dmusingu commited on
Commit
05530a3
·
verified ·
1 Parent(s): 27e2016

muk-luganda-digits-classification-clean

Browse files
Files changed (5) hide show
  1. README.md +111 -0
  2. config.json +141 -0
  3. model.safetensors +3 -0
  4. preprocessor_config.json +9 -0
  5. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,111 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: apache-2.0
4
+ base_model: facebook/wav2vec2-base
5
+ tags:
6
+ - generated_from_trainer
7
+ metrics:
8
+ - accuracy
9
+ model-index:
10
+ - name: muk-luganda-digits-classification-clean
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # muk-luganda-digits-classification-clean
18
+
19
+ This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on an unknown dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 3.2912
22
+ - Accuracy: 0.4444
23
+
24
+ ## Model description
25
+
26
+ More information needed
27
+
28
+ ## Intended uses & limitations
29
+
30
+ More information needed
31
+
32
+ ## Training and evaluation data
33
+
34
+ More information needed
35
+
36
+ ## Training procedure
37
+
38
+ ### Training hyperparameters
39
+
40
+ The following hyperparameters were used during training:
41
+ - learning_rate: 3e-05
42
+ - train_batch_size: 4
43
+ - eval_batch_size: 1
44
+ - seed: 42
45
+ - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
46
+ - lr_scheduler_type: linear
47
+ - lr_scheduler_warmup_ratio: 0.1
48
+ - num_epochs: 50
49
+
50
+ ### Training results
51
+
52
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
53
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
54
+ | 2.3031 | 1.0 | 54 | 2.3087 | 0.1111 |
55
+ | 2.2996 | 2.0 | 108 | 2.3192 | 0.0370 |
56
+ | 2.2895 | 3.0 | 162 | 2.3097 | 0.0741 |
57
+ | 2.2775 | 4.0 | 216 | 2.3196 | 0.0741 |
58
+ | 2.2559 | 5.0 | 270 | 2.3122 | 0.1852 |
59
+ | 2.2087 | 6.0 | 324 | 2.2864 | 0.1481 |
60
+ | 2.1098 | 7.0 | 378 | 2.2359 | 0.2222 |
61
+ | 2.0347 | 8.0 | 432 | 2.1798 | 0.2222 |
62
+ | 1.8935 | 9.0 | 486 | 2.0581 | 0.3333 |
63
+ | 1.8424 | 10.0 | 540 | 2.0617 | 0.3333 |
64
+ | 1.6853 | 11.0 | 594 | 2.0774 | 0.3333 |
65
+ | 1.6062 | 12.0 | 648 | 1.9688 | 0.4074 |
66
+ | 1.4425 | 13.0 | 702 | 1.8905 | 0.4444 |
67
+ | 1.3637 | 14.0 | 756 | 1.8887 | 0.4074 |
68
+ | 1.2574 | 15.0 | 810 | 1.8752 | 0.4444 |
69
+ | 1.1705 | 16.0 | 864 | 1.8439 | 0.4815 |
70
+ | 1.0439 | 17.0 | 918 | 1.8602 | 0.4074 |
71
+ | 0.9871 | 18.0 | 972 | 1.7512 | 0.4074 |
72
+ | 0.949 | 19.0 | 1026 | 2.0009 | 0.3704 |
73
+ | 0.8252 | 20.0 | 1080 | 2.0860 | 0.4444 |
74
+ | 0.7124 | 21.0 | 1134 | 1.8057 | 0.4815 |
75
+ | 0.6594 | 22.0 | 1188 | 1.9824 | 0.4074 |
76
+ | 0.5448 | 23.0 | 1242 | 2.3564 | 0.3333 |
77
+ | 0.502 | 24.0 | 1296 | 2.0142 | 0.4444 |
78
+ | 0.4108 | 25.0 | 1350 | 2.1864 | 0.4074 |
79
+ | 0.4095 | 26.0 | 1404 | 2.4376 | 0.4074 |
80
+ | 0.3495 | 27.0 | 1458 | 2.5146 | 0.4444 |
81
+ | 0.2818 | 28.0 | 1512 | 2.3985 | 0.4444 |
82
+ | 0.3269 | 29.0 | 1566 | 2.7282 | 0.4444 |
83
+ | 0.2856 | 30.0 | 1620 | 2.2764 | 0.4815 |
84
+ | 0.2542 | 31.0 | 1674 | 2.5317 | 0.4444 |
85
+ | 0.2205 | 32.0 | 1728 | 2.4492 | 0.4444 |
86
+ | 0.2585 | 33.0 | 1782 | 3.0590 | 0.4074 |
87
+ | 0.2258 | 34.0 | 1836 | 2.7979 | 0.4444 |
88
+ | 0.1818 | 35.0 | 1890 | 2.6064 | 0.4815 |
89
+ | 0.1975 | 36.0 | 1944 | 2.7659 | 0.4444 |
90
+ | 0.2536 | 37.0 | 1998 | 2.8510 | 0.4444 |
91
+ | 0.1456 | 38.0 | 2052 | 2.8677 | 0.4444 |
92
+ | 0.2197 | 39.0 | 2106 | 3.1799 | 0.4074 |
93
+ | 0.2195 | 40.0 | 2160 | 3.1353 | 0.4074 |
94
+ | 0.1617 | 41.0 | 2214 | 2.9508 | 0.4815 |
95
+ | 0.2328 | 42.0 | 2268 | 3.0263 | 0.4815 |
96
+ | 0.1884 | 43.0 | 2322 | 2.8975 | 0.4444 |
97
+ | 0.2127 | 44.0 | 2376 | 2.9817 | 0.4815 |
98
+ | 0.1545 | 45.0 | 2430 | 3.1183 | 0.4444 |
99
+ | 0.166 | 46.0 | 2484 | 3.1063 | 0.4444 |
100
+ | 0.1659 | 47.0 | 2538 | 3.2661 | 0.4444 |
