Hugging Face's logo

aubmindlab
/
bert-large-arabertv02-twitter

Fill-Mask
Transformers
PyTorch
TensorBoard
Safetensors
Arabic
bert
Inference Endpoints
Model card Files Metrics Community
1
wissamantoun commited on
Commit
707adf3
•
1 Parent(s): 13c5c6f

First commit

Browse files
Files changed (19) hide show
  1. arabert-tweet-large/all_results.json +14 -0
  2. arabert-tweet-large/config.json +26 -0
  3. arabert-tweet-large/eval_results.json +9 -0
  4. arabert-tweet-large/pytorch_model.bin +3 -0
  5. arabert-tweet-large/runs/1631980876.2870846/events.out.tfevents.1631980876.Wissam-PC.17448.1 +3 -0
  6. arabert-tweet-large/runs/1632039503.4908214/events.out.tfevents.1632039503.Wissam-PC.23248.1 +3 -0
  7. arabert-tweet-large/runs/1632064111.915584/events.out.tfevents.1632064111.Wissam-PC.25020.1 +3 -0
  8. arabert-tweet-large/runs/1632080429.9502525/events.out.tfevents.1632080429.Wissam-PC.24792.1 +3 -0
  9. arabert-tweet-large/runs/events.out.tfevents.1631980876.Wissam-PC.17448.0 +3 -0
  10. arabert-tweet-large/runs/events.out.tfevents.1632039503.Wissam-PC.23248.0 +3 -0
  11. arabert-tweet-large/runs/events.out.tfevents.1632064111.Wissam-PC.25020.0 +3 -0
  12. arabert-tweet-large/runs/events.out.tfevents.1632080429.Wissam-PC.24792.0 +3 -0
  13. arabert-tweet-large/runs/events.out.tfevents.1632515635.Wissam-PC.24792.2 +3 -0
  14. arabert-tweet-large/special_tokens_map.json +1 -0
  15. arabert-tweet-large/tokenizer_config.json +1 -0
  16. arabert-tweet-large/train_results.json +8 -0
  17. arabert-tweet-large/trainer_state.json +3339 -0
  18. arabert-tweet-large/training_args.bin +3 -0
  19. arabert-tweet-large/vocab.txt +0 -0
arabert-tweet-large/all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 1.9246433973312378,
4
+ "eval_runtime": 282.1545,
5
+ "eval_samples": 100000,
6
+ "eval_samples_per_second": 354.416,
7
+ "eval_steps_per_second": 44.302,
8
+ "perplexity": 6.852704537407429,
9
+ "train_loss": 1.8196985500431868,
10
+ "train_runtime": 434920.5349,
11
+ "train_samples": 64750662,
12
+ "train_samples_per_second": 148.879,
13
+ "train_steps_per_second": 0.62
14
+ }
arabert-tweet-large/config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "models/bert-large-arabertv02",
3
+ "architectures": [
4
+ "BertForMaskedLM"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 1024,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 4096,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 16,
18
+ "num_hidden_layers": 24,
19
+ "pad_token_id": 0,
20
+ "position_embedding_type": "absolute",
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.10.0",
23
+ "type_vocab_size": 2,
24
+ "use_cache": true,
25
+ "vocab_size": 64000
26
+ }
arabert-tweet-large/eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 1.9246433973312378,
4
+ "eval_runtime": 282.1545,
5
+ "eval_samples": 100000,
6
+ "eval_samples_per_second": 354.416,
7
+ "eval_steps_per_second": 44.302,
8
+ "perplexity": 6.852704537407429
9
+ }
arabert-tweet-large/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd509bec281aee5a6608353f2e7040ea9ae82620f6a66a0eeb5571d6b260e62a
3
+ size 1478119979
arabert-tweet-large/runs/1631980876.2870846/events.out.tfevents.1631980876.Wissam-PC.17448.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0413cde01ed45dfe53ff2fee63486996fc3f3d996f00535ef417da5ba402f48a
3
+ size 4156
arabert-tweet-large/runs/1632039503.4908214/events.out.tfevents.1632039503.Wissam-PC.23248.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad35064440853a44fe959d6e1253045ede3931b1b82a32342a4c165d1bd34a1d
3
+ size 4156
arabert-tweet-large/runs/1632064111.915584/events.out.tfevents.1632064111.Wissam-PC.25020.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ce0c61d8a131295f2d444a0b176461cd145b07c379b9ca389860c7fb769a8a45
3
+ size 4156
arabert-tweet-large/runs/1632080429.9502525/events.out.tfevents.1632080429.Wissam-PC.24792.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f135c1d7b0aa38b2ee0ebe780115b89cfafc7c3f1d44f6b50e6df70dacde508b
3
+ size 4156
arabert-tweet-large/runs/events.out.tfevents.1631980876.Wissam-PC.17448.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f1df2add2d772c948614e701573a925440fa872f2d0406ff2b0bfcc1d16eedff
3
+ size 19545
arabert-tweet-large/runs/events.out.tfevents.1632039503.Wissam-PC.23248.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7b21b03829c0b975aea497deb2171c6b1f1162306f8ab311e0e8fc75dca389a2
3
+ size 8049
arabert-tweet-large/runs/events.out.tfevents.1632064111.Wissam-PC.25020.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26eae1eca59833cbf67a0b8b6be8305d54f34cb90403fdf0bd1dd25f8f91a337
3
+ size 5329
arabert-tweet-large/runs/events.out.tfevents.1632080429.Wissam-PC.24792.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84183a9198eeea54d8a68b760c7558b751fa3f65652c2832eeee27462ce4b9fd
3
+ size 75739
arabert-tweet-large/runs/events.out.tfevents.1632515635.Wissam-PC.24792.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3da0407971c13552b5f0defdb5446d2cb5a35030718e7b8e0a40fae9b36f7fd
3
+ size 316
arabert-tweet-large/special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
1
+ {"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
arabert-tweet-large/tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
1
+ {"do_lower_case": false, "do_basic_tokenize": true, "never_split": ["[بريد]", "[مستخدم]", "[رابط]"], "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "max_len": 512, "special_tokens_map_file": null, "name_or_path": "models/bert-large-arabertv02", "use_fast": true, "tokenizer_file": null, "tokenizer_class": "BertTokenizer"}
arabert-tweet-large/train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 1.8196985500431868,
4
+ "train_runtime": 434920.5349,
5
+ "train_samples": 64750662,
6
+ "train_samples_per_second": 148.879,
7
+ "train_steps_per_second": 0.62
8
+ }
arabert-tweet-large/trainer_state.json ADDED
@@ -0,0 +1,3339 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9999981467376096,
5
+ "global_step": 269794,
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": 1.446997776130467e-06,
13
+ "loss": 4.2339,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 2.9295774647887324e-06,
19
+ "loss": 3.743,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.0,
24
+ "learning_rate": 4.4121571534469985e-06,
25
+ "loss": 3.5488,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 5.8947368421052634e-06,
31
+ "loss": 3.441,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 7.377316530763528e-06,
37
+ "loss": 3.3861,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 8.859896219421795e-06,
43
+ "loss": 3.3338,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.01,
48
+ "learning_rate": 1.034247590808006e-05,
49
+ "loss": 3.2739,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.01,
54
+ "learning_rate": 1.1825055596738326e-05,
55
+ "loss": 3.2384,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.01,
60
+ "learning_rate": 1.3307635285396591e-05,
61
+ "loss": 3.1955,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.01,
66
+ "learning_rate": 1.4790214974054856e-05,
67
+ "loss": 3.1733,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.02,
72
+ "learning_rate": 1.6272794662713124e-05,
73
+ "loss": 3.1648,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.02,
78
+ "learning_rate": 1.7755374351371386e-05,
79
+ "loss": 3.1262,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.02,
84
+ "learning_rate": 1.923795404002965e-05,
85
+ "loss": 3.112,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.02,
90
+ "learning_rate": 2.072053372868792e-05,
91
+ "loss": 3.0858,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.02,
96
+ "learning_rate": 2.2203113417346184e-05,
97
+ "loss": 3.0709,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.02,
102
+ "learning_rate": 2.368569310600445e-05,
103
+ "loss": 3.0446,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.03,
108
+ "learning_rate": 2.516827279466271e-05,
109
+ "loss": 3.0337,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.03,
114
+ "learning_rate": 2.6650852483320982e-05,
115
+ "loss": 3.0176,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.03,
120
+ "learning_rate": 2.8133432171979247e-05,
121
+ "loss": 2.9892,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.03,
126
+ "learning_rate": 2.9613046701260195e-05,
127
+ "loss": 2.9966,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.03,
132
+ "learning_rate": 3.1095626389918456e-05,
133
+ "loss": 2.9685,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.03,
138
+ "learning_rate": 3.257820607857672e-05,
139
+ "loss": 2.9461,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.03,
144
+ "learning_rate": 3.406078576723499e-05,
145
+ "loss": 2.9472,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.04,
150
+ "learning_rate": 3.554336545589326e-05,
151
+ "loss": 2.9346,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.04,
156
+ "learning_rate": 3.7022979985174206e-05,
157
+ "loss": 2.9418,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.04,
162
+ "learning_rate": 3.850555967383247e-05,
163
+ "loss": 2.9088,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.04,
168
+ "learning_rate": 3.9988139362490736e-05,
169
+ "loss": 2.8908,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.04,
174
+ "learning_rate": 4.147071905114901e-05,
175
+ "loss": 2.8886,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.04,
180
+ "learning_rate": 4.2953298739807265e-05,
181
+ "loss": 2.8681,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.04,
186
+ "learning_rate": 4.4432913269088213e-05,
187
+ "loss": 2.8626,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.05,
192
+ "learning_rate": 4.591549295774648e-05,
193
+ "loss": 2.8519,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.05,
198
+ "learning_rate": 4.739807264640474e-05,
199
+ "loss": 2.8402,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.05,
204
