yalhessi commited on
Commit
f0d39c1
·
verified ·
1 Parent(s): 72df6c4

End of training

Browse files
README.md ADDED
@@ -0,0 +1,89 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ license: other
4
+ base_model: deepseek-ai/deepseek-coder-1.3b-base
5
+ tags:
6
+ - generated_from_trainer
7
+ model-index:
8
+ - name: lemexp-processed-task1_min_symbols_template_small-deepseek-coder-1.3b-base
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # lemexp-processed-task1_min_symbols_template_small-deepseek-coder-1.3b-base
16
+
17
+ This model is a fine-tuned version of [deepseek-ai/deepseek-coder-1.3b-base](https://huggingface.co/deepseek-ai/deepseek-coder-1.3b-base) on an unknown dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 0.1407
20
+
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ ### Training hyperparameters
36
+
37
+ The following hyperparameters were used during training:
38
+ - learning_rate: 0.0002
39
+ - train_batch_size: 2
40
+ - eval_batch_size: 2
41
+ - seed: 42
42
+ - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
43
+ - lr_scheduler_type: linear
44
+ - lr_scheduler_warmup_steps: 100
45
+ - num_epochs: 6
46
+ - mixed_precision_training: Native AMP
47
+
48
+ ### Training results
49
+
50
+ | Training Loss | Epoch | Step | Validation Loss |
51
+ |:-------------:|:------:|:------:|:---------------:|
52
+ | 0.3044 | 0.2000 | 5030 | 0.2924 |
53
+ | 0.2643 | 0.4001 | 10060 | 0.2585 |
54
+ | 0.2396 | 0.6001 | 15090 | 0.2310 |
55
+ | 0.2277 | 0.8001 | 20120 | 0.2108 |
56
+ | 0.2103 | 1.0002 | 25150 | 0.2088 |
57
+ | 0.1942 | 1.2002 | 30180 | 0.2031 |
58
+ | 0.1846 | 1.4002 | 35210 | 0.1955 |
59
+ | 0.1814 | 1.6003 | 40240 | 0.1827 |
60
+ | 0.1688 | 1.8003 | 45270 | 0.1789 |
61
+ | 0.1772 | 2.0003 | 50300 | 0.1767 |
62
+ | 0.143 | 2.2003 | 55330 | 0.1748 |
63
+ | 0.1465 | 2.4004 | 60360 | 0.1691 |
64
+ | 0.1517 | 2.6004 | 65390 | 0.1702 |
65
+ | 0.1533 | 2.8004 | 70420 | 0.1651 |
66
+ | 0.1412 | 3.0005 | 75450 | 0.1569 |
67
+ | 0.1237 | 3.2005 | 80480 | 0.1581 |
68
+ | 0.131 | 3.4005 | 85510 | 0.1561 |
69
+ | 0.1221 | 3.6006 | 90540 | 0.1486 |
70
+ | 0.1161 | 3.8006 | 95570 | 0.1464 |
71
+ | 0.1165 | 4.0006 | 100600 | 0.1413 |
72
+ | 0.0994 | 4.2007 | 105630 | 0.1458 |
73
+ | 0.1113 | 4.4007 | 110660 | 0.1483 |
74
+ | 0.0996 | 4.6007 | 115690 | 0.1426 |
75
+ | 0.1116 | 4.8008 | 120720 | 0.1392 |
76
+ | 0.1004 | 5.0008 | 125750 | 0.1380 |
77
+ | 0.0814 | 5.2008 | 130780 | 0.1418 |
78
+ | 0.086 | 5.4009 | 135810 | 0.1406 |
79
+ | 0.0731 | 5.6009 | 140840 | 0.1416 |
80
+ | 0.0807 | 5.8009 | 145870 | 0.1407 |
81
+
82
+
83
+ ### Framework versions
84
+
85
+ - PEFT 0.14.0
86
+ - Transformers 4.47.0
87
+ - Pytorch 2.5.1+cu124
88
+ - Datasets 3.2.0
89
+ - Tokenizers 0.21.0
adapter_config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "deepseek-ai/deepseek-coder-1.3b-base",
5
+ "bias": "none",
6
+ "eva_config": null,
7
+ "exclude_modules": null,
8
+ "fan_in_fan_out": false,
9
+ "inference_mode": true,
10
+ "init_lora_weights": true,
11
+ "layer_replication": null,
12
+ "layers_pattern": null,
13
+ "layers_to_transform": null,
14
+ "loftq_config": {},
15
+ "lora_alpha": 32,
16
+ "lora_bias": false,
17
+ "lora_dropout": 0.05,
18
+ "megatron_config": null,
19
+ "megatron_core": "megatron.core",
20
+ "modules_to_save": null,
21
+ "peft_type": "LORA",
22
+ "r": 8,
23
+ "rank_pattern": {},
24
+ "revision": null,
25
+ "target_modules": [
26
+ "q_proj",
27
+ "v_proj"
28
+ ],
29
+ "task_type": "CAUSAL_LM",
30
+ "use_dora": false,
31
+ "use_rslora": false
32
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38baa1f41db86fa4aee1dd13df4734ab711ee7225bcf68937c9ad7caa9a64115
3
+ size 531035104
trainer_state.json ADDED
@@ -0,0 +1,2381 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 6.0,
5
+ "eval_steps": 5030,
6
+ "global_step": 150876,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.019883878151594685,
13
+ "grad_norm": 2.3147895336151123,
14
+ "learning_rate": 0.0001994694115774394,
15
+ "loss": 0.594,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.03976775630318937,
20
+ "grad_norm": 2.2017321586608887,
21
+ "learning_rate": 0.0001988061760492386,
22
+ "loss": 0.4251,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.05965163445478406,
27
+ "grad_norm": 2.0374693870544434,
28
+ "learning_rate": 0.00019814294052103784,
29
+ "loss": 0.386,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.07953551260637874,
34
+ "grad_norm": 1.3447909355163574,
35
+ "learning_rate": 0.00019747970499283706,
36
+ "loss": 0.3585,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.09941939075797343,
41
+ "grad_norm": 1.2458261251449585,
42
+ "learning_rate": 0.00019682044887780548,
43
+ "loss": 0.3545,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.11930326890956812,
48
+ "grad_norm": 1.844347357749939,
49
+ "learning_rate": 0.00019615721334960472,
50
+ "loss": 0.3329,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.1391871470611628,
55
+ "grad_norm": 1.3815340995788574,
56
+ "learning_rate": 0.00019549397782140396,
57
+ "loss": 0.3182,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 0.15907102521275748,
62
+ "grad_norm": 2.2595863342285156,
63
+ "learning_rate": 0.00019483074229320317,
64
+ "loss": 0.2985,
65
+ "step": 4000
66
+ },
67
+ {
68
+ "epoch": 0.17895490336435219,
69
+ "grad_norm": 1.8043708801269531,
70
+ "learning_rate": 0.00019416750676500238,
71
+ "loss": 0.3068,
72
+ "step": 4500
73
+ },
74
+ {
75
+ "epoch": 0.19883878151594686,
76
+ "grad_norm": 1.8568594455718994,
77
+ "learning_rate": 0.00019350427123680162,
78
+ "loss": 0.3044,
79
+ "step": 5000
80
+ },
81
+ {
82
+ "epoch": 0.20003181420504254,
83
+ "eval_loss": 0.29235559701919556,
84
+ "eval_runtime": 31.6112,
85
+ "eval_samples_per_second": 15.817,
86
+ "eval_steps_per_second": 7.909,
87
+ "step": 5030
88
+ },
89
+ {
90
+ "epoch": 0.21872265966754156,
91
+ "grad_norm": 1.780515193939209,
92
+ "learning_rate": 0.00019284103570860084,
93
+ "loss": 0.2884,
94
+ "step": 5500
95
+ },
96
+ {
97
+ "epoch": 0.23860653781913624,
98
+ "grad_norm": 1.3652188777923584,
99
+ "learning_rate": 0.00019217780018040008,
100
+ "loss": 0.2924,
101
+ "step": 6000
102
+ },
103
+ {
104
+ "epoch": 0.2584904159707309,
105
+ "grad_norm": 1.5280920267105103,
106
+ "learning_rate": 0.0001915158911232557,
107
+ "loss": 0.2714,
108
+ "step": 6500
109
+ },
110
+ {
111
+ "epoch": 0.2783742941223256,
112
+ "grad_norm": 1.8563861846923828,
113
+ "learning_rate": 0.00019085398206611135,
114
+ "loss": 0.2729,
115
+ "step": 7000
116
+ },
117
+ {
118
+ "epoch": 0.2982581722739203,
119
+ "grad_norm": 1.0958434343338013,
120
+ "learning_rate": 0.00019019074653791056,
121
+ "loss": 0.2735,
122
+ "step": 7500
123
+ },
124
+ {
125
+ "epoch": 0.31814205042551497,
126
+ "grad_norm": 1.145859956741333,
127
+ "learning_rate": 0.00018952751100970977,
128
+ "loss": 0.2737,
129
+ "step": 8000
130
+ },
131
+ {
132
+ "epoch": 0.33802592857710967,
133
+ "grad_norm": 1.2784571647644043,
134
+ "learning_rate": 0.000188864275481509,
135
+ "loss": 0.2611,
136
+ "step": 8500
137
+ },
138
+ {
139
+ "epoch": 0.35790980672870437,
140
+ "grad_norm": 1.3518733978271484,
141
+ "learning_rate": 0.00018820103995330822,
142
+ "loss": 0.2612,
143
+ "step": 9000
144
+ },
145
+ {
146
+ "epoch": 0.3777936848802991,
147
+ "grad_norm": 1.4692599773406982,
148
+ "learning_rate": 0.00018753780442510743,
149
+ "loss": 0.2609,
150
+ "step": 9500
151
+ },
152
+ {
153
+ "epoch": 0.3976775630318937,
154
+ "grad_norm": 2.682018518447876,
155
+ "learning_rate": 0.00018687456889690667,
156
+ "loss": 0.2643,
157
+ "step": 10000
158
+ },
159
+ {
160
+ "epoch": 0.4000636284100851,
161
+ "eval_loss": 0.2585141062736511,
162
+ "eval_runtime": 31.9094,
163
+ "eval_samples_per_second": 15.669,
164
+ "eval_steps_per_second": 7.835,
165
+ "step": 10060
166
+ },
167
+ {
168
+ "epoch": 0.4175614411834884,
169
+ "grad_norm": 1.7332996129989624,
170
+ "learning_rate": 0.00018621133336870591,
171
+ "loss": 0.2523,
172
+ "step": 10500
173
+ },
174
+ {
175
+ "epoch": 0.4374453193350831,
176
+ "grad_norm": 2.1415417194366455,
177
+ "learning_rate": 0.00018554942431156152,
178
+ "loss": 0.2585,
179
+ "step": 11000
180
+ },
181
+ {
182
+ "epoch": 0.4573291974866778,
183
+ "grad_norm": 1.2841447591781616,
184
+ "learning_rate": 0.00018488618878336073,