101
+ | 0.1923 | 48.0 | 2592 | 3.2802 | 0.4074 |
102
+ | 0.2008 | 49.0 | 2646 | 3.2903 | 0.4444 |
103
+ | 0.1849 | 50.0 | 2700 | 3.2912 | 0.4444 |
104
+
105
+
106
+ ### Framework versions
107
+
108
+ - Transformers 4.56.2
109
+ - Pytorch 2.8.0+cu128
110
+ - Datasets 3.6.0
111
+ - Tokenizers 0.22.1
config.json ADDED
@@ -0,0 +1,141 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "adapter_attn_dim": null,
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": false,
7
+ "apply_spec_augment": true,
8
+ "architectures": [
9
+ "Wav2Vec2ForSequenceClassification"
10
+ ],
11
+ "attention_dropout": 0.1,
12
+ "bos_token_id": 1,
13
+ "classifier_proj_size": 256,
14
+ "codevector_dim": 256,
15
+ "contrastive_logits_temperature": 0.1,
16
+ "conv_bias": false,
17
+ "conv_dim": [
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512,
22
+ 512,
23
+ 512,
24
+ 512
25
+ ],
26
+ "conv_kernel": [
27
+ 10,
28
+ 3,
29
+ 3,
30
+ 3,
31
+ 3,
32
+ 2,
33
+ 2
34
+ ],
35
+ "conv_stride": [
36
+ 5,
37
+ 2,
38
+ 2,
39
+ 2,
40
+ 2,
41
+ 2,
42
+ 2
43
+ ],
44
+ "ctc_loss_reduction": "sum",
45
+ "ctc_zero_infinity": false,
46
+ "diversity_loss_weight": 0.1,
47
+ "do_stable_layer_norm": false,
48
+ "dtype": "float32",
49
+ "eos_token_id": 2,
50
+ "feat_extract_activation": "gelu",
51
+ "feat_extract_norm": "group",
52
+ "feat_proj_dropout": 0.1,
53
+ "feat_quantizer_dropout": 0.0,
54
+ "final_dropout": 0.0,
55
+ "freeze_feat_extract_train": true,
56
+ "hidden_act": "gelu",
57
+ "hidden_dropout": 0.1,
58
+ "hidden_size": 768,
59
+ "id2label": {
60
+ "0": "notti",
61
+ "1": "emu",
62
+ "2": "bbiri",
63
+ "3": "ssatu",
64
+ "4": "nnya",
65
+ "5": "ttaano",
66
+ "6": "mukaaga",
67
+ "7": "musanvu",
68
+ "8": "munaana",
69
+ "9": "mwenda"
70
+ },
71
+ "initializer_range": 0.02,
72
+ "intermediate_size": 3072,
73
+ "label2id": {
74
+ "bbiri": 2,
75
+ "emu": 1,
76
+ "mukaaga": 6,
77
+ "munaana": 8,
78
+ "musanvu": 7,
79
+ "mwenda": 9,
80
+ "nnya": 4,
81
+ "notti": 0,
82
+ "ssatu": 3,
83
+ "ttaano": 5
84
+ },
85
+ "layer_norm_eps": 1e-05,
86
+ "layerdrop": 0.0,
87
+ "mask_channel_length": 10,
88
+ "mask_channel_min_space": 1,
89
+ "mask_channel_other": 0.0,
90
+ "mask_channel_prob": 0.0,
91
+ "mask_channel_selection": "static",
92
+ "mask_feature_length": 10,
93
+ "mask_feature_min_masks": 0,
94
+ "mask_feature_prob": 0.0,
95
+ "mask_time_length": 10,
96
+ "mask_time_min_masks": 2,
97
+ "mask_time_min_space": 1,
98
+ "mask_time_other": 0.0,
99
+ "mask_time_prob": 0.05,
100
+ "mask_time_selection": "static",
101
+ "model_type": "wav2vec2",
102
+ "no_mask_channel_overlap": false,
103
+ "no_mask_time_overlap": false,
104
+ "num_adapter_layers": 3,
105
+ "num_attention_heads": 12,
106
+ "num_codevector_groups": 2,
107
+ "num_codevectors_per_group": 320,
108
+ "num_conv_pos_embedding_groups": 16,
109
+ "num_conv_pos_embeddings": 128,
110
+ "num_feat_extract_layers": 7,
111
+ "num_hidden_layers": 12,
112
+ "num_negatives": 100,
113
+ "output_hidden_size": 768,
114
+ "pad_token_id": 0,
115
+ "proj_codevector_dim": 256,
116
+ "tdnn_dilation": [
117
+ 1,
118
+ 2,
119
+ 3,
120
+ 1,
121
+ 1
122
+ ],
123
+ "tdnn_dim": [
124
+ 512,
125
+ 512,
126
+ 512,
127
+ 512,
128
+ 1500
129
+ ],
130
+ "tdnn_kernel": [
131
+ 5,
132
+ 3,
133
+ 3,
134
+ 1,
135
+ 1
136
+ ],
137
+ "transformers_version": "4.56.2",
138
+ "use_weighted_layer_sum": false,
139
+ "vocab_size": 32,
140
+ "xvector_output_dim": 512
141
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:49dd6410b582dc09dc72d8f4fc85ce22a9b88e2764fae23fe2a763c8240ee4fe
3
+ size 378310592
preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0.0,
7
+ "return_attention_mask": false,
8
+ "sampling_rate": 16000
9
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55bcf331244893a24d3f31a7684942dce2937919831b33593940feba99dcb87e
3
+ size 5841