Text Classification
Transformers
PyTorch
English
deberta-v2
reward-model
reward_model
RLHF
Inference Endpoints
theblackcat102 commited on
Commit
c5f0c32
1 Parent(s): 7582796

Upload 13 files

Browse files
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
config.json ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/deberta-v3-base",
3
+ "architectures": [
4
+ "DebertaV2ForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "hidden_act": "gelu",
8
+ "hidden_dropout_prob": 0.1,
9
+ "hidden_size": 768,
10
+ "id2label": {
11
+ "0": "REWARD"
12
+ },
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 3072,
15
+ "label2id": {
16
+ "REWARD": 0
17
+ },
18
+ "layer_norm_eps": 1e-07,
19
+ "max_position_embeddings": 512,
20
+ "max_relative_positions": -1,
21
+ "model_type": "deberta-v2",
22
+ "norm_rel_ebd": "layer_norm",
23
+ "num_attention_heads": 12,
24
+ "num_hidden_layers": 12,
25
+ "pad_token_id": 0,
26
+ "pooler_dropout": 0,
27
+ "pooler_hidden_act": "gelu",
28
+ "pooler_hidden_size": 768,
29
+ "pos_att_type": [
30
+ "p2c",
31
+ "c2p"
32
+ ],
33
+ "position_biased_input": false,
34
+ "position_buckets": 256,
35
+ "problem_type": "regression",
36
+ "relative_attention": true,
37
+ "share_att_key": true,
38
+ "torch_dtype": "float32",
39
+ "transformers_version": "4.25.1",
40
+ "type_vocab_size": 0,
41
+ "vocab_size": 128100
42
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c51e84122732847167b0a9f8184f3097cc9a26a23cb3292d287d7e432736c480
3
+ size 1475500313
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:39f57de8c6e51be66e7023367b0b4fa0122b8f4eb481b509e795e7ebfc226865
3
+ size 737763883
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aea50855e4d0949ea23f93e66d195e6490b97c8de68e1c07fb65d4dba061fdb8
3
+ size 15587
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2349a271c10b7c9d605f61c613cb980d22d19b8cdd9f5c88d8eab60988c322bc
3
+ size 559
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea848812c4d01e9c7b1b400921d7f207693ac100d4c729e572925c6e111b123f
3
+ size 559
special_tokens_map.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
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": "[UNK]"
9
+ }
spm.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
3
+ size 2464616
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "cls_token": "[CLS]",
4
+ "do_lower_case": false,
5
+ "eos_token": "[SEP]",
6
+ "mask_token": "[MASK]",
7
+ "model_max_length": 1000000000000000019884624838656,
8
+ "name_or_path": "microsoft/deberta-v3-base",
9
+ "pad_token": "[PAD]",
10
+ "sep_token": "[SEP]",
11
+ "sp_model_kwargs": {},
12
+ "special_tokens_map_file": null,
13
+ "split_by_punct": false,
14
+ "tokenizer_class": "DebertaV2Tokenizer",
15
+ "unk_token": "[UNK]",
16
+ "vocab_type": "spm"
17
+ }
trainer_state.json ADDED
@@ -0,0 +1,2956 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.9083181542197936,
5
+ "global_step": 4000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 1.8e-07,
13
+ "loss": 0.692,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 3.8e-07,
19
+ "loss": 0.6931,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 5.800000000000001e-07,
25
+ "loss": 0.6938,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 7.8e-07,
31
+ "loss": 0.6928,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 9.800000000000001e-07,
37
+ "loss": 0.6936,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.03,
42
+ "learning_rate": 1.1800000000000001e-06,
43
+ "loss": 0.6944,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 1.3800000000000001e-06,
49
+ "loss": 0.6936,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.04,
54
+ "learning_rate": 1.5800000000000001e-06,
55
+ "loss": 0.6926,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.04,
60
+ "learning_rate": 1.7800000000000001e-06,
61
+ "loss": 0.6921,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.05,
66
+ "learning_rate": 1.98e-06,
67
+ "loss": 0.6921,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.05,
72
+ "learning_rate": 2.1800000000000003e-06,
73
+ "loss": 0.6924,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.06,
78
+ "learning_rate": 2.38e-06,
79
+ "loss": 0.6924,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.06,
84
+ "learning_rate": 2.5800000000000003e-06,
85
+ "loss": 0.6902,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.07,
90
+ "learning_rate": 2.7800000000000005e-06,
91
+ "loss": 0.6914,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.07,
96
+ "learning_rate": 2.9800000000000003e-06,
97
+ "loss": 0.6907,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.08,
102
+ "learning_rate": 3.1800000000000005e-06,
103
+ "loss": 0.6893,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.08,
108
+ "learning_rate": 3.3800000000000007e-06,
109
+ "loss": 0.6867,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.09,
114
+ "learning_rate": 3.58e-06,
115
+ "loss": 0.683,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.09,
120
+ "learning_rate": 3.7800000000000002e-06,
121
+ "loss": 0.677,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.1,
126
+ "learning_rate": 3.980000000000001e-06,
127
+ "loss": 0.6627,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.1,
132
+ "eval_webgpt_accuracy": 0.5160878447395302,
133
+ "eval_webgpt_loss": 0.69077467918396,
134
+ "eval_webgpt_runtime": 62.8216,
135
+ "eval_webgpt_samples_per_second": 62.335,
136
+ "eval_webgpt_steps_per_second": 6.24,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 0.1,
141
+ "eval_hfsummary_accuracy": 0.6221624399238279,
142
+ "eval_hfsummary_loss": 0.6861704587936401,
143
+ "eval_hfsummary_runtime": 1239.8418,
144
+ "eval_hfsummary_samples_per_second": 26.683,
145
+ "eval_hfsummary_steps_per_second": 2.669,
146
+ "step": 200
147
+ },
148
+ {
149
+ "epoch": 0.1,
150
+ "eval_gptsynthetic_accuracy": 0.9957767722473605,
151
+ "eval_gptsynthetic_loss": 0.3938244879245758,
152
+ "eval_gptsynthetic_runtime": 45.6545,
153
+ "eval_gptsynthetic_samples_per_second": 72.611,
154
+ "eval_gptsynthetic_steps_per_second": 7.272,
155
+ "step": 200
156
+ },
157
+ {
158
+ "epoch": 0.1,
159
+ "learning_rate": 4.18e-06,
160
+ "loss": 0.6481,
161
+ "step": 210
162
+ },
163
+ {
164
+ "epoch": 0.1,
165
+ "learning_rate": 4.38e-06,
166
+ "loss": 0.624,
167
+ "step": 220
168
+ },
169
+ {
170
+ "epoch": 0.11,
171
+ "learning_rate": 4.58e-06,
172
+ "loss": 0.6219,
173
+ "step": 230
174
+ },
175
+ {
176
+ "epoch": 0.11,
177
+ "learning_rate": 4.78e-06,
178
+ "loss": 0.6079,
179
+ "step": 240
180
+ },
181
+ {
182
+ "epoch": 0.12,
183
+ "learning_rate": 4.980000000000001e-06,
184
+ "loss": 0.5815,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.12,
189
+ "learning_rate": 5.18e-06,
190
+ "loss": 0.5714,
191
+ "step": 260
192
+ },
193
+ {
194
+ "epoch": 0.13,
195
+ "learning_rate": 5.380000000000001e-06,
196
+ "loss": 0.5904,
197
+ "step": 270
198
+ },
199
+ {
200
+ "epoch": 0.13,
201
+ "learning_rate": 5.580000000000001e-06,
202
+ "loss": 0.5796,
203
+ "step": 280
204
+ },
205
+ {
206
+ "epoch": 0.14,
207
+ "learning_rate": 5.78e-06,
208
+ "loss": 0.5649,
209
+ "step": 290
210
+ },
211
+ {
212
+ "epoch": 0.14,
213
+ "learning_rate": 5.98e-06,
214
+ "loss": 0.581,
215
+ "step": 300
216
+ },
217
+ {
218
+ "epoch": 0.15,
219
+ "learning_rate": 6.18e-06,
220
+ "loss": 0.5481,
221
+ "step": 310
222
+ },
223
+ {
224
+ "epoch": 0.15,
225
+ "learning_rate": 6.380000000000001e-06,
226
+ "loss": 0.5839,
227
+ "step": 320
228
+ },
229
+ {
230
+ "epoch": 0.16,
231
+ "learning_rate": 6.5800000000000005e-06,
232
+ "loss": 0.5552,
233
+ "step": 330
234
+ },
235
+ {
236
+ "epoch": 0.16,
237
+ "learning_rate": 6.780000000000001e-06,
238
+ "loss": 0.5654,
239
+ "step": 340
240
+ },
241
+ {
242
+ "epoch": 0.17,
243
+ "learning_rate": 6.98e-06,
244
+ "loss": 0.5618,
245
+ "step": 350
246
+ },
247
+ {
248
+ "epoch": 0.17,
249
+ "learning_rate": 7.180000000000001e-06,
250
+ "loss": 0.5528,
251
+ "step": 360
252
+ },
253
+ {
254
+ "epoch": 0.18,
255
+ "learning_rate": 7.3800000000000005e-06,
256
+ "loss": 0.5341,
257
+ "step": 370
258
+ },
259
+ {
260
+ "epoch": 0.18,
261
+ "learning_rate": 7.58e-06,
262
+ "loss": 0.5458,
263
+ "step": 380
264
+ },
265
+ {
266
+ "epoch": 0.19,
267
+ "learning_rate": 7.78e-06,
268
+ "loss": 0.5335,
269
+ "step": 390
270
+ },
271
+ {
272
+ "epoch": 0.19,
273
+ "learning_rate": 7.980000000000002e-06,
274
+ "loss": 0.5285,
275
+ "step": 400
276
+ },
277
+ {
278
+ "epoch": 0.19,
279
+ "eval_webgpt_accuracy": 0.5617977528089888,
280
+ "eval_webgpt_loss": 0.7310989499092102,
281
+ "eval_webgpt_runtime": 63.0205,
282
+ "eval_webgpt_samples_per_second": 62.139,
283
+ "eval_webgpt_steps_per_second": 6.22,
284
+ "step": 400
285
+ },
286
+ {
287
+ "epoch": 0.19,
288
+ "eval_hfsummary_accuracy": 0.6381525254662516,
289
+ "eval_hfsummary_loss": 0.6372680068016052,
290
+ "eval_hfsummary_runtime": 1238.3216,
291
+ "eval_hfsummary_samples_per_second": 26.716,
292
+ "eval_hfsummary_steps_per_second": 2.672,
293
+ "step": 400
294
+ },
295
+ {
296
+ "epoch": 0.19,
297
+ "eval_gptsynthetic_accuracy": 0.9981900452488688,
