Hugging Face's logo

unlearning-cleanslate
/
gemma-3-12b-simnpo-baseline

Safetensors
gemma3
Model card Files
xet
 Community
arpandeepk commited on
Commit
f480de5
·
verified ·
1 Parent(s): e1c9861

upload simnpo_gemma-3-12b-pt_20260427_203900

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +10 -0
  2. checkpoint-1000/config.json +122 -0
  3. checkpoint-1000/generation_config.json +14 -0
  4. checkpoint-1000/model.safetensors +3 -0
  5. checkpoint-1000/optimizer.pt +3 -0
  6. checkpoint-1000/rng_state.pth +3 -0
  7. checkpoint-1000/scheduler.pt +3 -0
  8. checkpoint-1000/tokenizer.json +3 -0
  9. checkpoint-1000/tokenizer_config.json +25 -0
  10. checkpoint-1000/trainer_state.json +3141 -0
  11. checkpoint-1000/training_args.bin +3 -0
  12. checkpoint-1200/config.json +122 -0
  13. checkpoint-1200/generation_config.json +14 -0
  14. checkpoint-1200/model.safetensors +3 -0
  15. checkpoint-1200/optimizer.pt +3 -0
  16. checkpoint-1200/rng_state.pth +3 -0
  17. checkpoint-1200/scheduler.pt +3 -0
  18. checkpoint-1200/tokenizer.json +3 -0
  19. checkpoint-1200/tokenizer_config.json +25 -0
  20. checkpoint-1200/trainer_state.json +3761 -0
  21. checkpoint-1200/training_args.bin +3 -0
  22. checkpoint-1400/config.json +122 -0
  23. checkpoint-1400/generation_config.json +14 -0
  24. checkpoint-1400/model.safetensors +3 -0
  25. checkpoint-1400/optimizer.pt +3 -0
  26. checkpoint-1400/rng_state.pth +3 -0
  27. checkpoint-1400/scheduler.pt +3 -0
  28. checkpoint-1400/tokenizer.json +3 -0
  29. checkpoint-1400/tokenizer_config.json +25 -0
  30. checkpoint-1400/trainer_state.json +0 -0
  31. checkpoint-1400/training_args.bin +3 -0
  32. checkpoint-1600/config.json +122 -0
  33. checkpoint-1600/generation_config.json +14 -0
  34. checkpoint-1600/model.safetensors +3 -0
  35. checkpoint-1600/optimizer.pt +3 -0
  36. checkpoint-1600/rng_state.pth +3 -0
  37. checkpoint-1600/scheduler.pt +3 -0
  38. checkpoint-1600/tokenizer.json +3 -0
  39. checkpoint-1600/tokenizer_config.json +25 -0
  40. checkpoint-1600/trainer_state.json +0 -0
  41. checkpoint-1600/training_args.bin +3 -0
  42. checkpoint-1746/config.json +122 -0
  43. checkpoint-1746/generation_config.json +14 -0
  44. checkpoint-1746/model.safetensors +3 -0
  45. checkpoint-1746/optimizer.pt +3 -0
  46. checkpoint-1746/rng_state.pth +3 -0
  47. checkpoint-1746/scheduler.pt +3 -0
  48. checkpoint-1746/tokenizer.json +3 -0
  49. checkpoint-1746/tokenizer_config.json +25 -0
  50. checkpoint-1746/trainer_state.json +0 -0
.gitattributes CHANGED
@@ -33,3 +33,13 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ checkpoint-1000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
+ checkpoint-1200/tokenizer.json filter=lfs diff=lfs merge=lfs -text
38
+ checkpoint-1400/tokenizer.json filter=lfs diff=lfs merge=lfs -text
39
+ checkpoint-1600/tokenizer.json filter=lfs diff=lfs merge=lfs -text
40
+ checkpoint-1746/tokenizer.json filter=lfs diff=lfs merge=lfs -text
41
+ checkpoint-200/tokenizer.json filter=lfs diff=lfs merge=lfs -text
42
+ checkpoint-400/tokenizer.json filter=lfs diff=lfs merge=lfs -text
43
+ checkpoint-600/tokenizer.json filter=lfs diff=lfs merge=lfs -text
44
+ checkpoint-800/tokenizer.json filter=lfs diff=lfs merge=lfs -text
45
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
checkpoint-1000/config.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "bos_token_id": 2,
7
+ "dtype": "bfloat16",
8
+ "eoi_token_index": 256000,
9
+ "eos_token_id": 1,
10
+ "image_token_index": 262144,
11
+ "initializer_range": 0.02,
12
+ "mm_tokens_per_image": 256,
13
+ "model_type": "gemma3",
14
+ "pad_token_id": 0,
15
+ "text_config": {
16
+ "_sliding_window_pattern": 6,
17
+ "attention_bias": false,
18
+ "attention_dropout": 0.0,
19
+ "attn_logit_softcapping": null,
20
+ "bos_token_id": 2,
21
+ "dtype": "bfloat16",
22
+ "eos_token_id": 1,
23
+ "final_logit_softcapping": null,
24
+ "head_dim": 256,
25
+ "hidden_activation": "gelu_pytorch_tanh",
26
+ "hidden_size": 3840,
27
+ "initializer_range": 0.02,
28
+ "intermediate_size": 15360,
29
+ "layer_types": [
30
+ "sliding_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "full_attention",
36
+ "sliding_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "sliding_attention",
40
+ "sliding_attention",
41
+ "full_attention",
42
+ "sliding_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "sliding_attention",
46
+ "sliding_attention",
47
+ "full_attention",
48
+ "sliding_attention",
49
+ "sliding_attention",
50
+ "sliding_attention",
51
+ "sliding_attention",
52
+ "sliding_attention",
53
+ "full_attention",
54
+ "sliding_attention",
55
+ "sliding_attention",
56
+ "sliding_attention",
57
+ "sliding_attention",
58
+ "sliding_attention",
59
+ "full_attention",
60
+ "sliding_attention",
61
+ "sliding_attention",
62
+ "sliding_attention",
63
+ "sliding_attention",
64
+ "sliding_attention",
65
+ "full_attention",
66
+ "sliding_attention",
67
+ "sliding_attention",
68
+ "sliding_attention",
69
+ "sliding_attention",
70
+ "sliding_attention",
71
+ "full_attention",
72
+ "sliding_attention",
73
+ "sliding_attention",
74
+ "sliding_attention",
75
+ "sliding_attention",
76
+ "sliding_attention",
77
+ "full_attention"
78
+ ],
79
+ "max_position_embeddings": 131072,
80
+ "model_type": "gemma3_text",
81
+ "num_attention_heads": 16,
82
+ "num_hidden_layers": 48,
83
+ "num_key_value_heads": 8,
84
+ "pad_token_id": 0,
85
+ "query_pre_attn_scalar": 256,
86
+ "rms_norm_eps": 1e-06,
87
+ "rope_parameters": {
88
+ "full_attention": {
89
+ "factor": 8.0,
90
+ "rope_theta": 1000000.0,
91
+ "rope_type": "linear"
92
+ },
93
+ "sliding_attention": {
94
+ "rope_theta": 10000.0,
95
+ "rope_type": "default"
96
+ }
97
+ },
98
+ "sliding_window": 1024,
99
+ "tie_word_embeddings": true,
100
+ "use_bidirectional_attention": false,
101
+ "use_cache": true,
102
+ "vocab_size": 262208
103
+ },
104
+ "tie_word_embeddings": true,
105
+ "transformers_version": "5.6.0",
106
+ "use_cache": false,
107
+ "vision_config": {
108
+ "attention_dropout": 0.0,
109
+ "dtype": "bfloat16",
110
+ "hidden_act": "gelu_pytorch_tanh",
111
+ "hidden_size": 1152,
112
+ "image_size": 896,
113
+ "intermediate_size": 4304,
114
+ "layer_norm_eps": 1e-06,
115
+ "model_type": "siglip_vision_model",
116
+ "num_attention_heads": 16,
117
+ "num_channels": 3,
118
+ "num_hidden_layers": 27,
119
+ "patch_size": 14,
120
+ "vision_use_head": false
121
+ }
122
+ }
checkpoint-1000/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "cache_implementation": "hybrid",
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 1,
7
+ 1,
8
+ 106
9
+ ],
10
+ "pad_token_id": 0,
11
+ "top_k": 64,
12
+ "top_p": 0.95,
13
+ "transformers_version": "5.6.0"
14
+ }
checkpoint-1000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a2c662a233b11db43d6169e65e53101e68a0e964fa18324c569223bbe1562c2
3
+ size 24374794824
checkpoint-1000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e430e983b4ed23fe593a57275c26d5074aedbdce31987894acdc03689c7f6b5a
3
+ size 47064687765
checkpoint-1000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfcd3156281d94ef5b5d5f3408b76801576ff242a500f78ee7918e0ad564348f
3
+ size 14645
checkpoint-1000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d85c4803dcd4c1ce3eca614857a9057cc4bb3028b9db257ff602b7ceff24c6b
3
+ size 1465
checkpoint-1000/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:daab2354f8a74e70d70b4d1f804939b68a8c9624dd06cb7858e52dd8970e9726
3
+ size 33384567
checkpoint-1000/tokenizer_config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "boi_token": "<start_of_image>",
4
+ "bos_token": "<bos>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eoi_token": "<end_of_image>",
7
+ "eos_token": "<eos>",
8
+ "image_token": "<image_soft_token>",
9
+ "is_local": false,
10
+ "local_files_only": false,
11
+ "mask_token": "<mask>",
12
+ "model_max_length": 1000000000000000019884624838656,
13
+ "model_specific_special_tokens": {
14
+ "boi_token": "<start_of_image>",
15
+ "eoi_token": "<end_of_image>",
16
+ "image_token": "<image_soft_token>"
17
+ },
18
+ "pad_token": "<pad>",
19
+ "processor_class": "Gemma3Processor",
20
+ "sp_model_kwargs": null,
21
+ "spaces_between_special_tokens": false,
22
+ "tokenizer_class": "GemmaTokenizer",
23
+ "unk_token": "<unk>",
24
+ "use_default_system_prompt": false
25
+ }
checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,3141 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.145475372279496,
6
+ "eval_steps": 500,
7
+ "global_step": 1000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0,
14
+ "forget_loss": 3.44928662343591e-06,
15
+ "retain_loss": 6.609266757965088,
16
+ "step": 0
17
+ },
18
+ {
19
+ "epoch": 0,
20
+ "forget_loss": 0.038371894508600235,
21
+ "retain_loss": 6.274988651275635,
22
+ "step": 0
23
+ },
24
+ {
25
+ "epoch": 0,
26
+ "forget_loss": 0.001581120421178639,
27
+ "retain_loss": 6.232261657714844,
28
+ "step": 0
29
+ },
30
+ {
31
+ "epoch": 0,
32
+ "forget_loss": 8.447099389741197e-06,
33
+ "retain_loss": 6.230651378631592,
34
+ "step": 0
35
+ },
36
+ {
37
+ "epoch": 0.001145475372279496,
38
+ "grad_norm": 282.0,
39
+ "learning_rate": 0.0,
40
+ "loss": 6.386748313903809,
41
+ "step": 1
42
+ },
43
+ {
44
+ "epoch": 0.011454753722794959,
45
+ "grad_norm": 9.4375,
46
+ "learning_rate": 4.5e-05,
47
+ "loss": 3.339502970377604,
48
+ "step": 10
49
+ },
50
+ {
51
+ "epoch": 0.011454753722794959,
52
+ "forget_loss": 2.795343243633397e-05,
53
+ "retain_loss": 1.8692482709884644,
54
+ "step": 10
55
+ },
56
+ {
57
+ "epoch": 0.011454753722794959,
58
+ "forget_loss": 5.640117888106033e-05,
59
+ "retain_loss": 2.053846597671509,
60
+ "step": 10
61
+ },
62
+ {
63
+ "epoch": 0.011454753722794959,
64
+ "forget_loss": 3.110460966126993e-05,
65
+ "retain_loss": 2.0232813358306885,
66
+ "step": 10
67
+ },
68
+ {
69
+ "epoch": 0.011454753722794959,
70
+ "forget_loss": 4.845933744945796e-06,
71
+ "retain_loss": 2.4105024337768555,
72
+ "step": 10
73
+ },
74
+ {
75
+ "epoch": 0.022909507445589918,
76
+ "grad_norm": 38.5,
77
+ "learning_rate": 4.9740783410138256e-05,
78
+ "loss": 2.1140087127685545,
79
+ "step": 20
80
+ },
81
+ {
82
+ "epoch": 0.022909507445589918,
83
+ "forget_loss": 1.082867765944684e-05,
84
+ "retain_loss": 2.002471446990967,
85
+ "step": 20
86
+ },
87
+ {
88
+ "epoch": 0.022909507445589918,
89
+ "forget_loss": 3.829125489573926e-06,
90
+ "retain_loss": 2.0777769088745117,
91
+ "step": 20
92
+ },
93
+ {
94
+ "epoch": 0.022909507445589918,
95
+ "forget_loss": 2.8280664992053062e-05,
96
+ "retain_loss": 2.1336162090301514,
97
+ "step": 20
98
+ },
99
+ {
100
+ "epoch": 0.022909507445589918,
101
+ "forget_loss": 1.5830282791284844e-05,
102
+ "retain_loss": 1.7852996587753296,
103
+ "step": 20
104
+ },
105
+ {
106
+ "epoch": 0.03436426116838488,
107
+ "grad_norm": 5.6875,
108
+ "learning_rate": 4.945276497695853e-05,
109
+ "loss": 1.8668533325195313,
110
+ "step": 30
111
+ },
112
+ {
113
+ "epoch": 0.03436426116838488,
114
+ "forget_loss": 1.3028584362473339e-05,
115
+ "retain_loss": 1.870590090751648,
116
+ "step": 30
117
+ },
118
+ {
119
+ "epoch": 0.03436426116838488,
120
+ "forget_loss": 9.773620149644557e-06,
121
+ "retain_loss": 2.1787960529327393,
122
+ "step": 30
123
+ },
124
+ {
125
+ "epoch": 0.03436426116838488,
126
+ "forget_loss": 2.0277722796890885e-05,
127
+ "retain_loss": 1.7472785711288452,
128
+ "step": 30
129
+ },
130
+ {
131
+ "epoch": 0.03436426116838488,
132
+ "forget_loss": 1.0735096111602616e-05,
133
+ "retain_loss": 2.227604866027832,
134
+ "step": 30
135
+ },
136
+ {
137
+ "epoch": 0.045819014891179836,
138
+ "grad_norm": 5.4375,
139
+ "learning_rate": 4.916474654377881e-05,
140
+ "loss": 1.695387840270996,
141
+ "step": 40
142
+ },
143
+ {
144
+ "epoch": 0.045819014891179836,
145
+ "forget_loss": 5.423281436378602e-06,
146
+ "retain_loss": 1.2835420370101929,
147
+ "step": 40
148
+ },
149
+ {
150
+ "epoch": 0.045819014891179836,
151
+ "forget_loss": 9.37562845138018e-07,
152
+ "retain_loss": 0.8797236680984497,
153
+ "step": 40
154
+ },
155
+ {
156
+ "epoch": 0.045819014891179836,
157
+ "forget_loss": 4.068289854330942e-06,
158
+ "retain_loss": 1.9588595628738403,
159
+ "step": 40
160
+ },
161
+ {
162
+ "epoch": 0.045819014891179836,
163
+ "forget_loss": 2.925765329564456e-06,
164
+ "retain_loss": 1.7313635349273682,
165
+ "step": 40
166
+ },
167
+ {
168
+ "epoch": 0.0572737686139748,
169
+ "grad_norm": 24.5,
170
+ "learning_rate": 4.887672811059908e-05,
171
+ "loss": 1.3742843627929688,
172
+ "step": 50
173
+ },
174
+ {
175
+ "epoch": 0.0572737686139748,
176
+ "forget_loss": 1.3801378372590989e-05,
177
+ "retain_loss": 1.4735946655273438,
178
+ "step": 50
179
+ },
180
+ {
181
+ "epoch": 0.0572737686139748,
182
+ "forget_loss": 8.779801646596752e-06,
183
+ "retain_loss": 1.0075594186782837,
184
+ "step": 50
185
+ },
186
+ {
187
+ "epoch": 0.0572737686139748,
188
+ "forget_loss": 1.5394653019029647e-05,
189
+ "retain_loss": 1.8036051988601685,
190
+ "step": 50
191
+ },
192
+ {
193
+ "epoch": 0.0572737686139748,
194
+ "forget_loss": 8.239669114118442e-06,
195
+ "retain_loss": 1.379029393196106,
196
+ "step": 50
197
+ },
198
+ {
199
+ "epoch": 0.06872852233676977,
200
+ "grad_norm": 4.375,
201
+ "learning_rate": 4.858870967741936e-05,
202
+ "loss": 1.3121252059936523,
203
+ "step": 60
204
+ },
205
+ {
206
+ "epoch": 0.06872852233676977,
207
+ "forget_loss": 3.8279208638414275e-06,
208
+ "retain_loss": 1.7327319383621216,
209
+ "step": 60
210
+ },
211
+ {
212
+ "epoch": 0.06872852233676977,
213
+ "forget_loss": 3.3966937280638376e-06,
214
+ "retain_loss": 1.0556788444519043,
215
+ "step": 60
216
+ },
217
+ {
218
+ "epoch": 0.06872852233676977,
219
+ "forget_loss": 3.698090495163342e-06,
220
+ "retain_loss": 0.9602264761924744,
221
+ "step": 60
222
+ },
223
+ {
224
+ "epoch": 0.06872852233676977,
225
+ "forget_loss": 4.131216337555088e-05,
226
+ "retain_loss": 1.1526304483413696,
227
+ "step": 60
228
+ },
229
+ {
230
+ "epoch": 0.08018327605956473,
231
+ "grad_norm": 4.5,
232
+ "learning_rate": 4.830069124423963e-05,
233
+ "loss": 1.1931480407714843,
234
+ "step": 70
235
+ },
236
+ {
237
+ "epoch": 0.08018327605956473,
238
+ "forget_loss": 3.9989358810998965e-06,
239
+ "retain_loss": 1.1984384059906006,
240
+ "step": 70
241
+ },
242
+ {
243
+ "epoch": 0.08018327605956473,
244
+ "forget_loss": 2.2779448045184836e-05,
245
+ "retain_loss": 1.3697820901870728,
246
+ "step": 70
247
+ },
248
+ {
249
+ "epoch": 0.08018327605956473,
250
+ "forget_loss": 7.756965715088882e-06,
251
+ "retain_loss": 0.8010083436965942,
252
+ "step": 70
253
+ },
254
+ {
255
+ "epoch": 0.08018327605956473,
256
+ "forget_loss": 1.9700873963302e-06,
257
+ "retain_loss": 1.1254682540893555,
258
+ "step": 70
259
+ },
260
+ {
261
+ "epoch": 0.09163802978235967,
262
+ "grad_norm": 12.625,
263
+ "learning_rate": 4.801267281105991e-05,
264
+ "loss": 1.0133052825927735,
265
+ "step": 80
266
+ },
267
+ {
268
+ "epoch": 0.09163802978235967,
269
+ "forget_loss": 7.642759555892553e-06,
270
+ "retain_loss": 0.8252509236335754,
271
+ "step": 80
272
+ },
273
+ {
274
+ "epoch": 0.09163802978235967,
275
+ "forget_loss": 8.59803822095273e-06,
276
+ "retain_loss": 0.7506107687950134,
277
+ "step": 80
278
+ },
279
+ {
280
+ "epoch": 0.09163802978235967,
281
+ "forget_loss": 4.804069703823188e-06,
282
+ "retain_loss": 0.7315255403518677,
283
+ "step": 80
284
+ },
285
+ {
286
+ "epoch": 0.09163802978235967,
287
+ "forget_loss": 4.970052486896748e-06,
288
+ "retain_loss": 0.7012558579444885,
289
+ "step": 80
290
+ },
291
+ {
292
+ "epoch": 0.10309278350515463,
293
+ "grad_norm": 4.0625,
294
+ "learning_rate": 4.772465437788018e-05,
295
+ "loss": 0.9705428123474121,
296
+ "step": 90
297
+ },
298
+ {
299
+ "epoch": 0.10309278350515463,
300
+ "forget_loss": 1.4775166619074298e-06,
301
+ "retain_loss": 0.7782251834869385,
302
+ "step": 90
303
+ },
304
+ {
305
+ "epoch": 0.10309278350515463,
306
+ "forget_loss": 1.5886946584942052e-06,
307
+ "retain_loss": 0.8455419540405273,
308
+ "step": 90
309
+ },
310
+ {
311
+ "epoch": 0.10309278350515463,
312
+ "forget_loss": 2.0495062926784158e-06,
313
+ "retain_loss": 0.6329978704452515,
314
+ "step": 90
315
+ },
316
+ {
317
+ "epoch": 0.10309278350515463,
318
+ "forget_loss": 3.828020635410212e-05,
319
+ "retain_loss": 0.8053420186042786,
320
+ "step": 90
321
+ },
322
+ {
323
+ "epoch": 0.1145475372279496,
324
+ "grad_norm": 11.0,
325
+ "learning_rate": 4.743663594470046e-05,
326
+ "loss": 0.8036625862121582,
327
+ "step": 100
328
+ },
329
+ {
330
+ "epoch": 0.1145475372279496,
331
+ "forget_loss": 4.424137387104565e-06,
332
+ "retain_loss": 0.949962317943573,
333
+ "step": 100
334
+ },
335
+ {
336
+ "epoch": 0.1145475372279496,
337
+ "forget_loss": 2.453501565469196e-06,
338
+ "retain_loss": 0.8932625651359558,
339
+ "step": 100
340
+ },
341
+ {
342
+ "epoch": 0.1145475372279496,
343
+ "forget_loss": 1.6591504845564486e-06,
344
+ "retain_loss": 1.121445894241333,
345
+ "step": 100
346
+ },
347
+ {
348
+ "epoch": 0.1145475372279496,
349
+ "forget_loss": 4.8642609726812225e-06,
350
+ "retain_loss": 0.37143388390541077,
351
+ "step": 100
352
+ },
353
+ {
354
+ "epoch": 0.12600229095074456,
355
+ "grad_norm": 3.453125,
356
+ "learning_rate": 4.7148617511520735e-05,
357
+ "loss": 0.6938334941864014,
358
+ "step": 110
359
+ },
360
+ {
361
+ "epoch": 0.12600229095074456,
362
+ "forget_loss": 4.921608706354164e-06,
363
+ "retain_loss": 0.3767385482788086,
364
+ "step": 110
365
+ },
366
+ {
367
+ "epoch": 0.12600229095074456,
368
+ "forget_loss": 4.79454456581152e-07,
369
+ "retain_loss": 0.6782515048980713,
370
+ "step": 110
371
+ },
372
+ {
373
+ "epoch": 0.12600229095074456,
374
+ "forget_loss": 3.918839865946211e-06,
375
+ "retain_loss": 0.8204460144042969,
376
+ "step": 110
377
+ },
378
+ {
379
+ "epoch": 0.12600229095074456,
380
+ "forget_loss": 4.693280516221421e-06,
381
+ "retain_loss": 0.841111958026886,
382
+ "step": 110
383
+ },
384
+ {
385
+ "epoch": 0.13745704467353953,
386
+ "grad_norm": 3.4375,
387
+ "learning_rate": 4.6860599078341014e-05,
388
+ "loss": 0.6417326927185059,
389
+ "step": 120
390
+ },
391
+ {
392
+ "epoch": 0.13745704467353953,
393
+ "forget_loss": 1.713265555736143e-05,
394
+ "retain_loss": 0.3679219186306,
395
+ "step": 120
396
+ },
397
+ {
398
+ "epoch": 0.13745704467353953,
399
+ "forget_loss": 8.306015843118075e-06,
400
+ "retain_loss": 0.6136802434921265,
401
+ "step": 120
402
+ },
403
+ {
404
+ "epoch": 0.13745704467353953,
405
+ "forget_loss": 2.449452495056903e-06,
406
+ "retain_loss": 1.2731798887252808,
407
+ "step": 120
408
+ },
409
+ {
410
+ "epoch": 0.13745704467353953,
411
+ "forget_loss": 4.083998192072613e-06,
412
+ "retain_loss": 0.5807819962501526,
413
+ "step": 120
414
+ },
415
+ {
416
+ "epoch": 0.14891179839633448,
417
+ "grad_norm": 4.03125,
418
+ "learning_rate": 4.657258064516129e-05,
419
+ "loss": 0.5295979499816894,
420
+ "step": 130
421
+ },
422
+ {
423
+ "epoch": 0.14891179839633448,
424
+ "forget_loss": 8.430975867668167e-06,
425
+ "retain_loss": 0.7869871854782104,
426
+ "step": 130
427
+ },
428
+ {
429
+ "epoch": 0.14891179839633448,
430
+ "forget_loss": 9.652731023379602e-06,
431
+ "retain_loss": 0.9147917628288269,
432
+ "step": 130
433
+ },
434
+ {
435
+ "epoch": 0.14891179839633448,
436
+ "forget_loss": 1.1524152796482667e-06,
437
+ "retain_loss": 0.48021697998046875,
438
+ "step": 130
439
+ },
440
+ {
441
+ "epoch": 0.14891179839633448,
442
+ "forget_loss": 2.4387713892792817e-06,
443
+ "retain_loss": 0.8574126362800598,
444
+ "step": 130
445
+ },
446
+ {
447
+ "epoch": 0.16036655211912945,
448
+ "grad_norm": 4.5625,
449
+ "learning_rate": 4.628456221198157e-05,
450
+ "loss": 0.5009561061859131,
451
+ "step": 140
452
+ },
453
+ {
454
+ "epoch": 0.16036655211912945,
455
+ "forget_loss": 5.176518243388273e-06,
456
+ "retain_loss": 0.2740374803543091,
457
+ "step": 140
458
+ },
459
+ {
460
+ "epoch": 0.16036655211912945,
461
+ "forget_loss": 4.7920007091306616e-06,
462
+ "retain_loss": 0.6662377119064331,
463
+ "step": 140
464
+ },
465
+ {
466
+ "epoch": 0.16036655211912945,
467
+ "forget_loss": 2.156430582544999e-06,
468
+ "retain_loss": 0.3841518461704254,
469
+ "step": 140
470
+ },
471
+ {
472
+ "epoch": 0.16036655211912945,
473
+ "forget_loss": 1.6250938642770052e-05,
474
+ "retain_loss": 0.3435481786727905,
475
+ "step": 140
476
+ },
477
+ {
478
+ "epoch": 0.1718213058419244,
479
+ "grad_norm": 2.9375,
480
+ "learning_rate": 4.599654377880185e-05,
481
+ "loss": 0.5061944961547852,
482
+ "step": 150
483
+ },
484
+ {
485
+ "epoch": 0.1718213058419244,
486
+ "forget_loss": 2.413504262221977e-06,
487
+ "retain_loss": 0.4674696624279022,
488
+ "step": 150
489
+ },
490
+ {
491
+ "epoch": 0.1718213058419244,
492
+ "forget_loss": 9.788258466869593e-06,
493
+ "retain_loss": 0.32738229632377625,
494
+ "step": 150
495
+ },
496
+ {
497
+ "epoch": 0.1718213058419244,
498
+ "forget_loss": 1.659399458731059e-06,
499
+ "retain_loss": 0.4860354959964752,
500
+ "step": 150
501
+ },
502
+ {
503
+ "epoch": 0.1718213058419244,
504
+ "forget_loss": 6.014466293891019e-07,
505
+ "retain_loss": 0.25174200534820557,
506
+ "step": 150
507
+ },
508
+ {
509
+ "epoch": 0.18327605956471935,
510
+ "grad_norm": 3.140625,
511
+ "learning_rate": 4.5708525345622124e-05,
512
+ "loss": 0.33492887020111084,
513
+ "step": 160
514
+ },
515
+ {
516
+ "epoch": 0.18327605956471935,
517
+ "forget_loss": 4.615132638718933e-06,
518
+ "retain_loss": 0.30095958709716797,
519
+ "step": 160
520
+ },
521
+ {
522
+ "epoch": 0.18327605956471935,
523
+ "forget_loss": 1.7731171055856976e-06,
524
+ "retain_loss": 0.19373854994773865,
525
+ "step": 160
526
+ },
527
+ {
528
+ "epoch": 0.18327605956471935,
529
+ "forget_loss": 2.2210085717233596e-06,
530
+ "retain_loss": 0.23821769654750824,
531
+ "step": 160
532
+ },
533
+ {
534
+ "epoch": 0.18327605956471935,
535
+ "forget_loss": 1.3112939996062778e-06,
536
+ "retain_loss": 0.3393084406852722,
537
+ "step": 160
538
+ },
539
+ {
540
+ "epoch": 0.19473081328751432,
541
+ "grad_norm": 2.671875,
542
+ "learning_rate": 4.54205069124424e-05,
543
+ "loss": 0.3298675775527954,
544
+ "step": 170
545
+ },
546
+ {
547
+ "epoch": 0.19473081328751432,
548
+ "forget_loss": 4.927760528516956e-06,
549
+ "retain_loss": 0.44287964701652527,
550
+ "step": 170
551
+ },
552
+ {
553
+ "epoch": 0.19473081328751432,
554
+ "forget_loss": 4.030034688184969e-06,
555
+ "retain_loss": 0.4538065195083618,
556
+ "step": 170
557
+ },
558
+ {
559
+ "epoch": 0.19473081328751432,
560
+ "forget_loss": 2.3950892682478298e-06,
561
+ "retain_loss": 0.31112024188041687,
562
+ "step": 170
563
+ },
564
+ {
565
+ "epoch": 0.19473081328751432,
566
+ "forget_loss": 2.15224758903787e-06,
567
+ "retain_loss": 0.9548900127410889,
568
+ "step": 170
569
+ },
570
+ {
571
+ "epoch": 0.20618556701030927,
572
+ "grad_norm": 2.953125,
573
+ "learning_rate": 4.5132488479262676e-05,
574
+ "loss": 0.3629490852355957,
575
+ "step": 180
576
+ },
577
+ {
578
+ "epoch": 0.20618556701030927,
579
+ "forget_loss": 7.814773198333569e-06,
580
+ "retain_loss": 0.2744402289390564,
581
+ "step": 180
582
+ },
583
+ {
584
+ "epoch": 0.20618556701030927,
585
+ "forget_loss": 9.177201150123437e-07,
586
+ "retain_loss": 0.16577409207820892,
587
+ "step": 180
588
+ },
589
+ {
590
+ "epoch": 0.20618556701030927,
591
+ "forget_loss": 2.789876816677861e-06,
592
+ "retain_loss": 0.26117604970932007,
593
+ "step": 180
594
+ },
595
+ {
596
+ "epoch": 0.20618556701030927,
597
+ "forget_loss": 3.266710336902179e-06,
598
+ "retain_loss": 0.2821800708770752,
599
+ "step": 180
600
+ },
601
+ {
602
+ "epoch": 0.21764032073310424,
603
+ "grad_norm": 3.125,
604
+ "learning_rate": 4.4844470046082955e-05,
605
+ "loss": 0.3086943864822388,
606
+ "step": 190
607
+ },
608
+ {
609
+ "epoch": 0.21764032073310424,
610
+ "forget_loss": 2.877425231417874e-06,
611
+ "retain_loss": 0.4397953450679779,
612
+ "step": 190
613
+ },
614
+ {
615
+ "epoch": 0.21764032073310424,
616
+ "forget_loss": 5.625155608868226e-06,
617
+ "retain_loss": 0.40479227900505066,
618
+ "step": 190
619
+ },
620
+ {
621
+ "epoch": 0.21764032073310424,
622
+ "forget_loss": 7.261318387463689e-06,
623
+ "retain_loss": 0.08186227083206177,
624
+ "step": 190
625
+ },
626
+ {
627
+ "epoch": 0.21764032073310424,
628
+ "forget_loss": 1.0345914915887988e-06,
629
+ "retain_loss": 0.5473820567131042,
630
+ "step": 190
631
+ },
632
+ {
633
+ "epoch": 0.2290950744558992,
634
+ "grad_norm": 2.59375,
635
+ "learning_rate": 4.455645161290323e-05,
636
+ "loss": 0.30754408836364744,
637
+ "step": 200
638
+ },
639
+ {
640
+ "epoch": 0.2290950744558992,
641
+ "forget_loss": 1.8709645246417494e-06,
642
+ "retain_loss": 0.15519601106643677,
643
+ "step": 200
644
+ },
645
+ {
646
+ "epoch": 0.2290950744558992,
647
+ "forget_loss": 5.221798460297578e-07,
648
+ "retain_loss": 0.0977967157959938,
649
+ "step": 200
650
+ },
651
+ {
652
+ "epoch": 0.2290950744558992,
653
+ "forget_loss": 4.533095307124313e-06,
654
+ "retain_loss": 0.12215632200241089,
655
+ "step": 200
656
+ },
657
+ {
658
+ "epoch": 0.2290950744558992,
659
+ "forget_loss": 2.3582092580909375e-06,
660
+ "retain_loss": 0.14002011716365814,
661
+ "step": 200
662
+ },
663
+ {
664
+ "epoch": 0.24054982817869416,
665
+ "grad_norm": 3.140625,
666
+ "learning_rate": 4.4268433179723506e-05,
667
+ "loss": 0.25385417938232424,
668
+ "step": 210
669
+ },
670
+ {
671
+ "epoch": 0.24054982817869416,
672
+ "forget_loss": 4.970596364728408e-06,
673
+ "retain_loss": 0.17109914124011993,
674
+ "step": 210
675
+ },
676
+ {
677
+ "epoch": 0.24054982817869416,
678
+ "forget_loss": 1.6977497580228373e-05,
679
+ "retain_loss": 0.2036435753107071,
680
+ "step": 210
681
+ },
682
+ {
683
+ "epoch": 0.24054982817869416,
684
+ "forget_loss": 3.1074764592631254e-06,
685
+ "retain_loss": 0.13067637383937836,
686
+ "step": 210
687
+ },
688
+ {
689
+ "epoch": 0.24054982817869416,
690
+ "forget_loss": 2.0813706669287058e-06,
691
+ "retain_loss": 0.1853303760290146,
692
+ "step": 210
693
+ },
694
+ {
695
+ "epoch": 0.2520045819014891,
696
+ "grad_norm": 2.5,
697
+ "learning_rate": 4.398041474654378e-05,
698
+ "loss": 0.2122504711151123,
699
+ "step": 220
700
+ },
701
+ {
702
+ "epoch": 0.2520045819014891,
703
+ "forget_loss": 1.3193698578106705e-06,
704
+ "retain_loss": 0.14352917671203613,
705
+ "step": 220
706
+ },
707
+ {
708
+ "epoch": 0.2520045819014891,
709
+ "forget_loss": 1.1001748134731315e-06,
710
+ "retain_loss": 0.30384859442710876,
711
+ "step": 220
712
+ },
713
+ {
714
+ "epoch": 0.2520045819014891,
715
+ "forget_loss": 1.4734857813891722e-06,
716
+ "retain_loss": 0.06934046000242233,
717
+ "step": 220
718
+ },
719
+ {
720
+ "epoch": 0.2520045819014891,
721
+ "forget_loss": 2.7343555757397553e-06,