185
+ "loss": 0.2557,
186
+ "step": 11500
187
+ },
188
+ {
189
+ "epoch": 0.4772130756382725,
190
+ "grad_norm": 3.2847816944122314,
191
+ "learning_rate": 0.00018422295325516,
192
+ "loss": 0.2609,
193
+ "step": 12000
194
+ },
195
+ {
196
+ "epoch": 0.4970969537898672,
197
+ "grad_norm": 0.7331926822662354,
198
+ "learning_rate": 0.00018355971772695921,
199
+ "loss": 0.2477,
200
+ "step": 12500
201
+ },
202
+ {
203
+ "epoch": 0.5169808319414618,
204
+ "grad_norm": 1.4415650367736816,
205
+ "learning_rate": 0.00018289648219875843,
206
+ "loss": 0.2482,
207
+ "step": 13000
208
+ },
209
+ {
210
+ "epoch": 0.5368647100930566,
211
+ "grad_norm": 1.7630778551101685,
212
+ "learning_rate": 0.00018223457314161406,
213
+ "loss": 0.2421,
214
+ "step": 13500
215
+ },
216
+ {
217
+ "epoch": 0.5567485882446512,
218
+ "grad_norm": 2.0509490966796875,
219
+ "learning_rate": 0.0001815713376134133,
220
+ "loss": 0.2444,
221
+ "step": 14000
222
+ },
223
+ {
224
+ "epoch": 0.5766324663962459,
225
+ "grad_norm": 1.959215760231018,
226
+ "learning_rate": 0.0001809081020852125,
227
+ "loss": 0.2398,
228
+ "step": 14500
229
+ },
230
+ {
231
+ "epoch": 0.5965163445478406,
232
+ "grad_norm": 1.6416336297988892,
233
+ "learning_rate": 0.00018024486655701173,
234
+ "loss": 0.2396,
235
+ "step": 15000
236
+ },
237
+ {
238
+ "epoch": 0.6000954426151277,
239
+ "eval_loss": 0.23102714121341705,
240
+ "eval_runtime": 32.4432,
241
+ "eval_samples_per_second": 15.412,
242
+ "eval_steps_per_second": 7.706,
243
+ "step": 15090
244
+ },
245
+ {
246
+ "epoch": 0.6164002226994353,
247
+ "grad_norm": 0.5961350798606873,
248
+ "learning_rate": 0.00017958163102881097,
249
+ "loss": 0.2291,
250
+ "step": 15500
251
+ },
252
+ {
253
+ "epoch": 0.6362841008510299,
254
+ "grad_norm": 1.3288564682006836,
255
+ "learning_rate": 0.00017891972197166657,
256
+ "loss": 0.2368,
257
+ "step": 16000
258
+ },
259
+ {
260
+ "epoch": 0.6561679790026247,
261
+ "grad_norm": 1.7699204683303833,
262
+ "learning_rate": 0.0001782564864434658,
263
+ "loss": 0.2397,
264
+ "step": 16500
265
+ },
266
+ {
267
+ "epoch": 0.6760518571542193,
268
+ "grad_norm": 1.3349616527557373,
269
+ "learning_rate": 0.00017759325091526505,
270
+ "loss": 0.2353,
271
+ "step": 17000
272
+ },
273
+ {
274
+ "epoch": 0.6959357353058141,
275
+ "grad_norm": 1.6228617429733276,
276
+ "learning_rate": 0.00017693001538706426,
277
+ "loss": 0.2314,
278
+ "step": 17500
279
+ },
280
+ {
281
+ "epoch": 0.7158196134574087,
282
+ "grad_norm": 1.349108338356018,
283
+ "learning_rate": 0.00017626677985886348,
284
+ "loss": 0.2339,
285
+ "step": 18000
286
+ },
287
+ {
288
+ "epoch": 0.7357034916090034,
289
+ "grad_norm": 2.2128469944000244,
290
+ "learning_rate": 0.00017560354433066272,
291
+ "loss": 0.2252,
292
+ "step": 18500
293
+ },
294
+ {
295
+ "epoch": 0.7555873697605981,
296
+ "grad_norm": 2.038177013397217,
297
+ "learning_rate": 0.00017494163527351835,
298
+ "loss": 0.2242,
299
+ "step": 19000
300
+ },
301
+ {
302
+ "epoch": 0.7754712479121928,
303
+ "grad_norm": 1.1377464532852173,
304
+ "learning_rate": 0.00017427839974531756,
305
+ "loss": 0.2197,
306
+ "step": 19500
307
+ },
308
+ {
309
+ "epoch": 0.7953551260637874,
310
+ "grad_norm": 1.2686738967895508,
311
+ "learning_rate": 0.00017361516421711678,
312
+ "loss": 0.2277,
313
+ "step": 20000
314
+ },
315
+ {
316
+ "epoch": 0.8001272568201702,
317
+ "eval_loss": 0.21075737476348877,
318
+ "eval_runtime": 32.819,
319
+ "eval_samples_per_second": 15.235,
320
+ "eval_steps_per_second": 7.618,
321
+ "step": 20120
322
+ },
323
+ {
324
+ "epoch": 0.8152390042153822,
325
+ "grad_norm": 2.1816446781158447,
326
+ "learning_rate": 0.00017295192868891602,
327
+ "loss": 0.2245,
328
+ "step": 20500
329
+ },
330
+ {
331
+ "epoch": 0.8351228823669768,
332
+ "grad_norm": 2.224292755126953,
333
+ "learning_rate": 0.00017229001963177165,
334
+ "loss": 0.2176,
335
+ "step": 21000
336
+ },
337
+ {
338
+ "epoch": 0.8550067605185715,
339
+ "grad_norm": 0.7720803618431091,
340
+ "learning_rate": 0.00017162811057462726,
341
+ "loss": 0.2097,
342
+ "step": 21500
343
+ },
344
+ {
345
+ "epoch": 0.8748906386701663,
346
+ "grad_norm": 1.716422200202942,
347
+ "learning_rate": 0.0001709648750464265,
348
+ "loss": 0.2104,
349
+ "step": 22000
350
+ },
351
+ {
352
+ "epoch": 0.8947745168217609,
353
+ "grad_norm": 1.5487083196640015,
354
+ "learning_rate": 0.00017030163951822574,
355
+ "loss": 0.2179,
356
+ "step": 22500
357
+ },
358
+ {
359
+ "epoch": 0.9146583949733555,
360
+ "grad_norm": 1.0968928337097168,
361
+ "learning_rate": 0.00016963840399002495,
362
+ "loss": 0.2194,
363
+ "step": 23000
364
+ },
365
+ {
366
+ "epoch": 0.9345422731249503,
367
+ "grad_norm": 0.36388519406318665,
368
+ "learning_rate": 0.00016897516846182416,
369
+ "loss": 0.2259,
370
+ "step": 23500
371
+ },
372
+ {
373
+ "epoch": 0.954426151276545,
374
+ "grad_norm": 1.2857156991958618,
375
+ "learning_rate": 0.0001683119329336234,
376
+ "loss": 0.2195,
377
+ "step": 24000
378
+ },
379
+ {
380
+ "epoch": 0.9743100294281397,
381
+ "grad_norm": 1.7136821746826172,
382
+ "learning_rate": 0.00016764869740542261,
383
+ "loss": 0.2036,
384
+ "step": 24500
385
+ },
386
+ {
387
+ "epoch": 0.9941939075797344,
388
+ "grad_norm": 2.1068880558013916,
389
+ "learning_rate": 0.00016698546187722183,
390
+ "loss": 0.2103,
391
+ "step": 25000
392
+ },
393
+ {
394
+ "epoch": 1.0001590710252128,
395
+ "eval_loss": 0.2087584286928177,
396
+ "eval_runtime": 34.3797,
397
+ "eval_samples_per_second": 14.543,
398
+ "eval_steps_per_second": 7.272,
399
+ "step": 25150
400
+ },
401
+ {
402
+ "epoch": 1.014077785731329,
403
+ "grad_norm": 0.9411464929580688,
404
+ "learning_rate": 0.0001663222263490211,
405
+ "loss": 0.1902,
406
+ "step": 25500
407
+ },
408
+ {
409
+ "epoch": 1.0339616638829237,
410
+ "grad_norm": 1.8499776124954224,
411
+ "learning_rate": 0.0001656589908208203,
412
+ "loss": 0.1961,
413
+ "step": 26000
414
+ },
415
+ {
416
+ "epoch": 1.0538455420345185,
417
+ "grad_norm": 2.804515838623047,
418
+ "learning_rate": 0.00016499708176367591,
419
+ "loss": 0.1975,
420
+ "step": 26500
421
+ },
422
+ {
423
+ "epoch": 1.0737294201861132,
424
+ "grad_norm": 1.0124012231826782,
425
+ "learning_rate": 0.00016433384623547515,
426
+ "loss": 0.1965,
427
+ "step": 27000
428
+ },
429
+ {
430
+ "epoch": 1.0936132983377078,
431
+ "grad_norm": 2.1755409240722656,
432
+ "learning_rate": 0.0001636706107072744,
433
+ "loss": 0.2032,
434
+ "step": 27500
435
+ },
436
+ {
437
+ "epoch": 1.1134971764893025,
438
+ "grad_norm": 1.9173059463500977,
439
+ "learning_rate": 0.0001630073751790736,
440
+ "loss": 0.1998,
441
+ "step": 28000
442
+ },
443
+ {
444
+ "epoch": 1.1333810546408971,
445
+ "grad_norm": 2.6213159561157227,
446
+ "learning_rate": 0.0001623454661219292,
447
+ "loss": 0.1944,
448
+ "step": 28500
449
+ },
450
+ {
451
+ "epoch": 1.1532649327924918,
452
+ "grad_norm": 1.979064702987671,
453
+ "learning_rate": 0.00016168223059372845,
454
+ "loss": 0.1947,
455
+ "step": 29000
456
+ },
457
+ {
458
+ "epoch": 1.1731488109440866,
459
+ "grad_norm": 3.864307403564453,
460
+ "learning_rate": 0.00016101899506552767,
461
+ "loss": 0.1948,
462
+ "step": 29500
463
+ },
464
+ {
465
+ "epoch": 1.1930326890956813,
466
+ "grad_norm": 1.3037413358688354,
467
+ "learning_rate": 0.0001603557595373269,
468
+ "loss": 0.1942,
469
+ "step": 30000
470
+ },
471
+ {
472
+ "epoch": 1.2001908852302554,
473
+ "eval_loss": 0.20310936868190765,
474
+ "eval_runtime": 33.1759,
475
+ "eval_samples_per_second": 15.071,
476
+ "eval_steps_per_second": 7.536,
477
+ "step": 30180
478
+ },
479
+ {
480
+ "epoch": 1.212916567247276,
481
+ "grad_norm": 1.4742060899734497,
482
+ "learning_rate": 0.00015969385048018254,
483
+ "loss": 0.1873,
484
+ "step": 30500
485
+ },
486
+ {
487
+ "epoch": 1.2328004453988706,
488
+ "grad_norm": 0.8921416401863098,
489
+ "learning_rate": 0.00015903061495198175,
490
+ "loss": 0.1888,
491
+ "step": 31000
492
+ },
493
+ {
494
+ "epoch": 1.2526843235504652,
495
+ "grad_norm": 2.814047336578369,
496
+ "learning_rate": 0.00015836737942378096,
497
+ "loss": 0.1933,
498
+ "step": 31500
499
+ },
500
+ {
501
+ "epoch": 1.27256820170206,
502
+ "grad_norm": 1.6011271476745605,
503
+ "learning_rate": 0.0001577041438955802,
504
+ "loss": 0.1943,
505
+ "step": 32000
506
