tamarher commited on
Commit
68f976d
·
verified ·
1 Parent(s): 9993540

Add files using upload-large-folder tool

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 +2 -0
  2. mf/mlx-base/config.json +120 -0
  3. mf/mlx-base/core.safetensors +3 -0
  4. mf/mlx-base/latent_stats.safetensors +3 -0
  5. mf/mlx-base/llm_config.json +114 -0
  6. mf/mlx-base/mlx_config.json +33 -0
  7. mf/mlx-base/speaker.safetensors +3 -0
  8. mf/mlx-base/tokenizer/added_tokens.json +31 -0
  9. mf/mlx-base/tokenizer/chat_template.jinja +54 -0
  10. mf/mlx-base/tokenizer/merges.txt +0 -0
  11. mf/mlx-base/tokenizer/special_tokens_map.json +67 -0
  12. mf/mlx-base/tokenizer/tokenizer_config.json +257 -0
  13. mf/mlx-base/tokenizer/vocab.json +0 -0
  14. mf/mlx-base/vocoder.safetensors +3 -0
  15. mf/mlx-int8/config.json +120 -0
  16. mf/mlx-int8/llm_config.json +114 -0
  17. mf/mlx-int8/mlx_config.json +243 -0
  18. mf/mlx-int8/tokenizer/added_tokens.json +31 -0
  19. mf/mlx-int8/tokenizer/chat_template.jinja +54 -0
  20. mf/mlx-int8/tokenizer/merges.txt +0 -0
  21. mf/mlx-int8/tokenizer/special_tokens_map.json +67 -0
  22. mf/mlx-int8/tokenizer/tokenizer_config.json +257 -0
  23. mf/mlx-int8/tokenizer/vocab.json +0 -0
  24. soar/mlx-base/config.json +113 -0
  25. soar/mlx-base/core.safetensors +3 -0
  26. soar/mlx-base/latent_stats.safetensors +3 -0
  27. soar/mlx-base/llm_config.json +114 -0
  28. soar/mlx-base/mlx_config.json +33 -0
  29. soar/mlx-base/speaker.safetensors +3 -0
  30. soar/mlx-base/tokenizer/added_tokens.json +31 -0
  31. soar/mlx-base/tokenizer/chat_template.jinja +54 -0
  32. soar/mlx-base/tokenizer/merges.txt +0 -0
  33. soar/mlx-base/tokenizer/special_tokens_map.json +67 -0
  34. soar/mlx-base/tokenizer/tokenizer.json +3 -0
  35. soar/mlx-base/tokenizer/tokenizer_config.json +257 -0
  36. soar/mlx-base/tokenizer/vocab.json +0 -0
  37. soar/mlx-base/vocoder.safetensors +3 -0
  38. soar/mlx-int8/config.json +113 -0
  39. soar/mlx-int8/core.safetensors +3 -0
  40. soar/mlx-int8/latent_stats.safetensors +3 -0
  41. soar/mlx-int8/llm_config.json +114 -0
  42. soar/mlx-int8/mlx_config.json +243 -0
  43. soar/mlx-int8/speaker.safetensors +3 -0
  44. soar/mlx-int8/tokenizer/added_tokens.json +31 -0
  45. soar/mlx-int8/tokenizer/chat_template.jinja +54 -0
  46. soar/mlx-int8/tokenizer/merges.txt +0 -0
  47. soar/mlx-int8/tokenizer/special_tokens_map.json +67 -0
  48. soar/mlx-int8/tokenizer/tokenizer.json +3 -0
  49. soar/mlx-int8/tokenizer/tokenizer_config.json +257 -0
  50. soar/mlx-int8/tokenizer/vocab.json +0 -0
.gitattributes CHANGED
@@ -33,3 +33,5 @@ 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
+ soar/mlx-int8/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
+ soar/mlx-base/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
mf/mlx-base/config.json ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "dots_tts",
3
+ "latent_dim": 128,
4
+ "patch_size": 4,
5
+ "cfg_droprate": 0.2,
6
+ "LLM": {
7
+ "model_name_or_path": "/prodcpfs/user/jiatu/logs/qwen25_15b"
8
+ },
9
+ "PatchEncoder": {
10
+ "num_layers": 24,
11
+ "num_heads": 16,
12
+ "hidden_size": 1024,
13
+ "ffn_hidden_size": 4096,
14
+ "modulation": false,
15
+ "qkv_bias": false,
16
+ "qk_norm": true,
17
+ "attn_dropout": 0.0,
18
+ "dropout": 0.1,
19
+ "norm_layer": "RMSNorm",
20
+ "alibi_bias": false,
21
+ "rotary_bias": true,
22
+ "rotary_theta": 10000.0,
23
+ "input_dim": 128,
24
+ "causal": true
25
+ },
26
+ "DiT": {
27
+ "num_layers": 18,
28
+ "num_heads": 16,
29
+ "hidden_size": 1024,
30
+ "ffn_hidden_size": 4096,
31
+ "modulation": true,
32
+ "qkv_bias": false,
33
+ "qk_norm": true,
34
+ "attn_dropout": 0.0,
35
+ "dropout": 0.0,
36
+ "norm_layer": "RMSNorm",
37
+ "alibi_bias": false,
38
+ "rotary_bias": true,
39
+ "rotary_theta": 10000.0
40
+ },
41
+ "vocoder": {
42
+ "sample_rate": 48000,
43
+ "upsample_rates": [
44
+ 10,
45
+ 6,
46
+ 4,
47
+ 2,
48
+ 2,
49
+ 2
50
+ ],
51
+ "upsample_kernel_sizes": [
52
+ 20,
53
+ 12,
54
+ 8,
55
+ 4,
56
+ 4,
57
+ 4
58
+ ],
59
+ "upsample_initial_channel": 1536,
60
+ "resblock": "1",
61
+ "resblock_kernel_sizes": [
62
+ 3,
63
+ 7,
64
+ 11
65
+ ],
66
+ "resblock_dilation_sizes": [
67
+ [
68
+ 1,
69
+ 3,
70
+ 5
71
+ ],
72
+ [
73
+ 1,
74
+ 3,
75
+ 5
76
+ ],
77
+ [
78
+ 1,
79
+ 3,
80
+ 5
81
+ ]
82
+ ],
83
+ "downsample_rates": [
84
+ 2,
85
+ 2,
86
+ 2,
87
+ 4,
88
+ 6,
89
+ 10
90
+ ],
91
+ "downsample_channels": [
92
+ 12,
93
+ 24,
94
+ 48,
95
+ 96,
96
+ 192,
97
+ 384,
98
+ 768
99
+ ],
100
+ "activation": "snakebeta",
101
+ "snake_logscale": true,
102
+ "latent_dim": 128,
103
+ "causal": true,
104
+ "mi_num_layers": 4,
105
+ "causal_encoder": true,
106
+ "use_bias_at_final": false,
107
+ "use_tanh_at_final": false
108
+ },
109
+ "fm_sigma": 0.0,
110
+ "xvec_drop_rate": 0.2,
111
+ "campplus_embedding_size": 512,
112
+ "xvec_max_audio_seconds": 10.0,
113
+ "meanflow": {
114
+ "enabled": true,
115
+ "use_duration_embedding": true
116
+ },
117
+ "architectures": [
118
+ "DotsTTSForConditionalGeneration"
119
+ ]
120
+ }
mf/mlx-base/core.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bbfca574cc825d7256783f20a4c1ea1c483f49c4ca6b163468a1ea66429d217e
3
+ size 4398911066
mf/mlx-base/latent_stats.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:370ea7fe3e54c06336ff050eaf14f9554417f39b002b2dad5228f67c86c7260f
3
+ size 1160
mf/mlx-base/llm_config.json ADDED
@@ -0,0 +1,114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "vocab_size": 151672,
3
+ "max_position_embeddings": 131072,
4
+ "hidden_size": 1536,
5
+ "intermediate_size": 8960,
6
+ "num_hidden_layers": 28,
7
+ "num_attention_heads": 12,
8
+ "use_sliding_window": false,
9
+ "sliding_window": null,
10
+ "max_window_layers": 28,
11
+ "num_key_value_heads": 2,
12
+ "hidden_act": "silu",
13
+ "initializer_range": 0.02,
14
+ "rms_norm_eps": 1e-06,
15
+ "use_cache": true,
16
+ "rope_theta": 1000000.0,
17
+ "rope_scaling": null,
18
+ "attention_dropout": 0.0,
19
+ "layer_types": [
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention"
48
+ ],
49
+ "return_dict": true,
50
+ "output_hidden_states": false,
51
+ "torchscript": false,
52
+ "dtype": "bfloat16",
53
+ "pruned_heads": {},
54
+ "tie_word_embeddings": true,
55
+ "chunk_size_feed_forward": 0,
56
+ "is_encoder_decoder": false,
57
+ "is_decoder": false,
58
+ "cross_attention_hidden_size": null,
59
+ "add_cross_attention": false,
60
+ "tie_encoder_decoder": false,
61
+ "architectures": [
62
+ "Qwen2ForCausalLM"
63
+ ],
64
+ "finetuning_task": null,
65
+ "id2label": {
66
+ "0": "LABEL_0",
67
+ "1": "LABEL_1"
68
+ },
69
+ "label2id": {
70
+ "LABEL_0": 0,
71
+ "LABEL_1": 1
72
+ },
73
+ "task_specific_params": null,
74
+ "problem_type": null,
75
+ "tokenizer_class": null,
76
+ "prefix": null,
77
+ "bos_token_id": 151643,
78
+ "pad_token_id": null,
79
+ "eos_token_id": 151643,
80
+ "sep_token_id": null,
81
+ "decoder_start_token_id": null,
82
+ "max_length": 20,
83
+ "min_length": 0,
84
+ "do_sample": false,
85
+ "early_stopping": false,
86
+ "num_beams": 1,
87
+ "temperature": 1.0,
88
+ "top_k": 50,
89
+ "top_p": 1.0,
90
+ "typical_p": 1.0,
91
+ "repetition_penalty": 1.0,
92
+ "length_penalty": 1.0,
93
+ "no_repeat_ngram_size": 0,
94
+ "encoder_no_repeat_ngram_size": 0,
95
+ "bad_words_ids": null,
96
+ "num_return_sequences": 1,
97
+ "output_scores": false,
98
+ "return_dict_in_generate": false,
99
+ "forced_bos_token_id": null,
100
+ "forced_eos_token_id": null,
101
+ "remove_invalid_values": false,
102
+ "exponential_decay_length_penalty": null,
103
+ "suppress_tokens": null,
104
+ "begin_suppress_tokens": null,
105
+ "num_beam_groups": 1,
106
+ "diversity_penalty": 0.0,
107
+ "_name_or_path": "",
108
+ "transformers_version": "4.57.0",
109
+ "model_type": "qwen2",
110
+ "use_mrope": false,
111
+ "tf_legacy_loss": false,
112
+ "use_bfloat16": false,
113
+ "output_attentions": false
114
+ }
mf/mlx-base/mlx_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "artifact_class": "base",
3
+ "dtype_policy": {
4
+ "core": {
5
+ "*": "bfloat16"
6
+ },
7
+ "latent_stats": {
8
+ "*": "float32"
9
+ },
10
+ "speaker": {
11
+ "*": "float32"
12
+ },
13
+ "vocoder": {
14
+ "audio_encoder.*": "float32",
15
+ "dec_mi_layer.*": "bfloat16",
16
