“Sara commited on
Commit
af5686d
1 Parent(s): df939db

adding model files

Browse files
added_tokens.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</headline>": 32008,
3
+ "</scratchpad>": 32006,
4
+ "<</SYS>>": 32004,
5
+ "<<SYS>>": 32003,
6
+ "<headline>": 32007,
7
+ "<pad>": 32000,
8
+ "<scratchpad>": 32005,
9
+ "[/INST]": 32002,
10
+ "[INST]": 32001
11
+ }
model-00001-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e39b7016acbc2a1d61317d75ef997a4d54deb0ecd8cb2f6d762d8f37f1c5c591
3
  size 4939116424
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d017cffe1e1a0a9c993743ae7c4b068fb672261441000ef2871b32c370df8b05
3
  size 4939116424
model-00002-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0ee5ccede844f101c62feced4d903e1caa53d31f847158a586af5f43a28bda79
3
  size 4947390880
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:79444a38aa98ab7ec492f96e8d34e4304bd6fe10c83623c279a4d374e1f43e25
3
  size 4947390880
model-00003-of-00003.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bebcfb114996140ad9b484d81e2ea9889ee47224919cc971a95b7313e87ff817
3
  size 3590619888
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b86201c526622980efed2ec5a7e85d5a6aaec18df1b0da14ce36c886b85584f
3
  size 3590619888
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:548f5da498253bace2643b1cfd48368f349b9461f62b66b5e0fb1ad02b8b93a1
3
+ size 1064
special_tokens_map.json CHANGED
@@ -1,4 +1,62 @@
1
  {
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2
  "bos_token": {
3
  "content": "<s>",
4
  "lstrip": false,
@@ -13,7 +71,7 @@
13
  "rstrip": false,
14
  "single_word": false
15
  },
16
- "pad_token": "</s>",
17
  "unk_token": {
18
  "content": "<unk>",
19
  "lstrip": false,
 
1
  {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "[INST]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": true
9
+ },
10
+ {
11
+ "content": "[/INST]",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": true
16
+ },
17
+ {
18
+ "content": "<<SYS>>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": true
23
+ },
24
+ {
25
+ "content": "<</SYS>>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": true
30
+ },
31
+ {
32
+ "content": "<scratchpad>",
33
+ "lstrip": false,
34
+ "normalized": false,
35
+ "rstrip": false,
36
+ "single_word": true
37
+ },
38
+ {
39
+ "content": "</scratchpad>",
40
+ "lstrip": false,
41
+ "normalized": false,
42
+ "rstrip": false,
43
+ "single_word": true
44
+ },
45
+ {
46
+ "content": "<headline>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": true
51
+ },
52
+ {
53
+ "content": "</headline>",
54
+ "lstrip": false,
55
+ "normalized": false,
56
+ "rstrip": false,
57
+ "single_word": true
58
+ }
59
+ ],
60
  "bos_token": {
61
  "content": "<s>",
62
  "lstrip": false,
 
71
  "rstrip": false,
72
  "single_word": false
73
  },
74
+ "pad_token": "<pad>",
75
  "unk_token": {
76
  "content": "<unk>",
77
  "lstrip": false,
tokenizer.json CHANGED
@@ -1,7 +1,14 @@
1
  {
2
  "version": "1.0",
3
  "truncation": null,
4
- "padding": null,
 
 
 
 
 
 
 
5
  "added_tokens": [
6
  {
7
  "id": 0,
@@ -29,6 +36,87 @@
29
  "rstrip": false,
30
  "normalized": false,
31
  "special": true
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
32
  }
33
  ],
34
  "normalizer": {
 
1
  {
2
  "version": "1.0",
3
  "truncation": null,
4
+ "padding": {
5
+ "strategy": "BatchLongest",
6
+ "direction": "Left",
7
+ "pad_to_multiple_of": null,
8
+ "pad_id": 32000,
9
+ "pad_type_id": 0,
10
+ "pad_token": "<pad>"
11
+ },
12
  "added_tokens": [
13
  {
14
  "id": 0,
 
36
  "rstrip": false,
37
  "normalized": false,
38
  "special": true
39
+ },
40
+ {
41
+ "id": 32000,
42
+ "content": "<pad>",
43
+ "single_word": false,
44
+ "lstrip": false,
45
+ "rstrip": false,
46
+ "normalized": false,
47
+ "special": true
48
+ },
49
+ {
50
+ "id": 32001,
51
+ "content": "[INST]",
52
+ "single_word": true,
53
+ "lstrip": false,
54
+ "rstrip": false,
55
+ "normalized": false,
56
+ "special": true
57
+ },
58
+ {
59
+ "id": 32002,
60
+ "content": "[/INST]",
61
+ "single_word": true,
62
+ "lstrip": false,
63
+ "rstrip": false,
64
+ "normalized": false,
65
+ "special": true
66
+ },
67
+ {
68
+ "id": 32003,
69
+ "content": "<<SYS>>",
70
+ "single_word": true,
71
+ "lstrip": false,
72
+ "rstrip": false,
73
+ "normalized": false,
74
+ "special": true
75
+ },
76
+ {
77
+ "id": 32004,
78
+ "content": "<</SYS>>",
79
+ "single_word": true,
80
+ "lstrip": false,
81
+ "rstrip": false,
82
+ "normalized": false,
83
+ "special": true
84
+ },
85
+ {
86
+ "id": 32005,
87
+ "content": "<scratchpad>",
88
+ "single_word": true,
89
+ "lstrip": false,
90
+ "rstrip": false,
91
+ "normalized": false,
92
+ "special": true
93
+ },
94
+ {
95
+ "id": 32006,
96
+ "content": "</scratchpad>",
97
+ "single_word": true,
98
+ "lstrip": false,
99
+ "rstrip": false,
100
+ "normalized": false,
101
+ "special": true
102
+ },
103
+ {
104
+ "id": 32007,
105
+ "content": "<headline>",
106
+ "single_word": true,
107
+ "lstrip": false,
108
+ "rstrip": false,
109
+ "normalized": false,
110
+ "special": true
111
+ },
112
+ {
113
+ "id": 32008,
114
+ "content": "</headline>",
115
+ "single_word": true,
116
+ "lstrip": false,
117
+ "rstrip": false,
118
+ "normalized": false,
119
+ "special": true
120
  }
121
  ],
122
  "normalizer": {
tokenizer_config.json CHANGED
@@ -25,15 +25,97 @@
25
  "rstrip": false,
26
  "single_word": false,
27
  "special": true
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
28
  }
29
  },
 
 
 
 
 
 
 
 
 
 
30
  "bos_token": "<s>",
31
  "clean_up_tokenization_spaces": false,
32
  "eos_token": "</s>",
33
  "legacy": false,
34
  "model_max_length": 1000000000000000019884624838656,
35
- "pad_token": "</s>",
36
- "padding_side": "right",
37
  "sp_model_kwargs": {},
38
  "tokenizer_class": "LlamaTokenizer",
39
  "unk_token": "<unk>",
 