+ "learning_rate": 4.888065233506301e-05,
205
+ "loss": 2.8259,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.05,
210
+ "learning_rate": 5.036323202372127e-05,
211
+ "loss": 2.836,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.05,
216
+ "learning_rate": 5.1845811712379545e-05,
217
+ "loss": 2.8166,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.05,
222
+ "learning_rate": 5.3328391401037816e-05,
223
+ "loss": 2.796,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.05,
228
+ "learning_rate": 5.4810971089696074e-05,
229
+ "loss": 2.7931,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.06,
234
+ "learning_rate": 5.629058561897702e-05,
235
+ "loss": 2.7947,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.06,
240
+ "learning_rate": 5.777020014825797e-05,
241
+ "loss": 2.7701,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.06,
246
+ "learning_rate": 5.925277983691624e-05,
247
+ "loss": 2.7674,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.06,
252
+ "learning_rate": 6.07353595255745e-05,
253
+ "loss": 2.7625,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.06,
258
+ "learning_rate": 6.221793921423277e-05,
259
+ "loss": 2.754,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.06,
264
+ "learning_rate": 6.370051890289104e-05,
265
+ "loss": 2.7621,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.07,
270
+ "learning_rate": 6.51830985915493e-05,
271
+ "loss": 2.7329,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.07,
276
+ "learning_rate": 6.666567828020757e-05,
277
+ "loss": 2.7392,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.07,
282
+ "learning_rate": 6.814825796886583e-05,
283
+ "loss": 2.7446,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.07,
288
+ "learning_rate": 6.96308376575241e-05,
289
+ "loss": 2.7346,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.07,
294
+ "learning_rate": 7.111341734618236e-05,
295
+ "loss": 2.7246,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.07,
300
+ "learning_rate": 7.259599703484063e-05,
301
+ "loss": 2.7393,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.07,
306
+ "learning_rate": 7.407857672349889e-05,
307
+ "loss": 2.7271,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.07,
312
+ "eval_loss": 2.5492873191833496,
313
+ "eval_runtime": 278.1602,
314
+ "eval_samples_per_second": 359.505,
315
+ "eval_steps_per_second": 44.938,
316
+ "step": 25000
317
+ },
318
+ {
319
+ "epoch": 0.08,
320
+ "learning_rate": 7.556115641215716e-05,
321
+ "loss": 2.7127,
322
+ "step": 25500
323
+ },
324
+ {
325
+ "epoch": 0.08,
326
+ "learning_rate": 7.704373610081542e-05,
327
+ "loss": 2.713,
328
+ "step": 26000
329
+ },
330
+ {
331
+ "epoch": 0.08,
332
+ "learning_rate": 7.852631578947369e-05,
333
+ "loss": 2.7002,
334
+ "step": 26500
335
+ },
336
+ {
337
+ "epoch": 0.08,
338
+ "learning_rate": 8.000593031875464e-05,
339
+ "loss": 2.7042,
340
+ "step": 27000
341
+ },
342
+ {
343
+ "epoch": 0.08,
344
+ "learning_rate": 8.14885100074129e-05,
345
+ "loss": 2.7054,
346
+ "step": 27500
347
+ },
348
+ {
349
+ "epoch": 0.08,
350
+ "learning_rate": 8.297108969607117e-05,
351
+ "loss": 2.6925,
352
+ "step": 28000
353
+ },
354
+ {
355
+ "epoch": 0.08,
356
+ "learning_rate": 8.445366938472942e-05,
357
+ "loss": 2.6927,
358
+ "step": 28500
359
+ },
360
+ {
361
+ "epoch": 0.09,
362
+ "learning_rate": 8.59362490733877e-05,
363
+ "loss": 2.6843,
364
+ "step": 29000
365
+ },
366
+ {
367
+ "epoch": 0.09,
368
+ "learning_rate": 8.741882876204596e-05,
369
+ "loss": 2.6848,
370
+ "step": 29500
371
+ },
372
+ {
373
+ "epoch": 0.09,
374
+ "learning_rate": 8.890140845070423e-05,
375
+ "loss": 2.6851,
376
+ "step": 30000
377
+ },
378
+ {
379
+ "epoch": 0.09,
380
+ "learning_rate": 9.03839881393625e-05,
381
+ "loss": 2.676,
382
+ "step": 30500
383
+ },
384
+ {
385
+ "epoch": 0.09,
386
+ "learning_rate": 9.186656782802076e-05,
387
+ "loss": 2.6811,
388
+ "step": 31000
389
+ },
390
+ {
391
+ "epoch": 0.09,
392
+ "learning_rate": 9.334914751667902e-05,
393
+ "loss": 2.6687,
394
+ "step": 31500
395
+ },
396
+ {
397
+ "epoch": 0.09,
398
+ "learning_rate": 9.483172720533729e-05,
399
+ "loss": 2.6751,
400
+ "step": 32000
401
+ },
402
+ {
403
+ "epoch": 0.1,
404
+ "learning_rate": 9.631430689399555e-05,
405
+ "loss": 2.6746,
406
+ "step": 32500
407
+ },
408
+ {
409
+ "epoch": 0.1,
410
+ "learning_rate": 9.779688658265382e-05,
411
+ "loss": 2.6662,
412
+ "step": 33000
413
+ },
414
+ {
415
+ "epoch": 0.1,
416
+ "learning_rate": 9.927946627131208e-05,
417
+ "loss": 2.6685,
418
+ "step": 33500
419
+ },
420
+ {
421
+ "epoch": 0.1,
422
+ "learning_rate": 9.999982309633072e-05,
423
+ "loss": 2.6702,
424
+ "step": 34000
425
+ },
426
+ {
427
+ "epoch": 0.1,
428
+ "learning_rate": 9.999846516961786e-05,
429
+ "loss": 2.6692,
430
+ "step": 34500
431
+ },
432
+ {
433
+ "epoch": 0.1,
434
+ "learning_rate": 9.999577483170183e-05,
435
+ "loss": 2.6732,
436
+ "step": 35000
437
+ },
438
+ {
439
+ "epoch": 0.11,
440
+ "learning_rate": 9.9991750774992e-05,
441
+ "loss": 2.6676,
442
+ "step": 35500
443
+ },
444
+ {
445
+ "epoch": 0.11,
446
+ "learning_rate": 9.998638103115459e-05,
447
+ "loss": 2.65,
448
+ "step": 36000
449
+ },
450
+ {
451
+ "epoch": 0.11,
452
+ "learning_rate": 9.997967246921306e-05,
453
+ "loss": 2.6493,
454
+ "step": 36500
455
+ },
456
+ {
457
+ "epoch": 0.11,
458
+ "learning_rate": 9.997162526884725e-05,
459
+ "loss": 2.657,
460
+ "step": 37000
461
+ },
462
+ {
463
+ "epoch": 0.11,
464
+ "learning_rate": 9.996223964559059e-05,
465
+ "loss": 2.6521,
466
+ "step": 37500
467
+ },
468
+ {
469
+ "epoch": 0.11,
470
+ "learning_rate": 9.995151585082442e-05,
471
+ "loss": 2.6412,
472
+ "step": 38000
473
+ },
474
+ {
475
+ "epoch": 0.11,
476
+ "learning_rate": 9.993947963012904e-05,
477
+ "loss": 2.6418,
478
+ "step": 38500
479
+ },
480
+ {
481
+ "epoch": 0.12,
482
+ "learning_rate": 9.992608306462086e-05,
483
+ "loss": 2.6442,
484
+ "step": 39000
485
+ },
486
+ {
487
+ "epoch": 0.12,
488
+ "learning_rate": 9.991134929600863e-05,
489
+ "loss": 2.6362,
490
+ "step": 39500
491
+ },
492
+ {
493
+ "epoch": 0.12,
494
+ "learning_rate": 9.989527871891659e-05,
495
+ "loss": 2.6354,
496
+ "step": 40000
497
+ },
498
+ {
499
+ "epoch": 0.12,
500
+ "learning_rate": 9.987787176377355e-05,
501
+ "loss": 2.6167,
502
+ "step": 40500
503
+ },
504
+ {
505
+ "epoch": 0.12,
506
+ "learning_rate": 9.985912889680143e-05,
507
+ "loss": 2.6316,
508
+ "step": 41000
509
+ },
510
+ {
511
+ "epoch": 0.12,
512
+ "learning_rate": 9.983905062000275e-05,
513
+ "loss": 2.6355,
514
+ "step": 41500
515
+ },
516
+ {
517
+ "epoch": 0.12,
518
+ "learning_rate": 9.981763747114724e-05,
519
+ "loss": 2.6183,
520
+ "step": 42000
521
+ },
522
+ {
523
+ "epoch": 0.13,
524
+ "learning_rate": 9.979493684988302e-05,
525
+ "loss": 2.6111,
526
+ "step": 42500
527
+ },
528
+ {
529
+ "epoch": 0.13,
530
+ "learning_rate": 9.977085837996527e-05,
531
+ "loss": 2.6053,
532
+ "step": 43000
533
+ },
534
+ {
535
+ "epoch": 0.13,
536
+ "learning_rate": 9.97454990173924e-05,
537
+ "loss": 2.6019,
538
+ "step": 43500
539
+ },
540
+ {
541
+ "epoch": 0.13,
542
+ "learning_rate": 9.971875780087649e-05,
543
+ "loss": 2.6166,
544
+ "step": 44000
545
+ },
546
+ {
547
+ "epoch": 0.13,
548
+ "learning_rate": 9.969074240841182e-05,
549
+ "loss": 2.6378,
550
+ "step": 44500
551
+ },
552
+ {
553
+ "epoch": 0.13,
554
+ "learning_rate": 9.966134130446202e-05,
555
+ "loss": 2.6069,
556
+ "step": 45000
557
+ },
558
+ {
559
+ "epoch": 0.13,
560
+ "learning_rate": 9.963067287750436e-05,
561
+ "loss": 2.6201,
562
+ "step": 45500
563
+ },
564
+ {
565
+ "epoch": 0.14,
566
+ "learning_rate": 9.959861502967945e-05,
567
+ "loss": 2.5973,
568
+ "step": 46000
569
+ },
570
+ {
571
+ "epoch": 0.14,
572
+ "learning_rate": 9.956522874954218e-05,
573
+ "loss": 2.5926,
574
+ "step": 46500
575
+ },
576
+ {
577
+ "epoch": 0.14,
578
+ "learning_rate": 9.95305149312992e-05,
579
+ "loss": 2.5856,
580
+ "step": 47000
581
+ },
582
+ {
583
+ "epoch": 0.14,
584
+ "learning_rate": 9.949447450471355e-05,
585
+ "loss": 2.5929,
586
+ "step": 47500
587
+ },
588
+ {
589
+ "epoch": 0.14,
590
+ "learning_rate": 9.945710843507967e-05,
591
+ "loss": 2.5851,
592
+ "step": 48000
593
+ },
594
+ {
595
+ "epoch": 0.14,
596
+ "learning_rate": 9.94184964259314e-05,
597
+ "loss": 2.5704,
598
+ "step": 48500
599
+ },
600
+ {
601
+ "epoch": 0.15,
602
+ "learning_rate": 9.9378484754234e-05,
603
+ "loss": 2.5823,
604
+ "step": 49000
605
+ },
606
+ {
607
+ "epoch": 0.15,
608
+ "learning_rate": 9.933715054611808e-05,
609
+ "loss": 2.5716,
610
+ "step": 49500
611
+ },
612
+ {
613
+ "epoch": 0.15,
614
+ "learning_rate": 9.929458153797284e-05,
615
+ "loss": 2.6217,
616
+ "step": 50000
617
+ },
618
+ {
619
+ "epoch": 0.15,
620
+ "eval_loss": 2.431061267852783,
621
+ "eval_runtime": 279.1709,
622
+ "eval_samples_per_second": 358.204,
623
+ "eval_steps_per_second": 44.775,
624
+ "step": 50000
625
+ },
626
+ {
627
+ "epoch": 0.19,
628
+ "learning_rate": 9.77079560136976e-05,
629
+ "loss": 2.5993,
630
+ "step": 50500
631
+ },
632
+ {
633
+ "epoch": 0.19,
634
+ "learning_rate": 9.76101478820416e-05,
635
+ "loss": 2.5537,
636
+ "step": 51000
637
+ },
638
+ {
639
+ "epoch": 0.19,
640
+ "learning_rate": 9.751034728712014e-05,
641
+ "loss": 2.5581,
642
+ "step": 51500
643
+ },
644
+ {
645
+ "epoch": 0.19,
646
+ "learning_rate": 9.740855840554321e-05,
647
+ "loss": 2.6137,
648
+ "step": 52000
649
+ },
650
+ {
651
+ "epoch": 0.19,
652
+ "learning_rate": 9.730478549712973e-05,
653
+ "loss": 2.5886,
654
+ "step": 52500
655
+ },
656
+ {
657
+ "epoch": 0.2,
658
+ "learning_rate": 9.719903290472921e-05,
659
+ "loss": 2.5534,
660
+ "step": 53000
661
+ },
662
+ {
663
+ "epoch": 0.2,
664
+ "learning_rate": 9.709130505404002e-05,
665
+ "loss": 2.5475,
666
+ "step": 53500
667
+ },
668
+ {
669
+ "epoch": 0.2,
670
+ "learning_rate": 9.698160645342426e-05,
671
+ "loss": 2.5613,
672
