Hugging Face's logo

LIT-TTS
/
voice_design

Safetensors
Model card Files
xet
 Community
suwon commited on
Commit
de1b1ab
·
verified ·
1 Parent(s): 3135c5e

Delete checkpoints/v7

Browse files
Files changed (9) hide show
  1. checkpoints/v7/chat_template.jinja +0 -89
  2. checkpoints/v7/config.json +0 -134
  3. checkpoints/v7/model.safetensors +0 -3
  4. checkpoints/v7/optimizer.bin +0 -3
  5. checkpoints/v7/random_states_0.pkl +0 -3
  6. checkpoints/v7/scheduler.bin +0 -3
  7. checkpoints/v7/tokenizer.json +0 -3
  8. checkpoints/v7/tokenizer_config.json +0 -23
  9. checkpoints/v7/train_config.json +0 -186
checkpoints/v7/chat_template.jinja DELETED
@@ -1,89 +0,0 @@
1
- {%- if tools %}
2
- {{- '<|im_start|>system\n' }}
3
- {%- if messages[0].role == 'system' %}
4
- {{- messages[0].content + '\n\n' }}
5
- {%- endif %}
6
- {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
7
- {%- for tool in tools %}
8
- {{- "\n" }}
9
- {{- tool | tojson }}
10
- {%- endfor %}
11
- {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
12
- {%- else %}
13
- {%- if messages[0].role == 'system' %}
14
- {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
- {%- endif %}
16
- {%- endif %}
17
- {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
18
- {%- for message in messages[::-1] %}
19
- {%- set index = (messages|length - 1) - loop.index0 %}
20
- {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
21
- {%- set ns.multi_step_tool = false %}
22
- {%- set ns.last_query_index = index %}
23
- {%- endif %}
24
- {%- endfor %}
25
- {%- for message in messages %}
26
- {%- if message.content is string %}
27
- {%- set content = message.content %}
28
- {%- else %}
29
- {%- set content = '' %}
30
- {%- endif %}
31
- {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
32
- {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
33
- {%- elif message.role == "assistant" %}
34
- {%- set reasoning_content = '' %}
35
- {%- if message.reasoning_content is string %}
36
- {%- set reasoning_content = message.reasoning_content %}
37
- {%- else %}
38
- {%- if '</think>' in content %}
39
- {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
40
- {%- set content = content.split('</think>')[-1].lstrip('\n') %}
41
- {%- endif %}
42
- {%- endif %}
43
- {%- if loop.index0 > ns.last_query_index %}
44
- {%- if loop.last or (not loop.last and reasoning_content) %}
45
- {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
46
- {%- else %}
47
- {{- '<|im_start|>' + message.role + '\n' + content }}
48
- {%- endif %}
49
- {%- else %}
50
- {{- '<|im_start|>' + message.role + '\n' + content }}
51
- {%- endif %}
52
- {%- if message.tool_calls %}
53
- {%- for tool_call in message.tool_calls %}
54
- {%- if (loop.first and content) or (not loop.first) %}
55
- {{- '\n' }}
56
- {%- endif %}
57
- {%- if tool_call.function %}
58
- {%- set tool_call = tool_call.function %}
59
- {%- endif %}
60
- {{- '<tool_call>\n{"name": "' }}
61
- {{- tool_call.name }}
62
- {{- '", "arguments": ' }}
63
- {%- if tool_call.arguments is string %}
64
- {{- tool_call.arguments }}
65
- {%- else %}
66
- {{- tool_call.arguments | tojson }}
67
- {%- endif %}
68
- {{- '}\n</tool_call>' }}
69
- {%- endfor %}
70
- {%- endif %}
71
- {{- '<|im_end|>\n' }}
72
- {%- elif message.role == "tool" %}
73
- {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
74
- {{- '<|im_start|>user' }}
75
- {%- endif %}
76
- {{- '\n<tool_response>\n' }}
77
- {{- content }}
78
- {{- '\n</tool_response>' }}
79
- {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
80
- {{- '<|im_end|>\n' }}
81
- {%- endif %}
82
- {%- endif %}
83
- {%- endfor %}
84
- {%- if add_generation_prompt %}
85
- {{- '<|im_start|>assistant\n' }}
86