298
+ "eval_gptsynthetic_loss": 0.026522908359766006,
299
+ "eval_gptsynthetic_runtime": 45.6333,
300
+ "eval_gptsynthetic_samples_per_second": 72.644,
301
+ "eval_gptsynthetic_steps_per_second": 7.275,
302
+ "step": 400
303
+ },
304
+ {
305
+ "epoch": 0.2,
306
+ "learning_rate": 8.18e-06,
307
+ "loss": 0.5663,
308
+ "step": 410
309
+ },
310
+ {
311
+ "epoch": 0.2,
312
+ "learning_rate": 8.380000000000001e-06,
313
+ "loss": 0.55,
314
+ "step": 420
315
+ },
316
+ {
317
+ "epoch": 0.21,
318
+ "learning_rate": 8.580000000000001e-06,
319
+ "loss": 0.5465,
320
+ "step": 430
321
+ },
322
+ {
323
+ "epoch": 0.21,
324
+ "learning_rate": 8.78e-06,
325
+ "loss": 0.5409,
326
+ "step": 440
327
+ },
328
+ {
329
+ "epoch": 0.21,
330
+ "learning_rate": 8.98e-06,
331
+ "loss": 0.5496,
332
+ "step": 450
333
+ },
334
+ {
335
+ "epoch": 0.22,
336
+ "learning_rate": 9.180000000000002e-06,
337
+ "loss": 0.5334,
338
+ "step": 460
339
+ },
340
+ {
341
+ "epoch": 0.22,
342
+ "learning_rate": 9.38e-06,
343
+ "loss": 0.5108,
344
+ "step": 470
345
+ },
346
+ {
347
+ "epoch": 0.23,
348
+ "learning_rate": 9.58e-06,
349
+ "loss": 0.5306,
350
+ "step": 480
351
+ },
352
+ {
353
+ "epoch": 0.23,
354
+ "learning_rate": 9.780000000000001e-06,
355
+ "loss": 0.5379,
356
+ "step": 490
357
+ },
358
+ {
359
+ "epoch": 0.24,
360
+ "learning_rate": 9.980000000000001e-06,
361
+ "loss": 0.5288,
362
+ "step": 500
363
+ },
364
+ {
365
+ "epoch": 0.24,
366
+ "learning_rate": 9.999967852222371e-06,
367
+ "loss": 0.5362,
368
+ "step": 510
369
+ },
370
+ {
371
+ "epoch": 0.25,
372
+ "learning_rate": 9.999856724632902e-06,
373
+ "loss": 0.5245,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.25,
378
+ "learning_rate": 9.99966622210923e-06,
379
+ "loss": 0.5424,
380
+ "step": 530
381
+ },
382
+ {
383
+ "epoch": 0.26,
384
+ "learning_rate": 9.999396347675668e-06,
385
+ "loss": 0.5161,
386
+ "step": 540
387
+ },
388
+ {
389
+ "epoch": 0.26,
390
+ "learning_rate": 9.99904710561659e-06,
391
+ "loss": 0.5295,
392
+ "step": 550
393
+ },
394
+ {
395
+ "epoch": 0.27,
396
+ "learning_rate": 9.998618501476367e-06,
397
+ "loss": 0.5049,
398
+ "step": 560
399
+ },
400
+ {
401
+ "epoch": 0.27,
402
+ "learning_rate": 9.998110542059283e-06,
403
+ "loss": 0.5136,
404
+ "step": 570
405
+ },
406
+ {
407
+ "epoch": 0.28,
408
+ "learning_rate": 9.997523235429414e-06,
409
+ "loss": 0.5059,
410
+ "step": 580
411
+ },
412
+ {
413
+ "epoch": 0.28,
414
+ "learning_rate": 9.996856590910513e-06,
415
+ "loss": 0.5268,
416
+ "step": 590
417
+ },
418
+ {
419
+ "epoch": 0.29,
420
+ "learning_rate": 9.996110619085859e-06,
421
+ "loss": 0.515,
422
+ "step": 600
423
+ },
424
+ {
425
+ "epoch": 0.29,
426
+ "eval_webgpt_accuracy": 0.5640960163432074,
427
+ "eval_webgpt_loss": 0.7018715143203735,
428
+ "eval_webgpt_runtime": 63.203,
429
+ "eval_webgpt_samples_per_second": 61.959,
430
+ "eval_webgpt_steps_per_second": 6.202,
431
+ "step": 600
432
+ },
433
+ {
434
+ "epoch": 0.29,
435
+ "eval_hfsummary_accuracy": 0.6488831121724148,
436
+ "eval_hfsummary_loss": 0.6238470077514648,
437
+ "eval_hfsummary_runtime": 1239.6473,
438
+ "eval_hfsummary_samples_per_second": 26.687,
439
+ "eval_hfsummary_steps_per_second": 2.669,
440
+ "step": 600
441
+ },
442
+ {
443
+ "epoch": 0.29,
444
+ "eval_gptsynthetic_accuracy": 0.9978883861236802,
445
+ "eval_gptsynthetic_loss": 0.031191527843475342,
446
+ "eval_gptsynthetic_runtime": 45.7943,
447
+ "eval_gptsynthetic_samples_per_second": 72.389,
448
+ "eval_gptsynthetic_steps_per_second": 7.25,
449
+ "step": 600
450
+ },
451
+ {
452
+ "epoch": 0.29,
453
+ "learning_rate": 9.995285331798082e-06,
454
+ "loss": 0.4913,
455
+ "step": 610
456
+ },
457
+ {
458
+ "epoch": 0.3,
459
+ "learning_rate": 9.994380742148981e-06,
460
+ "loss": 0.4957,
461
+ "step": 620
462
+ },
463
+ {
464
+ "epoch": 0.3,
465
+ "learning_rate": 9.993396864499321e-06,
466
+ "loss": 0.5041,
467
+ "step": 630
468
+ },
469
+ {
470
+ "epoch": 0.31,
471
+ "learning_rate": 9.992333714468589e-06,
472
+ "loss": 0.5204,
473
+ "step": 640
474
+ },
475
+ {
476
+ "epoch": 0.31,
477
+ "learning_rate": 9.991191308934765e-06,
478
+ "loss": 0.4964,
479
+ "step": 650
480
+ },
481
+ {
482
+ "epoch": 0.31,
483
+ "learning_rate": 9.989969666034039e-06,
484
+ "loss": 0.5054,
485
+ "step": 660
486
+ },
487
+ {
488
+ "epoch": 0.32,
489
+ "learning_rate": 9.988668805160533e-06,
490
+ "loss": 0.5156,
491
+ "step": 670
492
+ },
493
+ {
494
+ "epoch": 0.32,
495
+ "learning_rate": 9.987288746965987e-06,
496
+ "loss": 0.4875,
497
+ "step": 680
498
+ },
499
+ {
500
+ "epoch": 0.33,
501
+ "learning_rate": 9.985829513359437e-06,
502
+ "loss": 0.5002,
503
+ "step": 690
504
+ },
505
+ {
506
+ "epoch": 0.33,
507
+ "learning_rate": 9.984291127506857e-06,
508
+ "loss": 0.5103,
509
+ "step": 700
510
+ },
511
+ {
512
+ "epoch": 0.34,
513
+ "learning_rate": 9.982673613830807e-06,
514
+ "loss": 0.488,
515
+ "step": 710
516
+ },
517
+ {
518
+ "epoch": 0.34,
519
+ "learning_rate": 9.980976998010028e-06,
520
+ "loss": 0.5091,
521
+ "step": 720
522
+ },
523
+ {
524
+ "epoch": 0.35,
525
+ "learning_rate": 9.979201306979047e-06,
526
+ "loss": 0.5008,
527
+ "step": 730
528
+ },
529
+ {
530
+ "epoch": 0.35,
531
+ "learning_rate": 9.977346568927742e-06,
532
+ "loss": 0.4933,
533
+ "step": 740
534
+ },
535
+ {
536
+ "epoch": 0.36,
537
+ "learning_rate": 9.9754128133009e-06,
538
+ "loss": 0.5208,
539
+ "step": 750
540
+ },
541
+ {
542
+ "epoch": 0.36,
543
+ "learning_rate": 9.973400070797745e-06,
544
+ "loss": 0.5036,
545
+ "step": 760
546
+ },
547
+ {
548
+ "epoch": 0.37,
549
+ "learning_rate": 9.971308373371453e-06,
550
+ "loss": 0.4935,
551
+ "step": 770
552
+ },
553
+ {
554
+ "epoch": 0.37,
555
+ "learning_rate": 9.969137754228648e-06,
556
+ "loss": 0.5026,
557
+ "step": 780
558
+ },
559
+ {
560
+ "epoch": 0.38,
561
+ "learning_rate": 9.966888247828864e-06,
562
+ "loss": 0.5047,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.38,
567
+ "learning_rate": 9.964559889884011e-06,
568
+ "loss": 0.4785,
569
+ "step": 800
570
+ },
571
+ {
572
+ "epoch": 0.38,
573
+ "eval_webgpt_accuracy": 0.5684371807967313,
574
+ "eval_webgpt_loss": 0.6995264887809753,
575
+ "eval_webgpt_runtime": 63.093,
576
+ "eval_webgpt_samples_per_second": 62.067,
577
+ "eval_webgpt_steps_per_second": 6.213,
578
+ "step": 800
579
+ },
580
+ {
581
+ "epoch": 0.38,
582
+ "eval_hfsummary_accuracy": 0.6444397424659191,
583
+ "eval_hfsummary_loss": 0.6311432719230652,
584
+ "eval_hfsummary_runtime": 1241.3962,
585
+ "eval_hfsummary_samples_per_second": 26.65,
586
+ "eval_hfsummary_steps_per_second": 2.666,
587
+ "step": 800
588
+ },
589
+ {
590
+ "epoch": 0.38,
591
+ "eval_gptsynthetic_accuracy": 0.9978883861236802,
592
+ "eval_gptsynthetic_loss": 0.016028160229325294,
593
+ "eval_gptsynthetic_runtime": 45.7378,
594
+ "eval_gptsynthetic_samples_per_second": 72.478,
595
+ "eval_gptsynthetic_steps_per_second": 7.259,
596
+ "step": 800
597
+ },
598
+ {
599
+ "epoch": 0.39,
600
+ "learning_rate": 9.962152717357796e-06,
601
+ "loss": 0.4752,
602
+ "step": 810
603
+ },
604
+ {
605
+ "epoch": 0.39,
606
+ "learning_rate": 9.959666768465153e-06,
607
+ "loss": 0.4657,
608
+ "step": 820
609
+ },
610
+ {
611
+ "epoch": 0.4,
612
+ "learning_rate": 9.957102082671612e-06,
613
+ "loss": 0.4935,
614
+ "step": 830
615
+ },
616
+ {
617
+ "epoch": 0.4,
618
+ "learning_rate": 9.954458700692699e-06,
619
+ "loss": 0.4821,
620
+ "step": 840
621
+ },
622
+ {
623
+ "epoch": 0.41,
624
+ "learning_rate": 9.951736664493266e-06,
625
+ "loss": 0.4949,
626
+ "step": 850
627
+ },
628
+ {
629
+ "epoch": 0.41,
630
+ "learning_rate": 9.948936017286844e-06,
631
+ "loss": 0.4894,
632
+ "step": 860
633
+ },
634
+ {
635
+ "epoch": 0.42,
636
+ "learning_rate": 9.94605680353494e-06,
637
+ "loss": 0.5032,
638
+ "step": 870
639
+ },
640
+ {
641
+ "epoch": 0.42,
642
+ "learning_rate": 9.943099068946348e-06,
643
+ "loss": 0.4724,
644
+ "step": 880
645
+ },
646
+ {
647
+ "epoch": 0.42,
648
+ "learning_rate": 9.940062860476406e-06,
649
+ "loss": 0.4871,
650
+ "step": 890
651
+ },
652
+ {
653
+ "epoch": 0.43,
654
+ "learning_rate": 9.936948226326273e-06,
655
+ "loss": 0.4919,
656
+ "step": 900
657
+ },
658
+ {
659
+ "epoch": 0.43,
660
+ "learning_rate": 9.933755215942137e-06,
661
+ "loss": 0.4763,
662
+ "step": 910
663
+ },
664
+ {
665
+ "epoch": 0.44,
666
+ "learning_rate": 9.930483880014448e-06,
667
+ "loss": 0.4928,
668
+ "step": 920
669
+ },
670
+ {
671
+ "epoch": 0.44,
672
+ "learning_rate": 9.92713427047711e-06,
673
+ "loss": 0.4608,
674
+ "step": 930
675
+ },
676
+ {
677
+ "epoch": 0.45,
678
+ "learning_rate": 9.923706440506654e-06,
679
+ "loss": 0.491,
680
+ "step": 940
681
+ },
682
+ {
683
+ "epoch": 0.45,
684
+ "learning_rate": 9.920200444521395e-06,
685
+ "loss": 0.4624,
686
+ "step": 950
687
+ },
688
+ {
689
+ "epoch": 0.46,
690
+ "learning_rate": 9.916616338180569e-06,
691
+ "loss": 0.5028,
692
+ "step": 960
693
+ },
694
+ {
695
+ "epoch": 0.46,
696
+ "learning_rate": 9.912954178383445e-06,
697
+ "loss": 0.4814,
698
+ "step": 970
699
+ },
700
+ {
701
+ "epoch": 0.47,
702
+ "learning_rate": 9.90921402326843e-06,
