0.6439,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.3,
145
+ "learning_rate": 2.9730608686500093e-05,
146
+ "loss": 0.4829,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.31,
151
+ "learning_rate": 2.9689678003180932e-05,
152
+ "loss": 0.5554,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.32,
157
+ "learning_rate": 2.964588608187556e-05,
158
+ "loss": 0.5177,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.33,
163
+ "learning_rate": 2.9599241452322243e-05,
164
+ "loss": 0.5375,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.35,
169
+ "learning_rate": 2.9549753199906383e-05,
170
+ "loss": 0.5527,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.36,
175
+ "learning_rate": 2.9497430963890838e-05,
176
+ "loss": 0.5166,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.37,
181
+ "learning_rate": 2.9442284935538424e-05,
182
+ "loss": 0.5353,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.39,
187
+ "learning_rate": 2.9384325856126867e-05,
188
+ "loss": 0.531,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.4,
193
+ "learning_rate": 2.932356501485663e-05,
194
+ "loss": 0.5583,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.41,
199
+ "learning_rate": 2.9260014246652e-05,
200
+ "loss": 0.5355,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.42,
205
+ "learning_rate": 2.919368592985593e-05,
206
+ "loss": 0.5351,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.44,
211
+ "learning_rate": 2.9124592983818963e-05,
212
+ "loss": 0.5349,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.45,
217
+ "learning_rate": 2.9052748866382848e-05,
218
+ "loss": 0.5094,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.46,
223
+ "learning_rate": 2.8978167571259202e-05,
224
+ "loss": 0.538,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.48,
229
+ "learning_rate": 2.8900863625303867e-05,
230
+ "loss": 0.5041,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.49,
235
+ "learning_rate": 2.8820852085687373e-05,
236
+ "loss": 0.4906,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.5,
241
+ "learning_rate": 2.873814853696211e-05,
242
+ "loss": 0.5246,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.51,
247
+ "learning_rate": 2.8652769088026798e-05,
248
+ "loss": 0.5484,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.53,
253
+ "learning_rate": 2.856473036898882e-05,
254
+ "loss": 0.4325,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.54,
259
+ "learning_rate": 2.847404952792501e-05,
260
+ "loss": 0.4965,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.55,
265
+ "learning_rate": 2.8380744227541592e-05,
266
+ "loss": 0.509,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.57,
271
+ "learning_rate": 2.8284832641733857e-05,
272
+ "loss": 0.5052,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.58,
277
+ "learning_rate": 2.8186333452046256e-05,
278
+ "loss": 0.5119,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.59,
283
+ "learning_rate": 2.808526584403364e-05,
284
+ "loss": 0.4206,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.6,
289
+ "learning_rate": 2.7981649503524323e-05,
290
+ "loss": 0.5834,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.62,
295
+ "learning_rate": 2.7875504612785717e-05,
296
+ "loss": 0.4789,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.63,
301
+ "learning_rate": 2.7766851846593215e-05,
302
+ "loss": 0.4196,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.64,
307
+ "learning_rate": 2.7655712368203213e-05,
308
+ "loss": 0.4959,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.66,
313
+ "learning_rate": 2.7542107825230944e-05,
314
+ "loss": 0.4784,
315
+ "step": 510
316
+ },
317
+ {
318
+ "epoch": 0.67,
319
+ "learning_rate": 2.7426060345433984e-05,
320
+ "loss": 0.52,
321
+ "step": 520
322
+ },
323
+ {
324
+ "epoch": 0.68,
325
+ "learning_rate": 2.7307592532402248e-05,
326
+ "loss": 0.517,
327
+ "step": 530
328
+ },
329
+ {
330
+ "epoch": 0.69,
331
+ "learning_rate": 2.7186727461155252e-05,
332
+ "loss": 0.4751,
333
+ "step": 540
334
+ },
335
+ {
336
+ "epoch": 0.71,
337
+ "learning_rate": 2.706348867364763e-05,
338
+ "loss": 0.4711,
339
+ "step": 550
340
+ },
341
+ {
342
+ "epoch": 0.72,
343
+ "learning_rate": 2.6937900174183637e-05,
344
+ "loss": 0.4961,
345
+ "step": 560
346
+ },
347
+ {
348
+ "epoch": 0.73,
349
+ "learning_rate": 2.6809986424741623e-05,
350
+ "loss": 0.5276,
351
+ "step": 570
352
+ },
353
+ {
354
+ "epoch": 0.75,
355
+ "learning_rate": 2.6679772340209362e-05,
356
+ "loss": 0.5341,
357
+ "step": 580
358
+ },
359
+ {
360
+ "epoch": 0.76,
361
+ "learning_rate": 2.6547283283531153e-05,
362
+ "loss": 0.5208,
363
+ "step": 590
364
+ },
365
+ {
366
+ "epoch": 0.77,
367
+ "learning_rate": 2.6412545060767663e-05,
368
+ "loss": 0.4345,
369
+ "step": 600
370
+ },
371
+ {
372
+ "epoch": 0.78,
373
+ "learning_rate": 2.6275583916069456e-05,
374
+ "loss": 0.4814,
375
+ "step": 610
376
+ },
377
+ {
378
+ "epoch": 0.8,
379
+ "learning_rate": 2.6136426526565168e-05,
380
+ "loss": 0.4222,
381
+ "step": 620
382
+ },
383
+ {
384
+ "epoch": 0.81,
385
+ "learning_rate": 2.5995099997165395e-05,
386
+ "loss": 0.5428,
387
+ "step": 630
388
+ },
389
+ {
390
+ "epoch": 0.82,
391
+ "learning_rate": 2.585163185528322e-05,
392
+ "loss": 0.4636,
393
+ "step": 640
394
+ },
395
+ {
396
+ "epoch": 0.84,
397
+ "learning_rate": 2.5706050045472446e-05,
398
+ "loss": 0.4903,
399
+ "step": 650
400
+ },
401
+ {
402
+ "epoch": 0.85,
403
