Update classifier weights and label config
Browse files- checkpoint-5500/added_tokens.json +3 -0
- checkpoint-5500/model.safetensors +3 -0
- checkpoint-5500/optimizer.pt +3 -0
- checkpoint-5500/rng_state.pth +3 -0
- checkpoint-5500/scheduler.pt +3 -0
- checkpoint-5500/special_tokens_map.json +15 -0
- checkpoint-5500/spm.model +3 -0
- checkpoint-5500/tokenizer.json +0 -0
- checkpoint-5500/tokenizer_config.json +59 -0
- checkpoint-5500/trainer_state.json +1959 -0
- checkpoint-5500/training_args.bin +3 -0
- checkpoint-5541/added_tokens.json +3 -0
- checkpoint-5541/model.safetensors +3 -0
- checkpoint-5541/optimizer.pt +3 -0
- checkpoint-5541/rng_state.pth +3 -0
- checkpoint-5541/scheduler.pt +3 -0
- checkpoint-5541/special_tokens_map.json +15 -0
- checkpoint-5541/spm.model +3 -0
- checkpoint-5541/tokenizer.json +0 -0
- checkpoint-5541/tokenizer_config.json +59 -0
- checkpoint-5541/trainer_state.json +1973 -0
- checkpoint-5541/training_args.bin +3 -0
- model.safetensors +2 -2
- training_args.bin +1 -1
checkpoint-5500/added_tokens.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[MASK]": 128000
|
3 |
+
}
|
checkpoint-5500/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4f5c4739f5f26bad57099620123c44397bec8f0eb7f170819d28d867c36c420c
|
3 |
+
size 736593656
|
checkpoint-5500/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d423f368de91513666c8ac3732b3197a8c6f5f8b3cf9218b26c99812ba5d0c4f
|
3 |
+
size 1473311563
|
checkpoint-5500/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:423c3a04c39691e8158243c93ccc6132af1697bbe789ac7748ff3ad1b6110c5b
|
3 |
+
size 14645
|
checkpoint-5500/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:093b671269028e040ef3879754a3025d10c6372d98679638fcb877a22bea5ddf
|
3 |
+
size 1465
|
checkpoint-5500/special_tokens_map.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "[SEP]",
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"pad_token": "[PAD]",
|
7 |
+
"sep_token": "[SEP]",
|
8 |
+
"unk_token": {
|
9 |
+
"content": "[UNK]",
|
10 |
+
"lstrip": false,
|
11 |
+
"normalized": true,
|
12 |
+
"rstrip": false,
|
13 |
+
"single_word": false
|
14 |
+
}
|
15 |
+
}
|
checkpoint-5500/spm.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
3 |
+
size 2464616
|
checkpoint-5500/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-5500/tokenizer_config.json
ADDED
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "[CLS]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "[SEP]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "[UNK]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": true,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"128000": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"bos_token": "[CLS]",
|
45 |
+
"clean_up_tokenization_spaces": false,
|
46 |
+
"cls_token": "[CLS]",
|
47 |
+
"do_lower_case": false,
|
48 |
+
"eos_token": "[SEP]",
|
49 |
+
"extra_special_tokens": {},
|
50 |
+
"mask_token": "[MASK]",
|
51 |
+
"model_max_length": 1000000000000000019884624838656,
|
52 |
+
"pad_token": "[PAD]",
|
53 |
+
"sep_token": "[SEP]",
|
54 |
+
"sp_model_kwargs": {},
|
55 |
+
"split_by_punct": false,
|
56 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
57 |
+
"unk_token": "[UNK]",
|
58 |
+
"vocab_type": "spm"
|
59 |
+
}
|
checkpoint-5500/trainer_state.json
ADDED
@@ -0,0 +1,1959 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": null,
|
3 |
+
"best_metric": null,
|
4 |
+
"best_model_checkpoint": null,
|
5 |
+
"epoch": 2.977801840822956,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 5500,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.010828370330265295,
|
14 |
+
"grad_norm": 8.006661415100098,
|
15 |
+
"learning_rate": 1.9931420321241655e-05,
|
16 |
+
"loss": 4.0423,
|
17 |
+
"step": 20
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.02165674066053059,
|
21 |
+
"grad_norm": 1.674608588218689,
|
22 |
+
"learning_rate": 1.9859231185706552e-05,
|
23 |
+
"loss": 1.275,
|
24 |
+
"step": 40
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.03248511099079589,
|
28 |
+
"grad_norm": 1.2871980667114258,
|
29 |
+
"learning_rate": 1.978704205017145e-05,
|
30 |
+
"loss": 0.879,
|
31 |
+
"step": 60
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 0.04331348132106118,
|
35 |
+
"grad_norm": 1.0388418436050415,
|
36 |
+
"learning_rate": 1.971485291463635e-05,
|
37 |
+
"loss": 0.6456,
|
38 |
+
"step": 80
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.05414185165132648,
|
42 |
+
"grad_norm": 0.8618717193603516,
|
43 |
+
"learning_rate": 1.9642663779101247e-05,
|
44 |
+
"loss": 0.4888,
|
45 |
+
"step": 100
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.06497022198159177,
|
49 |
+
"grad_norm": 0.7138980627059937,
|
50 |
+
"learning_rate": 1.9570474643566144e-05,
|
51 |
+
"loss": 0.382,
|
52 |
+
"step": 120
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 0.07579859231185707,
|
56 |
+
"grad_norm": 0.6197817921638489,
|
57 |
+
"learning_rate": 1.9498285508031044e-05,
|
58 |
+
"loss": 0.3027,
|
59 |
+
"step": 140
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 0.08662696264212236,
|
63 |
+
"grad_norm": 0.5373287796974182,
|
64 |
+
"learning_rate": 1.942609637249594e-05,
|
65 |
+
"loss": 0.247,
|
66 |
+
"step": 160
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.09745533297238766,
|
70 |
+
"grad_norm": 0.46481892466545105,
|
71 |
+
"learning_rate": 1.935390723696084e-05,
|
72 |
+
"loss": 0.2073,
|
73 |
+
"step": 180
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 0.10828370330265295,
|
77 |
+
"grad_norm": 0.40900182723999023,
|
78 |
+
"learning_rate": 1.9281718101425736e-05,
|
79 |
+
"loss": 0.1771,
|
80 |
+
"step": 200
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.11911207363291824,
|
84 |
+
"grad_norm": 0.35755395889282227,
|
85 |
+
"learning_rate": 1.9209528965890636e-05,
|
86 |
+
"loss": 0.151,
|
87 |
+
"step": 220
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.12994044396318355,
|
91 |
+
"grad_norm": 0.32475224137306213,
|
92 |
+
"learning_rate": 1.9137339830355533e-05,
|
93 |
+
"loss": 0.1327,
|
94 |
+
"step": 240
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.14076881429344884,
|
98 |
+
"grad_norm": 0.2906845211982727,
|
99 |
+
"learning_rate": 1.906515069482043e-05,
|
100 |
+
"loss": 0.1179,
|
101 |
+
"step": 260
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 0.15159718462371413,
|
105 |
+
"grad_norm": 0.26396697759628296,
|
106 |
+
"learning_rate": 1.899296155928533e-05,
|
107 |
+
"loss": 0.1058,
|
108 |
+
"step": 280
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 0.16242555495397942,
|
112 |
+
"grad_norm": 0.24280692636966705,
|
113 |
+
"learning_rate": 1.8920772423750228e-05,
|
114 |
+
"loss": 0.0963,
|
115 |
+
"step": 300
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 0.17325392528424471,
|
119 |
+
"grad_norm": 0.21794995665550232,
|
120 |
+
"learning_rate": 1.8848583288215125e-05,
|
121 |
+
"loss": 0.0851,
|
122 |
+
"step": 320
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.18408229561451,
|
126 |
+
"grad_norm": 0.2041829228401184,
|
127 |
+
"learning_rate": 1.8776394152680025e-05,
|
128 |
+
"loss": 0.0786,
|
129 |
+
"step": 340
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.19491066594477532,
|
133 |
+
"grad_norm": 0.18174469470977783,
|
134 |
+
"learning_rate": 1.870420501714492e-05,
|
135 |
+
"loss": 0.0731,
|
136 |
+
"step": 360
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 0.20573903627504062,
|
140 |
+
"grad_norm": 0.17271022498607635,
|
141 |
+
"learning_rate": 1.863201588160982e-05,
|
142 |
+
"loss": 0.0676,
|
143 |
+
"step": 380
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 0.2165674066053059,
|
147 |
+
"grad_norm": 0.1657329499721527,
|
148 |
+
"learning_rate": 1.8559826746074717e-05,
|
149 |
+
"loss": 0.0616,
|
150 |
+
"step": 400
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 0.2273957769355712,
|
154 |
+
"grad_norm": 0.16118521988391876,
|
155 |
+
"learning_rate": 1.8487637610539614e-05,
|
156 |
+
"loss": 0.0594,
|
157 |
+
"step": 420
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 0.2382241472658365,
|
161 |
+
"grad_norm": 0.14374467730522156,
|
162 |
+
"learning_rate": 1.8415448475004514e-05,
|
163 |
+
"loss": 0.0542,
|
164 |
+
"step": 440
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 0.24905251759610178,
|
168 |
+
"grad_norm": 0.1439092755317688,
|
169 |
+
"learning_rate": 1.834325933946941e-05,
|
170 |
+
"loss": 0.0512,
|
171 |
+
"step": 460
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.2598808879263671,
|
175 |
+
"grad_norm": 0.1367517113685608,
|
176 |
+
"learning_rate": 1.827107020393431e-05,
|
177 |
+
"loss": 0.0482,
|
178 |
+
"step": 480
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.2707092582566324,
|
182 |
+
"grad_norm": 0.13729216158390045,
|
183 |
+
"learning_rate": 1.819888106839921e-05,
|
184 |
+
"loss": 0.0458,
|
185 |
+
"step": 500
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 0.2815376285868977,
|
189 |
+
"grad_norm": 0.11813738197088242,
|
190 |
+
"learning_rate": 1.8126691932864103e-05,
|
191 |
+
"loss": 0.0442,
|
192 |
+
"step": 520
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 0.292365998917163,
|
196 |
+
"grad_norm": 0.10173018276691437,
|
197 |
+
"learning_rate": 1.8054502797329003e-05,
|
198 |
+
"loss": 0.0397,
|
199 |
+
"step": 540
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 0.30319436924742826,
|
203 |
+
"grad_norm": 0.10882216691970825,
|
204 |
+
"learning_rate": 1.79823136617939e-05,
|
205 |
+
"loss": 0.0377,
|
206 |
+
"step": 560
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.31402273957769355,
|
210 |
+
"grad_norm": 0.10021235793828964,
|
211 |
+
"learning_rate": 1.7910124526258797e-05,
|
212 |
+
"loss": 0.0366,
|
213 |
+
"step": 580
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.32485110990795885,
|
217 |
+
"grad_norm": 0.11167757213115692,
|
218 |
+
"learning_rate": 1.7837935390723698e-05,
|
219 |
+
"loss": 0.0354,
|
220 |
+
"step": 600
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 0.33567948023822414,
|
224 |
+
"grad_norm": 0.11719091236591339,
|
225 |
+
"learning_rate": 1.7765746255188595e-05,
|
226 |
+
"loss": 0.0361,
|
227 |
+
"step": 620
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 0.34650785056848943,
|
231 |
+
"grad_norm": 0.08965040743350983,
|
232 |
+
"learning_rate": 1.7693557119653492e-05,
|
233 |
+
"loss": 0.0328,
|
234 |
+
"step": 640
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.3573362208987547,
|
238 |
+
"grad_norm": 0.09095824509859085,
|
239 |
+
"learning_rate": 1.7621367984118393e-05,
|
240 |
+
"loss": 0.0324,
|
241 |
+
"step": 660
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.36816459122902,
|
245 |
+
"grad_norm": 0.09133260697126389,
|
246 |
+
"learning_rate": 1.754917884858329e-05,
|
247 |
+
"loss": 0.0305,
|
248 |
+
"step": 680
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 0.3789929615592853,
|
252 |
+
"grad_norm": 0.10949356853961945,
|
253 |
+
"learning_rate": 1.7476989713048187e-05,
|
254 |
+
"loss": 0.0298,
|
255 |
+
"step": 700
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 0.38982133188955065,
|
259 |
+
"grad_norm": 0.09127330780029297,
|
260 |
+
"learning_rate": 1.7404800577513087e-05,
|
261 |
+
"loss": 0.0283,
|
262 |
+
"step": 720
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 0.40064970221981594,
|
266 |
+
"grad_norm": 0.07634657621383667,
|
267 |
+
"learning_rate": 1.7332611441977984e-05,
|
268 |
+
"loss": 0.027,
|
269 |
+
"step": 740
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 0.41147807255008123,
|
273 |
+
"grad_norm": 0.1150558814406395,
|
274 |
+
"learning_rate": 1.726042230644288e-05,
|
275 |
+
"loss": 0.0256,
|
276 |
+
"step": 760
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 0.4223064428803465,
|
280 |
+
"grad_norm": 0.07289394736289978,
|
281 |
+
"learning_rate": 1.718823317090778e-05,
|
282 |
+
"loss": 0.0241,
|
283 |
+
"step": 780
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 0.4331348132106118,
|
287 |
+
"grad_norm": 0.07605121284723282,
|
288 |
+
"learning_rate": 1.711604403537268e-05,
|
289 |
+
"loss": 0.0241,
|
290 |
+
"step": 800
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 0.4439631835408771,
|
294 |
+
"grad_norm": 0.07429105043411255,
|
295 |
+
"learning_rate": 1.7043854899837576e-05,
|
296 |
+
"loss": 0.0233,
|
297 |
+
"step": 820
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 0.4547915538711424,
|
301 |
+
"grad_norm": 0.06406375020742416,
|
302 |
+
"learning_rate": 1.6971665764302473e-05,
|
303 |
+
"loss": 0.023,
|
304 |
+
"step": 840
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 0.4656199242014077,
|
308 |
+
"grad_norm": 0.06342233717441559,
|
309 |
+
"learning_rate": 1.6899476628767374e-05,
|
310 |
+
"loss": 0.0202,
|
311 |
+
"step": 860
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 0.476448294531673,
|
315 |
+
"grad_norm": 0.06463098526000977,
|
316 |
+
"learning_rate": 1.682728749323227e-05,
|
317 |
+
"loss": 0.0204,
|
318 |
+
"step": 880
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 0.48727666486193827,
|
322 |
+
"grad_norm": 0.05809445306658745,
|
323 |
+
"learning_rate": 1.6755098357697168e-05,
|
324 |
+
"loss": 0.0192,
|
325 |
+
"step": 900
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 0.49810503519220356,
|
329 |
+
"grad_norm": 0.06424301117658615,
|
330 |
+
"learning_rate": 1.668290922216207e-05,
|
331 |
+
"loss": 0.0193,
|
332 |
+
"step": 920
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 0.5089334055224689,
|
336 |
+
"grad_norm": 0.07586075365543365,
|
337 |
+
"learning_rate": 1.6610720086626962e-05,
|
338 |
+
"loss": 0.0188,
|
339 |
+
"step": 940
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.5197617758527342,
|
343 |
+
"grad_norm": 0.05757075175642967,
|
344 |
+
"learning_rate": 1.6538530951091863e-05,
|
345 |
+
"loss": 0.018,
|
346 |
+
"step": 960
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 0.5305901461829995,
|
350 |
+
"grad_norm": 0.05930281803011894,
|
351 |
+
"learning_rate": 1.646634181555676e-05,
|
352 |
+
"loss": 0.0157,
|
353 |
+
"step": 980
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 0.5414185165132648,
|
357 |
+
"grad_norm": 0.04882697016000748,
|
358 |
+
"learning_rate": 1.6394152680021657e-05,
|
359 |
+
"loss": 0.016,
|
360 |
+
"step": 1000
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 0.5522468868435301,
|
364 |
+
"grad_norm": 0.047697920352220535,
|
365 |
+
"learning_rate": 1.6321963544486557e-05,
|
366 |
+
"loss": 0.0161,
|
367 |
+
"step": 1020
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 0.5630752571737954,
|
371 |
+
"grad_norm": 0.0575709231197834,
|
372 |
+
"learning_rate": 1.6249774408951454e-05,
|
373 |
+
"loss": 0.0144,
|
374 |
+
"step": 1040
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 0.5739036275040607,
|
378 |
+
"grad_norm": 0.07101713865995407,
|
379 |
+
"learning_rate": 1.617758527341635e-05,
|
380 |
+
"loss": 0.0146,
|
381 |
+
"step": 1060
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 0.584731997834326,
|
385 |
+
"grad_norm": 0.04585053771734238,
|
386 |
+
"learning_rate": 1.6105396137881252e-05,
|
387 |
+
"loss": 0.0138,
|
388 |
+
"step": 1080
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 0.5955603681645912,
|
392 |
+
"grad_norm": 0.05117352306842804,
|
393 |
+
"learning_rate": 1.6033207002346146e-05,
|
394 |
+
"loss": 0.0126,
|
395 |
+
"step": 1100
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 0.6063887384948565,
|
399 |
+
"grad_norm": 0.047820210456848145,
|
400 |
+
"learning_rate": 1.5961017866811046e-05,
|
401 |
+
"loss": 0.0129,
|
402 |
+
"step": 1120
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 0.6172171088251218,
|
406 |
+
"grad_norm": 0.04630187153816223,
|
407 |
+
"learning_rate": 1.5888828731275943e-05,
|
408 |
+
"loss": 0.0129,
|
409 |
+
"step": 1140
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"epoch": 0.6280454791553871,
|
413 |
+
"grad_norm": 0.04581563547253609,
|
414 |
+
"learning_rate": 1.581663959574084e-05,
|
415 |
+
"loss": 0.013,
|
416 |
+
"step": 1160