+ "decoder.*": "bfloat16",
17
+ "enc_mi_layer.*": "float32",
18
+ "post_proj.*": "bfloat16",
19
+ "pre_proj.*": "float32"
20
+ }
21
+ },
22
+ "mode": "meanflow",
23
+ "model_family": "dots_tts",
24
+ "quantization": null,
25
+ "schema_version": 1,
26
+ "source": {
27
+ "manifest_sha256": "c5b8638c5083a06b052a2472685ae5933a8f6b2bcbe96cc00cd11624bc12a6af",
28
+ "repo_id": "rednote-hilab/dots.tts-mf",
29
+ "resolved_repo_id": "dots-studio/dots.tts-mf",
30
+ "revision": "25c53fb462e57087e52237daa5ea30df1c5cc328"
31
+ },
32
+ "variant": "mf"
33
+ }
mf/mlx-base/speaker.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:105356b29dfe40c986412db337c252ab2d227703f433de64b525078de6ca5858
3
+ size 29122718
mf/mlx-base/tokenizer/added_tokens.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|audio_comp_end|>": 151667,
5
+ "<|audio_comp_span|>": 151666,
6
+ "<|audio_comp_start|>": 151665,
7
+ "<|audio_gen_end|>": 151670,
8
+ "<|audio_gen_span|>": 151669,
9
+ "<|audio_gen_start|>": 151668,
10
+ "<|box_end|>": 151649,
11
+ "<|box_start|>": 151648,
12
+ "<|endoftext|>": 151643,
13
+ "<|file_sep|>": 151664,
14
+ "<|fim_middle|>": 151660,
15
+ "<|fim_pad|>": 151662,
16
+ "<|fim_prefix|>": 151659,
17
+ "<|fim_suffix|>": 151661,
18
+ "<|im_end|>": 151645,
19
+ "<|im_start|>": 151644,
20
+ "<|image_pad|>": 151655,
21
+ "<|object_ref_end|>": 151647,
22
+ "<|object_ref_start|>": 151646,
23
+ "<|quad_end|>": 151651,
24
+ "<|quad_start|>": 151650,
25
+ "<|repo_name|>": 151663,
26
+ "<|text_cond_end|>": 151671,
27
+ "<|video_pad|>": 151656,
28
+ "<|vision_end|>": 151653,
29
+ "<|vision_pad|>": 151654,
30
+ "<|vision_start|>": 151652
31
+ }
mf/mlx-base/tokenizer/chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# 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>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\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" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
mf/mlx-base/tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
mf/mlx-base/tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "<|audio_comp_start|>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ {
11
+ "content": "<|audio_comp_span|>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ },
17
+ {
18
+ "content": "<|audio_comp_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ {
25
+ "content": "<|audio_gen_start|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ },
31
+ {
32
+ "content": "<|audio_gen_span|>",
33
+ "lstrip": false,
34
+ "normalized": false,
35
+ "rstrip": false,
36
+ "single_word": false
37
+ },
38
+ {
39
+ "content": "<|audio_gen_end|>",
40
+ "lstrip": false,
41
+ "normalized": false,
42
+ "rstrip": false,
43
+ "single_word": false
44
+ },
45
+ {
46
+ "content": "<|text_cond_end|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false
51
+ }
52
+ ],
53
+ "eos_token": {
54
+ "content": "<|endoftext|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false
59
+ },
60
+ "pad_token": {
61
+ "content": "<|endoftext|>",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false
66
+ }
67
+ }
mf/mlx-base/tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,257 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ },
181
+ "151665": {
182
+ "content": "<|audio_comp_start|>",
183
+ "lstrip": false,
184
+ "normalized": false,
185
+ "rstrip": false,
186
+ "single_word": false,
187
+ "special": true
188
+ },
189
+ "151666": {
190
+ "content": "<|audio_comp_span|>",
191
+ "lstrip": false,
192
+ "normalized": false,
193
+ "rstrip": false,
194
+ "single_word": false,
195
+ "special": true
196
+ },
197
+ "151667": {
198
+ "content": "<|audio_comp_end|>",
199
+ "lstrip": false,
200
+ "normalized": false,
201
+ "rstrip": false,
202
+ "single_word": false,
203
+ "special": true
204
+ },
205
+ "151668": {
206
+ "content": "<|audio_gen_start|>",
207
+ "lstrip": false,
208
+ "normalized": false,
209
+ "rstrip": false,
210
+ "single_word": false,
211
+ "special": true
212
+ },
213
+ "151669": {
214
+ "content": "<|audio_gen_span|>",
215
+ "lstrip": false,
216
+ "normalized": false,
217
+ "rstrip": false,
218
+ "single_word": false,
219
+ "special": true
220
+ },
221
+ "151670": {
222
+ "content": "<|audio_gen_end|>",
223
+ "lstrip": false,
224
+ "normalized": false,
225
+ "rstrip": false,
226
+ "single_word": false,
227
+ "special": true
228
+ },
229
+ "151671": {
230
+ "content": "<|text_cond_end|>",
231
+ "lstrip": false,
232
+ "normalized": false,
233
+ "rstrip": false,
234
+ "single_word": false,
235
+ "special": true
236
+ }
237
+ },
238
+ "additional_special_tokens": [
239
+ "<|audio_comp_start|>",
240
+ "<|audio_comp_span|>",
241
+ "<|audio_comp_end|>",
242
+ "<|audio_gen_start|>",
243
+ "<|audio_gen_span|>",
244
+ "<|audio_gen_end|>",
245
+ "<|text_cond_end|>"
246
+ ],
247
+ "bos_token": null,
248
+ "clean_up_tokenization_spaces": false,
249
+ "eos_token": "<|endoftext|>",
250
+ "errors": "replace",
251
+ "extra_special_tokens": {},
252
+ "model_max_length": 131072,
253
+ "pad_token": "<|endoftext|>",
254
+ "split_special_tokens": false,
255
+ "tokenizer_class": "Qwen2Tokenizer",
256
+ "unk_token": null
257
+ }
mf/mlx-base/tokenizer/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
mf/mlx-base/vocoder.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5b229d076aec43c20f7f259d2a2bbc75a4384b34adf7a669eefc69bff25e1c0
3
+ size 451779177
mf/mlx-int8/config.json ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "dots_tts",
3
+ "latent_dim": 128,
4
+ "patch_size": 4,
5
+ "cfg_droprate": 0.2,
6
+ "LLM": {
7
+ "model_name_or_path": "/prodcpfs/user/jiatu/logs/qwen25_15b"
8
+ },
9
+ "PatchEncoder": {
10
+ "num_layers": 24,
11
+ "num_heads": 16,
12
+ "hidden_size": 1024,
13
+ "ffn_hidden_size": 4096,
14
+ "modulation": false,
15
+ "qkv_bias": false,
16
+ "qk_norm": true,
17
+ "attn_dropout": 0.0,
18
+ "dropout": 0.1,
19
+ "norm_layer": "RMSNorm",
20
+ "alibi_bias": false,
21
+ "rotary_bias": true,
22
+ "rotary_theta": 10000.0,
23
+ "input_dim": 128,
24
+ "causal": true
25
+ },
26
+ "DiT": {
27
+ "num_layers": 18,
28
+ "num_heads": 16,
29
+ "hidden_size": 1024,
30
+ "ffn_hidden_size": 4096,
31
+ "modulation": true,
32
+ "qkv_bias": false,
33
+ "qk_norm": true,
34
+ "attn_dropout": 0.0,
35
+ "dropout": 0.0,
36
+ "norm_layer": "RMSNorm",
37
+ "alibi_bias": false,
38
+ "rotary_bias": true,
39
+ "rotary_theta": 10000.0
40
+ },
41
+ "vocoder": {
42
+ "sample_rate": 48000,
43
+ "upsample_rates": [
44
+ 10,
45
+ 6,
46
+ 4,
47
+ 2,
48
+ 2,
49
+ 2
50
+ ],
51
+ "upsample_kernel_sizes": [
52
+ 20,
53
+ 12,
54
+ 8,
55
+ 4,
56
+ 4,
57
+ 4
58
+ ],
59
+ "upsample_initial_channel": 1536,
60
+ "resblock": "1",
61
+ "resblock_kernel_sizes": [
62
+ 3,
63
+ 7,
64
+ 11
65
+ ],
66
+ "resblock_dilation_sizes": [
67
+ [
68
+ 1,
69
+ 3,
70
+ 5
71
+ ],
72
+ [
73
+ 1,
74
+ 3,
75
+ 5
76
+ ],
77
+ [
78
+ 1,
79
+ 3,
80
+ 5
81
+ ]
82
+ ],
83
+ "downsample_rates": [
84
+ 2,
85
+ 2,
86
+ 2,
87
+ 4,
88
+ 6,
89
+ 10
90
+ ],
91
+ "downsample_channels": [
92
+ 12,
93
+ 24,
94
+ 48,
95
+ 96,
96
+ 192,
97
+ 384,
98
+ 768
99
+ ],
100
+ "activation": "snakebeta",
101
+ "snake_logscale": true,
102
+ "latent_dim": 128,
103
+ "causal": true,
104
+ "mi_num_layers": 4,
105
+ "causal_encoder": true,
106
+ "use_bias_at_final": false,
107
+ "use_tanh_at_final": false
108
+ },
109
+ "fm_sigma": 0.0,
110
+ "xvec_drop_rate": 0.2,
111
+ "campplus_embedding_size": 512,
112
+ "xvec_max_audio_seconds": 10.0,
113
+ "meanflow": {
114
+ "enabled": true,
115
+ "use_duration_embedding": true
116
+ },
117
+ "architectures": [
118
+ "DotsTTSForConditionalGeneration"
119
+ ]
120
+ }
mf/mlx-int8/llm_config.json ADDED
@@ -0,0 +1,114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "vocab_size": 151672,
3
+ "max_position_embeddings": 131072,
4
+ "hidden_size": 1536,
5
+ "intermediate_size": 8960,
6
+ "num_hidden_layers": 28,
7
+ "num_attention_heads": 12,
8
+ "use_sliding_window": false,
9
+ "sliding_window": null,
10
+ "max_window_layers": 28,
11
+ "num_key_value_heads": 2,
12
+ "hidden_act": "silu",
13
+ "initializer_range": 0.02,
14
+ "rms_norm_eps": 1e-06,
15
+ "use_cache": true,
16
+ "rope_theta": 1000000.0,
17
+ "rope_scaling": null,
18
+ "attention_dropout": 0.0,
19
+ "layer_types": [
