File size: 33,598 Bytes
cc3cebf
1f3c3f5
 
3c71c8d
a1da39c
6b3b271
3c71c8d
52dd92a
 
cc3cebf
52dd92a
8cec513
cc3cebf
52dd92a
0f10080
a1da39c
52dd92a
6b3b271
 
52dd92a
cc3cebf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1f3c3f5
0f10080
1f3c3f5
 
3c71c8d
 
 
 
 
 
0f10080
af29d81
cc3cebf
0f10080
 
 
 
cc3cebf
3c71c8d
 
cc3cebf
 
 
 
 
3c71c8d
 
 
cc3cebf
 
 
 
3c71c8d
 
cc3cebf
 
52dd92a
cc3cebf
 
 
 
52dd92a
 
 
cc3cebf
 
 
52dd92a
 
cc3cebf
 
52dd92a
cc3cebf
 
 
 
 
 
52dd92a
 
 
cc3cebf
 
 
 
52dd92a
 
cc3cebf
 
52dd92a
cc3cebf
 
 
 
 
 
 
 
 
 
 
52dd92a
 
 
cc3cebf
52dd92a
cc3cebf
 
52dd92a
cc3cebf
 
 
 
 
 
 
 
 
 
52dd92a
 
 
 
 
cc3cebf
 
52dd92a
cc3cebf
 
 
 
 
 
 
 
 
 
52dd92a
 
 
cc3cebf
 
 
 
 
 
52dd92a
cc3cebf
 
 
 
 
 
 
 
 
52dd92a
 
 
cc3cebf
 
 
0f985e1
 
 
 
 
 
 
 
 
 
 
 
 
 
6b3b271
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52dd92a
 
 
cc3cebf
 
 
 
 
 
 
52dd92a
 
 
 
 
cc3cebf
 
 
 
 
 
 
52dd92a
 
 
 
1c33eb8
cc3cebf
 
 
 
 
 
 
1c33eb8
 
cc3cebf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1c33eb8
 
cc3cebf
 
 
 
 
 
 
1c33eb8
3aad5f3
cc3cebf
 
 
 
 
 
 
 
 
 
3aad5f3
 
cc3cebf
3aad5f3
 
cc3cebf
 
 
 
 
 
 
 
3aad5f3
 
cc3cebf
 
 
 
 
 
 
 
 
 
 
 
3aad5f3
 
cc3cebf
3aad5f3
 
cc3cebf
 
 
 
 
 
 
 
3aad5f3
cc3cebf
3aad5f3
 
 
cc3cebf
 
 
 
babf0fd
cc3cebf
babf0fd
 
 
cc3cebf
 
 
 
 
 
 
 
babf0fd
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
babf0fd
 
 
 
 
cc3cebf
 
 
 
 
 
 
babf0fd
 
 
eea2731
cc3cebf
14668fa
 
 
cc3cebf
 
 
 
 
 
 
 
14668fa
 
 
 
fd2c981
cc3cebf
 
 
 
 
 
 
 
 
 
eea2731
 
 
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
eea2731
 
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
eea2731
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
eea2731
 
 
 
 
cb9d3af
cc3cebf
 
 
 
 
 
 
 
 
cb9d3af
 
 
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
cb9d3af
 
 
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
 
 
cb9d3af
 
 
ad5ca4f
cc3cebf
 
 
 
 
 
 
ad5ca4f
 
 
2bb0b78
cc3cebf
 
 
 
 
 
 
 
2bb0b78
2ce5c0d
 
 
 
2bb0b78
 
cc3cebf
 
 
 
 
 
 
 
 
3437149
 
 
 
 
 
 
 
 
cc3cebf
cfbce02
 
 
 
cc3cebf
 
 
 
 
 
 
 
cfbce02
 
 
cc3cebf
 
 
 
 
 
 
 
 
 
 
cfbce02
 
 
e7d3e2d
cc3cebf
 
 
 