722
+ "retain_loss": 0.1081199198961258,
723
+ "step": 220
724
+ },
725
+ {
726
+ "epoch": 0.2634593356242841,
727
+ "grad_norm": 3.09375,
728
+ "learning_rate": 4.369239631336406e-05,
729
+ "loss": 0.20068085193634033,
730
+ "step": 230
731
+ },
732
+ {
733
+ "epoch": 0.2634593356242841,
734
+ "forget_loss": 2.154215962946182e-06,
735
+ "retain_loss": 0.10300299525260925,
736
+ "step": 230
737
+ },
738
+ {
739
+ "epoch": 0.2634593356242841,
740
+ "forget_loss": 9.424740710528567e-06,
741
+ "retain_loss": 0.13396301865577698,
742
+ "step": 230
743
+ },
744
+ {
745
+ "epoch": 0.2634593356242841,
746
+ "forget_loss": 3.0522624001605436e-06,
747
+ "retain_loss": 0.33339884877204895,
748
+ "step": 230
749
+ },
750
+ {
751
+ "epoch": 0.2634593356242841,
752
+ "forget_loss": 8.897523912310135e-06,
753
+ "retain_loss": 0.1534104347229004,
754
+ "step": 230
755
+ },
756
+ {
757
+ "epoch": 0.27491408934707906,
758
+ "grad_norm": 2.671875,
759
+ "learning_rate": 4.340437788018433e-05,
760
+ "loss": 0.1963863492012024,
761
+ "step": 240
762
+ },
763
+ {
764
+ "epoch": 0.27491408934707906,
765
+ "forget_loss": 1.7897947373057832e-06,
766
+ "retain_loss": 0.11809217184782028,
767
+ "step": 240
768
+ },
769
+ {
770
+ "epoch": 0.27491408934707906,
771
+ "forget_loss": 5.25018833741342e-07,
772
+ "retain_loss": 0.15055039525032043,
773
+ "step": 240
774
+ },
775
+ {
776
+ "epoch": 0.27491408934707906,
777
+ "forget_loss": 6.8833096520393156e-06,
778
+ "retain_loss": 0.16997316479682922,
779
+ "step": 240
780
+ },
781
+ {
782
+ "epoch": 0.27491408934707906,
783
+ "forget_loss": 2.313455297553446e-06,
784
+ "retain_loss": 0.10752896964550018,
785
+ "step": 240
786
+ },
787
+ {
788
+ "epoch": 0.286368843069874,
789
+ "grad_norm": 2.59375,
790
+ "learning_rate": 4.311635944700461e-05,
791
+ "loss": 0.20680031776428223,
792
+ "step": 250
793
+ },
794
+ {
795
+ "epoch": 0.286368843069874,
796
+ "forget_loss": 2.1803207346238196e-06,
797
+ "retain_loss": 0.07074207067489624,
798
+ "step": 250
799
+ },
800
+ {
801
+ "epoch": 0.286368843069874,
802
+ "forget_loss": 1.8291285641680588e-06,
803
+ "retain_loss": 0.09072300791740417,
804
+ "step": 250
805
+ },
806
+ {
807
+ "epoch": 0.286368843069874,
808
+ "forget_loss": 1.0191804449277697e-06,
809
+ "retain_loss": 0.4518914818763733,
810
+ "step": 250
811
+ },
812
+ {
813
+ "epoch": 0.286368843069874,
814
+ "forget_loss": 1.1188744792889338e-06,
815
+ "retain_loss": 0.09803809970617294,
816
+ "step": 250
817
+ },
818
+ {
819
+ "epoch": 0.29782359679266895,
820
+ "grad_norm": 2.703125,
821
+ "learning_rate": 4.282834101382488e-05,
822
+ "loss": 0.16371082067489623,
823
+ "step": 260
824
+ },
825
+ {
826
+ "epoch": 0.29782359679266895,
827
+ "forget_loss": 3.8669927562295925e-06,
828
+ "retain_loss": 0.17645332217216492,
829
+ "step": 260
830
+ },
831
+ {
832
+ "epoch": 0.29782359679266895,
833
+ "forget_loss": 5.920769581280183e-06,
834
+ "retain_loss": 0.18589472770690918,
835
+ "step": 260
836
+ },
837
+ {
838
+ "epoch": 0.29782359679266895,
839
+ "forget_loss": 5.846962267241906e-06,
840
+ "retain_loss": 0.1563558578491211,
841
+ "step": 260
842
+ },
843
+ {
844
+ "epoch": 0.29782359679266895,
845
+ "forget_loss": 2.4713845050428063e-06,
846
+ "retain_loss": 0.08506585657596588,
847
+ "step": 260
848
+ },
849
+ {
850
+ "epoch": 0.30927835051546393,
851
+ "grad_norm": 1.984375,
852
+ "learning_rate": 4.254032258064516e-05,
853
+ "loss": 0.1610316514968872,
854
+ "step": 270
855
+ },
856
+ {
857
+ "epoch": 0.30927835051546393,
858
+ "forget_loss": 2.003027475439012e-06,
859
+ "retain_loss": 0.1364717036485672,
860
+ "step": 270
861
+ },
862
+ {
863
+ "epoch": 0.30927835051546393,
864
+ "forget_loss": 2.4025737275223946e-06,
865
+ "retain_loss": 0.06489767134189606,
866
+ "step": 270
867
+ },
868
+ {
869
+ "epoch": 0.30927835051546393,
870
+ "forget_loss": 1.9498058918543393e-06,
871
+ "retain_loss": 0.15014496445655823,
872
+ "step": 270
873
+ },
874
+ {
875
+ "epoch": 0.30927835051546393,
876
+ "forget_loss": 1.4814335145274526e-06,
877
+ "retain_loss": 0.10888711363077164,
878
+ "step": 270
879
+ },
880
+ {
881
+ "epoch": 0.3207331042382589,
882
+ "grad_norm": 2.03125,
883
+ "learning_rate": 4.225230414746544e-05,
884
+ "loss": 0.1513701558113098,
885
+ "step": 280
886
+ },
887
+ {
888
+ "epoch": 0.3207331042382589,
889
+ "forget_loss": 3.564718099369202e-06,
890
+ "retain_loss": 0.18767109513282776,
891
+ "step": 280
892
+ },
893
+ {
894
+ "epoch": 0.3207331042382589,
895
+ "forget_loss": 1.5525797607551794e-06,
896
+ "retain_loss": 0.1132984533905983,
897
+ "step": 280
898
+ },
899
+ {
900
+ "epoch": 0.3207331042382589,
901
+ "forget_loss": 2.530495294195134e-06,
902
+ "retain_loss": 0.14477676153182983,
903
+ "step": 280
904
+ },
905
+ {
906
+ "epoch": 0.3207331042382589,
907
+ "forget_loss": 1.2278740086912876e-06,
908
+ "retain_loss": 0.22871914505958557,
909
+ "step": 280
910
+ },
911
+ {
912
+ "epoch": 0.3321878579610538,
913
+ "grad_norm": 1.9375,
914
+ "learning_rate": 4.196428571428572e-05,
915
+ "loss": 0.13684289455413817,
916
+ "step": 290
917
+ },
918
+ {
919
+ "epoch": 0.3321878579610538,
920
+ "forget_loss": 6.087240080887568e-07,
921
+ "retain_loss": 0.12418590486049652,
922
+ "step": 290
923
+ },
924
+ {
925
+ "epoch": 0.3321878579610538,
926
+ "forget_loss": 1.3446937145999982e-06,
927
+ "retain_loss": 0.07405225187540054,
928
+ "step": 290
929
+ },
930
+ {
931
+ "epoch": 0.3321878579610538,
932
+ "forget_loss": 5.7531210586603265e-06,
933
+ "retain_loss": 0.18760758638381958,
934
+ "step": 290
935
+ },
936
+ {
937
+ "epoch": 0.3321878579610538,
938
+ "forget_loss": 3.2198163353314158e-06,
939
+ "retain_loss": 0.06644970923662186,
940
+ "step": 290
941
+ },
942
+ {
943
+ "epoch": 0.3436426116838488,
944
+ "grad_norm": 1.734375,
945
+ "learning_rate": 4.167626728110599e-05,
946
+ "loss": 0.12020657062530518,
947
+ "step": 300
948
+ },
949
+ {
950
+ "epoch": 0.3436426116838488,
951
+ "forget_loss": 3.8907467114768224e-07,
952
+ "retain_loss": 0.09106294065713882,
953
+ "step": 300
954
+ },
955
+ {
956
+ "epoch": 0.3436426116838488,
957
+ "forget_loss": 1.2785067156073637e-06,
958
+ "retain_loss": 0.07731538265943527,
959
+ "step": 300
960
+ },
961
+ {
962
+ "epoch": 0.3436426116838488,
963
+ "forget_loss": 7.883311923251313e-07,
964
+ "retain_loss": 0.0779929980635643,
965
+ "step": 300
966
+ },
967
+ {
968
+ "epoch": 0.3436426116838488,
969
+ "forget_loss": 1.6741282706789207e-06,
970
+ "retain_loss": 0.11640097200870514,
971
+ "step": 300
972
+ },
973
+ {
974
+ "epoch": 0.3550973654066438,
975
+ "grad_norm": 2.09375,
976
+ "learning_rate": 4.138824884792627e-05,
977
+ "loss": 0.11634387969970703,
978
+ "step": 310
979
+ },
980
+ {
981
+ "epoch": 0.3550973654066438,
982
+ "forget_loss": 1.1909932027265313e-06,
983
+ "retain_loss": 0.09301010519266129,
984
+ "step": 310
985
+ },
986
+ {
987
+ "epoch": 0.3550973654066438,
988
+ "forget_loss": 5.385973622651363e-07,
989
+ "retain_loss": 0.13168799877166748,
990
+ "step": 310
991
+ },
992
+ {
993
+ "epoch": 0.3550973654066438,
994
+ "forget_loss": 4.287514457246289e-06,
995
+ "retain_loss": 0.07441989332437515,
996
+ "step": 310
997
+ },
998
+ {
999
+ "epoch": 0.3550973654066438,
1000
+ "forget_loss": 5.496154244610807e-06,
1001
+ "retain_loss": 0.0625639408826828,
1002
+ "step": 310
1003
+ },
1004
+ {
1005
+ "epoch": 0.3665521191294387,
1006
+ "grad_norm": 2.09375,
1007
+ "learning_rate": 4.1100230414746544e-05,
1008
+ "loss": 0.13164939880371093,
1009
+ "step": 320
1010
+ },
1011
+ {
1012
+ "epoch": 0.3665521191294387,
1013
+ "forget_loss": 1.5266854234141647e-06,
1014
+ "retain_loss": 0.07643962651491165,
1015
+ "step": 320
1016
+ },
1017
+ {
1018
+ "epoch": 0.3665521191294387,
1019
+ "forget_loss": 5.235110052126402e-07,
1020
+ "retain_loss": 0.10404697060585022,
1021
+ "step": 320
1022
+ },
1023
+ {
1024
+ "epoch": 0.3665521191294387,
1025
+ "forget_loss": 3.643459649538272e-06,
1026
+ "retain_loss": 0.11882194131612778,
1027
+ "step": 320
1028
+ },
1029
+ {
1030
+ "epoch": 0.3665521191294387,
1031
+ "forget_loss": 9.903519639919978e-07,
1032
+ "retain_loss": 0.07760447263717651,
1033
+ "step": 320
1034
+ },
1035
+ {
1036
+ "epoch": 0.37800687285223367,
1037
+ "grad_norm": 1.7265625,
1038
+ "learning_rate": 4.081221198156682e-05,
1039
+ "loss": 0.11806095838546753,
1040
+ "step": 330
1041
+ },
1042
+ {
1043
+ "epoch": 0.37800687285223367,
1044
+ "forget_loss": 2.186560550399008e-06,
1045
+ "retain_loss": 0.05915055423974991,
1046
+ "step": 330
1047
+ },
1048
+ {
1049
+ "epoch": 0.37800687285223367,
1050
+ "forget_loss": 2.5664467102615163e-06,
1051
+ "retain_loss": 0.06395165622234344,
1052
+ "step": 330
1053
+ },
1054
+ {
1055
+ "epoch": 0.37800687285223367,
1056
+ "forget_loss": 1.2531793345260667e-06,
1057
+ "retain_loss": 0.1056695505976677,
1058
+ "step": 330
1059
+ },
1060
+ {
1061
+ "epoch": 0.37800687285223367,
1062
+ "forget_loss": 7.011165507719852e-06,
1063
+ "retain_loss": 0.10078335553407669,
1064
+ "step": 330
1065
+ },
1066
+ {
1067
+ "epoch": 0.38946162657502864,
1068
+ "grad_norm": 1.734375,
1069
+ "learning_rate": 4.0524193548387096e-05,
1070
+ "loss": 0.10047016143798829,
1071
+ "step": 340
1072
+ },
1073
+ {
1074
+ "epoch": 0.38946162657502864,
1075
+ "forget_loss": 1.4236219385566073e-06,
1076
+ "retain_loss": 0.5226035714149475,
1077
+ "step": 340
1078
+ },
1079
+ {
1080
+ "epoch": 0.38946162657502864,
1081
+ "forget_loss": 1.951645344888675e-06,
1082
+ "retain_loss": 0.10346213728189468,
1083
+ "step": 340
1084
+ },
1085
+ {
1086
+ "epoch": 0.38946162657502864,
1087
+ "forget_loss": 2.398945980530698e-06,
1088
+ "retain_loss": 0.03990686312317848,
1089
+ "step": 340
1090
+ },
1091
+ {
1092
+ "epoch": 0.38946162657502864,
1093
+ "forget_loss": 4.940415237797424e-06,
1094
+ "retain_loss": 0.08106988668441772,
1095
+ "step": 340
1096
+ },
1097
+ {
1098
+ "epoch": 0.4009163802978236,
1099
+ "grad_norm": 1.9296875,
1100
+ "learning_rate": 4.0236175115207375e-05,
1101
+ "loss": 0.10264077186584472,
1102
+ "step": 350
1103
+ },
1104
+ {
1105
+ "epoch": 0.4009163802978236,
1106
+ "forget_loss": 1.1442448339948896e-06,
1107
+ "retain_loss": 0.07710669934749603,
1108
+ "step": 350
1109
+ },
1110
+ {
1111
+ "epoch": 0.4009163802978236,
1112
+ "forget_loss": 1.1809590887423838e-06,
1113
+ "retain_loss": 0.384318470954895,
1114
+ "step": 350
1115
+ },
1116
+ {
1117
+ "epoch": 0.4009163802978236,
1118
+ "forget_loss": 1.8739415281743277e-06,
1119
+ "retain_loss": 0.1440838724374771,
1120
+ "step": 350
1121
+ },
1122
+ {
1123
+ "epoch": 0.4009163802978236,
1124
+ "forget_loss": 1.596398078618222e-06,
1125
+ "retain_loss": 0.08889831602573395,
1126
+ "step": 350
1127
+ },
1128
+ {
1129
+ "epoch": 0.41237113402061853,
1130
+ "grad_norm": 1.40625,
1131
+ "learning_rate": 3.9948156682027654e-05,
1132
+ "loss": 0.09905218482017517,
1133
+ "step": 360
1134
+ },
1135
+ {
1136
+ "epoch": 0.41237113402061853,
1137
+ "forget_loss": 6.401254722732119e-06,
1138
+ "retain_loss": 0.06261590123176575,
1139
+ "step": 360
1140
+ },
1141
+ {
1142
+ "epoch": 0.41237113402061853,
1143
+ "forget_loss": 3.825871317530982e-07,
1144
+ "retain_loss": 0.15589363873004913,
1145
+ "step": 360
1146
+ },
1147
+ {
1148
+ "epoch": 0.41237113402061853,
1149
+ "forget_loss": 1.1340009677951457e-06,
1150
+ "retain_loss": 0.026375912129878998,
1151
+ "step": 360
1152
+ },
1153
+ {
1154
+ "epoch": 0.41237113402061853,
1155
+ "forget_loss": 1.092132379199029e-06,
1156
+ "retain_loss": 0.07016332447528839,
1157
+ "step": 360
1158
+ },
1159
+ {
1160
+ "epoch": 0.4238258877434135,
1161
+ "grad_norm": 2.28125,
1162
+ "learning_rate": 3.9660138248847926e-05,
1163
+ "loss": 0.10010771751403809,
1164
+ "step": 370
1165
+ },
1166
+ {
1167
+ "epoch": 0.4238258877434135,
1168
+ "forget_loss": 5.463415959638951e-07,
1169
+ "retain_loss": 0.09104490280151367,
1170
+ "step": 370
1171
+ },
1172
+ {
1173
+ "epoch": 0.4238258877434135,
1174
+ "forget_loss": 7.711361149631557e-07,
1175
+ "retain_loss": 0.07780595868825912,
1176
+ "step": 370
1177
+ },
1178
+ {
1179
+ "epoch": 0.4238258877434135,
1180
+ "forget_loss": 8.610635973127501e-07,
1181
+ "retain_loss": 0.0850876048207283,
1182
+ "step": 370
1183
+ },
1184
+ {
1185
+ "epoch": 0.4238258877434135,
1186
+ "forget_loss": 2.193863338106894e-06,
1187
+ "retain_loss": 0.02873000130057335,
1188
+ "step": 370
1189
+ },
1190
+ {
1191
+ "epoch": 0.4352806414662085,
1192
+ "grad_norm": 1.8125,
1193
+ "learning_rate": 3.9372119815668206e-05,
1194
+ "loss": 0.07582432627677918,
1195
+ "step": 380
1196
+ },
1197
+ {
1198
+ "epoch": 0.4352806414662085,
1199
+ "forget_loss": 5.40978362550959e-07,
1200
+ "retain_loss": 0.03535467013716698,
1201
+ "step": 380
1202
+ },
1203
+ {
1204
+ "epoch": 0.4352806414662085,
1205
+ "forget_loss": 8.922065717342775e-07,
1206
+ "retain_loss": 0.057049620896577835,
1207
+ "step": 380
1208
+ },
1209
+ {
1210
+ "epoch": 0.4352806414662085,
1211
+ "forget_loss": 5.560720524044882e-07,
1212
+ "retain_loss": 0.1527683138847351,
1213
+ "step": 380
1214
+ },
1215
+ {
1216
+ "epoch": 0.4352806414662085,
1217
+ "forget_loss": 4.7091802457543963e-07,
1218
+ "retain_loss": 0.1109684631228447,
1219
+ "step": 380
1220
+ },
1221
+ {
1222
+ "epoch": 0.44673539518900346,
1223
+ "grad_norm": 1.7734375,
1224
+ "learning_rate": 3.908410138248848e-05,
1225
+ "loss": 0.09133011102676392,
1226
+ "step": 390
1227
+ },
1228
+ {
1229
+ "epoch": 0.44673539518900346,
1230
+ "forget_loss": 8.056795763877744e-07,
1231
+ "retain_loss": 0.09070051461458206,
1232
+ "step": 390
1233
+ },
1234
+ {
1235
+ "epoch": 0.44673539518900346,
1236
+ "forget_loss": 1.1127366406071815e-06,
1237
+ "retain_loss": 0.0616828016936779,
1238
+ "step": 390
1239
+ },
1240
+ {
1241
+ "epoch": 0.44673539518900346,
1242
+ "forget_loss": 3.5524210488802055e-06,
1243
+ "retain_loss": 0.06886927038431168,
1244
+ "step": 390
1245
+ },
1246
+ {
1247
+ "epoch": 0.44673539518900346,
1248
+ "forget_loss": 2.0374077394080814e-06,
1249
+ "retain_loss": 0.10105489939451218,
1250
+ "step": 390
1251
+ },
1252
+ {
1253
+ "epoch": 0.4581901489117984,
1254
+ "grad_norm": 1.59375,
1255
+ "learning_rate": 3.879608294930876e-05,
1256
+ "loss": 0.07695338726043702,
1257
+ "step": 400
1258
+ },
1259
+ {
1260
+ "epoch": 0.4581901489117984,
1261
+ "forget_loss": 1.342506493529072e-06,
1262
+ "retain_loss": 0.03703423589468002,
1263
+ "step": 400
1264
+ },
1265
+ {
1266
+ "epoch": 0.4581901489117984,
1267
+ "forget_loss": 4.241804276716721e-07,
1268
+ "retain_loss": 0.07702656090259552,
1269
+ "step": 400
1270
+ },
1271
+ {
1272
+ "epoch": 0.4581901489117984,
1273
+ "forget_loss": 2.735993405167392e-07,
1274
+ "retain_loss": 0.04419003427028656,
1275
+ "step": 400
1276
+ },
1277
+ {
1278
+ "epoch": 0.4581901489117984,
1279
+ "forget_loss": 3.817524429905461e-07,
1280
+ "retain_loss": 0.040069010108709335,
1281
+ "step": 400
1282
+ },
1283
+ {
1284
+ "epoch": 0.46964490263459335,
1285
+ "grad_norm": 1.671875,
1286
+ "learning_rate": 3.8508064516129036e-05,
1287
+ "loss": 0.07269055247306824,
1288
+ "step": 410
1289
+ },
1290
+ {
1291
+ "epoch": 0.46964490263459335,
1292
+ "forget_loss": 2.8240863230166724e-06,
1293
+ "retain_loss": 0.04700532928109169,
1294
+ "step": 410
1295
+ },
1296
+ {
1297
+ "epoch": 0.46964490263459335,
1298
+ "forget_loss": 1.747673792351634e-07,
1299
+ "retain_loss": 0.033581238240003586,
1300
+ "step": 410
1301
+ },
1302
+ {
1303
+ "epoch": 0.46964490263459335,
1304
+ "forget_loss": 5.92311096170306e-07,
1305
+ "retain_loss": 0.09985388815402985,
1306
+ "step": 410
1307
+ },
1308
+ {
1309
+ "epoch": 0.46964490263459335,
1310
+ "forget_loss": 1.7826670273279888e-07,
1311
+ "retain_loss": 0.07349472492933273,
1312
+ "step": 410
1313
+ },
1314
+ {
1315
+ "epoch": 0.48109965635738833,
1316
+ "grad_norm": 1.8046875,
1317
+ "learning_rate": 3.8220046082949316e-05,
1318
+ "loss": 0.07912165522575379,
1319
+ "step": 420
1320
+ },
1321
+ {
1322
+ "epoch": 0.48109965635738833,
1323
+ "forget_loss": 1.2650438918626605e-07,
1324
+ "retain_loss": 0.06736782938241959,
1325
+ "step": 420
1326
+ },
1327
+ {
1328
+ "epoch": 0.48109965635738833,
1329
+ "forget_loss": 1.5735973875052878e-06,
1330
+ "retain_loss": 0.03057463839650154,
1331
+ "step": 420
1332
+ },
1333
+ {
1334
+ "epoch": 0.48109965635738833,
1335
+ "forget_loss": 4.024736313112953e-07,
1336
+ "retain_loss": 0.05045014247298241,
1337
+ "step": 420
1338
+ },
1339
+ {
1340
+ "epoch": 0.48109965635738833,
1341
+ "forget_loss": 2.5390670543856686e-06,
1342
+ "retain_loss": 0.07917766273021698,
1343
+ "step": 420
1344
+ },
1345
+ {
1346
+ "epoch": 0.4925544100801833,
1347
+ "grad_norm": 1.1953125,
1348
+ "learning_rate": 3.793202764976959e-05,
1349
+ "loss": 0.0656080961227417,
1350
+ "step": 430
1351
+ },
1352
+ {
1353
+ "epoch": 0.4925544100801833,
1354
+ "forget_loss": 5.468404538078175e-07,
1355
+ "retain_loss": 0.02683027647435665,
1356
+ "step": 430
1357
+ },
1358
+ {
1359
+ "epoch": 0.4925544100801833,
1360
+ "forget_loss": 5.585835083365964e-07,
1361
+ "retain_loss": 0.0377357117831707,
1362
+ "step": 430
1363
+ },
1364
+ {
1365
+ "epoch": 0.4925544100801833,
1366
+ "forget_loss": 3.003877679930156e-07,
1367
+ "retain_loss": 0.052259303629398346,
1368
+ "step": 430
1369
+ },
1370
+ {
1371
+ "epoch": 0.4925544100801833,
1372
+ "forget_loss": 1.2751459053106373e-06,
1373
+ "retain_loss": 0.10837093740701675,
1374
+ "step": 430
1375
+ },
1376
+ {
1377
+ "epoch": 0.5040091638029782,
1378
+ "grad_norm": 1.8046875,
1379
+ "learning_rate": 3.764400921658987e-05,
1380
+ "loss": 0.06297182440757751,
1381
+ "step": 440
1382
+ },
1383
+ {
1384
+ "epoch": 0.5040091638029782,
1385
+ "forget_loss": 2.636994622662314e-07,
1386
+ "retain_loss": 0.04880654811859131,
1387
+ "step": 440
1388
+ },
1389
+ {
1390
+ "epoch": 0.5040091638029782,
1391
+ "forget_loss": 2.0067361674591666e-06,
1392
+ "retain_loss": 0.027269020676612854,
1393
+ "step": 440
1394
+ },
1395
+ {
1396
+ "epoch": 0.5040091638029782,
1397
+ "forget_loss": 1.8749557284536422e-06,
1398
+ "retain_loss": 0.04770921915769577,
1399
+ "step": 440
1400
+ },
1401
+ {
1402
+ "epoch": 0.5040091638029782,
1403
+ "forget_loss": 7.214839001790097e-07,
1404
+ "retain_loss": 0.07100628316402435,
1405
+ "step": 440
1406
+ },
1407
+ {
1408
+ "epoch": 0.5154639175257731,
1409
+ "grad_norm": 1.0859375,
1410
+ "learning_rate": 3.735599078341014e-05,
1411
+ "loss": 0.0665643870830536,
1412
+ "step": 450
1413
+ },
1414
+ {
1415
+ "epoch": 0.5154639175257731,
1416
+ "forget_loss": 8.713016796946249e-08,
1417
+ "retain_loss": 0.042148686945438385,
1418
+ "step": 450
1419
+ },
1420
+ {
1421
+ "epoch": 0.5154639175257731,
1422
+ "forget_loss": 6.566732508872519e-07,
1423
+ "retain_loss": 0.05762806907296181,
1424
+ "step": 450
1425
+ },
1426
+ {
1427
+ "epoch": 0.5154639175257731,
1428
+ "forget_loss": 2.520494888358371e-07,
1429
+ "retain_loss": 0.0422804020345211,
1430
+ "step": 450
1431
+ },
1432
+ {
1433
+ "epoch": 0.5154639175257731,
1434
+ "forget_loss": 2.0052689251315314e-06,
1435
+ "retain_loss": 0.05088704079389572,
1436
+ "step": 450
1437
+ },
1438
+ {
1439
+ "epoch": 0.5269186712485682,
1440
+ "grad_norm": 1.140625,
1441
+ "learning_rate": 3.706797235023042e-05,
1442
+ "loss": 0.06394548416137695,
1443
+ "step": 460
1444
+ },
1445
+ {
1446
+ "epoch": 0.5269186712485682,
1447
+ "forget_loss": 4.606073389368248e-07,
1448
+ "retain_loss": 0.0739588811993599,
1449
+ "step": 460
1450
+ },
1451
+ {
1452
+ "epoch": 0.5269186712485682,
1453
+ "forget_loss": 4.4586619196707034e-07,
1454
+ "retain_loss": 0.059775300323963165,
1455
+ "step": 460
1456
+ },
1457
+ {
1458
+ "epoch": 0.5269186712485682,
1459
+ "forget_loss": 4.324638211983256e-07,
1460
+ "retain_loss": 0.08108869194984436,
1461
+ "step": 460
1462
+ },
1463
+ {
1464
+ "epoch": 0.5269186712485682,
1465
+ "forget_loss": 5.558591738008545e-07,
1466
+ "retain_loss": 0.027880918234586716,
1467
+ "step": 460
1468
+ },
1469
+ {
1470
+ "epoch": 0.5383734249713631,
1471
+ "grad_norm": 1.171875,
1472
+ "learning_rate": 3.677995391705069e-05,
1473
+ "loss": 0.05814751386642456,
1474
+ "step": 470
1475
+ },
1476
+ {
1477
+ "epoch": 0.5383734249713631,
1478
+ "forget_loss": 3.032934046132141e-07,
1479
+ "retain_loss": 0.024084806442260742,
1480
+ "step": 470
1481
+ },
1482
+ {
1483
+ "epoch": 0.5383734249713631,
1484
+ "forget_loss": 8.804982485344226e-07,
1485
+ "retain_loss": 0.03435460478067398,
1486
+ "step": 470
1487
+ },
1488
+ {
1489
+ "epoch": 0.5383734249713631,
1490
+ "forget_loss": 1.1724860087269917e-06,
1491
+ "retain_loss": 0.048975635319948196,
1492
+ "step": 470
1493
+ },
1494
+ {
1495
+ "epoch": 0.5383734249713631,
1496
+ "forget_loss": 3.8122900036796636e-07,
1497
+ "retain_loss": 0.0677463635802269,
1498
+ "step": 470
1499
+ },
1500
+ {
1501
+ "epoch": 0.5498281786941581,
1502
+ "grad_norm": 1.734375,
1503
+ "learning_rate": 3.649193548387097e-05,
1504
+ "loss": 0.060336172580718994,
1505
+ "step": 480
1506
+ },
1507
+ {
1508
+ "epoch": 0.5498281786941581,
1509
+ "forget_loss": 9.458409522267175e-07,
1510
+ "retain_loss": 0.07034104317426682,
1511
+ "step": 480
1512
+ },
1513
+ {
1514
+ "epoch": 0.5498281786941581,
1515
+ "forget_loss": 6.857035828033986e-07,
1516
+ "retain_loss": 0.03801558166742325,
1517
+ "step": 480
1518
+ },
1519
+ {
1520
+ "epoch": 0.5498281786941581,
1521
+ "forget_loss": 1.3080871212878264e-06,
1522
+ "retain_loss": 0.07545764744281769,
1523
+ "step": 480
1524
+ },
1525
+ {
1526
+ "epoch": 0.5498281786941581,
1527
+ "forget_loss": 3.9946274910107604e-07,
1528
+ "retain_loss": 0.03869559243321419,
1529
+ "step": 480
1530
+ },
1531
+ {
1532
+ "epoch": 0.561282932416953,
1533
+ "grad_norm": 1.0546875,
1534
+ "learning_rate": 3.620391705069124e-05,
1535
+ "loss": 0.05845548510551453,
1536
+ "step": 490
1537
+ },
1538
+ {
1539
+ "epoch": 0.561282932416953,
1540
+ "forget_loss": 9.76757291937247e-07,
1541
+ "retain_loss": 0.04530367627739906,
1542
+ "step": 490
1543
+ },
1544
+ {
1545
+ "epoch": 0.561282932416953,
1546
+ "forget_loss": 5.240474933998485e-07,
1547
+ "retain_loss": 0.03550443798303604,
1548
+ "step": 490
1549
+ },
1550
+ {
1551
+ "epoch": 0.561282932416953,
1552
+ "forget_loss": 7.904294534455403e-07,
1553
+ "retain_loss": 0.08905749022960663,
1554
+ "step": 490
1555
+ },
1556
+ {
1557
+ "epoch": 0.561282932416953,
1558
+ "forget_loss": 3.536492840794381e-07,
1559
+ "retain_loss": 0.05929813161492348,
1560
+ "step": 490
1561
+ },
1562
+ {
1563
+ "epoch": 0.572737686139748,
1564
+ "grad_norm": 1.1640625,
1565
+ "learning_rate": 3.591589861751152e-05,
1566
+ "loss": 0.06306434273719788,
1567
+ "step": 500
1568
+ },
1569
+ {
1570
+ "epoch": 0.572737686139748,
1571
+ "forget_loss": 4.770619170813006e-07,
1572
+ "retain_loss": 0.0518912598490715,
1573
+ "step": 500
1574
+ },
1575
+ {
1576
+ "epoch": 0.572737686139748,
1577
+ "forget_loss": 7.711427940648719e-08,
1578
+ "retain_loss": 0.06265456974506378,
1579
+ "step": 500
1580
+ },
1581
+ {
1582
+ "epoch": 0.572737686139748,
1583
+ "forget_loss": 1.8670939425646793e-06,
1584
+ "retain_loss": 0.07513444125652313,
1585
+ "step": 500
1586
+ },
1587
+ {
1588
+ "epoch": 0.572737686139748,
1589
+ "forget_loss": 4.7316902396232763e-07,
1590
+ "retain_loss": 0.08540341258049011,
1591
+ "step": 500
1592
+ },
1593
+ {
1594
+ "epoch": 0.584192439862543,
1595
+ "grad_norm": 0.984375,
1596
+ "learning_rate": 3.5627880184331795e-05,
1597
+ "loss": 0.05606219172477722,
1598
+ "step": 510
1599
+ },
1600
+ {
1601
+ "epoch": 0.584192439862543,
1602
+ "forget_loss": 8.002095910342177e-07,
1603
+ "retain_loss": 0.04273177683353424,
1604
+ "step": 510
1605
+ },
1606
+ {
1607
+ "epoch": 0.584192439862543,
1608
+ "forget_loss": 1.8704803039781837e-07,
1609
+ "retain_loss": 0.053921207785606384,
1610
+ "step": 510
1611
+ },
1612
+ {
1613
+ "epoch": 0.584192439862543,
1614
+ "forget_loss": 3.775904247049766e-07,
1615
+ "retain_loss": 0.0322711281478405,
1616
+ "step": 510
1617
+ },
1618
+ {
1619
+ "epoch": 0.584192439862543,
1620
+ "forget_loss": 6.053579681974952e-07,
1621
+ "retain_loss": 0.05659232288599014,
1622
+ "step": 510
1623
+ },
1624
+ {
1625
+ "epoch": 0.5956471935853379,
1626
+ "grad_norm": 1.2265625,
1627
+ "learning_rate": 3.5339861751152074e-05,
1628
+ "loss": 0.049942076206207275,
1629
+ "step": 520
1630
+ },
1631
+ {
1632
+ "epoch": 0.5956471935853379,
1633
+ "forget_loss": 2.9679642921109917e-07,
1634
+ "retain_loss": 0.08346150070428848,
1635
+ "step": 520
1636
+ },
1637
+ {
1638
+ "epoch": 0.5956471935853379,
1639
+ "forget_loss": 2.005875501254195e-07,
1640
+ "retain_loss": 0.030936188995838165,
1641
+ "step": 520
1642
+ },
1643
+ {
1644
+ "epoch": 0.5956471935853379,
1645
+ "forget_loss": 1.0443299061080324e-06,
1646
+ "retain_loss": 0.09520356357097626,
1647
+ "step": 520
1648
+ },
1649
+ {
1650
+ "epoch": 0.5956471935853379,
1651
+ "forget_loss": 2.0819126689275436e-07,
1652
+ "retain_loss": 0.0474083349108696,
1653
+ "step": 520
1654
+ },
1655
+ {
1656
+ "epoch": 0.6071019473081328,
1657
+ "grad_norm": 1.0625,
1658
+ "learning_rate": 3.5051843317972346e-05,
1659
+ "loss": 0.05396345257759094,
1660
+ "step": 530
1661
+ },
1662
+ {
1663
+ "epoch": 0.6071019473081328,
1664
+ "forget_loss": 8.294673534692265e-07,
1665
+ "retain_loss": 0.042203839868307114,
1666
+ "step": 530
1667
+ },
1668
+ {
1669
+ "epoch": 0.6071019473081328,
1670
+ "forget_loss": 7.164442195062293e-08,
1671
+ "retain_loss": 0.030783096328377724,
1672
+ "step": 530
1673
+ },
1674
+ {
1675
+ "epoch": 0.6071019473081328,
1676
+ "forget_loss": 1.3861478691978846e-05,
1677
+ "retain_loss": 0.04198005795478821,
1678
+ "step": 530
1679
+ },
1680
+ {
1681
+ "epoch": 0.6071019473081328,
1682
+ "forget_loss": 4.225911993671616e-07,
1683
+ "retain_loss": 0.06045946478843689,
1684
+ "step": 530
1685
+ },
1686
+ {
1687
+ "epoch": 0.6185567010309279,
1688
+ "grad_norm": 1.2578125,
1689
+ "learning_rate": 3.4763824884792625e-05,
1690
+ "loss": 0.05722784399986267,
1691
+ "step": 540
1692
+ },
1693
+ {
1694
+ "epoch": 0.6185567010309279,
1695
+ "forget_loss": 3.0147746201691916e-06,
1696
+ "retain_loss": 0.0411076694726944,
1697
+ "step": 540
1698
+ },
1699
+ {
1700
+ "epoch": 0.6185567010309279,
1701
+ "forget_loss": 2.582411582352506e-07,
1702
+ "retain_loss": 0.023279590532183647,
1703
+ "step": 540
1704
+ },
1705
+ {
1706
+ "epoch": 0.6185567010309279,
1707
+ "forget_loss": 2.0758065488735156e-07,
1708
+ "retain_loss": 0.04550226032733917,
1709
+ "step": 540
1710
+ },
1711
+ {
1712
+ "epoch": 0.6185567010309279,
1713
+ "forget_loss": 9.63341904025583e-07,
1714
+ "retain_loss": 0.0418274812400341,
1715
+ "step": 540
1716
+ },
1717
+ {
1718
+ "epoch": 0.6300114547537228,
1719
