Hugging Face's logo

aubmindlab
/
bert-base-arabertv02-twitter

Fill-Mask Transformers PyTorch TensorBoard Safetensors Arabic bert Inference Endpoints
Model card Files Metrics Community
1
wissamantoun commited on
Commit
c7be764
•
1 Parent(s): 1928744

First commit

Browse files
Files changed (24) hide show
  1. all_results.json +14 -0
  2. config.json +26 -0
  3. eval_results.json +9 -0
  4. pytorch_model.bin +3 -0
  5. runs/Aug29_21-32-23_660e80775921/1630284783.908466/events.out.tfevents.1630284783.660e80775921.3048.1 +3 -0
  6. runs/Aug29_21-32-23_660e80775921/events.out.tfevents.1630284783.660e80775921.3048.0 +3 -0
  7. runs/Aug31_02-06-24_6f9851eb76f4/1630387784.4171464/events.out.tfevents.1630387784.6f9851eb76f4.2942.1 +3 -0
  8. runs/Aug31_02-06-24_6f9851eb76f4/events.out.tfevents.1630387784.6f9851eb76f4.2942.0 +3 -0
  9. runs/Sep04_22-51-06_0fb508bd6187/1630808570.5717866/events.out.tfevents.1630808570.0fb508bd6187.3288.1 +3 -0
  10. runs/Sep04_22-51-06_0fb508bd6187/events.out.tfevents.1630808570.0fb508bd6187.3288.0 +3 -0
  11. runs/Sep06_02-34-50_00002cecfaa1/1630908011.3030038/events.out.tfevents.1630908011.00002cecfaa1.3087.1 +3 -0
  12. runs/Sep06_02-34-50_00002cecfaa1/events.out.tfevents.1630908011.00002cecfaa1.3087.0 +3 -0
  13. runs/Sep07_10-51-50_571b66b4ee56/1631024270.8993132/events.out.tfevents.1631024270.571b66b4ee56.2999.1 +3 -0
  14. runs/Sep07_10-51-50_571b66b4ee56/events.out.tfevents.1631024270.571b66b4ee56.2999.0 +3 -0
  15. runs/Sep08_11-24-31_84a764423b26/1631112740.8695328/events.out.tfevents.1631112740.84a764423b26.3000.1 +3 -0
  16. runs/Sep08_11-24-31_84a764423b26/events.out.tfevents.1631112740.84a764423b26.3000.0 +3 -0
  17. runs/Sep08_11-24-31_84a764423b26/events.out.tfevents.1631157768.84a764423b26.3000.2 +3 -0
  18. special_tokens_map.json +1 -0
  19. tokenizer.json +0 -0
  20. tokenizer_config.json +1 -0
  21. train_results.json +8 -0
  22. trainer_state.json +3135 -0
  23. training_args.bin +3 -0
  24. vocab.txt +0 -0
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 2.060576915740967,
4
+ "eval_runtime": 187.7576,
5
+ "eval_samples": 100000,
6
+ "eval_samples_per_second": 532.602,
7
+ "eval_steps_per_second": 66.575,
8
+ "perplexity": 7.850497579750318,
9
+ "train_loss": 0.24544810387602325,
10
+ "train_runtime": 44838.2229,
11
+ "train_samples": 64750662,
12
+ "train_samples_per_second": 1444.095,
13
+ "train_steps_per_second": 5.641
14
+ }
config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bert-base-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": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "bert",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
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
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 2.060576915740967,
4
+ "eval_runtime": 187.7576,
5
+ "eval_samples": 100000,
6
+ "eval_samples_per_second": 532.602,
7
+ "eval_steps_per_second": 66.575,
8
+ "perplexity": 7.850497579750318
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:923804e1baf1b1256807610f188ca3b8d76657264a16770bacea92fd7ed5ba58
3
+ size 541120363
runs/Aug29_21-32-23_660e80775921/1630284783.908466/events.out.tfevents.1630284783.660e80775921.3048.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5921f7573cf61430392326cdacdbf8b0af559988747051acbc43541bac4abc28
3
+ size 4168
runs/Aug29_21-32-23_660e80775921/events.out.tfevents.1630284783.660e80775921.3048.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:066c4cf09425d1da3aa40b840a76a55e6d33f00f7aa12b7bebcc0d585e9393e7
3
+ size 12202
runs/Aug31_02-06-24_6f9851eb76f4/1630387784.4171464/events.out.tfevents.1630387784.6f9851eb76f4.2942.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:527839a7cca3a80990bf3d7a92cc8c6c063fccf1772a9b0524bdeee666418cf0
3
+ size 4370
runs/Aug31_02-06-24_6f9851eb76f4/events.out.tfevents.1630387784.6f9851eb76f4.2942.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd216a39cb6e54ac2bd0f5c6fcd9e31d46e392c085d7c837105a13269a48b2ee
3
+ size 19658
runs/Sep04_22-51-06_0fb508bd6187/1630808570.5717866/events.out.tfevents.1630808570.0fb508bd6187.3288.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a95234da54bc311794d650f3d7a6b5ce7c98a23cc5d1bb76b46225dc068b7894
3
+ size 4370
runs/Sep04_22-51-06_0fb508bd6187/events.out.tfevents.1630808570.0fb508bd6187.3288.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efe0858f62e6ba70ecb64e41713acac98360a6e7424179f5aa7af1a0612e115e
3
+ size 20649
runs/Sep06_02-34-50_00002cecfaa1/1630908011.3030038/events.out.tfevents.1630908011.00002cecfaa1.3087.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31763d9f73ada7f8f58e730a651f28a1e81b35d309a6137c497acccdb20b943e
3
+ size 4371
runs/Sep06_02-34-50_00002cecfaa1/events.out.tfevents.1630908011.00002cecfaa1.3087.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a31b8a7c5bdb2536354c01221268d1e7b140407831a0683444c4f2132001b03c
3
+ size 20650
runs/Sep07_10-51-50_571b66b4ee56/1631024270.8993132/events.out.tfevents.1631024270.571b66b4ee56.2999.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1e00f2d646afb836c63c10ac7bf5a81097406467f14c97e1c5a71fa9725cc5ab
3
+ size 4371
runs/Sep07_10-51-50_571b66b4ee56/events.out.tfevents.1631024270.571b66b4ee56.2999.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de8b06bc2ca7587ab657cf3c64ec35a99ba224c3b025d3f3748217f2d62522e1
3
+ size 20170
runs/Sep08_11-24-31_84a764423b26/1631112740.8695328/events.out.tfevents.1631112740.84a764423b26.3000.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:49e08745fe4622e12fe227c78a4f2c73a688678a126e9fb903ad86998cc49ba6
3
+ size 4371
runs/Sep08_11-24-31_84a764423b26/events.out.tfevents.1631112740.84a764423b26.3000.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c550d87f7bf41907d20df6328718f9018cac06f4b3ed510a8fe4e895e3a0299c
3
+ size 12574
runs/Sep08_11-24-31_84a764423b26/events.out.tfevents.1631157768.84a764423b26.3000.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02bca86ad1dad1c5e3816403a8ff8cd5c238bc7dd0bd91a10696c9c090cdb998
3
+ size 316
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
1
+ {"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
1
+ {"do_lower_case": false, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "do_basic_tokenize": true, "never_split": ["[بريد]", "[مستخدم]", "[رابط]"], "max_len": 512, "special_tokens_map_file": null, "name_or_path": "/content/drive/MyDrive/arabert-twitter/training/train", "use_fast": true, "from_slow": true, "tokenizer_class": "BertTokenizer"}
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.24544810387602325,
4
+ "train_runtime": 44838.2229,
5
+ "train_samples": 64750662,
6
+ "train_samples_per_second": 1444.095,
7
+ "train_steps_per_second": 5.641
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,3135 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9999980231880047,
5
+ "global_step": 252932,
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": 9.980231840969114e-05,
13
+ "loss": 3.4601,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 9.960463681938229e-05,
19
+ "loss": 3.3095,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 9.940695522907343e-05,
25
+ "loss": 3.2164,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 9.920927363876458e-05,
31
+ "loss": 3.1537,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 9.901159204845572e-05,
37
+ "loss": 3.111,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 9.881391045814687e-05,
43
+ "loss": 3.0745,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.01,
48
+ "learning_rate": 9.8616228867838e-05,
49
+ "loss": 3.042,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 9.841854727752914e-05,
55
+ "loss": 3.031,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.02,
60
+ "learning_rate": 9.822086568722029e-05,
61
+ "loss": 3.0045,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.02,
66
+ "learning_rate": 9.802318409691143e-05,
67
+ "loss": 2.9784,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.02,
72
+ "learning_rate": 9.782550250660258e-05,
73
+ "loss": 2.9605,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.02,
78
+ "learning_rate": 9.762782091629372e-05,
79
+ "loss": 2.9598,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.03,
84
+ "learning_rate": 9.743013932598487e-05,
85
+ "loss": 2.9421,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.03,
90
+ "learning_rate": 9.7232457735676e-05,
91
+ "loss": 2.9273,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.03,
96
+ "learning_rate": 9.703477614536714e-05,
97
+ "loss": 2.9111,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.03,
102
+ "learning_rate": 9.683709455505828e-05,
103
+ "loss": 2.9034,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.03,
108
+ "learning_rate": 9.663941296474943e-05,
109
+ "loss": 2.8881,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "learning_rate": 9.644173137444056e-05,
115
+ "loss": 2.8827,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.04,
120
+ "learning_rate": 9.62440497841317e-05,
121
+ "loss": 2.8804,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.04,
126
+ "learning_rate": 9.604636819382285e-05,
127
+ "loss": 2.8645,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.04,
132
+ "learning_rate": 9.584868660351399e-05,
133
+ "loss": 2.8559,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.04,
138
+ "learning_rate": 9.565100501320512e-05,
139
