Hugging Face's logo

IrwinD
/
log_sage_reward_model

Text Classification
Transformers
Safetensors
distilbert
trl
reward-trainer
Generated from Trainer
Eval Results
Model card Files
xet
Community
IrwinD commited on
Commit
b65d966
·
verified ·
1 Parent(s): c387cfc

End of training

Browse files
Files changed (3) hide show
  1. README.md +64 -27
  2. model.safetensors +1 -1
  3. training_args.bin +1 -1
README.md CHANGED
@@ -34,7 +34,7 @@ should probably proofread and complete it, then remove this comment. -->
34
 
35
  This model is a fine-tuned version of [distilbert/distilbert-base-uncased](https://huggingface.co/distilbert/distilbert-base-uncased) on the hdfs_rlhf_log_summary_dataset dataset.
36
  It achieves the following results on the evaluation set:
37
- - Loss: 0.4696
38
  - Accuracy: 1.0
39
 
40
  ## Model description
@@ -62,37 +62,74 @@ The following hyperparameters were used during training:
62
  - total_train_batch_size: 64
63
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
64
  - lr_scheduler_type: linear
65
- - num_epochs: 40
66
 
67
  ### Training results
68
 
69
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
70
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
71
- | No log | 1.0 | 1 | 0.6955 | 0.4 |
72
- | No log | 2.0 | 3 | 0.6932 | 0.8 |
73
- | No log | 3.0 | 5 | 0.6909 | 0.8 |
74
- | No log | 4.0 | 6 | 0.6899 | 0.8 |
75
- | No log | 5.0 | 8 | 0.6880 | 0.8 |
76
- | 0.2853 | 6.0 | 10 | 0.6853 | 0.8 |
77
- | 0.2853 | 7.0 | 11 | 0.6831 | 0.8 |
78
- | 0.2853 | 8.0 | 13 | 0.6762 | 0.8 |
79
- | 0.2853 | 9.0 | 15 | 0.6632 | 1.0 |
80
- | 0.2853 | 10.0 | 16 | 0.6531 | 1.0 |
81
- | 0.2853 | 11.0 | 18 | 0.6293 | 1.0 |
82
- | 0.2623 | 12.0 | 20 | 0.5976 | 1.0 |
83
- | 0.2623 | 13.0 | 21 | 0.5817 | 1.0 |
84
- | 0.2623 | 14.0 | 23 | 0.5552 | 1.0 |
85
- | 0.2623 | 15.0 | 25 | 0.5370 | 1.0 |
86
- | 0.2623 | 16.0 | 26 | 0.5283 | 1.0 |
87
- | 0.2623 | 17.0 | 27 | 0.5198 | 1.0 |
88
- | 0.2623 | 18.0 | 29 | 0.5038 | 1.0 |
89
- | 0.233 | 19.0 | 31 | 0.4913 | 1.0 |
90
- | 0.233 | 20.0 | 32 | 0.4860 | 1.0 |
91
- | 0.233 | 21.0 | 34 | 0.4773 | 1.0 |
92
- | 0.233 | 22.0 | 36 | 0.4719 | 1.0 |
93
- | 0.233 | 23.0 | 37 | 0.4708 | 1.0 |
94
- | 0.233 | 24.0 | 39 | 0.4696 | 1.0 |
95
- | 0.1635 | 24.73 | 40 | 0.4696 | 1.0 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
96
 
97
 
98
  ### Framework versions
 
34
 
35
  This model is a fine-tuned version of [distilbert/distilbert-base-uncased](https://huggingface.co/distilbert/distilbert-base-uncased) on the hdfs_rlhf_log_summary_dataset dataset.
36
  It achieves the following results on the evaluation set:
37
+ - Loss: 0.4300
38
  - Accuracy: 1.0
39
 
40
  ## Model description
 
62
  - total_train_batch_size: 64
63
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
64
  - lr_scheduler_type: linear
65
+ - num_epochs: 100
66
 