+ "grad_norm": 1.3359375,
1720
+ "learning_rate": 3.4475806451612905e-05,
1721
+ "loss": 0.04913765788078308,
1722
+ "step": 550
1723
+ },
1724
+ {
1725
+ "epoch": 0.6300114547537228,
1726
+ "forget_loss": 2.0537379441520898e-07,
1727
+ "retain_loss": 0.046085719019174576,
1728
+ "step": 550
1729
+ },
1730
+ {
1731
+ "epoch": 0.6300114547537228,
1732
+ "forget_loss": 4.5911775714557734e-07,
1733
+ "retain_loss": 0.041781309992074966,
1734
+ "step": 550
1735
+ },
1736
+ {
1737
+ "epoch": 0.6300114547537228,
1738
+ "forget_loss": 1.873733936008648e-06,
1739
+ "retain_loss": 0.030963057652115822,
1740
+ "step": 550
1741
+ },
1742
+ {
1743
+ "epoch": 0.6300114547537228,
1744
+ "forget_loss": 2.3070414556514152e-07,
1745
+ "retain_loss": 0.036035358905792236,
1746
+ "step": 550
1747
+ },
1748
+ {
1749
+ "epoch": 0.6414662084765178,
1750
+ "grad_norm": 1.2109375,
1751
+ "learning_rate": 3.4187788018433184e-05,
1752
+ "loss": 0.046972605586051944,
1753
+ "step": 560
1754
+ },
1755
+ {
1756
+ "epoch": 0.6414662084765178,
1757
+ "forget_loss": 1.888665224214492e-07,
1758
+ "retain_loss": 0.04609245806932449,
1759
+ "step": 560
1760
+ },
1761
+ {
1762
+ "epoch": 0.6414662084765178,
1763
+ "forget_loss": 1.1302165603410685e-06,
1764
+ "retain_loss": 0.04581566900014877,
1765
+ "step": 560
1766
+ },
1767
+ {
1768
+ "epoch": 0.6414662084765178,
1769
+ "forget_loss": 3.713363128099445e-07,
1770
+ "retain_loss": 0.05701315402984619,
1771
+ "step": 560
1772
+ },
1773
+ {
1774
+ "epoch": 0.6414662084765178,
1775
+ "forget_loss": 1.0438467512585703e-07,
1776
+ "retain_loss": 0.04488621652126312,
1777
+ "step": 560
1778
+ },
1779
+ {
1780
+ "epoch": 0.6529209621993127,
1781
+ "grad_norm": 1.640625,
1782
+ "learning_rate": 3.389976958525346e-05,
1783
+ "loss": 0.04600127041339874,
1784
+ "step": 570
1785
+ },
1786
+ {
1787
+ "epoch": 0.6529209621993127,
1788
+ "forget_loss": 1.2589470088641974e-07,
1789
+ "retain_loss": 0.04046102985739708,
1790
+ "step": 570
1791
+ },
1792
+ {
1793
+ "epoch": 0.6529209621993127,
1794
+ "forget_loss": 2.068235062324675e-06,
1795
+ "retain_loss": 0.06527687609195709,
1796
+ "step": 570
1797
+ },
1798
+ {
1799
+ "epoch": 0.6529209621993127,
1800
+ "forget_loss": 4.6239173911999387e-07,
1801
+ "retain_loss": 0.02964375726878643,
1802
+ "step": 570
1803
+ },
1804
+ {
1805
+ "epoch": 0.6529209621993127,
1806
+ "forget_loss": 1.7551890323375119e-06,
1807
+ "retain_loss": 0.04939418286085129,
1808
+ "step": 570
1809
+ },
1810
+ {
1811
+ "epoch": 0.6643757159221076,
1812
+ "grad_norm": 1.4921875,
1813
+ "learning_rate": 3.3611751152073736e-05,
1814
+ "loss": 0.04693392515182495,
1815
+ "step": 580
1816
+ },
1817
+ {
1818
+ "epoch": 0.6643757159221076,
1819
+ "forget_loss": 1.162212441840893e-07,
1820
+ "retain_loss": 0.01956232823431492,
1821
+ "step": 580
1822
+ },
1823
+ {
1824
+ "epoch": 0.6643757159221076,
1825
+ "forget_loss": 5.034058858655044e-07,
1826
+ "retain_loss": 0.029308225959539413,
1827
+ "step": 580
1828
+ },
1829
+ {
1830
+ "epoch": 0.6643757159221076,
1831
+ "forget_loss": 1.8678618118883605e-07,
1832
+ "retain_loss": 0.057615552097558975,
1833
+ "step": 580
1834
+ },
1835
+ {
1836
+ "epoch": 0.6643757159221076,
1837
+ "forget_loss": 5.4416375405708095e-08,
1838
+ "retain_loss": 0.06805437058210373,
1839
+ "step": 580
1840
+ },
1841
+ {
1842
+ "epoch": 0.6758304696449027,
1843
+ "grad_norm": 0.85546875,
1844
+ "learning_rate": 3.3323732718894015e-05,
1845
+ "loss": 0.044681668281555176,
1846
+ "step": 590
1847
+ },
1848
+ {
1849
+ "epoch": 0.6758304696449027,
1850
+ "forget_loss": 6.884723688926897e-07,
1851
+ "retain_loss": 0.032636139541864395,
1852
+ "step": 590
1853
+ },
1854
+ {
1855
+ "epoch": 0.6758304696449027,
1856
+ "forget_loss": 2.1161808660963288e-07,
1857
+ "retain_loss": 0.03656921535730362,
1858
+ "step": 590
1859
+ },
1860
+ {
1861
+ "epoch": 0.6758304696449027,
1862
+ "forget_loss": 3.788201468069019e-07,
1863
+ "retain_loss": 0.0688297301530838,
1864
+ "step": 590
1865
+ },
1866
+ {
1867
+ "epoch": 0.6758304696449027,
1868
+ "forget_loss": 8.874022228155809e-08,
1869
+ "retain_loss": 0.021696442738175392,
1870
+ "step": 590
1871
+ },
1872
+ {
1873
+ "epoch": 0.6872852233676976,
1874
+ "grad_norm": 0.94140625,
1875
+ "learning_rate": 3.303571428571429e-05,
1876
+ "loss": 0.04318673312664032,
1877
+ "step": 600
1878
+ },
1879
+ {
1880
+ "epoch": 0.6872852233676976,
1881
+ "forget_loss": 3.341031913350889e-07,
1882
+ "retain_loss": 0.03312600031495094,
1883
+ "step": 600
1884
+ },
1885
+ {
1886
+ "epoch": 0.6872852233676976,
1887
+ "forget_loss": 1.59767705554259e-06,
1888
+ "retain_loss": 0.037248943001031876,
1889
+ "step": 600
1890
+ },
1891
+ {
1892
+ "epoch": 0.6872852233676976,
1893
+ "forget_loss": 1.2864059328876465e-07,
1894
+ "retain_loss": 0.051617469638586044,
1895
+ "step": 600
1896
+ },
1897
+ {
1898
+ "epoch": 0.6872852233676976,
1899
+ "forget_loss": 1.3946912247320142e-07,
1900
+ "retain_loss": 0.07654328644275665,
1901
+ "step": 600
1902
+ },
1903
+ {
1904
+ "epoch": 0.6987399770904925,
1905
+ "grad_norm": 0.6484375,
1906
+ "learning_rate": 3.2747695852534566e-05,
1907
+ "loss": 0.043347522616386414,
1908
+ "step": 610
1909
+ },
1910
+ {
1911
+ "epoch": 0.6987399770904925,
1912
+ "forget_loss": 1.6136347369410942e-07,
1913
+ "retain_loss": 0.06624381989240646,
1914
+ "step": 610
1915
+ },
1916
+ {
1917
+ "epoch": 0.6987399770904925,
1918
+ "forget_loss": 1.5072893688738986e-07,
1919
+ "retain_loss": 0.025636199861764908,
1920
+ "step": 610
1921
+ },
1922
+ {
1923
+ "epoch": 0.6987399770904925,
1924
+ "forget_loss": 3.572060336409777e-07,
1925
+ "retain_loss": 0.029844949021935463,
1926
+ "step": 610
1927
+ },
1928
+ {
1929
+ "epoch": 0.6987399770904925,
1930
+ "forget_loss": 3.3666503895801725e-07,
1931
+ "retain_loss": 0.08041810244321823,
1932
+ "step": 610
1933
+ },
1934
+ {
1935
+ "epoch": 0.7101947308132875,
1936
+ "grad_norm": 0.76953125,
1937
+ "learning_rate": 3.245967741935484e-05,
1938
+ "loss": 0.04078985154628754,
1939
+ "step": 620
1940
+ },
1941
+ {
1942
+ "epoch": 0.7101947308132875,
1943
+ "forget_loss": 2.2686955958306498e-07,
1944
+ "retain_loss": 0.03770332783460617,
1945
+ "step": 620
1946
+ },
1947
+ {
1948
+ "epoch": 0.7101947308132875,
1949
+ "forget_loss": 1.2471460308915994e-07,
1950
+ "retain_loss": 0.04125113785266876,
1951
+ "step": 620
1952
+ },
1953
+ {
1954
+ "epoch": 0.7101947308132875,
1955
+ "forget_loss": 5.980033392916084e-07,
1956
+ "retain_loss": 0.028695637360215187,
1957
+ "step": 620
1958
+ },
1959
+ {
1960
+ "epoch": 0.7101947308132875,
1961
+ "forget_loss": 2.1789082893519662e-07,
1962
+ "retain_loss": 0.021218575537204742,
1963
+ "step": 620
1964
+ },
1965
+ {
1966
+ "epoch": 0.7216494845360825,
1967
+ "grad_norm": 0.72265625,
1968
+ "learning_rate": 3.217165898617512e-05,
1969
+ "loss": 0.03691001832485199,
1970
+ "step": 630
1971
+ },
1972
+ {
1973
+ "epoch": 0.7216494845360825,
1974
+ "forget_loss": 5.882121172362531e-07,
1975
+ "retain_loss": 0.028744207695126534,
1976
+ "step": 630
1977
+ },
1978
+ {
1979
+ "epoch": 0.7216494845360825,
1980
+ "forget_loss": 3.706446420892462e-07,
1981
+ "retain_loss": 0.037669021636247635,
1982
+ "step": 630
1983
+ },
1984
+ {
1985
+ "epoch": 0.7216494845360825,
1986
+ "forget_loss": 9.750963272381341e-07,
1987
+ "retain_loss": 0.055296238511800766,
1988
+ "step": 630
1989
+ },
1990
+ {
1991
+ "epoch": 0.7216494845360825,
1992
+ "forget_loss": 2.988974188156135e-07,
1993
+ "retain_loss": 0.04488551616668701,
1994
+ "step": 630
1995
+ },
1996
+ {
1997
+ "epoch": 0.7331042382588774,
1998
+ "grad_norm": 0.8671875,
1999
+ "learning_rate": 3.188364055299539e-05,
2000
+ "loss": 0.038949227333068846,
2001
+ "step": 640
2002
+ },
2003
+ {
2004
+ "epoch": 0.7331042382588774,
2005
+ "forget_loss": 1.2395511816976068e-07,
2006
+ "retain_loss": 0.027795039117336273,
2007
+ "step": 640
2008
+ },
2009
+ {
2010
+ "epoch": 0.7331042382588774,
2011
+ "forget_loss": 5.673878149536904e-07,
2012
+ "retain_loss": 0.06736742705106735,
2013
+ "step": 640
2014
+ },
2015
+ {
2016
+ "epoch": 0.7331042382588774,
2017
+ "forget_loss": 1.1004679834059061e-07,
2018
+ "retain_loss": 0.0364450179040432,
2019
+ "step": 640
2020
+ },
2021
+ {
2022
+ "epoch": 0.7331042382588774,
2023
+ "forget_loss": 1.2864478549090563e-07,
2024
+ "retain_loss": 0.04605640098452568,
2025
+ "step": 640
2026
+ },
2027
+ {
2028
+ "epoch": 0.7445589919816724,
2029
+ "grad_norm": 1.2578125,
2030
+ "learning_rate": 3.159562211981567e-05,
2031
+ "loss": 0.03986696600914001,
2032
+ "step": 650
2033
+ },
2034
+ {
2035
+ "epoch": 0.7445589919816724,
2036
+ "forget_loss": 5.023429139328073e-07,
2037
+ "retain_loss": 0.035111404955387115,
2038
+ "step": 650
2039
+ },
2040
+ {
2041
+ "epoch": 0.7445589919816724,
2042
+ "forget_loss": 4.809768370250822e-07,
2043
+ "retain_loss": 0.02187221683561802,
2044
+ "step": 650
2045
+ },
2046
+ {
2047
+ "epoch": 0.7445589919816724,
2048
+ "forget_loss": 4.244728188496083e-06,
2049
+ "retain_loss": 0.08197719603776932,
2050
+ "step": 650
2051
+ },
2052
+ {
2053
+ "epoch": 0.7445589919816724,
2054
+ "forget_loss": 1.1673816402435477e-07,
2055
+ "retain_loss": 0.08870132267475128,
2056
+ "step": 650
2057
+ },
2058
+ {
2059
+ "epoch": 0.7560137457044673,
2060
+ "grad_norm": 0.6640625,
2061
+ "learning_rate": 3.130760368663594e-05,
2062
+ "loss": 0.039774364233016966,
2063
+ "step": 660
2064
+ },
2065
+ {
2066
+ "epoch": 0.7560137457044673,
2067
+ "forget_loss": 1.1249169062921283e-07,
2068
+ "retain_loss": 0.038397908210754395,
2069
+ "step": 660
2070
+ },
2071
+ {
2072
+ "epoch": 0.7560137457044673,
2073
+ "forget_loss": 5.782552534583374e-07,
2074
+ "retain_loss": 0.030945003032684326,
2075
+ "step": 660
2076
+ },
2077
+ {
2078
+ "epoch": 0.7560137457044673,
2079
+ "forget_loss": 2.596835315671342e-07,
2080
+ "retain_loss": 0.04453842341899872,
2081
+ "step": 660
2082
+ },
2083
+ {
2084
+ "epoch": 0.7560137457044673,
2085
+ "forget_loss": 8.396180533054576e-07,
2086
+ "retain_loss": 0.04271215945482254,
2087
+ "step": 660
2088
+ },
2089
+ {
2090
+ "epoch": 0.7674684994272624,
2091
+ "grad_norm": 0.89453125,
2092
+ "learning_rate": 3.101958525345622e-05,
2093
+ "loss": 0.03887006938457489,
2094
+ "step": 670
2095
+ },
2096
+ {
2097
+ "epoch": 0.7674684994272624,
2098
+ "forget_loss": 8.253943661884477e-08,
2099
+ "retain_loss": 0.03154648467898369,
2100
+ "step": 670
2101
+ },
2102
+ {
2103
+ "epoch": 0.7674684994272624,
2104
+ "forget_loss": 4.430251010489883e-07,
2105
+ "retain_loss": 0.047183044254779816,
2106
+ "step": 670
2107
+ },
2108
+ {
2109
+ "epoch": 0.7674684994272624,
2110
+ "forget_loss": 5.662621447299898e-07,
2111
+ "retain_loss": 0.0315515510737896,
2112
+ "step": 670
2113
+ },
2114
+ {
2115
+ "epoch": 0.7674684994272624,
2116
+ "forget_loss": 2.8115960049035493e-07,
2117
+ "retain_loss": 0.024359095841646194,
2118
+ "step": 670
2119
+ },
2120
+ {
2121
+ "epoch": 0.7789232531500573,
2122
+ "grad_norm": 0.890625,
2123
+ "learning_rate": 3.07315668202765e-05,
2124
+ "loss": 0.03558169901371002,
2125
+ "step": 680
2126
+ },
2127
+ {
2128
+ "epoch": 0.7789232531500573,
2129
+ "forget_loss": 2.95549085649327e-07,
2130
+ "retain_loss": 0.03642899915575981,
2131
+ "step": 680
2132
+ },
2133
+ {
2134
+ "epoch": 0.7789232531500573,
2135
+ "forget_loss": 7.595866691190167e-08,
2136
+ "retain_loss": 0.026957420632243156,
2137
+ "step": 680
2138
+ },
2139
+ {
2140
+ "epoch": 0.7789232531500573,
2141
+ "forget_loss": 3.971479429765168e-07,
2142
+ "retain_loss": 0.026554634794592857,
2143
+ "step": 680
2144
+ },
2145
+ {
2146
+ "epoch": 0.7789232531500573,
2147
+ "forget_loss": 1.0031957572209649e-05,
2148
+ "retain_loss": 0.055837519466876984,
2149
+ "step": 680
2150
+ },
2151
+ {
2152
+ "epoch": 0.7903780068728522,
2153
+ "grad_norm": 0.72265625,
2154
+ "learning_rate": 3.0443548387096776e-05,
2155
+ "loss": 0.032911685109138486,
2156
+ "step": 690
2157
+ },
2158
+ {
2159
+ "epoch": 0.7903780068728522,
2160
+ "forget_loss": 2.3226121470543148e-07,
2161
+ "retain_loss": 0.028471192345023155,
2162
+ "step": 690
2163
+ },
2164
+ {
2165
+ "epoch": 0.7903780068728522,
2166
+ "forget_loss": 2.1240731484795106e-07,
2167
+ "retain_loss": 0.061532407999038696,
2168
+ "step": 690
2169
+ },
2170
+ {
2171
+ "epoch": 0.7903780068728522,
2172
+ "forget_loss": 3.4258829373357e-07,
2173
+ "retain_loss": 0.045654185116291046,
2174
+ "step": 690
2175
+ },
2176
+ {
2177
+ "epoch": 0.7903780068728522,
2178
+ "forget_loss": 1.3982165114612144e-07,
2179
+ "retain_loss": 0.04419733211398125,
2180
+ "step": 690
2181
+ },
2182
+ {
2183
+ "epoch": 0.8018327605956472,
2184
+ "grad_norm": 0.369140625,
2185
+ "learning_rate": 3.015552995391705e-05,
2186
+ "loss": 0.03598496913909912,
2187
+ "step": 700
2188
+ },
2189
+ {
2190
+ "epoch": 0.8018327605956472,
2191
+ "forget_loss": 3.2066935773400473e-07,
2192
+ "retain_loss": 0.020428789779543877,
2193
+ "step": 700
2194
+ },
2195
+ {
2196
+ "epoch": 0.8018327605956472,
2197
+ "forget_loss": 9.008678603095177e-07,
2198
+ "retain_loss": 0.03619084134697914,
2199
+ "step": 700
2200
+ },
2201
+ {
2202
+ "epoch": 0.8018327605956472,
2203
+ "forget_loss": 3.202090113063605e-07,
2204
+ "retain_loss": 0.04894010350108147,
2205
+ "step": 700
2206
+ },
2207
+ {
2208
+ "epoch": 0.8018327605956472,
2209
+ "forget_loss": 3.2703263741495903e-07,
2210
+ "retain_loss": 0.039299145340919495,
2211
+ "step": 700
2212
+ },
2213
+ {
2214
+ "epoch": 0.8132875143184422,
2215
+ "grad_norm": 0.48046875,
2216
+ "learning_rate": 2.9867511520737328e-05,
2217
+ "loss": 0.03154398202896118,
2218
+ "step": 710
2219
+ },
2220
+ {
2221
+ "epoch": 0.8132875143184422,
2222
+ "forget_loss": 1.6021775195440568e-07,
2223
+ "retain_loss": 0.05344302952289581,
2224
+ "step": 710
2225
+ },
2226
+ {
2227
+ "epoch": 0.8132875143184422,
2228
+ "forget_loss": 3.327061222080374e-07,
2229
+ "retain_loss": 0.018441231921315193,
2230
+ "step": 710
2231
+ },
2232
+ {
2233
+ "epoch": 0.8132875143184422,
2234
+ "forget_loss": 1.2161015661149577e-07,
2235
+ "retain_loss": 0.01777707226574421,
2236
+ "step": 710
2237
+ },
2238
+ {
2239
+ "epoch": 0.8132875143184422,
2240
+ "forget_loss": 2.601665585189039e-07,
2241
+ "retain_loss": 0.021317142993211746,
2242
+ "step": 710
2243
+ },
2244
+ {
2245
+ "epoch": 0.8247422680412371,
2246
+ "grad_norm": 0.71484375,
2247
+ "learning_rate": 2.9579493087557604e-05,
2248
+ "loss": 0.031114420294761656,
2249
+ "step": 720
2250
+ },
2251
+ {
2252
+ "epoch": 0.8247422680412371,
2253
+ "forget_loss": 3.656723990275168e-08,
2254
+ "retain_loss": 0.02041560783982277,
2255
+ "step": 720
2256
+ },
2257
+ {
2258
+ "epoch": 0.8247422680412371,
2259
+ "forget_loss": 7.133304791295814e-08,
2260
+ "retain_loss": 0.0420437790453434,
2261
+ "step": 720
2262
+ },
2263
+ {
2264
+ "epoch": 0.8247422680412371,
2265
+ "forget_loss": 1.2686764705449605e-07,
2266
+ "retain_loss": 0.03707875683903694,
2267
+ "step": 720
2268
+ },
2269
+ {
2270
+ "epoch": 0.8247422680412371,
2271
+ "forget_loss": 3.098398337897379e-07,
2272
+ "retain_loss": 0.03388817980885506,
2273
+ "step": 720
2274
+ },
2275
+ {
2276
+ "epoch": 0.8361970217640321,
2277
+ "grad_norm": 0.62109375,
2278
+ "learning_rate": 2.9291474654377883e-05,
2279
+ "loss": 0.031438523530960084,
2280
+ "step": 730
2281
+ },
2282
+ {
2283
+ "epoch": 0.8361970217640321,
2284
+ "forget_loss": 6.143314408291189e-07,
2285
+ "retain_loss": 0.037241023033857346,
2286
+ "step": 730
2287
+ },
2288
+ {
2289
+ "epoch": 0.8361970217640321,
2290
+ "forget_loss": 1.567052549944492e-07,
2291
+ "retain_loss": 0.029914623126387596,
2292
+ "step": 730
2293
+ },
2294
+ {
2295
+ "epoch": 0.8361970217640321,
2296
+ "forget_loss": 1.7319891298939183e-07,
2297
+ "retain_loss": 0.026813646778464317,
2298
+ "step": 730
2299
+ },
2300
+ {
2301
+ "epoch": 0.8361970217640321,
2302
+ "forget_loss": 5.76246179662121e-08,
2303
+ "retain_loss": 0.036752790212631226,
2304
+ "step": 730
2305
+ },
2306
+ {
2307
+ "epoch": 0.847651775486827,
2308
+ "grad_norm": 0.447265625,
2309
+ "learning_rate": 2.9003456221198155e-05,
2310
+ "loss": 0.028349018096923827,
2311
+ "step": 740
2312
+ },
2313
+ {
2314
+ "epoch": 0.847651775486827,
2315
+ "forget_loss": 1.0741296136984602e-06,
2316
+ "retain_loss": 0.028334422037005424,
2317
+ "step": 740
2318
+ },
2319
+ {
2320
+ "epoch": 0.847651775486827,
2321
+ "forget_loss": 5.47009335605253e-07,
2322
+ "retain_loss": 0.03212451562285423,
2323
+ "step": 740
2324
+ },
2325
+ {
2326
+ "epoch": 0.847651775486827,
2327
+ "forget_loss": 7.226554998851498e-08,
2328
+ "retain_loss": 0.038880933076143265,
2329
+ "step": 740
2330
+ },
2331
+ {
2332
+ "epoch": 0.847651775486827,
2333
+ "forget_loss": 1.0503694625185744e-07,
2334
+ "retain_loss": 0.020103227347135544,
2335
+ "step": 740
2336
+ },
2337
+ {
2338
+ "epoch": 0.8591065292096219,
2339
+ "grad_norm": 0.62109375,
2340
+ "learning_rate": 2.8715437788018435e-05,
2341
+ "loss": 0.027798715233802795,
2342
+ "step": 750
2343
+ },
2344
+ {
2345
+ "epoch": 0.8591065292096219,
2346
+ "forget_loss": 9.740627149312786e-08,
2347
+ "retain_loss": 0.03145676106214523,
2348
+ "step": 750
2349
+ },
2350
+ {
2351
+ "epoch": 0.8591065292096219,
2352
+ "forget_loss": 8.817877272804253e-08,
2353
+ "retain_loss": 0.0294309351593256,
2354
+ "step": 750
2355
+ },
2356
+ {
2357
+ "epoch": 0.8591065292096219,
2358
+ "forget_loss": 2.1280260398270912e-07,
2359
+ "retain_loss": 0.01636165753006935,
2360
+ "step": 750
2361
+ },
2362
+ {
2363
+ "epoch": 0.8591065292096219,
2364
+ "forget_loss": 2.8950364594493294e-07,
2365
+ "retain_loss": 0.03548318147659302,
2366
+ "step": 750
2367
+ },
2368
+ {
2369
+ "epoch": 0.870561282932417,
2370
+ "grad_norm": 0.26171875,
2371
+ "learning_rate": 2.8427419354838714e-05,
2372
+ "loss": 0.027926307916641236,
2373
+ "step": 760
2374
+ },
2375
+ {
2376
+ "epoch": 0.870561282932417,
2377
+ "forget_loss": 1.0444756526339916e-07,
2378
+ "retain_loss": 0.035280242562294006,
2379
+ "step": 760
2380
+ },
2381
+ {
2382
+ "epoch": 0.870561282932417,
2383
+ "forget_loss": 1.4455953589731507e-07,
2384
+ "retain_loss": 0.03736939653754234,
2385
+ "step": 760
2386
+ },
2387
+ {
2388
+ "epoch": 0.870561282932417,
2389
+ "forget_loss": 8.093392267483068e-08,
2390
+ "retain_loss": 0.02323060669004917,
2391
+ "step": 760
2392
+ },
2393
+ {
2394
+ "epoch": 0.870561282932417,
2395
+ "forget_loss": 4.825432711186295e-07,
2396
+ "retain_loss": 0.028354808688163757,
2397
+ "step": 760
2398
+ },
2399
+ {
2400
+ "epoch": 0.8820160366552119,
2401
+ "grad_norm": 0.419921875,
2402
+ "learning_rate": 2.8139400921658986e-05,
2403
+ "loss": 0.02746644616127014,
2404
+ "step": 770
2405
+ },
2406
+ {
2407
+ "epoch": 0.8820160366552119,
2408
+ "forget_loss": 6.515568173881547e-08,
2409
+ "retain_loss": 0.022168956696987152,
2410
+ "step": 770
2411
+ },
2412
+ {
2413
+ "epoch": 0.8820160366552119,
2414
+ "forget_loss": 4.0877694118535146e-08,
2415
+ "retain_loss": 0.02454095333814621,
2416
+ "step": 770
2417
+ },
2418
+ {
2419
+ "epoch": 0.8820160366552119,
2420
+ "forget_loss": 4.9038700211667674e-08,
2421
+ "retain_loss": 0.009829733520746231,
2422
+ "step": 770
2423
+ },
2424
+ {
2425
+ "epoch": 0.8820160366552119,
2426
+ "forget_loss": 2.664446583366953e-07,
2427
+ "retain_loss": 0.01737787388265133,
2428
+ "step": 770
2429
+ },
2430
+ {
2431
+ "epoch": 0.8934707903780069,
2432
+ "grad_norm": 0.482421875,
2433
+ "learning_rate": 2.7851382488479265e-05,
2434
+ "loss": 0.025814527273178102,
2435
+ "step": 780
2436
+ },
2437
+ {
2438
+ "epoch": 0.8934707903780069,
2439
+ "forget_loss": 3.281249050246515e-08,
2440
+ "retain_loss": 0.020622696727514267,
2441
+ "step": 780
2442
+ },
2443
+ {
2444
+ "epoch": 0.8934707903780069,
2445
+ "forget_loss": 5.478144515791428e-08,
2446
+ "retain_loss": 0.020266670733690262,
2447
+ "step": 780
2448
+ },
2449
+ {
2450
+ "epoch": 0.8934707903780069,
2451
+ "forget_loss": 6.129197061000013e-08,
2452
+ "retain_loss": 0.018652314320206642,
2453
+ "step": 780
2454
+ },
2455
+ {
2456
+ "epoch": 0.8934707903780069,
2457
+ "forget_loss": 2.284050850676067e-07,
2458
+ "retain_loss": 0.013459251262247562,
2459
+ "step": 780
2460
+ },
2461
+ {
2462
+ "epoch": 0.9049255441008018,
2463
+ "grad_norm": 0.51953125,
2464
+ "learning_rate": 2.756336405529954e-05,
2465
+ "loss": 0.027006852626800536,
2466
+ "step": 790
2467
+ },
2468
+ {
2469
+ "epoch": 0.9049255441008018,
2470
+ "forget_loss": 2.6094051364111692e-08,
2471
+ "retain_loss": 0.018663065508008003,
2472
+ "step": 790
2473
+ },
2474
+ {
2475
+ "epoch": 0.9049255441008018,
2476
+ "forget_loss": 2.7085914666713506e-07,
2477
+ "retain_loss": 0.009133810177445412,
2478
+ "step": 790
2479
+ },
2480
+ {
2481
+ "epoch": 0.9049255441008018,
2482
+ "forget_loss": 8.054676214896972e-08,
2483
+ "retain_loss": 0.02572845108807087,
2484
+ "step": 790
2485
+ },
2486
+ {
2487
+ "epoch": 0.9049255441008018,
2488
+ "forget_loss": 1.0726026289376023e-07,
2489
+ "retain_loss": 0.024219941347837448,
2490
+ "step": 790
2491
+ },
2492
+ {
2493
+ "epoch": 0.9163802978235968,
2494
+ "grad_norm": 0.328125,
2495
+ "learning_rate": 2.727534562211982e-05,
2496
+ "loss": 0.02423024773597717,
2497
+ "step": 800
2498
+ },
2499
+ {
2500
+ "epoch": 0.9163802978235968,
2501
+ "forget_loss": 7.722936601339825e-08,
2502
+ "retain_loss": 0.04125501587986946,
2503
+ "step": 800
2504
+ },
2505
+ {
2506
+ "epoch": 0.9163802978235968,
2507
+ "forget_loss": 9.053938043734888e-08,
2508
+ "retain_loss": 0.02275395393371582,
2509
+ "step": 800
2510
+ },
2511
+ {
2512
+ "epoch": 0.9163802978235968,
2513
+ "forget_loss": 1.4687527993828553e-07,
2514
+ "retain_loss": 0.026991432532668114,
2515
+ "step": 800
2516
+ },
2517
+ {
2518
+ "epoch": 0.9163802978235968,
2519
+ "forget_loss": 3.380711177669582e-07,
2520
+ "retain_loss": 0.025429554283618927,
2521
+ "step": 800
2522
+ },
2523
+ {
2524
+ "epoch": 0.9278350515463918,
2525
+ "grad_norm": 0.74609375,
2526
+ "learning_rate": 2.6987327188940093e-05,
2527
+ "loss": 0.027974823117256166,
2528
+ "step": 810
2529
+ },
2530
+ {
2531
+ "epoch": 0.9278350515463918,
2532
+ "forget_loss": 2.2884126593680776e-08,
2533
+ "retain_loss": 0.034118637442588806,
2534
+ "step": 810
2535
+ },
2536
+ {
2537
+ "epoch": 0.9278350515463918,
2538
+ "forget_loss": 2.2747957473256974e-07,
2539
+ "retain_loss": 0.016463886946439743,
2540
+ "step": 810
2541
+ },
2542
+ {
2543
+ "epoch": 0.9278350515463918,
2544
+ "forget_loss": 2.8712694444266162e-08,
2545
+ "retain_loss": 0.05120648816227913,
2546
+ "step": 810
2547
+ },
2548
+ {
2549
+ "epoch": 0.9278350515463918,
2550
+ "forget_loss": 3.9451157363146194e-07,
2551
+ "retain_loss": 0.015402361750602722,
2552
+ "step": 810
2553
+ },
2554
+ {
2555
+ "epoch": 0.9392898052691867,
2556
+ "grad_norm": 0.232421875,
2557
+ "learning_rate": 2.6699308755760372e-05,
2558
+ "loss": 0.025883400440216066,
2559
+ "step": 820
2560
+ },
2561
+ {
2562
+ "epoch": 0.9392898052691867,
2563
+ "forget_loss": 6.696375720594006e-08,
2564
+ "retain_loss": 0.03956631198525429,
2565
+ "step": 820
2566
+ },
2567
+ {
2568
+ "epoch": 0.9392898052691867,
2569
+ "forget_loss": 3.5951167376424564e-08,
2570
+ "retain_loss": 0.024411948397755623,
2571
+ "step": 820
2572
+ },
2573
+ {
2574
+ "epoch": 0.9392898052691867,
2575
+ "forget_loss": 1.2076303335106786e-07,
2576
+ "retain_loss": 0.03519268333911896,
2577
+ "step": 820
2578
+ },
2579
+ {
2580
+ "epoch": 0.9392898052691867,
2581
+ "forget_loss": 3.650887236972267e-08,
2582
+ "retain_loss": 0.022809229791164398,
2583
+ "step": 820
2584
+ },
2585
+ {
2586
+ "epoch": 0.9507445589919816,
2587
+ "grad_norm": 0.181640625,
2588
+ "learning_rate": 2.6411290322580645e-05,
2589
+ "loss": 0.026767241954803466,
2590
+ "step": 830
2591
+ },
2592
+ {
2593
+ "epoch": 0.9507445589919816,
2594
+ "forget_loss": 1.1937875399326003e-07,
2595
+ "retain_loss": 0.009928344748914242,
2596
+ "step": 830
2597
+ },
2598
+ {
2599
+ "epoch": 0.9507445589919816,
2600
+ "forget_loss": 4.2366110619695974e-07,
2601
+ "retain_loss": 0.010477675125002861,
2602
+ "step": 830
2603
+ },
2604
+ {
2605
+ "epoch": 0.9507445589919816,
2606
+ "forget_loss": 9.538210576920392e-08,
2607
+ "retain_loss": 0.030232908204197884,
2608
+ "step": 830
2609
+ },
2610
+ {
2611
+ "epoch": 0.9507445589919816,
2612
+ "forget_loss": 4.3475404254422756e-07,
2613
+ "retain_loss": 0.027472183108329773,
2614
+ "step": 830
2615
+ },
2616
+ {
2617
+ "epoch": 0.9621993127147767,
2618
+ "grad_norm": 0.38671875,
2619
+ "learning_rate": 2.6123271889400924e-05,
2620
+ "loss": 0.026079732179641723,
2621
+ "step": 840
2622
+ },
2623
+ {
2624
+ "epoch": 0.9621993127147767,
2625
+ "forget_loss": 9.560527303165145e-08,
2626
+ "retain_loss": 0.035694848746061325,
2627
+ "step": 840
2628
+ },
2629
+ {
2630
+ "epoch": 0.9621993127147767,
2631
+ "forget_loss": 9.010035739720479e-08,
2632
+ "retain_loss": 0.02770412340760231,
2633
+ "step": 840
2634
+ },
2635
+ {
2636
+ "epoch": 0.9621993127147767,
2637
+ "forget_loss": 1.2952777979080565e-06,
2638
+ "retain_loss": 0.025091225281357765,
2639
+ "step": 840
2640
+ },
2641
+ {
2642
+ "epoch": 0.9621993127147767,
2643
+ "forget_loss": 1.8988541228281974e-07,
2644
+ "retain_loss": 0.02703693136572838,
2645
+ "step": 840
2646
+ },
2647
+ {
2648
+ "epoch": 0.9736540664375716,
2649
+ "grad_norm": 0.62109375,
2650
+ "learning_rate": 2.5835253456221196e-05,
2651
+ "loss": 0.024774086475372315,
2652
+ "step": 850
2653
+ },
2654
+ {
2655
+ "epoch": 0.9736540664375716,
2656
+ "forget_loss": 2.6555783350090678e-08,
2657
+ "retain_loss": 0.033134713768959045,
2658
+ "step": 850
2659
+ },
2660
+ {
2661
+ "epoch": 0.9736540664375716,
2662
+ "forget_loss": 2.7541537406250427e-07,
2663
+ "retain_loss": 0.018386946991086006,
2664
+ "step": 850
2665
+ },
2666
+ {
2667
+ "epoch": 0.9736540664375716,
2668
+ "forget_loss": 2.7093907561948072e-08,
2669
+ "retain_loss": 0.022385219112038612,
2670
+ "step": 850
2671
+ },
2672
+ {
2673
+ "epoch": 0.9736540664375716,
2674
+ "forget_loss": 6.318795442439296e-08,
2675
+ "retain_loss": 0.03880416601896286,
2676
+ "step": 850
2677
+ },
2678
+ {
2679
+ "epoch": 0.9851088201603666,
2680
+ "grad_norm": 0.57421875,
2681
+ "learning_rate": 2.5547235023041475e-05,
2682
+ "loss": 0.02385828346014023,
2683
+ "step": 860
2684
+ },
2685
+ {
2686
+ "epoch": 0.9851088201603666,
2687
+ "forget_loss": 3.6007147485861424e-08,
2688
+ "retain_loss": 0.03307569772005081,
2689
+ "step": 860
2690
+ },
2691
+ {
2692
+ "epoch": 0.9851088201603666,
2693
+ "forget_loss": 8.178755450671815e-08,
2694
+ "retain_loss": 0.02462468110024929,
2695
+ "step": 860
2696
+ },
2697
+ {
2698
+ "epoch": 0.9851088201603666,
2699
+ "forget_loss": 6.628091853144724e-08,
2700