+ },
507
+ {
508
+ "epoch": 1.2924520798536547,
509
+ "grad_norm": 2.1085023880004883,
510
+ "learning_rate": 0.00015704090836737944,
511
+ "loss": 0.1939,
512
+ "step": 32500
513
+ },
514
+ {
515
+ "epoch": 1.3123359580052494,
516
+ "grad_norm": 1.4254250526428223,
517
+ "learning_rate": 0.00015637899931023505,
518
+ "loss": 0.1898,
519
+ "step": 33000
520
+ },
521
+ {
522
+ "epoch": 1.332219836156844,
523
+ "grad_norm": 2.6085307598114014,
524
+ "learning_rate": 0.00015571576378203426,
525
+ "loss": 0.1869,
526
+ "step": 33500
527
+ },
528
+ {
529
+ "epoch": 1.3521037143084387,
530
+ "grad_norm": 1.4234007596969604,
531
+ "learning_rate": 0.00015505252825383353,
532
+ "loss": 0.1924,
533
+ "step": 34000
534
+ },
535
+ {
536
+ "epoch": 1.3719875924600333,
537
+ "grad_norm": 1.5566816329956055,
538
+ "learning_rate": 0.00015439061919668914,
539
+ "loss": 0.1932,
540
+ "step": 34500
541
+ },
542
+ {
543
+ "epoch": 1.391871470611628,
544
+ "grad_norm": 8.811753273010254,
545
+ "learning_rate": 0.00015372738366848835,
546
+ "loss": 0.1846,
547
+ "step": 35000
548
+ },
549
+ {
550
+ "epoch": 1.400222699435298,
551
+ "eval_loss": 0.19551397860050201,
552
+ "eval_runtime": 35.3626,
553
+ "eval_samples_per_second": 14.139,
554
+ "eval_steps_per_second": 7.07,
555
+ "step": 35210
556
+ },
557
+ {
558
+ "epoch": 1.4117553487632228,
559
+ "grad_norm": 1.463742971420288,
560
+ "learning_rate": 0.0001530641481402876,
561
+ "loss": 0.1924,
562
+ "step": 35500
563
+ },
564
+ {
565
+ "epoch": 1.4316392269148175,
566
+ "grad_norm": 2.8610801696777344,
567
+ "learning_rate": 0.00015240091261208683,
568
+ "loss": 0.1938,
569
+ "step": 36000
570
+ },
571
+ {
572
+ "epoch": 1.4515231050664121,
573
+ "grad_norm": 1.2369511127471924,
574
+ "learning_rate": 0.00015173767708388604,
575
+ "loss": 0.1817,
576
+ "step": 36500
577
+ },
578
+ {
579
+ "epoch": 1.4714069832180068,
580
+ "grad_norm": 1.4233882427215576,
581
+ "learning_rate": 0.00015107444155568526,
582
+ "loss": 0.1817,
583
+ "step": 37000
584
+ },
585
+ {
586
+ "epoch": 1.4912908613696017,
587
+ "grad_norm": 1.5385404825210571,
588
+ "learning_rate": 0.0001504112060274845,
589
+ "loss": 0.1849,
590
+ "step": 37500
591
+ },
592
+ {
593
+ "epoch": 1.5111747395211963,
594
+ "grad_norm": 1.2022610902786255,
595
+ "learning_rate": 0.0001497479704992837,
596
+ "loss": 0.183,
597
+ "step": 38000
598
+ },
599
+ {
600
+ "epoch": 1.531058617672791,
601
+ "grad_norm": 1.5988807678222656,
602
+ "learning_rate": 0.00014908606144213934,
603
+ "loss": 0.1821,
604
+ "step": 38500
605
+ },
606
+ {
607
+ "epoch": 1.5509424958243856,
608
+ "grad_norm": 1.480351209640503,
609
+ "learning_rate": 0.00014842282591393855,
610
+ "loss": 0.1787,
611
+ "step": 39000
612
+ },
613
+ {
614
+ "epoch": 1.5708263739759802,
615
+ "grad_norm": 1.9051077365875244,
616
+ "learning_rate": 0.0001477595903857378,
617
+ "loss": 0.1826,
618
+ "step": 39500
619
+ },
620
+ {
621
+ "epoch": 1.5907102521275749,
622
+ "grad_norm": 2.0775561332702637,
623
+ "learning_rate": 0.000147096354857537,
624
+ "loss": 0.1814,
625
+ "step": 40000
626
+ },
627
+ {
628
+ "epoch": 1.6002545136403405,
629
+ "eval_loss": 0.18265177309513092,
630
+ "eval_runtime": 31.3228,
631
+ "eval_samples_per_second": 15.963,
632
+ "eval_steps_per_second": 7.981,
633
+ "step": 40240
634
+ },
635
+ {
636
+ "epoch": 1.6105941302791695,
637
+ "grad_norm": 2.4270079135894775,
638
+ "learning_rate": 0.00014643444580039264,
639
+ "loss": 0.1728,
640
+ "step": 40500
641
+ },
642
+ {
643
+ "epoch": 1.6304780084307642,
644
+ "grad_norm": 1.4693158864974976,
645
+ "learning_rate": 0.00014577121027219188,
646
+ "loss": 0.1835,
647
+ "step": 41000
648
+ },
649
+ {
650
+ "epoch": 1.650361886582359,
651
+ "grad_norm": 2.014503240585327,
652
+ "learning_rate": 0.0001451079747439911,
653
+ "loss": 0.1825,
654
+ "step": 41500
655
+ },
656
+ {
657
+ "epoch": 1.6702457647339537,
658
+ "grad_norm": 2.4731264114379883,
659
+ "learning_rate": 0.0001444447392157903,
660
+ "loss": 0.1803,
661
+ "step": 42000
662
+ },
663
+ {
664
+ "epoch": 1.6901296428855483,
665
+ "grad_norm": 1.043967366218567,
666
+ "learning_rate": 0.00014378150368758955,
667
+ "loss": 0.1831,
668
+ "step": 42500
669
+ },
670
+ {
671
+ "epoch": 1.7100135210371432,
672
+ "grad_norm": 2.0062990188598633,
673
+ "learning_rate": 0.00014311826815938879,
674
+ "loss": 0.1851,
675
+ "step": 43000
676
+ },
677
+ {
678
+ "epoch": 1.7298973991887379,
679
+ "grad_norm": 2.958763837814331,
680
+ "learning_rate": 0.000142455032631188,
681
+ "loss": 0.1825,
682
+ "step": 43500
683
+ },
684
+ {
685
+ "epoch": 1.7497812773403325,
686
+ "grad_norm": 1.6624678373336792,
687
+ "learning_rate": 0.0001417917971029872,
688
+ "loss": 0.1834,
689
+ "step": 44000
690
+ },
691
+ {
692
+ "epoch": 1.7696651554919272,
693
+ "grad_norm": 1.1731349229812622,
694
+ "learning_rate": 0.00014112856157478645,
695
+ "loss": 0.1775,
696
+ "step": 44500
697
+ },
698
+ {
699
+ "epoch": 1.7895490336435218,
700
+ "grad_norm": 1.2256383895874023,
701
+ "learning_rate": 0.00014046665251764206,
702
+ "loss": 0.1688,
703
+ "step": 45000
704
+ },
705
+ {
706
+ "epoch": 1.8002863278453831,
707
+ "eval_loss": 0.1788606494665146,
708
+ "eval_runtime": 31.0353,
709
+ "eval_samples_per_second": 16.111,
710
+ "eval_steps_per_second": 8.055,
711
+ "step": 45270
712
+ },
713
+ {
714
+ "epoch": 1.8094329117951164,
715
+ "grad_norm": 5.187135219573975,
716
+ "learning_rate": 0.0001398034169894413,
717
+ "loss": 0.1757,
718
+ "step": 45500
719
+ },
720
+ {
721
+ "epoch": 1.829316789946711,
722
+ "grad_norm": 3.0221340656280518,
723
+ "learning_rate": 0.00013914018146124054,
724
+ "loss": 0.169,
725
+ "step": 46000
726
+ },
727
+ {
728
+ "epoch": 1.8492006680983057,
729
+ "grad_norm": 1.780038833618164,
730
+ "learning_rate": 0.00013847694593303975,
731
+ "loss": 0.1815,
732
+ "step": 46500
733
+ },
734
+ {
735
+ "epoch": 1.8690845462499006,
736
+ "grad_norm": 1.5590816736221313,
737
+ "learning_rate": 0.00013781503687589536,
738
+ "loss": 0.1754,
739
+ "step": 47000
740
+ },
741
+ {
742
+ "epoch": 1.8889684244014953,
743
+ "grad_norm": 2.1302263736724854,
744
+ "learning_rate": 0.0001371518013476946,
745
+ "loss": 0.1737,
746
+ "step": 47500
747
+ },
748
+ {
749
+ "epoch": 1.90885230255309,
750
+ "grad_norm": 2.173957109451294,
751
+ "learning_rate": 0.00013648856581949384,
752
+ "loss": 0.1696,
753
+ "step": 48000
754
+ },
755
+ {
756
+ "epoch": 1.9287361807046848,
757
+ "grad_norm": 1.0864589214324951,
758
+ "learning_rate": 0.00013582533029129305,
759
+ "loss": 0.1832,
760
+ "step": 48500
761
+ },
762
+ {
763
+ "epoch": 1.9486200588562794,
764
+ "grad_norm": 1.9979732036590576,
765
+ "learning_rate": 0.00013516209476309226,
766
+ "loss": 0.1758,
767
+ "step": 49000
768
+ },
769
+ {
770
+ "epoch": 1.968503937007874,
771
+ "grad_norm": 2.2656006813049316,
772
+ "learning_rate": 0.00013450018570594792,
773
+ "loss": 0.1728,
774
+ "step": 49500
775
+ },
776
+ {
777
+ "epoch": 1.9883878151594687,
778
+ "grad_norm": 2.077143669128418,
779
+ "learning_rate": 0.00013383695017774714,
780
+ "loss": 0.1772,
781
+ "step": 50000
782
+ },
783
+ {
784
+ "epoch": 2.0003181420504257,
785
+ "eval_loss": 0.1767009198665619,
786
+ "eval_runtime": 32.0718,
787
+ "eval_samples_per_second": 15.59,
788
+ "eval_steps_per_second": 7.795,
789
+ "step": 50300
790
+ },
791
+ {
792
+ "epoch": 2.0082716933110634,
793
+ "grad_norm": 1.5150219202041626,
794
+ "learning_rate": 0.00013317371464954635,
795
+ "loss": 0.164,
796
+ "step": 50500
797
+ },
798
+ {
799
+ "epoch": 2.028155571462658,
800
+ "grad_norm": 1.2225052118301392,
801
+ "learning_rate": 0.0001325104791213456,
802
+ "loss": 0.1534,
803
+ "step": 51000
804
+ },
805
+ {
806
+ "epoch": 2.0480394496142527,
807
+ "grad_norm": 1.9311732053756714,
808
+ "learning_rate": 0.0001318472435931448,
809
+ "loss": 0.1584,
810
+ "step": 51500
811
+ },
812
+ {
813
+ "epoch": 2.0679233277658473,
814
+ "grad_norm": 2.399226188659668,
815
+ "learning_rate": 0.000131184008064944,
816
+ "loss": 0.153,
817
+ "step": 52000
818
+ },
819
+ {
820
+ "epoch": 2.087807205917442,
821
+ "grad_norm": 2.9771511554718018,
822
+ "learning_rate": 0.00013052077253674325,
823
+ "loss": 0.1528,
824
+ "step": 52500
825
+ },
826
+ {
827
+ "epoch": 2.107691084069037,
828
+ "grad_norm": 1.5730229616165161,