+ "learning_rate": 2.55583829239846e-05,
404
+ "loss": 0.4702,
405
+ "step": 660
406
+ },
407
+ {
408
+ "epoch": 0.86,
409
+ "learning_rate": 2.540865925324574e-05,
410
+ "loss": 0.5231,
411
+ "step": 670
412
+ },
413
+ {
414
+ "epoch": 0.87,
415
+ "learning_rate": 2.5256908196254126e-05,
416
+ "loss": 0.5089,
417
+ "step": 680
418
+ },
419
+ {
420
+ "epoch": 0.89,
421
+ "learning_rate": 2.510315931089991e-05,
422
+ "loss": 0.4572,
423
+ "step": 690
424
+ },
425
+ {
426
+ "epoch": 0.9,
427
+ "learning_rate": 2.494744254420787e-05,
428
+ "loss": 0.471,
429
+ "step": 700
430
+ },
431
+ {
432
+ "epoch": 0.91,
433
+ "learning_rate": 2.478978822650436e-05,
434
+ "loss": 0.3929,
435
+ "step": 710
436
+ },
437
+ {
438
+ "epoch": 0.93,
439
+ "learning_rate": 2.4630227065509627e-05,
440
+ "loss": 0.476,
441
+ "step": 720
442
+ },
443
+ {
444
+ "epoch": 0.94,
445
+ "learning_rate": 2.446879014035656e-05,
446
+ "loss": 0.5348,
447
+ "step": 730
448
+ },
449
+ {
450
+ "epoch": 0.95,
451
+ "learning_rate": 2.4305508895537186e-05,
452
+ "loss": 0.4399,
453
+ "step": 740
454
+ },
455
+ {
456
+ "epoch": 0.96,
457
+ "learning_rate": 2.4140415134777895e-05,
458
+ "loss": 0.4443,
459
+ "step": 750
460
+ },
461
+ {
462
+ "epoch": 0.98,
463
+ "learning_rate": 2.3973541014844783e-05,
464
+ "loss": 0.5046,
465
+ "step": 760
466
+ },
467
+ {
468
+ "epoch": 0.99,
469
+ "learning_rate": 2.380491903928019e-05,
470
+ "loss": 0.4318,
471
+ "step": 770
472
+ },
473
+ {
474
+ "epoch": 1.0,
475
+ "learning_rate": 2.3634582052071682e-05,
476
+ "loss": 0.5244,
477
+ "step": 780
478
+ },
479
+ {
480
+ "epoch": 1.02,
481
+ "learning_rate": 2.3462563231254764e-05,
482
+ "loss": 0.5364,
483
+ "step": 790
484
+ },
485
+ {
486
+ "epoch": 1.03,
487
+ "learning_rate": 2.328889608245048e-05,
488
+ "loss": 0.5278,
489
+ "step": 800
490
+ },
491
+ {
492
+ "epoch": 1.04,
493
+ "learning_rate": 2.3113614432339247e-05,
494
+ "loss": 0.5021,
495
+ "step": 810
496
+ },
497
+ {
498
+ "epoch": 1.05,
499
+ "learning_rate": 2.2936752422072105e-05,
500
+ "loss": 0.4363,
501
+ "step": 820
502
+ },
503
+ {
504
+ "epoch": 1.07,
505
+ "learning_rate": 2.275834450062077e-05,
506
+ "loss": 0.4657,
507
+ "step": 830
508
+ },
509
+ {
510
+ "epoch": 1.08,
511
+ "learning_rate": 2.2578425418067713e-05,
512
+ "loss": 0.4355,
513
+ "step": 840
514
+ },
515
+ {
516
+ "epoch": 1.09,
517
+ "learning_rate": 2.239703021883754e-05,
518
+ "loss": 0.402,
519
+ "step": 850
520
+ },
521
+ {
522
+ "epoch": 1.11,
523
+ "learning_rate": 2.221419423487113e-05,
524
+ "loss": 0.5007,
525
+ "step": 860
526
+ },
527
+ {
528
+ "epoch": 1.12,
529
+ "learning_rate": 2.202995307874367e-05,
530
+ "loss": 0.4388,
531
+ "step": 870
532
+ },
533
+ {
534
+ "epoch": 1.13,
535
+ "learning_rate": 2.1844342636728133e-05,
536
+ "loss": 0.5182,
537
+ "step": 880
538
+ },
539
+ {
540
+ "epoch": 1.14,
541
+ "learning_rate": 2.165739906180535e-05,
542
+ "loss": 0.5118,
543
+ "step": 890
544
+ },
545
+ {
546
+ "epoch": 1.16,
547
+ "learning_rate": 2.1469158766622195e-05,
548
+ "loss": 0.3888,
549
+ "step": 900
550
+ },
551
+ {
552
+ "epoch": 1.17,
553
+ "learning_rate": 2.127965841639918e-05,
554
+ "loss": 0.3978,
555
+ "step": 910
556
+ },
557
+ {
558
+ "epoch": 1.18,
559
+ "learning_rate": 2.1088934921788813e-05,
560
+ "loss": 0.5071,
561
+ "step": 920
562
+ },
563
+ {
564
+ "epoch": 1.2,
565
+ "learning_rate": 2.0897025431686234e-05,
566
+ "loss": 0.4999,
567
+ "step": 930
568
+ },
569
+ {
570
+ "epoch": 1.21,
571
+ "learning_rate": 2.0703967325993373e-05,
572
+ "loss": 0.4734,
573
+ "step": 940
574
+ },
575
+ {
576
+ "epoch": 1.22,
577
+ "learning_rate": 2.0509798208338107e-05,
578
+ "loss": 0.5421,
579
+ "step": 950
580
+ },
581
+ {
582
+ "epoch": 1.23,
583
+ "learning_rate": 2.0314555898749934e-05,
584
+ "loss": 0.4867,
585
+ "step": 960
586
+ },
587
+ {
588
+ "epoch": 1.25,
589
+ "learning_rate": 2.0118278426293376e-05,
590
+ "loss": 0.4662,
591
+ "step": 970
592
+ },
593
+ {
594
+ "epoch": 1.26,
595
+ "learning_rate": 1.9921004021660762e-05,
596
+ "loss": 0.4709,
597
+ "step": 980
598
+ },
599
+ {
600
+ "epoch": 1.27,
601
+ "learning_rate": 1.972277110972567e-05,
602
+ "loss": 0.5178,
603
+ "step": 990
604
+ },
605
+ {
606
+ "epoch": 1.29,
607
+ "learning_rate": 1.952361830205859e-05,
608
+ "loss": 0.5188,
609
+ "step": 1000
610
+ },
611
+ {
612
+ "epoch": 1.3,
613
+ "learning_rate": 1.932358438940622e-05,
614
+ "loss": 0.412,
615
+ "step": 1010
616
+ },
617
+ {
618
+ "epoch": 1.31,
619
+ "learning_rate": 1.9122708334135816e-05,
620
+ "loss": 0.4236,
621
+ "step": 1020
622
+ },
623
+ {
624
+ "epoch": 1.32,
625
+ "learning_rate": 1.8921029262646187e-05,
626
+ "loss": 0.4594,
627
+ "step": 1030
628
+ },
629
+ {
630
+ "epoch": 1.34,
631
+ "learning_rate": 1.8718586457746694e-05,
632
+ "loss": 0.4521,
633
+ "step": 1040
634
+ },
635
+ {
636
+ "epoch": 1.35,
637
+ "learning_rate": 1.8515419351005745e-05,
638
+ "loss": 0.4427,
639
+ "step": 1050
640
+ },
641
+ {
642
+ "epoch": 1.36,
643
+ "learning_rate": 1.8311567515070438e-05,
644
+ "loss": 0.4509,
645
+ "step": 1060
646
+ },
647
+ {
648
+ "epoch": 1.38,
649
+ "learning_rate": 1.8107070655958584e-05,
650
+ "loss": 0.4548,
651
+ "step": 1070
652
+ },
653
+ {
654
+ "epoch": 1.39,
655
+ "learning_rate": 1.7901968605324847e-05,
656
+ "loss": 0.479,
657
+ "step": 1080
658
+ },
659
+ {
660
+ "epoch": 1.4,
661
+ "learning_rate": 1.7696301312702383e-05,
662
+ "loss": 0.4867,
663
+ "step": 1090
664
+ },
