Hugging Face's logo

eustlb
/
LFM3-Audio-2B

Any-to-Any
Transformers
Safetensors
lfm3_audio
automatic-speech-recognition
liquid
lfm3
audio
speech
edge
conversational
multilingual
custom_code
Model card Files
xet
 Community
1
eustlb HF Staff commited on
Commit
3e19d07
·
verified ·
1 Parent(s): 7db76aa

Flat audio_encoder_config (Lfm3AudioEncoderConfig(Lfm3Config))

Browse files
Files changed (1) hide show
  1. config.json +42 -67
config.json CHANGED
@@ -3,77 +3,52 @@
3
  "Lfm3AudioForConditionalGeneration"
4
  ],
5
  "audio_encoder_config": {
 
 
 
 
6
  "conv_channels": 256,
 
 
7
  "initializer_range": 0.02,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8
  "model_type": "lfm3_audio_encoder",
 
 
 
 
9
  "num_mel_bins": 128,
10
- "text_config": {
11
- "_name_or_path": "",
12
- "architectures": null,
13
- "attention_bias": false,
14
- "attention_dropout": 0.0,
15
- "bos_token_id": 1,
16
- "chunk_size_feed_forward": 0,
17
- "conv_L_cache": 3,
18
- "conv_bias": false,
19
- "dtype": null,
20
- "eos_token_id": null,
21
- "head_dim": 96,
22
- "hidden_size": 768,
23
- "id2label": {
24
- "0": "LABEL_0",
25
- "1": "LABEL_1"
26
- },
27
- "initializer_range": 0.02,
28
- "intermediate_size": 1792,
29
- "is_encoder_decoder": false,
30
- "label2id": {
31
- "LABEL_0": 0,
32
- "LABEL_1": 1
33
- },
34
- "layer_types": [
35
- "conv",
36
- "conv",
37
- "sliding_attention",
38
- "conv",
39
- "conv",
40
- "sliding_attention",
41
- "conv",
42
- "conv",
43
- "sliding_attention",
44
- "conv",
45
- "conv",
46
- "sliding_attention",
47
- "conv",
48
- "conv",
49
- "sliding_attention",
50
- "sliding_attention",
51
- "sliding_attention"
52
- ],
53
- "max_position_embeddings": 128000,
54
- "model_type": "lfm3",
55
- "norm_eps": 1e-05,
56
- "num_attention_heads": 8,
57
- "num_hidden_layers": 17,
58
- "num_key_value_heads": 4,
59
- "output_attentions": false,
60
- "output_hidden_states": false,
61
- "pad_token_id": 0,
62
- "problem_type": null,
63
- "return_dict": true,
64
- "rope_parameters": {
65
- "sliding_attention": {
66
- "rope_theta": 1000000.0,
67
- "rope_type": "default"
68
- }
69
- },
70
- "sliding_window": 32,
71
- "tie_word_embeddings": true,
72
- "use_attention_gate": false,
73
- "use_cache": true,
74
- "use_sandwich_norm": false,
75
- "vocab_size": 2
76
- }
77
  },
78
  "audio_eos_code": 2016,
79
  "audio_loss_weight": 1.0,
 
3
  "Lfm3AudioForConditionalGeneration"
4
  ],
5
  "audio_encoder_config": {
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "conv_L_cache": 3,
9
+ "conv_bias": false,
10
  "conv_channels": 256,
11
+ "head_dim": 96,
12
+ "hidden_size": 768,
13
  "initializer_range": 0.02,
14
+ "intermediate_size": 1792,
15
+ "layer_types": [
16
+ "conv",
17
+ "conv",
18
+ "sliding_attention",
19
+ "conv",
20
+ "conv",
21
+ "sliding_attention",
22
+ "conv",
23
+ "conv",
24
+ "sliding_attention",
25
+ "conv",
26
+ "conv",
27
+ "sliding_attention",
28
+ "conv",
29
+ "conv",
30
+ "sliding_attention",
31
+ "sliding_attention",
32
+ "sliding_attention"
33
+ ],
34
+ "max_position_embeddings": 128000,
35
  "model_type": "lfm3_audio_encoder",
36
+ "norm_eps": 1e-05,
37
+ "num_attention_heads": 8,
38
+ "num_hidden_layers": 17,
39
+ "num_key_value_heads": 4,
40
  "num_mel_bins": 128,
41
+ "rope_parameters": {
42
+ "sliding_attention": {
43
+ "rope_theta": 1000000.0,
44
+ "rope_type": "default"
45
+ }
46
+ },
47
+ "sliding_window": 32,
48
+ "subsampling_factor": 8,
49
+ "use_attention_gate": false,
50
+ "use_cache": true,
51
+ "use_sandwich_norm": false
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52
  },
53
  "audio_eos_code": 2016,
54
  "audio_loss_weight": 1.0,