 
 
e7d3e2d
 
cc3cebf
e7d3e2d
 
 
 
cc3cebf
 
 
 
 
 
 
 
 
 
 
e7d3e2d
 
cc3cebf
e7d3e2d
 
 
 
cc3cebf
 
 
 
 
 
 
 
 
 
 
383f88d
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
383f88d
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
 
 
383f88d
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
383f88d
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
 
383f88d
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
383f88d
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
 
383f88d
 
 
 
cc3cebf
 
 
 
 
 
 
 
383f88d
 
 
9923b72
cc3cebf
 
 
 
 
 
 
 
 
9923b72
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
9923b72
 
 
 
cc3cebf
 
 
 
 
 
 
 
9923b72
 
 
 
cc3cebf
 
 
 
 
 
 
 
 
9923b72
 
 
44c9d01
cc3cebf
 
 
 
 
 
 
 
44c9d01
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
44c9d01
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
44c9d01
 
 
 
cc3cebf
 
 
 
 
 
 
 
44c9d01
 
 
fb12895
cc3cebf
 
 
 
 
 
 
 
 
fb12895
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
fb12895
 
 
 
cc3cebf
 
 
 
 
 
 
fb12895
 
 
a1da39c
cc3cebf
 
 
 
 
 
 
 
 
 
 
 
8487b97
 
 
 
 
 
 
 
cc3cebf
 
af29d81
 
 
 
 
 
 
cc3cebf
 
af29d81
 
 
 
 
0f10080
cc3cebf
0f10080
 
 
 
cc3cebf
 
 
 
 
 
1ffa386
0f10080
1ffa386
 
 
0f10080
1ffa386
0f10080
1ffa386
cc3cebf
 
 
 
 
 
 
 
 
 
1ffa386
 
 
 
0f10080
1ffa386
0f10080
1ffa386
cc3cebf
 
 
 
 
 
 
 
 
 
1ffa386
 
0f10080
 
cc3cebf
 
 
 
 
 
0f10080
814aee6
0f10080
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
 
 
0f10080
 
 
 
 
 
 
 
cc3cebf
 
 
 
 
 
 
 
 
 
0f10080
 
 
1ffa386
a1da39c
cc3cebf
a1da39c
 
 
 
cc3cebf
 
 
 
 
 
 
 
a1da39c
 
 
cc3cebf
a1da39c
 
 
 
 
 
cc3cebf
a1da39c
cc3cebf
 
 
 
 
 
 
a1da39c
 
cc3cebf
a1da39c
cc3cebf
a1da39c
cc3cebf
 
 
 
 
 
 
 
 
 
 
 
 
 
a1da39c
 
cc3cebf
a1da39c
cc3cebf
a1da39c
 
 
 
 
 
 
 
 
 
788649f
 
 
 
 
 
 
 
 
cc3cebf
 
a1da39c
cc3cebf
a1da39c
cc3cebf
a1da39c
 
 
cc3cebf
 
 
 
 
 
 
a1da39c
 
cc3cebf
a1da39c
cc3cebf
a1da39c
 
788649f
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
# pylint: disable=too-many-lines
"""Module for testing the validation module"""

import logging
import os
import warnings
from typing import Optional

import pytest
from pydantic import ValidationError

from axolotl.utils.config import validate_config
from axolotl.utils.config.models.input.v0_4_1 import AxolotlConfigWCapabilities
from axolotl.utils.dict import DictDefault
from axolotl.utils.models import check_model_config
from axolotl.utils.wandb_ import setup_wandb_env_vars

warnings.filterwarnings("error")


@pytest.fixture(name="minimal_cfg")
def fixture_cfg():
    return DictDefault(
        {
            "base_model": "TinyLlama/TinyLlama-1.1B-Chat-v0.6",
            "learning_rate": 0.000001,
            "datasets": [
                {
                    "path": "mhenrichsen/alpaca_2k_test",
                    "type": "alpaca",
                }
            ],
            "micro_batch_size": 1,
            "gradient_accumulation_steps": 1,
        }
    )


class BaseValidation:
    """
    Base validation module to setup the log capture
    """