+ "loss": 2.8503,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.05,
144
+ "learning_rate": 9.545332342289628e-05,
145
+ "loss": 2.8343,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.05,
150
+ "learning_rate": 9.525564183258741e-05,
151
+ "loss": 2.8318,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.05,
156
+ "learning_rate": 9.505796024227856e-05,
157
+ "loss": 2.833,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.05,
162
+ "learning_rate": 9.48602786519697e-05,
163
+ "loss": 2.8119,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.05,
168
+ "learning_rate": 9.466259706166085e-05,
169
+ "loss": 2.8128,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.06,
174
+ "learning_rate": 9.446491547135199e-05,
175
+ "loss": 2.8052,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.06,
180
+ "learning_rate": 9.426723388104312e-05,
181
+ "loss": 2.7941,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.06,
186
+ "learning_rate": 9.406955229073427e-05,
187
+ "loss": 2.794,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.06,
192
+ "learning_rate": 9.387187070042541e-05,
193
+ "loss": 2.7877,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.06,
198
+ "learning_rate": 9.367418911011656e-05,
199
+ "loss": 2.7838,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.07,
204
+ "learning_rate": 9.34765075198077e-05,
205
+ "loss": 2.7682,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.07,
210
+ "learning_rate": 9.327882592949885e-05,
211
+ "loss": 2.7722,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.07,
216
+ "learning_rate": 9.308114433918999e-05,
217
+ "loss": 2.7626,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.07,
222
+ "learning_rate": 9.288346274888112e-05,
223
+ "loss": 2.7602,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.07,
228
+ "learning_rate": 9.268578115857227e-05,
229
+ "loss": 2.7625,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.08,
234
+ "learning_rate": 9.248809956826341e-05,
235
+ "loss": 2.7533,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.08,
240
+ "learning_rate": 9.229041797795456e-05,
241
+ "loss": 2.7464,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.08,
246
+ "learning_rate": 9.20927363876457e-05,
247
+ "loss": 2.7335,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.08,
252
+ "learning_rate": 9.189505479733683e-05,
253
+ "loss": 2.7307,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.08,
258
+ "learning_rate": 9.169737320702798e-05,
259
+ "loss": 2.7359,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.09,
264
+ "learning_rate": 9.149969161671912e-05,
265
+ "loss": 2.7351,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.09,
270
+ "learning_rate": 9.130201002641027e-05,
271
+ "loss": 2.7193,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.09,
276
+ "learning_rate": 9.110432843610141e-05,
277
+ "loss": 2.7194,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.09,
282
+ "learning_rate": 9.090664684579256e-05,
283
+ "loss": 2.7235,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.09,
288
+ "learning_rate": 9.07089652554837e-05,
289
+ "loss": 2.7255,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.09,
294
+ "learning_rate": 9.051128366517483e-05,
295
+ "loss": 2.7085,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.1,
300
+ "learning_rate": 9.031360207486598e-05,
301
+ "loss": 2.7122,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.1,
306
+ "learning_rate": 9.011592048455712e-05,
307
+ "loss": 2.7037,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.1,
312
+ "eval_loss": 2.510829210281372,
313
+ "eval_runtime": 324.1549,
314
+ "eval_samples_per_second": 308.495,
315
+ "eval_steps_per_second": 38.562,
316
+ "step": 25000
317
+ },
318
+ {
319
+ "epoch": 0.1,
320
+ "learning_rate": 8.991823889424826e-05,
321
+ "loss": 2.6986,
322
+ "step": 25500
323
+ },
324
+ {
325
+ "epoch": 0.1,
326
+ "learning_rate": 8.97205573039394e-05,
327
+ "loss": 2.6976,
328
+ "step": 26000
329
+ },
330
+ {
331
+ "epoch": 0.1,
332
+ "learning_rate": 8.952287571363054e-05,
333
+ "loss": 2.6908,
334
+ "step": 26500
335
+ },
336
+ {
337
+ "epoch": 0.11,
338
+ "learning_rate": 8.932519412332168e-05,
339
+ "loss": 2.7012,
340
+ "step": 27000
341
+ },
342
+ {
343
+ "epoch": 0.11,
344
+ "learning_rate": 8.912751253301282e-05,
345
+ "loss": 2.6839,
346
+ "step": 27500
347
+ },
348
+ {
349
+ "epoch": 0.11,
350
+ "learning_rate": 8.892983094270397e-05,
351
+ "loss": 2.6843,
352
+ "step": 28000
353
+ },
354
+ {
355
+ "epoch": 0.11,
356
+ "learning_rate": 8.87321493523951e-05,
357
+ "loss": 2.6721,
358
+ "step": 28500
359
+ },
360
+ {
361
+ "epoch": 0.11,
362
+ "learning_rate": 8.853446776208626e-05,
363
+ "loss": 2.6877,
364
+ "step": 29000
365
+ },
366
+ {
367
+ "epoch": 0.12,
368
+ "learning_rate": 8.83367861717774e-05,
369
+ "loss": 2.6742,
370
+ "step": 29500
371
+ },
372
+ {
373
+ "epoch": 0.12,
374
+ "learning_rate": 8.813910458146854e-05,
375
+ "loss": 2.6696,
376
+ "step": 30000
377
+ },
378
+ {
379
+ "epoch": 0.12,
380
+ "learning_rate": 8.794142299115968e-05,
381
+ "loss": 2.6703,
382
+ "step": 30500
383
+ },
384
+ {
385
+ "epoch": 0.12,
386
+ "learning_rate": 8.774374140085082e-05,
387
+ "loss": 2.6715,
388
+ "step": 31000
389
+ },
390
+ {
391
+ "epoch": 0.12,
392
+ "learning_rate": 8.754605981054197e-05,
393
+ "loss": 2.6629,
394
+ "step": 31500
395
+ },
396
+ {
397
+ "epoch": 0.13,
398
+ "learning_rate": 8.73483782202331e-05,
399
+ "loss": 2.6651,
400
+ "step": 32000
401
+ },
402
+ {
403
+ "epoch": 0.13,
404
+ "learning_rate": 8.715069662992426e-05,
405
+ "loss": 2.6481,
406
+ "step": 32500
407
+ },
408
+ {
409
+ "epoch": 0.13,
410
+ "learning_rate": 8.695301503961539e-05,
411
+ "loss": 2.656,
412
+ "step": 33000
413
+ },
414
+ {
415
+ "epoch": 0.13,
416
+ "learning_rate": 8.675533344930654e-05,
417
+ "loss": 2.6485,
418
+ "step": 33500
419
+ },
420
+ {
421
+ "epoch": 0.13,
422
+ "learning_rate": 8.655765185899768e-05,
423
+ "loss": 2.6547,
424
+ "step": 34000
425
+ },
426
+ {
427
+ "epoch": 0.14,
428
+ "learning_rate": 8.635997026868882e-05,
429
+ "loss": 2.6508,
430
+ "step": 34500
431
+ },
432
+ {
433
+ "epoch": 0.14,
434
+ "learning_rate": 8.616228867837997e-05,
435
+ "loss": 2.6453,
436
+ "step": 35000
437
+ },
438
+ {
439
+ "epoch": 0.14,
440
+ "learning_rate": 8.59646070880711e-05,
441
+ "loss": 2.6458,
442
+ "step": 35500
443
+ },
444
+ {
445
+ "epoch": 0.14,
446
+ "learning_rate": 8.576692549776225e-05,
447
+ "loss": 2.6443,
448
+ "step": 36000
449
+ },
450
+ {
451
+ "epoch": 0.14,
452
+ "learning_rate": 8.556924390745339e-05,
453
+ "loss": 2.6434,
454
+ "step": 36500
455
+ },
456
+ {
457
+ "epoch": 0.15,
458
+ "learning_rate": 8.537156231714454e-05,
459
+ "loss": 2.638,
460
+ "step": 37000
461
+ },
462
+ {
463
+ "epoch": 0.15,
464
+ "learning_rate": 8.517388072683568e-05,
465
+ "loss": 2.6238,
466
+ "step": 37500
467
+ },
468
+ {
469
+ "epoch": 0.15,
470
+ "learning_rate": 8.497619913652682e-05,
471
+ "loss": 2.6288,
472
+ "step": 38000
473
+ },
474
+ {
475
+ "epoch": 0.15,
476
+ "learning_rate": 8.477851754621797e-05,
477
+ "loss": 2.6309,
478
+ "step": 38500
479
+ },
480
+ {
481
+ "epoch": 0.15,
482
+ "learning_rate": 8.45808359559091e-05,
483
+ "loss": 2.6227,
484
+ "step": 39000
485
+ },
486
+ {
487
+ "epoch": 0.16,
488
+ "learning_rate": 8.438315436560025e-05,
489
+ "loss": 2.6248,
490
+ "step": 39500
491
+ },
492
+ {
493
+ "epoch": 0.16,
494
+ "learning_rate": 8.418547277529139e-05,
495
+ "loss": 2.6211,
496
+ "step": 40000
497
+ },
498
+ {
499
+ "epoch": 0.16,
500
+ "learning_rate": 8.398779118498254e-05,
501
+ "loss": 2.6227,
502
+ "step": 40500
503
+ },
504
+ {
505
+ "epoch": 0.16,
506
+ "learning_rate": 8.379010959467368e-05,
507
+ "loss": 2.6152,
508
+ "step": 41000
509
+ },
510
+ {
511
+ "epoch": 0.16,
512
+ "learning_rate": 8.359242800436481e-05,
513
+ "loss": 2.6184,
514
+ "step": 41500
515
+ },
516
+ {
517
+ "epoch": 0.17,
518
+ "learning_rate": 8.339474641405595e-05,
519
+ "loss": 2.6167,
520
+ "step": 42000
521
+ },
522
+ {
523
+ "epoch": 0.17,
524
+ "learning_rate": 8.31970648237471e-05,
525
+ "loss": 2.6119,
526
+ "step": 42500
527
+ },
528
+ {
529
+ "epoch": 0.17,
530
+ "learning_rate": 8.299938323343824e-05,
531
+ "loss": 2.619,
532
+ "step": 43000
533
+ },
534
+ {
535
+ "epoch": 0.17,
536
+ "learning_rate": 8.280170164312938e-05,
537
+ "loss": 2.6119,
538
+ "step": 43500
539
+ },
540
+ {
541
+ "epoch": 0.17,
542
+ "learning_rate": 8.260402005282053e-05,
543
+ "loss": 2.6075,
544
+ "step": 44000
545
+ },
546
+ {
547
+ "epoch": 0.18,
548
+ "learning_rate": 8.240633846251166e-05,
549
+ "loss": 2.6067,
550
+ "step": 44500
551
+ },
552
+ {
553
+ "epoch": 0.18,
554
+ "learning_rate": 8.22086568722028e-05,
555
+ "loss": 2.6144,
556
+ "step": 45000
557
+ },
558
+ {
559
+ "epoch": 0.18,
560
+ "learning_rate": 8.201097528189395e-05,
561
+ "loss": 2.5938,
562
+ "step": 45500
563
+ },
564
+ {
565
+ "epoch": 0.18,
566
+ "learning_rate": 8.181329369158509e-05,
567
+ "loss": 2.5939,
568
+ "step": 46000
569
+ },
570
+ {
571
+ "epoch": 0.18,
572
+ "learning_rate": 8.161561210127624e-05,
573
+ "loss": 2.596,
574
+ "step": 46500
575
+ },
576
+ {
577
+ "epoch": 0.19,
578
+ "learning_rate": 8.141793051096737e-05,
579
+ "loss": 2.5901,
580
+ "step": 47000
581
+ },
582
+ {
583
+ "epoch": 0.19,
584
+ "learning_rate": 8.122024892065852e-05,
585
+ "loss": 2.5888,
586
+ "step": 47500
587
+ },
588
+ {
589