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention"
48
+ ],
49
+ "return_dict": true,
50
+ "output_hidden_states": false,
51
+ "torchscript": false,
52
+ "dtype": "bfloat16",
53
+ "pruned_heads": {},
54
+ "tie_word_embeddings": true,
55
+ "chunk_size_feed_forward": 0,
56
+ "is_encoder_decoder": false,
57
+ "is_decoder": false,
58
+ "cross_attention_hidden_size": null,
59
+ "add_cross_attention": false,
60
+ "tie_encoder_decoder": false,
61
+ "architectures": [
62
+ "Qwen2ForCausalLM"
63
+ ],
64
+ "finetuning_task": null,
65
+ "id2label": {
66
+ "0": "LABEL_0",
67
+ "1": "LABEL_1"
68
+ },
69
+ "label2id": {
70
+ "LABEL_0": 0,
71
+ "LABEL_1": 1
72
+ },
73
+ "task_specific_params": null,
74
+ "problem_type": null,
75
+ "tokenizer_class": null,
76
+ "prefix": null,
77
+ "bos_token_id": 151643,
78
+ "pad_token_id": null,
79
+ "eos_token_id": 151643,
80
+ "sep_token_id": null,
81
+ "decoder_start_token_id": null,
82
+ "max_length": 20,
83
+ "min_length": 0,
84
+ "do_sample": false,
85
+ "early_stopping": false,
86
+ "num_beams": 1,
87
+ "temperature": 1.0,
88
+ "top_k": 50,
89
+ "top_p": 1.0,
90
+ "typical_p": 1.0,
91
+ "repetition_penalty": 1.0,
92
+ "length_penalty": 1.0,
93
+ "no_repeat_ngram_size": 0,
94
+ "encoder_no_repeat_ngram_size": 0,
95
+ "bad_words_ids": null,
96
+ "num_return_sequences": 1,
97
+ "output_scores": false,
98
+ "return_dict_in_generate": false,
99
+ "forced_bos_token_id": null,
100
+ "forced_eos_token_id": null,
101
+ "remove_invalid_values": false,
102
+ "exponential_decay_length_penalty": null,
103
+ "suppress_tokens": null,
104
+ "begin_suppress_tokens": null,
105
+ "num_beam_groups": 1,
106
+ "diversity_penalty": 0.0,
107
+ "_name_or_path": "",
108
+ "transformers_version": "4.57.0",
109
+ "model_type": "qwen2",
110
+ "use_mrope": false,
111
+ "tf_legacy_loss": false,
112
+ "use_bfloat16": false,
113
+ "output_attentions": false
114
+ }
mf/mlx-int8/mlx_config.json ADDED
@@ -0,0 +1,243 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "artifact_class": "int8",
3
+ "dtype_policy": {
4
+ "core": {
5
+ "*": "bfloat16"
6
+ },
7
+ "latent_stats": {
8
+ "*": "float32"
9
+ },
10
+ "speaker": {
11
+ "*": "float32"
12
+ },
13
+ "vocoder": {
14
+ "audio_encoder.*": "float32",
15
+ "dec_mi_layer.*": "bfloat16",
16
+ "decoder.*": "bfloat16",
17
+ "enc_mi_layer.*": "float32",
18
+ "post_proj.*": "bfloat16",
19
+ "pre_proj.*": "float32"
20
+ }
21
+ },
22
+ "mode": "meanflow",
23
+ "model_family": "dots_tts",
24
+ "quantization": {
25
+ "bits": 8,
26
+ "group_size": 64,
27
+ "mode": "affine",
28
+ "module_types": [
29
+ "Linear",
30
+ "Embedding"
31
+ ],
32
+ "path_prefixes": [
33
+ "qwen.model."
34
+ ],
35
+ "quantized_paths": [
36
+ "qwen.model.embed_tokens",
37
+ "qwen.model.layers.0.mlp.down_proj",
38
+ "qwen.model.layers.0.mlp.gate_proj",
39
+ "qwen.model.layers.0.mlp.up_proj",
40
+ "qwen.model.layers.0.self_attn.k_proj",
41
+ "qwen.model.layers.0.self_attn.o_proj",
42
+ "qwen.model.layers.0.self_attn.q_proj",
43
+ "qwen.model.layers.0.self_attn.v_proj",
44
+ "qwen.model.layers.1.mlp.down_proj",
45
+ "qwen.model.layers.1.mlp.gate_proj",
46
+ "qwen.model.layers.1.mlp.up_proj",
47
+ "qwen.model.layers.1.self_attn.k_proj",
48
+ "qwen.model.layers.1.self_attn.o_proj",
49
+ "qwen.model.layers.1.self_attn.q_proj",
50
+ "qwen.model.layers.1.self_attn.v_proj",
51
+ "qwen.model.layers.10.mlp.down_proj",
52
+ "qwen.model.layers.10.mlp.gate_proj",
53
+ "qwen.model.layers.10.mlp.up_proj",
54
+ "qwen.model.layers.10.self_attn.k_proj",
55
+ "qwen.model.layers.10.self_attn.o_proj",
56
+ "qwen.model.layers.10.self_attn.q_proj",
57
+ "qwen.model.layers.10.self_attn.v_proj",
58
+ "qwen.model.layers.11.mlp.down_proj",
59
+ "qwen.model.layers.11.mlp.gate_proj",
60
+ "qwen.model.layers.11.mlp.up_proj",
61
+ "qwen.model.layers.11.self_attn.k_proj",
62
+ "qwen.model.layers.11.self_attn.o_proj",
63
+ "qwen.model.layers.11.self_attn.q_proj",
64
+ "qwen.model.layers.11.self_attn.v_proj",
65
+ "qwen.model.layers.12.mlp.down_proj",
66
+ "qwen.model.layers.12.mlp.gate_proj",
67
+ "qwen.model.layers.12.mlp.up_proj",
68
+ "qwen.model.layers.12.self_attn.k_proj",
69
+ "qwen.model.layers.12.self_attn.o_proj",
70
+ "qwen.model.layers.12.self_attn.q_proj",
71
+ "qwen.model.layers.12.self_attn.v_proj",
72
+ "qwen.model.layers.13.mlp.down_proj",
73
+ "qwen.model.layers.13.mlp.gate_proj",
74
+ "qwen.model.layers.13.mlp.up_proj",
75
+ "qwen.model.layers.13.self_attn.k_proj",
76
+ "qwen.model.layers.13.self_attn.o_proj",
77
+ "qwen.model.layers.13.self_attn.q_proj",
78
+ "qwen.model.layers.13.self_attn.v_proj",
79
+ "qwen.model.layers.14.mlp.down_proj",
80
+ "qwen.model.layers.14.mlp.gate_proj",
81
+ "qwen.model.layers.14.mlp.up_proj",
82
+ "qwen.model.layers.14.self_attn.k_proj",
83
+ "qwen.model.layers.14.self_attn.o_proj",
84
+ "qwen.model.layers.14.self_attn.q_proj",
85
+ "qwen.model.layers.14.self_attn.v_proj",
86
+ "qwen.model.layers.15.mlp.down_proj",
87
+ "qwen.model.layers.15.mlp.gate_proj",
88
+ "qwen.model.layers.15.mlp.up_proj",
89
+ "qwen.model.layers.15.self_attn.k_proj",
90
+ "qwen.model.layers.15.self_attn.o_proj",
91
+ "qwen.model.layers.15.self_attn.q_proj",
92
+ "qwen.model.layers.15.self_attn.v_proj",
93
+ "qwen.model.layers.16.mlp.down_proj",
94
+ "qwen.model.layers.16.mlp.gate_proj",
95
+ "qwen.model.layers.16.mlp.up_proj",
96
+ "qwen.model.layers.16.self_attn.k_proj",
97
+ "qwen.model.layers.16.self_attn.o_proj",
98
+ "qwen.model.layers.16.self_attn.q_proj",
99
+ "qwen.model.layers.16.self_attn.v_proj",
100
+ "qwen.model.layers.17.mlp.down_proj",
101
+ "qwen.model.layers.17.mlp.gate_proj",
102
+ "qwen.model.layers.17.mlp.up_proj",
103
+ "qwen.model.layers.17.self_attn.k_proj",
104
+ "qwen.model.layers.17.self_attn.o_proj",
105
+ "qwen.model.layers.17.self_attn.q_proj",
106
+ "qwen.model.layers.17.self_attn.v_proj",
107
+ "qwen.model.layers.18.mlp.down_proj",
108
+ "qwen.model.layers.18.mlp.gate_proj",
109
+ "qwen.model.layers.18.mlp.up_proj",
110
+ "qwen.model.layers.18.self_attn.k_proj",
111
+ "qwen.model.layers.18.self_attn.o_proj",
112
+ "qwen.model.layers.18.self_attn.q_proj",
113
+ "qwen.model.layers.18.self_attn.v_proj",
114
+ "qwen.model.layers.19.mlp.down_proj",
115
+ "qwen.model.layers.19.mlp.gate_proj",
116
+ "qwen.model.layers.19.mlp.up_proj",
117
+ "qwen.model.layers.19.self_attn.k_proj",
118
+ "qwen.model.layers.19.self_attn.o_proj",
119
+ "qwen.model.layers.19.self_attn.q_proj",
120
+ "qwen.model.layers.19.self_attn.v_proj",
121
+ "qwen.model.layers.2.mlp.down_proj",
122
+ "qwen.model.layers.2.mlp.gate_proj",
123
+ "qwen.model.layers.2.mlp.up_proj",
124
+ "qwen.model.layers.2.self_attn.k_proj",
125
+ "qwen.model.layers.2.self_attn.o_proj",
126
+ "qwen.model.layers.2.self_attn.q_proj",
127
+ "qwen.model.layers.2.self_attn.v_proj",
128
+ "qwen.model.layers.20.mlp.down_proj",
129
+ "qwen.model.layers.20.mlp.gate_proj",
130
+ "qwen.model.layers.20.mlp.up_proj",
131
+ "qwen.model.layers.20.self_attn.k_proj",
132
+ "qwen.model.layers.20.self_attn.o_proj",
133
+ "qwen.model.layers.20.self_attn.q_proj",
134
+ "qwen.model.layers.20.self_attn.v_proj",
135
+ "qwen.model.layers.21.mlp.down_proj",
136
+ "qwen.model.layers.21.mlp.gate_proj",
137
+ "qwen.model.layers.21.mlp.up_proj",
138
+ "qwen.model.layers.21.self_attn.k_proj",
139
+ "qwen.model.layers.21.self_attn.o_proj",
140
+ "qwen.model.layers.21.self_attn.q_proj",
141
+ "qwen.model.layers.21.self_attn.v_proj",
142
+ "qwen.model.layers.22.mlp.down_proj",
143
+ "qwen.model.layers.22.mlp.gate_proj",
144
+ "qwen.model.layers.22.mlp.up_proj",
145
+ "qwen.model.layers.22.self_attn.k_proj",
146
+ "qwen.model.layers.22.self_attn.o_proj",
147
+ "qwen.model.layers.22.self_attn.q_proj",
148
+ "qwen.model.layers.22.self_attn.v_proj",
149
+ "qwen.model.layers.23.mlp.down_proj",
150
+ "qwen.model.layers.23.mlp.gate_proj",
151
+ "qwen.model.layers.23.mlp.up_proj",
152
+ "qwen.model.layers.23.self_attn.k_proj",
153
+ "qwen.model.layers.23.self_attn.o_proj",
154
+ "qwen.model.layers.23.self_attn.q_proj",
155
+ "qwen.model.layers.23.self_attn.v_proj",
156
+ "qwen.model.layers.24.mlp.down_proj",
157
+ "qwen.model.layers.24.mlp.gate_proj",
158
+ "qwen.model.layers.24.mlp.up_proj",
159