665
+ {
666
+ "epoch": 1.41,
667
+ "learning_rate": 1.7490108837721493e-05,
668
+ "loss": 0.4894,
669
+ "step": 1100
670
+ },
671
+ {
672
+ "epoch": 1.43,
673
+ "learning_rate": 1.7283431342306884e-05,
674
+ "loss": 0.4646,
675
+ "step": 1110
676
+ },
677
+ {
678
+ "epoch": 1.44,
679
+ "learning_rate": 1.7076309082854963e-05,
680
+ "loss": 0.4802,
681
+ "step": 1120
682
+ },
683
+ {
684
+ "epoch": 1.45,
685
+ "learning_rate": 1.686878240239274e-05,
686
+ "loss": 0.4287,
687
+ "step": 1130
688
+ },
689
+ {
690
+ "epoch": 1.47,
691
+ "learning_rate": 1.666089172271988e-05,
692
+ "loss": 0.4808,
693
+ "step": 1140
694
+ },
695
+ {
696
+ "epoch": 1.48,
697
+ "learning_rate": 1.6452677536535375e-05,
698
+ "loss": 0.4639,
699
+ "step": 1150
700
+ },
701
+ {
702
+ "epoch": 1.49,
703
+ "learning_rate": 1.624418039955042e-05,
704
+ "loss": 0.4263,
705
+ "step": 1160
706
+ },
707
+ {
708
+ "epoch": 1.5,
709
+ "learning_rate": 1.6035440922589054e-05,
710
+ "loss": 0.492,
711
+ "step": 1170
712
+ },
713
+ {
714
+ "epoch": 1.52,
715
+ "learning_rate": 1.5826499763677982e-05,
716
+ "loss": 0.487,
717
+ "step": 1180
718
+ },
719
+ {
720
+ "epoch": 1.53,
721
+ "learning_rate": 1.561739762012729e-05,
722
+ "loss": 0.3992,
723
+ "step": 1190
724
+ },
725
+ {
726
+ "epoch": 1.54,
727
+ "learning_rate": 1.5408175220603464e-05,
728
+ "loss": 0.4221,
729
+ "step": 1200
730
+ },
731
+ {
732
+ "epoch": 1.56,
733
+ "learning_rate": 1.5198873317196304e-05,
734
+ "loss": 0.4954,
735
+ "step": 1210
736
+ },
737
+ {
738
+ "epoch": 1.57,
739
+ "learning_rate": 1.4989532677481273e-05,
740
+ "loss": 0.4486,
741
+ "step": 1220
742
+ },
743
+ {
744
+ "epoch": 1.58,
745
+ "learning_rate": 1.4780194076578861e-05,
746
+ "loss": 0.4657,
747
+ "step": 1230
748
+ },
749
+ {
750
+ "epoch": 1.59,
751
+ "learning_rate": 1.4570898289212423e-05,
752
+ "loss": 0.3984,
753
+ "step": 1240
754
+ },
755
+ {
756
+ "epoch": 1.61,
757
+ "learning_rate": 1.4361686081766154e-05,
758
+ "loss": 0.512,
759
+ "step": 1250
760
+ },
761
+ {
762
+ "epoch": 1.62,
763
+ "learning_rate": 1.4152598204344656e-05,
764
+ "loss": 0.4178,
765
+ "step": 1260
766
+ },
767
+ {
768
+ "epoch": 1.63,
769
+ "learning_rate": 1.3943675382835672e-05,
770
+ "loss": 0.3971,
771
+ "step": 1270
772
+ },
773
+ {
774
+ "epoch": 1.65,
775
+ "learning_rate": 1.3734958310977542e-05,
776
+ "loss": 0.4343,
777
+ "step": 1280
778
+ },
779
+ {
780
+ "epoch": 1.66,
781
+ "learning_rate": 1.3526487642432953e-05,
782
+ "loss": 0.4675,
783
+ "step": 1290
784
+ },
785
+ {
786
+ "epoch": 1.67,
787
+ "learning_rate": 1.331830398287044e-05,
788
+ "loss": 0.469,
789
+ "step": 1300
790
+ },
791
+ {
792
+ "epoch": 1.68,
793
+ "learning_rate": 1.3110447882055255e-05,
794
+ "loss": 0.4457,
795
+ "step": 1310
796
+ },
797
+ {
798
+ "epoch": 1.7,
799
+ "learning_rate": 1.290295982595117e-05,
800
+ "loss": 0.4495,
801
+ "step": 1320
802
+ },
803
+ {
804
+ "epoch": 1.71,
805
+ "learning_rate": 1.2695880228834631e-05,
806
+ "loss": 0.4329,
807
+ "step": 1330
808
+ },
809
+ {
810
+ "epoch": 1.72,
811
+ "learning_rate": 1.2489249425422943e-05,
812
+ "loss": 0.4681,
813
+ "step": 1340
814
+ },
815
+ {
816
+ "epoch": 1.74,
817
+ "learning_rate": 1.2283107663017894e-05,
818
+ "loss": 0.4718,
819
+ "step": 1350
820
+ },
821
+ {
822
+ "epoch": 1.75,
823
+ "learning_rate": 1.207749509366645e-05,
824
+ "loss": 0.5212,
825
+ "step": 1360
826
+ },
827
+ {
828
+ "epoch": 1.76,
829
+ "learning_rate": 1.1872451766340002e-05,
830
+ "loss": 0.4652,
831
+ "step": 1370
832
+ },
833
+ {
834
+ "epoch": 1.77,
835
+ "learning_rate": 1.1668017619133619e-05,
836
+ "loss": 0.3881,
837
+ "step": 1380
838
+ },
839
+ {
840
+ "epoch": 1.79,
841
+ "learning_rate": 1.1464232471487033e-05,
842
+ "loss": 0.4384,
843
+ "step": 1390
844
+ },
845
+ {
846
+ "epoch": 1.8,
847
+ "learning_rate": 1.126113601642859e-05,
848
+ "loss": 0.4238,
849
+ "step": 1400
850
+ },
851
+ {
852
+ "epoch": 1.81,
853
+ "learning_rate": 1.1058767812843915e-05,
854
+ "loss": 0.4394,
855
+ "step": 1410
856
+ },
857
+ {
858
+ "epoch": 1.83,
859
+ "learning_rate": 1.0857167277770699e-05,
860
+ "loss": 0.4405,
861
+ "step": 1420
862
+ },
863
+ {
864
+ "epoch": 1.84,
865
+ "learning_rate": 1.0656373678721051e-05,
866
+ "loss": 0.4231,
867
+ "step": 1430
868
+ },
869
+ {
870
+ "epoch": 1.85,
871
+ "learning_rate": 1.0456426126033047e-05,
872
+ "loss": 0.4997,
873
+ "step": 1440
874
+ },
875
+ {
876
+ "epoch": 1.86,
877
+ "learning_rate": 1.0257363565252867e-05,
878
+ "loss": 0.4339,
879
+ "step": 1450
880
+ },
881
+ {
882
+ "epoch": 1.88,
883
+ "learning_rate": 1.0059224769549022e-05,
884
+ "loss": 0.4579,
885
+ "step": 1460
886
+ },
887
+ {
888
+ "epoch": 1.89,
889
+ "learning_rate": 9.862048332160156e-06,
890
+ "loss": 0.491,
891
+ "step": 1470
892
+ },
893
+ {
894
+ "epoch": 1.9,
895
+ "learning_rate": 9.665872658877904e-06,
896
+ "loss": 0.4215,
897
+ "step": 1480
898
+ },
899
+ {
900
+ "epoch": 1.92,
901
+ "learning_rate": 9.470735960566274e-06,
902
+ "loss": 0.3476,
903
+ "step": 1490
904
+ },
905
+ {
906
+ "epoch": 1.93,
907
+ "learning_rate": 9.276676245718955e-06,
908
+ "loss": 0.4651,
909
+ "step": 1500
910
+ },
911
+ {
912
+ "epoch": 1.94,
913
+ "learning_rate": 9.083731313056076e-06,
914
+ "loss": 0.4833,
915
+ "step": 1510
916
+ },
917
+ {
918
+ "epoch": 1.95,
919
+ "learning_rate": 8.891938744161832e-06,
920
+ "loss": 0.3936,
921
+ "step": 1520
922
+ },
923
+ {
924
+ "epoch": 1.97,
925
+ "learning_rate": 8.701335896164365e-06,
