Hugging Face's logo

amirabdullah19852020
/
interpreting_reward_models

Safetensors
English
Model card Files Community
amirabdullah19852020 commited on
Commit
d07f1d2
•
1 Parent(s): fc1854f

Upload folder using huggingface_hub

Browse files
Files changed (9) hide show
  1. data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/chosen/data-00000-of-00002.arrow +3 -0
  2. data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/chosen/data-00001-of-00002.arrow +3 -0
  3. data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/chosen/dataset_info.json +133 -0
  4. data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/chosen/state.json +16 -0
  5. data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/dataset_dict.json +1 -0
  6. data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/new_rejected/data-00000-of-00002.arrow +3 -0
  7. data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/new_rejected/data-00001-of-00002.arrow +3 -0
  8. data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/new_rejected/dataset_info.json +133 -0
  9. data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/new_rejected/state.json +16 -0
data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/chosen/data-00000-of-00002.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4443a84d3ec602669a00a2fccf1ac3e5de49824ed9059daadd9302e65c53248a
3
+ size 360170544
data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/chosen/data-00001-of-00002.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ae384b369dc6247248830d2574352d3f0c2f25f1b530de49b864d103e7ca3fc0
3
+ size 360237736
data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/chosen/dataset_info.json ADDED
@@ -0,0 +1,133 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "builder_name": "json",
3
+ "citation": "",
4
+ "config_name": "default",
5
+ "dataset_name": "json",
6
+ "dataset_size": 720377750,
7
+ "description": "",
8
+ "download_checksums": {
9
+ "/data/home/amir/work/codes/Interpreting-Reward-Models/Notebooks/./pythia_70m_from_chosen_task_hh_rlhf_activations_dataset.jsonl": {
10
+ "num_bytes": 1431046981,
11
+ "checksum": null
12
+ }
13
+ },
14
+ "download_size": 1431046981,
15
+ "features": {
16
+ "texts": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "token_ids": {
21
+ "feature": {
22
+ "dtype": "int64",
23
+ "_type": "Value"
24
+ },
25
+ "_type": "Sequence"
26
+ },
27
+ "tokens": {
28
+ "feature": {
29
+ "dtype": "string",
30
+ "_type": "Value"
31
+ },
32
+ "_type": "Sequence"
33
+ },
34
+ "averaged_token_repr": {
35
+ "feature": {
36
+ "dtype": "float64",
37
+ "_type": "Value"
38
+ },
39
+ "_type": "Sequence"
40
+ },
41
+ "averaged_activations_1": {
42
+ "feature": {
43
+ "dtype": "float64",
44
+ "_type": "Value"
45
+ },
46
+ "_type": "Sequence"
47
+ },
48
+ "averaged_reprs_1": {
49
+ "feature": {
50
+ "dtype": "float64",
51
+ "_type": "Value"
52
+ },
53
+ "_type": "Sequence"
54
+ },
55
+ "averaged_activations_2": {
56
+ "feature": {
57
+ "dtype": "float64",
58
+ "_type": "Value"
59
+ },
60
+ "_type": "Sequence"
61
+ },
62
+ "averaged_reprs_2": {
63
+ "feature": {
64
+ "dtype": "float64",
65
+ "_type": "Value"
66
+ },
67
+ "_type": "Sequence"
68
+ },
69
+ "averaged_activations_3": {
70
+ "feature": {
71
+ "dtype": "float64",
72
+ "_type": "Value"
73
+ },
74
+ "_type": "Sequence"
75
+ },
76
+ "averaged_reprs_3": {
77
+ "feature": {
78
+ "dtype": "float64",
79
+ "_type": "Value"
80
+ },
81
+ "_type": "Sequence"
82
+ },
83
+ "averaged_activations_4": {
84
+ "feature": {
85
+ "dtype": "float64",
86
+ "_type": "Value"
87
+ },
88
+ "_type": "Sequence"
89
+ },
90
+ "averaged_reprs_4": {
91
+ "feature": {
92
+ "dtype": "float64",
93
+ "_type": "Value"
94
+ },
95
+ "_type": "Sequence"
96
+ },
97
+ "averaged_activations_5": {
98
+ "feature": {
99
+ "dtype": "float64",
100
+ "_type": "Value"
101
+ },
102
+ "_type": "Sequence"
103
+ },
104
+ "averaged_reprs_5": {
105
+ "feature": {
106
+ "dtype": "float64",
107
+ "_type": "Value"
108
+ },
109
+ "_type": "Sequence"
110
+ }
111
+ },
112
+ "homepage": "",
113
+ "license": "",
114
+ "size_in_bytes": 2151424731,
115
+ "splits": {
116
+ "train": {
117
+ "name": "train",
118
+ "num_bytes": 720377750,
119
+ "num_examples": 15000,
120
+ "shard_lengths": [
121
+ 10492,
122
+ 4508
123
+ ],
124
+ "dataset_name": "json"
125
+ }
126
+ },
127
+ "version": {
128
+ "version_str": "0.0.0",
