Hugging Face's logo

OpenMOSS-Team
/
MOSS-Audio-4B-Thinking

Audio-Text-to-Text
Safetensors
English
Chinese
moss_audio
audio
speech
music
understanding
multimodal
reasoning
chain-of-thought
custom_code
Model card Files
xet
 Community
1
kiiic commited on
Commit
3dfd141
·
verified ·
1 Parent(s): 9ca172d

Upload folder using huggingface_hub

Browse files
Files changed (15) hide show
  1. added_tokens.json +28 -0
  2. chat_template.jinja +85 -0
  3. config.json +109 -0
  4. configuration_moss_audio.py +31 -0
  5. generation_config.json +6 -0
  6. merges.txt +0 -0
  7. model-00001-of-00003.safetensors +3 -0
  8. model-00002-of-00003.safetensors +3 -0
  9. model-00003-of-00003.safetensors +3 -0
  10. model.safetensors.index.json +910 -0
  11. preprocessor_config.json +18 -0
  12. processing_moss_audio.py +371 -0
  13. special_tokens_map.json +31 -0
  14. tokenizer_config.json +239 -0
  15. vocab.json +0 -0
added_tokens.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</think>": 151668,
3
+ "</tool_call>": 151658,
4
+ "</tool_response>": 151666,
5
+ "<think>": 151667,
6
+ "<tool_call>": 151657,
7
+ "<tool_response>": 151665,
8
+ "<|box_end|>": 151649,
9
+ "<|box_start|>": 151648,
10
+ "<|endoftext|>": 151643,
11
+ "<|file_sep|>": 151664,
12
+ "<|fim_middle|>": 151660,
13
+ "<|fim_pad|>": 151662,
14
+ "<|fim_prefix|>": 151659,
15
+ "<|fim_suffix|>": 151661,
16
+ "<|im_end|>": 151645,
17
+ "<|im_start|>": 151644,
18
+ "<|image_pad|>": 151655,
19
+ "<|object_ref_end|>": 151647,
20
+ "<|object_ref_start|>": 151646,
21
+ "<|quad_end|>": 151651,
22
+ "<|quad_start|>": 151650,
23
+ "<|repo_name|>": 151663,
24
+ "<|video_pad|>": 151656,
25
+ "<|vision_end|>": 151653,
26
+ "<|vision_pad|>": 151654,
27
+ "<|vision_start|>": 151652
28
+ }
chat_template.jinja ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {{- messages[0].content + '\n\n' }}
5
+ {%- endif %}
6
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
7
+ {%- for tool in tools %}
8
+ {{- "\n" }}
9
+ {{- tool | tojson }}
10
+ {%- endfor %}
11
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
12
+ {%- else %}
13
+ {%- if messages[0].role == 'system' %}
14
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
+ {%- endif %}
16
+ {%- endif %}
17
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
18
+ {%- for message in messages[::-1] %}
19
+ {%- set index = (messages|length - 1) - loop.index0 %}
20
+ {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
21
+ {%- set ns.multi_step_tool = false %}
22
+ {%- set ns.last_query_index = index %}
23
+ {%- endif %}
24
+ {%- endfor %}
25
+ {%- for message in messages %}
26
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
27
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
28
+ {%- elif message.role == "assistant" %}
29
+ {%- set content = message.content %}
30
+ {%- set reasoning_content = '' %}
31
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
32
+ {%- set reasoning_content = message.reasoning_content %}
33
+ {%- else %}
34
+ {%- if '</think>' in message.content %}
35
+ {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
36
+ {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
37
+ {%- endif %}
38
+ {%- endif %}
39
+ {%- if loop.index0 > ns.last_query_index %}
40
+ {%- if loop.last or (not loop.last and reasoning_content) %}
41
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
42
+ {%- else %}
43
+ {{- '<|im_start|>' + message.role + '\n' + content }}
44
+ {%- endif %}
45
+ {%- else %}
46
+ {{- '<|im_start|>' + message.role + '\n' + content }}
47
+ {%- endif %}
48
+ {%- if message.tool_calls %}
49
+ {%- for tool_call in message.tool_calls %}
50
+ {%- if (loop.first and content) or (not loop.first) %}
51
+ {{- '\n' }}
52
+ {%- endif %}
53
+ {%- if tool_call.function %}
54
+ {%- set tool_call = tool_call.function %}
55
+ {%- endif %}
56
+ {{- '<tool_call>\n{"name": "' }}
57
+ {{- tool_call.name }}
58
+ {{- '", "arguments": ' }}
59
+ {%- if tool_call.arguments is string %}
60
+ {{- tool_call.arguments }}
61
+ {%- else %}
62
+ {{- tool_call.arguments | tojson }}
63
+ {%- endif %}
64
+ {{- '}\n</tool_call>' }}
65
+ {%- endfor %}
66
+ {%- endif %}
67
+ {{- '<|im_end|>\n' }}
68
+ {%- elif message.role == "tool" %}
69
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
70
+ {{- '<|im_start|>user' }}
71
+ {%- endif %}
72
+ {{- '\n<tool_response>\n' }}
73
+ {{- message.content }}
74
+ {{- '\n</tool_response>' }}
75
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
76
+ {{- '<|im_end|>\n' }}
77
+ {%- endif %}
78
+ {%- endif %}
79
+ {%- endfor %}
80
+ {%- if add_generation_prompt %}
81
+ {{- '<|im_start|>assistant\n' }}
82
+ {%- if enable_thinking is defined and enable_thinking is false %}
83
+ {{- '<think>\n\n</think>\n\n' }}
84
+ {%- endif %}
85
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "adapter_hidden_size": 8192,
3
+ "architectures": [
4
+ "MossAudioModel"
5
+ ],
6
+ "audio_config": {
7
+ "_attn_implementation": "eager",
8
+ "activation_dropout": 0.0,
9
+ "activation_function": "gelu",
10
+ "attention_dropout": 0.1,
11
+ "d_model": 1280,
12
+ "deepstack_encoder_layer_indexes": [
13
+ 8,
14
+ 16,
15
+ 24
16
+ ],
17
+ "downsample_hidden_size": 480,
18
+ "downsample_rate": 8,
19
+ "dropout": 0.1,
20
+ "encoder_attention_heads": 20,
21
+ "encoder_attention_window_size": 100,
22
+ "encoder_ffn_dim": 5120,
23
+ "encoder_layers": 32,
24
+ "layer_norm_eps": 1e-05,
25
+ "max_source_positions": 1500,
26
+ "num_mel_bins": 128,
27
+ "output_dim": 1280,
28
+ "pretrained_path": ""
29
+ },
30
+ "auto_map": {
31
+ "AutoConfig": "configuration_moss_audio.MossAudioConfig",
32
+ "AutoProcessor": "processing_moss_audio.MossAudioProcessor"
33
+ },
34
+ "bos_token_id": 151643,
35
+ "deepstack_num_inject_layers": 3,
36
+ "dtype": "bfloat16",
37
+ "eos_token_id": 151645,
38
+ "ignore_index": -100,
39
+ "language_config": {
40
+ "architectures": [
41
+ "Qwen3ForCausalLM"
42
+ ],
43
+ "attention_bias": false,
44
+ "attention_dropout": 0.0,
45
+ "bos_token_id": 151643,
46
+ "eos_token_id": 151645,
47
+ "head_dim": 128,
48
+ "hidden_act": "silu",
49
+ "hidden_size": 2560,
50
+ "initializer_range": 0.02,
51
+ "intermediate_size": 9728,
52
+ "layer_types": [
53
+ "full_attention",
54
+ "full_attention",
55
+ "full_attention",
56
+ "full_attention",
57
+ "full_attention",
58
+ "full_attention",
59
+ "full_attention",
60
+ "full_attention",
61
+ "full_attention",
62
+ "full_attention",
63
+ "full_attention",
64
+ "full_attention",
65
+ "full_attention",
66
+ "full_attention",
67
+ "full_attention",
68
+ "full_attention",
69
+ "full_attention",
70
+ "full_attention",
71
+ "full_attention",
72
+ "full_attention",
73
+ "full_attention",
74
+ "full_attention",
75
+ "full_attention",
76
+ "full_attention",
77
+ "full_attention",
78
+ "full_attention",
79
+ "full_attention",
80
+ "full_attention",
81
+ "full_attention",
82
+ "full_attention",
83
+ "full_attention",
84
+ "full_attention",
85
+ "full_attention",
86
+ "full_attention",
87
+ "full_attention",
88
+ "full_attention"
89
+ ],
90
+ "max_position_embeddings": 40960,
91
+ "max_window_layers": 36,
92
+ "model_type": "qwen3",
93
+ "num_attention_heads": 32,
94
+ "num_hidden_layers": 36,
95
+ "num_key_value_heads": 8,
96
+ "rms_norm_eps": 1e-06,
97
+ "rope_scaling": null,
98
+ "rope_theta": 1000000,
99
+ "sliding_window": null,
100
+ "use_cache": true,
101
+ "use_sliding_window": false,
102
+ "vocab_size": 151936
103
+ },
104
+ "model_type": "moss_audio",
105
+ "num_hidden_layers": 36,
106
+ "tie_word_embeddings": false,
107
+ "transformers_version": "4.57.1",
108
+ "vocab_size": 151936
109
+ }
configuration_moss_audio.py ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from transformers import PretrainedConfig, Qwen3Config
2
+
3
+
4
+ class MossAudioConfig(PretrainedConfig):
5
+ model_type = "moss_audio"
6
+ is_composition = True
7
+
8
+ def __init__(
9
+ self,
10
+ audio_config=None,
11
+ language_config=None,
12
+ adapter_hidden_size=8192,
13
+ ignore_index=-100,
14
+ deepstack_num_inject_layers=None,
15
+ **kwargs,
16
+ ):
17
+ if isinstance(language_config, dict):
18
+ language_config = Qwen3Config(**language_config)
19
+ elif language_config is None:
20
+ language_config = Qwen3Config()
21
+
22
+ self.audio_config = audio_config
23
+ self.language_config = language_config
24
+ self.adapter_hidden_size = adapter_hidden_size
25
+ self.ignore_index = ignore_index
26
+ self.deepstack_num_inject_layers = deepstack_num_inject_layers
27
+
28
+ for key in ("num_hidden_layers", "eos_token_id", "bos_token_id", "vocab_size"):
29
+ kwargs.setdefault(key, getattr(language_config, key, None))
30
+
31
+ super().__init__(**kwargs)
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 151643,
4
+ "eos_token_id": 151645,
5
+ "transformers_version": "4.57.1"
6
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4f8c8893858a5adafb4e875b0f87499c6805d4f08b9a58be0aca7f42553e05a
3
+ size 4993540568
model-00002-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56010b47b02a338327eadb6b845bfc35c1866ee295ea44dd94676284633cda29
3
+ size 4674299440
model-00003-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6dc41baca9724f45bb0be81f71b84ac380d0c2e59a37e15ecd8c879c9117b6d8
3
+ size 777912448
model.safetensors.index.json ADDED
@@ -0,0 +1,910 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_parameters": 5222820736,
4
+ "total_size": 10445644032
5
+ },
6
+ "weight_map": {
7
+ "audio_adapter.down_proj.weight": "model-00002-of-00003.safetensors",
8
+ "audio_adapter.gate_proj.weight": "model-00002-of-00003.safetensors",
9
+ "audio_adapter.up_proj.weight": "model-00002-of-00003.safetensors",
10
+ "audio_encoder.conv1.bias": "model-00001-of-00003.safetensors",
11
+ "audio_encoder.conv1.weight": "model-00001-of-00003.safetensors",
12
+ "audio_encoder.conv2.bias": "model-00001-of-00003.safetensors",
13
+ "audio_encoder.conv2.weight": "model-00001-of-00003.safetensors",
14
+ "audio_encoder.conv3.bias": "model-00001-of-00003.safetensors",
15
+ "audio_encoder.conv3.weight": "model-00001-of-00003.safetensors",
16
+ "audio_encoder.embed_positions.inv_timescales": "model-00001-of-00003.safetensors",
17
+ "audio_encoder.layer_norm.bias": "model-00001-of-00003.safetensors",
18
+ "audio_encoder.layer_norm.weight": "model-00001-of-00003.safetensors",
19
+ "audio_encoder.layers.0.fc1.bias": "model-00001-of-00003.safetensors",
20
+ "audio_encoder.layers.0.fc1.weight": "model-00001-of-00003.safetensors",
21
+ "audio_encoder.layers.0.fc2.bias": "model-00001-of-00003.safetensors",
22
+ "audio_encoder.layers.0.fc2.weight": "model-00001-of-00003.safetensors",
23
+ "audio_encoder.layers.0.final_layer_norm.bias": "model-00001-of-00003.safetensors",
24
+ "audio_encoder.layers.0.final_layer_norm.weight": "model-00001-of-00003.safetensors",
25
+ "audio_encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
26
+ "audio_encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
27
+ "audio_encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
28
+ "audio_encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
29
+ "audio_encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
30
+ "audio_encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
31
+ "audio_encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
32
+ "audio_encoder.layers.0.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
33
+ "audio_encoder.layers.0.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
34
+ "audio_encoder.layers.1.fc1.bias": "model-00001-of-00003.safetensors",
35
+ "audio_encoder.layers.1.fc1.weight": "model-00001-of-00003.safetensors",
36
+ "audio_encoder.layers.1.fc2.bias": "model-00001-of-00003.safetensors",
37
+ "audio_encoder.layers.1.fc2.weight": "model-00001-of-00003.safetensors",
38
+ "audio_encoder.layers.1.final_layer_norm.bias": "model-00001-of-00003.safetensors",
39
+ "audio_encoder.layers.1.final_layer_norm.weight": "model-00001-of-00003.safetensors",
40
+ "audio_encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
41
+ "audio_encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
42
+ "audio_encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
43
+ "audio_encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
44
+ "audio_encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
45
+ "audio_encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
46
+ "audio_encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
47
+ "audio_encoder.layers.1.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
48
+ "audio_encoder.layers.1.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
49
+ "audio_encoder.layers.10.fc1.bias": "model-00001-of-00003.safetensors",
50
+ "audio_encoder.layers.10.fc1.weight": "model-00001-of-00003.safetensors",
51
+ "audio_encoder.layers.10.fc2.bias": "model-00001-of-00003.safetensors",
52
+ "audio_encoder.layers.10.fc2.weight": "model-00001-of-00003.safetensors",
53
+ "audio_encoder.layers.10.final_layer_norm.bias": "model-00001-of-00003.safetensors",
54
+ "audio_encoder.layers.10.final_layer_norm.weight": "model-00001-of-00003.safetensors",
55
+ "audio_encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
56
+ "audio_encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
57
+ "audio_encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
58
+ "audio_encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
59
+ "audio_encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
60
+ "audio_encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
61
+ "audio_encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
62
+ "audio_encoder.layers.10.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
63
+ "audio_encoder.layers.10.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
64
+ "audio_encoder.layers.11.fc1.bias": "model-00001-of-00003.safetensors",
65
+ "audio_encoder.layers.11.fc1.weight": "model-00001-of-00003.safetensors",
66
