Xenova HF Staff commited on
Commit
5a7ca3b
·
verified ·
1 Parent(s): 567fad1

Upload folder using huggingface_hub

Browse files
.gitattributes CHANGED
@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ onnx/model.onnx_data filter=lfs diff=lfs merge=lfs -text
37
+ onnx/model_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
config.json ADDED
@@ -0,0 +1,61 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Nemotron3DiarizationForAudioFrameClassification"
4
+ ],
5
+ "audio_config": {
6
+ "attention_dropout": 0.0,
7
+ "dtype": "float32",
8
+ "hidden_act": "gelu",
9
+ "hidden_size": 32,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 64,
12
+ "max_position_embeddings": 512,
13
+ "model_type": "nemotron3_diarization_audio",
14
+ "num_attention_heads": 2,
15
+ "num_hidden_layers": 2,
16
+ "num_key_value_heads": 2,
17
+ "num_mel_bins": 128,
18
+ "partial_rotary_factor": 1.0,
19
+ "rope_parameters": {
20
+ "partial_rotary_factor": 1.0,
21
+ "rope_theta": 10000.0,
22
+ "rope_type": "default"
23
+ },
24
+ "subsampling_factor": 8
25
+ },
26
+ "chunk_length": 4,
27
+ "chunk_right_context": 1,
28
+ "dtype": "float32",
29
+ "fifo_length": 4,
30
+ "head_config": {
31
+ "audio_hidden_size": 32,
32
+ "dtype": "float32",
33
+ "hidden_size": 16,
34
+ "model_type": "",
35
+ "num_speakers": 4,
36
+ "subsampling_factor": 8
37
+ },
38
+ "initializer_range": 0.02,
39
+ "model_type": "nemotron3_diarization",
40
+ "speaker_cache_update_period": 3,
41
+ "streaming_config": {
42
+ "dtype": "float32",
43
+ "fifo_length": 4,
44
+ "latest_frames_score_boost": 0.05,
45
+ "min_positive_scores_rate": 0.5,
46
+ "model_type": "",
47
+ "num_speakers": 4,
48
+ "prediction_score_threshold": 0.25,
49
+ "speaker_cache_length": 12,
50
+ "speaker_cache_silence_frames_per_speaker": 1,
51
+ "speaker_cache_update_period": 3,
52
+ "strong_boost_rate": 0.75,
53
+ "subsampling_factor": 8,
54
+ "weak_boost_rate": 1.5
55
+ },
56
+ "transformers.js_config": {
57
+ "dtype": "fp32",
58
+ "use_external_data_format": true
59
+ },
60
+ "transformers_version": "5.18.0.dev0"
61
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0ce1274fdcf36e25b57bc1c2e744108134b66376ebb514a44f7ec4220ef54bf3
3
+ size 232000
onnx/model.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f69a4115be1b8150ab2484d5ddc8bfe6ff59d0008c5d87660fb8cfc769bc8902
3
+ size 30063
onnx/model.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e1fa4864a150bccaaa9a640e3ac80186aa835776801524dec4a8fae43772c95
3
+ size 257024
onnx/model_fp16.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:64ad72ce432d14a0b42a5cca9aa713319a1928f9ae0b709e1b805a79de030702
3
+ size 35356
onnx/model_fp16.onnx_data ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:549b10f87cdf5b4e1ce821ef1d17a9fc38a59d5ddcfc92ece2e5a03499deea8a
3
+ size 128000
preprocessor_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "feature_extractor_type": "NemotronAsrStreamingFeatureExtractor",
3
+ "feature_size": 128,
4
+ "hop_length": 160,
5
+ "n_fft": 512,
6
+ "padding_side": "right",
7
+ "padding_value": 0.0,
8
+ "preemphasis": 0.97,
9
+ "return_attention_mask": true,
10
+ "sampling_rate": 16000,
11
+ "win_length": 400,
12
+ "processor_class": "Nemotron3DiarizationProcessor"
13
+ }
processor_config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "feature_extractor": {
3
+ "feature_extractor_type": "NemotronAsrStreamingFeatureExtractor",
4
+ "feature_size": 128,
5
+ "hop_length": 160,
6
+ "n_fft": 512,
7
+ "padding_side": "right",
8
+ "padding_value": 0.0,
9
+ "preemphasis": 0.97,
10
+ "return_attention_mask": true,
11
+ "sampling_rate": 16000,
12
+ "win_length": 400
13
+ },
14
+ "processor_class": "Nemotron3DiarizationProcessor",
15
+ "streaming_mode": "low_latency",
16
+ "streaming_modes": {
17
+ "low_latency": [
18
+ 9,
19
+ 4
20
+ ],
21
+ "ultra_low_latency": [
22
+ 3,
23
+ 1
24
+ ],
25
+ "very_low_latency": [
26
+ 6,
27
+ 2
28
+ ]
29
+ },
30
+ "subsampling_factor": 8
31
+ }