829
+ "learning_rate": 0.0001298588634795989,
830
+ "loss": 0.154,
831
+ "step": 53000
832
+ },
833
+ {
834
+ "epoch": 2.1275749622206317,
835
+ "grad_norm": 1.2603771686553955,
836
+ "learning_rate": 0.0001291956279513981,
837
+ "loss": 0.1575,
838
+ "step": 53500
839
+ },
840
+ {
841
+ "epoch": 2.1474588403722263,
842
+ "grad_norm": 2.315129280090332,
843
+ "learning_rate": 0.00012853371889425373,
844
+ "loss": 0.1603,
845
+ "step": 54000
846
+ },
847
+ {
848
+ "epoch": 2.167342718523821,
849
+ "grad_norm": 1.5500929355621338,
850
+ "learning_rate": 0.00012787048336605297,
851
+ "loss": 0.1544,
852
+ "step": 54500
853
+ },
854
+ {
855
+ "epoch": 2.1872265966754156,
856
+ "grad_norm": 1.2831624746322632,
857
+ "learning_rate": 0.0001272072478378522,
858
+ "loss": 0.143,
859
+ "step": 55000
860
+ },
861
+ {
862
+ "epoch": 2.2003499562554683,
863
+ "eval_loss": 0.1747935712337494,
864
+ "eval_runtime": 31.3745,
865
+ "eval_samples_per_second": 15.936,
866
+ "eval_steps_per_second": 7.968,
867
+ "step": 55330
868
+ },
869
+ {
870
+ "epoch": 2.2071104748270103,
871
+ "grad_norm": 1.4339005947113037,
872
+ "learning_rate": 0.0001265440123096514,
873
+ "loss": 0.1587,
874
+ "step": 55500
875
+ },
876
+ {
877
+ "epoch": 2.226994352978605,
878
+ "grad_norm": 2.5719878673553467,
879
+ "learning_rate": 0.00012588077678145064,
880
+ "loss": 0.1527,
881
+ "step": 56000
882
+ },
883
+ {
884
+ "epoch": 2.2468782311301996,
885
+ "grad_norm": 2.1503918170928955,
886
+ "learning_rate": 0.00012521754125324988,
887
+ "loss": 0.1585,
888
+ "step": 56500
889
+ },
890
+ {
891
+ "epoch": 2.2667621092817942,
892
+ "grad_norm": 3.443103551864624,
893
+ "learning_rate": 0.0001245543057250491,
894
+ "loss": 0.1593,
895
+ "step": 57000
896
+ },
897
+ {
898
+ "epoch": 2.286645987433389,
899
+ "grad_norm": 1.2123481035232544,
900
+ "learning_rate": 0.0001238910701968483,
901
+ "loss": 0.1624,
902
+ "step": 57500
903
+ },
904
+ {
905
+ "epoch": 2.3065298655849835,
906
+ "grad_norm": 1.7718604803085327,
907
+ "learning_rate": 0.00012322783466864754,
908
+ "loss": 0.1569,
909
+ "step": 58000
910
+ },
911
+ {
912
+ "epoch": 2.3264137437365786,
913
+ "grad_norm": 4.025502681732178,
914
+ "learning_rate": 0.00012256592561150315,
915
+ "loss": 0.156,
916
+ "step": 58500
917
+ },
918
+ {
919
+ "epoch": 2.3462976218881733,
920
+ "grad_norm": 2.876533269882202,
921
+ "learning_rate": 0.00012190401655435879,
922
+ "loss": 0.1616,
923
+ "step": 59000
924
+ },
925
+ {
926
+ "epoch": 2.366181500039768,
927
+ "grad_norm": 1.680629849433899,
928
+ "learning_rate": 0.00012124078102615801,
929
+ "loss": 0.1543,
930
+ "step": 59500
931
+ },
932
+ {
933
+ "epoch": 2.3860653781913626,
934
+ "grad_norm": 1.0101946592330933,
935
+ "learning_rate": 0.00012057754549795724,
936
+ "loss": 0.1465,
937
+ "step": 60000
938
+ },
939
+ {
940
+ "epoch": 2.400381770460511,
941
+ "eval_loss": 0.16907282173633575,
942
+ "eval_runtime": 31.0907,
943
+ "eval_samples_per_second": 16.082,
944
+ "eval_steps_per_second": 8.041,
945
+ "step": 60360
946
+ },
947
+ {
948
+ "epoch": 2.405949256342957,
949
+ "grad_norm": 1.9501550197601318,
950
+ "learning_rate": 0.00011991430996975645,
951
+ "loss": 0.1535,
952
+ "step": 60500
953
+ },
954
+ {
955
+ "epoch": 2.425833134494552,
956
+ "grad_norm": 3.678605556488037,
957
+ "learning_rate": 0.0001192510744415557,
958
+ "loss": 0.1453,
959
+ "step": 61000
960
+ },
961
+ {
962
+ "epoch": 2.4457170126461465,
963
+ "grad_norm": 4.075408458709717,
964
+ "learning_rate": 0.00011858783891335492,
965
+ "loss": 0.1494,
966
+ "step": 61500
967
+ },
968
+ {
969
+ "epoch": 2.465600890797741,
970
+ "grad_norm": 1.3122295141220093,
971
+ "learning_rate": 0.00011792460338515414,
972
+ "loss": 0.1593,
973
+ "step": 62000
974
+ },
975
+ {
976
+ "epoch": 2.485484768949336,
977
+ "grad_norm": 2.0857038497924805,
978
+ "learning_rate": 0.00011726136785695337,
979
+ "loss": 0.1514,
980
+ "step": 62500
981
+ },
982
+ {
983
+ "epoch": 2.5053686471009304,
984
+ "grad_norm": 1.6657792329788208,
985
+ "learning_rate": 0.000116599458799809,
986
+ "loss": 0.1501,
987
+ "step": 63000
988
+ },
989
+ {
990
+ "epoch": 2.525252525252525,
991
+ "grad_norm": 1.8100333213806152,
992
+ "learning_rate": 0.00011593622327160823,
993
+ "loss": 0.1519,
994
+ "step": 63500
995
+ },
996
+ {
997
+ "epoch": 2.54513640340412,
998
+ "grad_norm": 2.504148006439209,
999
+ "learning_rate": 0.00011527298774340744,
1000
+ "loss": 0.1496,
1001
+ "step": 64000
1002
+ },
1003
+ {
1004
+ "epoch": 2.5650202815557144,
1005
+ "grad_norm": 1.1805058717727661,
1006
+ "learning_rate": 0.00011461107868626306,
1007
+ "loss": 0.144,
1008
+ "step": 64500
1009
+ },
1010
+ {
1011
+ "epoch": 2.5849041597073095,
1012
+ "grad_norm": 3.109494686126709,
1013
+ "learning_rate": 0.0001139478431580623,
1014
+ "loss": 0.1517,
1015
+ "step": 65000
1016
+ },
1017
+ {
1018
+ "epoch": 2.600413584665553,
1019
+ "eval_loss": 0.17017363011837006,
1020
+ "eval_runtime": 31.1173,
1021
+ "eval_samples_per_second": 16.068,
1022
+ "eval_steps_per_second": 8.034,
1023
+ "step": 65390
1024
+ },
1025
+ {
1026
+ "epoch": 2.604788037858904,
1027
+ "grad_norm": 2.1431307792663574,
1028
+ "learning_rate": 0.00011328460762986153,
1029
+ "loss": 0.1561,
1030
+ "step": 65500
1031
+ },
1032
+ {
1033
+ "epoch": 2.6246719160104988,
1034
+ "grad_norm": 1.5804929733276367,
1035
+ "learning_rate": 0.00011262137210166075,
1036
+ "loss": 0.1536,
1037
+ "step": 66000
1038
+ },
1039
+ {
1040
+ "epoch": 2.6445557941620934,
1041
+ "grad_norm": 3.0483312606811523,
1042
+ "learning_rate": 0.00011195813657345997,
1043
+ "loss": 0.1497,
1044
+ "step": 66500
1045
+ },
1046
+ {
1047
+ "epoch": 2.664439672313688,
1048
+ "grad_norm": 2.7714364528656006,
1049
+ "learning_rate": 0.00011129490104525919,
1050
+ "loss": 0.1493,
1051
+ "step": 67000
1052
+ },
1053
+ {
1054
+ "epoch": 2.6843235504652827,
1055
+ "grad_norm": 1.1538747549057007,
1056
+ "learning_rate": 0.00011063299198811483,
1057
+ "loss": 0.1586,
1058
+ "step": 67500
1059
+ },
1060
+ {
1061
+ "epoch": 2.7042074286168774,
1062
+ "grad_norm": 1.946218729019165,
1063
+ "learning_rate": 0.00010996975645991405,
1064
+ "loss": 0.1484,
1065
+ "step": 68000
1066
+ },
1067
+ {
1068
+ "epoch": 2.724091306768472,
1069
+ "grad_norm": 1.453487753868103,
1070
+ "learning_rate": 0.00010930652093171328,
1071
+ "loss": 0.1545,
1072
+ "step": 68500
1073
+ },
1074
+ {
1075
+ "epoch": 2.7439751849200666,
1076
+ "grad_norm": 1.9838638305664062,
1077
+ "learning_rate": 0.00010864328540351249,
1078
+ "loss": 0.1443,
1079
+ "step": 69000
1080
+ },
1081
+ {
1082
+ "epoch": 2.7638590630716617,
1083
+ "grad_norm": 0.83545982837677,
1084
+ "learning_rate": 0.00010798004987531172,
1085
+ "loss": 0.148,
1086
+ "step": 69500
1087
+ },
1088
+ {
1089
+ "epoch": 2.783742941223256,
1090
+ "grad_norm": 2.1687493324279785,
1091
+ "learning_rate": 0.00010731681434711096,
1092
+ "loss": 0.1533,
1093
+ "step": 70000
1094
+ },
1095
+ {
1096
+ "epoch": 2.800445398870596,
1097
+ "eval_loss": 0.16508887708187103,
1098
+ "eval_runtime": 31.1958,
1099
+ "eval_samples_per_second": 16.028,
1100
+ "eval_steps_per_second": 8.014,
1101
+ "step": 70420
1102
+ },
1103
+ {
1104
+ "epoch": 2.803626819374851,
1105
+ "grad_norm": 1.3648791313171387,
1106
+ "learning_rate": 0.00010665357881891018,
1107
+ "loss": 0.1419,
1108
+ "step": 70500
1109
+ },
1110
+ {
1111
+ "epoch": 2.8235106975264457,
1112
+ "grad_norm": 1.7863802909851074,
1113
+ "learning_rate": 0.00010599034329070941,
1114
+ "loss": 0.1482,
1115
+ "step": 71000
1116
+ },
1117
+ {
1118
+ "epoch": 2.8433945756780403,
1119
+ "grad_norm": 3.2346866130828857,
1120
+ "learning_rate": 0.00010532710776250862,
1121
+ "loss": 0.1425,
1122
+ "step": 71500
1123
+ },
1124
+ {
1125
+ "epoch": 2.863278453829635,
1126
+ "grad_norm": 0.9211582541465759,
1127
+ "learning_rate": 0.00010466519870536427,
1128
+ "loss": 0.1501,
1129
+ "step": 72000
1130
+ },
1131
+ {
1132
+ "epoch": 2.8831623319812296,
1133
+ "grad_norm": 0.1923113614320755,
1134
+ "learning_rate": 0.00010400196317716348,
1135
+ "loss": 0.1469,
1136
+ "step": 72500
1137
+ },
1138
+ {
1139
+ "epoch": 2.9030462101328243,
1140
+ "grad_norm": 4.506950855255127,
1141
+ "learning_rate": 0.00010333872764896271,
1142
+ "loss": 0.1465,
1143
+ "step": 73000
1144
+ },
1145
+ {