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 0.6388738494856524,
|
420 |
+
"grad_norm": 0.04178199544548988,
|
421 |
+
"learning_rate": 1.574445046020574e-05,
|
422 |
+
"loss": 0.0118,
|
423 |
+
"step": 1180
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 0.6497022198159177,
|
427 |
+
"grad_norm": 0.0400426909327507,
|
428 |
+
"learning_rate": 1.5672261324670638e-05,
|
429 |
+
"loss": 0.0117,
|
430 |
+
"step": 1200
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 0.660530590146183,
|
434 |
+
"grad_norm": 0.03873790055513382,
|
435 |
+
"learning_rate": 1.5600072189135535e-05,
|
436 |
+
"loss": 0.0114,
|
437 |
+
"step": 1220
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 0.6713589604764483,
|
441 |
+
"grad_norm": 0.05929722636938095,
|
442 |
+
"learning_rate": 1.5527883053600436e-05,
|
443 |
+
"loss": 0.0113,
|
444 |
+
"step": 1240
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 0.6821873308067136,
|
448 |
+
"grad_norm": 0.03565024584531784,
|
449 |
+
"learning_rate": 1.5455693918065333e-05,
|
450 |
+
"loss": 0.0103,
|
451 |
+
"step": 1260
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 0.6930157011369789,
|
455 |
+
"grad_norm": 0.04300345107913017,
|
456 |
+
"learning_rate": 1.538350478253023e-05,
|
457 |
+
"loss": 0.0109,
|
458 |
+
"step": 1280
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 0.7038440714672441,
|
462 |
+
"grad_norm": 0.04502050206065178,
|
463 |
+
"learning_rate": 1.5311315646995127e-05,
|
464 |
+
"loss": 0.0102,
|
465 |
+
"step": 1300
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 0.7146724417975094,
|
469 |
+
"grad_norm": 0.0350734144449234,
|
470 |
+
"learning_rate": 1.5239126511460027e-05,
|
471 |
+
"loss": 0.01,
|
472 |
+
"step": 1320
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"epoch": 0.7255008121277747,
|
476 |
+
"grad_norm": 0.03979852795600891,
|
477 |
+
"learning_rate": 1.5166937375924925e-05,
|
478 |
+
"loss": 0.0093,
|
479 |
+
"step": 1340
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 0.73632918245804,
|
483 |
+
"grad_norm": 0.032689835876226425,
|
484 |
+
"learning_rate": 1.5094748240389823e-05,
|
485 |
+
"loss": 0.0093,
|
486 |
+
"step": 1360
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 0.7471575527883053,
|
490 |
+
"grad_norm": 0.03613670915365219,
|
491 |
+
"learning_rate": 1.5022559104854722e-05,
|
492 |
+
"loss": 0.0089,
|
493 |
+
"step": 1380
|
494 |
+
},
|
495 |
+
{
|
496 |
+
"epoch": 0.7579859231185706,
|
497 |
+
"grad_norm": 0.034867897629737854,
|
498 |
+
"learning_rate": 1.4950369969319618e-05,
|
499 |
+
"loss": 0.0089,
|
500 |
+
"step": 1400
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 0.7688142934488359,
|
504 |
+
"grad_norm": 0.0316634476184845,
|
505 |
+
"learning_rate": 1.4878180833784516e-05,
|
506 |
+
"loss": 0.0083,
|
507 |
+
"step": 1420
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 0.7796426637791013,
|
511 |
+
"grad_norm": 0.030254648998379707,
|
512 |
+
"learning_rate": 1.4805991698249415e-05,
|
513 |
+
"loss": 0.0081,
|
514 |
+
"step": 1440
|
515 |
+
},
|
516 |
+
{
|
517 |
+
"epoch": 0.7904710341093666,
|
518 |
+
"grad_norm": 0.031286563724279404,
|
519 |
+
"learning_rate": 1.4733802562714312e-05,
|
520 |
+
"loss": 0.0084,
|
521 |
+
"step": 1460
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 0.8012994044396319,
|
525 |
+
"grad_norm": 0.04057060182094574,
|
526 |
+
"learning_rate": 1.4661613427179211e-05,
|
527 |
+
"loss": 0.0083,
|
528 |
+
"step": 1480
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 0.8121277747698972,
|
532 |
+
"grad_norm": 0.027568308636546135,
|
533 |
+
"learning_rate": 1.458942429164411e-05,
|
534 |
+
"loss": 0.0079,
|
535 |
+
"step": 1500
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"epoch": 0.8229561451001625,
|
539 |
+
"grad_norm": 0.0294258501380682,
|
540 |
+
"learning_rate": 1.4517235156109007e-05,
|
541 |
+
"loss": 0.0079,
|
542 |
+
"step": 1520
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 0.8337845154304278,
|
546 |
+
"grad_norm": 0.023077772930264473,
|
547 |
+
"learning_rate": 1.4445046020573906e-05,
|
548 |
+
"loss": 0.0075,
|
549 |
+
"step": 1540
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 0.844612885760693,
|
553 |
+
"grad_norm": 0.02644222415983677,
|
554 |
+
"learning_rate": 1.4372856885038805e-05,
|
555 |
+
"loss": 0.0071,
|
556 |
+
"step": 1560
|
557 |
+
},
|
558 |
+
{
|
559 |
+
"epoch": 0.8554412560909583,
|
560 |
+
"grad_norm": 0.027747342362999916,
|
561 |
+
"learning_rate": 1.43006677495037e-05,
|
562 |
+
"loss": 0.0073,
|
563 |
+
"step": 1580
|
564 |
+
},
|
565 |
+
{
|
566 |
+
"epoch": 0.8662696264212236,
|
567 |
+
"grad_norm": 0.02826724573969841,
|
568 |
+
"learning_rate": 1.4228478613968599e-05,
|
569 |
+
"loss": 0.0075,
|
570 |
+
"step": 1600
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 0.8770979967514889,
|
574 |
+
"grad_norm": 0.04404623061418533,
|
575 |
+
"learning_rate": 1.4156289478433496e-05,
|
576 |
+
"loss": 0.0067,
|
577 |
+
"step": 1620
|
578 |
+
},
|
579 |
+
{
|
580 |
+
"epoch": 0.8879263670817542,
|
581 |
+
"grad_norm": 0.032218191772699356,
|
582 |
+
"learning_rate": 1.4084100342898395e-05,
|
583 |
+
"loss": 0.0069,
|
584 |
+
"step": 1640
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 0.8987547374120195,
|
588 |
+
"grad_norm": 0.0255202054977417,
|
589 |
+
"learning_rate": 1.4011911207363293e-05,
|
590 |
+
"loss": 0.0065,
|
591 |
+
"step": 1660
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 0.9095831077422848,
|
595 |
+
"grad_norm": 0.027019904926419258,
|
596 |
+
"learning_rate": 1.393972207182819e-05,
|
597 |
+
"loss": 0.0067,
|
598 |
+
"step": 1680
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 0.9204114780725501,
|
602 |
+
"grad_norm": 0.032661229372024536,
|
603 |
+
"learning_rate": 1.386753293629309e-05,
|
604 |
+
"loss": 0.0063,
|
605 |
+
"step": 1700
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 0.9312398484028154,
|
609 |
+
"grad_norm": 0.02468557097017765,
|
610 |
+
"learning_rate": 1.3795343800757988e-05,
|
611 |
+
"loss": 0.0064,
|
612 |
+
"step": 1720
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 0.9420682187330807,
|
616 |
+
"grad_norm": 0.02315792441368103,
|
617 |
+
"learning_rate": 1.3723154665222884e-05,
|
618 |
+
"loss": 0.006,
|
619 |
+
"step": 1740
|
620 |
+
},
|
621 |
+
{
|
622 |
+
"epoch": 0.952896589063346,
|
623 |
+
"grad_norm": 0.02668326534330845,
|
624 |
+
"learning_rate": 1.3650965529687782e-05,
|
625 |
+
"loss": 0.0059,
|
626 |
+
"step": 1760
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 0.9637249593936112,
|
630 |
+
"grad_norm": 0.03477375581860542,
|
631 |
+
"learning_rate": 1.3578776394152681e-05,
|
632 |
+
"loss": 0.0064,
|
633 |
+
"step": 1780
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 0.9745533297238765,
|
637 |
+
"grad_norm": 0.03006615862250328,
|
638 |
+
"learning_rate": 1.3506587258617578e-05,
|
639 |
+
"loss": 0.0061,
|
640 |
+
"step": 1800
|
641 |
+
},
|
642 |
+
{
|
643 |
+
"epoch": 0.9853817000541418,
|
644 |
+
"grad_norm": 0.024881117045879364,
|
645 |
+
"learning_rate": 1.3434398123082477e-05,
|
646 |
+
"loss": 0.0056,
|
647 |
+
"step": 1820
|
648 |
+
},
|
649 |
+
{
|
650 |
+
"epoch": 0.9962100703844071,
|
651 |
+
"grad_norm": 0.019141867756843567,
|
652 |
+
"learning_rate": 1.3362208987547376e-05,
|
653 |
+
"loss": 0.0057,
|
654 |
+
"step": 1840
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 1.0070384407146724,
|
658 |
+
"grad_norm": 0.03261445090174675,
|
659 |
+
"learning_rate": 1.3290019852012273e-05,
|
660 |
+
"loss": 0.0053,
|
661 |
+
"step": 1860
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"epoch": 1.0178668110449378,
|
665 |
+
"grad_norm": 0.022652000188827515,
|
666 |
+
"learning_rate": 1.3217830716477172e-05,
|
667 |
+
"loss": 0.0054,
|
668 |
+
"step": 1880
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 1.028695181375203,
|
672 |
+
"grad_norm": 0.02832789532840252,
|
673 |
+
"learning_rate": 1.314564158094207e-05,
|
674 |
+
"loss": 0.0055,
|
675 |
+
"step": 1900
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 1.0395235517054684,
|
679 |
+
"grad_norm": 0.019258731976151466,
|
680 |
+
"learning_rate": 1.3073452445406966e-05,
|
681 |
+
"loss": 0.0054,
|
682 |
+
"step": 1920
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 1.0503519220357336,
|
686 |
+
"grad_norm": 0.017714599147439003,
|
687 |
+
"learning_rate": 1.3001263309871865e-05,
|
688 |
+
"loss": 0.0053,
|
689 |
+
"step": 1940
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 1.061180292365999,
|
693 |
+
"grad_norm": 0.02675594761967659,
|
694 |
+
"learning_rate": 1.2929074174336763e-05,
|
695 |
+
"loss": 0.0049,
|
696 |
+
"step": 1960
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 1.0720086626962642,
|
700 |
+
"grad_norm": 0.018239008262753487,
|
701 |
+
"learning_rate": 1.285688503880166e-05,
|
702 |
+
"loss": 0.0051,
|
703 |
+
"step": 1980
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 1.0828370330265296,
|
707 |
+
"grad_norm": 0.024490129202604294,
|
708 |
+
"learning_rate": 1.278469590326656e-05,
|
709 |
+
"loss": 0.005,
|
710 |
+
"step": 2000
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 1.0936654033567947,
|
714 |
+
"grad_norm": 0.02328975684940815,
|
715 |
+
"learning_rate": 1.2712506767731458e-05,
|
716 |
+
"loss": 0.0049,
|
717 |
+
"step": 2020
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 1.1044937736870601,
|
721 |
+
"grad_norm": 0.027897818014025688,
|
722 |
+
"learning_rate": 1.2640317632196355e-05,
|
723 |
+
"loss": 0.005,
|
724 |
+
"step": 2040
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 1.1153221440173253,
|
728 |
+
"grad_norm": 0.015555203892290592,
|
729 |
+
"learning_rate": 1.2568128496661254e-05,
|
730 |
+
"loss": 0.0045,
|
731 |
+
"step": 2060
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 1.1261505143475907,
|
735 |
+
"grad_norm": 0.016391275450587273,
|
736 |
+
"learning_rate": 1.2495939361126153e-05,
|
737 |
+
"loss": 0.0046,
|
738 |
+
"step": 2080
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 1.136978884677856,
|
742 |
+
"grad_norm": 0.017521198838949203,
|
743 |
+
"learning_rate": 1.242375022559105e-05,
|
744 |
+
"loss": 0.0045,
|
745 |
+
"step": 2100
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 1.1478072550081213,
|
749 |
+
"grad_norm": 0.01668621227145195,
|
750 |
+
"learning_rate": 1.2351561090055949e-05,
|
751 |
+
"loss": 0.0044,
|
752 |
+
"step": 2120
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 1.1586356253383865,
|
756 |
+
"grad_norm": 0.015800384804606438,
|
757 |
+
"learning_rate": 1.2279371954520844e-05,
|
758 |
+
"loss": 0.0045,
|
759 |
+
"step": 2140
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 1.169463995668652,
|
763 |
+
"grad_norm": 0.028260117396712303,
|
764 |
+
"learning_rate": 1.2207182818985743e-05,
|
765 |
+
"loss": 0.0045,
|
766 |
+
"step": 2160
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 1.180292365998917,
|
770 |
+
"grad_norm": 0.015655677765607834,
|
771 |
+
"learning_rate": 1.2134993683450642e-05,
|
772 |
+
"loss": 0.0041,
|
773 |
+
"step": 2180
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 1.1911207363291825,
|
777 |
+
"grad_norm": 0.01357643585652113,
|
778 |
+
"learning_rate": 1.2062804547915539e-05,
|
779 |
+
"loss": 0.0042,
|
780 |
+
"step": 2200
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 1.2019491066594479,
|
784 |
+
"grad_norm": 0.013204825110733509,
|
785 |
+
"learning_rate": 1.1990615412380438e-05,
|
786 |
+
"loss": 0.0043,
|
787 |
+
"step": 2220
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 1.212777476989713,
|
791 |
+
"grad_norm": 0.01776127517223358,
|
792 |
+
"learning_rate": 1.1918426276845336e-05,
|
793 |
+
"loss": 0.0044,
|
794 |
+
"step": 2240
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 1.2236058473199782,
|
798 |
+
"grad_norm": 0.02663789503276348,
|
799 |
+
"learning_rate": 1.1846237141310234e-05,
|
800 |
+
"loss": 0.0041,
|
801 |
+
"step": 2260
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 1.2344342176502436,
|
805 |
+
"grad_norm": 0.016948629170656204,
|
806 |
+
"learning_rate": 1.1774048005775132e-05,
|
807 |
+
"loss": 0.0038,
|
808 |
+
"step": 2280
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 1.245262587980509,
|
812 |
+
"grad_norm": 0.030479852110147476,
|
813 |
+
"learning_rate": 1.1701858870240031e-05,
|
814 |
+
"loss": 0.004,
|
815 |
+
"step": 2300
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 1.2560909583107742,
|
819 |
+
"grad_norm": 0.020524190738797188,
|
820 |
+
"learning_rate": 1.1629669734704927e-05,
|
821 |
+
"loss": 0.0039,
|
822 |
+
"step": 2320
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 1.2669193286410394,
|
826 |
+
"grad_norm": 0.027631113305687904,
|
827 |
+
"learning_rate": 1.1557480599169825e-05,
|
828 |
+
"loss": 0.0038,
|
829 |
+
"step": 2340
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 1.2777476989713048,
|
833 |
+
"grad_norm": 0.017368769273161888,
|
834 |
+
"learning_rate": 1.1485291463634724e-05,
|
835 |
+
"loss": 0.0038,
|
836 |
+
"step": 2360
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 1.2885760693015702,
|
840 |
+
"grad_norm": 0.013879314996302128,
|
841 |
+
"learning_rate": 1.1413102328099621e-05,
|
842 |
+
"loss": 0.0038,
|
843 |
+
"step": 2380
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 1.2994044396318354,
|
847 |
+
"grad_norm": 0.03698274865746498,
|
848 |
+
"learning_rate": 1.134091319256452e-05,
|
849 |
+
"loss": 0.0039,
|
850 |
+
"step": 2400
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"epoch": 1.3102328099621008,
|
854 |
+
"grad_norm": 0.015918811783194542,
|
855 |
+
"learning_rate": 1.1268724057029419e-05,
|
856 |
+
"loss": 0.0038,
|
857 |
+
"step": 2420
|
858 |
+
},
|
859 |
+
{
|
860 |
+
"epoch": 1.321061180292366,
|
861 |
+
"grad_norm": 0.01589634083211422,
|
862 |
+
"learning_rate": 1.1196534921494316e-05,
|
863 |
+
"loss": 0.0039,
|
864 |
+
"step": 2440
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 1.3318895506226314,
|
868 |
+
"grad_norm": 0.01514554861932993,
|
869 |
+
"learning_rate": 1.1124345785959215e-05,
|
870 |
+
"loss": 0.0034,
|
871 |
+
"step": 2460
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"epoch": 1.3427179209528965,
|
875 |
+
"grad_norm": 0.013614562340080738,
|
876 |
+
"learning_rate": 1.1052156650424114e-05,
|
877 |
+
"loss": 0.0037,
|
878 |
+
"step": 2480
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 1.353546291283162,
|
882 |
+
"grad_norm": 0.018234172835946083,
|
883 |
+
"learning_rate": 1.0979967514889009e-05,
|
884 |
+
"loss": 0.0034,
|
885 |
+
"step": 2500
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 1.3643746616134271,
|
889 |
+
"grad_norm": 0.013683984987437725,
|
890 |
+
"learning_rate": 1.0907778379353908e-05,
|
891 |
+
"loss": 0.0034,
|
892 |
+
"step": 2520
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 1.3752030319436925,
|
896 |
+
"grad_norm": 0.014102173037827015,
|
897 |
+
"learning_rate": 1.0835589243818807e-05,
|
898 |
+
"loss": 0.0033,
|
899 |
+
"step": 2540
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 1.3860314022739577,
|
903 |
+
"grad_norm": 0.012239793315529823,
|
904 |
+
"learning_rate": 1.0763400108283704e-05,
|
905 |
+
"loss": 0.0032,
|
906 |
+
"step": 2560
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 1.3968597726042231,
|
910 |
+
"grad_norm": 0.013394709676504135,
|
911 |
+
"learning_rate": 1.0691210972748602e-05,
|
912 |
+
"loss": 0.0033,
|
913 |
+
"step": 2580
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 1.4076881429344883,
|
917 |
+
"grad_norm": 0.012910670600831509,
|
918 |
+
"learning_rate": 1.0619021837213501e-05,
|
919 |
+
"loss": 0.0033,
|
920 |
+
"step": 2600
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 1.4185165132647537,
|
924 |
+
"grad_norm": 0.0153429564088583,
|
925 |
+
"learning_rate": 1.0546832701678398e-05,
|
926 |
+
"loss": 0.0033,
|
927 |
+
"step": 2620
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 1.4293448835950189,
|
931 |
+
"grad_norm": 0.01714710146188736,
|
932 |
+
"learning_rate": 1.0474643566143297e-05,
|
933 |
+
"loss": 0.003,
|
934 |
+
"step": 2640
