Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- checkpoint-100/config.json +47 -0
- checkpoint-100/generation_config.json +175 -0
- checkpoint-100/model.safetensors +3 -0
- checkpoint-100/optimizer.pt +3 -0
- checkpoint-100/preprocessor_config.json +15 -0
- checkpoint-100/rng_state.pth +3 -0
- checkpoint-100/scaler.pt +3 -0
- checkpoint-100/scheduler.pt +3 -0
- checkpoint-100/trainer_state.json +78 -0
- checkpoint-100/training_args.bin +3 -0
- checkpoint-200/model.safetensors +1 -1
- checkpoint-200/optimizer.pt +1 -1
- checkpoint-200/rng_state.pth +1 -1
- checkpoint-200/scaler.pt +1 -1
- checkpoint-200/scheduler.pt +1 -1
- checkpoint-200/trainer_state.json +49 -40
- checkpoint-200/training_args.bin +1 -1
- checkpoint-300/config.json +47 -0
- checkpoint-300/generation_config.json +175 -0
- checkpoint-300/model.safetensors +3 -0
- checkpoint-300/optimizer.pt +3 -0
- checkpoint-300/preprocessor_config.json +15 -0
- checkpoint-300/rng_state.pth +3 -0
- checkpoint-300/scaler.pt +3 -0
- checkpoint-300/scheduler.pt +3 -0
- checkpoint-300/trainer_state.json +166 -0
- checkpoint-300/training_args.bin +3 -0
- checkpoint-400/model.safetensors +1 -1
- checkpoint-400/optimizer.pt +1 -1
- checkpoint-400/rng_state.pth +1 -1
- checkpoint-400/scaler.pt +1 -1
- checkpoint-400/scheduler.pt +1 -1
- checkpoint-400/trainer_state.json +91 -73
- checkpoint-400/training_args.bin +1 -1
- checkpoint-500/config.json +47 -0
- checkpoint-500/generation_config.json +175 -0
- checkpoint-500/model.safetensors +3 -0
- checkpoint-500/optimizer.pt +3 -0
- checkpoint-500/preprocessor_config.json +15 -0
- checkpoint-500/rng_state.pth +3 -0
- checkpoint-500/scaler.pt +3 -0
- checkpoint-500/scheduler.pt +3 -0
- checkpoint-500/trainer_state.json +254 -0
- checkpoint-500/training_args.bin +3 -0
- checkpoint-600/config.json +47 -0
- checkpoint-600/generation_config.json +175 -0
- checkpoint-600/model.safetensors +3 -0
- checkpoint-600/optimizer.pt +3 -0
- checkpoint-600/preprocessor_config.json +15 -0
- checkpoint-600/rng_state.pth +3 -0
checkpoint-100/config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"WhisperForConditionalGeneration"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"begin_suppress_tokens": null,
|
| 10 |
+
"bos_token_id": 50257,
|
| 11 |
+
"classifier_proj_size": 256,
|
| 12 |
+
"d_model": 768,
|
| 13 |
+
"decoder_attention_heads": 12,
|
| 14 |
+
"decoder_ffn_dim": 3072,
|
| 15 |
+
"decoder_layerdrop": 0.0,
|
| 16 |
+
"decoder_layers": 12,
|
| 17 |
+
"decoder_start_token_id": 50258,
|
| 18 |
+
"dropout": 0.0,
|
| 19 |
+
"encoder_attention_heads": 12,
|
| 20 |
+
"encoder_ffn_dim": 3072,
|
| 21 |
+
"encoder_layerdrop": 0.0,
|
| 22 |
+
"encoder_layers": 12,
|
| 23 |
+
"eos_token_id": 50257,
|
| 24 |
+
"forced_decoder_ids": null,
|
| 25 |
+
"init_std": 0.02,
|
| 26 |
+
"is_encoder_decoder": true,
|
| 27 |
+
"mask_feature_length": 10,
|
| 28 |
+
"mask_feature_min_masks": 0,
|
| 29 |
+
"mask_feature_prob": 0.0,
|
| 30 |
+
"mask_time_length": 10,
|
| 31 |
+
"mask_time_min_masks": 2,
|
| 32 |
+
"mask_time_prob": 0.05,
|
| 33 |
+
"max_length": null,
|
| 34 |
+
"max_source_positions": 1500,
|
| 35 |
+
"max_target_positions": 448,
|
| 36 |
+
"median_filter_width": 7,
|
| 37 |
+
"model_type": "whisper",
|
| 38 |
+
"num_hidden_layers": 12,
|
| 39 |
+
"num_mel_bins": 80,
|
| 40 |
+
"pad_token_id": 50257,
|
| 41 |
+
"scale_embedding": false,
|
| 42 |
+
"torch_dtype": "float32",
|
| 43 |
+
"transformers_version": "4.51.3",
|
| 44 |
+
"use_cache": false,
|
| 45 |
+
"use_weighted_layer_sum": false,
|
| 46 |
+
"vocab_size": 51865
|
| 47 |
+
}
|
checkpoint-100/generation_config.json
ADDED
|
@@ -0,0 +1,175 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alignment_heads": [
|
| 3 |
+
[
|
| 4 |
+
5,
|
| 5 |
+
3
|
| 6 |
+
],
|
| 7 |
+
[
|
| 8 |
+
5,
|
| 9 |
+
9
|
| 10 |
+
],
|
| 11 |
+
[
|
| 12 |
+
8,
|
| 13 |
+
0
|
| 14 |
+
],
|
| 15 |
+
[
|
| 16 |
+
8,
|
| 17 |
+
4
|
| 18 |
+
],
|
| 19 |
+
[
|
| 20 |
+
8,
|
| 21 |
+
7
|
| 22 |
+
],
|
| 23 |
+
[
|
| 24 |
+
8,
|
| 25 |
+
8
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
9,
|
| 29 |
+
0
|
| 30 |
+
],
|
| 31 |
+
[
|
| 32 |
+
9,
|
| 33 |
+
7
|
| 34 |
+
],
|
| 35 |
+
[
|
| 36 |
+
9,
|
| 37 |
+
9
|
| 38 |
+
],
|
| 39 |
+
[
|
| 40 |
+
10,
|
| 41 |
+
5
|
| 42 |
+
]
|
| 43 |
+
],
|
| 44 |
+
"begin_suppress_tokens": [
|
| 45 |
+
220,
|
| 46 |
+
50257
|
| 47 |
+
],
|
| 48 |
+
"bos_token_id": 50257,
|
| 49 |
+
"decoder_start_token_id": 50258,
|
| 50 |
+
"eos_token_id": 50257,
|
| 51 |
+
"forced_decoder_ids": [
|
| 52 |
+
[
|
| 53 |
+
1,
|
| 54 |
+
null
|
| 55 |
+
],
|
| 56 |
+
[
|
| 57 |
+
2,
|
| 58 |
+
50359
|
| 59 |
+
]
|
| 60 |
+
],
|
| 61 |
+
"is_multilingual": true,
|
| 62 |
+
"lang_to_id": {
|
| 63 |
+
"<|af|>": 50327,
|
| 64 |
+
"<|am|>": 50334,
|
| 65 |
+
"<|ar|>": 50272,
|
| 66 |
+
"<|as|>": 50350,
|
| 67 |
+
"<|az|>": 50304,
|
| 68 |
+
"<|ba|>": 50355,
|
| 69 |
+
"<|be|>": 50330,
|
| 70 |
+
"<|bg|>": 50292,
|
| 71 |
+
"<|bn|>": 50302,
|
| 72 |
+
"<|bo|>": 50347,
|
| 73 |
+
"<|br|>": 50309,
|
| 74 |
+
"<|bs|>": 50315,
|
| 75 |
+
"<|ca|>": 50270,
|
| 76 |
+
"<|cs|>": 50283,
|
| 77 |
+
"<|cy|>": 50297,
|
| 78 |
+
"<|da|>": 50285,
|
| 79 |
+
"<|de|>": 50261,
|
| 80 |
+
"<|el|>": 50281,
|
| 81 |
+
"<|en|>": 50259,
|
| 82 |
+
"<|es|>": 50262,
|
| 83 |
+
"<|et|>": 50307,
|
| 84 |
+
"<|eu|>": 50310,
|
| 85 |
+
"<|fa|>": 50300,
|
| 86 |
+
"<|fi|>": 50277,
|
| 87 |
+
"<|fo|>": 50338,
|
| 88 |
+
"<|fr|>": 50265,
|
| 89 |
+
"<|gl|>": 50319,
|
| 90 |
+
"<|gu|>": 50333,
|
| 91 |
+
"<|haw|>": 50352,
|
| 92 |
+
"<|ha|>": 50354,
|
| 93 |
+
"<|he|>": 50279,
|
| 94 |
+
"<|hi|>": 50276,
|
| 95 |
+
"<|hr|>": 50291,
|
| 96 |
+
"<|ht|>": 50339,
|
| 97 |
+
"<|hu|>": 50286,
|
| 98 |
+
"<|hy|>": 50312,
|
| 99 |
+
"<|id|>": 50275,
|
| 100 |
+
"<|is|>": 50311,
|
| 101 |
+
"<|it|>": 50274,
|
| 102 |
+
"<|ja|>": 50266,
|
| 103 |
+
"<|jw|>": 50356,
|
| 104 |
+
"<|ka|>": 50329,
|
| 105 |
+
"<|kk|>": 50316,
|
| 106 |
+
"<|km|>": 50323,
|
| 107 |
+
"<|kn|>": 50306,
|
| 108 |
+
"<|ko|>": 50264,
|
| 109 |
+
"<|la|>": 50294,
|
| 110 |
+
"<|lb|>": 50345,
|
| 111 |
+
"<|ln|>": 50353,
|
| 112 |
+
"<|lo|>": 50336,
|
| 113 |
+
"<|lt|>": 50293,
|
| 114 |
+
"<|lv|>": 50301,
|
| 115 |
+
"<|mg|>": 50349,
|
| 116 |
+
"<|mi|>": 50295,
|
| 117 |
+
"<|mk|>": 50308,
|
| 118 |
+
"<|ml|>": 50296,
|
| 119 |
+
"<|mn|>": 50314,
|
| 120 |
+
"<|mr|>": 50320,
|
| 121 |
+
"<|ms|>": 50282,
|
| 122 |
+
"<|mt|>": 50343,
|
| 123 |
+
"<|my|>": 50346,
|
| 124 |
+
"<|ne|>": 50313,
|
| 125 |
+
"<|nl|>": 50271,
|
| 126 |
+
"<|nn|>": 50342,
|
| 127 |
+
"<|no|>": 50288,
|
| 128 |
+
"<|oc|>": 50328,
|
| 129 |
+
"<|pa|>": 50321,
|
| 130 |
+
"<|pl|>": 50269,
|
| 131 |
+
"<|ps|>": 50340,
|
| 132 |
+
"<|pt|>": 50267,
|
| 133 |
+
"<|ro|>": 50284,
|
| 134 |
+
"<|ru|>": 50263,
|
| 135 |
+
"<|sa|>": 50344,
|
| 136 |
+
"<|sd|>": 50332,
|
| 137 |
+
"<|si|>": 50322,
|
| 138 |
+
"<|sk|>": 50298,
|
| 139 |
+
"<|sl|>": 50305,
|
| 140 |
+
"<|sn|>": 50324,
|
| 141 |
+
"<|so|>": 50326,
|
| 142 |
+
"<|sq|>": 50317,
|
| 143 |
+
"<|sr|>": 50303,
|
| 144 |
+
"<|su|>": 50357,
|
| 145 |
+
"<|sv|>": 50273,
|
| 146 |
+
"<|sw|>": 50318,
|
| 147 |
+
"<|ta|>": 50287,
|
| 148 |
+
"<|te|>": 50299,
|
| 149 |
+
"<|tg|>": 50331,
|
| 150 |
+
"<|th|>": 50289,
|
| 151 |
+
"<|tk|>": 50341,
|
| 152 |
+
"<|tl|>": 50348,
|
| 153 |
+
"<|tr|>": 50268,
|
| 154 |
+
"<|tt|>": 50351,
|
| 155 |
+
"<|uk|>": 50280,
|
| 156 |
+
"<|ur|>": 50290,
|
| 157 |
+
"<|uz|>": 50337,