+ "audio_encoder.layers.11.fc2.bias": "model-00001-of-00003.safetensors",
67
+ "audio_encoder.layers.11.fc2.weight": "model-00001-of-00003.safetensors",
68
+ "audio_encoder.layers.11.final_layer_norm.bias": "model-00001-of-00003.safetensors",
69
+ "audio_encoder.layers.11.final_layer_norm.weight": "model-00001-of-00003.safetensors",
70
+ "audio_encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
71
+ "audio_encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
72
+ "audio_encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
73
+ "audio_encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
74
+ "audio_encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
75
+ "audio_encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
76
+ "audio_encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
77
+ "audio_encoder.layers.11.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
78
+ "audio_encoder.layers.11.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
79
+ "audio_encoder.layers.12.fc1.bias": "model-00001-of-00003.safetensors",
80
+ "audio_encoder.layers.12.fc1.weight": "model-00001-of-00003.safetensors",
81
+ "audio_encoder.layers.12.fc2.bias": "model-00001-of-00003.safetensors",
82
+ "audio_encoder.layers.12.fc2.weight": "model-00001-of-00003.safetensors",
83
+ "audio_encoder.layers.12.final_layer_norm.bias": "model-00001-of-00003.safetensors",
84
+ "audio_encoder.layers.12.final_layer_norm.weight": "model-00001-of-00003.safetensors",
85
+ "audio_encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
86
+ "audio_encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
87
+ "audio_encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
88
+ "audio_encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
89
+ "audio_encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
90
+ "audio_encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
91
+ "audio_encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
92
+ "audio_encoder.layers.12.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
93
+ "audio_encoder.layers.12.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
94
+ "audio_encoder.layers.13.fc1.bias": "model-00001-of-00003.safetensors",
95
+ "audio_encoder.layers.13.fc1.weight": "model-00001-of-00003.safetensors",
96
+ "audio_encoder.layers.13.fc2.bias": "model-00001-of-00003.safetensors",
97
+ "audio_encoder.layers.13.fc2.weight": "model-00001-of-00003.safetensors",
98
+ "audio_encoder.layers.13.final_layer_norm.bias": "model-00001-of-00003.safetensors",
99
+ "audio_encoder.layers.13.final_layer_norm.weight": "model-00001-of-00003.safetensors",
100
+ "audio_encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
101
+ "audio_encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
102
+ "audio_encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
103
+ "audio_encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
104
+ "audio_encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
105
+ "audio_encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
106
+ "audio_encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
107
+ "audio_encoder.layers.13.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
108
+ "audio_encoder.layers.13.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
109
+ "audio_encoder.layers.14.fc1.bias": "model-00001-of-00003.safetensors",
110
+ "audio_encoder.layers.14.fc1.weight": "model-00001-of-00003.safetensors",
111
+ "audio_encoder.layers.14.fc2.bias": "model-00001-of-00003.safetensors",
112
+ "audio_encoder.layers.14.fc2.weight": "model-00001-of-00003.safetensors",
113
+ "audio_encoder.layers.14.final_layer_norm.bias": "model-00001-of-00003.safetensors",
114
+ "audio_encoder.layers.14.final_layer_norm.weight": "model-00001-of-00003.safetensors",
115
+ "audio_encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
116
+ "audio_encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
117
+ "audio_encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
118
+ "audio_encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
119
+ "audio_encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
120
+ "audio_encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
121
+ "audio_encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
122
+ "audio_encoder.layers.14.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
123
+ "audio_encoder.layers.14.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
124
+ "audio_encoder.layers.15.fc1.bias": "model-00001-of-00003.safetensors",
125
+ "audio_encoder.layers.15.fc1.weight": "model-00001-of-00003.safetensors",
126
+ "audio_encoder.layers.15.fc2.bias": "model-00001-of-00003.safetensors",
127
+ "audio_encoder.layers.15.fc2.weight": "model-00001-of-00003.safetensors",
128
+ "audio_encoder.layers.15.final_layer_norm.bias": "model-00001-of-00003.safetensors",
129
+ "audio_encoder.layers.15.final_layer_norm.weight": "model-00001-of-00003.safetensors",
130
+ "audio_encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
131
+ "audio_encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
132
+ "audio_encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
133
+ "audio_encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
134
+ "audio_encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
135
+ "audio_encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
136
+ "audio_encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
137
+ "audio_encoder.layers.15.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
138
+ "audio_encoder.layers.15.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
139
+ "audio_encoder.layers.16.fc1.bias": "model-00001-of-00003.safetensors",
140
+ "audio_encoder.layers.16.fc1.weight": "model-00001-of-00003.safetensors",
141
+ "audio_encoder.layers.16.fc2.bias": "model-00001-of-00003.safetensors",
142
+ "audio_encoder.layers.16.fc2.weight": "model-00001-of-00003.safetensors",
143
+ "audio_encoder.layers.16.final_layer_norm.bias": "model-00001-of-00003.safetensors",
144
+ "audio_encoder.layers.16.final_layer_norm.weight": "model-00001-of-00003.safetensors",
145
+ "audio_encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
146
+ "audio_encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
147
+ "audio_encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
148
+ "audio_encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
149
+ "audio_encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
150
+ "audio_encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
151
+ "audio_encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
152
+ "audio_encoder.layers.16.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
153
+ "audio_encoder.layers.16.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
154
+ "audio_encoder.layers.17.fc1.bias": "model-00001-of-00003.safetensors",
155
+ "audio_encoder.layers.17.fc1.weight": "model-00001-of-00003.safetensors",
156
+ "audio_encoder.layers.17.fc2.bias": "model-00001-of-00003.safetensors",
157
+ "audio_encoder.layers.17.fc2.weight": "model-00001-of-00003.safetensors",
158
+ "audio_encoder.layers.17.final_layer_norm.bias": "model-00001-of-00003.safetensors",
159
+ "audio_encoder.layers.17.final_layer_norm.weight": "model-00001-of-00003.safetensors",
160
+ "audio_encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
161
+ "audio_encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
162
+ "audio_encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
163
+ "audio_encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
164
+ "audio_encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
165
+ "audio_encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
166
+ "audio_encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
167
+ "audio_encoder.layers.17.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
168
+ "audio_encoder.layers.17.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
169
+ "audio_encoder.layers.18.fc1.bias": "model-00001-of-00003.safetensors",
170
+ "audio_encoder.layers.18.fc1.weight": "model-00001-of-00003.safetensors",
171
+ "audio_encoder.layers.18.fc2.bias": "model-00001-of-00003.safetensors",
172
+ "audio_encoder.layers.18.fc2.weight": "model-00001-of-00003.safetensors",
173
+ "audio_encoder.layers.18.final_layer_norm.bias": "model-00001-of-00003.safetensors",
174
+ "audio_encoder.layers.18.final_layer_norm.weight": "model-00001-of-00003.safetensors",
175
+ "audio_encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
176
+ "audio_encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
177
+ "audio_encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
178
+ "audio_encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
179
+ "audio_encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
180
+ "audio_encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
181
+ "audio_encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
182
+ "audio_encoder.layers.18.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
183
+ "audio_encoder.layers.18.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
184
+ "audio_encoder.layers.19.fc1.bias": "model-00001-of-00003.safetensors",
185
+ "audio_encoder.layers.19.fc1.weight": "model-00001-of-00003.safetensors",
186
+ "audio_encoder.layers.19.fc2.bias": "model-00001-of-00003.safetensors",
187
+ "audio_encoder.layers.19.fc2.weight": "model-00001-of-00003.safetensors",
188
+ "audio_encoder.layers.19.final_layer_norm.bias": "model-00001-of-00003.safetensors",
189
+ "audio_encoder.layers.19.final_layer_norm.weight": "model-00001-of-00003.safetensors",
190
+ "audio_encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
191
+ "audio_encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
192
+ "audio_encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
193
+ "audio_encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
194
+ "audio_encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
195
+ "audio_encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
196
+ "audio_encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
197
+ "audio_encoder.layers.19.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
198
+ "audio_encoder.layers.19.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
199
+ "audio_encoder.layers.2.fc1.bias": "model-00001-of-00003.safetensors",
200
+ "audio_encoder.layers.2.fc1.weight": "model-00001-of-00003.safetensors",
201
+ "audio_encoder.layers.2.fc2.bias": "model-00001-of-00003.safetensors",
202
+ "audio_encoder.layers.2.fc2.weight": "model-00001-of-00003.safetensors",
203
+ "audio_encoder.layers.2.final_layer_norm.bias": "model-00001-of-00003.safetensors",
204
+ "audio_encoder.layers.2.final_layer_norm.weight": "model-00001-of-00003.safetensors",
205
+ "audio_encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
206
+ "audio_encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
207
+ "audio_encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
208
+ "audio_encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
209
+ "audio_encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
210
+ "audio_encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
211
+ "audio_encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
212
+ "audio_encoder.layers.2.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
213
+ "audio_encoder.layers.2.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
214
+ "audio_encoder.layers.20.fc1.bias": "model-00001-of-00003.safetensors",
215
+ "audio_encoder.layers.20.fc1.weight": "model-00001-of-00003.safetensors",
216
+ "audio_encoder.layers.20.fc2.bias": "model-00001-of-00003.safetensors",
217
+ "audio_encoder.layers.20.fc2.weight": "model-00001-of-00003.safetensors",
218
+ "audio_encoder.layers.20.final_layer_norm.bias": "model-00001-of-00003.safetensors",
219
+ "audio_encoder.layers.20.final_layer_norm.weight": "model-00001-of-00003.safetensors",
220
+ "audio_encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
221
+ "audio_encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
222
+ "audio_encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
223
+ "audio_encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
224
+ "audio_encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
225
+ "audio_encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
226
+ "audio_encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
227
+ "audio_encoder.layers.20.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
228
+ "audio_encoder.layers.20.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
229
+ "audio_encoder.layers.21.fc1.bias": "model-00001-of-00003.safetensors",
230
+ "audio_encoder.layers.21.fc1.weight": "model-00001-of-00003.safetensors",
231
+ "audio_encoder.layers.21.fc2.bias": "model-00001-of-00003.safetensors",
232
+ "audio_encoder.layers.21.fc2.weight": "model-00001-of-00003.safetensors",
233
+ "audio_encoder.layers.21.final_layer_norm.bias": "model-00001-of-00003.safetensors",
234
+ "audio_encoder.layers.21.final_layer_norm.weight": "model-00001-of-00003.safetensors",
235
+ "audio_encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
236
+ "audio_encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
237
+ "audio_encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
238
+ "audio_encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
239
+ "audio_encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
240
+ "audio_encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
241
+ "audio_encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
242
+ "audio_encoder.layers.21.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
243
+ "audio_encoder.layers.21.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
244
+ "audio_encoder.layers.22.fc1.bias": "model-00001-of-00003.safetensors",
245
+ "audio_encoder.layers.22.fc1.weight": "model-00001-of-00003.safetensors",