926
+ "loss": 0.4424,
927
+ "step": 1530
928
+ },
929
+ {
930
+ "epoch": 1.98,
931
+ "learning_rate": 8.511959894459403e-06,
932
+ "loss": 0.4624,
933
+ "step": 1540
934
+ },
935
+ {
936
+ "epoch": 1.99,
937
+ "learning_rate": 8.32384762547901e-06,
938
+ "loss": 0.3945,
939
+ "step": 1550
940
+ },
941
+ {
942
+ "epoch": 2.01,
943
+ "learning_rate": 8.13703572950691e-06,
944
+ "loss": 0.5025,
945
+ "step": 1560
946
+ }
947
+ ],
948
+ "logging_steps": 10,
949
+ "max_steps": 2345,
950
+ "num_train_epochs": 4,
951
+ "save_steps": 260,
952
+ "total_flos": 8.940392149669786e+17,
953
+ "trial_name": null,
954
+ "trial_params": null
955
+ }
checkpoint-1560/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c490f195e94c140704efa846a450585480b78d2abae7738c4ecc5de88d1d30a
3
+ size 4091
checkpoint-1820/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0
checkpoint-1820/adapter_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Qwen/Qwen-14B",
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,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 128,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "w2",
18
+ "c_attn",
19
+ "w1",
20
+ "c_proj"
21
+ ],
22
+ "task_type": "CAUSAL_LM"
23
+ }
checkpoint-1820/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:294d49f3f2ec901e0c0c683965394198f6fd7a7f7fb3bbb8a6689838ab2933fa
3
+ size 892742669
checkpoint-1820/qwen.tiktoken ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1820/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a9b97eb02b9ff49b9659396a151540eb1f764ec0092e4b17867eb2ed629fbc8
3
+ size 14575
checkpoint-1820/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cf9e311845992c4b988daf9837ac1b0af9be6ce0fc5d8a11d75ef0ef21b719c
3
+ size 627
checkpoint-1820/special_tokens_map.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "pad_token": "<|endoftext|>"
3
+ }
checkpoint-1820/tokenizer_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {},
3
+ "additional_special_tokens": [],
4
+ "auto_map": {
5
+ "AutoTokenizer": [
6
+ "Qwen/Qwen-14B--tokenization_qwen.QWenTokenizer",
7
+ null
8
+ ]
9
+ },
10
+ "clean_up_tokenization_spaces": true,
11
+ "legacy": false,
12
+ "model_max_length": 1600,
13
+ "pad_token": "<|endoftext|>",
14
+ "padding_side": "right",
15
+ "tokenizer_class": "QWenTokenizer",
16
+ "tokenizer_file": null
17
+ }
checkpoint-1820/trainer_state.json ADDED
@@ -0,0 +1,1111 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.3393316195372753,
5
+ "eval_steps": 15000,
6
+ "global_step": 1820,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 3.1914893617021277e-06,
14
+ "loss": 0.8544,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.03,
19
+ "learning_rate": 6.3829787234042555e-06,
20
+ "loss": 0.8342,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.04,
25
+ "learning_rate": 9.574468085106385e-06,
26
+ "loss": 0.8016,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.05,
31
+ "learning_rate": 1.2765957446808511e-05,
32
+ "loss": 0.7802,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.06,
37
+ "learning_rate": 1.5957446808510637e-05,
38
+ "loss": 0.7127,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.08,
43
+ "learning_rate": 1.914893617021277e-05,
44
+ "loss": 0.6504,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.09,
49
+ "learning_rate": 2.2340425531914894e-05,
50
+ "loss": 0.6011,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.1,
55
+ "learning_rate": 2.5531914893617022e-05,
56
+ "loss": 0.6859,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.12,
61
+ "learning_rate": 2.872340425531915e-05,
62
+ "loss": 0.5766,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.13,
67
+ "learning_rate": 2.9999474091752356e-05,
68
+ "loss": 0.6617,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.14,
73
+ "learning_rate": 2.9996260341563078e-05,
74
+ "loss": 0.6804,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.15,
79
+ "learning_rate": 2.999012563764485e-05,
80
+ "loss": 0.5635,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.17,
85
+ "learning_rate": 2.9981071174907998e-05,
86
+ "loss": 0.5021,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.18,
91
+ "learning_rate": 2.9969098716969983e-05,
92
+ "loss": 0.575,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.19,
97
+ "learning_rate": 2.995421059581191e-05,
98
+ "loss": 0.6085,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.21,
103
+ "learning_rate": 2.993640971132428e-05,
104
+ "loss": 0.6241,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.22,
109
+ "learning_rate": 2.991569953074218e-05,
110
+ "loss": 0.6405,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.23,
115
+ "learning_rate": 2.9892084087969912e-05,
116
+ "loss": 0.6318,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.24,
121
+ "learning_rate": 2.98655679827953e-05,
122
+ "loss": 0.5592,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.26,
127
+ "learning_rate": 2.9836156379993726e-05,
128
+ "loss": 0.5625,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.27,
133
+ "learning_rate": 2.980385500832215e-05,
134
+ "loss": 0.5857,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.28,
139
+ "learning_rate": 2.9768670159403283e-05,
140
+ "loss": 0.6439,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.3,
145
+ "learning_rate": 2.9730608686500093e-05,
146
+ "loss": 0.4829,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.31,
151
+ "learning_rate": 2.9689678003180932e-05,