1146
+ "epoch": 2.922930088284419,
1147
+ "grad_norm": 3.244807004928589,
1148
+ "learning_rate": 0.00010267549212076194,
1149
+ "loss": 0.1425,
1150
+ "step": 73500
1151
+ },
1152
+ {
1153
+ "epoch": 2.9428139664360136,
1154
+ "grad_norm": 1.8501616716384888,
1155
+ "learning_rate": 0.00010201225659256115,
1156
+ "loss": 0.1446,
1157
+ "step": 74000
1158
+ },
1159
+ {
1160
+ "epoch": 2.962697844587608,
1161
+ "grad_norm": 3.9013025760650635,
1162
+ "learning_rate": 0.0001013503475354168,
1163
+ "loss": 0.1508,
1164
+ "step": 74500
1165
+ },
1166
+ {
1167
+ "epoch": 2.9825817227392033,
1168
+ "grad_norm": 3.1799283027648926,
1169
+ "learning_rate": 0.00010068711200721601,
1170
+ "loss": 0.1412,
1171
+ "step": 75000
1172
+ },
1173
+ {
1174
+ "epoch": 3.000477213075638,
1175
+ "eval_loss": 0.1569422334432602,
1176
+ "eval_runtime": 33.8027,
1177
+ "eval_samples_per_second": 14.792,
1178
+ "eval_steps_per_second": 7.396,
1179
+ "step": 75450
1180
+ },
1181
+ {
1182
+ "epoch": 3.002465600890798,
1183
+ "grad_norm": 1.1885572671890259,
1184
+ "learning_rate": 0.00010002520295007163,
1185
+ "loss": 0.1334,
1186
+ "step": 75500
1187
+ },
1188
+ {
1189
+ "epoch": 3.0223494790423926,
1190
+ "grad_norm": 0.8424841165542603,
1191
+ "learning_rate": 9.936196742187086e-05,
1192
+ "loss": 0.1261,
1193
+ "step": 76000
1194
+ },
1195
+ {
1196
+ "epoch": 3.0422333571939872,
1197
+ "grad_norm": 1.2642099857330322,
1198
+ "learning_rate": 9.869873189367008e-05,
1199
+ "loss": 0.1233,
1200
+ "step": 76500
1201
+ },
1202
+ {
1203
+ "epoch": 3.062117235345582,
1204
+ "grad_norm": 3.7216506004333496,
1205
+ "learning_rate": 9.803549636546931e-05,
1206
+ "loss": 0.1257,
1207
+ "step": 77000
1208
+ },
1209
+ {
1210
+ "epoch": 3.0820011134971765,
1211
+ "grad_norm": 1.576157808303833,
1212
+ "learning_rate": 9.737358730832493e-05,
1213
+ "loss": 0.1257,
1214
+ "step": 77500
1215
+ },
1216
+ {
1217
+ "epoch": 3.101884991648771,
1218
+ "grad_norm": 3.2462401390075684,
1219
+ "learning_rate": 9.671035178012417e-05,
1220
+ "loss": 0.1387,
1221
+ "step": 78000
1222
+ },
1223
+ {
1224
+ "epoch": 3.121768869800366,
1225
+ "grad_norm": 0.38061007857322693,
1226
+ "learning_rate": 9.604711625192338e-05,
1227
+ "loss": 0.1288,
1228
+ "step": 78500
1229
+ },
1230
+ {
1231
+ "epoch": 3.1416527479519605,
1232
+ "grad_norm": 1.4168367385864258,
1233
+ "learning_rate": 9.538388072372262e-05,
1234
+ "loss": 0.125,
1235
+ "step": 79000
1236
+ },
1237
+ {
1238
+ "epoch": 3.161536626103555,
1239
+ "grad_norm": 1.8440510034561157,
1240
+ "learning_rate": 9.472064519552183e-05,
1241
+ "loss": 0.1203,
1242
+ "step": 79500
1243
+ },
1244
+ {
1245
+ "epoch": 3.1814205042551498,
1246
+ "grad_norm": 0.012771493755280972,
1247
+ "learning_rate": 9.405873613837747e-05,
1248
+ "loss": 0.1237,
1249
+ "step": 80000
1250
+ },
1251
+ {
1252
+ "epoch": 3.2005090272806807,
1253
+ "eval_loss": 0.15806268155574799,
1254
+ "eval_runtime": 31.0713,
1255
+ "eval_samples_per_second": 16.092,
1256
+ "eval_steps_per_second": 8.046,
1257
+ "step": 80480
1258
+ },
1259
+ {
1260
+ "epoch": 3.2013043824067444,
1261
+ "grad_norm": 2.36724853515625,
1262
+ "learning_rate": 9.33955006101767e-05,
1263
+ "loss": 0.1253,
1264
+ "step": 80500
1265
+ },
1266
+ {
1267
+ "epoch": 3.221188260558339,
1268
+ "grad_norm": 1.2032071352005005,
1269
+ "learning_rate": 9.273226508197591e-05,
1270
+ "loss": 0.1321,
1271
+ "step": 81000
1272
+ },
1273
+ {
1274
+ "epoch": 3.241072138709934,
1275
+ "grad_norm": 1.6413310766220093,
1276
+ "learning_rate": 9.206902955377515e-05,
1277
+ "loss": 0.1256,
1278
+ "step": 81500
1279
+ },
1280
+ {
1281
+ "epoch": 3.260956016861529,
1282
+ "grad_norm": 1.263128638267517,
1283
+ "learning_rate": 9.140579402557436e-05,
1284
+ "loss": 0.1281,
1285
+ "step": 82000
1286
+ },
1287
+ {
1288
+ "epoch": 3.2808398950131235,
1289
+ "grad_norm": 2.12233304977417,
1290
+ "learning_rate": 9.074255849737359e-05,
1291
+ "loss": 0.1299,
1292
+ "step": 82500
1293
+ },
1294
+ {
1295
+ "epoch": 3.300723773164718,
1296
+ "grad_norm": 1.265188217163086,
1297
+ "learning_rate": 9.007932296917282e-05,
1298
+ "loss": 0.1226,
1299
+ "step": 83000
1300
+ },
1301
+ {
1302
+ "epoch": 3.3206076513163127,
1303
+ "grad_norm": 1.391507625579834,
1304
+ "learning_rate": 8.941608744097204e-05,
1305
+ "loss": 0.1269,
1306
+ "step": 83500
1307
+ },
1308
+ {
1309
+ "epoch": 3.3404915294679074,
1310
+ "grad_norm": 1.941706657409668,
1311
+ "learning_rate": 8.875285191277126e-05,
1312
+ "loss": 0.1306,
1313
+ "step": 84000
1314
+ },
1315
+ {
1316
+ "epoch": 3.360375407619502,
1317
+ "grad_norm": 1.1882116794586182,
1318
+ "learning_rate": 8.809094285562688e-05,
1319
+ "loss": 0.1287,
1320
+ "step": 84500
1321
+ },
1322
+ {
1323
+ "epoch": 3.3802592857710967,
1324
+ "grad_norm": 2.2689340114593506,
1325
+ "learning_rate": 8.742903379848252e-05,
1326
+ "loss": 0.1281,
1327
+ "step": 85000
1328
+ },
1329
+ {
1330
+ "epoch": 3.4001431639226913,
1331
+ "grad_norm": 2.9482505321502686,
1332
+ "learning_rate": 8.676579827028175e-05,
1333
+ "loss": 0.131,
1334
+ "step": 85500
1335
+ },
1336
+ {
1337
+ "epoch": 3.400540841485723,
1338
+ "eval_loss": 0.15611666440963745,
1339
+ "eval_runtime": 31.2043,
1340
+ "eval_samples_per_second": 16.023,
1341
+ "eval_steps_per_second": 8.012,
1342
+ "step": 85510
1343
+ },
1344
+ {
1345
+ "epoch": 3.420027042074286,
1346
+ "grad_norm": 1.8892121315002441,
1347
+ "learning_rate": 8.610256274208097e-05,
1348
+ "loss": 0.1261,
1349
+ "step": 86000
1350
+ },
1351
+ {
1352
+ "epoch": 3.4399109202258806,
1353
+ "grad_norm": 2.923980236053467,
1354
+ "learning_rate": 8.54393272138802e-05,
1355
+ "loss": 0.1304,
1356
+ "step": 86500
1357
+ },
1358
+ {
1359
+ "epoch": 3.4597947983774757,
1360
+ "grad_norm": 0.46090424060821533,
1361
+ "learning_rate": 8.477609168567942e-05,
1362
+ "loss": 0.1263,
1363
+ "step": 87000
1364
+ },
1365
+ {
1366
+ "epoch": 3.4796786765290704,
1367
+ "grad_norm": 1.307573676109314,
1368
+ "learning_rate": 8.411285615747865e-05,
1369
+ "loss": 0.1217,
1370
+ "step": 87500
1371
+ },
1372
+ {
1373
+ "epoch": 3.499562554680665,
1374
+ "grad_norm": 1.9948647022247314,
1375
+ "learning_rate": 8.345094710033427e-05,
1376
+ "loss": 0.1298,
1377
+ "step": 88000
1378
+ },
1379
+ {
1380
+ "epoch": 3.5194464328322597,
1381
+ "grad_norm": 1.2264705896377563,
1382
+ "learning_rate": 8.27877115721335e-05,
1383
+ "loss": 0.1254,
1384
+ "step": 88500
1385
+ },
1386
+ {
1387
+ "epoch": 3.5393303109838543,
1388
+ "grad_norm": 1.249614953994751,
1389
+ "learning_rate": 8.212447604393274e-05,
1390
+ "loss": 0.1217,
1391
+ "step": 89000
1392
+ },
1393
+ {
1394
+ "epoch": 3.559214189135449,
1395
+ "grad_norm": 1.5197207927703857,
1396
+ "learning_rate": 8.146124051573195e-05,
1397
+ "loss": 0.1276,
1398
+ "step": 89500
1399
+ },
1400
+ {
1401
+ "epoch": 3.5790980672870436,
1402
+ "grad_norm": 0.8200188875198364,
1403
+ "learning_rate": 8.079800498753118e-05,
1404
+ "loss": 0.1283,
1405
+ "step": 90000
1406
+ },
1407
+ {
1408
+ "epoch": 3.5989819454386383,
1409
+ "grad_norm": 1.3526813983917236,
1410
+ "learning_rate": 8.01347694593304e-05,
1411
+ "loss": 0.1221,
1412
+ "step": 90500
1413
+ },
1414
+ {
1415
+ "epoch": 3.600572655690766,
1416
+ "eval_loss": 0.14859890937805176,
1417
+ "eval_runtime": 32.6446,
1418
+ "eval_samples_per_second": 15.316,
1419
+ "eval_steps_per_second": 7.658,
1420
+ "step": 90540
1421
+ },
1422
+ {
1423
+ "epoch": 3.618865823590233,
1424
+ "grad_norm": 1.9000824689865112,
1425
+ "learning_rate": 7.947153393112963e-05,
1426
+ "loss": 0.1215,
1427
+ "step": 91000
1428
+ },
1429
+ {
1430
+ "epoch": 3.638749701741828,
1431
+ "grad_norm": 1.2409000396728516,
1432
+ "learning_rate": 7.880829840292885e-05,
1433
+ "loss": 0.1259,
1434
+ "step": 91500
1435
+ },
1436
+ {
1437
+ "epoch": 3.658633579893422,
1438
+ "grad_norm": 1.7296925783157349,
1439
+ "learning_rate": 7.814506287472808e-05,
1440
+ "loss": 0.1245,
1441
+ "step": 92000
1442
+ },
1443
+ {
1444
+ "epoch": 3.6785174580450173,
1445
+ "grad_norm": 1.3685667514801025,
1446
+ "learning_rate": 7.748315381758371e-05,
1447
+ "loss": 0.127,
1448
+ "step": 92500
1449
+ },
1450
+ {
1451
+ "epoch": 3.698401336196612,
1452
+ "grad_norm": 2.338923692703247,
1453
+ "learning_rate": 7.681991828938293e-05,
1454
+ "loss": 0.119,
1455
