Upload 9 files
Browse files- README.md +70 -1
- all_results.json +12 -0
- config.json +246 -0
- eval_results.json +8 -0
- preprocessor_config.json +14 -0
- pytorch_model.bin +3 -0
- train_results.json +7 -0
- trainer_state.json +529 -0
- training_args.bin +3 -0
README.md
CHANGED
@@ -1,3 +1,72 @@
|
|
1 |
---
|
2 |
-
license:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- audio-classification
|
5 |
+
- generated_from_trainer
|
6 |
+
datasets:
|
7 |
+
- common_language
|
8 |
+
metrics:
|
9 |
+
- accuracy
|
10 |
+
model-index:
|
11 |
+
- name: out
|
12 |
+
results: []
|
13 |
---
|
14 |
+
|
15 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
16 |
+
should probably proofread and complete it, then remove this comment. -->
|
17 |
+
|
18 |
+
# out
|
19 |
+
|
20 |
+
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the common_language dataset.
|
21 |
+
It achieves the following results on the evaluation set:
|
22 |
+
- Loss: 0.6339
|
23 |
+
- Accuracy: 0.8898
|
24 |
+
|
25 |
+
## Model description
|
26 |
+
|
27 |
+
More information needed
|
28 |
+
|
29 |
+
## Intended uses & limitations
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Training and evaluation data
|
34 |
+
|
35 |
+
More information needed
|
36 |
+
|
37 |
+
## Training procedure
|
38 |
+
|
39 |
+
### Training hyperparameters
|
40 |
+
|
41 |
+
The following hyperparameters were used during training:
|
42 |
+
- learning_rate: 1e-05
|
43 |
+
- train_batch_size: 32
|
44 |
+
- eval_batch_size: 32
|
45 |
+
- seed: 0
|
46 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
+
- lr_scheduler_type: linear
|
48 |
+
- lr_scheduler_warmup_ratio: 0.1
|
49 |
+
- num_epochs: 10.0
|
50 |
+
|
51 |
+
### Training results
|
52 |
+
|
53 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
54 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
55 |
+
| 1.4103 | 1.0 | 694 | 1.1299 | 0.7471 |
|
56 |
+
| 0.5324 | 2.0 | 1388 | 0.6167 | 0.8495 |
|
57 |
+
| 0.3164 | 3.0 | 2082 | 0.5136 | 0.8679 |
|
58 |
+
| 0.1535 | 4.0 | 2776 | 0.5264 | 0.8758 |
|
59 |
+
| 0.1018 | 5.0 | 3470 | 0.5890 | 0.8767 |
|
60 |
+
| 0.0555 | 6.0 | 4164 | 0.6121 | 0.8772 |
|
61 |
+
| 0.031 | 7.0 | 4858 | 0.6624 | 0.8777 |
|
62 |
+
| 0.0264 | 8.0 | 5552 | 0.6172 | 0.8869 |
|
63 |
+
| 0.0079 | 9.0 | 6246 | 0.6225 | 0.8884 |
|
64 |
+
| 0.0058 | 10.0 | 6940 | 0.6339 | 0.8898 |
|
65 |
+
|
66 |
+
|
67 |
+
### Framework versions
|
68 |
+
|
69 |
+
- Transformers 4.30.2
|
70 |
+
- Pytorch 1.8.0+cu111
|
71 |
+
- Datasets 2.11.0
|
72 |
+
- Tokenizers 0.13.3
|
all_results.json
ADDED
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"eval_accuracy": 0.8897758152173914,
|
4 |
+
"eval_loss": 0.6338901519775391,
|
5 |
+
"eval_runtime": 233.4489,
|
6 |
+
"eval_samples_per_second": 25.222,
|
7 |
+
"eval_steps_per_second": 0.788,
|
8 |
+
"train_loss": 0.3924562724076362,
|
9 |
+
"train_runtime": 18647.2594,
|
10 |
+