152
+ "loss": 0.5554,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.32,
157
+ "learning_rate": 2.964588608187556e-05,
158
+ "loss": 0.5177,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 0.33,
163
+ "learning_rate": 2.9599241452322243e-05,
164
+ "loss": 0.5375,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 0.35,
169
+ "learning_rate": 2.9549753199906383e-05,
170
+ "loss": 0.5527,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 0.36,
175
+ "learning_rate": 2.9497430963890838e-05,
176
+ "loss": 0.5166,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 0.37,
181
+ "learning_rate": 2.9442284935538424e-05,
182
+ "loss": 0.5353,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 0.39,
187
+ "learning_rate": 2.9384325856126867e-05,
188
+ "loss": 0.531,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 0.4,
193
+ "learning_rate": 2.932356501485663e-05,
194
+ "loss": 0.5583,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 0.41,
199
+ "learning_rate": 2.9260014246652e-05,
200
+ "loss": 0.5355,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 0.42,
205
+ "learning_rate": 2.919368592985593e-05,
206
+ "loss": 0.5351,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 0.44,
211
+ "learning_rate": 2.9124592983818963e-05,
212
+ "loss": 0.5349,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 0.45,
217
+ "learning_rate": 2.9052748866382848e-05,
218
+ "loss": 0.5094,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 0.46,
223
+ "learning_rate": 2.8978167571259202e-05,
224
+ "loss": 0.538,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 0.48,
229
+ "learning_rate": 2.8900863625303867e-05,
230
+ "loss": 0.5041,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 0.49,
235
+ "learning_rate": 2.8820852085687373e-05,
236
+ "loss": 0.4906,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 0.5,
241
+ "learning_rate": 2.873814853696211e-05,
242
+ "loss": 0.5246,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 0.51,
247
+ "learning_rate": 2.8652769088026798e-05,
248
+ "loss": 0.5484,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 0.53,
253
+ "learning_rate": 2.856473036898882e-05,
254
+ "loss": 0.4325,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 0.54,
259
+ "learning_rate": 2.847404952792501e-05,
260
+ "loss": 0.4965,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 0.55,
265
+ "learning_rate": 2.8380744227541592e-05,
266
+ "loss": 0.509,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 0.57,
271
+ "learning_rate": 2.8284832641733857e-05,
272
+ "loss": 0.5052,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 0.58,
277
+ "learning_rate": 2.8186333452046256e-05,
278
+ "loss": 0.5119,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 0.59,
283
+ "learning_rate": 2.808526584403364e-05,
284
+ "loss": 0.4206,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 0.6,
289
+ "learning_rate": 2.7981649503524323e-05,
290
+ "loss": 0.5834,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 0.62,
295
+ "learning_rate": 2.7875504612785717e-05,
296
+ "loss": 0.4789,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 0.63,
301
+ "learning_rate": 2.7766851846593215e-05,
302
+ "loss": 0.4196,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 0.64,
307
+ "learning_rate": 2.7655712368203213e-05,
308
+ "loss": 0.4959,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 0.66,
313
+ "learning_rate": 2.7542107825230944e-05,
314
+ "loss": 0.4784,
315
+ "step": 510
316
+ },
317
+ {
318
+ "epoch": 0.67,
319
+ "learning_rate": 2.7426060345433984e-05,
320
+ "loss": 0.52,
321
+ "step": 520
322
+ },
323
+ {
324
+ "epoch": 0.68,
325
+ "learning_rate": 2.7307592532402248e-05,
326
+ "loss": 0.517,
327
+ "step": 530
328
+ },
329
+ {
330
+ "epoch": 0.69,
331
+ "learning_rate": 2.7186727461155252e-05,
332
+ "loss": 0.4751,
333
+ "step": 540
334
+ },
335
+ {
336
+ "epoch": 0.71,
337
+ "learning_rate": 2.706348867364763e-05,
338
+ "loss": 0.4711,
339
+ "step": 550
340
+ },
341
+ {
342
+ "epoch": 0.72,
343
+ "learning_rate": 2.6937900174183637e-05,
344
+ "loss": 0.4961,
345
+ "step": 560
346
+ },
347
+ {
348
+ "epoch": 0.73,
349
+ "learning_rate": 2.6809986424741623e-05,
350
+ "loss": 0.5276,
351
+ "step": 570
352
+ },
353
+ {
354
+ "epoch": 0.75,
355
+ "learning_rate": 2.6679772340209362e-05,
356
+ "loss": 0.5341,
357
+ "step": 580
358
+ },
359
+ {
360
+ "epoch": 0.76,
361
+ "learning_rate": 2.6547283283531153e-05,
362
+ "loss": 0.5208,
363
+ "step": 590
364
+ },
365
+ {
366
+ "epoch": 0.77,
367
+ "learning_rate": 2.6412545060767663e-05,
368
+ "loss": 0.4345,
369
+ "step": 600
370
+ },
371
+ {
372
+ "epoch": 0.78,
373
+ "learning_rate": 2.6275583916069456e-05,
374
+ "loss": 0.4814,
375
+ "step": 610
376
+ },
377
+ {
378
+ "epoch": 0.8,
379
+ "learning_rate": 2.6136426526565168e-05,
380
+ "loss": 0.4222,
381
+ "step": 620
382
+ },
383
+ {
384
+ "epoch": 0.81,
385
+ "learning_rate": 2.5995099997165395e-05,
386
+ "loss": 0.5428,
387
+ "step": 630
388
+ },
389
+ {
390
+ "epoch": 0.82,
391
+ "learning_rate": 2.585163185528322e-05,
392
+ "loss": 0.4636,
393
+ "step": 640
394
+ },
395
+ {
396
+ "epoch": 0.84,
397
+ "learning_rate": 2.5706050045472446e-05,
398
+ "loss": 0.4903,
399
+ "step": 650
400
+ },
401
+ {
402
+ "epoch": 0.85,
403
+ "learning_rate": 2.55583829239846e-05,
404
+ "loss": 0.4702,
405
+ "step": 660
406
+ },
407
+ {
408
+ "epoch": 0.86,
409
+ "learning_rate": 2.540865925324574e-05,
410
+ "loss": 0.5231,