+ "epoch": 0.19,
590
+ "learning_rate": 8.102256733034966e-05,
591
+ "loss": 2.5927,
592
+ "step": 48000
593
+ },
594
+ {
595
+ "epoch": 0.19,
596
+ "learning_rate": 8.08248857400408e-05,
597
+ "loss": 2.5894,
598
+ "step": 48500
599
+ },
600
+ {
601
+ "epoch": 0.19,
602
+ "learning_rate": 8.062720414973195e-05,
603
+ "loss": 2.5829,
604
+ "step": 49000
605
+ },
606
+ {
607
+ "epoch": 0.2,
608
+ "learning_rate": 8.042952255942309e-05,
609
+ "loss": 2.5738,
610
+ "step": 49500
611
+ },
612
+ {
613
+ "epoch": 0.2,
614
+ "learning_rate": 8.023184096911424e-05,
615
+ "loss": 2.5794,
616
+ "step": 50000
617
+ },
618
+ {
619
+ "epoch": 0.2,
620
+ "eval_loss": 2.399911880493164,
621
+ "eval_runtime": 187.1469,
622
+ "eval_samples_per_second": 534.34,
623
+ "eval_steps_per_second": 66.792,
624
+ "step": 50000
625
+ },
626
+ {
627
+ "epoch": 0.2,
628
+ "learning_rate": 8.003415937880537e-05,
629
+ "loss": 2.5825,
630
+ "step": 50500
631
+ },
632
+ {
633
+ "epoch": 0.2,
634
+ "learning_rate": 7.983647778849652e-05,
635
+ "loss": 2.5859,
636
+ "step": 51000
637
+ },
638
+ {
639
+ "epoch": 0.2,
640
+ "learning_rate": 7.963879619818766e-05,
641
+ "loss": 2.5812,
642
+ "step": 51500
643
+ },
644
+ {
645
+ "epoch": 0.21,
646
+ "learning_rate": 7.94411146078788e-05,
647
+ "loss": 2.5763,
648
+ "step": 52000
649
+ },
650
+ {
651
+ "epoch": 0.21,
652
+ "learning_rate": 7.924343301756995e-05,
653
+ "loss": 2.5803,
654
+ "step": 52500
655
+ },
656
+ {
657
+ "epoch": 0.21,
658
+ "learning_rate": 7.904575142726108e-05,
659
+ "loss": 2.5743,
660
+ "step": 53000
661
+ },
662
+ {
663
+ "epoch": 0.21,
664
+ "learning_rate": 7.884806983695224e-05,
665
+ "loss": 2.572,
666
+ "step": 53500
667
+ },
668
+ {
669
+ "epoch": 0.21,
670
+ "learning_rate": 7.865038824664337e-05,
671
+ "loss": 2.5717,
672
+ "step": 54000
673
+ },
674
+ {
675
+ "epoch": 0.22,
676
+ "learning_rate": 7.845270665633452e-05,
677
+ "loss": 2.5614,
678
+ "step": 54500
679
+ },
680
+ {
681
+ "epoch": 0.22,
682
+ "learning_rate": 7.825502506602566e-05,
683
+ "loss": 2.5678,
684
+ "step": 55000
685
+ },
686
+ {
687
+ "epoch": 0.22,
688
+ "learning_rate": 7.80573434757168e-05,
689
+ "loss": 2.5539,
690
+ "step": 55500
691
+ },
692
+ {
693
+ "epoch": 0.22,
694
+ "learning_rate": 7.785966188540795e-05,
695
+ "loss": 2.5632,
696
+ "step": 56000
697
+ },
698
+ {
699
+ "epoch": 0.22,
700
+ "learning_rate": 7.766198029509908e-05,
701
+ "loss": 2.5617,
702
+ "step": 56500
703
+ },
704
+ {
705
+ "epoch": 0.23,
706
+ "learning_rate": 7.746429870479023e-05,
707
+ "loss": 2.5533,
708
+ "step": 57000
709
+ },
710
+ {
711
+ "epoch": 0.23,
712
+ "learning_rate": 7.726661711448137e-05,
713
+ "loss": 2.5501,
714
+ "step": 57500
715
+ },
716
+ {
717
+ "epoch": 0.23,
718
+ "learning_rate": 7.706893552417251e-05,
719
+ "loss": 2.5566,
720
+ "step": 58000
721
+ },
722
+ {
723
+ "epoch": 0.23,
724
+ "learning_rate": 7.687125393386365e-05,
725
+ "loss": 2.5597,
726
+ "step": 58500
727
+ },
728
+ {
729
+ "epoch": 0.23,
730
+ "learning_rate": 7.66735723435548e-05,
731
+ "loss": 2.5513,
732
+ "step": 59000
733
+ },
734
+ {
735
+ "epoch": 0.24,
736
+ "learning_rate": 7.647589075324593e-05,
737
+ "loss": 2.5424,
738
+ "step": 59500
739
+ },
740
+ {
741
+ "epoch": 0.24,
742
+ "learning_rate": 7.627820916293707e-05,
743
+ "loss": 2.5553,
744
+ "step": 60000
745
+ },
746
+ {
747
+ "epoch": 0.24,
748
+ "learning_rate": 7.608052757262822e-05,
749
+ "loss": 2.5458,
750
+ "step": 60500
751
+ },
752
+ {
753
+ "epoch": 0.24,
754
+ "learning_rate": 7.588284598231936e-05,
755
+ "loss": 2.5371,
756
+ "step": 61000
757
+ },
758
+ {
759
+ "epoch": 0.24,
760
+ "learning_rate": 7.56851643920105e-05,
761
+ "loss": 2.5474,
762
+ "step": 61500
763
+ },
764
+ {
765
+ "epoch": 0.25,
766
+ "learning_rate": 7.548748280170164e-05,
767
+ "loss": 2.539,
768
+ "step": 62000
769
+ },
770
+ {
771
+ "epoch": 0.25,
772
+ "learning_rate": 7.528980121139278e-05,
773
+ "loss": 2.5432,
774
+ "step": 62500
775
+ },
776
+ {
777
+ "epoch": 0.25,
778
+ "learning_rate": 7.509211962108393e-05,
779
+ "loss": 2.5308,
780
+ "step": 63000
781
+ },
782
+ {
783
+ "epoch": 0.25,
784
+ "learning_rate": 7.489443803077507e-05,
785
+ "loss": 2.5442,
786
+ "step": 63500
787
+ },
788
+ {
789
+ "epoch": 0.25,
790
+ "learning_rate": 7.469675644046622e-05,
791
+ "loss": 2.5345,
792
+ "step": 64000
793
+ },
794
+ {
795
+ "epoch": 0.26,
796
+ "learning_rate": 7.449907485015736e-05,
797
+ "loss": 2.5381,
798
+ "step": 64500
799
+ },
800
+ {
801
+ "epoch": 0.26,
802
+ "learning_rate": 7.430139325984849e-05,
803
+ "loss": 2.5296,
804
+ "step": 65000
805
+ },
806
+ {
807
+ "epoch": 0.26,
808
+ "learning_rate": 7.410371166953964e-05,
809
+ "loss": 2.5322,
810
+ "step": 65500
811
+ },
812
+ {
813
+ "epoch": 0.26,
814
+ "learning_rate": 7.390603007923078e-05,
815
+ "loss": 2.5296,
816
+ "step": 66000
817
+ },
818
+ {
819
+ "epoch": 0.26,
820
+ "learning_rate": 7.370834848892193e-05,
821
+ "loss": 2.5342,
822
+ "step": 66500
823
+ },
824
+ {
825
+ "epoch": 0.26,
826
+ "learning_rate": 7.351066689861307e-05,
827
+ "loss": 2.526,
828
+ "step": 67000
829
+ },
830
+ {
831
+ "epoch": 0.27,
832
+ "learning_rate": 7.331298530830422e-05,
833
+ "loss": 2.521,
834
+ "step": 67500
835
+ },
836
+ {
837
+ "epoch": 0.27,
838
+ "learning_rate": 7.311530371799535e-05,
839
+ "loss": 2.521,
840
+ "step": 68000
841
+ },
842
+ {
843
+ "epoch": 0.27,
844
+ "learning_rate": 7.291762212768649e-05,
845
+ "loss": 2.5145,
846
+ "step": 68500
847
+ },
848
+ {
849
+ "epoch": 0.27,
850
+ "learning_rate": 7.271994053737764e-05,
851
+ "loss": 2.5204,
852
+ "step": 69000
853
+ },
854
+ {
855
+ "epoch": 0.27,
856
+ "learning_rate": 7.252225894706878e-05,
857
+ "loss": 2.5146,
858
+ "step": 69500
859
+ },
860
+ {
861
+ "epoch": 0.28,
862
+ "learning_rate": 7.232457735675993e-05,
863
+ "loss": 2.5164,
864
+ "step": 70000
865
+ },
866
+ {
867
+ "epoch": 0.28,
868
+ "learning_rate": 7.212689576645107e-05,
869
+ "loss": 2.5156,
870
+ "step": 70500
871
+ },
872
+ {
873
+ "epoch": 0.28,
874
+ "learning_rate": 7.192921417614222e-05,
875
+ "loss": 2.5119,
876
+ "step": 71000
877
+ },
878
+ {
879
+ "epoch": 0.28,
880
+ "learning_rate": 7.173153258583335e-05,
881
+ "loss": 2.5172,
882
+ "step": 71500
883
+ },
884
+ {
885
+ "epoch": 0.28,
886
+ "learning_rate": 7.153385099552449e-05,
887
+ "loss": 2.4984,
888
+ "step": 72000
889
+ },
890
+ {
891
+ "epoch": 0.29,
892
+ "learning_rate": 7.133616940521564e-05,
893
+ "loss": 2.5078,
894
+ "step": 72500
895
+ },
896
+ {
897
+ "epoch": 0.29,
898
+ "learning_rate": 7.113848781490678e-05,
899
+ "loss": 2.5097,
900
+ "step": 73000
901
+ },
902
+ {
903
+ "epoch": 0.29,
904
+ "learning_rate": 7.094080622459793e-05,
905
+ "loss": 2.5116,
906
+ "step": 73500
907
+ },
908
+ {
909
+ "epoch": 0.29,
910
+ "learning_rate": 7.074312463428906e-05,
911
+ "loss": 2.5024,
912
+ "step": 74000
913
+ },
914
+ {
915
+ "epoch": 0.29,
916
+ "learning_rate": 7.05454430439802e-05,
917
+ "loss": 2.5133,
918
+ "step": 74500
919
+ },
920
+ {
921
+ "epoch": 0.3,
922
+ "learning_rate": 7.034776145367135e-05,
923
+ "loss": 2.5061,
924
+ "step": 75000
925
+ },
926
+ {
927
+ "epoch": 0.3,
928
+ "eval_loss": 2.319611072540283,
929
+ "eval_runtime": 187.4687,
930
+ "eval_samples_per_second": 533.422,
931
+ "eval_steps_per_second": 66.678,
932
+ "step": 75000
933
+ },
934
+ {
935
+ "epoch": 0.3,
936
+ "learning_rate": 7.015007986336249e-05,
937
+ "loss": 2.4962,
938
+ "step": 75500
939
+ },
940
+ {
941
+ "epoch": 0.3,
942
+ "learning_rate": 6.995239827305363e-05,
943
+ "loss": 2.4973,
944
+ "step": 76000
945
+ },
946
+ {
947
+ "epoch": 0.3,
948
+ "learning_rate": 6.975471668274476e-05,
949
+ "loss": 2.497,
950
+ "step": 76500
951
+ },
952
+ {
953
+ "epoch": 0.3,
954
+ "learning_rate": 6.955703509243591e-05,
955
+ "loss": 2.5022,
956
+ "step": 77000
957
+ },
958
+ {
959
+ "epoch": 0.31,
960
+ "learning_rate": 6.935935350212705e-05,
961
+ "loss": 2.4994,
962
+ "step": 77500
963
+ },
964
+ {
965
+ "epoch": 0.31,
966
+ "learning_rate": 6.91616719118182e-05,
967
+ "loss": 2.4915,
968
+ "step": 78000
969
+ },
970
+ {
971
+ "epoch": 0.31,
972
+ "learning_rate": 6.896399032150934e-05,
973
+ "loss": 2.4979,
974
+ "step": 78500
975
+ },
976
+ {
977
+ "epoch": 0.31,
978
+ "learning_rate": 6.876630873120047e-05,
979
+ "loss": 2.4881,
980
+ "step": 79000
981
+ },
982
+ {
983
+ "epoch": 0.31,
984
+ "learning_rate": 6.856862714089163e-05,
985
+ "loss": 2.4981,
986
+ "step": 79500
987
+ },
988
+ {
989
+ "epoch": 0.32,
990
+ "learning_rate": 6.837094555058276e-05,
991
+ "loss": 2.4873,
992
+ "step": 80000
993
+ },
994
+ {
995
+ "epoch": 0.32,
996
+ "learning_rate": 6.817326396027391e-05,
997
+ "loss": 2.4886,
998
+ "step": 80500
999
+ },
1000
+ {
1001
+ "epoch": 0.32,
1002
+ "learning_rate": 6.797558236996505e-05,
1003
+ "loss": 2.4869,
1004
+ "step": 81000
1005
+ },
1006
+ {
1007
+ "epoch": 0.32,
1008
+ "learning_rate": 6.77779007796562e-05,
1009
+ "loss": 2.4893,
1010
+ "step": 81500
1011
+ },
1012
+ {
1013
+ "epoch": 0.32,
1014
+ "learning_rate": 6.758021918934734e-05,
1015
+ "loss": 2.4847,
1016
+ "step": 82000
1017
+ },
1018
+ {
1019
+ "epoch": 0.33,
1020
+ "learning_rate": 6.738253759903847e-05,
1021
+ "loss": 2.4856,
1022
+ "step": 82500
1023
+ },
1024
+ {
1025
+ "epoch": 0.33,
1026
+ "learning_rate": 6.718485600872962e-05,
1027
+ "loss": 2.4811,
1028
+ "step": 83000
1029
+ },
1030
+ {
1031
+ "epoch": 0.33,