|
| 158 |
+
"<|vi|>": 50278,
|
| 159 |
+
"<|yi|>": 50335,
|
| 160 |
+
"<|yo|>": 50325,
|
| 161 |
+
"<|zh|>": 50260
|
| 162 |
+
},
|
| 163 |
+
"max_initial_timestamp_index": 50,
|
| 164 |
+
"max_length": 448,
|
| 165 |
+
"no_timestamps_token_id": 50363,
|
| 166 |
+
"pad_token_id": 50257,
|
| 167 |
+
"prev_sot_token_id": 50361,
|
| 168 |
+
"return_timestamps": false,
|
| 169 |
+
"suppress_tokens": [],
|
| 170 |
+
"task_to_id": {
|
| 171 |
+
"transcribe": 50359,
|
| 172 |
+
"translate": 50358
|
| 173 |
+
},
|
| 174 |
+
"transformers_version": "4.51.3"
|
| 175 |
+
}
|
checkpoint-100/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:657c075db432a88a3f9904c83183b312d2d62c7593c49e6ff59061f775381796
|
| 3 |
+
size 966995080
|
checkpoint-100/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb528f0a4edb75abac73e8218838a7cbc024cadebb3b4d6f30bcce2305919f3d
|
| 3 |
+
size 1925064044
|
checkpoint-100/preprocessor_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"chunk_length": 30,
|
| 3 |
+
"dither": 0.0,
|
| 4 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 5 |
+
"feature_size": 80,
|
| 6 |
+
"hop_length": 160,
|
| 7 |
+
"n_fft": 400,
|
| 8 |
+
"n_samples": 480000,
|
| 9 |
+
"nb_max_frames": 3000,
|
| 10 |
+
"padding_side": "right",
|
| 11 |
+
"padding_value": 0.0,
|
| 12 |
+
"processor_class": "WhisperProcessor",
|
| 13 |
+
"return_attention_mask": false,
|
| 14 |
+
"sampling_rate": 16000
|
| 15 |
+
}
|
checkpoint-100/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:56891aaf8ac3e28df502c97b7baca2742be4c09167d04576400c329248b881d2
|
| 3 |
+
size 14244
|
checkpoint-100/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac4a558c5b93581a5c41e2922404490319bcd15a10296c927a72a41012ff7f27
|
| 3 |
+
size 988
|
checkpoint-100/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e3f011953770dc18a51194e919361544c1e41f047fac26af2e17015642e43c7d
|
| 3 |
+
size 1064
|
checkpoint-100/trainer_state.json
ADDED
|
@@ -0,0 +1,78 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 100,
|
| 3 |
+
"best_metric": 36.6100851831126,
|
| 4 |
+
"best_model_checkpoint": "./JUDIC/checkpoint-100",
|
| 5 |
+
"epoch": 1.1764705882352942,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 100,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.23529411764705882,
|
| 14 |
+
"grad_norm": 4.239101886749268,
|
| 15 |
+
"learning_rate": 1.9000000000000002e-06,
|
| 16 |
+
"loss": 0.4078,
|
| 17 |
+
"step": 20
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.47058823529411764,
|
| 21 |
+
"grad_norm": 2.235217809677124,
|
| 22 |
+
"learning_rate": 3.900000000000001e-06,
|
| 23 |
+
"loss": 0.4217,
|
| 24 |
+
"step": 40
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.7058823529411765,
|
| 28 |
+
"grad_norm": 3.711195707321167,
|
| 29 |
+
"learning_rate": 5.9e-06,
|
| 30 |
+
"loss": 0.4252,
|
| 31 |
+
"step": 60
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.9411764705882353,
|
| 35 |
+
"grad_norm": 4.5320820808410645,
|
| 36 |
+
"learning_rate": 7.9e-06,
|
| 37 |
+
"loss": 0.3812,
|
| 38 |
+
"step": 80
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.1764705882352942,
|
| 42 |
+
"grad_norm": 3.9703948497772217,
|
| 43 |
+
"learning_rate": 9.9e-06,
|
| 44 |
+
"loss": 0.4223,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.1764705882352942,
|
| 49 |
+
"eval_loss": 0.40899839997291565,
|
| 50 |
+
"eval_runtime": 323.6745,
|
| 51 |
+
"eval_samples_per_second": 2.104,
|
| 52 |
+
"eval_steps_per_second": 0.266,
|
| 53 |
+
"eval_wer": 36.6100851831126,
|
| 54 |
+
"step": 100
|
| 55 |
+
}
|
| 56 |
+
],
|
| 57 |
+
"logging_steps": 20,
|
| 58 |
+
"max_steps": 850,
|
| 59 |
+
"num_input_tokens_seen": 0,
|
| 60 |
+
"num_train_epochs": 10,
|
| 61 |
+
"save_steps": 100,
|
| 62 |
+
"stateful_callbacks": {
|
| 63 |
+
"TrainerControl": {
|
| 64 |
+
"args": {
|
| 65 |
+
"should_epoch_stop": false,
|
| 66 |
+
"should_evaluate": false,
|
| 67 |
+
"should_log": false,
|
| 68 |
+
"should_save": true,
|
| 69 |
+
"should_training_stop": false
|
| 70 |
+
},
|
| 71 |
+
"attributes": {}
|
| 72 |
+
}
|
| 73 |
+
},
|
| 74 |
+
"total_flos": 9.23473281024e+17,
|
| 75 |
+
"train_batch_size": 16,
|
| 76 |
+
"trial_name": null,
|
| 77 |
+
"trial_params": null
|
| 78 |
+
}
|
checkpoint-100/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a582364a1defc1cdd820215263292cba4d325748959fa5eb84ed58c1bcccb33d
|
| 3 |
+
size 5432
|
checkpoint-200/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 966995080
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f81c82d6052baacb7e66c24e8c708a630cd8cedf72c4ff7a9ec6f231ecd905d
|
| 3 |
size 966995080
|
checkpoint-200/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1925064044
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:38b02504b8786f500125ee911e1e4fe3881605e2517b98180d456037a54f5715
|
| 3 |
size 1925064044
|
checkpoint-200/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b7e1937d511f9c3420c5c21f44a372adbc882c5626121f8a8c3d5d9e8ade15b
|
| 3 |
size 14244
|
checkpoint-200/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 988
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aa75768e65458cb8b0fa3d934398492a1aa75cd3ae318e75b37395db60dc10a2
|
| 3 |
size 988
|
checkpoint-200/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:23427674df6bb33784b590ae4aa53899078e927f5e43e1c513019f2bf257b7de
|
| 3 |
size 1064
|
checkpoint-200/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 200,
|
| 3 |
-
"best_metric":
|
| 4 |
"best_model_checkpoint": "./JUDIC/checkpoint-200",
|
| 5 |
"epoch": 2.3529411764705883,
|
| 6 |
-
"eval_steps":
|
| 7 |
"global_step": 200,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
|
@@ -11,89 +11,98 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.23529411764705882,
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate":
|
| 16 |
-
"loss":
|
| 17 |
"step": 20
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.47058823529411764,
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss":
|
| 24 |
"step": 40
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.7058823529411765,
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"loss":
|
| 31 |
"step": 60
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.9411764705882353,
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss":
|
| 38 |
"step": 80
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 1.1764705882352942,
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45 |
"step": 100
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 1.4117647058823528,
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss":
|
| 52 |
"step": 120
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 1.6470588235294117,
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss":
|
| 59 |
"step": 140
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 1.8823529411764706,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss":
|
| 66 |
"step": 160
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 2.1176470588235294,
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate": 8.
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 180
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 2.3529411764705883,
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 200
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 2.3529411764705883,
|
| 84 |
-
"eval_loss": 0.
|
| 85 |
-
"eval_runtime":
|
| 86 |
-
"eval_samples_per_second": 2.
|
| 87 |
-
"eval_steps_per_second": 0.