411
+ "step": 670
412
+ },
413
+ {
414
+ "epoch": 0.87,
415
+ "learning_rate": 2.5256908196254126e-05,
416
+ "loss": 0.5089,
417
+ "step": 680
418
+ },
419
+ {
420
+ "epoch": 0.89,
421
+ "learning_rate": 2.510315931089991e-05,
422
+ "loss": 0.4572,
423
+ "step": 690
424
+ },
425
+ {
426
+ "epoch": 0.9,
427
+ "learning_rate": 2.494744254420787e-05,
428
+ "loss": 0.471,
429
+ "step": 700
430
+ },
431
+ {
432
+ "epoch": 0.91,
433
+ "learning_rate": 2.478978822650436e-05,
434
+ "loss": 0.3929,
435
+ "step": 710
436
+ },
437
+ {
438
+ "epoch": 0.93,
439
+ "learning_rate": 2.4630227065509627e-05,
440
+ "loss": 0.476,
441
+ "step": 720
442
+ },
443
+ {
444
+ "epoch": 0.94,
445
+ "learning_rate": 2.446879014035656e-05,
446
+ "loss": 0.5348,
447
+ "step": 730
448
+ },
449
+ {
450
+ "epoch": 0.95,
451
+ "learning_rate": 2.4305508895537186e-05,
452
+ "loss": 0.4399,
453
+ "step": 740
454
+ },
455
+ {
456
+ "epoch": 0.96,
457
+ "learning_rate": 2.4140415134777895e-05,
458
+ "loss": 0.4443,
459
+ "step": 750
460
+ },
461
+ {
462
+ "epoch": 0.98,
463
+ "learning_rate": 2.3973541014844783e-05,
464
+ "loss": 0.5046,
465
+ "step": 760
466
+ },
467
+ {
468
+ "epoch": 0.99,
469
+ "learning_rate": 2.380491903928019e-05,
470
+ "loss": 0.4318,
471
+ "step": 770
472
+ },
473
+ {
474
+ "epoch": 1.0,
475
+ "learning_rate": 2.3634582052071682e-05,
476
+ "loss": 0.5244,
477
+ "step": 780
478
+ },
479
+ {
480
+ "epoch": 1.02,
481
+ "learning_rate": 2.3462563231254764e-05,
482
+ "loss": 0.5364,
483
+ "step": 790
484
+ },
485
+ {
486
+ "epoch": 1.03,
487
+ "learning_rate": 2.328889608245048e-05,
488
+ "loss": 0.5278,
489
+ "step": 800
490
+ },
491
+ {
492
+ "epoch": 1.04,
493
+ "learning_rate": 2.3113614432339247e-05,
494
+ "loss": 0.5021,
495
+ "step": 810
496
+ },
497
+ {
498
+ "epoch": 1.05,
499
+ "learning_rate": 2.2936752422072105e-05,
500
+ "loss": 0.4363,
501
+ "step": 820
502
+ },
503
+ {
504
+ "epoch": 1.07,
505
+ "learning_rate": 2.275834450062077e-05,
506
+ "loss": 0.4657,
507
+ "step": 830
508
+ },
509
+ {
510
+ "epoch": 1.08,
511
+ "learning_rate": 2.2578425418067713e-05,
512
+ "loss": 0.4355,
513
+ "step": 840
514
+ },
515
+ {
516
+ "epoch": 1.09,
517
+ "learning_rate": 2.239703021883754e-05,
518
+ "loss": 0.402,
519
+ "step": 850
520
+ },
521
+ {
522
+ "epoch": 1.11,
523
+ "learning_rate": 2.221419423487113e-05,
524
+ "loss": 0.5007,
525
+ "step": 860
526
+ },
527
+ {
528
+ "epoch": 1.12,
529
+ "learning_rate": 2.202995307874367e-05,
530
+ "loss": 0.4388,
531
+ "step": 870
532
+ },
533
+ {
534
+ "epoch": 1.13,
535
+ "learning_rate": 2.1844342636728133e-05,
536
+ "loss": 0.5182,
537
+ "step": 880
538
+ },
539
+ {
540
+ "epoch": 1.14,
541
+ "learning_rate": 2.165739906180535e-05,
542
+ "loss": 0.5118,
543
+ "step": 890
544
+ },
545
+ {
546
+ "epoch": 1.16,
547
+ "learning_rate": 2.1469158766622195e-05,
548
+ "loss": 0.3888,
549
+ "step": 900
550
+ },
551
+ {
552
+ "epoch": 1.17,
553
+ "learning_rate": 2.127965841639918e-05,
554
+ "loss": 0.3978,
555
+ "step": 910
556
+ },
557
+ {
558
+ "epoch": 1.18,
559
+ "learning_rate": 2.1088934921788813e-05,
560
+ "loss": 0.5071,
561
+ "step": 920
562
+ },
563
+ {
564
+ "epoch": 1.2,
565
+ "learning_rate": 2.0897025431686234e-05,
566
+ "loss": 0.4999,
567
+ "step": 930
568
+ },
569
+ {
570
+ "epoch": 1.21,
571
+ "learning_rate": 2.0703967325993373e-05,
572
+ "loss": 0.4734,
573
+ "step": 940
574
+ },
575
+ {
576
+ "epoch": 1.22,
577
+ "learning_rate": 2.0509798208338107e-05,
578
+ "loss": 0.5421,
579
+ "step": 950
580
+ },
581
+ {
582
+ "epoch": 1.23,
583
+ "learning_rate": 2.0314555898749934e-05,
584
+ "loss": 0.4867,
585
+ "step": 960
586
+ },
587
+ {
588
+ "epoch": 1.25,
589
+ "learning_rate": 2.0118278426293376e-05,
590
+ "loss": 0.4662,
591
+ "step": 970
592
+ },
593
+ {
594
+ "epoch": 1.26,
595
+ "learning_rate": 1.9921004021660762e-05,
596
+ "loss": 0.4709,
597
+ "step": 980
598
+ },
599
+ {
600
+ "epoch": 1.27,
601
+ "learning_rate": 1.972277110972567e-05,
602
+ "loss": 0.5178,
603
+ "step": 990
604
+ },
605
+ {
606
+ "epoch": 1.29,
607
+ "learning_rate": 1.952361830205859e-05,
608
+ "loss": 0.5188,
609
+ "step": 1000
610
+ },
611
+ {
612
+ "epoch": 1.3,
613
+ "learning_rate": 1.932358438940622e-05,
614
+ "loss": 0.412,
615
+ "step": 1010
616
+ },
617
+ {
618
+ "epoch": 1.31,
619
+ "learning_rate": 1.9122708334135816e-05,
620
+ "loss": 0.4236,
621
+ "step": 1020
622
+ },
623
+ {
624
+ "epoch": 1.32,
625
+ "learning_rate": 1.8921029262646187e-05,
626
+ "loss": 0.4594,
627
+ "step": 1030
628
+ },
629
+ {
630
+ "epoch": 1.34,
631
+ "learning_rate": 1.8718586457746694e-05,
632
+ "loss": 0.4521,
633
+ "step": 1040
634
+ },
635
+ {
636
+ "epoch": 1.35,
637
+ "learning_rate": 1.8515419351005745e-05,
638
+ "loss": 0.4427,
639
+ "step": 1050
640
+ },
641
+ {
642
+ "epoch": 1.36,
643
+ "learning_rate": 1.8311567515070438e-05,
644
+ "loss": 0.4509,
645
+ "step": 1060
646
+ },
647
+ {
648
+ "epoch": 1.38,
649
+ "learning_rate": 1.8107070655958584e-05,
650
+ "loss": 0.4548,
651
+ "step": 1070
652
+ },
653
+ {
654
+ "epoch": 1.39,
655
+ "learning_rate": 1.7901968605324847e-05,
656
+ "loss": 0.479,
657
+ "step": 1080
658
+ },
659
+ {
660
+ "epoch": 1.4,
661
+ "learning_rate": 1.7696301312702383e-05,
662
+ "loss": 0.4867,
663
+ "step": 1090
664
+ },
665
+ {
666
+ "epoch": 1.41,
667
+ "learning_rate": 1.7490108837721493e-05,
668
+ "loss": 0.4894,
669
+ "step": 1100
670
+ },
671
+ {
672
+ "epoch": 1.43,
673
+ "learning_rate": 1.7283431342306884e-05,