246
+ "audio_encoder.layers.22.fc2.bias": "model-00001-of-00003.safetensors",
247
+ "audio_encoder.layers.22.fc2.weight": "model-00001-of-00003.safetensors",
248
+ "audio_encoder.layers.22.final_layer_norm.bias": "model-00001-of-00003.safetensors",
249
+ "audio_encoder.layers.22.final_layer_norm.weight": "model-00001-of-00003.safetensors",
250
+ "audio_encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
251
+ "audio_encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
252
+ "audio_encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
253
+ "audio_encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
254
+ "audio_encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
255
+ "audio_encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
256
+ "audio_encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
257
+ "audio_encoder.layers.22.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
258
+ "audio_encoder.layers.22.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
259
+ "audio_encoder.layers.23.fc1.bias": "model-00001-of-00003.safetensors",
260
+ "audio_encoder.layers.23.fc1.weight": "model-00001-of-00003.safetensors",
261
+ "audio_encoder.layers.23.fc2.bias": "model-00001-of-00003.safetensors",
262
+ "audio_encoder.layers.23.fc2.weight": "model-00001-of-00003.safetensors",
263
+ "audio_encoder.layers.23.final_layer_norm.bias": "model-00001-of-00003.safetensors",
264
+ "audio_encoder.layers.23.final_layer_norm.weight": "model-00001-of-00003.safetensors",
265
+ "audio_encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
266
+ "audio_encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
267
+ "audio_encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
268
+ "audio_encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
269
+ "audio_encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
270
+ "audio_encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
271
+ "audio_encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
272
+ "audio_encoder.layers.23.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
273
+ "audio_encoder.layers.23.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
274
+ "audio_encoder.layers.24.fc1.bias": "model-00001-of-00003.safetensors",
275
+ "audio_encoder.layers.24.fc1.weight": "model-00001-of-00003.safetensors",
276
+ "audio_encoder.layers.24.fc2.bias": "model-00001-of-00003.safetensors",
277
+ "audio_encoder.layers.24.fc2.weight": "model-00001-of-00003.safetensors",
278
+ "audio_encoder.layers.24.final_layer_norm.bias": "model-00001-of-00003.safetensors",
279
+ "audio_encoder.layers.24.final_layer_norm.weight": "model-00001-of-00003.safetensors",
280
+ "audio_encoder.layers.24.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
281
+ "audio_encoder.layers.24.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
282
+ "audio_encoder.layers.24.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
283
+ "audio_encoder.layers.24.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
284
+ "audio_encoder.layers.24.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
285
+ "audio_encoder.layers.24.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
286
+ "audio_encoder.layers.24.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
287
+ "audio_encoder.layers.24.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
288
+ "audio_encoder.layers.24.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
289
+ "audio_encoder.layers.25.fc1.bias": "model-00001-of-00003.safetensors",
290
+ "audio_encoder.layers.25.fc1.weight": "model-00001-of-00003.safetensors",
291
+ "audio_encoder.layers.25.fc2.bias": "model-00001-of-00003.safetensors",
292
+ "audio_encoder.layers.25.fc2.weight": "model-00001-of-00003.safetensors",
293
+ "audio_encoder.layers.25.final_layer_norm.bias": "model-00001-of-00003.safetensors",
294
+ "audio_encoder.layers.25.final_layer_norm.weight": "model-00001-of-00003.safetensors",
295
+ "audio_encoder.layers.25.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
296
+ "audio_encoder.layers.25.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
297
+ "audio_encoder.layers.25.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
298
+ "audio_encoder.layers.25.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
299
+ "audio_encoder.layers.25.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
300
+ "audio_encoder.layers.25.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
301
+ "audio_encoder.layers.25.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
302
+ "audio_encoder.layers.25.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
303
+ "audio_encoder.layers.25.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
304
+ "audio_encoder.layers.26.fc1.bias": "model-00001-of-00003.safetensors",
305
+ "audio_encoder.layers.26.fc1.weight": "model-00001-of-00003.safetensors",
306
+ "audio_encoder.layers.26.fc2.bias": "model-00001-of-00003.safetensors",
307
+ "audio_encoder.layers.26.fc2.weight": "model-00001-of-00003.safetensors",
308
+ "audio_encoder.layers.26.final_layer_norm.bias": "model-00001-of-00003.safetensors",
309
+ "audio_encoder.layers.26.final_layer_norm.weight": "model-00001-of-00003.safetensors",
310
+ "audio_encoder.layers.26.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
311
+ "audio_encoder.layers.26.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
312
+ "audio_encoder.layers.26.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
313
+ "audio_encoder.layers.26.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
314
+ "audio_encoder.layers.26.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
315
+ "audio_encoder.layers.26.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
316
+ "audio_encoder.layers.26.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
317
+ "audio_encoder.layers.26.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
318
+ "audio_encoder.layers.26.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
319
+ "audio_encoder.layers.27.fc1.bias": "model-00001-of-00003.safetensors",
320
+ "audio_encoder.layers.27.fc1.weight": "model-00001-of-00003.safetensors",
321
+ "audio_encoder.layers.27.fc2.bias": "model-00001-of-00003.safetensors",
322
+ "audio_encoder.layers.27.fc2.weight": "model-00001-of-00003.safetensors",
323
+ "audio_encoder.layers.27.final_layer_norm.bias": "model-00001-of-00003.safetensors",
324
+ "audio_encoder.layers.27.final_layer_norm.weight": "model-00001-of-00003.safetensors",
325
+ "audio_encoder.layers.27.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
326
+ "audio_encoder.layers.27.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
327
+ "audio_encoder.layers.27.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
328
+ "audio_encoder.layers.27.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
329
+ "audio_encoder.layers.27.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
330
+ "audio_encoder.layers.27.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
331
+ "audio_encoder.layers.27.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
332
+ "audio_encoder.layers.27.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
333
+ "audio_encoder.layers.27.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
334
+ "audio_encoder.layers.28.fc1.bias": "model-00001-of-00003.safetensors",
335
+ "audio_encoder.layers.28.fc1.weight": "model-00001-of-00003.safetensors",
336
+ "audio_encoder.layers.28.fc2.bias": "model-00001-of-00003.safetensors",
337
+ "audio_encoder.layers.28.fc2.weight": "model-00001-of-00003.safetensors",
338
+ "audio_encoder.layers.28.final_layer_norm.bias": "model-00001-of-00003.safetensors",
339
+ "audio_encoder.layers.28.final_layer_norm.weight": "model-00001-of-00003.safetensors",
340
+ "audio_encoder.layers.28.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
341
+ "audio_encoder.layers.28.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
342
+ "audio_encoder.layers.28.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
343
+ "audio_encoder.layers.28.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
344
+ "audio_encoder.layers.28.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
345
+ "audio_encoder.layers.28.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
346
+ "audio_encoder.layers.28.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
347
+ "audio_encoder.layers.28.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
348
+ "audio_encoder.layers.28.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
349
+ "audio_encoder.layers.29.fc1.bias": "model-00001-of-00003.safetensors",
350
+ "audio_encoder.layers.29.fc1.weight": "model-00001-of-00003.safetensors",
351
+ "audio_encoder.layers.29.fc2.bias": "model-00001-of-00003.safetensors",
352
+ "audio_encoder.layers.29.fc2.weight": "model-00001-of-00003.safetensors",
353
+ "audio_encoder.layers.29.final_layer_norm.bias": "model-00001-of-00003.safetensors",
354
+ "audio_encoder.layers.29.final_layer_norm.weight": "model-00001-of-00003.safetensors",
355
+ "audio_encoder.layers.29.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
356
+ "audio_encoder.layers.29.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
357
+ "audio_encoder.layers.29.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
358
+ "audio_encoder.layers.29.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
359
+ "audio_encoder.layers.29.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
360
+ "audio_encoder.layers.29.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
361
+ "audio_encoder.layers.29.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
362
+ "audio_encoder.layers.29.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
363
+ "audio_encoder.layers.29.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
364
+ "audio_encoder.layers.3.fc1.bias": "model-00001-of-00003.safetensors",
365
+ "audio_encoder.layers.3.fc1.weight": "model-00001-of-00003.safetensors",
366
+ "audio_encoder.layers.3.fc2.bias": "model-00001-of-00003.safetensors",
367
+ "audio_encoder.layers.3.fc2.weight": "model-00001-of-00003.safetensors",
368
+ "audio_encoder.layers.3.final_layer_norm.bias": "model-00001-of-00003.safetensors",
369
+ "audio_encoder.layers.3.final_layer_norm.weight": "model-00001-of-00003.safetensors",
370
+ "audio_encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
371
+ "audio_encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
372
+ "audio_encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
373
+ "audio_encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
374
+ "audio_encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
375
+ "audio_encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
376
+ "audio_encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
377
+ "audio_encoder.layers.3.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
378
+ "audio_encoder.layers.3.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
379
+ "audio_encoder.layers.30.fc1.bias": "model-00001-of-00003.safetensors",
380
+ "audio_encoder.layers.30.fc1.weight": "model-00001-of-00003.safetensors",
381
+ "audio_encoder.layers.30.fc2.bias": "model-00001-of-00003.safetensors",
382
+ "audio_encoder.layers.30.fc2.weight": "model-00001-of-00003.safetensors",
383
+ "audio_encoder.layers.30.final_layer_norm.bias": "model-00001-of-00003.safetensors",
384
+ "audio_encoder.layers.30.final_layer_norm.weight": "model-00001-of-00003.safetensors",
385
+ "audio_encoder.layers.30.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
386
+ "audio_encoder.layers.30.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
387
+ "audio_encoder.layers.30.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
388
+ "audio_encoder.layers.30.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
389
+ "audio_encoder.layers.30.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
390
+ "audio_encoder.layers.30.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
391
+ "audio_encoder.layers.30.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
392
+ "audio_encoder.layers.30.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
393
+ "audio_encoder.layers.30.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
394
+ "audio_encoder.layers.31.fc1.bias": "model-00001-of-00003.safetensors",
395
+ "audio_encoder.layers.31.fc1.weight": "model-00001-of-00003.safetensors",
396
+ "audio_encoder.layers.31.fc2.bias": "model-00001-of-00003.safetensors",
397
+ "audio_encoder.layers.31.fc2.weight": "model-00001-of-00003.safetensors",
398
+ "audio_encoder.layers.31.final_layer_norm.bias": "model-00001-of-00003.safetensors",
399
+ "audio_encoder.layers.31.final_layer_norm.weight": "model-00001-of-00003.safetensors",
400
+ "audio_encoder.layers.31.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
401
+ "audio_encoder.layers.31.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
402
+ "audio_encoder.layers.31.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
403
+ "audio_encoder.layers.31.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
404
+ "audio_encoder.layers.31.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
405
+ "audio_encoder.layers.31.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
406
+ "audio_encoder.layers.31.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
407
+ "audio_encoder.layers.31.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
408
+ "audio_encoder.layers.31.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
409
+ "audio_encoder.layers.4.fc1.bias": "model-00001-of-00003.safetensors",
410
+ "audio_encoder.layers.4.fc1.weight": "model-00001-of-00003.safetensors",
411
+ "audio_encoder.layers.4.fc2.bias": "model-00001-of-00003.safetensors",
412
+ "audio_encoder.layers.4.fc2.weight": "model-00001-of-00003.safetensors",
413
+ "audio_encoder.layers.4.final_layer_norm.bias": "model-00001-of-00003.safetensors",