129
+ "major": 0,
130
+ "minor": 0,
131
+ "patch": 0
132
+ }
133
+ }
data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/chosen/state.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00002.arrow"
5
+ },
6
+ {
7
+ "filename": "data-00001-of-00002.arrow"
8
+ }
9
+ ],
10
+ "_fingerprint": "1f230f60e092956c",
11
+ "_format_columns": null,
12
+ "_format_kwargs": {},
13
+ "_format_type": null,
14
+ "_output_all_columns": false,
15
+ "_split": "train"
16
+ }
data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/dataset_dict.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"splits": ["chosen", "new_rejected"]}
data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/new_rejected/data-00000-of-00002.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4041e6e96c6b264c7eb112f776aa8e46495feca66a544b0a3c960a29a00b365
3
+ size 341368000
data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/new_rejected/data-00001-of-00002.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:713a325ef16faae00ed8d750bce9426a60cd9a9fa14fca54f9aacdb65c81df29
3
+ size 341361632
data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/new_rejected/dataset_info.json ADDED
@@ -0,0 +1,133 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "builder_name": "json",
3
+ "citation": "",
4
+ "config_name": "default",
5
+ "dataset_name": "json",
6
+ "dataset_size": 682699101,
7
+ "description": "",
8
+ "download_checksums": {
9
+ "/data/home/amir/work/codes/Interpreting-Reward-Models/Notebooks/./pythia_70m_from_new_rejected_task_hh_rlhf_activations_dataset.jsonl": {
10
+ "num_bytes": 1301164959,
11
+ "checksum": null
12
+ }
13
+ },
14
+ "download_size": 1301164959,
15
+ "features": {
16
+ "texts": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "token_ids": {
21
+ "feature": {
22
+ "dtype": "int64",
23
+ "_type": "Value"
24
+ },
25
+ "_type": "Sequence"
26
+ },
27
+ "tokens": {
28
+ "feature": {
29
+ "dtype": "string",
30
+ "_type": "Value"
31
+ },
32
+ "_type": "Sequence"
33
+ },
34
+ "averaged_token_repr": {
35
+ "feature": {
36
+ "dtype": "float64",
37
+ "_type": "Value"
38
+ },
39
+ "_type": "Sequence"
40
+ },
41
+ "averaged_activations_1": {
42
+ "feature": {
43
+ "dtype": "float64",
44
+ "_type": "Value"
45
+ },
46
+ "_type": "Sequence"
47
+ },
48
+ "averaged_reprs_1": {
49
+ "feature": {
50
+ "dtype": "float64",
51
+ "_type": "Value"
52
+ },
53
+ "_type": "Sequence"
54
+ },
55
+ "averaged_activations_2": {
56
+ "feature": {
57
+ "dtype": "float64",
58
+ "_type": "Value"
59
+ },
60
+ "_type": "Sequence"
61
+ },
62
+ "averaged_reprs_2": {
63
+ "feature": {
64
+ "dtype": "float64",
65
+ "_type": "Value"
66
+ },
67
+ "_type": "Sequence"
68
+ },
69
+ "averaged_activations_3": {
70
+ "feature": {
71
+ "dtype": "float64",
72
+ "_type": "Value"
73
+ },
74
+ "_type": "Sequence"
75
+ },
76
+ "averaged_reprs_3": {
77
+ "feature": {
78
+ "dtype": "float64",
79
+ "_type": "Value"
80
+ },
81
+ "_type": "Sequence"
82
+ },
83
+ "averaged_activations_4": {
84
+ "feature": {
85
+ "dtype": "float64",
86
+ "_type": "Value"
87
+ },
88
+ "_type": "Sequence"
89
+ },
90
+ "averaged_reprs_4": {
91
+ "feature": {
92
+ "dtype": "float64",
93
+ "_type": "Value"
94
+ },
95
+ "_type": "Sequence"
96
+ },
97
+ "averaged_activations_5": {
98
+ "feature": {
99
+ "dtype": "float64",
100
+ "_type": "Value"
101
+ },
102
+ "_type": "Sequence"
103
+ },
104
+ "averaged_reprs_5": {
105
+ "feature": {
106
+ "dtype": "float64",
107
+ "_type": "Value"
108
+ },
109
+ "_type": "Sequence"
110
+ }
111
+ },
112
+ "homepage": "",
113
+ "license": "",
114
+ "size_in_bytes": 1983864060,
115
+ "splits": {
116
+ "train": {
117
+ "name": "train",
118
+ "num_bytes": 682699101,
119
+ "num_examples": 15000,
120
+ "shard_lengths": [
121
+ 11040,
122
+ 3960
123
+ ],
124
+ "dataset_name": "json"
125
+ }
126
+ },
127
+ "version": {
128
+ "version_str": "0.0.0",
129
+ "major": 0,
130
+ "minor": 0,
131
+ "patch": 0
132
+ }
133
+ }
data/merged_contrastive_pythia_70m_from_model_rlhf_on_task_hh_rlhf_activations_dataset.hf/new_rejected/state.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00002.arrow"
5
+ },
6
+ {
7
+ "filename": "data-00001-of-00002.arrow"
8
+ }
9
+ ],
10
+ "_fingerprint": "e01fae6d8a21cad2",
11
+ "_format_columns": null,
12
+ "_format_kwargs": {},
13
+ "_format_type": null,
14
+ "_output_all_columns": false,
15
+ "_split": "train"
16
+ }