"train_samples_per_second": 11.902,
|
11 |
+
"train_steps_per_second": 0.372
|
12 |
+
}
|
config.json
ADDED
@@ -0,0 +1,246 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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 |
+
"WhisperForAudioClassification"
|
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 |
+
"finetuning_task": "audio-classification",
|
29 |
+
"forced_decoder_ids": [
|
30 |
+
[
|
31 |
+
1,
|
32 |
+
50259
|
33 |
+
],
|
34 |
+
[
|
35 |
+
2,
|
36 |
+
50359
|
37 |
+
],
|
38 |
+
[
|
39 |
+
3,
|
40 |
+
50363
|
41 |
+
]
|
42 |
+
],
|
43 |
+
"id2label": {
|
44 |
+
"0": "Arabic",
|
45 |
+
"1": "Basque",
|
46 |
+
"10": "Dutch",
|
47 |
+
"11": "English",
|
48 |
+
"12": "Esperanto",
|
49 |
+
"13": "Estonian",
|
50 |
+
"14": "French",
|
51 |
+
"15": "Frisian",
|
52 |
+
"16": "Georgian",
|
53 |
+
"17": "German",
|
54 |
+
"18": "Greek",
|
55 |
+
"19": "Hakha_Chin",
|
56 |
+
"2": "Breton",
|
57 |
+
"20": "Indonesian",
|
58 |
+
"21": "Interlingua",
|
59 |
+
"22": "Italian",
|
60 |
+
"23": "Japanese",
|
61 |
+
"24": "Kabyle",
|
62 |
+
"25": "Kinyarwanda",
|
63 |
+
"26": "Kyrgyz",
|
64 |
+
"27": "Latvian",
|
65 |
+
"28": "Maltese",
|
66 |
+
"29": "Mangolian",
|
67 |
+
"3": "Catalan",
|
68 |
+
"30": "Persian",
|
69 |
+
"31": "Polish",
|
70 |
+
"32": "Portuguese",
|
71 |
+
"33": "Romanian",
|
72 |
+
"34": "Romansh_Sursilvan",
|
73 |
+
"35": "Russian",
|
74 |
+
"36": "Sakha",
|
75 |
+
"37": "Slovenian",
|
76 |
+
"38": "Spanish",
|
77 |
+
"39": "Swedish",
|
78 |
+
"4": "Chinese_China",
|
79 |
+
"40": "Tamil",
|
80 |
+
"41": "Tatar",
|
81 |
+
"42": "Turkish",
|
82 |
+
"43": "Ukranian",
|
83 |
+
"44": "Welsh",
|
84 |
+
"5": "Chinese_Hongkong",
|
85 |
+
"6": "Chinese_Taiwan",
|
86 |
+
"7": "Chuvash",
|
87 |
+
"8": "Czech",
|
88 |
+
"9": "Dhivehi"
|
89 |
+
},
|
90 |
+
"init_std": 0.02,
|
91 |
+
"is_encoder_decoder": true,
|
92 |
+
"label2id": {
|
93 |
+
"Arabic": "0",
|
94 |
+
"Basque": "1",
|
95 |
+
"Breton": "2",
|
96 |
+
"Catalan": "3",
|
97 |
+
"Chinese_China": "4",
|
98 |
+
"Chinese_Hongkong": "5",
|
99 |
+
"Chinese_Taiwan": "6",
|
100 |
+
"Chuvash": "7",
|
101 |
+
"Czech": "8",
|
102 |
+
"Dhivehi": "9",
|
103 |
+
"Dutch": "10",
|
104 |
+
"English": "11",
|
105 |
+
"Esperanto": "12",
|
106 |
+
"Estonian": "13",
|
107 |
+
"French": "14",
|
108 |
+
"Frisian": "15",
|
109 |
+
"Georgian": "16",
|
110 |
+
"German": "17",
|
111 |
+
"Greek": "18",
|
112 |
+
"Hakha_Chin": "19",
|
113 |
+
"Indonesian": "20",
|
114 |
+
"Interlingua": "21",
|
115 |
+
"Italian": "22",
|
116 |
+
"Japanese": "23",
|
117 |
+
"Kabyle": "24",
|
118 |
+
"Kinyarwanda": "25",
|
119 |
+
"Kyrgyz": "26",
|
120 |
+
"Latvian": "27",
|
121 |
+
"Maltese": "28",
|
122 |
+
"Mangolian": "29",
|
123 |
+
"Persian": "30",
|
124 |
+
"Polish": "31",
|
125 |
+
"Portuguese": "32",
|
126 |
+
"Romanian": "33",
|
127 |
+
"Romansh_Sursilvan": "34",
|
128 |
+
"Russian": "35",
|
129 |
+
"Sakha": "36",
|
130 |