414
+ "audio_encoder.layers.4.final_layer_norm.weight": "model-00001-of-00003.safetensors",
415
+ "audio_encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
416
+ "audio_encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
417
+ "audio_encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
418
+ "audio_encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
419
+ "audio_encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
420
+ "audio_encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
421
+ "audio_encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
422
+ "audio_encoder.layers.4.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
423
+ "audio_encoder.layers.4.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
424
+ "audio_encoder.layers.5.fc1.bias": "model-00001-of-00003.safetensors",
425
+ "audio_encoder.layers.5.fc1.weight": "model-00001-of-00003.safetensors",
426
+ "audio_encoder.layers.5.fc2.bias": "model-00001-of-00003.safetensors",
427
+ "audio_encoder.layers.5.fc2.weight": "model-00001-of-00003.safetensors",
428
+ "audio_encoder.layers.5.final_layer_norm.bias": "model-00001-of-00003.safetensors",
429
+ "audio_encoder.layers.5.final_layer_norm.weight": "model-00001-of-00003.safetensors",
430
+ "audio_encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
431
+ "audio_encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
432
+ "audio_encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
433
+ "audio_encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
434
+ "audio_encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
435
+ "audio_encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
436
+ "audio_encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
437
+ "audio_encoder.layers.5.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
438
+ "audio_encoder.layers.5.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
439
+ "audio_encoder.layers.6.fc1.bias": "model-00001-of-00003.safetensors",
440
+ "audio_encoder.layers.6.fc1.weight": "model-00001-of-00003.safetensors",
441
+ "audio_encoder.layers.6.fc2.bias": "model-00001-of-00003.safetensors",
442
+ "audio_encoder.layers.6.fc2.weight": "model-00001-of-00003.safetensors",
443
+ "audio_encoder.layers.6.final_layer_norm.bias": "model-00001-of-00003.safetensors",
444
+ "audio_encoder.layers.6.final_layer_norm.weight": "model-00001-of-00003.safetensors",
445
+ "audio_encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
446
+ "audio_encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
447
+ "audio_encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
448
+ "audio_encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
449
+ "audio_encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
450
+ "audio_encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
451
+ "audio_encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
452
+ "audio_encoder.layers.6.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
453
+ "audio_encoder.layers.6.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
454
+ "audio_encoder.layers.7.fc1.bias": "model-00001-of-00003.safetensors",
455
+ "audio_encoder.layers.7.fc1.weight": "model-00001-of-00003.safetensors",
456
+ "audio_encoder.layers.7.fc2.bias": "model-00001-of-00003.safetensors",
457
+ "audio_encoder.layers.7.fc2.weight": "model-00001-of-00003.safetensors",
458
+ "audio_encoder.layers.7.final_layer_norm.bias": "model-00001-of-00003.safetensors",
459
+ "audio_encoder.layers.7.final_layer_norm.weight": "model-00001-of-00003.safetensors",
460
+ "audio_encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
461
+ "audio_encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
462
+ "audio_encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
463
+ "audio_encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
464
+ "audio_encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
465
+ "audio_encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
466
+ "audio_encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
467
+ "audio_encoder.layers.7.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
468
+ "audio_encoder.layers.7.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
469
+ "audio_encoder.layers.8.fc1.bias": "model-00001-of-00003.safetensors",
470
+ "audio_encoder.layers.8.fc1.weight": "model-00001-of-00003.safetensors",
471
+ "audio_encoder.layers.8.fc2.bias": "model-00001-of-00003.safetensors",
472
+ "audio_encoder.layers.8.fc2.weight": "model-00001-of-00003.safetensors",
473
+ "audio_encoder.layers.8.final_layer_norm.bias": "model-00001-of-00003.safetensors",
474
+ "audio_encoder.layers.8.final_layer_norm.weight": "model-00001-of-00003.safetensors",
475
+ "audio_encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
476
+ "audio_encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
477
+ "audio_encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
478
+ "audio_encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
479
+ "audio_encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
480
+ "audio_encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
481
+ "audio_encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
482
+ "audio_encoder.layers.8.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
483
+ "audio_encoder.layers.8.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
484
+ "audio_encoder.layers.9.fc1.bias": "model-00001-of-00003.safetensors",
485
+ "audio_encoder.layers.9.fc1.weight": "model-00001-of-00003.safetensors",
486
+ "audio_encoder.layers.9.fc2.bias": "model-00001-of-00003.safetensors",
487
+ "audio_encoder.layers.9.fc2.weight": "model-00001-of-00003.safetensors",
488
+ "audio_encoder.layers.9.final_layer_norm.bias": "model-00001-of-00003.safetensors",
489
+ "audio_encoder.layers.9.final_layer_norm.weight": "model-00001-of-00003.safetensors",
490
+ "audio_encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
491
+ "audio_encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00003.safetensors",
492
+ "audio_encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00003.safetensors",
493
+ "audio_encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00003.safetensors",
494
+ "audio_encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
495
+ "audio_encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00003.safetensors",
496
+ "audio_encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
497
+ "audio_encoder.layers.9.self_attn_layer_norm.bias": "model-00001-of-00003.safetensors",
498
+ "audio_encoder.layers.9.self_attn_layer_norm.weight": "model-00001-of-00003.safetensors",
499
+ "audio_encoder.stem_proj.bias": "model-00001-of-00003.safetensors",
500
+ "audio_encoder.stem_proj.weight": "model-00001-of-00003.safetensors",
501
+ "deepstack_audio_merger_list.0.down_proj.weight": "model-00002-of-00003.safetensors",
502
+ "deepstack_audio_merger_list.0.gate_proj.weight": "model-00002-of-00003.safetensors",
503
+ "deepstack_audio_merger_list.0.up_proj.weight": "model-00002-of-00003.safetensors",
504
+ "deepstack_audio_merger_list.1.down_proj.weight": "model-00002-of-00003.safetensors",
505
+ "deepstack_audio_merger_list.1.gate_proj.weight": "model-00002-of-00003.safetensors",
506
+ "deepstack_audio_merger_list.1.up_proj.weight": "model-00002-of-00003.safetensors",
507
+ "deepstack_audio_merger_list.2.down_proj.weight": "model-00002-of-00003.safetensors",
508
+ "deepstack_audio_merger_list.2.gate_proj.weight": "model-00002-of-00003.safetensors",
509
+ "deepstack_audio_merger_list.2.up_proj.weight": "model-00002-of-00003.safetensors",
510
+ "language_model.embed_tokens.weight": "model-00001-of-00003.safetensors",
511
+ "language_model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
512
+ "language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
513
+ "language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
514
+ "language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
515
+ "language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
516
+ "language_model.layers.0.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
517
+ "language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
518
+ "language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
519
+ "language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
520
+ "language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
521
+ "language_model.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
522
+ "language_model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
523
+ "language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
524
+ "language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
525
+ "language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
526
+ "language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
527
+ "language_model.layers.1.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
528
+ "language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
529
+ "language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
530
+ "language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
531
+ "language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
532
+ "language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
533
+ "language_model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors",
534
+ "language_model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
535
+ "language_model.layers.10.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
536
+ "language_model.layers.10.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
537
+ "language_model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
538
+ "language_model.layers.10.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
539
+ "language_model.layers.10.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
540
+ "language_model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
541
+ "language_model.layers.10.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
542
+ "language_model.layers.10.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
543
+ "language_model.layers.10.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
544
+ "language_model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors",
545
+ "language_model.layers.11.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
546
+ "language_model.layers.11.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
547
+ "language_model.layers.11.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
548
+ "language_model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
549
+ "language_model.layers.11.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
550
+ "language_model.layers.11.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
551
+ "language_model.layers.11.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
552
+ "language_model.layers.11.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
553
+ "language_model.layers.11.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
554
+ "language_model.layers.11.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
555
+ "language_model.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors",
556
+ "language_model.layers.12.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
557
+ "language_model.layers.12.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
558
+ "language_model.layers.12.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
559
+ "language_model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
560
+ "language_model.layers.12.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
561
+ "language_model.layers.12.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
562
+ "language_model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
563
+ "language_model.layers.12.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
564
+ "language_model.layers.12.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
565
+ "language_model.layers.12.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
566
+ "language_model.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors",
567
+ "language_model.layers.13.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
568
+ "language_model.layers.13.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
569
+ "language_model.layers.13.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
570
+ "language_model.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
571
+ "language_model.layers.13.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
572
+ "language_model.layers.13.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
573
+ "language_model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
574
+ "language_model.layers.13.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
575
+ "language_model.layers.13.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
576
+ "language_model.layers.13.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
577
+ "language_model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors",
578
+ "language_model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
579
+ "language_model.layers.14.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
580
+ "language_model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
581
+ "language_model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
582
+ "language_model.layers.14.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
583
+ "language_model.layers.14.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
584
+ "language_model.layers.14.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
585
+ "language_model.layers.14.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
586
+ "language_model.layers.14.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
587
+ "language_model.layers.14.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
588