25
  "rstrip": false,
26
  "single_word": false,
27
  "special": true
28
+ },
29
+ "32000": {
30
+ "content": "<pad>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "32001": {
38
+ "content": "[INST]",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": true,
43
+ "special": true
44
+ },
45
+ "32002": {
46
+ "content": "[/INST]",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": true,
51
+ "special": true
52
+ },
53
+ "32003": {
54
+ "content": "<<SYS>>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": true,
59
+ "special": true
60
+ },
61
+ "32004": {
62
+ "content": "<</SYS>>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": true,
67
+ "special": true
68
+ },
69
+ "32005": {
70
+ "content": "<scratchpad>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": true,
75
+ "special": true
76
+ },
77
+ "32006": {
78
+ "content": "</scratchpad>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": true,
83
+ "special": true
84
+ },
85
+ "32007": {
86
+ "content": "<headline>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": true,
91
+ "special": true
92
+ },
93
+ "32008": {
94
+ "content": "</headline>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": true,
99
+ "special": true
100
  }
101
  },
102
+ "additional_special_tokens": [
103
+ "[INST]",
104
+ "[/INST]",
105
+ "<<SYS>>",
106
+ "<</SYS>>",
107
+ "<scratchpad>",
108
+ "</scratchpad>",
109
+ "<headline>",
110
+ "</headline>"
111
+ ],
112
  "bos_token": "<s>",
113
  "clean_up_tokenization_spaces": false,
114
  "eos_token": "</s>",
115
  "legacy": false,
116
  "model_max_length": 1000000000000000019884624838656,
117
+ "pad_token": "<pad>",
118
+ "padding_side": "left",
119
  "sp_model_kwargs": {},
120
  "tokenizer_class": "LlamaTokenizer",
121
  "unk_token": "<unk>",
trainer_state.json ADDED
@@ -0,0 +1,1521 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.0,
5
+ "eval_steps": 10,
6
+ "global_step": 1000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "grad_norm": 22.5,
14
+ "learning_rate": 4.0000000000000003e-07,
15
+ "loss": 1.9102,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.02,
20
+ "eval_loss": 1.9285269975662231,
21
+ "eval_runtime": 2.9533,
22
+ "eval_samples_per_second": 52.822,
23
+ "eval_steps_per_second": 2.709,
24
+ "step": 10
25
+ },
26
+ {
27
+ "epoch": 0.04,
28
+ "grad_norm": 14.0,
29
+ "learning_rate": 8.000000000000001e-07,
30
+ "loss": 1.8734,
31
+ "step": 20
32
+ },
33
+ {
34
+ "epoch": 0.04,
35
+ "eval_loss": 1.907572627067566,
36
+ "eval_runtime": 2.9586,
37
+ "eval_samples_per_second": 52.728,
38
+ "eval_steps_per_second": 2.704,
39
+ "step": 20
40
+ },
41
+ {
42
+ "epoch": 0.06,
43
+ "grad_norm": 9.625,
44
+ "learning_rate": 1.2000000000000002e-06,
45
+ "loss": 1.77,
46
+ "step": 30
47
+ },
48
+ {
49
+ "epoch": 0.06,
50
+ "eval_loss": 1.8689841032028198,
51
+ "eval_runtime": 2.9565,
52
+ "eval_samples_per_second": 52.765,
53
+ "eval_steps_per_second": 2.706,
54
+ "step": 30
55
+ },
56
+ {
57
+ "epoch": 0.08,
58
+ "grad_norm": 12.75,
59
+ "learning_rate": 1.6000000000000001e-06,
60
+ "loss": 1.8568,
61
+ "step": 40
62
+ },
63
+ {
64
+ "epoch": 0.08,
65
+ "eval_loss": 1.812107801437378,
66
+ "eval_runtime": 2.9619,
67
+ "eval_samples_per_second": 52.669,
68
+ "eval_steps_per_second": 2.701,
69
+ "step": 40
70
+ },
71
+ {
72
+ "epoch": 0.1,
73
+ "grad_norm": 12.875,
74
+ "learning_rate": 2.0000000000000003e-06,
75
+ "loss": 1.7562,
76
+ "step": 50
77
+ },
78
+ {
79
+ "epoch": 0.1,
80
+ "eval_loss": 1.7258977890014648,
81
+ "eval_runtime": 2.9652,
82
+ "eval_samples_per_second": 52.61,
83
+ "eval_steps_per_second": 2.698,
84
+ "step": 50
85
+ },
86
+ {
87
+ "epoch": 0.12,
88
+ "grad_norm": 7.8125,
89
+ "learning_rate": 2.4000000000000003e-06,
90
+ "loss": 1.6208,
91
+ "step": 60
92
+ },
93
+ {
94
+ "epoch": 0.12,
95
+ "eval_loss": 1.5986799001693726,
96
+ "eval_runtime": 2.9727,
97
+ "eval_samples_per_second": 52.478,
98
+ "eval_steps_per_second": 2.691,
99
+ "step": 60
100
+ },
101
+ {
102
+ "epoch": 0.14,
103
+ "grad_norm": 8.25,
104
+ "learning_rate": 2.8000000000000003e-06,
105
+ "loss": 1.4408,
106
+ "step": 70
107
+ },
108
+ {
109
+ "epoch": 0.14,
110
+ "eval_loss": 1.3833414316177368,
111
+ "eval_runtime": 2.9613,
112
+ "eval_samples_per_second": 52.679,
113
+ "eval_steps_per_second": 2.701,
114
+ "step": 70
115
+ },
116
+ {
117
+ "epoch": 0.16,
118
+ "grad_norm": 5.65625,
119
+ "learning_rate": 3.2000000000000003e-06,
120
+ "loss": 1.1332,
121
+ "step": 80
122
+ },
123
+ {
124
+ "epoch": 0.16,
125
+ "eval_loss": 1.1267226934432983,
126
+ "eval_runtime": 2.97,
127
+ "eval_samples_per_second": 52.525,
128
+ "eval_steps_per_second": 2.694,
129
+ "step": 80
130
+ },
131
+ {
132
+ "epoch": 0.18,
133
+ "grad_norm": 7.28125,
134
+ "learning_rate": 3.6000000000000003e-06,
135
+ "loss": 1.0659,
136
+ "step": 90
137
+ },
138
+ {
139
+ "epoch": 0.18,
140
+ "eval_loss": 1.0400179624557495,
141
+ "eval_runtime": 2.9774,
142
+ "eval_samples_per_second": 52.395,
143
+ "eval_steps_per_second": 2.687,
144
+ "step": 90
145
+ },
146
+ {
147
+ "epoch": 0.2,
148
+ "grad_norm": 4.9375,
149
+ "learning_rate": 4.000000000000001e-06,
150
+ "loss": 0.9587,
151
+ "step": 100
152
+ },
153
+ {
154
+ "epoch": 0.2,
155
+ "eval_loss": 0.9716721773147583,
156
+ "eval_runtime": 2.9692,
157
+ "eval_samples_per_second": 52.54,
158
+ "eval_steps_per_second": 2.694,
159
+ "step": 100
160
+ },
161
+ {
162
+ "epoch": 0.22,
163
+ "grad_norm": 4.78125,
164