+
"Slovenian": "37",
|
131 |
+
"Spanish": "38",
|
132 |
+
"Swedish": "39",
|
133 |
+
"Tamil": "40",
|
134 |
+
"Tatar": "41",
|
135 |
+
"Turkish": "42",
|
136 |
+
"Ukranian": "43",
|
137 |
+
"Welsh": "44"
|
138 |
+
},
|
139 |
+
"mask_feature_length": 10,
|
140 |
+
"mask_feature_min_masks": 0,
|
141 |
+
"mask_feature_prob": 0.0,
|
142 |
+
"mask_time_length": 10,
|
143 |
+
"mask_time_min_masks": 2,
|
144 |
+
"mask_time_prob": 0.05,
|
145 |
+
"max_length": 448,
|
146 |
+
"max_source_positions": 1500,
|
147 |
+
"max_target_positions": 448,
|
148 |
+
"model_type": "whisper",
|
149 |
+
"num_hidden_layers": 12,
|
150 |
+
"num_mel_bins": 80,
|
151 |
+
"pad_token_id": 50257,
|
152 |
+
"scale_embedding": false,
|
153 |
+
"suppress_tokens": [
|
154 |
+
1,
|
155 |
+
2,
|
156 |
+
7,
|
157 |
+
8,
|
158 |
+
9,
|
159 |
+
10,
|
160 |
+
14,
|
161 |
+
25,
|
162 |
+
26,
|
163 |
+
27,
|
164 |
+
28,
|
165 |
+
29,
|
166 |
+
31,
|
167 |
+
58,
|
168 |
+
59,
|
169 |
+
60,
|
170 |
+
61,
|
171 |
+
62,
|
172 |
+
63,
|
173 |
+
90,
|
174 |
+
91,
|
175 |
+
92,
|
176 |
+
93,
|
177 |
+
359,
|
178 |
+
503,
|
179 |
+
522,
|
180 |
+
542,
|
181 |
+
873,
|
182 |
+
893,
|
183 |
+
902,
|
184 |
+
918,
|
185 |
+
922,
|
186 |
+
931,
|
187 |
+
1350,
|
188 |
+
1853,
|
189 |
+
1982,
|
190 |
+
2460,
|
191 |
+
2627,
|
192 |
+
3246,
|
193 |
+
3253,
|
194 |
+
3268,
|
195 |
+
3536,
|
196 |
+
3846,
|
197 |
+
3961,
|
198 |
+
4183,
|
199 |
+
4667,
|
200 |
+
6585,
|
201 |
+
6647,
|
202 |
+
7273,
|
203 |
+
9061,
|
204 |
+
9383,
|
205 |
+
10428,
|
206 |
+
10929,
|
207 |
+
11938,
|
208 |
+
12033,
|
209 |
+
12331,
|
210 |
+
12562,
|
211 |
+
13793,
|
212 |
+
14157,
|
213 |
+
14635,
|
214 |
+
15265,
|
215 |
+
15618,
|
216 |
+
16553,
|
217 |
+
16604,
|
218 |
+
18362,
|
219 |
+
18956,
|
220 |
+
20075,
|
221 |
+
21675,
|
222 |
+
22520,
|
223 |
+
26130,
|
224 |
+
26161,
|
225 |
+
26435,
|
226 |
+
28279,
|
227 |
+
29464,
|
228 |
+
31650,
|
229 |
+
32302,
|
230 |
+
32470,
|
231 |
+
36865,
|
232 |
+
42863,
|
233 |
+
47425,
|
234 |
+
49870,
|
235 |
+
50254,
|
236 |
+
50258,
|
237 |
+
50360,
|
238 |
+
50361,
|
239 |
+
50362
|
240 |
+
],
|
241 |
+
"torch_dtype": "float32",
|
242 |
+
"transformers_version": "4.30.2",
|
243 |
+
"use_cache": true,
|
244 |
+
"use_weighted_layer_sum": false,
|
245 |
+
"vocab_size": 51865
|
246 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"eval_accuracy": 0.8897758152173914,
|
4 |
+
"eval_loss": 0.6338901519775391,
|
5 |
+
"eval_runtime": 233.4489,
|
6 |
+
"eval_samples_per_second": 25.222,
|
7 |
+
"eval_steps_per_second": 0.788
|
8 |
+
}
|
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 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:aac1acd3e553000ddee100565d65162516314472e48628a49a5b34de8c8cd0cd
|
3 |
+
size 353514605
|
train_results.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"train_loss": 0.3924562724076362,
|
4 |
+
"train_runtime": 18647.2594,
|
5 |
+