+ "language_model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors",
589
+ "language_model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
590
+ "language_model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
591
+ "language_model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
592
+ "language_model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
593
+ "language_model.layers.15.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
594
+ "language_model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
595
+ "language_model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
596
+ "language_model.layers.15.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
597
+ "language_model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
598
+ "language_model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
599
+ "language_model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors",
600
+ "language_model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
601
+ "language_model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
602
+ "language_model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
603
+ "language_model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
604
+ "language_model.layers.16.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
605
+ "language_model.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
606
+ "language_model.layers.16.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
607
+ "language_model.layers.16.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
608
+ "language_model.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
609
+ "language_model.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
610
+ "language_model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors",
611
+ "language_model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
612
+ "language_model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
613
+ "language_model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
614
+ "language_model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
615
+ "language_model.layers.17.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
616
+ "language_model.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
617
+ "language_model.layers.17.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
618
+ "language_model.layers.17.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
619
+ "language_model.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
620
+ "language_model.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
621
+ "language_model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors",
622
+ "language_model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
623
+ "language_model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
624
+ "language_model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
625
+ "language_model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
626
+ "language_model.layers.18.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
627
+ "language_model.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
628
+ "language_model.layers.18.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
629
+ "language_model.layers.18.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
630
+ "language_model.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
631
+ "language_model.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
632
+ "language_model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors",
633
+ "language_model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
634
+ "language_model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
635
+ "language_model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
636
+ "language_model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
637
+ "language_model.layers.19.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
638
+ "language_model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
639
+ "language_model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
640
+ "language_model.layers.19.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
641
+ "language_model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
642
+ "language_model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
643
+ "language_model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
644
+ "language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
645
+ "language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
646
+ "language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
647
+ "language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
648
+ "language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
649
+ "language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
650
+ "language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
651
+ "language_model.layers.2.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
652
+ "language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
653
+ "language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
654
+ "language_model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors",
655
+ "language_model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
656
+ "language_model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
657
+ "language_model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
658
+ "language_model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
659
+ "language_model.layers.20.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
660
+ "language_model.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
661
+ "language_model.layers.20.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
662
+ "language_model.layers.20.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
663
+ "language_model.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
664
+ "language_model.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
665
+ "language_model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors",
666
+ "language_model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
667
+ "language_model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
668
+ "language_model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
669
+ "language_model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
670
+ "language_model.layers.21.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
671
+ "language_model.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
672
+ "language_model.layers.21.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
673
+ "language_model.layers.21.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
674
+ "language_model.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
675
+ "language_model.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
676
+ "language_model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors",
677
+ "language_model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
678
+ "language_model.layers.22.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
679
+ "language_model.layers.22.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
680
+ "language_model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
681
+ "language_model.layers.22.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
682
+ "language_model.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
683
+ "language_model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
684
+ "language_model.layers.22.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
685
+ "language_model.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
686
+ "language_model.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
687
+ "language_model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors",
688
+ "language_model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
689
+ "language_model.layers.23.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
690
+ "language_model.layers.23.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
691
+ "language_model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
692
+ "language_model.layers.23.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
693
+ "language_model.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
694
+ "language_model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
695
+ "language_model.layers.23.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
696
+ "language_model.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
697
+ "language_model.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
698
+ "language_model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors",
699
+ "language_model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
700
+ "language_model.layers.24.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
701
+ "language_model.layers.24.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
702
+ "language_model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
703
+ "language_model.layers.24.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
704
+ "language_model.layers.24.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
705
+ "language_model.layers.24.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
706
+ "language_model.layers.24.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
707
+ "language_model.layers.24.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
708
+ "language_model.layers.24.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
709
+ "language_model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors",
710
+ "language_model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
711
+ "language_model.layers.25.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
712
+ "language_model.layers.25.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
713
+ "language_model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
714
+ "language_model.layers.25.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
715
+ "language_model.layers.25.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
716
+ "language_model.layers.25.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
717
+ "language_model.layers.25.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
718
+ "language_model.layers.25.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
719
+ "language_model.layers.25.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
720
+ "language_model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors",
721
+ "language_model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
722
+ "language_model.layers.26.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
723
+ "language_model.layers.26.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
724
+ "language_model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
725
+ "language_model.layers.26.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
726
+ "language_model.layers.26.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
727
+ "language_model.layers.26.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
728
+ "language_model.layers.26.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
729
+ "language_model.layers.26.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
730
+ "language_model.layers.26.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
731
+ "language_model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors",
732
+ "language_model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
733
+ "language_model.layers.27.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
734
+ "language_model.layers.27.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
735
+ "language_model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
736
+ "language_model.layers.27.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
737
+ "language_model.layers.27.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
738
+ "language_model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
739
+ "language_model.layers.27.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
740
+ "language_model.layers.27.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
741
+ "language_model.layers.27.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
742
+ "language_model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors",
743
+ "language_model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
744
+ "language_model.layers.28.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
745
+ "language_model.layers.28.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
746
+ "language_model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
747
+ "language_model.layers.28.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
748
+ "language_model.layers.28.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
749
+ "language_model.layers.28.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
750
+ "language_model.layers.28.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
751
+ "language_model.layers.28.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
752
+ "language_model.layers.28.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
753
+ "language_model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors",
754
+ "language_model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
755
+ "language_model.layers.29.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
756
+ "language_model.layers.29.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
757
+ "language_model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
758
+ "language_model.layers.29.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
759
+ "language_model.layers.29.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
760
+ "language_model.layers.29.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
761
+ "language_model.layers.29.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