1032
+ "learning_rate": 6.698717441842076e-05,
1033
+ "loss": 2.4909,
1034
+ "step": 83500
1035
+ },
1036
+ {
1037
+ "epoch": 0.33,
1038
+ "learning_rate": 6.678949282811191e-05,
1039
+ "loss": 2.477,
1040
+ "step": 84000
1041
+ },
1042
+ {
1043
+ "epoch": 0.33,
1044
+ "learning_rate": 6.659181123780305e-05,
1045
+ "loss": 2.4754,
1046
+ "step": 84500
1047
+ },
1048
+ {
1049
+ "epoch": 0.34,
1050
+ "learning_rate": 6.63941296474942e-05,
1051
+ "loss": 2.4826,
1052
+ "step": 85000
1053
+ },
1054
+ {
1055
+ "epoch": 0.34,
1056
+ "learning_rate": 6.619644805718534e-05,
1057
+ "loss": 2.4672,
1058
+ "step": 85500
1059
+ },
1060
+ {
1061
+ "epoch": 0.34,
1062
+ "learning_rate": 6.599876646687647e-05,
1063
+ "loss": 2.4728,
1064
+ "step": 86000
1065
+ },
1066
+ {
1067
+ "epoch": 0.34,
1068
+ "learning_rate": 6.580108487656762e-05,
1069
+ "loss": 2.4758,
1070
+ "step": 86500
1071
+ },
1072
+ {
1073
+ "epoch": 0.34,
1074
+ "learning_rate": 6.560340328625876e-05,
1075
+ "loss": 2.471,
1076
+ "step": 87000
1077
+ },
1078
+ {
1079
+ "epoch": 0.35,
1080
+ "learning_rate": 6.540572169594991e-05,
1081
+ "loss": 2.4684,
1082
+ "step": 87500
1083
+ },
1084
+ {
1085
+ "epoch": 0.35,
1086
+ "learning_rate": 6.520804010564105e-05,
1087
+ "loss": 2.47,
1088
+ "step": 88000
1089
+ },
1090
+ {
1091
+ "epoch": 0.35,
1092
+ "learning_rate": 6.50103585153322e-05,
1093
+ "loss": 2.4677,
1094
+ "step": 88500
1095
+ },
1096
+ {
1097
+ "epoch": 0.35,
1098
+ "learning_rate": 6.481267692502333e-05,
1099
+ "loss": 2.4696,
1100
+ "step": 89000
1101
+ },
1102
+ {
1103
+ "epoch": 0.35,
1104
+ "learning_rate": 6.461499533471447e-05,
1105
+ "loss": 2.4647,
1106
+ "step": 89500
1107
+ },
1108
+ {
1109
+ "epoch": 0.36,
1110
+ "learning_rate": 6.441731374440562e-05,
1111
+ "loss": 2.4599,
1112
+ "step": 90000
1113
+ },
1114
+ {
1115
+ "epoch": 0.36,
1116
+ "learning_rate": 6.421963215409676e-05,
1117
+ "loss": 2.4605,
1118
+ "step": 90500
1119
+ },
1120
+ {
1121
+ "epoch": 0.36,
1122
+ "learning_rate": 6.402195056378791e-05,
1123
+ "loss": 2.4648,
1124
+ "step": 91000
1125
+ },
1126
+ {
1127
+ "epoch": 0.36,
1128
+ "learning_rate": 6.382426897347905e-05,
1129
+ "loss": 2.4569,
1130
+ "step": 91500
1131
+ },
1132
+ {
1133
+ "epoch": 0.36,
1134
+ "learning_rate": 6.362658738317018e-05,
1135
+ "loss": 2.4677,
1136
+ "step": 92000
1137
+ },
1138
+ {
1139
+ "epoch": 0.37,
1140
+ "learning_rate": 6.342890579286132e-05,
1141
+ "loss": 2.4602,
1142
+ "step": 92500
1143
+ },
1144
+ {
1145
+ "epoch": 0.37,
1146
+ "learning_rate": 6.323122420255247e-05,
1147
+ "loss": 2.4582,
1148
+ "step": 93000
1149
+ },
1150
+ {
1151
+ "epoch": 0.37,
1152
+ "learning_rate": 6.303354261224361e-05,
1153
+ "loss": 2.4601,
1154
+ "step": 93500
1155
+ },
1156
+ {
1157
+ "epoch": 0.37,
1158
+ "learning_rate": 6.283586102193474e-05,
1159
+ "loss": 2.4598,
1160
+ "step": 94000
1161
+ },
1162
+ {
1163
+ "epoch": 0.37,
1164
+ "learning_rate": 6.26381794316259e-05,
1165
+ "loss": 2.4597,
1166
+ "step": 94500
1167
+ },
1168
+ {
1169
+ "epoch": 0.38,
1170
+ "learning_rate": 6.244049784131703e-05,
1171
+ "loss": 2.4466,
1172
+ "step": 95000
1173
+ },
1174
+ {
1175
+ "epoch": 0.38,
1176
+ "learning_rate": 6.224281625100817e-05,
1177
+ "loss": 2.4488,
1178
+ "step": 95500
1179
+ },
1180
+ {
1181
+ "epoch": 0.38,
1182
+ "learning_rate": 6.204513466069932e-05,
1183
+ "loss": 2.4557,
1184
+ "step": 96000
1185
+ },
1186
+ {
1187
+ "epoch": 0.38,
1188
+ "learning_rate": 6.184745307039046e-05,
1189
+ "loss": 2.4555,
1190
+ "step": 96500
1191
+ },
1192
+ {
1193
+ "epoch": 0.38,
1194
+ "learning_rate": 6.16497714800816e-05,
1195
+ "loss": 2.4514,
1196
+ "step": 97000
1197
+ },
1198
+ {
1199
+ "epoch": 0.39,
1200
+ "learning_rate": 6.145208988977274e-05,
1201
+ "loss": 2.4436,
1202
+ "step": 97500
1203
+ },
1204
+ {
1205
+ "epoch": 0.39,
1206
+ "learning_rate": 6.12544082994639e-05,
1207
+ "loss": 2.4497,
1208
+ "step": 98000
1209
+ },
1210
+ {
1211
+ "epoch": 0.39,
1212
+ "learning_rate": 6.105672670915503e-05,
1213
+ "loss": 2.4527,
1214
+ "step": 98500
1215
+ },
1216
+ {
1217
+ "epoch": 0.39,
1218
+ "learning_rate": 6.085904511884617e-05,
1219
+ "loss": 2.4448,
1220
+ "step": 99000
1221
+ },
1222
+ {
1223
+ "epoch": 0.39,
1224
+ "learning_rate": 6.066136352853732e-05,
1225
+ "loss": 2.4407,
1226
+ "step": 99500
1227
+ },
1228
+ {
1229
+ "epoch": 0.4,
1230
+ "learning_rate": 6.0463681938228455e-05,
1231
+ "loss": 2.4431,
1232
+ "step": 100000
1233
+ },
1234
+ {
1235
+ "epoch": 0.4,
1236
+ "eval_loss": 2.261350393295288,
1237
+ "eval_runtime": 191.0766,
1238
+ "eval_samples_per_second": 523.35,
1239
+ "eval_steps_per_second": 65.419,
1240
+ "step": 100000
1241
+ },
1242
+ {
1243
+ "epoch": 0.4,
1244
+ "learning_rate": 6.0266000347919605e-05,
1245
+ "loss": 2.441,
1246
+ "step": 100500
1247
+ },
1248
+ {
1249
+ "epoch": 0.4,
1250
+ "learning_rate": 6.006831875761074e-05,
1251
+ "loss": 2.4422,
1252
+ "step": 101000
1253
+ },
1254
+ {
1255
+ "epoch": 0.4,
1256
+ "learning_rate": 5.987063716730189e-05,
1257
+ "loss": 2.4335,
1258
+ "step": 101500
1259
+ },
1260
+ {
1261
+ "epoch": 0.4,
1262
+ "learning_rate": 5.967295557699303e-05,
1263
+ "loss": 2.4346,
1264
+ "step": 102000
1265
+ },
1266
+ {
1267
+ "epoch": 0.41,
1268
+ "learning_rate": 5.9475273986684166e-05,
1269
+ "loss": 2.4278,
1270
+ "step": 102500
1271
+ },
1272
+ {
1273
+ "epoch": 0.41,
1274
+ "learning_rate": 5.927759239637532e-05,
1275
+ "loss": 2.4443,
1276
+ "step": 103000
1277
+ },
1278
+ {
1279
+ "epoch": 0.41,
1280
+ "learning_rate": 5.9079910806066454e-05,
1281
+ "loss": 2.4433,
1282
+ "step": 103500
1283
+ },
1284
+ {
1285
+ "epoch": 0.41,
1286
+ "learning_rate": 5.88822292157576e-05,
1287
+ "loss": 2.4365,
1288
+ "step": 104000
1289
+ },
1290
+ {
1291
+ "epoch": 0.41,
1292
+ "learning_rate": 5.868454762544874e-05,
1293
+ "loss": 2.425,
1294
+ "step": 104500
1295
+ },
1296
+ {
1297
+ "epoch": 0.42,
1298
+ "learning_rate": 5.8486866035139885e-05,
1299
+ "loss": 2.4187,
1300
+ "step": 105000
1301
+ },
1302
+ {
1303
+ "epoch": 0.42,
1304
+ "learning_rate": 5.828918444483102e-05,
1305
+ "loss": 2.4294,
1306
+ "step": 105500
1307
+ },
1308
+ {
1309
+ "epoch": 0.42,
1310
+ "learning_rate": 5.8091502854522165e-05,
1311
+ "loss": 2.425,
1312
+ "step": 106000
1313
+ },
1314
+ {
1315
+ "epoch": 0.42,
1316
+ "learning_rate": 5.789382126421331e-05,
1317
+ "loss": 2.4356,
1318
+ "step": 106500
1319
+ },
1320
+ {
1321
+ "epoch": 0.42,
1322
+ "learning_rate": 5.7696139673904446e-05,
1323
+ "loss": 2.425,
1324
+ "step": 107000
1325
+ },
1326
+ {
1327
+ "epoch": 0.43,
1328
+ "learning_rate": 5.7498458083595596e-05,
1329
+ "loss": 2.4223,
1330
+ "step": 107500
1331
+ },
1332
+ {
1333
+ "epoch": 0.43,
1334
+ "learning_rate": 5.730077649328673e-05,
1335
+ "loss": 2.4156,
1336
+ "step": 108000
1337
+ },
1338
+ {
1339
+ "epoch": 0.43,
1340
+ "learning_rate": 5.7103094902977883e-05,
1341
+ "loss": 2.4187,
1342
+ "step": 108500
1343
+ },
1344
+ {
1345
+ "epoch": 0.43,
1346
+ "learning_rate": 5.690541331266902e-05,
1347
+ "loss": 2.4212,
1348
+ "step": 109000
1349
+ },
1350
+ {
1351
+ "epoch": 0.43,
1352
+ "learning_rate": 5.670773172236016e-05,
1353
+ "loss": 2.4218,
1354
+ "step": 109500
1355
+ },
1356
+ {
1357
+ "epoch": 0.43,
1358
+ "learning_rate": 5.651005013205131e-05,
1359
+ "loss": 2.417,
1360
+ "step": 110000
1361
+ },
1362
+ {
1363
+ "epoch": 0.44,
1364
+ "learning_rate": 5.6312368541742445e-05,
1365
+ "loss": 2.416,
1366
+ "step": 110500
1367
+ },
1368
+ {
1369
+ "epoch": 0.44,
1370
+ "learning_rate": 5.6114686951433595e-05,
1371
+ "loss": 2.4142,
1372
+ "step": 111000
1373
+ },
1374
+ {
1375
+ "epoch": 0.44,
1376
+ "learning_rate": 5.591700536112473e-05,
1377
+ "loss": 2.4197,
1378
+ "step": 111500
1379
+ },
1380
+ {
1381
+ "epoch": 0.44,
1382
+ "learning_rate": 5.5719323770815876e-05,
1383
+ "loss": 2.4167,
1384
+ "step": 112000
1385
+ },
1386
+ {
1387
+ "epoch": 0.44,
1388
+ "learning_rate": 5.552164218050702e-05,
1389
+ "loss": 2.4136,
1390
+ "step": 112500
1391
+ },
1392
+ {
1393
+ "epoch": 0.45,
1394
+ "learning_rate": 5.5323960590198156e-05,
1395
+ "loss": 2.409,
1396
+ "step": 113000
1397
+ },
1398
+ {
1399
+ "epoch": 0.45,
1400
+ "learning_rate": 5.51262789998893e-05,
1401
+ "loss": 2.4108,
1402
+ "step": 113500
1403
+ },
1404
+ {
1405
+ "epoch": 0.45,
1406
+ "learning_rate": 5.492859740958044e-05,
1407
+ "loss": 2.4148,
1408
+ "step": 114000
1409
+ },
1410
+ {
1411
+ "epoch": 0.45,
1412
+ "learning_rate": 5.473091581927159e-05,
1413
+ "loss": 2.4143,
1414
+ "step": 114500
1415
+ },
1416
+ {
1417
+ "epoch": 0.45,
1418
+ "learning_rate": 5.4533234228962724e-05,
1419
+ "loss": 2.4139,
1420
+ "step": 115000
1421
+ },
1422
+ {
1423
+ "epoch": 0.46,
1424
+ "learning_rate": 5.4335552638653875e-05,
1425
+ "loss": 2.4114,
1426
+ "step": 115500
1427
+ },
1428
+ {
1429
+ "epoch": 0.46,
1430
+ "learning_rate": 5.413787104834501e-05,
1431
+ "loss": 2.4064,
1432
+ "step": 116000
1433
+ },
1434
+ {
1435
+ "epoch": 0.46,
1436
+ "learning_rate": 5.394018945803615e-05,
1437
+ "loss": 2.41,
1438
+ "step": 116500
1439
+ },
1440
+ {
1441
+ "epoch": 0.46,
1442
+ "learning_rate": 5.37425078677273e-05,
1443
+ "loss": 2.3971,
1444
+ "step": 117000
1445
+ },
1446
+ {
1447
+ "epoch": 0.46,
1448
+ "learning_rate": 5.3544826277418436e-05,
1449
+ "loss": 2.4097,
1450
+ "step": 117500
1451
+ },
1452
+ {
1453
+ "epoch": 0.47,
1454