"train_samples_per_second": 11.902,
|
6 |
+
"train_steps_per_second": 0.372
|
7 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,529 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.8897758152173914,
|
3 |
+
"best_model_checkpoint": "out/checkpoint-6940",
|
4 |
+
"epoch": 10.0,
|
5 |
+
"global_step": 6940,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.14,
|
12 |
+
"learning_rate": 1.4409221902017292e-06,
|
13 |
+
"loss": 3.8183,
|
14 |
+
"step": 100
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.29,
|
18 |
+
"learning_rate": 2.8818443804034585e-06,
|
19 |
+
"loss": 3.7614,
|
20 |
+
"step": 200
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.43,
|
24 |
+
"learning_rate": 4.322766570605187e-06,
|
25 |
+
"loss": 3.2003,
|
26 |
+
"step": 300
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.58,
|
30 |
+
"learning_rate": 5.749279538904899e-06,
|
31 |
+
"loss": 2.3107,
|
32 |
+
"step": 400
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.72,
|
36 |
+
"learning_rate": 7.190201729106629e-06,
|
37 |
+
"loss": 1.7861,
|
38 |
+
"step": 500
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.86,
|
42 |
+
"learning_rate": 8.631123919308358e-06,
|
43 |
+
"loss": 1.4103,
|
44 |
+
"step": 600
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 1.0,
|
48 |
+
"eval_accuracy": 0.7471127717391305,
|
49 |
+
"eval_loss": 1.1298972368240356,
|
50 |
+
"eval_runtime": 237.0543,
|
51 |
+
"eval_samples_per_second": 24.838,
|
52 |
+
"eval_steps_per_second": 0.776,
|
53 |
+
"step": 694
|
54 |
+
},
|
55 |
+
{
|
56 |
+
"epoch": 1.01,
|
57 |
+
"learning_rate": 9.993595901376882e-06,
|
58 |
+
"loss": 1.1941,
|
59 |
+
"step": 700
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 1.15,
|
63 |
+
"learning_rate": 9.833493435798911e-06,
|
64 |
+
"loss": 0.9417,
|
65 |
+
"step": 800
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 1.3,
|
69 |
+
"learning_rate": 9.673390970220942e-06,
|
70 |
+
"loss": 0.8036,
|
71 |
+
"step": 900
|
72 |
+
},
|
73 |
+
{
|
74 |
+
"epoch": 1.44,
|
75 |
+
"learning_rate": 9.513288504642972e-06,
|
76 |
+
"loss": 0.7488,
|
77 |
+
"step": 1000
|
78 |
+
},
|
79 |
+
{
|
80 |
+
"epoch": 1.59,
|
81 |
+
"learning_rate": 9.353186039065003e-06,
|
82 |
+
"loss": 0.6777,
|
83 |
+
"step": 1100
|
84 |
+
},
|
85 |
+
{
|
86 |
+
"epoch": 1.73,
|
87 |
+
"learning_rate": 9.193083573487034e-06,
|
88 |
+
"loss": 0.6306,
|
89 |
+
"step": 1200
|
90 |
+
},
|
91 |
+
{
|
92 |
+
"epoch": 1.87,
|
93 |
+
"learning_rate": 9.032981107909063e-06,
|
94 |
+
"loss": 0.5324,
|
95 |
+
"step": 1300
|
96 |
+
},
|
97 |
+
{
|
98 |
+
"epoch": 2.0,
|
99 |
+
"eval_accuracy": 0.8495244565217391,
|
100 |
+
"eval_loss": 0.6167322397232056,
|
101 |
+
"eval_runtime": 234.135,
|
102 |
+
"eval_samples_per_second": 25.148,
|
103 |
+
"eval_steps_per_second": 0.786,
|
104 |
+
"step": 1388
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 2.02,
|
108 |