    _caplog: Optional[pytest.LogCaptureFixture] = None

    @pytest.fixture(autouse=True)
    def inject_fixtures(self, caplog):
        self._caplog = caplog


# pylint: disable=too-many-public-methods
class TestValidation(BaseValidation):
    """
    Test the validation module
    """

    def test_datasets_min_length(self):
        cfg = DictDefault(
            {
                "base_model": "TinyLlama/TinyLlama-1.1B-Chat-v0.6",
                "learning_rate": 0.000001,
                "datasets": [],
                "micro_batch_size": 1,
                "gradient_accumulation_steps": 1,
            }
        )

        with pytest.raises(
            ValidationError,
            match=r".*List should have at least 1 item after validation*",
        ):
            validate_config(cfg)

    def test_datasets_min_length_empty(self):
        cfg = DictDefault(
            {
                "base_model": "TinyLlama/TinyLlama-1.1B-Chat-v0.6",
                "learning_rate": 0.000001,
                "micro_batch_size": 1,
                "gradient_accumulation_steps": 1,
            }
        )

        with pytest.raises(
            ValueError, match=r".*either datasets or pretraining_dataset is required*"
        ):
            validate_config(cfg)

    def test_pretrain_dataset_min_length(self):
        cfg = DictDefault(
            {
                "base_model": "TinyLlama/TinyLlama-1.1B-Chat-v0.6",
                "learning_rate": 0.000001,
                "pretraining_dataset": [],
                "micro_batch_size": 1,
                "gradient_accumulation_steps": 1,
                "max_steps": 100,
            }
        )

        with pytest.raises(
            ValidationError,
            match=r".*List should have at least 1 item after validation*",
        ):
            validate_config(cfg)

    def test_valid_pretrain_dataset(self):
        cfg = DictDefault(
            {
                "base_model": "TinyLlama/TinyLlama-1.1B-Chat-v0.6",
                "learning_rate": 0.000001,
                "pretraining_dataset": [
                    {
                        "path": "mhenrichsen/alpaca_2k_test",
                        "type": "alpaca",
                    }
                ],
                "micro_batch_size": 1,
                "gradient_accumulation_steps": 1,
                "max_steps": 100,
            }
        )

        validate_config(cfg)

    def test_valid_sft_dataset(self):
        cfg = DictDefault(
            {
                "base_model": "TinyLlama/TinyLlama-1.1B-Chat-v0.6",
                "learning_rate": 0.000001,
                "datasets": [
                    {
                        "path": "mhenrichsen/alpaca_2k_test",
                        "type": "alpaca",
                    }
                ],
                "micro_batch_size": 1,
                "gradient_accumulation_steps": 1,
            }
        )

        validate_config(cfg)

    def test_batch_size_unused_warning(self):
        cfg = DictDefault(
            {
                "base_model": "TinyLlama/TinyLlama-1.1B-Chat-v0.6",
                "learning_rate": 0.000001,
                "datasets": [
                    {
                        "path": "mhenrichsen/alpaca_2k_test",
                        "type": "alpaca",
                    }
                ],
                "micro_batch_size": 4,
                "batch_size": 32,
            }
        )

        with self._caplog.at_level(logging.WARNING):
            validate_config(cfg)
            assert "batch_size is not recommended" in self._caplog.records[0].message

    def test_batch_size_more_params(self):
        cfg = DictDefault(
            {
                "base_model": "TinyLlama/TinyLlama-1.1B-Chat-v0.6",
                "learning_rate": 0.000001,
                "datasets": [
                    {
                        "path": "mhenrichsen/alpaca_2k_test",
                        "type": "alpaca",
                    }
                ],
                "batch_size": 32,
            }
        )