703
+ "loss": 0.4901,
704
+ "step": 980
705
+ },
706
+ {
707
+ "epoch": 0.47,
708
+ "learning_rate": 9.905395932212139e-06,
709
+ "loss": 0.459,
710
+ "step": 990
711
+ },
712
+ {
713
+ "epoch": 0.48,
714
+ "learning_rate": 9.901499965828451e-06,
715
+ "loss": 0.4624,
716
+ "step": 1000
717
+ },
718
+ {
719
+ "epoch": 0.48,
720
+ "eval_webgpt_accuracy": 0.5661389172625128,
721
+ "eval_webgpt_loss": 0.6934247016906738,
722
+ "eval_webgpt_runtime": 63.1076,
723
+ "eval_webgpt_samples_per_second": 62.053,
724
+ "eval_webgpt_steps_per_second": 6.212,
725
+ "step": 1000
726
+ },
727
+ {
728
+ "epoch": 0.48,
729
+ "eval_hfsummary_accuracy": 0.6628177613880241,
730
+ "eval_hfsummary_loss": 0.6276853680610657,
731
+ "eval_hfsummary_runtime": 1241.5998,
732
+ "eval_hfsummary_samples_per_second": 26.645,
733
+ "eval_hfsummary_steps_per_second": 2.665,
734
+ "step": 1000
735
+ },
736
+ {
737
+ "epoch": 0.48,
738
+ "eval_gptsynthetic_accuracy": 0.9975867269984917,
739
+ "eval_gptsynthetic_loss": 0.015120746567845345,
740
+ "eval_gptsynthetic_runtime": 45.6555,
741
+ "eval_gptsynthetic_samples_per_second": 72.609,
742
+ "eval_gptsynthetic_steps_per_second": 7.272,
743
+ "step": 1000
744
+ },
745
+ {
746
+ "epoch": 0.48,
747
+ "learning_rate": 9.897526185967557e-06,
748
+ "loss": 0.4729,
749
+ "step": 1010
750
+ },
751
+ {
752
+ "epoch": 0.49,
753
+ "learning_rate": 9.893474655714962e-06,
754
+ "loss": 0.4767,
755
+ "step": 1020
756
+ },
757
+ {
758
+ "epoch": 0.49,
759
+ "learning_rate": 9.889345439390508e-06,
760
+ "loss": 0.4737,
761
+ "step": 1030
762
+ },
763
+ {
764
+ "epoch": 0.5,
765
+ "learning_rate": 9.88513860254732e-06,
766
+ "loss": 0.4795,
767
+ "step": 1040
768
+ },
769
+ {
770
+ "epoch": 0.5,
771
+ "learning_rate": 9.880854211970794e-06,
772
+ "loss": 0.4674,
773
+ "step": 1050
774
+ },
775
+ {
776
+ "epoch": 0.51,
777
+ "learning_rate": 9.876492335677523e-06,
778
+ "loss": 0.4605,
779
+ "step": 1060
780
+ },
781
+ {
782
+ "epoch": 0.51,
783
+ "learning_rate": 9.872053042914223e-06,
784
+ "loss": 0.4347,
785
+ "step": 1070
786
+ },
787
+ {
788
+ "epoch": 0.52,
789
+ "learning_rate": 9.867536404156621e-06,
790
+ "loss": 0.4614,
791
+ "step": 1080
792
+ },
793
+ {
794
+ "epoch": 0.52,
795
+ "learning_rate": 9.862942491108358e-06,
796
+ "loss": 0.4849,
797
+ "step": 1090
798
+ },
799
+ {
800
+ "epoch": 0.52,
801
+ "learning_rate": 9.858271376699832e-06,
802
+ "loss": 0.4715,
803
+ "step": 1100
804
+ },
805
+ {
806
+ "epoch": 0.53,
807
+ "learning_rate": 9.853523135087047e-06,
808
+ "loss": 0.4768,
809
+ "step": 1110
810
+ },
811
+ {
812
+ "epoch": 0.53,
813
+ "learning_rate": 9.848697841650435e-06,
814
+ "loss": 0.4686,
815
+ "step": 1120
816
+ },
817
+ {
818
+ "epoch": 0.54,
819
+ "learning_rate": 9.843795572993664e-06,
820
+ "loss": 0.4521,
821
+ "step": 1130
822
+ },
823
+ {
824
+ "epoch": 0.54,
825
+ "learning_rate": 9.83881640694241e-06,
826
+ "loss": 0.4765,
827
+ "step": 1140
828
+ },
829
+ {
830
+ "epoch": 0.55,
831
+ "learning_rate": 9.833760422543137e-06,
832
+ "loss": 0.4829,
833
+ "step": 1150
834
+ },
835
+ {
836
+ "epoch": 0.55,
837
+ "learning_rate": 9.828627700061833e-06,
838
+ "loss": 0.4587,
839
+ "step": 1160
840
+ },
841
+ {
842
+ "epoch": 0.56,
843
+ "learning_rate": 9.82341832098273e-06,
844
+ "loss": 0.4684,
845
+ "step": 1170
846
+ },
847
+ {
848
+ "epoch": 0.56,
849
+ "learning_rate": 9.818132368007022e-06,
850
+ "loss": 0.4443,
851
+ "step": 1180
852
+ },
853
+ {
854
+ "epoch": 0.57,
855
+ "learning_rate": 9.812769925051553e-06,
856
+ "loss": 0.4898,
857
+ "step": 1190
858
+ },
859
+ {
860
+ "epoch": 0.57,
861
+ "learning_rate": 9.807331077247468e-06,
862
+ "loss": 0.44,
863
+ "step": 1200
864
+ },
865
+ {
866
+ "epoch": 0.57,
867
+ "eval_webgpt_accuracy": 0.5605209397344229,
868
+ "eval_webgpt_loss": 0.6905099749565125,
869
+ "eval_webgpt_runtime": 63.2036,
870
+ "eval_webgpt_samples_per_second": 61.959,
871
+ "eval_webgpt_steps_per_second": 6.202,
872
+ "step": 1200
873
+ },
874
+ {
875
+ "epoch": 0.57,
876
+ "eval_hfsummary_accuracy": 0.664419792642747,
877
+ "eval_hfsummary_loss": 0.6330422163009644,
878
+ "eval_hfsummary_runtime": 1242.5416,
879
+ "eval_hfsummary_samples_per_second": 26.625,
880
+ "eval_hfsummary_steps_per_second": 2.663,
881
+ "step": 1200
882
+ },
883
+ {
884
+ "epoch": 0.57,
885
+ "eval_gptsynthetic_accuracy": 0.9981900452488688,
886
+ "eval_gptsynthetic_loss": 0.011154625564813614,
887
+ "eval_gptsynthetic_runtime": 45.6845,
888
+ "eval_gptsynthetic_samples_per_second": 72.563,
889
+ "eval_gptsynthetic_steps_per_second": 7.267,
890
+ "step": 1200
891
+ },
892
+ {
893
+ "epoch": 0.58,
894
+ "learning_rate": 9.80181591093888e-06,
895
+ "loss": 0.4648,
896
+ "step": 1210
897
+ },
898
+ {
899
+ "epoch": 0.58,
900
+ "learning_rate": 9.796224513681495e-06,
901
+ "loss": 0.4483,
902
+ "step": 1220
903
+ },
904
+ {
905
+ "epoch": 0.59,
906
+ "learning_rate": 9.79055697424121e-06,
907
+ "loss": 0.4707,
908
+ "step": 1230
909
+ },
910
+ {
911
+ "epoch": 0.59,
912
+ "learning_rate": 9.784813382592716e-06,
913
+ "loss": 0.4638,
914
+ "step": 1240
915
+ },
916
+ {
917
+ "epoch": 0.6,
918
+ "learning_rate": 9.77899382991807e-06,
919
+ "loss": 0.4366,
920
+ "step": 1250
921
+ },
922
+ {
923
+ "epoch": 0.6,
924
+ "learning_rate": 9.773098408605242e-06,
925
+ "loss": 0.4457,
926
+ "step": 1260
927
+ },
928
+ {
929
+ "epoch": 0.61,
930
+ "learning_rate": 9.76712721224665e-06,
931
+ "loss": 0.4363,
932
+ "step": 1270
933
+ },
934
+ {
935
+ "epoch": 0.61,
936
+ "learning_rate": 9.761080335637672e-06,
937
+ "loss": 0.4577,
938
+ "step": 1280
939
+ },
940
+ {
941
+ "epoch": 0.62,
942
+ "learning_rate": 9.754957874775146e-06,
943
+ "loss": 0.4538,
944
+ "step": 1290
945
+ },
946
+ {
947
+ "epoch": 0.62,
948
+ "learning_rate": 9.748759926855845e-06,
949
+ "loss": 0.4499,
950
+ "step": 1300
951
+ },
952
+ {
953
+ "epoch": 0.62,
954
+ "learning_rate": 9.742486590274927e-06,
955
+ "loss": 0.4466,
956
+ "step": 1310
957
+ },
958
+ {
959
+ "epoch": 0.63,
960
+ "learning_rate": 9.736137964624381e-06,
961
+ "loss": 0.4459,
962
+ "step": 1320
963
+ },
964
+ {
965
+ "epoch": 0.63,
966
+ "learning_rate": 9.729714150691448e-06,
967
+ "loss": 0.4544,
968
+ "step": 1330
969
+ },
970
+ {
971
+ "epoch": 0.64,
972
+ "learning_rate": 9.723215250457009e-06,
973
+ "loss": 0.4453,
974
+ "step": 1340
975
+ },
976
+ {
977
+ "epoch": 0.64,
978
+ "learning_rate": 9.716641367093976e-06,
979
+ "loss": 0.4614,
980
+ "step": 1350
981
+ },
982
+ {
983
+ "epoch": 0.65,
984
+ "learning_rate": 9.709992604965652e-06,
985
+ "loss": 0.4863,
986
+ "step": 1360
987
+ },
988
+ {
989
+ "epoch": 0.65,
990
+ "learning_rate": 9.703269069624075e-06,
991
+ "loss": 0.4432,
992
+ "step": 1370
993
+ },
994
+ {
995
+ "epoch": 0.66,
996
+ "learning_rate": 9.696470867808333e-06,
997
+ "loss": 0.438,
998
+ "step": 1380
999
+ },
1000
+ {
1001
+ "epoch": 0.66,
1002
+ "learning_rate": 9.68959810744289e-06,
1003
+ "loss": 0.4593,
1004
+ "step": 1390
1005
+ },
1006
+ {
1007
+ "epoch": 0.67,
1008
+ "learning_rate": 9.682650897635845e-06,
1009
+ "loss": 0.4463,
1010
+ "step": 1400
1011
+ },
1012
+ {
1013
+ "epoch": 0.67,
1014
+ "eval_webgpt_accuracy": 0.5617977528089888,
1015
+ "eval_webgpt_loss": 0.6875464916229248,
1016
+ "eval_webgpt_runtime": 63.2336,
1017
+ "eval_webgpt_samples_per_second": 61.929,
1018
+ "eval_webgpt_steps_per_second": 6.199,
1019
+ "step": 1400
1020
+ },
1021
+ {
1022
+ "epoch": 0.67,
1023
+ "eval_hfsummary_accuracy": 0.6741528881903093,
1024
+ "eval_hfsummary_loss": 0.6355233788490295,
1025
+ "eval_hfsummary_runtime": 1241.0977,
1026
+ "eval_hfsummary_samples_per_second": 26.656,
1027
+ "eval_hfsummary_steps_per_second": 2.666,
1028
+ "step": 1400
1029
+ },
1030
+ {
1031
+ "epoch": 0.67,
1032
+ "eval_gptsynthetic_accuracy": 0.9975867269984917,
1033
+ "eval_gptsynthetic_loss": 0.010252352803945541,
1034
+ "eval_gptsynthetic_runtime": 45.7898,
1035
+ "eval_gptsynthetic_samples_per_second": 72.396,
1036
+ "eval_gptsynthetic_steps_per_second": 7.251,
1037
+ "step": 1400
1038
+ },
1039
+ {
1040
+ "epoch": 0.67,
1041
+ "learning_rate": 9.675629348677232e-06,
1042
+ "loss": 0.4454,
1043
+ "step": 1410
1044
+ },
1045
+ {
1046
+ "epoch": 0.68,
1047
+ "learning_rate": 9.668533572037236e-06,
1048
+ "loss": 0.4444,
1049
+ "step": 1420
1050
+ },
1051
+ {
1052
+ "epoch": 0.68,
1053
+ "learning_rate": 9.661363680364445e-06,
1054
+ "loss": 0.4622,
1055
+ "step": 1430
1056
+ },
1057
+ {
1058
+ "epoch": 0.69,
1059
+ "learning_rate": 9.65411978748406e-06,
1060
+ "loss": 0.4591,
1061
+ "step": 1440
1062
+ },
1063
+ {
1064
+ "epoch": 0.69,
1065
+ "learning_rate": 9.64680200839608e-06,
1066
+ "loss": 0.4462,
1067
+ "step": 1450
1068
+ },
1069
+ {
1070
+ "epoch": 0.7,
1071
+ "learning_rate": 9.63941045927348e-06,
1072
+ "loss": 0.4423,
1073
+ "step": 1460
1074
+ },
1075
+ {
1076
+ "epoch": 0.7,
1077
+ "learning_rate": 9.631945257460369e-06,
1078
+ "loss": 0.4302,
1079
+ "step": 1470
1080
+ },
1081
+ {
1082
+ "epoch": 0.71,
1083
+ "learning_rate": 9.62440652147012e-06,
1084
+ "loss": 0.4668,
1085
