Commit
•
1f39d64
1
Parent(s):
e4c6b7b
Upload folder using huggingface_hub
Browse files- config.json +52 -0
- generation_config.json +264 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- preprocessor_config.json +14 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1257 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "openai/whisper-small",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "gelu",
|
| 5 |
+
"apply_spec_augment": false,
|
| 6 |
+
"architectures": [
|
| 7 |
+
"WhisperForConditionalGeneration"
|
| 8 |
+
],
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"begin_suppress_tokens": [
|
| 11 |
+
220,
|
| 12 |
+
50257
|
| 13 |
+
],
|
| 14 |
+
"bos_token_id": 50257,
|
| 15 |
+
"classifier_proj_size": 256,
|
| 16 |
+
"d_model": 768,
|
| 17 |
+
"decoder_attention_heads": 12,
|
| 18 |
+
"decoder_ffn_dim": 3072,
|
| 19 |
+
"decoder_layerdrop": 0.0,
|
| 20 |
+
"decoder_layers": 12,
|
| 21 |
+
"decoder_start_token_id": 50258,
|
| 22 |
+
"dropout": 0.0,
|
| 23 |
+
"encoder_attention_heads": 12,
|
| 24 |
+
"encoder_ffn_dim": 3072,
|
| 25 |
+
"encoder_layerdrop": 0.0,
|
| 26 |
+
"encoder_layers": 12,
|
| 27 |
+
"eos_token_id": 50257,
|
| 28 |
+
"forced_decoder_ids": null,
|
| 29 |
+
"init_std": 0.02,
|
| 30 |
+
"is_encoder_decoder": true,
|
| 31 |
+
"mask_feature_length": 10,
|
| 32 |
+
"mask_feature_min_masks": 0,
|
| 33 |
+
"mask_feature_prob": 0.0,
|
| 34 |
+
"mask_time_length": 10,
|
| 35 |
+
"mask_time_min_masks": 2,
|
| 36 |
+
"mask_time_prob": 0.05,
|
| 37 |
+
"max_length": 448,
|
| 38 |
+
"max_source_positions": 1500,
|
| 39 |
+
"max_target_positions": 448,
|
| 40 |
+
"median_filter_width": 7,
|
| 41 |
+
"model_type": "whisper",
|
| 42 |
+
"num_hidden_layers": 12,
|
| 43 |
+
"num_mel_bins": 80,
|
| 44 |
+
"pad_token_id": 50257,
|
| 45 |
+
"scale_embedding": false,
|
| 46 |
+
"suppress_tokens": [],
|
| 47 |
+
"torch_dtype": "float32",
|
| 48 |
+
"transformers_version": "4.38.0.dev0",
|
| 49 |
+
"use_cache": true,
|
| 50 |
+
"use_weighted_layer_sum": false,
|
| 51 |
+
"vocab_size": 51865
|
| 52 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,264 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
1,
|
| 171 |
+
2,
|
| 172 |
+
7,
|
| 173 |
+
8,
|
| 174 |
+
9,
|
| 175 |
+
10,
|
| 176 |
+
14,
|
| 177 |
+
25,
|
| 178 |
+
26,
|
| 179 |
+
27,
|
| 180 |
+
28,
|
| 181 |
+
29,
|
| 182 |
+
31,
|
| 183 |
+
58,
|
| 184 |
+
59,
|
| 185 |
+
60,
|
| 186 |
+
61,
|
| 187 |
+
62,
|
| 188 |
+
63,
|
| 189 |
+
90,
|
| 190 |
+
91,
|
| 191 |
+
92,
|
| 192 |
+
93,
|
| 193 |
+
359,
|
| 194 |
+
503,
|
| 195 |
+
522,
|
| 196 |
+
542,
|
| 197 |
+
873,
|
| 198 |
+
893,
|
| 199 |
+
902,
|
| 200 |
+
918,
|
| 201 |
+
922,
|
| 202 |
+
931,
|
| 203 |
+
1350,
|
| 204 |
+
1853,
|
| 205 |
+
1982,
|
| 206 |
+
2460,
|
| 207 |
+
2627,
|
| 208 |
+
3246,
|
| 209 |
+
3253,
|
| 210 |
+
3268,
|
| 211 |
+
3536,
|
| 212 |
+
3846,
|
| 213 |
+
3961,
|
| 214 |
+
4183,
|
| 215 |
+
4667,
|
| 216 |
+
6585,
|
| 217 |
+
6647,
|
| 218 |
+
7273,
|
| 219 |
+
9061,
|
| 220 |
+
9383,
|
| 221 |
+
10428,
|
| 222 |
+
10929,
|
| 223 |
+
11938,
|
| 224 |
+
12033,
|
| 225 |
+
12331,
|
| 226 |
+
12562,
|
| 227 |
+
13793,
|
| 228 |
+
14157,
|
| 229 |
+
14635,
|
| 230 |
+
15265,
|
| 231 |
+
15618,
|
| 232 |
+
16553,
|
| 233 |
+
16604,
|
| 234 |
+
18362,
|
| 235 |
+
18956,
|
| 236 |
+
20075,
|
| 237 |
+
21675,
|
| 238 |
+
22520,
|
| 239 |
+
26130,
|
| 240 |
+
26161,
|
| 241 |
+
26435,
|
| 242 |
+
28279,
|
| 243 |
+
29464,
|
| 244 |
+
31650,
|
| 245 |
+
32302,
|
| 246 |
+
32470,
|
| 247 |
+
36865,
|
| 248 |
+
42863,
|
| 249 |
+
47425,
|
| 250 |
+
49870,
|
| 251 |
+
50254,
|
| 252 |
+
50258,
|
| 253 |
+
50358,
|
| 254 |
+
50359,
|
| 255 |
+
50360,
|
| 256 |
+
50361,
|
| 257 |
+
50362
|
| 258 |
+
],
|
| 259 |
+
"task_to_id": {
|
| 260 |
+
"transcribe": 50359,
|
| 261 |
+
"translate": 50358
|
| 262 |
+
},
|
| 263 |
+
"transformers_version": "4.38.0.dev0"
|
| 264 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:819208b8f73b694ced19c0707f3a18847a2cf1468c79409c06ca1530090f80ea
|
| 3 |
+
size 966995080
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ea8eb5f9acf4f75733de6bd252be226c7ba3c2ece3f4bd84d340407cbe2ecb6
|
| 3 |
+
size 1925064044
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"chunk_length": 30,
|
| 3 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 4 |
+
"feature_size": 80,
|
| 5 |
+
"hop_length": 160,
|
| 6 |
+
"n_fft": 400,
|
| 7 |
+
"n_samples": 480000,
|
| 8 |
+
"nb_max_frames": 3000,
|
| 9 |
+
"padding_side": "right",
|
| 10 |
+
"padding_value": 0.0,
|
| 11 |
+
"processor_class": "WhisperProcessor",
|
| 12 |
+
"return_attention_mask": false,
|
| 13 |
+
"sampling_rate": 16000
|
| 14 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:afe9cb20ebfcddaeeb71af090bd3ea2a6e3690d6a96eda45b9a5076bc0242894
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7efbf80617c96c78286826ce59d9a12c86da62d7631874b3d6364a8e993ada60