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 1.4401732539252843,
|
938 |
+
"grad_norm": 0.010895042680203915,
|
939 |
+
"learning_rate": 1.0402454430608196e-05,
|
940 |
+
"loss": 0.0033,
|
941 |
+
"step": 2660
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 1.4510016242555495,
|
945 |
+
"grad_norm": 0.012086061760783195,
|
946 |
+
"learning_rate": 1.0330265295073091e-05,
|
947 |
+
"loss": 0.0031,
|
948 |
+
"step": 2680
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 1.4618299945858149,
|
952 |
+
"grad_norm": 0.012154944241046906,
|
953 |
+
"learning_rate": 1.025807615953799e-05,
|
954 |
+
"loss": 0.003,
|
955 |
+
"step": 2700
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 1.47265836491608,
|
959 |
+
"grad_norm": 0.016563011333346367,
|
960 |
+
"learning_rate": 1.0185887024002887e-05,
|
961 |
+
"loss": 0.0032,
|
962 |
+
"step": 2720
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 1.4834867352463454,
|
966 |
+
"grad_norm": 0.011688679456710815,
|
967 |
+
"learning_rate": 1.0113697888467786e-05,
|
968 |
+
"loss": 0.003,
|
969 |
+
"step": 2740
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 1.4943151055766108,
|
973 |
+
"grad_norm": 0.010492874309420586,
|
974 |
+
"learning_rate": 1.0041508752932685e-05,
|
975 |
+
"loss": 0.0029,
|
976 |
+
"step": 2760
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 1.505143475906876,
|
980 |
+
"grad_norm": 0.01102561503648758,
|
981 |
+
"learning_rate": 9.969319617397582e-06,
|
982 |
+
"loss": 0.0029,
|
983 |
+
"step": 2780
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 1.5159718462371412,
|
987 |
+
"grad_norm": 0.013833099976181984,
|
988 |
+
"learning_rate": 9.89713048186248e-06,
|
989 |
+
"loss": 0.003,
|
990 |
+
"step": 2800
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 1.5268002165674066,
|
994 |
+
"grad_norm": 0.012262464500963688,
|
995 |
+
"learning_rate": 9.82494134632738e-06,
|
996 |
+
"loss": 0.0029,
|
997 |
+
"step": 2820
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 1.537628586897672,
|
1001 |
+
"grad_norm": 0.010000896640121937,
|
1002 |
+
"learning_rate": 9.752752210792277e-06,
|
1003 |
+
"loss": 0.0028,
|
1004 |
+
"step": 2840
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 1.5484569572279372,
|
1008 |
+
"grad_norm": 0.008956516161561012,
|
1009 |
+
"learning_rate": 9.680563075257174e-06,
|
1010 |
+
"loss": 0.003,
|
1011 |
+
"step": 2860
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 1.5592853275582024,
|
1015 |
+
"grad_norm": 0.01270230021327734,
|
1016 |
+
"learning_rate": 9.608373939722073e-06,
|
1017 |
+
"loss": 0.0028,
|
1018 |
+
"step": 2880
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 1.5701136978884678,
|
1022 |
+
"grad_norm": 0.012823579832911491,
|
1023 |
+
"learning_rate": 9.536184804186971e-06,
|
1024 |
+
"loss": 0.0028,
|
1025 |
+
"step": 2900
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 1.5809420682187332,
|
1029 |
+
"grad_norm": 0.009921281598508358,
|
1030 |
+
"learning_rate": 9.463995668651868e-06,
|
1031 |
+
"loss": 0.0027,
|
1032 |
+
"step": 2920
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 1.5917704385489984,
|
1036 |
+
"grad_norm": 0.021674729883670807,
|
1037 |
+
"learning_rate": 9.391806533116766e-06,
|
1038 |
+
"loss": 0.0027,
|
1039 |
+
"step": 2940
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"epoch": 1.6025988088792635,
|
1043 |
+
"grad_norm": 0.012086630798876286,
|
1044 |
+
"learning_rate": 9.319617397581664e-06,
|
1045 |
+
"loss": 0.0027,
|
1046 |
+
"step": 2960
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"epoch": 1.613427179209529,
|
1050 |
+
"grad_norm": 0.013828184455633163,
|
1051 |
+
"learning_rate": 9.247428262046563e-06,
|
1052 |
+
"loss": 0.0028,
|
1053 |
+
"step": 2980
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 1.6242555495397943,
|
1057 |
+
"grad_norm": 0.01388394832611084,
|
1058 |
+
"learning_rate": 9.17523912651146e-06,
|
1059 |
+
"loss": 0.0027,
|
1060 |
+
"step": 3000
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 1.6350839198700595,
|
1064 |
+
"grad_norm": 0.009276646189391613,
|
1065 |
+
"learning_rate": 9.103049990976359e-06,
|
1066 |
+
"loss": 0.0026,
|
1067 |
+
"step": 3020
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"epoch": 1.6459122902003247,
|
1071 |
+
"grad_norm": 0.010650601238012314,
|
1072 |
+
"learning_rate": 9.030860855441256e-06,
|
1073 |
+
"loss": 0.0026,
|
1074 |
+
"step": 3040
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 1.65674066053059,
|
1078 |
+
"grad_norm": 0.009121935814619064,
|
1079 |
+
"learning_rate": 8.958671719906155e-06,
|
1080 |
+
"loss": 0.0026,
|
1081 |
+
"step": 3060
|
1082 |
+
},
|
1083 |
+
{
|
1084 |
+
"epoch": 1.6675690308608555,
|
1085 |
+
"grad_norm": 0.012495677918195724,
|
1086 |
+
"learning_rate": 8.886482584371054e-06,
|
1087 |
+
"loss": 0.0026,
|
1088 |
+
"step": 3080
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"epoch": 1.6783974011911207,
|
1092 |
+
"grad_norm": 0.01117989607155323,
|
1093 |
+
"learning_rate": 8.81429344883595e-06,
|
1094 |
+
"loss": 0.0026,
|
1095 |
+
"step": 3100
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 1.6892257715213859,
|
1099 |
+
"grad_norm": 0.010890294797718525,
|
1100 |
+
"learning_rate": 8.74210431330085e-06,
|
1101 |
+
"loss": 0.0024,
|
1102 |
+
"step": 3120
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 1.7000541418516515,
|
1106 |
+
"grad_norm": 0.009491957724094391,
|
1107 |
+
"learning_rate": 8.669915177765747e-06,
|
1108 |
+
"loss": 0.0025,
|
1109 |
+
"step": 3140
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 1.7108825121819167,
|
1113 |
+
"grad_norm": 0.011014943942427635,
|
1114 |
+
"learning_rate": 8.597726042230645e-06,
|
1115 |
+
"loss": 0.0026,
|
1116 |
+
"step": 3160
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"epoch": 1.7217108825121819,
|
1120 |
+
"grad_norm": 0.015149922110140324,
|
1121 |
+
"learning_rate": 8.525536906695543e-06,
|
1122 |
+
"loss": 0.0024,
|
1123 |
+
"step": 3180
|
1124 |
+
},
|
1125 |
+
{
|
1126 |
+
"epoch": 1.7325392528424473,
|
1127 |
+
"grad_norm": 0.011757172644138336,
|
1128 |
+
"learning_rate": 8.453347771160441e-06,
|
1129 |
+
"loss": 0.0023,
|
1130 |
+
"step": 3200
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 1.7433676231727127,
|
1134 |
+
"grad_norm": 0.007756380829960108,
|
1135 |
+
"learning_rate": 8.38115863562534e-06,
|
1136 |
+
"loss": 0.0023,
|
1137 |
+
"step": 3220
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 1.7541959935029778,
|
1141 |
+
"grad_norm": 0.010520155541598797,
|
1142 |
+
"learning_rate": 8.308969500090237e-06,
|
1143 |
+
"loss": 0.0023,
|
1144 |
+
"step": 3240
|
1145 |
+
},
|
1146 |
+
{
|
1147 |
+
"epoch": 1.765024363833243,
|
1148 |
+
"grad_norm": 0.009943107143044472,
|
1149 |
+
"learning_rate": 8.236780364555134e-06,
|
1150 |
+
"loss": 0.0023,
|
1151 |
+
"step": 3260
|
1152 |
+
},
|
1153 |
+
{
|
1154 |
+
"epoch": 1.7758527341635084,
|
1155 |
+
"grad_norm": 0.011330407112836838,
|
1156 |
+
"learning_rate": 8.164591229020033e-06,
|
1157 |
+
"loss": 0.0023,
|
1158 |
+
"step": 3280
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"epoch": 1.7866811044937738,
|
1162 |
+
"grad_norm": 0.013512694276869297,
|
1163 |
+
"learning_rate": 8.092402093484932e-06,
|
1164 |
+
"loss": 0.0023,
|
1165 |
+
"step": 3300
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"epoch": 1.797509474824039,
|
1169 |
+
"grad_norm": 0.010237209498882294,
|
1170 |
+
"learning_rate": 8.020212957949829e-06,
|
1171 |
+
"loss": 0.0023,
|
1172 |
+
"step": 3320
|
1173 |
+
},
|
1174 |
+
{
|
1175 |
+
"epoch": 1.8083378451543042,
|
1176 |
+
"grad_norm": 0.013071142137050629,
|
1177 |
+
"learning_rate": 7.948023822414728e-06,
|
1178 |
+
"loss": 0.0022,
|
1179 |
+
"step": 3340
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 1.8191662154845696,
|
1183 |
+
"grad_norm": 0.0089937848970294,
|
1184 |
+
"learning_rate": 7.875834686879625e-06,
|
1185 |
+
"loss": 0.0024,
|
1186 |
+
"step": 3360
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 1.829994585814835,
|
1190 |
+
"grad_norm": 0.01620866358280182,
|
1191 |
+
"learning_rate": 7.803645551344524e-06,
|
1192 |
+
"loss": 0.0022,
|
1193 |
+
"step": 3380
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 1.8408229561451002,
|
1197 |
+
"grad_norm": 0.009367492981255054,
|
1198 |
+
"learning_rate": 7.731456415809423e-06,
|
1199 |
+
"loss": 0.0021,
|
1200 |
+
"step": 3400
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 1.8516513264753653,
|
1204 |
+
"grad_norm": 0.006648328620940447,
|
1205 |
+
"learning_rate": 7.65926728027432e-06,
|
1206 |
+
"loss": 0.0021,
|
1207 |
+
"step": 3420
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 1.8624796968056307,
|
1211 |
+
"grad_norm": 0.013929731212556362,
|
1212 |
+
"learning_rate": 7.587078144739218e-06,
|
1213 |
+
"loss": 0.0021,
|
1214 |
+
"step": 3440
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 1.8733080671358961,
|
1218 |
+
"grad_norm": 0.014050954952836037,
|
1219 |
+
"learning_rate": 7.514889009204115e-06,
|
1220 |
+
"loss": 0.0023,
|
1221 |
+
"step": 3460
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 1.8841364374661613,
|
1225 |
+
"grad_norm": 0.011463869363069534,
|
1226 |
+
"learning_rate": 7.4426998736690135e-06,
|
1227 |
+
"loss": 0.002,
|
1228 |
+
"step": 3480
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 1.8949648077964265,
|
1232 |
+
"grad_norm": 0.008514183573424816,
|
1233 |
+
"learning_rate": 7.3705107381339114e-06,
|
1234 |
+
"loss": 0.0021,
|
1235 |
+
"step": 3500
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 1.905793178126692,
|
1239 |
+
"grad_norm": 0.010145863518118858,
|
1240 |
+
"learning_rate": 7.298321602598809e-06,
|
1241 |
+
"loss": 0.0021,
|
1242 |
+
"step": 3520
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 1.9166215484569573,
|
1246 |
+
"grad_norm": 0.008266255259513855,
|
1247 |
+
"learning_rate": 7.226132467063708e-06,
|
1248 |
+
"loss": 0.0023,
|
1249 |
+
"step": 3540
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 1.9274499187872225,
|
1253 |
+
"grad_norm": 0.008439561352133751,
|
1254 |
+
"learning_rate": 7.153943331528605e-06,
|
1255 |
+
"loss": 0.002,
|
1256 |
+
"step": 3560
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 1.9382782891174877,
|
1260 |
+
"grad_norm": 0.007920574396848679,
|
1261 |
+
"learning_rate": 7.081754195993503e-06,
|
1262 |
+
"loss": 0.0021,
|
1263 |
+
"step": 3580
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 1.949106659447753,
|
1267 |
+
"grad_norm": 0.01343179028481245,
|
1268 |
+
"learning_rate": 7.009565060458402e-06,
|
1269 |
+
"loss": 0.0021,
|
1270 |
+
"step": 3600
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 1.9599350297780185,
|
1274 |
+
"grad_norm": 0.008394268341362476,
|
1275 |
+
"learning_rate": 6.9373759249233e-06,
|
1276 |
+
"loss": 0.002,
|
1277 |
+
"step": 3620
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 1.9707634001082837,
|
1281 |
+
"grad_norm": 0.007805489934980869,
|
1282 |
+
"learning_rate": 6.865186789388197e-06,
|
1283 |
+
"loss": 0.0019,
|
1284 |
+
"step": 3640
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 1.9815917704385488,
|
1288 |
+
"grad_norm": 0.00863006990402937,
|
1289 |
+
"learning_rate": 6.792997653853096e-06,
|
1290 |
+
"loss": 0.0021,
|
1291 |
+
"step": 3660
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 1.9924201407688142,
|
1295 |
+
"grad_norm": 0.011791090480983257,
|
1296 |
+
"learning_rate": 6.720808518317994e-06,
|
1297 |
+
"loss": 0.002,
|
1298 |
+
"step": 3680
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 2.0032485110990796,
|
1302 |
+
"grad_norm": 0.008283623494207859,
|
1303 |
+
"learning_rate": 6.648619382782892e-06,
|
1304 |
+
"loss": 0.0018,
|
1305 |
+
"step": 3700
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 2.014076881429345,
|
1309 |
+
"grad_norm": 0.006815766915678978,
|
1310 |
+
"learning_rate": 6.576430247247789e-06,
|
1311 |
+
"loss": 0.0018,
|
1312 |
+
"step": 3720
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 2.02490525175961,
|
1316 |
+
"grad_norm": 0.009198131039738655,
|
1317 |
+
"learning_rate": 6.504241111712688e-06,
|
1318 |
+
"loss": 0.0019,
|
1319 |
+
"step": 3740
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 2.0357336220898756,
|
1323 |
+
"grad_norm": 0.007912855595350266,
|
1324 |
+
"learning_rate": 6.432051976177586e-06,
|
1325 |
+
"loss": 0.0019,
|
1326 |
+
"step": 3760
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 2.046561992420141,
|
1330 |
+
"grad_norm": 0.008341687731444836,
|
1331 |
+
"learning_rate": 6.3598628406424836e-06,
|
1332 |
+
"loss": 0.0018,
|
1333 |
+
"step": 3780
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 2.057390362750406,
|
1337 |
+
"grad_norm": 0.007925598882138729,
|
1338 |
+
"learning_rate": 6.287673705107382e-06,
|
1339 |
+
"loss": 0.0018,
|
1340 |
+
"step": 3800
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 2.068218733080671,
|
1344 |
+
"grad_norm": 0.006331819575279951,
|
1345 |
+
"learning_rate": 6.2154845695722795e-06,
|
1346 |
+
"loss": 0.0018,
|
1347 |
+
"step": 3820
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 2.079047103410937,
|
1351 |
+
"grad_norm": 0.00803130678832531,
|
1352 |
+
"learning_rate": 6.143295434037177e-06,
|
1353 |
+
"loss": 0.0019,
|
1354 |
+
"step": 3840
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 2.089875473741202,
|
1358 |
+
"grad_norm": 0.009752047248184681,
|
1359 |
+
"learning_rate": 6.071106298502076e-06,
|
1360 |
+
"loss": 0.0019,
|
1361 |
+
"step": 3860
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 2.100703844071467,
|
1365 |
+
"grad_norm": 0.007367837708443403,
|
1366 |
+
"learning_rate": 5.998917162966974e-06,
|
1367 |
+
"loss": 0.0019,
|
1368 |
+
"step": 3880
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 2.1115322144017323,
|
1372 |
+
"grad_norm": 0.008053092285990715,
|
1373 |
+
"learning_rate": 5.926728027431872e-06,
|
1374 |
+
"loss": 0.0018,
|
1375 |
+
"step": 3900
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 2.122360584731998,
|
1379 |
+
"grad_norm": 0.008594776503741741,
|
1380 |
+
"learning_rate": 5.854538891896771e-06,
|
1381 |
+
"loss": 0.0017,
|
1382 |
+
"step": 3920
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 2.133188955062263,
|
1386 |
+
"grad_norm": 0.006970062851905823,
|
1387 |
+
"learning_rate": 5.782349756361668e-06,
|
1388 |
+
"loss": 0.0017,
|
1389 |
+
"step": 3940
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 2.1440173253925283,
|
1393 |
+
"grad_norm": 0.008627238683402538,
|
1394 |
+
"learning_rate": 5.710160620826566e-06,
|
1395 |
+
"loss": 0.0018,
|
1396 |
+
"step": 3960
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 2.154845695722794,
|
1400 |
+
"grad_norm": 0.008501927368342876,
|
1401 |
+
"learning_rate": 5.637971485291464e-06,
|
1402 |
+
"loss": 0.0018,
|
1403 |
+
"step": 3980
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 2.165674066053059,
|
1407 |
+
"grad_norm": 0.005817601457238197,
|
1408 |
+
"learning_rate": 5.565782349756363e-06,
|
1409 |
+
"loss": 0.0016,
|
1410 |
+
"step": 4000
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 2.1765024363833243,
|
1414 |
+
"grad_norm": 0.0064091249369084835,
|
1415 |
+
"learning_rate": 5.49359321422126e-06,
|
1416 |
+
"loss": 0.0018,
|
1417 |
+
"step": 4020
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 2.1873308067135895,
|
1421 |
+
"grad_norm": 0.005200565326958895,
|
1422 |
+
"learning_rate": 5.421404078686158e-06,
|
1423 |
+
"loss": 0.0016,
|
1424 |
+
"step": 4040
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 2.198159177043855,
|
1428 |
+
"grad_norm": 0.006780552677810192,
|
1429 |
+
"learning_rate": 5.3492149431510565e-06,
|
1430 |
+
"loss": 0.0018,
|
1431 |
+
"step": 4060
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 2.2089875473741203,
|
1435 |
+
"grad_norm": 0.006558467634022236,
|
1436 |
+
"learning_rate": 5.2770258076159545e-06,
|
1437 |
+
"loss": 0.0016,
|
1438 |
+
"step": 4080
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 2.2198159177043855,
|
1442 |
+
"grad_norm": 0.007784364745020866,
|
1443 |
+