+ "step": 54000
673
+ },
674
+ {
675
+ "epoch": 0.2,
676
+ "learning_rate": 9.686994169371903e-05,
677
+ "loss": 2.5487,
678
+ "step": 54500
679
+ },
680
+ {
681
+ "epoch": 0.2,
682
+ "learning_rate": 9.675631544804424e-05,
683
+ "loss": 2.5659,
684
+ "step": 55000
685
+ },
686
+ {
687
+ "epoch": 0.21,
688
+ "learning_rate": 9.664073247160717e-05,
689
+ "loss": 2.5319,
690
+ "step": 55500
691
+ },
692
+ {
693
+ "epoch": 0.21,
694
+ "learning_rate": 9.652319760150334e-05,
695
+ "loss": 2.5536,
696
+ "step": 56000
697
+ },
698
+ {
699
+ "epoch": 0.21,
700
+ "learning_rate": 9.640371575651422e-05,
701
+ "loss": 2.53,
702
+ "step": 56500
703
+ },
704
+ {
705
+ "epoch": 0.21,
706
+ "learning_rate": 9.628229193690122e-05,
707
+ "loss": 2.524,
708
+ "step": 57000
709
+ },
710
+ {
711
+ "epoch": 0.21,
712
+ "learning_rate": 9.615893122419657e-05,
713
+ "loss": 2.5219,
714
+ "step": 57500
715
+ },
716
+ {
717
+ "epoch": 0.21,
718
+ "learning_rate": 9.60336387809906e-05,
719
+ "loss": 2.51,
720
+ "step": 58000
721
+ },
722
+ {
723
+ "epoch": 0.22,
724
+ "learning_rate": 9.590641985071566e-05,
725
+ "loss": 2.5131,
726
+ "step": 58500
727
+ },
728
+ {
729
+ "epoch": 0.22,
730
+ "learning_rate": 9.577727975742674e-05,
731
+ "loss": 2.506,
732
+ "step": 59000
733
+ },
734
+ {
735
+ "epoch": 0.22,
736
+ "learning_rate": 9.564622390557863e-05,
737
+ "loss": 2.5086,
738
+ "step": 59500
739
+ },
740
+ {
741
+ "epoch": 0.22,
742
+ "learning_rate": 9.551325777979978e-05,
743
+ "loss": 2.4997,
744
+ "step": 60000
745
+ },
746
+ {
747
+ "epoch": 0.22,
748
+ "learning_rate": 9.53783869446627e-05,
749
+ "loss": 2.4888,
750
+ "step": 60500
751
+ },
752
+ {
753
+ "epoch": 0.23,
754
+ "learning_rate": 9.524161704445116e-05,
755
+ "loss": 2.4952,
756
+ "step": 61000
757
+ },
758
+ {
759
+ "epoch": 0.23,
760
+ "learning_rate": 9.510295380292393e-05,
761
+ "loss": 2.4868,
762
+ "step": 61500
763
+ },
764
+ {
765
+ "epoch": 0.23,
766
+ "learning_rate": 9.496240302307531e-05,
767
+ "loss": 2.4906,
768
+ "step": 62000
769
+ },
770
+ {
771
+ "epoch": 0.23,
772
+ "learning_rate": 9.481997058689214e-05,
773
+ "loss": 2.4815,
774
+ "step": 62500
775
+ },
776
+ {
777
+ "epoch": 0.23,
778
+ "learning_rate": 9.467566245510782e-05,
779
+ "loss": 2.4772,
780
+ "step": 63000
781
+ },
782
+ {
783
+ "epoch": 0.24,
784
+ "learning_rate": 9.45294846669527e-05,
785
+ "loss": 2.4919,
786
+ "step": 63500
787
+ },
788
+ {
789
+ "epoch": 0.24,
790
+ "learning_rate": 9.438144333990148e-05,
791
+ "loss": 2.4889,
792
+ "step": 64000
793
+ },
794
+ {
795
+ "epoch": 0.24,
796
+ "learning_rate": 9.423154466941704e-05,
797
+ "loss": 2.4811,
798
+ "step": 64500
799
+ },
800
+ {
801
+ "epoch": 0.24,
802
+ "learning_rate": 9.407979492869132e-05,
803
+ "loss": 2.4721,
804
+ "step": 65000
805
+ },
806
+ {
807
+ "epoch": 0.24,
808
+ "learning_rate": 9.392620046838267e-05,
809
+ "loss": 2.4696,
810
+ "step": 65500
811
+ },
812
+ {
813
+ "epoch": 0.24,
814
+ "learning_rate": 9.377076771635013e-05,
815
+ "loss": 2.4739,
816
+ "step": 66000
817
+ },
818
+ {
819
+ "epoch": 0.25,
820
+ "learning_rate": 9.361350317738446e-05,
821
+ "loss": 2.4704,
822
+ "step": 66500
823
+ },
824
+ {
825
+ "epoch": 0.25,
826
+ "learning_rate": 9.345441343293581e-05,
827
+ "loss": 2.4727,
828
+ "step": 67000
829
+ },
830
+ {
831
+ "epoch": 0.25,
832
+ "learning_rate": 9.329350514083845e-05,
833
+ "loss": 2.4649,
834
+ "step": 67500
835
+ },
836
+ {
837
+ "epoch": 0.25,
838
+ "learning_rate": 9.313078503503196e-05,
839
+ "loss": 2.4619,
840
+ "step": 68000
841
+ },
842
+ {
843
+ "epoch": 0.25,
844
+ "learning_rate": 9.296625992527957e-05,
845
+ "loss": 2.4577,
846
+ "step": 68500
847
+ },
848
+ {
849
+ "epoch": 0.26,
850
+ "learning_rate": 9.279993669688308e-05,
851
+ "loss": 2.46,
852
+ "step": 69000
853
+ },
854
+ {
855
+ "epoch": 0.26,
856
+ "learning_rate": 9.263182231039476e-05,
857
+ "loss": 2.4562,
858
+ "step": 69500
859
+ },
860
+ {
861
+ "epoch": 0.26,
862
+ "learning_rate": 9.246192380132604e-05,
863
+ "loss": 2.4431,
864
+ "step": 70000
865
+ },
866
+ {
867
+ "epoch": 0.26,
868
+ "learning_rate": 9.229024827985306e-05,
869
+ "loss": 2.4523,
870
+ "step": 70500
871
+ },
872
+ {
873
+ "epoch": 0.26,
874
+ "learning_rate": 9.211680293051915e-05,
875
+ "loss": 2.4558,
876
+ "step": 71000
877
+ },
878
+ {
879
+ "epoch": 0.27,
880
+ "learning_rate": 9.194159501193414e-05,
881
+ "loss": 2.4508,
882
+ "step": 71500
883
+ },
884
+ {
885
+ "epoch": 0.27,
886
+ "learning_rate": 9.17646318564706e-05,
887
+ "loss": 2.4526,
888
+ "step": 72000
889
+ },
890
+ {
891
+ "epoch": 0.27,
892
+ "learning_rate": 9.158592086995692e-05,
893
+ "loss": 2.4455,
894
+ "step": 72500
895
+ },
896
+ {
897
+ "epoch": 0.27,
898
+ "learning_rate": 9.140546953136748e-05,
899
+ "loss": 2.4423,
900
+ "step": 73000
901
+ },
902
+ {
903
+ "epoch": 0.27,
904
+ "learning_rate": 9.122328539250962e-05,
905
+ "loss": 2.4291,
906
+ "step": 73500
907
+ },
908
+ {
909
+ "epoch": 0.27,
910
+ "learning_rate": 9.103937607770756e-05,
911
+ "loss": 2.4521,
912
+ "step": 74000
913
+ },
914
+ {
915
+ "epoch": 0.28,
916
+ "learning_rate": 9.085374928348338e-05,
917
+ "loss": 2.4427,
918
+ "step": 74500
919
+ },
920
+ {
921
+ "epoch": 0.28,
922
+ "learning_rate": 9.066641277823487e-05,
923
+ "loss": 2.4378,
924
+ "step": 75000
925
+ },
926
+ {
927
+ "epoch": 0.28,
928
+ "eval_loss": 2.274932622909546,
929
+ "eval_runtime": 278.4521,
930
+ "eval_samples_per_second": 359.128,
931
+ "eval_steps_per_second": 44.891,
932
+ "step": 75000
933
+ },
934
+ {
935
+ "epoch": 0.28,
936
+ "learning_rate": 9.047737440191048e-05,
937
+ "loss": 2.4375,
938
+ "step": 75500
939
+ },
940
+ {
941
+ "epoch": 0.28,
942
+ "learning_rate": 9.028664206568123e-05,
943
+ "loss": 2.4262,
944
+ "step": 76000
945
+ },
946
+ {
947
+ "epoch": 0.28,
948
+ "learning_rate": 9.009422375160955e-05,
949
+ "loss": 2.4342,
950
+ "step": 76500
951
+ },
952
+ {
953
+ "epoch": 0.29,
954
+ "learning_rate": 8.990012751231527e-05,
955
+ "loss": 2.427,
956
+ "step": 77000
957
+ },
958
+ {
959
+ "epoch": 0.29,
960
+ "learning_rate": 8.970436147063869e-05,
961
+ "loss": 2.419,
962
+ "step": 77500
963
+ },
964
+ {
965
+ "epoch": 0.29,
966
+ "learning_rate": 8.950693381930058e-05,
967
+ "loss": 2.4241,
968
+ "step": 78000
969
+ },
970
+ {
971
+ "epoch": 0.29,
972
+ "learning_rate": 8.930785282055928e-05,
973
+ "loss": 2.4192,
974
+ "step": 78500
975
+ },
976
+ {
977
+ "epoch": 0.29,
978
+ "learning_rate": 8.910712680586502e-05,
979
+ "loss": 2.4179,
980
+ "step": 79000
981
+ },
982
+ {
983
+ "epoch": 0.29,
984
+ "learning_rate": 8.890476417551119e-05,
985
+ "loss": 2.4176,
986
+ "step": 79500
987
+ },
988
+ {
989
+ "epoch": 0.3,
990
+ "learning_rate": 8.870077339828284e-05,
991
+ "loss": 2.4179,
992
+ "step": 80000
993
+ },
994
+ {
995
+ "epoch": 0.3,
996
+ "learning_rate": 8.849516301110216e-05,
997
+ "loss": 2.418,
998
+ "step": 80500
999
+ },
1000
+ {
1001
+ "epoch": 0.3,
1002
+ "learning_rate": 8.828794161867136e-05,
1003
+ "loss": 2.4116,
1004
+ "step": 81000
1005
+ },
1006
+ {
1007
+ "epoch": 0.3,
1008
+ "learning_rate": 8.807911789311245e-05,
1009
+ "loss": 2.4114,
1010
+ "step": 81500
1011
+ },
1012
+ {
1013
+ "epoch": 0.3,
1014
+ "learning_rate": 8.786870057360441e-05,
1015
+ "loss": 2.4091,
1016
+ "step": 82000
1017
+ },
1018
+ {
1019
+ "epoch": 0.31,
1020
+ "learning_rate": 8.765669846601735e-05,
1021
+ "loss": 2.4125,
1022
+ "step": 82500
1023
+ },
1024
+ {
1025
+ "epoch": 0.31,
1026
+ "learning_rate": 8.744312044254409e-05,
1027
+ "loss": 2.4115,
1028
+ "step": 83000
1029
+ },
1030
+ {
1031
+ "epoch": 0.31,
1032
+ "learning_rate": 8.722797544132881e-05,
1033
+ "loss": 2.4195,
1034
+ "step": 83500
1035
+ },
1036
+ {
1037
+ "epoch": 0.31,
1038
+ "learning_rate": 8.701127246609299e-05,
1039
+ "loss": 2.4048,
1040
+ "step": 84000
1041
+ },
1042
+ {
1043
+ "epoch": 0.31,
1044
+ "learning_rate": 8.679302058575865e-05,
1045
+ "loss": 2.4075,
1046
+ "step": 84500
1047
+ },
1048
+ {
1049
+ "epoch": 0.32,
1050
+ "learning_rate": 8.657322893406876e-05,
1051
+ "loss": 2.3983,
1052
+ "step": 85000
1053
+ },
1054
+ {
1055
+ "epoch": 0.32,
1056
+ "learning_rate": 8.635190670920503e-05,
1057
+ "loss": 2.3999,
1058
+ "step": 85500
1059
+ },
1060
+ {
1061
+ "epoch": 0.32,
1062
+ "learning_rate": 8.6129063173403e-05,
1063
+ "loss": 2.4055,
1064
+ "step": 86000
1065
+ },
1066
+ {
1067
+ "epoch": 0.32,
1068
+ "learning_rate": 8.590470765256439e-05,
1069
+ "loss": 2.3899,
1070
+ "step": 86500
1071
+ },
1072
+ {
1073
+ "epoch": 0.32,
1074
+ "learning_rate": 8.567884953586675e-05,
1075
+ "loss": 2.3935,
1076
+ "step": 87000
1077
+ },
1078
+ {
1079
+ "epoch": 0.32,
1080
+ "learning_rate": 8.545149827537065e-05,
1081
+ "loss": 2.3872,
1082
+ "step": 87500
1083
+ },
1084
+ {
1085
+ "epoch": 0.33,
1086
+ "learning_rate": 8.522266338562404e-05,
1087
+ "loss": 2.3877,
1088
+ "step": 88000
1089
+ },
1090
+ {
1091
+ "epoch": 0.33,
1092
+ "learning_rate": 8.499235444326407e-05,
1093
+ "loss": 2.3939,
1094
+ "step": 88500
1095
+ },
1096
+ {
1097
+ "epoch": 0.33,
1098
+ "learning_rate": 8.476058108661639e-05,
1099
+ "loss": 2.3855,
1100
+ "step": 89000
1101
+ },
1102
+ {
1103
+ "epoch": 0.33,
1104
+ "learning_rate": 8.452735301529164e-05,
1105
+ "loss": 2.388,
1106
+ "step": 89500
1107
+ },
1108
+ {
1109
+ "epoch": 0.33,
1110
+ "learning_rate": 8.429267998977967e-05,
1111
+ "loss": 2.3801,
1112
+ "step": 90000
1113
+ },
1114
+ {
1115
+ "epoch": 0.34,
1116
+ "learning_rate": 8.4056571831041e-05,
1117
+ "loss": 2.3723,
1118
+ "step": 90500
1119
+ },
1120
+ {
1121
+ "epoch": 0.34,
1122
+ "learning_rate": 8.381903842009583e-05,
1123
+ "loss": 2.3838,
1124
+ "step": 91000
1125
+ },
1126
+ {
1127
+ "epoch": 0.34,
1128
+ "learning_rate": 8.358008969761053e-05,
1129
+ "loss": 2.3857,
1130
+ "step": 91500
1131
+ },
1132
+ {
1133
+ "epoch": 0.34,
1134
+ "learning_rate": 8.333973566348161e-05,
1135
+ "loss": 2.3678,
1136