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1257 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 66.17647058823529,
|
| 3 |
+
"best_model_checkpoint": "./whisper-small-bn/checkpoint-4000",
|
| 4 |
+
"epoch": 40.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 4000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.2,
|
| 13 |
+
"learning_rate": 3.2e-07,
|
| 14 |
+
"loss": 2.3459,
|
| 15 |
+
"step": 20
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.4,
|
| 19 |
+
"learning_rate": 7.2e-07,
|
| 20 |
+
"loss": 2.1757,
|
| 21 |
+
"step": 40
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.6,
|
| 25 |
+
"learning_rate": 1.12e-06,
|
| 26 |
+
"loss": 1.8784,
|
| 27 |
+
"step": 60
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.8,
|
| 31 |
+
"learning_rate": 1.52e-06,
|
| 32 |
+
"loss": 1.7012,
|
| 33 |
+
"step": 80
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 1.0,
|
| 37 |
+
"learning_rate": 1.9200000000000003e-06,
|
| 38 |
+
"loss": 1.5541,
|
| 39 |
+
"step": 100
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.2,
|
| 43 |
+
"learning_rate": 2.3200000000000002e-06,
|
| 44 |
+
"loss": 1.3971,
|
| 45 |
+
"step": 120
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 1.4,
|
| 49 |
+
"learning_rate": 2.7200000000000002e-06,
|
| 50 |
+
"loss": 1.3046,
|
| 51 |
+
"step": 140
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 1.6,
|
| 55 |
+
"learning_rate": 3.12e-06,
|
| 56 |
+
"loss": 1.1632,
|
| 57 |
+
"step": 160
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 1.8,
|
| 61 |
+
"learning_rate": 3.52e-06,
|
| 62 |
+
"loss": 1.0418,
|
| 63 |
+
"step": 180
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 2.0,
|
| 67 |
+
"learning_rate": 3.920000000000001e-06,
|
| 68 |
+
"loss": 0.9148,
|
| 69 |
+
"step": 200
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 2.2,
|
| 73 |
+
"learning_rate": 4.32e-06,
|
| 74 |
+
"loss": 0.7515,
|
| 75 |
+
"step": 220
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 2.4,
|
| 79 |
+
"learning_rate": 4.7200000000000005e-06,
|
| 80 |
+
"loss": 0.7166,
|
| 81 |
+
"step": 240
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 2.6,
|
| 85 |
+
"learning_rate": 5.12e-06,
|
| 86 |
+
"loss": 0.6553,
|
| 87 |
+
"step": 260
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 2.8,
|
| 91 |
+
"learning_rate": 5.5200000000000005e-06,
|
| 92 |
+
"loss": 0.5869,
|
| 93 |
+
"step": 280
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 3.0,
|
| 97 |
+
"learning_rate": 5.92e-06,
|
| 98 |
+
"loss": 0.58,
|
| 99 |
+
"step": 300
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 3.2,
|
| 103 |
+
"learning_rate": 6.3200000000000005e-06,
|
| 104 |
+
"loss": 0.4109,
|
| 105 |
+
"step": 320
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 3.4,
|
| 109 |
+
"learning_rate": 6.720000000000001e-06,
|
| 110 |
+
"loss": 0.3971,
|
| 111 |
+
"step": 340
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 3.6,
|
| 115 |
+
"learning_rate": 7.1200000000000004e-06,
|
| 116 |
+
"loss": 0.3937,
|
| 117 |
+
"step": 360
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 3.8,
|
| 121 |
+
"learning_rate": 7.520000000000001e-06,
|
| 122 |
+
"loss": 0.3944,
|
| 123 |
+
"step": 380
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 4.0,
|
| 127 |
+
"learning_rate": 7.92e-06,
|
| 128 |
+
"loss": 0.3928,
|
| 129 |
+
"step": 400
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 4.2,
|
| 133 |
+
"learning_rate": 8.32e-06,
|
| 134 |
+
"loss": 0.2465,
|
| 135 |
+
"step": 420
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 4.4,
|
| 139 |
+
"learning_rate": 8.720000000000001e-06,
|
| 140 |
+
"loss": 0.2513,
|
| 141 |
+
"step": 440
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 4.6,
|
| 145 |
+
"learning_rate": 9.12e-06,
|
| 146 |
+
"loss": 0.2313,
|
| 147 |
+
"step": 460
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 4.8,
|
| 151 |
+
"learning_rate": 9.52e-06,
|
| 152 |
+
"loss": 0.2216,
|
| 153 |
+
"step": 480
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 5.0,
|
| 157 |
+
"learning_rate": 9.920000000000002e-06,
|
| 158 |
+
"loss": 0.2286,
|
| 159 |
+
"step": 500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 5.2,
|
| 163 |
+
"learning_rate": 9.954285714285715e-06,
|
| 164 |
+
"loss": 0.1615,
|
| 165 |
+
"step": 520
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 5.4,
|
| 169 |
+
"learning_rate": 9.897142857142858e-06,
|
| 170 |
+
"loss": 0.1365,
|
| 171 |
+
"step": 540
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 5.6,
|
| 175 |
+
"learning_rate": 9.84e-06,
|
| 176 |
+
"loss": 0.1372,
|
| 177 |
+
"step": 560
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 5.8,
|
| 181 |
+
"learning_rate": 9.782857142857145e-06,
|
| 182 |
+
"loss": 0.125,
|
| 183 |
+
"step": 580
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 6.0,
|
| 187 |
+
"learning_rate": 9.725714285714287e-06,
|
| 188 |
+