"learning_rate": 5.204836672080852e-06,
|
1444 |
+
"loss": 0.0019,
|
1445 |
+
"step": 4100
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 2.2306442880346506,
|
1449 |
+
"grad_norm": 0.00484019098803401,
|
1450 |
+
"learning_rate": 5.13264753654575e-06,
|
1451 |
+
"loss": 0.0017,
|
1452 |
+
"step": 4120
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 2.2414726583649163,
|
1456 |
+
"grad_norm": 0.007469740230590105,
|
1457 |
+
"learning_rate": 5.060458401010648e-06,
|
1458 |
+
"loss": 0.0018,
|
1459 |
+
"step": 4140
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 2.2523010286951815,
|
1463 |
+
"grad_norm": 0.006617364473640919,
|
1464 |
+
"learning_rate": 4.988269265475546e-06,
|
1465 |
+
"loss": 0.0016,
|
1466 |
+
"step": 4160
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 2.2631293990254466,
|
1470 |
+
"grad_norm": 0.007724768482148647,
|
1471 |
+
"learning_rate": 4.916080129940444e-06,
|
1472 |
+
"loss": 0.0017,
|
1473 |
+
"step": 4180
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 2.273957769355712,
|
1477 |
+
"grad_norm": 0.0065683526918292046,
|
1478 |
+
"learning_rate": 4.843890994405342e-06,
|
1479 |
+
"loss": 0.0016,
|
1480 |
+
"step": 4200
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 2.2847861396859774,
|
1484 |
+
"grad_norm": 0.006633590906858444,
|
1485 |
+
"learning_rate": 4.77170185887024e-06,
|
1486 |
+
"loss": 0.0016,
|
1487 |
+
"step": 4220
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 2.2956145100162426,
|
1491 |
+
"grad_norm": 0.01162402518093586,
|
1492 |
+
"learning_rate": 4.699512723335138e-06,
|
1493 |
+
"loss": 0.0016,
|
1494 |
+
"step": 4240
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 2.306442880346508,
|
1498 |
+
"grad_norm": 0.008504554629325867,
|
1499 |
+
"learning_rate": 4.627323587800037e-06,
|
1500 |
+
"loss": 0.0015,
|
1501 |
+
"step": 4260
|
1502 |
+
},
|
1503 |
+
{
|
1504 |
+
"epoch": 2.317271250676773,
|
1505 |
+
"grad_norm": 0.00829976238310337,
|
1506 |
+
"learning_rate": 4.555134452264934e-06,
|
1507 |
+
"loss": 0.0015,
|
1508 |
+
"step": 4280
|
1509 |
+
},
|
1510 |
+
{
|
1511 |
+
"epoch": 2.3280996210070386,
|
1512 |
+
"grad_norm": 0.00552743673324585,
|
1513 |
+
"learning_rate": 4.482945316729833e-06,
|
1514 |
+
"loss": 0.0014,
|
1515 |
+
"step": 4300
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 2.338927991337304,
|
1519 |
+
"grad_norm": 0.005899062845855951,
|
1520 |
+
"learning_rate": 4.410756181194731e-06,
|
1521 |
+
"loss": 0.0016,
|
1522 |
+
"step": 4320
|
1523 |
+
},
|
1524 |
+
{
|
1525 |
+
"epoch": 2.349756361667569,
|
1526 |
+
"grad_norm": 0.008548817597329617,
|
1527 |
+
"learning_rate": 4.338567045659629e-06,
|
1528 |
+
"loss": 0.0015,
|
1529 |
+
"step": 4340
|
1530 |
+
},
|
1531 |
+
{
|
1532 |
+
"epoch": 2.360584731997834,
|
1533 |
+
"grad_norm": 0.00820871815085411,
|
1534 |
+
"learning_rate": 4.266377910124527e-06,
|
1535 |
+
"loss": 0.0017,
|
1536 |
+
"step": 4360
|
1537 |
+
},
|
1538 |
+
{
|
1539 |
+
"epoch": 2.3714131023280998,
|
1540 |
+
"grad_norm": 0.010167138651013374,
|
1541 |
+
"learning_rate": 4.1941887745894246e-06,
|
1542 |
+
"loss": 0.0016,
|
1543 |
+
"step": 4380
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 2.382241472658365,
|
1547 |
+
"grad_norm": 0.006324863061308861,
|
1548 |
+
"learning_rate": 4.1219996390543225e-06,
|
1549 |
+
"loss": 0.0016,
|
1550 |
+
"step": 4400
|
1551 |
+
},
|
1552 |
+
{
|
1553 |
+
"epoch": 2.39306984298863,
|
1554 |
+
"grad_norm": 0.00801610853523016,
|
1555 |
+
"learning_rate": 4.049810503519221e-06,
|
1556 |
+
"loss": 0.0015,
|
1557 |
+
"step": 4420
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 2.4038982133188957,
|
1561 |
+
"grad_norm": 0.006985550280660391,
|
1562 |
+
"learning_rate": 3.977621367984118e-06,
|
1563 |
+
"loss": 0.0016,
|
1564 |
+
"step": 4440
|
1565 |
+
},
|
1566 |
+
{
|
1567 |
+
"epoch": 2.414726583649161,
|
1568 |
+
"grad_norm": 0.00809464231133461,
|
1569 |
+
"learning_rate": 3.905432232449017e-06,
|
1570 |
+
"loss": 0.0015,
|
1571 |
+
"step": 4460
|
1572 |
+
},
|
1573 |
+
{
|
1574 |
+
"epoch": 2.425554953979426,
|
1575 |
+
"grad_norm": 0.0049890028312802315,
|
1576 |
+
"learning_rate": 3.833243096913914e-06,
|
1577 |
+
"loss": 0.0016,
|
1578 |
+
"step": 4480
|
1579 |
+
},
|
1580 |
+
{
|
1581 |
+
"epoch": 2.4363833243096913,
|
1582 |
+
"grad_norm": 0.0060018254444003105,
|
1583 |
+
"learning_rate": 3.7610539613788127e-06,
|
1584 |
+
"loss": 0.0015,
|
1585 |
+
"step": 4500
|
1586 |
+
},
|
1587 |
+
{
|
1588 |
+
"epoch": 2.4472116946399565,
|
1589 |
+
"grad_norm": 0.011303539387881756,
|
1590 |
+
"learning_rate": 3.688864825843711e-06,
|
1591 |
+
"loss": 0.0014,
|
1592 |
+
"step": 4520
|
1593 |
+
},
|
1594 |
+
{
|
1595 |
+
"epoch": 2.458040064970222,
|
1596 |
+
"grad_norm": 0.006434119306504726,
|
1597 |
+
"learning_rate": 3.6166756903086086e-06,
|
1598 |
+
"loss": 0.0015,
|
1599 |
+
"step": 4540
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 2.4688684353004873,
|
1603 |
+
"grad_norm": 0.01008700393140316,
|
1604 |
+
"learning_rate": 3.544486554773507e-06,
|
1605 |
+
"loss": 0.0015,
|
1606 |
+
"step": 4560
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"epoch": 2.4796968056307525,
|
1610 |
+
"grad_norm": 0.004655952099710703,
|
1611 |
+
"learning_rate": 3.4722974192384053e-06,
|
1612 |
+
"loss": 0.0015,
|
1613 |
+
"step": 4580
|
1614 |
+
},
|
1615 |
+
{
|
1616 |
+
"epoch": 2.490525175961018,
|
1617 |
+
"grad_norm": 0.009297892451286316,
|
1618 |
+
"learning_rate": 3.400108283703303e-06,
|
1619 |
+
"loss": 0.0016,
|
1620 |
+
"step": 4600
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 2.5013535462912833,
|
1624 |
+
"grad_norm": 0.007229143753647804,
|
1625 |
+
"learning_rate": 3.327919148168201e-06,
|
1626 |
+
"loss": 0.0017,
|
1627 |
+
"step": 4620
|
1628 |
+
},
|
1629 |
+
{
|
1630 |
+
"epoch": 2.5121819166215484,
|
1631 |
+
"grad_norm": 0.007766417693346739,
|
1632 |
+
"learning_rate": 3.2557300126330987e-06,
|
1633 |
+
"loss": 0.0015,
|
1634 |
+
"step": 4640
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 2.5230102869518136,
|
1638 |
+
"grad_norm": 0.010671594180166721,
|
1639 |
+
"learning_rate": 3.183540877097997e-06,
|
1640 |
+
"loss": 0.0016,
|
1641 |
+
"step": 4660
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 2.533838657282079,
|
1645 |
+
"grad_norm": 0.008643384091556072,
|
1646 |
+
"learning_rate": 3.111351741562895e-06,
|
1647 |
+
"loss": 0.0015,
|
1648 |
+
"step": 4680
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 2.5446670276123444,
|
1652 |
+
"grad_norm": 0.0042122360318899155,
|
1653 |
+
"learning_rate": 3.039162606027793e-06,
|
1654 |
+
"loss": 0.0015,
|
1655 |
+
"step": 4700
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 2.5554953979426096,
|
1659 |
+
"grad_norm": 0.00437292642891407,
|
1660 |
+
"learning_rate": 2.966973470492691e-06,
|
1661 |
+
"loss": 0.0014,
|
1662 |
+
"step": 4720
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 2.566323768272875,
|
1666 |
+
"grad_norm": 0.005281255580484867,
|
1667 |
+
"learning_rate": 2.8947843349575893e-06,
|
1668 |
+
"loss": 0.0014,
|
1669 |
+
"step": 4740
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 2.5771521386031404,
|
1673 |
+
"grad_norm": 0.007413716986775398,
|
1674 |
+
"learning_rate": 2.8225951994224873e-06,
|
1675 |
+
"loss": 0.0015,
|
1676 |
+
"step": 4760
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 2.5879805089334056,
|
1680 |
+
"grad_norm": 0.005595244932919741,
|
1681 |
+
"learning_rate": 2.750406063887385e-06,
|
1682 |
+
"loss": 0.0014,
|
1683 |
+
"step": 4780
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 2.5988088792636708,
|
1687 |
+
"grad_norm": 0.004735792987048626,
|
1688 |
+
"learning_rate": 2.678216928352283e-06,
|
1689 |
+
"loss": 0.0013,
|
1690 |
+
"step": 4800
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 2.609637249593936,
|
1694 |
+
"grad_norm": 0.005141600966453552,
|
1695 |
+
"learning_rate": 2.606027792817181e-06,
|
1696 |
+
"loss": 0.0013,
|
1697 |
+
"step": 4820
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 2.6204656199242016,
|
1701 |
+
"grad_norm": 0.0045648557133972645,
|
1702 |
+
"learning_rate": 2.5338386572820795e-06,
|
1703 |
+
"loss": 0.0014,
|
1704 |
+
"step": 4840
|
1705 |
+
},
|
1706 |
+
{
|
1707 |
+
"epoch": 2.6312939902544668,
|
1708 |
+
"grad_norm": 0.006522185634821653,
|
1709 |
+
"learning_rate": 2.4616495217469774e-06,
|
1710 |
+
"loss": 0.0014,
|
1711 |
+
"step": 4860
|
1712 |
+
},
|
1713 |
+
{
|
1714 |
+
"epoch": 2.642122360584732,
|
1715 |
+
"grad_norm": 0.010215125046670437,
|
1716 |
+
"learning_rate": 2.3894603862118754e-06,
|
1717 |
+
"loss": 0.0014,
|
1718 |
+
"step": 4880
|
1719 |
+
},
|
1720 |
+
{
|
1721 |
+
"epoch": 2.6529507309149976,
|
1722 |
+
"grad_norm": 0.007124153897166252,
|
1723 |
+
"learning_rate": 2.3172712506767733e-06,
|
1724 |
+
"loss": 0.0014,
|
1725 |
+
"step": 4900
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"epoch": 2.6637791012452627,
|
1729 |
+
"grad_norm": 0.0049623060040175915,
|
1730 |
+
"learning_rate": 2.2450821151416713e-06,
|
1731 |
+
"loss": 0.0013,
|
1732 |
+
"step": 4920
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"epoch": 2.674607471575528,
|
1736 |
+
"grad_norm": 0.005493966396898031,
|
1737 |
+
"learning_rate": 2.1728929796065692e-06,
|
1738 |
+
"loss": 0.0013,
|
1739 |
+
"step": 4940
|
1740 |
+
},
|
1741 |
+
{
|
1742 |
+
"epoch": 2.685435841905793,
|
1743 |
+
"grad_norm": 0.006936436053365469,
|
1744 |
+
"learning_rate": 2.1007038440714676e-06,
|
1745 |
+
"loss": 0.0013,
|
1746 |
+
"step": 4960
|
1747 |
+
},
|
1748 |
+
{
|
1749 |
+
"epoch": 2.6962642122360583,
|
1750 |
+
"grad_norm": 0.006746096536517143,
|
1751 |
+
"learning_rate": 2.0285147085363655e-06,
|
1752 |
+
"loss": 0.0013,
|
1753 |
+
"step": 4980
|
1754 |
+
},
|
1755 |
+
{
|
1756 |
+
"epoch": 2.707092582566324,
|
1757 |
+
"grad_norm": 0.00648007495328784,
|
1758 |
+
"learning_rate": 1.9563255730012635e-06,
|
1759 |
+
"loss": 0.0013,
|
1760 |
+
"step": 5000
|
1761 |
+
},
|
1762 |
+
{
|
1763 |
+
"epoch": 2.717920952896589,
|
1764 |
+
"grad_norm": 0.006984813138842583,
|
1765 |
+
"learning_rate": 1.8841364374661614e-06,
|
1766 |
+
"loss": 0.0015,
|
1767 |
+
"step": 5020
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 2.7287493232268543,
|
1771 |
+
"grad_norm": 0.007297712843865156,
|
1772 |
+
"learning_rate": 1.8119473019310596e-06,
|
1773 |
+
"loss": 0.0014,
|
1774 |
+
"step": 5040
|
1775 |
+
},
|
1776 |
+
{
|
1777 |
+
"epoch": 2.73957769355712,
|
1778 |
+
"grad_norm": 0.006131511647254229,
|
1779 |
+
"learning_rate": 1.7397581663959575e-06,
|
1780 |
+
"loss": 0.0013,
|
1781 |
+
"step": 5060
|
1782 |
+
},
|
1783 |
+
{
|
1784 |
+
"epoch": 2.750406063887385,
|
1785 |
+
"grad_norm": 0.0050123510882258415,
|
1786 |
+
"learning_rate": 1.6675690308608555e-06,
|
1787 |
+
"loss": 0.0013,
|
1788 |
+
"step": 5080
|
1789 |
+
},
|
1790 |
+
{
|
1791 |
+
"epoch": 2.7612344342176502,
|
1792 |
+
"grad_norm": 0.005957981571555138,
|
1793 |
+
"learning_rate": 1.5953798953257534e-06,
|
1794 |
+
"loss": 0.0013,
|
1795 |
+
"step": 5100
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"epoch": 2.7720628045479154,
|
1799 |
+
"grad_norm": 0.005641666240990162,
|
1800 |
+
"learning_rate": 1.5231907597906518e-06,
|
1801 |
+
"loss": 0.0013,
|
1802 |
+
"step": 5120
|
1803 |
+
},
|
1804 |
+
{
|
1805 |
+
"epoch": 2.7828911748781806,
|
1806 |
+
"grad_norm": 0.005417118314653635,
|
1807 |
+
"learning_rate": 1.4510016242555498e-06,
|
1808 |
+
"loss": 0.0014,
|
1809 |
+
"step": 5140
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 2.7937195452084462,
|
1813 |
+
"grad_norm": 0.006614298094063997,
|
1814 |
+
"learning_rate": 1.3788124887204477e-06,
|
1815 |
+
"loss": 0.0014,
|
1816 |
+
"step": 5160
|
1817 |
+
},
|
1818 |
+
{
|
1819 |
+
"epoch": 2.8045479155387114,
|
1820 |
+
"grad_norm": 0.007049723528325558,
|
1821 |
+
"learning_rate": 1.3066233531853457e-06,
|
1822 |
+
"loss": 0.0014,
|
1823 |
+
"step": 5180
|
1824 |
+
},
|
1825 |
+
{
|
1826 |
+
"epoch": 2.8153762858689766,
|
1827 |
+
"grad_norm": 0.004931174218654633,
|
1828 |
+
"learning_rate": 1.2344342176502438e-06,
|
1829 |
+
"loss": 0.0013,
|
1830 |
+
"step": 5200
|
1831 |
+
},
|
1832 |
+
{
|
1833 |
+
"epoch": 2.826204656199242,
|
1834 |
+
"grad_norm": 0.007498623337596655,
|
1835 |
+
"learning_rate": 1.1622450821151418e-06,
|
1836 |
+
"loss": 0.0014,
|
1837 |
+
"step": 5220
|
1838 |
+
},
|
1839 |
+
{
|
1840 |
+
"epoch": 2.8370330265295074,
|
1841 |
+
"grad_norm": 0.006211725063621998,
|
1842 |
+
"learning_rate": 1.0900559465800397e-06,
|
1843 |
+
"loss": 0.0014,
|
1844 |
+
"step": 5240
|
1845 |
+
},
|
1846 |
+
{
|
1847 |
+
"epoch": 2.8478613968597726,
|
1848 |
+
"grad_norm": 0.008632579818367958,
|
1849 |
+
"learning_rate": 1.0178668110449379e-06,
|
1850 |
+
"loss": 0.0013,
|
1851 |
+
"step": 5260
|
1852 |
+
},
|
1853 |
+
{
|
1854 |
+
"epoch": 2.8586897671900378,
|
1855 |
+
"grad_norm": 0.008940057829022408,
|
1856 |
+
"learning_rate": 9.456776755098358e-07,
|
1857 |
+
"loss": 0.0013,
|
1858 |
+
"step": 5280
|
1859 |
+
},
|
1860 |
+
{
|
1861 |
+
"epoch": 2.869518137520303,
|
1862 |
+
"grad_norm": 0.011228718794882298,
|
1863 |
+
"learning_rate": 8.73488539974734e-07,
|
1864 |
+
"loss": 0.0014,
|
1865 |
+
"step": 5300
|
1866 |
+
},
|
1867 |
+
{
|
1868 |
+
"epoch": 2.8803465078505686,
|
1869 |
+
"grad_norm": 0.007089744322001934,
|
1870 |
+
"learning_rate": 8.012994044396319e-07,
|
1871 |
+
"loss": 0.0015,
|
1872 |
+
"step": 5320
|
1873 |
+
},
|
1874 |
+
{
|
1875 |
+
"epoch": 2.8911748781808337,
|
1876 |
+
"grad_norm": 0.006037184968590736,
|
1877 |
+
"learning_rate": 7.2911026890453e-07,
|
1878 |
+
"loss": 0.0013,
|
1879 |
+
"step": 5340
|
1880 |
+
},
|
1881 |
+
{
|
1882 |
+
"epoch": 2.902003248511099,
|
1883 |
+
"grad_norm": 0.005260369274765253,
|
1884 |
+
"learning_rate": 6.569211333694279e-07,
|
1885 |
+
"loss": 0.0012,
|
1886 |
+
"step": 5360
|
1887 |
+
},
|
1888 |
+
{
|
1889 |
+
"epoch": 2.9128316188413645,
|
1890 |
+
"grad_norm": 0.004701657686382532,
|
1891 |
+
"learning_rate": 5.84731997834326e-07,
|
1892 |
+
"loss": 0.0015,
|
1893 |
+
"step": 5380
|
1894 |
+
},
|
1895 |
+
{
|
1896 |
+
"epoch": 2.9236599891716297,
|
1897 |
+
"grad_norm": 0.009111477993428707,
|
1898 |
+
"learning_rate": 5.12542862299224e-07,
|
1899 |
+
"loss": 0.0013,
|
1900 |
+
"step": 5400
|
1901 |
+
},
|
1902 |
+
{
|
1903 |
+
"epoch": 2.934488359501895,
|
1904 |
+
"grad_norm": 0.0066655525006353855,
|
1905 |
+
"learning_rate": 4.4035372676412204e-07,
|
1906 |
+
"loss": 0.0013,
|
1907 |
+
"step": 5420
|
1908 |
+
},
|
1909 |
+
{
|
1910 |
+
"epoch": 2.94531672983216,
|
1911 |
+
"grad_norm": 0.007916849106550217,
|
1912 |
+
"learning_rate": 3.681645912290201e-07,
|
1913 |
+
"loss": 0.0013,
|
1914 |
+
"step": 5440
|
1915 |
+
},
|
1916 |
+
{
|
1917 |
+
"epoch": 2.9561451001624257,
|
1918 |
+
"grad_norm": 0.006292740814387798,
|
1919 |
+
"learning_rate": 2.959754556939181e-07,
|
1920 |
+
"loss": 0.0013,
|
1921 |
+
"step": 5460
|
1922 |
+
},
|
1923 |
+
{
|
1924 |
+
"epoch": 2.966973470492691,
|
1925 |
+
"grad_norm": 0.007718087639659643,
|
1926 |
+
"learning_rate": 2.2378632015881612e-07,
|
1927 |
+
"loss": 0.0013,
|
1928 |
+
"step": 5480
|
1929 |
+
},
|
1930 |
+
{
|
1931 |
+
"epoch": 2.977801840822956,
|
1932 |
+
"grad_norm": 0.005025485064834356,
|
1933 |
+
"learning_rate": 1.5159718462371415e-07,
|
1934 |
+
"loss": 0.0013,
|
1935 |
+
"step": 5500
|
1936 |
+
}
|
1937 |
+
],
|
1938 |
+
"logging_steps": 20,
|
1939 |
+
"max_steps": 5541,
|
1940 |
+
"num_input_tokens_seen": 0,
|
1941 |
+
"num_train_epochs": 3,
|
1942 |
+
"save_steps": 500,
|
1943 |
+
"stateful_callbacks": {
|
1944 |
+
"TrainerControl": {
|
1945 |
+
"args": {
|
1946 |
+
"should_epoch_stop": false,
|
1947 |
+
"should_evaluate": false,
|
1948 |
+
"should_log": false,
|
1949 |
+
"should_save": true,
|
1950 |
+
"should_training_stop": false
|
1951 |
+
},
|
1952 |
+
"attributes": {}
|
1953 |
+
}
|
1954 |
+
},
|
1955 |
+
"total_flos": 0.0,
|
1956 |
+
"train_batch_size": 16,
|
1957 |
+
"trial_name": null,
|
1958 |
+
"trial_params": null
|
1959 |
+
}
|
checkpoint-5500/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0214d33554592467921c6cf9ea4c66d3b830194ad3c4677046bb7d47de8435dc
|
3 |
+
size 5777
|
checkpoint-5541/added_tokens.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[MASK]": 128000
|
3 |
+
}
|
checkpoint-5541/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6aa9c0850bf7cfc7c43401b0061d3d3556b7a50742184ccb1159d53662b196fa
|
3 |
+
size 736593656
|
checkpoint-5541/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f6196029548a505d9e8588f8d67dcb0f209f2733bf8b2b3bd57fab1ab85550ad
|
3 |
+
size 1473311563
|
checkpoint-5541/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3d61577f3db55c65349f461a27d525ceb8471b141a7ad0c24606bf8c9517199d
|
3 |
+
size 14645
|
checkpoint-5541/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3665f383604f7568d9a9b7c135da62256c3ea65f2ddb8e6903404bcc5468f2e1
|
3 |
+
size 1465
|
checkpoint-5541/special_tokens_map.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "[SEP]",
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"pad_token": "[PAD]",
|
7 |
+
"sep_token": "[SEP]",
|
8 |
+
"unk_token": {
|
9 |
+
"content": "[UNK]",
|
10 |
+
"lstrip": false,
|
11 |
+
"normalized": true,
|
12 |
+
"rstrip": false,
|
13 |
+
"single_word": false
|
14 |
+
}
|
15 |
+
}
|
checkpoint-5541/spm.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
3 |
+
size 2464616
|
checkpoint-5541/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-5541/tokenizer_config.json
ADDED
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "[CLS]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "[SEP]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "[UNK]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": true,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"128000": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"bos_token": "[CLS]",
|
45 |
+
"clean_up_tokenization_spaces": false,
|
46 |
+
"cls_token": "[CLS]",
|
47 |
+
"do_lower_case": false,
|
48 |
+
"eos_token": "[SEP]",
|
49 |
+
"extra_special_tokens": {},
|
50 |
+
"mask_token": "[MASK]",
|
51 |
+
"model_max_length": 1000000000000000019884624838656,
|
52 |
+
"pad_token": "[PAD]",
|
53 |
+
"sep_token": "[SEP]",
|
54 |
+
"sp_model_kwargs": {},
|
55 |
+
"split_by_punct": false,
|
56 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
57 |
+
"unk_token": "[UNK]",
|
58 |
+
"vocab_type": "spm"
|
59 |
+
}
|
checkpoint-5541/trainer_state.json
ADDED
@@ -0,0 +1,1973 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": null,
|
3 |
+
"best_metric": null,
|
4 |
+
"best_model_checkpoint": null,
|
5 |
+
"epoch": 3.0,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 5541,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.010828370330265295,
|
14 |
+
"grad_norm": 8.006661415100098,
|
15 |
+
"learning_rate": 1.9931420321241655e-05,
|
16 |
+
"loss": 4.0423,
|
17 |
+
"step": 20
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.02165674066053059,
|
21 |
+
"grad_norm": 1.674608588218689,
|
22 |
+
"learning_rate": 1.9859231185706552e-05,
|
23 |
+
"loss": 1.275,
|
24 |
+
"step": 40
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.03248511099079589,
|
28 |
+
"grad_norm": 1.2871980667114258,
|
29 |
+
"learning_rate": 1.978704205017145e-05,
|
30 |
+
"loss": 0.879,
|
31 |
+
"step": 60
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 0.04331348132106118,
|
35 |
+
"grad_norm": 1.0388418436050415,
|
36 |
+
"learning_rate": 1.971485291463635e-05,
|
37 |
+
"loss": 0.6456,
|
38 |
+
"step": 80
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.05414185165132648,
|
42 |
+
"grad_norm": 0.8618717193603516,
|
43 |
+
"learning_rate": 1.9642663779101247e-05,
|
44 |
+
"loss": 0.4888,
|
45 |
+
"step": 100
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.06497022198159177,
|
49 |
+
"grad_norm": 0.7138980627059937,
|
50 |
+
"learning_rate": 1.9570474643566144e-05,
|
51 |
+
"loss": 0.382,
|
52 |
+
"step": 120
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 0.07579859231185707,
|
56 |
+
"grad_norm": 0.6197817921638489,
|
57 |
+
"learning_rate": 1.9498285508031044e-05,
|
58 |
+
"loss": 0.3027,
|
59 |
+
"step": 140
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 0.08662696264212236,
|
63 |
+
"grad_norm": 0.5373287796974182,
|
64 |
+
"learning_rate": 1.942609637249594e-05,
|
65 |
+
"loss": 0.247,
|
66 |
+
"step": 160
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.09745533297238766,
|
70 |
+
"grad_norm": 0.46481892466545105,
|
71 |
+
"learning_rate": 1.935390723696084e-05,
|
72 |
+
"loss": 0.2073,
|
73 |
+
"step": 180
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 0.10828370330265295,
|
77 |
+
"grad_norm": 0.40900182723999023,
|
78 |
+
"learning_rate": 1.9281718101425736e-05,
|
79 |
+
"loss": 0.1771,
|
80 |
+
"step": 200
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.11911207363291824,
|
84 |
+
"grad_norm": 0.35755395889282227,
|
85 |
+
"learning_rate": 1.9209528965890636e-05,
|
86 |
+
"loss": 0.151,
|
87 |
+
"step": 220
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.12994044396318355,
|
91 |
+
"grad_norm": 0.32475224137306213,
|
92 |
+
"learning_rate": 1.9137339830355533e-05,
|
93 |
+
"loss": 0.1327,
|
94 |
+
"step": 240
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.14076881429344884,
|
98 |
+
"grad_norm": 0.2906845211982727,
|
99 |
+
"learning_rate": 1.906515069482043e-05,
|
100 |
+
"loss": 0.1179,
|
101 |
+
"step": 260
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 0.15159718462371413,
|
105 |
+
"grad_norm": 0.26396697759628296,
|
106 |
+
"learning_rate": 1.899296155928533e-05,
|
107 |
+
"loss": 0.1058,
|
108 |
+
"step": 280
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 0.16242555495397942,
|
112 |
+
"grad_norm": 0.24280692636966705,
|
113 |
+
"learning_rate": 1.8920772423750228e-05,
|
114 |
+
"loss": 0.0963,
|
115 |
+
"step": 300
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 0.17325392528424471,
|
119 |
+
"grad_norm": 0.21794995665550232,
|
120 |
+
"learning_rate": 1.8848583288215125e-05,
|
121 |
+
"loss": 0.0851,
|
122 |
+
"step": 320
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.18408229561451,
|
126 |
+
"grad_norm": 0.2041829228401184,
|
127 |
+
"learning_rate": 1.8776394152680025e-05,
|
128 |
+
"loss": 0.0786,
|
129 |
+
"step": 340
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.19491066594477532,
|
133 |
+
"grad_norm": 0.18174469470977783,
|
134 |
+
"learning_rate": 1.870420501714492e-05,
|
135 |
+
"loss": 0.0731,
|
136 |
+
"step": 360
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 0.20573903627504062,
|
140 |
+
"grad_norm": 0.17271022498607635,
|
141 |
+
"learning_rate": 1.863201588160982e-05,
|
142 |
+
"loss": 0.0676,
|
143 |
+
"step": 380
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 0.2165674066053059,
|
147 |
+
"grad_norm": 0.1657329499721527,
|
148 |
+
"learning_rate": 1.8559826746074717e-05,
|
149 |
+
"loss": 0.0616,
|
150 |
+
"step": 400
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 0.2273957769355712,
|
154 |
+
"grad_norm": 0.16118521988391876,
|
155 |
+
"learning_rate": 1.8487637610539614e-05,
|
156 |
+
"loss": 0.0594,
|
157 |
+
"step": 420
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 0.2382241472658365,
|
161 |
+
"grad_norm": 0.14374467730522156,
|
162 |
+
"learning_rate": 1.8415448475004514e-05,
|
163 |
+
"loss": 0.0542,
|
164 |
+
"step": 440
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 0.24905251759610178,
|
168 |
+
"grad_norm": 0.1439092755317688,
|
169 |
+
"learning_rate": 1.834325933946941e-05,
|
170 |
+
"loss": 0.0512,
|
171 |
+
"step": 460
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.2598808879263671,
|
175 |
+
"grad_norm": 0.1367517113685608,
|
176 |
+
"learning_rate": 1.827107020393431e-05,
|
177 |
+
"loss": 0.0482,
|
178 |
+
"step": 480
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.2707092582566324,
|
182 |
+
"grad_norm": 0.13729216158390045,
|
183 |
+
"learning_rate": 1.819888106839921e-05,
|
184 |
+
"loss": 0.0458,
|
185 |
+
"step": 500
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 0.2815376285868977,
|
189 |
+
"grad_norm": 0.11813738197088242,
|
190 |
+
"learning_rate": 1.8126691932864103e-05,
|
191 |
+
"loss": 0.0442,
|
192 |
+
"step": 520
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 0.292365998917163,
|
196 |
+
"grad_norm": 0.10173018276691437,
|
197 |
+
"learning_rate": 1.8054502797329003e-05,
|
198 |
+
"loss": 0.0397,
|
199 |
+
"step": 540
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 0.30319436924742826,
|
203 |
+
"grad_norm": 0.10882216691970825,
|
204 |
+
"learning_rate": 1.79823136617939e-05,
|
205 |
+
"loss": 0.0377,
|
206 |
+
"step": 560
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.31402273957769355,
|
210 |
+
"grad_norm": 0.10021235793828964,
|
211 |
+
"learning_rate": 1.7910124526258797e-05,
|
212 |
+
"loss": 0.0366,
|
213 |
+
"step": 580
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.32485110990795885,
|
217 |
+
"grad_norm": 0.11167757213115692,
|
218 |
+
"learning_rate": 1.7837935390723698e-05,
|
219 |
+
"loss": 0.0354,
|
220 |
+
"step": 600
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 0.33567948023822414,
|
224 |
+
"grad_norm": 0.11719091236591339,
|
225 |
+
"learning_rate": 1.7765746255188595e-05,
|
226 |
+
"loss": 0.0361,
|
227 |
+
"step": 620
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 0.34650785056848943,
|
231 |
+
"grad_norm": 0.08965040743350983,
|
232 |
+
"learning_rate": 1.7693557119653492e-05,
|
233 |
+
"loss": 0.0328,
|
234 |
+
"step": 640
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.3573362208987547,
|
238 |
+
"grad_norm": 0.09095824509859085,
|
239 |
+
"learning_rate": 1.7621367984118393e-05,
|
240 |
+
"loss": 0.0324,
|
241 |
+
"step": 660
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.36816459122902,
|
245 |
+
"grad_norm": 0.09133260697126389,
|
246 |
+
"learning_rate": 1.754917884858329e-05,
|
247 |
+
"loss": 0.0305,
|
248 |
+
"step": 680
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 0.3789929615592853,
|
252 |
+
"grad_norm": 0.10949356853961945,
|
253 |
+
"learning_rate": 1.7476989713048187e-05,
|
254 |
+
"loss": 0.0298,
|
255 |
+
"step": 700
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 0.38982133188955065,
|
259 |
+
"grad_norm": 0.09127330780029297,
|
260 |
+
"learning_rate": 1.7404800577513087e-05,
|
261 |
+
"loss": 0.0283,
|
262 |
+
"step": 720
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 0.40064970221981594,
|
266 |
+
"grad_norm": 0.07634657621383667,
|
267 |
+
"learning_rate": 1.7332611441977984e-05,
|
268 |
+
"loss": 0.027,
|
269 |
+
"step": 740
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 0.41147807255008123,
|
273 |
+
"grad_norm": 0.1150558814406395,
|
274 |
+
"learning_rate": 1.726042230644288e-05,
|
275 |
+
"loss": 0.0256,
|
276 |
+
"step": 760
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 0.4223064428803465,
|
280 |
+
"grad_norm": 0.07289394736289978,
|
281 |
+
"learning_rate": 1.718823317090778e-05,
|
282 |
+
"loss": 0.0241,
|
283 |
+
"step": 780
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 0.4331348132106118,
|
287 |
+
"grad_norm": 0.07605121284723282,
|
288 |
+
"learning_rate": 1.711604403537268e-05,
|
289 |
+
"loss": 0.0241,
|
290 |
+
"step": 800
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 0.4439631835408771,
|
294 |
+
"grad_norm": 0.07429105043411255,
|
295 |
+
"learning_rate": 1.7043854899837576e-05,
|
296 |
+
"loss": 0.0233,
|
297 |
+
"step": 820
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 0.4547915538711424,
|
301 |
+
"grad_norm": 0.06406375020742416,
|
302 |
+
"learning_rate": 1.6971665764302473e-05,
|
303 |
+
"loss": 0.023,
|
304 |
+
"step": 840
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 0.4656199242014077,
|
308 |
+
"grad_norm": 0.06342233717441559,
|
309 |
+
"learning_rate": 1.6899476628767374e-05,
|
310 |
+
"loss": 0.0202,
|
311 |
+
"step": 860
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 0.476448294531673,
|
315 |
+
"grad_norm": 0.06463098526000977,
|
316 |
+
"learning_rate": 1.682728749323227e-05,
|
317 |
+
"loss": 0.0204,
|
318 |
+
"step": 880
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 0.48727666486193827,
|
322 |
+
"grad_norm": 0.05809445306658745,
|
323 |
+
"learning_rate": 1.6755098357697168e-05,
|
324 |
+
"loss": 0.0192,
|
325 |
+
"step": 900
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 0.49810503519220356,
|
329 |
+
"grad_norm": 0.06424301117658615,
|
330 |
+
"learning_rate": 1.668290922216207e-05,
|
331 |
+
"loss": 0.0193,
|
332 |
+
"step": 920
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 0.5089334055224689,
|
336 |
+
"grad_norm": 0.07586075365543365,
|
337 |
+
"learning_rate": 1.6610720086626962e-05,
|
338 |
+
"loss": 0.0188,
|
339 |
+
"step": 940
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.5197617758527342,
|
343 |
+
"grad_norm": 0.05757075175642967,
|
344 |
+
"learning_rate": 1.6538530951091863e-05,
|
345 |
+
"loss": 0.018,
|
346 |
+
"step": 960
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 0.5305901461829995,
|
350 |
+
"grad_norm": 0.05930281803011894,
|
351 |
+
"learning_rate": 1.646634181555676e-05,
|
352 |
+
"loss": 0.0157,
|
353 |
+
"step": 980
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 0.5414185165132648,
|
357 |
+
"grad_norm": 0.04882697016000748,
|
358 |
+
"learning_rate": 1.6394152680021657e-05,
|
359 |
+
"loss": 0.016,
|
360 |
+
"step": 1000
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 0.5522468868435301,
|
364 |
+
"grad_norm": 0.047697920352220535,
|
365 |
+
"learning_rate": 1.6321963544486557e-05,
|
366 |
+
"loss": 0.0161,
|
367 |
+
"step": 1020
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 0.5630752571737954,
|
371 |
+
"grad_norm": 0.0575709231197834,
|
372 |
+
"learning_rate": 1.6249774408951454e-05,
|
373 |
+
"loss": 0.0144,
|
374 |
+
"step": 1040
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 0.5739036275040607,
|
378 |
+
"grad_norm": 0.07101713865995407,
|
379 |
+
"learning_rate": 1.617758527341635e-05,
|
380 |
+
"loss": 0.0146,
|
381 |
+
"step": 1060
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 0.584731997834326,
|
385 |
+
"grad_norm": 0.04585053771734238,
|
386 |
+
"learning_rate": 1.6105396137881252e-05,
|
387 |
+
"loss": 0.0138,
|
388 |
+
"step": 1080
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 0.5955603681645912,
|
392 |
+
"grad_norm": 0.05117352306842804,
|
393 |
+
"learning_rate": 1.6033207002346146e-05,
|
394 |
+
"loss": 0.0126,
|
395 |
+
"step": 1100
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 0.6063887384948565,
|
399 |
+
"grad_norm": 0.047820210456848145,
|
400 |
+
"learning_rate": 1.5961017866811046e-05,
|
401 |
+
"loss": 0.0129,
|
402 |
+
"step": 1120
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 0.6172171088251218,
|
406 |
+
"grad_norm": 0.04630187153816223,
|
407 |
+
"learning_rate": 1.5888828731275943e-05,
|
408 |
+
"loss": 0.0129,
|
409 |
+
"step": 1140
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"epoch": 0.6280454791553871,
|
413 |
+
"grad_norm": 0.04581563547253609,
|
414 |
+
"learning_rate": 1.581663959574084e-05,
|
415 |
+
"loss": 0.013,
|
416 |
+
"step": 1160
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 0.6388738494856524,
|
420 |
+
"grad_norm": 0.04178199544548988,
|
421 |
+
"learning_rate": 1.574445046020574e-05,
|
422 |
+
"loss": 0.0118,
|
423 |
+
"step": 1180
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 0.6497022198159177,
|
427 |
+
"grad_norm": 0.0400426909327507,
|
428 |
+
"learning_rate": 1.5672261324670638e-05,
|
429 |
+
"loss": 0.0117,
|
430 |
+
"step": 1200
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 0.660530590146183,
|
434 |
+
"grad_norm": 0.03873790055513382,
|
435 |
+
"learning_rate": 1.5600072189135535e-05,
|
436 |
+
"loss": 0.0114,
|
437 |
+
"step": 1220
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 0.6713589604764483,
|
441 |
+
"grad_norm": 0.05929722636938095,
|
442 |
+
"learning_rate": 1.5527883053600436e-05,
|
443 |
+
"loss": 0.0113,
|
444 |
+
"step": 1240
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 0.6821873308067136,
|
448 |
+
"grad_norm": 0.03565024584531784,
|
449 |
+
"learning_rate": 1.5455693918065333e-05,
|
450 |
+
"loss": 0.0103,
|
451 |
+
"step": 1260
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 0.6930157011369789,
|
455 |
+
"grad_norm": 0.04300345107913017,
|
456 |
+
"learning_rate": 1.538350478253023e-05,
|
457 |
+
"loss": 0.0109,
|
458 |
+
"step": 1280
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 0.7038440714672441,
|
462 |
+
"grad_norm": 0.04502050206065178,
|
463 |
+
"learning_rate": 1.5311315646995127e-05,
|
464 |
+
"loss": 0.0102,
|
465 |
+
"step": 1300
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 0.7146724417975094,
|
469 |
+
"grad_norm": 0.0350734144449234,
|
470 |
+
"learning_rate": 1.5239126511460027e-05,
|
471 |
+
"loss": 0.01,
|
472 |
+
"step": 1320
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"epoch": 0.7255008121277747,
|
476 |
+
"grad_norm": 0.03979852795600891,
|
477 |
+
"learning_rate": 1.5166937375924925e-05,
|
478 |
+
"loss": 0.0093,
|
479 |
+
"step": 1340
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 0.73632918245804,
|
483 |
+
"grad_norm": 0.032689835876226425,
|
484 |
+
"learning_rate": 1.5094748240389823e-05,
|
485 |
+
"loss": 0.0093,
|
486 |
+
"step": 1360
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 0.7471575527883053,
|
490 |
+
"grad_norm": 0.03613670915365219,
|
491 |
+
"learning_rate": 1.5022559104854722e-05,
|
492 |
+
"loss": 0.0089,
|
493 |
+
"step": 1380
|
494 |
+
},
|
495 |
+
{
|
496 |
+
"epoch": 0.7579859231185706,
|
497 |
+
"grad_norm": 0.034867897629737854,
|
498 |
+
"learning_rate": 1.4950369969319618e-05,
|
499 |
+
"loss": 0.0089,
|
500 |
+
"step": 1400
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 0.7688142934488359,
|
504 |
+
"grad_norm": 0.0316634476184845,
|
505 |
+
"learning_rate": 1.4878180833784516e-05,
|
506 |
+
"loss": 0.0083,
|
507 |
+
"step": 1420
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 0.7796426637791013,
|
511 |
+
"grad_norm": 0.030254648998379707,
|
512 |
+
"learning_rate": 1.4805991698249415e-05,
|
513 |
+
"loss": 0.0081,
|
514 |
+
"step": 1440
|
515 |
+
},
|
516 |
+
{
|
517 |
+
"epoch": 0.7904710341093666,
|
518 |
+
"grad_norm": 0.031286563724279404,
|
519 |
+
"learning_rate": 1.4733802562714312e-05,
|
520 |
+
"loss": 0.0084,
|
521 |
+
"step": 1460
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 0.8012994044396319,
|
525 |
+
"grad_norm": 0.04057060182094574,
|
526 |
+
"learning_rate": 1.4661613427179211e-05,
|
527 |
+
"loss": 0.0083,
|
528 |
+
"step": 1480
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 0.8121277747698972,
|
532 |
+
"grad_norm": 0.027568308636546135,
|
533 |
+
"learning_rate": 1.458942429164411e-05,
|
534 |
+
"loss": 0.0079,
|
535 |
+
"step": 1500
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"epoch": 0.8229561451001625,
|
539 |
+
"grad_norm": 0.0294258501380682,
|
540 |
+
"learning_rate": 1.4517235156109007e-05,
|
541 |
+
"loss": 0.0079,
|
542 |
+
"step": 1520
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 0.8337845154304278,
|
546 |
+
"grad_norm": 0.023077772930264473,
|
547 |
+
"learning_rate": 1.4445046020573906e-05,
|
548 |
+
"loss": 0.0075,
|
549 |
+
"step": 1540
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 0.844612885760693,
|
553 |
+
"grad_norm": 0.02644222415983677,
|
554 |
+
"learning_rate": 1.4372856885038805e-05,
|
555 |
+
"loss": 0.0071,
|
556 |
+
"step": 1560
|
557 |
+
},
|
558 |
+
{
|
559 |
+
"epoch": 0.8554412560909583,
|
560 |
+
"grad_norm": 0.027747342362999916,
|
561 |
+
"learning_rate": 1.43006677495037e-05,
|
562 |
+
"loss": 0.0073,
|
563 |
+
"step": 1580
|
564 |
+
},
|
565 |
+
{
|
566 |
+
"epoch": 0.8662696264212236,
|
567 |
+
"grad_norm": 0.02826724573969841,
|
568 |
+
"learning_rate": 1.4228478613968599e-05,
|
569 |
+
"loss": 0.0075,
|
570 |
+
"step": 1600
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 0.8770979967514889,
|
574 |
+
"grad_norm": 0.04404623061418533,
|
575 |
+
"learning_rate": 1.4156289478433496e-05,
|
576 |
+
"loss": 0.0067,
|
577 |
+
"step": 1620
|
578 |
+
},
|
579 |
+
{
|
580 |
+
"epoch": 0.8879263670817542,
|
581 |
+
"grad_norm": 0.032218191772699356,
|
582 |
+
"learning_rate": 1.4084100342898395e-05,
|
583 |
+
"loss": 0.0069,
|
584 |
+
"step": 1640
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 0.8987547374120195,
|
588 |
+
"grad_norm": 0.0255202054977417,
|
589 |
+
"learning_rate": 1.4011911207363293e-05,
|
590 |
+
"loss": 0.0065,
|
591 |
+
"step": 1660
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 0.9095831077422848,
|
595 |
+
"grad_norm": 0.027019904926419258,
|
596 |
+
"learning_rate": 1.393972207182819e-05,
|
597 |
+
"loss": 0.0067,
|
598 |
+
"step": 1680
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 0.9204114780725501,
|
602 |
+
"grad_norm": 0.032661229372024536,
|
603 |
+
"learning_rate": 1.386753293629309e-05,
|
604 |
+
"loss": 0.0063,
|
605 |
+
"step": 1700
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 0.9312398484028154,
|
609 |
+
"grad_norm": 0.02468557097017765,
|
610 |
+
"learning_rate": 1.3795343800757988e-05,
|
611 |
+
"loss": 0.0064,
|
612 |
+
"step": 1720
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 0.9420682187330807,
|
616 |
+
"grad_norm": 0.02315792441368103,
|
617 |
+
"learning_rate": 1.3723154665222884e-05,
|
618 |
+
"loss": 0.006,
|
619 |
+
"step": 1740
|
620 |
+
},
|
621 |
+
{
|
622 |
+
"epoch": 0.952896589063346,
|
623 |
+
"grad_norm": 0.02668326534330845,
|
624 |
+
"learning_rate": 1.3650965529687782e-05,
|
625 |
+
"loss": 0.0059,
|
626 |
+
"step": 1760
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 0.9637249593936112,
|
630 |
+
"grad_norm": 0.03477375581860542,
|
631 |
+
"learning_rate": 1.3578776394152681e-05,
|
632 |
+
"loss": 0.0064,
|
633 |
+
"step": 1780
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 0.9745533297238765,
|
637 |
+
"grad_norm": 0.03006615862250328,
|
638 |
+
"learning_rate": 1.3506587258617578e-05,
|
639 |
+
"loss": 0.0061,
|
640 |
+
"step": 1800
|
641 |
+
},
|
642 |
+
{
|
643 |
+
"epoch": 0.9853817000541418,
|
644 |
+
"grad_norm": 0.024881117045879364,
|
645 |
+
"learning_rate": 1.3434398123082477e-05,
|
646 |
+
"loss": 0.0056,
|
647 |
+
"step": 1820
|
648 |
+
},
|
649 |
+
{
|
650 |
+
"epoch": 0.9962100703844071,
|
651 |
+
"grad_norm": 0.019141867756843567,
|
652 |
+
"learning_rate": 1.3362208987547376e-05,
|
653 |
+
"loss": 0.0057,
|
654 |
+
"step": 1840
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 1.0070384407146724,
|
658 |
+
"grad_norm": 0.03261445090174675,
|
659 |
+
"learning_rate": 1.3290019852012273e-05,
|
660 |
+
"loss": 0.0053,
|
661 |
+
"step": 1860
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"epoch": 1.0178668110449378,
|
665 |
+
"grad_norm": 0.022652000188827515,
|
666 |
+
"learning_rate": 1.3217830716477172e-05,
|
667 |
+
"loss": 0.0054,
|
668 |
+
"step": 1880
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 1.028695181375203,
|
672 |
+
"grad_norm": 0.02832789532840252,
|
673 |
+
"learning_rate": 1.314564158094207e-05,
|
674 |
+
"loss": 0.0055,
|
675 |
+
"step": 1900
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 1.0395235517054684,
|
679 |
+
"grad_norm": 0.019258731976151466,
|
680 |
+
"learning_rate": 1.3073452445406966e-05,
|
681 |
+
"loss": 0.0054,
|
682 |
+
"step": 1920
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 1.0503519220357336,
|
686 |
+
"grad_norm": 0.017714599147439003,
|
687 |
+
"learning_rate": 1.3001263309871865e-05,
|
688 |
+
"loss": 0.0053,
|
689 |
+
"step": 1940
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 1.061180292365999,
|
693 |
+
"grad_norm": 0.02675594761967659,
|
694 |
+
"learning_rate": 1.2929074174336763e-05,
|
695 |
+
"loss": 0.0049,
|
696 |
+
"step": 1960
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 1.0720086626962642,
|
700 |
+
"grad_norm": 0.018239008262753487,
|
701 |
+
"learning_rate": 1.285688503880166e-05,
|
702 |
+
"loss": 0.0051,
|
703 |
+
"step": 1980
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 1.0828370330265296,
|
707 |
+
"grad_norm": 0.024490129202604294,
|
708 |
+
"learning_rate": 1.278469590326656e-05,
|
709 |
+
"loss": 0.005,
|
710 |
+
"step": 2000
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 1.0936654033567947,
|
714 |
+
"grad_norm": 0.02328975684940815,
|
715 |
+
"learning_rate": 1.2712506767731458e-05,
|
716 |
+
"loss": 0.0049,
|
717 |
+
"step": 2020
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 1.1044937736870601,
|
721 |
+
"grad_norm": 0.027897818014025688,
|
722 |
+
"learning_rate": 1.2640317632196355e-05,
|
723 |
+
"loss": 0.005,
|
724 |
+
"step": 2040
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 1.1153221440173253,
|
728 |
+
"grad_norm": 0.015555203892290592,
|
729 |
+
"learning_rate": 1.2568128496661254e-05,
|
730 |
+
"loss": 0.0045,
|
731 |
+
"step": 2060
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 1.1261505143475907,
|
735 |
+
"grad_norm": 0.016391275450587273,
|
736 |
+
"learning_rate": 1.2495939361126153e-05,
|
737 |
+
"loss": 0.0046,
|
738 |
+
"step": 2080
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 1.136978884677856,
|
742 |
+
"grad_norm": 0.017521198838949203,
|
743 |
+
"learning_rate": 1.242375022559105e-05,
|
744 |
+
"loss": 0.0045,
|
745 |
+
"step": 2100
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 1.1478072550081213,
|
749 |
+
"grad_norm": 0.01668621227145195,
|
750 |
+
"learning_rate": 1.2351561090055949e-05,
|
751 |
+
"loss": 0.0044,
|
752 |
+
"step": 2120
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 1.1586356253383865,
|
756 |
+
"grad_norm": 0.015800384804606438,
|
757 |
+
"learning_rate": 1.2279371954520844e-05,
|
758 |
+
"loss": 0.0045,
|
759 |
+
"step": 2140
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 1.169463995668652,
|
763 |
+
"grad_norm": 0.028260117396712303,
|
764 |
+
"learning_rate": 1.2207182818985743e-05,
|
765 |
+
"loss": 0.0045,
|
766 |
+
"step": 2160
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 1.180292365998917,
|
770 |
+
"grad_norm": 0.015655677765607834,
|
771 |
+
"learning_rate": 1.2134993683450642e-05,
|
772 |
+
"loss": 0.0041,
|
773 |
+
"step": 2180
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 1.1911207363291825,
|
777 |
+
"grad_norm": 0.01357643585652113,
|
778 |
+
"learning_rate": 1.2062804547915539e-05,
|
779 |
+
"loss": 0.0042,
|
780 |
+
"step": 2200
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 1.2019491066594479,
|
784 |
+
"grad_norm": 0.013204825110733509,
|
785 |
+
"learning_rate": 1.1990615412380438e-05,
|
786 |
+
"loss": 0.0043,
|
787 |
+
"step": 2220
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 1.212777476989713,
|
791 |
+
"grad_norm": 0.01776127517223358,
|
792 |
+
"learning_rate": 1.1918426276845336e-05,
|
793 |
+
"loss": 0.0044,
|
794 |
+
"step": 2240
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 1.2236058473199782,
|
798 |
+
"grad_norm": 0.02663789503276348,
|
799 |
+
"learning_rate": 1.1846237141310234e-05,
|
800 |
+
"loss": 0.0041,
|
801 |
+
"step": 2260
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 1.2344342176502436,
|
805 |
+
"grad_norm": 0.016948629170656204,
|
806 |
+
"learning_rate": 1.1774048005775132e-05,
|
807 |
+
"loss": 0.0038,
|
808 |
+
"step": 2280
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 1.245262587980509,
|
812 |
+
"grad_norm": 0.030479852110147476,
|
813 |
+
"learning_rate": 1.1701858870240031e-05,
|
814 |
+
"loss": 0.004,
|
815 |
+
"step": 2300
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 1.2560909583107742,
|
819 |
+
"grad_norm": 0.020524190738797188,
|
820 |
+
"learning_rate": 1.1629669734704927e-05,
|
821 |
+
"loss": 0.0039,
|
822 |
+
"step": 2320
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 1.2669193286410394,
|
826 |
+
"grad_norm": 0.027631113305687904,
|
827 |
+
"learning_rate": 1.1557480599169825e-05,
|
828 |
+
"loss": 0.0038,
|
829 |
+
"step": 2340
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 1.2777476989713048,
|
833 |
+
"grad_norm": 0.017368769273161888,
|
834 |
+
"learning_rate": 1.1485291463634724e-05,
|
835 |
+
"loss": 0.0038,
|
836 |
+
"step": 2360
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 1.2885760693015702,
|
840 |
+
"grad_norm": 0.013879314996302128,
|
841 |
+
"learning_rate": 1.1413102328099621e-05,
|
842 |
+
"loss": 0.0038,
|
843 |
+
"step": 2380
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 1.2994044396318354,
|
847 |
+
"grad_norm": 0.03698274865746498,
|
848 |
+
"learning_rate": 1.134091319256452e-05,
|
849 |
+
"loss": 0.0039,
|
850 |
+
"step": 2400
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"epoch": 1.3102328099621008,
|
854 |
+
"grad_norm": 0.015918811783194542,
|
855 |
+
"learning_rate": 1.1268724057029419e-05,
|
856 |
+
"loss": 0.0038,
|
857 |
+
"step": 2420
|
858 |
+
},
|
859 |
+
{
|
860 |
+
"epoch": 1.321061180292366,
|
861 |
+
"grad_norm": 0.01589634083211422,
|
862 |
+
"learning_rate": 1.1196534921494316e-05,
|
863 |
+
"loss": 0.0039,
|
864 |
+
"step": 2440
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 1.3318895506226314,
|
868 |
+
"grad_norm": 0.01514554861932993,
|
869 |
+
"learning_rate": 1.1124345785959215e-05,
|
870 |
+
"loss": 0.0034,
|
871 |
+
"step": 2460
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"epoch": 1.3427179209528965,
|
875 |
+
"grad_norm": 0.013614562340080738,
|
876 |
+
"learning_rate": 1.1052156650424114e-05,
|
877 |
+
"loss": 0.0037,
|
878 |
+
"step": 2480
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 1.353546291283162,
|
882 |
+
"grad_norm": 0.018234172835946083,
|
883 |
+
"learning_rate": 1.0979967514889009e-05,
|
884 |
+
"loss": 0.0034,
|
885 |
+
"step": 2500
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 1.3643746616134271,
|
889 |
+
"grad_norm": 0.013683984987437725,
|
890 |
+
"learning_rate": 1.0907778379353908e-05,
|
891 |
+
"loss": 0.0034,
|
892 |
+
"step": 2520
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 1.3752030319436925,
|
896 |
+
"grad_norm": 0.014102173037827015,
|
897 |
+
"learning_rate": 1.0835589243818807e-05,
|
898 |
+
"loss": 0.0033,
|
899 |
+
"step": 2540
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 1.3860314022739577,
|
903 |
+
"grad_norm": 0.012239793315529823,
|
904 |
+
"learning_rate": 1.0763400108283704e-05,
|
905 |
+
"loss": 0.0032,
|
906 |
+
"step": 2560
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 1.3968597726042231,
|
910 |
+
"grad_norm": 0.013394709676504135,
|
911 |
+
"learning_rate": 1.0691210972748602e-05,
|
912 |
+
"loss": 0.0033,
|
913 |
+
"step": 2580
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 1.4076881429344883,
|
917 |
+
"grad_norm": 0.012910670600831509,
|
918 |
+
"learning_rate": 1.0619021837213501e-05,
|
919 |
+
"loss": 0.0033,
|
920 |
+
"step": 2600
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 1.4185165132647537,
|
924 |
+
"grad_norm": 0.0153429564088583,
|
925 |
+
"learning_rate": 1.0546832701678398e-05,
|
926 |
+
"loss": 0.0033,
|
927 |
+
"step": 2620
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 1.4293448835950189,
|
931 |
+
"grad_norm": 0.01714710146188736,
|
932 |
+
"learning_rate": 1.0474643566143297e-05,
|
933 |
+
"loss": 0.003,
|
934 |
+
"step": 2640
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 1.4401732539252843,
|
938 |
+
"grad_norm": 0.010895042680203915,
|
939 |
+
"learning_rate": 1.0402454430608196e-05,
|
940 |
+
"loss": 0.0033,
|
941 |
+
"step": 2660
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 1.4510016242555495,
|
945 |
+
"grad_norm": 0.012086061760783195,
|
946 |
+
"learning_rate": 1.0330265295073091e-05,
|
947 |
+
"loss": 0.0031,
|
948 |
+
"step": 2680
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 1.4618299945858149,
|
952 |
+
"grad_norm": 0.012154944241046906,
|
953 |
+
"learning_rate": 1.025807615953799e-05,
|
954 |
+
"loss": 0.003,
|
955 |
+
"step": 2700
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 1.47265836491608,
|
959 |
+
"grad_norm": 0.016563011333346367,
|
960 |
+
"learning_rate": 1.0185887024002887e-05,
|
961 |
+
"loss": 0.0032,
|
962 |
+
"step": 2720
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 1.4834867352463454,
|
966 |
+
"grad_norm": 0.011688679456710815,
|
967 |
+
"learning_rate": 1.0113697888467786e-05,
|
968 |
+
"loss": 0.003,
|
969 |
+
"step": 2740
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 1.4943151055766108,
|
973 |
+
"grad_norm": 0.010492874309420586,
|
974 |
+
"learning_rate": 1.0041508752932685e-05,
|
975 |
+
"loss": 0.0029,
|
976 |
+
"step": 2760
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 1.505143475906876,
|
980 |
+
"grad_norm": 0.01102561503648758,
|
981 |
+
"learning_rate": 9.969319617397582e-06,
|
982 |
+
"loss": 0.0029,
|
983 |
+
"step": 2780
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 1.5159718462371412,
|
987 |
+
"grad_norm": 0.013833099976181984,
|
988 |
+
"learning_rate": 9.89713048186248e-06,
|
989 |
+
"loss": 0.003,
|
990 |
+
"step": 2800
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 1.5268002165674066,
|
994 |
+
"grad_norm": 0.012262464500963688,
|
995 |
+
"learning_rate": 9.82494134632738e-06,
|
996 |
+
"loss": 0.0029,
|
997 |
+
"step": 2820
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 1.537628586897672,
|
1001 |
+
"grad_norm": 0.010000896640121937,
|
1002 |
+
"learning_rate": 9.752752210792277e-06,
|
1003 |
+
"loss": 0.0028,
|
1004 |
+
"step": 2840
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 1.5484569572279372,
|
1008 |
+
"grad_norm": 0.008956516161561012,
|
1009 |
+
"learning_rate": 9.680563075257174e-06,
|
1010 |
+
"loss": 0.003,
|
1011 |
+
"step": 2860
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 1.5592853275582024,
|
1015 |
+
"grad_norm": 0.01270230021327734,
|
1016 |
+
"learning_rate": 9.608373939722073e-06,
|
1017 |
+
"loss": 0.0028,
|
1018 |
+
"step": 2880
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 1.5701136978884678,
|
1022 |
+
"grad_norm": 0.012823579832911491,
|
1023 |
+
"learning_rate": 9.536184804186971e-06,
|
1024 |
+
"loss": 0.0028,
|
1025 |
+
"step": 2900
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 1.5809420682187332,
|
1029 |
+
"grad_norm": 0.009921281598508358,
|
1030 |
+
"learning_rate": 9.463995668651868e-06,
|
1031 |
+
"loss": 0.0027,
|
1032 |
+
"step": 2920
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 1.5917704385489984,
|
1036 |
+
"grad_norm": 0.021674729883670807,
|
1037 |
+
"learning_rate": 9.391806533116766e-06,
|
1038 |
+
"loss": 0.0027,
|
1039 |
+
"step": 2940
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"epoch": 1.6025988088792635,
|
1043 |
+
"grad_norm": 0.012086630798876286,
|
1044 |
+
"learning_rate": 9.319617397581664e-06,
|
1045 |
+
"loss": 0.0027,
|
1046 |
+
"step": 2960
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"epoch": 1.613427179209529,
|
1050 |
+
"grad_norm": 0.013828184455633163,
|
1051 |
+
"learning_rate": 9.247428262046563e-06,
|
1052 |
+
"loss": 0.0028,
|
1053 |
+
"step": 2980
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 1.6242555495397943,
|
1057 |
+
"grad_norm": 0.01388394832611084,
|
1058 |
+
"learning_rate": 9.17523912651146e-06,
|
1059 |
+
"loss": 0.0027,
|
1060 |
+
"step": 3000
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 1.6350839198700595,
|
1064 |
+
"grad_norm": 0.009276646189391613,
|
1065 |
+
"learning_rate": 9.103049990976359e-06,
|
1066 |
+
"loss": 0.0026,
|
1067 |
+
"step": 3020
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"epoch": 1.6459122902003247,
|
1071 |
+
"grad_norm": 0.010650601238012314,
|
1072 |
+
"learning_rate": 9.030860855441256e-06,
|
1073 |
+
"loss": 0.0026,
|
1074 |
+
"step": 3040
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 1.65674066053059,
|
1078 |
+
"grad_norm": 0.009121935814619064,
|
1079 |
+
"learning_rate": 8.958671719906155e-06,
|
1080 |
+
"loss": 0.0026,
|
1081 |
+
"step": 3060
|
1082 |
+
},
|
1083 |
+
{
|
1084 |
+
"epoch": 1.6675690308608555,
|
1085 |
+
"grad_norm": 0.012495677918195724,
|
1086 |
+
"learning_rate": 8.886482584371054e-06,
|
1087 |
+
"loss": 0.0026,
|
1088 |
+
"step": 3080
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"epoch": 1.6783974011911207,
|
1092 |
+
"grad_norm": 0.01117989607155323,
|
1093 |
+
"learning_rate": 8.81429344883595e-06,
|
1094 |
+
"loss": 0.0026,
|
1095 |
+
"step": 3100
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 1.6892257715213859,
|
1099 |
+
"grad_norm": 0.010890294797718525,
|
1100 |
+
"learning_rate": 8.74210431330085e-06,
|
1101 |
+
"loss": 0.0024,
|
1102 |
+
"step": 3120
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 1.7000541418516515,
|
1106 |
+
"grad_norm": 0.009491957724094391,
|
1107 |
+
"learning_rate": 8.669915177765747e-06,
|
1108 |
+
"loss": 0.0025,
|
1109 |
+
"step": 3140
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 1.7108825121819167,
|
1113 |
+
"grad_norm": 0.011014943942427635,
|
1114 |
+
"learning_rate": 8.597726042230645e-06,
|
1115 |
+
"loss": 0.0026,
|
1116 |
+
"step": 3160
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"epoch": 1.7217108825121819,
|
1120 |
+
"grad_norm": 0.015149922110140324,
|
1121 |
+
"learning_rate": 8.525536906695543e-06,
|
1122 |
+
"loss": 0.0024,
|
1123 |
+
"step": 3180
|
1124 |
+
},
|
1125 |
+
{
|
1126 |
+
"epoch": 1.7325392528424473,
|
1127 |
+
"grad_norm": 0.011757172644138336,
|
1128 |
+
"learning_rate": 8.453347771160441e-06,
|
1129 |
+
"loss": 0.0023,
|
1130 |
+
"step": 3200
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 1.7433676231727127,
|
1134 |
+
"grad_norm": 0.007756380829960108,
|
1135 |
+
"learning_rate": 8.38115863562534e-06,
|
1136 |
+
"loss": 0.0023,
|
1137 |
+
"step": 3220
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 1.7541959935029778,
|
1141 |
+
"grad_norm": 0.010520155541598797,
|
1142 |
+
"learning_rate": 8.308969500090237e-06,
|
1143 |
+
"loss": 0.0023,
|
1144 |
+
"step": 3240
|
1145 |
+
},
|
1146 |
+
{
|
1147 |
+
"epoch": 1.765024363833243,
|
1148 |
+
"grad_norm": 0.009943107143044472,
|
1149 |
+
"learning_rate": 8.236780364555134e-06,
|
1150 |
+
"loss": 0.0023,
|
1151 |
+
"step": 3260
|
1152 |
+
},
|
1153 |
+
{
|
1154 |
+
"epoch": 1.7758527341635084,
|
1155 |
+
"grad_norm": 0.011330407112836838,
|
1156 |
+
"learning_rate": 8.164591229020033e-06,
|
1157 |
+
"loss": 0.0023,
|
1158 |
+
"step": 3280
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"epoch": 1.7866811044937738,
|
1162 |
+
"grad_norm": 0.013512694276869297,
|
1163 |
+
"learning_rate": 8.092402093484932e-06,
|
1164 |
+
"loss": 0.0023,
|
1165 |
+
"step": 3300
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"epoch": 1.797509474824039,
|
1169 |
+
"grad_norm": 0.010237209498882294,
|
1170 |
+
"learning_rate": 8.020212957949829e-06,
|
1171 |
+
"loss": 0.0023,
|
1172 |
+
"step": 3320
|
1173 |
+
},
|
1174 |
+
{
|
1175 |
+
"epoch": 1.8083378451543042,
|
1176 |
+
"grad_norm": 0.013071142137050629,
|
1177 |
+
"learning_rate": 7.948023822414728e-06,
|
1178 |
+
"loss": 0.0022,
|
1179 |
+
"step": 3340
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 1.8191662154845696,
|
1183 |
+
"grad_norm": 0.0089937848970294,
|
1184 |
+
"learning_rate": 7.875834686879625e-06,
|
1185 |
+
"loss": 0.0024,
|
1186 |
+
"step": 3360
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 1.829994585814835,
|
1190 |
+
"grad_norm": 0.01620866358280182,
|
1191 |
+
"learning_rate": 7.803645551344524e-06,
|
1192 |
+
"loss": 0.0022,
|
1193 |
+
"step": 3380
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 1.8408229561451002,
|
1197 |
+
"grad_norm": 0.009367492981255054,
|
1198 |
+
"learning_rate": 7.731456415809423e-06,
|
1199 |
+
"loss": 0.0021,
|
1200 |
+
"step": 3400
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 1.8516513264753653,
|
1204 |
+
"grad_norm": 0.006648328620940447,
|
1205 |
+
"learning_rate": 7.65926728027432e-06,
|
1206 |
+
"loss": 0.0021,
|
1207 |
+
"step": 3420
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 1.8624796968056307,
|
1211 |
+
"grad_norm": 0.013929731212556362,
|
1212 |
+
"learning_rate": 7.587078144739218e-06,
|
1213 |
+
"loss": 0.0021,
|
1214 |
+
"step": 3440
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 1.8733080671358961,
|
1218 |
+
"grad_norm": 0.014050954952836037,
|
1219 |
+
"learning_rate": 7.514889009204115e-06,
|
1220 |
+
"loss": 0.0023,
|
1221 |
+
"step": 3460
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 1.8841364374661613,
|
1225 |
+
"grad_norm": 0.011463869363069534,
|
1226 |
+
"learning_rate": 7.4426998736690135e-06,
|
1227 |
+
"loss": 0.002,
|
1228 |
+
"step": 3480
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 1.8949648077964265,
|
1232 |
+
"grad_norm": 0.008514183573424816,
|
1233 |
+
"learning_rate": 7.3705107381339114e-06,
|
1234 |
+
"loss": 0.0021,
|
1235 |
+
"step": 3500
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 1.905793178126692,
|
1239 |
+
"grad_norm": 0.010145863518118858,
|
1240 |
+
"learning_rate": 7.298321602598809e-06,
|
1241 |
+
"loss": 0.0021,
|
1242 |
+
"step": 3520
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 1.9166215484569573,
|
1246 |
+
"grad_norm": 0.008266255259513855,
|
1247 |
+
"learning_rate": 7.226132467063708e-06,
|
1248 |
+
"loss": 0.0023,
|
1249 |
+
"step": 3540
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 1.9274499187872225,
|
1253 |
+
"grad_norm": 0.008439561352133751,
|
1254 |
+
"learning_rate": 7.153943331528605e-06,
|
1255 |
+
"loss": 0.002,
|
1256 |
+
"step": 3560
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 1.9382782891174877,
|
1260 |
+
"grad_norm": 0.007920574396848679,
|
1261 |
+
"learning_rate": 7.081754195993503e-06,
|
1262 |
+
"loss": 0.0021,
|
1263 |
+
"step": 3580
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 1.949106659447753,
|
1267 |
+
"grad_norm": 0.01343179028481245,
|
1268 |
+
"learning_rate": 7.009565060458402e-06,
|
1269 |
+
"loss": 0.0021,
|
1270 |
+
"step": 3600
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 1.9599350297780185,
|
1274 |
+
"grad_norm": 0.008394268341362476,
|
1275 |
+
"learning_rate": 6.9373759249233e-06,
|
1276 |
+
"loss": 0.002,
|
1277 |
+
"step": 3620
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 1.9707634001082837,
|
1281 |
+
"grad_norm": 0.007805489934980869,
|
1282 |
+
"learning_rate": 6.865186789388197e-06,
|
1283 |
+
"loss": 0.0019,
|
1284 |
+
"step": 3640
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 1.9815917704385488,
|
1288 |
+
"grad_norm": 0.00863006990402937,
|
1289 |
+
"learning_rate": 6.792997653853096e-06,
|
1290 |
+
"loss": 0.0021,
|
1291 |
+
"step": 3660
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 1.9924201407688142,
|
1295 |
+
"grad_norm": 0.011791090480983257,
|
1296 |
+
"learning_rate": 6.720808518317994e-06,
|
1297 |
+
"loss": 0.002,
|
1298 |
+
"step": 3680
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 2.0032485110990796,
|
1302 |
+
"grad_norm": 0.008283623494207859,
|
1303 |
+
"learning_rate": 6.648619382782892e-06,
|
1304 |
+
"loss": 0.0018,
|
1305 |
+
"step": 3700
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 2.014076881429345,
|
1309 |
+
"grad_norm": 0.006815766915678978,
|
1310 |
+
"learning_rate": 6.576430247247789e-06,
|
1311 |
+
"loss": 0.0018,
|
1312 |
+
"step": 3720
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 2.02490525175961,
|
1316 |
+
"grad_norm": 0.009198131039738655,
|
1317 |
+
"learning_rate": 6.504241111712688e-06,
|
1318 |
+
"loss": 0.0019,
|
1319 |
+
"step": 3740
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 2.0357336220898756,
|
1323 |
+
"grad_norm": 0.007912855595350266,
|
1324 |
+
"learning_rate": 6.432051976177586e-06,
|
1325 |
+
"loss": 0.0019,
|
1326 |
+
"step": 3760
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 2.046561992420141,
|
1330 |
+
"grad_norm": 0.008341687731444836,
|
1331 |
+
"learning_rate": 6.3598628406424836e-06,
|
1332 |
+
"loss": 0.0018,
|
1333 |
+
"step": 3780
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 2.057390362750406,
|
1337 |
+
"grad_norm": 0.007925598882138729,
|
1338 |
+
"learning_rate": 6.287673705107382e-06,
|
1339 |
+
"loss": 0.0018,
|
1340 |
+
"step": 3800
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 2.068218733080671,
|
1344 |
+
"grad_norm": 0.006331819575279951,
|
1345 |
+
"learning_rate": 6.2154845695722795e-06,
|
1346 |
+
"loss": 0.0018,
|
1347 |
+
"step": 3820
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 2.079047103410937,
|
1351 |
+
"grad_norm": 0.00803130678832531,
|
1352 |
+
"learning_rate": 6.143295434037177e-06,
|
1353 |
+
"loss": 0.0019,
|
1354 |
+
"step": 3840
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 2.089875473741202,
|
1358 |
+
"grad_norm": 0.009752047248184681,
|
1359 |
+
"learning_rate": 6.071106298502076e-06,
|
1360 |
+
"loss": 0.0019,
|
1361 |
+
"step": 3860
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 2.100703844071467,
|
1365 |
+
"grad_norm": 0.007367837708443403,
|
1366 |
+
"learning_rate": 5.998917162966974e-06,
|
1367 |
+
"loss": 0.0019,
|
1368 |
+
"step": 3880
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 2.1115322144017323,
|
1372 |
+
"grad_norm": 0.008053092285990715,
|
1373 |
+
"learning_rate": 5.926728027431872e-06,
|
1374 |
+
"loss": 0.0018,
|
1375 |
+
"step": 3900
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 2.122360584731998,
|
1379 |
+
"grad_norm": 0.008594776503741741,
|
1380 |
+
"learning_rate": 5.854538891896771e-06,
|
1381 |
+
"loss": 0.0017,
|
1382 |
+
"step": 3920
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 2.133188955062263,
|
1386 |
+
"grad_norm": 0.006970062851905823,
|
1387 |
+
"learning_rate": 5.782349756361668e-06,
|
1388 |
+
"loss": 0.0017,
|
1389 |
+
"step": 3940
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 2.1440173253925283,
|
1393 |
+
"grad_norm": 0.008627238683402538,
|
1394 |
+
"learning_rate": 5.710160620826566e-06,
|
1395 |
+
"loss": 0.0018,
|
1396 |
+
"step": 3960
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 2.154845695722794,
|
1400 |
+
"grad_norm": 0.008501927368342876,
|
1401 |
+
"learning_rate": 5.637971485291464e-06,
|
1402 |
+
"loss": 0.0018,
|
1403 |
+
"step": 3980
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 2.165674066053059,
|
1407 |
+
"grad_norm": 0.005817601457238197,
|
1408 |
+
"learning_rate": 5.565782349756363e-06,
|
1409 |
+
"loss": 0.0016,
|
1410 |
+
"step": 4000
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 2.1765024363833243,
|
1414 |
+
"grad_norm": 0.0064091249369084835,
|
1415 |
+
"learning_rate": 5.49359321422126e-06,
|
1416 |
+
"loss": 0.0018,
|
1417 |
+
"step": 4020
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 2.1873308067135895,
|
1421 |
+
"grad_norm": 0.005200565326958895,
|
1422 |
+
"learning_rate": 5.421404078686158e-06,
|
1423 |
+
"loss": 0.0016,
|
1424 |
+
"step": 4040
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 2.198159177043855,
|
1428 |
+
"grad_norm": 0.006780552677810192,
|
1429 |
+
"learning_rate": 5.3492149431510565e-06,
|
1430 |
+
"loss": 0.0018,
|
1431 |
+
"step": 4060
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 2.2089875473741203,
|
1435 |
+
"grad_norm": 0.006558467634022236,
|
1436 |
+
"learning_rate": 5.2770258076159545e-06,
|
1437 |
+
"loss": 0.0016,
|
1438 |
+
"step": 4080
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 2.2198159177043855,
|
1442 |
+
"grad_norm": 0.007784364745020866,
|
1443 |
+
"learning_rate": 5.204836672080852e-06,
|
1444 |
+
"loss": 0.0019,
|
1445 |
+
"step": 4100
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 2.2306442880346506,
|
1449 |
+
"grad_norm": 0.00484019098803401,
|
1450 |
+
"learning_rate": 5.13264753654575e-06,
|
1451 |
+
"loss": 0.0017,
|
1452 |
+
"step": 4120
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 2.2414726583649163,
|
1456 |
+
"grad_norm": 0.007469740230590105,
|
1457 |
+
"learning_rate": 5.060458401010648e-06,
|
1458 |
+
"loss": 0.0018,
|
1459 |
+
"step": 4140
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 2.2523010286951815,
|
1463 |
+
"grad_norm": 0.006617364473640919,
|
1464 |
+
"learning_rate": 4.988269265475546e-06,
|
1465 |
+
"loss": 0.0016,
|
1466 |
+
"step": 4160
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 2.2631293990254466,
|
1470 |
+
"grad_norm": 0.007724768482148647,
|
1471 |
+
