Hugging Face's logo

Uni-MoE
/
Uni-MoE-speech-base

Text Generation
Transformers
PyTorch
llava
Inference Endpoints
Model card Files Community
Uni-MoE commited on
Commit
bf2b75d
•
1 Parent(s): 30d8534

Upload 10 files

Browse files
Files changed (10) hide show
  1. 4config.json +55 -0
  2. 8config.json +72 -0
  3. config.json +72 -0
  4. generation_config.json +14 -0
  5. mm_audio_aligner.bin +3 -0
  6. mm_projector.bin +3 -0
  7. pytorch_model.bin.index.json +0 -0
  8. special_tokens_map.json +24 -0
  9. tokenizer.model +3 -0
  10. tokenizer_config.json +35 -0
4config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/path/to/Uni-MoE-speech-base",
3
+ "architectures": [
4
+ "LlavaLlamaForCausalLM"
5
+ ],
6
+ "bos_token_id": 1,
7
+ "eos_token_id": 2,
8
+ "freeze_mm_mlp_adapter": false,
9
+ "freeze_mm_vision_resampler": false,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 4096,
12
+ "image_aspect_ratio": "pad",
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
15
+ "language": "English",
16
+ "local_files_only": false,
17
+ "max_length": 4096,
18
+ "max_position_embeddings": 4096,
19
+ "mm_audio_hidden_size": 768,
20
+ "mm_audio_select_feature": "patch",
21
+ "mm_audio_select_layer": 0,
22
+ "mm_audio_tower": "/path/to/whisper-small",
23
+ "mm_hidden_size": 1024,
24
+ "mm_projector_type": "mlp2x_gelu",
25
+ "mm_query_tokens": true,
26
+ "mm_resampler_type": null,
27
+ "mm_use_im_patch_token": false,
28
+ "mm_use_im_start_end": false,
29
+ "mm_vision_select_feature": "patch",
30
+ "mm_vision_select_layer": -2,
31
+ "mm_vision_tower": "/path/to/clip-vit-large-patch14-336",
32
+ "model_type": "llava",
33
+ "moe": true,
34
+ "num_attention_heads": 32,
35
+ "num_experts": 4,
36
+ "num_experts_per_tok": 2,
37
+ "num_hidden_layers": 32,
38
+ "num_key_value_heads": 32,
39
+ "pad_token_id": 0,
40
+ "pretraining_tp": 1,
41
+ "query_tokens_size": 50,
42
+ "rms_norm_eps": 1e-05,
43
+ "rope_scaling": null,
44
+ "task": "transcribe",
45
+ "tie_word_embeddings": false,
46
+ "torch_dtype": "float32",
47
+ "transformers_version": "4.31.0",
48
+ "tune_mm_audio_aligner": true,
49
+ "tune_mm_mlp_adapter": false,
50
+ "tune_mm_vision_resampler": false,
51
+ "unfreeze_mm_vision_tower": false,
52
+ "use_cache": true,
53
+ "use_mm_proj": true,
54
+ "vocab_size": 32000
55
+ }
8config.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/path/to/Uni-MoE-speech-base",
3
+ "architectures": [
4
+ "LlavaLlamaForCausalLM"
5
+ ],
6
+ "bos_token_id": 1,
7
+ "eos_token_id": 2,
8
+ "freeze_mm_mlp_adapter": false,
9
+ "freeze_mm_vision_resampler": false,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 4096,
12
+ "image_aspect_ratio": "pad",
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
15
+ "language": "English",
16
+ "local_files_only": false,
17
+ "max_length": 4096,
18
+ "max_position_embeddings": 4096,
19
+ "mm_audio_hidden_size": 768,
20
+ "mm_audio_select_feature": "patch",
21
+ "mm_audio_select_layer": 0,
22
+ "mm_audio_tower": "/path/to/whisper-small",
23
+ "mm_hidden_size": 1024,
24
+ "mm_projector_type": "mlp2x_gelu",
25
+ "mm_query_tokens": true,
26
+ "mm_resampler_type": null,
27
+ "mm_use_im_patch_token": false,
28
+ "mm_use_im_start_end": false,
29
+ "mm_vision_select_feature": "patch",
30
+ "mm_vision_select_layer": -2,
31
+ "mm_vision_tower": "/path/to/clip-vit-large-patch14-336",
32
+ "model_type": "llava_llama",
33
+ "moe": true,
34
+ "num_attention_heads": 32,
35
+ "num_experts": 8,
36
+ "num_experts_per_tok": 2,
37
+ "moe_type": "dense",
38
+ "ep_size": 8,
39
+ "capacity_factor": 1.5,
40
+ "lora_r": 8,
41
+ "lora_alpha": 16,
42
+ "image_expert_indices": [2, 3, 4, 6],
43
+ "video_expert_indices": [2, 3, 4, 6],
44
+ "audio_expert_indices": [0, 4, 5, 7],
45
+ "use_fm_block": false,
46
+ "use_flash_attn": false,
47
+ "aux_balance_loss": true,
48
+ "aux_balance_loss_coef": 0.01,
49
+ "aux_multimodal_loss": false,
50
+ "aux_multimodal_loss_coef": 0.01,
51
+ "aux_kl_loss": false,
52
+ "aux_kl_loss_coef": 0.01,
53
+ "lora_dropout": 0.05,
54
+ "num_hidden_layers": 32,
55
+ "num_key_value_heads": 32,
56
+ "pad_token_id": 0,
57
+ "pretraining_tp": 1,
58
+ "query_tokens_size": 50,
59
+ "rms_norm_eps": 1e-05,
60
+ "rope_scaling": null,
61
+ "task": "transcribe",
62
+ "tie_word_embeddings": false,
63
+ "torch_dtype": "float32",
64
+ "transformers_version": "4.31.0",
65
+ "tune_mm_audio_aligner": true,
66
+ "tune_mm_mlp_adapter": false,
67
+ "tune_mm_vision_resampler": false,
68
+ "unfreeze_mm_vision_tower": false,
69
+ "use_cache": true,
70
+ "use_mm_proj": true,
71
+ "vocab_size": 32000
72
+ }
config.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/path/to/Uni-MoE-speech-base",
3
+ "architectures": [
4
+ "LlavaLlamaForCausalLM"
5
+ ],
6
+ "bos_token_id": 1,
7
+ "eos_token_id": 2,
8
+ "freeze_mm_mlp_adapter": false,
9
+ "freeze_mm_vision_resampler": false,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 4096,
12
+ "image_aspect_ratio": "pad",
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
15
+ "language": "English",
16
+ "local_files_only": false,
17
+ "max_length": 4096,
18
+ "max_position_embeddings": 4096,
19
+ "mm_audio_hidden_size": 768,
20
+ "mm_audio_select_feature": "patch",
21
+ "mm_audio_select_layer": 0,
22
+ "mm_audio_tower": "/path/to/whisper-small",
23
+ "mm_hidden_size": 1024,
24
+ "mm_projector_type": "mlp2x_gelu",
25
+ "mm_query_tokens": true,
26
+ "mm_resampler_type": null,
27
+ "mm_use_im_patch_token": false,
28
+ "mm_use_im_start_end": false,
29
+ "mm_vision_select_feature": "patch",
30
+ "mm_vision_select_layer": -2,
31
+ "mm_vision_tower": "/path/to/clip-vit-large-patch14-336",
32
+ "model_type": "llava_llama",
33
+ "moe": true,
34
+ "num_attention_heads": 32,
35
+ "num_experts": 8,
36
+ "num_experts_per_tok": 2,
37
+ "moe_type": "dense",
38
+ "ep_size": 8,
39
+ "capacity_factor": 1.5,
40
+ "lora_r": 8,
41
+ "lora_alpha": 16,
42
+ "image_expert_indices": [2, 3, 4, 6],
43
+ "video_expert_indices": [2, 3, 4, 6],
44
+ "audio_expert_indices": [0, 4, 5, 7],
45
+ "use_fm_block": false,
46
+ "use_flash_attn": false,
47
+ "aux_balance_loss": true,
48
+ "aux_balance_loss_coef": 0.01,
49
+ "aux_multimodal_loss": false,
50
+ "aux_multimodal_loss_coef": 0.01,
51
+ "aux_kl_loss": false,
52
+ "aux_kl_loss_coef": 0.01,
53
+ "lora_dropout": 0.05,
54
+ "num_hidden_layers": 32,
55
+ "num_key_value_heads": 32,
56
+ "pad_token_id": 0,
57
+ "pretraining_tp": 1,
58
+ "query_tokens_size": 50,
59
+ "rms_norm_eps": 1e-05,
60
+ "rope_scaling": null,
61
+ "task": "transcribe",
62
+ "tie_word_embeddings": false,
63
+ "torch_dtype": "float32",
64
+ "transformers_version": "4.31.0",
65
+ "tune_mm_audio_aligner": true,
66
+ "tune_mm_mlp_adapter": false,
67
+ "tune_mm_vision_resampler": false,
68
+ "unfreeze_mm_vision_tower": false,
69
+ "use_cache": true,
70
+ "use_mm_proj": true,
71
+ "vocab_size": 32000
72
+ }
generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 1,
3
+ "eos_token_id": 2,
4
+ "max_length": 4096,
5
+ "max_memory": {
6
+ "cpu": 99304017920
7
+ },
8
+ "no_split_module_classes": [
9
+ "LlamaDecoderLayer"
10
+ ],
11
+ "pad_token_id": 0,
12
+ "special_dtypes": {},
13
+ "transformers_version": "4.31.0"
14
+ }
mm_audio_aligner.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b4f0e83d3db72c0d8251e681a8cb3f585e52c8930ea5e5aa32dfcdf815f344d
3
+ size 313644388
mm_projector.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0df1fd04dd34b57e0e4f6965e9f9a0cf8e7c1a6974db93550cb7f5a7f279a96a
3
+ size 41961085
pytorch_model.bin.index.json ADDED
The diff for this file is too large to render. See raw diff
 
special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "<unk>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
tokenizer_config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "bos_token": {
5
+ "__type": "AddedToken",
6
+ "content": "<s>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "clean_up_tokenization_spaces": false,
13
+ "eos_token": {
14
+ "__type": "AddedToken",
15
+ "content": "</s>",
16
+ "lstrip": false,
17
+ "normalized": false,
18
+ "rstrip": false,
19
+ "single_word": false
20
+ },
21
+ "legacy": false,
22
+ "model_max_length": 2048,
23
+ "pad_token": null,
24
+ "padding_side": "right",
25
+ "sp_model_kwargs": {},
26
+ "tokenizer_class": "LlamaTokenizer",
27
+ "unk_token": {
28
+ "__type": "AddedToken",
29
+ "content": "<unk>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false
34
+ }
35
+ }