"loss": 0.122,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 6.2,
|
| 193 |
+
"learning_rate": 9.668571428571429e-06,
|
| 194 |
+
"loss": 0.0747,
|
| 195 |
+
"step": 620
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 6.4,
|
| 199 |
+
"learning_rate": 9.611428571428573e-06,
|
| 200 |
+
"loss": 0.0728,
|
| 201 |
+
"step": 640
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 6.6,
|
| 205 |
+
"learning_rate": 9.554285714285715e-06,
|
| 206 |
+
"loss": 0.0837,
|
| 207 |
+
"step": 660
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 6.8,
|
| 211 |
+
"learning_rate": 9.497142857142859e-06,
|
| 212 |
+
"loss": 0.0774,
|
| 213 |
+
"step": 680
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 7.0,
|
| 217 |
+
"learning_rate": 9.440000000000001e-06,
|
| 218 |
+
"loss": 0.0883,
|
| 219 |
+
"step": 700
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 7.2,
|
| 223 |
+
"learning_rate": 9.382857142857143e-06,
|
| 224 |
+
"loss": 0.0398,
|
| 225 |
+
"step": 720
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 7.4,
|
| 229 |
+
"learning_rate": 9.325714285714287e-06,
|
| 230 |
+
"loss": 0.0558,
|
| 231 |
+
"step": 740
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 7.6,
|
| 235 |
+
"learning_rate": 9.26857142857143e-06,
|
| 236 |
+
"loss": 0.0618,
|
| 237 |
+
"step": 760
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 7.8,
|
| 241 |
+
"learning_rate": 9.211428571428572e-06,
|
| 242 |
+
"loss": 0.04,
|
| 243 |
+
"step": 780
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 8.0,
|
| 247 |
+
"learning_rate": 9.154285714285715e-06,
|
| 248 |
+
"loss": 0.0534,
|
| 249 |
+
"step": 800
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 8.2,
|
| 253 |
+
"learning_rate": 9.097142857142858e-06,
|
| 254 |
+
"loss": 0.046,
|
| 255 |
+
"step": 820
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 8.4,
|
| 259 |
+
"learning_rate": 9.040000000000002e-06,
|
| 260 |
+
"loss": 0.0452,
|
| 261 |
+
"step": 840
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 8.6,
|
| 265 |
+
"learning_rate": 8.982857142857144e-06,
|
| 266 |
+
"loss": 0.0442,
|
| 267 |
+
"step": 860
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 8.8,
|
| 271 |
+
"learning_rate": 8.925714285714286e-06,
|
| 272 |
+
"loss": 0.0315,
|
| 273 |
+
"step": 880
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 9.0,
|
| 277 |
+
"learning_rate": 8.86857142857143e-06,
|
| 278 |
+
"loss": 0.0484,
|
| 279 |
+
"step": 900
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 9.2,
|
| 283 |
+
"learning_rate": 8.811428571428572e-06,
|
| 284 |
+
"loss": 0.0214,
|
| 285 |
+
"step": 920
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 9.4,
|
| 289 |
+
"learning_rate": 8.754285714285716e-06,
|
| 290 |
+
"loss": 0.0351,
|
| 291 |
+
"step": 940
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 9.6,
|
| 295 |
+
"learning_rate": 8.697142857142858e-06,
|
| 296 |
+
"loss": 0.0276,
|
| 297 |
+
"step": 960
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 9.8,
|
| 301 |
+
"learning_rate": 8.64e-06,
|
| 302 |
+
"loss": 0.0271,
|
| 303 |
+
"step": 980
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 10.0,
|
| 307 |
+
"learning_rate": 8.582857142857144e-06,
|
| 308 |
+
"loss": 0.0271,
|
| 309 |
+
"step": 1000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 10.0,
|
| 313 |
+
"eval_loss": 0.5339307188987732,
|
| 314 |
+
"eval_runtime": 341.9642,
|
| 315 |
+
"eval_samples_per_second": 0.292,
|
| 316 |
+
"eval_steps_per_second": 0.292,
|
| 317 |
+
"eval_wer": 76.58371040723982,
|
| 318 |
+
"step": 1000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 10.2,
|
| 322 |
+
"learning_rate": 8.525714285714286e-06,
|
| 323 |
+
"loss": 0.0298,
|
| 324 |
+
"step": 1020
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 10.4,
|
| 328 |
+
"learning_rate": 8.468571428571429e-06,
|
| 329 |
+
"loss": 0.0196,
|
| 330 |
+
"step": 1040
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 10.6,
|
| 334 |
+
"learning_rate": 8.411428571428572e-06,
|
| 335 |
+
"loss": 0.0203,
|
| 336 |
+
"step": 1060
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 10.8,
|
| 340 |
+
"learning_rate": 8.354285714285715e-06,
|
| 341 |
+
"loss": 0.0258,
|
| 342 |
+
"step": 1080
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 11.0,
|
| 346 |
+
"learning_rate": 8.297142857142859e-06,
|
| 347 |
+
"loss": 0.0154,
|
| 348 |
+
"step": 1100
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 11.2,
|
| 352 |
+
"learning_rate": 8.24e-06,
|
| 353 |
+
"loss": 0.0118,
|
| 354 |
+
"step": 1120
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 11.4,
|
| 358 |
+
"learning_rate": 8.182857142857143e-06,
|
| 359 |
+
"loss": 0.0157,
|
| 360 |
+
"step": 1140
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 11.6,
|
| 364 |
+
"learning_rate": 8.125714285714287e-06,
|
| 365 |
+
"loss": 0.0187,
|
| 366 |
+
"step": 1160
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 11.8,
|
| 370 |
+