+ "retain_loss": 0.020657110959291458,
2701
+ "step": 860
2702
+ },
2703
+ {
2704
+ "epoch": 0.9851088201603666,
2705
+ "forget_loss": 6.628161486332829e-08,
2706
+ "retain_loss": 0.032617855817079544,
2707
+ "step": 860
2708
+ },
2709
+ {
2710
+ "epoch": 0.9965635738831615,
2711
+ "grad_norm": 0.236328125,
2712
+ "learning_rate": 2.525921658986175e-05,
2713
+ "loss": 0.024864019453525545,
2714
+ "step": 870
2715
+ },
2716
+ {
2717
+ "epoch": 0.9965635738831615,
2718
+ "forget_loss": 7.505889243475394e-08,
2719
+ "retain_loss": 0.01775062270462513,
2720
+ "step": 870
2721
+ },
2722
+ {
2723
+ "epoch": 0.9965635738831615,
2724
+ "forget_loss": 1.0310161258075823e-07,
2725
+ "retain_loss": 0.02414717711508274,
2726
+ "step": 870
2727
+ },
2728
+ {
2729
+ "epoch": 0.9965635738831615,
2730
+ "forget_loss": 2.3729118936444138e-07,
2731
+ "retain_loss": 0.027325890958309174,
2732
+ "step": 870
2733
+ },
2734
+ {
2735
+ "epoch": 0.9965635738831615,
2736
+ "forget_loss": 2.6968228539203665e-08,
2737
+ "retain_loss": 0.022969461977481842,
2738
+ "step": 870
2739
+ },
2740
+ {
2741
+ "epoch": 1.0080183276059564,
2742
+ "grad_norm": 0.2041015625,
2743
+ "learning_rate": 2.497119815668203e-05,
2744
+ "loss": 0.025278234481811525,
2745
+ "step": 880
2746
+ },
2747
+ {
2748
+ "epoch": 1.0080183276059564,
2749
+ "forget_loss": 1.2440519014944584e-07,
2750
+ "retain_loss": 0.028113704174757004,
2751
+ "step": 880
2752
+ },
2753
+ {
2754
+ "epoch": 1.0080183276059564,
2755
+ "forget_loss": 6.923085038579302e-08,
2756
+ "retain_loss": 0.035193175077438354,
2757
+ "step": 880
2758
+ },
2759
+ {
2760
+ "epoch": 1.0080183276059564,
2761
+ "forget_loss": 3.75761644022532e-08,
2762
+ "retain_loss": 0.018518565222620964,
2763
+ "step": 880
2764
+ },
2765
+ {
2766
+ "epoch": 1.0080183276059564,
2767
+ "forget_loss": 2.242090957338405e-08,
2768
+ "retain_loss": 0.017261430621147156,
2769
+ "step": 880
2770
+ },
2771
+ {
2772
+ "epoch": 1.0194730813287514,
2773
+ "grad_norm": 0.37109375,
2774
+ "learning_rate": 2.4683179723502306e-05,
2775
+ "loss": 0.024524253606796265,
2776
+ "step": 890
2777
+ },
2778
+ {
2779
+ "epoch": 1.0194730813287514,
2780
+ "forget_loss": 3.599013496113912e-07,
2781
+ "retain_loss": 0.034332528710365295,
2782
+ "step": 890
2783
+ },
2784
+ {
2785
+ "epoch": 1.0194730813287514,
2786
+ "forget_loss": 1.7287271703025908e-07,
2787
+ "retain_loss": 0.01914825290441513,
2788
+ "step": 890
2789
+ },
2790
+ {
2791
+ "epoch": 1.0194730813287514,
2792
+ "forget_loss": 9.782493748389243e-08,
2793
+ "retain_loss": 0.023516574874520302,
2794
+ "step": 890
2795
+ },
2796
+ {
2797
+ "epoch": 1.0194730813287514,
2798
+ "forget_loss": 1.5106013506738236e-07,
2799
+ "retain_loss": 0.019009225070476532,
2800
+ "step": 890
2801
+ },
2802
+ {
2803
+ "epoch": 1.0309278350515463,
2804
+ "grad_norm": 0.1787109375,
2805
+ "learning_rate": 2.4395161290322582e-05,
2806
+ "loss": 0.024021840095520018,
2807
+ "step": 900
2808
+ },
2809
+ {
2810
+ "epoch": 1.0309278350515463,
2811
+ "forget_loss": 2.6426999966133735e-07,
2812
+ "retain_loss": 0.029771650210022926,
2813
+ "step": 900
2814
+ },
2815
+ {
2816
+ "epoch": 1.0309278350515463,
2817
+ "forget_loss": 5.678279535459296e-08,
2818
+ "retain_loss": 0.02570934407413006,
2819
+ "step": 900
2820
+ },
2821
+ {
2822
+ "epoch": 1.0309278350515463,
2823
+ "forget_loss": 2.8987988898165895e-08,
2824
+ "retain_loss": 0.02868933044373989,
2825
+ "step": 900
2826
+ },
2827
+ {
2828
+ "epoch": 1.0309278350515463,
2829
+ "forget_loss": 1.6226587717937946e-07,
2830
+ "retain_loss": 0.024329733103513718,
2831
+ "step": 900
2832
+ },
2833
+ {
2834
+ "epoch": 1.0423825887743414,
2835
+ "grad_norm": 0.259765625,
2836
+ "learning_rate": 2.4107142857142858e-05,
2837
+ "loss": 0.02456314116716385,
2838
+ "step": 910
2839
+ },
2840
+ {
2841
+ "epoch": 1.0423825887743414,
2842
+ "forget_loss": 3.358028379807365e-07,
2843
+ "retain_loss": 0.014405038207769394,
2844
+ "step": 910
2845
+ },
2846
+ {
2847
+ "epoch": 1.0423825887743414,
2848
+ "forget_loss": 8.003183893379173e-08,
2849
+ "retain_loss": 0.012783446349203587,
2850
+ "step": 910
2851
+ },
2852
+ {
2853
+ "epoch": 1.0423825887743414,
2854
+ "forget_loss": 3.088370874593238e-07,
2855
+ "retain_loss": 0.04180136322975159,
2856
+ "step": 910
2857
+ },
2858
+ {
2859
+ "epoch": 1.0423825887743414,
2860
+ "forget_loss": 1.6710265526853618e-07,
2861
+ "retain_loss": 0.015366771258413792,
2862
+ "step": 910
2863
+ },
2864
+ {
2865
+ "epoch": 1.0538373424971363,
2866
+ "grad_norm": 0.3359375,
2867
+ "learning_rate": 2.3819124423963134e-05,
2868
+ "loss": 0.023695509135723113,
2869
+ "step": 920
2870
+ },
2871
+ {
2872
+ "epoch": 1.0538373424971363,
2873
+ "forget_loss": 1.9374034820884845e-07,
2874
+ "retain_loss": 0.0453634150326252,
2875
+ "step": 920
2876
+ },
2877
+ {
2878
+ "epoch": 1.0538373424971363,
2879
+ "forget_loss": 4.01810851258233e-08,
2880
+ "retain_loss": 0.027023399248719215,
2881
+ "step": 920
2882
+ },
2883
+ {
2884
+ "epoch": 1.0538373424971363,
2885
+ "forget_loss": 3.384187508004288e-08,
2886
+ "retain_loss": 0.011590844951570034,
2887
+ "step": 920
2888
+ },
2889
+ {
2890
+ "epoch": 1.0538373424971363,
2891
+ "forget_loss": 2.813683330771255e-08,
2892
+ "retain_loss": 0.02153298258781433,
2893
+ "step": 920
2894
+ },
2895
+ {
2896
+ "epoch": 1.0652920962199313,
2897
+ "grad_norm": 0.228515625,
2898
+ "learning_rate": 2.3531105990783413e-05,
2899
+ "loss": 0.023248623311519622,
2900
+ "step": 930
2901
+ },
2902
+ {
2903
+ "epoch": 1.0652920962199313,
2904
+ "forget_loss": 3.62256820096718e-08,
2905
+ "retain_loss": 0.04065975919365883,
2906
+ "step": 930
2907
+ },
2908
+ {
2909
+ "epoch": 1.0652920962199313,
2910
+ "forget_loss": 1.747634712501167e-07,
2911
+ "retain_loss": 0.022159142419695854,
2912
+ "step": 930
2913
+ },
2914
+ {
2915
+ "epoch": 1.0652920962199313,
2916
+ "forget_loss": 8.5983149489266e-09,
2917
+ "retain_loss": 0.011597687378525734,
2918
+ "step": 930
2919
+ },
2920
+ {
2921
+ "epoch": 1.0652920962199313,
2922
+ "forget_loss": 2.4163021805634344e-08,
2923
+ "retain_loss": 0.015060486271977425,
2924
+ "step": 930
2925
+ },
2926
+ {
2927
+ "epoch": 1.0767468499427262,
2928
+ "grad_norm": 0.216796875,
2929
+ "learning_rate": 2.324308755760369e-05,
2930
+ "loss": 0.022622266411781312,
2931
+ "step": 940
2932
+ },
2933
+ {
2934
+ "epoch": 1.0767468499427262,
2935
+ "forget_loss": 2.2849988567941182e-08,
2936
+ "retain_loss": 0.014040598645806313,
2937
+ "step": 940
2938
+ },
2939
+ {
2940
+ "epoch": 1.0767468499427262,
2941
+ "forget_loss": 1.0765910829491077e-08,
2942
+ "retain_loss": 0.021398184821009636,
2943
+ "step": 940
2944
+ },
2945
+ {
2946
+ "epoch": 1.0767468499427262,
2947
+ "forget_loss": 3.679618885144009e-07,
2948
+ "retain_loss": 0.01932062953710556,
2949
+ "step": 940
2950
+ },
2951
+ {
2952
+ "epoch": 1.0767468499427262,
2953
+ "forget_loss": 4.7389018931198734e-08,
2954
+ "retain_loss": 0.03855086490511894,
2955
+ "step": 940
2956
+ },
2957
+ {
2958
+ "epoch": 1.088201603665521,
2959
+ "grad_norm": 0.1669921875,
2960
+ "learning_rate": 2.2955069124423965e-05,
2961
+ "loss": 0.022816181182861328,
2962
+ "step": 950
2963
+ },
2964
+ {
2965
+ "epoch": 1.088201603665521,
2966
+ "forget_loss": 5.09719377816964e-08,
2967
+ "retain_loss": 0.01380919013172388,
2968
+ "step": 950
2969
+ },
2970
+ {
2971
+ "epoch": 1.088201603665521,
2972
+ "forget_loss": 2.481048966274102e-08,
2973
+ "retain_loss": 0.02061321772634983,
2974
+ "step": 950
2975
+ },
2976
+ {
2977
+ "epoch": 1.088201603665521,
2978
+ "forget_loss": 3.024410943908151e-08,
2979
+ "retain_loss": 0.021798167377710342,
2980
+ "step": 950
2981
+ },
2982
+ {
2983
+ "epoch": 1.088201603665521,
2984
+ "forget_loss": 6.814799036192198e-08,
2985
+ "retain_loss": 0.02376674860715866,
2986
+ "step": 950
2987
+ },
2988
+ {
2989
+ "epoch": 1.0996563573883162,
2990
+ "grad_norm": 0.2099609375,
2991
+ "learning_rate": 2.266705069124424e-05,
2992
+ "loss": 0.023587432503700257,
2993
+ "step": 960
2994
+ },
2995
+ {
2996
+ "epoch": 1.0996563573883162,
2997
+ "forget_loss": 7.863557982545899e-08,
2998
+ "retain_loss": 0.034126508980989456,
2999
+ "step": 960
3000
+ },
3001
+ {
3002
+ "epoch": 1.0996563573883162,
3003
+ "forget_loss": 1.2243476454898428e-08,
3004
+ "retain_loss": 0.025233590975403786,
3005
+ "step": 960
3006
+ },
3007
+ {
3008
+ "epoch": 1.0996563573883162,
3009
+ "forget_loss": 2.9529656941917892e-08,
3010
+ "retain_loss": 0.026653651148080826,
3011
+ "step": 960
3012
+ },
3013
+ {
3014
+ "epoch": 1.0996563573883162,
3015
+ "forget_loss": 1.9638031290014624e-07,
3016
+ "retain_loss": 0.021439678966999054,
3017
+ "step": 960
3018
+ },
3019
+ {
3020
+ "epoch": 1.1111111111111112,
3021
+ "grad_norm": 0.1318359375,
3022
+ "learning_rate": 2.2379032258064516e-05,
3023
+ "loss": 0.02087740898132324,
3024
+ "step": 970
3025
+ },
3026
+ {
3027
+ "epoch": 1.1111111111111112,
3028
+ "forget_loss": 5.724940876916662e-08,
3029
+ "retain_loss": 0.031759753823280334,
3030
+ "step": 970
3031
+ },
3032
+ {
3033
+ "epoch": 1.1111111111111112,
3034
+ "forget_loss": 2.719896485814388e-08,
3035
+ "retain_loss": 0.02280360274016857,
3036
+ "step": 970
3037
+ },
3038
+ {
3039
+ "epoch": 1.1111111111111112,
3040
+ "forget_loss": 1.3328020997960266e-07,
3041
+ "retain_loss": 0.034389056265354156,
3042
+ "step": 970
3043
+ },
3044
+ {
3045
+ "epoch": 1.1111111111111112,
3046
+ "forget_loss": 2.331986337367198e-08,
3047
+ "retain_loss": 0.027264058589935303,
3048
+ "step": 970
3049
+ },
3050
+ {
3051
+ "epoch": 1.122565864833906,
3052
+ "grad_norm": 0.2255859375,
3053
+ "learning_rate": 2.2091013824884792e-05,
3054
+ "loss": 0.02323000878095627,
3055
+ "step": 980
3056
+ },
3057
+ {
3058
+ "epoch": 1.122565864833906,
3059
+ "forget_loss": 6.373924179570167e-07,
3060
+ "retain_loss": 0.022899720817804337,
3061
+ "step": 980
3062
+ },
3063
+ {
3064
+ "epoch": 1.122565864833906,
3065
+ "forget_loss": 8.226674452771476e-08,
3066
+ "retain_loss": 0.017953844740986824,
3067
+ "step": 980
3068
+ },
3069
+ {
3070
+ "epoch": 1.122565864833906,
3071
+ "forget_loss": 3.995974395820667e-08,
3072
+ "retain_loss": 0.023313166573643684,
3073
+ "step": 980
3074
+ },
3075
+ {
3076
+ "epoch": 1.122565864833906,
3077
+ "forget_loss": 5.536628577829106e-08,
3078
+ "retain_loss": 0.02542729862034321,
3079
+ "step": 980
3080
+ },
3081
+ {
3082
+ "epoch": 1.134020618556701,
3083
+ "grad_norm": 0.142578125,
3084
+ "learning_rate": 2.1802995391705068e-05,
3085
+ "loss": 0.02163739502429962,
3086
+ "step": 990
3087
+ },
3088
+ {
3089
+ "epoch": 1.134020618556701,
3090
+ "forget_loss": 4.036724732259245e-08,
3091
+ "retain_loss": 0.03384210169315338,
3092
+ "step": 990
3093
+ },
3094
+ {
3095
+ "epoch": 1.134020618556701,
3096
+ "forget_loss": 1.0333982913834916e-07,
3097
+ "retain_loss": 0.021868150681257248,
3098
+ "step": 990
3099
+ },
3100
+ {
3101
+ "epoch": 1.134020618556701,
3102
+ "forget_loss": 5.9504564831058815e-08,
3103
+ "retain_loss": 0.010602345690131187,
3104
+ "step": 990
3105
+ },
3106
+ {
3107
+ "epoch": 1.134020618556701,
3108
+ "forget_loss": 4.604233794225365e-08,
3109
+ "retain_loss": 0.03545906022191048,
3110
+ "step": 990
3111
+ },
3112
+ {
3113
+ "epoch": 1.145475372279496,
3114
+ "grad_norm": 0.1796875,
3115
+ "learning_rate": 2.1514976958525347e-05,
3116
+ "loss": 0.023172636330127717,
3117
+ "step": 1000
3118
+ }
3119
+ ],
3120
+ "logging_steps": 10,
3121
+ "max_steps": 1746,
3122
+ "num_input_tokens_seen": 0,
3123
+ "num_train_epochs": 2,
3124
+ "save_steps": 200,
3125
+ "stateful_callbacks": {
3126
+ "TrainerControl": {
3127
+ "args": {
3128
+ "should_epoch_stop": false,
3129
+ "should_evaluate": false,
3130
+ "should_log": false,
3131
+ "should_save": true,
3132
+ "should_training_stop": false
3133
+ },
3134
+ "attributes": {}
3135
+ }
3136
+ },
3137
+ "total_flos": 0.0,
3138
+ "train_batch_size": 8,
3139
+ "trial_name": null,
3140
+ "trial_params": null
3141
+ }
checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70e0b39ebb6ac634552b14ecd3387db7085650b2028306aaf485cd5ea3c517f8
3
+ size 5329
checkpoint-1200/config.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "bos_token_id": 2,
7
+ "dtype": "bfloat16",
8
+ "eoi_token_index": 256000,
9
+ "eos_token_id": 1,
10
+ "image_token_index": 262144,
11
+ "initializer_range": 0.02,
12
+ "mm_tokens_per_image": 256,
13
+ "model_type": "gemma3",
14
+ "pad_token_id": 0,
15
+ "text_config": {
16
+ "_sliding_window_pattern": 6,
17
+ "attention_bias": false,
18
+ "attention_dropout": 0.0,
19
+ "attn_logit_softcapping": null,
20
+ "bos_token_id": 2,
21
+ "dtype": "bfloat16",
22
+ "eos_token_id": 1,
23
+ "final_logit_softcapping": null,
24
+ "head_dim": 256,
25
+ "hidden_activation": "gelu_pytorch_tanh",
26
+ "hidden_size": 3840,
27
+ "initializer_range": 0.02,
28
+ "intermediate_size": 15360,
29
+ "layer_types": [
30
+ "sliding_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "full_attention",
36
+ "sliding_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "sliding_attention",
40
+ "sliding_attention",
41
+ "full_attention",
42
+ "sliding_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "sliding_attention",
46
+ "sliding_attention",
47
+ "full_attention",
48
+ "sliding_attention",
49
+ "sliding_attention",
50
+ "sliding_attention",
51
+ "sliding_attention",
52
+ "sliding_attention",
53
+ "full_attention",
54
+ "sliding_attention",
55
+ "sliding_attention",
56
+ "sliding_attention",
57
+ "sliding_attention",
58
+ "sliding_attention",
59
+ "full_attention",
60
+ "sliding_attention",
61
+ "sliding_attention",
62
+ "sliding_attention",
63
+ "sliding_attention",
64
+ "sliding_attention",
65
+ "full_attention",
66
+ "sliding_attention",
67
+ "sliding_attention",
68
+ "sliding_attention",
69
+ "sliding_attention",
70
+ "sliding_attention",
71
+ "full_attention",
72
+ "sliding_attention",
73
+ "sliding_attention",
74
+ "sliding_attention",
75
+ "sliding_attention",
76
+ "sliding_attention",
77
+ "full_attention"
78
+ ],
79
+ "max_position_embeddings": 131072,
80
+ "model_type": "gemma3_text",
81
+ "num_attention_heads": 16,
82
+ "num_hidden_layers": 48,
83
+ "num_key_value_heads": 8,
84
+ "pad_token_id": 0,
85
+ "query_pre_attn_scalar": 256,
86
+ "rms_norm_eps": 1e-06,
87
+ "rope_parameters": {
88
+ "full_attention": {
89
+ "factor": 8.0,
90
+ "rope_theta": 1000000.0,
91
+ "rope_type": "linear"
92
+ },
93
+ "sliding_attention": {
94
+ "rope_theta": 10000.0,
95
+ "rope_type": "default"
96
+ }
97
+ },
98
+ "sliding_window": 1024,
99
+ "tie_word_embeddings": true,
100
+ "use_bidirectional_attention": false,
101
+ "use_cache": true,
102
+ "vocab_size": 262208
103
+ },
104
+ "tie_word_embeddings": true,
105
+ "transformers_version": "5.6.0",
106
+ "use_cache": false,
107
+ "vision_config": {
108
+ "attention_dropout": 0.0,
109
+ "dtype": "bfloat16",
110
+ "hidden_act": "gelu_pytorch_tanh",
111
+ "hidden_size": 1152,
112
+ "image_size": 896,
113
+ "intermediate_size": 4304,
114
+ "layer_norm_eps": 1e-06,
115
+ "model_type": "siglip_vision_model",
116
+ "num_attention_heads": 16,
117
+ "num_channels": 3,
118
+ "num_hidden_layers": 27,
119
+ "patch_size": 14,
120
+ "vision_use_head": false
121
+ }
122
+ }
checkpoint-1200/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "cache_implementation": "hybrid",
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 1,
7
+ 1,
8
+ 106
9
+ ],
10
+ "pad_token_id": 0,
11
+ "top_k": 64,
12
+ "top_p": 0.95,
13
+ "transformers_version": "5.6.0"
14
+ }
checkpoint-1200/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b367dd7cb6a925bd54853db13b38c09a890802775fd53a9b5f92f229d269db92
3
+ size 24374794824
checkpoint-1200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c0a7981a6ad8e023656c878516f66bc11559c78ceab0c152f0b30223e7f7204
3
+ size 47064687765
checkpoint-1200/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2bef1a4d7370346e7d4bef1dc2c84c11387174800ef6b9ed831e2c65cc526332
3
+ size 14645
checkpoint-1200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f849fcebf11665780b3a911798e4034726eb1d6441b5dba8e011a58776190ca
3
+ size 1465
checkpoint-1200/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:daab2354f8a74e70d70b4d1f804939b68a8c9624dd06cb7858e52dd8970e9726
3
+ size 33384567
checkpoint-1200/tokenizer_config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "boi_token": "<start_of_image>",
4
+ "bos_token": "<bos>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eoi_token": "<end_of_image>",
7
+ "eos_token": "<eos>",
8
+ "image_token": "<image_soft_token>",
9
+ "is_local": false,
10
+ "local_files_only": false,
11
+ "mask_token": "<mask>",
12
+ "model_max_length": 1000000000000000019884624838656,
13
+ "model_specific_special_tokens": {
14
+ "boi_token": "<start_of_image>",
15
+ "eoi_token": "<end_of_image>",
16
+ "image_token": "<image_soft_token>"
17
+ },
18
+ "pad_token": "<pad>",
19
+ "processor_class": "Gemma3Processor",
20
+ "sp_model_kwargs": null,
21
+ "spaces_between_special_tokens": false,
22
+ "tokenizer_class": "GemmaTokenizer",
23
+ "unk_token": "<unk>",
24
+ "use_default_system_prompt": false
25
+ }
checkpoint-1200/trainer_state.json ADDED
@@ -0,0 +1,3761 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.3745704467353952,
6
+ "eval_steps": 500,
7
+ "global_step": 1200,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0,
14
+ "forget_loss": 3.44928662343591e-06,
15
+ "retain_loss": 6.609266757965088,
16
+ "step": 0
17
+ },
18
+ {
19
+ "epoch": 0,
20
+ "forget_loss": 0.038371894508600235,
21
+ "retain_loss": 6.274988651275635,
22
+ "step": 0
23
+ },
24
+ {
25
+ "epoch": 0,
26
+ "forget_loss": 0.001581120421178639,
27
+ "retain_loss": 6.232261657714844,
28
+ "step": 0
29
+ },
30
+ {
31
+ "epoch": 0,
32
+ "forget_loss": 8.447099389741197e-06,
33
+ "retain_loss": 6.230651378631592,
34
+ "step": 0
35
+ },
36
+ {
37
+ "epoch": 0.001145475372279496,
38
+ "grad_norm": 282.0,
39
+ "learning_rate": 0.0,
40
+ "loss": 6.386748313903809,
41
+ "step": 1
42
+ },
43
+ {
44
+ "epoch": 0.011454753722794959,
45
+ "grad_norm": 9.4375,
46
+ "learning_rate": 4.5e-05,
47
+ "loss": 3.339502970377604,
48
+ "step": 10
49
+ },
50
+ {
51
+ "epoch": 0.011454753722794959,
52
+ "forget_loss": 2.795343243633397e-05,
53
+ "retain_loss": 1.8692482709884644,
54
+ "step": 10
55
+ },
56
+ {
57
+ "epoch": 0.011454753722794959,
58
+ "forget_loss": 5.640117888106033e-05,
59
+ "retain_loss": 2.053846597671509,
60
+ "step": 10
61
+ },
62
+ {
63
+ "epoch": 0.011454753722794959,
64
+ "forget_loss": 3.110460966126993e-05,
65
+ "retain_loss": 2.0232813358306885,
66
+ "step": 10
67
+ },
68
+ {
69
+ "epoch": 0.011454753722794959,
70
+ "forget_loss": 4.845933744945796e-06,
71
+ "retain_loss": 2.4105024337768555,
72
+ "step": 10
73
+ },
74
+ {
75
+ "epoch": 0.022909507445589918,
76
+ "grad_norm": 38.5,
77
+ "learning_rate": 4.9740783410138256e-05,
78
+ "loss": 2.1140087127685545,
79
+ "step": 20
80
+ },
81
+ {
82
+ "epoch": 0.022909507445589918,
83
+ "forget_loss": 1.082867765944684e-05,
84
+ "retain_loss": 2.002471446990967,
85
+ "step": 20
86
+ },
87
+ {
88
+ "epoch": 0.022909507445589918,
89
+ "forget_loss": 3.829125489573926e-06,
90
+ "retain_loss": 2.0777769088745117,
91
+ "step": 20
92
+ },
93
+ {
94
+ "epoch": 0.022909507445589918,
95
+ "forget_loss": 2.8280664992053062e-05,
96
+ "retain_loss": 2.1336162090301514,
97
+ "step": 20
98
+ },
99
+ {
100
+ "epoch": 0.022909507445589918,
101
+ "forget_loss": 1.5830282791284844e-05,
102
+ "retain_loss": 1.7852996587753296,
103
+ "step": 20
104
+ },
105
+ {
106
+ "epoch": 0.03436426116838488,
107
+ "grad_norm": 5.6875,
108
+ "learning_rate": 4.945276497695853e-05,
109
+ "loss": 1.8668533325195313,
110
+ "step": 30
111
+ },
112
+ {
113
+ "epoch": 0.03436426116838488,
114
+ "forget_loss": 1.3028584362473339e-05,
115
+ "retain_loss": 1.870590090751648,
116
+ "step": 30
117
+ },
118
+ {
119
+ "epoch": 0.03436426116838488,
120
+ "forget_loss": 9.773620149644557e-06,
121
+ "retain_loss": 2.1787960529327393,
122
+ "step": 30
123
+ },
124
+ {
125
+ "epoch": 0.03436426116838488,
126
+ "forget_loss": 2.0277722796890885e-05,
127
+ "retain_loss": 1.7472785711288452,
128
+ "step": 30
129
+ },
130
+ {
131
+ "epoch": 0.03436426116838488,
132
+ "forget_loss": 1.0735096111602616e-05,
133
+ "retain_loss": 2.227604866027832,
134
+ "step": 30
135
+ },
136
+ {
137
+ "epoch": 0.045819014891179836,
138
+ "grad_norm": 5.4375,
139
+ "learning_rate": 4.916474654377881e-05,
140
+ "loss": 1.695387840270996,
141
+ "step": 40
142
+ },
143
+ {
144
+ "epoch": 0.045819014891179836,
145
+ "forget_loss": 5.423281436378602e-06,
146
+ "retain_loss": 1.2835420370101929,
147
+ "step": 40
148
+ },
149
+ {
150
+ "epoch": 0.045819014891179836,
151
+ "forget_loss": 9.37562845138018e-07,
152
+ "retain_loss": 0.8797236680984497,
153
+ "step": 40
154
+ },
155
+ {
156
+ "epoch": 0.045819014891179836,
157
+ "forget_loss": 4.068289854330942e-06,
158
+ "retain_loss": 1.9588595628738403,
159
+ "step": 40
160
+ },
161
+ {
162
+ "epoch": 0.045819014891179836,
163
+ "forget_loss": 2.925765329564456e-06,
164
+ "retain_loss": 1.7313635349273682,
165
+ "step": 40
166
+ },
167
+ {
168
+ "epoch": 0.0572737686139748,
169
+ "grad_norm": 24.5,
170
+ "learning_rate": 4.887672811059908e-05,
171
+ "loss": 1.3742843627929688,
172
+ "step": 50
173
+ },
174
+ {
175
+ "epoch": 0.0572737686139748,
176
+ "forget_loss": 1.3801378372590989e-05,
177
+ "retain_loss": 1.4735946655273438,
178
+ "step": 50
179
+ },
180
+ {
181
+ "epoch": 0.0572737686139748,
182
+ "forget_loss": 8.779801646596752e-06,
183
+ "retain_loss": 1.0075594186782837,
184
+ "step": 50
185
+ },
186
+ {
187
+ "epoch": 0.0572737686139748,
188
+ "forget_loss": 1.5394653019029647e-05,
189
+ "retain_loss": 1.8036051988601685,
190
+ "step": 50
191
+ },
192
+ {
193
+ "epoch": 0.0572737686139748,
194
+ "forget_loss": 8.239669114118442e-06,
195
+ "retain_loss": 1.379029393196106,
196
+ "step": 50
197
+ },
198
+ {
199
+ "epoch": 0.06872852233676977,
200
+ "grad_norm": 4.375,
201
+ "learning_rate": 4.858870967741936e-05,
202
+ "loss": 1.3121252059936523,
203
+ "step": 60
204
+ },
205
+ {
206
+ "epoch": 0.06872852233676977,
207
+ "forget_loss": 3.8279208638414275e-06,
208
+ "retain_loss": 1.7327319383621216,
209
+ "step": 60
210
+ },
211
+ {
212
+ "epoch": 0.06872852233676977,
213
+ "forget_loss": 3.3966937280638376e-06,
214
+ "retain_loss": 1.0556788444519043,
215
+ "step": 60
216
+ },
217
+ {
218
+ "epoch": 0.06872852233676977,
219
+ "forget_loss": 3.698090495163342e-06,
220
+ "retain_loss": 0.9602264761924744,
221
+ "step": 60
222
+ },
223
+ {
224
+ "epoch": 0.06872852233676977,
225
+ "forget_loss": 4.131216337555088e-05,
226
+ "retain_loss": 1.1526304483413696,
227
+ "step": 60
228
+ },
229
+ {
230
+ "epoch": 0.08018327605956473,
231
+ "grad_norm": 4.5,
232
+ "learning_rate": 4.830069124423963e-05,
233
+ "loss": 1.1931480407714843,
234
+ "step": 70
235
+ },
236
+ {
237
+ "epoch": 0.08018327605956473,
238
+ "forget_loss": 3.9989358810998965e-06,
239
+ "retain_loss": 1.1984384059906006,
240
+ "step": 70
241
+ },
242
+ {
243
+ "epoch": 0.08018327605956473,
244
+ "forget_loss": 2.2779448045184836e-05,
245
+ "retain_loss": 1.3697820901870728,
246
+ "step": 70
247
+ },
248
+ {
249
+ "epoch": 0.08018327605956473,
250
+ "forget_loss": 7.756965715088882e-06,
251
+ "retain_loss": 0.8010083436965942,
252
+ "step": 70
253
+ },
254
+ {
255
+ "epoch": 0.08018327605956473,
256
+ "forget_loss": 1.9700873963302e-06,
257
+ "retain_loss": 1.1254682540893555,
258
+ "step": 70
259
+ },
260
+ {
261
+ "epoch": 0.09163802978235967,
262
+ "grad_norm": 12.625,
263
+ "learning_rate": 4.801267281105991e-05,
264
+ "loss": 1.0133052825927735,
265
+ "step": 80
266
+ },
267
+ {
268
+ "epoch": 0.09163802978235967,
269
+ "forget_loss": 7.642759555892553e-06,
270
+ "retain_loss": 0.8252509236335754,
271
+ "step": 80
272
+ },
273
+ {
274
+ "epoch": 0.09163802978235967,
275
+ "forget_loss": 8.59803822095273e-06,
276
+ "retain_loss": 0.7506107687950134,
277
+ "step": 80
278
+ },
279
+ {
280
+ "epoch": 0.09163802978235967,
281
+ "forget_loss": 4.804069703823188e-06,
282
+ "retain_loss": 0.7315255403518677,
283
+ "step": 80
284
+ },
285
+ {
286
+ "epoch": 0.09163802978235967,
287
+ "forget_loss": 4.970052486896748e-06,
288
+ "retain_loss": 0.7012558579444885,
289
+ "step": 80
290
+ },
291
+ {
292
+ "epoch": 0.10309278350515463,
293
+ "grad_norm": 4.0625,
294
+ "learning_rate": 4.772465437788018e-05,
295
+ "loss": 0.9705428123474121,
296
+ "step": 90
297
+ },
298
+ {
299
+ "epoch": 0.10309278350515463,
300
+ "forget_loss": 1.4775166619074298e-06,
301
+ "retain_loss": 0.7782251834869385,
302
+ "step": 90
303
+ },
304
+ {
305
+ "epoch": 0.10309278350515463,
306
+ "forget_loss": 1.5886946584942052e-06,
307
+ "retain_loss": 0.8455419540405273,
308
+ "step": 90
309
+ },
310
+ {
311
+ "epoch": 0.10309278350515463,
312
+ "forget_loss": 2.0495062926784158e-06,
313
+ "retain_loss": 0.6329978704452515,
314
+ "step": 90
315
+ },
316
+ {
317
+ "epoch": 0.10309278350515463,
318
+ "forget_loss": 3.828020635410212e-05,
319
+ "retain_loss": 0.8053420186042786,
320
+ "step": 90
321
+ },
322
+ {
323
+ "epoch": 0.1145475372279496,
324
+ "grad_norm": 11.0,
325
+ "learning_rate": 4.743663594470046e-05,
326
+ "loss": 0.8036625862121582,
327
+ "step": 100
328
+ },
329
+ {
330
+ "epoch": 0.1145475372279496,
331
+ "forget_loss": 4.424137387104565e-06,
332
+ "retain_loss": 0.949962317943573,
333
+ "step": 100
334
+ },
335
+ {
336
+ "epoch": 0.1145475372279496,
337
+ "forget_loss": 2.453501565469196e-06,
338
+ "retain_loss": 0.8932625651359558,
339
+ "step": 100
340
+ },
341
+ {
342
+ "epoch": 0.1145475372279496,
343
+ "forget_loss": 1.6591504845564486e-06,
344
+ "retain_loss": 1.121445894241333,
345
+ "step": 100
346
+ },
347
+ {
348
+ "epoch": 0.1145475372279496,
349
+ "forget_loss": 4.8642609726812225e-06,
350
+ "retain_loss": 0.37143388390541077,
351
+ "step": 100
352
+ },
353
+ {
354
+ "epoch": 0.12600229095074456,
355
+ "grad_norm": 3.453125,
356
+ "learning_rate": 4.7148617511520735e-05,
357
+ "loss": 0.6938334941864014,
358
+ "step": 110
359
+ },
360
+ {
361
+ "epoch": 0.12600229095074456,
362
+ "forget_loss": 4.921608706354164e-06,
363
+ "retain_loss": 0.3767385482788086,
364
+ "step": 110
365
+ },
366
+ {
367
+ "epoch": 0.12600229095074456,
368
+ "forget_loss": 4.79454456581152e-07,
369
+ "retain_loss": 0.6782515048980713,
370
+ "step": 110
371
+ },
372
+ {
373
+ "epoch": 0.12600229095074456,
374
+ "forget_loss": 3.918839865946211e-06,
375
+ "retain_loss": 0.8204460144042969,
376
+ "step": 110
377
+ },
378
+ {
379
+ "epoch": 0.12600229095074456,
380
+ "forget_loss": 4.693280516221421e-06,
381
+ "retain_loss": 0.841111958026886,
382
+ "step": 110
383
+ },
384
+ {