+ "qwen.model.layers.24.self_attn.k_proj",
160
+ "qwen.model.layers.24.self_attn.o_proj",
161
+ "qwen.model.layers.24.self_attn.q_proj",
162
+ "qwen.model.layers.24.self_attn.v_proj",
163
+ "qwen.model.layers.25.mlp.down_proj",
164
+ "qwen.model.layers.25.mlp.gate_proj",
165
+ "qwen.model.layers.25.mlp.up_proj",
166
+ "qwen.model.layers.25.self_attn.k_proj",
167
+ "qwen.model.layers.25.self_attn.o_proj",
168
+ "qwen.model.layers.25.self_attn.q_proj",
169
+ "qwen.model.layers.25.self_attn.v_proj",
170
+ "qwen.model.layers.26.mlp.down_proj",
171
+ "qwen.model.layers.26.mlp.gate_proj",
172
+ "qwen.model.layers.26.mlp.up_proj",
173
+ "qwen.model.layers.26.self_attn.k_proj",
174
+ "qwen.model.layers.26.self_attn.o_proj",
175
+ "qwen.model.layers.26.self_attn.q_proj",
176
+ "qwen.model.layers.26.self_attn.v_proj",
177
+ "qwen.model.layers.27.mlp.down_proj",
178
+ "qwen.model.layers.27.mlp.gate_proj",
179
+ "qwen.model.layers.27.mlp.up_proj",
180
+ "qwen.model.layers.27.self_attn.k_proj",
181
+ "qwen.model.layers.27.self_attn.o_proj",
182
+ "qwen.model.layers.27.self_attn.q_proj",
183
+ "qwen.model.layers.27.self_attn.v_proj",
184
+ "qwen.model.layers.3.mlp.down_proj",
185
+ "qwen.model.layers.3.mlp.gate_proj",
186
+ "qwen.model.layers.3.mlp.up_proj",
187
+ "qwen.model.layers.3.self_attn.k_proj",
188
+ "qwen.model.layers.3.self_attn.o_proj",
189
+ "qwen.model.layers.3.self_attn.q_proj",
190
+ "qwen.model.layers.3.self_attn.v_proj",
191
+ "qwen.model.layers.4.mlp.down_proj",
192
+ "qwen.model.layers.4.mlp.gate_proj",
193
+ "qwen.model.layers.4.mlp.up_proj",
194
+ "qwen.model.layers.4.self_attn.k_proj",
195
+ "qwen.model.layers.4.self_attn.o_proj",
196
+ "qwen.model.layers.4.self_attn.q_proj",
197
+ "qwen.model.layers.4.self_attn.v_proj",
198
+ "qwen.model.layers.5.mlp.down_proj",
199
+ "qwen.model.layers.5.mlp.gate_proj",
200
+ "qwen.model.layers.5.mlp.up_proj",
201
+ "qwen.model.layers.5.self_attn.k_proj",
202
+ "qwen.model.layers.5.self_attn.o_proj",
203
+ "qwen.model.layers.5.self_attn.q_proj",
204
+ "qwen.model.layers.5.self_attn.v_proj",
205
+ "qwen.model.layers.6.mlp.down_proj",
206
+ "qwen.model.layers.6.mlp.gate_proj",
207
+ "qwen.model.layers.6.mlp.up_proj",
208
+ "qwen.model.layers.6.self_attn.k_proj",
209
+ "qwen.model.layers.6.self_attn.o_proj",
210
+ "qwen.model.layers.6.self_attn.q_proj",
211
+ "qwen.model.layers.6.self_attn.v_proj",
212
+ "qwen.model.layers.7.mlp.down_proj",
213
+ "qwen.model.layers.7.mlp.gate_proj",
214
+ "qwen.model.layers.7.mlp.up_proj",
215
+ "qwen.model.layers.7.self_attn.k_proj",
216
+ "qwen.model.layers.7.self_attn.o_proj",
217
+ "qwen.model.layers.7.self_attn.q_proj",
218
+ "qwen.model.layers.7.self_attn.v_proj",
219
+ "qwen.model.layers.8.mlp.down_proj",
220
+ "qwen.model.layers.8.mlp.gate_proj",
221
+ "qwen.model.layers.8.mlp.up_proj",
222
+ "qwen.model.layers.8.self_attn.k_proj",
223
+ "qwen.model.layers.8.self_attn.o_proj",
224
+ "qwen.model.layers.8.self_attn.q_proj",
225
+ "qwen.model.layers.8.self_attn.v_proj",
226
+ "qwen.model.layers.9.mlp.down_proj",
227
+ "qwen.model.layers.9.mlp.gate_proj",
228
+ "qwen.model.layers.9.mlp.up_proj",
229
+ "qwen.model.layers.9.self_attn.k_proj",
230
+ "qwen.model.layers.9.self_attn.o_proj",
231
+ "qwen.model.layers.9.self_attn.q_proj",
232
+ "qwen.model.layers.9.self_attn.v_proj"
233
+ ]
234
+ },
235
+ "schema_version": 1,
236
+ "source": {
237
+ "manifest_sha256": "c5b8638c5083a06b052a2472685ae5933a8f6b2bcbe96cc00cd11624bc12a6af",
238
+ "repo_id": "rednote-hilab/dots.tts-mf",
239
+ "resolved_repo_id": "dots-studio/dots.tts-mf",
240
+ "revision": "25c53fb462e57087e52237daa5ea30df1c5cc328"
241
+ },
242
+ "variant": "mf"
243
+ }
mf/mlx-int8/tokenizer/added_tokens.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|audio_comp_end|>": 151667,
5
+ "<|audio_comp_span|>": 151666,
6
+ "<|audio_comp_start|>": 151665,
7
+ "<|audio_gen_end|>": 151670,
8
+ "<|audio_gen_span|>": 151669,
9
+ "<|audio_gen_start|>": 151668,
10
+ "<|box_end|>": 151649,
11
+ "<|box_start|>": 151648,
12
+ "<|endoftext|>": 151643,
13
+ "<|file_sep|>": 151664,
14
+ "<|fim_middle|>": 151660,
15
+ "<|fim_pad|>": 151662,
16
+ "<|fim_prefix|>": 151659,
17
+ "<|fim_suffix|>": 151661,
18
+ "<|im_end|>": 151645,
19
+ "<|im_start|>": 151644,
20
+ "<|image_pad|>": 151655,
21
+ "<|object_ref_end|>": 151647,
22
+ "<|object_ref_start|>": 151646,
23
+ "<|quad_end|>": 151651,
24
+ "<|quad_start|>": 151650,
25
+ "<|repo_name|>": 151663,
26
+ "<|text_cond_end|>": 151671,
27
+ "<|video_pad|>": 151656,
28
+ "<|vision_end|>": 151653,
29
+ "<|vision_pad|>": 151654,
30
+ "<|vision_start|>": 151652
31
+ }
mf/mlx-int8/tokenizer/chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# 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>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\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" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
mf/mlx-int8/tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
mf/mlx-int8/tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "<|audio_comp_start|>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ {
11
+ "content": "<|audio_comp_span|>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ },
17
+ {
18
+ "content": "<|audio_comp_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ {
25
+ "content": "<|audio_gen_start|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ },
31
+ {
32
+ "content": "<|audio_gen_span|>",
33
+ "lstrip": false,
34
+ "normalized": false,
35
+ "rstrip": false,
36
+ "single_word": false
37
+ },
38
+ {
39
+ "content": "<|audio_gen_end|>",
40
+ "lstrip": false,
41
+ "normalized": false,
42
+ "rstrip": false,
43
+ "single_word": false
44
+ },
45
+ {
46
+ "content": "<|text_cond_end|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false
51
+ }
52
+ ],
53
+ "eos_token": {
54
+ "content": "<|endoftext|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false
59
+ },
60
+ "pad_token": {
61
+ "content": "<|endoftext|>",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false
66
+ }
67
+ }
mf/mlx-int8/tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,257 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ },
181
+ "151665": {
182
+ "content": "<|audio_comp_start|>",
183
+ "lstrip": false,
184
+ "normalized": false,
185
+ "rstrip": false,
186
+ "single_word": false,
187
+ "special": true
188
+ },
189
+ "151666": {
190
+ "content": "<|audio_comp_span|>",
191
+ "lstrip": false,
192
+ "normalized": false,
193
+ "rstrip": false,
194
+ "single_word": false,
195
+ "special": true
196
+ },
197
+ "151667": {
198
+ "content": "<|audio_comp_end|>",
199
+ "lstrip": false,
200
+ "normalized": false,
201
+ "rstrip": false,
202
+ "single_word": false,
203
+ "special": true
204
+ },
205
+ "151668": {
206
+ "content": "<|audio_gen_start|>",
207
+ "lstrip": false,
208
+ "normalized": false,
209
+ "rstrip": false,
210
+ "single_word": false,
211
+ "special": true
212
+ },
213
+ "151669": {
214
+ "content": "<|audio_gen_span|>",
215
+ "lstrip": false,
216
+ "normalized": false,
217
+ "rstrip": false,
218
+ "single_word": false,
219
+ "special": true
220
+ },
221
+ "151670": {
222
+ "content": "<|audio_gen_end|>",
223
+ "lstrip": false,
224
+ "normalized": false,
225
+ "rstrip": false,
226
+ "single_word": false,
227
+ "special": true
228
+ },
229
+ "151671": {
230
+ "content": "<|text_cond_end|>",
231
+ "lstrip": false,
232
+ "normalized": false,
233
+ "rstrip": false,
234
+ "single_word": false,
235
+ "special": true
236
+ }
237
+ },
238
+ "additional_special_tokens": [
239
+ "<|audio_comp_start|>",
240
+ "<|audio_comp_span|>",
241
+ "<|audio_comp_end|>",
242
+ "<|audio_gen_start|>",
243
+ "<|audio_gen_span|>",
244
+ "<|audio_gen_end|>",
245
+ "<|text_cond_end|>"
246
+ ],
247
+ "bos_token": null,
248
+ "clean_up_tokenization_spaces": false,
249
+ "eos_token": "<|endoftext|>",
250
+ "errors": "replace",
251
+ "extra_special_tokens": {},
252
+ "model_max_length": 131072,
253
+ "pad_token": "<|endoftext|>",
254
+ "split_special_tokens": false,
255
+ "tokenizer_class": "Qwen2Tokenizer",
256
+ "unk_token": null
257
+ }
mf/mlx-int8/tokenizer/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