"learning_rate": 4.916080129940444e-06,
|
1472 |
+
"loss": 0.0017,
|
1473 |
+
"step": 4180
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 2.273957769355712,
|
1477 |
+
"grad_norm": 0.0065683526918292046,
|
1478 |
+
"learning_rate": 4.843890994405342e-06,
|
1479 |
+
"loss": 0.0016,
|
1480 |
+
"step": 4200
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 2.2847861396859774,
|
1484 |
+
"grad_norm": 0.006633590906858444,
|
1485 |
+
"learning_rate": 4.77170185887024e-06,
|
1486 |
+
"loss": 0.0016,
|
1487 |
+
"step": 4220
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 2.2956145100162426,
|
1491 |
+
"grad_norm": 0.01162402518093586,
|
1492 |
+
"learning_rate": 4.699512723335138e-06,
|
1493 |
+
"loss": 0.0016,
|
1494 |
+
"step": 4240
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 2.306442880346508,
|
1498 |
+
"grad_norm": 0.008504554629325867,
|
1499 |
+
"learning_rate": 4.627323587800037e-06,
|
1500 |
+
"loss": 0.0015,
|
1501 |
+
"step": 4260
|
1502 |
+
},
|
1503 |
+
{
|
1504 |
+
"epoch": 2.317271250676773,
|
1505 |
+
"grad_norm": 0.00829976238310337,
|
1506 |
+
"learning_rate": 4.555134452264934e-06,
|
1507 |
+
"loss": 0.0015,
|
1508 |
+
"step": 4280
|
1509 |
+
},
|
1510 |
+
{
|
1511 |
+
"epoch": 2.3280996210070386,
|
1512 |
+
"grad_norm": 0.00552743673324585,
|
1513 |
+
"learning_rate": 4.482945316729833e-06,
|
1514 |
+
"loss": 0.0014,
|
1515 |
+
"step": 4300
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 2.338927991337304,
|
1519 |
+
"grad_norm": 0.005899062845855951,
|
1520 |
+
"learning_rate": 4.410756181194731e-06,
|
1521 |
+
"loss": 0.0016,
|
1522 |
+
"step": 4320
|
1523 |
+
},
|
1524 |
+
{
|
1525 |
+
"epoch": 2.349756361667569,
|
1526 |
+
"grad_norm": 0.008548817597329617,
|
1527 |
+
"learning_rate": 4.338567045659629e-06,
|
1528 |
+
"loss": 0.0015,
|
1529 |
+
"step": 4340
|
1530 |
+
},
|
1531 |
+
{
|
1532 |
+
"epoch": 2.360584731997834,
|
1533 |
+
"grad_norm": 0.00820871815085411,
|
1534 |
+
"learning_rate": 4.266377910124527e-06,
|
1535 |
+
"loss": 0.0017,
|
1536 |
+
"step": 4360
|
1537 |
+
},
|
1538 |
+
{
|
1539 |
+
"epoch": 2.3714131023280998,
|
1540 |
+
"grad_norm": 0.010167138651013374,
|
1541 |
+
"learning_rate": 4.1941887745894246e-06,
|
1542 |
+
"loss": 0.0016,
|
1543 |
+
"step": 4380
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 2.382241472658365,
|
1547 |
+
"grad_norm": 0.006324863061308861,
|
1548 |
+
"learning_rate": 4.1219996390543225e-06,
|
1549 |
+
"loss": 0.0016,
|
1550 |
+
"step": 4400
|
1551 |
+
},
|
1552 |
+
{
|
1553 |
+
"epoch": 2.39306984298863,
|
1554 |
+
"grad_norm": 0.00801610853523016,
|
1555 |
+
"learning_rate": 4.049810503519221e-06,
|
1556 |
+
"loss": 0.0015,
|
1557 |
+
"step": 4420
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 2.4038982133188957,
|
1561 |
+
"grad_norm": 0.006985550280660391,
|
1562 |
+
"learning_rate": 3.977621367984118e-06,
|
1563 |
+
"loss": 0.0016,
|
1564 |
+
"step": 4440
|
1565 |
+
},
|
1566 |
+
{
|
1567 |
+
"epoch": 2.414726583649161,
|
1568 |
+
"grad_norm": 0.00809464231133461,
|
1569 |
+
"learning_rate": 3.905432232449017e-06,
|
1570 |
+
"loss": 0.0015,
|
1571 |
+
"step": 4460
|
1572 |
+
},
|
1573 |
+
{
|
1574 |
+
"epoch": 2.425554953979426,
|
1575 |
+
"grad_norm": 0.0049890028312802315,
|
1576 |
+
"learning_rate": 3.833243096913914e-06,
|
1577 |
+
"loss": 0.0016,
|
1578 |
+
"step": 4480
|
1579 |
+
},
|
1580 |
+
{
|
1581 |
+
"epoch": 2.4363833243096913,
|
1582 |
+
"grad_norm": 0.0060018254444003105,
|
1583 |
+
"learning_rate": 3.7610539613788127e-06,
|
1584 |
+
"loss": 0.0015,
|
1585 |
+
"step": 4500
|
1586 |
+
},
|
1587 |
+
{
|
1588 |
+
"epoch": 2.4472116946399565,
|
1589 |
+
"grad_norm": 0.011303539387881756,
|
1590 |
+
"learning_rate": 3.688864825843711e-06,
|
1591 |
+
"loss": 0.0014,
|
1592 |
+
"step": 4520
|
1593 |
+
},
|
1594 |
+
{
|
1595 |
+
"epoch": 2.458040064970222,
|
1596 |
+
"grad_norm": 0.006434119306504726,
|
1597 |
+
"learning_rate": 3.6166756903086086e-06,
|
1598 |
+
"loss": 0.0015,
|
1599 |
+
"step": 4540
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 2.4688684353004873,
|
1603 |
+
"grad_norm": 0.01008700393140316,
|
1604 |
+
"learning_rate": 3.544486554773507e-06,
|
1605 |
+
"loss": 0.0015,
|
1606 |
+
"step": 4560
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"epoch": 2.4796968056307525,
|
1610 |
+
"grad_norm": 0.004655952099710703,
|
1611 |
+
"learning_rate": 3.4722974192384053e-06,
|
1612 |
+
"loss": 0.0015,
|
1613 |
+
"step": 4580
|
1614 |
+
},
|
1615 |
+
{
|
1616 |
+
"epoch": 2.490525175961018,
|
1617 |
+
"grad_norm": 0.009297892451286316,
|
1618 |
+
"learning_rate": 3.400108283703303e-06,
|
1619 |
+
"loss": 0.0016,
|
1620 |
+
"step": 4600
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 2.5013535462912833,
|
1624 |
+
"grad_norm": 0.007229143753647804,
|
1625 |
+
"learning_rate": 3.327919148168201e-06,
|
1626 |
+
"loss": 0.0017,
|
1627 |
+
"step": 4620
|
1628 |
+
},
|
1629 |
+
{
|
1630 |
+
"epoch": 2.5121819166215484,
|
1631 |
+
"grad_norm": 0.007766417693346739,
|
1632 |
+
"learning_rate": 3.2557300126330987e-06,
|
1633 |
+
"loss": 0.0015,
|
1634 |
+
"step": 4640
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 2.5230102869518136,
|
1638 |
+
"grad_norm": 0.010671594180166721,
|
1639 |
+
"learning_rate": 3.183540877097997e-06,
|
1640 |
+
"loss": 0.0016,
|
1641 |
+
"step": 4660
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 2.533838657282079,
|
1645 |
+
"grad_norm": 0.008643384091556072,
|
1646 |
+
"learning_rate": 3.111351741562895e-06,
|
1647 |
+
"loss": 0.0015,
|
1648 |
+
"step": 4680
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 2.5446670276123444,
|
1652 |
+
"grad_norm": 0.0042122360318899155,
|
1653 |
+
"learning_rate": 3.039162606027793e-06,
|
1654 |
+
"loss": 0.0015,
|
1655 |
+
"step": 4700
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 2.5554953979426096,
|
1659 |
+
"grad_norm": 0.00437292642891407,
|
1660 |
+
"learning_rate": 2.966973470492691e-06,
|
1661 |
+
"loss": 0.0014,
|
1662 |
+
"step": 4720
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 2.566323768272875,
|
1666 |
+
"grad_norm": 0.005281255580484867,
|
1667 |
+
"learning_rate": 2.8947843349575893e-06,
|
1668 |
+
"loss": 0.0014,
|
1669 |
+
"step": 4740
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 2.5771521386031404,
|
1673 |
+
"grad_norm": 0.007413716986775398,
|
1674 |
+
"learning_rate": 2.8225951994224873e-06,
|
1675 |
+
"loss": 0.0015,
|
1676 |
+
"step": 4760
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 2.5879805089334056,
|
1680 |
+
"grad_norm": 0.005595244932919741,
|
1681 |
+
"learning_rate": 2.750406063887385e-06,
|
1682 |
+
"loss": 0.0014,
|
1683 |
+
"step": 4780
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 2.5988088792636708,
|
1687 |
+
"grad_norm": 0.004735792987048626,
|
1688 |
+
"learning_rate": 2.678216928352283e-06,
|
1689 |
+
"loss": 0.0013,
|
1690 |
+
"step": 4800
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 2.609637249593936,
|
1694 |
+
"grad_norm": 0.005141600966453552,
|
1695 |
+
"learning_rate": 2.606027792817181e-06,
|
1696 |
+
"loss": 0.0013,
|
1697 |
+
"step": 4820
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 2.6204656199242016,
|
1701 |
+
"grad_norm": 0.0045648557133972645,
|
1702 |
+
"learning_rate": 2.5338386572820795e-06,
|
1703 |
+
"loss": 0.0014,
|
1704 |
+
"step": 4840
|
1705 |
+
},
|
1706 |
+
{
|
1707 |
+
"epoch": 2.6312939902544668,
|
1708 |
+
"grad_norm": 0.006522185634821653,
|
1709 |
+
"learning_rate": 2.4616495217469774e-06,
|
1710 |
+
"loss": 0.0014,
|
1711 |
+
"step": 4860
|
1712 |
+
},
|
1713 |
+
{
|
1714 |
+
"epoch": 2.642122360584732,
|
1715 |
+
"grad_norm": 0.010215125046670437,
|
1716 |
+
"learning_rate": 2.3894603862118754e-06,
|
1717 |
+
"loss": 0.0014,
|
1718 |
+
"step": 4880
|
1719 |
+
},
|
1720 |
+
{
|
1721 |
+
"epoch": 2.6529507309149976,
|
1722 |
+
"grad_norm": 0.007124153897166252,
|
1723 |
+
"learning_rate": 2.3172712506767733e-06,
|
1724 |
+
"loss": 0.0014,
|
1725 |
+
"step": 4900
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"epoch": 2.6637791012452627,
|
1729 |
+
"grad_norm": 0.0049623060040175915,
|
1730 |
+
"learning_rate": 2.2450821151416713e-06,
|
1731 |
+
"loss": 0.0013,
|
1732 |
+
"step": 4920
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"epoch": 2.674607471575528,
|
1736 |
+
"grad_norm": 0.005493966396898031,
|
1737 |
+
"learning_rate": 2.1728929796065692e-06,
|
1738 |
+
"loss": 0.0013,
|
1739 |
+
"step": 4940
|
1740 |
+
},
|
1741 |
+
{
|
1742 |
+
"epoch": 2.685435841905793,
|
1743 |
+
"grad_norm": 0.006936436053365469,
|
1744 |
+
"learning_rate": 2.1007038440714676e-06,
|
1745 |
+
"loss": 0.0013,
|
1746 |
+
"step": 4960
|
1747 |
+
},
|
1748 |
+
{
|
1749 |
+
"epoch": 2.6962642122360583,
|
1750 |
+
"grad_norm": 0.006746096536517143,
|
1751 |
+
"learning_rate": 2.0285147085363655e-06,
|
1752 |
+
"loss": 0.0013,
|
1753 |
+
"step": 4980
|
1754 |
+
},
|
1755 |
+
{
|
1756 |
+
"epoch": 2.707092582566324,
|
1757 |
+
"grad_norm": 0.00648007495328784,
|
1758 |
+
"learning_rate": 1.9563255730012635e-06,
|
1759 |
+
"loss": 0.0013,
|
1760 |
+
"step": 5000
|
1761 |
+
},
|
1762 |
+
{
|
1763 |
+
"epoch": 2.717920952896589,
|
1764 |
+
"grad_norm": 0.006984813138842583,
|
1765 |
+
"learning_rate": 1.8841364374661614e-06,
|
1766 |
+
"loss": 0.0015,
|
1767 |
+
"step": 5020
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 2.7287493232268543,
|
1771 |
+
"grad_norm": 0.007297712843865156,
|
1772 |
+
"learning_rate": 1.8119473019310596e-06,
|
1773 |
+
"loss": 0.0014,
|
1774 |
+
"step": 5040
|
1775 |
+
},
|
1776 |
+
{
|
1777 |
+
"epoch": 2.73957769355712,
|
1778 |
+
"grad_norm": 0.006131511647254229,
|
1779 |
+
"learning_rate": 1.7397581663959575e-06,
|
1780 |
+
"loss": 0.0013,
|
1781 |
+
"step": 5060
|
1782 |
+
},
|
1783 |
+
{
|
1784 |
+
"epoch": 2.750406063887385,
|
1785 |
+
"grad_norm": 0.0050123510882258415,
|
1786 |
+
"learning_rate": 1.6675690308608555e-06,
|
1787 |
+
"loss": 0.0013,
|
1788 |
+
"step": 5080
|
1789 |
+
},
|
1790 |
+
{
|
1791 |
+
"epoch": 2.7612344342176502,
|
1792 |
+
"grad_norm": 0.005957981571555138,
|
1793 |
+
"learning_rate": 1.5953798953257534e-06,
|
1794 |
+
"loss": 0.0013,
|
1795 |
+
"step": 5100
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"epoch": 2.7720628045479154,
|
1799 |
+
"grad_norm": 0.005641666240990162,
|
1800 |
+
"learning_rate": 1.5231907597906518e-06,
|
1801 |
+
"loss": 0.0013,
|
1802 |
+
"step": 5120
|
1803 |
+
},
|
1804 |
+
{
|
1805 |
+
"epoch": 2.7828911748781806,
|
1806 |
+
"grad_norm": 0.005417118314653635,
|
1807 |
+
"learning_rate": 1.4510016242555498e-06,
|
1808 |
+
"loss": 0.0014,
|
1809 |
+
"step": 5140
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 2.7937195452084462,
|
1813 |
+
"grad_norm": 0.006614298094063997,
|
1814 |
+
"learning_rate": 1.3788124887204477e-06,
|
1815 |
+
"loss": 0.0014,
|
1816 |
+
"step": 5160
|
1817 |
+
},
|
1818 |
+
{
|
1819 |
+
"epoch": 2.8045479155387114,
|
1820 |
+
"grad_norm": 0.007049723528325558,
|
1821 |
+
"learning_rate": 1.3066233531853457e-06,
|
1822 |
+
"loss": 0.0014,
|
1823 |
+
"step": 5180
|
1824 |
+
},
|
1825 |
+
{
|
1826 |
+
"epoch": 2.8153762858689766,
|
1827 |
+
"grad_norm": 0.004931174218654633,
|
1828 |
+
"learning_rate": 1.2344342176502438e-06,
|
1829 |
+
"loss": 0.0013,
|
1830 |
+
"step": 5200
|
1831 |
+
},
|
1832 |
+
{
|
1833 |
+
"epoch": 2.826204656199242,
|
1834 |
+
"grad_norm": 0.007498623337596655,
|
1835 |
+
"learning_rate": 1.1622450821151418e-06,
|
1836 |
+
"loss": 0.0014,
|
1837 |
+
"step": 5220
|
1838 |
+
},
|
1839 |
+
{
|
1840 |
+
"epoch": 2.8370330265295074,
|
1841 |
+
"grad_norm": 0.006211725063621998,
|
1842 |
+
"learning_rate": 1.0900559465800397e-06,
|
1843 |
+
"loss": 0.0014,
|
1844 |
+
"step": 5240
|
1845 |
+
},
|
1846 |
+
{
|
1847 |
+
"epoch": 2.8478613968597726,
|
1848 |
+
"grad_norm": 0.008632579818367958,
|
1849 |
+
"learning_rate": 1.0178668110449379e-06,
|
1850 |
+
"loss": 0.0013,
|
1851 |
+
"step": 5260
|
1852 |
+
},
|
1853 |
+
{
|
1854 |
+
"epoch": 2.8586897671900378,
|
1855 |
+
"grad_norm": 0.008940057829022408,
|
1856 |
+
"learning_rate": 9.456776755098358e-07,
|
1857 |
+
"loss": 0.0013,
|
1858 |
+
"step": 5280
|
1859 |
+
},
|
1860 |
+
{
|
1861 |
+
"epoch": 2.869518137520303,
|
1862 |
+
"grad_norm": 0.011228718794882298,
|
1863 |
+
"learning_rate": 8.73488539974734e-07,
|
1864 |
+
"loss": 0.0014,
|
1865 |
+
"step": 5300
|
1866 |
+
},
|
1867 |
+
{
|
1868 |
+
"epoch": 2.8803465078505686,
|
1869 |
+
"grad_norm": 0.007089744322001934,
|
1870 |
+
"learning_rate": 8.012994044396319e-07,
|
1871 |
+
"loss": 0.0015,
|
1872 |
+
"step": 5320
|
1873 |
+
},
|
1874 |
+
{
|
1875 |
+
"epoch": 2.8911748781808337,
|
1876 |
+
"grad_norm": 0.006037184968590736,
|
1877 |
+
"learning_rate": 7.2911026890453e-07,
|
1878 |
+
"loss": 0.0013,
|
1879 |
+
"step": 5340
|
1880 |
+
},
|
1881 |
+
{
|
1882 |
+
"epoch": 2.902003248511099,
|
1883 |
+
"grad_norm": 0.005260369274765253,
|
1884 |
+
"learning_rate": 6.569211333694279e-07,
|
1885 |
+
"loss": 0.0012,
|
1886 |
+
"step": 5360
|
1887 |
+
},
|
1888 |
+
{
|
1889 |
+
"epoch": 2.9128316188413645,
|
1890 |
+
"grad_norm": 0.004701657686382532,
|
1891 |
+
"learning_rate": 5.84731997834326e-07,
|
1892 |
+
"loss": 0.0015,
|
1893 |
+
"step": 5380
|
1894 |
+
},
|
1895 |
+
{
|
1896 |
+
"epoch": 2.9236599891716297,
|
1897 |
+
"grad_norm": 0.009111477993428707,
|
1898 |
+
"learning_rate": 5.12542862299224e-07,
|
1899 |
+
"loss": 0.0013,
|
1900 |
+
"step": 5400
|
1901 |
+
},
|
1902 |
+
{
|
1903 |
+
"epoch": 2.934488359501895,
|
1904 |
+
"grad_norm": 0.0066655525006353855,
|
1905 |
+
"learning_rate": 4.4035372676412204e-07,
|
1906 |
+
"loss": 0.0013,
|
1907 |
+
"step": 5420
|
1908 |
+
},
|
1909 |
+
{
|
1910 |
+
"epoch": 2.94531672983216,
|
1911 |
+
"grad_norm": 0.007916849106550217,
|
1912 |
+
"learning_rate": 3.681645912290201e-07,
|
1913 |
+
"loss": 0.0013,
|
1914 |
+
"step": 5440
|
1915 |
+
},
|
1916 |
+
{
|
1917 |
+
"epoch": 2.9561451001624257,
|
1918 |
+
"grad_norm": 0.006292740814387798,
|
1919 |
+
"learning_rate": 2.959754556939181e-07,
|
1920 |
+
"loss": 0.0013,
|
1921 |
+
"step": 5460
|
1922 |
+
},
|
1923 |
+
{
|
1924 |
+
"epoch": 2.966973470492691,
|
1925 |
+
"grad_norm": 0.007718087639659643,
|
1926 |
+
"learning_rate": 2.2378632015881612e-07,
|
1927 |
+
"loss": 0.0013,
|
1928 |
+
"step": 5480
|
1929 |
+
},
|
1930 |
+
{
|
1931 |
+
"epoch": 2.977801840822956,
|
1932 |
+
"grad_norm": 0.005025485064834356,
|
1933 |
+
"learning_rate": 1.5159718462371415e-07,
|
1934 |
+
"loss": 0.0013,
|
1935 |
+
"step": 5500
|
1936 |
+
},
|
1937 |
+
{
|
1938 |
+
"epoch": 2.9886302111532217,
|
1939 |
+
"grad_norm": 0.006124204955995083,
|
1940 |
+
"learning_rate": 7.940804908861216e-08,
|
1941 |
+
"loss": 0.0013,
|
1942 |
+
"step": 5520
|
1943 |
+
},
|
1944 |
+
{
|
1945 |
+
"epoch": 2.999458581483487,
|
1946 |
+
"grad_norm": 0.00655241496860981,
|
1947 |
+
"learning_rate": 7.218913553510197e-09,
|
1948 |
+
"loss": 0.0013,
|
1949 |
+
"step": 5540
|
1950 |
+
}
|
1951 |
+
],
|
1952 |
+
"logging_steps": 20,
|
1953 |
+
"max_steps": 5541,
|
1954 |
+
"num_input_tokens_seen": 0,
|
1955 |
+
"num_train_epochs": 3,
|
1956 |
+
"save_steps": 500,
|
1957 |
+
"stateful_callbacks": {
|
1958 |
+
"TrainerControl": {
|
1959 |
+
"args": {
|
1960 |
+
"should_epoch_stop": false,
|
1961 |
+
"should_evaluate": false,
|
1962 |
+
"should_log": false,
|
1963 |
+
"should_save": true,
|
1964 |
+
"should_training_stop": true
|
1965 |
+
},
|
1966 |
+
"attributes": {}
|
1967 |
+
}
|
1968 |
+
},
|
1969 |
+
"total_flos": 0.0,
|
1970 |
+
"train_batch_size": 16,
|
1971 |
+
"trial_name": null,
|
1972 |
+
"trial_params": null
|
1973 |
+
}
|
checkpoint-5541/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0214d33554592467921c6cf9ea4c66d3b830194ad3c4677046bb7d47de8435dc
|
3 |
+
size 5777
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6aa9c0850bf7cfc7c43401b0061d3d3556b7a50742184ccb1159d53662b196fa
|
3 |
+
size 736593656
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5777
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0214d33554592467921c6cf9ea4c66d3b830194ad3c4677046bb7d47de8435dc
|
3 |
size 5777
|