674
+ "loss": 0.4646,
675
+ "step": 1110
676
+ },
677
+ {
678
+ "epoch": 1.44,
679
+ "learning_rate": 1.7076309082854963e-05,
680
+ "loss": 0.4802,
681
+ "step": 1120
682
+ },
683
+ {
684
+ "epoch": 1.45,
685
+ "learning_rate": 1.686878240239274e-05,
686
+ "loss": 0.4287,
687
+ "step": 1130
688
+ },
689
+ {
690
+ "epoch": 1.47,
691
+ "learning_rate": 1.666089172271988e-05,
692
+ "loss": 0.4808,
693
+ "step": 1140
694
+ },
695
+ {
696
+ "epoch": 1.48,
697
+ "learning_rate": 1.6452677536535375e-05,
698
+ "loss": 0.4639,
699
+ "step": 1150
700
+ },
701
+ {
702
+ "epoch": 1.49,
703
+ "learning_rate": 1.624418039955042e-05,
704
+ "loss": 0.4263,
705
+ "step": 1160
706
+ },
707
+ {
708
+ "epoch": 1.5,
709
+ "learning_rate": 1.6035440922589054e-05,
710
+ "loss": 0.492,
711
+ "step": 1170
712
+ },
713
+ {
714
+ "epoch": 1.52,
715
+ "learning_rate": 1.5826499763677982e-05,
716
+ "loss": 0.487,
717
+ "step": 1180
718
+ },
719
+ {
720
+ "epoch": 1.53,
721
+ "learning_rate": 1.561739762012729e-05,
722
+ "loss": 0.3992,
723
+ "step": 1190
724
+ },
725
+ {
726
+ "epoch": 1.54,
727
+ "learning_rate": 1.5408175220603464e-05,
728
+ "loss": 0.4221,
729
+ "step": 1200
730
+ },
731
+ {
732
+ "epoch": 1.56,
733
+ "learning_rate": 1.5198873317196304e-05,
734
+ "loss": 0.4954,
735
+ "step": 1210
736
+ },
737
+ {
738
+ "epoch": 1.57,
739
+ "learning_rate": 1.4989532677481273e-05,
740
+ "loss": 0.4486,
741
+ "step": 1220
742
+ },
743
+ {
744
+ "epoch": 1.58,
745
+ "learning_rate": 1.4780194076578861e-05,
746
+ "loss": 0.4657,
747
+ "step": 1230
748
+ },
749
+ {
750
+ "epoch": 1.59,
751
+ "learning_rate": 1.4570898289212423e-05,
752
+ "loss": 0.3984,
753
+ "step": 1240
754
+ },
755
+ {
756
+ "epoch": 1.61,
757
+ "learning_rate": 1.4361686081766154e-05,
758
+ "loss": 0.512,
759
+ "step": 1250
760
+ },
761
+ {
762
+ "epoch": 1.62,
763
+ "learning_rate": 1.4152598204344656e-05,
764
+ "loss": 0.4178,
765
+ "step": 1260
766
+ },
767
+ {
768
+ "epoch": 1.63,
769
+ "learning_rate": 1.3943675382835672e-05,
770
+ "loss": 0.3971,
771
+ "step": 1270
772
+ },
773
+ {
774
+ "epoch": 1.65,
775
+ "learning_rate": 1.3734958310977542e-05,
776
+ "loss": 0.4343,
777
+ "step": 1280
778
+ },
779
+ {
780
+ "epoch": 1.66,
781
+ "learning_rate": 1.3526487642432953e-05,
782
+ "loss": 0.4675,
783
+ "step": 1290
784
+ },
785
+ {
786
+ "epoch": 1.67,
787
+ "learning_rate": 1.331830398287044e-05,
788
+ "loss": 0.469,
789
+ "step": 1300
790
+ },
791
+ {
792
+ "epoch": 1.68,
793
+ "learning_rate": 1.3110447882055255e-05,
794
+ "loss": 0.4457,
795
+ "step": 1310
796
+ },
797
+ {
798
+ "epoch": 1.7,
799
+ "learning_rate": 1.290295982595117e-05,
800
+ "loss": 0.4495,
801
+ "step": 1320
802
+ },
803
+ {
804
+ "epoch": 1.71,
805
+ "learning_rate": 1.2695880228834631e-05,
806
+ "loss": 0.4329,
807
+ "step": 1330
808
+ },
809
+ {
810
+ "epoch": 1.72,
811
+ "learning_rate": 1.2489249425422943e-05,
812
+ "loss": 0.4681,
813
+ "step": 1340
814
+ },
815
+ {
816
+ "epoch": 1.74,
817
+ "learning_rate": 1.2283107663017894e-05,
818
+ "loss": 0.4718,
819
+ "step": 1350
820
+ },
821
+ {
822
+ "epoch": 1.75,
823
+ "learning_rate": 1.207749509366645e-05,
824
+ "loss": 0.5212,
825
+ "step": 1360
826
+ },
827
+ {
828
+ "epoch": 1.76,
829
+ "learning_rate": 1.1872451766340002e-05,
830
+ "loss": 0.4652,
831
+ "step": 1370
832
+ },
833
+ {
834
+ "epoch": 1.77,
835
+ "learning_rate": 1.1668017619133619e-05,
836
+ "loss": 0.3881,
837
+ "step": 1380
838
+ },
839
+ {
840
+ "epoch": 1.79,
841
+ "learning_rate": 1.1464232471487033e-05,
842
+ "loss": 0.4384,
843
+ "step": 1390
844
+ },
845
+ {
846
+ "epoch": 1.8,
847
+ "learning_rate": 1.126113601642859e-05,
848
+ "loss": 0.4238,
849
+ "step": 1400
850
+ },
851
+ {
852
+ "epoch": 1.81,
853
+ "learning_rate": 1.1058767812843915e-05,
854
+ "loss": 0.4394,
855
+ "step": 1410
856
+ },
857
+ {
858
+ "epoch": 1.83,
859
+ "learning_rate": 1.0857167277770699e-05,
860
+ "loss": 0.4405,
861
+ "step": 1420
862
+ },
863
+ {
864
+ "epoch": 1.84,
865
+ "learning_rate": 1.0656373678721051e-05,
866
+ "loss": 0.4231,
867
+ "step": 1430
868
+ },
869
+ {
870
+ "epoch": 1.85,
871
+ "learning_rate": 1.0456426126033047e-05,
872
+ "loss": 0.4997,
873
+ "step": 1440
874
+ },
875
+ {
876
+ "epoch": 1.86,
877
+ "learning_rate": 1.0257363565252867e-05,
878
+ "loss": 0.4339,
879
+ "step": 1450
880
+ },
881
+ {
882
+ "epoch": 1.88,
883
+ "learning_rate": 1.0059224769549022e-05,
884
+ "loss": 0.4579,
885
+ "step": 1460
886
+ },
887
+ {
888
+ "epoch": 1.89,
889
+ "learning_rate": 9.862048332160156e-06,
890
+ "loss": 0.491,
891
+ "step": 1470
892
+ },
893
+ {
894
+ "epoch": 1.9,
895
+ "learning_rate": 9.665872658877904e-06,
896
+ "loss": 0.4215,
897
+ "step": 1480
898
+ },
899
+ {
900
+ "epoch": 1.92,
901
+ "learning_rate": 9.470735960566274e-06,
902
+ "loss": 0.3476,
903
+ "step": 1490
904
+ },
905
+ {
906
+ "epoch": 1.93,
907
+ "learning_rate": 9.276676245718955e-06,
908
+ "loss": 0.4651,
909
+ "step": 1500
910
+ },
911
+ {
912
+ "epoch": 1.94,
913
+ "learning_rate": 9.083731313056076e-06,
914
+ "loss": 0.4833,
915
+ "step": 1510
916
+ },
917
+ {
918
+ "epoch": 1.95,
919
+ "learning_rate": 8.891938744161832e-06,
920
+ "loss": 0.3936,
921
+ "step": 1520
922
+ },
923
+ {
924
+ "epoch": 1.97,
925
+ "learning_rate": 8.701335896164365e-06,
926
+ "loss": 0.4424,
927
+ "step": 1530
928
+ },
929
+ {
930
+ "epoch": 1.98,
931
+ "learning_rate": 8.511959894459403e-06,
932
+ "loss": 0.4624,
933
+ "step": 1540