"learning_rate": 8.068571428571429e-06,
|
| 371 |
+
"loss": 0.0246,
|
| 372 |
+
"step": 1180
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 12.0,
|
| 376 |
+
"learning_rate": 8.011428571428573e-06,
|
| 377 |
+
"loss": 0.012,
|
| 378 |
+
"step": 1200
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 12.2,
|
| 382 |
+
"learning_rate": 7.954285714285715e-06,
|
| 383 |
+
"loss": 0.0159,
|
| 384 |
+
"step": 1220
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 12.4,
|
| 388 |
+
"learning_rate": 7.897142857142857e-06,
|
| 389 |
+
"loss": 0.0166,
|
| 390 |
+
"step": 1240
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 12.6,
|
| 394 |
+
"learning_rate": 7.840000000000001e-06,
|
| 395 |
+
"loss": 0.0105,
|
| 396 |
+
"step": 1260
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 12.8,
|
| 400 |
+
"learning_rate": 7.782857142857143e-06,
|
| 401 |
+
"loss": 0.0097,
|
| 402 |
+
"step": 1280
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 13.0,
|
| 406 |
+
"learning_rate": 7.725714285714286e-06,
|
| 407 |
+
"loss": 0.0066,
|
| 408 |
+
"step": 1300
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 13.2,
|
| 412 |
+
"learning_rate": 7.66857142857143e-06,
|
| 413 |
+
"loss": 0.0068,
|
| 414 |
+
"step": 1320
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 13.4,
|
| 418 |
+
"learning_rate": 7.611428571428572e-06,
|
| 419 |
+
"loss": 0.0083,
|
| 420 |
+
"step": 1340
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 13.6,
|
| 424 |
+
"learning_rate": 7.5542857142857155e-06,
|
| 425 |
+
"loss": 0.0076,
|
| 426 |
+
"step": 1360
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 13.8,
|
| 430 |
+
"learning_rate": 7.497142857142857e-06,
|
| 431 |
+
"loss": 0.0119,
|
| 432 |
+
"step": 1380
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 14.0,
|
| 436 |
+
"learning_rate": 7.440000000000001e-06,
|
| 437 |
+
"loss": 0.0097,
|
| 438 |
+
"step": 1400
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 14.2,
|
| 442 |
+
"learning_rate": 7.382857142857144e-06,
|
| 443 |
+
"loss": 0.0075,
|
| 444 |
+
"step": 1420
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 14.4,
|
| 448 |
+
"learning_rate": 7.325714285714286e-06,
|
| 449 |
+
"loss": 0.013,
|
| 450 |
+
"step": 1440
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 14.6,
|
| 454 |
+
"learning_rate": 7.268571428571429e-06,
|
| 455 |
+
"loss": 0.0088,
|
| 456 |
+
"step": 1460
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 14.8,
|
| 460 |
+
"learning_rate": 7.211428571428573e-06,
|
| 461 |
+
"loss": 0.01,
|
| 462 |
+
"step": 1480
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 15.0,
|
| 466 |
+
"learning_rate": 7.154285714285715e-06,
|
| 467 |
+
"loss": 0.0116,
|
| 468 |
+
"step": 1500
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 15.2,
|
| 472 |
+
"learning_rate": 7.097142857142858e-06,
|
| 473 |
+
"loss": 0.0053,
|
| 474 |
+
"step": 1520
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 15.4,
|
| 478 |
+
"learning_rate": 7.04e-06,
|
| 479 |
+
"loss": 0.005,
|
| 480 |
+
"step": 1540
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 15.6,
|
| 484 |
+
"learning_rate": 6.982857142857143e-06,
|
| 485 |
+
"loss": 0.0045,
|
| 486 |
+
"step": 1560
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 15.8,
|
| 490 |
+
"learning_rate": 6.9257142857142864e-06,
|
| 491 |
+
"loss": 0.013,
|
| 492 |
+
"step": 1580
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 16.0,
|
| 496 |
+
"learning_rate": 6.868571428571429e-06,
|
| 497 |
+
"loss": 0.0093,
|
| 498 |
+
"step": 1600
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 16.2,
|
| 502 |
+
"learning_rate": 6.8114285714285725e-06,
|
| 503 |
+
"loss": 0.0053,
|
| 504 |
+
"step": 1620
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 16.4,
|
| 508 |
+
"learning_rate": 6.754285714285715e-06,
|
| 509 |
+
"loss": 0.009,
|
| 510 |
+
"step": 1640
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 16.6,
|
| 514 |
+
"learning_rate": 6.697142857142858e-06,
|
| 515 |
+
"loss": 0.0094,
|
| 516 |
+
"step": 1660
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 16.8,
|
| 520 |
+
"learning_rate": 6.640000000000001e-06,
|
| 521 |
+
"loss": 0.003,
|
| 522 |
+
"step": 1680
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 17.0,
|
| 526 |
+
"learning_rate": 6.582857142857143e-06,
|
| 527 |
+
"loss": 0.0037,
|
| 528 |
+
"step": 1700
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 17.2,
|
| 532 |
+
"learning_rate": 6.525714285714286e-06,
|
| 533 |
+
"loss": 0.002,
|
| 534 |
+
"step": 1720
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 17.4,
|
| 538 |
+
"learning_rate": 6.46857142857143e-06,
|
| 539 |
+
"loss": 0.0084,
|
| 540 |
+
"step": 1740
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 17.6,
|
| 544 |
+
"learning_rate": 6.411428571428572e-06,
|
| 545 |
+
"loss": 0.0032,
|
| 546 |
+
"step": 1760
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 17.8,