        with pytest.raises(ValueError, match=r".*At least two of*"):
            validate_config(cfg)

    def test_lr_as_float(self, minimal_cfg):
        cfg = (
            DictDefault(  # pylint: disable=unsupported-binary-operation
                {
                    "learning_rate": "5e-5",
                }
            )
            | minimal_cfg
        )

        new_cfg = validate_config(cfg)

        assert new_cfg.learning_rate == 0.00005

    def test_model_config_remap(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "model_config": {"model_type": "mistral"},
                }
            )
            | minimal_cfg
        )

        new_cfg = validate_config(cfg)
        assert new_cfg.overrides_of_model_config["model_type"] == "mistral"

    def test_model_type_remap(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "model_type": "AutoModelForCausalLM",
                }
            )
            | minimal_cfg
        )

        new_cfg = validate_config(cfg)
        assert new_cfg.type_of_model == "AutoModelForCausalLM"

    def test_model_revision_remap(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "model_revision": "main",
                }
            )
            | minimal_cfg
        )

        new_cfg = validate_config(cfg)
        assert new_cfg.revision_of_model == "main"

    def test_qlora(self, minimal_cfg):
        base_cfg = (
            DictDefault(
                {
                    "adapter": "qlora",
                }
            )
            | minimal_cfg
        )

        cfg = (
            DictDefault(  # pylint: disable=unsupported-binary-operation
                {
                    "load_in_8bit": True,
                }
            )
            | base_cfg
        )

        with pytest.raises(ValueError, match=r".*8bit.*"):
            validate_config(cfg)

        cfg = (
            DictDefault(  # pylint: disable=unsupported-binary-operation
                {
                    "gptq": True,
                }
            )
            | base_cfg
        )

        with pytest.raises(ValueError, match=r".*gptq.*"):
            validate_config(cfg)

        cfg = (
            DictDefault(  # pylint: disable=unsupported-binary-operation
                {
                    "load_in_4bit": False,
                }
            )
            | base_cfg
        )

        with pytest.raises(ValueError, match=r".*4bit.*"):
            validate_config(cfg)

        cfg = (
            DictDefault(  # pylint: disable=unsupported-binary-operation
                {
                    "load_in_4bit": True,
                }
            )
            | base_cfg
        )

        validate_config(cfg)

    def test_qlora_merge(self, minimal_cfg):
        base_cfg = (
            DictDefault(
                {
                    "adapter": "qlora",
                    "merge_lora": True,
                }
            )
            | minimal_cfg
        )

        cfg = (
            DictDefault(  # pylint: disable=unsupported-binary-operation
                {
                    "load_in_8bit": True,
                }
            )
            | base_cfg
        )

        with pytest.raises(ValueError, match=r".*8bit.*"):
            validate_config(cfg)

        cfg = (
            DictDefault(  # pylint: disable=unsupported-binary-operation
                {
                    "gptq": True,
                }
            )
            | base_cfg
        )

        with pytest.raises(ValueError, match=r".*gptq.*"):
            validate_config(cfg)

        cfg = (
            DictDefault(  # pylint: disable=unsupported-binary-operation
                {
                    "load_in_4bit": True,
                }
            )
            | base_cfg
        )

        with pytest.raises(ValueError, match=r".*4bit.*"):
            validate_config(cfg)

    def test_hf_use_auth_token(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "push_dataset_to_hub": "namespace/repo",
                }
            )
            | minimal_cfg
        )

        with pytest.raises(ValueError, match=r".*hf_use_auth_token.*"):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "push_dataset_to_hub": "namespace/repo",
                    "hf_use_auth_token": True,
                }
            )
            | minimal_cfg
        )
        validate_config(cfg)

    def test_gradient_accumulations_or_batch_size(self):
        cfg = DictDefault(
            {
                "base_model": "TinyLlama/TinyLlama-1.1B-Chat-v0.6",
                "learning_rate": 0.000001,
                "datasets": [
                    {
                        "path": "mhenrichsen/alpaca_2k_test",
                        "type": "alpaca",
                    }
                ],
                "gradient_accumulation_steps": 1,
                "batch_size": 1,
            }
        )

        with pytest.raises(
            ValueError, match=r".*gradient_accumulation_steps or batch_size.*"
        ):
            validate_config(cfg)

    def test_falcon_fsdp(self, minimal_cfg):
        regex_exp = r".*FSDP is not supported for falcon models.*"