934
+ },
935
+ {
936
+ "epoch": 1.99,
937
+ "learning_rate": 8.32384762547901e-06,
938
+ "loss": 0.3945,
939
+ "step": 1550
940
+ },
941
+ {
942
+ "epoch": 2.01,
943
+ "learning_rate": 8.13703572950691e-06,
944
+ "loss": 0.5025,
945
+ "step": 1560
946
+ },
947
+ {
948
+ "epoch": 2.02,
949
+ "learning_rate": 7.951560593541705e-06,
950
+ "loss": 0.5045,
951
+ "step": 1570
952
+ },
953
+ {
954
+ "epoch": 2.03,
955
+ "learning_rate": 7.767458344209473e-06,
956
+ "loss": 0.5225,
957
+ "step": 1580
958
+ },
959
+ {
960
+ "epoch": 2.04,
961
+ "learning_rate": 7.584764840727073e-06,
962
+ "loss": 0.4701,
963
+ "step": 1590
964
+ },
965
+ {
966
+ "epoch": 2.06,
967
+ "learning_rate": 7.403515667917537e-06,
968
+ "loss": 0.3575,
969
+ "step": 1600
970
+ },
971
+ {
972
+ "epoch": 2.07,
973
+ "learning_rate": 7.223746129278879e-06,
974
+ "loss": 0.447,
975
+ "step": 1610
976
+ },
977
+ {
978
+ "epoch": 2.08,
979
+ "learning_rate": 7.045491240107726e-06,
980
+ "loss": 0.4191,
981
+ "step": 1620
982
+ },
983
+ {
984
+ "epoch": 2.1,
985
+ "learning_rate": 6.868785720679093e-06,
986
+ "loss": 0.3947,
987
+ "step": 1630
988
+ },
989
+ {
990
+ "epoch": 2.11,
991
+ "learning_rate": 6.6936639894836e-06,
992
+ "loss": 0.4557,
993
+ "step": 1640
994
+ },
995
+ {
996
+ "epoch": 2.12,
997
+ "learning_rate": 6.520160156523481e-06,
998
+ "loss": 0.404,
999
+ "step": 1650
1000
+ },
1001
+ {
1002
+ "epoch": 2.13,
1003
+ "learning_rate": 6.348308016668679e-06,
1004
+ "loss": 0.4951,
1005
+ "step": 1660
1006
+ },
1007
+ {
1008
+ "epoch": 2.15,
1009
+ "learning_rate": 6.178141043074322e-06,
1010
+ "loss": 0.4677,
1011
+ "step": 1670
1012
+ },
1013
+ {
1014
+ "epoch": 2.16,
1015
+ "learning_rate": 6.0096923806608595e-06,
1016
+ "loss": 0.3628,
1017
+ "step": 1680
1018
+ },
1019
+ {
1020
+ "epoch": 2.17,
1021
+ "learning_rate": 5.842994839658134e-06,
1022
+ "loss": 0.4757,
1023
+ "step": 1690
1024
+ },
1025
+ {
1026
+ "epoch": 2.19,
1027
+ "learning_rate": 5.6780808892146444e-06,
1028
+ "loss": 0.387,
1029
+ "step": 1700
1030
+ },
1031
+ {
1032
+ "epoch": 2.2,
1033
+ "learning_rate": 5.514982651073259e-06,
1034
+ "loss": 0.4685,
1035
+ "step": 1710
1036
+ },
1037
+ {
1038
+ "epoch": 2.21,
1039
+ "learning_rate": 5.3537318933145445e-06,
1040
+ "loss": 0.4819,
1041
+ "step": 1720
1042
+ },
1043
+ {
1044
+ "epoch": 2.22,
1045
+ "learning_rate": 5.19436002416905e-06,
1046
+ "loss": 0.4856,
1047
+ "step": 1730
1048
+ },
1049
+ {
1050
+ "epoch": 2.24,
1051
+ "learning_rate": 5.036898085899639e-06,
1052
+ "loss": 0.4924,
1053
+ "step": 1740
1054
+ },
1055
+ {
1056
+ "epoch": 2.25,
1057
+ "learning_rate": 4.881376748755111e-06,
1058
+ "loss": 0.4058,
1059
+ "step": 1750
1060
+ },
1061
+ {
1062
+ "epoch": 2.26,
1063
+ "learning_rate": 4.727826304996294e-06,
1064
+ "loss": 0.4738,
1065
+ "step": 1760
1066
+ },
1067
+ {
1068
+ "epoch": 2.28,
1069
+ "learning_rate": 4.576276662995751e-06,
1070
+ "loss": 0.519,
1071
+ "step": 1770
1072
+ },
1073
+ {
1074
+ "epoch": 2.29,
1075
+ "learning_rate": 4.426757341412246e-06,
1076
+ "loss": 0.4648,
1077
+ "step": 1780
1078
+ },
1079
+ {
1080
+ "epoch": 2.3,
1081
+ "learning_rate": 4.279297463441161e-06,
1082
+ "loss": 0.3798,
1083
+ "step": 1790
1084
+ },
1085
+ {
1086
+ "epoch": 2.31,
1087
+ "learning_rate": 4.133925751141913e-06,
1088
+ "loss": 0.4013,
1089
+ "step": 1800
1090
+ },
1091
+ {
1092
+ "epoch": 2.33,
1093
+ "learning_rate": 3.990670519843484e-06,
1094
+ "loss": 0.4445,
1095
+ "step": 1810
1096
+ },
1097
+ {
1098
+ "epoch": 2.34,
1099
+ "learning_rate": 3.849559672629211e-06,
1100
+ "loss": 0.4292,
1101
+ "step": 1820
1102
+ }
1103
+ ],
1104
+ "logging_steps": 10,
1105
+ "max_steps": 2345,
1106
+ "num_train_epochs": 4,
1107
+ "save_steps": 260,
1108
+ "total_flos": 1.0397345454789427e+18,
1109
+ "trial_name": null,
1110
+ "trial_params": null
1111
+ }
checkpoint-1820/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c490f195e94c140704efa846a450585480b78d2abae7738c4ecc5de88d1d30a
3
+ size 4091
checkpoint-2080/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - quant_method: bitsandbytes
9
+ - load_in_8bit: False
10
+ - load_in_4bit: True
11
+ - llm_int8_threshold: 6.0
12
+ - llm_int8_skip_modules: None
13
+ - llm_int8_enable_fp32_cpu_offload: False
14
+ - llm_int8_has_fp16_weight: False
15
+ - bnb_4bit_quant_type: nf4
16
+ - bnb_4bit_use_double_quant: True
17
+ - bnb_4bit_compute_dtype: bfloat16
18
+ ### Framework versions
19
+
20
+
21
+ - PEFT 0.5.0
checkpoint-2080/adapter_config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "Qwen/Qwen-14B",
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,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 128,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "w2",
18
+ "c_attn",
19
+ "w1",
20
+ "c_proj"
21
+ ],
22
+ "task_type": "CAUSAL_LM"
23
+ }
checkpoint-2080/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f04695ec066bc0a9acc4152162aa478f8b9295b64d100ffa48168eec1b4068f3
3
+ size 892742669
checkpoint-2080/qwen.tiktoken ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-2080/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af9f268cae4eb87e82fe5c0738becca8d5ca5385a639c7f07a301060b6a88141
3
+ size 14575
checkpoint-2080/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c1ba9b4fc24f5b38c53da2c35532e0b8a290dd02a21b3f2cb8b532ed2703f0c
3
+ size 627
checkpoint-2080/special_tokens_map.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "pad_token": "<|endoftext|>"
3
+ }