|
| 550 |
+
"learning_rate": 6.354285714285715e-06,
|
| 551 |
+
"loss": 0.0032,
|
| 552 |
+
"step": 1780
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 18.0,
|
| 556 |
+
"learning_rate": 6.297142857142857e-06,
|
| 557 |
+
"loss": 0.0031,
|
| 558 |
+
"step": 1800
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 18.2,
|
| 562 |
+
"learning_rate": 6.24e-06,
|
| 563 |
+
"loss": 0.0027,
|
| 564 |
+
"step": 1820
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 18.4,
|
| 568 |
+
"learning_rate": 6.1828571428571434e-06,
|
| 569 |
+
"loss": 0.0006,
|
| 570 |
+
"step": 1840
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 18.6,
|
| 574 |
+
"learning_rate": 6.125714285714286e-06,
|
| 575 |
+
"loss": 0.0009,
|
| 576 |
+
"step": 1860
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 18.8,
|
| 580 |
+
"learning_rate": 6.0685714285714295e-06,
|
| 581 |
+
"loss": 0.0058,
|
| 582 |
+
"step": 1880
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 19.0,
|
| 586 |
+
"learning_rate": 6.011428571428572e-06,
|
| 587 |
+
"loss": 0.002,
|
| 588 |
+
"step": 1900
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 19.2,
|
| 592 |
+
"learning_rate": 5.954285714285715e-06,
|
| 593 |
+
"loss": 0.0009,
|
| 594 |
+
"step": 1920
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 19.4,
|
| 598 |
+
"learning_rate": 5.897142857142858e-06,
|
| 599 |
+
"loss": 0.0046,
|
| 600 |
+
"step": 1940
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 19.6,
|
| 604 |
+
"learning_rate": 5.84e-06,
|
| 605 |
+
"loss": 0.0008,
|
| 606 |
+
"step": 1960
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 19.8,
|
| 610 |
+
"learning_rate": 5.782857142857143e-06,
|
| 611 |
+
"loss": 0.0026,
|
| 612 |
+
"step": 1980
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 20.0,
|
| 616 |
+
"learning_rate": 5.725714285714287e-06,
|
| 617 |
+
"loss": 0.0039,
|
| 618 |
+
"step": 2000
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 20.0,
|
| 622 |
+
"eval_loss": 0.654371976852417,
|
| 623 |
+
"eval_runtime": 329.4667,
|
| 624 |
+
"eval_samples_per_second": 0.304,
|
| 625 |
+
"eval_steps_per_second": 0.304,
|
| 626 |
+
"eval_wer": 66.9683257918552,
|
| 627 |
+
"step": 2000
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 20.2,
|
| 631 |
+
"learning_rate": 5.668571428571429e-06,
|
| 632 |
+
"loss": 0.0011,
|
| 633 |
+
"step": 2020
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 20.4,
|
| 637 |
+
"learning_rate": 5.611428571428572e-06,
|
| 638 |
+
"loss": 0.0023,
|
| 639 |
+
"step": 2040
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 20.6,
|
| 643 |
+
"learning_rate": 5.554285714285714e-06,
|
| 644 |
+
"loss": 0.0012,
|
| 645 |
+
"step": 2060
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 20.8,
|
| 649 |
+
"learning_rate": 5.497142857142857e-06,
|
| 650 |
+
"loss": 0.0003,
|
| 651 |
+
"step": 2080
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 21.0,
|
| 655 |
+
"learning_rate": 5.4400000000000004e-06,
|
| 656 |
+
"loss": 0.0006,
|
| 657 |
+
"step": 2100
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 21.2,
|
| 661 |
+
"learning_rate": 5.382857142857143e-06,
|
| 662 |
+
"loss": 0.0013,
|
| 663 |
+
"step": 2120
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 21.4,
|
| 667 |
+
"learning_rate": 5.3257142857142865e-06,
|
| 668 |
+
"loss": 0.0009,
|
| 669 |
+
"step": 2140
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 21.6,
|
| 673 |
+
"learning_rate": 5.268571428571429e-06,
|
| 674 |
+
"loss": 0.0023,
|
| 675 |
+
"step": 2160
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 21.8,
|
| 679 |
+
"learning_rate": 5.211428571428572e-06,
|
| 680 |
+
"loss": 0.0071,
|
| 681 |
+
"step": 2180
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 22.0,
|
| 685 |
+
"learning_rate": 5.154285714285715e-06,
|
| 686 |
+
"loss": 0.0034,
|
| 687 |
+
"step": 2200
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 22.2,
|
| 691 |
+
"learning_rate": 5.097142857142857e-06,
|
| 692 |
+
"loss": 0.0064,
|
| 693 |
+
"step": 2220
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 22.4,
|
| 697 |
+
"learning_rate": 5.04e-06,
|
| 698 |
+
"loss": 0.0033,
|
| 699 |
+
"step": 2240
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 22.6,
|
| 703 |
+
"learning_rate": 4.982857142857143e-06,
|
| 704 |
+
"loss": 0.0011,
|
| 705 |
+
"step": 2260
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 22.8,
|
| 709 |
+
"learning_rate": 4.925714285714286e-06,
|
| 710 |
+
"loss": 0.0001,
|
| 711 |
+
"step": 2280
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 23.0,
|
| 715 |
+
"learning_rate": 4.868571428571429e-06,
|
| 716 |
+
"loss": 0.0001,
|
| 717 |
+
"step": 2300
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 23.2,
|
| 721 |
+
"learning_rate": 4.811428571428572e-06,
|
| 722 |
+
"loss": 0.0001,
|
| 723 |
+
"step": 2320
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 23.4,
|
| 727 |
+