385
+ "epoch": 0.13745704467353953,
386
+ "grad_norm": 3.4375,
387
+ "learning_rate": 4.6860599078341014e-05,
388
+ "loss": 0.6417326927185059,
389
+ "step": 120
390
+ },
391
+ {
392
+ "epoch": 0.13745704467353953,
393
+ "forget_loss": 1.713265555736143e-05,
394
+ "retain_loss": 0.3679219186306,
395
+ "step": 120
396
+ },
397
+ {
398
+ "epoch": 0.13745704467353953,
399
+ "forget_loss": 8.306015843118075e-06,
400
+ "retain_loss": 0.6136802434921265,
401
+ "step": 120
402
+ },
403
+ {
404
+ "epoch": 0.13745704467353953,
405
+ "forget_loss": 2.449452495056903e-06,
406
+ "retain_loss": 1.2731798887252808,
407
+ "step": 120
408
+ },
409
+ {
410
+ "epoch": 0.13745704467353953,
411
+ "forget_loss": 4.083998192072613e-06,
412
+ "retain_loss": 0.5807819962501526,
413
+ "step": 120
414
+ },
415
+ {
416
+ "epoch": 0.14891179839633448,
417
+ "grad_norm": 4.03125,
418
+ "learning_rate": 4.657258064516129e-05,
419
+ "loss": 0.5295979499816894,
420
+ "step": 130
421
+ },
422
+ {
423
+ "epoch": 0.14891179839633448,
424
+ "forget_loss": 8.430975867668167e-06,
425
+ "retain_loss": 0.7869871854782104,
426
+ "step": 130
427
+ },
428
+ {
429
+ "epoch": 0.14891179839633448,
430
+ "forget_loss": 9.652731023379602e-06,
431
+ "retain_loss": 0.9147917628288269,
432
+ "step": 130
433
+ },
434
+ {
435
+ "epoch": 0.14891179839633448,
436
+ "forget_loss": 1.1524152796482667e-06,
437
+ "retain_loss": 0.48021697998046875,
438
+ "step": 130
439
+ },
440
+ {
441
+ "epoch": 0.14891179839633448,
442
+ "forget_loss": 2.4387713892792817e-06,
443
+ "retain_loss": 0.8574126362800598,
444
+ "step": 130
445
+ },
446
+ {
447
+ "epoch": 0.16036655211912945,
448
+ "grad_norm": 4.5625,
449
+ "learning_rate": 4.628456221198157e-05,
450
+ "loss": 0.5009561061859131,
451
+ "step": 140
452
+ },
453
+ {
454
+ "epoch": 0.16036655211912945,
455
+ "forget_loss": 5.176518243388273e-06,
456
+ "retain_loss": 0.2740374803543091,
457
+ "step": 140
458
+ },
459
+ {
460
+ "epoch": 0.16036655211912945,
461
+ "forget_loss": 4.7920007091306616e-06,
462
+ "retain_loss": 0.6662377119064331,
463
+ "step": 140
464
+ },
465
+ {
466
+ "epoch": 0.16036655211912945,
467
+ "forget_loss": 2.156430582544999e-06,
468
+ "retain_loss": 0.3841518461704254,
469
+ "step": 140
470
+ },
471
+ {
472
+ "epoch": 0.16036655211912945,
473
+ "forget_loss": 1.6250938642770052e-05,
474
+ "retain_loss": 0.3435481786727905,
475
+ "step": 140
476
+ },
477
+ {
478
+ "epoch": 0.1718213058419244,
479
+ "grad_norm": 2.9375,
480
+ "learning_rate": 4.599654377880185e-05,
481
+ "loss": 0.5061944961547852,
482
+ "step": 150
483
+ },
484
+ {
485
+ "epoch": 0.1718213058419244,
486
+ "forget_loss": 2.413504262221977e-06,
487
+ "retain_loss": 0.4674696624279022,
488
+ "step": 150
489
+ },
490
+ {
491
+ "epoch": 0.1718213058419244,
492
+ "forget_loss": 9.788258466869593e-06,
493
+ "retain_loss": 0.32738229632377625,
494
+ "step": 150
495
+ },
496
+ {
497
+ "epoch": 0.1718213058419244,
498
+ "forget_loss": 1.659399458731059e-06,
499
+ "retain_loss": 0.4860354959964752,
500
+ "step": 150
501
+ },
502
+ {
503
+ "epoch": 0.1718213058419244,
504
+ "forget_loss": 6.014466293891019e-07,
505
+ "retain_loss": 0.25174200534820557,
506
+ "step": 150
507
+ },
508
+ {
509
+ "epoch": 0.18327605956471935,
510
+ "grad_norm": 3.140625,
511
+ "learning_rate": 4.5708525345622124e-05,
512
+ "loss": 0.33492887020111084,
513
+ "step": 160
514
+ },
515
+ {
516
+ "epoch": 0.18327605956471935,
517
+ "forget_loss": 4.615132638718933e-06,
518
+ "retain_loss": 0.30095958709716797,
519
+ "step": 160
520
+ },
521
+ {
522
+ "epoch": 0.18327605956471935,
523
+ "forget_loss": 1.7731171055856976e-06,
524
+ "retain_loss": 0.19373854994773865,
525
+ "step": 160
526
+ },
527
+ {
528
+ "epoch": 0.18327605956471935,
529
+ "forget_loss": 2.2210085717233596e-06,
530
+ "retain_loss": 0.23821769654750824,
531
+ "step": 160
532
+ },
533
+ {
534
+ "epoch": 0.18327605956471935,
535
+ "forget_loss": 1.3112939996062778e-06,
536
+ "retain_loss": 0.3393084406852722,
537
+ "step": 160
538
+ },
539
+ {
540
+ "epoch": 0.19473081328751432,
541
+ "grad_norm": 2.671875,
542
+ "learning_rate": 4.54205069124424e-05,
543
+ "loss": 0.3298675775527954,
544
+ "step": 170
545
+ },
546
+ {
547
+ "epoch": 0.19473081328751432,
548
+ "forget_loss": 4.927760528516956e-06,
549
+ "retain_loss": 0.44287964701652527,
550
+ "step": 170
551
+ },
552
+ {
553
+ "epoch": 0.19473081328751432,
554
+ "forget_loss": 4.030034688184969e-06,
555
+ "retain_loss": 0.4538065195083618,
556
+ "step": 170
557
+ },
558
+ {
559
+ "epoch": 0.19473081328751432,
560
+ "forget_loss": 2.3950892682478298e-06,
561
+ "retain_loss": 0.31112024188041687,
562
+ "step": 170
563
+ },
564
+ {
565
+ "epoch": 0.19473081328751432,
566
+ "forget_loss": 2.15224758903787e-06,
567
+ "retain_loss": 0.9548900127410889,
568
+ "step": 170
569
+ },
570
+ {
571
+ "epoch": 0.20618556701030927,
572
+ "grad_norm": 2.953125,
573
+ "learning_rate": 4.5132488479262676e-05,
574
+ "loss": 0.3629490852355957,
575
+ "step": 180
576
+ },
577
+ {
578
+ "epoch": 0.20618556701030927,
579
+ "forget_loss": 7.814773198333569e-06,
580
+ "retain_loss": 0.2744402289390564,
581
+ "step": 180
582
+ },
583
+ {
584
+ "epoch": 0.20618556701030927,
585
+ "forget_loss": 9.177201150123437e-07,
586
+ "retain_loss": 0.16577409207820892,
587
+ "step": 180
588
+ },
589
+ {
590
+ "epoch": 0.20618556701030927,
591
+ "forget_loss": 2.789876816677861e-06,
592
+ "retain_loss": 0.26117604970932007,
593
+ "step": 180
594
+ },
595
+ {
596
+ "epoch": 0.20618556701030927,
597
+ "forget_loss": 3.266710336902179e-06,
598
+ "retain_loss": 0.2821800708770752,
599
+ "step": 180
600
+ },
601
+ {
602
+ "epoch": 0.21764032073310424,
603
+ "grad_norm": 3.125,
604
+ "learning_rate": 4.4844470046082955e-05,
605
+ "loss": 0.3086943864822388,
606
+ "step": 190
607
+ },
608
+ {
609
+ "epoch": 0.21764032073310424,
610
+ "forget_loss": 2.877425231417874e-06,
611
+ "retain_loss": 0.4397953450679779,
612
+ "step": 190
613
+ },
614
+ {
615
+ "epoch": 0.21764032073310424,
616
+ "forget_loss": 5.625155608868226e-06,
617
+ "retain_loss": 0.40479227900505066,
618
+ "step": 190
619
+ },
620
+ {
621
+ "epoch": 0.21764032073310424,
622
+ "forget_loss": 7.261318387463689e-06,
623
+ "retain_loss": 0.08186227083206177,
624
+ "step": 190
625
+ },
626
+ {
627
+ "epoch": 0.21764032073310424,
628
+ "forget_loss": 1.0345914915887988e-06,
629
+ "retain_loss": 0.5473820567131042,
630
+ "step": 190
631
+ },
632
+ {
633
+ "epoch": 0.2290950744558992,
634
+ "grad_norm": 2.59375,
635
+ "learning_rate": 4.455645161290323e-05,
636
+ "loss": 0.30754408836364744,
637
+ "step": 200
638
+ },
639
+ {
640
+ "epoch": 0.2290950744558992,
641
+ "forget_loss": 1.8709645246417494e-06,
642
+ "retain_loss": 0.15519601106643677,
643
+ "step": 200
644
+ },
645
+ {
646
+ "epoch": 0.2290950744558992,
647
+ "forget_loss": 5.221798460297578e-07,
648
+ "retain_loss": 0.0977967157959938,
649
+ "step": 200
650
+ },
651
+ {
652
+ "epoch": 0.2290950744558992,
653
+ "forget_loss": 4.533095307124313e-06,
654
+ "retain_loss": 0.12215632200241089,
655
+ "step": 200
656
+ },
657
+ {
658
+ "epoch": 0.2290950744558992,
659
+ "forget_loss": 2.3582092580909375e-06,
660
+ "retain_loss": 0.14002011716365814,
661
+ "step": 200
662
+ },
663
+ {
664
+ "epoch": 0.24054982817869416,
665
+ "grad_norm": 3.140625,
666
+ "learning_rate": 4.4268433179723506e-05,
667
+ "loss": 0.25385417938232424,
668
+ "step": 210
669
+ },
670
+ {
671
+ "epoch": 0.24054982817869416,
672
+ "forget_loss": 4.970596364728408e-06,
673
+ "retain_loss": 0.17109914124011993,
674
+ "step": 210
675
+ },
676
+ {
677
+ "epoch": 0.24054982817869416,
678
+ "forget_loss": 1.6977497580228373e-05,
679
+ "retain_loss": 0.2036435753107071,
680
+ "step": 210
681
+ },
682
+ {
683
+ "epoch": 0.24054982817869416,
684
+ "forget_loss": 3.1074764592631254e-06,
685
+ "retain_loss": 0.13067637383937836,
686
+ "step": 210
687
+ },
688
+ {
689
+ "epoch": 0.24054982817869416,
690
+ "forget_loss": 2.0813706669287058e-06,
691
+ "retain_loss": 0.1853303760290146,
692
+ "step": 210
693
+ },
694
+ {
695
+ "epoch": 0.2520045819014891,
696
+ "grad_norm": 2.5,
697
+ "learning_rate": 4.398041474654378e-05,
698
+ "loss": 0.2122504711151123,
699
+ "step": 220
700
+ },
701
+ {
702
+ "epoch": 0.2520045819014891,
703
+ "forget_loss": 1.3193698578106705e-06,
704
+ "retain_loss": 0.14352917671203613,
705
+ "step": 220
706
+ },
707
+ {
708
+ "epoch": 0.2520045819014891,
709
+ "forget_loss": 1.1001748134731315e-06,
710
+ "retain_loss": 0.30384859442710876,
711
+ "step": 220
712
+ },
713
+ {
714
+ "epoch": 0.2520045819014891,
715
+ "forget_loss": 1.4734857813891722e-06,
716
+ "retain_loss": 0.06934046000242233,
717
+ "step": 220
718
+ },
719
+ {
720
+ "epoch": 0.2520045819014891,
721
+ "forget_loss": 2.7343555757397553e-06,
722
+ "retain_loss": 0.1081199198961258,
723
+ "step": 220
724
+ },
725
+ {
726
+ "epoch": 0.2634593356242841,
727
+ "grad_norm": 3.09375,
728
+ "learning_rate": 4.369239631336406e-05,
729
+ "loss": 0.20068085193634033,
730
+ "step": 230
731
+ },
732
+ {
733
+ "epoch": 0.2634593356242841,
734
+ "forget_loss": 2.154215962946182e-06,
735
+ "retain_loss": 0.10300299525260925,
736
+ "step": 230
737
+ },
738
+ {
739
+ "epoch": 0.2634593356242841,
740
+ "forget_loss": 9.424740710528567e-06,
741
+ "retain_loss": 0.13396301865577698,
742
+ "step": 230
743
+ },
744
+ {
745
+ "epoch": 0.2634593356242841,
746
+ "forget_loss": 3.0522624001605436e-06,
747
+ "retain_loss": 0.33339884877204895,
748
+ "step": 230
749
+ },
750
+ {
751
+ "epoch": 0.2634593356242841,
752
+ "forget_loss": 8.897523912310135e-06,
753
+ "retain_loss": 0.1534104347229004,
754
+ "step": 230
755
+ },
756
+ {
757
+ "epoch": 0.27491408934707906,
758
+ "grad_norm": 2.671875,
759
+ "learning_rate": 4.340437788018433e-05,
760
+ "loss": 0.1963863492012024,
761
+ "step": 240
762
+ },
763
+ {
764
+ "epoch": 0.27491408934707906,
765
+ "forget_loss": 1.7897947373057832e-06,
766
+ "retain_loss": 0.11809217184782028,
767
+ "step": 240
768
+ },
769
+ {
770
+ "epoch": 0.27491408934707906,
771
+ "forget_loss": 5.25018833741342e-07,
772
+ "retain_loss": 0.15055039525032043,
773
+ "step": 240
774
+ },
775
+ {
776
+ "epoch": 0.27491408934707906,
777
+ "forget_loss": 6.8833096520393156e-06,
778
+ "retain_loss": 0.16997316479682922,
779
+ "step": 240
780
+ },
781
+ {
782
+ "epoch": 0.27491408934707906,
783
+ "forget_loss": 2.313455297553446e-06,
784
+ "retain_loss": 0.10752896964550018,
785
+ "step": 240
786
+ },
787
+ {
788
+ "epoch": 0.286368843069874,
789
+ "grad_norm": 2.59375,
790
+ "learning_rate": 4.311635944700461e-05,
791
+ "loss": 0.20680031776428223,
792
+ "step": 250
793
+ },
794
+ {
795
+ "epoch": 0.286368843069874,
796
+ "forget_loss": 2.1803207346238196e-06,
797
+ "retain_loss": 0.07074207067489624,
798
+ "step": 250
799
+ },
800
+ {
801
+ "epoch": 0.286368843069874,
802
+ "forget_loss": 1.8291285641680588e-06,
803
+ "retain_loss": 0.09072300791740417,
804
+ "step": 250
805
+ },
806
+ {
807
+ "epoch": 0.286368843069874,
808
+ "forget_loss": 1.0191804449277697e-06,
809
+ "retain_loss": 0.4518914818763733,
810
+ "step": 250
811
+ },
812
+ {
813
+ "epoch": 0.286368843069874,
814
+ "forget_loss": 1.1188744792889338e-06,
815
+ "retain_loss": 0.09803809970617294,
816
+ "step": 250
817
+ },
818
+ {
819
+ "epoch": 0.29782359679266895,
820
+ "grad_norm": 2.703125,
821
+ "learning_rate": 4.282834101382488e-05,
822
+ "loss": 0.16371082067489623,
823
+ "step": 260
824
+ },
825
+ {
826
+ "epoch": 0.29782359679266895,
827
+ "forget_loss": 3.8669927562295925e-06,
828
+ "retain_loss": 0.17645332217216492,
829
+ "step": 260
830
+ },
831
+ {
832
+ "epoch": 0.29782359679266895,
833
+ "forget_loss": 5.920769581280183e-06,
834
+ "retain_loss": 0.18589472770690918,
835
+ "step": 260
836
+ },
837
+ {
838
+ "epoch": 0.29782359679266895,
839
+ "forget_loss": 5.846962267241906e-06,
840
+ "retain_loss": 0.1563558578491211,
841
+ "step": 260
842
+ },
843
+ {
844
+ "epoch": 0.29782359679266895,
845
+ "forget_loss": 2.4713845050428063e-06,
846
+ "retain_loss": 0.08506585657596588,
847
+ "step": 260
848
+ },
849
+ {
850
+ "epoch": 0.30927835051546393,
851
+ "grad_norm": 1.984375,
852
+ "learning_rate": 4.254032258064516e-05,
853
+ "loss": 0.1610316514968872,
854
+ "step": 270
855
+ },
856
+ {
857
+ "epoch": 0.30927835051546393,
858
+ "forget_loss": 2.003027475439012e-06,
859
+ "retain_loss": 0.1364717036485672,
860
+ "step": 270
861
+ },
862
+ {
863
+ "epoch": 0.30927835051546393,
864
+ "forget_loss": 2.4025737275223946e-06,
865
+ "retain_loss": 0.06489767134189606,
866
+ "step": 270
867
+ },
868
+ {
869
+ "epoch": 0.30927835051546393,
870
+ "forget_loss": 1.9498058918543393e-06,
871
+ "retain_loss": 0.15014496445655823,
872
+ "step": 270
873
+ },
874
+ {
875
+ "epoch": 0.30927835051546393,
876
+ "forget_loss": 1.4814335145274526e-06,
877
+ "retain_loss": 0.10888711363077164,
878
+ "step": 270
879
+ },
880
+ {
881
+ "epoch": 0.3207331042382589,
882
+ "grad_norm": 2.03125,
883
+ "learning_rate": 4.225230414746544e-05,
884
+ "loss": 0.1513701558113098,
885
+ "step": 280
886
+ },
887
+ {
888
+ "epoch": 0.3207331042382589,
889
+ "forget_loss": 3.564718099369202e-06,
890
+ "retain_loss": 0.18767109513282776,
891
+ "step": 280
892
+ },
893
+ {
894
+ "epoch": 0.3207331042382589,
895
+ "forget_loss": 1.5525797607551794e-06,
896
+ "retain_loss": 0.1132984533905983,
897
+ "step": 280
898
+ },
899
+ {
900
+ "epoch": 0.3207331042382589,
901
+ "forget_loss": 2.530495294195134e-06,
902
+ "retain_loss": 0.14477676153182983,
903
+ "step": 280
904
+ },
905
+ {
906
+ "epoch": 0.3207331042382589,
907
+ "forget_loss": 1.2278740086912876e-06,
908
+ "retain_loss": 0.22871914505958557,
909
+ "step": 280
910
+ },
911
+ {
912
+ "epoch": 0.3321878579610538,
913
+ "grad_norm": 1.9375,
914
+ "learning_rate": 4.196428571428572e-05,
915
+ "loss": 0.13684289455413817,
916
+ "step": 290
917
+ },
918
+ {
919
+ "epoch": 0.3321878579610538,
920
+ "forget_loss": 6.087240080887568e-07,
921
+ "retain_loss": 0.12418590486049652,
922
+ "step": 290
923
+ },
924
+ {
925
+ "epoch": 0.3321878579610538,
926
+ "forget_loss": 1.3446937145999982e-06,
927
+ "retain_loss": 0.07405225187540054,
928
+ "step": 290
929
+ },
930
+ {
931
+ "epoch": 0.3321878579610538,
932
+ "forget_loss": 5.7531210586603265e-06,
933
+ "retain_loss": 0.18760758638381958,
934
+ "step": 290
935
+ },
936
+ {
937
+ "epoch": 0.3321878579610538,
938
+ "forget_loss": 3.2198163353314158e-06,
939
+ "retain_loss": 0.06644970923662186,
940
+ "step": 290
941
+ },
942
+ {
943
+ "epoch": 0.3436426116838488,
944
+ "grad_norm": 1.734375,
945
+ "learning_rate": 4.167626728110599e-05,
946
+ "loss": 0.12020657062530518,
947
+ "step": 300
948
+ },
949
+ {
950
+ "epoch": 0.3436426116838488,
951
+ "forget_loss": 3.8907467114768224e-07,
952
+ "retain_loss": 0.09106294065713882,
953
+ "step": 300
954
+ },
955
+ {
956
+ "epoch": 0.3436426116838488,
957
+ "forget_loss": 1.2785067156073637e-06,
958
+ "retain_loss": 0.07731538265943527,
959
+ "step": 300
960
+ },
961
+ {
962
+ "epoch": 0.3436426116838488,
963
+ "forget_loss": 7.883311923251313e-07,
964
+ "retain_loss": 0.0779929980635643,
965
+ "step": 300
966
+ },
967
+ {
968
+ "epoch": 0.3436426116838488,
969
+ "forget_loss": 1.6741282706789207e-06,
970
+ "retain_loss": 0.11640097200870514,
971
+ "step": 300
972
+ },
973
+ {
974
+ "epoch": 0.3550973654066438,
975
+ "grad_norm": 2.09375,
976
+ "learning_rate": 4.138824884792627e-05,
977
+ "loss": 0.11634387969970703,
978
+ "step": 310
979
+ },
980
+ {
981
+ "epoch": 0.3550973654066438,
982
+ "forget_loss": 1.1909932027265313e-06,
983
+ "retain_loss": 0.09301010519266129,
984
+ "step": 310
985
+ },
986
+ {
987
+ "epoch": 0.3550973654066438,
988
+ "forget_loss": 5.385973622651363e-07,
989
+ "retain_loss": 0.13168799877166748,
990
+ "step": 310
991
+ },
992
+ {
993
+ "epoch": 0.3550973654066438,
994
+ "forget_loss": 4.287514457246289e-06,
995
+ "retain_loss": 0.07441989332437515,
996
+ "step": 310
997
+ },
998
+ {
999
+ "epoch": 0.3550973654066438,
1000
+ "forget_loss": 5.496154244610807e-06,
1001
+ "retain_loss": 0.0625639408826828,
1002
+ "step": 310
1003
+ },
1004
+ {
1005
+ "epoch": 0.3665521191294387,
1006
+ "grad_norm": 2.09375,
1007
+ "learning_rate": 4.1100230414746544e-05,
1008
+ "loss": 0.13164939880371093,
1009
+ "step": 320
1010
+ },
1011
+ {
1012
+ "epoch": 0.3665521191294387,
1013
+ "forget_loss": 1.5266854234141647e-06,
1014
+ "retain_loss": 0.07643962651491165,
1015
+ "step": 320
1016
+ },
1017
+ {
1018
+ "epoch": 0.3665521191294387,
1019
+ "forget_loss": 5.235110052126402e-07,
1020
+ "retain_loss": 0.10404697060585022,
1021
+ "step": 320
1022
+ },
1023
+ {
1024
+ "epoch": 0.3665521191294387,
1025
+ "forget_loss": 3.643459649538272e-06,
1026
+ "retain_loss": 0.11882194131612778,
1027
+ "step": 320
1028
+ },
1029
+ {
1030
+ "epoch": 0.3665521191294387,
1031
+ "forget_loss": 9.903519639919978e-07,
1032
+ "retain_loss": 0.07760447263717651,
1033
+ "step": 320
1034
+ },
1035
+ {
1036
+ "epoch": 0.37800687285223367,
1037
+ "grad_norm": 1.7265625,
1038
+ "learning_rate": 4.081221198156682e-05,
1039
+ "loss": 0.11806095838546753,
1040
+ "step": 330
1041
+ },
1042
+ {
1043
+ "epoch": 0.37800687285223367,
1044
+ "forget_loss": 2.186560550399008e-06,
1045
+ "retain_loss": 0.05915055423974991,
1046
+ "step": 330
1047
+ },
1048
+ {
1049
+ "epoch": 0.37800687285223367,
1050
+ "forget_loss": 2.5664467102615163e-06,
1051
+ "retain_loss": 0.06395165622234344,
1052
+ "step": 330
1053
+ },
1054
+ {
1055
+ "epoch": 0.37800687285223367,
1056
+ "forget_loss": 1.2531793345260667e-06,
1057
+ "retain_loss": 0.1056695505976677,
1058
+ "step": 330
1059
+ },
1060
+ {
1061
+ "epoch": 0.37800687285223367,
1062
+ "forget_loss": 7.011165507719852e-06,
1063
+ "retain_loss": 0.10078335553407669,
1064
+ "step": 330
1065
+ },
1066
+ {
1067
+ "epoch": 0.38946162657502864,
1068
+ "grad_norm": 1.734375,
1069
+ "learning_rate": 4.0524193548387096e-05,
1070
+ "loss": 0.10047016143798829,
1071
+ "step": 340
1072
+ },
1073
+ {
1074
+ "epoch": 0.38946162657502864,
1075
+ "forget_loss": 1.4236219385566073e-06,
1076
+ "retain_loss": 0.5226035714149475,
1077
+ "step": 340
1078
+ },
1079
+ {
1080
+ "epoch": 0.38946162657502864,
1081
+ "forget_loss": 1.951645344888675e-06,
1082
+ "retain_loss": 0.10346213728189468,
1083
+ "step": 340
1084
+ },
1085
+ {
1086
+ "epoch": 0.38946162657502864,
1087
+ "forget_loss": 2.398945980530698e-06,
1088
+ "retain_loss": 0.03990686312317848,
1089
+ "step": 340
1090
+ },
1091
+ {
1092
+ "epoch": 0.38946162657502864,
1093
+ "forget_loss": 4.940415237797424e-06,
1094
+ "retain_loss": 0.08106988668441772,
1095
+ "step": 340
1096
+ },
1097
+ {
1098
+ "epoch": 0.4009163802978236,
1099
+ "grad_norm": 1.9296875,
1100
+ "learning_rate": 4.0236175115207375e-05,
1101
+ "loss": 0.10264077186584472,
1102
+ "step": 350
1103
+ },
1104
+ {
1105
+ "epoch": 0.4009163802978236,
1106
+ "forget_loss": 1.1442448339948896e-06,
1107
+ "retain_loss": 0.07710669934749603,
1108
+ "step": 350
1109
+ },
1110
+ {
1111
+ "epoch": 0.4009163802978236,
1112
+ "forget_loss": 1.1809590887423838e-06,
1113
+ "retain_loss": 0.384318470954895,
1114
+ "step": 350
1115
+ },
1116
+ {
1117
+ "epoch": 0.4009163802978236,
1118
+ "forget_loss": 1.8739415281743277e-06,
1119
+ "retain_loss": 0.1440838724374771,
1120
+ "step": 350
1121
+ },
1122
+ {
1123
+ "epoch": 0.4009163802978236,
1124
+ "forget_loss": 1.596398078618222e-06,
1125
+ "retain_loss": 0.08889831602573395,
1126
+ "step": 350
1127
+ },
1128
+ {
1129
+ "epoch": 0.41237113402061853,
1130
+ "grad_norm": 1.40625,
1131
+ "learning_rate": 3.9948156682027654e-05,
1132
+ "loss": 0.09905218482017517,
1133
+ "step": 360
1134
+ },
1135
+ {
1136
+ "epoch": 0.41237113402061853,
1137
+ "forget_loss": 6.401254722732119e-06,
1138
+ "retain_loss": 0.06261590123176575,
1139
+ "step": 360
1140
+ },
1141
+ {
1142
+ "epoch": 0.41237113402061853,
1143
+ "forget_loss": 3.825871317530982e-07,
1144
+ "retain_loss": 0.15589363873004913,
1145
+ "step": 360
1146
+ },
1147
+ {
1148
+ "epoch": 0.41237113402061853,
1149
+ "forget_loss": 1.1340009677951457e-06,
1150
+ "retain_loss": 0.026375912129878998,
1151
+ "step": 360
1152
+ },
1153
+ {
1154
+ "epoch": 0.41237113402061853,
1155
+ "forget_loss": 1.092132379199029e-06,
1156
+ "retain_loss": 0.07016332447528839,
1157
+ "step": 360
1158
+ },
1159
+ {
1160
+ "epoch": 0.4238258877434135,
1161
+ "grad_norm": 2.28125,
1162
+ "learning_rate": 3.9660138248847926e-05,
1163
+ "loss": 0.10010771751403809,
1164
+ "step": 370
1165
+ },
1166
+ {
1167
+ "epoch": 0.4238258877434135,
1168
+ "forget_loss": 5.463415959638951e-07,
1169
+ "retain_loss": 0.09104490280151367,
1170
+ "step": 370
1171
+ },
1172
+ {
1173
+ "epoch": 0.4238258877434135,
1174
+ "forget_loss": 7.711361149631557e-07,
1175
+ "retain_loss": 0.07780595868825912,
1176
+ "step": 370
1177
+ },
1178
+ {
1179
+ "epoch": 0.4238258877434135,
1180
+ "forget_loss": 8.610635973127501e-07,
1181
+ "retain_loss": 0.0850876048207283,
1182
+ "step": 370
1183
+ },
1184
+ {
1185
+ "epoch": 0.4238258877434135,
1186
+ "forget_loss": 2.193863338106894e-06,
1187
+ "retain_loss": 0.02873000130057335,
1188
+ "step": 370
1189
+ },
1190
+ {
1191
+ "epoch": 0.4352806414662085,
1192
+ "grad_norm": 1.8125,
1193
+ "learning_rate": 3.9372119815668206e-05,
1194
+ "loss": 0.07582432627677918,
1195
+ "step": 380
1196
+ },
1197
+ {
1198
+ "epoch": 0.4352806414662085,
1199
+ "forget_loss": 5.40978362550959e-07,
1200
+ "retain_loss": 0.03535467013716698,
1201
+ "step": 380
1202
+ },
1203
+ {
1204
+ "epoch": 0.4352806414662085,
1205
+ "forget_loss": 8.922065717342775e-07,
1206
+ "retain_loss": 0.057049620896577835,
1207
+ "step": 380
1208
+ },
1209
+ {
1210
+ "epoch": 0.4352806414662085,
1211
+ "forget_loss": 5.560720524044882e-07,
1212
+ "retain_loss": 0.1527683138847351,
1213
+ "step": 380
1214
+ },
1215
+ {
1216
+ "epoch": 0.4352806414662085,
1217
+ "forget_loss": 4.7091802457543963e-07,
1218
+ "retain_loss": 0.1109684631228447,
1219
+ "step": 380
1220
+ },
1221
+ {
1222
+ "epoch": 0.44673539518900346,
1223
+ "grad_norm": 1.7734375,
1224
+ "learning_rate": 3.908410138248848e-05,
1225
+ "loss": 0.09133011102676392,
1226
+ "step": 390
1227
+ },
1228
+ {
1229
+ "epoch": 0.44673539518900346,
1230
+ "forget_loss": 8.056795763877744e-07,
1231
+ "retain_loss": 0.09070051461458206,
1232
+ "step": 390
1233
+ },
1234
+ {
1235
+ "epoch": 0.44673539518900346,
1236
+ "forget_loss": 1.1127366406071815e-06,
1237
+ "retain_loss": 0.0616828016936779,
1238
+ "step": 390
1239
+ },
1240
+ {
1241
+ "epoch": 0.44673539518900346,
1242
+ "forget_loss": 3.5524210488802055e-06,
1243
+ "retain_loss": 0.06886927038431168,
1244
+ "step": 390
1245
+ },
1246
+ {
1247
+ "epoch": 0.44673539518900346,
1248
+ "forget_loss": 2.0374077394080814e-06,
1249
+ "retain_loss": 0.10105489939451218,
1250
+ "step": 390
1251
+ },
1252
+ {
1253
+ "epoch": 0.4581901489117984,
1254
+ "grad_norm": 1.59375,
1255
+ "learning_rate": 3.879608294930876e-05,
1256
+ "loss": 0.07695338726043702,
1257
+ "step": 400
1258
+ },
1259
+ {
1260
+ "epoch": 0.4581901489117984,
1261
+ "forget_loss": 1.342506493529072e-06,
1262
+ "retain_loss": 0.03703423589468002,
1263
+ "step": 400
1264
+ },
1265
+ {
1266
+ "epoch": 0.4581901489117984,
1267
+ "forget_loss": 4.241804276716721e-07,
1268
+ "retain_loss": 0.07702656090259552,
1269
+ "step": 400
1270
+ },
1271
+ {
1272
+ "epoch": 0.4581901489117984,
1273
+ "forget_loss": 2.735993405167392e-07,
1274
+ "retain_loss": 0.04419003427028656,
1275
+ "step": 400
1276
+ },
1277
+ {
1278
+ "epoch": 0.4581901489117984,
1279
+ "forget_loss": 3.817524429905461e-07,
1280
+ "retain_loss": 0.040069010108709335,
1281
+ "step": 400
1282
+ },
1283
+ {
1284
+ "epoch": 0.46964490263459335,
1285
+ "grad_norm": 1.671875,
1286
+ "learning_rate": 3.8508064516129036e-05,
1287
+ "loss": 0.07269055247306824,
1288
+ "step": 410
1289
+ },
1290
+ {
1291
+ "epoch": 0.46964490263459335,
1292
+ "forget_loss": 2.8240863230166724e-06,
1293
+ "retain_loss": 0.04700532928109169,
1294
+ "step": 410
1295
+ },
1296
+ {
1297
+ "epoch": 0.46964490263459335,
1298
+ "forget_loss": 1.747673792351634e-07,
1299
+ "retain_loss": 0.033581238240003586,
1300
+ "step": 410
1301
+ },
1302
+ {
1303
+ "epoch": 0.46964490263459335,
1304
+ "forget_loss": 5.92311096170306e-07,
1305
+ "retain_loss": 0.09985388815402985,
1306
+ "step": 410
1307
+ },
1308
+ {
1309
+ "epoch": 0.46964490263459335,
1310
+ "forget_loss": 1.7826670273279888e-07,
1311
+ "retain_loss": 0.07349472492933273,
1312
+ "step": 410
1313
+ },
1314
+ {
1315
+ "epoch": 0.48109965635738833,
1316
+ "grad_norm": 1.8046875,
1317
+ "learning_rate": 3.8220046082949316e-05,
1318
+ "loss": 0.07912165522575379,
1319
+ "step": 420
1320
+ },
1321
+ {
1322
+ "epoch": 0.48109965635738833,
1323
+ "forget_loss": 1.2650438918626605e-07,
1324
+ "retain_loss": 0.06736782938241959,
1325
+ "step": 420
1326
+ },
1327
+ {
1328
+ "epoch": 0.48109965635738833,
1329
+ "forget_loss": 1.5735973875052878e-06,
1330
+ "retain_loss": 0.03057463839650154,
1331
+ "step": 420
1332
+ },
1333
+ {
1334
+ "epoch": 0.48109965635738833,
1335
+ "forget_loss": 4.024736313112953e-07,
1336
+ "retain_loss": 0.05045014247298241,
1337
+ "step": 420
1338
+ },
1339
+ {
1340
+ "epoch": 0.48109965635738833,
1341
+ "forget_loss": 2.5390670543856686e-06,
1342
+ "retain_loss": 0.07917766273021698,
1343
+ "step": 420
1344
+ },
1345
+ {
1346
+ "epoch": 0.4925544100801833,
1347
+ "grad_norm": 1.1953125,
1348
+ "learning_rate": 3.793202764976959e-05,
1349
+ "loss": 0.0656080961227417,
1350
+ "step": 430
1351
+ },
1352
+ {
1353
+ "epoch": 0.4925544100801833,
1354
+ "forget_loss": 5.468404538078175e-07,
1355
+ "retain_loss": 0.02683027647435665,
1356
+ "step": 430
1357
+ },
1358
+ {
1359
+ "epoch": 0.4925544100801833,
1360
+ "forget_loss": 5.585835083365964e-07,
1361
+ "retain_loss": 0.0377357117831707,
1362
+ "step": 430
1363
+ },
1364
+ {
1365
+ "epoch": 0.4925544100801833,
1366
+ "forget_loss": 3.003877679930156e-07,
1367
+ "retain_loss": 0.052259303629398346,
1368
+ "step": 430
1369
+ },
1370
+ {
1371
+ "epoch": 0.4925544100801833,
1372
+ "forget_loss": 1.2751459053106373e-06,
1373
+ "retain_loss": 0.10837093740701675,
1374
+ "step": 430
1375
+ },
1376
+ {
1377
+ "epoch": 0.5040091638029782,
1378
+ "grad_norm": 1.8046875,
1379
+ "learning_rate": 3.764400921658987e-05,
1380
+ "loss": 0.06297182440757751,
1381
+ "step": 440
1382
+ },
1383
+ {
1384
+ "epoch": 0.5040091638029782,
1385
+ "forget_loss": 2.636994622662314e-07,
1386
+ "retain_loss": 0.04880654811859131,
1387
+ "step": 440
1388
+ },
1389
+ {
1390
+ "epoch": 0.5040091638029782,
1391
+ "forget_loss": 2.0067361674591666e-06,
1392
+ "retain_loss": 0.027269020676612854,
