Update tiny models for MoonshineStreamingModel
Browse files- config.json +73 -0
- model.safetensors +3 -0
- preprocessor_config.json +10 -0
- tokenizer.json +0 -0
- tokenizer_config.json +11 -0
config.json
ADDED
|
@@ -0,0 +1,73 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"MoonshineStreamingModel"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 1,
|
| 8 |
+
"decoder_start_token_id": 85,
|
| 9 |
+
"dtype": "float32",
|
| 10 |
+
"encoder_config": {
|
| 11 |
+
"attention_bias": false,
|
| 12 |
+
"attention_dropout": 0.0,
|
| 13 |
+
"frame_ms": 5.0,
|
| 14 |
+
"head_dim": 4,
|
| 15 |
+
"hidden_act": "gelu",
|
| 16 |
+
"hidden_size": 8,
|
| 17 |
+
"intermediate_size": 32,
|
| 18 |
+
"max_position_embeddings": 4096,
|
| 19 |
+
"model_type": "moonshine_streaming_encoder",
|
| 20 |
+
"num_attention_heads": 2,
|
| 21 |
+
"num_hidden_layers": 2,
|
| 22 |
+
"num_key_value_heads": 2,
|
| 23 |
+
"sample_rate": 16000,
|
| 24 |
+
"sliding_windows": [
|
| 25 |
+
[
|
| 26 |
+
16,
|
| 27 |
+
4
|
| 28 |
+
],
|
| 29 |
+
[
|
| 30 |
+
16,
|
| 31 |
+
4
|
| 32 |
+
],
|
| 33 |
+
[
|
| 34 |
+
16,
|
| 35 |
+
0
|
| 36 |
+
],
|
| 37 |
+
[
|
| 38 |
+
16,
|
| 39 |
+
0
|
| 40 |
+
],
|
| 41 |
+
[
|
| 42 |
+
16,
|
| 43 |
+
4
|
| 44 |
+
],
|
| 45 |
+
[
|
| 46 |
+
16,
|
| 47 |
+
4
|
| 48 |
+
]
|
| 49 |
+
]
|
| 50 |
+
},
|
| 51 |
+
"eos_token_id": 2,
|
| 52 |
+
"head_dim": 4,
|
| 53 |
+
"hidden_act": "silu",
|
| 54 |
+
"hidden_size": 8,
|
| 55 |
+
"intermediate_size": 32,
|
| 56 |
+
"is_encoder_decoder": true,
|
| 57 |
+
"max_position_embeddings": 4096,
|
| 58 |
+
"model_type": "moonshine_streaming",
|
| 59 |
+
"num_attention_heads": 2,
|
| 60 |
+
"num_hidden_layers": 2,
|
| 61 |
+
"num_key_value_heads": 2,
|
| 62 |
+
"pad_head_dim_to_multiple_of": null,
|
| 63 |
+
"pad_token_id": 0,
|
| 64 |
+
"rope_parameters": {
|
| 65 |
+
"partial_rotary_factor": 0.8,
|
| 66 |
+
"rope_theta": 10000.0,
|
| 67 |
+
"rope_type": "default"
|
| 68 |
+
},
|
| 69 |
+
"tie_word_embeddings": false,
|
| 70 |
+
"transformers_version": "5.16.0.dev0",
|
| 71 |
+
"use_cache": true,
|
| 72 |
+
"vocab_size": 32768
|
| 73 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e558b72862445c543dc908de7574f1b37d18e23f9f47bce6176e3de649355e3
|
| 3 |
+
size 1211220
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_normalize": false,
|
| 3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
| 4 |
+
"feature_size": 1,
|
| 5 |
+
"pad_to_multiple_of": 80,
|
| 6 |
+
"padding_side": "right",
|
| 7 |
+
"padding_value": 0.0,
|
| 8 |
+
"return_attention_mask": true,
|
| 9 |
+
"sampling_rate": 16000
|
| 10 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"bos_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"is_local": true,
|
| 6 |
+
"local_files_only": false,
|
| 7 |
+
"model_max_length": 4096,
|
| 8 |
+
"pad_token": "<unk>",
|
| 9 |
+
"tokenizer_class": "TokenizersBackend",
|
| 10 |
+
"unk_token": "<unk>"
|
| 11 |
+
}
|