+ "step": 92000
1137
+ },
1138
+ {
1139
+ "epoch": 0.34,
1140
+ "learning_rate": 8.309798637641725e-05,
1141
+ "loss": 2.3804,
1142
+ "step": 92500
1143
+ },
1144
+ {
1145
+ "epoch": 0.34,
1146
+ "learning_rate": 8.285485195351632e-05,
1147
+ "loss": 2.3667,
1148
+ "step": 93000
1149
+ },
1150
+ {
1151
+ "epoch": 0.35,
1152
+ "learning_rate": 8.261034256984503e-05,
1153
+ "loss": 2.3648,
1154
+ "step": 93500
1155
+ },
1156
+ {
1157
+ "epoch": 0.35,
1158
+ "learning_rate": 8.236446845801104e-05,
1159
+ "loss": 2.4219,
1160
+ "step": 94000
1161
+ },
1162
+ {
1163
+ "epoch": 0.35,
1164
+ "learning_rate": 8.211723990773533e-05,
1165
+ "loss": 2.4274,
1166
+ "step": 94500
1167
+ },
1168
+ {
1169
+ "epoch": 0.35,
1170
+ "learning_rate": 8.186866726542143e-05,
1171
+ "loss": 2.4006,
1172
+ "step": 95000
1173
+ },
1174
+ {
1175
+ "epoch": 0.35,
1176
+ "learning_rate": 8.161876093372264e-05,
1177
+ "loss": 2.3849,
1178
+ "step": 95500
1179
+ },
1180
+ {
1181
+ "epoch": 0.36,
1182
+ "learning_rate": 8.136753137110643e-05,
1183
+ "loss": 2.3937,
1184
+ "step": 96000
1185
+ },
1186
+ {
1187
+ "epoch": 0.36,
1188
+ "learning_rate": 8.111498909141696e-05,
1189
+ "loss": 2.3867,
1190
+ "step": 96500
1191
+ },
1192
+ {
1193
+ "epoch": 0.36,
1194
+ "learning_rate": 8.086114466343502e-05,
1195
+ "loss": 2.3914,
1196
+ "step": 97000
1197
+ },
1198
+ {
1199
+ "epoch": 0.36,
1200
+ "learning_rate": 8.060600871043566e-05,
1201
+ "loss": 2.3946,
1202
+ "step": 97500
1203
+ },
1204
+ {
1205
+ "epoch": 0.36,
1206
+ "learning_rate": 8.034959190974374e-05,
1207
+ "loss": 2.397,
1208
+ "step": 98000
1209
+ },
1210
+ {
1211
+ "epoch": 0.37,
1212
+ "learning_rate": 8.009190499228698e-05,
1213
+ "loss": 2.3854,
1214
+ "step": 98500
1215
+ },
1216
+ {
1217
+ "epoch": 0.37,
1218
+ "learning_rate": 7.983295874214692e-05,
1219
+ "loss": 2.4079,
1220
+ "step": 99000
1221
+ },
1222
+ {
1223
+ "epoch": 0.37,
1224
+ "learning_rate": 7.95727639961076e-05,
1225
+ "loss": 2.394,
1226
+ "step": 99500
1227
+ },
1228
+ {
1229
+ "epoch": 0.37,
1230
+ "learning_rate": 7.931133164320208e-05,
1231
+ "loss": 2.394,
1232
+ "step": 100000
1233
+ },
1234
+ {
1235
+ "epoch": 0.37,
1236
+ "eval_loss": 2.2282919883728027,
1237
+ "eval_runtime": 285.8926,
1238
+ "eval_samples_per_second": 349.782,
1239
+ "eval_steps_per_second": 43.723,
1240
+ "step": 100000
1241
+ },
1242
+ {
1243
+ "epoch": 0.37,
1244
+ "learning_rate": 7.904867262425669e-05,
1245
+ "loss": 2.3871,
1246
+ "step": 100500
1247
+ },
1248
+ {
1249
+ "epoch": 0.37,
1250
+ "learning_rate": 7.878479793143314e-05,
1251
+ "loss": 2.3768,
1252
+ "step": 101000
1253
+ },
1254
+ {
1255
+ "epoch": 0.38,
1256
+ "learning_rate": 7.851971860776864e-05,
1257
+ "loss": 2.3783,
1258
+ "step": 101500
1259
+ },
1260
+ {
1261
+ "epoch": 0.38,
1262
+ "learning_rate": 7.82534457467136e-05,
1263
+ "loss": 2.3733,
1264
+ "step": 102000
1265
+ },
1266
+ {
1267
+ "epoch": 0.38,
1268
+ "learning_rate": 7.798599049166741e-05,
1269
+ "loss": 2.3717,
1270
+ "step": 102500
1271
+ },
1272
+ {
1273
+ "epoch": 0.38,
1274
+ "learning_rate": 7.771736403551216e-05,
1275
+ "loss": 2.3689,
1276
+ "step": 103000
1277
+ },
1278
+ {
1279
+ "epoch": 0.38,
1280
+ "learning_rate": 7.744757762014416e-05,
1281
+ "loss": 2.3714,
1282
+ "step": 103500
1283
+ },
1284
+ {
1285
+ "epoch": 0.39,
1286
+ "learning_rate": 7.717664253600352e-05,
1287
+ "loss": 2.3792,
1288
+ "step": 104000
1289
+ },
1290
+ {
1291
+ "epoch": 0.39,
1292
+ "learning_rate": 7.690457012160156e-05,
1293
+ "loss": 2.3711,
1294
+ "step": 104500
1295
+ },
1296
+ {
1297
+ "epoch": 0.39,
1298
+ "learning_rate": 7.663137176304642e-05,
1299
+ "loss": 2.3829,
1300
+ "step": 105000
1301
+ },
1302
+ {
1303
+ "epoch": 0.39,
1304
+ "learning_rate": 7.635705889356646e-05,
1305
+ "loss": 2.4541,
1306
+ "step": 105500
1307
+ },
1308
+ {
1309
+ "epoch": 0.39,
1310
+ "learning_rate": 7.608164299303187e-05,
1311
+ "loss": 2.386,
1312
+ "step": 106000
1313
+ },
1314
+ {
1315
+ "epoch": 0.39,
1316
+ "learning_rate": 7.580513558747409e-05,
1317
+ "loss": 2.3839,
1318
+ "step": 106500
1319
+ },
1320
+ {
1321
+ "epoch": 0.4,
1322
+ "learning_rate": 7.552754824860368e-05,
1323
+ "loss": 2.3896,
1324
+ "step": 107000
1325
+ },
1326
+ {
1327
+ "epoch": 0.4,
1328
+ "learning_rate": 7.524889259332584e-05,
1329
+ "loss": 2.3839,
1330
+ "step": 107500
1331
+ },
1332
+ {
1333
+ "epoch": 0.4,
1334
+ "learning_rate": 7.496918028325434e-05,
1335
+ "loss": 2.3821,
1336
+ "step": 108000
1337
+ },
1338
+ {
1339
+ "epoch": 0.4,
1340
+ "learning_rate": 7.468842302422355e-05,
1341
+ "loss": 2.3797,
1342
+ "step": 108500
1343
+ },
1344
+ {
1345
+ "epoch": 0.4,
1346
+ "learning_rate": 7.440663256579836e-05,
1347
+ "loss": 2.348,
1348
+ "step": 109000
1349
+ },
1350
+ {
1351
+ "epoch": 0.41,
1352
+ "learning_rate": 7.412382070078269e-05,
1353
+ "loss": 2.3413,
1354
+ "step": 109500
1355
+ },
1356
+ {
1357
+ "epoch": 0.41,
1358
+ "learning_rate": 7.383999926472585e-05,
1359
+ "loss": 2.3497,
1360
+ "step": 110000
1361
+ },
1362
+ {
1363
+ "epoch": 0.41,
1364
+ "learning_rate": 7.355518013542717e-05,
1365
+ "loss": 2.3303,
1366
+ "step": 110500
1367
+ },
1368
+ {
1369
+ "epoch": 0.41,
1370
+ "learning_rate": 7.326937523243908e-05,
1371
+ "loss": 2.3348,
1372
+ "step": 111000
1373
+ },
1374
+ {
1375
+ "epoch": 0.41,
1376
+ "learning_rate": 7.29825965165682e-05,
1377
+ "loss": 2.3375,
1378
+ "step": 111500
1379
+ },
1380
+ {
1381
+ "epoch": 0.42,
1382
+ "learning_rate": 7.269485598937468e-05,
1383
+ "loss": 2.3271,
1384
+ "step": 112000
1385
+ },
1386
+ {
1387
+ "epoch": 0.42,
1388
+ "learning_rate": 7.240616569267015e-05,
1389
+ "loss": 2.3184,
1390
+ "step": 112500
1391
+ },
1392
+ {
1393
+ "epoch": 0.42,
1394
+ "learning_rate": 7.211653770801363e-05,
1395
+ "loss": 2.333,
1396
+ "step": 113000
1397
+ },
1398
+ {
1399
+ "epoch": 0.42,
1400
+ "learning_rate": 7.182598415620591e-05,
1401
+ "loss": 2.3712,
1402
+ "step": 113500
1403
+ },
1404
+ {
1405
+ "epoch": 0.42,
1406
+ "learning_rate": 7.153451719678243e-05,
1407
+ "loss": 2.3387,
1408
+ "step": 114000
1409
+ },
1410
+ {
1411
+ "epoch": 0.42,
1412
+ "learning_rate": 7.124214902750428e-05,
1413
+ "loss": 2.3274,
1414
+ "step": 114500
1415
+ },
1416
+ {
1417
+ "epoch": 0.43,
1418
+ "learning_rate": 7.094889188384774e-05,
1419
+ "loss": 2.326,
1420
+ "step": 115000
1421
+ },
1422
+ {
1423
+ "epoch": 0.43,
1424
+ "learning_rate": 7.065475803849241e-05,
1425
+ "loss": 2.3131,
1426
+ "step": 115500
1427
+ },
1428
+ {
1429
+ "epoch": 0.43,
1430
+ "learning_rate": 7.035975980080728e-05,
1431
+ "loss": 2.3186,
1432
+ "step": 116000
1433
+ },
1434
+ {
1435
+ "epoch": 0.43,
1436
+ "learning_rate": 7.006390951633589e-05,
1437
+ "loss": 2.3156,
1438
+ "step": 116500
1439
+ },
1440
+ {
1441
+ "epoch": 0.43,
1442
+ "learning_rate": 6.976721956627952e-05,
1443
+ "loss": 2.3211,
1444
+ "step": 117000
1445
+ },
1446
+ {
1447
+ "epoch": 0.44,
1448
+ "learning_rate": 6.946970236697905e-05,
1449
+ "loss": 2.3116,
1450
+ "step": 117500
1451
+ },
1452
+ {
1453
+ "epoch": 0.44,
1454
+ "learning_rate": 6.917137036939542e-05,
1455
+ "loss": 2.3189,
1456
+ "step": 118000
1457
+ },
1458
+ {
1459
+ "epoch": 0.44,
1460
+ "learning_rate": 6.88722360585885e-05,
1461
+ "loss": 2.3286,
1462
+ "step": 118500
1463
+ },
1464
+ {
1465
+ "epoch": 0.44,
1466
+ "learning_rate": 6.857231195319457e-05,
1467
+ "loss": 2.3186,
1468
+ "step": 119000
1469
+ },
1470
+ {
1471
+ "epoch": 0.44,
1472
+ "learning_rate": 6.827161060490248e-05,
1473
+ "loss": 2.3102,
1474
+ "step": 119500
1475
+ },
1476
+ {
1477
+ "epoch": 0.44,
1478
+ "learning_rate": 6.797014459792836e-05,
1479
+ "loss": 2.3146,
1480
+ "step": 120000
1481
+ },
1482
+ {
1483
+ "epoch": 0.45,
1484
+ "learning_rate": 6.766792654848896e-05,
1485
+ "loss": 2.2962,
1486
+ "step": 120500
1487
+ },
1488
+ {
1489
+ "epoch": 0.45,
1490
+ "learning_rate": 6.736496910427364e-05,
1491
+ "loss": 2.2984,
1492
+ "step": 121000
1493
+ },
1494
+ {
1495
+ "epoch": 0.45,
1496
+ "learning_rate": 6.70612849439151e-05,
1497
+ "loss": 2.308,
1498
+ "step": 121500
1499
+ },
1500
+ {
1501
+ "epoch": 0.45,
1502
+ "learning_rate": 6.675688677645883e-05,
1503
+ "loss": 2.3029,
1504
+ "step": 122000
1505
+ },
1506
+ {
1507
+ "epoch": 0.45,
1508
+ "learning_rate": 6.645178734083117e-05,
1509
+ "loss": 2.2933,
1510
+ "step": 122500
1511
+ },
1512
+ {
1513
+ "epoch": 0.46,
1514
+ "learning_rate": 6.614599940530619e-05,
1515
+ "loss": 2.3026,
1516
+ "step": 123000
1517
+ },
1518
+ {
1519
+ "epoch": 0.46,
1520
+ "learning_rate": 6.583953576697141e-05,
1521
+ "loss": 2.3084,
1522
+ "step": 123500
1523
+ },
1524
+ {
1525
+ "epoch": 0.46,
1526
+ "learning_rate": 6.553240925119219e-05,
1527
+ "loss": 2.2985,
1528
+ "step": 124000
1529
+ },
1530
+ {
1531
+ "epoch": 0.46,
1532
+ "learning_rate": 6.522463271107502e-05,
1533
+ "loss": 2.3035,
1534
+ "step": 124500
1535
+ },
1536
+ {
1537
+ "epoch": 0.46,
1538
+ "learning_rate": 6.49162190269296e-05,
1539
+ "loss": 2.2891,
1540
+ "step": 125000
1541
+ },
1542
+ {
1543
+ "epoch": 0.46,
1544
+ "eval_loss": 2.1396679878234863,
1545
+ "eval_runtime": 305.6778,
1546
+ "eval_samples_per_second": 327.142,
1547
+ "eval_steps_per_second": 40.893,
1548
+ "step": 125000
1549
+ },
1550
+ {
1551
+ "epoch": 0.47,
1552
+ "learning_rate": 6.460718110572983e-05,
1553
+ "loss": 2.2987,
1554
+ "step": 125500
1555
+ },
1556
+ {
1557
+ "epoch": 0.47,
1558
+ "learning_rate": 6.429753188057368e-05,
1559
+ "loss": 2.2899,
1560
+ "step": 126000
1561
+ },
1562
+ {
1563
+ "epoch": 0.47,
1564
+ "learning_rate": 6.398728431014187e-05,
1565
+ "loss": 2.2919,
1566
+ "step": 126500
1567
+ },
1568
+ {
1569
+ "epoch": 0.47,
1570
+ "learning_rate": 6.367645137815561e-05,
1571
+ "loss": 2.2878,
1572
+ "step": 127000
1573
+ },
1574
+ {
1575
+ "epoch": 0.47,
1576
+ "learning_rate": 6.336504609283325e-05,
1577
+ "loss": 2.2848,
1578
+ "step": 127500
1579
+ },
1580
+ {
1581
+ "epoch": 0.47,