+ "step": 93000
1456
+ },
1457
+ {
1458
+ "epoch": 3.7182852143482066,
1459
+ "grad_norm": 2.3641035556793213,
1460
+ "learning_rate": 7.615668276118215e-05,
1461
+ "loss": 0.1256,
1462
+ "step": 93500
1463
+ },
1464
+ {
1465
+ "epoch": 3.7381690924998012,
1466
+ "grad_norm": 1.444541335105896,
1467
+ "learning_rate": 7.549477370403777e-05,
1468
+ "loss": 0.1218,
1469
+ "step": 94000
1470
+ },
1471
+ {
1472
+ "epoch": 3.758052970651396,
1473
+ "grad_norm": 2.2175920009613037,
1474
+ "learning_rate": 7.4831538175837e-05,
1475
+ "loss": 0.1181,
1476
+ "step": 94500
1477
+ },
1478
+ {
1479
+ "epoch": 3.7779368488029905,
1480
+ "grad_norm": 2.533275604248047,
1481
+ "learning_rate": 7.416830264763624e-05,
1482
+ "loss": 0.1203,
1483
+ "step": 95000
1484
+ },
1485
+ {
1486
+ "epoch": 3.797820726954585,
1487
+ "grad_norm": 2.5117480754852295,
1488
+ "learning_rate": 7.350506711943545e-05,
1489
+ "loss": 0.1161,
1490
+ "step": 95500
1491
+ },
1492
+ {
1493
+ "epoch": 3.8006044698958084,
1494
+ "eval_loss": 0.14638979732990265,
1495
+ "eval_runtime": 33.193,
1496
+ "eval_samples_per_second": 15.063,
1497
+ "eval_steps_per_second": 7.532,
1498
+ "step": 95570
1499
+ },
1500
+ {
1501
+ "epoch": 3.81770460510618,
1502
+ "grad_norm": 0.9872569441795349,
1503
+ "learning_rate": 7.284183159123469e-05,
1504
+ "loss": 0.1257,
1505
+ "step": 96000
1506
+ },
1507
+ {
1508
+ "epoch": 3.8375884832577745,
1509
+ "grad_norm": 2.7399699687957764,
1510
+ "learning_rate": 7.21785960630339e-05,
1511
+ "loss": 0.1309,
1512
+ "step": 96500
1513
+ },
1514
+ {
1515
+ "epoch": 3.8574723614093696,
1516
+ "grad_norm": 1.9438740015029907,
1517
+ "learning_rate": 7.151536053483313e-05,
1518
+ "loss": 0.1196,
1519
+ "step": 97000
1520
+ },
1521
+ {
1522
+ "epoch": 3.8773562395609638,
1523
+ "grad_norm": 3.683006763458252,
1524
+ "learning_rate": 7.085212500663237e-05,
1525
+ "loss": 0.1207,
1526
+ "step": 97500
1527
+ },
1528
+ {
1529
+ "epoch": 3.897240117712559,
1530
+ "grad_norm": 2.3406124114990234,
1531
+ "learning_rate": 7.019154242054439e-05,
1532
+ "loss": 0.1202,
1533
+ "step": 98000
1534
+ },
1535
+ {
1536
+ "epoch": 3.9171239958641535,
1537
+ "grad_norm": 1.3052023649215698,
1538
+ "learning_rate": 6.952830689234361e-05,
1539
+ "loss": 0.1233,
1540
+ "step": 98500
1541
+ },
1542
+ {
1543
+ "epoch": 3.937007874015748,
1544
+ "grad_norm": 1.8666225671768188,
1545
+ "learning_rate": 6.886507136414284e-05,
1546
+ "loss": 0.1217,
1547
+ "step": 99000
1548
+ },
1549
+ {
1550
+ "epoch": 3.956891752167343,
1551
+ "grad_norm": 1.7979810237884521,
1552
+ "learning_rate": 6.820183583594206e-05,
1553
+ "loss": 0.1194,
1554
+ "step": 99500
1555
+ },
1556
+ {
1557
+ "epoch": 3.9767756303189374,
1558
+ "grad_norm": 0.9724803566932678,
1559
+ "learning_rate": 6.753860030774129e-05,
1560
+ "loss": 0.1276,
1561
+ "step": 100000
1562
+ },
1563
+ {
1564
+ "epoch": 3.996659508470532,
1565
+ "grad_norm": 3.251120090484619,
1566
+ "learning_rate": 6.687536477954052e-05,
1567
+ "loss": 0.1165,
1568
+ "step": 100500
1569
+ },
1570
+ {
1571
+ "epoch": 4.000636284100851,
1572
+ "eval_loss": 0.1412682831287384,
1573
+ "eval_runtime": 31.0761,
1574
+ "eval_samples_per_second": 16.09,
1575
+ "eval_steps_per_second": 8.045,
1576
+ "step": 100600
1577
+ },
1578
+ {
1579
+ "epoch": 4.016543386622127,
1580
+ "grad_norm": 3.6826343536376953,
1581
+ "learning_rate": 6.621212925133974e-05,
1582
+ "loss": 0.1046,
1583
+ "step": 101000
1584
+ },
1585
+ {
1586
+ "epoch": 4.036427264773722,
1587
+ "grad_norm": 5.360718250274658,
1588
+ "learning_rate": 6.554889372313896e-05,
1589
+ "loss": 0.1032,
1590
+ "step": 101500
1591
+ },
1592
+ {
1593
+ "epoch": 4.056311142925316,
1594
+ "grad_norm": 1.780490756034851,
1595
+ "learning_rate": 6.488698466599459e-05,
1596
+ "loss": 0.1014,
1597
+ "step": 102000
1598
+ },
1599
+ {
1600
+ "epoch": 4.076195021076911,
1601
+ "grad_norm": 2.984694480895996,
1602
+ "learning_rate": 6.422374913779382e-05,
1603
+ "loss": 0.1044,
1604
+ "step": 102500
1605
+ },
1606
+ {
1607
+ "epoch": 4.096078899228505,
1608
+ "grad_norm": 0.6237806677818298,
1609
+ "learning_rate": 6.356184008064945e-05,
1610
+ "loss": 0.1009,
1611
+ "step": 103000
1612
+ },
1613
+ {
1614
+ "epoch": 4.1159627773801,
1615
+ "grad_norm": 3.3006651401519775,
1616
+ "learning_rate": 6.289860455244868e-05,
1617
+ "loss": 0.1017,
1618
+ "step": 103500
1619
+ },
1620
+ {
1621
+ "epoch": 4.135846655531695,
1622
+ "grad_norm": 1.9449480772018433,
1623
+ "learning_rate": 6.223536902424789e-05,
1624
+ "loss": 0.1043,
1625
+ "step": 104000
1626
+ },
1627
+ {
1628
+ "epoch": 4.15573053368329,
1629
+ "grad_norm": 2.495291233062744,
1630
+ "learning_rate": 6.157213349604713e-05,
1631
+ "loss": 0.1024,
1632
+ "step": 104500
1633
+ },
1634
+ {
1635
+ "epoch": 4.175614411834884,
1636
+ "grad_norm": 0.23489686846733093,
1637
+ "learning_rate": 6.090889796784635e-05,
1638
+ "loss": 0.1031,
1639
+ "step": 105000
1640
+ },
1641
+ {
1642
+ "epoch": 4.195498289986479,
1643
+ "grad_norm": 3.0352611541748047,
1644
+ "learning_rate": 6.024566243964557e-05,
1645
+ "loss": 0.0994,
1646
+ "step": 105500
1647
+ },
1648
+ {
1649
+ "epoch": 4.2006680983058935,
1650
+ "eval_loss": 0.14581650495529175,
1651
+ "eval_runtime": 35.7264,
1652
+ "eval_samples_per_second": 13.995,
1653
+ "eval_steps_per_second": 6.998,
1654
+ "step": 105630
1655
+ },
1656
+ {
1657
+ "epoch": 4.215382168138074,
1658
+ "grad_norm": 1.1670618057250977,
1659
+ "learning_rate": 5.95824269114448e-05,
1660
+ "loss": 0.1045,
1661
+ "step": 106000
1662
+ },
1663
+ {
1664
+ "epoch": 4.235266046289668,
1665
+ "grad_norm": 2.0033414363861084,
1666
+ "learning_rate": 5.891919138324402e-05,
1667
+ "loss": 0.1016,
1668
+ "step": 106500
1669
+ },
1670
+ {
1671
+ "epoch": 4.255149924441263,
1672
+ "grad_norm": 1.4979524612426758,
1673
+ "learning_rate": 5.825728232609965e-05,
1674
+ "loss": 0.1079,
1675
+ "step": 107000
1676
+ },
1677
+ {
1678
+ "epoch": 4.275033802592858,
1679
+ "grad_norm": 1.7079240083694458,
1680
+ "learning_rate": 5.7594046797898874e-05,
1681
+ "loss": 0.1016,
1682
+ "step": 107500
1683
+ },
1684
+ {
1685
+ "epoch": 4.294917680744453,
1686
+ "grad_norm": 1.1605221033096313,
1687
+ "learning_rate": 5.6930811269698107e-05,
1688
+ "loss": 0.1008,
1689
+ "step": 108000
1690
+ },
1691
+ {
1692
+ "epoch": 4.314801558896047,
1693
+ "grad_norm": 1.4162037372589111,
1694
+ "learning_rate": 5.6267575741497326e-05,
1695
+ "loss": 0.1083,
1696
+ "step": 108500
1697
+ },
1698
+ {
1699
+ "epoch": 4.334685437047642,
1700
+ "grad_norm": 2.373020887374878,
1701
+ "learning_rate": 5.5604340213296545e-05,
1702
+ "loss": 0.1073,
1703
+ "step": 109000
1704
+ },
1705
+ {
1706
+ "epoch": 4.354569315199236,
1707
+ "grad_norm": 2.836911916732788,
1708
+ "learning_rate": 5.494375762720858e-05,
1709
+ "loss": 0.1019,
1710
+ "step": 109500
1711
+ },
1712
+ {
1713
+ "epoch": 4.374453193350831,
1714
+ "grad_norm": 3.7747974395751953,
1715
+ "learning_rate": 5.42805220990078e-05,
1716
+ "loss": 0.1038,
1717
+ "step": 110000
1718
+ },
1719
+ {
1720
+ "epoch": 4.3943370715024255,
1721
+ "grad_norm": 1.4646281003952026,
1722
+ "learning_rate": 5.361728657080703e-05,
1723
+ "loss": 0.1113,
1724
+ "step": 110500
1725
+ },
1726
+ {
1727
+ "epoch": 4.4006999125109365,
1728
+ "eval_loss": 0.14834672212600708,
1729
+ "eval_runtime": 37.228,
1730
+ "eval_samples_per_second": 13.431,
1731
+ "eval_steps_per_second": 6.715,
1732
+ "step": 110660
1733
+ },
1734
+ {
1735
+ "epoch": 4.414220949654021,
1736
+ "grad_norm": 2.5483474731445312,
1737
+ "learning_rate": 5.295405104260626e-05,
1738
+ "loss": 0.1006,
1739
+ "step": 111000
1740
+ },
1741
+ {
1742
+ "epoch": 4.434104827805616,
1743
+ "grad_norm": 2.5747101306915283,
1744
+ "learning_rate": 5.229081551440548e-05,
1745
+ "loss": 0.1082,
1746
+ "step": 111500
1747
+ },
1748
+ {
1749
+ "epoch": 4.45398870595721,
1750
+ "grad_norm": 0.23354358971118927,
1751
+ "learning_rate": 5.16275799862047e-05,
1752
+ "loss": 0.1032,
1753
+ "step": 112000
1754
+ },
1755
+ {
1756
+ "epoch": 4.473872584108805,
1757
+ "grad_norm": 2.4420604705810547,
1758
+ "learning_rate": 5.096434445800393e-05,
1759
+ "loss": 0.106,
1760
+ "step": 112500
1761
+ },
1762
+ {
1763
+ "epoch": 4.493756462260399,
1764
+ "grad_norm": 1.2000492811203003,
1765
+ "learning_rate": 5.030110892980315e-05,
1766
+ "loss": 0.0961,
1767
+ "step": 113000
1768
+ },
1769
+ {
1770
+ "epoch": 4.513640340411994,
1771
+ "grad_norm": 1.8864790201187134,
1772
+ "learning_rate": 4.963919987265878e-05,
1773
+ "loss": 0.0978,
1774
+ "step": 113500
1775
+ },
1776
+ {
1777
+ "epoch": 4.5335242185635884,
1778
+ "grad_norm": 0.45711418986320496,
1779
+ "learning_rate": 4.897596434445801e-05,
1780
+ "loss": 0.094,
1781
+ "step": 114000
1782
+ },
1783
+ {
1784
+ "epoch": 4.5534080967151835,
1785
+ "grad_norm": 7.053537845611572,
1786
+ "learning_rate": 4.831272881625723e-05,
1787
+ "loss": 0.1,
1788
+ "step": 114500
1789
+ },
1790
+ {
1791
+ "epoch": 4.573291974866778,
1792
+ "grad_norm": 3.276334047317505,
1793
+ "learning_rate": 4.765081975911286e-05,
1794
+ "loss": 0.0996,
1795
+ "step": 115000
1796
+ },
1797
+ {
1798
+ "epoch": 4.593175853018373,
1799
+ "grad_norm": 1.4245741367340088,
1800
+ "learning_rate": 4.6987584230912085e-05,
1801
+ "loss": 0.0996,
1802
+ "step": 115500
1803
+ },
1804
+ {
1805
+ "epoch": 4.600731726715979,
1806
+ "eval_loss": 0.14262481033802032,
1807
+ "eval_runtime": 32.3953,
1808
+ "eval_samples_per_second": 15.434,
1809
+ "eval_steps_per_second": 7.717,
1810
+ "step": 115690
1811
+ },
1812
+ {
1813
+ "epoch": 4.613059731169967,
1814
+ "grad_norm": 2.3254101276397705,
1815
+ "learning_rate": 4.6324348702711304e-05,
1816
+ "loss": 0.1017,
1817
+ "step": 116000
1818
+ },
1819
+ {
1820
+ "epoch": 4.632943609321562,
1821
+ "grad_norm": 1.167358636856079,
1822
+ "learning_rate": 4.566111317451054e-05,
1823
+ "loss": 0.0981,
1824
+ "step": 116500
1825
+ },
1826
+ {
1827
+ "epoch": 4.652827487473157,
1828
+ "grad_norm": 4.5853376388549805,
1829
+ "learning_rate": 4.499787764630976e-05,
1830
+ "loss": 0.1016,
1831
+ "step": 117000
1832
+ },
1833
+ {
1834
+ "epoch": 4.672711365624751,
1835
+ "grad_norm": 1.4917422533035278,
1836
+ "learning_rate": 4.433464211810899e-05,
1837
+ "loss": 0.1018,
1838
+ "step": 117500
1839
+ },
1840
+ {
1841
+ "epoch": 4.6925952437763465,
1842
+ "grad_norm": 0.012467560358345509,
1843
+ "learning_rate": 4.367273306096461e-05,
1844
+ "loss": 0.0975,
1845
+ "step": 118000
1846
+ },
1847
+ {
1848
+ "epoch": 4.712479121927941,
1849
+ "grad_norm": 0.8470689058303833,
1850
+ "learning_rate": 4.3009497532763836e-05,
1851
+ "loss": 0.0965,
1852
+ "step": 118500
1853
+ },
1854
+ {
1855
+ "epoch": 4.732363000079536,
1856
+ "grad_norm": 0.43946540355682373,
1857
+ "learning_rate": 4.234626200456306e-05,
1858
+ "loss": 0.1014,
1859
+ "step": 119000
1860
+ },
1861
+ {
1862
+ "epoch": 4.75224687823113,
1863
+ "grad_norm": 2.230210304260254,
1864
+ "learning_rate": 4.168302647636229e-05,
1865
+ "loss": 0.0996,
1866
+ "step": 119500
1867
+ },
1868
+ {
1869
+ "epoch": 4.772130756382725,
1870
+ "grad_norm": 3.4537596702575684,
1871
+ "learning_rate": 4.1019790948161514e-05,
1872
+ "loss": 0.0966,
1873
+ "step": 120000
1874
+ },
1875
+ {
1876
+ "epoch": 4.792014634534319,
1877
+ "grad_norm": 1.7490534782409668,
1878
+ "learning_rate": 4.035655541996074e-05,
1879
+ "loss": 0.1116,
1880
+ "step": 120500
1881
+ },
1882
+ {
1883
+ "epoch": 4.800763540921022,
1884
+ "eval_loss": 0.1392485797405243,
1885
+ "eval_runtime": 34.3723,
1886
+ "eval_samples_per_second": 14.547,
1887
+ "eval_steps_per_second": 7.273,
1888
+ "step": 120720
1889
+ },
1890
+ {
1891
+ "epoch": 4.811898512685914,
1892
+ "grad_norm": 0.9216163158416748,
1893
+ "learning_rate": 3.969331989175996e-05,
1894
+ "loss": 0.0965,
1895
+ "step": 121000
1896
+ },
1897
+ {
1898
+ "epoch": 4.831782390837509,
1899
+ "grad_norm": 4.56651496887207,
1900
+ "learning_rate": 3.9030084363559186e-05,
1901
+ "loss": 0.0965,
1902
+ "step": 121500
1903
+ },
1904
+ {
1905
+ "epoch": 4.851666268989104,
1906
+ "grad_norm": 1.2061914205551147,
1907
+ "learning_rate": 3.8368175306414814e-05,
1908
+ "loss": 0.1059,
1909
+ "step": 122000
1910
+ },
1911
+ {
1912
+ "epoch": 4.871550147140699,
1913
+ "grad_norm": 1.8895795345306396,
1914
+ "learning_rate": 3.770493977821405e-05,
1915
+ "loss": 0.0988,
1916
+ "step": 122500
1917
+ },
1918
+ {
1919
+ "epoch": 4.891434025292293,
1920
+ "grad_norm": 2.730050563812256,
1921
+ "learning_rate": 3.7041704250013266e-05,
1922
+ "loss": 0.1022,
1923
+ "step": 123000
1924
+ },
1925
+ {
1926
+ "epoch": 4.911317903443888,
1927
+ "grad_norm": 2.6674962043762207,
1928
+ "learning_rate": 3.637846872181249e-05,
1929
+ "loss": 0.0966,
1930
+ "step": 123500
1931
+ },
1932
+ {
1933
+ "epoch": 4.931201781595482,
1934
+ "grad_norm": 5.602296352386475,
1935
+ "learning_rate": 3.571655966466812e-05,
1936
+ "loss": 0.0944,
1937
+ "step": 124000
1938
+ },
1939
+ {
1940
+ "epoch": 4.951085659747077,
1941
+ "grad_norm": 2.389179229736328,
1942
+ "learning_rate": 3.505332413646734e-05,
1943
+ "loss": 0.0975,
1944
+ "step": 124500
1945
+ },
1946
+ {
1947
+ "epoch": 4.970969537898672,
1948
+ "grad_norm": 3.1410694122314453,
1949
+ "learning_rate": 3.4390088608266565e-05,
1950
+ "loss": 0.0923,
1951
+ "step": 125000
1952
+ },
1953
+ {
1954
+ "epoch": 4.990853416050267,
1955
+ "grad_norm": 0.3184365928173065,
1956
+ "learning_rate": 3.37268530800658e-05,
1957
+ "loss": 0.1004,
1958
+ "step": 125500
1959
+ },
1960
+ {
1961
+ "epoch": 5.000795355126064,
1962
+ "eval_loss": 0.13804183900356293,
1963
+ "eval_runtime": 34.1527,
1964
+ "eval_samples_per_second": 14.64,
1965
+ "eval_steps_per_second": 7.32,
1966
+ "step": 125750
1967
+ },
1968
+ {
1969
+ "epoch": 5.010737294201861,
1970
+ "grad_norm": 1.9156638383865356,
1971
+ "learning_rate": 3.3063617551865024e-05,
1972
+ "loss": 0.0899,
1973
+ "step": 126000
1974
+ },
1975
+ {
1976
+ "epoch": 5.030621172353456,
1977
+ "grad_norm": 1.1794829368591309,
1978
+ "learning_rate": 3.2401708494720645e-05,
1979
+ "loss": 0.0762,
1980
+ "step": 126500
1981
+ },
1982
+ {
1983
+ "epoch": 5.05050505050505,
1984
+ "grad_norm": 2.2873692512512207,
1985
+ "learning_rate": 3.173847296651987e-05,
1986
+ "loss": 0.0823,
1987
+ "step": 127000
1988
+ },
1989
+ {
1990
+ "epoch": 5.070388928656645,
1991
+ "grad_norm": 2.4998435974121094,
1992
+ "learning_rate": 3.10752374383191e-05,
1993
+ "loss": 0.0747,
1994
+ "step": 127500
1995
+ },
1996
+ {
1997
+ "epoch": 5.0902728068082395,
1998
+ "grad_norm": 3.9110782146453857,
1999
+ "learning_rate": 3.041200191011832e-05,
2000
+ "loss": 0.0853,
2001
+ "step": 128000
2002
+ },
2003
+ {
2004
+ "epoch": 5.1101566849598345,
2005
+ "grad_norm": 3.3292062282562256,
2006
+ "learning_rate": 2.9748766381917546e-05,
2007
+ "loss": 0.0839,
2008
+ "step": 128500
2009
+ },
2010
+ {
2011
+ "epoch": 5.13004056311143,
2012
+ "grad_norm": 4.433184623718262,
2013
+ "learning_rate": 2.9085530853716776e-05,
2014
+ "loss": 0.0802,
2015
+ "step": 129000
2016
+ },
2017
+ {
2018
+ "epoch": 5.149924441263024,
2019
+ "grad_norm": 3.4676051139831543,
2020
+ "learning_rate": 2.8422295325515995e-05,
2021
+ "loss": 0.0838,
2022
+ "step": 129500
2023
+ },
2024
+ {
2025
+ "epoch": 5.169808319414619,
2026
+ "grad_norm": 1.5169734954833984,
2027
+ "learning_rate": 2.7759059797315225e-05,
2028
+ "loss": 0.0783,
2029
+ "step": 130000
2030
+ },
2031
+ {
2032
+ "epoch": 5.189692197566213,
2033
+ "grad_norm": 1.637904405593872,
2034
+ "learning_rate": 2.709582426911445e-05,
2035
+ "loss": 0.0814,
2036
+ "step": 130500
2037
+ },
2038
+ {
2039
+ "epoch": 5.200827169331106,
2040
+ "eval_loss": 0.141804039478302,
2041
+ "eval_runtime": 32.6973,
2042
+ "eval_samples_per_second": 15.292,
2043
+ "eval_steps_per_second": 7.646,
2044
+ "step": 130780
2045
+ },
2046
+ {
2047
+ "epoch": 5.209576075717808,
2048
+ "grad_norm": 2.086423873901367,
2049
+ "learning_rate": 2.643391521197008e-05,
2050
+ "loss": 0.0831,
2051
+ "step": 131000
2052
+ },
2053
+ {
2054
+ "epoch": 5.229459953869402,
2055
+ "grad_norm": 0.6672555804252625,
2056
+ "learning_rate": 2.57720061548257e-05,
2057
+ "loss": 0.0862,
2058
+ "step": 131500
2059
+ },
2060
+ {
2061
+ "epoch": 5.2493438320209975,
2062
+ "grad_norm": 0.04646410793066025,
2063
+ "learning_rate": 2.510877062662493e-05,
2064
+ "loss": 0.0792,
2065
+ "step": 132000
2066
+ },
2067
+ {
2068
+ "epoch": 5.269227710172592,
2069
+ "grad_norm": 4.182920932769775,
2070
+ "learning_rate": 2.4445535098424155e-05,
2071
+ "loss": 0.0802,
2072
+ "step": 132500
2073
+ },
2074
+ {
2075
+ "epoch": 5.289111588324187,
2076