soar/mlx-base/config.json ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "dots_tts",
3
+ "latent_dim": 128,
4
+ "patch_size": 4,
5
+ "cfg_droprate": 0.2,
6
+ "PatchEncoder": {
7
+ "num_layers": 24,
8
+ "num_heads": 16,
9
+ "hidden_size": 1024,
10
+ "ffn_hidden_size": 4096,
11
+ "modulation": false,
12
+ "qkv_bias": false,
13
+ "qk_norm": true,
14
+ "attn_dropout": 0.0,
15
+ "dropout": 0.1,
16
+ "norm_layer": "RMSNorm",
17
+ "alibi_bias": false,
18
+ "rotary_bias": true,
19
+ "rotary_theta": 10000.0,
20
+ "input_dim": 128,
21
+ "causal": true
22
+ },
23
+ "DiT": {
24
+ "num_layers": 18,
25
+ "num_heads": 16,
26
+ "hidden_size": 1024,
27
+ "ffn_hidden_size": 4096,
28
+ "modulation": true,
29
+ "qkv_bias": false,
30
+ "qk_norm": true,
31
+ "attn_dropout": 0.0,
32
+ "dropout": 0.0,
33
+ "norm_layer": "RMSNorm",
34
+ "alibi_bias": false,
35
+ "rotary_bias": true,
36
+ "rotary_theta": 10000.0
37
+ },
38
+ "vocoder": {
39
+ "sample_rate": 48000,
40
+ "upsample_rates": [
41
+ 10,
42
+ 6,
43
+ 4,
44
+ 2,
45
+ 2,
46
+ 2
47
+ ],
48
+ "upsample_kernel_sizes": [
49
+ 20,
50
+ 12,
51
+ 8,
52
+ 4,
53
+ 4,
54
+ 4
55
+ ],
56
+ "upsample_initial_channel": 1536,
57
+ "resblock": "1",
58
+ "resblock_kernel_sizes": [
59
+ 3,
60
+ 7,
61
+ 11
62
+ ],
63
+ "resblock_dilation_sizes": [
64
+ [
65
+ 1,
66
+ 3,
67
+ 5
68
+ ],
69
+ [
70
+ 1,
71
+ 3,
72
+ 5
73
+ ],
74
+ [
75
+ 1,
76
+ 3,
77
+ 5
78
+ ]
79
+ ],
80
+ "downsample_rates": [
81
+ 2,
82
+ 2,
83
+ 2,
84
+ 4,
85
+ 6,
86
+ 10
87
+ ],
88
+ "downsample_channels": [
89
+ 12,
90
+ 24,
91
+ 48,
92
+ 96,
93
+ 192,
94
+ 384,
95
+ 768
96
+ ],
97
+ "activation": "snakebeta",
98
+ "snake_logscale": true,
99
+ "latent_dim": 128,
100
+ "causal": true,
101
+ "mi_num_layers": 4,
102
+ "causal_encoder": true,
103
+ "use_bias_at_final": false,
104
+ "use_tanh_at_final": false
105
+ },
106
+ "fm_sigma": 0.0,
107
+ "xvec_drop_rate": 0.2,
108
+ "campplus_embedding_size": 512,
109
+ "xvec_max_audio_seconds": 10.0,
110
+ "architectures": [
111
+ "DotsTTSForConditionalGeneration"
112
+ ]
113
+ }
soar/mlx-base/core.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d16bce6f079ac3b3162ff25bc7ee92d70e8724fccd17791e15bf5cc6b817d43a
3
+ size 4396285089
soar/mlx-base/latent_stats.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:370ea7fe3e54c06336ff050eaf14f9554417f39b002b2dad5228f67c86c7260f
3
+ size 1160
soar/mlx-base/llm_config.json ADDED
@@ -0,0 +1,114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "vocab_size": 151672,
3
+ "max_position_embeddings": 131072,
4
+ "hidden_size": 1536,
5
+ "intermediate_size": 8960,
6
+ "num_hidden_layers": 28,
7
+ "num_attention_heads": 12,
8
+ "use_sliding_window": false,
9
+ "sliding_window": null,
10
+ "max_window_layers": 28,
11
+ "num_key_value_heads": 2,
12
+ "hidden_act": "silu",
13
+ "initializer_range": 0.02,
14
+ "rms_norm_eps": 1e-06,
15
+ "use_cache": true,
16
+ "rope_theta": 1000000.0,
17
+ "rope_scaling": null,
18
+ "attention_dropout": 0.0,
19
+ "layer_types": [
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention"
48
+ ],
49
+ "return_dict": true,
50
+ "output_hidden_states": false,
51
+ "torchscript": false,
52
+ "dtype": "bfloat16",
53
+ "pruned_heads": {},
54
+ "tie_word_embeddings": true,
55
+ "chunk_size_feed_forward": 0,
56
+ "is_encoder_decoder": false,
57
+ "is_decoder": false,
58
+ "cross_attention_hidden_size": null,
59
+ "add_cross_attention": false,
60
+ "tie_encoder_decoder": false,
61
+ "architectures": [
62
+ "Qwen2ForCausalLM"
63
+ ],
64
+ "finetuning_task": null,
65
+ "id2label": {
66
+ "0": "LABEL_0",
67
+ "1": "LABEL_1"
68
+ },
69
+ "label2id": {
70
+ "LABEL_0": 0,
71
+ "LABEL_1": 1
72
+ },
73
+ "task_specific_params": null,
74
+ "problem_type": null,
75
+ "tokenizer_class": null,
76
+ "prefix": null,
77
+ "bos_token_id": 151643,
78
+ "pad_token_id": null,
79
+ "eos_token_id": 151643,
80
+ "sep_token_id": null,
81
+ "decoder_start_token_id": null,
82
+ "max_length": 20,
83
+ "min_length": 0,
84
+ "do_sample": false,
85
+ "early_stopping": false,
86
+ "num_beams": 1,
87
+ "temperature": 1.0,
88
+ "top_k": 50,
89
+ "top_p": 1.0,
90
+ "typical_p": 1.0,
91
+ "repetition_penalty": 1.0,
92
+ "length_penalty": 1.0,
93
+ "no_repeat_ngram_size": 0,
94
+ "encoder_no_repeat_ngram_size": 0,
95
+ "bad_words_ids": null,
96
+ "num_return_sequences": 1,
97
+ "output_scores": false,
98
+ "return_dict_in_generate": false,
99
+ "forced_bos_token_id": null,
100
+ "forced_eos_token_id": null,
101
+ "remove_invalid_values": false,
102
+ "exponential_decay_length_penalty": null,
103
+ "suppress_tokens": null,
104
+ "begin_suppress_tokens": null,
105
+ "num_beam_groups": 1,
106
+ "diversity_penalty": 0.0,
107
+ "_name_or_path": "",
108
+ "transformers_version": "4.57.0",
109
+ "model_type": "qwen2",
110
+ "use_mrope": false,
111
+ "tf_legacy_loss": false,
112
+ "use_bfloat16": false,
113
+ "output_attentions": false
114
+ }
soar/mlx-base/mlx_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "artifact_class": "base",
3
+ "dtype_policy": {
4
+ "core": {
5
+ "*": "bfloat16"
6
+ },
7
+ "latent_stats": {
8
+ "*": "float32"
9
+ },
10
+ "speaker": {
11
+ "*": "float32"
12
+ },
13
+ "vocoder": {
14
+ "audio_encoder.*": "float32",
15
+ "dec_mi_layer.*": "bfloat16",
16
+ "decoder.*": "bfloat16",
17
+ "enc_mi_layer.*": "float32",
18
+ "post_proj.*": "bfloat16",
19
+ "pre_proj.*": "float32"
20
+ }
21
+ },
22
+ "mode": "flow_matching",
23
+ "model_family": "dots_tts",
24
+ "quantization": null,
25
+ "schema_version": 1,
26
+ "source": {
27
+ "manifest_sha256": "c5b8638c5083a06b052a2472685ae5933a8f6b2bcbe96cc00cd11624bc12a6af",
28
+ "repo_id": "rednote-hilab/dots.tts-soar",
29
+ "resolved_repo_id": "dots-studio/dots.tts-soar",
30
+ "revision": "e3520f75254d0020a0406db31c51a79d00d22d55"
31
+ },
32
+ "variant": "soar"
33
+ }
soar/mlx-base/speaker.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:105356b29dfe40c986412db337c252ab2d227703f433de64b525078de6ca5858
3
+ size 29122718
soar/mlx-base/tokenizer/added_tokens.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|audio_comp_end|>": 151667,
5
+ "<|audio_comp_span|>": 151666,
6
+ "<|audio_comp_start|>": 151665,
7
+ "<|audio_gen_end|>": 151670,
8
+ "<|audio_gen_span|>": 151669,
9
+ "<|audio_gen_start|>": 151668,
10
+ "<|box_end|>": 151649,
11
+ "<|box_start|>": 151648,
12
+ "<|endoftext|>": 151643,
13
+ "<|file_sep|>": 151664,
14
+ "<|fim_middle|>": 151660,
15
+ "<|fim_pad|>": 151662,
16
+ "<|fim_prefix|>": 151659,
17
+ "<|fim_suffix|>": 151661,
18
+ "<|im_end|>": 151645,
19
+ "<|im_start|>": 151644,
20
+ "<|image_pad|>": 151655,
21
+ "<|object_ref_end|>": 151647,
22
+ "<|object_ref_start|>": 151646,
23
+ "<|quad_end|>": 151651,
24
+ "<|quad_start|>": 151650,
25
+ "<|repo_name|>": 151663,
26
+ "<|text_cond_end|>": 151671,
27
+ "<|video_pad|>": 151656,
28
+ "<|vision_end|>": 151653,
29
+ "<|vision_pad|>": 151654,
30
+ "<|vision_start|>": 151652
31
+ }
soar/mlx-base/tokenizer/chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# 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>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\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" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
soar/mlx-base/tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
soar/mlx-base/tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "<|audio_comp_start|>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ {
11
+ "content": "<|audio_comp_span|>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ },
17
+ {
18
+ "content": "<|audio_comp_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ {
25
+ "content": "<|audio_gen_start|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ },
31
+ {
32
+ "content": "<|audio_gen_span|>",
33
+ "lstrip": false,
34
+ "normalized": false,
35
+ "rstrip": false,
36
+ "single_word": false
37
+ },
38
+ {
39
+ "content": "<|audio_gen_end|>",
40
+ "lstrip": false,
41
+ "normalized": false,
42
+ "rstrip": false,
43
+ "single_word": false
44
+ },
45
+ {
46
+ "content": "<|text_cond_end|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false
51
+ }
52
+ ],
53
+ "eos_token": {
54
+ "content": "<|endoftext|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false
59
+ },
60
+ "pad_token": {
61