1393
+ "step": 440
1394
+ },
1395
+ {
1396
+ "epoch": 0.5040091638029782,
1397
+ "forget_loss": 1.8749557284536422e-06,
1398
+ "retain_loss": 0.04770921915769577,
1399
+ "step": 440
1400
+ },
1401
+ {
1402
+ "epoch": 0.5040091638029782,
1403
+ "forget_loss": 7.214839001790097e-07,
1404
+ "retain_loss": 0.07100628316402435,
1405
+ "step": 440
1406
+ },
1407
+ {
1408
+ "epoch": 0.5154639175257731,
1409
+ "grad_norm": 1.0859375,
1410
+ "learning_rate": 3.735599078341014e-05,
1411
+ "loss": 0.0665643870830536,
1412
+ "step": 450
1413
+ },
1414
+ {
1415
+ "epoch": 0.5154639175257731,
1416
+ "forget_loss": 8.713016796946249e-08,
1417
+ "retain_loss": 0.042148686945438385,
1418
+ "step": 450
1419
+ },
1420
+ {
1421
+ "epoch": 0.5154639175257731,
1422
+ "forget_loss": 6.566732508872519e-07,
1423
+ "retain_loss": 0.05762806907296181,
1424
+ "step": 450
1425
+ },
1426
+ {
1427
+ "epoch": 0.5154639175257731,
1428
+ "forget_loss": 2.520494888358371e-07,
1429
+ "retain_loss": 0.0422804020345211,
1430
+ "step": 450
1431
+ },
1432
+ {
1433
+ "epoch": 0.5154639175257731,
1434
+ "forget_loss": 2.0052689251315314e-06,
1435
+ "retain_loss": 0.05088704079389572,
1436
+ "step": 450
1437
+ },
1438
+ {
1439
+ "epoch": 0.5269186712485682,
1440
+ "grad_norm": 1.140625,
1441
+ "learning_rate": 3.706797235023042e-05,
1442
+ "loss": 0.06394548416137695,
1443
+ "step": 460
1444
+ },
1445
+ {
1446
+ "epoch": 0.5269186712485682,
1447
+ "forget_loss": 4.606073389368248e-07,
1448
+ "retain_loss": 0.0739588811993599,
1449
+ "step": 460
1450
+ },
1451
+ {
1452
+ "epoch": 0.5269186712485682,
1453
+ "forget_loss": 4.4586619196707034e-07,
1454
+ "retain_loss": 0.059775300323963165,
1455
+ "step": 460
1456
+ },
1457
+ {
1458
+ "epoch": 0.5269186712485682,
1459
+ "forget_loss": 4.324638211983256e-07,
1460
+ "retain_loss": 0.08108869194984436,
1461
+ "step": 460
1462
+ },
1463
+ {
1464
+ "epoch": 0.5269186712485682,
1465
+ "forget_loss": 5.558591738008545e-07,
1466
+ "retain_loss": 0.027880918234586716,
1467
+ "step": 460
1468
+ },
1469
+ {
1470
+ "epoch": 0.5383734249713631,
1471
+ "grad_norm": 1.171875,
1472
+ "learning_rate": 3.677995391705069e-05,
1473
+ "loss": 0.05814751386642456,
1474
+ "step": 470
1475
+ },
1476
+ {
1477
+ "epoch": 0.5383734249713631,
1478
+ "forget_loss": 3.032934046132141e-07,
1479
+ "retain_loss": 0.024084806442260742,
1480
+ "step": 470
1481
+ },
1482
+ {
1483
+ "epoch": 0.5383734249713631,
1484
+ "forget_loss": 8.804982485344226e-07,
1485
+ "retain_loss": 0.03435460478067398,
1486
+ "step": 470
1487
+ },
1488
+ {
1489
+ "epoch": 0.5383734249713631,
1490
+ "forget_loss": 1.1724860087269917e-06,
1491
+ "retain_loss": 0.048975635319948196,
1492
+ "step": 470
1493
+ },
1494
+ {
1495
+ "epoch": 0.5383734249713631,
1496
+ "forget_loss": 3.8122900036796636e-07,
1497
+ "retain_loss": 0.0677463635802269,
1498
+ "step": 470
1499
+ },
1500
+ {
1501
+ "epoch": 0.5498281786941581,
1502
+ "grad_norm": 1.734375,
1503
+ "learning_rate": 3.649193548387097e-05,
1504
+ "loss": 0.060336172580718994,
1505
+ "step": 480
1506
+ },
1507
+ {
1508
+ "epoch": 0.5498281786941581,
1509
+ "forget_loss": 9.458409522267175e-07,
1510
+ "retain_loss": 0.07034104317426682,
1511
+ "step": 480
1512
+ },
1513
+ {
1514
+ "epoch": 0.5498281786941581,
1515
+ "forget_loss": 6.857035828033986e-07,
1516
+ "retain_loss": 0.03801558166742325,
1517
+ "step": 480
1518
+ },
1519
+ {
1520
+ "epoch": 0.5498281786941581,
1521
+ "forget_loss": 1.3080871212878264e-06,
1522
+ "retain_loss": 0.07545764744281769,
1523
+ "step": 480
1524
+ },
1525
+ {
1526
+ "epoch": 0.5498281786941581,
1527
+ "forget_loss": 3.9946274910107604e-07,
1528
+ "retain_loss": 0.03869559243321419,
1529
+ "step": 480
1530
+ },
1531
+ {
1532
+ "epoch": 0.561282932416953,
1533
+ "grad_norm": 1.0546875,
1534
+ "learning_rate": 3.620391705069124e-05,
1535
+ "loss": 0.05845548510551453,
1536
+ "step": 490
1537
+ },
1538
+ {
1539
+ "epoch": 0.561282932416953,
1540
+ "forget_loss": 9.76757291937247e-07,
1541
+ "retain_loss": 0.04530367627739906,
1542
+ "step": 490
1543
+ },
1544
+ {
1545
+ "epoch": 0.561282932416953,
1546
+ "forget_loss": 5.240474933998485e-07,
1547
+ "retain_loss": 0.03550443798303604,
1548
+ "step": 490
1549
+ },
1550
+ {
1551
+ "epoch": 0.561282932416953,
1552
+ "forget_loss": 7.904294534455403e-07,
1553
+ "retain_loss": 0.08905749022960663,
1554
+ "step": 490
1555
+ },
1556
+ {
1557
+ "epoch": 0.561282932416953,
1558
+ "forget_loss": 3.536492840794381e-07,
1559
+ "retain_loss": 0.05929813161492348,
1560
+ "step": 490
1561
+ },
1562
+ {
1563
+ "epoch": 0.572737686139748,
1564
+ "grad_norm": 1.1640625,
1565
+ "learning_rate": 3.591589861751152e-05,
1566
+ "loss": 0.06306434273719788,
1567
+ "step": 500
1568
+ },
1569
+ {
1570
+ "epoch": 0.572737686139748,
1571
+ "forget_loss": 4.770619170813006e-07,
1572
+ "retain_loss": 0.0518912598490715,
1573
+ "step": 500
1574
+ },
1575
+ {
1576
+ "epoch": 0.572737686139748,
1577
+ "forget_loss": 7.711427940648719e-08,
1578
+ "retain_loss": 0.06265456974506378,
1579
+ "step": 500
1580
+ },
1581
+ {
1582
+ "epoch": 0.572737686139748,
1583
+ "forget_loss": 1.8670939425646793e-06,
1584
+ "retain_loss": 0.07513444125652313,
1585
+ "step": 500
1586
+ },
1587
+ {
1588
+ "epoch": 0.572737686139748,
1589
+ "forget_loss": 4.7316902396232763e-07,
1590
+ "retain_loss": 0.08540341258049011,
1591
+ "step": 500
1592
+ },
1593
+ {
1594
+ "epoch": 0.584192439862543,
1595
+ "grad_norm": 0.984375,
1596
+ "learning_rate": 3.5627880184331795e-05,
1597
+ "loss": 0.05606219172477722,
1598
+ "step": 510
1599
+ },
1600
+ {
1601
+ "epoch": 0.584192439862543,
1602
+ "forget_loss": 8.002095910342177e-07,
1603
+ "retain_loss": 0.04273177683353424,
1604
+ "step": 510
1605
+ },
1606
+ {
1607
+ "epoch": 0.584192439862543,
1608
+ "forget_loss": 1.8704803039781837e-07,
1609
+ "retain_loss": 0.053921207785606384,
1610
+ "step": 510
1611
+ },
1612
+ {
1613
+ "epoch": 0.584192439862543,
1614
+ "forget_loss": 3.775904247049766e-07,
1615
+ "retain_loss": 0.0322711281478405,
1616
+ "step": 510
1617
+ },
1618
+ {
1619
+ "epoch": 0.584192439862543,
1620
+ "forget_loss": 6.053579681974952e-07,
1621
+ "retain_loss": 0.05659232288599014,
1622
+ "step": 510
1623
+ },
1624
+ {
1625
+ "epoch": 0.5956471935853379,
1626
+ "grad_norm": 1.2265625,
1627
+ "learning_rate": 3.5339861751152074e-05,
1628
+ "loss": 0.049942076206207275,
1629
+ "step": 520
1630
+ },
1631
+ {
1632
+ "epoch": 0.5956471935853379,
1633
+ "forget_loss": 2.9679642921109917e-07,
1634
+ "retain_loss": 0.08346150070428848,
1635
+ "step": 520
1636
+ },
1637
+ {
1638
+ "epoch": 0.5956471935853379,
1639
+ "forget_loss": 2.005875501254195e-07,
1640
+ "retain_loss": 0.030936188995838165,
1641
+ "step": 520
1642
+ },
1643
+ {
1644
+ "epoch": 0.5956471935853379,
1645
+ "forget_loss": 1.0443299061080324e-06,
1646
+ "retain_loss": 0.09520356357097626,
1647
+ "step": 520
1648
+ },
1649
+ {
1650
+ "epoch": 0.5956471935853379,
1651
+ "forget_loss": 2.0819126689275436e-07,
1652
+ "retain_loss": 0.0474083349108696,
1653
+ "step": 520
1654
+ },
1655
+ {
1656
+ "epoch": 0.6071019473081328,
1657
+ "grad_norm": 1.0625,
1658
+ "learning_rate": 3.5051843317972346e-05,
1659
+ "loss": 0.05396345257759094,
1660
+ "step": 530
1661
+ },
1662
+ {
1663
+ "epoch": 0.6071019473081328,
1664
+ "forget_loss": 8.294673534692265e-07,
1665
+ "retain_loss": 0.042203839868307114,
1666
+ "step": 530
1667
+ },
1668
+ {
1669
+ "epoch": 0.6071019473081328,
1670
+ "forget_loss": 7.164442195062293e-08,
1671
+ "retain_loss": 0.030783096328377724,
1672
+ "step": 530
1673
+ },
1674
+ {
1675
+ "epoch": 0.6071019473081328,
1676
+ "forget_loss": 1.3861478691978846e-05,
1677
+ "retain_loss": 0.04198005795478821,
1678
+ "step": 530
1679
+ },
1680
+ {
1681
+ "epoch": 0.6071019473081328,
1682
+ "forget_loss": 4.225911993671616e-07,
1683
+ "retain_loss": 0.06045946478843689,
1684
+ "step": 530
1685
+ },
1686
+ {
1687
+ "epoch": 0.6185567010309279,
1688
+ "grad_norm": 1.2578125,
1689
+ "learning_rate": 3.4763824884792625e-05,
1690
+ "loss": 0.05722784399986267,
1691
+ "step": 540
1692
+ },
1693
+ {
1694
+ "epoch": 0.6185567010309279,
1695
+ "forget_loss": 3.0147746201691916e-06,
1696
+ "retain_loss": 0.0411076694726944,
1697
+ "step": 540
1698
+ },
1699
+ {
1700
+ "epoch": 0.6185567010309279,
1701
+ "forget_loss": 2.582411582352506e-07,
1702
+ "retain_loss": 0.023279590532183647,
1703
+ "step": 540
1704
+ },
1705
+ {
1706
+ "epoch": 0.6185567010309279,
1707
+ "forget_loss": 2.0758065488735156e-07,
1708
+ "retain_loss": 0.04550226032733917,
1709
+ "step": 540
1710
+ },
1711
+ {
1712
+ "epoch": 0.6185567010309279,
1713
+ "forget_loss": 9.63341904025583e-07,
1714
+ "retain_loss": 0.0418274812400341,
1715
+ "step": 540
1716
+ },
1717
+ {
1718
+ "epoch": 0.6300114547537228,
1719
+ "grad_norm": 1.3359375,
1720
+ "learning_rate": 3.4475806451612905e-05,
1721
+ "loss": 0.04913765788078308,
1722
+ "step": 550
1723
+ },
1724
+ {
1725
+ "epoch": 0.6300114547537228,
1726
+ "forget_loss": 2.0537379441520898e-07,
1727
+ "retain_loss": 0.046085719019174576,
1728
+ "step": 550
1729
+ },
1730
+ {
1731
+ "epoch": 0.6300114547537228,
1732
+ "forget_loss": 4.5911775714557734e-07,
1733
+ "retain_loss": 0.041781309992074966,
1734
+ "step": 550
1735
+ },
1736
+ {
1737
+ "epoch": 0.6300114547537228,
1738
+ "forget_loss": 1.873733936008648e-06,
1739
+ "retain_loss": 0.030963057652115822,
1740
+ "step": 550
1741
+ },
1742
+ {
1743
+ "epoch": 0.6300114547537228,
1744
+ "forget_loss": 2.3070414556514152e-07,
1745
+ "retain_loss": 0.036035358905792236,
1746
+ "step": 550
1747
+ },
1748
+ {
1749
+ "epoch": 0.6414662084765178,
1750
+ "grad_norm": 1.2109375,
1751
+ "learning_rate": 3.4187788018433184e-05,
1752
+ "loss": 0.046972605586051944,
1753
+ "step": 560
1754
+ },
1755
+ {
1756
+ "epoch": 0.6414662084765178,
1757
+ "forget_loss": 1.888665224214492e-07,
1758
+ "retain_loss": 0.04609245806932449,
1759
+ "step": 560
1760
+ },
1761
+ {
1762
+ "epoch": 0.6414662084765178,
1763
+ "forget_loss": 1.1302165603410685e-06,
1764
+ "retain_loss": 0.04581566900014877,
1765
+ "step": 560
1766
+ },
1767
+ {
1768
+ "epoch": 0.6414662084765178,
1769
+ "forget_loss": 3.713363128099445e-07,
1770
+ "retain_loss": 0.05701315402984619,
1771
+ "step": 560
1772
+ },
1773
+ {
1774
+ "epoch": 0.6414662084765178,
1775
+ "forget_loss": 1.0438467512585703e-07,
1776
+ "retain_loss": 0.04488621652126312,
1777
+ "step": 560
1778
+ },
1779
+ {
1780
+ "epoch": 0.6529209621993127,
1781
+ "grad_norm": 1.640625,
1782
+ "learning_rate": 3.389976958525346e-05,
1783
+ "loss": 0.04600127041339874,
1784
+ "step": 570
1785
+ },
1786
+ {
1787
+ "epoch": 0.6529209621993127,
1788
+ "forget_loss": 1.2589470088641974e-07,
1789
+ "retain_loss": 0.04046102985739708,
1790
+ "step": 570
1791
+ },
1792
+ {
1793
+ "epoch": 0.6529209621993127,
1794
+ "forget_loss": 2.068235062324675e-06,
1795
+ "retain_loss": 0.06527687609195709,
1796
+ "step": 570
1797
+ },
1798
+ {
1799
+ "epoch": 0.6529209621993127,
1800
+ "forget_loss": 4.6239173911999387e-07,
1801
+ "retain_loss": 0.02964375726878643,
1802
+ "step": 570
1803
+ },
1804
+ {
1805
+ "epoch": 0.6529209621993127,
1806
+ "forget_loss": 1.7551890323375119e-06,
1807
+ "retain_loss": 0.04939418286085129,
1808
+ "step": 570
1809
+ },
1810
+ {
1811
+ "epoch": 0.6643757159221076,
1812
+ "grad_norm": 1.4921875,
1813
+ "learning_rate": 3.3611751152073736e-05,
1814
+ "loss": 0.04693392515182495,
1815
+ "step": 580
1816
+ },
1817
+ {
1818
+ "epoch": 0.6643757159221076,
1819
+ "forget_loss": 1.162212441840893e-07,
1820
+ "retain_loss": 0.01956232823431492,
1821
+ "step": 580
1822
+ },
1823
+ {
1824
+ "epoch": 0.6643757159221076,
1825
+ "forget_loss": 5.034058858655044e-07,
1826
+ "retain_loss": 0.029308225959539413,
1827
+ "step": 580
1828
+ },
1829
+ {
1830
+ "epoch": 0.6643757159221076,
1831
+ "forget_loss": 1.8678618118883605e-07,
1832
+ "retain_loss": 0.057615552097558975,
1833
+ "step": 580
1834
+ },
1835
+ {
1836
+ "epoch": 0.6643757159221076,
1837
+ "forget_loss": 5.4416375405708095e-08,
1838
+ "retain_loss": 0.06805437058210373,
1839
+ "step": 580
1840
+ },
1841
+ {
1842
+ "epoch": 0.6758304696449027,
1843
+ "grad_norm": 0.85546875,
1844
+ "learning_rate": 3.3323732718894015e-05,
1845
+ "loss": 0.044681668281555176,
1846
+ "step": 590
1847
+ },
1848
+ {
1849
+ "epoch": 0.6758304696449027,
1850
+ "forget_loss": 6.884723688926897e-07,
1851
+ "retain_loss": 0.032636139541864395,
1852
+ "step": 590
1853
+ },
1854
+ {
1855
+ "epoch": 0.6758304696449027,
1856
+ "forget_loss": 2.1161808660963288e-07,
1857
+ "retain_loss": 0.03656921535730362,
1858
+ "step": 590
1859
+ },
1860
+ {
1861
+ "epoch": 0.6758304696449027,
1862
+ "forget_loss": 3.788201468069019e-07,
1863
+ "retain_loss": 0.0688297301530838,
1864
+ "step": 590
1865
+ },
1866
+ {
1867
+ "epoch": 0.6758304696449027,
1868
+ "forget_loss": 8.874022228155809e-08,
1869
+ "retain_loss": 0.021696442738175392,
1870
+ "step": 590
1871
+ },
1872
+ {
1873
+ "epoch": 0.6872852233676976,
1874
+ "grad_norm": 0.94140625,
1875
+ "learning_rate": 3.303571428571429e-05,
1876
+ "loss": 0.04318673312664032,
1877
+ "step": 600
1878
+ },
1879
+ {
1880
+ "epoch": 0.6872852233676976,
1881
+ "forget_loss": 3.341031913350889e-07,
1882
+ "retain_loss": 0.03312600031495094,
1883
+ "step": 600
1884
+ },
1885
+ {
1886
+ "epoch": 0.6872852233676976,
1887
+ "forget_loss": 1.59767705554259e-06,
1888
+ "retain_loss": 0.037248943001031876,
1889
+ "step": 600
1890
+ },
1891
+ {
1892
+ "epoch": 0.6872852233676976,
1893
+ "forget_loss": 1.2864059328876465e-07,
1894
+ "retain_loss": 0.051617469638586044,
1895
+ "step": 600
1896
+ },
1897
+ {
1898
+ "epoch": 0.6872852233676976,
1899
+ "forget_loss": 1.3946912247320142e-07,
1900
+ "retain_loss": 0.07654328644275665,
1901
+ "step": 600
1902
+ },
1903
+ {
1904
+ "epoch": 0.6987399770904925,
1905
+ "grad_norm": 0.6484375,
1906
+ "learning_rate": 3.2747695852534566e-05,
1907
+ "loss": 0.043347522616386414,
1908
+ "step": 610
1909
+ },
1910
+ {
1911
+ "epoch": 0.6987399770904925,
1912
+ "forget_loss": 1.6136347369410942e-07,
1913
+ "retain_loss": 0.06624381989240646,
1914
+ "step": 610
1915
+ },
1916
+ {
1917
+ "epoch": 0.6987399770904925,
1918
+ "forget_loss": 1.5072893688738986e-07,
1919
+ "retain_loss": 0.025636199861764908,
1920
+ "step": 610
1921
+ },
1922
+ {
1923
+ "epoch": 0.6987399770904925,
1924
+ "forget_loss": 3.572060336409777e-07,
1925
+ "retain_loss": 0.029844949021935463,
1926
+ "step": 610
1927
+ },
1928
+ {
1929
+ "epoch": 0.6987399770904925,
1930
+ "forget_loss": 3.3666503895801725e-07,
1931
+ "retain_loss": 0.08041810244321823,
1932
+ "step": 610
1933
+ },
1934
+ {
1935
+ "epoch": 0.7101947308132875,
1936
+ "grad_norm": 0.76953125,
1937
+ "learning_rate": 3.245967741935484e-05,
1938
+ "loss": 0.04078985154628754,
1939
+ "step": 620
1940
+ },
1941
+ {
1942
+ "epoch": 0.7101947308132875,
1943
+ "forget_loss": 2.2686955958306498e-07,
1944
+ "retain_loss": 0.03770332783460617,
1945
+ "step": 620
1946
+ },
1947
+ {
1948
+ "epoch": 0.7101947308132875,
1949
+ "forget_loss": 1.2471460308915994e-07,
1950
+ "retain_loss": 0.04125113785266876,
1951
+ "step": 620
1952
+ },
1953
+ {
1954
+ "epoch": 0.7101947308132875,
1955
+ "forget_loss": 5.980033392916084e-07,
1956
+ "retain_loss": 0.028695637360215187,
1957
+ "step": 620
1958
+ },
1959
+ {
1960
+ "epoch": 0.7101947308132875,
1961
+ "forget_loss": 2.1789082893519662e-07,
1962
+ "retain_loss": 0.021218575537204742,
1963
+ "step": 620
1964
+ },
1965
+ {
1966
+ "epoch": 0.7216494845360825,
1967
+ "grad_norm": 0.72265625,
1968
+ "learning_rate": 3.217165898617512e-05,
1969
+ "loss": 0.03691001832485199,
1970
+ "step": 630
1971
+ },
1972
+ {
1973
+ "epoch": 0.7216494845360825,
1974
+ "forget_loss": 5.882121172362531e-07,
1975
+ "retain_loss": 0.028744207695126534,
1976
+ "step": 630
1977
+ },
1978
+ {
1979
+ "epoch": 0.7216494845360825,
1980
+ "forget_loss": 3.706446420892462e-07,
1981
+ "retain_loss": 0.037669021636247635,
1982
+ "step": 630
1983
+ },
1984
+ {
1985
+ "epoch": 0.7216494845360825,
1986
+ "forget_loss": 9.750963272381341e-07,
1987
+ "retain_loss": 0.055296238511800766,
1988
+ "step": 630
1989
+ },
1990
+ {
1991
+ "epoch": 0.7216494845360825,
1992
+ "forget_loss": 2.988974188156135e-07,
1993
+ "retain_loss": 0.04488551616668701,
1994
+ "step": 630
1995
+ },
1996
+ {
1997
+ "epoch": 0.7331042382588774,
1998
+ "grad_norm": 0.8671875,
1999
+ "learning_rate": 3.188364055299539e-05,
2000
+ "loss": 0.038949227333068846,
2001
+ "step": 640
2002
+ },
2003
+ {
2004
+ "epoch": 0.7331042382588774,
2005
+ "forget_loss": 1.2395511816976068e-07,
2006
+ "retain_loss": 0.027795039117336273,
2007
+ "step": 640
2008
+ },
2009
+ {
2010
+ "epoch": 0.7331042382588774,
2011
+ "forget_loss": 5.673878149536904e-07,
2012
+ "retain_loss": 0.06736742705106735,
2013
+ "step": 640
2014
+ },
2015
+ {
2016
+ "epoch": 0.7331042382588774,
2017
+ "forget_loss": 1.1004679834059061e-07,
2018
+ "retain_loss": 0.0364450179040432,
2019
+ "step": 640
2020
+ },
2021
+ {
2022
+ "epoch": 0.7331042382588774,
2023
+ "forget_loss": 1.2864478549090563e-07,
2024
+ "retain_loss": 0.04605640098452568,
2025
+ "step": 640
2026
+ },
2027
+ {
2028
+ "epoch": 0.7445589919816724,
2029
+ "grad_norm": 1.2578125,
2030
+ "learning_rate": 3.159562211981567e-05,
2031
+ "loss": 0.03986696600914001,
2032
+ "step": 650
2033
+ },
2034
+ {
2035
+ "epoch": 0.7445589919816724,
2036
+ "forget_loss": 5.023429139328073e-07,
2037
+ "retain_loss": 0.035111404955387115,
2038
+ "step": 650
2039
+ },
2040
+ {
2041
+ "epoch": 0.7445589919816724,
2042
+ "forget_loss": 4.809768370250822e-07,
2043
+ "retain_loss": 0.02187221683561802,
2044
+ "step": 650
2045
+ },
2046
+ {
2047
+ "epoch": 0.7445589919816724,
2048
+ "forget_loss": 4.244728188496083e-06,
2049
+ "retain_loss": 0.08197719603776932,
2050
+ "step": 650
2051
+ },
2052
+ {
2053
+ "epoch": 0.7445589919816724,
2054
+ "forget_loss": 1.1673816402435477e-07,
2055
+ "retain_loss": 0.08870132267475128,
2056
+ "step": 650
2057
+ },
2058
+ {
2059
+ "epoch": 0.7560137457044673,
2060
+ "grad_norm": 0.6640625,
2061
+ "learning_rate": 3.130760368663594e-05,
2062
+ "loss": 0.039774364233016966,
2063
+ "step": 660
2064
+ },
2065
+ {
2066
+ "epoch": 0.7560137457044673,
2067
+ "forget_loss": 1.1249169062921283e-07,
2068
+ "retain_loss": 0.038397908210754395,
2069
+ "step": 660
2070
+ },
2071
+ {
2072
+ "epoch": 0.7560137457044673,
2073
+ "forget_loss": 5.782552534583374e-07,
2074
+ "retain_loss": 0.030945003032684326,
2075
+ "step": 660
2076
+ },
2077
+ {
2078
+ "epoch": 0.7560137457044673,
2079
+ "forget_loss": 2.596835315671342e-07,
2080
+ "retain_loss": 0.04453842341899872,
2081
+ "step": 660
2082
+ },
2083
+ {
2084
+ "epoch": 0.7560137457044673,
2085
+ "forget_loss": 8.396180533054576e-07,
2086
+ "retain_loss": 0.04271215945482254,
2087
+ "step": 660
2088
+ },
2089
+ {
2090
+ "epoch": 0.7674684994272624,
2091
+ "grad_norm": 0.89453125,
2092
+ "learning_rate": 3.101958525345622e-05,
2093
+ "loss": 0.03887006938457489,
2094
+ "step": 670
2095
+ },
2096
+ {
2097
+ "epoch": 0.7674684994272624,
2098
+ "forget_loss": 8.253943661884477e-08,
2099
+ "retain_loss": 0.03154648467898369,
2100
+ "step": 670
2101
+ },
2102
+ {
2103
+ "epoch": 0.7674684994272624,
2104
+ "forget_loss": 4.430251010489883e-07,
2105
+ "retain_loss": 0.047183044254779816,
2106
+ "step": 670
2107
+ },
2108
+ {
2109
+ "epoch": 0.7674684994272624,
2110
+ "forget_loss": 5.662621447299898e-07,
2111
+ "retain_loss": 0.0315515510737896,
2112
+ "step": 670
2113
+ },
2114
+ {
2115
+ "epoch": 0.7674684994272624,
2116
+ "forget_loss": 2.8115960049035493e-07,
2117
+ "retain_loss": 0.024359095841646194,
2118
+ "step": 670
2119
+ },
2120
+ {
2121
+ "epoch": 0.7789232531500573,
2122
+ "grad_norm": 0.890625,
2123
+ "learning_rate": 3.07315668202765e-05,
2124
+ "loss": 0.03558169901371002,
2125
+ "step": 680
2126
+ },
2127
+ {
2128
+ "epoch": 0.7789232531500573,
2129
+ "forget_loss": 2.95549085649327e-07,
2130
+ "retain_loss": 0.03642899915575981,
2131
+ "step": 680
2132
+ },
2133
+ {
2134
+ "epoch": 0.7789232531500573,
2135
+ "forget_loss": 7.595866691190167e-08,
2136
+ "retain_loss": 0.026957420632243156,
2137
+ "step": 680
2138
+ },
2139
+ {
2140
+ "epoch": 0.7789232531500573,
2141
+ "forget_loss": 3.971479429765168e-07,
2142
+ "retain_loss": 0.026554634794592857,
2143
+ "step": 680
2144
+ },
2145
+ {
2146
+ "epoch": 0.7789232531500573,
2147
+ "forget_loss": 1.0031957572209649e-05,
2148
+ "retain_loss": 0.055837519466876984,
2149
+ "step": 680
2150
+ },
2151
+ {
2152
+ "epoch": 0.7903780068728522,
2153
+ "grad_norm": 0.72265625,
2154
+ "learning_rate": 3.0443548387096776e-05,
2155
+ "loss": 0.032911685109138486,
2156
+ "step": 690
2157
+ },
2158
+ {
2159
+ "epoch": 0.7903780068728522,
2160
+ "forget_loss": 2.3226121470543148e-07,
2161
+ "retain_loss": 0.028471192345023155,
2162
+ "step": 690
2163
+ },
2164
+ {
2165
+ "epoch": 0.7903780068728522,
2166
+ "forget_loss": 2.1240731484795106e-07,
2167
+ "retain_loss": 0.061532407999038696,
2168
+ "step": 690
2169
+ },
2170
+ {
2171
+ "epoch": 0.7903780068728522,
2172
+ "forget_loss": 3.4258829373357e-07,
2173
+ "retain_loss": 0.045654185116291046,
2174
+ "step": 690
2175
+ },
2176
+ {
2177
+ "epoch": 0.7903780068728522,
2178
+ "forget_loss": 1.3982165114612144e-07,
2179
+ "retain_loss": 0.04419733211398125,
2180
+ "step": 690
2181
+ },
2182
+ {
2183
+ "epoch": 0.8018327605956472,
2184
+ "grad_norm": 0.369140625,
2185
+ "learning_rate": 3.015552995391705e-05,
2186
+ "loss": 0.03598496913909912,
2187
+ "step": 700
2188
+ },
2189
+ {
2190
+ "epoch": 0.8018327605956472,
2191
+ "forget_loss": 3.2066935773400473e-07,
2192
+ "retain_loss": 0.020428789779543877,
2193
+ "step": 700
2194
+ },
2195
+ {
2196
+ "epoch": 0.8018327605956472,
2197
+ "forget_loss": 9.008678603095177e-07,
2198
+ "retain_loss": 0.03619084134697914,
2199
+ "step": 700
2200
+ },
2201
+ {
2202
+ "epoch": 0.8018327605956472,
2203
+ "forget_loss": 3.202090113063605e-07,
2204
+ "retain_loss": 0.04894010350108147,
2205
+ "step": 700
2206
+ },
2207
+ {
2208
+ "epoch": 0.8018327605956472,
2209
+ "forget_loss": 3.2703263741495903e-07,
2210
+ "retain_loss": 0.039299145340919495,
2211
+ "step": 700
2212
+ },
2213
+ {
2214
+ "epoch": 0.8132875143184422,
2215
+ "grad_norm": 0.48046875,
2216
+ "learning_rate": 2.9867511520737328e-05,
2217
+ "loss": 0.03154398202896118,
2218
+ "step": 710
2219
+ },
2220
+ {
2221
+ "epoch": 0.8132875143184422,
2222
+ "forget_loss": 1.6021775195440568e-07,
2223
+ "retain_loss": 0.05344302952289581,
2224
+ "step": 710
2225
+ },
2226
+ {
2227
+ "epoch": 0.8132875143184422,
2228
+ "forget_loss": 3.327061222080374e-07,
2229
+ "retain_loss": 0.018441231921315193,
2230
+ "step": 710
2231
+ },
2232
+ {
2233
+ "epoch": 0.8132875143184422,
2234
+ "forget_loss": 1.2161015661149577e-07,
2235
+ "retain_loss": 0.01777707226574421,
2236
+ "step": 710
2237
+ },
2238
+ {
2239
+ "epoch": 0.8132875143184422,
2240
+ "forget_loss": 2.601665585189039e-07,
2241
+ "retain_loss": 0.021317142993211746,
2242
+ "step": 710
2243
+ },
2244
+ {
2245
+ "epoch": 0.8247422680412371,
2246
+ "grad_norm": 0.71484375,
2247
+ "learning_rate": 2.9579493087557604e-05,
2248
+ "loss": 0.031114420294761656,
2249
+ "step": 720
2250
+ },
2251
+ {
2252
+ "epoch": 0.8247422680412371,
2253
+ "forget_loss": 3.656723990275168e-08,
2254
+ "retain_loss": 0.02041560783982277,
2255
+ "step": 720
2256
+ },
2257
+ {
2258
+ "epoch": 0.8247422680412371,
2259
+ "forget_loss": 7.133304791295814e-08,
2260
+ "retain_loss": 0.0420437790453434,
2261
+ "step": 720
2262
+ },
2263
+ {
2264
+ "epoch": 0.8247422680412371,
2265
+ "forget_loss": 1.2686764705449605e-07,
2266
+ "retain_loss": 0.03707875683903694,
2267
+ "step": 720
2268
+ },
2269
+ {
2270
+ "epoch": 0.8247422680412371,
2271
+ "forget_loss": 3.098398337897379e-07,
2272
+ "retain_loss": 0.03388817980885506,
2273
+ "step": 720
2274
+ },
2275
+ {
2276
+ "epoch": 0.8361970217640321,
2277
+ "grad_norm": 0.62109375,
2278
+ "learning_rate": 2.9291474654377883e-05,
2279
+ "loss": 0.031438523530960084,
2280
+ "step": 730
2281
+ },
2282
+ {
2283
+ "epoch": 0.8361970217640321,
2284
+ "forget_loss": 6.143314408291189e-07,
2285
+ "retain_loss": 0.037241023033857346,
2286
+ "step": 730
2287
+ },
2288
+ {
2289
+ "epoch": 0.8361970217640321,
2290
+ "forget_loss": 1.567052549944492e-07,
2291
+ "retain_loss": 0.029914623126387596,
2292
+ "step": 730
2293
+ },
2294
+ {
2295
+ "epoch": 0.8361970217640321,
2296
+ "forget_loss": 1.7319891298939183e-07,
2297
+ "retain_loss": 0.026813646778464317,
2298
+ "step": 730
2299
+ },
2300
+ {
2301
+ "epoch": 0.8361970217640321,
2302
+ "forget_loss": 5.76246179662121e-08,
2303
+ "retain_loss": 0.036752790212631226,
2304
+ "step": 730
2305
+ },
2306
+ {
2307
+ "epoch": 0.847651775486827,
2308
+ "grad_norm": 0.447265625,
2309
+ "learning_rate": 2.9003456221198155e-05,
2310
+ "loss": 0.028349018096923827,
2311
+ "step": 740
2312
+ },
2313
+ {
2314
+ "epoch": 0.847651775486827,
2315
+ "forget_loss": 1.0741296136984602e-06,
2316
+ "retain_loss": 0.028334422037005424,
2317
+ "step": 740
2318
+ },
2319
+ {
2320
+ "epoch": 0.847651775486827,
2321
+ "forget_loss": 5.47009335605253e-07,
2322
+ "retain_loss": 0.03212451562285423,
2323
+ "step": 740
2324
+ },
2325
+ {
2326
+ "epoch": 0.847651775486827,
2327
+ "forget_loss": 7.226554998851498e-08,
2328
+ "retain_loss": 0.038880933076143265,
2329
+ "step": 740
2330
+ },
2331
+ {
2332
+ "epoch": 0.847651775486827,
2333
+ "forget_loss": 1.0503694625185744e-07,
2334
+ "retain_loss": 0.020103227347135544,
2335
+ "step": 740
2336
+ },
2337
+ {
2338
+ "epoch": 0.8591065292096219,
2339
+ "grad_norm": 0.62109375,
2340
+ "learning_rate": 2.8715437788018435e-05,
2341
+ "loss": 0.027798715233802795,
2342
+ "step": 750
2343
+ },
2344
+ {
2345
+ "epoch": 0.8591065292096219,
2346
+ "forget_loss": 9.740627149312786e-08,
2347
+ "retain_loss": 0.03145676106214523,
2348
+ "step": 750
2349
+ },
2350
+ {
2351
+ "epoch": 0.8591065292096219,
2352
+ "forget_loss": 8.817877272804253e-08,
2353
+ "retain_loss": 0.0294309351593256,
2354
+ "step": 750
2355
+ },
2356
+ {
2357
+ "epoch": 0.8591065292096219,
2358
+ "forget_loss": 2.1280260398270912e-07,
2359
+ "retain_loss": 0.01636165753006935,
2360
+ "step": 750
2361
+ },
2362
+ {
2363
+ "epoch": 0.8591065292096219,
2364
+ "forget_loss": 2.8950364594493294e-07,
2365
+ "retain_loss": 0.03548318147659302,
2366
+ "step": 750
2367
+ },
2368
+ {
2369
+ "epoch": 0.870561282932417,
2370
+ "grad_norm": 0.26171875,
2371
+ "learning_rate": 2.8427419354838714e-05,
2372
+ "loss": 0.027926307916641236,
2373
+ "step": 760
2374
+ },
2375
+ {
2376
+ "epoch": 0.870561282932417,
2377
+ "forget_loss": 1.0444756526339916e-07,
2378