+
"learning_rate": 8.874479666986872e-06,
|
109 |
+
"loss": 0.5331,
|
110 |
+
"step": 1400
|
111 |
+
},
|
112 |
+
{
|
113 |
+
"epoch": 2.16,
|
114 |
+
"learning_rate": 8.714377201408901e-06,
|
115 |
+
"loss": 0.3651,
|
116 |
+
"step": 1500
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 2.31,
|
120 |
+
"learning_rate": 8.554274735830932e-06,
|
121 |
+
"loss": 0.3298,
|
122 |
+
"step": 1600
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 2.45,
|
126 |
+
"learning_rate": 8.394172270252963e-06,
|
127 |
+
"loss": 0.3177,
|
128 |
+
"step": 1700
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 2.59,
|
132 |
+
"learning_rate": 8.234069804674993e-06,
|
133 |
+
"loss": 0.3093,
|
134 |
+
"step": 1800
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 2.74,
|
138 |
+
"learning_rate": 8.073967339097024e-06,
|
139 |
+
"loss": 0.317,
|
140 |
+
"step": 1900
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"epoch": 2.88,
|
144 |
+
"learning_rate": 7.913864873519053e-06,
|
145 |
+
"loss": 0.3164,
|
146 |
+
"step": 2000
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 3.0,
|
150 |
+
"eval_accuracy": 0.8678668478260869,
|
151 |
+
"eval_loss": 0.5136280655860901,
|
152 |
+
"eval_runtime": 234.6734,
|
153 |
+
"eval_samples_per_second": 25.09,
|
154 |
+
"eval_steps_per_second": 0.784,
|
155 |
+
"step": 2082
|
156 |
+
},
|
157 |
+
{
|
158 |
+
"epoch": 3.03,
|
159 |
+
"learning_rate": 7.753762407941084e-06,
|
160 |
+
"loss": 0.2798,
|
161 |
+
"step": 2100
|
162 |
+
},
|
163 |
+
{
|
164 |
+
"epoch": 3.17,
|
165 |
+
"learning_rate": 7.593659942363113e-06,
|
166 |
+
"loss": 0.1962,
|
167 |
+
"step": 2200
|
168 |
+
},
|
169 |
+
{
|
170 |
+
"epoch": 3.31,
|
171 |
+
"learning_rate": 7.4335574767851426e-06,
|
172 |
+
"loss": 0.1898,
|
173 |
+
"step": 2300
|
174 |
+
},
|
175 |
+
{
|
176 |
+
"epoch": 3.46,
|
177 |
+
"learning_rate": 7.2734550112071735e-06,
|
178 |
+
"loss": 0.1615,
|
179 |
+
"step": 2400
|
180 |
+
},
|
181 |
+
{
|
182 |
+
"epoch": 3.6,
|
183 |
+
"learning_rate": 7.113352545629203e-06,
|
184 |
+
"loss": 0.1534,
|
185 |
+
"step": 2500
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 3.75,
|
189 |
+
"learning_rate": 6.953250080051234e-06,
|
190 |
+
"loss": 0.162,
|
191 |
+
"step": 2600
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 3.89,
|
195 |
+
"learning_rate": 6.793147614473263e-06,
|
196 |
+
"loss": 0.1535,
|
197 |
+
"step": 2700
|
198 |
+
},
|
199 |
+
{
|
200 |
+
"epoch": 4.0,
|
201 |
+
"eval_accuracy": 0.8758491847826086,
|
202 |
+
"eval_loss": 0.5263639092445374,
|
203 |
+
"eval_runtime": 233.6811,
|
204 |
+
"eval_samples_per_second": 25.197,
|
205 |
+
"eval_steps_per_second": 0.787,
|
206 |
+
"step": 2776
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 4.03,
|
210 |
+
"learning_rate": 6.633045148895294e-06,
|
211 |
+
"loss": 0.1458,
|
212 |
+
"step": 2800
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 4.18,