+ "content": "<|endoftext|>",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false
66
+ }
67
+ }
soar/mlx-base/tokenizer/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c16521f66774c7a4774e5303b7c8ec5c99830c0be5aef6c6edde3ca2a5e05dd0
3
+ size 11423263
soar/mlx-base/tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,257 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ },
181
+ "151665": {
182
+ "content": "<|audio_comp_start|>",
183
+ "lstrip": false,
184
+ "normalized": false,
185
+ "rstrip": false,
186
+ "single_word": false,
187
+ "special": true
188
+ },
189
+ "151666": {
190
+ "content": "<|audio_comp_span|>",
191
+ "lstrip": false,
192
+ "normalized": false,
193
+ "rstrip": false,
194
+ "single_word": false,
195
+ "special": true
196
+ },
197
+ "151667": {
198
+ "content": "<|audio_comp_end|>",
199
+ "lstrip": false,
200
+ "normalized": false,
201
+ "rstrip": false,
202
+ "single_word": false,
203
+ "special": true
204
+ },
205
+ "151668": {
206
+ "content": "<|audio_gen_start|>",
207
+ "lstrip": false,
208
+ "normalized": false,
209
+ "rstrip": false,
210
+ "single_word": false,
211
+ "special": true
212
+ },
213
+ "151669": {
214
+ "content": "<|audio_gen_span|>",
215
+ "lstrip": false,
216
+ "normalized": false,
217
+ "rstrip": false,
218
+ "single_word": false,
219
+ "special": true
220
+ },
221
+ "151670": {
222
+ "content": "<|audio_gen_end|>",
223
+ "lstrip": false,
224
+ "normalized": false,
225
+ "rstrip": false,
226
+ "single_word": false,
227
+ "special": true
228
+ },
229
+ "151671": {
230
+ "content": "<|text_cond_end|>",
231
+ "lstrip": false,
232
+ "normalized": false,
233
+ "rstrip": false,
234
+ "single_word": false,
235
+ "special": true
236
+ }
237
+ },
238
+ "additional_special_tokens": [
239
+ "<|audio_comp_start|>",
240
+ "<|audio_comp_span|>",
241
+ "<|audio_comp_end|>",
242
+ "<|audio_gen_start|>",
243
+ "<|audio_gen_span|>",
244
+ "<|audio_gen_end|>",
245
+ "<|text_cond_end|>"
246
+ ],
247
+ "bos_token": null,
248
+ "clean_up_tokenization_spaces": false,
249
+ "eos_token": "<|endoftext|>",
250
+ "errors": "replace",
251
+ "extra_special_tokens": {},
252
+ "model_max_length": 131072,
253
+ "pad_token": "<|endoftext|>",
254
+ "split_special_tokens": false,
255
+ "tokenizer_class": "Qwen2Tokenizer",
256
+ "unk_token": null
257
+ }
soar/mlx-base/tokenizer/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
soar/mlx-base/vocoder.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5b229d076aec43c20f7f259d2a2bbc75a4384b34adf7a669eefc69bff25e1c0
3
+ size 451779177
soar/mlx-int8/config.json ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "dots_tts",
3
+ "latent_dim": 128,
4
+ "patch_size": 4,
5
+ "cfg_droprate": 0.2,
6
+ "PatchEncoder": {
7
+ "num_layers": 24,
8
+ "num_heads": 16,
9
+ "hidden_size": 1024,
10
+ "ffn_hidden_size": 4096,
11
+ "modulation": false,
12
+ "qkv_bias": false,
13
+ "qk_norm": true,
14
+ "attn_dropout": 0.0,
15
+ "dropout": 0.1,
16
+ "norm_layer": "RMSNorm",
17
+ "alibi_bias": false,
18
+ "rotary_bias": true,
19
+ "rotary_theta": 10000.0,
20
+ "input_dim": 128,
21
+ "causal": true
22
+ },
23
+ "DiT": {
24
+ "num_layers": 18,
25
+ "num_heads": 16,
26
+ "hidden_size": 1024,
27
+ "ffn_hidden_size": 4096,
28
+ "modulation": true,
29
+ "qkv_bias": false,
30
+ "qk_norm": true,
31
+ "attn_dropout": 0.0,
32
+ "dropout": 0.0,
33
+ "norm_layer": "RMSNorm",
34
+ "alibi_bias": false,
35
+ "rotary_bias": true,
36
+ "rotary_theta": 10000.0
37
+ },
38
+ "vocoder": {
39
+ "sample_rate": 48000,
40
+ "upsample_rates": [
41
+ 10,
42
+ 6,
43
+ 4,
44
+ 2,
45
+ 2,
46
+ 2
47
+ ],
48
+ "upsample_kernel_sizes": [
49
+ 20,
50
+ 12,
51
+ 8,
52
+ 4,
53
+ 4,
54
+ 4
55
+ ],
56
+ "upsample_initial_channel": 1536,
57
+ "resblock": "1",
58
+ "resblock_kernel_sizes": [
59
+ 3,
60
+ 7,
61
+ 11
62
+ ],
63
+ "resblock_dilation_sizes": [
64
+ [
65
+ 1,
66
+ 3,
67
+ 5
68
+ ],
69
+ [
70
+ 1,
71
+ 3,
72
+ 5
73
+ ],
74
+ [
75
+ 1,
76
+ 3,
77
+ 5
78
+ ]
79
+ ],
80
+ "downsample_rates": [
81
+ 2,
82
+ 2,
83
+ 2,
84
+ 4,
85
+ 6,
86
+ 10
87
+ ],
88
+ "downsample_channels": [
89
+ 12,
90
+ 24,
91
+ 48,
92
+ 96,
93
+ 192,
94
+ 384,
95
+ 768
96
+ ],
97
+ "activation": "snakebeta",
98
+ "snake_logscale": true,
99
+ "latent_dim": 128,
100
+ "causal": true,
101
+ "mi_num_layers": 4,
102
+ "causal_encoder": true,
103
+ "use_bias_at_final": false,
104
+ "use_tanh_at_final": false
105
+ },
106
+ "fm_sigma": 0.0,
107
+ "xvec_drop_rate": 0.2,
108
+ "campplus_embedding_size": 512,
109
+ "xvec_max_audio_seconds": 10.0,
110
+ "architectures": [
111
+ "DotsTTSForConditionalGeneration"
112
+ ]
113
+ }
soar/mlx-int8/core.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9eb78ce331cc116b990dfc4276f49fd548de7a232a710375a02fe9d296a9bdb9
3
+ size 2949614907
soar/mlx-int8/latent_stats.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:370ea7fe3e54c06336ff050eaf14f9554417f39b002b2dad5228f67c86c7260f
3
+ size 1160
soar/mlx-int8/llm_config.json ADDED
@@ -0,0 +1,114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "vocab_size": 151672,
3
+ "max_position_embeddings": 131072,
4
+ "hidden_size": 1536,
5
+ "intermediate_size": 8960,
6
+ "num_hidden_layers": 28,
7
+ "num_attention_heads": 12,
8
+ "use_sliding_window": false,
9
+ "sliding_window": null,
10
+ "max_window_layers": 28,
11
+ "num_key_value_heads": 2,
12
+ "hidden_act": "silu",
13
+ "initializer_range": 0.02,
14
+ "rms_norm_eps": 1e-06,
15
+ "use_cache": true,
16
+ "rope_theta": 1000000.0,
17
+ "rope_scaling": null,
18
+ "attention_dropout": 0.0,
19
+ "layer_types": [
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention"
48
+ ],
49
+ "return_dict": true,
50
+ "output_hidden_states": false,
51
+ "torchscript": false,
52
+ "dtype": "bfloat16",
53
+ "pruned_heads": {},
54
+ "tie_word_embeddings": true,
55
+ "chunk_size_feed_forward": 0,
56
+ "is_encoder_decoder": false,
57
+ "is_decoder": false,
58
+ "cross_attention_hidden_size": null,
59
+ "add_cross_attention": false,
60
+ "tie_encoder_decoder": false,
61
+ "architectures": [
62
+ "Qwen2ForCausalLM"
63
+ ],
64
+ "finetuning_task": null,
65
+ "id2label": {
66
+ "0": "LABEL_0",
67
+ "1": "LABEL_1"
68
+ },
69
+ "label2id": {
70
+ "LABEL_0": 0,
71
+ "LABEL_1": 1
72
+ },
73
+ "task_specific_params": null,
74
+ "problem_type": null,
75
+ "tokenizer_class": null,
76
+ "prefix": null,
77
+ "bos_token_id": 151643,
78
+ "pad_token_id": null,
79
+ "eos_token_id": 151643,
80
+ "sep_token_id": null,
81
+ "decoder_start_token_id": null,
82
+ "max_length": 20,
83
+ "min_length": 0,
84
+ "do_sample": false,
85
+ "early_stopping": false,
86
+ "num_beams": 1,
87
+ "temperature": 1.0,
88
+ "top_k": 50,
89
+ "top_p": 1.0,
90
+ "typical_p": 1.0,
91
+ "repetition_penalty": 1.0,
92
+ "length_penalty": 1.0,
93
+ "no_repeat_ngram_size": 0,
94
+ "encoder_no_repeat_ngram_size": 0,
95
+ "bad_words_ids": null,
96
+ "num_return_sequences": 1,
97
+ "output_scores": false,
98
+ "return_dict_in_generate": false,
99
+ "forced_bos_token_id": null,
100
+ "forced_eos_token_id": null,
101
+ "remove_invalid_values": false,
102
+ "exponential_decay_length_penalty": null,
103
+ "suppress_tokens": null,
104
+ "begin_suppress_tokens": null,
105
+ "num_beam_groups": 1,
106
+ "diversity_penalty": 0.0,
107
+ "_name_or_path": "",
108
+ "transformers_version": "4.57.0",
109
+ "model_type": "qwen2",
110
+ "use_mrope": false,
111
+ "tf_legacy_loss": false,
112
+ "use_bfloat16": false,
113
+ "output_attentions": false
114
+ }
soar/mlx-int8/mlx_config.json ADDED
@@ -0,0 +1,243 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "artifact_class": "int8",
3
+ "dtype_policy": {
4
+ "core": {
5
+ "*": "bfloat16"
6
+ },
7
+ "latent_stats": {
8
+ "*": "float32"
9
+ },
10
+ "speaker": {
11
+ "*": "float32"
12
+ },
13
+ "vocoder": {
14
+ "audio_encoder.*": "float32",
15
+ "dec_mi_layer.*": "bfloat16",
16
+ "decoder.*": "bfloat16",
17
+ "enc_mi_layer.*": "float32",
18
+ "post_proj.*": "bfloat16",
19
+ "pre_proj.*": "float32"
20
+ }
21
+ },
22
+ "mode": "flow_matching",
23
+ "model_family": "dots_tts",
24
+ "quantization": {
25
+ "bits": 8,
26
+ "group_size": 64,
27
+ "mode": "affine",
28
+ "module_types": [
29
+ "Linear",
30
+ "Embedding"
31
+ ],
32
+ "path_prefixes": [
33
+ "qwen.model."
