XuHuang commited on
Commit
28474da
·
verified ·
1 Parent(s): 1308e42

Upload folder using huggingface_hub

Browse files
dataset_dict.json CHANGED
@@ -1 +1 @@
1
- {"splits": ["train"]}
 
1
+ {"splits": ["train", "test"]}
test/cache-dafba0fe1a069236.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ff4c7f5b52898f528234746ccec7f1671703480bb4f2f97c89df8810ee00421
3
+ size 25166872
test/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c718fdf8dfa94ff76d054e0d2679584fc35debfd035d70d1d9d0c5d33f3f4b5a
3
+ size 25617864
test/dataset_info.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "builder_name": "json",
3
+ "citation": "",
4
+ "config_name": "default",
5
+ "dataset_name": "json",
6
+ "dataset_size": 8556388,
7
+ "description": "",
8
+ "download_checksums": {
9
+ "/hai/scratch/fangwu97/xu/SimPO_slurm/data/gemma2_ufb_part2_test.jsonl": {
10
+ "num_bytes": 8964456,
11
+ "checksum": null
12
+ }
13
+ },
14
+ "download_size": 8964456,
15
+ "features": {
16
+ "prompt_id": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "prompt": {
21
+ "dtype": "string",
22
+ "_type": "Value"
23
+ },
24
+ "all_generated_responses": {
25
+ "feature": {
26
+ "dtype": "string",
27
+ "_type": "Value"
28
+ },
29
+ "_type": "Sequence"
30
+ },
31
+ "all_rm_scores": {
32
+ "feature": {
33
+ "dtype": "float64",
34
+ "_type": "Value"
35
+ },
36
+ "_type": "Sequence"
37
+ },
38
+ "chosen": {
39
+ "dtype": "string",
40
+ "_type": "Value"
41
+ },
42
+ "rejected": {
43
+ "dtype": "string",
44
+ "_type": "Value"
45
+ },
46
+ "chosen_input_ids": {
47
+ "feature": {
48
+ "dtype": "int64",
49
+ "_type": "Value"
50
+ },
51
+ "_type": "Sequence"
52
+ },
53
+ "chosen_attention_mask": {
54
+ "feature": {
55
+ "dtype": "int64",
56
+ "_type": "Value"
57
+ },
58
+ "_type": "Sequence"
59
+ },
60
+ "chosen_labels": {
61
+ "feature": {
62
+ "dtype": "int64",
63
+ "_type": "Value"
64
+ },
65
+ "_type": "Sequence"
66
+ },
67
+ "rejected_input_ids": {
68
+ "feature": {
69
+ "dtype": "int64",
70
+ "_type": "Value"
71
+ },
72
+ "_type": "Sequence"
73
+ },
74
+ "rejected_attention_mask": {
75
+ "feature": {
76
+ "dtype": "int64",
77
+ "_type": "Value"
78
+ },
79
+ "_type": "Sequence"
80
+ },
81
+ "rejected_labels": {
82
+ "feature": {
83
+ "dtype": "int64",
84
+ "_type": "Value"
85
+ },
86
+ "_type": "Sequence"
87
+ },
88
+ "reference_chosen_logps": {
89
+ "dtype": "float32",
90
+ "_type": "Value"
91
+ },
92
+ "reference_rejected_logps": {
93
+ "dtype": "float32",
94
+ "_type": "Value"
95
+ },
96
+ "history0_chosen_logps": {
97
+ "dtype": "float32",
98
+ "_type": "Value"
99
+ },
100
+ "history0_rejected_logps": {
101
+ "dtype": "float32",
102
+ "_type": "Value"
103
+ }
104
+ },
105
+ "homepage": "",
106
+ "license": "",
107
+ "size_in_bytes": 17520844,
108
+ "splits": {
109
+ "train": {
110
+ "name": "train",
111
+ "num_bytes": 8556388,
112
+ "num_examples": 647,
113
+ "dataset_name": "json"
114
+ }
115
+ },
116
+ "version": {
117
+ "version_str": "0.0.0",
118
+ "major": 0,
119
+ "minor": 0,
120
+ "patch": 0
121
+ }
122
+ }
test/state.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "0450ab0d809aebe0",
8
+ "_format_columns": null,
9
+ "_format_kwargs": {},
10
+ "_format_type": null,
11
+ "_output_all_columns": false,
12
+ "_split": null
13
+ }
train/cache-ebdd986512ba8955.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c793fa6491f4b2bd96737dca20bd6d3c7c47eb1c42c70a67e4a162250c45b800
3
+ size 732126608
train/data-00000-of-00002.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0faa1c8ce97a78b75c6c9172e43a8f3e7c59e5ac0357d9ea8778f961f962136a
3
+ size 340797176
train/data-00001-of-00002.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a846812c8ec12833ba1d1b9f86d1df3717b0c7a5d6183de579d6afcd49a4581
3
+ size 347019280
train/dataset_info.json ADDED
@@ -0,0 +1,92 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "prompt": {
6
+ "dtype": "string",
7
+ "_type": "Value"
8
+ },
9
+ "all_generated_responses": {
10
+ "feature": {
11
+ "dtype": "string",
12
+ "_type": "Value"
13
+ },
14
+ "_type": "Sequence"
15
+ },
16
+ "all_rm_scores": {
17
+ "feature": {
18
+ "dtype": "float64",
19
+ "_type": "Value"
20
+ },
21
+ "_type": "Sequence"
22
+ },
23
+ "chosen": {
24
+ "dtype": "string",
25
+ "_type": "Value"
26
+ },
27
+ "rejected": {
28
+ "dtype": "string",
29
+ "_type": "Value"
30
+ },
31
+ "chosen_input_ids": {
32
+ "feature": {
33
+ "dtype": "int64",
34
+ "_type": "Value"
35
+ },
36
+ "_type": "Sequence"
37
+ },
38
+ "chosen_attention_mask": {
39
+ "feature": {
40
+ "dtype": "int64",
41
+ "_type": "Value"
42
+ },
43
+ "_type": "Sequence"
44
+ },
45
+ "chosen_labels": {
46
+ "feature": {
47
+ "dtype": "int64",
48
+ "_type": "Value"
49
+ },
50
+ "_type": "Sequence"
51
+ },
52
+ "rejected_input_ids": {
53
+ "feature": {
54
+ "dtype": "int64",
55
+ "_type": "Value"
56
+ },
57
+ "_type": "Sequence"
58
+ },
59
+ "rejected_attention_mask": {
60
+ "feature": {
61
+ "dtype": "int64",
62
+ "_type": "Value"
63
+ },
64
+ "_type": "Sequence"
65
+ },
66
+ "rejected_labels": {
67
+ "feature": {
68
+ "dtype": "int64",
69
+ "_type": "Value"
70
+ },
71
+ "_type": "Sequence"
72
+ },
73
+ "reference_chosen_logps": {
74
+ "dtype": "float32",
75
+ "_type": "Value"
76
+ },
77
+ "reference_rejected_logps": {
78
+ "dtype": "float32",
79
+ "_type": "Value"
80
+ },
81
+ "history0_chosen_logps": {
82
+ "dtype": "float32",
83
+ "_type": "Value"
84
+ },
85
+ "history0_rejected_logps": {
86
+ "dtype": "float32",
87
+ "_type": "Value"
88
+ }
89
+ },
90
+ "homepage": "",
91
+ "license": ""
92
+ }
train/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": "16715b50e5b3cc4a",
11
+ "_format_columns": null,
12
+ "_format_kwargs": {},
13
+ "_format_type": null,
14
+ "_output_all_columns": false,
15
+ "_split": null
16
+ }