Hugging Face's logo

marma
/
whisper-small-sv

Automatic Speech Recognition
Transformers
PyTorch
whisper
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
KB commited on
Commit
78d2835
•
1 Parent(s): 4e83fbe

New 20k steps version

Browse files
Files changed (2) hide show
  1. README.md +86 -26
  2. pytorch_model.bin +1 -1
README.md CHANGED
@@ -21,7 +21,7 @@ model-index:
21
  metrics:
22
  - name: Wer
23
  type: wer
24
- value: 0.2426515530366172
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -31,8 +31,8 @@ should probably proofread and complete it, then remove this comment. -->
31
 
32
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the dataset/riksdagen audiofolder dataset.
33
  It achieves the following results on the evaluation set:
34
- - Loss: 0.3479
35
- - Wer: 0.2427
36
 
37
  ## Model description
38
 
@@ -58,32 +58,92 @@ The following hyperparameters were used during training:
58
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
59
  - lr_scheduler_type: linear
60
  - lr_scheduler_warmup_steps: 100
61
- - training_steps: 5000
62
 
63
  ### Training results
64
 
65
- | Training Loss | Epoch | Step | Validation Loss | Wer |
66
- |:-------------:|:-----:|:----:|:---------------:|:------:|
67
- | 0.5024 | 0.04 | 250 | 0.5073 | 0.2948 |
68
- | 0.4684 | 0.08 | 500 | 0.4639 | 0.2784 |
69
- | 0.4246 | 0.12 | 750 | 0.4396 | 0.2758 |
70
- | 0.4132 | 0.17 | 1000 | 0.4222 | 0.2664 |
71
- | 0.4021 | 0.21 | 1250 | 0.4101 | 0.2633 |
72
- | 0.3871 | 0.25 | 1500 | 0.3982 | 0.2619 |
73
- | 0.3813 | 0.29 | 1750 | 0.3895 | 0.2577 |
74
- | 0.3878 | 0.33 | 2000 | 0.3827 | 0.2533 |
75
- | 0.3704 | 0.37 | 2250 | 0.3770 | 0.2533 |
76
- | 0.3516 | 0.42 | 2500 | 0.3714 | 0.2540 |
77
- | 0.3792 | 0.46 | 2750 | 0.3675 | 0.2495 |
78
- | 0.3476 | 0.5 | 3000 | 0.3636 | 0.2456 |
79
- | 0.3522 | 0.54 | 3250 | 0.3611 | 0.2462 |
80
- | 0.3545 | 0.58 | 3500 | 0.3560 | 0.2440 |
81
- | 0.3426 | 0.62 | 3750 | 0.3543 | 0.2464 |
82
- | 0.3437 | 0.66 | 4000 | 0.3524 | 0.2464 |
83
- | 0.3562 | 0.71 | 4250 | 0.3507 | 0.2452 |
84
- | 0.3555 | 0.75 | 4500 | 0.3491 | 0.2426 |
85
- | 0.3397 | 0.79 | 4750 | 0.3483 | 0.2419 |
86
- | 0.3516 | 0.83 | 5000 | 0.3479 | 0.2427 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
87
 
88
 
89
  ### Framework versions
 
21
  metrics:
22
  - name: Wer
23
  type: wer
24
+ value: 0.22405586116204554
25
  ---
26
 
27
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
31
 
32
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the dataset/riksdagen audiofolder dataset.
33
  It achieves the following results on the evaluation set:
34
+ - Loss: 0.2917
35
+ - Wer: 0.2241
36
 
37
  ## Model description
38
 
 
58
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
59
  - lr_scheduler_type: linear
60
  - lr_scheduler_warmup_steps: 100
61
+ - training_steps: 20000
62
 