+ "learning_rate": 5.3347144687109586e-05,
1455
+ "loss": 2.3987,
1456
+ "step": 118000
1457
+ },
1458
+ {
1459
+ "epoch": 0.47,
1460
+ "learning_rate": 5.314946309680072e-05,
1461
+ "loss": 2.4049,
1462
+ "step": 118500
1463
+ },
1464
+ {
1465
+ "epoch": 0.47,
1466
+ "learning_rate": 5.2951781506491873e-05,
1467
+ "loss": 2.4072,
1468
+ "step": 119000
1469
+ },
1470
+ {
1471
+ "epoch": 0.47,
1472
+ "learning_rate": 5.275409991618301e-05,
1473
+ "loss": 2.3864,
1474
+ "step": 119500
1475
+ },
1476
+ {
1477
+ "epoch": 0.47,
1478
+ "learning_rate": 5.255641832587415e-05,
1479
+ "loss": 2.3971,
1480
+ "step": 120000
1481
+ },
1482
+ {
1483
+ "epoch": 0.48,
1484
+ "learning_rate": 5.23587367355653e-05,
1485
+ "loss": 2.397,
1486
+ "step": 120500
1487
+ },
1488
+ {
1489
+ "epoch": 0.48,
1490
+ "learning_rate": 5.2161055145256435e-05,
1491
+ "loss": 2.3943,
1492
+ "step": 121000
1493
+ },
1494
+ {
1495
+ "epoch": 0.48,
1496
+ "learning_rate": 5.196337355494758e-05,
1497
+ "loss": 2.395,
1498
+ "step": 121500
1499
+ },
1500
+ {
1501
+ "epoch": 0.48,
1502
+ "learning_rate": 5.1765691964638715e-05,
1503
+ "loss": 2.3998,
1504
+ "step": 122000
1505
+ },
1506
+ {
1507
+ "epoch": 0.48,
1508
+ "learning_rate": 5.1568010374329866e-05,
1509
+ "loss": 2.3948,
1510
+ "step": 122500
1511
+ },
1512
+ {
1513
+ "epoch": 0.49,
1514
+ "learning_rate": 5.1370328784021e-05,
1515
+ "loss": 2.3955,
1516
+ "step": 123000
1517
+ },
1518
+ {
1519
+ "epoch": 0.49,
1520
+ "learning_rate": 5.117264719371214e-05,
1521
+ "loss": 2.39,
1522
+ "step": 123500
1523
+ },
1524
+ {
1525
+ "epoch": 0.49,
1526
+ "learning_rate": 5.097496560340329e-05,
1527
+ "loss": 2.3935,
1528
+ "step": 124000
1529
+ },
1530
+ {
1531
+ "epoch": 0.49,
1532
+ "learning_rate": 5.077728401309443e-05,
1533
+ "loss": 2.3905,
1534
+ "step": 124500
1535
+ },
1536
+ {
1537
+ "epoch": 0.49,
1538
+ "learning_rate": 5.057960242278558e-05,
1539
+ "loss": 2.3815,
1540
+ "step": 125000
1541
+ },
1542
+ {
1543
+ "epoch": 0.49,
1544
+ "eval_loss": 2.2064666748046875,
1545
+ "eval_runtime": 190.7509,
1546
+ "eval_samples_per_second": 524.244,
1547
+ "eval_steps_per_second": 65.53,
1548
+ "step": 125000
1549
+ },
1550
+ {
1551
+ "epoch": 0.5,
1552
+ "learning_rate": 5.0381920832476714e-05,
1553
+ "loss": 2.3912,
1554
+ "step": 125500
1555
+ },
1556
+ {
1557
+ "epoch": 0.5,
1558
+ "learning_rate": 5.0184239242167865e-05,
1559
+ "loss": 2.3816,
1560
+ "step": 126000
1561
+ },
1562
+ {
1563
+ "epoch": 0.5,
1564
+ "learning_rate": 4.9986557651859e-05,
1565
+ "loss": 2.3818,
1566
+ "step": 126500
1567
+ },
1568
+ {
1569
+ "epoch": 0.5,
1570
+ "learning_rate": 4.9788876061550145e-05,
1571
+ "loss": 2.3937,
1572
+ "step": 127000
1573
+ },
1574
+ {
1575
+ "epoch": 0.5,
1576
+ "learning_rate": 4.959119447124129e-05,
1577
+ "loss": 2.376,
1578
+ "step": 127500
1579
+ },
1580
+ {
1581
+ "epoch": 0.51,
1582
+ "learning_rate": 4.9393512880932426e-05,
1583
+ "loss": 2.3803,
1584
+ "step": 128000
1585
+ },
1586
+ {
1587
+ "epoch": 0.51,
1588
+ "learning_rate": 4.919583129062357e-05,
1589
+ "loss": 2.3896,
1590
+ "step": 128500
1591
+ },
1592
+ {
1593
+ "epoch": 0.51,
1594
+ "learning_rate": 4.899814970031471e-05,
1595
+ "loss": 2.3814,
1596
+ "step": 129000
1597
+ },
1598
+ {
1599
+ "epoch": 0.51,
1600
+ "learning_rate": 4.880046811000586e-05,
1601
+ "loss": 2.3786,
1602
+ "step": 129500
1603
+ },
1604
+ {
1605
+ "epoch": 0.51,
1606
+ "learning_rate": 4.8602786519696994e-05,
1607
+ "loss": 2.3813,
1608
+ "step": 130000
1609
+ },
1610
+ {
1611
+ "epoch": 0.52,
1612
+ "learning_rate": 4.840510492938814e-05,
1613
+ "loss": 2.381,
1614
+ "step": 130500
1615
+ },
1616
+ {
1617
+ "epoch": 0.52,
1618
+ "learning_rate": 4.820742333907928e-05,
1619
+ "loss": 2.3753,
1620
+ "step": 131000
1621
+ },
1622
+ {
1623
+ "epoch": 0.52,
1624
+ "learning_rate": 4.800974174877042e-05,
1625
+ "loss": 2.3792,
1626
+ "step": 131500
1627
+ },
1628
+ {
1629
+ "epoch": 0.52,
1630
+ "learning_rate": 4.781206015846156e-05,
1631
+ "loss": 2.3823,
1632
+ "step": 132000
1633
+ },
1634
+ {
1635
+ "epoch": 0.52,
1636
+ "learning_rate": 4.7614378568152705e-05,
1637
+ "loss": 2.3755,
1638
+ "step": 132500
1639
+ },
1640
+ {
1641
+ "epoch": 0.53,
1642
+ "learning_rate": 4.741669697784385e-05,
1643
+ "loss": 2.3744,
1644
+ "step": 133000
1645
+ },
1646
+ {
1647
+ "epoch": 0.53,
1648
+ "learning_rate": 4.721901538753499e-05,
1649
+ "loss": 2.3698,
1650
+ "step": 133500
1651
+ },
1652
+ {
1653
+ "epoch": 0.53,
1654
+ "learning_rate": 4.7021333797226136e-05,
1655
+ "loss": 2.3671,
1656
+ "step": 134000
1657
+ },
1658
+ {
1659
+ "epoch": 0.53,
1660
+ "learning_rate": 4.682365220691728e-05,
1661
+ "loss": 2.3682,
1662
+ "step": 134500
1663
+ },
1664
+ {
1665
+ "epoch": 0.53,
1666
+ "learning_rate": 4.662597061660842e-05,
1667
+ "loss": 2.3754,
1668
+ "step": 135000
1669
+ },
1670
+ {
1671
+ "epoch": 0.54,
1672
+ "learning_rate": 4.642828902629956e-05,
1673
+ "loss": 2.3734,
1674
+ "step": 135500
1675
+ },
1676
+ {
1677
+ "epoch": 0.54,
1678
+ "learning_rate": 4.6230607435990704e-05,
1679
+ "loss": 2.3701,
1680
+ "step": 136000
1681
+ },
1682
+ {
1683
+ "epoch": 0.54,
1684
+ "learning_rate": 4.603292584568185e-05,
1685
+ "loss": 2.3612,
1686
+ "step": 136500
1687
+ },
1688
+ {
1689
+ "epoch": 0.54,
1690
+ "learning_rate": 4.583524425537299e-05,
1691
+ "loss": 2.3665,
1692
+ "step": 137000
1693
+ },
1694
+ {
1695
+ "epoch": 0.54,
1696
+ "learning_rate": 4.5637562665064135e-05,
1697
+ "loss": 2.3608,
1698
+ "step": 137500
1699
+ },
1700
+ {
1701
+ "epoch": 0.55,
1702
+ "learning_rate": 4.543988107475527e-05,
1703
+ "loss": 2.3632,
1704
+ "step": 138000
1705
+ },
1706
+ {
1707
+ "epoch": 0.55,
1708
+ "learning_rate": 4.5242199484446416e-05,
1709
+ "loss": 2.3652,
1710
+ "step": 138500
1711
+ },
1712
+ {
1713
+ "epoch": 0.55,
1714
+ "learning_rate": 4.504451789413755e-05,
1715
+ "loss": 2.3618,
1716
+ "step": 139000
1717
+ },
1718
+ {
1719
+ "epoch": 0.55,
1720
+ "learning_rate": 4.4846836303828696e-05,
1721
+ "loss": 2.3618,
1722
+ "step": 139500
1723
+ },
1724
+ {
1725
+ "epoch": 0.55,
1726
+ "learning_rate": 4.464915471351984e-05,
1727
+ "loss": 2.3665,
1728
+ "step": 140000
1729
+ },
1730
+ {
1731
+ "epoch": 0.56,
1732
+ "learning_rate": 4.4451473123210984e-05,
1733
+ "loss": 2.3625,
1734
+ "step": 140500
1735
+ },
1736
+ {
1737
+ "epoch": 0.56,
1738
+ "learning_rate": 4.425379153290213e-05,
1739
+ "loss": 2.3635,
1740
+ "step": 141000
1741
+ },
1742
+ {
1743
+ "epoch": 0.56,
1744
+ "learning_rate": 4.4056109942593264e-05,
1745
+ "loss": 2.365,
1746
+ "step": 141500
1747
+ },
1748
+ {
1749
+ "epoch": 0.56,
1750
+ "learning_rate": 4.385842835228441e-05,
1751
+ "loss": 2.3606,
1752
+ "step": 142000
1753
+ },
1754
+ {
1755
+ "epoch": 0.56,
1756
+ "learning_rate": 4.366074676197555e-05,
1757
+ "loss": 2.3566,
1758
+ "step": 142500
1759
+ },
1760
+ {
1761
+ "epoch": 0.57,
1762
+ "learning_rate": 4.3463065171666695e-05,
1763
+ "loss": 2.3492,
1764
+ "step": 143000
1765
+ },
1766
+ {
1767
+ "epoch": 0.57,
1768
+ "learning_rate": 4.326538358135784e-05,
1769
+ "loss": 2.3548,
1770
+ "step": 143500
1771
+ },
1772
+ {
1773
+ "epoch": 0.57,
1774
+ "learning_rate": 4.306770199104898e-05,
1775
+ "loss": 2.3466,
1776
+ "step": 144000
1777
+ },
1778
+ {
1779
+ "epoch": 0.57,
1780
+ "learning_rate": 4.2870020400740126e-05,
1781
+ "loss": 2.3456,
1782
+ "step": 144500
1783
+ },
1784
+ {
1785
+ "epoch": 0.57,
1786
+ "learning_rate": 4.267233881043126e-05,
1787
+ "loss": 2.3461,
1788
+ "step": 145000
1789
+ },
1790
+ {
1791
+ "epoch": 0.58,
1792
+ "learning_rate": 4.247465722012241e-05,
1793
+ "loss": 2.35,
1794
+ "step": 145500
1795
+ },
1796
+ {
1797
+ "epoch": 0.58,
1798
+ "learning_rate": 4.227697562981355e-05,
1799
+ "loss": 2.3541,
1800
+ "step": 146000
1801
+ },
1802
+ {
1803
+ "epoch": 0.58,
1804
+ "learning_rate": 4.2079294039504694e-05,
1805
+ "loss": 2.3475,
1806
+ "step": 146500
1807
+ },
1808
+ {
1809
+ "epoch": 0.58,
1810
+ "learning_rate": 4.188161244919583e-05,
1811
+ "loss": 2.3557,
1812
+ "step": 147000
1813
+ },
1814
+ {
1815
+ "epoch": 0.58,
1816
+ "learning_rate": 4.1683930858886975e-05,
1817
+ "loss": 2.3398,
1818
+ "step": 147500
1819
+ },
1820
+ {
1821
+ "epoch": 0.59,
1822
+ "learning_rate": 4.148624926857812e-05,
1823
+ "loss": 2.3409,
1824
+ "step": 148000
1825
+ },
1826
+ {
1827
+ "epoch": 0.59,
1828
+ "learning_rate": 4.1288567678269255e-05,
1829
+ "loss": 2.3476,
1830
+ "step": 148500
1831
+ },
1832
+ {
1833
+ "epoch": 0.59,
1834
+ "learning_rate": 4.10908860879604e-05,
1835
+ "loss": 2.3381,
1836
+ "step": 149000
1837
+ },
1838
+ {
1839
+ "epoch": 0.59,
1840
+ "learning_rate": 4.089320449765154e-05,
1841
+ "loss": 2.3415,
1842
+ "step": 149500
1843
+ },
1844
+ {
1845
+ "epoch": 0.59,
1846
+ "learning_rate": 4.0695522907342686e-05,
1847
+ "loss": 2.3467,
1848
+ "step": 150000
1849
+ },
1850
+ {
1851
+ "epoch": 0.59,
1852
+ "eval_loss": 2.168158531188965,
1853
+ "eval_runtime": 189.775,
1854
+ "eval_samples_per_second": 526.94,
1855
+ "eval_steps_per_second": 65.867,
1856
+ "step": 150000
1857
+ },
1858
+ {
1859
+ "epoch": 0.6,
1860
+ "learning_rate": 4.049784131703383e-05,
1861
+ "loss": 2.337,
1862
+ "step": 150500
1863
+ },
1864
+ {
1865
+ "epoch": 0.6,
1866
+ "learning_rate": 4.0300159726724974e-05,
1867
+ "loss": 2.3416,
1868
+ "step": 151000
1869
+ },
1870
+ {
1871
+ "epoch": 0.6,
1872
+ "learning_rate": 4.010247813641612e-05,
1873