762
+ "language_model.layers.29.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
763
+ "language_model.layers.29.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
764
+ "language_model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
765
+ "language_model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
766
+ "language_model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
767
+ "language_model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
768
+ "language_model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
769
+ "language_model.layers.3.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
770
+ "language_model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
771
+ "language_model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
772
+ "language_model.layers.3.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
773
+ "language_model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
774
+ "language_model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
775
+ "language_model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors",
776
+ "language_model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
777
+ "language_model.layers.30.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
778
+ "language_model.layers.30.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
779
+ "language_model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
780
+ "language_model.layers.30.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
781
+ "language_model.layers.30.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
782
+ "language_model.layers.30.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
783
+ "language_model.layers.30.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
784
+ "language_model.layers.30.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
785
+ "language_model.layers.30.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
786
+ "language_model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors",
787
+ "language_model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
788
+ "language_model.layers.31.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
789
+ "language_model.layers.31.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
790
+ "language_model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
791
+ "language_model.layers.31.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
792
+ "language_model.layers.31.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
793
+ "language_model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
794
+ "language_model.layers.31.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
795
+ "language_model.layers.31.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
796
+ "language_model.layers.31.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
797
+ "language_model.layers.32.input_layernorm.weight": "model-00002-of-00003.safetensors",
798
+ "language_model.layers.32.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
799
+ "language_model.layers.32.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
800
+ "language_model.layers.32.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
801
+ "language_model.layers.32.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
802
+ "language_model.layers.32.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
803
+ "language_model.layers.32.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
804
+ "language_model.layers.32.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
805
+ "language_model.layers.32.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
806
+ "language_model.layers.32.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
807
+ "language_model.layers.32.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
808
+ "language_model.layers.33.input_layernorm.weight": "model-00002-of-00003.safetensors",
809
+ "language_model.layers.33.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
810
+ "language_model.layers.33.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
811
+ "language_model.layers.33.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
812
+ "language_model.layers.33.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
813
+ "language_model.layers.33.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
814
+ "language_model.layers.33.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
815
+ "language_model.layers.33.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
816
+ "language_model.layers.33.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
817
+ "language_model.layers.33.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
818
+ "language_model.layers.33.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
819
+ "language_model.layers.34.input_layernorm.weight": "model-00002-of-00003.safetensors",
820
+ "language_model.layers.34.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
821
+ "language_model.layers.34.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
822
+ "language_model.layers.34.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
823
+ "language_model.layers.34.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
824
+ "language_model.layers.34.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
825
+ "language_model.layers.34.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
826
+ "language_model.layers.34.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
827
+ "language_model.layers.34.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
828
+ "language_model.layers.34.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
829
+ "language_model.layers.34.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
830
+ "language_model.layers.35.input_layernorm.weight": "model-00002-of-00003.safetensors",
831
+ "language_model.layers.35.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
832
+ "language_model.layers.35.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
833
+ "language_model.layers.35.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
834
+ "language_model.layers.35.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
835
+ "language_model.layers.35.self_attn.k_norm.weight": "model-00002-of-00003.safetensors",
836
+ "language_model.layers.35.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
837
+ "language_model.layers.35.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
838
+ "language_model.layers.35.self_attn.q_norm.weight": "model-00002-of-00003.safetensors",
839
+ "language_model.layers.35.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
840
+ "language_model.layers.35.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
841
+ "language_model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
842
+ "language_model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
843
+ "language_model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
844
+ "language_model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
845
+ "language_model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
846
+ "language_model.layers.4.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
847
+ "language_model.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
848
+ "language_model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
849
+ "language_model.layers.4.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
850
+ "language_model.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
851
+ "language_model.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
852
+ "language_model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
853
+ "language_model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
854
+ "language_model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
855
+ "language_model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
856
+ "language_model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
857
+ "language_model.layers.5.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
858
+ "language_model.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
859
+ "language_model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
860
+ "language_model.layers.5.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
861
+ "language_model.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
862
+ "language_model.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
863
+ "language_model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors",
864
+ "language_model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
865
+ "language_model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
866
+ "language_model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
867
+ "language_model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
868
+ "language_model.layers.6.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
869
+ "language_model.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
870
+ "language_model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
871
+ "language_model.layers.6.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
872
+ "language_model.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
873
+ "language_model.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
874
+ "language_model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors",
875
+ "language_model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
876
+ "language_model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
877
+ "language_model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
878
+ "language_model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
879
+ "language_model.layers.7.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
880
+ "language_model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
881
+ "language_model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
882
+ "language_model.layers.7.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
883
+ "language_model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
884
+ "language_model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
885
+ "language_model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors",
886
+ "language_model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
887
+ "language_model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
888
+ "language_model.layers.8.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
889
+ "language_model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
890
+ "language_model.layers.8.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
891
+ "language_model.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
892
+ "language_model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
893
+ "language_model.layers.8.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
894
+ "language_model.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
895
+ "language_model.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
896
+ "language_model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors",
897
+ "language_model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
898
+ "language_model.layers.9.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
899
+ "language_model.layers.9.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
900
+ "language_model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
901
+ "language_model.layers.9.self_attn.k_norm.weight": "model-00001-of-00003.safetensors",
902
+ "language_model.layers.9.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
903
+ "language_model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
904
+ "language_model.layers.9.self_attn.q_norm.weight": "model-00001-of-00003.safetensors",
905
+ "language_model.layers.9.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
906
+ "language_model.layers.9.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
907
+ "language_model.norm.weight": "model-00002-of-00003.safetensors",
908
+ "lm_head.weight": "model-00003-of-00003.safetensors"
909
+ }
910
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "processor_class": "MossAudioProcessor",
3
+ "auto_map": {
4
+ "AutoProcessor": "processing_moss_audio.MossAudioProcessor"
5
+ },
6
+ "mel_config": {
7
+ "mel_sr": 16000,
8
+ "mel_dim": 128,
9
+ "mel_n_fft": 400,
10
+ "mel_hop_length": 160,
11
+ "mel_dtype": "bfloat16",
12
+ "use_whisper_feature_extractor": true
13
+ },
14
+ "enable_time_marker": true,
15
+ "audio_token_id": 151654,
16
+ "audio_start_id": 151669,
17
+ "audio_end_id": 151670
18
+ }
processing_moss_audio.py ADDED
@@ -0,0 +1,371 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import re
2
+ import types
3
+ from dataclasses import dataclass
4
+ from typing import List, Optional, Sequence, Union
5
+
6
+ import numpy as np
7
+ import torch
8
+ from transformers import BatchFeature
9
+ from transformers.processing_utils import ProcessorMixin
10
+ from transformers.models.whisper.feature_extraction_whisper import (
11
+ WhisperFeatureExtractor,
12
+ )
13
+
14
+
15
+ @dataclass
16
+ class MelConfig:
17
+ mel_sr: int = 16000
18
+ mel_dim: int = 128
19
+ mel_n_fft: int = 400
20
+ mel_hop_length: int = 160
21
+ mel_dtype: torch.dtype = torch.bfloat16
22
+ use_whisper_feature_extractor: bool = True
23
+
24
+
25
+ def _normalize_mel_config(mel_config) -> dict[str, object]:
26
+ default_config = MelConfig()
27
+ if mel_config is None:
28
+ source = {}
29
+ elif isinstance(mel_config, MelConfig):
30
+ source = {
31
+ key: getattr(mel_config, key) for key in MelConfig.__dataclass_fields__.keys()
32
+ }
33
+ else:
34
+ source = dict(mel_config)
35
+
36
+ normalized = {}
37
+ for key in MelConfig.__dataclass_fields__.keys():
38
+ value = source.get(key, getattr(default_config, key))
39
+ if key == "mel_dtype":
40
+ if isinstance(value, torch.dtype):
41
+ value = str(value).removeprefix("torch.")