+ "learning_rate": 4.4e-06,
165
+ "loss": 0.9241,
166
+ "step": 110
167
+ },
168
+ {
169
+ "epoch": 0.22,
170
+ "eval_loss": 0.9269797205924988,
171
+ "eval_runtime": 2.9883,
172
+ "eval_samples_per_second": 52.204,
173
+ "eval_steps_per_second": 2.677,
174
+ "step": 110
175
+ },
176
+ {
177
+ "epoch": 0.24,
178
+ "grad_norm": 8.5,
179
+ "learning_rate": 4.800000000000001e-06,
180
+ "loss": 0.8071,
181
+ "step": 120
182
+ },
183
+ {
184
+ "epoch": 0.24,
185
+ "eval_loss": 0.8944205045700073,
186
+ "eval_runtime": 2.9805,
187
+ "eval_samples_per_second": 52.341,
188
+ "eval_steps_per_second": 2.684,
189
+ "step": 120
190
+ },
191
+ {
192
+ "epoch": 0.26,
193
+ "grad_norm": 3.34375,
194
+ "learning_rate": 5.2e-06,
195
+ "loss": 0.7872,
196
+ "step": 130
197
+ },
198
+ {
199
+ "epoch": 0.26,
200
+ "eval_loss": 0.874414324760437,
201
+ "eval_runtime": 2.9734,
202
+ "eval_samples_per_second": 52.465,
203
+ "eval_steps_per_second": 2.691,
204
+ "step": 130
205
+ },
206
+ {
207
+ "epoch": 0.28,
208
+ "grad_norm": 4.0625,
209
+ "learning_rate": 5.600000000000001e-06,
210
+ "loss": 0.7182,
211
+ "step": 140
212
+ },
213
+ {
214
+ "epoch": 0.28,
215
+ "eval_loss": 0.8589532375335693,
216
+ "eval_runtime": 2.9762,
217
+ "eval_samples_per_second": 52.415,
218
+ "eval_steps_per_second": 2.688,
219
+ "step": 140
220
+ },
221
+ {
222
+ "epoch": 0.3,
223
+ "grad_norm": 4.8125,
224
+ "learning_rate": 6e-06,
225
+ "loss": 0.7139,
226
+ "step": 150
227
+ },
228
+ {
229
+ "epoch": 0.3,
230
+ "eval_loss": 0.8482792973518372,
231
+ "eval_runtime": 2.974,
232
+ "eval_samples_per_second": 52.455,
233
+ "eval_steps_per_second": 2.69,
234
+ "step": 150
235
+ },
236
+ {
237
+ "epoch": 0.32,
238
+ "grad_norm": 4.4375,
239
+ "learning_rate": 6.4000000000000006e-06,
240
+ "loss": 0.7533,
241
+ "step": 160
242
+ },
243
+ {
244
+ "epoch": 0.32,
245
+ "eval_loss": 0.8366221785545349,
246
+ "eval_runtime": 2.9775,
247
+ "eval_samples_per_second": 52.393,
248
+ "eval_steps_per_second": 2.687,
249
+ "step": 160
250
+ },
251
+ {
252
+ "epoch": 0.34,
253
+ "grad_norm": 3.625,
254
+ "learning_rate": 6.800000000000001e-06,
255
+ "loss": 0.808,
256
+ "step": 170
257
+ },
258
+ {
259
+ "epoch": 0.34,
260
+ "eval_loss": 0.8267260789871216,
261
+ "eval_runtime": 2.9805,
262
+ "eval_samples_per_second": 52.341,
263
+ "eval_steps_per_second": 2.684,
264
+ "step": 170
265
+ },
266
+ {
267
+ "epoch": 0.36,
268
+ "grad_norm": 4.9375,
269
+ "learning_rate": 7.2000000000000005e-06,
270
+ "loss": 0.5991,
271
+ "step": 180
272
+ },
273
+ {
274
+ "epoch": 0.36,
275
+ "eval_loss": 0.8209083676338196,
276
+ "eval_runtime": 2.9739,
277
+ "eval_samples_per_second": 52.456,
278
+ "eval_steps_per_second": 2.69,
279
+ "step": 180
280
+ },
281
+ {
282
+ "epoch": 0.38,
283
+ "grad_norm": 3.671875,
284
+ "learning_rate": 7.600000000000001e-06,
285
+ "loss": 0.6681,
286
+ "step": 190
287
+ },
288
+ {
289
+ "epoch": 0.38,
290
+ "eval_loss": 0.8131274580955505,
291
+ "eval_runtime": 2.9722,
292
+ "eval_samples_per_second": 52.486,
293
+ "eval_steps_per_second": 2.692,
294
+ "step": 190
295
+ },
296
+ {
297
+ "epoch": 0.4,
298
+ "grad_norm": 3.578125,
299
+ "learning_rate": 8.000000000000001e-06,
300
+ "loss": 0.6707,
301
+ "step": 200
302
+ },
303
+ {
304
+ "epoch": 0.4,
305
+ "eval_loss": 0.8074705600738525,
306
+ "eval_runtime": 2.9825,
307
+ "eval_samples_per_second": 52.305,
308
+ "eval_steps_per_second": 2.682,
309
+ "step": 200
310
+ },
311
+ {
312
+ "epoch": 0.42,
313
+ "grad_norm": 3.71875,
314
+ "learning_rate": 8.400000000000001e-06,
315
+ "loss": 0.7301,
316
+ "step": 210
317
+ },
318
+ {
319
+ "epoch": 0.42,
320
+ "eval_loss": 0.8011156916618347,
321
+ "eval_runtime": 2.9798,
322
+ "eval_samples_per_second": 52.352,
323
+ "eval_steps_per_second": 2.685,
324
+ "step": 210
325
+ },
326
+ {
327
+ "epoch": 0.44,
328
+ "grad_norm": 3.734375,
329
+ "learning_rate": 8.8e-06,
330
+ "loss": 0.6486,
331
+ "step": 220
332
+ },
333
+ {
334
+ "epoch": 0.44,
335
+ "eval_loss": 0.7965356111526489,
336
+ "eval_runtime": 2.9734,
337
+ "eval_samples_per_second": 52.465,
338
+ "eval_steps_per_second": 2.69,
339
+ "step": 220
340
+ },
341
+ {
342
+ "epoch": 0.46,
343
+ "grad_norm": 3.34375,
344
+ "learning_rate": 9.200000000000002e-06,
345
+ "loss": 0.6744,
346
+ "step": 230
347
+ },
348
+ {
349
+ "epoch": 0.46,
350
+ "eval_loss": 0.7928165197372437,
351
+ "eval_runtime": 2.9833,
352
+ "eval_samples_per_second": 52.29,
353
+ "eval_steps_per_second": 2.682,
354
+ "step": 230
355
+ },
356
+ {
357
+ "epoch": 0.48,
358
+ "grad_norm": 3.8125,
359
+ "learning_rate": 9.600000000000001e-06,
360
+ "loss": 0.7228,
361
+ "step": 240
362
+ },
363
+ {
364
+ "epoch": 0.48,
365
+ "eval_loss": 0.7894927859306335,
366
+ "eval_runtime": 2.9769,
367
+ "eval_samples_per_second": 52.403,
368
+ "eval_steps_per_second": 2.687,
369
+ "step": 240
370
+ },
371
+ {
372
+ "epoch": 0.5,
373
+ "grad_norm": 4.5,
374
+ "learning_rate": 1e-05,
375
+ "loss": 0.7532,
376
+ "step": 250
377
+ },
378
+ {
379
+ "epoch": 0.5,
380
+ "eval_loss": 0.7849324345588684,
381
+ "eval_runtime": 2.9712,
382
+ "eval_samples_per_second": 52.504,
383
+ "eval_steps_per_second": 2.693,
384
+ "step": 250
385
+ },
386
+ {
387
+ "epoch": 0.52,
388
+ "grad_norm": 5.8125,
389
+ "learning_rate": 1.04e-05,
390
+ "loss": 0.7019,
391
+ "step": 260
392
+ },
393
+ {
394
+ "epoch": 0.52,
395
+ "eval_loss": 0.7828314304351807,
396
+ "eval_runtime": 2.979,
397
+ "eval_samples_per_second": 52.366,
398