+ "step": 1480
1086
+ },
1087
+ {
1088
+ "epoch": 0.71,
1089
+ "learning_rate": 9.616794370983498e-06,
1090
+ "loss": 0.4252,
1091
+ "step": 1490
1092
+ },
1093
+ {
1094
+ "epoch": 0.72,
1095
+ "learning_rate": 9.609108926846759e-06,
1096
+ "loss": 0.4563,
1097
+ "step": 1500
1098
+ },
1099
+ {
1100
+ "epoch": 0.72,
1101
+ "learning_rate": 9.60135031106972e-06,
1102
+ "loss": 0.4396,
1103
+ "step": 1510
1104
+ },
1105
+ {
1106
+ "epoch": 0.73,
1107
+ "learning_rate": 9.593518646823833e-06,
1108
+ "loss": 0.463,
1109
+ "step": 1520
1110
+ },
1111
+ {
1112
+ "epoch": 0.73,
1113
+ "learning_rate": 9.58561405844023e-06,
1114
+ "loss": 0.4292,
1115
+ "step": 1530
1116
+ },
1117
+ {
1118
+ "epoch": 0.73,
1119
+ "learning_rate": 9.577636671407743e-06,
1120
+ "loss": 0.4669,
1121
+ "step": 1540
1122
+ },
1123
+ {
1124
+ "epoch": 0.74,
1125
+ "learning_rate": 9.569586612370913e-06,
1126
+ "loss": 0.4648,
1127
+ "step": 1550
1128
+ },
1129
+ {
1130
+ "epoch": 0.74,
1131
+ "learning_rate": 9.561464009127988e-06,
1132
+ "loss": 0.4318,
1133
+ "step": 1560
1134
+ },
1135
+ {
1136
+ "epoch": 0.75,
1137
+ "learning_rate": 9.553268990628877e-06,
1138
+ "loss": 0.4274,
1139
+ "step": 1570
1140
+ },
1141
+ {
1142
+ "epoch": 0.75,
1143
+ "learning_rate": 9.545001686973123e-06,
1144
+ "loss": 0.4532,
1145
+ "step": 1580
1146
+ },
1147
+ {
1148
+ "epoch": 0.76,
1149
+ "learning_rate": 9.53666222940782e-06,
1150
+ "loss": 0.4476,
1151
+ "step": 1590
1152
+ },
1153
+ {
1154
+ "epoch": 0.76,
1155
+ "learning_rate": 9.528250750325545e-06,
1156
+ "loss": 0.4477,
1157
+ "step": 1600
1158
+ },
1159
+ {
1160
+ "epoch": 0.76,
1161
+ "eval_webgpt_accuracy": 0.5801838610827375,
1162
+ "eval_webgpt_loss": 0.6804662346839905,
1163
+ "eval_webgpt_runtime": 63.2449,
1164
+ "eval_webgpt_samples_per_second": 61.918,
1165
+ "eval_webgpt_steps_per_second": 6.198,
1166
+ "step": 1600
1167
+ },
1168
+ {
1169
+ "epoch": 0.76,
1170
+ "eval_hfsummary_accuracy": 0.6604298280083426,
1171
+ "eval_hfsummary_loss": 0.6250379085540771,
1172
+ "eval_hfsummary_runtime": 1240.5251,
1173
+ "eval_hfsummary_samples_per_second": 26.669,
1174
+ "eval_hfsummary_steps_per_second": 2.667,
1175
+ "step": 1600
1176
+ },
1177
+ {
1178
+ "epoch": 0.76,
1179
+ "eval_gptsynthetic_accuracy": 0.9975867269984917,
1180
+ "eval_gptsynthetic_loss": 0.01225399598479271,
1181
+ "eval_gptsynthetic_runtime": 45.6829,
1182
+ "eval_gptsynthetic_samples_per_second": 72.565,
1183
+ "eval_gptsynthetic_steps_per_second": 7.267,
1184
+ "step": 1600
1185
+ },
1186
+ {
1187
+ "epoch": 0.77,
1188
+ "learning_rate": 9.519767383262244e-06,
1189
+ "loss": 0.4265,
1190
+ "step": 1610
1191
+ },
1192
+ {
1193
+ "epoch": 0.77,
1194
+ "learning_rate": 9.511212262895113e-06,
1195
+ "loss": 0.4509,
1196
+ "step": 1620
1197
+ },
1198
+ {
1199
+ "epoch": 0.78,
1200
+ "learning_rate": 9.502585525040469e-06,
1201
+ "loss": 0.4159,
1202
+ "step": 1630
1203
+ },
1204
+ {
1205
+ "epoch": 0.78,
1206
+ "learning_rate": 9.493887306651585e-06,
1207
+ "loss": 0.4241,
1208
+ "step": 1640
1209
+ },
1210
+ {
1211
+ "epoch": 0.79,
1212
+ "learning_rate": 9.485117745816523e-06,
1213
+ "loss": 0.4186,
1214
+ "step": 1650
1215
+ },
1216
+ {
1217
+ "epoch": 0.79,
1218
+ "learning_rate": 9.476276981755933e-06,
1219
+ "loss": 0.4334,
1220
+ "step": 1660
1221
+ },
1222
+ {
1223
+ "epoch": 0.8,
1224
+ "learning_rate": 9.467365154820853e-06,
1225
+ "loss": 0.4508,
1226
+ "step": 1670
1227
+ },
1228
+ {
1229
+ "epoch": 0.8,
1230
+ "learning_rate": 9.45838240649047e-06,
1231
+ "loss": 0.4227,
1232
+ "step": 1680
1233
+ },
1234
+ {
1235
+ "epoch": 0.81,
1236
+ "learning_rate": 9.44932887936988e-06,
1237
+ "loss": 0.4391,
1238
+ "step": 1690
1239
+ },
1240
+ {
1241
+ "epoch": 0.81,
1242
+ "learning_rate": 9.440204717187829e-06,
1243
+ "loss": 0.4461,
1244
+ "step": 1700
1245
+ },
1246
+ {
1247
+ "epoch": 0.82,
1248
+ "learning_rate": 9.43101006479442e-06,
1249
+ "loss": 0.4334,
1250
+ "step": 1710
1251
+ },
1252
+ {
1253
+ "epoch": 0.82,
1254
+ "learning_rate": 9.421745068158823e-06,
1255
+ "loss": 0.4095,
1256
+ "step": 1720
1257
+ },
1258
+ {
1259
+ "epoch": 0.83,
1260
+ "learning_rate": 9.41240987436695e-06,
1261
+ "loss": 0.4629,
1262
+ "step": 1730
1263
+ },
1264
+ {
1265
+ "epoch": 0.83,
1266
+ "learning_rate": 9.40300463161913e-06,
1267
+ "loss": 0.433,
1268
+ "step": 1740
1269
+ },
1270
+ {
1271
+ "epoch": 0.83,
1272
+ "learning_rate": 9.393529489227745e-06,
1273
+ "loss": 0.4323,
1274
+ "step": 1750
1275
+ },
1276
+ {
1277
+ "epoch": 0.84,
1278
+ "learning_rate": 9.383984597614867e-06,
1279
+ "loss": 0.4339,
1280
+ "step": 1760
1281
+ },
1282
+ {
1283
+ "epoch": 0.84,
1284
+ "learning_rate": 9.374370108309862e-06,
1285
+ "loss": 0.4263,
1286
+ "step": 1770
1287
+ },
1288
+ {
1289
+ "epoch": 0.85,
1290
+ "learning_rate": 9.364686173947e-06,
1291
+ "loss": 0.4095,
1292
+ "step": 1780
1293
+ },
1294
+ {
1295
+ "epoch": 0.85,
1296
+ "learning_rate": 9.354932948263014e-06,
1297
+ "loss": 0.4284,
1298
+ "step": 1790
1299
+ },
1300
+ {
1301
+ "epoch": 0.86,
1302
+ "learning_rate": 9.345110586094672e-06,
1303
+ "loss": 0.4398,
1304
+ "step": 1800
1305
+ },
1306
+ {
1307
+ "epoch": 0.86,
1308
+ "eval_webgpt_accuracy": 0.5873340143003064,
1309
+ "eval_webgpt_loss": 0.6802140474319458,
1310
+ "eval_webgpt_runtime": 63.2638,
1311
+ "eval_webgpt_samples_per_second": 61.9,
1312
+ "eval_webgpt_steps_per_second": 6.196,
1313
+ "step": 1800
1314
+ },
1315
+ {
1316
+ "epoch": 0.86,
1317
+ "eval_hfsummary_accuracy": 0.6609739140948523,
1318
+ "eval_hfsummary_loss": 0.6455074548721313,
1319
+ "eval_hfsummary_runtime": 1241.9993,
1320
+ "eval_hfsummary_samples_per_second": 26.637,
1321
+ "eval_hfsummary_steps_per_second": 2.664,
1322
+ "step": 1800
1323
+ },
1324
+ {
1325
+ "epoch": 0.86,
1326
+ "eval_gptsynthetic_accuracy": 0.9975867269984917,
1327
+ "eval_gptsynthetic_loss": 0.007792285177856684,
1328
+ "eval_gptsynthetic_runtime": 45.7802,
1329
+ "eval_gptsynthetic_samples_per_second": 72.411,
1330
+ "eval_gptsynthetic_steps_per_second": 7.252,
1331
+ "step": 1800
1332
+ },
1333
+ {
1334
+ "epoch": 0.86,
1335
+ "learning_rate": 9.335219243376313e-06,
1336
+ "loss": 0.4483,
1337
+ "step": 1810
1338
+ },
1339
+ {
1340
+ "epoch": 0.87,
1341
+ "learning_rate": 9.325259077137371e-06,
1342
+ "loss": 0.4317,
1343
+ "step": 1820
1344
+ },
1345
+ {
1346
+ "epoch": 0.87,
1347
+ "learning_rate": 9.315230245499886e-06,
1348
+ "loss": 0.4285,
1349
+ "step": 1830
1350
+ },
1351
+ {
1352
+ "epoch": 0.88,
1353
+ "learning_rate": 9.305132907675987e-06,
1354
+ "loss": 0.4044,
1355
+ "step": 1840
1356
+ },
1357
+ {
1358
+ "epoch": 0.88,
1359
+ "learning_rate": 9.294967223965377e-06,
1360
+ "loss": 0.4418,
1361
+ "step": 1850
1362
+ },
1363
+ {
1364
+ "epoch": 0.89,
1365
+ "learning_rate": 9.284733355752775e-06,
1366
+ "loss": 0.4336,
1367
+ "step": 1860
1368
+ },
1369
+ {
1370
+ "epoch": 0.89,
1371
+ "learning_rate": 9.274431465505357e-06,
1372
+ "loss": 0.4481,
1373
+ "step": 1870
1374
+ },
1375
+ {
1376
+ "epoch": 0.9,
1377
+ "learning_rate": 9.264061716770183e-06,
1378
+ "loss": 0.4294,
1379
+ "step": 1880
1380
+ },
1381
+ {
1382
+ "epoch": 0.9,
1383
+ "learning_rate": 9.253624274171595e-06,
1384
+ "loss": 0.4442,
1385
+ "step": 1890
1386
+ },
1387
+ {
1388
+ "epoch": 0.91,
1389
+ "learning_rate": 9.243119303408606e-06,
1390
+ "loss": 0.4618,
1391
+ "step": 1900
1392
+ },
1393
+ {
1394
+ "epoch": 0.91,
1395
+ "learning_rate": 9.232546971252265e-06,
1396
+ "loss": 0.4286,
1397
+ "step": 1910
1398
+ },
1399
+ {
1400
+ "epoch": 0.92,
1401
+ "learning_rate": 9.22190744554302e-06,
1402
+ "loss": 0.4247,
1403
+ "step": 1920
1404
+ },
1405
+ {
1406
+ "epoch": 0.92,
1407
+ "learning_rate": 9.211200895188035e-06,
1408
+ "loss": 0.4145,
1409
+ "step": 1930
1410
+ },
1411
+ {
1412
+ "epoch": 0.93,
1413
+ "learning_rate": 9.200427490158532e-06,
1414
+ "loss": 0.4019,
1415
+ "step": 1940
1416
+ },
1417
+ {
1418
+ "epoch": 0.93,
1419
+ "learning_rate": 9.189587401487073e-06,
1420
+ "loss": 0.4154,
1421
+ "step": 1950
1422
+ },
1423
+ {
1424
+ "epoch": 0.94,
1425
+ "learning_rate": 9.178680801264855e-06,
1426
+ "loss": 0.4267,
1427
+ "step": 1960
1428
+ },
1429
+ {
1430
+ "epoch": 0.94,
1431
+ "learning_rate": 9.167707862638973e-06,
1432
+ "loss": 0.4344,
1433
+ "step": 1970
1434
+ },
1435
+ {
1436
+ "epoch": 0.94,
1437
+ "learning_rate": 9.156668759809678e-06,
1438
+ "loss": 0.4095,
1439
+ "step": 1980
1440
+ },
1441
+ {
1442
+ "epoch": 0.95,
1443
+ "learning_rate": 9.145563668027599e-06,
1444
+ "loss": 0.4053,
1445
+ "step": 1990
1446
+ },
1447
+ {
1448
+ "epoch": 0.95,
1449
+ "learning_rate": 9.134392763590978e-06,
1450
+ "loss": 0.4244,
1451
+ "step": 2000
1452
+ },
1453
+ {
1454
+ "epoch": 0.95,
1455
+ "eval_webgpt_accuracy": 0.585291113381001,
1456
+ "eval_webgpt_loss": 0.6761194467544556,
1457
+ "eval_webgpt_runtime": 63.2755,
1458
+ "eval_webgpt_samples_per_second": 61.888,
1459
+ "eval_webgpt_steps_per_second": 6.195,
1460
+ "step": 2000
1461
+ },
1462
+ {
1463
+ "epoch": 0.95,
1464
+ "eval_hfsummary_accuracy": 0.6690142973732733,
1465