+ "retain_loss": 0.035280242562294006,
2379
+ "step": 760
2380
+ },
2381
+ {
2382
+ "epoch": 0.870561282932417,
2383
+ "forget_loss": 1.4455953589731507e-07,
2384
+ "retain_loss": 0.03736939653754234,
2385
+ "step": 760
2386
+ },
2387
+ {
2388
+ "epoch": 0.870561282932417,
2389
+ "forget_loss": 8.093392267483068e-08,
2390
+ "retain_loss": 0.02323060669004917,
2391
+ "step": 760
2392
+ },
2393
+ {
2394
+ "epoch": 0.870561282932417,
2395
+ "forget_loss": 4.825432711186295e-07,
2396
+ "retain_loss": 0.028354808688163757,
2397
+ "step": 760
2398
+ },
2399
+ {
2400
+ "epoch": 0.8820160366552119,
2401
+ "grad_norm": 0.419921875,
2402
+ "learning_rate": 2.8139400921658986e-05,
2403
+ "loss": 0.02746644616127014,
2404
+ "step": 770
2405
+ },
2406
+ {
2407
+ "epoch": 0.8820160366552119,
2408
+ "forget_loss": 6.515568173881547e-08,
2409
+ "retain_loss": 0.022168956696987152,
2410
+ "step": 770
2411
+ },
2412
+ {
2413
+ "epoch": 0.8820160366552119,
2414
+ "forget_loss": 4.0877694118535146e-08,
2415
+ "retain_loss": 0.02454095333814621,
2416
+ "step": 770
2417
+ },
2418
+ {
2419
+ "epoch": 0.8820160366552119,
2420
+ "forget_loss": 4.9038700211667674e-08,
2421
+ "retain_loss": 0.009829733520746231,
2422
+ "step": 770
2423
+ },
2424
+ {
2425
+ "epoch": 0.8820160366552119,
2426
+ "forget_loss": 2.664446583366953e-07,
2427
+ "retain_loss": 0.01737787388265133,
2428
+ "step": 770
2429
+ },
2430
+ {
2431
+ "epoch": 0.8934707903780069,
2432
+ "grad_norm": 0.482421875,
2433
+ "learning_rate": 2.7851382488479265e-05,
2434
+ "loss": 0.025814527273178102,
2435
+ "step": 780
2436
+ },
2437
+ {
2438
+ "epoch": 0.8934707903780069,
2439
+ "forget_loss": 3.281249050246515e-08,
2440
+ "retain_loss": 0.020622696727514267,
2441
+ "step": 780
2442
+ },
2443
+ {
2444
+ "epoch": 0.8934707903780069,
2445
+ "forget_loss": 5.478144515791428e-08,
2446
+ "retain_loss": 0.020266670733690262,
2447
+ "step": 780
2448
+ },
2449
+ {
2450
+ "epoch": 0.8934707903780069,
2451
+ "forget_loss": 6.129197061000013e-08,
2452
+ "retain_loss": 0.018652314320206642,
2453
+ "step": 780
2454
+ },
2455
+ {
2456
+ "epoch": 0.8934707903780069,
2457
+ "forget_loss": 2.284050850676067e-07,
2458
+ "retain_loss": 0.013459251262247562,
2459
+ "step": 780
2460
+ },
2461
+ {
2462
+ "epoch": 0.9049255441008018,
2463
+ "grad_norm": 0.51953125,
2464
+ "learning_rate": 2.756336405529954e-05,
2465
+ "loss": 0.027006852626800536,
2466
+ "step": 790
2467
+ },
2468
+ {
2469
+ "epoch": 0.9049255441008018,
2470
+ "forget_loss": 2.6094051364111692e-08,
2471
+ "retain_loss": 0.018663065508008003,
2472
+ "step": 790
2473
+ },
2474
+ {
2475
+ "epoch": 0.9049255441008018,
2476
+ "forget_loss": 2.7085914666713506e-07,
2477
+ "retain_loss": 0.009133810177445412,
2478
+ "step": 790
2479
+ },
2480
+ {
2481
+ "epoch": 0.9049255441008018,
2482
+ "forget_loss": 8.054676214896972e-08,
2483
+ "retain_loss": 0.02572845108807087,
2484
+ "step": 790
2485
+ },
2486
+ {
2487
+ "epoch": 0.9049255441008018,
2488
+ "forget_loss": 1.0726026289376023e-07,
2489
+ "retain_loss": 0.024219941347837448,
2490
+ "step": 790
2491
+ },
2492
+ {
2493
+ "epoch": 0.9163802978235968,
2494
+ "grad_norm": 0.328125,
2495
+ "learning_rate": 2.727534562211982e-05,
2496
+ "loss": 0.02423024773597717,
2497
+ "step": 800
2498
+ },
2499
+ {
2500
+ "epoch": 0.9163802978235968,
2501
+ "forget_loss": 7.722936601339825e-08,
2502
+ "retain_loss": 0.04125501587986946,
2503
+ "step": 800
2504
+ },
2505
+ {
2506
+ "epoch": 0.9163802978235968,
2507
+ "forget_loss": 9.053938043734888e-08,
2508
+ "retain_loss": 0.02275395393371582,
2509
+ "step": 800
2510
+ },
2511
+ {
2512
+ "epoch": 0.9163802978235968,
2513
+ "forget_loss": 1.4687527993828553e-07,
2514
+ "retain_loss": 0.026991432532668114,
2515
+ "step": 800
2516
+ },
2517
+ {
2518
+ "epoch": 0.9163802978235968,
2519
+ "forget_loss": 3.380711177669582e-07,
2520
+ "retain_loss": 0.025429554283618927,
2521
+ "step": 800
2522
+ },
2523
+ {
2524
+ "epoch": 0.9278350515463918,
2525
+ "grad_norm": 0.74609375,
2526
+ "learning_rate": 2.6987327188940093e-05,
2527
+ "loss": 0.027974823117256166,
2528
+ "step": 810
2529
+ },
2530
+ {
2531
+ "epoch": 0.9278350515463918,
2532
+ "forget_loss": 2.2884126593680776e-08,
2533
+ "retain_loss": 0.034118637442588806,
2534
+ "step": 810
2535
+ },
2536
+ {
2537
+ "epoch": 0.9278350515463918,
2538
+ "forget_loss": 2.2747957473256974e-07,
2539
+ "retain_loss": 0.016463886946439743,
2540
+ "step": 810
2541
+ },
2542
+ {
2543
+ "epoch": 0.9278350515463918,
2544
+ "forget_loss": 2.8712694444266162e-08,
2545
+ "retain_loss": 0.05120648816227913,
2546
+ "step": 810
2547
+ },
2548
+ {
2549
+ "epoch": 0.9278350515463918,
2550
+ "forget_loss": 3.9451157363146194e-07,
2551
+ "retain_loss": 0.015402361750602722,
2552
+ "step": 810
2553
+ },
2554
+ {
2555
+ "epoch": 0.9392898052691867,
2556
+ "grad_norm": 0.232421875,
2557
+ "learning_rate": 2.6699308755760372e-05,
2558
+ "loss": 0.025883400440216066,
2559
+ "step": 820
2560
+ },
2561
+ {
2562
+ "epoch": 0.9392898052691867,
2563
+ "forget_loss": 6.696375720594006e-08,
2564
+ "retain_loss": 0.03956631198525429,
2565
+ "step": 820
2566
+ },
2567
+ {
2568
+ "epoch": 0.9392898052691867,
2569
+ "forget_loss": 3.5951167376424564e-08,
2570
+ "retain_loss": 0.024411948397755623,
2571
+ "step": 820
2572
+ },
2573
+ {
2574
+ "epoch": 0.9392898052691867,
2575
+ "forget_loss": 1.2076303335106786e-07,
2576
+ "retain_loss": 0.03519268333911896,
2577
+ "step": 820
2578
+ },
2579
+ {
2580
+ "epoch": 0.9392898052691867,
2581
+ "forget_loss": 3.650887236972267e-08,
2582
+ "retain_loss": 0.022809229791164398,
2583
+ "step": 820
2584
+ },
2585
+ {
2586
+ "epoch": 0.9507445589919816,
2587
+ "grad_norm": 0.181640625,
2588
+ "learning_rate": 2.6411290322580645e-05,
2589
+ "loss": 0.026767241954803466,
2590
+ "step": 830
2591
+ },
2592
+ {
2593
+ "epoch": 0.9507445589919816,
2594
+ "forget_loss": 1.1937875399326003e-07,
2595
+ "retain_loss": 0.009928344748914242,
2596
+ "step": 830
2597
+ },
2598
+ {
2599
+ "epoch": 0.9507445589919816,
2600
+ "forget_loss": 4.2366110619695974e-07,
2601
+ "retain_loss": 0.010477675125002861,
2602
+ "step": 830
2603
+ },
2604
+ {
2605
+ "epoch": 0.9507445589919816,
2606
+ "forget_loss": 9.538210576920392e-08,
2607
+ "retain_loss": 0.030232908204197884,
2608
+ "step": 830
2609
+ },
2610
+ {
2611
+ "epoch": 0.9507445589919816,
2612
+ "forget_loss": 4.3475404254422756e-07,
2613
+ "retain_loss": 0.027472183108329773,
2614
+ "step": 830
2615
+ },
2616
+ {
2617
+ "epoch": 0.9621993127147767,
2618
+ "grad_norm": 0.38671875,
2619
+ "learning_rate": 2.6123271889400924e-05,
2620
+ "loss": 0.026079732179641723,
2621
+ "step": 840
2622
+ },
2623
+ {
2624
+ "epoch": 0.9621993127147767,
2625
+ "forget_loss": 9.560527303165145e-08,
2626
+ "retain_loss": 0.035694848746061325,
2627
+ "step": 840
2628
+ },
2629
+ {
2630
+ "epoch": 0.9621993127147767,
2631
+ "forget_loss": 9.010035739720479e-08,
2632
+ "retain_loss": 0.02770412340760231,
2633
+ "step": 840
2634
+ },
2635
+ {
2636
+ "epoch": 0.9621993127147767,
2637
+ "forget_loss": 1.2952777979080565e-06,
2638
+ "retain_loss": 0.025091225281357765,
2639
+ "step": 840
2640
+ },
2641
+ {
2642
+ "epoch": 0.9621993127147767,
2643
+ "forget_loss": 1.8988541228281974e-07,
2644
+ "retain_loss": 0.02703693136572838,
2645
+ "step": 840
2646
+ },
2647
+ {
2648
+ "epoch": 0.9736540664375716,
2649
+ "grad_norm": 0.62109375,
2650
+ "learning_rate": 2.5835253456221196e-05,
2651
+ "loss": 0.024774086475372315,
2652
+ "step": 850
2653
+ },
2654
+ {
2655
+ "epoch": 0.9736540664375716,
2656
+ "forget_loss": 2.6555783350090678e-08,
2657
+ "retain_loss": 0.033134713768959045,
2658
+ "step": 850
2659
+ },
2660
+ {
2661
+ "epoch": 0.9736540664375716,
2662
+ "forget_loss": 2.7541537406250427e-07,
2663
+ "retain_loss": 0.018386946991086006,
2664
+ "step": 850
2665
+ },
2666
+ {
2667
+ "epoch": 0.9736540664375716,
2668
+ "forget_loss": 2.7093907561948072e-08,
2669
+ "retain_loss": 0.022385219112038612,
2670
+ "step": 850
2671
+ },
2672
+ {
2673
+ "epoch": 0.9736540664375716,
2674
+ "forget_loss": 6.318795442439296e-08,
2675
+ "retain_loss": 0.03880416601896286,
2676
+ "step": 850
2677
+ },
2678
+ {
2679
+ "epoch": 0.9851088201603666,
2680
+ "grad_norm": 0.57421875,
2681
+ "learning_rate": 2.5547235023041475e-05,
2682
+ "loss": 0.02385828346014023,
2683
+ "step": 860
2684
+ },
2685
+ {
2686
+ "epoch": 0.9851088201603666,
2687
+ "forget_loss": 3.6007147485861424e-08,
2688
+ "retain_loss": 0.03307569772005081,
2689
+ "step": 860
2690
+ },
2691
+ {
2692
+ "epoch": 0.9851088201603666,
2693
+ "forget_loss": 8.178755450671815e-08,
2694
+ "retain_loss": 0.02462468110024929,
2695
+ "step": 860
2696
+ },
2697
+ {
2698
+ "epoch": 0.9851088201603666,
2699
+ "forget_loss": 6.628091853144724e-08,
2700
+ "retain_loss": 0.020657110959291458,
2701
+ "step": 860
2702
+ },
2703
+ {
2704
+ "epoch": 0.9851088201603666,
2705
+ "forget_loss": 6.628161486332829e-08,
2706
+ "retain_loss": 0.032617855817079544,
2707
+ "step": 860
2708
+ },
2709
+ {
2710
+ "epoch": 0.9965635738831615,
2711
+ "grad_norm": 0.236328125,
2712
+ "learning_rate": 2.525921658986175e-05,
2713
+ "loss": 0.024864019453525545,
2714
+ "step": 870
2715
+ },
2716
+ {
2717
+ "epoch": 0.9965635738831615,
2718
+ "forget_loss": 7.505889243475394e-08,
2719
+ "retain_loss": 0.01775062270462513,
2720
+ "step": 870
2721
+ },
2722
+ {
2723
+ "epoch": 0.9965635738831615,
2724
+ "forget_loss": 1.0310161258075823e-07,
2725
+ "retain_loss": 0.02414717711508274,
2726
+ "step": 870
2727
+ },
2728
+ {
2729
+ "epoch": 0.9965635738831615,
2730
+ "forget_loss": 2.3729118936444138e-07,
2731
+ "retain_loss": 0.027325890958309174,
2732
+ "step": 870
2733
+ },
2734
+ {
2735
+ "epoch": 0.9965635738831615,
2736
+ "forget_loss": 2.6968228539203665e-08,
2737
+ "retain_loss": 0.022969461977481842,
2738
+ "step": 870
2739
+ },
2740
+ {
2741
+ "epoch": 1.0080183276059564,
2742
+ "grad_norm": 0.2041015625,
2743
+ "learning_rate": 2.497119815668203e-05,
2744
+ "loss": 0.025278234481811525,
2745
+ "step": 880
2746
+ },
2747
+ {
2748
+ "epoch": 1.0080183276059564,
2749
+ "forget_loss": 1.2440519014944584e-07,
2750
+ "retain_loss": 0.028113704174757004,
2751
+ "step": 880
2752
+ },
2753
+ {
2754
+ "epoch": 1.0080183276059564,
2755
+ "forget_loss": 6.923085038579302e-08,
2756
+ "retain_loss": 0.035193175077438354,
2757
+ "step": 880
2758
+ },
2759
+ {
2760
+ "epoch": 1.0080183276059564,
2761
+ "forget_loss": 3.75761644022532e-08,
2762
+ "retain_loss": 0.018518565222620964,
2763
+ "step": 880
2764
+ },
2765
+ {
2766
+ "epoch": 1.0080183276059564,
2767
+ "forget_loss": 2.242090957338405e-08,
2768
+ "retain_loss": 0.017261430621147156,
2769
+ "step": 880
2770
+ },
2771
+ {
2772
+ "epoch": 1.0194730813287514,
2773
+ "grad_norm": 0.37109375,
2774
+ "learning_rate": 2.4683179723502306e-05,
2775
+ "loss": 0.024524253606796265,
2776
+ "step": 890
2777
+ },
2778
+ {
2779
+ "epoch": 1.0194730813287514,
2780
+ "forget_loss": 3.599013496113912e-07,
2781
+ "retain_loss": 0.034332528710365295,
2782
+ "step": 890
2783
+ },
2784
+ {
2785
+ "epoch": 1.0194730813287514,
2786
+ "forget_loss": 1.7287271703025908e-07,
2787
+ "retain_loss": 0.01914825290441513,
2788
+ "step": 890
2789
+ },
2790
+ {
2791
+ "epoch": 1.0194730813287514,
2792
+ "forget_loss": 9.782493748389243e-08,
2793
+ "retain_loss": 0.023516574874520302,
2794
+ "step": 890
2795
+ },
2796
+ {
2797
+ "epoch": 1.0194730813287514,
2798
+ "forget_loss": 1.5106013506738236e-07,
2799
+ "retain_loss": 0.019009225070476532,
2800
+ "step": 890
2801
+ },
2802
+ {
2803
+ "epoch": 1.0309278350515463,
2804
+ "grad_norm": 0.1787109375,
2805
+ "learning_rate": 2.4395161290322582e-05,
2806
+ "loss": 0.024021840095520018,
2807
+ "step": 900
2808
+ },
2809
+ {
2810
+ "epoch": 1.0309278350515463,
2811
+ "forget_loss": 2.6426999966133735e-07,
2812
+ "retain_loss": 0.029771650210022926,
2813
+ "step": 900
2814
+ },
2815
+ {
2816
+ "epoch": 1.0309278350515463,
2817
+ "forget_loss": 5.678279535459296e-08,
2818
+ "retain_loss": 0.02570934407413006,
2819
+ "step": 900
2820
+ },
2821
+ {
2822
+ "epoch": 1.0309278350515463,
2823
+ "forget_loss": 2.8987988898165895e-08,
2824
+ "retain_loss": 0.02868933044373989,
2825
+ "step": 900
2826
+ },
2827
+ {
2828
+ "epoch": 1.0309278350515463,
2829
+ "forget_loss": 1.6226587717937946e-07,
2830
+ "retain_loss": 0.024329733103513718,
2831
+ "step": 900
2832
+ },
2833
+ {
2834
+ "epoch": 1.0423825887743414,
2835
+ "grad_norm": 0.259765625,
2836
+ "learning_rate": 2.4107142857142858e-05,
2837
+ "loss": 0.02456314116716385,
2838
+ "step": 910
2839
+ },
2840
+ {
2841
+ "epoch": 1.0423825887743414,
2842
+ "forget_loss": 3.358028379807365e-07,
2843
+ "retain_loss": 0.014405038207769394,
2844
+ "step": 910
2845
+ },
2846
+ {
2847
+ "epoch": 1.0423825887743414,
2848
+ "forget_loss": 8.003183893379173e-08,
2849
+ "retain_loss": 0.012783446349203587,
2850
+ "step": 910
2851
+ },
2852
+ {
2853
+ "epoch": 1.0423825887743414,
2854
+ "forget_loss": 3.088370874593238e-07,
2855
+ "retain_loss": 0.04180136322975159,
2856
+ "step": 910
2857
+ },
2858
+ {
2859
+ "epoch": 1.0423825887743414,
2860
+ "forget_loss": 1.6710265526853618e-07,
2861
+ "retain_loss": 0.015366771258413792,
2862
+ "step": 910
2863
+ },
2864
+ {
2865
+ "epoch": 1.0538373424971363,
2866
+ "grad_norm": 0.3359375,
2867
+ "learning_rate": 2.3819124423963134e-05,
2868
+ "loss": 0.023695509135723113,
2869
+ "step": 920
2870
+ },
2871
+ {
2872
+ "epoch": 1.0538373424971363,
2873
+ "forget_loss": 1.9374034820884845e-07,
2874
+ "retain_loss": 0.0453634150326252,
2875
+ "step": 920
2876
+ },
2877
+ {
2878
+ "epoch": 1.0538373424971363,
2879
+ "forget_loss": 4.01810851258233e-08,
2880
+ "retain_loss": 0.027023399248719215,
2881
+ "step": 920
2882
+ },
2883
+ {
2884
+ "epoch": 1.0538373424971363,
2885
+ "forget_loss": 3.384187508004288e-08,
2886
+ "retain_loss": 0.011590844951570034,
2887
+ "step": 920
2888
+ },
2889
+ {
2890
+ "epoch": 1.0538373424971363,
2891
+ "forget_loss": 2.813683330771255e-08,
2892
+ "retain_loss": 0.02153298258781433,
2893
+ "step": 920
2894
+ },
2895
+ {
2896
+ "epoch": 1.0652920962199313,
2897
+ "grad_norm": 0.228515625,
2898
+ "learning_rate": 2.3531105990783413e-05,
2899
+ "loss": 0.023248623311519622,
2900
+ "step": 930
2901
+ },
2902
+ {
2903
+ "epoch": 1.0652920962199313,
2904
+ "forget_loss": 3.62256820096718e-08,
2905
+ "retain_loss": 0.04065975919365883,
2906
+ "step": 930
2907
+ },
2908
+ {
2909
+ "epoch": 1.0652920962199313,
2910
+ "forget_loss": 1.747634712501167e-07,
2911
+ "retain_loss": 0.022159142419695854,
2912
+ "step": 930
2913
+ },
2914
+ {
2915
+ "epoch": 1.0652920962199313,
2916
+ "forget_loss": 8.5983149489266e-09,
2917
+ "retain_loss": 0.011597687378525734,
2918
+ "step": 930
2919
+ },
2920
+ {
2921
+ "epoch": 1.0652920962199313,
2922
+ "forget_loss": 2.4163021805634344e-08,
2923
+ "retain_loss": 0.015060486271977425,
2924
+ "step": 930
2925
+ },
2926
+ {
2927
+ "epoch": 1.0767468499427262,
2928
+ "grad_norm": 0.216796875,
2929
+ "learning_rate": 2.324308755760369e-05,
2930
+ "loss": 0.022622266411781312,
2931
+ "step": 940
2932
+ },
2933
+ {
2934
+ "epoch": 1.0767468499427262,
2935
+ "forget_loss": 2.2849988567941182e-08,
2936
+ "retain_loss": 0.014040598645806313,
2937
+ "step": 940
2938
+ },
2939
+ {
2940
+ "epoch": 1.0767468499427262,
2941
+ "forget_loss": 1.0765910829491077e-08,
2942
+ "retain_loss": 0.021398184821009636,
2943
+ "step": 940
2944
+ },
2945
+ {
2946
+ "epoch": 1.0767468499427262,
2947
+ "forget_loss": 3.679618885144009e-07,
2948
+ "retain_loss": 0.01932062953710556,
2949
+ "step": 940
2950
+ },
2951
+ {
2952
+ "epoch": 1.0767468499427262,
2953
+ "forget_loss": 4.7389018931198734e-08,
2954
+ "retain_loss": 0.03855086490511894,
2955
+ "step": 940
2956
+ },
2957
+ {
2958
+ "epoch": 1.088201603665521,
2959
+ "grad_norm": 0.1669921875,
2960
+ "learning_rate": 2.2955069124423965e-05,
2961
+ "loss": 0.022816181182861328,
2962
+ "step": 950
2963
+ },
2964
+ {
2965
+ "epoch": 1.088201603665521,
2966
+ "forget_loss": 5.09719377816964e-08,
2967
+ "retain_loss": 0.01380919013172388,
2968
+ "step": 950
2969
+ },
2970
+ {
2971
+ "epoch": 1.088201603665521,
2972
+ "forget_loss": 2.481048966274102e-08,
2973
+ "retain_loss": 0.02061321772634983,
2974
+ "step": 950
2975
+ },
2976
+ {
2977
+ "epoch": 1.088201603665521,
2978
+ "forget_loss": 3.024410943908151e-08,
2979
+ "retain_loss": 0.021798167377710342,
2980
+ "step": 950
2981
+ },
2982
+ {
2983
+ "epoch": 1.088201603665521,
2984
+ "forget_loss": 6.814799036192198e-08,
2985
+ "retain_loss": 0.02376674860715866,
2986
+ "step": 950
2987
+ },
2988
+ {
2989
+ "epoch": 1.0996563573883162,
2990
+ "grad_norm": 0.2099609375,
2991
+ "learning_rate": 2.266705069124424e-05,
2992
+ "loss": 0.023587432503700257,
2993
+ "step": 960
2994
+ },
2995
+ {
2996
+ "epoch": 1.0996563573883162,
2997
+ "forget_loss": 7.863557982545899e-08,
2998
+ "retain_loss": 0.034126508980989456,
2999
+ "step": 960
3000
+ },
3001
+ {
3002
+ "epoch": 1.0996563573883162,
3003
+ "forget_loss": 1.2243476454898428e-08,
3004
+ "retain_loss": 0.025233590975403786,
3005
+ "step": 960
3006
+ },
3007
+ {
3008
+ "epoch": 1.0996563573883162,
3009
+ "forget_loss": 2.9529656941917892e-08,
3010
+ "retain_loss": 0.026653651148080826,
3011
+ "step": 960
3012
+ },
3013
+ {
3014
+ "epoch": 1.0996563573883162,
3015
+ "forget_loss": 1.9638031290014624e-07,
3016
+ "retain_loss": 0.021439678966999054,
3017
+ "step": 960
3018
+ },
3019
+ {
3020
+ "epoch": 1.1111111111111112,
3021
+ "grad_norm": 0.1318359375,
3022
+ "learning_rate": 2.2379032258064516e-05,
3023
+ "loss": 0.02087740898132324,
3024
+ "step": 970
3025
+ },
3026
+ {
3027
+ "epoch": 1.1111111111111112,
3028
+ "forget_loss": 5.724940876916662e-08,
3029
+ "retain_loss": 0.031759753823280334,
3030
+ "step": 970
3031
+ },
3032
+ {
3033
+ "epoch": 1.1111111111111112,
3034
+ "forget_loss": 2.719896485814388e-08,
3035
+ "retain_loss": 0.02280360274016857,
3036
+ "step": 970
3037
+ },
3038
+ {
3039
+ "epoch": 1.1111111111111112,
3040
+ "forget_loss": 1.3328020997960266e-07,
3041
+ "retain_loss": 0.034389056265354156,
3042
+ "step": 970
3043
+ },
3044
+ {
3045
+ "epoch": 1.1111111111111112,
3046
+ "forget_loss": 2.331986337367198e-08,
3047
+ "retain_loss": 0.027264058589935303,
3048
+ "step": 970
3049
+ },
3050
+ {
3051
+ "epoch": 1.122565864833906,
3052
+ "grad_norm": 0.2255859375,
3053
+ "learning_rate": 2.2091013824884792e-05,
3054
+ "loss": 0.02323000878095627,
3055
+ "step": 980
3056
+ },
3057
+ {
3058
+ "epoch": 1.122565864833906,
3059
+ "forget_loss": 6.373924179570167e-07,
3060
+ "retain_loss": 0.022899720817804337,
3061
+ "step": 980
3062
+ },
3063
+ {
3064
+ "epoch": 1.122565864833906,
3065
+ "forget_loss": 8.226674452771476e-08,
3066
+ "retain_loss": 0.017953844740986824,
3067
+ "step": 980
3068
+ },
3069
+ {
3070
+ "epoch": 1.122565864833906,
3071
+ "forget_loss": 3.995974395820667e-08,
3072
+ "retain_loss": 0.023313166573643684,
3073
+ "step": 980
3074
+ },
3075
+ {
3076
+ "epoch": 1.122565864833906,
3077
+ "forget_loss": 5.536628577829106e-08,
3078
+ "retain_loss": 0.02542729862034321,
3079
+ "step": 980
3080
+ },
3081
+ {
3082
+ "epoch": 1.134020618556701,
3083
+ "grad_norm": 0.142578125,
3084
+ "learning_rate": 2.1802995391705068e-05,
3085
+ "loss": 0.02163739502429962,
3086
+ "step": 990
3087
+ },
3088
+ {
3089
+ "epoch": 1.134020618556701,
3090
+ "forget_loss": 4.036724732259245e-08,
3091
+ "retain_loss": 0.03384210169315338,
3092
+ "step": 990
3093
+ },
3094
+ {
3095
+ "epoch": 1.134020618556701,
3096
+ "forget_loss": 1.0333982913834916e-07,
3097
+ "retain_loss": 0.021868150681257248,
3098
+ "step": 990
3099
+ },
3100
+ {
3101
+ "epoch": 1.134020618556701,
3102
+ "forget_loss": 5.9504564831058815e-08,
3103
+ "retain_loss": 0.010602345690131187,
3104
+ "step": 990
3105
+ },
3106
+ {
3107
+ "epoch": 1.134020618556701,
3108
+ "forget_loss": 4.604233794225365e-08,
3109
+ "retain_loss": 0.03545906022191048,
3110
+ "step": 990
3111
+ },
3112
+ {
3113
+ "epoch": 1.145475372279496,
3114
+ "grad_norm": 0.1796875,
3115
+ "learning_rate": 2.1514976958525347e-05,
3116
+ "loss": 0.023172636330127717,
3117
+ "step": 1000
3118
+ },
3119
+ {
3120
+ "epoch": 1.145475372279496,
3121
+ "forget_loss": 5.493423032021383e-07,
3122
+ "retain_loss": 0.032305389642715454,
3123
+ "step": 1000
3124
+ },
3125
+ {
3126
+ "epoch": 1.145475372279496,
3127
+ "forget_loss": 2.8481522917900293e-07,
3128
+ "retain_loss": 0.028948910534381866,
3129
+ "step": 1000
3130
+ },
3131
+ {
3132
+ "epoch": 1.145475372279496,
3133
+ "forget_loss": 1.4940823689357785e-08,
3134
+ "retain_loss": 0.051252782344818115,
3135
+ "step": 1000
3136
+ },
3137
+ {
3138
+ "epoch": 1.145475372279496,
3139
+ "forget_loss": 2.7136126234950098e-08,
3140
+ "retain_loss": 0.013740372844040394,
3141
+ "step": 1000
3142
+ },
3143
+ {
3144
+ "epoch": 1.1569301260022908,
3145
+ "grad_norm": 0.1748046875,
3146
+ "learning_rate": 2.1226958525345623e-05,
3147
+ "loss": 0.023918163776397706,
3148
+ "step": 1010
3149
+ },
3150
+ {
3151
+ "epoch": 1.1569301260022908,
3152
+ "forget_loss": 5.731934393793381e-08,
3153
+ "retain_loss": 0.022277191281318665,
3154
+ "step": 1010
3155
+ },
3156
+ {
3157
+ "epoch": 1.1569301260022908,
3158
+ "forget_loss": 6.457765522327463e-08,
3159
+ "retain_loss": 0.01483561284840107,
3160
+ "step": 1010
3161
+ },
3162
+ {
3163
+ "epoch": 1.1569301260022908,
3164
+ "forget_loss": 2.3290189332669797e-08,
3165
+ "retain_loss": 0.011415845714509487,
3166
+ "step": 1010
3167
+ },
3168
+ {
3169
+ "epoch": 1.1569301260022908,
3170
+ "forget_loss": 8.277501706288604e-08,
3171
+ "retain_loss": 0.0220047514885664,
3172
+ "step": 1010
3173
+ },
3174
+ {
3175
+ "epoch": 1.168384879725086,
3176
+ "grad_norm": 0.1982421875,
3177
+ "learning_rate": 2.09389400921659e-05,
3178
+ "loss": 0.024459759891033172,
3179
+ "step": 1020
3180
+ },
3181
+ {
3182
+ "epoch": 1.168384879725086,
3183
+ "forget_loss": 2.648368280233626e-08,
3184
+ "retain_loss": 0.017724836245179176,
3185
+ "step": 1020
3186
+ },
3187
+ {
3188
+ "epoch": 1.168384879725086,
3189
+ "forget_loss": 7.936857571166911e-08,
3190
+ "retain_loss": 0.016395514830946922,
3191
+ "step": 1020
3192
+ },
3193
+ {
3194
+ "epoch": 1.168384879725086,
3195
+ "forget_loss": 3.1398002420246485e-08,
3196
+ "retain_loss": 0.024634411558508873,
3197
+ "step": 1020
3198
+ },
3199
+ {
3200
+ "epoch": 1.168384879725086,
3201
+ "forget_loss": 4.1696470276519904e-08,
3202
+ "retain_loss": 0.02067175693809986,
3203
+ "step": 1020
3204
+ },
3205
+ {
3206
+ "epoch": 1.179839633447881,
3207
+ "grad_norm": 0.2109375,
3208
+ "learning_rate": 2.0650921658986175e-05,
3209
+ "loss": 0.021289190649986266,
3210
+ "step": 1030
3211
+ },
3212
+ {
3213
+ "epoch": 1.179839633447881,
3214
+ "forget_loss": 7.644550947816242e-08,
3215
+ "retain_loss": 0.018215566873550415,
3216
+ "step": 1030
3217
+ },
3218
+ {
3219
+ "epoch": 1.179839633447881,
3220
+ "forget_loss": 9.573097869974845e-09,
3221
+ "retain_loss": 0.02367250621318817,
3222
+ "step": 1030
3223
+ },
3224
+ {
3225
+ "epoch": 1.179839633447881,
3226
+ "forget_loss": 1.727073062340878e-08,
3227
+ "retain_loss": 0.029460415244102478,
3228
+ "step": 1030
3229
+ },
3230
+ {
3231
+ "epoch": 1.179839633447881,
3232
+ "forget_loss": 2.794078568513214e-08,
3233
+ "retain_loss": 0.015188030898571014,
3234
+ "step": 1030
3235
+ },
3236
+ {
3237
+ "epoch": 1.1912943871706758,
3238
+ "grad_norm": 0.1923828125,
3239
+ "learning_rate": 2.0362903225806454e-05,
3240
+ "loss": 0.021110835671424865,
3241
+ "step": 1040
3242
+ },
3243
+ {
3244
+ "epoch": 1.1912943871706758,
3245
+ "forget_loss": 2.4121279196265277e-08,
3246
+ "retain_loss": 0.017713716253638268,
3247
+ "step": 1040
3248
+ },
3249
+ {
3250
+ "epoch": 1.1912943871706758,
3251
+ "forget_loss": 9.602725015156466e-08,
3252
+ "retain_loss": 0.025713184848427773,
3253
+ "step": 1040
3254
+ },
3255
+ {
3256
+ "epoch": 1.1912943871706758,
3257
+ "forget_loss": 2.1239122816041345e-07,
3258
+ "retain_loss": 0.017294256016612053,
3259
+ "step": 1040
3260
+ },
3261
+ {
3262
+ "epoch": 1.1912943871706758,
3263
+ "forget_loss": 2.7992754780825635e-08,
3264
+ "retain_loss": 0.01869981549680233,
3265
+ "step": 1040
3266
+ },
3267
+ {
3268
+ "epoch": 1.2027491408934707,
3269
+ "grad_norm": 0.1611328125,
3270
+ "learning_rate": 2.007488479262673e-05,
3271
+ "loss": 0.02240833044052124,
3272
+ "step": 1050
3273
+ },
3274
+ {
3275
+ "epoch": 1.2027491408934707,
3276
+ "forget_loss": 5.0892207781316756e-08,
3277
+ "retain_loss": 0.02070891670882702,
3278
+ "step": 1050
3279
+ },
3280
+ {
3281
+ "epoch": 1.2027491408934707,
3282
+ "forget_loss": 2.7021894055678786e-08,
3283
+ "retain_loss": 0.016385093331336975,
3284
+ "step": 1050
3285
+ },
3286
+ {
3287
+ "epoch": 1.2027491408934707,
3288
+ "forget_loss": 1.3184821057166118e-07,
3289
+ "retain_loss": 0.012673438526690006,
3290
+ "step": 1050
3291
+ },
3292
+ {
3293
+ "epoch": 1.2027491408934707,
3294
+ "forget_loss": 2.6013424303528154e-07,
3295
+ "retain_loss": 0.048716556280851364,
3296
+ "step": 1050
3297
+ },
3298
+ {
3299
+ "epoch": 1.2142038946162657,
3300
+ "grad_norm": 0.1552734375,
3301
+ "learning_rate": 1.9786866359447005e-05,
3302
+ "loss": 0.02106960415840149,
3303
+ "step": 1060
3304
+ },
3305
+ {
3306
+ "epoch": 1.2142038946162657,
3307
+ "forget_loss": 6.823061937666353e-08,
3308
+ "retain_loss": 0.0206641536206007,
3309
+ "step": 1060
3310
+ },
3311
+ {
3312
+ "epoch": 1.2142038946162657,
3313
+ "forget_loss": 7.832964854515012e-08,
3314
+ "retain_loss": 0.017997639253735542,
3315
+ "step": 1060
3316
+ },
3317
+ {
3318
+ "epoch": 1.2142038946162657,
3319
+ "forget_loss": 1.1000237165603721e-08,
3320
+ "retain_loss": 0.026349617168307304,
3321
+ "step": 1060
3322
+ },
3323
+ {
3324
+ "epoch": 1.2142038946162657,
3325
+ "forget_loss": 2.8729951750960936e-08,
3326
+ "retain_loss": 0.03591274842619896,
3327
+ "step": 1060
3328
+ },
3329
+ {
3330
+ "epoch": 1.2256586483390608,
3331
+ "grad_norm": 0.1796875,
3332
+ "learning_rate": 1.9498847926267285e-05,
3333
+ "loss": 0.023325277864933013,
3334
+ "step": 1070
3335
+ },
3336
+ {
3337
+ "epoch": 1.2256586483390608,
3338
+ "forget_loss": 5.677220116240278e-08,
3339
+ "retain_loss": 0.017587844282388687,
3340
+ "step": 1070
3341
+ },
3342
+ {
3343
+ "epoch": 1.2256586483390608,
3344
+ "forget_loss": 1.0990573429126016e-07,
3345
+ "retain_loss": 0.014925125055015087,
3346
+ "step": 1070
3347
+ },
3348
+ {
3349
+ "epoch": 1.2256586483390608,
3350
+ "forget_loss": 6.773029781470541e-08,
3351
+ "retain_loss": 0.023775167763233185,
3352
+ "step": 1070
3353
+ },
3354
+ {
3355
+ "epoch": 1.2256586483390608,
3356
+ "forget_loss": 2.3923700354089306e-08,
3357