+ "eval_steps_per_second": 2.685,
399
+ "step": 260
400
+ },
401
+ {
402
+ "epoch": 0.54,
403
+ "grad_norm": 3.75,
404
+ "learning_rate": 1.0800000000000002e-05,
405
+ "loss": 0.7255,
406
+ "step": 270
407
+ },
408
+ {
409
+ "epoch": 0.54,
410
+ "eval_loss": 0.7806087136268616,
411
+ "eval_runtime": 2.9708,
412
+ "eval_samples_per_second": 52.512,
413
+ "eval_steps_per_second": 2.693,
414
+ "step": 270
415
+ },
416
+ {
417
+ "epoch": 0.56,
418
+ "grad_norm": 3.75,
419
+ "learning_rate": 1.1200000000000001e-05,
420
+ "loss": 0.6541,
421
+ "step": 280
422
+ },
423
+ {
424
+ "epoch": 0.56,
425
+ "eval_loss": 0.7776769995689392,
426
+ "eval_runtime": 2.972,
427
+ "eval_samples_per_second": 52.489,
428
+ "eval_steps_per_second": 2.692,
429
+ "step": 280
430
+ },
431
+ {
432
+ "epoch": 0.58,
433
+ "grad_norm": 4.09375,
434
+ "learning_rate": 1.16e-05,
435
+ "loss": 0.6201,
436
+ "step": 290
437
+ },
438
+ {
439
+ "epoch": 0.58,
440
+ "eval_loss": 0.7737365961074829,
441
+ "eval_runtime": 2.9739,
442
+ "eval_samples_per_second": 52.456,
443
+ "eval_steps_per_second": 2.69,
444
+ "step": 290
445
+ },
446
+ {
447
+ "epoch": 0.6,
448
+ "grad_norm": 2.390625,
449
+ "learning_rate": 1.2e-05,
450
+ "loss": 0.6149,
451
+ "step": 300
452
+ },
453
+ {
454
+ "epoch": 0.6,
455
+ "eval_loss": 0.7721969485282898,
456
+ "eval_runtime": 2.9719,
457
+ "eval_samples_per_second": 52.492,
458
+ "eval_steps_per_second": 2.692,
459
+ "step": 300
460
+ },
461
+ {
462
+ "epoch": 0.62,
463
+ "grad_norm": 4.59375,
464
+ "learning_rate": 1.2400000000000002e-05,
465
+ "loss": 0.6927,
466
+ "step": 310
467
+ },
468
+ {
469
+ "epoch": 0.62,
470
+ "eval_loss": 0.7704524397850037,
471
+ "eval_runtime": 2.9722,
472
+ "eval_samples_per_second": 52.487,
473
+ "eval_steps_per_second": 2.692,
474
+ "step": 310
475
+ },
476
+ {
477
+ "epoch": 0.64,
478
+ "grad_norm": 2.953125,
479
+ "learning_rate": 1.2800000000000001e-05,
480
+ "loss": 0.6573,
481
+ "step": 320
482
+ },
483
+ {
484
+ "epoch": 0.64,
485
+ "eval_loss": 0.7695561051368713,
486
+ "eval_runtime": 2.9675,
487
+ "eval_samples_per_second": 52.57,
488
+ "eval_steps_per_second": 2.696,
489
+ "step": 320
490
+ },
491
+ {
492
+ "epoch": 0.66,
493
+ "grad_norm": 4.65625,
494
+ "learning_rate": 1.3200000000000002e-05,
495
+ "loss": 0.6786,
496
+ "step": 330
497
+ },
498
+ {
499
+ "epoch": 0.66,
500
+ "eval_loss": 0.7669094800949097,
501
+ "eval_runtime": 2.9692,
502
+ "eval_samples_per_second": 52.54,
503
+ "eval_steps_per_second": 2.694,
504
+ "step": 330
505
+ },
506
+ {
507
+ "epoch": 0.68,
508
+ "grad_norm": 3.359375,
509
+ "learning_rate": 1.3600000000000002e-05,
510
+ "loss": 0.7484,
511
+ "step": 340
512
+ },
513
+ {
514
+ "epoch": 0.68,
515
+ "eval_loss": 0.7643001079559326,
516
+ "eval_runtime": 2.9737,
517
+ "eval_samples_per_second": 52.46,
518
+ "eval_steps_per_second": 2.69,
519
+ "step": 340
520
+ },
521
+ {
522
+ "epoch": 0.7,
523
+ "grad_norm": 4.4375,
524
+ "learning_rate": 1.4e-05,
525
+ "loss": 0.5991,
526
+ "step": 350
527
+ },
528
+ {
529
+ "epoch": 0.7,
530
+ "eval_loss": 0.7622258067131042,
531
+ "eval_runtime": 2.9746,
532
+ "eval_samples_per_second": 52.444,
533
+ "eval_steps_per_second": 2.689,
534
+ "step": 350
535
+ },
536
+ {
537
+ "epoch": 0.72,
538
+ "grad_norm": 3.359375,
539
+ "learning_rate": 1.4400000000000001e-05,
540
+ "loss": 0.6621,
541
+ "step": 360
542
+ },
543
+ {
544
+ "epoch": 0.72,
545
+ "eval_loss": 0.761843204498291,
546
+ "eval_runtime": 2.9802,
547
+ "eval_samples_per_second": 52.345,
548
+ "eval_steps_per_second": 2.684,
549
+ "step": 360
550
+ },
551
+ {
552
+ "epoch": 0.74,
553
+ "grad_norm": 3.890625,
554
+ "learning_rate": 1.48e-05,
555
+ "loss": 0.697,
556
+ "step": 370
557
+ },
558
+ {
559
+ "epoch": 0.74,
560
+ "eval_loss": 0.7606173753738403,
561
+ "eval_runtime": 2.9702,
562
+ "eval_samples_per_second": 52.522,
563
+ "eval_steps_per_second": 2.693,
564
+ "step": 370
565
+ },
566
+ {
567
+ "epoch": 0.76,
568
+ "grad_norm": 3.15625,
569
+ "learning_rate": 1.5200000000000002e-05,
570
+ "loss": 0.6213,
571
+ "step": 380
572
+ },
573
+ {
574
+ "epoch": 0.76,
575
+ "eval_loss": 0.7589120268821716,
576
+ "eval_runtime": 2.9752,
577
+ "eval_samples_per_second": 52.434,
578
+ "eval_steps_per_second": 2.689,
579
+ "step": 380
580
+ },
581
+ {
582
+ "epoch": 0.78,
583
+ "grad_norm": 3.96875,
584
+ "learning_rate": 1.5600000000000003e-05,
585
+ "loss": 0.6767,
586
+ "step": 390
587
+ },
588
+ {
589
+ "epoch": 0.78,
590
+ "eval_loss": 0.7586512565612793,
591
+ "eval_runtime": 2.9725,
592
+ "eval_samples_per_second": 52.482,
593
+ "eval_steps_per_second": 2.691,
594
+ "step": 390
595
+ },
596
+ {
597
+ "epoch": 0.8,
598
+ "grad_norm": 4.34375,
599
+ "learning_rate": 1.6000000000000003e-05,
600
+ "loss": 0.8044,
601
+ "step": 400
602
+ },
603
+ {
604
+ "epoch": 0.8,
605
+ "eval_loss": 0.7563453316688538,
606
+ "eval_runtime": 2.9763,
607
+ "eval_samples_per_second": 52.414,
608
+ "eval_steps_per_second": 2.688,
609
+ "step": 400
610
+ },
611
+ {
612
+ "epoch": 0.82,
613
+ "grad_norm": 3.203125,
614
+ "learning_rate": 1.64e-05,
615
+ "loss": 0.6338,
616
+ "step": 410
617
+ },
618
+ {
619
+ "epoch": 0.82,
620
+ "eval_loss": 0.7531482577323914,
621
+ "eval_runtime": 2.9746,
622
+ "eval_samples_per_second": 52.444,
623
+ "eval_steps_per_second": 2.689,
624
+ "step": 410
625
+ },
626
+ {
627
+ "epoch": 0.84,
628
+ "grad_norm": 3.140625,
629
+ "learning_rate": 1.6800000000000002e-05,
630
+ "loss": 0.6242,
631