+ "loss": 2.3464,
1874
+ "step": 151500
1875
+ },
1876
+ {
1877
+ "epoch": 0.6,
1878
+ "learning_rate": 3.9904796546107254e-05,
1879
+ "loss": 2.3419,
1880
+ "step": 152000
1881
+ },
1882
+ {
1883
+ "epoch": 0.6,
1884
+ "learning_rate": 3.97071149557984e-05,
1885
+ "loss": 2.3455,
1886
+ "step": 152500
1887
+ },
1888
+ {
1889
+ "epoch": 0.6,
1890
+ "learning_rate": 3.950943336548954e-05,
1891
+ "loss": 2.341,
1892
+ "step": 153000
1893
+ },
1894
+ {
1895
+ "epoch": 0.61,
1896
+ "learning_rate": 3.9311751775180685e-05,
1897
+ "loss": 2.3386,
1898
+ "step": 153500
1899
+ },
1900
+ {
1901
+ "epoch": 0.61,
1902
+ "learning_rate": 3.911407018487183e-05,
1903
+ "loss": 2.3374,
1904
+ "step": 154000
1905
+ },
1906
+ {
1907
+ "epoch": 0.61,
1908
+ "learning_rate": 3.891638859456297e-05,
1909
+ "loss": 2.3338,
1910
+ "step": 154500
1911
+ },
1912
+ {
1913
+ "epoch": 0.61,
1914
+ "learning_rate": 3.871870700425411e-05,
1915
+ "loss": 2.3376,
1916
+ "step": 155000
1917
+ },
1918
+ {
1919
+ "epoch": 0.61,
1920
+ "learning_rate": 3.852102541394525e-05,
1921
+ "loss": 2.3428,
1922
+ "step": 155500
1923
+ },
1924
+ {
1925
+ "epoch": 0.62,
1926
+ "learning_rate": 3.832334382363639e-05,
1927
+ "loss": 2.3384,
1928
+ "step": 156000
1929
+ },
1930
+ {
1931
+ "epoch": 0.62,
1932
+ "learning_rate": 3.8125662233327534e-05,
1933
+ "loss": 2.3302,
1934
+ "step": 156500
1935
+ },
1936
+ {
1937
+ "epoch": 0.62,
1938
+ "learning_rate": 3.792798064301868e-05,
1939
+ "loss": 2.3278,
1940
+ "step": 157000
1941
+ },
1942
+ {
1943
+ "epoch": 0.62,
1944
+ "learning_rate": 3.773029905270982e-05,
1945
+ "loss": 2.3283,
1946
+ "step": 157500
1947
+ },
1948
+ {
1949
+ "epoch": 0.62,
1950
+ "learning_rate": 3.7532617462400965e-05,
1951
+ "loss": 2.3339,
1952
+ "step": 158000
1953
+ },
1954
+ {
1955
+ "epoch": 0.63,
1956
+ "learning_rate": 3.733493587209211e-05,
1957
+ "loss": 2.327,
1958
+ "step": 158500
1959
+ },
1960
+ {
1961
+ "epoch": 0.63,
1962
+ "learning_rate": 3.7137254281783245e-05,
1963
+ "loss": 2.3264,
1964
+ "step": 159000
1965
+ },
1966
+ {
1967
+ "epoch": 0.63,
1968
+ "learning_rate": 3.693957269147439e-05,
1969
+ "loss": 2.3307,
1970
+ "step": 159500
1971
+ },
1972
+ {
1973
+ "epoch": 0.63,
1974
+ "learning_rate": 3.674189110116553e-05,
1975
+ "loss": 2.331,
1976
+ "step": 160000
1977
+ },
1978
+ {
1979
+ "epoch": 0.63,
1980
+ "learning_rate": 3.6544209510856676e-05,
1981
+ "loss": 2.3277,
1982
+ "step": 160500
1983
+ },
1984
+ {
1985
+ "epoch": 0.64,
1986
+ "learning_rate": 3.634652792054782e-05,
1987
+ "loss": 2.3251,
1988
+ "step": 161000
1989
+ },
1990
+ {
1991
+ "epoch": 0.64,
1992
+ "learning_rate": 3.6148846330238964e-05,
1993
+ "loss": 2.3175,
1994
+ "step": 161500
1995
+ },
1996
+ {
1997
+ "epoch": 0.64,
1998
+ "learning_rate": 3.59511647399301e-05,
1999
+ "loss": 2.3192,
2000
+ "step": 162000
2001
+ },
2002
+ {
2003
+ "epoch": 0.64,
2004
+ "learning_rate": 3.5753483149621244e-05,
2005
+ "loss": 2.3174,
2006
+ "step": 162500
2007
+ },
2008
+ {
2009
+ "epoch": 0.64,
2010
+ "learning_rate": 3.555580155931239e-05,
2011
+ "loss": 2.3218,
2012
+ "step": 163000
2013
+ },
2014
+ {
2015
+ "epoch": 0.65,
2016
+ "learning_rate": 3.5358119969003525e-05,
2017
+ "loss": 2.3192,
2018
+ "step": 163500
2019
+ },
2020
+ {
2021
+ "epoch": 0.65,
2022
+ "learning_rate": 3.516043837869467e-05,
2023
+ "loss": 2.3215,
2024
+ "step": 164000
2025
+ },
2026
+ {
2027
+ "epoch": 0.65,
2028
+ "learning_rate": 3.496275678838581e-05,
2029
+ "loss": 2.3203,
2030
+ "step": 164500
2031
+ },
2032
+ {
2033
+ "epoch": 0.65,
2034
+ "learning_rate": 3.4765075198076956e-05,
2035
+ "loss": 2.3133,
2036
+ "step": 165000
2037
+ },
2038
+ {
2039
+ "epoch": 0.65,
2040
+ "learning_rate": 3.456739360776809e-05,
2041
+ "loss": 2.312,
2042
+ "step": 165500
2043
+ },
2044
+ {
2045
+ "epoch": 0.66,
2046
+ "learning_rate": 3.4369712017459236e-05,
2047
+ "loss": 2.3161,
2048
+ "step": 166000
2049
+ },
2050
+ {
2051
+ "epoch": 0.66,
2052
+ "learning_rate": 3.417203042715038e-05,
2053
+ "loss": 2.3173,
2054
+ "step": 166500
2055
+ },
2056
+ {
2057
+ "epoch": 0.66,
2058
+ "learning_rate": 3.3974348836841524e-05,
2059
+ "loss": 2.3089,
2060
+ "step": 167000
2061
+ },
2062
+ {
2063
+ "epoch": 0.66,
2064
+ "learning_rate": 3.377666724653267e-05,
2065
+ "loss": 2.309,
2066
+ "step": 167500
2067
+ },
2068
+ {
2069
+ "epoch": 0.66,
2070
+ "learning_rate": 3.357898565622381e-05,
2071
+ "loss": 2.3217,
2072
+ "step": 168000
2073
+ },
2074
+ {
2075
+ "epoch": 0.67,
2076
+ "learning_rate": 3.3381304065914955e-05,
2077
+ "loss": 2.3093,
2078
+ "step": 168500
2079
+ },
2080
+ {
2081
+ "epoch": 0.67,
2082
+ "learning_rate": 3.318362247560609e-05,
2083
+ "loss": 2.3054,
2084
+ "step": 169000
2085
+ },
2086
+ {
2087
+ "epoch": 0.67,
2088
+ "learning_rate": 3.2985940885297235e-05,
2089
+ "loss": 2.3117,
2090
+ "step": 169500
2091
+ },
2092
+ {
2093
+ "epoch": 0.67,
2094
+ "learning_rate": 3.278825929498838e-05,
2095
+ "loss": 2.31,
2096
+ "step": 170000
2097
+ },
2098
+ {
2099
+ "epoch": 0.67,
2100
+ "learning_rate": 3.259057770467952e-05,
2101
+ "loss": 2.3081,
2102
+ "step": 170500
2103
+ },
2104
+ {
2105
+ "epoch": 0.68,
2106
+ "learning_rate": 3.2392896114370666e-05,
2107
+ "loss": 2.3143,
2108
+ "step": 171000
2109
+ },
2110
+ {
2111
+ "epoch": 0.68,
2112
+ "learning_rate": 3.21952145240618e-05,
2113
+ "loss": 2.3133,
2114
+ "step": 171500
2115
+ },
2116
+ {
2117
+ "epoch": 0.68,
2118
+ "learning_rate": 3.199753293375295e-05,
2119
+ "loss": 2.3057,
2120
+ "step": 172000
2121
+ },
2122
+ {
2123
+ "epoch": 0.68,
2124
+ "learning_rate": 3.1799851343444084e-05,
2125
+ "loss": 2.3143,
2126
+ "step": 172500
2127
+ },
2128
+ {
2129
+ "epoch": 0.68,
2130
+ "learning_rate": 3.160216975313523e-05,
2131
+ "loss": 2.303,
2132
+ "step": 173000
2133
+ },
2134
+ {
2135
+ "epoch": 0.69,
2136
+ "learning_rate": 3.140448816282637e-05,
2137
+ "loss": 2.308,
2138
+ "step": 173500
2139
+ },
2140
+ {
2141
+ "epoch": 0.69,
2142
+ "learning_rate": 3.1206806572517515e-05,
2143
+ "loss": 2.3083,
2144
+ "step": 174000
2145
+ },
2146
+ {
2147
+ "epoch": 0.69,
2148
+ "learning_rate": 3.100912498220866e-05,
2149
+ "loss": 2.2954,
2150
+ "step": 174500
2151
+ },
2152
+ {
2153
+ "epoch": 0.69,
2154
+ "learning_rate": 3.08114433918998e-05,
2155
+ "loss": 2.3085,
2156
+ "step": 175000
2157
+ },
2158
+ {
2159
+ "epoch": 0.69,
2160
+ "eval_loss": 2.1400139331817627,
2161
+ "eval_runtime": 190.1966,
2162
+ "eval_samples_per_second": 525.772,
2163
+ "eval_steps_per_second": 65.721,
2164
+ "step": 175000
2165
+ },
2166
+ {
2167
+ "epoch": 0.69,
2168
+ "learning_rate": 3.0613761801590946e-05,
2169
+ "loss": 2.2955,
2170
+ "step": 175500
2171
+ },
2172
+ {
2173
+ "epoch": 0.7,
2174
+ "learning_rate": 3.0416080211282083e-05,
2175
+ "loss": 2.2923,
2176
+ "step": 176000
2177
+ },
2178
+ {
2179
+ "epoch": 0.7,
2180
+ "learning_rate": 3.0218398620973226e-05,
2181
+ "loss": 2.3059,
2182
+ "step": 176500
2183
+ },
2184
+ {
2185
+ "epoch": 0.7,
2186
+ "learning_rate": 3.002071703066437e-05,
2187
+ "loss": 2.2998,
2188
+ "step": 177000
2189
+ },
2190
+ {
2191
+ "epoch": 0.7,
2192
+ "learning_rate": 2.9823035440355514e-05,
2193
+ "loss": 2.2972,
2194
+ "step": 177500
2195
+ },
2196
+ {
2197
+ "epoch": 0.7,
2198
+ "learning_rate": 2.9625353850046654e-05,
2199
+ "loss": 2.2964,
2200
+ "step": 178000
2201
+ },
2202
+ {
2203
+ "epoch": 0.71,
2204
+ "learning_rate": 2.9427672259737798e-05,
2205
+ "loss": 2.2996,
2206
+ "step": 178500
2207
+ },
2208
+ {
2209
+ "epoch": 0.71,
2210
+ "learning_rate": 2.9229990669428935e-05,
2211
+ "loss": 2.2992,
2212
+ "step": 179000
2213
+ },
2214
+ {
2215
+ "epoch": 0.71,
2216
+ "learning_rate": 2.9032309079120078e-05,
2217
+ "loss": 2.2903,
2218
+ "step": 179500
2219
+ },
2220
+ {
2221
+ "epoch": 0.71,
2222
+ "learning_rate": 2.8834627488811222e-05,
2223
+ "loss": 2.2928,
2224
+ "step": 180000
2225
+ },
2226
+ {
2227
+ "epoch": 0.71,
2228
+ "learning_rate": 2.8636945898502366e-05,
2229
+ "loss": 2.3004,
2230
+ "step": 180500
2231
+ },
2232
+ {
2233
+ "epoch": 0.72,
2234
+ "learning_rate": 2.843926430819351e-05,
2235
+ "loss": 2.3031,
2236
+ "step": 181000
2237
+ },
2238
+ {
2239
+ "epoch": 0.72,
2240
+ "learning_rate": 2.8241582717884653e-05,
2241
+ "loss": 2.2985,
2242
+ "step": 181500
2243
+ },
2244
+ {
2245
+ "epoch": 0.72,
2246
+ "learning_rate": 2.8043901127575793e-05,
2247
+ "loss": 2.2885,
2248
+ "step": 182000
2249
+ },
2250
+ {
2251
+ "epoch": 0.72,
2252
+ "learning_rate": 2.7846219537266933e-05,
2253
+ "loss": 2.2954,
2254
+ "step": 182500
2255
+ },
2256
+ {
2257
+ "epoch": 0.72,
2258
+ "learning_rate": 2.7648537946958074e-05,
2259
+ "loss": 2.2983,
2260
+ "step": 183000
2261
+ },
2262
+ {
2263
+ "epoch": 0.73,
2264
+ "learning_rate": 2.7450856356649217e-05,
2265
+ "loss": 2.2888,
2266
+ "step": 183500
2267
+ },
2268
+ {
2269
+ "epoch": 0.73,
2270
+ "learning_rate": 2.725317476634036e-05,
2271
+ "loss": 2.2814,
2272
+ "step": 184000
2273
+ },
2274
+ {
2275
+ "epoch": 0.73,
2276
+ "learning_rate": 2.7055493176031505e-05,
2277
+ "loss": 2.282,
2278
+ "step": 184500
2279
+ },
2280
+ {
2281
+ "epoch": 0.73,
2282
+ "learning_rate": 2.685781158572265e-05,
2283
+ "loss": 2.2866,
2284
+ "step": 185000
2285
+ },
2286
+ {
2287
+ "epoch": 0.73,
2288
+ "learning_rate": 2.6660129995413792e-05,
2289
+ "loss": 2.2841,
2290
+ "step": 185500
2291
+ },
2292
+ {
2293
+ "epoch": 0.74,
2294