|
| 88 |
-
"eval_wer":
|
| 89 |
"step": 200
|
| 90 |
}
|
| 91 |
],
|
| 92 |
"logging_steps": 20,
|
| 93 |
-
"max_steps":
|
| 94 |
"num_input_tokens_seen": 0,
|
| 95 |
-
"num_train_epochs":
|
| 96 |
-
"save_steps":
|
| 97 |
"stateful_callbacks": {
|
| 98 |
"TrainerControl": {
|
| 99 |
"args": {
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 200,
|
| 3 |
+
"best_metric": 29.938444320089534,
|
| 4 |
"best_model_checkpoint": "./JUDIC/checkpoint-200",
|
| 5 |
"epoch": 2.3529411764705883,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
"global_step": 200,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.23529411764705882,
|
| 14 |
+
"grad_norm": 4.239101886749268,
|
| 15 |
+
"learning_rate": 1.9000000000000002e-06,
|
| 16 |
+
"loss": 0.4078,
|
| 17 |
"step": 20
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.47058823529411764,
|
| 21 |
+
"grad_norm": 2.235217809677124,
|
| 22 |
+
"learning_rate": 3.900000000000001e-06,
|
| 23 |
+
"loss": 0.4217,
|
| 24 |
"step": 40
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.7058823529411765,
|
| 28 |
+
"grad_norm": 3.711195707321167,
|
| 29 |
+
"learning_rate": 5.9e-06,
|
| 30 |
+
"loss": 0.4252,
|
| 31 |
"step": 60
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.9411764705882353,
|
| 35 |
+
"grad_norm": 4.5320820808410645,
|
| 36 |
+
"learning_rate": 7.9e-06,
|
| 37 |
+
"loss": 0.3812,
|
| 38 |
"step": 80
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 1.1764705882352942,
|
| 42 |
+
"grad_norm": 3.9703948497772217,
|
| 43 |
+
"learning_rate": 9.9e-06,
|
| 44 |
+
"loss": 0.4223,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.1764705882352942,
|
| 49 |
+
"eval_loss": 0.40899839997291565,
|
| 50 |
+
"eval_runtime": 323.6745,
|
| 51 |
+
"eval_samples_per_second": 2.104,
|
| 52 |
+
"eval_steps_per_second": 0.266,
|
| 53 |
+
"eval_wer": 36.6100851831126,
|
| 54 |
"step": 100
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 1.4117647058823528,
|
| 58 |
+
"grad_norm": 3.8265273571014404,
|
| 59 |
+
"learning_rate": 9.746666666666668e-06,
|
| 60 |
+
"loss": 0.2872,
|
| 61 |
"step": 120
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 1.6470588235294117,
|
| 65 |
+
"grad_norm": 4.8625168800354,
|
| 66 |
+
"learning_rate": 9.48e-06,
|
| 67 |
+
"loss": 0.3207,
|
| 68 |
"step": 140
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 1.8823529411764706,
|
| 72 |
+
"grad_norm": 4.235057830810547,
|
| 73 |
+
"learning_rate": 9.213333333333334e-06,
|
| 74 |
+
"loss": 0.3585,
|
| 75 |
"step": 160
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 2.1176470588235294,
|
| 79 |
+
"grad_norm": 2.1225504875183105,
|
| 80 |
+
"learning_rate": 8.973333333333334e-06,
|
| 81 |
+
"loss": 0.2658,
|
| 82 |
"step": 180
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"epoch": 2.3529411764705883,
|
| 86 |
+
"grad_norm": 8.930822372436523,
|
| 87 |
+
"learning_rate": 8.706666666666667e-06,
|
| 88 |
+
"loss": 0.2576,
|
| 89 |
"step": 200
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 2.3529411764705883,
|
| 93 |
+
"eval_loss": 0.35160648822784424,
|
| 94 |
+
"eval_runtime": 296.116,
|
| 95 |
+
"eval_samples_per_second": 2.3,
|
| 96 |
+
"eval_steps_per_second": 0.29,
|
| 97 |
+
"eval_wer": 29.938444320089534,
|
| 98 |
"step": 200
|
| 99 |
}
|
| 100 |
],
|
| 101 |
"logging_steps": 20,
|
| 102 |
+
"max_steps": 850,
|
| 103 |
"num_input_tokens_seen": 0,
|
| 104 |
+
"num_train_epochs": 10,
|
| 105 |
+
"save_steps": 100,
|
| 106 |
"stateful_callbacks": {
|
| 107 |
"TrainerControl": {
|
| 108 |
"args": {
|
checkpoint-200/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5432
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a582364a1defc1cdd820215263292cba4d325748959fa5eb84ed58c1bcccb33d
|
| 3 |
size 5432
|
checkpoint-300/config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"WhisperForConditionalGeneration"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"begin_suppress_tokens": null,
|
| 10 |
+
"bos_token_id": 50257,
|
| 11 |
+
"classifier_proj_size": 256,
|
| 12 |
+
"d_model": 768,
|
| 13 |
+
"decoder_attention_heads": 12,
|
| 14 |
+
"decoder_ffn_dim": 3072,
|
| 15 |
+
"decoder_layerdrop": 0.0,
|
| 16 |
+
"decoder_layers": 12,
|
| 17 |
+
"decoder_start_token_id": 50258,
|
| 18 |
+
"dropout": 0.0,
|
| 19 |
+
"encoder_attention_heads": 12,
|
| 20 |
+
"encoder_ffn_dim": 3072,
|
| 21 |
+
"encoder_layerdrop": 0.0,
|
| 22 |
+
"encoder_layers": 12,
|
| 23 |
+
"eos_token_id": 50257,
|
| 24 |
+
"forced_decoder_ids": null,
|
| 25 |
+
"init_std": 0.02,
|
| 26 |
+
"is_encoder_decoder": true,
|
| 27 |
+
"mask_feature_length": 10,
|
| 28 |
+
"mask_feature_min_masks": 0,
|
| 29 |
+
"mask_feature_prob": 0.0,
|
| 30 |
+
"mask_time_length": 10,
|
| 31 |
+
"mask_time_min_masks": 2,
|
| 32 |
+
"mask_time_prob": 0.05,
|
| 33 |
+
"max_length": null,
|
| 34 |
+
"max_source_positions": 1500,
|
| 35 |
+
"max_target_positions": 448,
|
| 36 |
+
"median_filter_width": 7,
|
| 37 |
+
"model_type": "whisper",
|
| 38 |
+
"num_hidden_layers": 12,
|
| 39 |
+
"num_mel_bins": 80,
|
| 40 |
+
"pad_token_id": 50257,
|
| 41 |
+
"scale_embedding": false,
|
| 42 |
+
"torch_dtype": "float32",
|
| 43 |
+
"transformers_version": "4.51.3",
|
| 44 |
+
"use_cache": false,
|
| 45 |
+
"use_weighted_layer_sum": false,
|
| 46 |
+
"vocab_size": 51865
|
| 47 |
+
}
|
checkpoint-300/generation_config.json
ADDED
|
@@ -0,0 +1,175 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alignment_heads": [
|
| 3 |
+
[
|
| 4 |
+
5,
|
| 5 |
+
3
|
| 6 |
+
],
|
| 7 |
+
[
|
| 8 |
+
5,
|
| 9 |
+
9
|
| 10 |
+
],
|
| 11 |
+
[
|
| 12 |
+
8,
|
| 13 |
+
0
|
| 14 |
+
],
|
| 15 |
+
[
|
| 16 |
+
8,
|
| 17 |
+
4
|
| 18 |
+
],
|
| 19 |
+
[
|
| 20 |
+
8,
|
| 21 |
+
7
|
| 22 |
+
],
|
| 23 |
+
[
|
| 24 |
+
8,
|
| 25 |
+
8
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
9,
|
| 29 |
+
0
|
| 30 |
+
],
|
| 31 |
+
[
|
| 32 |
+
9,
|
| 33 |
+
7
|
| 34 |
+
],
|
| 35 |
+
[
|
| 36 |
+
9,
|
| 37 |
+
9
|
| 38 |
+
],
|
| 39 |
+
[
|
| 40 |
+
10,
|
| 41 |
+
5
|
| 42 |
+
]
|
| 43 |
+
],
|
| 44 |
+
"begin_suppress_tokens": [
|
| 45 |
+
220,
|
| 46 |
+
50257
|
| 47 |
+
],
|
| 48 |
+
"bos_token_id": 50257,
|
| 49 |
+
"decoder_start_token_id": 50258,
|
| 50 |
+
"eos_token_id": 50257,
|
| 51 |
+
"forced_decoder_ids": [
|
| 52 |
+
[
|
| 53 |
+
1,
|
| 54 |
+
null
|
| 55 |
+
],
|
| 56 |
+
[
|
| 57 |
+
2,
|
| 58 |
+
50359
|
| 59 |
+
]
|
| 60 |
+
],
|
| 61 |
+
"is_multilingual": true,
|
| 62 |
+
"lang_to_id": {
|
| 63 |
+
"<|af|>": 50327,
|
| 64 |
+
"<|am|>": 50334,
|
| 65 |
+
"<|ar|>": 50272,
|
| 66 |
+
"<|as|>": 50350,
|
| 67 |
+
"<|az|>": 50304,
|
| 68 |
+
"<|ba|>": 50355,
|
| 69 |
+
"<|be|>": 50330,
|
| 70 |
+
"<|bg|>": 50292,
|
| 71 |
+
"<|bn|>": 50302,
|
| 72 |
+
"<|bo|>": 50347,
|
| 73 |
+
"<|br|>": 50309,
|
| 74 |
+
"<|bs|>": 50315,
|
| 75 |
+
"<|ca|>": 50270,
|
| 76 |
+
"<|cs|>": 50283,
|
| 77 |
+
"<|cy|>": 50297,
|
| 78 |
+
"<|da|>": 50285,
|
| 79 |
+
"<|de|>": 50261,
|
| 80 |
+
"<|el|>": 50281,
|
| 81 |
+
"<|en|>": 50259,
|
| 82 |
+
"<|es|>": 50262,
|
| 83 |
+
"<|et|>": 50307,
|
| 84 |
+
"<|eu|>": 50310,
|
| 85 |
+
"<|fa|>": 50300,
|
| 86 |
+
"<|fi|>": 50277,
|
| 87 |
+
"<|fo|>": 50338,
|
| 88 |
+
"<|fr|>": 50265,
|
| 89 |
+
"<|gl|>": 50319,
|
| 90 |
+
"<|gu|>": 50333,
|
| 91 |
+
"<|haw|>": 50352,
|
| 92 |
+
"<|ha|>": 50354,
|
| 93 |
+
"<|he|>": 50279,
|
| 94 |
+
"<|hi|>": 50276,
|
| 95 |
+
"<|hr|>": 50291,
|
| 96 |
+
"<|ht|>": 50339,
|
| 97 |
+
"<|hu|>": 50286,
|
| 98 |
+
"<|hy|>": 50312,
|
| 99 |
+
"<|id|>": 50275,
|
| 100 |
+
"<|is|>": 50311,
|
| 101 |
+
"<|it|>": 50274,
|
| 102 |
+
"<|ja|>": 50266,
|
| 103 |
+
"<|jw|>": 50356,
|
| 104 |
+
"<|ka|>": 50329,
|
| 105 |
+
"<|kk|>": 50316,
|
| 106 |
+
"<|km|>": 50323,
|
| 107 |
+
"<|kn|>": 50306,
|
| 108 |
+
"<|ko|>": 50264,
|
| 109 |
+
"<|la|>": 50294,
|
| 110 |
+
"<|lb|>": 50345,
|
| 111 |
+
"<|ln|>": 50353,
|
| 112 |
+
"<|lo|>": 50336,
|
| 113 |
+
"<|lt|>": 50293,
|
| 114 |
+
"<|lv|>": 50301,
|
| 115 |
+
"<|mg|>": 50349,
|
| 116 |
+
"<|mi|>": 50295,
|
| 117 |
+
"<|mk|>": 50308,
|
| 118 |
+
"<|ml|>": 50296,
|
| 119 |
+
"<|mn|>": 50314,
|
| 120 |
+
"<|mr|>": 50320,
|
| 121 |
+
"<|ms|>": 50282,
|
| 122 |
+
"<|mt|>": 50343,
|
| 123 |
+
"<|my|>": 50346,
|
| 124 |
+
"<|ne|>": 50313,
|
| 125 |
+
"<|nl|>": 50271,
|
| 126 |
+
"<|nn|>": 50342,
|
| 127 |
+
"<|no|>": 50288,
|
| 128 |
+
"<|oc|>": 50328,
|
| 129 |
+
"<|pa|>": 50321,
|
| 130 |
+
"<|pl|>": 50269,
|
| 131 |
+
"<|ps|>": 50340,
|
| 132 |
+
"<|pt|>": 50267,
|
| 133 |
+
"<|ro|>": 50284,
|
| 134 |
+
"<|ru|>": 50263,
|
| 135 |
+
"<|sa|>": 50344,
|
| 136 |
+
"<|sd|>": 50332,
|
| 137 |
+
"<|si|>": 50322,
|
| 138 |
+
"<|sk|>": 50298,
|
| 139 |
+
"<|sl|>": 50305,
|
| 140 |
+
"<|sn|>": 50324,
|
| 141 |
+
"<|so|>": 50326,
|
| 142 |
+
"<|sq|>": 50317,
|
| 143 |
+
"<|sr|>": 50303,
|
| 144 |
+
"<|su|>": 50357,
|
| 145 |
+
"<|sv|>": 50273,
|
| 146 |
+
"<|sw|>": 50318,
|
| 147 |
+
"<|ta|>": 50287,
|
| 148 |
+
"<|te|>": 50299,
|
| 149 |
+
"<|tg|>": 50331,
|
| 150 |
+
"<|th|>": 50289,
|
| 151 |
+
"<|tk|>": 50341,
|
| 152 |
+
"<|tl|>": 50348,
|
| 153 |
+
"<|tr|>": 50268,
|
| 154 |
+
"<|tt|>": 50351,
|
| 155 |
+
"<|uk|>": 50280,
|
| 156 |
+
"<|ur|>": 50290,
|
| 157 |
+
"<|uz|>": 50337,
|
| 158 |
+
"<|vi|>": 50278,
|
| 159 |
+
"<|yi|>": 50335,
|
| 160 |
+
"<|yo|>": 50325,
|
| 161 |
+
"<|zh|>": 50260
|
| 162 |
+
},
|
| 163 |
+
"max_initial_timestamp_index": 50,
|
| 164 |
+
"max_length": 448,
|
| 165 |
+
"no_timestamps_token_id": 50363,
|
| 166 |
+
"pad_token_id": 50257,
|
| 167 |
+
"prev_sot_token_id": 50361,
|
| 168 |
+
