Add auto classes
Browse files- config.json +7 -6
- preprocessor_config.json +8 -1
- tokenizer_config.json +4 -1
config.json
CHANGED
@@ -5,8 +5,8 @@
|
|
5 |
"AutoProcessor": "gigaam_transformers.GigaAMProcessor",
|
6 |
"AutoTokenizer": "gigaam_transformers.GigaAMCTCTokenizer",
|
7 |
"AutoFeatureExtractor": "gigaam_transformers.GigaAMFeatureExtractor"
|
8 |
-
|
9 |
-
|
10 |
"encoder": {
|
11 |
"feat_in": 64,
|
12 |
"n_layers": 16,
|
@@ -18,11 +18,11 @@
|
|
18 |
"n_heads": 16,
|
19 |
"conv_kernel_size": 31,
|
20 |
"flash_attn": false
|
21 |
-
|
22 |
"head": {
|
23 |
"feat_in": 768,
|
24 |
"num_classes": 34
|
25 |
-
|
26 |
"labels": [
|
27 |
" ",
|
28 |
"а",
|
@@ -57,6 +57,7 @@
|
|
57 |
"э",
|
58 |
"ю",
|
59 |
"я"
|
60 |
-
|
61 |
-
|
|
|
62 |
}
|
|
|
5 |
"AutoProcessor": "gigaam_transformers.GigaAMProcessor",
|
6 |
"AutoTokenizer": "gigaam_transformers.GigaAMCTCTokenizer",
|
7 |
"AutoFeatureExtractor": "gigaam_transformers.GigaAMFeatureExtractor"
|
8 |
+
},
|
9 |
+
|
10 |
"encoder": {
|
11 |
"feat_in": 64,
|
12 |
"n_layers": 16,
|
|
|
18 |
"n_heads": 16,
|
19 |
"conv_kernel_size": 31,
|
20 |
"flash_attn": false
|
21 |
+
},
|
22 |
"head": {
|
23 |
"feat_in": 768,
|
24 |
"num_classes": 34
|
25 |
+
},
|
26 |
"labels": [
|
27 |
" ",
|
28 |
"а",
|
|
|
57 |
"э",
|
58 |
"ю",
|
59 |
"я"
|
60 |
+
],
|
61 |
+
"blank_id": 33,
|
62 |
+
"model_type": "gigaam-ctc"
|
63 |
}
|
preprocessor_config.json
CHANGED
@@ -1,11 +1,18 @@
|
|
1 |
{
|
2 |
"chunk_length": 30,
|
3 |
"feature_extractor_type": "GigaAMFeatureExtractor",
|
|
|
4 |
"feature_size": 64,
|
5 |
"hop_length": 160,
|
6 |
"n_samples": 480000,
|
7 |
"padding_side": "right",
|
8 |
"padding_value": 0.0,
|
9 |
"return_attention_mask": true,
|
10 |
-
"sampling_rate": 16000
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
}
|
|
|
1 |
{
|
2 |
"chunk_length": 30,
|
3 |
"feature_extractor_type": "GigaAMFeatureExtractor",
|
4 |
+
"feature_extractor_class": "GigaAMFeatureExtractor",
|
5 |
"feature_size": 64,
|
6 |
"hop_length": 160,
|
7 |
"n_samples": 480000,
|
8 |
"padding_side": "right",
|
9 |
"padding_value": 0.0,
|
10 |
"return_attention_mask": true,
|
11 |
+
"sampling_rate": 16000,
|
12 |
+
"auto_map": {
|
13 |
+
"AutoFeatureExtractor": "gigaam_transformers.GigaAMFeatureExtractor",
|
14 |
+
"AutoProcessor": "gigaam_transformers.GigaAMProcessor"
|
15 |
+
},
|
16 |
+
"processor_class": "GigaAMProcessor",
|
17 |
+
"model_type": "gigaam-ctc"
|
18 |
}
|
tokenizer_config.json
CHANGED
@@ -19,5 +19,8 @@
|
|
19 |
"target_lang": null,
|
20 |
"tokenizer_class": "GigaAMCTCTokenizer",
|
21 |
"unk_token": "[BLANK]",
|
22 |
-
"word_delimiter_token": " "
|
|
|
|
|
|
|
23 |
}
|
|
|
19 |
"target_lang": null,
|
20 |
"tokenizer_class": "GigaAMCTCTokenizer",
|
21 |
"unk_token": "[BLANK]",
|
22 |
+
"word_delimiter_token": " ",
|
23 |
+
"auto_map": {
|
24 |
+
"AutoTokenizer": ["gigaam_transformers.GigaAMCTCTokenizer", null]
|
25 |
+
}
|
26 |
}
|