+ "eval_hfsummary_loss": 0.6501129269599915,
1466
+ "eval_hfsummary_runtime": 1243.8141,
1467
+ "eval_hfsummary_samples_per_second": 26.598,
1468
+ "eval_hfsummary_steps_per_second": 2.66,
1469
+ "step": 2000
1470
+ },
1471
+ {
1472
+ "epoch": 0.95,
1473
+ "eval_gptsynthetic_accuracy": 0.9978883861236802,
1474
+ "eval_gptsynthetic_loss": 0.0076271467842161655,
1475
+ "eval_gptsynthetic_runtime": 45.8118,
1476
+ "eval_gptsynthetic_samples_per_second": 72.361,
1477
+ "eval_gptsynthetic_steps_per_second": 7.247,
1478
+ "step": 2000
1479
+ },
1480
+ {
1481
+ "epoch": 0.96,
1482
+ "learning_rate": 9.124282826330867e-06,
1483
+ "loss": 0.4132,
1484
+ "step": 2010
1485
+ },
1486
+ {
1487
+ "epoch": 0.96,
1488
+ "learning_rate": 9.112987367291958e-06,
1489
+ "loss": 0.4089,
1490
+ "step": 2020
1491
+ },
1492
+ {
1493
+ "epoch": 0.97,
1494
+ "learning_rate": 9.101626612761673e-06,
1495
+ "loss": 0.42,
1496
+ "step": 2030
1497
+ },
1498
+ {
1499
+ "epoch": 0.97,
1500
+ "learning_rate": 9.090200743097004e-06,
1501
+ "loss": 0.3961,
1502
+ "step": 2040
1503
+ },
1504
+ {
1505
+ "epoch": 0.98,
1506
+ "learning_rate": 9.078709939688683e-06,
1507
+ "loss": 0.4139,
1508
+ "step": 2050
1509
+ },
1510
+ {
1511
+ "epoch": 0.98,
1512
+ "learning_rate": 9.06715438495829e-06,
1513
+ "loss": 0.4021,
1514
+ "step": 2060
1515
+ },
1516
+ {
1517
+ "epoch": 0.99,
1518
+ "learning_rate": 9.055534262355358e-06,
1519
+ "loss": 0.4246,
1520
+ "step": 2070
1521
+ },
1522
+ {
1523
+ "epoch": 0.99,
1524
+ "learning_rate": 9.043849756354469e-06,
1525
+ "loss": 0.4233,
1526
+ "step": 2080
1527
+ },
1528
+ {
1529
+ "epoch": 1.0,
1530
+ "learning_rate": 9.032101052452318e-06,
1531
+ "loss": 0.4121,
1532
+ "step": 2090
1533
+ },
1534
+ {
1535
+ "epoch": 1.0,
1536
+ "learning_rate": 9.020288337164766e-06,
1537
+ "loss": 0.4299,
1538
+ "step": 2100
1539
+ },
1540
+ {
1541
+ "epoch": 1.01,
1542
+ "learning_rate": 9.00841179802389e-06,
1543
+ "loss": 0.3955,
1544
+ "step": 2110
1545
+ },
1546
+ {
1547
+ "epoch": 1.01,
1548
+ "learning_rate": 8.996471623574997e-06,
1549
+ "loss": 0.4185,
1550
+ "step": 2120
1551
+ },
1552
+ {
1553
+ "epoch": 1.02,
1554
+ "learning_rate": 8.984468003373625e-06,
1555
+ "loss": 0.4086,
1556
+ "step": 2130
1557
+ },
1558
+ {
1559
+ "epoch": 1.02,
1560
+ "learning_rate": 8.97240112798255e-06,
1561
+ "loss": 0.4061,
1562
+ "step": 2140
1563
+ },
1564
+ {
1565
+ "epoch": 1.03,
1566
+ "learning_rate": 8.960271188968753e-06,
1567
+ "loss": 0.3859,
1568
+ "step": 2150
1569
+ },
1570
+ {
1571
+ "epoch": 1.03,
1572
+ "learning_rate": 8.948078378900368e-06,
1573
+ "loss": 0.3951,
1574
+ "step": 2160
1575
+ },
1576
+ {
1577
+ "epoch": 1.04,
1578
+ "learning_rate": 8.935822891343647e-06,
1579
+ "loss": 0.3914,
1580
+ "step": 2170
1581
+ },
1582
+ {
1583
+ "epoch": 1.04,
1584
+ "learning_rate": 8.923504920859864e-06,
1585
+ "loss": 0.3672,
1586
+ "step": 2180
1587
+ },
1588
+ {
1589
+ "epoch": 1.04,
1590
+ "learning_rate": 8.911124663002243e-06,
1591
+ "loss": 0.385,
1592
+ "step": 2190
1593
+ },
1594
+ {
1595
+ "epoch": 1.05,
1596
+ "learning_rate": 8.898682314312852e-06,
1597
+ "loss": 0.3932,
1598
+ "step": 2200
1599
+ },
1600
+ {
1601
+ "epoch": 1.05,
1602
+ "eval_webgpt_accuracy": 0.5681818181818182,
1603
+ "eval_webgpt_loss": 0.6895765662193298,
1604
+ "eval_webgpt_runtime": 63.5181,
1605
+ "eval_webgpt_samples_per_second": 61.652,
1606
+ "eval_webgpt_steps_per_second": 6.171,
1607
+ "step": 2200
1608
+ },
1609
+ {
1610
+ "epoch": 1.05,
1611
+ "eval_hfsummary_accuracy": 0.6638757065562373,
1612
+ "eval_hfsummary_loss": 0.6736522912979126,
1613
+ "eval_hfsummary_runtime": 1274.8625,
1614
+ "eval_hfsummary_samples_per_second": 25.95,
1615
+ "eval_hfsummary_steps_per_second": 2.596,
1616
+ "step": 2200
1617
+ },
1618
+ {
1619
+ "epoch": 1.05,
1620
+ "eval_gptsynthetic_accuracy": 0.9984917043740573,
1621
+ "eval_gptsynthetic_loss": 0.006684631574898958,
1622
+ "eval_gptsynthetic_runtime": 46.0527,
1623
+ "eval_gptsynthetic_samples_per_second": 71.983,
1624
+ "eval_gptsynthetic_steps_per_second": 7.209,
1625
+ "step": 2200
1626
+ },
1627
+ {
1628
+ "epoch": 1.05,
1629
+ "learning_rate": 8.886178072319464e-06,
1630
+ "loss": 0.3868,
1631
+ "step": 2210
1632
+ },
1633
+ {
1634
+ "epoch": 1.06,
1635
+ "learning_rate": 8.873612135532451e-06,
1636
+ "loss": 0.3712,
1637
+ "step": 2220
1638
+ },
1639
+ {
1640
+ "epoch": 1.06,
1641
+ "learning_rate": 8.860984703441608e-06,
1642
+ "loss": 0.4153,
1643
+ "step": 2230
1644
+ },
1645
+ {
1646
+ "epoch": 1.07,
1647
+ "learning_rate": 8.848295976512996e-06,
1648
+ "loss": 0.4364,
1649
+ "step": 2240
1650
+ },
1651
+ {
1652
+ "epoch": 1.07,
1653
+ "learning_rate": 8.835546156185765e-06,
1654
+ "loss": 0.3733,
1655
+ "step": 2250
1656
+ },
1657
+ {
1658
+ "epoch": 1.08,
1659
+ "learning_rate": 8.822735444868941e-06,
1660
+ "loss": 0.389,
1661
+ "step": 2260
1662
+ },
1663
+ {
1664
+ "epoch": 1.08,
1665
+ "learning_rate": 8.809864045938227e-06,
1666
+ "loss": 0.3869,
1667
+ "step": 2270
1668
+ },
1669
+ {
1670
+ "epoch": 1.09,
1671
+ "learning_rate": 8.796932163732768e-06,
1672
+ "loss": 0.3782,
1673
+ "step": 2280
1674
+ },
1675
+ {
1676
+ "epoch": 1.09,
1677
+ "learning_rate": 8.783940003551902e-06,
1678
+ "loss": 0.4087,
1679
+ "step": 2290
1680
+ },
1681
+ {
1682
+ "epoch": 1.1,
1683
+ "learning_rate": 8.770887771651916e-06,
1684
+ "loss": 0.3849,
1685
+ "step": 2300
1686
+ },
1687
+ {
1688
+ "epoch": 1.1,
1689
+ "learning_rate": 8.757775675242757e-06,
1690
+ "loss": 0.3848,
1691
+ "step": 2310
1692
+ },
1693
+ {
1694
+ "epoch": 1.11,
1695
+ "learning_rate": 8.744603922484746e-06,
1696
+ "loss": 0.4036,
1697
+ "step": 2320
1698
+ },
1699
+ {
1700
+ "epoch": 1.11,
1701
+ "learning_rate": 8.731372722485276e-06,
1702
+ "loss": 0.3931,
1703
+ "step": 2330
1704
+ },
1705
+ {
1706
+ "epoch": 1.12,
1707
+ "learning_rate": 8.718082285295496e-06,
1708
+ "loss": 0.3909,
1709
+ "step": 2340
1710
+ },
1711
+ {
1712
+ "epoch": 1.12,
1713
+ "learning_rate": 8.70473282190696e-06,
1714
+ "loss": 0.3912,
1715
+ "step": 2350
1716
+ },
1717
+ {
1718
+ "epoch": 1.13,
1719
+ "learning_rate": 8.691324544248303e-06,
1720
+ "loss": 0.3874,
1721
+ "step": 2360
1722
+ },
1723
+ {
1724
+ "epoch": 1.13,
1725
+ "learning_rate": 8.677857665181854e-06,
1726
+ "loss": 0.3878,
1727
+ "step": 2370
1728
+ },
1729
+ {
1730
+ "epoch": 1.14,
1731
+ "learning_rate": 8.664332398500269e-06,
1732
+ "loss": 0.4063,
1733
+ "step": 2380
1734
+ },
1735
+ {
1736
+ "epoch": 1.14,
1737
+ "learning_rate": 8.650748958923128e-06,
1738
+ "loss": 0.3794,
1739
+ "step": 2390
1740
+ },
1741
+ {
1742
+ "epoch": 1.15,
1743
+ "learning_rate": 8.637107562093537e-06,
1744
+ "loss": 0.3514,
1745
+ "step": 2400
1746
+ },
1747
+ {
1748
+ "epoch": 1.15,
1749
+ "eval_webgpt_accuracy": 0.5840143003064351,
1750
+ "eval_webgpt_loss": 0.6855098605155945,
1751
+ "eval_webgpt_runtime": 63.4383,
1752
+ "eval_webgpt_samples_per_second": 61.729,
1753
+ "eval_webgpt_steps_per_second": 6.179,
1754
+ "step": 2400
1755
+ },
1756
+ {
1757
+ "epoch": 1.15,
1758
+ "eval_hfsummary_accuracy": 0.6732158510413203,
1759
+ "eval_hfsummary_loss": 0.6792841553688049,
1760
+ "eval_hfsummary_runtime": 1256.3167,
1761
+ "eval_hfsummary_samples_per_second": 26.333,
1762
+ "eval_hfsummary_steps_per_second": 2.634,
1763
+ "step": 2400
1764
+ },
1765
+ {
1766
+ "epoch": 1.15,
1767
+ "eval_gptsynthetic_accuracy": 0.9978883861236802,
1768
+ "eval_gptsynthetic_loss": 0.00653717340901494,
1769
+ "eval_gptsynthetic_runtime": 45.9265,
1770
+ "eval_gptsynthetic_samples_per_second": 72.181,
1771
+ "eval_gptsynthetic_steps_per_second": 7.229,
1772
+ "step": 2400
1773
+ },
1774
+ {
1775
+ "epoch": 1.15,
1776
+ "learning_rate": 8.623408424574695e-06,
1777
+ "loss": 0.3742,
1778
+ "step": 2410
1779
+ },
1780
+ {
1781
+ "epoch": 1.15,
1782
+ "learning_rate": 8.609651763846462e-06,
1783
+ "loss": 0.3696,
1784
+ "step": 2420
1785
+ },
1786
+ {
1787
+ "epoch": 1.16,
1788
+ "learning_rate": 8.595837798301908e-06,
1789
+ "loss": 0.3943,
1790
+ "step": 2430
1791
+ },
1792
+ {
1793
+ "epoch": 1.16,
1794
+ "learning_rate": 8.581966747243835e-06,
1795
+ "loss": 0.3928,
1796
+ "step": 2440
1797
+ },
1798
+ {
1799
+ "epoch": 1.17,
1800
+ "learning_rate": 8.568038830881305e-06,
1801
+ "loss": 0.3753,
1802
+ "step": 2450
1803
+ },
1804
+ {
1805
+ "epoch": 1.17,
1806
+ "learning_rate": 8.554054270326144e-06,
1807
+ "loss": 0.3594,
1808
+ "step": 2460
1809
+ },
1810
+ {
1811
+ "epoch": 1.18,
1812
+ "learning_rate": 8.54001328758943e-06,
1813
+ "loss": 0.3674,
1814
+ "step": 2470
1815
+ },
1816
+ {
1817
+ "epoch": 1.18,
1818
+ "learning_rate": 8.525916105577967e-06,
1819
+ "loss": 0.3714,
1820
+ "step": 2480
1821
+ },
1822
+ {
1823
+ "epoch": 1.19,
1824
+ "learning_rate": 8.511762948090745e-06,
1825
+ "loss": 0.3937,
1826
+ "step": 2490
1827
+ },
1828
+ {
1829
+ "epoch": 1.19,
1830
+ "learning_rate": 8.49755403981539e-06,
1831
+ "loss": 0.3848,
1832
+ "step": 2500
1833
+ },
1834
+ {
1835
+ "epoch": 1.2,
1836
+ "learning_rate": 8.4832896063246e-06,
1837
+ "loss": 0.3709,
1838
+ "step": 2510
1839
+ },
1840
+ {