42
+ elif isinstance(value, str) and value.startswith("torch."):
43
+ value = value.removeprefix("torch.")
44
+ normalized[key] = value
45
+ return normalized
46
+
47
+
48
+ def _build_mel_config(mel_config_dict: dict[str, object]) -> MelConfig:
49
+ default_config = MelConfig()
50
+
51
+ def _int_value(key: str, default: int) -> int:
52
+ value = mel_config_dict.get(key, default)
53
+ if isinstance(value, bool):
54
+ return int(value)
55
+ if isinstance(value, (int, str)):
56
+ return int(value)
57
+ return default
58
+
59
+ def _bool_value(key: str, default: bool) -> bool:
60
+ value = mel_config_dict.get(key, default)
61
+ if isinstance(value, bool):
62
+ return value
63
+ if isinstance(value, str):
64
+ return value.lower() in {"1", "true", "yes", "on"}
65
+ if isinstance(value, int):
66
+ return bool(value)
67
+ return default
68
+
69
+ mel_dtype_value = mel_config_dict.get("mel_dtype", default_config.mel_dtype)
70
+ if isinstance(mel_dtype_value, str):
71
+ mel_dtype = getattr(torch, mel_dtype_value.removeprefix("torch."))
72
+ elif isinstance(mel_dtype_value, torch.dtype):
73
+ mel_dtype = mel_dtype_value
74
+ else:
75
+ mel_dtype = default_config.mel_dtype
76
+
77
+ return MelConfig(
78
+ mel_sr=_int_value("mel_sr", default_config.mel_sr),
79
+ mel_dim=_int_value("mel_dim", default_config.mel_dim),
80
+ mel_n_fft=_int_value("mel_n_fft", default_config.mel_n_fft),
81
+ mel_hop_length=_int_value("mel_hop_length", default_config.mel_hop_length),
82
+ mel_dtype=mel_dtype,
83
+ use_whisper_feature_extractor=_bool_value(
84
+ "use_whisper_feature_extractor",
85
+ default_config.use_whisper_feature_extractor,
86
+ ),
87
+ )
88
+
89
+
90
+ class MossAudioProcessor(ProcessorMixin):
91
+ attributes = ["tokenizer"]
92
+ tokenizer_class = ("Qwen2Tokenizer", "Qwen2TokenizerFast")
93
+
94
+ _AUDIO_SPAN_RE = re.compile(r"<\|audio_bos\|>(?:<\|AUDIO\|>)+<\|audio_eos\|>")
95
+
96
+ def __init__(
97
+ self,
98
+ tokenizer=None,
99
+ mel_config=None,
100
+ enable_time_marker: bool = False,
101
+ audio_token_id: int = 151654,
102
+ audio_start_id: int = 151669,
103
+ audio_end_id: int = 151670,
104
+ chat_template=None,
105
+ ):
106
+ super().__init__(tokenizer, chat_template=chat_template)
107
+ if tokenizer is None:
108
+ raise ValueError("MossAudioProcessor requires a tokenizer.")
109
+
110
+ self._base_tokenizer = tokenizer
111
+ self.mel_config = _normalize_mel_config(mel_config)
112
+ self.config = _build_mel_config(self.mel_config)
113
+ self.enable_time_marker = bool(enable_time_marker)
114
+ self.audio_token_id = int(audio_token_id)
115
+ self.audio_start_id = int(audio_start_id)
116
+ self.audio_end_id = int(audio_end_id)
117
+ self._whisper_feature_extractor = None
118
+
119
+ alias_map = {
120
+ "<|AUDIO|>": self.audio_token_id,
121
+ "<|audio_bos|>": self.audio_start_id,
122
+ "<|audio_eos|>": self.audio_end_id,
123
+ }
124
+ orig_convert_tokens_to_ids = tokenizer.convert_tokens_to_ids
125
+
126
+ def _patched_convert_tokens_to_ids(tokenizer_self, tokens):
127
+ if isinstance(tokens, (list, tuple)):
128
+ converted = [
129
+ _patched_convert_tokens_to_ids(tokenizer_self, token)
130
+ for token in tokens
131
+ ]
132
+ return converted if isinstance(tokens, list) else tuple(converted)
133
+ if isinstance(tokens, str) and tokens in alias_map:
134
+ return alias_map[tokens]
135
+ return orig_convert_tokens_to_ids(tokens)
136
+
137
+ tokenizer.convert_tokens_to_ids = types.MethodType(
138
+ _patched_convert_tokens_to_ids, tokenizer
139
+ )
140
+
141
+ self._digit_token_ids = {
142
+ "0": 15,
143
+ "1": 16,
144
+ "2": 17,
145
+ "3": 18,
146
+ "4": 19,
147
+ "5": 20,
148
+ "6": 21,
149
+ "7": 22,
150
+ "8": 23,
151
+ "9": 24,
152
+ }
153
+ self.audio_tokens_per_second = 12.5
154
+ self.time_marker_every_seconds = 2
155
+ self.time_marker_every_audio_tokens = int(
156
+ self.audio_tokens_per_second * self.time_marker_every_seconds
157
+ )
158
+
159
+ @property
160
+ def model_input_names(self):
161
+ return [
162
+ "input_ids",
163
+ "attention_mask",
164
+ "audio_data",
165
+ "audio_data_seqlens",
166
+ ]
167
+
168
+ @staticmethod
169
+ def _conv3_downsample_len(raw_mel_len: int) -> int:
170
+ def conv_out_len(length: int) -> int:
171
+ return (length - 1) // 2 + 1
172
+
173
+ length1 = conv_out_len(int(raw_mel_len))
174
+ length2 = conv_out_len(length1)
175
+ length3 = conv_out_len(length2)
176
+ return int(length3)
177
+
178
+ def _get_whisper_feature_extractor(self):
179
+ if self._whisper_feature_extractor is not None:
180
+ return self._whisper_feature_extractor
181
+
182
+ self._whisper_feature_extractor = WhisperFeatureExtractor(
183
+ feature_size=int(self.config.mel_dim),
184
+ sampling_rate=int(self.config.mel_sr),
185
+ hop_length=int(self.config.mel_hop_length),
186
+ n_fft=int(self.config.mel_n_fft),
187
+ )
188
+ return self._whisper_feature_extractor
189
+
190
+ def _extract_mel(self, audio: Union[np.ndarray, torch.Tensor]) -> torch.Tensor:
191
+ if isinstance(audio, np.ndarray):
192
+ wav = torch.from_numpy(audio)
193
+ else:
194
+ wav = audio
195
+ wav = wav.to(dtype=torch.float32)
196
+ if wav.dim() == 1:
197
+ wav = wav.unsqueeze(0)
198
+
199
+ if bool(getattr(self.config, "use_whisper_feature_extractor", False)):
200
+ fe = self._get_whisper_feature_extractor()
201
+ wav_np = wav.detach().to("cpu", torch.float32).contiguous().numpy()
202
+ if wav_np.ndim == 2:
203
+ wav_np = wav_np[0]
204
+ feats = fe._np_extract_fbank_features(wav_np[None, ...], device="cpu")
205
+ mel = torch.from_numpy(feats[0])
206
+ else:
207
+ raise ValueError("MossAudioProcessor requires whisper feature extraction.")