34
+ ],
35
+ "quantized_paths": [
36
+ "qwen.model.embed_tokens",
37
+ "qwen.model.layers.0.mlp.down_proj",
38
+ "qwen.model.layers.0.mlp.gate_proj",
39
+ "qwen.model.layers.0.mlp.up_proj",
40
+ "qwen.model.layers.0.self_attn.k_proj",
41
+ "qwen.model.layers.0.self_attn.o_proj",
42
+ "qwen.model.layers.0.self_attn.q_proj",
43
+ "qwen.model.layers.0.self_attn.v_proj",
44
+ "qwen.model.layers.1.mlp.down_proj",
45
+ "qwen.model.layers.1.mlp.gate_proj",
46
+ "qwen.model.layers.1.mlp.up_proj",
47
+ "qwen.model.layers.1.self_attn.k_proj",
48
+ "qwen.model.layers.1.self_attn.o_proj",
49
+ "qwen.model.layers.1.self_attn.q_proj",
50
+ "qwen.model.layers.1.self_attn.v_proj",
51
+ "qwen.model.layers.10.mlp.down_proj",
52
+ "qwen.model.layers.10.mlp.gate_proj",
53
+ "qwen.model.layers.10.mlp.up_proj",
54
+ "qwen.model.layers.10.self_attn.k_proj",
55
+ "qwen.model.layers.10.self_attn.o_proj",
56
+ "qwen.model.layers.10.self_attn.q_proj",
57
+ "qwen.model.layers.10.self_attn.v_proj",
58
+ "qwen.model.layers.11.mlp.down_proj",
59
+ "qwen.model.layers.11.mlp.gate_proj",
60
+ "qwen.model.layers.11.mlp.up_proj",
61
+ "qwen.model.layers.11.self_attn.k_proj",
62
+ "qwen.model.layers.11.self_attn.o_proj",
63
+ "qwen.model.layers.11.self_attn.q_proj",
64
+ "qwen.model.layers.11.self_attn.v_proj",
65
+ "qwen.model.layers.12.mlp.down_proj",
66
+ "qwen.model.layers.12.mlp.gate_proj",
67
+ "qwen.model.layers.12.mlp.up_proj",
68
+ "qwen.model.layers.12.self_attn.k_proj",
69
+ "qwen.model.layers.12.self_attn.o_proj",
70
+ "qwen.model.layers.12.self_attn.q_proj",
71
+ "qwen.model.layers.12.self_attn.v_proj",
72
+ "qwen.model.layers.13.mlp.down_proj",
73
+ "qwen.model.layers.13.mlp.gate_proj",
74
+ "qwen.model.layers.13.mlp.up_proj",
75
+ "qwen.model.layers.13.self_attn.k_proj",
76
+ "qwen.model.layers.13.self_attn.o_proj",
77
+ "qwen.model.layers.13.self_attn.q_proj",
78
+ "qwen.model.layers.13.self_attn.v_proj",
79
+ "qwen.model.layers.14.mlp.down_proj",
80
+ "qwen.model.layers.14.mlp.gate_proj",
81
+ "qwen.model.layers.14.mlp.up_proj",
82
+ "qwen.model.layers.14.self_attn.k_proj",
83
+ "qwen.model.layers.14.self_attn.o_proj",
84
+ "qwen.model.layers.14.self_attn.q_proj",
85
+ "qwen.model.layers.14.self_attn.v_proj",
86
+ "qwen.model.layers.15.mlp.down_proj",
87
+ "qwen.model.layers.15.mlp.gate_proj",
88
+ "qwen.model.layers.15.mlp.up_proj",
89
+ "qwen.model.layers.15.self_attn.k_proj",
90
+ "qwen.model.layers.15.self_attn.o_proj",
91
+ "qwen.model.layers.15.self_attn.q_proj",
92
+ "qwen.model.layers.15.self_attn.v_proj",
93
+ "qwen.model.layers.16.mlp.down_proj",
94
+ "qwen.model.layers.16.mlp.gate_proj",
95
+ "qwen.model.layers.16.mlp.up_proj",
96
+ "qwen.model.layers.16.self_attn.k_proj",
97
+ "qwen.model.layers.16.self_attn.o_proj",
98
+ "qwen.model.layers.16.self_attn.q_proj",
99
+ "qwen.model.layers.16.self_attn.v_proj",
100
+ "qwen.model.layers.17.mlp.down_proj",
101
+ "qwen.model.layers.17.mlp.gate_proj",
102
+ "qwen.model.layers.17.mlp.up_proj",
103
+ "qwen.model.layers.17.self_attn.k_proj",
104
+ "qwen.model.layers.17.self_attn.o_proj",
105
+ "qwen.model.layers.17.self_attn.q_proj",
106
+ "qwen.model.layers.17.self_attn.v_proj",
107
+ "qwen.model.layers.18.mlp.down_proj",
108
+ "qwen.model.layers.18.mlp.gate_proj",
109
+ "qwen.model.layers.18.mlp.up_proj",
110
+ "qwen.model.layers.18.self_attn.k_proj",
111
+ "qwen.model.layers.18.self_attn.o_proj",
112
+ "qwen.model.layers.18.self_attn.q_proj",
113
+ "qwen.model.layers.18.self_attn.v_proj",
114
+ "qwen.model.layers.19.mlp.down_proj",
115
+ "qwen.model.layers.19.mlp.gate_proj",
116
+ "qwen.model.layers.19.mlp.up_proj",
117
+ "qwen.model.layers.19.self_attn.k_proj",
118
+ "qwen.model.layers.19.self_attn.o_proj",
119
+ "qwen.model.layers.19.self_attn.q_proj",
120
+ "qwen.model.layers.19.self_attn.v_proj",
121
+ "qwen.model.layers.2.mlp.down_proj",
122
+ "qwen.model.layers.2.mlp.gate_proj",
123
+ "qwen.model.layers.2.mlp.up_proj",
124
+ "qwen.model.layers.2.self_attn.k_proj",
125
+ "qwen.model.layers.2.self_attn.o_proj",
126
+ "qwen.model.layers.2.self_attn.q_proj",
127
+ "qwen.model.layers.2.self_attn.v_proj",
128
+ "qwen.model.layers.20.mlp.down_proj",
129
+ "qwen.model.layers.20.mlp.gate_proj",
130
+ "qwen.model.layers.20.mlp.up_proj",
131
+ "qwen.model.layers.20.self_attn.k_proj",
132
+ "qwen.model.layers.20.self_attn.o_proj",
133
+ "qwen.model.layers.20.self_attn.q_proj",
134
+ "qwen.model.layers.20.self_attn.v_proj",
135
+ "qwen.model.layers.21.mlp.down_proj",
136
+ "qwen.model.layers.21.mlp.gate_proj",
137
+ "qwen.model.layers.21.mlp.up_proj",
138
+ "qwen.model.layers.21.self_attn.k_proj",
139
+ "qwen.model.layers.21.self_attn.o_proj",
140
+ "qwen.model.layers.21.self_attn.q_proj",
141
+ "qwen.model.layers.21.self_attn.v_proj",
142
+ "qwen.model.layers.22.mlp.down_proj",
143
+ "qwen.model.layers.22.mlp.gate_proj",
144
+ "qwen.model.layers.22.mlp.up_proj",
145
+ "qwen.model.layers.22.self_attn.k_proj",
146
+ "qwen.model.layers.22.self_attn.o_proj",
147
+ "qwen.model.layers.22.self_attn.q_proj",
148
+ "qwen.model.layers.22.self_attn.v_proj",
149
+ "qwen.model.layers.23.mlp.down_proj",
150
+ "qwen.model.layers.23.mlp.gate_proj",
151
+ "qwen.model.layers.23.mlp.up_proj",
152
+ "qwen.model.layers.23.self_attn.k_proj",
153
+ "qwen.model.layers.23.self_attn.o_proj",
154
+ "qwen.model.layers.23.self_attn.q_proj",
155
+ "qwen.model.layers.23.self_attn.v_proj",
156
+ "qwen.model.layers.24.mlp.down_proj",
157
+ "qwen.model.layers.24.mlp.gate_proj",
158
+ "qwen.model.layers.24.mlp.up_proj",
159
+ "qwen.model.layers.24.self_attn.k_proj",
160
+ "qwen.model.layers.24.self_attn.o_proj",
161
+ "qwen.model.layers.24.self_attn.q_proj",
162
+ "qwen.model.layers.24.self_attn.v_proj",
163
+ "qwen.model.layers.25.mlp.down_proj",
164
+ "qwen.model.layers.25.mlp.gate_proj",
165
+ "qwen.model.layers.25.mlp.up_proj",
166
+ "qwen.model.layers.25.self_attn.k_proj",
167
+ "qwen.model.layers.25.self_attn.o_proj",
168
+ "qwen.model.layers.25.self_attn.q_proj",
169
+ "qwen.model.layers.25.self_attn.v_proj",
170
+ "qwen.model.layers.26.mlp.down_proj",
171
+ "qwen.model.layers.26.mlp.gate_proj",
172
+ "qwen.model.layers.26.mlp.up_proj",
173
+ "qwen.model.layers.26.self_attn.k_proj",
174
+ "qwen.model.layers.26.self_attn.o_proj",
175
+ "qwen.model.layers.26.self_attn.q_proj",
176
+ "qwen.model.layers.26.self_attn.v_proj",
177
+ "qwen.model.layers.27.mlp.down_proj",
178
+ "qwen.model.layers.27.mlp.gate_proj",
179
+ "qwen.model.layers.27.mlp.up_proj",
180
+ "qwen.model.layers.27.self_attn.k_proj",
181
+ "qwen.model.layers.27.self_attn.o_proj",
182
+ "qwen.model.layers.27.self_attn.q_proj",
183
+ "qwen.model.layers.27.self_attn.v_proj",
184
+ "qwen.model.layers.3.mlp.down_proj",
185
+ "qwen.model.layers.3.mlp.gate_proj",
186
+ "qwen.model.layers.3.mlp.up_proj",
187
+ "qwen.model.layers.3.self_attn.k_proj",
188
+ "qwen.model.layers.3.self_attn.o_proj",
189
+ "qwen.model.layers.3.self_attn.q_proj",
190
+ "qwen.model.layers.3.self_attn.v_proj",
191
+ "qwen.model.layers.4.mlp.down_proj",
192