1841
+ "epoch": 1.2,
1842
+ "learning_rate": 8.46896987407256e-06,
1843
+ "loss": 0.4056,
1844
+ "step": 2520
1845
+ },
1846
+ {
1847
+ "epoch": 1.21,
1848
+ "learning_rate": 8.45459507039134e-06,
1849
+ "loss": 0.3794,
1850
+ "step": 2530
1851
+ },
1852
+ {
1853
+ "epoch": 1.21,
1854
+ "learning_rate": 8.440165423487302e-06,
1855
+ "loss": 0.3597,
1856
+ "step": 2540
1857
+ },
1858
+ {
1859
+ "epoch": 1.22,
1860
+ "learning_rate": 8.425681162437464e-06,
1861
+ "loss": 0.3773,
1862
+ "step": 2550
1863
+ },
1864
+ {
1865
+ "epoch": 1.22,
1866
+ "learning_rate": 8.41114251718587e-06,
1867
+ "loss": 0.3678,
1868
+ "step": 2560
1869
+ },
1870
+ {
1871
+ "epoch": 1.23,
1872
+ "learning_rate": 8.396549718539934e-06,
1873
+ "loss": 0.3831,
1874
+ "step": 2570
1875
+ },
1876
+ {
1877
+ "epoch": 1.23,
1878
+ "learning_rate": 8.381902998166782e-06,
1879
+ "loss": 0.3873,
1880
+ "step": 2580
1881
+ },
1882
+ {
1883
+ "epoch": 1.24,
1884
+ "learning_rate": 8.367202588589569e-06,
1885
+ "loss": 0.3775,
1886
+ "step": 2590
1887
+ },
1888
+ {
1889
+ "epoch": 1.24,
1890
+ "learning_rate": 8.352448723183792e-06,
1891
+ "loss": 0.3936,
1892
+ "step": 2600
1893
+ },
1894
+ {
1895
+ "epoch": 1.24,
1896
+ "eval_webgpt_accuracy": 0.5760980592441267,
1897
+ "eval_webgpt_loss": 0.6807132959365845,
1898
+ "eval_webgpt_runtime": 68.3724,
1899
+ "eval_webgpt_samples_per_second": 57.275,
1900
+ "eval_webgpt_steps_per_second": 5.733,
1901
+ "step": 2600
1902
+ },
1903
+ {
1904
+ "epoch": 1.24,
1905
+ "eval_hfsummary_accuracy": 0.6699815615270683,
1906
+ "eval_hfsummary_loss": 0.6664829254150391,
1907
+ "eval_hfsummary_runtime": 1323.7739,
1908
+ "eval_hfsummary_samples_per_second": 24.991,
1909
+ "eval_hfsummary_steps_per_second": 2.5,
1910
+ "step": 2600
1911
+ },
1912
+ {
1913
+ "epoch": 1.24,
1914
+ "eval_gptsynthetic_accuracy": 0.9984917043740573,
1915
+ "eval_gptsynthetic_loss": 0.006850704550743103,
1916
+ "eval_gptsynthetic_runtime": 48.9916,
1917
+ "eval_gptsynthetic_samples_per_second": 67.665,
1918
+ "eval_gptsynthetic_steps_per_second": 6.777,
1919
+ "step": 2600
1920
+ },
1921
+ {
1922
+ "epoch": 1.25,
1923
+ "learning_rate": 8.337641636173582e-06,
1924
+ "loss": 0.3647,
1925
+ "step": 2610
1926
+ },
1927
+ {
1928
+ "epoch": 1.25,
1929
+ "learning_rate": 8.322781562627987e-06,
1930
+ "loss": 0.3815,
1931
+ "step": 2620
1932
+ },
1933
+ {
1934
+ "epoch": 1.25,
1935
+ "learning_rate": 8.307868738457236e-06,
1936
+ "loss": 0.3979,
1937
+ "step": 2630
1938
+ },
1939
+ {
1940
+ "epoch": 1.26,
1941
+ "learning_rate": 8.292903400409005e-06,
1942
+ "loss": 0.3685,
1943
+ "step": 2640
1944
+ },
1945
+ {
1946
+ "epoch": 1.26,
1947
+ "learning_rate": 8.27788578606464e-06,
1948
+ "loss": 0.3498,
1949
+ "step": 2650
1950
+ },
1951
+ {
1952
+ "epoch": 1.27,
1953
+ "learning_rate": 8.262816133835412e-06,
1954
+ "loss": 0.3641,
1955
+ "step": 2660
1956
+ },
1957
+ {
1958
+ "epoch": 1.27,
1959
+ "learning_rate": 8.247694682958699e-06,
1960
+ "loss": 0.4006,
1961
+ "step": 2670
1962
+ },
1963
+ {
1964
+ "epoch": 1.28,
1965
+ "learning_rate": 8.232521673494216e-06,
1966
+ "loss": 0.3857,
1967
+ "step": 2680
1968
+ },
1969
+ {
1970
+ "epoch": 1.28,
1971
+ "learning_rate": 8.217297346320192e-06,
1972
+ "loss": 0.3838,
1973
+ "step": 2690
1974
+ },
1975
+ {
1976
+ "epoch": 1.29,
1977
+ "learning_rate": 8.202021943129548e-06,
1978
+ "loss": 0.3979,
1979
+ "step": 2700
1980
+ },
1981
+ {
1982
+ "epoch": 1.29,
1983
+ "learning_rate": 8.186695706426053e-06,
1984
+ "loss": 0.3589,
1985
+ "step": 2710
1986
+ },
1987
+ {
1988
+ "epoch": 1.3,
1989
+ "learning_rate": 8.17131887952049e-06,
1990
+ "loss": 0.3853,
1991
+ "step": 2720
1992
+ },
1993
+ {
1994
+ "epoch": 1.3,
1995
+ "learning_rate": 8.155891706526773e-06,
1996
+ "loss": 0.3698,
1997
+ "step": 2730
1998
+ },
1999
+ {
2000
+ "epoch": 1.31,
2001
+ "learning_rate": 8.140414432358094e-06,
2002
+ "loss": 0.3637,
2003
+ "step": 2740
2004
+ },
2005
+ {
2006
+ "epoch": 1.31,
2007
+ "learning_rate": 8.124887302723014e-06,
2008
+ "loss": 0.3742,
2009
+ "step": 2750
2010
+ },
2011
+ {
2012
+ "epoch": 1.32,
2013
+ "learning_rate": 8.109310564121577e-06,
2014
+ "loss": 0.3849,
2015
+ "step": 2760
2016
+ },
2017
+ {
2018
+ "epoch": 1.32,
2019
+ "learning_rate": 8.09368446384139e-06,
2020
+ "loss": 0.3721,
2021
+ "step": 2770
2022
+ },
2023
+ {
2024
+ "epoch": 1.33,
2025
+ "learning_rate": 8.078009249953693e-06,
2026
+ "loss": 0.3818,
2027
+ "step": 2780
2028
+ },
2029
+ {
2030
+ "epoch": 1.33,
2031
+ "learning_rate": 8.062285171309437e-06,
2032
+ "loss": 0.384,
2033
+ "step": 2790
2034
+ },
2035
+ {
2036
+ "epoch": 1.34,
2037
+ "learning_rate": 8.046512477535315e-06,
2038
+ "loss": 0.3567,
2039
+ "step": 2800
2040
+ },
2041
+ {
2042
+ "epoch": 1.34,
2043
+ "eval_webgpt_accuracy": 0.5822267620020429,
2044
+ "eval_webgpt_loss": 0.6815086007118225,
2045
+ "eval_webgpt_runtime": 68.1221,
2046
+ "eval_webgpt_samples_per_second": 57.485,
2047
+ "eval_webgpt_steps_per_second": 5.754,
2048
+ "step": 2800
2049
+ },
2050
+ {
2051
+ "epoch": 1.34,
2052
+ "eval_hfsummary_accuracy": 0.6710697337000877,
2053
+ "eval_hfsummary_loss": 0.6821005940437317,
2054
+ "eval_hfsummary_runtime": 1319.5773,
2055
+ "eval_hfsummary_samples_per_second": 25.071,
2056
+ "eval_hfsummary_steps_per_second": 2.508,
2057
+ "step": 2800
2058
+ },
2059
+ {
2060
+ "epoch": 1.34,
2061
+ "eval_gptsynthetic_accuracy": 0.9978883861236802,
2062
+ "eval_gptsynthetic_loss": 0.005701866466552019,
2063
+ "eval_gptsynthetic_runtime": 48.8957,
2064
+ "eval_gptsynthetic_samples_per_second": 67.797,
2065
+ "eval_gptsynthetic_steps_per_second": 6.79,
2066
+ "step": 2800
2067
+ },
2068
+ {
2069
+ "epoch": 1.34,
2070
+ "learning_rate": 8.030691419029806e-06,
2071
+ "loss": 0.3813,
2072
+ "step": 2810
2073
+ },
2074
+ {
2075
+ "epoch": 1.35,
2076
+ "learning_rate": 8.014822246959203e-06,
2077
+ "loss": 0.3645,
2078
+ "step": 2820
2079
+ },
2080
+ {
2081
+ "epoch": 1.35,
2082
+ "learning_rate": 7.998905213253627e-06,
2083
+ "loss": 0.3752,
2084
+ "step": 2830
2085
+ },
2086
+ {
2087
+ "epoch": 1.35,
2088
+ "learning_rate": 7.982940570603014e-06,
2089
+ "loss": 0.4075,
2090
+ "step": 2840
2091
+ },
2092
+ {
2093
+ "epoch": 1.36,
2094
+ "learning_rate": 7.966928572453123e-06,
2095
+ "loss": 0.3823,
2096
+ "step": 2850
2097
+ },
2098
+ {
2099
+ "epoch": 1.36,
2100
+ "learning_rate": 7.950869473001494e-06,
2101
+ "loss": 0.3776,
2102
+ "step": 2860
2103
+ },
2104
+ {
2105
+ "epoch": 1.37,
2106
+ "learning_rate": 7.93476352719343e-06,
2107
+ "loss": 0.3689,
2108
+ "step": 2870
2109
+ },
2110
+ {
2111
+ "epoch": 1.37,
2112
+ "learning_rate": 7.91861099071793e-06,
2113
+ "loss": 0.3883,
2114
+ "step": 2880
2115
+ },
2116
+ {
2117
+ "epoch": 1.38,
2118
+ "learning_rate": 7.902412120003649e-06,
2119
+ "loss": 0.3736,
2120
+ "step": 2890
2121
+ },
2122
+ {
2123
+ "epoch": 1.38,
2124
+ "learning_rate": 7.886167172214814e-06,
2125
+ "loss": 0.3363,
2126
+ "step": 2900
2127
+ },
2128
+ {
2129
+ "epoch": 1.39,
2130
+ "learning_rate": 7.86987640524715e-06,
2131
+ "loss": 0.3866,
2132
+ "step": 2910
2133
+ },
2134
+ {
2135
+ "epoch": 1.39,
2136
+ "learning_rate": 7.853540077723779e-06,
2137
+ "loss": 0.3717,
2138
+ "step": 2920
2139
+ },
2140
+ {
2141
+ "epoch": 1.4,
2142
+ "learning_rate": 7.837158448991112e-06,
2143
+ "loss": 0.3665,
2144
+ "step": 2930
2145
+ },
2146
+ {
2147
+ "epoch": 1.4,
2148
+ "learning_rate": 7.820731779114746e-06,
2149
+ "loss": 0.3814,
2150
+ "step": 2940
2151
+ },
2152
+ {
2153
+ "epoch": 1.41,
2154
+ "learning_rate": 7.804260328875322e-06,
2155
+ "loss": 0.3703,
2156
+ "step": 2950
2157
+ },
2158
+ {
2159
+ "epoch": 1.41,
2160
+ "learning_rate": 7.787744359764388e-06,
2161
+ "loss": 0.3896,
2162
+ "step": 2960
2163
+ },
2164
+ {
2165
+ "epoch": 1.42,
2166
+ "learning_rate": 7.77118413398025e-06,
2167
+ "loss": 0.3566,
2168
+ "step": 2970
2169
+ },
2170
+ {
2171
+ "epoch": 1.42,
2172
+ "learning_rate": 7.754579914423806e-06,
2173
+ "loss": 0.3743,
2174
+ "step": 2980
2175
+ },
2176
+ {
2177
+ "epoch": 1.43,
2178
+ "learning_rate": 7.737931964694378e-06,
2179
+ "loss": 0.3657,
2180
+ "step": 2990
2181
+ },
2182
+ {
2183
+ "epoch": 1.43,
2184
+ "learning_rate": 7.721240549085523e-06,
2185
+ "loss": 0.3956,
2186
+ "step": 3000
2187
+ },
2188
+ {
2189
+ "epoch": 1.43,
2190
+ "eval_webgpt_accuracy": 0.5799284984678243,
2191
+ "eval_webgpt_loss": 0.6848716139793396,
2192
+ "eval_webgpt_runtime": 67.9306,
2193
+ "eval_webgpt_samples_per_second": 57.647,
2194
+ "eval_webgpt_steps_per_second": 5.771,
2195
+ "step": 3000
2196
+ },
2197
+ {
2198
+ "epoch": 1.43,
2199
+ "eval_hfsummary_accuracy": 0.6674122661185503,
2200
+ "eval_hfsummary_loss": 0.6731451153755188,
2201
+ "eval_hfsummary_runtime": 1273.6757,
2202
+ "eval_hfsummary_samples_per_second": 25.974,
2203
+ "eval_hfsummary_steps_per_second": 2.598,
2204
+ "step": 3000
2205
+ },
2206
+ {
2207
+ "epoch": 1.43,
2208
+ "eval_gptsynthetic_accuracy": 0.9984917043740573,
2209
+ "eval_gptsynthetic_loss": 0.005726581439375877,
2210
+ "eval_gptsynthetic_runtime": 48.6337,
2211
