yalhessi commited on
Commit
2282226
·
verified ·
1 Parent(s): cb8dfb3

End of training

Browse files
Files changed (2) hide show
  1. README.md +122 -0
  2. loss_plot.png +0 -0
README.md ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ license: other
4
+ base_model: deepseek-ai/deepseek-coder-1.3b-base
5
+ tags:
6
+ - generated_from_trainer
7
+ model-index:
8
+ - name: lemexp-task1-v2-template_small-deepseek-coder-1.3b-base-8lr-12epochs-normal-eos
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # lemexp-task1-v2-template_small-deepseek-coder-1.3b-base-8lr-12epochs-normal-eos
16
+
17
+ This model is a fine-tuned version of [deepseek-ai/deepseek-coder-1.3b-base](https://huggingface.co/deepseek-ai/deepseek-coder-1.3b-base) on an unknown dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 0.1544
20
+
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ ### Training hyperparameters
36
+
37
+ The following hyperparameters were used during training:
38
+ - learning_rate: 0.0008
39
+ - train_batch_size: 2
40
+ - eval_batch_size: 2
41
+ - seed: 42
42
+ - distributed_type: multi-GPU
43
+ - num_devices: 8
44
+ - total_train_batch_size: 16
45
+ - total_eval_batch_size: 16
46
+ - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
47
+ - lr_scheduler_type: linear
48
+ - num_epochs: 12
49
+ - mixed_precision_training: Native AMP
50
+
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Validation Loss |
54
+ |:-------------:|:-------:|:-----:|:---------------:|
55
+ | 0.3895 | 0.2001 | 720 | 0.2991 |
56
+ | 0.2897 | 0.4001 | 1440 | 0.2637 |
57
+ | 0.2536 | 0.6002 | 2160 | 0.2559 |
58
+ | 0.2453 | 0.8002 | 2880 | 0.2426 |
59
+ | 0.2335 | 1.0003 | 3600 | 0.2423 |
60
+ | 0.2214 | 1.2003 | 4320 | 0.2344 |
61
+ | 0.217 | 1.4004 | 5040 | 0.2214 |
62
+ | 0.2154 | 1.6004 | 5760 | 0.2222 |
63
+ | 0.2133 | 1.8005 | 6480 | 0.2249 |
64
+ | 0.2095 | 2.0006 | 7200 | 0.2111 |
65
+ | 0.1984 | 2.2006 | 7920 | 0.2192 |
66
+ | 0.1973 | 2.4007 | 8640 | 0.2109 |
67
+ | 0.1938 | 2.6007 | 9360 | 0.1993 |
68
+ | 0.1941 | 2.8008 | 10080 | 0.2004 |
69
+ | 0.1962 | 3.0008 | 10800 | 0.1977 |
70
+ | 0.1813 | 3.2009 | 11520 | 0.2014 |
71
+ | 0.1822 | 3.4009 | 12240 | 0.2025 |
72
+ | 0.1814 | 3.6010 | 12960 | 0.1900 |
73
+ | 0.1788 | 3.8011 | 13680 | 0.1898 |
74
+ | 0.1795 | 4.0011 | 14400 | 0.1856 |
75
+ | 0.1678 | 4.2012 | 15120 | 0.1905 |
76
+ | 0.1693 | 4.4012 | 15840 | 0.1903 |
77
+ | 0.1675 | 4.6013 | 16560 | 0.1858 |
78
+ | 0.1681 | 4.8013 | 17280 | 0.1844 |
79
+ | 0.1635 | 5.0014 | 18000 | 0.1824 |
80
+ | 0.1537 | 5.2014 | 18720 | 0.1821 |
81
+ | 0.155 | 5.4015 | 19440 | 0.1762 |
82
+ | 0.1559 | 5.6016 | 20160 | 0.1837 |
83
+ | 0.1565 | 5.8016 | 20880 | 0.1725 |
84
+ | 0.151 | 6.0017 | 21600 | 0.1710 |
85
+ | 0.1444 | 6.2017 | 22320 | 0.1706 |
86
+ | 0.1447 | 6.4018 | 23040 | 0.1719 |
87
+ | 0.1429 | 6.6018 | 23760 | 0.1716 |
88
+ | 0.1428 | 6.8019 | 24480 | 0.1673 |
89
+ | 0.1407 | 7.0019 | 25200 | 0.1669 |
90
+ | 0.1335 | 7.2020 | 25920 | 0.1650 |
91
+ | 0.131 | 7.4021 | 26640 | 0.1670 |
92
+ | 0.1314 | 7.6021 | 27360 | 0.1617 |
93
+ | 0.1293 | 7.8022 | 28080 | 0.1615 |
94
+ | 0.1309 | 8.0022 | 28800 | 0.1615 |
95
+ | 0.1192 | 8.2023 | 29520 | 0.1590 |
96
+ | 0.1186 | 8.4023 | 30240 | 0.1567 |
97
+ | 0.1191 | 8.6024 | 30960 | 0.1597 |
98
+ | 0.1181 | 8.8024 | 31680 | 0.1533 |
99
+ | 0.1181 | 9.0025 | 32400 | 0.1540 |
100
+ | 0.1068 | 9.2026 | 33120 | 0.1538 |
101
+ | 0.1066 | 9.4026 | 33840 | 0.1517 |
102
+ | 0.1071 | 9.6027 | 34560 | 0.1535 |
103
+ | 0.107 | 9.8027 | 35280 | 0.1484 |
104
+ | 0.1038 | 10.0028 | 36000 | 0.1516 |
105
+ | 0.0952 | 10.2028 | 36720 | 0.1565 |
106
+ | 0.094 | 10.4029 | 37440 | 0.1533 |
107
+ | 0.0943 | 10.6029 | 38160 | 0.1556 |
108
+ | 0.094 | 10.8030 | 38880 | 0.1504 |
109
+ | 0.0942 | 11.0031 | 39600 | 0.1492 |
110
+ | 0.0856 | 11.2031 | 40320 | 0.1560 |
111
+ | 0.085 | 11.4032 | 41040 | 0.1553 |
112
+ | 0.0832 | 11.6032 | 41760 | 0.1560 |
113
+ | 0.0833 | 11.8033 | 42480 | 0.1544 |
114
+
115
+
116
+ ### Framework versions
117
+
118
+ - PEFT 0.14.0
119
+ - Transformers 4.47.0
120
+ - Pytorch 2.5.1+cu124
121
+ - Datasets 3.2.0
122
+ - Tokenizers 0.21.0
loss_plot.png ADDED