63
  ### Training results
64
 
65
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
66
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|
67
+ | 0.5023 | 0.04 | 250 | 0.5072 | 0.2949 |
68
+ | 0.4678 | 0.08 | 500 | 0.4632 | 0.2780 |
69
+ | 0.4233 | 0.12 | 750 | 0.4384 | 0.2749 |
70
+ | 0.4113 | 0.17 | 1000 | 0.4205 | 0.2673 |
71
+ | 0.3994 | 0.21 | 1250 | 0.4079 | 0.2649 |
72
+ | 0.3841 | 0.25 | 1500 | 0.3947 | 0.2609 |
73
+ | 0.3775 | 0.29 | 1750 | 0.3854 | 0.2564 |
74
+ | 0.383 | 0.33 | 2000 | 0.3781 | 0.2540 |
75
+ | 0.3651 | 0.37 | 2250 | 0.3721 | 0.2532 |
76
+ | 0.3456 | 0.42 | 2500 | 0.3651 | 0.2517 |
77
+ | 0.3719 | 0.46 | 2750 | 0.3612 | 0.2481 |
78
+ | 0.3399 | 0.5 | 3000 | 0.3561 | 0.2437 |
79
+ | 0.3428 | 0.54 | 3250 | 0.3522 | 0.2465 |
80
+ | 0.3442 | 0.58 | 3500 | 0.3451 | 0.2399 |
81
+ | 0.3315 | 0.62 | 3750 | 0.3431 | 0.2417 |
82
+ | 0.3299 | 0.66 | 4000 | 0.3404 | 0.2428 |
83
+ | 0.3417 | 0.71 | 4250 | 0.3373 | 0.2395 |
84
+ | 0.3399 | 0.75 | 4500 | 0.3332 | 0.2390 |
85
+ | 0.3222 | 0.79 | 4750 | 0.3310 | 0.2385 |
86
+ | 0.3319 | 0.83 | 5000 | 0.3291 | 0.2372 |
87
+ | 0.3188 | 0.87 | 5250 | 0.3265 | 0.2359 |
88
+ | 0.3197 | 0.91 | 5500 | 0.3240 | 0.2378 |
89
+ | 0.3099 | 0.96 | 5750 | 0.3215 | 0.2342 |
90
+ | 0.3132 | 1.0 | 6000 | 0.3195 | 0.2374 |
91
+ | 0.286 | 1.04 | 6250 | 0.3179 | 0.2348 |
92
+ | 0.2765 | 1.08 | 6500 | 0.3166 | 0.2354 |
93
+ | 0.2795 | 1.12 | 6750 | 0.3153 | 0.2324 |
94
+ | 0.2825 | 1.16 | 7000 | 0.3145 | 0.2316 |
95
+ | 0.2865 | 1.21 | 7250 | 0.3144 | 0.2329 |
96
+ | 0.2703 | 1.25 | 7500 | 0.3126 | 0.2326 |
97
+ | 0.2792 | 1.29 | 7750 | 0.3121 | 0.2324 |
98
+ | 0.2749 | 1.33 | 8000 | 0.3106 | 0.2325 |
99
+ | 0.2762 | 1.37 | 8250 | 0.3093 | 0.2315 |
100
+ | 0.2813 | 1.41 | 8500 | 0.3080 | 0.2302 |
101
+ | 0.2755 | 1.45 | 8750 | 0.3078 | 0.2321 |
102
+ | 0.2779 | 1.5 | 9000 | 0.3062 | 0.2305 |
103
+ | 0.2764 | 1.54 | 9250 | 0.3059 | 0.2336 |
104
+ | 0.2763 | 1.58 | 9500 | 0.3041 | 0.2310 |
105
+ | 0.2723 | 1.62 | 9750 | 0.3027 | 0.2292 |
106
+ | 0.2756 | 1.66 | 10000 | 0.3026 | 0.2301 |
107
+ | 0.2663 | 1.7 | 10250 | 0.3008 | 0.2262 |
108
+ | 0.269 | 1.75 | 10500 | 0.3006 | 0.2280 |
109