"return_timestamps": false,
|
| 169 |
+
"suppress_tokens": [],
|
| 170 |
+
"task_to_id": {
|
| 171 |
+
"transcribe": 50359,
|
| 172 |
+
"translate": 50358
|
| 173 |
+
},
|
| 174 |
+
"transformers_version": "4.51.3"
|
| 175 |
+
}
|
checkpoint-300/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b6fea0b87575944e3fad7d172cfb03dc2c5fbd22d6556d834ff0a0b0568706ad
|
| 3 |
+
size 966995080
|
checkpoint-300/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd683dced7bbbadf49aeaca0d526ec98fa5ac9eb42ff2852ce2c726b56120616
|
| 3 |
+
size 1925064044
|
checkpoint-300/preprocessor_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"chunk_length": 30,
|
| 3 |
+
"dither": 0.0,
|
| 4 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 5 |
+
"feature_size": 80,
|
| 6 |
+
"hop_length": 160,
|
| 7 |
+
"n_fft": 400,
|
| 8 |
+
"n_samples": 480000,
|
| 9 |
+
"nb_max_frames": 3000,
|
| 10 |
+
"padding_side": "right",
|
| 11 |
+
"padding_value": 0.0,
|
| 12 |
+
"processor_class": "WhisperProcessor",
|
| 13 |
+
"return_attention_mask": false,
|
| 14 |
+
"sampling_rate": 16000
|
| 15 |
+
}
|
checkpoint-300/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4c75de8aef036ad6437dee31b4cc33fdaf4d9234973688095b81d98166c4ca11
|
| 3 |
+
size 14244
|
checkpoint-300/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f55124c020962c0cda07fe10a52f1e374d6480f4931ceea4539f39f03b1ff2e
|
| 3 |
+
size 988
|
checkpoint-300/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9122dc6585a724b33eebfaf95afd97b180142b91a30efe1fe1ddc84cb2148c14
|
| 3 |
+
size 1064
|
checkpoint-300/trainer_state.json
ADDED
|
@@ -0,0 +1,166 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 200,
|
| 3 |
+
"best_metric": 29.938444320089534,
|
| 4 |
+
"best_model_checkpoint": "./JUDIC/checkpoint-200",
|
| 5 |
+
"epoch": 3.5294117647058822,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 300,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.23529411764705882,
|
| 14 |
+
"grad_norm": 4.239101886749268,
|
| 15 |
+
"learning_rate": 1.9000000000000002e-06,
|
| 16 |
+
"loss": 0.4078,
|
| 17 |
+
"step": 20
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.47058823529411764,
|
| 21 |
+
"grad_norm": 2.235217809677124,
|
| 22 |
+
"learning_rate": 3.900000000000001e-06,
|
| 23 |
+
"loss": 0.4217,
|
| 24 |
+
"step": 40
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.7058823529411765,
|
| 28 |
+
"grad_norm": 3.711195707321167,
|
| 29 |
+
"learning_rate": 5.9e-06,
|
| 30 |
+
"loss": 0.4252,
|
| 31 |
+
"step": 60
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.9411764705882353,
|
| 35 |
+
"grad_norm": 4.5320820808410645,
|
| 36 |
+
"learning_rate": 7.9e-06,
|
| 37 |
+
"loss": 0.3812,
|
| 38 |
+
"step": 80
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.1764705882352942,
|
| 42 |
+
"grad_norm": 3.9703948497772217,
|
| 43 |
+
"learning_rate": 9.9e-06,
|
| 44 |
+
"loss": 0.4223,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.1764705882352942,
|
| 49 |
+
"eval_loss": 0.40899839997291565,
|
| 50 |
+
"eval_runtime": 323.6745,
|
| 51 |
+
"eval_samples_per_second": 2.104,
|
| 52 |
+
"eval_steps_per_second": 0.266,
|
| 53 |
+
"eval_wer": 36.6100851831126,
|
| 54 |
+
"step": 100
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 1.4117647058823528,
|
| 58 |
+
"grad_norm": 3.8265273571014404,
|
| 59 |
+
"learning_rate": 9.746666666666668e-06,
|
| 60 |
+
"loss": 0.2872,
|
| 61 |
+
"step": 120
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 1.6470588235294117,
|
| 65 |
+
"grad_norm": 4.8625168800354,
|
| 66 |
+
"learning_rate": 9.48e-06,
|
| 67 |
+
"loss": 0.3207,
|
| 68 |
+
"step": 140
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.8823529411764706,
|
| 72 |
+
"grad_norm": 4.235057830810547,
|
| 73 |
+
"learning_rate": 9.213333333333334e-06,
|
| 74 |
+
"loss": 0.3585,
|
| 75 |
+
"step": 160
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 2.1176470588235294,
|
| 79 |
+
"grad_norm": 2.1225504875183105,
|
| 80 |
+
"learning_rate": 8.973333333333334e-06,
|
| 81 |
+
"loss": 0.2658,
|
| 82 |
+
"step": 180
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 2.3529411764705883,
|
| 86 |
+
"grad_norm": 8.930822372436523,
|
| 87 |
+
"learning_rate": 8.706666666666667e-06,
|
| 88 |
+
"loss": 0.2576,
|
| 89 |
+
"step": 200
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 2.3529411764705883,
|
| 93 |
+
"eval_loss": 0.35160648822784424,
|
| 94 |
+
"eval_runtime": 296.116,
|
| 95 |
+
"eval_samples_per_second": 2.3,
|
| 96 |
+
"eval_steps_per_second": 0.29,
|
| 97 |
+
"eval_wer": 29.938444320089534,
|
| 98 |
+
"step": 200
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.588235294117647,
|
| 102 |
+
"grad_norm": 1.942764163017273,
|
| 103 |
+
"learning_rate": 8.44e-06,
|
| 104 |
+
"loss": 0.283,
|
| 105 |
+
"step": 220
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 2.8235294117647056,
|
| 109 |
+
"grad_norm": 3.2245473861694336,
|
| 110 |
+
"learning_rate": 8.173333333333334e-06,
|
| 111 |
+
"loss": 0.2443,
|
| 112 |
+
"step": 240
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 3.0588235294117645,
|
| 116 |
+
"grad_norm": 3.0929267406463623,
|
| 117 |
+
"learning_rate": 7.906666666666667e-06,
|
| 118 |
+
"loss": 0.2489,
|
| 119 |
+
"step": 260
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 3.2941176470588234,
|
| 123 |
+
"grad_norm": 3.477473497390747,
|
| 124 |
+
"learning_rate": 7.640000000000001e-06,
|
| 125 |
+
"loss": 0.2006,
|
| 126 |
+
"step": 280
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 3.5294117647058822,
|
| 130 |
+
"grad_norm": 3.163804531097412,
|
| 131 |
+
"learning_rate": 7.373333333333334e-06,
|
| 132 |
+
"loss": 0.1837,
|
| 133 |
+
"step": 300
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 3.5294117647058822,
|
| 137 |
+
"eval_loss": 0.30386850237846375,
|
| 138 |
+
"eval_runtime": 329.1958,
|
| 139 |
+
"eval_samples_per_second": 2.069,
|
| 140 |
+
"eval_steps_per_second": 0.261,
|
| 141 |
+
"eval_wer": 31.555058135919918,
|
| 142 |
+
"step": 300
|
| 143 |
+
}
|
| 144 |
+
],
|
| 145 |
+
"logging_steps": 20,
|
| 146 |
+
"max_steps": 850,
|
| 147 |
+
"num_input_tokens_seen": 0,
|
| 148 |
+
"num_train_epochs": 10,
|
| 149 |
+
"save_steps": 100,
|
| 150 |
+
"stateful_callbacks": {
|
| 151 |
+
"TrainerControl": {
|
| 152 |
+
"args": {
|
| 153 |
+
"should_epoch_stop": false,
|
| 154 |
+
"should_evaluate": false,
|
| 155 |
+
"should_log": false,
|
| 156 |
+
"should_save": true,
|
| 157 |
+
"should_training_stop": false
|
| 158 |
+
},
|
| 159 |
+
"attributes": {}
|
| 160 |
+
}
|
| 161 |
+
},
|
| 162 |
+
"total_flos": 2.770419843072e+18,
|
| 163 |
+
"train_batch_size": 16,
|
| 164 |
+
"trial_name": null,
|
| 165 |
+
"trial_params": null
|
| 166 |
+
}
|
checkpoint-300/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a582364a1defc1cdd820215263292cba4d325748959fa5eb84ed58c1bcccb33d
|
| 3 |
+
size 5432
|
checkpoint-400/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 966995080
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:826cae9d99f0d4d549aa98fab1b85fc27755f00ce7b56ac7a10cf63a2a6fb584
|
| 3 |
size 966995080
|
checkpoint-400/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1925064044
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f32422f07da705bcb7284e1286b3f99b2f566ecd46a211462f9b454aff8d60a5
|
| 3 |
size 1925064044
|
checkpoint-400/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:30c47e162e68dbe940a216a2770ed84b7751c4b1e55b4de071db84c5e99904ed
|
| 3 |
size 14244
|
checkpoint-400/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 988
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ba4abf7693b716a8ac4dd2dada1c394abd2cf2ea9e18f42a3a5733573c24b77
|
| 3 |
size 988
|
checkpoint-400/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:79d838619a5e73d7fb3ccfd45f48dd11a5b5251829dba401a7a2d6341f3b7f2d
|
| 3 |
size 1064
|
checkpoint-400/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 400,
|
| 3 |
-
"best_metric":
|
| 4 |
"best_model_checkpoint": "./JUDIC/checkpoint-400",
|
| 5 |
"epoch": 4.705882352941177,
|
| 6 |
-
"eval_steps":
|
| 7 |
"global_step": 400,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
|
@@ -11,168 +11,186 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.23529411764705882,
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate":
|
| 16 |
-
"loss":
|
| 17 |
"step": 20
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.47058823529411764,
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss":
|
| 24 |
"step": 40
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.7058823529411765,
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"loss":
|
| 31 |
"step": 60
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.9411764705882353,
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss":
|
| 38 |
"step": 80
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 1.1764705882352942,
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45 |
"step": 100
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 1.4117647058823528,
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss":
|
| 52 |
"step": 120
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 1.6470588235294117,
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss":
|
| 59 |
"step": 140
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 1.8823529411764706,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss":
|
| 66 |
"step": 160
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 2.1176470588235294,
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate": 8.