|
216 |
+
"learning_rate": 6.472942683317323e-06,
|
217 |
+
"loss": 0.1132,
|
218 |
+
"step": 2900
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 4.32,
|
222 |
+
"learning_rate": 6.3128402177393535e-06,
|
223 |
+
"loss": 0.1162,
|
224 |
+
"step": 3000
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"epoch": 4.47,
|
228 |
+
"learning_rate": 6.152737752161384e-06,
|
229 |
+
"loss": 0.0797,
|
230 |
+
"step": 3100
|
231 |
+
},
|
232 |
+
{
|
233 |
+
"epoch": 4.61,
|
234 |
+
"learning_rate": 5.992635286583414e-06,
|
235 |
+
"loss": 0.0832,
|
236 |
+
"step": 3200
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"epoch": 4.76,
|
240 |
+
"learning_rate": 5.832532821005445e-06,
|
241 |
+
"loss": 0.0759,
|
242 |
+
"step": 3300
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 4.9,
|
246 |
+
"learning_rate": 5.672430355427474e-06,
|
247 |
+
"loss": 0.1018,
|
248 |
+
"step": 3400
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 5.0,
|
252 |
+
"eval_accuracy": 0.8766983695652174,
|
253 |
+
"eval_loss": 0.5890083909034729,
|
254 |
+
"eval_runtime": 236.5762,
|
255 |
+
"eval_samples_per_second": 24.888,
|
256 |
+
"eval_steps_per_second": 0.778,
|
257 |
+
"step": 3470
|
258 |
+
},
|
259 |
+
{
|
260 |
+
"epoch": 5.04,
|
261 |
+
"learning_rate": 5.512327889849503e-06,
|
262 |
+
"loss": 0.0816,
|
263 |
+
"step": 3500
|
264 |
+
},
|
265 |
+
{
|
266 |
+
"epoch": 5.19,
|
267 |
+
"learning_rate": 5.352225424271534e-06,
|
268 |
+
"loss": 0.0557,
|
269 |
+
"step": 3600
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 5.33,
|
273 |
+
"learning_rate": 5.1921229586935636e-06,
|
274 |
+
"loss": 0.0544,
|
275 |
+
"step": 3700
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 5.48,
|
279 |
+
"learning_rate": 5.0320204931155946e-06,
|
280 |
+
"loss": 0.045,
|
281 |
+
"step": 3800
|
282 |
+
},
|
283 |
+
{
|
284 |
+
"epoch": 5.62,
|
285 |
+
"learning_rate": 4.871918027537625e-06,
|
286 |
+
"loss": 0.0541,
|
287 |
+
"step": 3900
|
288 |
+
},
|
289 |
+
{
|
290 |
+
"epoch": 5.76,
|
291 |
+
"learning_rate": 4.713416586615434e-06,
|
292 |
+
"loss": 0.0656,
|
293 |
+
"step": 4000
|
294 |
+
},
|
295 |
+
{
|
296 |
+
"epoch": 5.91,
|
297 |
+
"learning_rate": 4.553314121037464e-06,
|
298 |
+
"loss": 0.0555,
|
299 |
+
"step": 4100
|
300 |
+
},
|
301 |
+
{
|
302 |
+
"epoch": 6.0,
|
303 |
+
"eval_accuracy": 0.8772078804347826,
|
304 |
+
"eval_loss": 0.612079381942749,
|
305 |
+
"eval_runtime": 232.3119,
|
306 |
+
"eval_samples_per_second": 25.345,
|
307 |
+
"eval_steps_per_second": 0.792,
|
308 |
+
"step": 4164
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 6.05,
|
312 |
+
"learning_rate": 4.393211655459494e-06,
|
313 |
+
"loss": 0.0439,
|
314 |
+
"step": 4200
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"epoch": 6.2,
|
318 |
+
"learning_rate": 4.2331091898815245e-06,
|
319 |
+
"loss": 0.0234,
|
320 |
+
"step": 4300
|
321 |
+
},
|