208
+
209
+ return mel.to(dtype=self.config.mel_dtype)
210
+
211
+ def _get_time_marker_token_ids(self, second: int) -> List[int]:
212
+ return [self._digit_token_ids[digit] for digit in str(second)]
213
+
214
+ def _build_audio_tokens_with_time_markers(self, audio_seq_len: int) -> List[int]:
215
+ total_duration_seconds = audio_seq_len / self.audio_tokens_per_second
216
+ num_full_seconds = int(total_duration_seconds)
217
+
218
+ token_ids: List[int] = []
219
+ audio_tokens_consumed = 0
220
+ for second in range(
221
+ self.time_marker_every_seconds,
222
+ num_full_seconds + 1,
223
+ self.time_marker_every_seconds,
224
+ ):
225
+ marker_pos = (
226
+ second // self.time_marker_every_seconds
227
+ ) * self.time_marker_every_audio_tokens
228
+ audio_segment_len = marker_pos - audio_tokens_consumed
229
+ if audio_segment_len > 0:
230
+ token_ids.extend([self.audio_token_id] * audio_segment_len)
231
+ audio_tokens_consumed += audio_segment_len
232
+ token_ids.extend(self._get_time_marker_token_ids(second))
233
+
234
+ remaining = audio_seq_len - audio_tokens_consumed
235
+ if remaining > 0:
236
+ token_ids.extend([self.audio_token_id] * remaining)
237
+ return token_ids
238
+
239
+ def _build_audio_placeholder_ids(self, num_audio_tokens: int) -> List[int]:
240
+ if self.enable_time_marker:
241
+ return self._build_audio_tokens_with_time_markers(num_audio_tokens)
242
+ return [self.audio_token_id] * num_audio_tokens
243
+
244
+ def _build_default_prompt(self, text: str, has_audio: bool) -> str:
245
+ if has_audio:
246
+ return (
247
+ "<|im_start|>system\n"
248
+ "You are a helpful assistant.<|im_end|>\n"
249
+ "<|im_start|>user\n"
250
+ "<|audio_bos|><|AUDIO|><|audio_eos|>\n"
251
+ f"{text}<|im_end|>\n"
252
+ "<|im_start|>assistant\n"
253
+ )
254
+ return (
255
+ "<|im_start|>system\n"
256
+ "You are a helpful assistant.<|im_end|>\n"
257
+ "<|im_start|>user\n"
258
+ f"{text}<|im_end|>\n"
259
+ "<|im_start|>assistant\n"
260
+ )
261
+
262
+ def _build_input_from_prompt(self, prompt: str, token_lens: List[int]) -> List[int]:
263
+ spans = list(self._AUDIO_SPAN_RE.finditer(prompt))
264
+ if len(spans) != len(token_lens):
265
+ raise ValueError(
266
+ f"Audio placeholder count mismatch: found {len(spans)} spans in text, "
267
+ f"but got {len(token_lens)} audio inputs."
268
+ )
269
+
270
+ input_ids: List[int] = []
271
+ cursor = 0
272
+ for index, match in enumerate(spans):
273
+ prefix = prompt[cursor : match.start()]
274
+ if prefix:
275
+ input_ids.extend(
276
+ self._base_tokenizer.encode(prefix, add_special_tokens=False)
277
+ )
278
+
279
+ input_ids.append(self.audio_start_id)
280
+ input_ids.extend(self._build_audio_placeholder_ids(int(token_lens[index])))
281
+ input_ids.append(self.audio_end_id)
282
+ cursor = match.end()
283
+
284
+ suffix = prompt[cursor:]
285
+ if suffix:
286
+ input_ids.extend(
287
+ self._base_tokenizer.encode(suffix, add_special_tokens=False)
288
+ )
289
+ return input_ids
290
+
291
+ def __call__(
292
+ self,
293
+ *args,
294
+ text: Union[str, Sequence[str], None] = None,
295
+ audios: Optional[Sequence[Union[np.ndarray, torch.Tensor]]] = None,
296
+ audio: Optional[Sequence[Union[np.ndarray, torch.Tensor]]] = None,
297
+ return_tensors: str = "pt",
298
+ **kwargs,
299
+ ) -> BatchFeature:
300
+ _ = args, kwargs
301
+
302
+ if isinstance(text, str):
303
+ prompt_text: Optional[str] = text
304
+ elif isinstance(text, (list, tuple)):
305
+ if len(text) != 1:
306
+ raise ValueError(f"Expected text batch size 1, got {len(text)}")
307
+ prompt_text = text[0]
308
+ if not isinstance(prompt_text, str):
309
+ raise TypeError("Expected text batch size 1 with string content.")
310
+ elif text is None:
311
+ prompt_text = None
312
+ else:
313
+ raise TypeError("MossAudioProcessor text must be a string or a batch of one string.")
314
+
315
+ audio_list = audios if audios is not None else audio
316
+ audio_list = [] if audio_list is None else list(audio_list)
317
+
318
+ mels: List[torch.Tensor] = []
319
+ raw_lengths: List[int] = []
320
+ token_lens: List[int] = []
321
+ for one_audio in audio_list:
322
+ mel = self._extract_mel(one_audio)
323
+ raw_len = int(mel.shape[-1])
324
+ mels.append(mel)
325
+ raw_lengths.append(raw_len)
326
+ token_lens.append(self._conv3_downsample_len(raw_len))
327
+
328
+ if mels:
329
+ max_length = max(raw_lengths)
330
+ audio_batch = torch.zeros(
331
+ (len(mels), self.config.mel_dim, max_length),
332
+ dtype=self.config.mel_dtype,
333
+ )
334
+ for index, mel in enumerate(mels):
335
+ audio_batch[index, :, : mel.shape[-1]] = mel
336
+ seqlens_tensor = torch.tensor(raw_lengths, dtype=torch.long)
337
+ else:
338
+ audio_batch = None
339
+ seqlens_tensor = None
340
+
341
+ if prompt_text is None:
342
+ raise ValueError(
343
+ "MossAudioProcessor requires text input. Apply a chat template before calling the processor if needed."
344
+ )
345
+
346
+ if self._AUDIO_SPAN_RE.search(prompt_text) is None and audio_list:
347
+ prompt_text = self._build_default_prompt(prompt_text, has_audio=True)
348
+ elif self._AUDIO_SPAN_RE.search(prompt_text) is None and not audio_list:
349
+ prompt_text = self._build_default_prompt(prompt_text, has_audio=False)
350
+ input_ids_list = self._build_input_from_prompt(prompt_text, token_lens)
351
+
352
+ input_ids_tensor = torch.tensor([input_ids_list], dtype=torch.long)
353
+ attention_mask_tensor = torch.ones_like(input_ids_tensor)
354
+
355
+ data = {
356
+ "input_ids": input_ids_tensor,
357
+ "attention_mask": attention_mask_tensor,
358
+ }
359
+ if audio_batch is not None and seqlens_tensor is not None:
360
+ data["audio_data"] = audio_batch
361
+ data["audio_data_seqlens"] = seqlens_tensor
362
+ return BatchFeature(data=data, tensor_type=return_tensors)
363
+
364
+ def batch_decode(self, *args, **kwargs):
365
+ return self._base_tokenizer.batch_decode(*args, **kwargs)
366
+
367
+ def decode(self, *args, **kwargs):
368
+ return self._base_tokenizer.decode(*args, **kwargs)
369
+
370
+
371
+ __all__ = ["MelConfig", "MossAudioProcessor"]
special_tokens_map.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>"
16
+ ],
17
+ "eos_token": {
18
+ "content": "<|im_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ "pad_token": {
25
+ "content": "<|endoftext|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ }
31
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,239 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "<tool_response>",
183
+ "lstrip": false,
184
+ "normalized": false,
185
+ "rstrip": false,
186
+ "single_word": false,
187
+ "special": false
188
+ },
189
+ "151666": {
190
+ "content": "</tool_response>",
191
+ "lstrip": false,
192
+ "normalized": false,
193
+ "rstrip": false,
194
+ "single_word": false,
195
+ "special": false
196
+ },
197
+ "151667": {
198
+ "content": "<think>",
199
+ "lstrip": false,
200
+ "normalized": false,
201
+ "rstrip": false,
202
+ "single_word": false,
203
+ "special": false
204
+ },
205
+ "151668": {
206
+ "content": "</think>",
207
+ "lstrip": false,
208
+ "normalized": false,
209
+ "rstrip": false,
210
+ "single_word": false,
211
+ "special": false
212
+ }
213
+ },
214
+ "additional_special_tokens": [
215
+ "<|im_start|>",
216
+ "<|im_end|>",
217
+ "<|object_ref_start|>",
218
+ "<|object_ref_end|>",
219
+ "<|box_start|>",
220
+ "<|box_end|>",
221
+ "<|quad_start|>",
222
+ "<|quad_end|>",
223
+ "<|vision_start|>",
224
+ "<|vision_end|>",
225
+ "<|vision_pad|>",
226
+ "<|image_pad|>",
227
+ "<|video_pad|>"
228
+ ],
229
+ "bos_token": null,
230
+ "clean_up_tokenization_spaces": false,
231
+ "eos_token": "<|im_end|>",
232
+ "errors": "replace",
233
+ "extra_special_tokens": {},
234
+ "model_max_length": 131072,
235
+ "pad_token": "<|endoftext|>",
236
+ "split_special_tokens": false,
237
+ "tokenizer_class": "Qwen2Tokenizer",
238
+ "unk_token": null
239
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff