Hugging Face's logo

onnx-community
/
granite-4.0-1b-speech-ONNX

Automatic Speech Recognition
Transformers.js
ONNX
granite_speech
Model card Files
xet
 Community
1
Xenova HF Staff commited on
Commit
618bb30
·
verified ·
1 Parent(s): 4e258fb

Upload folder using huggingface_hub

Browse files
Files changed (41) hide show
  1. .gitattributes +19 -0
  2. chat_template.jinja +2 -0
  3. config.json +98 -0
  4. onnx/audio_encoder.onnx +3 -0
  5. onnx/audio_encoder.onnx_data +3 -0
  6. onnx/audio_encoder_fp16.onnx +3 -0
  7. onnx/audio_encoder_fp16.onnx_data +3 -0
  8. onnx/audio_encoder_q4.onnx +3 -0
  9. onnx/audio_encoder_q4.onnx_data +3 -0
  10. onnx/audio_encoder_q4f16.onnx +3 -0
  11. onnx/audio_encoder_q4f16.onnx_data +3 -0
  12. onnx/audio_encoder_quantized.onnx +3 -0
  13. onnx/audio_encoder_quantized.onnx_data +3 -0
  14. onnx/decoder_model_merged.onnx +3 -0
  15. onnx/decoder_model_merged.onnx_data +3 -0
  16. onnx/decoder_model_merged.onnx_data_1 +3 -0
  17. onnx/decoder_model_merged.onnx_data_2 +3 -0
  18. onnx/decoder_model_merged.onnx_data_3 +3 -0
  19. onnx/decoder_model_merged_fp16.onnx +3 -0
  20. onnx/decoder_model_merged_fp16.onnx_data +3 -0
  21. onnx/decoder_model_merged_fp16.onnx_data_1 +3 -0
  22. onnx/decoder_model_merged_q4.onnx +3 -0
  23. onnx/decoder_model_merged_q4.onnx_data +3 -0
  24. onnx/decoder_model_merged_q4f16.onnx +3 -0
  25. onnx/decoder_model_merged_q4f16.onnx_data +3 -0
  26. onnx/decoder_model_merged_quantized.onnx +3 -0
  27. onnx/decoder_model_merged_quantized.onnx_data +3 -0
  28. onnx/embed_tokens.onnx +3 -0
  29. onnx/embed_tokens.onnx_data +3 -0
  30. onnx/embed_tokens_fp16.onnx +3 -0
  31. onnx/embed_tokens_fp16.onnx_data +3 -0
  32. onnx/embed_tokens_q4.onnx +3 -0
  33. onnx/embed_tokens_q4.onnx_data +3 -0
  34. onnx/embed_tokens_q4f16.onnx +3 -0
  35. onnx/embed_tokens_q4f16.onnx_data +3 -0
  36. onnx/embed_tokens_quantized.onnx +3 -0
  37. onnx/embed_tokens_quantized.onnx_data +3 -0
  38. preprocessor_config.json +14 -0
  39. processor_config.json +17 -0
  40. tokenizer.json +0 -0
  41. tokenizer_config.json +16 -0
.gitattributes CHANGED
@@ -33,3 +33,22 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ onnx/audio_encoder.onnx_data filter=lfs diff=lfs merge=lfs -text
37
+ onnx/audio_encoder_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
38
+ onnx/audio_encoder_q4.onnx_data filter=lfs diff=lfs merge=lfs -text
39
+ onnx/audio_encoder_q4f16.onnx_data filter=lfs diff=lfs merge=lfs -text
40
+ onnx/audio_encoder_quantized.onnx_data filter=lfs diff=lfs merge=lfs -text
41
+ onnx/decoder_model_merged.onnx_data filter=lfs diff=lfs merge=lfs -text
42
+ onnx/decoder_model_merged.onnx_data_1 filter=lfs diff=lfs merge=lfs -text
43
+ onnx/decoder_model_merged.onnx_data_2 filter=lfs diff=lfs merge=lfs -text
44
+ onnx/decoder_model_merged.onnx_data_3 filter=lfs diff=lfs merge=lfs -text
45
+ onnx/decoder_model_merged_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
46
+ onnx/decoder_model_merged_fp16.onnx_data_1 filter=lfs diff=lfs merge=lfs -text
47
+ onnx/decoder_model_merged_q4.onnx_data filter=lfs diff=lfs merge=lfs -text
48
+ onnx/decoder_model_merged_q4f16.onnx_data filter=lfs diff=lfs merge=lfs -text
49
+ onnx/decoder_model_merged_quantized.onnx_data filter=lfs diff=lfs merge=lfs -text
50
+ onnx/embed_tokens.onnx_data filter=lfs diff=lfs merge=lfs -text
51
+ onnx/embed_tokens_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
52
+ onnx/embed_tokens_q4.onnx_data filter=lfs diff=lfs merge=lfs -text
53
+ onnx/embed_tokens_q4f16.onnx_data filter=lfs diff=lfs merge=lfs -text
54
+ onnx/embed_tokens_quantized.onnx_data filter=lfs diff=lfs merge=lfs -text
chat_template.jinja ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ {% for message in messages %}{% if message['role'] == 'user' %}USER: {{ message['content'] }}
2
+ ASSISTANT:{% elif message['role'] == 'assistant' %}{{ message['content'] }}{% endif %}{% endfor %}
config.json ADDED
@@ -0,0 +1,98 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "GraniteSpeechForConditionalGeneration"
4
+ ],
5
+ "audio_token_index": 100352,
6
+ "downsample_rate": 5,
7
+ "dtype": "bfloat16",
8
+ "encoder_config": {
9
+ "context_size": 200,
10
+ "conv_expansion_factor": 2,
11
+ "conv_kernel_size": 15,
12
+ "dim_head": 128,
13
+ "dropout": 0.1,
14
+ "dtype": "bfloat16",
15
+ "feedforward_mult": 4,
16
+ "hidden_dim": 1024,
17
+ "input_dim": 160,
18
+ "max_pos_emb": 512,
19
+ "model_type": "granite_speech_encoder",
20
+ "num_heads": 8,
21
+ "num_layers": 16,
22
+ "output_dim": 348
23
+ },
24
+ "has_lora_adapter": false,
25
+ "initializer_range": 0.02,
26
+ "model_type": "granite_speech",
27
+ "projector_config": {
28
+ "_attn_implementation_autoset": true,
29
+ "attention_probs_dropout_prob": 0.1,
30
+ "cross_attention_frequency": 1,
31
+ "dtype": "bfloat16",
32
+ "encoder_hidden_size": 1024,
33
+ "hidden_act": "gelu",
34
+ "hidden_dropout_prob": 0.1,
35
+ "hidden_size": 1024,
36
+ "initializer_range": 0.02,
37
+ "intermediate_size": 4096,
38
+ "layer_norm_eps": 1e-12,
39
+ "max_position_embeddings": 2048,
40
+ "model_type": "blip_2_qformer",
41
+ "num_attention_heads": 16,
42
+ "num_hidden_layers": 2,
43
+ "pad_token_id": 0,
44
+ "position_embedding_type": "absolute",
45
+ "use_qformer_text_input": false,
46
+ "vocab_size": 30522
47
+ },
48
+ "text_config": {
49
+ "_name_or_path": "/proj/speech/saon/slam-llm/29.2-c/granite-4.0-1b-base",
50
+ "architectures": [
51
+ "GraniteForCausalLM"
52
+ ],
53
+ "attention_bias": false,
54
+ "attention_dropout": 0.0,
55
+ "attention_multiplier": 0.0078125,
56
+ "bos_token_id": 100257,
57
+ "dtype": "float32",
58
+ "embedding_multiplier": 12,
59
+ "eos_token_id": 100257,
60
+ "hidden_act": "silu",
61
+ "hidden_size": 2048,
62
+ "initializer_range": 0.1,
63
+ "intermediate_size": 4096,
64
+ "logits_scaling": 8,
65
+ "max_position_embeddings": 4096,
66
+ "mlp_bias": false,
67
+ "model_type": "granite",
68
+ "num_attention_heads": 16,
69
+ "num_hidden_layers": 40,
70
+ "num_key_value_heads": 4,
71
+ "pad_token_id": 100256,
72
+ "residual_multiplier": 0.22,
73
+ "rms_norm_eps": 1e-05,
74
+ "rope_parameters": {
75
+ "rope_theta": 10000,
76
+ "rope_type": "default"
77
+ },
78
+ "tie_word_embeddings": false,
79
+ "use_cache": true,
80
+ "vocab_size": 100353
81
+ },
82
+ "tie_word_embeddings": false,
83
+ "transformers_version": "5.3.0.dev0",
84
+ "window_size": 15,
85
+ "transformers.js_config": {
86
+ "use_external_data_format": {
87
+ "audio_encoder": 1,
88
+ "embed_tokens": 1,
89
+ "decoder_model_merged.onnx": 4,
90
+ "decoder_model_merged_fp16.onnx": 2,
91
+ "decoder_model_merged": 1
92
+ },
93
+ "kv_cache_dtype": {
94
+ "q4f16": "float16",
95
+ "fp16": "float16"
96
+ }
97
+ }
98
+ }
onnx/audio_encoder.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de2fdaac1fe03a9b854208191827d4a5650b0106e3138178faa581924d371ce7
3
+ size 303030
onnx/audio_encoder.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f255f3117c45703b088a54033575ede1dac6a57a2f0cdb33346acd847b06942
3
+ size 1895899568
onnx/audio_encoder_fp16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b6ee23c99771e10b5619cb28a85048b9572012373c6de552382404aa14148757
3
+ size 306875
onnx/audio_encoder_fp16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:659ec9bfa6e0145768075fca85d5bedafa461541df4b92776e1ee43334f911ae
3
+ size 948429888
onnx/audio_encoder_q4.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5bc191d3bf746d714ed79f256d43899367deb310dd778ab517260437fed738a6
3
+ size 354758
onnx/audio_encoder_q4.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2dc0e1c1975176e8c7f8743f7c91838f5c63eb1b120ab453951423a2c6b5fb7
3
+ size 658277008
onnx/audio_encoder_q4f16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15208e02405c0e2f4861fff6c73c9c33ee8003f82fbd871e312463bff889cd50
3
+ size 358553
onnx/audio_encoder_q4f16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d53d62e13c9ca1a949fcde14926ecffa3387841838b5472d175e3f95dae5e3ff
3
+ size 424598944
onnx/audio_encoder_quantized.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9581c6664ab0803aa5ee144b855df8239f7c88186ac4c059d04fcebf9e4d89b8
3
+ size 361220
onnx/audio_encoder_quantized.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e94fa56cb27edcb9380a8535833e69f8493c7fb5f30e2a376c4b5719faf324ff
3
+ size 847771760
onnx/decoder_model_merged.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21cc4a487c52ddf7f3d778f1908241184a71e8d18194d8694325ebe42d601145
3
+ size 298151
onnx/decoder_model_merged.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d68a7ec06f1f880a440cc0f7a96e1de7add91312a82282d8d6efa0e7679bba6
3
+ size 2074329088
onnx/decoder_model_merged.onnx_data_1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16f94a3e98618043103f1eb3b9a82c7bae94024ae424e5a45e8995409317ade8
3
+ size 2089000960
onnx/decoder_model_merged.onnx_data_2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9257ee80b0cb9c65ea366041d7f43eb2c49ca56c08e6b5e9fa988dadc207688e
3
+ size 1543684096
onnx/decoder_model_merged.onnx_data_3 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1587d0e86c810feb5ab0c7c5676e8b86f56e1d615001d1624d2e037dbf124b9
3
+ size 822091776
onnx/decoder_model_merged_fp16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5b5f7f9f6fcaafa62158fde44f8e51d1d8e882e044e1640edaf568c14d9398b
3
+ size 299991
onnx/decoder_model_merged_fp16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba1550fd1e2437725f644b97eff977ce872a26033da4fe555ee9078a47d3f172
3
+ size 2090057728
onnx/decoder_model_merged_fp16.onnx_data_1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4569c9a8a0eea0df5fa0f894a80e8b2b2ce939d303559f0eafe700888ebf371
3
+ size 1174495232
onnx/decoder_model_merged_q4.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c57ab5045fee2f0774ae1e9f017cd545e30feb8229bc03bd709ed05a3a2713fd
3
+ size 433604
onnx/decoder_model_merged_q4.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9ff213c2209277419881a77b048ad58f31ce86a3f96ddbacc6b5d601e6e7dc5
3
+ size 1047995680
onnx/decoder_model_merged_q4f16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8c4b7df6affa245dc66ccad596ab675f94d789d59fdb1a065eb8edb20b85331
3
+ size 436556
onnx/decoder_model_merged_q4f16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40630fc30057acf2332b9acea30ccdf214d12976e9fe485d314534b745b6c25d
3
+ size 944641184
onnx/decoder_model_merged_quantized.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8328c44ff3dbaa169543376ad8115b59a71879f6930654c5e53b0266a2ef3ce
3
+ size 440659
onnx/decoder_model_merged_quantized.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0700fcf06838b4b0bae897eb4ea97c7920b91a25209f906a5f67920caff7867e
3
+ size 1889282368
onnx/embed_tokens.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27876aee502105e2180423470fc02df8f352a91f5b775da2e4b0aac049a4c439
3
+ size 434
onnx/embed_tokens.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16e18f112dd9c51b019fa2d7346cee5b32a78dd69b10564b0577eebd2a70b6a2
3
+ size 822091776
onnx/embed_tokens_fp16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c7a044d964446c786562fc6bf8c005a4cdabfb8dca0e061bd34dfe1aa567f25
3
+ size 637
onnx/embed_tokens_fp16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f25206ea095c171ad7f6ebc2a44247d6b90c70c211a61cb1690a87c2b44c8c1f
3
+ size 411045888
onnx/embed_tokens_q4.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6411f2b4802e235e6ba28c8506787bfc92f0888266bbbc2c53cf67dd6763bc36
3
+ size 857
onnx/embed_tokens_q4.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc1e90dccf09d0323cd17e06df312c5ece70eb083a74cdc00a998697a0d4382f
3
+ size 131663136
onnx/embed_tokens_q4f16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5bd332cac37108126d86a686e1683bdcdc3ef62dc02caa99bdadad552a365cea
3
+ size 1064
onnx/embed_tokens_q4f16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b265c93bb0628ae21177033fd80bee89859210c9cc0674886ef9a8198ed5c687
3
+ size 118817952
onnx/embed_tokens_quantized.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:498d03831ead2fb4a21397c58a18f20c146bff920a79b34ddf05b118efda6679
3
+ size 879
onnx/embed_tokens_quantized.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70fbaebc6347b72bf3adb76296f901ee79e0a4b58c8cfaee77ec268921725b10
3
+ size 237635904
preprocessor_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "feature_extractor_type": "GraniteSpeechFeatureExtractor",
3
+ "melspec_kwargs": {
4
+ "hop_length": 160,
5
+ "n_fft": 512,
6
+ "n_mels": 80,
7
+ "sample_rate": 16000,
8
+ "win_length": 400
9
+ },
10
+ "processor_class": "GraniteSpeechProcessor",
11
+ "projector_downsample_rate": 5,
12
+ "projector_window_size": 15,
13
+ "sampling_rate": 16000
14
+ }
processor_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "audio_processor": {
3
+ "feature_extractor_type": "GraniteSpeechFeatureExtractor",
4
+ "melspec_kwargs": {
5
+ "hop_length": 160,
6
+ "n_fft": 512,
7
+ "n_mels": 80,
8
+ "sample_rate": 16000,
9
+ "win_length": 400
10
+ },
11
+ "projector_downsample_rate": 5,
12
+ "projector_window_size": 15,
13
+ "sampling_rate": 16000
14
+ },
15
+ "audio_token": "<|audio|>",
16
+ "processor_class": "GraniteSpeechProcessor"
17
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|end_of_text|>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|end_of_text|>",
7
+ "errors": "replace",
8
+ "is_local": false,
9
+ "model_max_length": 1000000000000000019884624838656,
10
+ "pad_token": "<|pad|>",
11
+ "padding_side": "left",
12
+ "processor_class": "GraniteSpeechProcessor",
13
+ "tokenizer_class": "GPT2Tokenizer",
14
+ "unk_token": "<|unk|>",
15
+ "chat_template": "{% for message in messages %}{% if message['role'] == 'user' %}USER: {{ message['content'] }}\n ASSISTANT:{% elif message['role'] == 'assistant' %}{{ message['content'] }}{% endif %}{% endfor %}"
16
+ }