"learning_rate": 4.754285714285714e-06,
|
| 728 |
+
"loss": 0.0001,
|
| 729 |
+
"step": 2340
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 23.6,
|
| 733 |
+
"learning_rate": 4.6971428571428574e-06,
|
| 734 |
+
"loss": 0.0001,
|
| 735 |
+
"step": 2360
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 23.8,
|
| 739 |
+
"learning_rate": 4.6400000000000005e-06,
|
| 740 |
+
"loss": 0.0009,
|
| 741 |
+
"step": 2380
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 24.0,
|
| 745 |
+
"learning_rate": 4.5828571428571435e-06,
|
| 746 |
+
"loss": 0.0003,
|
| 747 |
+
"step": 2400
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 24.2,
|
| 751 |
+
"learning_rate": 4.525714285714286e-06,
|
| 752 |
+
"loss": 0.0008,
|
| 753 |
+
"step": 2420
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 24.4,
|
| 757 |
+
"learning_rate": 4.468571428571429e-06,
|
| 758 |
+
"loss": 0.002,
|
| 759 |
+
"step": 2440
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 24.6,
|
| 763 |
+
"learning_rate": 4.411428571428572e-06,
|
| 764 |
+
"loss": 0.0001,
|
| 765 |
+
"step": 2460
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 24.8,
|
| 769 |
+
"learning_rate": 4.354285714285715e-06,
|
| 770 |
+
"loss": 0.0001,
|
| 771 |
+
"step": 2480
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 25.0,
|
| 775 |
+
"learning_rate": 4.297142857142858e-06,
|
| 776 |
+
"loss": 0.0001,
|
| 777 |
+
"step": 2500
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 25.2,
|
| 781 |
+
"learning_rate": 4.24e-06,
|
| 782 |
+
"loss": 0.0,
|
| 783 |
+
"step": 2520
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 25.4,
|
| 787 |
+
"learning_rate": 4.182857142857143e-06,
|
| 788 |
+
"loss": 0.0001,
|
| 789 |
+
"step": 2540
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 25.6,
|
| 793 |
+
"learning_rate": 4.125714285714286e-06,
|
| 794 |
+
"loss": 0.0,
|
| 795 |
+
"step": 2560
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 25.8,
|
| 799 |
+
"learning_rate": 4.068571428571429e-06,
|
| 800 |
+
"loss": 0.0002,
|
| 801 |
+
"step": 2580
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 26.0,
|
| 805 |
+
"learning_rate": 4.011428571428571e-06,
|
| 806 |
+
"loss": 0.0003,
|
| 807 |
+
"step": 2600
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 26.2,
|
| 811 |
+
"learning_rate": 3.954285714285714e-06,
|
| 812 |
+
"loss": 0.0001,
|
| 813 |
+
"step": 2620
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 26.4,
|
| 817 |
+
"learning_rate": 3.8971428571428575e-06,
|
| 818 |
+
"loss": 0.0001,
|
| 819 |
+
"step": 2640
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 26.6,
|
| 823 |
+
"learning_rate": 3.8400000000000005e-06,
|
| 824 |
+
"loss": 0.0001,
|
| 825 |
+
"step": 2660
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 26.8,
|
| 829 |
+
"learning_rate": 3.782857142857143e-06,
|
| 830 |
+
"loss": 0.0,
|
| 831 |
+
"step": 2680
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 27.0,
|
| 835 |
+
"learning_rate": 3.7257142857142857e-06,
|
| 836 |
+
"loss": 0.0,
|
| 837 |
+
"step": 2700
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 27.2,
|
| 841 |
+
"learning_rate": 3.668571428571429e-06,
|
| 842 |
+
"loss": 0.0,
|
| 843 |
+
"step": 2720
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 27.4,
|
| 847 |
+
"learning_rate": 3.611428571428572e-06,
|
| 848 |
+
"loss": 0.0,
|
| 849 |
+
"step": 2740
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 27.6,
|
| 853 |
+
"learning_rate": 3.5542857142857144e-06,
|
| 854 |
+
"loss": 0.0,
|
| 855 |
+
"step": 2760
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 27.8,
|
| 859 |
+
"learning_rate": 3.4971428571428575e-06,
|
| 860 |
+
"loss": 0.0,
|
| 861 |
+
"step": 2780
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 28.0,
|
| 865 |
+
"learning_rate": 3.44e-06,
|
| 866 |
+
"loss": 0.0,
|
| 867 |
+
"step": 2800
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 28.2,
|
| 871 |
+
"learning_rate": 3.382857142857143e-06,
|
| 872 |
+
"loss": 0.0,
|
| 873 |
+
"step": 2820
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 28.4,
|
| 877 |
+
"learning_rate": 3.325714285714286e-06,
|
| 878 |
+
"loss": 0.0,
|
| 879 |
+
"step": 2840
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 28.6,
|
| 883 |
+
"learning_rate": 3.268571428571429e-06,
|
| 884 |
+
"loss": 0.0,
|
| 885 |
+
"step": 2860
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 28.8,
|
| 889 |
+
"learning_rate": 3.2114285714285714e-06,
|
| 890 |
+
"loss": 0.0,
|
| 891 |
+
"step": 2880
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 29.0,
|
| 895 |
+
"learning_rate": 3.154285714285715e-06,
|
| 896 |
+
"loss": 0.0,
|
| 897 |
+
"step": 2900
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 29.2,
|
| 901 |
+
"learning_rate": 3.0971428571428575e-06,
|
| 902 |
+
"loss": 0.0,
|
| 903 |
+
"step": 2920
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 29.4,
|
| 907 |
+
"learning_rate": 3.04e-06,
|
| 908 |