+ | 0.2682 | 1.79 | 10750 | 0.3002 | 0.2291 |
110
+ | 0.2721 | 1.83 | 11000 | 0.2994 | 0.2267 |
111
+ | 0.2681 | 1.87 | 11250 | 0.2987 | 0.2288 |
112
+ | 0.278 | 1.91 | 11500 | 0.2978 | 0.2296 |
113
+ | 0.2625 | 1.95 | 11750 | 0.2978 | 0.2278 |
114
+ | 0.2583 | 1.99 | 12000 | 0.2967 | 0.2259 |
115
+ | 0.2403 | 2.04 | 12250 | 0.2976 | 0.2276 |
116
+ | 0.2414 | 2.08 | 12500 | 0.2972 | 0.2264 |
117
+ | 0.251 | 2.12 | 12750 | 0.2969 | 0.2256 |
118
+ | 0.2404 | 2.16 | 13000 | 0.2968 | 0.2253 |
119
+ | 0.2473 | 2.2 | 13250 | 0.2966 | 0.2253 |
120
+ | 0.2444 | 2.24 | 13500 | 0.2965 | 0.2262 |
121
+ | 0.2512 | 2.29 | 13750 | 0.2962 | 0.2253 |
122
+ | 0.2417 | 2.33 | 14000 | 0.2950 | 0.2280 |
123
+ | 0.2445 | 2.37 | 14250 | 0.2950 | 0.2256 |
124
+ | 0.2461 | 2.41 | 14500 | 0.2949 | 0.2262 |
125
+ | 0.2496 | 2.45 | 14750 | 0.2944 | 0.2261 |
126
+ | 0.2422 | 2.49 | 15000 | 0.2942 | 0.2248 |
127
+ | 0.2415 | 2.53 | 15250 | 0.2940 | 0.2252 |
128
+ | 0.2465 | 2.58 | 15500 | 0.2932 | 0.2269 |
129
+ | 0.2508 | 2.62 | 15750 | 0.2931 | 0.2245 |
130
+ | 0.2339 | 2.66 | 16000 | 0.2930 | 0.2257 |
131
+ | 0.2441 | 2.7 | 16250 | 0.2923 | 0.2247 |
132
+ | 0.2444 | 2.74 | 16500 | 0.2921 | 0.2246 |
133
+ | 0.2416 | 2.78 | 16750 | 0.2918 | 0.2264 |
134
+ | 0.2425 | 2.83 | 17000 | 0.2916 | 0.2251 |
135
+ | 0.2404 | 2.87 | 17250 | 0.2916 | 0.2234 |
136
+ | 0.2456 | 2.91 | 17500 | 0.2911 | 0.2238 |
137
+ | 0.2384 | 2.95 | 17750 | 0.2908 | 0.2252 |
138
+ | 0.244 | 2.99 | 18000 | 0.2905 | 0.2251 |
139
+ | 0.2197 | 3.03 | 18250 | 0.2919 | 0.2239 |
140
+ | 0.2194 | 3.08 | 18500 | 0.2919 | 0.2237 |
141
+ | 0.2294 | 3.12 | 18750 | 0.2919 | 0.2243 |
142
+ | 0.2225 | 3.16 | 19000 | 0.2918 | 0.2252 |
143
+ | 0.2229 | 3.2 | 19250 | 0.2919 | 0.2242 |
144
+ | 0.2153 | 3.24 | 19500 | 0.2917 | 0.2241 |
145
+ | 0.2137 | 3.28 | 19750 | 0.2917 | 0.2239 |
146
+ | 0.2194 | 3.32 | 20000 | 0.2917 | 0.2241 |
147
 
148
 
149
  ### Framework versions
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f5a4aa7b26913f28641ab05e515a09c91b47042cad88deac851dba122c6f49da
3
  size 967099139
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0b5b0c4522bedf6f5cd4fbeff0d53c538af8810548c22644eb2fe82c2f4fa615
3
  size 967099139