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 180
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 2.3529411764705883,
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 200
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 2.3529411764705883,
|
| 84 |
-
"eval_loss": 0.
|
| 85 |
-
"eval_runtime":
|
| 86 |
-
"eval_samples_per_second": 2.
|
| 87 |
-
"eval_steps_per_second": 0.
|
| 88 |
-
"eval_wer":
|
| 89 |
"step": 200
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 2.588235294117647,
|
| 93 |
-
"grad_norm":
|
| 94 |
-
"learning_rate":
|
| 95 |
-
"loss": 0.
|
| 96 |
"step": 220
|
| 97 |
},
|
| 98 |
{
|
| 99 |
"epoch": 2.8235294117647056,
|
| 100 |
-
"grad_norm":
|
| 101 |
-
"learning_rate": 8.
|
| 102 |
-
"loss": 0.
|
| 103 |
"step": 240
|
| 104 |
},
|
| 105 |
{
|
| 106 |
"epoch": 3.0588235294117645,
|
| 107 |
-
"grad_norm": 3.
|
| 108 |
-
"learning_rate": 7.
|
| 109 |
-
"loss": 0.
|
| 110 |
"step": 260
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 3.2941176470588234,
|
| 114 |
-
"grad_norm": 3.
|
| 115 |
-
"learning_rate":
|
| 116 |
-
"loss": 0.
|
| 117 |
"step": 280
|
| 118 |
},
|
| 119 |
{
|
| 120 |
"epoch": 3.5294117647058822,
|
| 121 |
-
"grad_norm": 3.
|
| 122 |
-
"learning_rate":
|
| 123 |
-
"loss": 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 124 |
"step": 300
|
| 125 |
},
|
| 126 |
{
|
| 127 |
"epoch": 3.764705882352941,
|
| 128 |
-
"grad_norm":
|
| 129 |
-
"learning_rate":
|
| 130 |
-
"loss": 0.
|
| 131 |
"step": 320
|
| 132 |
},
|
| 133 |
{
|
| 134 |
"epoch": 4.0,
|
| 135 |
-
"grad_norm":
|
| 136 |
-
"learning_rate":
|
| 137 |
-
"loss": 0.
|
| 138 |
"step": 340
|
| 139 |
},
|
| 140 |
{
|
| 141 |
"epoch": 4.235294117647059,
|
| 142 |
-
"grad_norm":
|
| 143 |
-
"learning_rate":
|
| 144 |
-
"loss": 0.
|
| 145 |
"step": 360
|
| 146 |
},
|
| 147 |
{
|
| 148 |
"epoch": 4.470588235294118,
|
| 149 |
-
"grad_norm": 3.
|
| 150 |
-
"learning_rate":
|
| 151 |
-
"loss": 0.
|
| 152 |
"step": 380
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"epoch": 4.705882352941177,
|
| 156 |
-
"grad_norm": 3.
|
| 157 |
-
"learning_rate":
|
| 158 |
-
"loss": 0.
|
| 159 |
"step": 400
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 4.705882352941177,
|
| 163 |
-
"eval_loss": 0.
|
| 164 |
-
"eval_runtime": 307.
|
| 165 |
"eval_samples_per_second": 2.215,
|
| 166 |
"eval_steps_per_second": 0.28,
|
| 167 |
-
"eval_wer":
|
| 168 |
"step": 400
|
| 169 |
}
|
| 170 |
],
|
| 171 |
"logging_steps": 20,
|
| 172 |
-
"max_steps":
|
| 173 |
"num_input_tokens_seen": 0,
|
| 174 |
-
"num_train_epochs":
|
| 175 |
-
"save_steps":
|
| 176 |
"stateful_callbacks": {
|
| 177 |
"TrainerControl": {
|
| 178 |
"args": {
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 400,
|
| 3 |
+
"best_metric": 25.511409562892496,
|
| 4 |
"best_model_checkpoint": "./JUDIC/checkpoint-400",
|
| 5 |
"epoch": 4.705882352941177,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
"global_step": 400,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.23529411764705882,
|
| 14 |
+
"grad_norm": 4.239101886749268,
|
| 15 |
+
"learning_rate": 1.9000000000000002e-06,
|
| 16 |
+
"loss": 0.4078,
|
| 17 |
"step": 20
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.47058823529411764,
|
| 21 |
+
"grad_norm": 2.235217809677124,
|
| 22 |
+
"learning_rate": 3.900000000000001e-06,
|
| 23 |
+
"loss": 0.4217,
|
| 24 |
"step": 40
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.7058823529411765,
|
| 28 |
+
"grad_norm": 3.711195707321167,
|
| 29 |
+
"learning_rate": 5.9e-06,
|
| 30 |
+
"loss": 0.4252,
|
| 31 |
"step": 60
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.9411764705882353,
|
| 35 |
+
"grad_norm": 4.5320820808410645,
|
| 36 |
+
"learning_rate": 7.9e-06,
|
| 37 |
+
"loss": 0.3812,
|
| 38 |
"step": 80
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 1.1764705882352942,
|
| 42 |
+
"grad_norm": 3.9703948497772217,
|
| 43 |
+
"learning_rate": 9.9e-06,
|
| 44 |
+
"loss": 0.4223,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.1764705882352942,
|
| 49 |
+
"eval_loss": 0.40899839997291565,
|
| 50 |
+
"eval_runtime": 323.6745,
|
| 51 |
+
"eval_samples_per_second": 2.104,
|
| 52 |
+
"eval_steps_per_second": 0.266,
|
| 53 |
+
"eval_wer": 36.6100851831126,
|
| 54 |
"step": 100
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 1.4117647058823528,
|
| 58 |
+
"grad_norm": 3.8265273571014404,
|
| 59 |
+
"learning_rate": 9.746666666666668e-06,
|
| 60 |
+
"loss": 0.2872,
|
| 61 |
"step": 120
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 1.6470588235294117,
|
| 65 |
+
"grad_norm": 4.8625168800354,
|
| 66 |
+
"learning_rate": 9.48e-06,
|
| 67 |
+
"loss": 0.3207,
|
| 68 |
"step": 140
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 1.8823529411764706,
|
| 72 |
+
"grad_norm": 4.235057830810547,
|
| 73 |
+
"learning_rate": 9.213333333333334e-06,
|
| 74 |
+
"loss": 0.3585,
|
| 75 |
"step": 160
|
| 76 |
},
|
| 77 |
{
|
| 78 |
"epoch": 2.1176470588235294,
|
| 79 |
+
"grad_norm": 2.1225504875183105,
|
| 80 |
+
"learning_rate": 8.973333333333334e-06,
|
| 81 |
+
"loss": 0.2658,
|
| 82 |
"step": 180
|
| 83 |
},
|
| 84 |
{
|
| 85 |
"epoch": 2.3529411764705883,
|
| 86 |
+
"grad_norm": 8.930822372436523,
|
| 87 |
+
"learning_rate": 8.706666666666667e-06,
|
| 88 |
+
"loss": 0.2576,
|
| 89 |
"step": 200
|
| 90 |
},
|
| 91 |
{
|
| 92 |
"epoch": 2.3529411764705883,
|
| 93 |
+
"eval_loss": 0.35160648822784424,
|
| 94 |
+
"eval_runtime": 296.116,
|
| 95 |
+
"eval_samples_per_second": 2.3,
|
| 96 |
+
"eval_steps_per_second": 0.29,
|
| 97 |
+
"eval_wer": 29.938444320089534,
|
| 98 |
"step": 200
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 2.588235294117647,
|
| 102 |
+
"grad_norm": 1.942764163017273,
|
| 103 |
+
"learning_rate": 8.44e-06,
|
| 104 |
+
"loss": 0.283,
|
| 105 |
"step": 220
|
| 106 |
},
|
| 107 |
{
|
| 108 |
"epoch": 2.8235294117647056,
|
| 109 |
+
"grad_norm": 3.2245473861694336,
|
| 110 |
+
"learning_rate": 8.173333333333334e-06,
|
| 111 |
+
"loss": 0.2443,
|
| 112 |
"step": 240
|
| 113 |
},
|
| 114 |
{
|
| 115 |
"epoch": 3.0588235294117645,
|
| 116 |
+
"grad_norm": 3.0929267406463623,
|
| 117 |
+
"learning_rate": 7.906666666666667e-06,
|
| 118 |
+
"loss": 0.2489,
|
| 119 |
"step": 260
|
| 120 |
},
|
| 121 |
{
|
| 122 |
"epoch": 3.2941176470588234,
|
| 123 |
+
"grad_norm": 3.477473497390747,
|
| 124 |
+
"learning_rate": 7.640000000000001e-06,
|
| 125 |
+
"loss": 0.2006,
|
| 126 |
"step": 280
|
| 127 |
},
|
| 128 |
{
|
| 129 |
"epoch": 3.5294117647058822,
|
| 130 |
+
"grad_norm": 3.163804531097412,
|
| 131 |
+
"learning_rate": 7.373333333333334e-06,
|
| 132 |
+
"loss": 0.1837,
|
| 133 |
+
"step": 300
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 3.5294117647058822,
|
| 137 |
+
"eval_loss": 0.30386850237846375,
|
| 138 |
+
"eval_runtime": 329.1958,
|
| 139 |
+
"eval_samples_per_second": 2.069,
|
| 140 |
+
"eval_steps_per_second": 0.261,
|
| 141 |
+
"eval_wer": 31.555058135919918,
|
| 142 |
"step": 300
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 3.764705882352941,
|
| 146 |
+
"grad_norm": 2.77553391456604,
|
| 147 |
+
"learning_rate": 7.1066666666666675e-06,
|
| 148 |
+
"loss": 0.2342,
|
| 149 |
"step": 320
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 4.0,
|
| 153 |
+
"grad_norm": 3.944244861602783,
|
| 154 |
+
"learning_rate": 6.8400000000000014e-06,
|
| 155 |
+
"loss": 0.1751,
|
| 156 |
"step": 340
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 4.235294117647059,
|
| 160 |
+
"grad_norm": 6.829845905303955,
|
| 161 |
+
"learning_rate": 6.573333333333334e-06,
|
| 162 |
+
"loss": 0.1431,
|
| 163 |
"step": 360
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 4.470588235294118,
|
| 167 |
+
"grad_norm": 3.4126970767974854,
|
| 168 |
+
"learning_rate": 6.3066666666666676e-06,
|
| 169 |
+