+ "learning_rate": 2.646244840510493e-05,
2295
+ "loss": 2.2944,
2296
+ "step": 186000
2297
+ },
2298
+ {
2299
+ "epoch": 0.74,
2300
+ "learning_rate": 2.6264766814796073e-05,
2301
+ "loss": 2.2881,
2302
+ "step": 186500
2303
+ },
2304
+ {
2305
+ "epoch": 0.74,
2306
+ "learning_rate": 2.6067085224487213e-05,
2307
+ "loss": 2.282,
2308
+ "step": 187000
2309
+ },
2310
+ {
2311
+ "epoch": 0.74,
2312
+ "learning_rate": 2.5869403634178357e-05,
2313
+ "loss": 2.2832,
2314
+ "step": 187500
2315
+ },
2316
+ {
2317
+ "epoch": 0.74,
2318
+ "learning_rate": 2.56717220438695e-05,
2319
+ "loss": 2.2799,
2320
+ "step": 188000
2321
+ },
2322
+ {
2323
+ "epoch": 0.75,
2324
+ "learning_rate": 2.5474040453560644e-05,
2325
+ "loss": 2.2819,
2326
+ "step": 188500
2327
+ },
2328
+ {
2329
+ "epoch": 0.75,
2330
+ "learning_rate": 2.5276358863251788e-05,
2331
+ "loss": 2.2857,
2332
+ "step": 189000
2333
+ },
2334
+ {
2335
+ "epoch": 0.75,
2336
+ "learning_rate": 2.5078677272942925e-05,
2337
+ "loss": 2.2719,
2338
+ "step": 189500
2339
+ },
2340
+ {
2341
+ "epoch": 0.75,
2342
+ "learning_rate": 2.488099568263407e-05,
2343
+ "loss": 2.2748,
2344
+ "step": 190000
2345
+ },
2346
+ {
2347
+ "epoch": 0.75,
2348
+ "learning_rate": 2.4683314092325212e-05,
2349
+ "loss": 2.2778,
2350
+ "step": 190500
2351
+ },
2352
+ {
2353
+ "epoch": 0.76,
2354
+ "learning_rate": 2.4485632502016352e-05,
2355
+ "loss": 2.283,
2356
+ "step": 191000
2357
+ },
2358
+ {
2359
+ "epoch": 0.76,
2360
+ "learning_rate": 2.4287950911707496e-05,
2361
+ "loss": 2.2724,
2362
+ "step": 191500
2363
+ },
2364
+ {
2365
+ "epoch": 0.76,
2366
+ "learning_rate": 2.4090269321398636e-05,
2367
+ "loss": 2.283,
2368
+ "step": 192000
2369
+ },
2370
+ {
2371
+ "epoch": 0.76,
2372
+ "learning_rate": 2.389258773108978e-05,
2373
+ "loss": 2.2799,
2374
+ "step": 192500
2375
+ },
2376
+ {
2377
+ "epoch": 0.76,
2378
+ "learning_rate": 2.3694906140780923e-05,
2379
+ "loss": 2.2738,
2380
+ "step": 193000
2381
+ },
2382
+ {
2383
+ "epoch": 0.77,
2384
+ "learning_rate": 2.3497224550472064e-05,
2385
+ "loss": 2.2766,
2386
+ "step": 193500
2387
+ },
2388
+ {
2389
+ "epoch": 0.77,
2390
+ "learning_rate": 2.3299542960163207e-05,
2391
+ "loss": 2.2735,
2392
+ "step": 194000
2393
+ },
2394
+ {
2395
+ "epoch": 0.77,
2396
+ "learning_rate": 2.310186136985435e-05,
2397
+ "loss": 2.2787,
2398
+ "step": 194500
2399
+ },
2400
+ {
2401
+ "epoch": 0.77,
2402
+ "learning_rate": 2.290417977954549e-05,
2403
+ "loss": 2.2754,
2404
+ "step": 195000
2405
+ },
2406
+ {
2407
+ "epoch": 0.77,
2408
+ "learning_rate": 2.270649818923663e-05,
2409
+ "loss": 2.2672,
2410
+ "step": 195500
2411
+ },
2412
+ {
2413
+ "epoch": 0.77,
2414
+ "learning_rate": 2.2508816598927775e-05,
2415
+ "loss": 2.2776,
2416
+ "step": 196000
2417
+ },
2418
+ {
2419
+ "epoch": 0.78,
2420
+ "learning_rate": 2.231113500861892e-05,
2421
+ "loss": 2.2706,
2422
+ "step": 196500
2423
+ },
2424
+ {
2425
+ "epoch": 0.78,
2426
+ "learning_rate": 2.211345341831006e-05,
2427
+ "loss": 2.2713,
2428
+ "step": 197000
2429
+ },
2430
+ {
2431
+ "epoch": 0.78,
2432
+ "learning_rate": 2.1915771828001203e-05,
2433
+ "loss": 2.2693,
2434
+ "step": 197500
2435
+ },
2436
+ {
2437
+ "epoch": 0.78,
2438
+ "learning_rate": 2.1718090237692347e-05,
2439
+ "loss": 2.2736,
2440
+ "step": 198000
2441
+ },
2442
+ {
2443
+ "epoch": 0.78,
2444
+ "learning_rate": 2.152040864738349e-05,
2445
+ "loss": 2.2696,
2446
+ "step": 198500
2447
+ },
2448
+ {
2449
+ "epoch": 0.79,
2450
+ "learning_rate": 2.132272705707463e-05,
2451
+ "loss": 2.2696,
2452
+ "step": 199000
2453
+ },
2454
+ {
2455
+ "epoch": 0.79,
2456
+ "learning_rate": 2.112504546676577e-05,
2457
+ "loss": 2.2663,
2458
+ "step": 199500
2459
+ },
2460
+ {
2461
+ "epoch": 0.79,
2462
+ "learning_rate": 2.0927363876456915e-05,
2463
+ "loss": 2.266,
2464
+ "step": 200000
2465
+ },
2466
+ {
2467
+ "epoch": 0.79,
2468
+ "eval_loss": 2.1008613109588623,
2469
+ "eval_runtime": 187.6272,
2470
+ "eval_samples_per_second": 532.972,
2471
+ "eval_steps_per_second": 66.621,
2472
+ "step": 200000
2473
+ },
2474
+ {
2475
+ "epoch": 0.79,
2476
+ "learning_rate": 2.0729682286148055e-05,
2477
+ "loss": 2.2594,
2478
+ "step": 200500
2479
+ },
2480
+ {
2481
+ "epoch": 0.79,
2482
+ "learning_rate": 2.05320006958392e-05,
2483
+ "loss": 2.2658,
2484
+ "step": 201000
2485
+ },
2486
+ {
2487
+ "epoch": 0.8,
2488
+ "learning_rate": 2.0334319105530342e-05,
2489
+ "loss": 2.2681,
2490
+ "step": 201500
2491
+ },
2492
+ {
2493
+ "epoch": 0.8,
2494
+ "learning_rate": 2.0136637515221486e-05,
2495
+ "loss": 2.2669,
2496
+ "step": 202000
2497
+ },
2498
+ {
2499
+ "epoch": 0.8,
2500
+ "learning_rate": 1.9938955924912626e-05,
2501
+ "loss": 2.2609,
2502
+ "step": 202500
2503
+ },
2504
+ {
2505
+ "epoch": 0.8,
2506
+ "learning_rate": 1.974127433460377e-05,
2507
+ "loss": 2.2639,
2508
+ "step": 203000
2509
+ },
2510
+ {
2511
+ "epoch": 0.8,
2512
+ "learning_rate": 1.954359274429491e-05,
2513
+ "loss": 2.2602,
2514
+ "step": 203500
2515
+ },
2516
+ {
2517
+ "epoch": 0.81,
2518
+ "learning_rate": 1.934591115398605e-05,
2519
+ "loss": 2.2636,
2520
+ "step": 204000
2521
+ },
2522
+ {
2523
+ "epoch": 0.81,
2524
+ "learning_rate": 1.9148229563677194e-05,
2525
+ "loss": 2.2611,
2526
+ "step": 204500
2527
+ },
2528
+ {
2529
+ "epoch": 0.81,
2530
+ "learning_rate": 1.8950547973368338e-05,
2531
+ "loss": 2.2573,
2532
+ "step": 205000
2533
+ },
2534
+ {
2535
+ "epoch": 0.81,
2536
+ "learning_rate": 1.8752866383059478e-05,
2537
+ "loss": 2.2539,
2538
+ "step": 205500
2539
+ },
2540
+ {
2541
+ "epoch": 0.81,
2542
+ "learning_rate": 1.855518479275062e-05,
2543
+ "loss": 2.2621,
2544
+ "step": 206000
2545
+ },
2546
+ {
2547
+ "epoch": 0.82,
2548
+ "learning_rate": 1.8357503202441765e-05,
2549
+ "loss": 2.2614,
2550
+ "step": 206500
2551
+ },
2552
+ {
2553
+ "epoch": 0.82,
2554
+ "learning_rate": 1.815982161213291e-05,
2555
+ "loss": 2.2549,
2556
+ "step": 207000
2557
+ },
2558
+ {
2559
+ "epoch": 0.82,
2560
+ "learning_rate": 1.796214002182405e-05,
2561
+ "loss": 2.247,
2562
+ "step": 207500
2563
+ },
2564
+ {
2565
+ "epoch": 0.82,
2566
+ "learning_rate": 1.776445843151519e-05,
2567
+ "loss": 2.243,
2568
+ "step": 208000
2569
+ },
2570
+ {
2571
+ "epoch": 0.82,
2572
+ "learning_rate": 1.7566776841206333e-05,
2573
+ "loss": 2.2571,
2574
+ "step": 208500
2575
+ },
2576
+ {
2577
+ "epoch": 0.83,
2578
+ "learning_rate": 1.7369095250897474e-05,
2579
+ "loss": 2.2594,
2580
+ "step": 209000
2581
+ },
2582
+ {
2583
+ "epoch": 0.83,
2584
+ "learning_rate": 1.7171413660588617e-05,
2585
+ "loss": 2.2589,
2586
+ "step": 209500
2587
+ },
2588
+ {
2589
+ "epoch": 0.83,
2590
+ "learning_rate": 1.697373207027976e-05,
2591
+ "loss": 2.2585,
2592
+ "step": 210000
2593
+ },
2594
+ {
2595
+ "epoch": 0.83,
2596
+ "learning_rate": 1.6776050479970905e-05,
2597
+ "loss": 2.2479,
2598
+ "step": 210500
2599
+ },
2600
+ {
2601
+ "epoch": 0.83,
2602
+ "learning_rate": 1.6578368889662045e-05,
2603
+ "loss": 2.2532,
2604
+ "step": 211000
2605
+ },
2606
+ {
2607
+ "epoch": 0.84,
2608
+ "learning_rate": 1.6380687299353185e-05,
2609
+ "loss": 2.2419,
2610
+ "step": 211500
2611
+ },
2612
+ {
2613
+ "epoch": 0.84,
2614
+ "learning_rate": 1.618300570904433e-05,
2615
+ "loss": 2.2498,
2616
+ "step": 212000
2617
+ },
2618
+ {
2619
+ "epoch": 0.84,
2620
+ "learning_rate": 1.598532411873547e-05,
2621
+ "loss": 2.2444,
2622
+ "step": 212500
2623
+ },
2624
+ {
2625
+ "epoch": 0.84,
2626
+ "learning_rate": 1.5787642528426613e-05,
2627
+ "loss": 2.2498,
2628
+ "step": 213000
2629
+ },
2630
+ {
2631
+ "epoch": 0.84,
2632
+ "learning_rate": 1.5589960938117756e-05,
2633
+ "loss": 2.2475,
2634
+ "step": 213500
2635
+ },
2636
+ {
2637
+ "epoch": 0.85,
2638
+ "learning_rate": 1.5392279347808897e-05,
2639
+ "loss": 2.2552,
2640
+ "step": 214000
2641
+ },
2642
+ {
2643
+ "epoch": 0.85,
2644
+ "learning_rate": 1.519459775750004e-05,
2645
+ "loss": 2.2422,
2646
+ "step": 214500
2647
+ },
2648
+ {
2649
+ "epoch": 0.85,
2650
+ "learning_rate": 1.4996916167191182e-05,
2651
+ "loss": 2.2537,
2652
+ "step": 215000
2653
+ },
2654
+ {
2655
+ "epoch": 0.85,
2656
+ "learning_rate": 1.4799234576882326e-05,
2657
+ "loss": 2.2395,
2658
+ "step": 215500
2659
+ },
2660
+ {
2661
+ "epoch": 0.85,
2662
+ "learning_rate": 1.4601552986573466e-05,
2663
+ "loss": 2.242,
2664
+ "step": 216000
2665
+ },
2666
+ {
2667
+ "epoch": 0.86,
2668
+ "learning_rate": 1.440387139626461e-05,
2669
+ "loss": 2.2437,
2670
+ "step": 216500
2671
+ },
2672
+ {
2673
+ "epoch": 0.86,
2674
+ "learning_rate": 1.4206189805955752e-05,
2675
+ "loss": 2.2392,
2676
+ "step": 217000
2677
+ },
2678
+ {
2679
+ "epoch": 0.86,
2680
+ "learning_rate": 1.4008508215646892e-05,
2681
+ "loss": 2.2444,
2682
+ "step": 217500
2683
+ },
2684
+ {
2685
+ "epoch": 0.86,
2686
+ "learning_rate": 1.3810826625338036e-05,
2687
+ "loss": 2.2448,
2688
+ "step": 218000
2689
+ },
2690
+ {
2691
+ "epoch": 0.86,
2692
+ "learning_rate": 1.361314503502918e-05,
2693
+ "loss": 2.2497,
2694
+ "step": 218500
2695
+ },
2696
+ {
2697
+ "epoch": 0.87,
2698
+ "learning_rate": 1.3415463444720322e-05,
2699
+ "loss": 2.2489,
2700
+ "step": 219000
2701
+ },
2702
+ {
2703
+ "epoch": 0.87,
2704
+ "learning_rate": 1.3217781854411462e-05,
2705
+ "loss": 2.2343,
2706
+ "step": 219500
2707
+ },
2708
+ {
2709
+ "epoch": 0.87,
2710
+ "learning_rate": 1.3020100264102606e-05,
2711
+ "loss": 2.2393,
2712
+ "step": 220000
2713
+ },
2714
+ {
