Hugging Face's logo

IrwinD
/
log_sage_reward_model

Text Classification
Transformers
Safetensors
distilbert
trl
reward-trainer
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
IrwinD commited on
Commit
bc8c057
•
1 Parent(s): 547ecfd

End of training

Browse files
Files changed (4) hide show
  1. README.md +45 -50
  2. config.json +1 -0
  3. model.safetensors +1 -1
  4. training_args.bin +2 -2
README.md CHANGED
@@ -2,29 +2,12 @@
2
  license: apache-2.0
3
  base_model: distilbert/distilbert-base-uncased
4
  tags:
5
- - trl
6
- - reward-trainer
7
  - generated_from_trainer
8
  datasets:
9
  - hdfs_rlhf_log_summary_dataset
10
- metrics:
11
- - accuracy
12
  model-index:
13
  - name: log_sage_reward_model
14
- results:
15
- - task:
16
- name: Text Classification
17
- type: text-classification
18
- dataset:
19
- name: hdfs_rlhf_log_summary_dataset
20
- type: hdfs_rlhf_log_summary_dataset
21
- config: default
22
- split: None
23
- args: default
24
- metrics:
25
- - name: Accuracy
26
- type: accuracy
27
- value: 0.8
28
  ---
29
 
30
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -34,8 +17,7 @@ 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: 0.7472
38
- - Accuracy: 0.8
39
 
40
  ## Model description
41
 
@@ -56,43 +38,56 @@ More information needed
56
  The following hyperparameters were used during training:
57
  - learning_rate: 1.41e-05
58
  - train_batch_size: 4
59
- - eval_batch_size: 16
60
  - seed: 42
61
- - gradient_accumulation_steps: 16
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 | 0.6909 | 1.0 |
72
- | No log | 2.0 | 3 | 0.6899 | 0.8 |
73
- | No log | 3.0 | 5 | 0.6896 | 0.8 |
74
- | No log | 4.0 | 6 | 0.6889 | 0.8 |
75
- | No log | 5.0 | 8 | 0.6890 | 0.8 |
76
- | 0.2839 | 6.0 | 10 | 0.6912 | 0.8 |
77
- | 0.2839 | 7.0 | 11 | 0.6931 | 0.8 |
78
- | 0.2839 | 8.0 | 13 | 0.6982 | 0.8 |
79
- | 0.2839 | 9.0 | 15 | 0.7055 | 0.8 |
80
- | 0.2839 | 10.0 | 16 | 0.7098 | 0.8 |
81
- | 0.2839 | 11.0 | 18 | 0.7184 | 0.8 |
82
- | 0.259 | 12.0 | 20 | 0.7245 | 0.8 |
83
- | 0.259 | 13.0 | 21 | 0.7259 | 0.8 |
84
- | 0.259 | 14.0 | 23 | 0.7268 | 0.8 |
85
- | 0.259 | 15.0 | 25 | 0.7285 | 0.8 |
86
- | 0.259 | 16.0 | 26 | 0.7294 | 0.8 |
87
- | 0.259 | 17.0 | 27 | 0.7304 | 0.8 |
88
- | 0.259 | 18.0 | 29 | 0.7333 | 0.8 |
89
- | 0.2339 | 19.0 | 31 | 0.7356 | 0.8 |
90
- | 0.2339 | 20.0 | 32 | 0.7373 | 0.8 |
91
- | 0.2339 | 21.0 | 34 | 0.7414 | 0.8 |
92
- | 0.2339 | 22.0 | 36 | 0.7442 | 0.8 |
93
- | 0.2339 | 23.0 | 37 | 0.7454 | 0.8 |
94
- | 0.2339 | 24.0 | 39 | 0.7469 | 0.8 |
95
- | 0.1594 | 24.73 | 40 | 0.7472 | 0.8 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
96
 
97
 
98
  ### Framework versions
 
2
  license: apache-2.0
3
  base_model: distilbert/distilbert-base-uncased
4
  tags:
 
 
5
  - generated_from_trainer
6
  datasets:
7
  - hdfs_rlhf_log_summary_dataset
 
 
8
  model-index:
9
  - name: log_sage_reward_model
10
+ results: []
 
 
 
 
 
 
 
 
 
 
 
 
 
11
  ---
12
 
13
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
17
 
18
  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.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 0.0005
 