"loss": 0.1468,
|
| 170 |
"step": 380
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 4.705882352941177,
|
| 174 |
+
"grad_norm": 3.5233240127563477,
|
| 175 |
+
"learning_rate": 6.040000000000001e-06,
|
| 176 |
+
"loss": 0.1376,
|
| 177 |
"step": 400
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 4.705882352941177,
|
| 181 |
+
"eval_loss": 0.27411767840385437,
|
| 182 |
+
"eval_runtime": 307.4101,
|
| 183 |
"eval_samples_per_second": 2.215,
|
| 184 |
"eval_steps_per_second": 0.28,
|
| 185 |
+
"eval_wer": 25.511409562892496,
|
| 186 |
"step": 400
|
| 187 |
}
|
| 188 |
],
|
| 189 |
"logging_steps": 20,
|
| 190 |
+
"max_steps": 850,
|
| 191 |
"num_input_tokens_seen": 0,
|
| 192 |
+
"num_train_epochs": 10,
|
| 193 |
+
"save_steps": 100,
|
| 194 |
"stateful_callbacks": {
|
| 195 |
"TrainerControl": {
|
| 196 |
"args": {
|
checkpoint-400/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5432
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a582364a1defc1cdd820215263292cba4d325748959fa5eb84ed58c1bcccb33d
|
| 3 |
size 5432
|
checkpoint-500/config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"WhisperForConditionalGeneration"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"begin_suppress_tokens": null,
|
| 10 |
+
"bos_token_id": 50257,
|
| 11 |
+
"classifier_proj_size": 256,
|
| 12 |
+
"d_model": 768,
|
| 13 |
+
"decoder_attention_heads": 12,
|
| 14 |
+
"decoder_ffn_dim": 3072,
|
| 15 |
+
"decoder_layerdrop": 0.0,
|
| 16 |
+
"decoder_layers": 12,
|
| 17 |
+
"decoder_start_token_id": 50258,
|
| 18 |
+
"dropout": 0.0,
|
| 19 |
+
"encoder_attention_heads": 12,
|
| 20 |
+
"encoder_ffn_dim": 3072,
|
| 21 |
+
"encoder_layerdrop": 0.0,
|
| 22 |
+
"encoder_layers": 12,
|
| 23 |
+
"eos_token_id": 50257,
|
| 24 |
+
"forced_decoder_ids": null,
|
| 25 |
+
"init_std": 0.02,
|
| 26 |
+
"is_encoder_decoder": true,
|
| 27 |
+
"mask_feature_length": 10,
|
| 28 |
+
"mask_feature_min_masks": 0,
|
| 29 |
+
"mask_feature_prob": 0.0,
|
| 30 |
+
"mask_time_length": 10,
|
| 31 |
+
"mask_time_min_masks": 2,
|
| 32 |
+
"mask_time_prob": 0.05,
|
| 33 |
+
"max_length": null,
|
| 34 |
+
"max_source_positions": 1500,
|
| 35 |
+
"max_target_positions": 448,
|
| 36 |
+
"median_filter_width": 7,
|
| 37 |
+
"model_type": "whisper",
|
| 38 |
+
"num_hidden_layers": 12,
|
| 39 |
+
"num_mel_bins": 80,
|
| 40 |
+
"pad_token_id": 50257,
|
| 41 |
+
"scale_embedding": false,
|
| 42 |
+
"torch_dtype": "float32",
|
| 43 |
+
"transformers_version": "4.51.3",
|
| 44 |
+
"use_cache": false,
|
| 45 |
+
"use_weighted_layer_sum": false,
|
| 46 |
+
"vocab_size": 51865
|
| 47 |
+
}
|
checkpoint-500/generation_config.json
ADDED
|
@@ -0,0 +1,175 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alignment_heads": [
|
| 3 |
+
[
|
| 4 |
+
5,
|
| 5 |
+
3
|
| 6 |
+
],
|
| 7 |
+
[
|
| 8 |
+
5,
|
| 9 |
+
9
|
| 10 |
+
],
|
| 11 |
+
[
|
| 12 |
+
8,
|
| 13 |
+
0
|
| 14 |
+
],
|
| 15 |
+
[
|
| 16 |
+
8,
|
| 17 |
+
4
|
| 18 |
+
],
|
| 19 |
+
[
|
| 20 |
+
8,
|
| 21 |
+
7
|
| 22 |
+
],
|
| 23 |
+
[
|
| 24 |
+
8,
|
| 25 |
+
8
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
9,
|
| 29 |
+
0
|
| 30 |
+
],
|
| 31 |
+
[
|
| 32 |
+
9,
|
| 33 |
+
7
|
| 34 |
+
],
|
| 35 |
+
[
|
| 36 |
+
9,
|
| 37 |
+
9
|
| 38 |
+
],
|
| 39 |
+
[
|
| 40 |
+
10,
|
| 41 |
+
5
|
| 42 |
+
]
|
| 43 |
+
],
|
| 44 |
+
"begin_suppress_tokens": [
|
| 45 |
+
220,
|
| 46 |
+
50257
|
| 47 |
+
],
|
| 48 |
+
"bos_token_id": 50257,
|
| 49 |
+
"decoder_start_token_id": 50258,
|
| 50 |
+
"eos_token_id": 50257,
|
| 51 |
+
"forced_decoder_ids": [
|
| 52 |
+
[
|
| 53 |
+
1,
|
| 54 |
+
null
|
| 55 |
+
],
|
| 56 |
+
[
|
| 57 |
+
2,
|
| 58 |
+
50359
|
| 59 |
+
]
|
| 60 |
+
],
|
| 61 |
+
"is_multilingual": true,
|
| 62 |
+
"lang_to_id": {
|
| 63 |
+
"<|af|>": 50327,
|
| 64 |
+
"<|am|>": 50334,
|
| 65 |
+
"<|ar|>": 50272,
|
| 66 |
+
"<|as|>": 50350,
|
| 67 |
+
"<|az|>": 50304,
|
| 68 |
+
"<|ba|>": 50355,
|
| 69 |
+
"<|be|>": 50330,
|
| 70 |
+
"<|bg|>": 50292,
|
| 71 |
+
"<|bn|>": 50302,
|
| 72 |
+
"<|bo|>": 50347,
|
| 73 |
+
"<|br|>": 50309,
|
| 74 |
+
"<|bs|>": 50315,
|
| 75 |
+
"<|ca|>": 50270,
|
| 76 |
+
"<|cs|>": 50283,
|
| 77 |
+
"<|cy|>": 50297,
|
| 78 |
+
"<|da|>": 50285,
|
| 79 |
+
"<|de|>": 50261,
|
| 80 |
+
"<|el|>": 50281,
|
| 81 |
+
"<|en|>": 50259,
|
| 82 |
+
"<|es|>": 50262,
|
| 83 |
+
"<|et|>": 50307,
|
| 84 |
+
"<|eu|>": 50310,
|
| 85 |
+
"<|fa|>": 50300,
|
| 86 |
+
"<|fi|>": 50277,
|
| 87 |
+
"<|fo|>": 50338,
|
| 88 |
+
"<|fr|>": 50265,
|
| 89 |
+
"<|gl|>": 50319,
|
| 90 |
+
"<|gu|>": 50333,
|
| 91 |
+
"<|haw|>": 50352,
|
| 92 |
+
"<|ha|>": 50354,
|
| 93 |
+
"<|he|>": 50279,
|
| 94 |
+
"<|hi|>": 50276,
|
| 95 |
+
"<|hr|>": 50291,
|
| 96 |
+
"<|ht|>": 50339,
|
| 97 |
+
"<|hu|>": 50286,
|
| 98 |
+
"<|hy|>": 50312,
|
| 99 |
+
"<|id|>": 50275,
|
| 100 |
+
"<|is|>": 50311,
|
| 101 |
+
"<|it|>": 50274,
|
| 102 |
+
"<|ja|>": 50266,
|
| 103 |
+
"<|jw|>": 50356,
|
| 104 |
+
"<|ka|>": 50329,
|
| 105 |
+
"<|kk|>": 50316,
|
| 106 |
+
"<|km|>": 50323,
|
| 107 |
+
"<|kn|>": 50306,
|
| 108 |
+
"<|ko|>": 50264,
|
| 109 |
+
"<|la|>": 50294,
|
| 110 |
+
"<|lb|>": 50345,
|
| 111 |
+
"<|ln|>": 50353,
|
| 112 |
+
"<|lo|>": 50336,
|
| 113 |
+
"<|lt|>": 50293,
|
| 114 |
+
"<|lv|>": 50301,
|
| 115 |
+
"<|mg|>": 50349,
|
| 116 |
+
"<|mi|>": 50295,
|
| 117 |
+
"<|mk|>": 50308,
|
| 118 |
+
"<|ml|>": 50296,
|
| 119 |
+
"<|mn|>": 50314,
|
| 120 |
+
"<|mr|>": 50320,
|
| 121 |
+
"<|ms|>": 50282,
|
| 122 |
+
"<|mt|>": 50343,
|
| 123 |
+
"<|my|>": 50346,
|
| 124 |
+
"<|ne|>": 50313,
|
| 125 |
+
"<|nl|>": 50271,
|
| 126 |
+
"<|nn|>": 50342,
|
| 127 |
+
"<|no|>": 50288,
|
| 128 |
+
"<|oc|>": 50328,
|
| 129 |
+
"<|pa|>": 50321,
|
| 130 |
+
"<|pl|>": 50269,
|
| 131 |
+
"<|ps|>": 50340,
|
| 132 |
+
"<|pt|>": 50267,
|
| 133 |
+
"<|ro|>": 50284,
|
| 134 |
+
"<|ru|>": 50263,
|
| 135 |
+
"<|sa|>": 50344,
|
| 136 |
+
"<|sd|>": 50332,
|
| 137 |
+
"<|si|>": 50322,
|
| 138 |
+
"<|sk|>": 50298,
|
| 139 |
+
"<|sl|>": 50305,
|
| 140 |
+
"<|sn|>": 50324,
|
| 141 |
+
"<|so|>": 50326,
|
| 142 |
+
"<|sq|>": 50317,
|
| 143 |
+
"<|sr|>": 50303,
|
| 144 |
+
"<|su|>": 50357,
|
| 145 |
+
"<|sv|>": 50273,
|
| 146 |
+
"<|sw|>": 50318,
|
| 147 |
+
"<|ta|>": 50287,
|
| 148 |
+
"<|te|>": 50299,
|
| 149 |
+
"<|tg|>": 50331,
|
| 150 |
+
"<|th|>": 50289,
|
| 151 |
+
"<|tk|>": 50341,
|
| 152 |
+
"<|tl|>": 50348,
|
| 153 |
+
"<|tr|>": 50268,
|
| 154 |
+
"<|tt|>": 50351,
|
| 155 |
+
"<|uk|>": 50280,
|
| 156 |
+
"<|ur|>": 50290,
|
| 157 |
+
"<|uz|>": 50337,
|
| 158 |
+
"<|vi|>": 50278,
|
| 159 |
+
"<|yi|>": 50335,
|
| 160 |
+
"<|yo|>": 50325,
|
| 161 |
+
"<|zh|>": 50260
|
| 162 |
+
},
|
| 163 |
+
"max_initial_timestamp_index": 50,
|
| 164 |
+
"max_length": 448,
|
| 165 |
+
"no_timestamps_token_id": 50363,
|
| 166 |
+
"pad_token_id": 50257,
|
| 167 |
+
"prev_sot_token_id": 50361,
|
| 168 |
+
"return_timestamps": false,
|
| 169 |
+
"suppress_tokens": [],
|
| 170 |
+
"task_to_id": {
|
| 171 |
+
"transcribe": 50359,
|
| 172 |
+
"translate": 50358
|
| 173 |
+
},
|
| 174 |
+
"transformers_version": "4.51.3"
|
| 175 |
+
}
|
checkpoint-500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d528589c3533907ce7d62b0f82eb8e4b39324a2ede9668d4637805894ab9c265
|
| 3 |
+
size 966995080
|
checkpoint-500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2312d2bf46c62764990aa1f900a0cc3f4aa4aceee8f2f20998ac84b07c5d89a6
|
| 3 |
+
size 1925064044
|
checkpoint-500/preprocessor_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"chunk_length": 30,
|
| 3 |
+
"dither": 0.0,
|
| 4 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 5 |
+
"feature_size": 80,
|
| 6 |
+
"hop_length": 160,
|
| 7 |
+
"n_fft": 400,
|
| 8 |
+
"n_samples": 480000,
|
| 9 |
+
"nb_max_frames": 3000,
|
| 10 |
+
"padding_side": "right",
|
| 11 |
+
"padding_value": 0.0,
|
| 12 |
+
"processor_class": "WhisperProcessor",
|
| 13 |