1582
+ "learning_rate": 6.305308148634585e-05,
1583
+ "loss": 2.2813,
1584
+ "step": 128000
1585
+ },
1586
+ {
1587
+ "epoch": 0.48,
1588
+ "learning_rate": 6.274057061427182e-05,
1589
+ "loss": 2.2804,
1590
+ "step": 128500
1591
+ },
1592
+ {
1593
+ "epoch": 0.48,
1594
+ "learning_rate": 6.242752655505053e-05,
1595
+ "loss": 2.2888,
1596
+ "step": 129000
1597
+ },
1598
+ {
1599
+ "epoch": 0.48,
1600
+ "learning_rate": 6.211396240943499e-05,
1601
+ "loss": 2.2884,
1602
+ "step": 129500
1603
+ },
1604
+ {
1605
+ "epoch": 0.48,
1606
+ "learning_rate": 6.17998912999436e-05,
1607
+ "loss": 2.28,
1608
+ "step": 130000
1609
+ },
1610
+ {
1611
+ "epoch": 0.48,
1612
+ "learning_rate": 6.148532637031098e-05,
1613
+ "loss": 2.2855,
1614
+ "step": 130500
1615
+ },
1616
+ {
1617
+ "epoch": 0.49,
1618
+ "learning_rate": 6.117028078493787e-05,
1619
+ "loss": 2.3124,
1620
+ "step": 131000
1621
+ },
1622
+ {
1623
+ "epoch": 0.49,
1624
+ "learning_rate": 6.085476772834029e-05,
1625
+ "loss": 2.2886,
1626
+ "step": 131500
1627
+ },
1628
+ {
1629
+ "epoch": 0.49,
1630
+ "learning_rate": 6.053880040459764e-05,
1631
+ "loss": 2.3043,
1632
+ "step": 132000
1633
+ },
1634
+ {
1635
+ "epoch": 0.49,
1636
+ "learning_rate": 6.022239203680027e-05,
1637
+ "loss": 2.3534,
1638
+ "step": 132500
1639
+ },
1640
+ {
1641
+ "epoch": 0.49,
1642
+ "learning_rate": 5.990555586649599e-05,
1643
+ "loss": 2.569,
1644
+ "step": 133000
1645
+ },
1646
+ {
1647
+ "epoch": 0.49,
1648
+ "learning_rate": 5.958830515313596e-05,
1649
+ "loss": 2.341,
1650
+ "step": 133500
1651
+ },
1652
+ {
1653
+ "epoch": 0.5,
1654
+ "learning_rate": 5.927065317351976e-05,
1655
+ "loss": 2.2866,
1656
+ "step": 134000
1657
+ },
1658
+ {
1659
+ "epoch": 0.5,
1660
+ "learning_rate": 5.8952613221239826e-05,
1661
+ "loss": 2.2786,
1662
+ "step": 134500
1663
+ },
1664
+ {
1665
+ "epoch": 0.5,
1666
+ "learning_rate": 5.863419860612506e-05,
1667
+ "loss": 2.2707,
1668
+ "step": 135000
1669
+ },
1670
+ {
1671
+ "epoch": 0.5,
1672
+ "learning_rate": 5.831542265368378e-05,
1673
+ "loss": 2.277,
1674
+ "step": 135500
1675
+ },
1676
+ {
1677
+ "epoch": 0.5,
1678
+ "learning_rate": 5.799629870454619e-05,
1679
+ "loss": 2.2702,
1680
+ "step": 136000
1681
+ },
1682
+ {
1683
+ "epoch": 0.51,
1684
+ "learning_rate": 5.7676840113905974e-05,
1685
+ "loss": 2.2673,
1686
+ "step": 136500
1687
+ },
1688
+ {
1689
+ "epoch": 0.51,
1690
+ "learning_rate": 5.735706025096136e-05,
1691
+ "loss": 2.2517,
1692
+ "step": 137000
1693
+ },
1694
+ {
1695
+ "epoch": 0.51,
1696
+ "learning_rate": 5.7036972498355744e-05,
1697
+ "loss": 2.2768,
1698
+ "step": 137500
1699
+ },
1700
+ {
1701
+ "epoch": 0.51,
1702
+ "learning_rate": 5.671659025161755e-05,
1703
+ "loss": 2.2569,
1704
+ "step": 138000
1705
+ },
1706
+ {
1707
+ "epoch": 0.51,
1708
+ "learning_rate": 5.6395926918599606e-05,
1709
+ "loss": 2.2527,
1710
+ "step": 138500
1711
+ },
1712
+ {
1713
+ "epoch": 0.52,
1714
+ "learning_rate": 5.607499591891816e-05,
1715
+ "loss": 2.2662,
1716
+ "step": 139000
1717
+ },
1718
+ {
1719
+ "epoch": 0.52,
1720
+ "learning_rate": 5.5753810683391104e-05,
1721
+ "loss": 2.2446,
1722
+ "step": 139500
1723
+ },
1724
+ {
1725
+ "epoch": 0.52,
1726
+ "learning_rate": 5.5432384653476e-05,
1727
+ "loss": 2.2506,
1728
+ "step": 140000
1729
+ },
1730
+ {
1731
+ "epoch": 0.52,
1732
+ "learning_rate": 5.5110731280707605e-05,
1733
+ "loss": 2.2495,
1734
+ "step": 140500
1735
+ },
1736
+ {
1737
+ "epoch": 0.52,
1738
+ "learning_rate": 5.4788864026134824e-05,
1739
+ "loss": 2.2513,
1740
+ "step": 141000
1741
+ },
1742
+ {
1743
+ "epoch": 0.52,
1744
+ "learning_rate": 5.446679635975741e-05,
1745
+ "loss": 2.2512,
1746
+ "step": 141500
1747
+ },
1748
+ {
1749
+ "epoch": 0.53,
1750
+ "learning_rate": 5.41445417599623e-05,
1751
+ "loss": 2.2497,
1752
+ "step": 142000
1753
+ },
1754
+ {
1755
+ "epoch": 0.53,
1756
+ "learning_rate": 5.3822113712959466e-05,
1757
+ "loss": 2.2491,
1758
+ "step": 142500
1759
+ },
1760
+ {
1761
+ "epoch": 0.53,
1762
+ "learning_rate": 5.349952571221761e-05,
1763
+ "loss": 2.2425,
1764
+ "step": 143000
1765
+ },
1766
+ {
1767
+ "epoch": 0.53,
1768
+ "learning_rate": 5.3176791257899405e-05,
1769
+ "loss": 2.2491,
1770
+ "step": 143500
1771
+ },
1772
+ {
1773
+ "epoch": 0.53,
1774
+ "learning_rate": 5.285392385629653e-05,
1775
+ "loss": 2.2542,
1776
+ "step": 144000
1777
+ },
1778
+ {
1779
+ "epoch": 0.54,
1780
+ "learning_rate": 5.253093701926446e-05,
1781
+ "loss": 2.2498,
1782
+ "step": 144500
1783
+ },
1784
+ {
1785
+ "epoch": 0.54,
1786
+ "learning_rate": 5.2207844263657e-05,
1787
+ "loss": 2.2456,
1788
+ "step": 145000
1789
+ },
1790
+ {
1791
+ "epoch": 0.54,
1792
+ "learning_rate": 5.188465911076059e-05,
1793
+ "loss": 2.235,
1794
+ "step": 145500
1795
+ },
1796
+ {
1797
+ "epoch": 0.54,
1798
+ "learning_rate": 5.156139508572844e-05,
1799
+ "loss": 2.2463,
1800
+ "step": 146000
1801
+ },
1802
+ {
1803
+ "epoch": 0.54,
1804
+ "learning_rate": 5.1238065717014526e-05,
1805
+ "loss": 2.232,
1806
+ "step": 146500
1807
+ },
1808
+ {
1809
+ "epoch": 0.54,
1810
+ "learning_rate": 5.091468453580748e-05,
1811
+ "loss": 2.227,
1812
+ "step": 147000
1813
+ },
1814
+ {
1815
+ "epoch": 0.55,
1816
+ "learning_rate": 5.0591265075464167e-05,
1817
+ "loss": 2.2354,
1818
+ "step": 147500
1819
+ },
1820
+ {
1821
+ "epoch": 0.55,
1822
+ "learning_rate": 5.026782087094353e-05,
1823
+ "loss": 2.2266,
1824
+ "step": 148000
1825
+ },
1826
+ {
1827
+ "epoch": 0.55,
1828
+ "learning_rate": 4.9944365458239946e-05,
1829
+ "loss": 2.2319,
1830
+ "step": 148500
1831
+ },
1832
+ {
1833
+ "epoch": 0.55,
1834
+ "learning_rate": 4.9620912373816894e-05,
1835
+ "loss": 2.2383,
1836
+ "step": 149000
1837
+ },
1838
+ {
1839
+ "epoch": 0.55,
1840
+ "learning_rate": 4.929747515404043e-05,
1841
+ "loss": 2.2315,
1842
+ "step": 149500
1843
+ },
1844
+ {
1845
+ "epoch": 0.56,
1846
+ "learning_rate": 4.897406733461264e-05,
1847
+ "loss": 2.2286,
1848
+ "step": 150000
1849
+ },
1850
+ {
1851
+ "epoch": 0.56,
1852
+ "eval_loss": 2.0796430110931396,
1853
+ "eval_runtime": 286.5966,
1854
+ "eval_samples_per_second": 348.923,
1855
+ "eval_steps_per_second": 43.615,
1856
+ "step": 150000
1857
+ },
1858
+ {
1859
+ "epoch": 0.56,
1860
+ "learning_rate": 4.8650702450005264e-05,
1861
+ "loss": 2.2269,
1862
+ "step": 150500
1863
+ },
1864
+ {
1865
+ "epoch": 0.56,
1866
+ "learning_rate": 4.832739403289318e-05,
1867
+ "loss": 2.2294,
1868
+ "step": 151000
1869
+ },
1870
+ {
1871
+ "epoch": 0.56,
1872
+ "learning_rate": 4.8004155613588214e-05,
1873
+ "loss": 2.216,
1874
+ "step": 151500
1875
+ },
1876
+ {
1877
+ "epoch": 0.56,
1878
+ "learning_rate": 4.7681000719472726e-05,
1879
+ "loss": 2.2128,
1880
+ "step": 152000
1881
+ },
1882
+ {
1883
+ "epoch": 0.57,
1884
+ "learning_rate": 4.7357942874433634e-05,
1885
+ "loss": 2.2088,
1886
+ "step": 152500
1887
+ },
1888
+ {
1889
+ "epoch": 0.57,
1890
+ "learning_rate": 4.703499559829639e-05,
1891
+ "loss": 2.2112,
1892
+ "step": 153000
1893
+ },
1894
+ {
1895
+ "epoch": 0.57,
1896
+ "learning_rate": 4.671217240625916e-05,
1897
+ "loss": 2.2168,
1898
+ "step": 153500
1899
+ },
1900
+ {
1901
+ "epoch": 0.57,
1902
+ "learning_rate": 4.6389486808327304e-05,
1903
+ "loss": 2.2034,
1904
+ "step": 154000
1905
+ },
1906
+ {
1907
+ "epoch": 0.57,
1908
+ "learning_rate": 4.606695230874788e-05,
1909
+ "loss": 2.2045,
1910
+ "step": 154500
1911
+ },
1912
+ {
1913
+ "epoch": 0.57,
1914
+ "learning_rate": 4.5744582405444544e-05,
1915
+ "loss": 2.2043,
1916
+ "step": 155000
1917
+ },
1918
+ {
1919
+ "epoch": 0.58,
1920
+ "learning_rate": 4.542239058945272e-05,
1921
+ "loss": 2.2116,
1922
+ "step": 155500
1923
+ },
1924
+ {
1925
+ "epoch": 0.58,
1926
+ "learning_rate": 4.51003903443549e-05,
1927
+ "loss": 2.2046,
1928
+ "step": 156000
1929
+ },
1930
+ {
1931
+ "epoch": 0.58,
1932
+ "learning_rate": 4.4778595145716465e-05,
1933
+ "loss": 2.1967,
1934
+ "step": 156500
1935
+ },
1936
+ {
1937
+ "epoch": 0.58,
1938
+ "learning_rate": 4.4457018460521684e-05,
1939
+ "loss": 2.2046,
1940
+ "step": 157000
1941
+ },
1942
+ {
1943
+ "epoch": 0.58,
1944
+ "learning_rate": 4.4135673746610115e-05,
1945
+ "loss": 2.1952,
1946
+ "step": 157500
1947
+ },
1948
+ {
1949
+ "epoch": 0.59,
1950
+ "learning_rate": 4.381457445211346e-05,
1951
+ "loss": 2.2071,
1952
+ "step": 158000
1953
+ },
1954
+ {
1955
+ "epoch": 0.59,
1956
+ "learning_rate": 4.349373401489269e-05,
1957
+ "loss": 2.1993,
1958
+ "step": 158500
1959
+ },
1960
+ {
1961
+ "epoch": 0.59,
1962
+ "learning_rate": 4.317316586197571e-05,
1963
+ "loss": 2.1998,
1964
+ "step": 159000
1965
+ },
1966
+ {
1967
+ "epoch": 0.59,
1968
+ "learning_rate": 4.2852883408995515e-05,
1969
+ "loss": 2.1938,
1970
+ "step": 159500
1971
+ },
1972
+ {
1973
+ "epoch": 0.59,
1974
+ "learning_rate": 4.253290005962863e-05,
1975
+ "loss": 2.1901,
1976
+ "step": 160000
1977
+ },
1978
+ {
1979
+ "epoch": 0.59,
1980
+ "learning_rate": 4.221322920503423e-05,
1981
+ "loss": 2.1933,
1982
+ "step": 160500
1983
+ },
1984
+ {
1985
+ "epoch": 0.6,
1986
+ "learning_rate": 4.1893884223293746e-05,
1987
+ "loss": 2.1884,
1988
+ "step": 161000
1989
+ },
1990
+ {
1991
+ "epoch": 0.6,
1992
+ "learning_rate": 4.157487847885094e-05,
1993
+ "loss": 2.1855,
1994
+ "step": 161500
1995
+ },
1996
+ {
1997
+ "epoch": 0.6,
1998
+ "learning_rate": 4.1256225321952705e-05,
1999
+ "loss": 2.1968,
2000
+ "step": 162000
2001
+ },
2002
+ {
2003
+ "epoch": 0.6,
2004
+ "learning_rate": 4.093793808809028e-05,
2005
+ "loss": 2.186,
2006
+ "step": 162500
2007
+ },
2008
+ {
2009
+ "epoch": 0.6,
2010
+ "learning_rate": 4.062003009744115e-05,
2011
+ "loss": 2.1905,
2012
+ "step": 163000
2013
+ },
2014
+ {
2015
+ "epoch": 0.61,
2016
+ "learning_rate": 4.0302514654311675e-05,
2017
+ "loss": 2.1828,
2018
+ "step": 163500
2019
+ },
2020
+ {
2021
+ "epoch": 0.61,
2022
+ "learning_rate": 3.998540504658027e-05,
2023
+ "loss": 2.1878,
2024
+ "step": 164000
2025
