mcuo commited on
Commit
ade06db
·
verified ·
1 Parent(s): 0502227

Upload 5 files

Browse files
adapter_config.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alora_invocation_tokens": null,
3
+ "alpha_pattern": {},
4
+ "arrow_config": null,
5
+ "auto_mapping": {
6
+ "base_model_class": "TextToLatentRFDiT",
7
+ "parent_library": "irodori_tts.model"
8
+ },
9
+ "base_model_name_or_path": null,
10
+ "bias": "none",
11
+ "corda_config": null,
12
+ "ensure_weight_tying": false,
13
+ "eva_config": null,
14
+ "exclude_modules": null,
15
+ "fan_in_fan_out": false,
16
+ "inference_mode": true,
17
+ "init_lora_weights": true,
18
+ "layer_replication": null,
19
+ "layers_pattern": null,
20
+ "layers_to_transform": null,
21
+ "loftq_config": {},
22
+ "lora_alpha": 32.0,
23
+ "lora_bias": false,
24
+ "lora_dropout": 0.05,
25
+ "lora_ga_config": null,
26
+ "megatron_config": null,
27
+ "megatron_core": "megatron.core",
28
+ "modules_to_save": null,
29
+ "peft_type": "LORA",
30
+ "peft_version": "0.19.1",
31
+ "qalora_group_size": 16,
32
+ "r": 16,
33
+ "rank_pattern": {},
34
+ "revision": null,
35
+ "target_modules": [
36
+ "wk",
37
+ "wo",
38
+ "wv",
39
+ "wq"
40
+ ],
41
+ "target_parameters": null,
42
+ "task_type": null,
43
+ "trainable_token_indices": null,
44
+ "use_bdlora": null,
45
+ "use_dora": false,
46
+ "use_qalora": false,
47
+ "use_rslora": false
48
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fc10166d27b4a7b1defee3c77664ced749196d72062eaec88cfc583840b10f1
3
+ size 6847184
ema_shadow.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d5d9103dc03853300da50d10c2782cfe68b8350832935ef40230859f1c4c73f
3
+ size 13719297
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93ca2ca9f6962dd73a72da87411292a803de2f6c32724e605b4422d0f40a7081
3
+ size 13836619
train_state.json ADDED
@@ -0,0 +1,102 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "step": 3000,
3
+ "base_model_path": "E:\\Emoji-TTS\\checkpoints\\Aratako_Irodori-TTS-500M-v2\\model.safetensors",
4
+ "base_model_sha256": "e5add885303babe328eae2c426475e41bc2f2aca2bcb010a2ceb6b6ddd1b8d9c",
5
+ "base_model_config": {
6
+ "latent_dim": 32,
7
+ "latent_patch_size": 1,
8
+ "model_dim": 1280,
9
+ "num_layers": 12,
10
+ "num_heads": 20,
11
+ "mlp_ratio": 2.875,
12
+ "text_mlp_ratio": 2.6,
13
+ "speaker_mlp_ratio": 2.6,
14
+ "dropout": 0.0,
15
+ "text_vocab_size": 99574,
16
+ "text_tokenizer_repo": "llm-jp/llm-jp-3-150m",
17
+ "text_add_bos": true,
18
+ "text_dim": 512,
19
+ "text_layers": 10,
20
+ "text_heads": 8,
21
+ "speaker_dim": 768,
22
+ "speaker_layers": 8,
23
+ "speaker_heads": 12,
24
+ "speaker_patch_size": 1,
25
+ "timestep_embed_dim": 512,
26
+ "adaln_rank": 192,
27
+ "norm_eps": 1e-05
28
+ },
29
+ "lora_config": {
30
+ "r": 16,
31
+ "lora_alpha": 32.0,
32
+ "lora_dropout": 0.05,
33
+ "target_modules": [
34
+ "wq",
35
+ "wk",
36
+ "wv",
37
+ "wo"
38
+ ]
39
+ },
40
+ "train_config": {
41
+ "manifest_path": "E:\\Emoji-TTS\\data\\train_manifest.jsonl",
42
+ "output_dir": "E:\\Emoji-TTS\\lora\\my_lora",
43
+ "batch_size": 4,
44
+ "num_workers": 2,
45
+ "dataloader_persistent_workers": false,
46
+ "dataloader_prefetch_factor": 2,
47
+ "allow_tf32": false,
48
+ "compile_model": false,
49
+ "learning_rate": 0.0001,
50
+ "weight_decay": 0.01,
51
+ "optimizer": "adamw",
52
+ "adam_beta1": 0.9,
53
+ "adam_beta2": 0.999,
54
+ "adam_eps": 1e-08,
55
+ "muon_momentum": 0.95,
56
+ "muon_adjust_lr_fn": "match_rms_adamw",
57
+ "lr_scheduler": "none",
58
+ "warmup_steps": 0,
59
+ "caption_warmup": false,
60
+ "caption_warmup_steps": 0,
61
+ "stable_steps": 0,
62
+ "min_lr_scale": 0.1,
63
+ "max_steps": 3000,
64
+ "log_every": 10,
65
+ "save_every": 500,
66
+ "checkpoint_best_n": 0,
67
+ "valid_ratio": 0.1,
68
+ "valid_every": 100,
69
+ "progress": true,
70
+ "progress_all_ranks": false,
71
+ "precision": "bf16",
72
+ "grad_clip_norm": 1.0,
73
+ "gradient_accumulation_steps": 1,
74
+ "max_text_len": 256,
75
+ "max_caption_len": null,
76
+ "text_condition_dropout": 0.1,
77
+ "caption_condition_dropout": 0.1,
78
+ "speaker_condition_dropout": 0.1,
79
+ "max_latent_steps": 750,
80
+ "fixed_target_latent_steps": null,
81
+ "fixed_target_full_mask": false,
82
+ "timestep_logit_mean": 0.0,
83
+ "timestep_logit_std": 1.0,
84
+ "timestep_stratified": false,
85
+ "timestep_min": 0.001,
86
+ "timestep_max": 0.999,
87
+ "wandb_enabled": false,
88
+ "wandb_project": null,
89
+ "wandb_entity": null,
90
+ "wandb_run_name": null,
91
+ "wandb_mode": "online",
92
+ "ddp_find_unused_parameters": false,
93
+ "lora_enabled": false,
94
+ "lora_r": 16,
95
+ "lora_alpha": 32,
96
+ "lora_dropout": 0.0,
97
+ "lora_bias": "none",
98
+ "lora_target_modules": "diffusion_attn",
99
+ "seed": 0
100
+ },
101
+ "ema_decay": 0.9999
102
+ }