        # Check for lower-case
        cfg = (
            DictDefault(
                {
                    "base_model": "tiiuae/falcon-7b",
                    "fsdp": ["full_shard", "auto_wrap"],
                }
            )
            | minimal_cfg
        )

        with pytest.raises(ValueError, match=regex_exp):
            validate_config(cfg)

        # Check for upper-case
        cfg = (
            DictDefault(
                {
                    "base_model": "Falcon-7b",
                    "fsdp": ["full_shard", "auto_wrap"],
                }
            )
            | minimal_cfg
        )

        with pytest.raises(ValueError, match=regex_exp):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "base_model": "tiiuae/falcon-7b",
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

    def test_mpt_gradient_checkpointing(self, minimal_cfg):
        regex_exp = r".*gradient_checkpointing is not supported for MPT models*"

        # Check for lower-case
        cfg = (
            DictDefault(
                {
                    "base_model": "mosaicml/mpt-7b",
                    "gradient_checkpointing": True,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(ValueError, match=regex_exp):
            validate_config(cfg)

    def test_flash_optimum(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "flash_optimum": True,
                    "adapter": "lora",
                    "bf16": False,
                }
            )
            | minimal_cfg
        )

        with self._caplog.at_level(logging.WARNING):
            validate_config(cfg)
            assert any(
                "BetterTransformers probably doesn't work with PEFT adapters"
                in record.message
                for record in self._caplog.records
            )

        cfg = (
            DictDefault(
                {
                    "flash_optimum": True,
                    "bf16": False,
                }
            )
            | minimal_cfg
        )

        with self._caplog.at_level(logging.WARNING):
            validate_config(cfg)
            assert any(
                "probably set bfloat16 or float16" in record.message
                for record in self._caplog.records
            )

        cfg = (
            DictDefault(
                {
                    "flash_optimum": True,
                    "fp16": True,
                }
            )
            | minimal_cfg
        )
        regex_exp = r".*AMP is not supported.*"

        with pytest.raises(ValueError, match=regex_exp):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "flash_optimum": True,
                    "bf16": True,
                }
            )
            | minimal_cfg
        )
        regex_exp = r".*AMP is not supported.*"

        with pytest.raises(ValueError, match=regex_exp):
            validate_config(cfg)

    def test_adamw_hyperparams(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "optimizer": None,
                    "adam_epsilon": 0.0001,
                }
            )
            | minimal_cfg
        )

        with self._caplog.at_level(logging.WARNING):
            validate_config(cfg)
            assert any(
                "adamw hyperparameters found, but no adamw optimizer set"
                in record.message
                for record in self._caplog.records
            )

        cfg = (
            DictDefault(
                {
                    "optimizer": "adafactor",
                    "adam_beta1": 0.0001,
                }
            )
            | minimal_cfg
        )

        with self._caplog.at_level(logging.WARNING):
            validate_config(cfg)
            assert any(
                "adamw hyperparameters found, but no adamw optimizer set"
                in record.message
                for record in self._caplog.records
            )

        cfg = (
            DictDefault(
                {
                    "optimizer": "adamw_bnb_8bit",
                    "adam_beta1": 0.9,
                    "adam_beta2": 0.99,
                    "adam_epsilon": 0.0001,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "optimizer": "adafactor",
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

    def test_deprecated_packing(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "max_packed_sequence_len": 1024,
                }
            )
            | minimal_cfg
        )
        with pytest.raises(
            DeprecationWarning,
            match=r"`max_packed_sequence_len` is no longer supported",
        ):
            validate_config(cfg)

    def test_packing(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "sample_packing": True,
                    "pad_to_sequence_len": None,
                }
            )
            | minimal_cfg
        )
        with self._caplog.at_level(logging.WARNING):
            validate_config(cfg)
            assert any(
                "`pad_to_sequence_len: true` is recommended when using sample_packing"
                in record.message
                for record in self._caplog.records
            )

    def test_merge_lora_no_bf16_fail(self, minimal_cfg):
        """
        This is assumed to be run on a CPU machine, so bf16 is not supported.
        """

        cfg = (
            DictDefault(
                {
                    "bf16": True,
                    "capabilities": {"bf16": False},
                }
            )
            | minimal_cfg
        )

        with pytest.raises(ValueError, match=r".*AMP is not supported on this GPU*"):
            AxolotlConfigWCapabilities(**cfg.to_dict())

        cfg = (
            DictDefault(
                {
                    "bf16": True,
                    "merge_lora": True,
                    "capabilities": {"bf16": False},
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

    def test_sharegpt_deprecation(self, minimal_cfg):
        cfg = (
            DictDefault(
                {"datasets": [{"path": "lorem/ipsum", "type": "sharegpt:chat"}]}
            )
            | minimal_cfg
        )
        with self._caplog.at_level(logging.WARNING):
            new_cfg = validate_config(cfg)
            assert any(
                "`type: sharegpt:chat` will soon be deprecated." in record.message
                for record in self._caplog.records
            )
        assert new_cfg.datasets[0].type == "sharegpt"

        cfg = (
            DictDefault(
                {
                    "datasets": [
                        {"path": "lorem/ipsum", "type": "sharegpt_simple:load_role"}
                    ]
                }
            )
            | minimal_cfg
        )
        with self._caplog.at_level(logging.WARNING):
            new_cfg = validate_config(cfg)
            assert any(
                "`type: sharegpt_simple` will soon be deprecated." in record.message
                for record in self._caplog.records
            )
        assert new_cfg.datasets[0].type == "sharegpt:load_role"

    def test_no_conflict_save_strategy(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "save_strategy": "epoch",
                    "save_steps": 10,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError, match=r".*save_strategy and save_steps mismatch.*"
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "save_strategy": "no",
                    "save_steps": 10,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError, match=r".*save_strategy and save_steps mismatch.*"
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "save_strategy": "steps",
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "save_strategy": "steps",
                    "save_steps": 10,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "save_steps": 10,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "save_strategy": "no",
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

    def test_no_conflict_eval_strategy(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "evaluation_strategy": "epoch",
                    "eval_steps": 10,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError, match=r".*evaluation_strategy and eval_steps mismatch.*"
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "evaluation_strategy": "no",
                    "eval_steps": 10,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError, match=r".*evaluation_strategy and eval_steps mismatch.*"
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "evaluation_strategy": "steps",
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "evaluation_strategy": "steps",
                    "eval_steps": 10,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "eval_steps": 10,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "evaluation_strategy": "no",
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "evaluation_strategy": "epoch",
                    "val_set_size": 0,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError,
            match=r".*eval_steps and evaluation_strategy are not supported with val_set_size == 0.*",
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "eval_steps": 10,
                    "val_set_size": 0,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError,
            match=r".*eval_steps and evaluation_strategy are not supported with val_set_size == 0.*",
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "val_set_size": 0,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "eval_steps": 10,
                    "val_set_size": 0.01,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "evaluation_strategy": "epoch",
                    "val_set_size": 0.01,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