- {%- if enable_thinking is defined and enable_thinking is false %}
87
- {{- '<think>\n\n</think>\n\n' }}
88
- {%- endif %}
89
- {%- endif %}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
checkpoints/v7/config.json DELETED
@@ -1,134 +0,0 @@
1
- {
2
- "architectures": [
3
- "OmniVoice"
4
- ],
5
- "attn_implementation": "flex_attention",
6
- "audio_codebook_weights": [
7
- 8,
8
- 8,
9
- 6,
10
- 6,
11
- 4,
12
- 4,
13
- 2,
14
- 2
15
- ],
16
- "audio_mask_id": 1024,
17
- "audio_vocab_size": 1025,
18
- "bos_token_id": null,
19
- "codebook_group_condition_scale": 1.0,
20
- "codebook_group_state_mode": "expected",
21
- "codebook_group_teacher_forcing_ratio": 1.0,
22
- "codebook_groups": [
23
- [
24
- 0,
25
- 1
26
- ],
27
- [
28
- 2,
29
- 3,
30
- 4,
31
- 5
32
- ],
33
- [
34
- 6,
35
- 7
36
- ]
37
- ],
38
- "codebook_head_type": "grouped",
39
- "compile_flex_attention_mask": false,
40
- "dtype": "float32",
41
- "eos_token_id": 151645,
42
- "llm_config": {
43
- "_name_or_path": "",
44
- "architectures": [
45
- "Qwen3ForCausalLM"
46
- ],
47
- "attention_bias": false,
48
- "attention_dropout": 0.0,
49
- "bos_token_id": 151643,
50
- "chunk_size_feed_forward": 0,
51
- "dtype": "float32",
52
- "eos_token_id": 151645,
53
- "head_dim": 128,
54
- "hidden_act": "silu",
55
- "hidden_size": 1024,
56
- "id2label": {
57
- "0": "LABEL_0",
58
- "1": "LABEL_1"
59
- },
60
- "initializer_range": 0.02,
61
- "intermediate_size": 3072,
62
- "is_encoder_decoder": false,
63
- "label2id": {
64
- "LABEL_0": 0,
65
- "LABEL_1": 1
66
- },
67
- "layer_types": [
68
- "full_attention",
69
- "full_attention",
70
- "full_attention",
71
- "full_attention",
72
- "full_attention",
73
- "full_attention",
74
- "full_attention",
75
- "full_attention",
76
- "full_attention",
77
- "full_attention",
78
- "full_attention",
79
- "full_attention",
80
- "full_attention",
81
- "full_attention",
82
- "full_attention",
83
- "full_attention",
84
- "full_attention",
85
- "full_attention",
86
- "full_attention",
87
- "full_attention",
88
- "full_attention",
89
- "full_attention",
90
- "full_attention",
91
- "full_attention",
92
- "full_attention",
93
- "full_attention",
94
- "full_attention",
95
- "full_attention"
96
- ],
97
- "max_position_embeddings": 40960,
98
- "max_window_layers": 28,
99
- "model_type": "qwen3",
100
- "num_attention_heads": 16,
101
- "num_hidden_layers": 28,
102
- "num_key_value_heads": 8,
103
- "output_attentions": false,
104
- "output_hidden_states": false,
105
- "pad_token_id": null,
106
- "problem_type": null,
107
- "return_dict": true,
108
- "rms_norm_eps": 1e-06,
109
- "rope_parameters": {
110
- "rope_theta": 1000000,
111
- "rope_type": "default"
112
- },
113
- "sliding_window": null,
114
- "tie_word_embeddings": true,
115
- "use_cache": true,
116
- "use_sliding_window": false,
117
- "vocab_size": 151683
118
- },
119
- "loss_codebook_weighting_mode": "static",
120
- "loss_entropy_regularization_weight": 0.0,
121
- "loss_focal_gamma": 0.0,
122
- "loss_label_smoothing": 0.01,
123
- "loss_mask_ratio_codebook_weight_buckets": null,
124
- "model_type": "omnivoice",
125
- "num_audio_codebook": 8,
126
- "pad_token_id": 151643,
127
- "paired_ref_target_base_loss_weight": 1.0,
128
- "paired_ref_target_normalize_loss_by_active_weights": true,
129
- "paired_ref_target_self_loss": true,
130
- "paired_ref_target_self_loss_weight": 0.2,
131
- "paired_ref_use_boundary_tokens": true,
132
- "transformers_version": "5.5.4",
133
- "validate_audio_token_range": false
134
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
checkpoints/v7/model.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:e59073a6fdba4289cbe7551d2c69392cbb982e04469a1b68fdb61c1b86161dba
3
- size 3264436260
 