+ },
2026
+ {
2027
+ "epoch": 0.61,
2028
+ "learning_rate": 3.966871454514137e-05,
2029
+ "loss": 2.1918,
2030
+ "step": 164500
2031
+ },
2032
+ {
2033
+ "epoch": 0.61,
2034
+ "learning_rate": 3.935245640334991e-05,
2035
+ "loss": 2.182,
2036
+ "step": 165000
2037
+ },
2038
+ {
2039
+ "epoch": 0.61,
2040
+ "learning_rate": 3.903664385646685e-05,
2041
+ "loss": 2.1872,
2042
+ "step": 165500
2043
+ },
2044
+ {
2045
+ "epoch": 0.62,
2046
+ "learning_rate": 3.872129012110515e-05,
2047
+ "loss": 2.18,
2048
+ "step": 166000
2049
+ },
2050
+ {
2051
+ "epoch": 0.62,
2052
+ "learning_rate": 3.8406408394676724e-05,
2053
+ "loss": 2.1794,
2054
+ "step": 166500
2055
+ },
2056
+ {
2057
+ "epoch": 0.62,
2058
+ "learning_rate": 3.8092011854840135e-05,
2059
+ "loss": 2.1778,
2060
+ "step": 167000
2061
+ },
2062
+ {
2063
+ "epoch": 0.62,
2064
+ "learning_rate": 3.7778113658949145e-05,
2065
+ "loss": 2.1764,
2066
+ "step": 167500
2067
+ },
2068
+ {
2069
+ "epoch": 0.62,
2070
+ "learning_rate": 3.7464726943501955e-05,
2071
+ "loss": 2.1727,
2072
+ "step": 168000
2073
+ },
2074
+ {
2075
+ "epoch": 0.62,
2076
+ "learning_rate": 3.715186482359162e-05,
2077
+ "loss": 2.1718,
2078
+ "step": 168500
2079
+ },
2080
+ {
2081
+ "epoch": 0.63,
2082
+ "learning_rate": 3.683954039235707e-05,
2083
+ "loss": 2.1766,
2084
+ "step": 169000
2085
+ },
2086
+ {
2087
+ "epoch": 0.63,
2088
+ "learning_rate": 3.6527766720435186e-05,
2089
+ "loss": 2.1838,
2090
+ "step": 169500
2091
+ },
2092
+ {
2093
+ "epoch": 0.63,
2094
+ "learning_rate": 3.6216556855413906e-05,
2095
+ "loss": 2.1786,
2096
+ "step": 170000
2097
+ },
2098
+ {
2099
+ "epoch": 0.63,
2100
+ "learning_rate": 3.5905923821286006e-05,
2101
+ "loss": 2.1748,
2102
+ "step": 170500
2103
+ },
2104
+ {
2105
+ "epoch": 0.63,
2106
+ "learning_rate": 3.559588061790419e-05,
2107
+ "loss": 2.1708,
2108
+ "step": 171000
2109
+ },
2110
+ {
2111
+ "epoch": 0.64,
2112
+ "learning_rate": 3.528644022043701e-05,
2113
+ "loss": 2.1749,
2114
+ "step": 171500
2115
+ },
2116
+ {
2117
+ "epoch": 0.64,
2118
+ "learning_rate": 3.497761557882584e-05,
2119
+ "loss": 2.1668,
2120
+ "step": 172000
2121
+ },
2122
+ {
2123
+ "epoch": 0.64,
2124
+ "learning_rate": 3.466941961724301e-05,
2125
+ "loss": 2.1703,
2126
+ "step": 172500
2127
+ },
2128
+ {
2129
+ "epoch": 0.64,
2130
+ "learning_rate": 3.4361865233550814e-05,
2131
+ "loss": 2.1682,
2132
+ "step": 173000
2133
+ },
2134
+ {
2135
+ "epoch": 0.64,
2136
+ "learning_rate": 3.40549652987618e-05,
2137
+ "loss": 2.1705,
2138
+ "step": 173500
2139
+ },
2140
+ {
2141
+ "epoch": 0.64,
2142
+ "learning_rate": 3.374873265650016e-05,
2143
+ "loss": 2.1671,
2144
+ "step": 174000
2145
+ },
2146
+ {
2147
+ "epoch": 0.65,
2148
+ "learning_rate": 3.3443180122464156e-05,
2149
+ "loss": 2.16,
2150
+ "step": 174500
2151
+ },
2152
+ {
2153
+ "epoch": 0.65,
2154
+ "learning_rate": 3.3138320483889874e-05,
2155
+ "loss": 2.1536,
2156
+ "step": 175000
2157
+ },
2158
+ {
2159
+ "epoch": 0.65,
2160
+ "eval_loss": 2.0123212337493896,
2161
+ "eval_runtime": 294.337,
2162
+ "eval_samples_per_second": 339.747,
2163
+ "eval_steps_per_second": 42.468,
2164
+ "step": 175000
2165
+ },
2166
+ {
2167
+ "epoch": 0.65,
2168
+ "learning_rate": 3.283416649901599e-05,
2169
+ "loss": 2.152,
2170
+ "step": 175500
2171
+ },
2172
+ {
2173
+ "epoch": 0.65,
2174
+ "learning_rate": 3.253073089654992e-05,
2175
+ "loss": 2.1519,
2176
+ "step": 176000
2177
+ },
2178
+ {
2179
+ "epoch": 0.65,
2180
+ "learning_rate": 3.222802637513508e-05,
2181
+ "loss": 2.1638,
2182
+ "step": 176500
2183
+ },
2184
+ {
2185
+ "epoch": 0.66,
2186
+ "learning_rate": 3.192606560281948e-05,
2187
+ "loss": 2.1527,
2188
+ "step": 177000
2189
+ },
2190
+ {
2191
+ "epoch": 0.66,
2192
+ "learning_rate": 3.162486121652556e-05,
2193
+ "loss": 2.1499,
2194
+ "step": 177500
2195
+ },
2196
+ {
2197
+ "epoch": 0.66,
2198
+ "learning_rate": 3.1324425821521375e-05,
2199
+ "loss": 2.1614,
2200
+ "step": 178000
2201
+ },
2202
+ {
2203
+ "epoch": 0.66,
2204
+ "learning_rate": 3.1024771990893e-05,
2205
+ "loss": 2.1556,
2206
+ "step": 178500
2207
+ },
2208
+ {
2209
+ "epoch": 0.66,
2210
+ "learning_rate": 3.072591226501842e-05,
2211
+ "loss": 2.1514,
2212
+ "step": 179000
2213
+ },
2214
+ {
2215
+ "epoch": 0.67,
2216
+ "learning_rate": 3.042785915104267e-05,
2217
+ "loss": 2.1533,
2218
+ "step": 179500
2219
+ },
2220
+ {
2221
+ "epoch": 0.67,
2222
+ "learning_rate": 3.013062512235445e-05,
2223
+ "loss": 2.1413,
2224
+ "step": 180000
2225
+ },
2226
+ {
2227
+ "epoch": 0.67,
2228
+ "learning_rate": 2.9834222618064146e-05,
2229
+ "loss": 2.1461,
2230
+ "step": 180500
2231
+ },
2232
+ {
2233
+ "epoch": 0.67,
2234
+ "learning_rate": 2.9538664042483145e-05,
2235
+ "loss": 2.1475,
2236
+ "step": 181000
2237
+ },
2238
+ {
2239
+ "epoch": 0.67,
2240
+ "learning_rate": 2.9243961764604878e-05,
2241
+ "loss": 2.1452,
2242
+ "step": 181500
2243
+ },
2244
+ {
2245
+ "epoch": 0.67,
2246
+ "learning_rate": 2.895012811758705e-05,
2247
+ "loss": 2.1488,
2248
+ "step": 182000
2249
+ },
2250
+ {
2251
+ "epoch": 0.68,
2252
+ "learning_rate": 2.8657175398235548e-05,
2253
+ "loss": 2.1496,
2254
+ "step": 182500
2255
+ },
2256
+ {
2257
+ "epoch": 0.68,
2258
+ "learning_rate": 2.8365115866489895e-05,
2259
+ "loss": 2.1348,
2260
+ "step": 183000
2261
+ },
2262
+ {
2263
+ "epoch": 0.68,
2264
+ "learning_rate": 2.8073961744910036e-05,
2265
+ "loss": 2.1406,
2266
+ "step": 183500
2267
+ },
2268
+ {
2269
+ "epoch": 0.68,
2270
+ "learning_rate": 2.7783725218164992e-05,
2271
+ "loss": 2.1484,
2272
+ "step": 184000
2273
+ },
2274
+ {
2275
+ "epoch": 0.68,
2276
+ "learning_rate": 2.7494418432522773e-05,
2277
+ "loss": 2.1482,
2278
+ "step": 184500
2279
+ },
2280
+ {
2281
+ "epoch": 0.69,
2282
+ "learning_rate": 2.7206053495342176e-05,
2283
+ "loss": 2.1491,
2284
+ "step": 185000
2285
+ },
2286
+ {
2287
+ "epoch": 0.69,
2288
+ "learning_rate": 2.691864247456609e-05,
2289
+ "loss": 2.1316,
2290
+ "step": 185500
2291
+ },
2292
+ {
2293
+ "epoch": 0.69,
2294
+ "learning_rate": 2.6632197398216403e-05,
2295
+ "loss": 2.1459,
2296
+ "step": 186000
2297
+ },
2298
+ {
2299
+ "epoch": 0.69,
2300
+ "learning_rate": 2.6346730253890626e-05,
2301
+ "loss": 2.1308,
2302
+ "step": 186500
2303
+ },
2304
+ {
2305
+ "epoch": 0.69,
2306
+ "learning_rate": 2.6062252988260348e-05,
2307
+ "loss": 2.1389,
2308
+ "step": 187000
2309
+ },
2310
+ {
2311
+ "epoch": 0.69,
2312
+ "learning_rate": 2.5778777506571112e-05,
2313
+ "loss": 2.1374,
2314
+ "step": 187500
2315
+ },
2316
+ {
2317
+ "epoch": 0.7,
2318
+ "learning_rate": 2.5496315672144238e-05,
2319
+ "loss": 2.1322,
2320
+ "step": 188000
2321
+ },
2322
+ {
2323
+ "epoch": 0.7,
2324
+ "learning_rate": 2.521487930588044e-05,
2325
+ "loss": 2.1235,
2326
+ "step": 188500
2327
+ },
2328
+ {
2329
+ "epoch": 0.7,
2330
+ "learning_rate": 2.4934480185764976e-05,
2331
+ "loss": 2.1291,
2332
+ "step": 189000
2333
+ },
2334
+ {
2335
+ "epoch": 0.7,
2336
+ "learning_rate": 2.465513004637487e-05,
2337
+ "loss": 2.1366,
2338
+ "step": 189500
2339
+ },
2340
+ {
2341
+ "epoch": 0.7,
2342
+ "learning_rate": 2.4376840578387754e-05,
2343
+ "loss": 2.1256,
2344
+ "step": 190000
2345
+ },
2346
+ {
2347
+ "epoch": 0.71,
2348
+ "learning_rate": 2.409962342809261e-05,
2349
+ "loss": 2.1304,
2350
+ "step": 190500
2351
+ },
2352
+ {
2353
+ "epoch": 0.71,
2354
+ "learning_rate": 2.382349019690248e-05,
2355
+ "loss": 2.1229,
2356
+ "step": 191000
2357
+ },
2358
+ {
2359
+ "epoch": 0.71,
2360
+ "learning_rate": 2.3548452440868816e-05,
2361
+ "loss": 2.1268,
2362
+ "step": 191500
2363
+ },
2364
+ {
2365
+ "epoch": 0.71,
2366
+ "learning_rate": 2.3274521670197923e-05,
2367
+ "loss": 2.1278,
2368
+ "step": 192000
2369
+ },
2370
+ {
2371
+ "epoch": 0.71,
2372
+ "learning_rate": 2.300170934876934e-05,
2373
+ "loss": 2.1277,
2374
+ "step": 192500
2375
+ },
2376
+ {
2377
+ "epoch": 0.72,
2378
+ "learning_rate": 2.2730026893655916e-05,
2379
+ "loss": 2.1195,
2380
+ "step": 193000
2381
+ },
2382
+ {
2383
+ "epoch": 0.72,
2384
+ "learning_rate": 2.2459485674646187e-05,
2385
+ "loss": 2.1253,
2386
+ "step": 193500
2387
+ },
2388
+ {
2389
+ "epoch": 0.72,
2390
+ "learning_rate": 2.2190097013768403e-05,
2391
+ "loss": 2.1252,
2392
+ "step": 194000
2393
+ },
2394
+ {
2395
+ "epoch": 0.72,
2396
+ "learning_rate": 2.1921872184816817e-05,
2397
+ "loss": 2.1168,
2398
+ "step": 194500
2399
+ },
2400
+ {
2401
+ "epoch": 0.72,
2402
+ "learning_rate": 2.1654822412879798e-05,
2403
+ "loss": 2.1288,
2404
+ "step": 195000
2405
+ },
2406
+ {
2407
+ "epoch": 0.72,
2408
+ "learning_rate": 2.138895887387018e-05,
2409
+ "loss": 2.1195,
2410
+ "step": 195500
2411
+ },
2412
+ {
2413
+ "epoch": 0.73,
2414
+ "learning_rate": 2.112429269405739e-05,
2415
+ "loss": 2.1196,
2416
+ "step": 196000
2417
+ },
2418
+ {
2419
+ "epoch": 0.73,
2420
+ "learning_rate": 2.0860834949602015e-05,
2421
+ "loss": 2.1175,
2422
+ "step": 196500
2423
+ },
2424
+ {
2425
+ "epoch": 0.73,
2426
+ "learning_rate": 2.0598596666092075e-05,
2427
+ "loss": 2.1234,
2428
+ "step": 197000
2429
+ },
2430
+ {
2431
+ "epoch": 0.73,
2432
+ "learning_rate": 2.0337588818081744e-05,
2433
+ "loss": 2.1122,
2434
+ "step": 197500
2435
+ },
2436
+ {
2437
+ "epoch": 0.73,
2438
+ "learning_rate": 2.007782232863199e-05,
2439
+ "loss": 2.1185,
2440
+ "step": 198000
2441
+ },
2442
+ {
2443
+ "epoch": 0.74,
2444
+ "learning_rate": 1.9819308068853526e-05,
2445
+ "loss": 2.1201,
2446
+ "step": 198500
2447
+ },
2448
+ {
2449
+ "epoch": 0.74,
2450
+ "learning_rate": 1.956205685745183e-05,
2451
+ "loss": 2.1079,
2452
+ "step": 199000
2453
+ },
2454
+ {
2455
+ "epoch": 0.74,
2456
+ "learning_rate": 1.9306079460274302e-05,
2457
+ "loss": 2.1137,
2458
+ "step": 199500
2459
+ },
2460
+ {
2461
+ "epoch": 0.74,
2462
+ "learning_rate": 1.9051386589859843e-05,
2463
+ "loss": 2.114,
2464
+ "step": 200000
2465
+ },
2466
+ {
2467
+ "epoch": 0.74,
2468
+ "eval_loss": 1.9818874597549438,