    def test_eval_table_size_conflict_eval_packing(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "sample_packing": True,
                    "eval_table_size": 100,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError, match=r".*Please set 'eval_sample_packing' to false.*"
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "sample_packing": True,
                    "eval_sample_packing": False,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "sample_packing": False,
                    "eval_table_size": 100,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "sample_packing": True,
                    "eval_table_size": 100,
                    "eval_sample_packing": False,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

    def test_load_in_x_bit_without_adapter(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "load_in_4bit": True,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError,
            match=r".*load_in_8bit and load_in_4bit are not supported without setting an adapter.*",
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "load_in_8bit": True,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError,
            match=r".*load_in_8bit and load_in_4bit are not supported without setting an adapter.*",
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "load_in_4bit": True,
                    "adapter": "qlora",
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "load_in_8bit": True,
                    "adapter": "lora",
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

    def test_warmup_step_no_conflict(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "warmup_steps": 10,
                    "warmup_ratio": 0.1,
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError,
            match=r".*warmup_steps and warmup_ratio are mutually exclusive*",
        ):
            validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "warmup_steps": 10,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

        cfg = (
            DictDefault(
                {
                    "warmup_ratio": 0.1,
                }
            )
            | minimal_cfg
        )

        validate_config(cfg)

    def test_unfrozen_parameters_w_peft_layers_to_transform(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "adapter": "lora",
                    "unfrozen_parameters": [
                        "model.layers.2[0-9]+.block_sparse_moe.gate.*"
                    ],
                    "peft_layers_to_transform": [0, 1],
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError,
            match=r".*can have unexpected behavior*",
        ):
            validate_config(cfg)

    def test_hub_model_id_save_value_warns(self, minimal_cfg):
        cfg = DictDefault({"hub_model_id": "test"}) | minimal_cfg

        with self._caplog.at_level(logging.WARNING):
            validate_config(cfg)
            assert (
                "set without any models being saved" in self._caplog.records[0].message
            )

    def test_hub_model_id_save_value(self, minimal_cfg):
        cfg = DictDefault({"hub_model_id": "test", "saves_per_epoch": 4}) | minimal_cfg

        with self._caplog.at_level(logging.WARNING):
            validate_config(cfg)
            assert len(self._caplog.records) == 0


class TestValidationCheckModelConfig(BaseValidation):
    """
    Test the validation for the config when the model config is available
    """

    def test_llama_add_tokens_adapter(self, minimal_cfg):
        cfg = (
            DictDefault(
                {"adapter": "qlora", "load_in_4bit": True, "tokens": ["<|imstart|>"]}
            )
            | minimal_cfg
        )
        model_config = DictDefault({"model_type": "llama"})

        with pytest.raises(
            ValueError,
            match=r".*`lora_modules_to_save` not properly set when adding new tokens*",
        ):
            check_model_config(cfg, model_config)

        cfg = (
            DictDefault(
                {
                    "adapter": "qlora",
                    "load_in_4bit": True,
                    "tokens": ["<|imstart|>"],
                    "lora_modules_to_save": ["embed_tokens"],
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError,
            match=r".*`lora_modules_to_save` not properly set when adding new tokens*",
        ):
            check_model_config(cfg, model_config)

        cfg = (
            DictDefault(
                {
                    "adapter": "qlora",
                    "load_in_4bit": True,
                    "tokens": ["<|imstart|>"],
                    "lora_modules_to_save": ["embed_tokens", "lm_head"],
                }
            )
            | minimal_cfg
        )

        check_model_config(cfg, model_config)

    def test_phi_add_tokens_adapter(self, minimal_cfg):
        cfg = (
            DictDefault(
                {"adapter": "qlora", "load_in_4bit": True, "tokens": ["<|imstart|>"]}
            )
            | minimal_cfg
        )
        model_config = DictDefault({"model_type": "phi"})

        with pytest.raises(
            ValueError,
            match=r".*`lora_modules_to_save` not properly set when adding new tokens*",
        ):
            check_model_config(cfg, model_config)