+
"loss": 0.0,
|
| 909 |
+
"step": 2940
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 29.6,
|
| 913 |
+
"learning_rate": 2.9828571428571427e-06,
|
| 914 |
+
"loss": 0.0,
|
| 915 |
+
"step": 2960
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 29.8,
|
| 919 |
+
"learning_rate": 2.925714285714286e-06,
|
| 920 |
+
"loss": 0.0,
|
| 921 |
+
"step": 2980
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 30.0,
|
| 925 |
+
"learning_rate": 2.868571428571429e-06,
|
| 926 |
+
"loss": 0.0,
|
| 927 |
+
"step": 3000
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 30.0,
|
| 931 |
+
"eval_loss": 0.6721540093421936,
|
| 932 |
+
"eval_runtime": 332.3176,
|
| 933 |
+
"eval_samples_per_second": 0.301,
|
| 934 |
+
"eval_steps_per_second": 0.301,
|
| 935 |
+
"eval_wer": 66.51583710407239,
|
| 936 |
+
"step": 3000
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 30.2,
|
| 940 |
+
"learning_rate": 2.8114285714285714e-06,
|
| 941 |
+
"loss": 0.0,
|
| 942 |
+
"step": 3020
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 30.4,
|
| 946 |
+
"learning_rate": 2.7542857142857145e-06,
|
| 947 |
+
"loss": 0.0,
|
| 948 |
+
"step": 3040
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 30.6,
|
| 952 |
+
"learning_rate": 2.6971428571428575e-06,
|
| 953 |
+
"loss": 0.0,
|
| 954 |
+
"step": 3060
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 30.8,
|
| 958 |
+
"learning_rate": 2.64e-06,
|
| 959 |
+
"loss": 0.0,
|
| 960 |
+
"step": 3080
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 31.0,
|
| 964 |
+
"learning_rate": 2.582857142857143e-06,
|
| 965 |
+
"loss": 0.0,
|
| 966 |
+
"step": 3100
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 31.2,
|
| 970 |
+
"learning_rate": 2.525714285714286e-06,
|
| 971 |
+
"loss": 0.0,
|
| 972 |
+
"step": 3120
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 31.4,
|
| 976 |
+
"learning_rate": 2.468571428571429e-06,
|
| 977 |
+
"loss": 0.0,
|
| 978 |
+
"step": 3140
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 31.6,
|
| 982 |
+
"learning_rate": 2.4114285714285715e-06,
|
| 983 |
+
"loss": 0.0,
|
| 984 |
+
"step": 3160
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 31.8,
|
| 988 |
+
"learning_rate": 2.3542857142857145e-06,
|
| 989 |
+
"loss": 0.0,
|
| 990 |
+
"step": 3180
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 32.0,
|
| 994 |
+
"learning_rate": 2.297142857142857e-06,
|
| 995 |
+
"loss": 0.0,
|
| 996 |
+
"step": 3200
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 32.2,
|
| 1000 |
+
"learning_rate": 2.24e-06,
|
| 1001 |
+
"loss": 0.0,
|
| 1002 |
+
"step": 3220
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 32.4,
|
| 1006 |
+
"learning_rate": 2.1828571428571428e-06,
|
| 1007 |
+
"loss": 0.0,
|
| 1008 |
+
"step": 3240
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 32.6,
|
| 1012 |
+
"learning_rate": 2.125714285714286e-06,
|
| 1013 |
+
"loss": 0.0,
|
| 1014 |
+
"step": 3260
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 32.8,
|
| 1018 |
+
"learning_rate": 2.068571428571429e-06,
|
| 1019 |
+
"loss": 0.0,
|
| 1020 |
+
"step": 3280
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 33.0,
|
| 1024 |
+
"learning_rate": 2.0114285714285715e-06,
|
| 1025 |
+
"loss": 0.0,
|
| 1026 |
+
"step": 3300
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 33.2,
|
| 1030 |
+
"learning_rate": 1.9542857142857145e-06,
|
| 1031 |
+
"loss": 0.0,
|
| 1032 |
+
"step": 3320
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 33.4,
|
| 1036 |
+
"learning_rate": 1.8971428571428573e-06,
|
| 1037 |
+
"loss": 0.0,
|
| 1038 |
+
"step": 3340
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 33.6,
|
| 1042 |
+
"learning_rate": 1.8400000000000002e-06,
|
| 1043 |
+
"loss": 0.0,
|
| 1044 |
+
"step": 3360
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 33.8,
|
| 1048 |
+
"learning_rate": 1.7828571428571432e-06,
|
| 1049 |
+
"loss": 0.0,
|
| 1050 |
+
"step": 3380
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 34.0,
|
| 1054 |
+
"learning_rate": 1.7257142857142858e-06,
|
| 1055 |
+
"loss": 0.0,
|
| 1056 |
+
"step": 3400
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 34.2,
|
| 1060 |
+
"learning_rate": 1.6685714285714289e-06,
|
| 1061 |
+
"loss": 0.0,
|
| 1062 |
+
"step": 3420
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 34.4,
|
| 1066 |
+
"learning_rate": 1.6114285714285715e-06,
|
| 1067 |
+
"loss": 0.0,
|
| 1068 |
+
"step": 3440
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 34.6,
|
| 1072 |
+
"learning_rate": 1.5542857142857143e-06,
|
| 1073 |
+
"loss": 0.0,
|
| 1074 |
+
"step": 3460
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 34.8,
|
| 1078 |
+
"learning_rate": 1.4971428571428574e-06,
|
| 1079 |
+
"loss": 0.0,
|
| 1080 |
+
"step": 3480
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 35.0,
|
| 1084 |
+
"learning_rate": 1.44e-06,
|
| 1085 |
+
"loss": 0.0,
|
| 1086 |
+
"step": 3500