67
  ### Training results
68
 
69
  | Training Loss | Epoch | Step | Validation Loss | Accuracy |
70
  |:-------------:|:-----:|:----:|:---------------:|:--------:|
71
+ | No log | 1.0 | 1 | 0.6905 | 1.0 |
72
+ | No log | 2.0 | 3 | 0.6854 | 1.0 |
73
+ | No log | 3.0 | 5 | 0.6802 | 1.0 |
74
+ | No log | 4.0 | 6 | 0.6772 | 1.0 |
75
+ | No log | 5.0 | 8 | 0.6692 | 1.0 |
76
+ | 0.2821 | 6.0 | 10 | 0.6568 | 1.0 |
77
+ | 0.2821 | 7.0 | 11 | 0.6470 | 1.0 |
78
+ | 0.2821 | 8.0 | 13 | 0.6235 | 1.0 |
79
+ | 0.2821 | 9.0 | 15 | 0.5924 | 1.0 |
80
+ | 0.2821 | 10.0 | 16 | 0.5745 | 1.0 |
81
+ | 0.2821 | 11.0 | 18 | 0.5410 | 1.0 |
82
+ | 0.2394 | 12.0 | 20 | 0.5131 | 1.0 |
83
+ | 0.2394 | 13.0 | 21 | 0.5048 | 1.0 |
84
+ | 0.2394 | 14.0 | 23 | 0.4948 | 1.0 |
85
+ | 0.2394 | 15.0 | 25 | 0.4873 | 1.0 |
86
+ | 0.2394 | 16.0 | 26 | 0.4790 | 1.0 |
87
+ | 0.2394 | 17.0 | 27 | 0.4710 | 1.0 |
88
+ | 0.2394 | 18.0 | 29 | 0.4541 | 1.0 |
89
+ | 0.17 | 19.0 | 31 | 0.4446 | 1.0 |
90
+ | 0.17 | 20.0 | 32 | 0.4420 | 1.0 |
91
+ | 0.17 | 21.0 | 34 | 0.4379 | 1.0 |
92
+ | 0.17 | 22.0 | 36 | 0.4341 | 1.0 |
93
+ | 0.17 | 23.0 | 37 | 0.4326 | 1.0 |
94
+ | 0.17 | 24.0 | 39 | 0.4303 | 1.0 |
95
+ | 0.0955 | 25.0 | 41 | 0.4285 | 1.0 |
96
+ | 0.0955 | 26.0 | 42 | 0.4278 | 1.0 |
97
+ | 0.0955 | 27.0 | 44 | 0.4267 | 1.0 |
98
+ | 0.0955 | 28.0 | 46 | 0.4260 | 1.0 |
99
+ | 0.0955 | 29.0 | 47 | 0.4260 | 1.0 |
100
+ | 0.0955 | 30.0 | 49 | 0.4269 | 1.0 |
101
+ | 0.089 | 31.0 | 51 | 0.4264 | 1.0 |
102
+ | 0.089 | 32.0 | 52 | 0.4260 | 1.0 |
103
+ | 0.089 | 33.0 | 53 | 0.4263 | 1.0 |
104
+ | 0.089 | 34.0 | 55 | 0.4248 | 1.0 |
105
+ | 0.089 | 35.0 | 57 | 0.4242 | 1.0 |
106
+ | 0.089 | 36.0 | 58 | 0.4239 | 1.0 |
107
+ | 0.0816 | 37.0 | 60 | 0.4238 | 1.0 |
108
+ | 0.0816 | 38.0 | 62 | 0.4234 | 1.0 |
109
+ | 0.0816 | 39.0 | 63 | 0.4234 | 1.0 |
110
+ | 0.0816 | 40.0 | 65 | 0.4245 | 1.0 |
111
+ | 0.0816 | 41.0 | 67 | 0.4274 | 1.0 |
112
+ | 0.0816 | 42.0 | 68 | 0.4276 | 1.0 |
113
+ | 0.0707 | 43.0 | 70 | 0.4265 | 1.0 |
114
+ | 0.0707 | 44.0 | 72 | 0.4238 | 1.0 |
115
+ | 0.0707 | 45.0 | 73 | 0.4231 | 1.0 |
116
+ | 0.0707 | 46.0 | 75 | 0.4224 | 1.0 |
117
+ | 0.0707 | 47.0 | 77 | 0.4224 | 1.0 |
118
+ | 0.0707 | 48.0 | 78 | 0.4225 | 1.0 |
119
+ | 0.0707 | 49.0 | 79 | 0.4228 | 1.0 |
120
+ | 0.0737 | 50.0 | 81 | 0.4232 | 1.0 |
121
+ | 0.0737 | 51.0 | 83 | 0.4235 | 1.0 |
122
+ | 0.0737 | 52.0 | 84 | 0.4238 | 1.0 |
123
+ | 0.0737 | 53.0 | 86 | 0.4253 | 1.0 |
124
+ | 0.0737 | 54.0 | 88 | 0.4278 | 1.0 |
125
+ | 0.0737 | 55.0 | 89 | 0.4285 | 1.0 |
126
+ | 0.0675 | 56.0 | 91 | 0.4293 | 1.0 |
127
+ | 0.0675 | 57.0 | 93 | 0.4292 | 1.0 |
128
+ | 0.0675 | 58.0 | 94 | 0.4294 | 1.0 |
129
+ | 0.0675 | 59.0 | 96 | 0.4299 | 1.0 |
130
+ | 0.0675 | 60.0 | 98 | 0.4299 | 1.0 |
131
+ | 0.0675 | 61.0 | 99 | 0.4299 | 1.0 |
132
+ | 0.0632 | 61.09 | 100 | 0.4300 | 1.0 |
133
 
134
 
135
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:db8b35e5702ea61852d635d9d3e1e14634850e9dbfe39ef97968e83a9de2d849
3
  size 267829484
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b3b100dc36c668801ba928048a0c92bf794d9c50bff464cdc84649634c7d265
3
  size 267829484
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:666eaf83950bf2d92962366f8e9e515efb8827b5b9872a09a864d6bdeb55895d
3
  size 4984
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f7687ebf79507638b6f9868c6e7cc2a23740b4f950ee0475453b816522577d6
3
  size 4984