+
"return_attention_mask": false,
|
| 14 |
+
"sampling_rate": 16000
|
| 15 |
+
}
|
checkpoint-500/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:825599fa0730d95fb5920cab5c6707dfe544fd8910dfd0d5035f628a04b4f17f
|
| 3 |
+
size 14244
|
checkpoint-500/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc4ce6f6531c680ba0e80ac9ab7f956e79be869228ce1b03c87da5b41fcf9d77
|
| 3 |
+
size 988
|
checkpoint-500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e37d0e3abe4f26a67104e2dc03695cbf6a36428ae4534496e8677a64b279559
|
| 3 |
+
size 1064
|
checkpoint-500/trainer_state.json
ADDED
|
@@ -0,0 +1,254 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 400,
|
| 3 |
+
"best_metric": 25.511409562892496,
|
| 4 |
+
"best_model_checkpoint": "./JUDIC/checkpoint-400",
|
| 5 |
+
"epoch": 5.882352941176471,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.23529411764705882,
|
| 14 |
+
"grad_norm": 4.239101886749268,
|
| 15 |
+
"learning_rate": 1.9000000000000002e-06,
|
| 16 |
+
"loss": 0.4078,
|
| 17 |
+
"step": 20
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.47058823529411764,
|
| 21 |
+
"grad_norm": 2.235217809677124,
|
| 22 |
+
"learning_rate": 3.900000000000001e-06,
|
| 23 |
+
"loss": 0.4217,
|
| 24 |
+
"step": 40
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.7058823529411765,
|
| 28 |
+
"grad_norm": 3.711195707321167,
|
| 29 |
+
"learning_rate": 5.9e-06,
|
| 30 |
+
"loss": 0.4252,
|
| 31 |
+
"step": 60
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.9411764705882353,
|
| 35 |
+
"grad_norm": 4.5320820808410645,
|
| 36 |
+
"learning_rate": 7.9e-06,
|
| 37 |
+
"loss": 0.3812,
|
| 38 |
+
"step": 80
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.1764705882352942,
|
| 42 |
+
"grad_norm": 3.9703948497772217,
|
| 43 |
+
"learning_rate": 9.9e-06,
|
| 44 |
+
"loss": 0.4223,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.1764705882352942,
|
| 49 |
+
"eval_loss": 0.40899839997291565,
|
| 50 |
+
"eval_runtime": 323.6745,
|
| 51 |
+
"eval_samples_per_second": 2.104,
|
| 52 |
+
"eval_steps_per_second": 0.266,
|
| 53 |
+
"eval_wer": 36.6100851831126,
|
| 54 |
+
"step": 100
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 1.4117647058823528,
|
| 58 |
+
"grad_norm": 3.8265273571014404,
|
| 59 |
+
"learning_rate": 9.746666666666668e-06,
|
| 60 |
+
"loss": 0.2872,
|
| 61 |
+
"step": 120
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 1.6470588235294117,
|
| 65 |
+
"grad_norm": 4.8625168800354,
|
| 66 |
+
"learning_rate": 9.48e-06,
|
| 67 |
+
"loss": 0.3207,
|
| 68 |
+
"step": 140
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.8823529411764706,
|
| 72 |
+
"grad_norm": 4.235057830810547,
|
| 73 |
+
"learning_rate": 9.213333333333334e-06,
|
| 74 |
+
"loss": 0.3585,
|
| 75 |
+
"step": 160
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 2.1176470588235294,
|
| 79 |
+
"grad_norm": 2.1225504875183105,
|
| 80 |
+
"learning_rate": 8.973333333333334e-06,
|
| 81 |
+
"loss": 0.2658,
|
| 82 |
+
"step": 180
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 2.3529411764705883,
|
| 86 |
+
"grad_norm": 8.930822372436523,
|
| 87 |
+
"learning_rate": 8.706666666666667e-06,
|
| 88 |
+
"loss": 0.2576,
|
| 89 |
+
"step": 200
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 2.3529411764705883,
|
| 93 |
+
"eval_loss": 0.35160648822784424,
|
| 94 |
+
"eval_runtime": 296.116,
|
| 95 |
+
"eval_samples_per_second": 2.3,
|
| 96 |
+
"eval_steps_per_second": 0.29,
|
| 97 |
+
"eval_wer": 29.938444320089534,
|
| 98 |
+
"step": 200
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 2.588235294117647,
|
| 102 |
+
"grad_norm": 1.942764163017273,
|
| 103 |
+
"learning_rate": 8.44e-06,
|
| 104 |
+
"loss": 0.283,
|
| 105 |
+
"step": 220
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 2.8235294117647056,
|
| 109 |
+
"grad_norm": 3.2245473861694336,
|
| 110 |
+
"learning_rate": 8.173333333333334e-06,
|
| 111 |
+
"loss": 0.2443,
|
| 112 |
+
"step": 240
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 3.0588235294117645,
|
| 116 |
+
"grad_norm": 3.0929267406463623,
|
| 117 |
+
"learning_rate": 7.906666666666667e-06,
|
| 118 |
+
"loss": 0.2489,
|
| 119 |
+
"step": 260
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 3.2941176470588234,
|
| 123 |
+
"grad_norm": 3.477473497390747,
|
| 124 |
+
"learning_rate": 7.640000000000001e-06,
|
| 125 |
+
"loss": 0.2006,
|
| 126 |
+
"step": 280
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 3.5294117647058822,
|
| 130 |
+
"grad_norm": 3.163804531097412,
|
| 131 |
+
"learning_rate": 7.373333333333334e-06,
|
| 132 |
+
"loss": 0.1837,
|
| 133 |
+
"step": 300
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 3.5294117647058822,
|
| 137 |
+
"eval_loss": 0.30386850237846375,
|
| 138 |
+
"eval_runtime": 329.1958,
|
| 139 |
+
"eval_samples_per_second": 2.069,
|
| 140 |
+
"eval_steps_per_second": 0.261,
|
| 141 |
+
"eval_wer": 31.555058135919918,
|
| 142 |
+
"step": 300
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 3.764705882352941,
|
| 146 |
+
"grad_norm": 2.77553391456604,
|
| 147 |
+
"learning_rate": 7.1066666666666675e-06,
|
| 148 |
+
"loss": 0.2342,
|
| 149 |
+
"step": 320
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 4.0,
|
| 153 |
+
"grad_norm": 3.944244861602783,
|
| 154 |
+
"learning_rate": 6.8400000000000014e-06,
|
| 155 |
+
"loss": 0.1751,
|
| 156 |
+
"step": 340
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 4.235294117647059,
|
| 160 |
+
"grad_norm": 6.829845905303955,
|
| 161 |
+
"learning_rate": 6.573333333333334e-06,
|
| 162 |
+
"loss": 0.1431,
|
| 163 |
+
"step": 360
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 4.470588235294118,
|
| 167 |
+
"grad_norm": 3.4126970767974854,
|
| 168 |
+
"learning_rate": 6.3066666666666676e-06,
|
| 169 |
+
"loss": 0.1468,
|
| 170 |
+
"step": 380
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.705882352941177,
|
| 174 |
+
"grad_norm": 3.5233240127563477,
|
| 175 |
+
"learning_rate": 6.040000000000001e-06,
|
| 176 |
+
"loss": 0.1376,
|
| 177 |
+
"step": 400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 4.705882352941177,
|
| 181 |
+
"eval_loss": 0.27411767840385437,
|
| 182 |
+
"eval_runtime": 307.4101,
|
| 183 |
+
"eval_samples_per_second": 2.215,
|
| 184 |
+
"eval_steps_per_second": 0.28,
|
| 185 |
+
"eval_wer": 25.511409562892496,
|
| 186 |
+
"step": 400
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 4.9411764705882355,
|
| 190 |
+
"grad_norm": 2.9697437286376953,
|
| 191 |
+
"learning_rate": 5.7733333333333345e-06,
|
| 192 |
+
"loss": 0.1395,
|
| 193 |
+
"step": 420
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 5.176470588235294,
|
| 197 |
+
"grad_norm": 3.495859384536743,
|
| 198 |
+
"learning_rate": 5.506666666666667e-06,
|
| 199 |
+
"loss": 0.1046,
|
| 200 |
+
"step": 440
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 5.411764705882353,
|
| 204 |
+
"grad_norm": 4.7592902183532715,
|
| 205 |
+
"learning_rate": 5.240000000000001e-06,
|
| 206 |
+
"loss": 0.1179,
|
| 207 |
+
"step": 460
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 5.647058823529412,
|
| 211 |
+
"grad_norm": 7.579683303833008,
|
| 212 |
+
"learning_rate": 4.973333333333334e-06,
|
| 213 |
+
"loss": 0.1361,
|
| 214 |
+
"step": 480
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 5.882352941176471,
|
| 218 |
+
"grad_norm": 4.949639797210693,
|
| 219 |
+
"learning_rate": 4.706666666666667e-06,
|
| 220 |
+
"loss": 0.0935,
|
| 221 |
+
"step": 500
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 5.882352941176471,
|
| 225 |
+
"eval_loss": 0.2585192918777466,
|
| 226 |
+
"eval_runtime": 321.1171,
|
| 227 |
+
"eval_samples_per_second": 2.121,
|
| 228 |
+
"eval_steps_per_second": 0.268,
|
| 229 |
+
"eval_wer": 26.26997450724367,
|
| 230 |
+
"step": 500
|
| 231 |
+
}
|
| 232 |
+
],
|
| 233 |