+ "grad_norm": 3.0408833026885986,
2077
+ "learning_rate": 2.3782299570223378e-05,
2078
+ "loss": 0.0836,
2079
+ "step": 133000
2080
+ },
2081
+ {
2082
+ "epoch": 5.308995466475781,
2083
+ "grad_norm": 1.9850901365280151,
2084
+ "learning_rate": 2.3120390513079006e-05,
2085
+ "loss": 0.0779,
2086
+ "step": 133500
2087
+ },
2088
+ {
2089
+ "epoch": 5.328879344627376,
2090
+ "grad_norm": 2.1532881259918213,
2091
+ "learning_rate": 2.2457154984878232e-05,
2092
+ "loss": 0.0813,
2093
+ "step": 134000
2094
+ },
2095
+ {
2096
+ "epoch": 5.348763222778971,
2097
+ "grad_norm": 2.2527785301208496,
2098
+ "learning_rate": 2.1793919456677455e-05,
2099
+ "loss": 0.0761,
2100
+ "step": 134500
2101
+ },
2102
+ {
2103
+ "epoch": 5.368647100930565,
2104
+ "grad_norm": 1.829942226409912,
2105
+ "learning_rate": 2.1130683928476684e-05,
2106
+ "loss": 0.0813,
2107
+ "step": 135000
2108
+ },
2109
+ {
2110
+ "epoch": 5.3885309790821605,
2111
+ "grad_norm": 2.5190794467926025,
2112
+ "learning_rate": 2.046877487133231e-05,
2113
+ "loss": 0.086,
2114
+ "step": 135500
2115
+ },
2116
+ {
2117
+ "epoch": 5.400858983536149,
2118
+ "eval_loss": 0.1405518800020218,
2119
+ "eval_runtime": 32.4801,
2120
+ "eval_samples_per_second": 15.394,
2121
+ "eval_steps_per_second": 7.697,
2122
+ "step": 135810
2123
+ },
2124
+ {
2125
+ "epoch": 5.408414857233755,
2126
+ "grad_norm": 0.9381058812141418,
2127
+ "learning_rate": 1.9805539343131535e-05,
2128
+ "loss": 0.0769,
2129
+ "step": 136000
2130
+ },
2131
+ {
2132
+ "epoch": 5.42829873538535,
2133
+ "grad_norm": 3.986593246459961,
2134
+ "learning_rate": 1.914230381493076e-05,
2135
+ "loss": 0.08,
2136
+ "step": 136500
2137
+ },
2138
+ {
2139
+ "epoch": 5.448182613536944,
2140
+ "grad_norm": 1.4495145082473755,
2141
+ "learning_rate": 1.848039475778639e-05,
2142
+ "loss": 0.0787,
2143
+ "step": 137000
2144
+ },
2145
+ {
2146
+ "epoch": 5.468066491688539,
2147
+ "grad_norm": 2.496209144592285,
2148
+ "learning_rate": 1.781715922958561e-05,
2149
+ "loss": 0.0808,
2150
+ "step": 137500
2151
+ },
2152
+ {
2153
+ "epoch": 5.487950369840133,
2154
+ "grad_norm": 1.4230667352676392,
2155
+ "learning_rate": 1.7153923701384837e-05,
2156
+ "loss": 0.0818,
2157
+ "step": 138000
2158
+ },
2159
+ {
2160
+ "epoch": 5.507834247991728,
2161
+ "grad_norm": 0.07668805867433548,
2162
+ "learning_rate": 1.6490688173184063e-05,
2163
+ "loss": 0.0786,
2164
+ "step": 138500
2165
+ },
2166
+ {
2167
+ "epoch": 5.5277181261433235,
2168
+ "grad_norm": 2.7259907722473145,
2169
+ "learning_rate": 1.5827452644983286e-05,
2170
+ "loss": 0.0839,
2171
+ "step": 139000
2172
+ },
2173
+ {
2174
+ "epoch": 5.547602004294918,
2175
+ "grad_norm": 1.5834600925445557,
2176
+ "learning_rate": 1.5164217116782512e-05,
2177
+ "loss": 0.0807,
2178
+ "step": 139500
2179
+ },
2180
+ {
2181
+ "epoch": 5.567485882446512,
2182
+ "grad_norm": 1.825166940689087,
2183
+ "learning_rate": 1.4500981588581738e-05,
2184
+ "loss": 0.0786,
2185
+ "step": 140000
2186
+ },
2187
+ {
2188
+ "epoch": 5.587369760598107,
2189
+ "grad_norm": 0.33820098638534546,
2190
+ "learning_rate": 1.3837746060380963e-05,
2191
+ "loss": 0.0731,
2192
+ "step": 140500
2193
+ },
2194
+ {
2195
+ "epoch": 5.600890797741192,
2196
+ "eval_loss": 0.1415776163339615,
2197
+ "eval_runtime": 32.1699,
2198
+ "eval_samples_per_second": 15.542,
2199
+ "eval_steps_per_second": 7.771,
2200
+ "step": 140840
2201
+ },
2202
+ {
2203
+ "epoch": 5.607253638749702,
2204
+ "grad_norm": 0.2709919214248657,
2205
+ "learning_rate": 1.317583700323659e-05,
2206
+ "loss": 0.0755,
2207
+ "step": 141000
2208
+ },
2209
+ {
2210
+ "epoch": 5.627137516901296,
2211
+ "grad_norm": 2.084345817565918,
2212
+ "learning_rate": 1.2512601475035815e-05,
2213
+ "loss": 0.0784,
2214
+ "step": 141500
2215
+ },
2216
+ {
2217
+ "epoch": 5.647021395052891,
2218
+ "grad_norm": 1.5787373781204224,
2219
+ "learning_rate": 1.1849365946835041e-05,
2220
+ "loss": 0.0724,
2221
+ "step": 142000
2222
+ },
2223
+ {
2224
+ "epoch": 5.666905273204486,
2225
+ "grad_norm": 2.0165510177612305,
2226
+ "learning_rate": 1.1186130418634267e-05,
2227
+ "loss": 0.0814,
2228
+ "step": 142500
2229
+ },
2230
+ {
2231
+ "epoch": 5.686789151356081,
2232
+ "grad_norm": 1.0443742275238037,
2233
+ "learning_rate": 1.052289489043349e-05,
2234
+ "loss": 0.0749,
2235
+ "step": 143000
2236
+ },
2237
+ {
2238
+ "epoch": 5.706673029507675,
2239
+ "grad_norm": 0.3988170921802521,
2240
+ "learning_rate": 9.859659362232716e-06,
2241
+ "loss": 0.0805,
2242
+ "step": 143500
2243
+ },
2244
+ {
2245
+ "epoch": 5.72655690765927,
2246
+ "grad_norm": 1.6842561960220337,
2247
+ "learning_rate": 9.196423834031942e-06,
2248
+ "loss": 0.0859,
2249
+ "step": 144000
2250
+ },
2251
+ {
2252
+ "epoch": 5.746440785810865,
2253
+ "grad_norm": 3.893829584121704,
2254
+ "learning_rate": 8.533188305831168e-06,
2255
+ "loss": 0.0834,
2256
+ "step": 144500
2257
+ },
2258
+ {
2259
+ "epoch": 5.766324663962459,
2260
+ "grad_norm": 3.017742156982422,
2261
+ "learning_rate": 7.871279248686794e-06,
2262
+ "loss": 0.0767,
2263
+ "step": 145000
2264
+ },
2265
+ {
2266
+ "epoch": 5.786208542114053,
2267
+ "grad_norm": 1.0755605697631836,
2268
+ "learning_rate": 7.2080437204860195e-06,
2269
+ "loss": 0.0807,
2270
+ "step": 145500
2271
+ },
2272
+ {
2273
+ "epoch": 5.800922611946234,
2274
+ "eval_loss": 0.14066436886787415,
2275
+ "eval_runtime": 37.1151,
2276
+ "eval_samples_per_second": 13.472,
2277
+ "eval_steps_per_second": 6.736,
2278
+ "step": 145870
2279
+ },
2280
+ {
2281
+ "epoch": 5.8060924202656485,
2282
+ "grad_norm": 2.315115451812744,
2283
+ "learning_rate": 6.544808192285244e-06,
2284
+ "loss": 0.0737,
2285
+ "step": 146000
2286
+ },
2287
+ {
2288
+ "epoch": 5.825976298417244,
2289
+ "grad_norm": 1.7753525972366333,
2290
+ "learning_rate": 5.88157266408447e-06,
2291
+ "loss": 0.0781,
2292
+ "step": 146500
2293
+ },
2294
+ {
2295
+ "epoch": 5.845860176568838,
2296
+ "grad_norm": 0.942420244216919,
2297
+ "learning_rate": 5.218337135883695e-06,
2298
+ "loss": 0.0785,
2299
+ "step": 147000
2300
+ },
2301
+ {
2302
+ "epoch": 5.865744054720433,
2303
+ "grad_norm": 1.6908499002456665,
2304
+ "learning_rate": 4.5551016076829205e-06,
2305
+ "loss": 0.0773,
2306
+ "step": 147500
2307
+ },
2308
+ {
2309
+ "epoch": 5.885627932872027,
2310
+ "grad_norm": 3.541743755340576,
2311
+ "learning_rate": 3.891866079482146e-06,
2312
+ "loss": 0.082,
2313
+ "step": 148000
2314
+ },
2315
+ {
2316
+ "epoch": 5.905511811023622,
2317
+ "grad_norm": 1.41182541847229,
2318
+ "learning_rate": 3.228630551281371e-06,
2319
+ "loss": 0.0755,
2320
+ "step": 148500
2321
+ },
2322
+ {
2323
+ "epoch": 5.925395689175216,
2324
+ "grad_norm": 1.457543134689331,
2325
+ "learning_rate": 2.5653950230805963e-06,
2326
+ "loss": 0.0721,
2327
+ "step": 149000
2328
+ },
2329
+ {
2330
+ "epoch": 5.9452795673268115,
2331
+ "grad_norm": 1.3256011009216309,
2332
+ "learning_rate": 1.9034859659362233e-06,
2333
+ "loss": 0.0756,
2334
+ "step": 149500
2335
+ },
2336
+ {
2337
+ "epoch": 5.965163445478407,
2338
+ "grad_norm": 4.112858295440674,
2339
+ "learning_rate": 1.2402504377354487e-06,
2340
+ "loss": 0.0792,
2341
+ "step": 150000
2342
+ },
2343
+ {
2344
+ "epoch": 5.985047323630001,
2345
+ "grad_norm": 9.184978485107422,
2346
+ "learning_rate": 5.77014909534674e-07,
2347
+ "loss": 0.0763,
2348
+ "step": 150500
2349
+ },
2350
+ {
2351
+ "epoch": 6.0,
2352
+ "step": 150876,
2353
+ "total_flos": 9.584515856152166e+17,
2354
+ "train_loss": 0.1514410440879876,
2355
+ "train_runtime": 79770.8244,
2356
+ "train_samples_per_second": 3.783,
2357
+ "train_steps_per_second": 1.891
2358
+ }
2359
+ ],
2360
+ "logging_steps": 500,
2361
+ "max_steps": 150876,
2362
+ "num_input_tokens_seen": 0,
2363
+ "num_train_epochs": 6,
2364
+ "save_steps": 500,
2365
+ "stateful_callbacks": {
2366
+ "TrainerControl": {
2367
+ "args": {
2368
+ "should_epoch_stop": false,
2369
+ "should_evaluate": false,
2370
+ "should_log": false,
2371
+ "should_save": true,
2372
+ "should_training_stop": true
2373
+ },
2374
+ "attributes": {}
2375
+ }
2376
+ },
2377
+ "total_flos": 9.584515856152166e+17,
2378
+ "train_batch_size": 2,
2379
+ "trial_name": null,
2380
+ "trial_params": null
2381
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e44c288a8a1d75715797591278118531f9151cb4d40d3f4f556b002e2023aa33
3
+ size 5496