        cfg = (
            DictDefault(
                {
                    "adapter": "qlora",
                    "load_in_4bit": True,
                    "tokens": ["<|imstart|>"],
                    "lora_modules_to_save": ["embd.wte", "lm_head.linear"],
                }
            )
            | minimal_cfg
        )

        with pytest.raises(
            ValueError,
            match=r".*`lora_modules_to_save` not properly set when adding new tokens*",
        ):
            check_model_config(cfg, model_config)

        cfg = (
            DictDefault(
                {
                    "adapter": "qlora",
                    "load_in_4bit": True,
                    "tokens": ["<|imstart|>"],
                    "lora_modules_to_save": ["embed_tokens", "lm_head"],
                }
            )
            | minimal_cfg
        )

        check_model_config(cfg, model_config)


class TestValidationWandb(BaseValidation):
    """
    Validation test for wandb
    """

    def test_wandb_set_run_id_to_name(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "wandb_run_id": "foo",
                }
            )
            | minimal_cfg
        )

        with self._caplog.at_level(logging.WARNING):
            new_cfg = validate_config(cfg)
            assert any(
                "wandb_run_id sets the ID of the run. If you would like to set the name, please use wandb_name instead."
                in record.message
                for record in self._caplog.records
            )

            assert new_cfg.wandb_name == "foo" and new_cfg.wandb_run_id == "foo"

        cfg = (
            DictDefault(
                {
                    "wandb_name": "foo",
                }
            )
            | minimal_cfg
        )

        new_cfg = validate_config(cfg)

        assert new_cfg.wandb_name == "foo" and new_cfg.wandb_run_id is None

    def test_wandb_sets_env(self, minimal_cfg):
        cfg = (
            DictDefault(
                {
                    "wandb_project": "foo",
                    "wandb_name": "bar",
                    "wandb_run_id": "bat",
                    "wandb_entity": "baz",
                    "wandb_mode": "online",
                    "wandb_watch": "false",
                    "wandb_log_model": "checkpoint",
                }
            )
            | minimal_cfg
        )

        new_cfg = validate_config(cfg)

        setup_wandb_env_vars(new_cfg)

        assert os.environ.get("WANDB_PROJECT", "") == "foo"
        assert os.environ.get("WANDB_NAME", "") == "bar"
        assert os.environ.get("WANDB_RUN_ID", "") == "bat"
        assert os.environ.get("WANDB_ENTITY", "") == "baz"
        assert os.environ.get("WANDB_MODE", "") == "online"
        assert os.environ.get("WANDB_WATCH", "") == "false"
        assert os.environ.get("WANDB_LOG_MODEL", "") == "checkpoint"
        assert os.environ.get("WANDB_DISABLED", "") != "true"

        os.environ.pop("WANDB_PROJECT", None)
        os.environ.pop("WANDB_NAME", None)
        os.environ.pop("WANDB_RUN_ID", None)
        os.environ.pop("WANDB_ENTITY", None)
        os.environ.pop("WANDB_MODE", None)
        os.environ.pop("WANDB_WATCH", None)
        os.environ.pop("WANDB_LOG_MODEL", None)
        os.environ.pop("WANDB_DISABLED", None)

    def test_wandb_set_disabled(self, minimal_cfg):
        cfg = DictDefault({}) | minimal_cfg

        new_cfg = validate_config(cfg)

        setup_wandb_env_vars(new_cfg)

        assert os.environ.get("WANDB_DISABLED", "") == "true"

        cfg = (
            DictDefault(
                {
                    "wandb_project": "foo",
                }
            )
            | minimal_cfg
        )

        new_cfg = validate_config(cfg)

        setup_wandb_env_vars(new_cfg)

        assert os.environ.get("WANDB_DISABLED", "") != "true"

        os.environ.pop("WANDB_PROJECT", None)
        os.environ.pop("WANDB_DISABLED", None)