322 |
+
{
|
323 |
+
"epoch": 6.34,
|
324 |
+
"learning_rate": 4.073006724303555e-06,
|
325 |
+
"loss": 0.035,
|
326 |
+
"step": 4400
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 6.48,
|
330 |
+
"learning_rate": 3.912904258725585e-06,
|
331 |
+
"loss": 0.0376,
|
332 |
+
"step": 4500
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 6.63,
|
336 |
+
"learning_rate": 3.7528017931476145e-06,
|
337 |
+
"loss": 0.0348,
|
338 |
+
"step": 4600
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 6.77,
|
342 |
+
"learning_rate": 3.5926993275696446e-06,
|
343 |
+
"loss": 0.0169,
|
344 |
+
"step": 4700
|
345 |
+
},
|
346 |
+
{
|
347 |
+
"epoch": 6.92,
|
348 |
+
"learning_rate": 3.432596861991675e-06,
|
349 |
+
"loss": 0.031,
|
350 |
+
"step": 4800
|
351 |
+
},
|
352 |
+
{
|
353 |
+
"epoch": 7.0,
|
354 |
+
"eval_accuracy": 0.8777173913043478,
|
355 |
+
"eval_loss": 0.6623581051826477,
|
356 |
+
"eval_runtime": 235.7029,
|
357 |
+
"eval_samples_per_second": 24.981,
|
358 |
+
"eval_steps_per_second": 0.781,
|
359 |
+
"step": 4858
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 7.06,
|
363 |
+
"learning_rate": 3.2724943964137053e-06,
|
364 |
+
"loss": 0.0305,
|
365 |
+
"step": 4900
|
366 |
+
},
|
367 |
+
{
|
368 |
+
"epoch": 7.2,
|
369 |
+
"learning_rate": 3.1123919308357354e-06,
|
370 |
+
"loss": 0.0205,
|
371 |
+
"step": 5000
|
372 |
+
},
|
373 |
+
{
|
374 |
+
"epoch": 7.35,
|
375 |
+
"learning_rate": 2.952289465257765e-06,
|
376 |
+
"loss": 0.0239,
|
377 |
+
"step": 5100
|
378 |
+
},
|
379 |
+
{
|
380 |
+
"epoch": 7.49,
|
381 |
+
"learning_rate": 2.7921869996797953e-06,
|
382 |
+
"loss": 0.0224,
|
383 |
+
"step": 5200
|
384 |
+
},
|
385 |
+
{
|
386 |
+
"epoch": 7.64,
|
387 |
+
"learning_rate": 2.6320845341018254e-06,
|
388 |
+
"loss": 0.0187,
|
389 |
+
"step": 5300
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 7.78,
|
393 |
+
"learning_rate": 2.4719820685238555e-06,
|
394 |
+
"loss": 0.0125,
|
395 |
+
"step": 5400
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 7.93,
|
399 |
+
"learning_rate": 2.3118796029458857e-06,
|
400 |
+
"loss": 0.0264,
|
401 |
+
"step": 5500
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 8.0,
|
405 |
+
"eval_accuracy": 0.8868885869565217,
|
406 |
+
"eval_loss": 0.6172327995300293,
|
407 |
+
"eval_runtime": 233.4791,
|
408 |
+
"eval_samples_per_second": 25.219,
|
409 |
+
"eval_steps_per_second": 0.788,
|
410 |
+
"step": 5552
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 8.07,
|
414 |
+
"learning_rate": 2.1517771373679154e-06,
|
415 |
+
"loss": 0.013,
|
416 |
+
"step": 5600
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 8.21,
|
420 |
+
"learning_rate": 1.9916746717899455e-06,
|
421 |
+
"loss": 0.0145,
|
422 |
+
"step": 5700
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 8.36,
|
426 |
+
"learning_rate": 1.8315722062119759e-06,
|
427 |
+
"loss": 0.0146,
|
428 |
+
"step": 5800
|
429 |
+
},
|
430 |
+
{
|
431 |
+
"epoch": 8.5,
|
432 |
+
"learning_rate": 1.6714697406340058e-06,
|
433 |
+
"loss": 0.0127,
|
434 |
+
"step": 5900
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 8.65,
|
438 |
+
"learning_rate": 1.511367275056036e-06,
|
439 |
+
"loss": 0.0112,
|
440 |
+
"step": 6000
|
441 |
+
},
|
442 |
+
{
|
443 |
+
"epoch": 8.79,
|
444 |
+
"learning_rate": 1.351264809478066e-06,
|
445 |
+
"loss": 0.0115,
|
446 |
+
"step": 6100
|
447 |
+
},
|
448 |
+
{
|
449 |
+
"epoch": 8.93,
|
450 |
+
"learning_rate": 1.1911623439000962e-06,
|
451 |
+
"loss": 0.0079,
|
452 |
+
"step": 6200
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"epoch": 9.0,
|
456 |
+
"eval_accuracy": 0.8884171195652174,
|
457 |
+
"eval_loss": 0.6224890947341919,
|
458 |
+
"eval_runtime": 234.5344,
|
459 |
+
"eval_samples_per_second": 25.105,
|
460 |
+
"eval_steps_per_second": 0.785,
|
461 |
+
"step": 6246
|
462 |
+
},
|
463 |
+
{
|
464 |
+
"epoch": 9.08,
|
465 |
+
"learning_rate": 1.0310598783221263e-06,
|
466 |
+
"loss": 0.0073,
|
467 |
+
"step": 6300
|
468 |
+
},
|
469 |
+
{
|
470 |
+
"epoch": 9.22,
|
471 |
+
"learning_rate": 8.709574127441562e-07,
|
472 |
+
"loss": 0.0084,
|
473 |
+
"step": 6400
|
474 |
+
},
|
475 |
+
{
|
476 |
+
"epoch": 9.37,
|
477 |
+
"learning_rate": 7.108549471661865e-07,
|
478 |
+
"loss": 0.0034,
|
479 |
+
"step": 6500
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 9.51,
|
483 |
+
"learning_rate": 5.507524815882165e-07,
|
484 |
+
"loss": 0.0058,
|
485 |
+
"step": 6600
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 9.65,
|
489 |
+
"learning_rate": 3.906500160102466e-07,
|
490 |
+
"loss": 0.0048,
|
491 |
+
"step": 6700
|
492 |
+
},
|
493 |
+
{
|
494 |
+
"epoch": 9.8,
|
495 |
+
"learning_rate": 2.3054755043227666e-07,
|
496 |
+
"loss": 0.0063,
|
497 |
+
"step": 6800
|
498 |
+
},
|
499 |
+
{
|
500 |
+
"epoch": 9.94,
|
501 |
+
"learning_rate": 7.044508485430676e-08,
|
502 |
+
"loss": 0.0058,
|
503 |
+
"step": 6900
|
504 |
+
},
|
505 |
+
{
|
506 |
+
"epoch": 10.0,
|
507 |
+
"eval_accuracy": 0.8897758152173914,
|
508 |
+
"eval_loss": 0.6338901519775391,
|
509 |
+
"eval_runtime": 234.8904,
|
510 |
+
"eval_samples_per_second": 25.067,
|
511 |
+
"eval_steps_per_second": 0.783,
|
512 |
+
"step": 6940
|
513 |
+
},
|
514 |
+
{
|
515 |
+
"epoch": 10.0,
|
516 |
+
"step": 6940,
|
517 |
+
"total_flos": 2.7871930756137603e+19,
|
518 |
+
"train_loss": 0.3924562724076362,
|
519 |
+
"train_runtime": 18647.2594,
|
520 |
+
"train_samples_per_second": 11.902,
|
521 |
+
"train_steps_per_second": 0.372
|
522 |
+
}
|
523 |
+
],
|
524 |
+
"max_steps": 6940,
|
525 |
+
"num_train_epochs": 10,
|
526 |
+
"total_flos": 2.7871930756137603e+19,
|
527 |
+
"trial_name": null,
|
528 |
+
"trial_params": null
|
529 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c72e4781587bade3156e9c5f3816e2fc2aa8f7e2db9c8ca61f95867ce197b660
|
3 |
+
size 3887
|