 
 
 
checkpoints/v7/optimizer.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:e32ca969dde2e49bc9326d89884f2c0fdfa876909a187605136e2feaaf39af48
3
- size 4917720171
 
 
 
 
checkpoints/v7/random_states_0.pkl DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:eac6e62c3c86e85ababdcf8d058fdd0fbc54b695334817035ced3b7dbf734cce
3
- size 14757
 
 
 
 
checkpoints/v7/scheduler.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d866b849e2c5e5972c9ebbab67a6523f27771103f7d18e68a2b08546ebe8c3b0
3
- size 1465
 
 
 
 
checkpoints/v7/tokenizer.json DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:4c16473266e9ae04e079c2c3ff62e899d71426108398c078c7e375b7b0e535dc
3
- size 11425357
 
 
 
 
checkpoints/v7/tokenizer_config.json DELETED
@@ -1,23 +0,0 @@
1
- {
2
- "add_prefix_space": false,
3
- "backend": "tokenizers",
4
- "bos_token": null,
5
- "clean_up_tokenization_spaces": false,
6
- "eos_token": "<|im_end|>",
7
- "errors": "replace",
8
- "extra_special_tokens": [
9
- "<|ref_text_start|>",
10
- "<|ref_text_end|>",
11
- "<|target_text_start|>",
12
- "<|target_text_end|>",
13
- "<|ref_audio_start|>",
14
- "<|ref_audio_end|>",
15
- "<|target_audio_start|>"
16
- ],
17
- "is_local": false,
18
- "model_max_length": 131072,
19
- "pad_token": "<|endoftext|>",
20
- "split_special_tokens": false,
21
- "tokenizer_class": "Qwen2Tokenizer",
22
- "unk_token": null
23
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
checkpoints/v7/train_config.json DELETED
@@ -1,186 +0,0 @@
1
- {
2
- "output_dir": "/home/infidea/suwon/omni-voice-advance-plus-pair/exp/omnivoice_0.6B-PAIR-FT-hccd",
3
- "data_config": "/home/infidea/suwon/omni-voice-advance-plus-pair/examples/config/data_config_pair_plus.json",
4
- "llm_name_or_path": "Qwen/Qwen3-0.6B",
5
- "tokenizer_name_or_path": null,
6
- "expected_llm_model_type": null,
7
- "expected_llm_hidden_size": null,
8
- "expected_llm_intermediate_size": null,
9
- "expected_llm_num_hidden_layers": null,
10
- "expected_llm_num_attention_heads": null,
11
- "expected_llm_num_key_value_heads": null,
12
- "expected_llm_vocab_size": null,
13
- "audio_vocab_size": 1025,
14
- "audio_mask_id": 1024,
15
- "num_audio_codebook": 8,
16
- "audio_codebook_weights": [
17
- 8,
18
- 8,
19
- 6,
20
- 6,
21
- 4,
22
- 4,
23
- 2,
24
- 2
25
- ],
26
- "drop_cond_ratio": 0.1,
27
- "prompt_ratio_range": [
28
- 0.05,
29
- 0.3
30
- ],
31
- "mask_ratio_range": [
32
- 0.3,
33
- 0.9
34
- ],
35
- "min_masked_audio_tokens": 8,
36
- "language_ratio": 0.0,
37
- "use_pinyin_ratio": 0.0,
38
- "instruct_ratio": 1.0,
39
- "only_instruct_ratio": 0.3,
40
- "log_codebook_losses": true,
41
- "loss_label_smoothing": 0.01,
42
- "loss_focal_gamma": 0.0,
43
- "loss_entropy_regularization_weight": 0.0,
44
- "loss_codebook_weighting_mode": "static",
45
- "loss_mask_ratio_codebook_weight_buckets": null,
46
- "use_paired_reference_training": true,
47
- "paired_ref_full_mask": true,
48
- "paired_ref_drop_cond_ratio": 0.0,
49
- "paired_ref_branch_weights": {
50
- "full": 0.55,
51
- "partial": 0.3,
52
- "span": 0.1,
53
- "ref_perturbed": 0.05,
54
- "no_ref": 0.0
55
- },
56
- "paired_ref_partial_mask_ratio_range": [
57
- 0.25,
58
- 0.75
59
- ],
60
- "paired_ref_span_mask_ratio_range": [
61
- 0.1,
62
- 0.35
63
- ],
64
- "paired_ref_perturbed_target_full_ratio": 0.5,
65
- "paired_ref_perturb_crop_ratio_range": [
66
- 0.6,
67
- 1.0
68
- ],
69
- "paired_ref_perturb_token_drop_ratio": 0.02,
70
- "paired_ref_perturb_token_replace_ratio": 0.01,
71
- "paired_ref_use_boundary_tokens": true,
72
- "paired_ref_target_self_loss": true,
73
- "paired_ref_target_base_loss_weight": 1.0,
74
- "paired_ref_target_self_loss_weight": 0.2,
75
- "paired_ref_target_normalize_loss_by_active_weights": true,
76
