Commit
•
7782f4f
1
Parent(s):
6c77bad
Training in progress, step 1000, checkpoint
Browse files- checkpoint-1000/added_tokens.json +4 -0
- checkpoint-1000/config.json +92 -0
- checkpoint-1000/generation_config.json +9 -0
- checkpoint-1000/model.safetensors +3 -0
- checkpoint-1000/optimizer.pt +3 -0
- checkpoint-1000/preprocessor_config.json +19 -0
- checkpoint-1000/rng_state.pth +3 -0
- checkpoint-1000/scheduler.pt +3 -0
- checkpoint-1000/special_tokens_map.json +13 -0
- checkpoint-1000/spm_char.model +3 -0
- checkpoint-1000/tokenizer_config.json +63 -0
- checkpoint-1000/trainer_state.json +189 -0
- checkpoint-1000/training_args.bin +3 -0
checkpoint-1000/added_tokens.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"<ctc_blank>": 80,
|
3 |
+
"<mask>": 79
|
4 |
+
}
|
checkpoint-1000/config.json
ADDED
@@ -0,0 +1,92 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "microsoft/speecht5_tts",
|
3 |
+
"activation_dropout": 0.1,
|
4 |
+
"apply_spec_augment": true,
|
5 |
+
"architectures": [
|
6 |
+
"SpeechT5ForTextToSpeech"
|
7 |
+
],
|
8 |
+
"attention_dropout": 0.1,
|
9 |
+
"bos_token_id": 0,
|
10 |
+
"conv_bias": false,
|
11 |
+
"conv_dim": [
|
12 |
+
512,
|
13 |
+
512,
|
14 |
+
512,
|
15 |
+
512,
|
16 |
+
512,
|
17 |
+
512,
|
18 |
+
512
|
19 |
+
],
|
20 |
+
"conv_kernel": [
|
21 |
+
10,
|
22 |
+
3,
|
23 |
+
3,
|
24 |
+
3,
|
25 |
+
3,
|
26 |
+
2,
|
27 |
+
2
|
28 |
+
],
|
29 |
+
"conv_stride": [
|
30 |
+
5,
|
31 |
+
2,
|
32 |
+
2,
|
33 |
+
2,
|
34 |
+
2,
|
35 |
+
2,
|
36 |
+
2
|
37 |
+
],
|
38 |
+
"decoder_attention_heads": 12,
|
39 |
+
"decoder_ffn_dim": 3072,
|
40 |
+
"decoder_layerdrop": 0.1,
|
41 |
+
"decoder_layers": 6,
|
42 |
+
"decoder_start_token_id": 2,
|
43 |
+
"encoder_attention_heads": 12,
|
44 |
+
"encoder_ffn_dim": 3072,
|
45 |
+
"encoder_layerdrop": 0.1,
|
46 |
+
"encoder_layers": 12,
|
47 |
+
"encoder_max_relative_position": 160,
|
48 |
+
"eos_token_id": 2,
|
49 |
+
"feat_extract_activation": "gelu",
|
50 |
+
"feat_extract_norm": "group",
|
51 |
+
"feat_proj_dropout": 0.0,
|
52 |
+
"guided_attention_loss_num_heads": 2,
|
53 |
+
"guided_attention_loss_scale": 10.0,
|
54 |
+
"guided_attention_loss_sigma": 0.4,
|
55 |
+
"hidden_act": "gelu",
|
56 |
+
"hidden_dropout": 0.1,
|
57 |
+
"hidden_size": 768,
|
58 |
+
"initializer_range": 0.02,
|
59 |
+
"is_encoder_decoder": true,
|
60 |
+
"layer_norm_eps": 1e-05,
|
61 |
+
"mask_feature_length": 10,
|
62 |
+
"mask_feature_min_masks": 0,
|
63 |
+
"mask_feature_prob": 0.0,
|
64 |
+
"mask_time_length": 10,
|
65 |
+
"mask_time_min_masks": 2,
|
66 |
+
"mask_time_prob": 0.05,
|
67 |
+
"max_length": 1876,
|
68 |
+
"max_speech_positions": 1876,
|
69 |
+
"max_text_positions": 600,
|
70 |
+
"model_type": "speecht5",
|
71 |
+
"num_conv_pos_embedding_groups": 16,
|
72 |
+
"num_conv_pos_embeddings": 128,
|
73 |
+
"num_feat_extract_layers": 7,
|
74 |
+
"num_mel_bins": 80,
|
75 |
+
"pad_token_id": 1,
|
76 |
+
"positional_dropout": 0.1,
|
77 |
+
"reduction_factor": 2,
|
78 |
+
"scale_embedding": false,
|
79 |
+
"speaker_embedding_dim": 512,
|
80 |
+
"speech_decoder_postnet_dropout": 0.5,
|
81 |
+
"speech_decoder_postnet_kernel": 5,
|
82 |
+
"speech_decoder_postnet_layers": 5,
|
83 |
+
"speech_decoder_postnet_units": 256,
|
84 |
+
"speech_decoder_prenet_dropout": 0.5,
|
85 |
+
"speech_decoder_prenet_layers": 2,
|
86 |
+
"speech_decoder_prenet_units": 256,
|
87 |
+
"torch_dtype": "float32",
|
88 |
+
"transformers_version": "4.41.0.dev0",
|
89 |
+
"use_cache": false,
|
90 |
+
"use_guided_attention_loss": true,
|
91 |
+
"vocab_size": 81
|
92 |
+
}
|
checkpoint-1000/generation_config.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_from_model_config": true,
|
3 |
+
"bos_token_id": 0,
|
4 |
+
"decoder_start_token_id": 2,
|
5 |
+
"eos_token_id": 2,
|
6 |
+
"max_length": 1876,
|
7 |
+
"pad_token_id": 1,
|
8 |
+
"transformers_version": "4.41.0.dev0"
|
9 |
+
}
|
checkpoint-1000/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fe64b04c5eaa2e5e32541e99ae462567b245fc0912df1c1b6dff2197c5029551
|
3 |
+
size 577789320
|
checkpoint-1000/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7a79dd6377c1487b587e581814368a709694e321da22d2d5f7e4f96b7f56a2b6
|
3 |
+
size 1155772233
|
checkpoint-1000/preprocessor_config.json
ADDED
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"do_normalize": false,
|
3 |
+
"feature_extractor_type": "SpeechT5FeatureExtractor",
|
4 |
+
"feature_size": 1,
|
5 |
+
"fmax": 7600,
|
6 |
+
"fmin": 80,
|
7 |
+
"frame_signal_scale": 1.0,
|
8 |
+
"hop_length": 16,
|
9 |
+
"mel_floor": 1e-10,
|
10 |
+
"num_mel_bins": 80,
|
11 |
+
"padding_side": "right",
|
12 |
+
"padding_value": 0.0,
|
13 |
+
"processor_class": "SpeechT5Processor",
|
14 |
+
"reduction_factor": 2,
|
15 |
+
"return_attention_mask": true,
|
16 |
+
"sampling_rate": 16000,
|
17 |
+
"win_function": "hann_window",
|
18 |
+
"win_length": 64
|
19 |
+
}
|
checkpoint-1000/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:40175b15300fc6815a708d967ff317cbd2e764a0ad982f7e73e9abe0e80eee4b
|
3 |
+
size 14244
|
checkpoint-1000/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1e91a0701321b0c1ce0908635dba5b9e0faa643c500a1c008fbe16e9fe1fb43e
|
3 |
+
size 1064
|
checkpoint-1000/special_tokens_map.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"eos_token": "</s>",
|
4 |
+
"mask_token": {
|
5 |
+
"content": "<mask>",
|
6 |
+
"lstrip": true,
|
7 |
+
"normalized": true,
|
8 |
+
"rstrip": false,
|
9 |
+
"single_word": false
|
10 |
+
},
|
11 |
+
"pad_token": "<pad>",
|
12 |
+
"unk_token": "<unk>"
|
13 |
+
}
|
checkpoint-1000/spm_char.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7fcc48f3e225f627b1641db410ceb0c8649bd2b0c982e150b03f8be3728ab560
|
3 |
+
size 238473
|
checkpoint-1000/tokenizer_config.json
ADDED
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "<s>",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "<pad>",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "</s>",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "<unk>",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"79": {
|
36 |
+
"content": "<mask>",
|
37 |
+
"lstrip": true,
|
38 |
+
"normalized": true,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
},
|
43 |
+
"80": {
|
44 |
+
"content": "<ctc_blank>",
|
45 |
+
"lstrip": false,
|
46 |
+
"normalized": true,
|
47 |
+
"rstrip": false,
|
48 |
+
"single_word": false,
|
49 |
+
"special": false
|
50 |
+
}
|
51 |
+
},
|
52 |
+
"bos_token": "<s>",
|
53 |
+
"clean_up_tokenization_spaces": true,
|
54 |
+
"eos_token": "</s>",
|
55 |
+
"mask_token": "<mask>",
|
56 |
+
"model_max_length": 600,
|
57 |
+
"normalize": false,
|
58 |
+
"pad_token": "<pad>",
|
59 |
+
"processor_class": "SpeechT5Processor",
|
60 |
+
"sp_model_kwargs": {},
|
61 |
+
"tokenizer_class": "SpeechT5Tokenizer",
|
62 |
+
"unk_token": "<unk>"
|
63 |
+
}
|
checkpoint-1000/trainer_state.json
ADDED
@@ -0,0 +1,189 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.40811634063720703,
|
3 |
+
"best_model_checkpoint": "mikhail_panzo/fil_b128_le4_s4000/checkpoint-1000",
|
4 |
+
"epoch": 88.88888888888889,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 4.444444444444445,
|
13 |
+
"grad_norm": 2.152308940887451,
|
14 |
+
"learning_rate": 2.5e-06,
|
15 |
+
"loss": 0.7738,
|
16 |
+
"step": 50
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 8.88888888888889,
|
20 |
+
"grad_norm": 1.326554536819458,
|
21 |
+
"learning_rate": 5e-06,
|
22 |
+
"loss": 0.6968,
|
23 |
+
"step": 100
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 13.333333333333334,
|
27 |
+
"grad_norm": 1.4387542009353638,
|
28 |
+
"learning_rate": 7.5e-06,
|
29 |
+
"loss": 0.636,
|
30 |
+
"step": 150
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 17.77777777777778,
|
34 |
+
"grad_norm": 1.033190131187439,
|
35 |
+
"learning_rate": 1e-05,
|
36 |
+
"loss": 0.5486,
|
37 |
+
"step": 200
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 22.22222222222222,
|
41 |
+
"grad_norm": 1.4383914470672607,
|
42 |
+
"learning_rate": 1.25e-05,
|
43 |
+
"loss": 0.516,
|
44 |
+
"step": 250
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 26.666666666666668,
|
48 |
+
"grad_norm": 1.8267806768417358,
|
49 |
+
"learning_rate": 1.5e-05,
|
50 |
+
"loss": 0.4989,
|
51 |
+
"step": 300
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 31.11111111111111,
|
55 |
+
"grad_norm": 1.2525017261505127,
|
56 |
+
"learning_rate": 1.75e-05,
|
57 |
+
"loss": 0.4908,
|
58 |
+
"step": 350
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 35.55555555555556,
|
62 |
+
"grad_norm": 1.1662768125534058,
|
63 |
+
"learning_rate": 2e-05,
|
64 |
+
"loss": 0.4766,
|
65 |
+
"step": 400
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 40.0,
|
69 |
+
"grad_norm": 1.110721230506897,
|
70 |
+
"learning_rate": 2.25e-05,
|
71 |
+
"loss": 0.4716,
|
72 |
+
"step": 450
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 44.44444444444444,
|
76 |
+
"grad_norm": 1.6693321466445923,
|
77 |
+
"learning_rate": 2.5e-05,
|
78 |
+
"loss": 0.4635,
|
79 |
+
"step": 500
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 44.44444444444444,
|
83 |
+
"eval_loss": 0.42072969675064087,
|
84 |
+
"eval_runtime": 9.5188,
|
85 |
+
"eval_samples_per_second": 16.704,
|
86 |
+
"eval_steps_per_second": 2.101,
|
87 |
+
"step": 500