+ "qwen.model.layers.4.mlp.gate_proj",
193
+ "qwen.model.layers.4.mlp.up_proj",
194
+ "qwen.model.layers.4.self_attn.k_proj",
195
+ "qwen.model.layers.4.self_attn.o_proj",
196
+ "qwen.model.layers.4.self_attn.q_proj",
197
+ "qwen.model.layers.4.self_attn.v_proj",
198
+ "qwen.model.layers.5.mlp.down_proj",
199
+ "qwen.model.layers.5.mlp.gate_proj",
200
+ "qwen.model.layers.5.mlp.up_proj",
201
+ "qwen.model.layers.5.self_attn.k_proj",
202
+ "qwen.model.layers.5.self_attn.o_proj",
203
+ "qwen.model.layers.5.self_attn.q_proj",
204
+ "qwen.model.layers.5.self_attn.v_proj",
205
+ "qwen.model.layers.6.mlp.down_proj",
206
+ "qwen.model.layers.6.mlp.gate_proj",
207
+ "qwen.model.layers.6.mlp.up_proj",
208
+ "qwen.model.layers.6.self_attn.k_proj",
209
+ "qwen.model.layers.6.self_attn.o_proj",
210
+ "qwen.model.layers.6.self_attn.q_proj",
211
+ "qwen.model.layers.6.self_attn.v_proj",
212
+ "qwen.model.layers.7.mlp.down_proj",
213
+ "qwen.model.layers.7.mlp.gate_proj",
214
+ "qwen.model.layers.7.mlp.up_proj",
215
+ "qwen.model.layers.7.self_attn.k_proj",
216
+ "qwen.model.layers.7.self_attn.o_proj",
217
+ "qwen.model.layers.7.self_attn.q_proj",
218
+ "qwen.model.layers.7.self_attn.v_proj",
219
+ "qwen.model.layers.8.mlp.down_proj",
220
+ "qwen.model.layers.8.mlp.gate_proj",
221
+ "qwen.model.layers.8.mlp.up_proj",
222
+ "qwen.model.layers.8.self_attn.k_proj",
223
+ "qwen.model.layers.8.self_attn.o_proj",
224
+ "qwen.model.layers.8.self_attn.q_proj",
225
+ "qwen.model.layers.8.self_attn.v_proj",
226
+ "qwen.model.layers.9.mlp.down_proj",
227
+ "qwen.model.layers.9.mlp.gate_proj",
228
+ "qwen.model.layers.9.mlp.up_proj",
229
+ "qwen.model.layers.9.self_attn.k_proj",
230
+ "qwen.model.layers.9.self_attn.o_proj",
231
+ "qwen.model.layers.9.self_attn.q_proj",
232
+ "qwen.model.layers.9.self_attn.v_proj"
233
+ ]
234
+ },
235
+ "schema_version": 1,
236
+ "source": {
237
+ "manifest_sha256": "c5b8638c5083a06b052a2472685ae5933a8f6b2bcbe96cc00cd11624bc12a6af",
238
+ "repo_id": "rednote-hilab/dots.tts-soar",
239
+ "resolved_repo_id": "dots-studio/dots.tts-soar",
240
+ "revision": "e3520f75254d0020a0406db31c51a79d00d22d55"
241
+ },
242
+ "variant": "soar"
243
+ }
soar/mlx-int8/speaker.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:105356b29dfe40c986412db337c252ab2d227703f433de64b525078de6ca5858
3
+ size 29122718
soar/mlx-int8/tokenizer/added_tokens.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|audio_comp_end|>": 151667,
5
+ "<|audio_comp_span|>": 151666,
6
+ "<|audio_comp_start|>": 151665,
7
+ "<|audio_gen_end|>": 151670,
8
+ "<|audio_gen_span|>": 151669,
9
+ "<|audio_gen_start|>": 151668,
10
+ "<|box_end|>": 151649,
11
+ "<|box_start|>": 151648,
12
+ "<|endoftext|>": 151643,
13
+ "<|file_sep|>": 151664,
14
+ "<|fim_middle|>": 151660,
15
+ "<|fim_pad|>": 151662,
16
+ "<|fim_prefix|>": 151659,
17
+ "<|fim_suffix|>": 151661,
18
+ "<|im_end|>": 151645,
19
+ "<|im_start|>": 151644,
20
+ "<|image_pad|>": 151655,
21
+ "<|object_ref_end|>": 151647,
22
+ "<|object_ref_start|>": 151646,
23
+ "<|quad_end|>": 151651,
24
+ "<|quad_start|>": 151650,
25
+ "<|repo_name|>": 151663,
26
+ "<|text_cond_end|>": 151671,
27
+ "<|video_pad|>": 151656,
28
+ "<|vision_end|>": 151653,
29
+ "<|vision_pad|>": 151654,
30
+ "<|vision_start|>": 151652
31
+ }
soar/mlx-int8/tokenizer/chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# 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>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\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" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
soar/mlx-int8/tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
soar/mlx-int8/tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "<|audio_comp_start|>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ {
11
+ "content": "<|audio_comp_span|>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ },
17
+ {
18
+ "content": "<|audio_comp_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ {
25
+ "content": "<|audio_gen_start|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ },
31
+ {
32
+ "content": "<|audio_gen_span|>",
33
+ "lstrip": false,
34
+ "normalized": false,
35
+ "rstrip": false,
36
+ "single_word": false
37
+ },
38
+ {
39
+ "content": "<|audio_gen_end|>",
40
+ "lstrip": false,
41
+ "normalized": false,
42
+ "rstrip": false,
43
+ "single_word": false
44
+ },
45
+ {
46
+ "content": "<|text_cond_end|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false
51
+ }
52
+ ],
53
+ "eos_token": {
54
+ "content": "<|endoftext|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false
59
+ },
60
+ "pad_token": {
61
+ "content": "<|endoftext|>",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false
66
+ }
67
+ }
soar/mlx-int8/tokenizer/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c16521f66774c7a4774e5303b7c8ec5c99830c0be5aef6c6edde3ca2a5e05dd0
3
+ size 11423263
soar/mlx-int8/tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,257 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ },
181
+ "151665": {
182
+ "content": "<|audio_comp_start|>",
183
+ "lstrip": false,
184
+ "normalized": false,
185
+ "rstrip": false,
186
+ "single_word": false,
187
+ "special": true
188
+ },
189
+ "151666": {
190
+ "content": "<|audio_comp_span|>",
191
+ "lstrip": false,
192
+ "normalized": false,
193
+ "rstrip": false,
194
+ "single_word": false,
195
+ "special": true
196
+ },
197
+ "151667": {
198
+ "content": "<|audio_comp_end|>",
199
+ "lstrip": false,
200
+ "normalized": false,
201
+ "rstrip": false,
202
+ "single_word": false,
203
+ "special": true
204
+ },
205
+ "151668": {
206
+ "content": "<|audio_gen_start|>",
207
+ "lstrip": false,
208
+ "normalized": false,
209
+ "rstrip": false,
210
+ "single_word": false,
211
+ "special": true
212
+ },
213
+ "151669": {
214
+ "content": "<|audio_gen_span|>",
215
+ "lstrip": false,
216
+ "normalized": false,
217
+ "rstrip": false,
218
+ "single_word": false,
219
+ "special": true
220
+ },
221
+ "151670": {
222
+ "content": "<|audio_gen_end|>",
223
+ "lstrip": false,
224
+ "normalized": false,
225
+ "rstrip": false,
226
+ "single_word": false,
227
+ "special": true
228
+ },
229
+ "151671": {
230
+ "content": "<|text_cond_end|>",
231
+ "lstrip": false,
232
+ "normalized": false,
233
+ "rstrip": false,
234
+ "single_word": false,
235
+ "special": true
236
+ }
237
+ },
238
+ "additional_special_tokens": [
239
+ "<|audio_comp_start|>",
240
+ "<|audio_comp_span|>",
241
+ "<|audio_comp_end|>",
242
+ "<|audio_gen_start|>",
243
+ "<|audio_gen_span|>",
244
+ "<|audio_gen_end|>",
245
+ "<|text_cond_end|>"
246
+ ],
247
+ "bos_token": null,
248
+ "clean_up_tokenization_spaces": false,
249
+ "eos_token": "<|endoftext|>",
250
+ "errors": "replace",
251
+ "extra_special_tokens": {},
252
+ "model_max_length": 131072,
253
+ "pad_token": "<|endoftext|>",
254
+ "split_special_tokens": false,
255
+ "tokenizer_class": "Qwen2Tokenizer",
256
+ "unk_token": null
257
+ }
soar/mlx-int8/tokenizer/vocab.json ADDED
The diff for this file is too large to render. See raw diff