+ "retain_loss": 0.028247609734535217,
3358
+ "step": 1070
3359
+ },
3360
+ {
3361
+ "epoch": 1.2371134020618557,
3362
+ "grad_norm": 0.189453125,
3363
+ "learning_rate": 1.921082949308756e-05,
3364
+ "loss": 0.022065652906894682,
3365
+ "step": 1080
3366
+ },
3367
+ {
3368
+ "epoch": 1.2371134020618557,
3369
+ "forget_loss": 6.123426654625064e-09,
3370
+ "retain_loss": 0.011591405607759953,
3371
+ "step": 1080
3372
+ },
3373
+ {
3374
+ "epoch": 1.2371134020618557,
3375
+ "forget_loss": 3.527255643120952e-08,
3376
+ "retain_loss": 0.015910839661955833,
3377
+ "step": 1080
3378
+ },
3379
+ {
3380
+ "epoch": 1.2371134020618557,
3381
+ "forget_loss": 9.751467189289542e-08,
3382
+ "retain_loss": 0.007821759209036827,
3383
+ "step": 1080
3384
+ },
3385
+ {
3386
+ "epoch": 1.2371134020618557,
3387
+ "forget_loss": 2.560214170443942e-06,
3388
+ "retain_loss": 0.021734777837991714,
3389
+ "step": 1080
3390
+ },
3391
+ {
3392
+ "epoch": 1.2485681557846506,
3393
+ "grad_norm": 0.1357421875,
3394
+ "learning_rate": 1.8922811059907836e-05,
3395
+ "loss": 0.021877190470695494,
3396
+ "step": 1090
3397
+ },
3398
+ {
3399
+ "epoch": 1.2485681557846506,
3400
+ "forget_loss": 8.14992162645467e-08,
3401
+ "retain_loss": 0.024061294272542,
3402
+ "step": 1090
3403
+ },
3404
+ {
3405
+ "epoch": 1.2485681557846506,
3406
+ "forget_loss": 5.5814986410496203e-08,
3407
+ "retain_loss": 0.027457626536488533,
3408
+ "step": 1090
3409
+ },
3410
+ {
3411
+ "epoch": 1.2485681557846506,
3412
+ "forget_loss": 7.33262709218252e-07,
3413
+ "retain_loss": 0.016484102234244347,
3414
+ "step": 1090
3415
+ },
3416
+ {
3417
+ "epoch": 1.2485681557846506,
3418
+ "forget_loss": 1.206726665259339e-07,
3419
+ "retain_loss": 0.018026303499937057,
3420
+ "step": 1090
3421
+ },
3422
+ {
3423
+ "epoch": 1.2600229095074456,
3424
+ "grad_norm": 0.1630859375,
3425
+ "learning_rate": 1.8634792626728112e-05,
3426
+ "loss": 0.023145848512649538,
3427
+ "step": 1100
3428
+ },
3429
+ {
3430
+ "epoch": 1.2600229095074456,
3431
+ "forget_loss": 4.808517672927337e-08,
3432
+ "retain_loss": 0.02503185160458088,
3433
+ "step": 1100
3434
+ },
3435
+ {
3436
+ "epoch": 1.2600229095074456,
3437
+ "forget_loss": 6.350063586069155e-08,
3438
+ "retain_loss": 0.01136860717087984,
3439
+ "step": 1100
3440
+ },
3441
+ {
3442
+ "epoch": 1.2600229095074456,
3443
+ "forget_loss": 7.402125845601404e-08,
3444
+ "retain_loss": 0.023490875959396362,
3445
+ "step": 1100
3446
+ },
3447
+ {
3448
+ "epoch": 1.2600229095074456,
3449
+ "forget_loss": 1.3122317454872245e-08,
3450
+ "retain_loss": 0.009062117896974087,
3451
+ "step": 1100
3452
+ },
3453
+ {
3454
+ "epoch": 1.2714776632302405,
3455
+ "grad_norm": 0.201171875,
3456
+ "learning_rate": 1.8346774193548388e-05,
3457
+ "loss": 0.025184395909309387,
3458
+ "step": 1110
3459
+ },
3460
+ {
3461
+ "epoch": 1.2714776632302405,
3462
+ "forget_loss": 6.530651130987053e-09,
3463
+ "retain_loss": 0.010407781228423119,
3464
+ "step": 1110
3465
+ },
3466
+ {
3467
+ "epoch": 1.2714776632302405,
3468
+ "forget_loss": 9.545535561983343e-08,
3469
+ "retain_loss": 0.01965162716805935,
3470
+ "step": 1110
3471
+ },
3472
+ {
3473
+ "epoch": 1.2714776632302405,
3474
+ "forget_loss": 8.817356622614625e-09,
3475
+ "retain_loss": 0.03341147303581238,
3476
+ "step": 1110
3477
+ },
3478
+ {
3479
+ "epoch": 1.2714776632302405,
3480
+ "forget_loss": 3.525239478108233e-08,
3481
+ "retain_loss": 0.025235410779714584,
3482
+ "step": 1110
3483
+ },
3484
+ {
3485
+ "epoch": 1.2829324169530354,
3486
+ "grad_norm": 0.181640625,
3487
+ "learning_rate": 1.8058755760368664e-05,
3488
+ "loss": 0.021285700798034667,
3489
+ "step": 1120
3490
+ },
3491
+ {
3492
+ "epoch": 1.2829324169530354,
3493
+ "forget_loss": 2.0350250906631118e-07,
3494
+ "retain_loss": 0.02628685161471367,
3495
+ "step": 1120
3496
+ },
3497
+ {
3498
+ "epoch": 1.2829324169530354,
3499
+ "forget_loss": 5.7581591050848147e-08,
3500
+ "retain_loss": 0.03228295221924782,
3501
+ "step": 1120
3502
+ },
3503
+ {
3504
+ "epoch": 1.2829324169530354,
3505
+ "forget_loss": 2.3273013738389636e-08,
3506
+ "retain_loss": 0.02587534859776497,
3507
+ "step": 1120
3508
+ },
3509
+ {
3510
+ "epoch": 1.2829324169530354,
3511
+ "forget_loss": 1.0459762478376433e-07,
3512
+ "retain_loss": 0.020380979403853416,
3513
+ "step": 1120
3514
+ },
3515
+ {
3516
+ "epoch": 1.2943871706758305,
3517
+ "grad_norm": 0.203125,
3518
+ "learning_rate": 1.777073732718894e-05,
3519
+ "loss": 0.02414794862270355,
3520
+ "step": 1130
3521
+ },
3522
+ {
3523
+ "epoch": 1.2943871706758305,
3524
+ "forget_loss": 2.0749947537979097e-08,
3525
+ "retain_loss": 0.027595985680818558,
3526
+ "step": 1130
3527
+ },
3528
+ {
3529
+ "epoch": 1.2943871706758305,
3530
+ "forget_loss": 1.0538060735143517e-07,
3531
+ "retain_loss": 0.03342067450284958,
3532
+ "step": 1130
3533
+ },
3534
+ {
3535
+ "epoch": 1.2943871706758305,
3536
+ "forget_loss": 1.0734373745435732e-07,
3537
+ "retain_loss": 0.0166862141340971,
3538
+ "step": 1130
3539
+ },
3540
+ {
3541
+ "epoch": 1.2943871706758305,
3542
+ "forget_loss": 1.4255629565695926e-08,
3543
+ "retain_loss": 0.013674532994627953,
3544
+ "step": 1130
3545
+ },
3546
+ {
3547
+ "epoch": 1.3058419243986255,
3548
+ "grad_norm": 0.1845703125,
3549
+ "learning_rate": 1.748271889400922e-05,
3550
+ "loss": 0.022767890989780427,
3551
+ "step": 1140
3552
+ },
3553
+ {
3554
+ "epoch": 1.3058419243986255,
3555
+ "forget_loss": 1.1125802501510407e-08,
3556
+ "retain_loss": 0.021995795890688896,
3557
+ "step": 1140
3558
+ },
3559
+ {
3560
+ "epoch": 1.3058419243986255,
3561
+ "forget_loss": 2.5488265720241543e-08,
3562
+ "retain_loss": 0.01751900650560856,
3563
+ "step": 1140
3564
+ },
3565
+ {
3566
+ "epoch": 1.3058419243986255,
3567
+ "forget_loss": 1.843478969476564e-08,
3568
+ "retain_loss": 0.026673439890146255,
3569
+ "step": 1140
3570
+ },
3571
+ {
3572
+ "epoch": 1.3058419243986255,
3573
+ "forget_loss": 1.5914393713956088e-07,
3574
+ "retain_loss": 0.016024751588702202,
3575
+ "step": 1140
3576
+ },
3577
+ {
3578
+ "epoch": 1.3172966781214204,
3579
+ "grad_norm": 0.2158203125,
3580
+ "learning_rate": 1.7194700460829495e-05,
3581
+ "loss": 0.0235839381814003,
3582
+ "step": 1150
3583
+ },
3584
+ {
3585
+ "epoch": 1.3172966781214204,
3586
+ "forget_loss": 1.8218237585188035e-08,
3587
+ "retain_loss": 0.014317602850496769,
3588
+ "step": 1150
3589
+ },
3590
+ {
3591
+ "epoch": 1.3172966781214204,
3592
+ "forget_loss": 1.9948769747202277e-08,
3593
+ "retain_loss": 0.021403903141617775,
3594
+ "step": 1150
3595
+ },
3596
+ {
3597
+ "epoch": 1.3172966781214204,
3598
+ "forget_loss": 7.099181242864461e-09,
3599
+ "retain_loss": 0.030618898570537567,
3600
+ "step": 1150
3601
+ },
3602
+ {
3603
+ "epoch": 1.3172966781214204,
3604
+ "forget_loss": 5.6788714175581845e-08,
3605
+ "retain_loss": 0.01884060725569725,
3606
+ "step": 1150
3607
+ },
3608
+ {
3609
+ "epoch": 1.3287514318442153,
3610
+ "grad_norm": 0.203125,
3611
+ "learning_rate": 1.690668202764977e-05,
3612
+ "loss": 0.02227586954832077,
3613
+ "step": 1160
3614
+ },
3615
+ {
3616
+ "epoch": 1.3287514318442153,
3617
+ "forget_loss": 1.0105923209380308e-08,
3618
+ "retain_loss": 0.028436832129955292,
3619
+ "step": 1160
3620
+ },
3621
+ {
3622
+ "epoch": 1.3287514318442153,
3623
+ "forget_loss": 5.427875180430419e-07,
3624
+ "retain_loss": 0.011484701186418533,
3625
+ "step": 1160
3626
+ },
3627
+ {
3628
+ "epoch": 1.3287514318442153,
3629
+ "forget_loss": 6.005623021110296e-08,
3630
+ "retain_loss": 0.007445945404469967,
3631
+ "step": 1160
3632
+ },
3633
+ {
3634
+ "epoch": 1.3287514318442153,
3635
+ "forget_loss": 3.227989253673513e-08,
3636
+ "retain_loss": 0.031919676810503006,
3637
+ "step": 1160
3638
+ },
3639
+ {
3640
+ "epoch": 1.3402061855670104,
3641
+ "grad_norm": 0.197265625,
3642
+ "learning_rate": 1.6618663594470046e-05,
3643
+ "loss": 0.023313036561012267,
3644
+ "step": 1170
3645
+ },
3646
+ {
3647
+ "epoch": 1.3402061855670104,
3648
+ "forget_loss": 2.445302449416431e-08,
3649
+ "retain_loss": 0.018606863915920258,
3650
+ "step": 1170
3651
+ },
3652
+ {
3653
+ "epoch": 1.3402061855670104,
3654
+ "forget_loss": 3.023370425125904e-07,
3655
+ "retain_loss": 0.0139198899269104,
3656
+ "step": 1170
3657
+ },
3658
+ {
3659
+ "epoch": 1.3402061855670104,
3660
+ "forget_loss": 4.2964636293163494e-08,
3661
+ "retain_loss": 0.020331142470240593,
3662
+ "step": 1170
3663
+ },
3664
+ {
3665
+ "epoch": 1.3402061855670104,
3666
+ "forget_loss": 1.9158818531650468e-08,
3667
+ "retain_loss": 0.029496945440769196,
3668
+ "step": 1170
3669
+ },
3670
+ {
3671
+ "epoch": 1.3516609392898054,
3672
+ "grad_norm": 0.1708984375,
3673
+ "learning_rate": 1.6330645161290322e-05,
3674
+ "loss": 0.021104805171489716,
3675
+ "step": 1180
3676
+ },
3677
+ {
3678
+ "epoch": 1.3516609392898054,
3679
+ "forget_loss": 4.3303703733954535e-08,
3680
+ "retain_loss": 0.014237130992114544,
3681
+ "step": 1180
3682
+ },
3683
+ {
3684
+ "epoch": 1.3516609392898054,
3685
+ "forget_loss": 7.990769823607025e-09,
3686
+ "retain_loss": 0.04001491144299507,
3687
+ "step": 1180
3688
+ },
3689
+ {
3690
+ "epoch": 1.3516609392898054,
3691
+ "forget_loss": 9.303267489713107e-09,
3692
+ "retain_loss": 0.027049237862229347,
3693
+ "step": 1180
3694
+ },
3695
+ {
3696
+ "epoch": 1.3516609392898054,
3697
+ "forget_loss": 5.501159971998959e-09,
3698
+ "retain_loss": 0.005112822633236647,
3699
+ "step": 1180
3700
+ },
3701
+ {
3702
+ "epoch": 1.3631156930126003,
3703
+ "grad_norm": 0.1884765625,
3704
+ "learning_rate": 1.6042626728110598e-05,
3705
+ "loss": 0.023045459389686586,
3706
+ "step": 1190
3707
+ },
3708
+ {
3709
+ "epoch": 1.3631156930126003,
3710
+ "forget_loss": 2.7066937136055458e-08,
3711
+ "retain_loss": 0.020094262436032295,
3712
+ "step": 1190
3713
+ },
3714
+ {
3715
+ "epoch": 1.3631156930126003,
3716
+ "forget_loss": 1.7727275647416718e-08,
3717
+ "retain_loss": 0.041554100811481476,
3718
+ "step": 1190
3719
+ },
3720
+ {
3721
+ "epoch": 1.3631156930126003,
3722
+ "forget_loss": 3.883479759991815e-09,
3723
+ "retain_loss": 0.01996440999209881,
3724
+ "step": 1190
3725
+ },
3726
+ {
3727
+ "epoch": 1.3631156930126003,
3728
+ "forget_loss": 8.269919860026675e-09,
3729
+ "retain_loss": 0.025666773319244385,
3730
+ "step": 1190
3731
+ },
3732
+ {
3733
+ "epoch": 1.3745704467353952,
3734
+ "grad_norm": 0.1923828125,
3735
+ "learning_rate": 1.5754608294930877e-05,
3736
+ "loss": 0.023423218727111818,
3737
+ "step": 1200
3738
+ }
3739
+ ],
3740
+ "logging_steps": 10,
3741
+ "max_steps": 1746,
3742
+ "num_input_tokens_seen": 0,
3743
+ "num_train_epochs": 2,
3744
+ "save_steps": 200,
3745
+ "stateful_callbacks": {
3746
+ "TrainerControl": {
3747
+ "args": {
3748
+ "should_epoch_stop": false,
3749
+ "should_evaluate": false,
3750
+ "should_log": false,
3751
+ "should_save": true,
3752
+ "should_training_stop": false
3753
+ },
3754
+ "attributes": {}
3755
+ }
3756
+ },
3757
+ "total_flos": 0.0,
3758
+ "train_batch_size": 8,
3759
+ "trial_name": null,
3760
+ "trial_params": null
3761
+ }
checkpoint-1200/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70e0b39ebb6ac634552b14ecd3387db7085650b2028306aaf485cd5ea3c517f8
3
+ size 5329
checkpoint-1400/config.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "bos_token_id": 2,
7
+ "dtype": "bfloat16",
8
+ "eoi_token_index": 256000,
9
+ "eos_token_id": 1,
10
+ "image_token_index": 262144,
11
+ "initializer_range": 0.02,
12
+ "mm_tokens_per_image": 256,
13
+ "model_type": "gemma3",
14
+ "pad_token_id": 0,
15
+ "text_config": {
16
+ "_sliding_window_pattern": 6,
17
+ "attention_bias": false,
18
+ "attention_dropout": 0.0,
19
+ "attn_logit_softcapping": null,
20
+ "bos_token_id": 2,
21
+ "dtype": "bfloat16",
22
+ "eos_token_id": 1,
23
+ "final_logit_softcapping": null,
24
+ "head_dim": 256,
25
+ "hidden_activation": "gelu_pytorch_tanh",
26
+ "hidden_size": 3840,
27
+ "initializer_range": 0.02,
28
+ "intermediate_size": 15360,
29
+ "layer_types": [
30
+ "sliding_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "full_attention",
36
+ "sliding_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "sliding_attention",
40
+ "sliding_attention",
41
+ "full_attention",
42
+ "sliding_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "sliding_attention",
46
+ "sliding_attention",
47
+ "full_attention",
48
+ "sliding_attention",
49
+ "sliding_attention",
50
+ "sliding_attention",
51
+ "sliding_attention",
52
+ "sliding_attention",
53
+ "full_attention",
54
+ "sliding_attention",
55
+ "sliding_attention",
56
+ "sliding_attention",
57
+ "sliding_attention",
58
+ "sliding_attention",
59
+ "full_attention",
60
+ "sliding_attention",
61
+ "sliding_attention",
62
+ "sliding_attention",
63
+ "sliding_attention",
64
+ "sliding_attention",
65
+ "full_attention",
66
+ "sliding_attention",
67
+ "sliding_attention",
68
+ "sliding_attention",
69
+ "sliding_attention",
70
+ "sliding_attention",
71
+ "full_attention",
72
+ "sliding_attention",
73
+ "sliding_attention",
74
+ "sliding_attention",
75
+ "sliding_attention",
76
+ "sliding_attention",
77
+ "full_attention"
78
+ ],
79
+ "max_position_embeddings": 131072,
80
+ "model_type": "gemma3_text",
81
+ "num_attention_heads": 16,
82
+ "num_hidden_layers": 48,
83
+ "num_key_value_heads": 8,
84
+ "pad_token_id": 0,
85
+ "query_pre_attn_scalar": 256,
86
+ "rms_norm_eps": 1e-06,
87
+ "rope_parameters": {
88
+ "full_attention": {
89
+ "factor": 8.0,
90
+ "rope_theta": 1000000.0,
91
+ "rope_type": "linear"
92
+ },
93
+ "sliding_attention": {
94
+ "rope_theta": 10000.0,
95
+ "rope_type": "default"
96
+ }
97
+ },
98
+ "sliding_window": 1024,
99
+ "tie_word_embeddings": true,
100
+ "use_bidirectional_attention": false,
101
+ "use_cache": true,
102
+ "vocab_size": 262208
103
+ },
104
+ "tie_word_embeddings": true,
105
+ "transformers_version": "5.6.0",
106
+ "use_cache": false,
107
+ "vision_config": {
108
+ "attention_dropout": 0.0,
109
+ "dtype": "bfloat16",
110
+ "hidden_act": "gelu_pytorch_tanh",
111
+ "hidden_size": 1152,
112
+ "image_size": 896,
113
+ "intermediate_size": 4304,
114
+ "layer_norm_eps": 1e-06,
115
+ "model_type": "siglip_vision_model",
116
+ "num_attention_heads": 16,
117
+ "num_channels": 3,
118
+ "num_hidden_layers": 27,
119
+ "patch_size": 14,
120
+ "vision_use_head": false
121
+ }
122
+ }
checkpoint-1400/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "cache_implementation": "hybrid",
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 1,
7
+ 1,
8
+ 106
9
+ ],
10
+ "pad_token_id": 0,
11
+ "top_k": 64,
12
+ "top_p": 0.95,
13
+ "transformers_version": "5.6.0"
14
+ }
checkpoint-1400/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd7e32d5781a89f7b9d9d37b8c986ada2059fe6a886f800f20addea39f02592d
3
+ size 24374794824
checkpoint-1400/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d17205a590529274d5911a29c63071680d3323900071945c36dbfbd75fd463f
3
+ size 47064687765
checkpoint-1400/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9a935bbe1855a06e1b7ce88ba2ca468cff3e9a911aaf86c1456c391fcd8fe19
3
+ size 14645
checkpoint-1400/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95b7deb80ca630485caba632693bf70bc582269083adb884d9a484d32edce4f6
3
+ size 1465
checkpoint-1400/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:daab2354f8a74e70d70b4d1f804939b68a8c9624dd06cb7858e52dd8970e9726
3
+ size 33384567
checkpoint-1400/tokenizer_config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "boi_token": "<start_of_image>",
4
+ "bos_token": "<bos>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eoi_token": "<end_of_image>",
7
+ "eos_token": "<eos>",
8
+ "image_token": "<image_soft_token>",
9
+ "is_local": false,
10
+ "local_files_only": false,
11
+ "mask_token": "<mask>",
12
+ "model_max_length": 1000000000000000019884624838656,
13
+ "model_specific_special_tokens": {
14
+ "boi_token": "<start_of_image>",
15
+ "eoi_token": "<end_of_image>",
16
+ "image_token": "<image_soft_token>"
17
+ },
18
+ "pad_token": "<pad>",
19
+ "processor_class": "Gemma3Processor",
20
+ "sp_model_kwargs": null,
21
+ "spaces_between_special_tokens": false,
22
+ "tokenizer_class": "GemmaTokenizer",
23
+ "unk_token": "<unk>",
24
+ "use_default_system_prompt": false
25
+ }
checkpoint-1400/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1400/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70e0b39ebb6ac634552b14ecd3387db7085650b2028306aaf485cd5ea3c517f8
3
+ size 5329
checkpoint-1600/config.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "bos_token_id": 2,
7
+ "dtype": "bfloat16",
8
+ "eoi_token_index": 256000,
9
+ "eos_token_id": 1,
10
+ "image_token_index": 262144,
11
+ "initializer_range": 0.02,
12
+ "mm_tokens_per_image": 256,
13
+ "model_type": "gemma3",
14
+ "pad_token_id": 0,
15
+ "text_config": {
16
+ "_sliding_window_pattern": 6,
17
+ "attention_bias": false,
18
+ "attention_dropout": 0.0,
19
+ "attn_logit_softcapping": null,
20
+ "bos_token_id": 2,
21
+ "dtype": "bfloat16",
22
+ "eos_token_id": 1,
23
+ "final_logit_softcapping": null,
24
+ "head_dim": 256,
25
+ "hidden_activation": "gelu_pytorch_tanh",
26
+ "hidden_size": 3840,
27
+ "initializer_range": 0.02,
28
+ "intermediate_size": 15360,
29
+ "layer_types": [
30
+ "sliding_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "full_attention",
36
+ "sliding_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "sliding_attention",
40
+ "sliding_attention",
41
+ "full_attention",
42
+ "sliding_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "sliding_attention",
46
+ "sliding_attention",
47
+ "full_attention",
48
+ "sliding_attention",
49
+ "sliding_attention",
50
+ "sliding_attention",
51
+ "sliding_attention",
52
+ "sliding_attention",
53
+ "full_attention",
54
+ "sliding_attention",
55
+ "sliding_attention",
56
+ "sliding_attention",
57
+ "sliding_attention",
58
+ "sliding_attention",
59
+ "full_attention",
60
+ "sliding_attention",
61
+ "sliding_attention",
62
+ "sliding_attention",
63
+ "sliding_attention",
64
+ "sliding_attention",
65
+ "full_attention",
66
+ "sliding_attention",
67
+ "sliding_attention",
68
+ "sliding_attention",
69
+ "sliding_attention",
70
+ "sliding_attention",
71
+ "full_attention",
72
+ "sliding_attention",
73
+ "sliding_attention",
74
+ "sliding_attention",
75
+ "sliding_attention",
76
+ "sliding_attention",
77
+ "full_attention"
78
+ ],
79
+ "max_position_embeddings": 131072,
80
+ "model_type": "gemma3_text",
81
+ "num_attention_heads": 16,
82
+ "num_hidden_layers": 48,
83
+ "num_key_value_heads": 8,
84
+ "pad_token_id": 0,
85
+ "query_pre_attn_scalar": 256,
86
+ "rms_norm_eps": 1e-06,
87
+ "rope_parameters": {
88
+ "full_attention": {
89
+ "factor": 8.0,
90
+ "rope_theta": 1000000.0,
91
+ "rope_type": "linear"
92
+ },
93
+ "sliding_attention": {
94
+ "rope_theta": 10000.0,
95
+ "rope_type": "default"
96
+ }
97
+ },
98
+ "sliding_window": 1024,
99
+ "tie_word_embeddings": true,
100
+ "use_bidirectional_attention": false,
101
+ "use_cache": true,
102
+ "vocab_size": 262208
103
+ },
104
+ "tie_word_embeddings": true,
105
+ "transformers_version": "5.6.0",
106
+ "use_cache": false,
107
+ "vision_config": {
108
+ "attention_dropout": 0.0,
109
+ "dtype": "bfloat16",
110
+ "hidden_act": "gelu_pytorch_tanh",
111
+ "hidden_size": 1152,
112
+ "image_size": 896,
113
+ "intermediate_size": 4304,
114
+ "layer_norm_eps": 1e-06,
115
+ "model_type": "siglip_vision_model",
116
+ "num_attention_heads": 16,
117
+ "num_channels": 3,
118
+ "num_hidden_layers": 27,
119
+ "patch_size": 14,
120
+ "vision_use_head": false
121
+ }
122
+ }
checkpoint-1600/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "cache_implementation": "hybrid",
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 1,
7
+ 1,
8
+ 106
9
+ ],
10
+ "pad_token_id": 0,
11
+ "top_k": 64,
12
+ "top_p": 0.95,
13
+ "transformers_version": "5.6.0"
14
+ }
checkpoint-1600/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1d2eaa2270f79539b49c2e0cab5bb76b5c92d82aee3b9e3b8d0acc3f0cd8a5f
3
+ size 24374794824
checkpoint-1600/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33212cfd36c16ef8c58c474e42035432d5a2c6c0ab36e0e83fdf371b3e19a9be
3
+ size 47064687765
checkpoint-1600/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e85e7082a69731e9ce0f0c7155749fd102f60ffcaf11611280b715623c501f98
3
+ size 14645
checkpoint-1600/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be2855c8735e63aa5ace393feb916e8726f457c4810ce13a70cfc84c1ab5c9a8
3
+ size 1465
checkpoint-1600/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:daab2354f8a74e70d70b4d1f804939b68a8c9624dd06cb7858e52dd8970e9726
3
+ size 33384567
checkpoint-1600/tokenizer_config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "boi_token": "<start_of_image>",
4
+ "bos_token": "<bos>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eoi_token": "<end_of_image>",
7
+ "eos_token": "<eos>",
8
+ "image_token": "<image_soft_token>",
9
+ "is_local": false,
10
+ "local_files_only": false,
11
+ "mask_token": "<mask>",
12
+ "model_max_length": 1000000000000000019884624838656,
13
+ "model_specific_special_tokens": {
14
+ "boi_token": "<start_of_image>",
15
+ "eoi_token": "<end_of_image>",
16
+ "image_token": "<image_soft_token>"
17
+ },
18
+ "pad_token": "<pad>",
19
+ "processor_class": "Gemma3Processor",
20
+ "sp_model_kwargs": null,
21
+ "spaces_between_special_tokens": false,
22
+ "tokenizer_class": "GemmaTokenizer",
23
+ "unk_token": "<unk>",
24
+ "use_default_system_prompt": false
25
+ }
checkpoint-1600/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1600/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70e0b39ebb6ac634552b14ecd3387db7085650b2028306aaf485cd5ea3c517f8
3
+ size 5329
checkpoint-1746/config.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "bos_token_id": 2,
7
+ "dtype": "bfloat16",
8
+ "eoi_token_index": 256000,
9
+ "eos_token_id": 1,
10
+ "image_token_index": 262144,
11
+ "initializer_range": 0.02,
12
+ "mm_tokens_per_image": 256,
13
+ "model_type": "gemma3",
14
+ "pad_token_id": 0,
15
+ "text_config": {
16
+ "_sliding_window_pattern": 6,
17
+ "attention_bias": false,
18
+ "attention_dropout": 0.0,
19
+ "attn_logit_softcapping": null,
20
+ "bos_token_id": 2,
21
+ "dtype": "bfloat16",
22
+ "eos_token_id": 1,
23
+ "final_logit_softcapping": null,
24
+ "head_dim": 256,
25
+ "hidden_activation": "gelu_pytorch_tanh",
26
+ "hidden_size": 3840,
27
+ "initializer_range": 0.02,
28
+ "intermediate_size": 15360,
29
+ "layer_types": [
30
+ "sliding_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "full_attention",
36
+ "sliding_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "sliding_attention",
40
+ "sliding_attention",
41
+ "full_attention",
42
+ "sliding_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "sliding_attention",
46
+ "sliding_attention",
47
+ "full_attention",
48
+ "sliding_attention",
49
+ "sliding_attention",
50
+ "sliding_attention",
51
+ "sliding_attention",
52
+ "sliding_attention",
53
+ "full_attention",
54
+ "sliding_attention",
55
+ "sliding_attention",
56
+ "sliding_attention",
57
+ "sliding_attention",
58
+ "sliding_attention",
59
+ "full_attention",
60
+ "sliding_attention",
61
+ "sliding_attention",
62
+ "sliding_attention",
63
+ "sliding_attention",
64
+ "sliding_attention",
65
+ "full_attention",
66
+ "sliding_attention",
67
+ "sliding_attention",
68
+ "sliding_attention",
69
+ "sliding_attention",
70
+ "sliding_attention",
71
+ "full_attention",
72
+ "sliding_attention",
73
+ "sliding_attention",
74
+ "sliding_attention",
75
+ "sliding_attention",
76
+ "sliding_attention",
77
+ "full_attention"
78
+ ],
79
+ "max_position_embeddings": 131072,
80
+ "model_type": "gemma3_text",
81
+ "num_attention_heads": 16,
82
+ "num_hidden_layers": 48,
83
+ "num_key_value_heads": 8,
84
+ "pad_token_id": 0,
85
+ "query_pre_attn_scalar": 256,
86
+ "rms_norm_eps": 1e-06,
87
+ "rope_parameters": {
88
+ "full_attention": {
89
+ "factor": 8.0,
90
+ "rope_theta": 1000000.0,
91
+ "rope_type": "linear"
92
+ },
93
+ "sliding_attention": {
94
+ "rope_theta": 10000.0,
95
+ "rope_type": "default"
96
+ }
97
+ },
98
+ "sliding_window": 1024,
99
+ "tie_word_embeddings": true,
100
+ "use_bidirectional_attention": false,
101
+ "use_cache": true,
102
+ "vocab_size": 262208
103
+ },
104
+ "tie_word_embeddings": true,
105
+ "transformers_version": "5.6.0",
106
+ "use_cache": false,
107
+ "vision_config": {
108
+ "attention_dropout": 0.0,
109
+ "dtype": "bfloat16",
110
+ "hidden_act": "gelu_pytorch_tanh",
111
+ "hidden_size": 1152,
112
+ "image_size": 896,
113
+ "intermediate_size": 4304,
114
+ "layer_norm_eps": 1e-06,
115
+ "model_type": "siglip_vision_model",
116
+ "num_attention_heads": 16,
117
+ "num_channels": 3,
118
+ "num_hidden_layers": 27,
119
+ "patch_size": 14,
120
+ "vision_use_head": false
121
+ }
122
+ }
checkpoint-1746/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "cache_implementation": "hybrid",
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 1,
7
+ 1,
8
+ 106
9
+ ],
10
+ "pad_token_id": 0,
11
+ "top_k": 64,
12
+ "top_p": 0.95,
13
+ "transformers_version": "5.6.0"
14
+ }
checkpoint-1746/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd63605701d62f69b318f170457db759c5d56b7ac1b815d87c573bc45f0a1570
3
+ size 24374794824
checkpoint-1746/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8807960bc80f271c8b2221a5586dd712b068d4061581ea148ff09ddad85fe119
3
+ size 47064687765
checkpoint-1746/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4aa35f8a1931635954e621c61b6ae1ad9db45c41670c099fdc2d76b7964e1bb
3
+ size 14645
checkpoint-1746/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5848f807c7b26e4fc92fea8bad88632cac0a9be5082f991a909e8d8f3b6ed37c
3
+ size 1465
checkpoint-1746/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:daab2354f8a74e70d70b4d1f804939b68a8c9624dd06cb7858e52dd8970e9726
3
+ size 33384567
checkpoint-1746/tokenizer_config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "boi_token": "<start_of_image>",
4
+ "bos_token": "<bos>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eoi_token": "<end_of_image>",
7
+ "eos_token": "<eos>",
8
+ "image_token": "<image_soft_token>",
9
+ "is_local": false,
10
+ "local_files_only": false,
11
+ "mask_token": "<mask>",
12
+ "model_max_length": 1000000000000000019884624838656,
13
+ "model_specific_special_tokens": {
14
+ "boi_token": "<start_of_image>",
15
+ "eoi_token": "<end_of_image>",
16
+ "image_token": "<image_soft_token>"
17
+ },
18
+ "pad_token": "<pad>",
19
+ "processor_class": "Gemma3Processor",
20
+ "sp_model_kwargs": null,
21
+ "spaces_between_special_tokens": false,
22
+ "tokenizer_class": "GemmaTokenizer",
23
+ "unk_token": "<unk>",
24
+ "use_default_system_prompt": false
25
+ }
checkpoint-1746/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff