Update README.md
Browse files
README.md
CHANGED
@@ -65,6 +65,19 @@ model = AutoModelForCausalLM.from_pretrained("supermy/couplet")
|
|
65 |
|
66 |
bpe分词:"vocab_size"=50000
|
67 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
68 |
[INFO|trainer.py:1608] 2022-11-30 12:51:36,357 >> ***** Running training *****
|
69 |
[INFO|trainer.py:1609] 2022-11-30 12:51:36,357 >> Num examples = 260926
|
70 |
[INFO|trainer.py:1610] 2022-11-30 12:51:36,357 >> Num Epochs = 81
|
@@ -106,4 +119,33 @@ bpe分词:"vocab_size"=50000
|
|
106 |
eval_steps_per_second = 3.883
|
107 |
perplexity = 26.6108
|
108 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
109 |
```
|
|
|
65 |
|
66 |
bpe分词:"vocab_size"=50000
|
67 |
```
|
68 |
+
[INFO|trainer.py:1608] 2022-12-07 02:32:58,307 >> ***** Running training *****
|
69 |
+
[INFO|trainer.py:1609] 2022-12-07 02:32:58,307 >> Num examples = 260926
|
70 |
+
[INFO|trainer.py:1610] 2022-12-07 02:32:58,307 >> Num Epochs = 160
|
71 |
+
[INFO|trainer.py:1611] 2022-12-07 02:32:58,307 >> Instantaneous batch size per device = 96
|
72 |
+
[INFO|trainer.py:1612] 2022-12-07 02:32:58,307 >> Total train batch size (w. parallel, distributed & accumulation) = 96
|
73 |
+
[INFO|trainer.py:1613] 2022-12-07 02:32:58,307 >> Gradient Accumulation steps = 1
|
74 |
+
[INFO|trainer.py:1614] 2022-12-07 02:32:58,307 >> Total optimization steps = 434880
|
75 |
+
[INFO|trainer.py:1616] 2022-12-07 02:32:58,308 >> Number of trainable parameters = 124439808
|
76 |
+
[INFO|trainer.py:1637] 2022-12-07 02:32:58,309 >> Continuing training from checkpoint, will skip to saved global_step
|
77 |
+
[INFO|trainer.py:1638] 2022-12-07 02:32:58,310 >> Continuing training from epoch 93
|
78 |
+
[INFO|trainer.py:1639] 2022-12-07 02:32:58,310 >> Continuing training from global step 253500
|
79 |
+
|
80 |
+
|
81 |
[INFO|trainer.py:1608] 2022-11-30 12:51:36,357 >> ***** Running training *****
|
82 |
[INFO|trainer.py:1609] 2022-11-30 12:51:36,357 >> Num examples = 260926
|
83 |
[INFO|trainer.py:1610] 2022-11-30 12:51:36,357 >> Num Epochs = 81
|
|
|
119 |
eval_steps_per_second = 3.883
|
120 |
perplexity = 26.6108
|
121 |
|
122 |
+
|
123 |
+
{'loss': 3.0967, 'learning_rate': 1.8027961736571009e-07, 'epoch': 159.49}
|
124 |
+
{'loss': 3.0922, 'learning_rate': 1.227924944812362e-07, 'epoch': 159.68}
|
125 |
+
{'loss': 3.0934, 'learning_rate': 6.530537159676233e-08, 'epoch': 159.86}
|
126 |
+
{'train_runtime': 120967.2394, 'train_samples_per_second': 345.12, 'train_steps_per_second': 3.595, 'train_loss': 1.3456422273861828, 'epoch': 160.0}
|
127 |
+
***** train metrics *****
|
128 |
+
epoch = 160.0
|
129 |
+
train_loss = 1.3456
|
130 |
+
train_runtime = 1 day, 9:36:07.23
|
131 |
+
train_samples = 260926
|
132 |
+
train_samples_per_second = 345.12
|
133 |
+
train_steps_per_second = 3.595
|
134 |
+
12/08/2022 12:09:08 - INFO - __main__ - *** Evaluate ***
|
135 |
+
[INFO|trainer.py:2929] 2022-12-08 12:09:08,522 >> ***** Running Evaluation *****
|
136 |
+
[INFO|trainer.py:2931] 2022-12-08 12:09:08,522 >> Num examples = 1350
|
137 |
+
[INFO|trainer.py:2934] 2022-12-08 12:09:08,522 >> Batch size = 96
|
138 |
+
100%|██████████| 15/15 [00:03<00:00, 4.16it/s]
|
139 |
+
[INFO|modelcard.py:449] 2022-12-08 12:09:13,448 >> Dropping the following result as it does not have all the necessary fields:
|
140 |
+
{'task': {'name': 'Causal Language Modeling', 'type': 'text-generation'}, 'metrics': [{'name': 'Accuracy', 'type': 'accuracy', 'value': 0.433615520282187}]}
|
141 |
+
***** eval metrics *****
|
142 |
+
epoch = 160.0
|
143 |
+
eval_accuracy = 0.4336
|
144 |
+
eval_loss = 3.3005
|
145 |
+
eval_runtime = 0:00:03.93
|
146 |
+
eval_samples = 1350
|
147 |
+
eval_samples_per_second = 343.164
|
148 |
+
eval_steps_per_second = 3.813
|
149 |
+
perplexity = 27.1257
|
150 |
+
|
151 |
```
|