Upload FlaxWhisperForConditionalGeneration
#28
by
Shir02021
- opened
- config.json +11 -3
- generation_config.json +2 -4
config.json
CHANGED
@@ -2,6 +2,7 @@
|
|
2 |
"_name_or_path": "openai/whisper-tiny",
|
3 |
"activation_dropout": 0.0,
|
4 |
"activation_function": "gelu",
|
|
|
5 |
"architectures": [
|
6 |
"WhisperForConditionalGeneration"
|
7 |
],
|
@@ -11,6 +12,7 @@
|
|
11 |
50257
|
12 |
],
|
13 |
"bos_token_id": 50257,
|
|
|
14 |
"d_model": 384,
|
15 |
"decoder_attention_heads": 6,
|
16 |
"decoder_ffn_dim": 1536,
|
@@ -39,9 +41,16 @@
|
|
39 |
],
|
40 |
"init_std": 0.02,
|
41 |
"is_encoder_decoder": true,
|
|
|
|
|
|
|
|
|
|
|
|
|
42 |
"max_length": 448,
|
43 |
"max_source_positions": 1500,
|
44 |
"max_target_positions": 448,
|
|
|
45 |
"model_type": "whisper",
|
46 |
"num_hidden_layers": 4,
|
47 |
"num_mel_bins": 80,
|
@@ -131,14 +140,13 @@
|
|
131 |
49870,
|
132 |
50254,
|
133 |
50258,
|
134 |
-
50358,
|
135 |
-
50359,
|
136 |
50360,
|
137 |
50361,
|
138 |
50362
|
139 |
],
|
140 |
"torch_dtype": "float32",
|
141 |
-
"transformers_version": "4.
|
142 |
"use_cache": true,
|
|
|
143 |
"vocab_size": 51865
|
144 |
}
|
|
|
2 |
"_name_or_path": "openai/whisper-tiny",
|
3 |
"activation_dropout": 0.0,
|
4 |
"activation_function": "gelu",
|
5 |
+
"apply_spec_augment": false,
|
6 |
"architectures": [
|
7 |
"WhisperForConditionalGeneration"
|
8 |
],
|
|
|
12 |
50257
|
13 |
],
|
14 |
"bos_token_id": 50257,
|
15 |
+
"classifier_proj_size": 256,
|
16 |
"d_model": 384,
|
17 |
"decoder_attention_heads": 6,
|
18 |
"decoder_ffn_dim": 1536,
|
|
|
41 |
],
|
42 |
"init_std": 0.02,
|
43 |
"is_encoder_decoder": true,
|
44 |
+
"mask_feature_length": 10,
|
45 |
+
"mask_feature_min_masks": 0,
|
46 |
+
"mask_feature_prob": 0.0,
|
47 |
+
"mask_time_length": 10,
|
48 |
+
"mask_time_min_masks": 2,
|
49 |
+
"mask_time_prob": 0.05,
|
50 |
"max_length": 448,
|
51 |
"max_source_positions": 1500,
|
52 |
"max_target_positions": 448,
|
53 |
+
"median_filter_width": 7,
|
54 |
"model_type": "whisper",
|
55 |
"num_hidden_layers": 4,
|
56 |
"num_mel_bins": 80,
|
|
|
140 |
49870,
|
141 |
50254,
|
142 |
50258,
|
|
|
|
|
143 |
50360,
|
144 |
50361,
|
145 |
50362
|
146 |
],
|
147 |
"torch_dtype": "float32",
|
148 |
+
"transformers_version": "4.31.0",
|
149 |
"use_cache": true,
|
150 |
+
"use_weighted_layer_sum": false,
|
151 |
"vocab_size": 51865
|
152 |
}
|
generation_config.json
CHANGED
@@ -207,8 +207,6 @@
|
|
207 |
49870,
|
208 |
50254,
|
209 |
50258,
|
210 |
-
50358,
|
211 |
-
50359,
|
212 |
50360,
|
213 |
50361,
|
214 |
50362
|
@@ -217,5 +215,5 @@
|
|
217 |
"transcribe": 50359,
|
218 |
"translate": 50358
|
219 |
},
|
220 |
-
"transformers_version": "4.
|
221 |
-
}
|
|
|
207 |
49870,
|
208 |
50254,
|
209 |
50258,
|
|
|
|
|
210 |
50360,
|
211 |
50361,
|
212 |
50362
|
|
|
215 |
"transcribe": 50359,
|
216 |
"translate": 50358
|
217 |
},
|
218 |
+
"transformers_version": "4.31.0"
|
219 |
+
}
|