2469
+ "eval_runtime": 287.1851,
2470
+ "eval_samples_per_second": 348.208,
2471
+ "eval_steps_per_second": 43.526,
2472
+ "step": 200000
2473
+ },
2474
+ {
2475
+ "epoch": 0.74,
2476
+ "learning_rate": 1.879798890499046e-05,
2477
+ "loss": 2.1094,
2478
+ "step": 200500
2479
+ },
2480
+ {
2481
+ "epoch": 0.75,
2482
+ "learning_rate": 1.8545897010245273e-05,
2483
+ "loss": 2.1185,
2484
+ "step": 201000
2485
+ },
2486
+ {
2487
+ "epoch": 0.75,
2488
+ "learning_rate": 1.8295121455556607e-05,
2489
+ "loss": 2.1187,
2490
+ "step": 201500
2491
+ },
2492
+ {
2493
+ "epoch": 0.75,
2494
+ "learning_rate": 1.8045672735768616e-05,
2495
+ "loss": 2.1096,
2496
+ "step": 202000
2497
+ },
2498
+ {
2499
+ "epoch": 0.75,
2500
+ "learning_rate": 1.7797561290197957e-05,
2501
+ "loss": 2.1058,
2502
+ "step": 202500
2503
+ },
2504
+ {
2505
+ "epoch": 0.75,
2506
+ "learning_rate": 1.755079750219699e-05,
2507
+ "loss": 2.1065,
2508
+ "step": 203000
2509
+ },
2510
+ {
2511
+ "epoch": 0.75,
2512
+ "learning_rate": 1.7305391698719187e-05,
2513
+ "loss": 2.104,
2514
+ "step": 203500
2515
+ },
2516
+ {
2517
+ "epoch": 0.76,
2518
+ "learning_rate": 1.706135414988701e-05,
2519
+ "loss": 2.1047,
2520
+ "step": 204000
2521
+ },
2522
+ {
2523
+ "epoch": 0.76,
2524
+ "learning_rate": 1.6818695068562084e-05,
2525
+ "loss": 2.1033,
2526
+ "step": 204500
2527
+ },
2528
+ {
2529
+ "epoch": 0.76,
2530
+ "learning_rate": 1.6577424609917756e-05,
2531
+ "loss": 2.1106,
2532
+ "step": 205000
2533
+ },
2534
+ {
2535
+ "epoch": 0.76,
2536
+ "learning_rate": 1.633755287101416e-05,
2537
+ "loss": 2.1068,
2538
+ "step": 205500
2539
+ },
2540
+ {
2541
+ "epoch": 0.76,
2542
+ "learning_rate": 1.6099089890375623e-05,
2543
+ "loss": 2.1013,
2544
+ "step": 206000
2545
+ },
2546
+ {
2547
+ "epoch": 0.77,
2548
+ "learning_rate": 1.5862045647570574e-05,
2549
+ "loss": 2.1019,
2550
+ "step": 206500
2551
+ },
2552
+ {
2553
+ "epoch": 0.77,
2554
+ "learning_rate": 1.562643006279392e-05,
2555
+ "loss": 2.0999,
2556
+ "step": 207000
2557
+ },
2558
+ {
2559
+ "epoch": 0.77,
2560
+ "learning_rate": 1.5392252996451884e-05,
2561
+ "loss": 2.1059,
2562
+ "step": 207500
2563
+ },
2564
+ {
2565
+ "epoch": 0.77,
2566
+ "learning_rate": 1.5159524248749296e-05,
2567
+ "loss": 2.1054,
2568
+ "step": 208000
2569
+ },
2570
+ {
2571
+ "epoch": 0.77,
2572
+ "learning_rate": 1.4928253559279532e-05,
2573
+ "loss": 2.1043,
2574
+ "step": 208500
2575
+ },
2576
+ {
2577
+ "epoch": 0.77,
2578
+ "learning_rate": 1.469845060661686e-05,
2579
+ "loss": 2.105,
2580
+ "step": 209000
2581
+ },
2582
+ {
2583
+ "epoch": 0.78,
2584
+ "learning_rate": 1.4470125007911478e-05,
2585
+ "loss": 2.1091,
2586
+ "step": 209500
2587
+ },
2588
+ {
2589
+ "epoch": 0.78,
2590
+ "learning_rate": 1.4243286318486915e-05,
2591
+ "loss": 2.101,
2592
+ "step": 210000
2593
+ },
2594
+ {
2595
+ "epoch": 0.78,
2596
+ "learning_rate": 1.4017944031440283e-05,
2597
+ "loss": 2.1054,
2598
+ "step": 210500
2599
+ },
2600
+ {
2601
+ "epoch": 0.78,
2602
+ "learning_rate": 1.3794107577244886e-05,
2603
+ "loss": 2.0959,
2604
+ "step": 211000
2605
+ },
2606
+ {
2607
+ "epoch": 0.78,
2608
+ "learning_rate": 1.3571786323355596e-05,
2609
+ "loss": 2.0913,
2610
+ "step": 211500
2611
+ },
2612
+ {
2613
+ "epoch": 0.79,
2614
+ "learning_rate": 1.335098957381687e-05,
2615
+ "loss": 2.0924,
2616
+ "step": 212000
2617
+ },
2618
+ {
2619
+ "epoch": 0.79,
2620
+ "learning_rate": 1.3131726568873315e-05,
2621
+ "loss": 2.0939,
2622
+ "step": 212500
2623
+ },
2624
+ {
2625
+ "epoch": 0.79,
2626
+ "learning_rate": 1.2914006484583013e-05,
2627
+ "loss": 2.0917,
2628
+ "step": 213000
2629
+ },
2630
+ {
2631
+ "epoch": 0.79,
2632
+ "learning_rate": 1.2697838432433545e-05,
2633
+ "loss": 2.0899,
2634
+ "step": 213500
2635
+ },
2636
+ {
2637
+ "epoch": 0.79,
2638
+ "learning_rate": 1.2483231458960599e-05,
2639
+ "loss": 2.0961,
2640
+ "step": 214000
2641
+ },
2642
+ {
2643
+ "epoch": 0.8,
2644
+ "learning_rate": 1.2270194545369473e-05,
2645
+ "loss": 2.0955,
2646
+ "step": 214500
2647
+ },
2648
+ {
2649
+ "epoch": 0.8,
2650
+ "learning_rate": 1.2058736607159133e-05,
2651
+ "loss": 2.0848,
2652
+ "step": 215000
2653
+ },
2654
+ {
2655
+ "epoch": 0.8,
2656
+ "learning_rate": 1.1848866493749111e-05,
2657
+ "loss": 2.0841,
2658
+ "step": 215500
2659
+ },
2660
+ {
2661
+ "epoch": 0.8,
2662
+ "learning_rate": 1.1640592988109261e-05,
2663
+ "loss": 2.088,
2664
+ "step": 216000
2665
+ },
2666
+ {
2667
+ "epoch": 0.8,
2668
+ "learning_rate": 1.1433924806392054e-05,
2669
+ "loss": 2.092,
2670
+ "step": 216500
2671
+ },
2672
+ {
2673
+ "epoch": 0.8,
2674
+ "learning_rate": 1.1228870597567887e-05,
2675
+ "loss": 2.0807,
2676
+ "step": 217000
2677
+ },
2678
+ {
2679
+ "epoch": 0.81,
2680
+ "learning_rate": 1.102543894306316e-05,
2681
+ "loss": 2.0881,
2682
+ "step": 217500
2683
+ },
2684
+ {
2685
+ "epoch": 0.81,
2686
+ "learning_rate": 1.0823638356401056e-05,
2687
+ "loss": 2.084,
2688
+ "step": 218000
2689
+ },
2690
+ {
2691
+ "epoch": 0.81,
2692
+ "learning_rate": 1.0623477282845312e-05,
2693
+ "loss": 2.0816,
2694
+ "step": 218500
2695
+ },
2696
+ {
2697
+ "epoch": 0.81,
2698
+ "learning_rate": 1.0424964099046813e-05,
2699
+ "loss": 2.0868,
2700
+ "step": 219000
2701
+ },
2702
+ {
2703
+ "epoch": 0.81,
2704
+ "learning_rate": 1.0228107112692953e-05,
2705
+ "loss": 2.0883,
2706
+ "step": 219500
2707
+ },
2708
+ {
2709
+ "epoch": 0.82,
2710
+ "learning_rate": 1.0032914562160051e-05,
2711
+ "loss": 2.0836,
2712
+ "step": 220000
2713
+ },
2714
+ {
2715
+ "epoch": 0.82,
2716
+ "learning_rate": 9.839394616168506e-06,
2717
+ "loss": 2.0877,
2718
+ "step": 220500
2719
+ },
2720
+ {
2721
+ "epoch": 0.82,
2722
+ "learning_rate": 9.647555373440976e-06,
2723
+ "loss": 2.082,
2724
+ "step": 221000
2725
+ },
2726
+ {
2727
+ "epoch": 0.82,
2728
+ "learning_rate": 9.457404862363428e-06,
2729
+ "loss": 2.0798,
2730
+ "step": 221500
2731
+ },
2732
+ {
2733
+ "epoch": 0.82,
2734
+ "learning_rate": 9.26895104064921e-06,
2735
+ "loss": 2.0855,
2736
+ "step": 222000
2737
+ },
2738
+ {
2739
+ "epoch": 0.82,
2740
+ "learning_rate": 9.082201795005968e-06,
2741
+ "loss": 2.0892,
2742
+ "step": 222500
2743
+ },
2744
+ {
2745
+ "epoch": 0.83,
2746
+ "learning_rate": 8.897164940805591e-06,
2747
+ "loss": 2.0816,
2748
+ "step": 223000
2749
+ },
2750
+ {
2751
+ "epoch": 0.83,
2752
+ "learning_rate": 8.71384822175716e-06,
2753
+ "loss": 2.0886,
2754
+ "step": 223500
2755
+ },
2756
+ {
2757
+ "epoch": 0.83,
2758
+ "learning_rate": 8.532259309582886e-06,
2759
+ "loss": 2.0848,
2760
+ "step": 224000
2761
+ },
2762
+ {
2763
+ "epoch": 0.83,
2764
+ "learning_rate": 8.35240580369701e-06,
2765
+ "loss": 2.0806,
2766
+ "step": 224500
2767
+ },
2768
+ {
2769
+ "epoch": 0.83,
2770
+ "learning_rate": 8.174295230887846e-06,
2771
+ "loss": 2.0845,
2772
+ "step": 225000
2773
+ },
2774
+ {
2775
+ "epoch": 0.83,
2776
+ "eval_loss": 1.9471186399459839,
2777
+ "eval_runtime": 284.8288,
2778
+ "eval_samples_per_second": 351.088,
2779
+ "eval_steps_per_second": 43.886,
2780
+ "step": 225000
2781
+ },
2782
+ {
2783
+ "epoch": 0.84,
2784
+ "learning_rate": 7.997935045002724e-06,
2785
+ "loss": 2.0879,
2786
+ "step": 225500
2787
+ },
2788
+ {
2789
+ "epoch": 0.84,
2790
+ "learning_rate": 7.823332626636065e-06,
2791
+ "loss": 2.0804,
2792
+ "step": 226000
2793
+ },
2794
+ {
2795
+ "epoch": 0.84,
2796
+ "learning_rate": 7.650495282820502e-06,
2797
+ "loss": 2.0835,
2798
+ "step": 226500
2799
+ },
2800
+ {
2801
+ "epoch": 0.84,
2802
+ "learning_rate": 7.479430246721092e-06,
2803
+ "loss": 2.0834,
2804
+ "step": 227000
2805
+ },
2806
+ {
2807
+ "epoch": 0.84,
2808
+ "learning_rate": 7.310144677332631e-06,
2809
+ "loss": 2.0755,
2810
+ "step": 227500
2811
+ },
2812
+ {
2813
+ "epoch": 0.85,
2814
+ "learning_rate": 7.142645659180036e-06,
2815
+ "loss": 2.0744,
2816
+ "step": 228000
2817
+ },
2818
+ {
2819
+ "epoch": 0.85,
2820
+ "learning_rate": 6.9769402020218314e-06,
2821
+ "loss": 2.073,
2822
+ "step": 228500
2823
+ },
2824
+ {
2825
+ "epoch": 0.85,
2826
+ "learning_rate": 6.813035240556842e-06,
2827
+ "loss": 2.075,
2828
+ "step": 229000
2829
+ },
2830
+ {
2831
+ "epoch": 0.85,
2832
+ "learning_rate": 6.650937634133952e-06,
2833
+ "loss": 2.0767,
2834
+ "step": 229500
2835
+ },
2836
+ {
2837
+ "epoch": 0.85,
2838
+ "learning_rate": 6.4906541664650325e-06,
2839
+ "loss": 2.0683,
2840
+ "step": 230000
2841
+ },
2842
+ {
2843
+ "epoch": 0.85,
2844
+ "learning_rate": 6.332191545341093e-06,
2845
+ "loss": 2.0783,
2846
+ "step": 230500
2847
+ },
2848
+ {
2849
+ "epoch": 0.86,
2850
+ "learning_rate": 6.175556402351546e-06,
2851
+ "loss": 2.0665,
2852
+ "step": 231000
2853
+ },
2854
+ {
2855
+ "epoch": 0.86,
2856
+ "learning_rate": 6.020755292606617e-06,
2857
+ "loss": 2.0818,
2858
+ "step": 231500
2859
+ },
2860
+ {
2861
+ "epoch": 0.86,
2862
+ "learning_rate": 5.867794694463102e-06,
2863
+ "loss": 2.0693,
2864
+ "step": 232000
2865
+ },
2866
+ {
2867
+ "epoch": 0.86,
2868
+ "learning_rate": 5.716681009253189e-06,
2869
+ "loss": 2.0705,
2870
+ "step": 232500
2871
+ },
2872
+ {
2873
+ "epoch": 0.86,
2874
+ "learning_rate": 5.567420561016629e-06,
2875
+ "loss": 2.0818,
2876
+ "step": 233000
2877
+ },
2878
+ {
2879
+ "epoch": 0.87,
2880
+ "learning_rate": 5.420019596236003e-06,
2881
+ "loss": 2.0723,
2882
+ "step": 233500
2883
+ },
2884
+ {
2885
+ "epoch": 0.87,
2886
+ "learning_rate": 5.274484283575371e-06,
2887
+ "loss": 2.0833,
2888
+ "step": 234000
2889
+ },
2890
+ {
2891
+ "epoch": 0.87,
2892
+ "learning_rate": 5.1308207136220755e-06,
2893
+ "loss": 2.0764,
2894
+ "step": 234500
2895
+ },
2896
+ {
2897
+ "epoch": 0.87,
2898
+ "learning_rate": 4.989034898631872e-06,
2899
+ "loss": 2.0724,
2900
+ "step": 235000
2901
+ },
2902
+ {
2903
+ "epoch": 0.87,
2904
+ "learning_rate": 4.84913277227731e-06,
2905
+ "loss": 2.0718,
2906
+ "step": 235500
2907
+ },
2908
+ {
2909
+ "epoch": 0.87,
2910
+ "learning_rate": 4.711120189399443e-06,