+ "step": 420
632
+ },
633
+ {
634
+ "epoch": 0.84,
635
+ "eval_loss": 0.7514854669570923,
636
+ "eval_runtime": 2.9804,
637
+ "eval_samples_per_second": 52.342,
638
+ "eval_steps_per_second": 2.684,
639
+ "step": 420
640
+ },
641
+ {
642
+ "epoch": 0.86,
643
+ "grad_norm": 3.203125,
644
+ "learning_rate": 1.72e-05,
645
+ "loss": 0.6462,
646
+ "step": 430
647
+ },
648
+ {
649
+ "epoch": 0.86,
650
+ "eval_loss": 0.7528823614120483,
651
+ "eval_runtime": 2.9744,
652
+ "eval_samples_per_second": 52.448,
653
+ "eval_steps_per_second": 2.69,
654
+ "step": 430
655
+ },
656
+ {
657
+ "epoch": 0.88,
658
+ "grad_norm": 4.9375,
659
+ "learning_rate": 1.76e-05,
660
+ "loss": 0.6832,
661
+ "step": 440
662
+ },
663
+ {
664
+ "epoch": 0.88,
665
+ "eval_loss": 0.7538066506385803,
666
+ "eval_runtime": 2.9692,
667
+ "eval_samples_per_second": 52.54,
668
+ "eval_steps_per_second": 2.694,
669
+ "step": 440
670
+ },
671
+ {
672
+ "epoch": 0.9,
673
+ "grad_norm": 6.28125,
674
+ "learning_rate": 1.8e-05,
675
+ "loss": 0.5959,
676
+ "step": 450
677
+ },
678
+ {
679
+ "epoch": 0.9,
680
+ "eval_loss": 0.7551948428153992,
681
+ "eval_runtime": 2.9739,
682
+ "eval_samples_per_second": 52.456,
683
+ "eval_steps_per_second": 2.69,
684
+ "step": 450
685
+ },
686
+ {
687
+ "epoch": 0.92,
688
+ "grad_norm": 7.5625,
689
+ "learning_rate": 1.8400000000000003e-05,
690
+ "loss": 0.6891,
691
+ "step": 460
692
+ },
693
+ {
694
+ "epoch": 0.92,
695
+ "eval_loss": 0.7493566870689392,
696
+ "eval_runtime": 2.976,
697
+ "eval_samples_per_second": 52.42,
698
+ "eval_steps_per_second": 2.688,
699
+ "step": 460
700
+ },
701
+ {
702
+ "epoch": 0.94,
703
+ "grad_norm": 5.75,
704
+ "learning_rate": 1.88e-05,
705
+ "loss": 0.6819,
706
+ "step": 470
707
+ },
708
+ {
709
+ "epoch": 0.94,
710
+ "eval_loss": 0.7475622296333313,
711
+ "eval_runtime": 2.9745,
712
+ "eval_samples_per_second": 52.445,
713
+ "eval_steps_per_second": 2.69,
714
+ "step": 470
715
+ },
716
+ {
717
+ "epoch": 0.96,
718
+ "grad_norm": 3.0,
719
+ "learning_rate": 1.9200000000000003e-05,
720
+ "loss": 0.6456,
721
+ "step": 480
722
+ },
723
+ {
724
+ "epoch": 0.96,
725
+ "eval_loss": 0.7493394613265991,
726
+ "eval_runtime": 2.968,
727
+ "eval_samples_per_second": 52.56,
728
+ "eval_steps_per_second": 2.695,
729
+ "step": 480
730
+ },
731
+ {
732
+ "epoch": 0.98,
733
+ "grad_norm": 3.65625,
734
+ "learning_rate": 1.9600000000000002e-05,
735
+ "loss": 0.6415,
736
+ "step": 490
737
+ },
738
+ {
739
+ "epoch": 0.98,
740
+ "eval_loss": 0.7468904256820679,
741
+ "eval_runtime": 2.972,
742
+ "eval_samples_per_second": 52.491,
743
+ "eval_steps_per_second": 2.692,
744
+ "step": 490
745
+ },
746
+ {
747
+ "epoch": 1.0,
748
+ "grad_norm": 3.421875,
749
+ "learning_rate": 2e-05,
750
+ "loss": 0.658,
751
+ "step": 500
752
+ },
753
+ {
754
+ "epoch": 1.0,
755
+ "eval_loss": 0.7486743330955505,
756
+ "eval_runtime": 2.9736,
757
+ "eval_samples_per_second": 52.462,
758
+ "eval_steps_per_second": 2.69,
759
+ "step": 500
760
+ },
761
+ {
762
+ "epoch": 1.02,
763
+ "grad_norm": 5.625,
764
+ "learning_rate": 1.9999756307053947e-05,
765
+ "loss": 0.547,
766
+ "step": 510
767
+ },
768
+ {
769
+ "epoch": 1.02,
770
+ "eval_loss": 0.7537599205970764,
771
+ "eval_runtime": 2.957,
772
+ "eval_samples_per_second": 52.757,
773
+ "eval_steps_per_second": 2.705,
774
+ "step": 510
775
+ },
776
+ {
777
+ "epoch": 1.04,
778
+ "grad_norm": 3.09375,
779
+ "learning_rate": 1.9999025240093045e-05,
780
+ "loss": 0.5551,
781
+ "step": 520
782
+ },
783
+ {
784
+ "epoch": 1.04,
785
+ "eval_loss": 0.7580664157867432,
786
+ "eval_runtime": 2.9622,
787
+ "eval_samples_per_second": 52.664,
788
+ "eval_steps_per_second": 2.701,
789
+ "step": 520
790
+ },
791
+ {
792
+ "epoch": 1.06,
793
+ "grad_norm": 2.828125,
794
+ "learning_rate": 1.9997806834748455e-05,
795
+ "loss": 0.596,
796
+ "step": 530
797
+ },
798
+ {
799
+ "epoch": 1.06,
800
+ "eval_loss": 0.7601476907730103,
801
+ "eval_runtime": 2.9678,
802
+ "eval_samples_per_second": 52.564,
803
+ "eval_steps_per_second": 2.696,
804
+ "step": 530
805
+ },
806
+ {
807
+ "epoch": 1.08,
808
+ "grad_norm": 2.3125,
809
+ "learning_rate": 1.9996101150403543e-05,
810
+ "loss": 0.5611,
811
+ "step": 540
812
+ },
813
+ {
814
+ "epoch": 1.08,
815
+ "eval_loss": 0.7624161243438721,
816
+ "eval_runtime": 2.9644,
817
+ "eval_samples_per_second": 52.624,
818
+ "eval_steps_per_second": 2.699,
819
+ "step": 540
820
+ },
821
+ {
822
+ "epoch": 1.1,
823
+ "grad_norm": 2.953125,
824
+ "learning_rate": 1.999390827019096e-05,
825
+ "loss": 0.5514,
826
+ "step": 550
827
+ },
828
+ {
829
+ "epoch": 1.1,
830
+ "eval_loss": 0.7590138912200928,
831
+ "eval_runtime": 2.9709,
832
+ "eval_samples_per_second": 52.51,
833
+ "eval_steps_per_second": 2.693,
834
+ "step": 550
835
+ },
836
+ {
837
+ "epoch": 1.12,
838
+ "grad_norm": 3.59375,
839
+ "learning_rate": 1.9991228300988586e-05,
840
+ "loss": 0.5109,
841
+ "step": 560
842
+ },
843
+ {
844
+ "epoch": 1.12,
845
+ "eval_loss": 0.754431426525116,
846
+ "eval_runtime": 2.9642,
847
+ "eval_samples_per_second": 52.628,
848
+ "eval_steps_per_second": 2.699,
849
+ "step": 560
850
+ },
851
+ {
852
+ "epoch": 1.14,
853
+ "grad_norm": 3.03125,
854
+ "learning_rate": 1.9988061373414342e-05,
855
+ "loss": 0.5984,
856
+ "step": 570
857
+ },
858
+ {
859
+ "epoch": 1.14,
860
+ "eval_loss": 0.7549035549163818,
861
+ "eval_runtime": 2.9676,
862
+ "eval_samples_per_second": 52.567,
863
+ "eval_steps_per_second": 2.696,
864
+ "step": 570
865
+ },
866
+ {
867
+ "epoch": 1.16,
868
+ "grad_norm": 1.9296875,
869
+ "learning_rate": 1.9984407641819812e-05,
870
+ "loss": 0.5346,
871
+ "step": 580
872
+ },
873
+ {
874
+ "epoch": 1.16,
875
+ "eval_loss": 0.7530075311660767,
876
+ "eval_runtime": 2.9726,
877
+ "eval_samples_per_second": 52.48,
878
+ "eval_steps_per_second": 2.691,
879
+ "step": 580
880
+ },
881
+ {
882
+ "epoch": 1.18,
883
+ "grad_norm": 3.15625,
884
+ "learning_rate": 1.9980267284282718e-05,
885
+ "loss": 0.6438,
886
+ "step": 590
887
+ },
888
+ {
889
+ "epoch": 1.18,
890
+ "eval_loss": 0.7559706568717957,
891
+ "eval_runtime": 2.9783,
892
+ "eval_samples_per_second": 52.378,
893
+ "eval_steps_per_second": 2.686,
894
+ "step": 590
895
+ },
896
+ {
897
+ "epoch": 1.2,
898
+ "grad_norm": 3.0,
899
+ "learning_rate": 1.9975640502598243e-05,
900
+ "loss": 0.5193,
901
+ "step": 600
902
+ },
903
+ {
904
+ "epoch": 1.2,
905
+ "eval_loss": 0.7533587217330933,
906
+ "eval_runtime": 2.9733,
907
+ "eval_samples_per_second": 52.468,
908
+ "eval_steps_per_second": 2.691,
909
+ "step": 600
910
+ },
911
+ {
912
+ "epoch": 1.22,
913
+ "grad_norm": 4.34375,
914
+ "learning_rate": 1.9970527522269204e-05,
915
+ "loss": 0.6232,
916
+ "step": 610
917
+ },
918
+ {
919
+ "epoch": 1.22,
920
+ "eval_loss": 0.7505587339401245,
921
+ "eval_runtime": 2.9729,
922
+ "eval_samples_per_second": 52.474,
923
+ "eval_steps_per_second": 2.691,
924
+ "step": 610
925
+ },
926
+ {
927
+ "epoch": 1.24,
928
+ "grad_norm": 2.125,
929
+ "learning_rate": 1.9964928592495046e-05,
930
+ "loss": 0.6118,
931
+ "step": 620
932
+ },
933
+ {
934
+ "epoch": 1.24,
935
+ "eval_loss": 0.747382402420044,
936
+ "eval_runtime": 2.9751,
937
+ "eval_samples_per_second": 52.436,
938
+ "eval_steps_per_second": 2.689,
939
+ "step": 620
940
+ },
941
+ {
942
+ "epoch": 1.26,
943
+ "grad_norm": 3.21875,
944
+ "learning_rate": 1.9958843986159705e-05,
945
+ "loss": 0.6072,
946
+ "step": 630
947
+ },
948
+ {
949
+ "epoch": 1.26,
950
+ "eval_loss": 0.7488105893135071,
951
+ "eval_runtime": 2.9725,
952
+ "eval_samples_per_second": 52.481,
953
+ "eval_steps_per_second": 2.691,
954
+ "step": 630
955
+ },
956
+ {
957
+ "epoch": 1.28,
958
+ "grad_norm": 2.71875,
959
+ "learning_rate": 1.9952273999818312e-05,
960
+ "loss": 0.6088,
961
+ "step": 640
962
+ },
963
+ {
964
+ "epoch": 1.28,
965
+ "eval_loss": 0.7461674809455872,
966
+ "eval_runtime": 2.9722,
967
+ "eval_samples_per_second": 52.486,
968
+ "eval_steps_per_second": 2.692,
969
+ "step": 640
970
+ },
971
+ {
972
+ "epoch": 1.3,
973
+ "grad_norm": 2.5,
974
+ "learning_rate": 1.9945218953682736e-05,
975
+ "loss": 0.6276,
976
+ "step": 650
977
+ },
978
+ {
979
+ "epoch": 1.3,
980
+ "eval_loss": 0.7464598417282104,
981
+ "eval_runtime": 2.9781,
982
+ "eval_samples_per_second": 52.383,
983
+ "eval_steps_per_second": 2.686,
984
+ "step": 650
985
+ },
986
+ {
987
+ "epoch": 1.32,
988
+ "grad_norm": 2.421875,
989
+ "learning_rate": 1.9937679191605964e-05,
990
+ "loss": 0.5798,
991
+ "step": 660
992
+ },
993
+ {
994
+ "epoch": 1.32,
995
+ "eval_loss": 0.7467732429504395,
996
+ "eval_runtime": 2.977,
997
+ "eval_samples_per_second": 52.403,
998
+ "eval_steps_per_second": 2.687,
999
+ "step": 660
1000
+ },
1001
+ {
1002
+ "epoch": 1.34,
1003
+ "grad_norm": 3.4375,
1004
+ "learning_rate": 1.992965508106537e-05,
1005
+ "loss": 0.6283,
1006
+ "step": 670
1007
+ },
1008
+ {
1009
+ "epoch": 1.34,
1010
+ "eval_loss": 0.7440893054008484,
1011
+ "eval_runtime": 2.9772,
1012
+ "eval_samples_per_second": 52.398,
1013
+ "eval_steps_per_second": 2.687,
1014
+ "step": 670
1015
+ },
1016
+ {
1017
+ "epoch": 1.36,
1018
+ "grad_norm": 2.78125,
1019
+ "learning_rate": 1.9921147013144782e-05,
1020
+ "loss": 0.5477,
1021
+ "step": 680
1022
+ },
1023
+ {
1024
+ "epoch": 1.36,
1025
+ "eval_loss": 0.745173454284668,
1026
+ "eval_runtime": 2.9776,
1027
+ "eval_samples_per_second": 52.392,
1028
+ "eval_steps_per_second": 2.687,
1029
+ "step": 680
1030
+ },
1031
+ {
1032
+ "epoch": 1.38,
1033
+ "grad_norm": 2.21875,
1034
+ "learning_rate": 1.991215540251542e-05,
1035
+ "loss": 0.5479,
1036
+ "step": 690
1037
+ },
1038
+ {
1039
+ "epoch": 1.38,
1040
+ "eval_loss": 0.7425445318222046,
1041
+ "eval_runtime": 2.9811,
1042
+ "eval_samples_per_second": 52.33,
1043
+ "eval_steps_per_second": 2.684,
1044
+ "step": 690
1045
+ },
1046
+ {
1047
+ "epoch": 1.4,
1048
+ "grad_norm": 3.171875,
1049
+ "learning_rate": 1.9902680687415704e-05,
1050
+ "loss": 0.6648,
1051
+ "step": 700
1052
+ },
1053
+ {
1054
+ "epoch": 1.4,
1055
+ "eval_loss": 0.7418652772903442,
1056
+ "eval_runtime": 2.9712,
1057
+ "eval_samples_per_second": 52.503,
1058
+ "eval_steps_per_second": 2.692,
1059
+ "step": 700
1060
+ },
1061
+ {
1062
+ "epoch": 1.42,
1063
+ "grad_norm": 3.5,
1064
+ "learning_rate": 1.9892723329629885e-05,
1065
+ "loss": 0.5988,
1066
+ "step": 710
1067
+ },
1068
+ {
1069
+ "epoch": 1.42,
1070
+ "eval_loss": 0.7394148707389832,
1071
+ "eval_runtime": 2.9689,
1072
+ "eval_samples_per_second": 52.544,
1073
+ "eval_steps_per_second": 2.695,
1074
+ "step": 710
1075
+ },
1076
+ {
1077
+ "epoch": 1.44,
1078
+ "grad_norm": 2.984375,
1079
+ "learning_rate": 1.988228381446553e-05,
1080
+ "loss": 0.5557,
1081
+ "step": 720
1082
+ },
1083
+ {
1084
+ "epoch": 1.44,
1085
+ "eval_loss": 0.7416422963142395,
1086
+ "eval_runtime": 2.9747,
1087
+ "eval_samples_per_second": 52.442,
1088
+ "eval_steps_per_second": 2.689,
1089
+ "step": 720
1090
+ },
1091
+ {
1092
+ "epoch": 1.46,
1093
+ "grad_norm": 2.421875,
1094
+ "learning_rate": 1.987136265072988e-05,
1095
+ "loss": 0.4506,
1096
+ "step": 730
1097
+ },
1098
+ {
1099
+ "epoch": 1.46,
1100
+ "eval_loss": 0.7449523210525513,
1101
+ "eval_runtime": 2.9679,
1102
+ "eval_samples_per_second": 52.563,
1103
+ "eval_steps_per_second": 2.696,
1104
+ "step": 730
1105
+ },
1106
+ {
1107
+ "epoch": 1.48,
1108
+ "grad_norm": 3.265625,
1109
+ "learning_rate": 1.985996037070505e-05,
1110
+ "loss": 0.6386,
1111
+ "step": 740
1112
+ },
1113
+ {
1114
+ "epoch": 1.48,
1115
+ "eval_loss": 0.7423770427703857,
1116
+ "eval_runtime": 2.9762,
1117
+ "eval_samples_per_second": 52.416,
1118
+ "eval_steps_per_second": 2.688,
1119
+ "step": 740
1120
+ },
1121
+ {
1122
+ "epoch": 1.5,
1123
+ "grad_norm": 4.1875,
1124
+ "learning_rate": 1.9848077530122083e-05,
1125
+ "loss": 0.5889,
1126
+ "step": 750
1127
+ },
1128
+ {
1129
+ "epoch": 1.5,
1130
+ "eval_loss": 0.7404635548591614,
1131
+ "eval_runtime": 2.9716,
1132
+ "eval_samples_per_second": 52.497,
1133
+ "eval_steps_per_second": 2.692,
1134
+ "step": 750
1135
+ },
1136
+ {
1137
+ "epoch": 1.52,
1138
+ "grad_norm": 2.546875,
1139
+ "learning_rate": 1.983571470813386e-05,
1140
+ "loss": 0.5561,
1141
+ "step": 760
1142
+ },
1143
+ {
1144
+ "epoch": 1.52,
1145
+ "eval_loss": 0.7413467764854431,
1146
+ "eval_runtime": 2.9834,
1147
+ "eval_samples_per_second": 52.29,
1148
+ "eval_steps_per_second": 2.682,
1149
+ "step": 760
1150
+ },
1151
+ {
1152
+ "epoch": 1.54,
1153
+ "grad_norm": 2.125,
1154
+ "learning_rate": 1.982287250728689e-05,
1155
+ "loss": 0.5307,
1156
+ "step": 770
1157
+ },
1158
+ {
1159
+ "epoch": 1.54,
1160
+ "eval_loss": 0.7405444383621216,
1161
+ "eval_runtime": 2.9797,
1162
+ "eval_samples_per_second": 52.355,
1163
+ "eval_steps_per_second": 2.685,
1164
+ "step": 770
1165
+ },
1166
+ {
1167
+ "epoch": 1.56,
1168
+ "grad_norm": 2.0625,
1169
+ "learning_rate": 1.9809551553491918e-05,
1170
+ "loss": 0.5464,
1171
+ "step": 780
1172
+ },
1173
+ {
1174
+ "epoch": 1.56,
1175
+ "eval_loss": 0.742436945438385,
1176
+ "eval_runtime": 2.973,
1177
+ "eval_samples_per_second": 52.471,
1178
+ "eval_steps_per_second": 2.691,
1179
+ "step": 780
1180
+ },
1181
+ {
1182
+ "epoch": 1.58,
1183
+ "grad_norm": 2.3125,
1184
+ "learning_rate": 1.979575249599344e-05,
1185
+ "loss": 0.5726,
1186
+ "step": 790
1187
+ },
1188
+ {
1189
+ "epoch": 1.58,
1190
+ "eval_loss": 0.7410677671432495,
1191
+ "eval_runtime": 2.968,
1192
+ "eval_samples_per_second": 52.561,
1193
+ "eval_steps_per_second": 2.695,
1194
+ "step": 790
1195
+ },
1196
+ {
1197
+ "epoch": 1.6,
1198
+ "grad_norm": 2.578125,
1199
+ "learning_rate": 1.9781476007338058e-05,
1200
+ "loss": 0.6039,
1201
+ "step": 800
1202
+ },
1203
+ {
1204
+ "epoch": 1.6,
1205
+ "eval_loss": 0.7446211576461792,
1206
+ "eval_runtime": 2.9788,
1207
+ "eval_samples_per_second": 52.37,
1208
+ "eval_steps_per_second": 2.686,
1209
+ "step": 800
1210
+ },
1211
+ {
1212
+ "epoch": 1.62,
1213
+ "grad_norm": 3.5,
1214
+ "learning_rate": 1.9766722783341682e-05,
1215
+ "loss": 0.5236,
1216
+ "step": 810
1217
+ },
1218
+ {
1219
+ "epoch": 1.62,
1220
+ "eval_loss": 0.7396934628486633,
1221
+ "eval_runtime": 2.9783,
1222
+ "eval_samples_per_second": 52.378,
1223
+ "eval_steps_per_second": 2.686,
1224
+ "step": 810
1225
+ },
1226
+ {
1227
+ "epoch": 1.64,
1228
+ "grad_norm": 3.703125,
1229
+ "learning_rate": 1.9751493543055634e-05,
1230
+ "loss": 0.5176,
1231
+ "step": 820
1232
+ },
1233
+ {
1234
+ "epoch": 1.64,
1235
+ "eval_loss": 0.7380808591842651,
1236
+ "eval_runtime": 2.9661,
1237
+ "eval_samples_per_second": 52.595,
1238
+ "eval_steps_per_second": 2.697,
1239
+ "step": 820
1240
+ },
1241
+ {
1242
+ "epoch": 1.66,
1243
+ "grad_norm": 2.25,
1244
+ "learning_rate": 1.9735789028731603e-05,
1245
+ "loss": 0.5991,
1246
+ "step": 830
1247
+ },
1248
+ {
1249
+ "epoch": 1.66,
1250
+ "eval_loss": 0.7352851629257202,
1251
+ "eval_runtime": 2.9801,
1252
+ "eval_samples_per_second": 52.348,
1253
+ "eval_steps_per_second": 2.685,
1254
+ "step": 830
1255
+ },
1256
+ {
1257
+ "epoch": 1.68,
1258
+ "grad_norm": 3.71875,
1259
+ "learning_rate": 1.9719610005785466e-05,
1260
+ "loss": 0.5451,
1261
+ "step": 840
1262
+ },
1263
+ {
1264
+ "epoch": 1.68,
1265
+ "eval_loss": 0.7372366786003113,
1266
+ "eval_runtime": 2.9742,
1267
+ "eval_samples_per_second": 52.452,
1268
+ "eval_steps_per_second": 2.69,
1269
+ "step": 840
1270
+ },
1271
+ {
1272
+ "epoch": 1.7,
1273
+ "grad_norm": 2.140625,
1274
+ "learning_rate": 1.9702957262759964e-05,
1275
+ "loss": 0.4484,
1276
+ "step": 850
1277
+ },
1278
+ {
1279
+ "epoch": 1.7,
1280
+ "eval_loss": 0.7352163791656494,
1281
+ "eval_runtime": 2.9703,
1282
+ "eval_samples_per_second": 52.52,
1283
+ "eval_steps_per_second": 2.693,
1284
+ "step": 850
1285
+ },
1286
+ {
1287
+ "epoch": 1.72,
1288
+ "grad_norm": 2.125,
1289
+ "learning_rate": 1.9685831611286312e-05,
1290
+ "loss": 0.5568,
1291
+ "step": 860
1292
+ },
1293
+ {
1294
+ "epoch": 1.72,
1295
+ "eval_loss": 0.7367239594459534,
1296
+ "eval_runtime": 2.9748,
1297
+ "eval_samples_per_second": 52.441,
1298
+ "eval_steps_per_second": 2.689,
1299
+ "step": 860
1300
+ },
1301
+ {
1302
+ "epoch": 1.74,
1303
+ "grad_norm": 2.890625,
1304
+ "learning_rate": 1.9668233886044597e-05,
1305
+ "loss": 0.5643,
1306
+ "step": 870
1307
+ },
1308
+ {
1309
+ "epoch": 1.74,
1310
+ "eval_loss": 0.7365850806236267,
1311
+ "eval_runtime": 2.9727,
1312
+ "eval_samples_per_second": 52.477,
