Upload config.json with huggingface_hub
Browse files- config.json +65 -0
config.json
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "THUDM/glm-4-voice-tokenizer",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "gelu",
|
| 5 |
+
"apply_spec_augment": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"WhisperVQEncoder"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"begin_suppress_tokens": [
|
| 11 |
+
220,
|
| 12 |
+
50257
|
| 13 |
+
],
|
| 14 |
+
"bos_token_id": 50257,
|
| 15 |
+
"classifier_proj_size": 256,
|
| 16 |
+
"d_model": 1280,
|
| 17 |
+
"decoder_attention_heads": 20,
|
| 18 |
+
"decoder_ffn_dim": 5120,
|
| 19 |
+
"decoder_layerdrop": 0.0,
|
| 20 |
+
"decoder_layers": 32,
|
| 21 |
+
"decoder_start_token_id": 50258,
|
| 22 |
+
"dropout": 0.0,
|
| 23 |
+
"encoder_attention_heads": 20,
|
| 24 |
+
"encoder_causal_attention": false,
|
| 25 |
+
"encoder_causal_convolution": true,
|
| 26 |
+
"encoder_ffn_dim": 5120,
|
| 27 |
+
"encoder_layerdrop": 0.0,
|
| 28 |
+
"encoder_layers": 32,
|
| 29 |
+
"eos_token_id": 50257,
|
| 30 |
+
"init_std": 0.02,
|
| 31 |
+
"is_encoder_decoder": true,
|
| 32 |
+
"mask_feature_length": 10,
|
| 33 |
+
"mask_feature_min_masks": 0,
|
| 34 |
+
"mask_feature_prob": 0.0,
|
| 35 |
+
"mask_time_length": 10,
|
| 36 |
+
"mask_time_min_masks": 2,
|
| 37 |
+
"mask_time_prob": 0.05,
|
| 38 |
+
"max_length": 448,
|
| 39 |
+
"max_source_positions": 1500,
|
| 40 |
+
"max_target_positions": 448,
|
| 41 |
+
"median_filter_width": 7,
|
| 42 |
+
"model_type": "whisper",
|
| 43 |
+
"num_hidden_layers": 32,
|
| 44 |
+
"num_mel_bins": 128,
|
| 45 |
+
"pad_token_id": 50256,
|
| 46 |
+
"pooling_kernel_size": 4,
|
| 47 |
+
"pooling_position": 16,
|
| 48 |
+
"pooling_type": "avg",
|
| 49 |
+
"quantize_causal_block_size": 200,
|
| 50 |
+
"quantize_causal_encoder": false,
|
| 51 |
+
"quantize_commit_coefficient": 0.25,
|
| 52 |
+
"quantize_ema_decay": 0.99,
|
| 53 |
+
"quantize_encoder_only": true,
|
| 54 |
+
"quantize_loss_scale": 10.0,
|
| 55 |
+
"quantize_position": 16,
|
| 56 |
+
"quantize_restart_interval": 100,
|
| 57 |
+
"quantize_vocab_size": 16384,
|
| 58 |
+
"scale_embedding": false,
|
| 59 |
+
"skip_language_detection": true,
|
| 60 |
+
"torch_dtype": "float32",
|
| 61 |
+
"transformers_version": "4.44.1",
|
| 62 |
+
"use_cache": true,
|
| 63 |
+
"use_weighted_layer_sum": false,
|
| 64 |
+
"vocab_size": 51866
|
| 65 |
+
}
|