2911
+ "loss": 2.0752,
2912
+ "step": 236000
2913
+ },
2914
+ {
2915
+ "epoch": 0.88,
2916
+ "learning_rate": 4.575002925762739e-06,
2917
+ "loss": 2.0718,
2918
+ "step": 236500
2919
+ },
2920
+ {
2921
+ "epoch": 0.88,
2922
+ "learning_rate": 4.440786677813458e-06,
2923
+ "loss": 2.0747,
2924
+ "step": 237000
2925
+ },
2926
+ {
2927
+ "epoch": 0.88,
2928
+ "learning_rate": 4.308477062441168e-06,
2929
+ "loss": 2.0714,
2930
+ "step": 237500
2931
+ },
2932
+ {
2933
+ "epoch": 0.88,
2934
+ "learning_rate": 4.17807961674373e-06,
2935
+ "loss": 2.0652,
2936
+ "step": 238000
2937
+ },
2938
+ {
2939
+ "epoch": 0.88,
2940
+ "learning_rate": 4.049599797795589e-06,
2941
+ "loss": 2.0726,
2942
+ "step": 238500
2943
+ },
2944
+ {
2945
+ "epoch": 0.89,
2946
+ "learning_rate": 3.923042982419334e-06,
2947
+ "loss": 2.0692,
2948
+ "step": 239000
2949
+ },
2950
+ {
2951
+ "epoch": 0.89,
2952
+ "learning_rate": 3.798414466960759e-06,
2953
+ "loss": 2.076,
2954
+ "step": 239500
2955
+ },
2956
+ {
2957
+ "epoch": 0.89,
2958
+ "learning_rate": 3.6757194670671513e-06,
2959
+ "loss": 2.0695,
2960
+ "step": 240000
2961
+ },
2962
+ {
2963
+ "epoch": 0.89,
2964
+ "learning_rate": 3.55496311746904e-06,
2965
+ "loss": 2.066,
2966
+ "step": 240500
2967
+ },
2968
+ {
2969
+ "epoch": 0.89,
2970
+ "learning_rate": 3.436150471765326e-06,
2971
+ "loss": 2.0671,
2972
+ "step": 241000
2973
+ },
2974
+ {
2975
+ "epoch": 0.9,
2976
+ "learning_rate": 3.319286502211766e-06,
2977
+ "loss": 2.0682,
2978
+ "step": 241500
2979
+ },
2980
+ {
2981
+ "epoch": 0.9,
2982
+ "learning_rate": 3.2043760995128903e-06,
2983
+ "loss": 2.0716,
2984
+ "step": 242000
2985
+ },
2986
+ {
2987
+ "epoch": 0.9,
2988
+ "learning_rate": 3.091424072617366e-06,
2989
+ "loss": 2.0684,
2990
+ "step": 242500
2991
+ },
2992
+ {
2993
+ "epoch": 0.9,
2994
+ "learning_rate": 2.9804351485166747e-06,
2995
+ "loss": 2.0631,
2996
+ "step": 243000
2997
+ },
2998
+ {
2999
+ "epoch": 0.9,
3000
+ "learning_rate": 2.8714139720473597e-06,
3001
+ "loss": 2.0686,
3002
+ "step": 243500
3003
+ },
3004
+ {
3005
+ "epoch": 0.9,
3006
+ "learning_rate": 2.7643651056965924e-06,
3007
+ "loss": 2.0651,
3008
+ "step": 244000
3009
+ },
3010
+ {
3011
+ "epoch": 0.91,
3012
+ "learning_rate": 2.659293029411264e-06,
3013
+ "loss": 2.0695,
3014
+ "step": 244500
3015
+ },
3016
+ {
3017
+ "epoch": 0.91,
3018
+ "learning_rate": 2.556202140410474e-06,
3019
+ "loss": 2.0704,
3020
+ "step": 245000
3021
+ },
3022
+ {
3023
+ "epoch": 0.91,
3024
+ "learning_rate": 2.4550967530015623e-06,
3025
+ "loss": 2.0668,
3026
+ "step": 245500
3027
+ },
3028
+ {
3029
+ "epoch": 0.91,
3030
+ "learning_rate": 2.3559810983994732e-06,
3031
+ "loss": 2.0656,
3032
+ "step": 246000
3033
+ },
3034
+ {
3035
+ "epoch": 0.91,
3036
+ "learning_rate": 2.258859324549778e-06,
3037
+ "loss": 2.0691,
3038
+ "step": 246500
3039
+ },
3040
+ {
3041
+ "epoch": 0.92,
3042
+ "learning_rate": 2.1637354959549884e-06,
3043
+ "loss": 2.0589,
3044
+ "step": 247000
3045
+ },
3046
+ {
3047
+ "epoch": 0.92,
3048
+ "learning_rate": 2.0706135935045333e-06,
3049
+ "loss": 2.067,
3050
+ "step": 247500
3051
+ },
3052
+ {
3053
+ "epoch": 0.92,
3054
+ "learning_rate": 1.9794975143081264e-06,
3055
+ "loss": 2.0654,
3056
+ "step": 248000
3057
+ },
3058
+ {
3059
+ "epoch": 0.92,
3060
+ "learning_rate": 1.8903910715326823e-06,
3061
+ "loss": 2.0635,
3062
+ "step": 248500
3063
+ },
3064
+ {
3065
+ "epoch": 0.92,
3066
+ "learning_rate": 1.8032979942427475e-06,
3067
+ "loss": 2.0673,
3068
+ "step": 249000
3069
+ },
3070
+ {
3071
+ "epoch": 0.92,
3072
+ "learning_rate": 1.7182219272444011e-06,
3073
+ "loss": 2.0662,
3074
+ "step": 249500
3075
+ },
3076
+ {
3077
+ "epoch": 0.93,
3078
+ "learning_rate": 1.635166430932772e-06,
3079
+ "loss": 2.0697,
3080
+ "step": 250000
3081
+ },
3082
+ {
3083
+ "epoch": 0.93,
3084
+ "eval_loss": 1.9344754219055176,
3085
+ "eval_runtime": 284.7729,
3086
+ "eval_samples_per_second": 351.157,
3087
+ "eval_steps_per_second": 43.895,
3088
+ "step": 250000
3089
+ },
3090
+ {
3091
+ "epoch": 0.93,
3092
+ "learning_rate": 1.5541349811430016e-06,
3093
+ "loss": 2.061,
3094
+ "step": 250500
3095
+ },
3096
+ {
3097
+ "epoch": 0.93,
3098
+ "learning_rate": 1.475130969004812e-06,
3099
+ "loss": 2.0696,
3100
+ "step": 251000
3101
+ },
3102
+ {
3103
+ "epoch": 0.93,
3104
+ "learning_rate": 1.3981577008005563e-06,
3105
+ "loss": 2.0691,
3106
+ "step": 251500
3107
+ },
3108
+ {
3109
+ "epoch": 0.93,
3110
+ "learning_rate": 1.3232183978268698e-06,
3111
+ "loss": 2.0622,
3112
+ "step": 252000
3113
+ },
3114
+ {
3115
+ "epoch": 0.94,
3116
+ "learning_rate": 1.2503161962598653e-06,
3117
+ "loss": 2.0714,
3118
+ "step": 252500
3119
+ },
3120
+ {
3121
+ "epoch": 0.94,
3122
+ "learning_rate": 1.1794541470238729e-06,
3123
+ "loss": 2.0664,
3124
+ "step": 253000
3125
+ },
3126
+ {
3127
+ "epoch": 0.94,
3128
+ "learning_rate": 1.1106352156637633e-06,
3129
+ "loss": 2.0606,
3130
+ "step": 253500
3131
+ },
3132
+ {
3133
+ "epoch": 0.94,
3134
+ "learning_rate": 1.0438622822208478e-06,
3135
+ "loss": 2.0645,
3136
+ "step": 254000
3137
+ },
3138
+ {
3139
+ "epoch": 0.94,
3140
+ "learning_rate": 9.791381411123513e-07,
3141
+ "loss": 2.0724,
3142
+ "step": 254500
3143
+ },
3144
+ {
3145
+ "epoch": 0.95,
3146
+ "learning_rate": 9.164655010144518e-07,
3147
+ "loss": 2.0629,
3148
+ "step": 255000
3149
+ },
3150
+ {
3151
+ "epoch": 0.95,
3152
+ "learning_rate": 8.558469847489314e-07,
3153
+ "loss": 2.0688,
3154
+ "step": 255500
3155
+ },
3156
+ {
3157
+ "epoch": 0.95,
3158
+ "learning_rate": 7.972851291734252e-07,
3159
+ "loss": 2.0566,
3160
+ "step": 256000
3161
+ },
3162
+ {
3163
+ "epoch": 0.95,
3164
+ "learning_rate": 7.407823850752338e-07,
3165
+ "loss": 2.0689,
3166
+ "step": 256500
3167
+ },
3168
+ {
3169
+ "epoch": 0.95,
3170
+ "learning_rate": 6.863411170687673e-07,
3171
+ "loss": 2.071,
3172
+ "step": 257000
3173
+ },
3174
+ {
3175
+ "epoch": 0.95,
3176
+ "learning_rate": 6.339636034966123e-07,
3177
+ "loss": 2.0665,
3178
+ "step": 257500
3179
+ },
3180
+ {
3181
+ "epoch": 0.96,
3182
+ "learning_rate": 5.836520363341258e-07,
3183
+ "loss": 2.0606,
3184
+ "step": 258000
3185
+ },
3186
+ {
3187
+ "epoch": 0.96,
3188
+ "learning_rate": 5.354085210977633e-07,
3189
+ "loss": 2.0646,
3190
+ "step": 258500
3191
+ },
3192
+ {
3193
+ "epoch": 0.96,
3194
+ "learning_rate": 4.892350767569276e-07,
3195
+ "loss": 2.0663,
3196
+ "step": 259000
3197
+ },
3198
+ {
3199
+ "epoch": 0.96,
3200
+ "learning_rate": 4.451336356494862e-07,
3201
+ "loss": 2.0683,
3202
+ "step": 259500
3203
+ },
3204
+ {
3205
+ "epoch": 0.96,
3206
+ "learning_rate": 4.031060434009026e-07,
3207
+ "loss": 2.0595,
3208
+ "step": 260000
3209
+ },
3210
+ {
3211
+ "epoch": 0.97,
3212
+ "learning_rate": 3.631540588470039e-07,
3213
+ "loss": 2.0691,
3214
+ "step": 260500
3215
+ },
3216
+ {
3217
+ "epoch": 0.97,
3218
+ "learning_rate": 3.252793539603671e-07,
3219
+ "loss": 2.0651,
3220
+ "step": 261000
3221
+ },
3222
+ {
3223
+ "epoch": 0.97,
3224
+ "learning_rate": 2.8948351378034753e-07,
3225
+ "loss": 2.0611,
3226
+ "step": 261500
3227
+ },
3228
+ {
3229
+ "epoch": 0.97,
3230
+ "learning_rate": 2.557680363467485e-07,
3231
+ "loss": 2.0733,
3232
+ "step": 262000
3233
+ },
3234
+ {
3235
+ "epoch": 0.97,
3236
+ "learning_rate": 2.2413433263713257e-07,
3237
+ "loss": 2.0591,
3238
+ "step": 262500
3239
+ },
3240
+ {
3241
+ "epoch": 0.97,
3242
+ "learning_rate": 1.9458372650776325e-07,
3243
+ "loss": 2.0626,
3244
+ "step": 263000
3245
+ },
3246
+ {
3247
+ "epoch": 0.98,
3248
+ "learning_rate": 1.6711745463821593e-07,
3249
+ "loss": 2.0669,
3250
+ "step": 263500
3251
+ },
3252
+ {
3253
+ "epoch": 0.98,
3254
+ "learning_rate": 1.4173666647959715e-07,
3255
+ "loss": 2.0701,
3256
+ "step": 264000
3257
+ },
3258
+ {
3259
+ "epoch": 0.98,
3260
+ "learning_rate": 1.1844242420647745e-07,
3261
+ "loss": 2.0679,
3262
+ "step": 264500
3263
+ },
3264
+ {
3265
+ "epoch": 0.98,
3266
+ "learning_rate": 9.723570267241577e-08,
3267
+ "loss": 2.0679,
3268
+ "step": 265000
3269
+ },
3270
+ {
3271
+ "epoch": 0.98,
3272
+ "learning_rate": 7.811738936916447e-08,
3273
+ "loss": 2.0643,
3274
+ "step": 265500
3275
+ },
3276
+ {
3277
+ "epoch": 0.99,
3278
+ "learning_rate": 6.108828438952663e-08,
3279
+ "loss": 2.0615,
3280
+ "step": 266000
3281
+ },
3282
+ {
3283
+ "epoch": 0.99,
3284
+ "learning_rate": 4.614910039389409e-08,
3285
+ "loss": 2.0511,
3286
+ "step": 266500
3287
+ },
3288
+ {
3289
+ "epoch": 0.99,
3290
+ "learning_rate": 3.3300462580387884e-08,
3291
+ "loss": 2.0594,
3292
+ "step": 267000
3293
+ },
3294
+ {
3295
+ "epoch": 0.99,
3296
+ "learning_rate": 2.254290865871811e-08,
3297
+ "loss": 2.0665,
3298
+ "step": 267500
3299
+ },
3300
+ {
3301
+ "epoch": 0.99,
3302
+ "learning_rate": 1.3876888827679679e-08,
3303
+ "loss": 2.0572,
3304
+ "step": 268000
3305
+ },
3306
+ {
3307
+ "epoch": 1.0,
3308
+ "learning_rate": 7.302765756300733e-09,
3309
+ "loss": 2.0682,
3310
+ "step": 268500
3311
+ },
3312
+ {
3313
+ "epoch": 1.0,
3314
+ "learning_rate": 2.820814568671448e-09,
3315
+ "loss": 2.0611,
3316
+ "step": 269000
3317
+ },
3318
+ {
3319
+ "epoch": 1.0,
3320
+ "learning_rate": 4.31222832436573e-10,
3321
+ "loss": 2.0658,
3322
+ "step": 269500
3323
+ },
3324
+ {
3325
+ "epoch": 1.0,
3326
+ "step": 269794,
3327
+ "total_flos": 7.264837577610363e+18,
3328
+ "train_loss": 1.8196985500431868,
3329
+ "train_runtime": 434920.5349,
3330
+ "train_samples_per_second": 148.879,
3331
+ "train_steps_per_second": 0.62
3332
+ }
3333
+ ],
3334
+ "max_steps": 269794,
3335
+ "num_train_epochs": 1,
3336
+ "total_flos": 7.264837577610363e+18,
3337
+ "trial_name": null,
3338
+ "trial_params": null
3339
+ }
arabert-tweet-large/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83aaee1e19b7930b32652274e3decd2676d32ecb73eeec35d9292eebddc4e962
3
+ size 2607
arabert-tweet-large/vocab.txt ADDED
The diff for this file is too large to render. See raw diff