21
 
22
  ## Model description
23
 
 
38
  The following hyperparameters were used during training:
39
  - learning_rate: 1.41e-05
40
  - train_batch_size: 4
41
+ - eval_batch_size: 4
42
  - seed: 42
 
 
43
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
44
  - lr_scheduler_type: linear
45
  - num_epochs: 40
46
 
47
  ### Training results
48
 
49
+ | Training Loss | Epoch | Step | Validation Loss |
50
+ |:-------------:|:-----:|:----:|:---------------:|
51
+ | No log | 1.0 | 11 | 0.0022 |
52
+ | No log | 2.0 | 22 | 0.0049 |
53
+ | No log | 3.0 | 33 | 0.0006 |
54
+ | No log | 4.0 | 44 | 0.0006 |
55
+ | No log | 5.0 | 55 | 0.0008 |
56
+ | No log | 6.0 | 66 | 0.0003 |
57
+ | No log | 7.0 | 77 | 0.0005 |
58
+ | No log | 8.0 | 88 | 0.0010 |
59
+ | No log | 9.0 | 99 | 0.0008 |
60
+ | No log | 10.0 | 110 | 0.0007 |
61
+ | No log | 11.0 | 121 | 0.0007 |
62
+ | No log | 12.0 | 132 | 0.0006 |
63
+ | No log | 13.0 | 143 | 0.0006 |
64
+ | No log | 14.0 | 154 | 0.0004 |
65
+ | No log | 15.0 | 165 | 0.0007 |
66
+ | No log | 16.0 | 176 | 0.0007 |
67
+ | No log | 17.0 | 187 | 0.0006 |
68
+ | No log | 18.0 | 198 | 0.0004 |
69
+ | No log | 19.0 | 209 | 0.0005 |
70
+ | No log | 20.0 | 220 | 0.0006 |
71
+ | No log | 21.0 | 231 | 0.0006 |
72
+ | No log | 22.0 | 242 | 0.0006 |
73
+ | No log | 23.0 | 253 | 0.0009 |
74
+ | No log | 24.0 | 264 | 0.0006 |
75
+ | No log | 25.0 | 275 | 0.0007 |
76
+ | No log | 26.0 | 286 | 0.0005 |
77
+ | No log | 27.0 | 297 | 0.0005 |
78
+ | No log | 28.0 | 308 | 0.0004 |
79
+ | No log | 29.0 | 319 | 0.0004 |
80
+ | No log | 30.0 | 330 | 0.0005 |
81
+ | No log | 31.0 | 341 | 0.0005 |
82
+ | No log | 32.0 | 352 | 0.0005 |
83
+ | No log | 33.0 | 363 | 0.0005 |
84
+ | No log | 34.0 | 374 | 0.0004 |
85
+ | No log | 35.0 | 385 | 0.0004 |
86
+ | No log | 36.0 | 396 | 0.0005 |
87
+ | No log | 37.0 | 407 | 0.0005 |
88
+ | No log | 38.0 | 418 | 0.0005 |
89
+ | No log | 39.0 | 429 | 0.0005 |
90
+ | No log | 40.0 | 440 | 0.0005 |
91
 
92
 
93
  ### Framework versions
config.json CHANGED
@@ -20,6 +20,7 @@
20
  "n_heads": 12,
21
  "n_layers": 6,
22
  "pad_token_id": 0,
 
23
  "qa_dropout": 0.1,
24
  "seq_classif_dropout": 0.2,
25
  "sinusoidal_pos_embds": false,
 
20
  "n_heads": 12,
21
  "n_layers": 6,
22
  "pad_token_id": 0,
23
+ "problem_type": "regression",
24
  "qa_dropout": 0.1,
25
  "seq_classif_dropout": 0.2,
26
  "sinusoidal_pos_embds": false,
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fd16d102d3bd0317100fa11c3a30af9375da5b27c584dec86620177cbc4db177
3
  size 267829484
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16869d0953d5b87a61040dee2caef1db68a03ed0adf7482b276d86381884e93c
3
  size 267829484
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:eca786d42752b9e3500b1efcaf37454a613b921da12e7e7d7739edd9f92291dc
3
- size 4984
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a519663b9a6387514f11ccce00d19ac348e481362fef0e7f53e66f3b08db7db
3
+ size 4920