+ "eval_gptsynthetic_samples_per_second": 68.163,
2212
+ "eval_gptsynthetic_steps_per_second": 6.827,
2213
+ "step": 3000
2214
+ },
2215
+ {
2216
+ "epoch": 1.44,
2217
+ "learning_rate": 7.704505932580835e-06,
2218
+ "loss": 0.3617,
2219
+ "step": 3010
2220
+ },
2221
+ {
2222
+ "epoch": 1.44,
2223
+ "learning_rate": 7.687728380849746e-06,
2224
+ "loss": 0.3944,
2225
+ "step": 3020
2226
+ },
2227
+ {
2228
+ "epoch": 1.45,
2229
+ "learning_rate": 7.6709081602433e-06,
2230
+ "loss": 0.347,
2231
+ "step": 3030
2232
+ },
2233
+ {
2234
+ "epoch": 1.45,
2235
+ "learning_rate": 7.654045537789929e-06,
2236
+ "loss": 0.3657,
2237
+ "step": 3040
2238
+ },
2239
+ {
2240
+ "epoch": 1.46,
2241
+ "learning_rate": 7.637140781191214e-06,
2242
+ "loss": 0.3874,
2243
+ "step": 3050
2244
+ },
2245
+ {
2246
+ "epoch": 1.46,
2247
+ "learning_rate": 7.6201941588176355e-06,
2248
+ "loss": 0.3875,
2249
+ "step": 3060
2250
+ },
2251
+ {
2252
+ "epoch": 1.46,
2253
+ "learning_rate": 7.603205939704308e-06,
2254
+ "loss": 0.371,
2255
+ "step": 3070
2256
+ },
2257
+ {
2258
+ "epoch": 1.47,
2259
+ "learning_rate": 7.586176393546714e-06,
2260
+ "loss": 0.4006,
2261
+ "step": 3080
2262
+ },
2263
+ {
2264
+ "epoch": 1.47,
2265
+ "learning_rate": 7.569105790696423e-06,
2266
+ "loss": 0.3721,
2267
+ "step": 3090
2268
+ },
2269
+ {
2270
+ "epoch": 1.48,
2271
+ "learning_rate": 7.551994402156794e-06,
2272
+ "loss": 0.3825,
2273
+ "step": 3100
2274
+ },
2275
+ {
2276
+ "epoch": 1.48,
2277
+ "learning_rate": 7.5348424995786814e-06,
2278
+ "loss": 0.3688,
2279
+ "step": 3110
2280
+ },
2281
+ {
2282
+ "epoch": 1.49,
2283
+ "learning_rate": 7.517650355256116e-06,
2284
+ "loss": 0.3698,
2285
+ "step": 3120
2286
+ },
2287
+ {
2288
+ "epoch": 1.49,
2289
+ "learning_rate": 7.500418242121982e-06,
2290
+ "loss": 0.3623,
2291
+ "step": 3130
2292
+ },
2293
+ {
2294
+ "epoch": 1.5,
2295
+ "learning_rate": 7.483146433743686e-06,
2296
+ "loss": 0.3739,
2297
+ "step": 3140
2298
+ },
2299
+ {
2300
+ "epoch": 1.5,
2301
+ "learning_rate": 7.465835204318822e-06,
2302
+ "loss": 0.3708,
2303
+ "step": 3150
2304
+ },
2305
+ {
2306
+ "epoch": 1.51,
2307
+ "learning_rate": 7.448484828670799e-06,
2308
+ "loss": 0.3344,
2309
+ "step": 3160
2310
+ },
2311
+ {
2312
+ "epoch": 1.51,
2313
+ "learning_rate": 7.431095582244495e-06,
2314
+ "loss": 0.3548,
2315
+ "step": 3170
2316
+ },
2317
+ {
2318
+ "epoch": 1.52,
2319
+ "learning_rate": 7.413667741101881e-06,
2320
+ "loss": 0.3702,
2321
+ "step": 3180
2322
+ },
2323
+ {
2324
+ "epoch": 1.52,
2325
+ "learning_rate": 7.3962015819176324e-06,
2326
+ "loss": 0.3361,
2327
+ "step": 3190
2328
+ },
2329
+ {
2330
+ "epoch": 1.53,
2331
+ "learning_rate": 7.378697381974744e-06,
2332
+ "loss": 0.4019,
2333
+ "step": 3200
2334
+ },
2335
+ {
2336
+ "epoch": 1.53,
2337
+ "eval_webgpt_accuracy": 0.5891215526046987,
2338
+ "eval_webgpt_loss": 0.6808570623397827,
2339
+ "eval_webgpt_runtime": 63.3758,
2340
+ "eval_webgpt_samples_per_second": 61.79,
2341
+ "eval_webgpt_steps_per_second": 6.185,
2342
+ "step": 3200
2343
+ },
2344
+ {
2345
+ "epoch": 1.53,
2346
+ "eval_hfsummary_accuracy": 0.6569839494604479,
2347
+ "eval_hfsummary_loss": 0.692561686038971,
2348
+ "eval_hfsummary_runtime": 1242.943,
2349
+ "eval_hfsummary_samples_per_second": 26.617,
2350
+ "eval_hfsummary_steps_per_second": 2.662,
2351
+ "step": 3200
2352
+ },
2353
+ {
2354
+ "epoch": 1.53,
2355
+ "eval_gptsynthetic_accuracy": 0.9987933634992459,
2356
+ "eval_gptsynthetic_loss": 0.005168143659830093,
2357
+ "eval_gptsynthetic_runtime": 45.8689,
2358
+ "eval_gptsynthetic_samples_per_second": 72.271,
2359
+ "eval_gptsynthetic_steps_per_second": 7.238,
2360
+ "step": 3200
2361
+ },
2362
+ {
2363
+ "epoch": 1.53,
2364
+ "learning_rate": 7.361155419160121e-06,
2365
+ "loss": 0.3573,
2366
+ "step": 3210
2367
+ },
2368
+ {
2369
+ "epoch": 1.54,
2370
+ "learning_rate": 7.343575971960171e-06,
2371
+ "loss": 0.3677,
2372
+ "step": 3220
2373
+ },
2374
+ {
2375
+ "epoch": 1.54,
2376
+ "learning_rate": 7.325959319456386e-06,
2377
+ "loss": 0.365,
2378
+ "step": 3230
2379
+ },
2380
+ {
2381
+ "epoch": 1.55,
2382
+ "learning_rate": 7.308305741320907e-06,
2383
+ "loss": 0.3241,
2384
+ "step": 3240
2385
+ },
2386
+ {
2387
+ "epoch": 1.55,
2388
+ "learning_rate": 7.290615517812084e-06,
2389
+ "loss": 0.3639,
2390
+ "step": 3250
2391
+ },
2392
+ {
2393
+ "epoch": 1.56,
2394
+ "learning_rate": 7.27288892977003e-06,
2395
+ "loss": 0.3535,
2396
+ "step": 3260
2397
+ },
2398
+ {
2399
+ "epoch": 1.56,
2400
+ "learning_rate": 7.255126258612163e-06,
2401
+ "loss": 0.364,
2402
+ "step": 3270
2403
+ },
2404
+ {
2405
+ "epoch": 1.56,
2406
+ "learning_rate": 7.237327786328732e-06,
2407
+ "loss": 0.3728,
2408
+ "step": 3280
2409
+ },
2410
+ {
2411
+ "epoch": 1.57,
2412
+ "learning_rate": 7.219493795478347e-06,
2413
+ "loss": 0.3598,
2414
+ "step": 3290
2415
+ },
2416
+ {
2417
+ "epoch": 1.57,
2418
+ "learning_rate": 7.201624569183492e-06,
2419
+ "loss": 0.3749,
2420
+ "step": 3300
2421
+ },
2422
+ {
2423
+ "epoch": 1.58,
2424
+ "learning_rate": 7.183720391126028e-06,
2425
+ "loss": 0.3361,
2426
+ "step": 3310
2427
+ },
2428
+ {
2429
+ "epoch": 1.58,
2430
+ "learning_rate": 7.16578154554269e-06,
2431
+ "loss": 0.3383,
2432
+ "step": 3320
2433
+ },
2434
+ {
2435
+ "epoch": 1.59,
2436
+ "learning_rate": 7.147808317220578e-06,
2437
+ "loss": 0.3424,
2438
+ "step": 3330
2439
+ },
2440
+ {
2441
+ "epoch": 1.59,
2442
+ "learning_rate": 7.12980099149263e-06,
2443
+ "loss": 0.3627,
2444
+ "step": 3340
2445
+ },
2446
+ {
2447
+ "epoch": 1.6,
2448
+ "learning_rate": 7.111759854233096e-06,
2449
+ "loss": 0.3498,
2450
+ "step": 3350
2451
+ },
2452
+ {
2453
+ "epoch": 1.6,
2454
+ "learning_rate": 7.093685191853003e-06,
2455
+ "loss": 0.3838,
2456
+ "step": 3360
2457
+ },
2458
+ {
2459
+ "epoch": 1.61,
2460
+ "learning_rate": 7.075577291295599e-06,
2461
+ "loss": 0.3638,
2462
+ "step": 3370
2463
+ },
2464
+ {
2465
+ "epoch": 1.61,
2466
+ "learning_rate": 7.0574364400318014e-06,
2467
+ "loss": 0.3737,
2468
+ "step": 3380
2469
+ },
2470
+ {
2471
+ "epoch": 1.62,
2472
+ "learning_rate": 7.0392629260556425e-06,
2473
+ "loss": 0.367,
2474
+ "step": 3390
2475
+ },
2476
+ {
2477
+ "epoch": 1.62,
2478
+ "learning_rate": 7.021057037879681e-06,
2479
+ "loss": 0.3778,
2480
+ "step": 3400
2481
+ },
2482
+ {
2483
+ "epoch": 1.62,
2484
+ "eval_webgpt_accuracy": 0.5845250255362615,
2485
+ "eval_webgpt_loss": 0.6846094727516174,
2486
+ "eval_webgpt_runtime": 63.3427,
2487
+ "eval_webgpt_samples_per_second": 61.822,
2488
+ "eval_webgpt_steps_per_second": 6.189,
2489
+ "step": 3400
2490
+ },
2491
+ {
2492
+ "epoch": 1.62,
2493
+ "eval_hfsummary_accuracy": 0.6690445243780794,
2494
+ "eval_hfsummary_loss": 0.6792688369750977,
2495
+ "eval_hfsummary_runtime": 1245.6911,
2496
+ "eval_hfsummary_samples_per_second": 26.558,
2497
+ "eval_hfsummary_steps_per_second": 2.656,
2498
+ "step": 3400
2499
+ },
2500
+ {
2501
+ "epoch": 1.62,
2502
+ "eval_gptsynthetic_accuracy": 0.9984917043740573,
2503
+ "eval_gptsynthetic_loss": 0.005154031794518232,
2504
+ "eval_gptsynthetic_runtime": 45.8488,
2505
+ "eval_gptsynthetic_samples_per_second": 72.303,
2506
+ "eval_gptsynthetic_steps_per_second": 7.241,
2507
+ "step": 3400
2508
+ },
2509
+ {
2510
+ "epoch": 1.63,
2511
+ "learning_rate": 7.002819064530437e-06,
2512
+ "loss": 0.3706,
2513
+ "step": 3410
2514
+ },
2515
+ {
2516
+ "epoch": 1.63,
2517
+ "learning_rate": 6.984549295543795e-06,
2518
+ "loss": 0.3743,
2519
+ "step": 3420
2520
+ },
2521
+ {
2522
+ "epoch": 1.64,
2523
+ "learning_rate": 6.966248020960404e-06,
2524
+ "loss": 0.3613,
2525
+ "step": 3430
2526
+ },
2527
+ {
2528
+ "epoch": 1.64,
2529
+ "learning_rate": 6.947915531321085e-06,
2530
+ "loss": 0.3375,
2531
+ "step": 3440
2532
+ },
2533
+ {
2534
+ "epoch": 1.65,
2535
+ "learning_rate": 6.929552117662211e-06,
2536
+ "loss": 0.3515,
2537
+ "step": 3450
2538
+ },
2539
+ {
2540
+ "epoch": 1.65,
2541
+ "learning_rate": 6.911158071511084e-06,
2542
+ "loss": 0.357,
2543
+ "step": 3460
2544
+ },
2545
+ {
2546
+ "epoch": 1.66,
2547
+ "learning_rate": 6.892733684881313e-06,
2548
+ "loss": 0.3267,
2549
+ "step": 3470
2550
+ },
2551
+ {
2552
+ "epoch": 1.66,
2553
+ "learning_rate": 6.874279250268174e-06,
2554
+ "loss": 0.3562,
2555
+ "step": 3480
2556
+ },
2557
+ {
2558
+ "epoch": 1.67,
2559
+ "learning_rate": 6.85579506064397e-06,
2560
+ "loss": 0.361,
2561
+ "step": 3490
2562
+ },
2563
+ {
2564
+ "epoch": 1.67,
2565
+ "learning_rate": 6.837281409453376e-06,
2566
+ "loss": 0.3452,
2567
+ "step": 3500
2568
+ },
2569
+ {
2570
+ "epoch": 1.67,
2571
+ "learning_rate": 6.818738590608782e-06,
2572
+ "loss": 0.348,
2573
+ "step": 3510
2574
+ },
2575
+ {
2576
+ "epoch": 1.68,
2577
+ "learning_rate": 6.800166898485632e-06,
2578
+ "loss": 0.3496,
2579
+ "step": 3520
2580
+ },
2581
+ {
2582
+ "epoch": 1.68,
2583
+ "learning_rate": 6.781566627917738e-06,
2584
+ "loss": 0.3548,
2585
+ "step": 3530
2586
+ },
2587
+ {
2588
+ "epoch": 1.69,
2589
+ "learning_rate": 6.762938074192615e-06,
2590
+ "loss": 0.3302,
2591
+ "step": 3540
2592
+ },
2593
+ {
2594
+ "epoch": 1.69,
2595