- "paired_ref_target_self_loss_mask_ratio_range": [
77
- 0.3,
78
- 0.9
79
- ],
80
- "paired_ref_target_self_loss_prompt_ratio_range": [
81
- 0.05,
82
- 0.25
83
- ],
84
- "paired_ref_fallback_to_prompt": true,
85
- "paired_ref_fallback_prompt_ratio_range": [
86
- 0.05,
87
- 0.3
88
- ],
89
- "paired_ref_fallback_min_ref_tokens": 1,
90
- "paired_ref_fallback_min_target_tokens": 1,
91
- "require_ref_audio_tokens": false,
92
- "codebook_head_type": "grouped",
93
- "codebook_groups": [
94
- [
95
- 0,
96
- 1
97
- ],
98
- [
99
- 2,
100
- 3,
101
- 4,
102
- 5
103
- ],
104
- [
105
- 6,
106
- 7
107
- ]
108
- ],
109
- "codebook_group_names": [
110
- "coarse",
111
- "mid",
112
- "fine"
113
- ],
114
- "codebook_group_state_mode": "expected",
115
- "codebook_group_teacher_forcing_ratio": 1.0,
116
- "codebook_group_condition_scale": 1.0,
117
- "resume_from_checkpoint": null,
118
- "init_from_checkpoint": "k2-fsa/OmniVoice",
119
- "learning_rate": 2e-05,
120
- "weight_decay": 0.01,
121
- "max_grad_norm": 1.0,
122
- "steps": 1000000,
123
- "seed": 42,
124
- "lr_scheduler_type": "cosine",
125
- "warmup_type": "steps",
126
- "warmup_ratio": 0.03,
127
- "warmup_steps": 10000,
128
- "batch_tokens": 8192,
129
- "gradient_accumulation_steps": 4,
130
- "num_workers": 2,
131
- "mixed_precision": "bf16",
132
- "allow_tf32": true,
133
- "require_cuda": true,
134
- "use_deepspeed": false,
135
- "deepspeed_config": null,
136
- "compile_flex_attention_mask": false,
137
- "validate_audio_token_range": false,
138
- "skip_bad_batches": true,
139
- "max_consecutive_batch_skips": 50,
140
- "attn_implementation": "flex_attention",
141
- "logging_steps": 100,
142
- "eval_steps": 500,
143
- "save_steps": 500,
144
- "keep_last_n_checkpoints": 3,
145
- "use_wandb": true,
146
- "wandb_project": "omnivoice_0.6B-FT-hccd",
147
- "wandb_entity": null,
148
- "wandb_run_name": null,
149
- "wandb_group": "hccd_pair",
150
- "wandb_tags": [
151
- "paired-zero-shot",
152
- "hccd",
153
- "grouped-head"
154
- ],
155
- "wandb_mode": null,
156
- "inference_logging_steps": 1000,
157
- "inference_logging_text": null,
158
- "inference_logging_language": null,
159
- "inference_logging_ref_audio": null,
160
- "inference_logging_ref_text": null,
161
- "inference_logging_num_step": 16,
162
- "inference_logging_guidance_scale": 2.0,
163
- "inference_logging_speed": 1.0,
164
- "inference_logging_duration": null,
165
- "inference_audio_tokenizer_path": null,
166
- "inference_logging_jsonl_dir": [
167
- "/home/infidea/tts-data/suwon/OmniVoice_data/zs_data/ch_data/txts",
168
- "/home/infidea/tts-data/suwon/OmniVoice_data/zs_data/en_data/txts",
169
- "/home/infidea/tts-data/suwon/OmniVoice_data/zs_data/ja_data/txts",
170
- "/home/infidea/tts-data/suwon/OmniVoice_data/zs_data/ko_data/txts",
171
- "/home/infidea/tts-data/suwon/OmniVoice_data/zs_data/pre_en_data/txts",
172
- "/home/infidea/tts-data/suwon/OmniVoice_data/zs_data/pre_ja_data/txts",
173
- "/home/infidea/tts-data/suwon/OmniVoice_data/zs_data/pre_ko_data/txts",
174
- "/home/infidea/tts-data/suwon/OmniVoice_data/zs_data/pre_vi_data/txts"
175
- ],
176
- "inference_logging_voice_design_jsonl_dir": [
177
- "/home/infidea/tts-data/suwon/OmniVoice_data/voice_design_enhance/ch/txts",
178
- "/home/infidea/tts-data/suwon/OmniVoice_data/voice_design_enhance/en/txts",
179
- "/home/infidea/tts-data/suwon/OmniVoice_data/voice_design_enhance/ja/txts",
180
- "/home/infidea/tts-data/suwon/OmniVoice_data/voice_design_enhance/ko/txts",
181
- "/home/infidea/tts-data/suwon/OmniVoice_data/voice_design_enhance/vi/txts",
182
- "/home/infidea/tts-data/suwon/OmniVoice_data/voice_design_enhance/ru/txts"
183
- ],
184
- "inference_logging_save_eval_artifacts": true,
185
- "inference_logging_eval_dir": null
186
- }