+
"logging_steps": 20,
|
| 234 |
+
"max_steps": 850,
|
| 235 |
+
"num_input_tokens_seen": 0,
|
| 236 |
+
"num_train_epochs": 10,
|
| 237 |
+
"save_steps": 100,
|
| 238 |
+
"stateful_callbacks": {
|
| 239 |
+
"TrainerControl": {
|
| 240 |
+
"args": {
|
| 241 |
+
"should_epoch_stop": false,
|
| 242 |
+
"should_evaluate": false,
|
| 243 |
+
"should_log": false,
|
| 244 |
+
"should_save": true,
|
| 245 |
+
"should_training_stop": false
|
| 246 |
+
},
|
| 247 |
+
"attributes": {}
|
| 248 |
+
}
|
| 249 |
+
},
|
| 250 |
+
"total_flos": 4.61736640512e+18,
|
| 251 |
+
"train_batch_size": 16,
|
| 252 |
+
"trial_name": null,
|
| 253 |
+
"trial_params": null
|
| 254 |
+
}
|
checkpoint-500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a582364a1defc1cdd820215263292cba4d325748959fa5eb84ed58c1bcccb33d
|
| 3 |
+
size 5432
|
checkpoint-600/config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"WhisperForConditionalGeneration"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"begin_suppress_tokens": null,
|
| 10 |
+
"bos_token_id": 50257,
|
| 11 |
+
"classifier_proj_size": 256,
|
| 12 |
+
"d_model": 768,
|
| 13 |
+
"decoder_attention_heads": 12,
|
| 14 |
+
"decoder_ffn_dim": 3072,
|
| 15 |
+
"decoder_layerdrop": 0.0,
|
| 16 |
+
"decoder_layers": 12,
|
| 17 |
+
"decoder_start_token_id": 50258,
|
| 18 |
+
"dropout": 0.0,
|
| 19 |
+
"encoder_attention_heads": 12,
|
| 20 |
+
"encoder_ffn_dim": 3072,
|
| 21 |
+
"encoder_layerdrop": 0.0,
|
| 22 |
+
"encoder_layers": 12,
|
| 23 |
+
"eos_token_id": 50257,
|
| 24 |
+
"forced_decoder_ids": null,
|
| 25 |
+
"init_std": 0.02,
|
| 26 |
+
"is_encoder_decoder": true,
|
| 27 |
+
"mask_feature_length": 10,
|
| 28 |
+
"mask_feature_min_masks": 0,
|
| 29 |
+
"mask_feature_prob": 0.0,
|
| 30 |
+
"mask_time_length": 10,
|
| 31 |
+
"mask_time_min_masks": 2,
|
| 32 |
+
"mask_time_prob": 0.05,
|
| 33 |
+
"max_length": null,
|
| 34 |
+
"max_source_positions": 1500,
|
| 35 |
+
"max_target_positions": 448,
|
| 36 |
+
"median_filter_width": 7,
|
| 37 |
+
"model_type": "whisper",
|
| 38 |
+
"num_hidden_layers": 12,
|
| 39 |
+
"num_mel_bins": 80,
|
| 40 |
+
"pad_token_id": 50257,
|
| 41 |
+
"scale_embedding": false,
|
| 42 |
+
"torch_dtype": "float32",
|
| 43 |
+
"transformers_version": "4.51.3",
|
| 44 |
+
"use_cache": false,
|
| 45 |
+
"use_weighted_layer_sum": false,
|
| 46 |
+
"vocab_size": 51865
|
| 47 |
+
}
|
checkpoint-600/generation_config.json
ADDED
|
@@ -0,0 +1,175 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alignment_heads": [
|
| 3 |
+
[
|
| 4 |
+
5,
|
| 5 |
+
3
|
| 6 |
+
],
|
| 7 |
+
[
|
| 8 |
+
5,
|
| 9 |
+
9
|
| 10 |
+
],
|
| 11 |
+
[
|
| 12 |
+
8,
|
| 13 |
+
0
|
| 14 |
+
],
|
| 15 |
+
[
|
| 16 |
+
8,
|
| 17 |
+
4
|
| 18 |
+
],
|
| 19 |
+
[
|
| 20 |
+
8,
|
| 21 |
+
7
|
| 22 |
+
],
|
| 23 |
+
[
|
| 24 |
+
8,
|
| 25 |
+
8
|
| 26 |
+
],
|
| 27 |
+
[
|
| 28 |
+
9,
|
| 29 |
+
0
|
| 30 |
+
],
|
| 31 |
+
[
|
| 32 |
+
9,
|
| 33 |
+
7
|
| 34 |
+
],
|
| 35 |
+
[
|
| 36 |
+
9,
|
| 37 |
+
9
|
| 38 |
+
],
|
| 39 |
+
[
|
| 40 |
+
10,
|
| 41 |
+
5
|
| 42 |
+
]
|
| 43 |
+
],
|
| 44 |
+
"begin_suppress_tokens": [
|
| 45 |
+
220,
|
| 46 |
+
50257
|
| 47 |
+
],
|
| 48 |
+
"bos_token_id": 50257,
|
| 49 |
+
"decoder_start_token_id": 50258,
|
| 50 |
+
"eos_token_id": 50257,
|
| 51 |
+
"forced_decoder_ids": [
|
| 52 |
+
[
|
| 53 |
+
1,
|
| 54 |
+
null
|
| 55 |
+
],
|
| 56 |
+
[
|
| 57 |
+
2,
|
| 58 |
+
50359
|
| 59 |
+
]
|
| 60 |
+
],
|
| 61 |
+
"is_multilingual": true,
|
| 62 |
+
"lang_to_id": {
|
| 63 |
+
"<|af|>": 50327,
|
| 64 |
+
"<|am|>": 50334,
|
| 65 |
+
"<|ar|>": 50272,
|
| 66 |
+
"<|as|>": 50350,
|
| 67 |
+
"<|az|>": 50304,
|
| 68 |
+
"<|ba|>": 50355,
|
| 69 |
+
"<|be|>": 50330,
|
| 70 |
+
"<|bg|>": 50292,
|
| 71 |
+
"<|bn|>": 50302,
|
| 72 |
+
"<|bo|>": 50347,
|
| 73 |
+
"<|br|>": 50309,
|
| 74 |
+
"<|bs|>": 50315,
|
| 75 |
+
"<|ca|>": 50270,
|
| 76 |
+
"<|cs|>": 50283,
|
| 77 |
+
"<|cy|>": 50297,
|
| 78 |
+
"<|da|>": 50285,
|
| 79 |
+
"<|de|>": 50261,
|
| 80 |
+
"<|el|>": 50281,
|
| 81 |
+
"<|en|>": 50259,
|
| 82 |
+
"<|es|>": 50262,
|
| 83 |
+
"<|et|>": 50307,
|
| 84 |
+
"<|eu|>": 50310,
|
| 85 |
+
"<|fa|>": 50300,
|
| 86 |
+
"<|fi|>": 50277,
|
| 87 |
+
"<|fo|>": 50338,
|
| 88 |
+
"<|fr|>": 50265,
|
| 89 |
+
"<|gl|>": 50319,
|
| 90 |
+
"<|gu|>": 50333,
|
| 91 |
+
"<|haw|>": 50352,
|
| 92 |
+
"<|ha|>": 50354,
|
| 93 |
+
"<|he|>": 50279,
|
| 94 |
+
"<|hi|>": 50276,
|
| 95 |
+
"<|hr|>": 50291,
|
| 96 |
+
"<|ht|>": 50339,
|
| 97 |
+
"<|hu|>": 50286,
|
| 98 |
+
"<|hy|>": 50312,
|
| 99 |
+
"<|id|>": 50275,
|
| 100 |
+
"<|is|>": 50311,
|
| 101 |
+
"<|it|>": 50274,
|
| 102 |
+
"<|ja|>": 50266,
|
| 103 |
+
"<|jw|>": 50356,
|
| 104 |
+
"<|ka|>": 50329,
|
| 105 |
+
"<|kk|>": 50316,
|
| 106 |
+
"<|km|>": 50323,
|
| 107 |
+
"<|kn|>": 50306,
|
| 108 |
+
"<|ko|>": 50264,
|
| 109 |
+
"<|la|>": 50294,
|
| 110 |
+
"<|lb|>": 50345,
|
| 111 |
+
"<|ln|>": 50353,
|
| 112 |
+
"<|lo|>": 50336,
|
| 113 |
+
"<|lt|>": 50293,
|
| 114 |
+
"<|lv|>": 50301,
|
| 115 |
+
"<|mg|>": 50349,
|
| 116 |
+
"<|mi|>": 50295,
|
| 117 |
+
"<|mk|>": 50308,
|
| 118 |
+
"<|ml|>": 50296,
|
| 119 |
+
"<|mn|>": 50314,
|
| 120 |
+
"<|mr|>": 50320,
|
| 121 |
+
"<|ms|>": 50282,
|
| 122 |
+
"<|mt|>": 50343,
|
| 123 |
+
"<|my|>": 50346,
|
| 124 |
+
"<|ne|>": 50313,
|
| 125 |
+
"<|nl|>": 50271,
|
| 126 |
+
"<|nn|>": 50342,
|
| 127 |
+
"<|no|>": 50288,
|
| 128 |
+
"<|oc|>": 50328,
|
| 129 |
+
"<|pa|>": 50321,
|
| 130 |
+
"<|pl|>": 50269,
|
| 131 |
+
"<|ps|>": 50340,
|
| 132 |
+
"<|pt|>": 50267,
|
| 133 |
+
"<|ro|>": 50284,
|
| 134 |
+
"<|ru|>": 50263,
|
| 135 |
+
"<|sa|>": 50344,
|
| 136 |
+
"<|sd|>": 50332,
|
| 137 |
+
"<|si|>": 50322,
|
| 138 |
+
"<|sk|>": 50298,
|
| 139 |
+
"<|sl|>": 50305,
|
| 140 |
+
"<|sn|>": 50324,
|
| 141 |
+
"<|so|>": 50326,
|
| 142 |
+
"<|sq|>": 50317,
|
| 143 |
+
"<|sr|>": 50303,
|
| 144 |
+
"<|su|>": 50357,
|
| 145 |
+
"<|sv|>": 50273,
|
| 146 |
+
"<|sw|>": 50318,
|
| 147 |
+
"<|ta|>": 50287,
|
| 148 |
+
"<|te|>": 50299,
|
| 149 |
+
"<|tg|>": 50331,
|
| 150 |
+
"<|th|>": 50289,
|
| 151 |
+
"<|tk|>": 50341,
|
| 152 |
+
"<|tl|>": 50348,
|
| 153 |
+
"<|tr|>": 50268,
|
| 154 |
+
"<|tt|>": 50351,
|
| 155 |
+
"<|uk|>": 50280,
|
| 156 |
+
"<|ur|>": 50290,
|
| 157 |
+
"<|uz|>": 50337,
|
| 158 |
+
"<|vi|>": 50278,
|
| 159 |
+
"<|yi|>": 50335,
|
| 160 |
+
"<|yo|>": 50325,
|
| 161 |
+
"<|zh|>": 50260
|
| 162 |
+
},
|
| 163 |
+
"max_initial_timestamp_index": 50,
|
| 164 |
+
"max_length": 448,
|
| 165 |
+
"no_timestamps_token_id": 50363,
|
| 166 |
+
"pad_token_id": 50257,
|
| 167 |
+
"prev_sot_token_id": 50361,
|
| 168 |
+
"return_timestamps": false,
|
| 169 |
+
"suppress_tokens": [],
|
| 170 |
+
"task_to_id": {
|
| 171 |
+
"transcribe": 50359,
|
| 172 |
+
"translate": 50358
|
| 173 |
+
},
|
| 174 |
+
"transformers_version": "4.51.3"
|
| 175 |
+
}
|
checkpoint-600/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3799349bbd9f1b5bc5b63fe1366f4ff31d0d06cabecdbfa7aef16ea012bf4d1d
|
| 3 |
+
size 966995080
|
checkpoint-600/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c86ac17197344a70407c4bcbd330f7eb44fbf135c27a3a8744bc30e21c62cf5
|
| 3 |
+
size 1925064044
|
checkpoint-600/preprocessor_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"chunk_length": 30,
|
| 3 |
+
"dither": 0.0,
|
| 4 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 5 |
+
"feature_size": 80,
|
| 6 |
+
"hop_length": 160,
|
| 7 |
+
"n_fft": 400,
|
| 8 |
+
"n_samples": 480000,
|
| 9 |
+
"nb_max_frames": 3000,
|
| 10 |
+
"padding_side": "right",
|
| 11 |
+
"padding_value": 0.0,
|
| 12 |
+
"processor_class": "WhisperProcessor",
|
| 13 |
+
"return_attention_mask": false,
|
| 14 |
+
"sampling_rate": 16000
|
| 15 |
+
}
|
checkpoint-600/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:480661d528573adc366803cd6192719058233ab699007a6e9d72a2a4445eab53
|
| 3 |
+
size 14244
|