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 35.2,
|
| 1090 |
+
"learning_rate": 1.382857142857143e-06,
|
| 1091 |
+
"loss": 0.0,
|
| 1092 |
+
"step": 3520
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 35.4,
|
| 1096 |
+
"learning_rate": 1.3257142857142856e-06,
|
| 1097 |
+
"loss": 0.0,
|
| 1098 |
+
"step": 3540
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 35.6,
|
| 1102 |
+
"learning_rate": 1.2685714285714287e-06,
|
| 1103 |
+
"loss": 0.0,
|
| 1104 |
+
"step": 3560
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 35.8,
|
| 1108 |
+
"learning_rate": 1.2114285714285715e-06,
|
| 1109 |
+
"loss": 0.0,
|
| 1110 |
+
"step": 3580
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 36.0,
|
| 1114 |
+
"learning_rate": 1.1542857142857143e-06,
|
| 1115 |
+
"loss": 0.0,
|
| 1116 |
+
"step": 3600
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 36.2,
|
| 1120 |
+
"learning_rate": 1.0971428571428572e-06,
|
| 1121 |
+
"loss": 0.0,
|
| 1122 |
+
"step": 3620
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 36.4,
|
| 1126 |
+
"learning_rate": 1.04e-06,
|
| 1127 |
+
"loss": 0.0,
|
| 1128 |
+
"step": 3640
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 36.6,
|
| 1132 |
+
"learning_rate": 9.828571428571428e-07,
|
| 1133 |
+
"loss": 0.0,
|
| 1134 |
+
"step": 3660
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 36.8,
|
| 1138 |
+
"learning_rate": 9.257142857142858e-07,
|
| 1139 |
+
"loss": 0.0,
|
| 1140 |
+
"step": 3680
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 37.0,
|
| 1144 |
+
"learning_rate": 8.685714285714286e-07,
|
| 1145 |
+
"loss": 0.0,
|
| 1146 |
+
"step": 3700
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 37.2,
|
| 1150 |
+
"learning_rate": 8.114285714285715e-07,
|
| 1151 |
+
"loss": 0.0,
|
| 1152 |
+
"step": 3720
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 37.4,
|
| 1156 |
+
"learning_rate": 7.542857142857144e-07,
|
| 1157 |
+
"loss": 0.0,
|
| 1158 |
+
"step": 3740
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 37.6,
|
| 1162 |
+
"learning_rate": 6.971428571428572e-07,
|
| 1163 |
+
"loss": 0.0,
|
| 1164 |
+
"step": 3760
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 37.8,
|
| 1168 |
+
"learning_rate": 6.4e-07,
|
| 1169 |
+
"loss": 0.0,
|
| 1170 |
+
"step": 3780
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 38.0,
|
| 1174 |
+
"learning_rate": 5.82857142857143e-07,
|
| 1175 |
+
"loss": 0.0,
|
| 1176 |
+
"step": 3800
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 38.2,
|
| 1180 |
+
"learning_rate": 5.257142857142858e-07,
|
| 1181 |
+
"loss": 0.0,
|
| 1182 |
+
"step": 3820
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 38.4,
|
| 1186 |
+
"learning_rate": 4.6857142857142855e-07,
|
| 1187 |
+
"loss": 0.0,
|
| 1188 |
+
"step": 3840
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 38.6,
|
| 1192 |
+
"learning_rate": 4.1142857142857144e-07,
|
| 1193 |
+
"loss": 0.0,
|
| 1194 |
+
"step": 3860
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 38.8,
|
| 1198 |
+
"learning_rate": 3.542857142857143e-07,
|
| 1199 |
+
"loss": 0.0,
|
| 1200 |
+
"step": 3880
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 39.0,
|
| 1204 |
+
"learning_rate": 2.9714285714285715e-07,
|
| 1205 |
+
"loss": 0.0,
|
| 1206 |
+
"step": 3900
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 39.2,
|
| 1210 |
+
"learning_rate": 2.4000000000000003e-07,
|
| 1211 |
+
"loss": 0.0,
|
| 1212 |
+
"step": 3920
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 39.4,
|
| 1216 |
+
"learning_rate": 1.828571428571429e-07,
|
| 1217 |
+
"loss": 0.0,
|
| 1218 |
+
"step": 3940
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 39.6,
|
| 1222 |
+
"learning_rate": 1.2571428571428572e-07,
|
| 1223 |
+
"loss": 0.0,
|
| 1224 |
+
"step": 3960
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 39.8,
|
| 1228 |
+
"learning_rate": 6.857142857142857e-08,
|
| 1229 |
+
"loss": 0.0,
|
| 1230 |
+
"step": 3980
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 40.0,
|
| 1234 |
+
"learning_rate": 1.142857142857143e-08,
|
| 1235 |
+
"loss": 0.0,
|
| 1236 |
+
"step": 4000
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 40.0,
|
| 1240 |
+
"eval_loss": 0.6841861009597778,
|
| 1241 |
+
"eval_runtime": 339.379,
|
| 1242 |
+
"eval_samples_per_second": 0.295,
|
| 1243 |
+
"eval_steps_per_second": 0.295,
|
| 1244 |
+
"eval_wer": 66.17647058823529,
|
| 1245 |
+
"step": 4000
|
| 1246 |
+
}
|
| 1247 |
+
],
|
| 1248 |
+
"logging_steps": 20,
|
| 1249 |
+
"max_steps": 4000,
|
| 1250 |
+
"num_input_tokens_seen": 0,
|
| 1251 |
+
"num_train_epochs": 40,
|
| 1252 |
+
"save_steps": 1000,
|
| 1253 |
+
"total_flos": 2.30868320256e+18,
|
| 1254 |
+
"train_batch_size": 2,
|
| 1255 |
+
"trial_name": null,
|
| 1256 |
+
"trial_params": null
|
| 1257 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b4a50ea23b25ee823a834a1111d8d875f60b801fd40911e285e3c6210cb4b7b
|
| 3 |
+
size 4856
|