1313
+ "eval_steps_per_second": 2.691,
1314
+ "step": 870
1315
+ },
1316
+ {
1317
+ "epoch": 1.76,
1318
+ "grad_norm": 3.3125,
1319
+ "learning_rate": 1.9650164944723116e-05,
1320
+ "loss": 0.5341,
1321
+ "step": 880
1322
+ },
1323
+ {
1324
+ "epoch": 1.76,
1325
+ "eval_loss": 0.7373999357223511,
1326
+ "eval_runtime": 2.9825,
1327
+ "eval_samples_per_second": 52.305,
1328
+ "eval_steps_per_second": 2.682,
1329
+ "step": 880
1330
+ },
1331
+ {
1332
+ "epoch": 1.78,
1333
+ "grad_norm": 3.625,
1334
+ "learning_rate": 1.9631625667976584e-05,
1335
+ "loss": 0.5255,
1336
+ "step": 890
1337
+ },
1338
+ {
1339
+ "epoch": 1.78,
1340
+ "eval_loss": 0.7373459935188293,
1341
+ "eval_runtime": 2.9729,
1342
+ "eval_samples_per_second": 52.474,
1343
+ "eval_steps_per_second": 2.691,
1344
+ "step": 890
1345
+ },
1346
+ {
1347
+ "epoch": 1.8,
1348
+ "grad_norm": 3.296875,
1349
+ "learning_rate": 1.961261695938319e-05,
1350
+ "loss": 0.5351,
1351
+ "step": 900
1352
+ },
1353
+ {
1354
+ "epoch": 1.8,
1355
+ "eval_loss": 0.7352133393287659,
1356
+ "eval_runtime": 2.9749,
1357
+ "eval_samples_per_second": 52.439,
1358
+ "eval_steps_per_second": 2.689,
1359
+ "step": 900
1360
+ },
1361
+ {
1362
+ "epoch": 1.82,
1363
+ "grad_norm": 3.15625,
1364
+ "learning_rate": 1.9593139745400575e-05,
1365
+ "loss": 0.6188,
1366
+ "step": 910
1367
+ },
1368
+ {
1369
+ "epoch": 1.82,
1370
+ "eval_loss": 0.7367925643920898,
1371
+ "eval_runtime": 2.9685,
1372
+ "eval_samples_per_second": 52.552,
1373
+ "eval_steps_per_second": 2.695,
1374
+ "step": 910
1375
+ },
1376
+ {
1377
+ "epoch": 1.84,
1378
+ "grad_norm": 2.71875,
1379
+ "learning_rate": 1.9573194975320672e-05,
1380
+ "loss": 0.4968,
1381
+ "step": 920
1382
+ },
1383
+ {
1384
+ "epoch": 1.84,
1385
+ "eval_loss": 0.7363703846931458,
1386
+ "eval_runtime": 2.9748,
1387
+ "eval_samples_per_second": 52.441,
1388
+ "eval_steps_per_second": 2.689,
1389
+ "step": 920
1390
+ },
1391
+ {
1392
+ "epoch": 1.86,
1393
+ "grad_norm": 2.5625,
1394
+ "learning_rate": 1.9552783621223437e-05,
1395
+ "loss": 0.543,
1396
+ "step": 930
1397
+ },
1398
+ {
1399
+ "epoch": 1.86,
1400
+ "eval_loss": 0.7352641224861145,
1401
+ "eval_runtime": 2.9728,
1402
+ "eval_samples_per_second": 52.476,
1403
+ "eval_steps_per_second": 2.691,
1404
+ "step": 930
1405
+ },
1406
+ {
1407
+ "epoch": 1.88,
1408
+ "grad_norm": 2.921875,
1409
+ "learning_rate": 1.9531906677929472e-05,
1410
+ "loss": 0.6299,
1411
+ "step": 940
1412
+ },
1413
+ {
1414
+ "epoch": 1.88,
1415
+ "eval_loss": 0.7344778776168823,
1416
+ "eval_runtime": 2.994,
1417
+ "eval_samples_per_second": 52.104,
1418
+ "eval_steps_per_second": 2.672,
1419
+ "step": 940
1420
+ },
1421
+ {
1422
+ "epoch": 1.9,
1423
+ "grad_norm": 2.703125,
1424
+ "learning_rate": 1.9510565162951538e-05,
1425
+ "loss": 0.6208,
1426
+ "step": 950
1427
+ },
1428
+ {
1429
+ "epoch": 1.9,
1430
+ "eval_loss": 0.7322725057601929,
1431
+ "eval_runtime": 2.9757,
1432
+ "eval_samples_per_second": 52.425,
1433
+ "eval_steps_per_second": 2.688,
1434
+ "step": 950
1435
+ },
1436
+ {
1437
+ "epoch": 1.92,
1438
+ "grad_norm": 2.453125,
1439
+ "learning_rate": 1.9488760116444966e-05,
1440
+ "loss": 0.5667,
1441
+ "step": 960
1442
+ },
1443
+ {
1444
+ "epoch": 1.92,
1445
+ "eval_loss": 0.7319491505622864,
1446
+ "eval_runtime": 2.9677,
1447
+ "eval_samples_per_second": 52.565,
1448
+ "eval_steps_per_second": 2.696,
1449
+ "step": 960
1450
+ },
1451
+ {
1452
+ "epoch": 1.94,
1453
+ "grad_norm": 3.0,
1454
+ "learning_rate": 1.9466492601156964e-05,
1455
+ "loss": 0.5686,
1456
+ "step": 970
1457
+ },
1458
+ {
1459
+ "epoch": 1.94,
1460
+ "eval_loss": 0.7332378625869751,
1461
+ "eval_runtime": 2.9718,
1462
+ "eval_samples_per_second": 52.494,
1463
+ "eval_steps_per_second": 2.692,
1464
+ "step": 970
1465
+ },
1466
+ {
1467
+ "epoch": 1.96,
1468
+ "grad_norm": 2.890625,
1469
+ "learning_rate": 1.944376370237481e-05,
1470
+ "loss": 0.5745,
1471
+ "step": 980
1472
+ },
1473
+ {
1474
+ "epoch": 1.96,
1475
+ "eval_loss": 0.7319948077201843,
1476
+ "eval_runtime": 2.9667,
1477
+ "eval_samples_per_second": 52.584,
1478
+ "eval_steps_per_second": 2.697,
1479
+ "step": 980
1480
+ },
1481
+ {
1482
+ "epoch": 1.98,
1483
+ "grad_norm": 2.71875,
1484
+ "learning_rate": 1.942057452787297e-05,
1485
+ "loss": 0.5632,
1486
+ "step": 990
1487
+ },
1488
+ {
1489
+ "epoch": 1.98,
1490
+ "eval_loss": 0.7325617074966431,
1491
+ "eval_runtime": 2.9757,
1492
+ "eval_samples_per_second": 52.425,
1493
+ "eval_steps_per_second": 2.688,
1494
+ "step": 990
1495
+ },
1496
+ {
1497
+ "epoch": 2.0,
1498
+ "grad_norm": 2.84375,
1499
+ "learning_rate": 1.9396926207859085e-05,
1500
+ "loss": 0.5558,
1501
+ "step": 1000
1502
+ },
1503
+ {
1504
+ "epoch": 2.0,
1505
+ "eval_loss": 0.7342491149902344,
1506
+ "eval_runtime": 2.9766,
1507
+ "eval_samples_per_second": 52.41,
1508
+ "eval_steps_per_second": 2.688,
1509
+ "step": 1000
1510
+ }
1511
+ ],
1512
+ "logging_steps": 10,
1513
+ "max_steps": 5000,
1514
+ "num_input_tokens_seen": 0,
1515
+ "num_train_epochs": 10,
1516
+ "save_steps": 500,
1517
+ "total_flos": 5.34460687754199e+16,
1518
+ "train_batch_size": 2,
1519
+ "trial_name": null,
1520
+ "trial_params": null
1521
+ }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:39d17cc997a441afa9f157f8181af6fd34321ef5f93ab005c5aa7fa59ffd6fab
3
- size 6392
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b6718b54d2889093102adabbeac5758f46fd9ed04669e1fad29d69d5343cbe3
3
+ size 5112