+ "learning_rate": 6.744281533046782e-06,
2596
+ "loss": 0.3447,
2597
+ "step": 3550
2598
+ },
2599
+ {
2600
+ "epoch": 1.7,
2601
+ "learning_rate": 6.725597300661071e-06,
2602
+ "loss": 0.3615,
2603
+ "step": 3560
2604
+ },
2605
+ {
2606
+ "epoch": 1.7,
2607
+ "learning_rate": 6.706885673655924e-06,
2608
+ "loss": 0.3263,
2609
+ "step": 3570
2610
+ },
2611
+ {
2612
+ "epoch": 1.71,
2613
+ "learning_rate": 6.688146949086688e-06,
2614
+ "loss": 0.3258,
2615
+ "step": 3580
2616
+ },
2617
+ {
2618
+ "epoch": 1.71,
2619
+ "learning_rate": 6.669381424438892e-06,
2620
+ "loss": 0.3483,
2621
+ "step": 3590
2622
+ },
2623
+ {
2624
+ "epoch": 1.72,
2625
+ "learning_rate": 6.650589397623528e-06,
2626
+ "loss": 0.3643,
2627
+ "step": 3600
2628
+ },
2629
+ {
2630
+ "epoch": 1.72,
2631
+ "eval_webgpt_accuracy": 0.5893769152196119,
2632
+ "eval_webgpt_loss": 0.6877950429916382,
2633
+ "eval_webgpt_runtime": 63.3493,
2634
+ "eval_webgpt_samples_per_second": 61.816,
2635
+ "eval_webgpt_steps_per_second": 6.188,
2636
+ "step": 3600
2637
+ },
2638
+ {
2639
+ "epoch": 1.72,
2640
+ "eval_hfsummary_accuracy": 0.6664147749599492,
2641
+ "eval_hfsummary_loss": 0.7294792532920837,
2642
+ "eval_hfsummary_runtime": 1242.1345,
2643
+ "eval_hfsummary_samples_per_second": 26.634,
2644
+ "eval_hfsummary_steps_per_second": 2.664,
2645
+ "step": 3600
2646
+ },
2647
+ {
2648
+ "epoch": 1.72,
2649
+ "eval_gptsynthetic_accuracy": 0.9993966817496229,
2650
+ "eval_gptsynthetic_loss": 0.005264561623334885,
2651
+ "eval_gptsynthetic_runtime": 45.9664,
2652
+ "eval_gptsynthetic_samples_per_second": 72.118,
2653
+ "eval_gptsynthetic_steps_per_second": 7.223,
2654
+ "step": 3600
2655
+ },
2656
+ {
2657
+ "epoch": 1.72,
2658
+ "learning_rate": 6.6317711669723265e-06,
2659
+ "loss": 0.3443,
2660
+ "step": 3610
2661
+ },
2662
+ {
2663
+ "epoch": 1.73,
2664
+ "learning_rate": 6.61292703123301e-06,
2665
+ "loss": 0.3539,
2666
+ "step": 3620
2667
+ },
2668
+ {
2669
+ "epoch": 1.73,
2670
+ "learning_rate": 6.594057289564557e-06,
2671
+ "loss": 0.3123,
2672
+ "step": 3630
2673
+ },
2674
+ {
2675
+ "epoch": 1.74,
2676
+ "learning_rate": 6.575162241532453e-06,
2677
+ "loss": 0.3047,
2678
+ "step": 3640
2679
+ },
2680
+ {
2681
+ "epoch": 1.74,
2682
+ "learning_rate": 6.556242187103934e-06,
2683
+ "loss": 0.3687,
2684
+ "step": 3650
2685
+ },
2686
+ {
2687
+ "epoch": 1.75,
2688
+ "learning_rate": 6.53729742664322e-06,
2689
+ "loss": 0.3652,
2690
+ "step": 3660
2691
+ },
2692
+ {
2693
+ "epoch": 1.75,
2694
+ "learning_rate": 6.518328260906753e-06,
2695
+ "loss": 0.3493,
2696
+ "step": 3670
2697
+ },
2698
+ {
2699
+ "epoch": 1.76,
2700
+ "learning_rate": 6.49933499103842e-06,
2701
+ "loss": 0.3637,
2702
+ "step": 3680
2703
+ },
2704
+ {
2705
+ "epoch": 1.76,
2706
+ "learning_rate": 6.480317918564767e-06,
2707
+ "loss": 0.3464,
2708
+ "step": 3690
2709
+ },
2710
+ {
2711
+ "epoch": 1.77,
2712
+ "learning_rate": 6.4612773453902244e-06,
2713
+ "loss": 0.3376,
2714
+ "step": 3700
2715
+ },
2716
+ {
2717
+ "epoch": 1.77,
2718
+ "learning_rate": 6.4422135737923e-06,
2719
+ "loss": 0.3473,
2720
+ "step": 3710
2721
+ },
2722
+ {
2723
+ "epoch": 1.77,
2724
+ "learning_rate": 6.423126906416789e-06,
2725
+ "loss": 0.3356,
2726
+ "step": 3720
2727
+ },
2728
+ {
2729
+ "epoch": 1.78,
2730
+ "learning_rate": 6.40401764627297e-06,
2731
+ "loss": 0.3374,
2732
+ "step": 3730
2733
+ },
2734
+ {
2735
+ "epoch": 1.78,
2736
+ "learning_rate": 6.384886096728789e-06,
2737
+ "loss": 0.3544,
2738
+ "step": 3740
2739
+ },
2740
+ {
2741
+ "epoch": 1.79,
2742
+ "learning_rate": 6.365732561506044e-06,
2743
+ "loss": 0.3361,
2744
+ "step": 3750
2745
+ },
2746
+ {
2747
+ "epoch": 1.79,
2748
+ "learning_rate": 6.346557344675568e-06,
2749
+ "loss": 0.3689,
2750
+ "step": 3760
2751
+ },
2752
+ {
2753
+ "epoch": 1.8,
2754
+ "learning_rate": 6.327360750652402e-06,
2755
+ "loss": 0.359,
2756
+ "step": 3770
2757
+ },
2758
+ {
2759
+ "epoch": 1.8,
2760
+ "learning_rate": 6.308143084190954e-06,
2761
+ "loss": 0.3423,
2762
+ "step": 3780
2763
+ },
2764
+ {
2765
+ "epoch": 1.81,
2766
+ "learning_rate": 6.2889046503801685e-06,
2767
+ "loss": 0.3392,
2768
+ "step": 3790
2769
+ },
2770
+ {
2771
+ "epoch": 1.81,
2772
+ "learning_rate": 6.269645754638682e-06,
2773
+ "loss": 0.342,
2774
+ "step": 3800
2775
+ },
2776
+ {
2777
+ "epoch": 1.81,
2778
+ "eval_webgpt_accuracy": 0.5842696629213483,
2779
+ "eval_webgpt_loss": 0.6836753487586975,
2780
+ "eval_webgpt_runtime": 63.4162,
2781
+ "eval_webgpt_samples_per_second": 61.751,
2782
+ "eval_webgpt_steps_per_second": 6.181,
2783
+ "step": 3800
2784
+ },
2785
+ {
2786
+ "epoch": 1.81,
2787
+ "eval_hfsummary_accuracy": 0.6713720037481486,
2788
+ "eval_hfsummary_loss": 0.7460022568702698,
2789
+ "eval_hfsummary_runtime": 1243.8848,
2790
+ "eval_hfsummary_samples_per_second": 26.597,
2791
+ "eval_hfsummary_steps_per_second": 2.66,
2792
+ "step": 3800
2793
+ },
2794
+ {
2795
+ "epoch": 1.81,
2796
+ "eval_gptsynthetic_accuracy": 0.9987933634992459,
2797
+ "eval_gptsynthetic_loss": 0.005401162896305323,
2798
+ "eval_gptsynthetic_runtime": 46.0774,
2799
+ "eval_gptsynthetic_samples_per_second": 71.944,
2800
+ "eval_gptsynthetic_steps_per_second": 7.205,
2801
+ "step": 3800
2802
+ },
2803
+ {
2804
+ "epoch": 1.82,
2805
+ "learning_rate": 6.250366702709971e-06,
2806
+ "loss": 0.3678,
2807
+ "step": 3810
2808
+ },
2809
+ {
2810
+ "epoch": 1.82,
2811
+ "learning_rate": 6.231067800657502e-06,
2812
+ "loss": 0.3446,
2813
+ "step": 3820
2814
+ },
2815
+ {
2816
+ "epoch": 1.83,
2817
+ "learning_rate": 6.211749354859871e-06,
2818
+ "loss": 0.3247,
2819
+ "step": 3830
2820
+ },
2821
+ {
2822
+ "epoch": 1.83,
2823
+ "learning_rate": 6.192411672005937e-06,
2824
+ "loss": 0.3639,
2825
+ "step": 3840
2826
+ },
2827
+ {
2828
+ "epoch": 1.84,
2829
+ "learning_rate": 6.173055059089958e-06,
2830
+ "loss": 0.3581,
2831
+ "step": 3850
2832
+ },
2833
+ {
2834
+ "epoch": 1.84,
2835
+ "learning_rate": 6.1536798234067175e-06,
2836
+ "loss": 0.3423,
2837
+ "step": 3860
2838
+ },
2839
+ {
2840
+ "epoch": 1.85,
2841
+ "learning_rate": 6.134286272546639e-06,
2842
+ "loss": 0.3623,
2843
+ "step": 3870
2844
+ },
2845
+ {
2846
+ "epoch": 1.85,
2847
+ "learning_rate": 6.11487471439091e-06,
2848
+ "loss": 0.321,
2849
+ "step": 3880
2850
+ },
2851
+ {
2852
+ "epoch": 1.86,
2853
+ "learning_rate": 6.0954454571065925e-06,
2854
+ "loss": 0.3363,
2855
+ "step": 3890
2856
+ },
2857
+ {
2858
+ "epoch": 1.86,
2859
+ "learning_rate": 6.0759988091417266e-06,
2860
+ "loss": 0.3392,
2861
+ "step": 3900
2862
+ },
2863
+ {
2864
+ "epoch": 1.87,
2865
+ "learning_rate": 6.0565350792204406e-06,
2866
+ "loss": 0.3458,
2867
+ "step": 3910
2868
+ },
2869
+ {
2870
+ "epoch": 1.87,
2871
+ "learning_rate": 6.037054576338044e-06,
2872
+ "loss": 0.3742,
2873
+ "step": 3920
2874
+ },
2875
+ {
2876
+ "epoch": 1.87,
2877
+ "learning_rate": 6.0175576097561265e-06,
2878
+ "loss": 0.3518,
2879
+ "step": 3930
2880
+ },
2881
+ {
2882
+ "epoch": 1.88,
2883
+ "learning_rate": 5.998044488997645e-06,
2884
+ "loss": 0.3489,
2885
+ "step": 3940
2886
+ },
2887
+ {
2888
+ "epoch": 1.88,
2889
+ "learning_rate": 5.9785155238420135e-06,
2890
+ "loss": 0.3411,
2891
+ "step": 3950
2892
+ },
2893
+ {
2894
+ "epoch": 1.89,
2895
+ "learning_rate": 5.958971024320179e-06,
2896
+ "loss": 0.3326,
2897
+ "step": 3960
2898
+ },
2899
+ {
2900
+ "epoch": 1.89,
2901
+ "learning_rate": 5.9394113007097045e-06,
2902
+ "loss": 0.322,
2903
+ "step": 3970
2904
+ },
2905
+ {
2906
+ "epoch": 1.9,
2907
+ "learning_rate": 5.919836663529845e-06,
2908
+ "loss": 0.3343,
2909
+ "step": 3980
2910
+ },
2911
+ {
2912
+ "epoch": 1.9,
2913
+ "learning_rate": 5.900247423536611e-06,
2914
+ "loss": 0.3551,
2915
+ "step": 3990
2916
+ },
2917
+ {
2918
+ "epoch": 1.91,
2919
+ "learning_rate": 5.880643891717842e-06,
2920
+ "loss": 0.3344,
2921
+ "step": 4000
2922
+ },
2923
+ {
2924
+ "epoch": 1.91,
2925
+ "eval_webgpt_accuracy": 0.5906537282941777,
2926
+ "eval_webgpt_loss": 0.6790869832038879,
2927
+ "eval_webgpt_runtime": 63.367,
2928
+ "eval_webgpt_samples_per_second": 61.799,
2929
+ "eval_webgpt_steps_per_second": 6.186,
2930
+ "step": 4000
2931
+ },
2932
+ {
2933
+ "epoch": 1.91,
2934
+ "eval_hfsummary_accuracy": 0.6683795302723453,
2935
+ "eval_hfsummary_loss": 0.7165785431861877,
2936
+ "eval_hfsummary_runtime": 1243.3678,
2937
+ "eval_hfsummary_samples_per_second": 26.608,
2938
+ "eval_hfsummary_steps_per_second": 2.661,
2939
+ "step": 4000
2940
+ },
2941
+ {
2942
+ "epoch": 1.91,
2943
+ "eval_gptsynthetic_accuracy": 0.9984917043740573,
2944
+ "eval_gptsynthetic_loss": 0.005581920500844717,
2945
+ "eval_gptsynthetic_runtime": 46.0581,
2946
+ "eval_gptsynthetic_samples_per_second": 71.974,
2947
+ "eval_gptsynthetic_steps_per_second": 7.208,
2948
+ "step": 4000
2949
+ }
2950
+ ],
2951
+ "max_steps": 4192,
2952
+ "num_train_epochs": 2,
2953
+ "total_flos": 2.4953197931002634e+17,
2954
+ "trial_name": null,
2955
+ "trial_params": null
2956
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30ca00b0e391e47b7926352fa66f9123794a61f02fe1fcfa905a2af50c701baf
3
+ size 3439