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 48.888888888888886,
|
91 |
+
"grad_norm": 1.0238850116729736,
|
92 |
+
"learning_rate": 2.7500000000000004e-05,
|
93 |
+
"loss": 0.4585,
|
94 |
+
"step": 550
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 53.333333333333336,
|
98 |
+
"grad_norm": 0.8729687929153442,
|
99 |
+
"learning_rate": 3e-05,
|
100 |
+
"loss": 0.4572,
|
101 |
+
"step": 600
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 57.77777777777778,
|
105 |
+
"grad_norm": 1.4291303157806396,
|
106 |
+
"learning_rate": 3.2500000000000004e-05,
|
107 |
+
"loss": 0.452,
|
108 |
+
"step": 650
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 62.22222222222222,
|
112 |
+
"grad_norm": 1.3374826908111572,
|
113 |
+
"learning_rate": 3.5e-05,
|
114 |
+
"loss": 0.447,
|
115 |
+
"step": 700
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 66.66666666666667,
|
119 |
+
"grad_norm": 1.7066700458526611,
|
120 |
+
"learning_rate": 3.7500000000000003e-05,
|
121 |
+
"loss": 0.4424,
|
122 |
+
"step": 750
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 71.11111111111111,
|
126 |
+
"grad_norm": 2.1874194145202637,
|
127 |
+
"learning_rate": 4e-05,
|
128 |
+
"loss": 0.4443,
|
129 |
+
"step": 800
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 75.55555555555556,
|
133 |
+
"grad_norm": 1.4660208225250244,
|
134 |
+
"learning_rate": 4.25e-05,
|
135 |
+
"loss": 0.4346,
|
136 |
+
"step": 850
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 80.0,
|
140 |
+
"grad_norm": 1.2421706914901733,
|
141 |
+
"learning_rate": 4.5e-05,
|
142 |
+
"loss": 0.4328,
|
143 |
+
"step": 900
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 84.44444444444444,
|
147 |
+
"grad_norm": 0.8712289929389954,
|
148 |
+
"learning_rate": 4.75e-05,
|
149 |
+
"loss": 0.4335,
|
150 |
+
"step": 950
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 88.88888888888889,
|
154 |
+
"grad_norm": 1.161516547203064,
|
155 |
+
"learning_rate": 5e-05,
|
156 |
+
"loss": 0.4317,
|
157 |
+
"step": 1000
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 88.88888888888889,
|
161 |
+
"eval_loss": 0.40811634063720703,
|
162 |
+
"eval_runtime": 9.5158,
|
163 |
+
"eval_samples_per_second": 16.709,
|
164 |
+
"eval_steps_per_second": 2.102,
|
165 |
+
"step": 1000
|
166 |
+
}
|
167 |
+
],
|
168 |
+
"logging_steps": 50,
|
169 |
+
"max_steps": 4000,
|
170 |
+
"num_input_tokens_seen": 0,
|
171 |
+
"num_train_epochs": 364,
|
172 |
+
"save_steps": 500,
|
173 |
+
"stateful_callbacks": {
|
174 |
+
"TrainerControl": {
|
175 |
+
"args": {
|
176 |
+
"should_epoch_stop": false,
|
177 |
+
"should_evaluate": false,
|
178 |
+
"should_log": false,
|
179 |
+
"should_save": true,
|
180 |
+
"should_training_stop": false
|
181 |
+
},
|
182 |
+
"attributes": {}
|
183 |
+
}
|
184 |
+
},
|
185 |
+
"total_flos": 2.748976043947373e+16,
|
186 |
+
"train_batch_size": 16,
|
187 |
+
"trial_name": null,
|
188 |
+
"trial_params": null
|
189 |
+
}
|
checkpoint-1000/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ebe21f66ae29351cf4a52c96ecd128baf630207d21c90cee80fdb06c7fcd7437
|
3 |
+
size 5304
|