2715
+ "epoch": 0.87,
2716
+ "learning_rate": 1.282241867379375e-05,
2717
+ "loss": 2.2432,
2718
+ "step": 220500
2719
+ },
2720
+ {
2721
+ "epoch": 0.87,
2722
+ "learning_rate": 1.262473708348489e-05,
2723
+ "loss": 2.2424,
2724
+ "step": 221000
2725
+ },
2726
+ {
2727
+ "epoch": 0.88,
2728
+ "learning_rate": 1.2427055493176031e-05,
2729
+ "loss": 2.2546,
2730
+ "step": 221500
2731
+ },
2732
+ {
2733
+ "epoch": 0.88,
2734
+ "learning_rate": 1.2229373902867175e-05,
2735
+ "loss": 2.2374,
2736
+ "step": 222000
2737
+ },
2738
+ {
2739
+ "epoch": 0.88,
2740
+ "learning_rate": 1.2031692312558317e-05,
2741
+ "loss": 2.2418,
2742
+ "step": 222500
2743
+ },
2744
+ {
2745
+ "epoch": 0.88,
2746
+ "learning_rate": 1.1834010722249459e-05,
2747
+ "loss": 2.2354,
2748
+ "step": 223000
2749
+ },
2750
+ {
2751
+ "epoch": 0.88,
2752
+ "learning_rate": 1.1636329131940601e-05,
2753
+ "loss": 2.2361,
2754
+ "step": 223500
2755
+ },
2756
+ {
2757
+ "epoch": 0.89,
2758
+ "learning_rate": 1.1438647541631743e-05,
2759
+ "loss": 2.2432,
2760
+ "step": 224000
2761
+ },
2762
+ {
2763
+ "epoch": 0.89,
2764
+ "learning_rate": 1.1240965951322887e-05,
2765
+ "loss": 2.2386,
2766
+ "step": 224500
2767
+ },
2768
+ {
2769
+ "epoch": 0.89,
2770
+ "learning_rate": 1.1043284361014029e-05,
2771
+ "loss": 2.2249,
2772
+ "step": 225000
2773
+ },
2774
+ {
2775
+ "epoch": 0.89,
2776
+ "eval_loss": 2.0728375911712646,
2777
+ "eval_runtime": 187.3688,
2778
+ "eval_samples_per_second": 533.707,
2779
+ "eval_steps_per_second": 66.713,
2780
+ "step": 225000
2781
+ },
2782
+ {
2783
+ "epoch": 0.89,
2784
+ "learning_rate": 1.0845602770705169e-05,
2785
+ "loss": 2.2418,
2786
+ "step": 225500
2787
+ },
2788
+ {
2789
+ "epoch": 0.89,
2790
+ "learning_rate": 1.0647921180396313e-05,
2791
+ "loss": 2.2308,
2792
+ "step": 226000
2793
+ },
2794
+ {
2795
+ "epoch": 0.9,
2796
+ "learning_rate": 1.0450239590087455e-05,
2797
+ "loss": 2.2325,
2798
+ "step": 226500
2799
+ },
2800
+ {
2801
+ "epoch": 0.9,
2802
+ "learning_rate": 1.0252557999778598e-05,
2803
+ "loss": 2.2238,
2804
+ "step": 227000
2805
+ },
2806
+ {
2807
+ "epoch": 0.9,
2808
+ "learning_rate": 1.0054876409469739e-05,
2809
+ "loss": 2.2336,
2810
+ "step": 227500
2811
+ },
2812
+ {
2813
+ "epoch": 0.9,
2814
+ "learning_rate": 9.857194819160882e-06,
2815
+ "loss": 2.2374,
2816
+ "step": 228000
2817
+ },
2818
+ {
2819
+ "epoch": 0.9,
2820
+ "learning_rate": 9.659513228852024e-06,
2821
+ "loss": 2.2312,
2822
+ "step": 228500
2823
+ },
2824
+ {
2825
+ "epoch": 0.91,
2826
+ "learning_rate": 9.461831638543166e-06,
2827
+ "loss": 2.2274,
2828
+ "step": 229000
2829
+ },
2830
+ {
2831
+ "epoch": 0.91,
2832
+ "learning_rate": 9.264150048234308e-06,
2833
+ "loss": 2.2265,
2834
+ "step": 229500
2835
+ },
2836
+ {
2837
+ "epoch": 0.91,
2838
+ "learning_rate": 9.06646845792545e-06,
2839
+ "loss": 2.2343,
2840
+ "step": 230000
2841
+ },
2842
+ {
2843
+ "epoch": 0.91,
2844
+ "learning_rate": 8.868786867616594e-06,
2845
+ "loss": 2.2292,
2846
+ "step": 230500
2847
+ },
2848
+ {
2849
+ "epoch": 0.91,
2850
+ "learning_rate": 8.671105277307736e-06,
2851
+ "loss": 2.2292,
2852
+ "step": 231000
2853
+ },
2854
+ {
2855
+ "epoch": 0.92,
2856
+ "learning_rate": 8.473423686998878e-06,
2857
+ "loss": 2.2317,
2858
+ "step": 231500
2859
+ },
2860
+ {
2861
+ "epoch": 0.92,
2862
+ "learning_rate": 8.27574209669002e-06,
2863
+ "loss": 2.2348,
2864
+ "step": 232000
2865
+ },
2866
+ {
2867
+ "epoch": 0.92,
2868
+ "learning_rate": 8.078060506381162e-06,
2869
+ "loss": 2.2243,
2870
+ "step": 232500
2871
+ },
2872
+ {
2873
+ "epoch": 0.92,
2874
+ "learning_rate": 7.880378916072305e-06,
2875
+ "loss": 2.2302,
2876
+ "step": 233000
2877
+ },
2878
+ {
2879
+ "epoch": 0.92,
2880
+ "learning_rate": 7.682697325763447e-06,
2881
+ "loss": 2.2187,
2882
+ "step": 233500
2883
+ },
2884
+ {
2885
+ "epoch": 0.93,
2886
+ "learning_rate": 7.4850157354545885e-06,
2887
+ "loss": 2.2321,
2888
+ "step": 234000
2889
+ },
2890
+ {
2891
+ "epoch": 0.93,
2892
+ "learning_rate": 7.287334145145731e-06,
2893
+ "loss": 2.2179,
2894
+ "step": 234500
2895
+ },
2896
+ {
2897
+ "epoch": 0.93,
2898
+ "learning_rate": 7.089652554836873e-06,
2899
+ "loss": 2.2292,
2900
+ "step": 235000
2901
+ },
2902
+ {
2903
+ "epoch": 0.93,
2904
+ "learning_rate": 6.891970964528016e-06,
2905
+ "loss": 2.2253,
2906
+ "step": 235500
2907
+ },
2908
+ {
2909
+ "epoch": 0.93,
2910
+ "learning_rate": 6.694289374219158e-06,
2911
+ "loss": 2.2198,
2912
+ "step": 236000
2913
+ },
2914
+ {
2915
+ "epoch": 0.94,
2916
+ "learning_rate": 6.496607783910301e-06,
2917
+ "loss": 2.222,
2918
+ "step": 236500
2919
+ },
2920
+ {
2921
+ "epoch": 0.94,
2922
+ "learning_rate": 6.298926193601443e-06,
2923
+ "loss": 2.2334,
2924
+ "step": 237000
2925
+ },
2926
+ {
2927
+ "epoch": 0.94,
2928
+ "learning_rate": 6.101244603292585e-06,
2929
+ "loss": 2.2309,
2930
+ "step": 237500
2931
+ },
2932
+ {
2933
+ "epoch": 0.94,
2934
+ "learning_rate": 5.903563012983728e-06,
2935
+ "loss": 2.223,
2936
+ "step": 238000
2937
+ },
2938
+ {
2939
+ "epoch": 0.94,
2940
+ "learning_rate": 5.705881422674869e-06,
2941
+ "loss": 2.2223,
2942
+ "step": 238500
2943
+ },
2944
+ {
2945
+ "epoch": 0.94,
2946
+ "learning_rate": 5.508199832366012e-06,
2947
+ "loss": 2.2224,
2948
+ "step": 239000
2949
+ },
2950
+ {
2951
+ "epoch": 0.95,
2952
+ "learning_rate": 5.310518242057154e-06,
2953
+ "loss": 2.2232,
2954
+ "step": 239500
2955
+ },
2956
+ {
2957
+ "epoch": 0.95,
2958
+ "learning_rate": 5.1128366517482965e-06,
2959
+ "loss": 2.2263,
2960
+ "step": 240000
2961
+ },
2962
+ {
2963
+ "epoch": 0.95,
2964
+ "learning_rate": 4.9151550614394385e-06,
2965
+ "loss": 2.2241,
2966
+ "step": 240500
2967
+ },
2968
+ {
2969
+ "epoch": 0.95,
2970
+ "learning_rate": 4.717473471130581e-06,
2971
+ "loss": 2.2265,
2972
+ "step": 241000
2973
+ },
2974
+ {
2975
+ "epoch": 0.95,
2976
+ "learning_rate": 4.5197918808217224e-06,
2977
+ "loss": 2.215,
2978
+ "step": 241500
2979
+ },
2980
+ {
2981
+ "epoch": 0.96,
2982
+ "learning_rate": 4.322110290512865e-06,
2983
+ "loss": 2.222,
2984
+ "step": 242000
2985
+ },
2986
+ {
2987
+ "epoch": 0.96,
2988
+ "learning_rate": 4.124428700204007e-06,
2989
+ "loss": 2.2281,
2990
+ "step": 242500
2991
+ },
2992
+ {
2993
+ "epoch": 0.96,
2994
+ "learning_rate": 3.92674710989515e-06,
2995
+ "loss": 2.2164,
2996
+ "step": 243000
2997
+ },
2998
+ {
2999
+ "epoch": 0.96,
3000
+ "learning_rate": 3.7290655195862925e-06,
3001
+ "loss": 2.216,
3002
+ "step": 243500
3003
+ },
3004
+ {
3005
+ "epoch": 0.96,
3006
+ "learning_rate": 3.531383929277434e-06,
3007
+ "loss": 2.2129,
3008
+ "step": 244000
3009
+ },
3010
+ {
3011
+ "epoch": 0.97,
3012
+ "learning_rate": 3.3337023389685764e-06,
3013
+ "loss": 2.2055,
3014
+ "step": 244500
3015
+ },
3016
+ {
3017
+ "epoch": 0.97,
3018
+ "learning_rate": 3.136020748659719e-06,
3019
+ "loss": 2.2171,
3020
+ "step": 245000
3021
+ },
3022
+ {
3023
+ "epoch": 0.97,
3024
+ "learning_rate": 2.9383391583508612e-06,
3025
+ "loss": 2.2211,
3026
+ "step": 245500
3027
+ },
3028
+ {
3029
+ "epoch": 0.97,
3030
+ "learning_rate": 2.7406575680420036e-06,
3031
+ "loss": 2.2162,
3032
+ "step": 246000
3033
+ },
3034
+ {
3035
+ "epoch": 0.97,
3036
+ "learning_rate": 2.5429759777331456e-06,
3037
+ "loss": 2.2107,
3038
+ "step": 246500
3039
+ },
3040
+ {
3041
+ "epoch": 0.98,
3042
+ "learning_rate": 2.345294387424288e-06,
3043
+ "loss": 2.211,
3044
+ "step": 247000
3045
+ },
3046
+ {
3047
+ "epoch": 0.98,
3048
+ "learning_rate": 2.1476127971154304e-06,
3049
+ "loss": 2.2221,
3050
+ "step": 247500
3051
+ },
3052
+ {
3053
+ "epoch": 0.98,
3054
+ "learning_rate": 1.9499312068065724e-06,
3055
+ "loss": 2.212,
3056
+ "step": 248000
3057
+ },
3058
+ {
3059
+ "epoch": 0.98,
3060
+ "learning_rate": 1.752249616497715e-06,
3061
+ "loss": 2.22,
3062
+ "step": 248500
3063
+ },
3064
+ {
3065
+ "epoch": 0.98,
3066
+ "learning_rate": 1.5545680261888572e-06,
3067
+ "loss": 2.2029,
3068
+ "step": 249000
3069
+ },
3070
+ {
3071
+ "epoch": 0.99,
3072
+ "learning_rate": 1.3568864358799994e-06,
3073
+ "loss": 2.2029,
3074
+ "step": 249500
3075
+ },
3076
+ {
3077
+ "epoch": 0.99,
3078
+ "learning_rate": 1.1592048455711418e-06,
3079
+ "loss": 2.2177,
3080
+ "step": 250000
3081
+ },
3082
+ {
3083
+ "epoch": 0.99,
3084
+ "eval_loss": 2.0536677837371826,
3085
+ "eval_runtime": 187.7144,
3086
+ "eval_samples_per_second": 532.724,
3087
+ "eval_steps_per_second": 66.591,
3088
+ "step": 250000
3089
+ },
3090
+ {
3091
+ "epoch": 0.99,
3092
+ "learning_rate": 9.61523255262284e-07,
3093
+ "loss": 2.2143,
3094
+ "step": 250500
3095
+ },
3096
+ {
3097
+ "epoch": 0.99,
3098
+ "learning_rate": 7.638416649534263e-07,
3099
+ "loss": 2.2206,
3100
+ "step": 251000
3101
+ },
3102
+ {
3103
+ "epoch": 0.99,
3104
+ "learning_rate": 5.661600746445686e-07,
3105
+ "loss": 2.2153,
3106
+ "step": 251500
3107
+ },
3108
+ {
3109
+ "epoch": 1.0,
3110
+ "learning_rate": 3.684784843357108e-07,
3111
+ "loss": 2.2198,
3112
+ "step": 252000
3113
+ },
3114
+ {
3115
+ "epoch": 1.0,
3116
+ "learning_rate": 1.7079689402685307e-07,
3117
+ "loss": 2.2138,
3118
+ "step": 252500
3119
+ },
3120
+ {
3121
+ "epoch": 1.0,
3122
+ "step": 252932,
3123
+ "total_flos": 2.131165835358634e+18,
3124
+ "train_loss": 0.24544810387602325,
3125
+ "train_runtime": 44838.2229,
3126
+ "train_samples_per_second": 1444.095,
3127
+ "train_steps_per_second": 5.641
3128
+ }
3129
+ ],
3130
+ "max_steps": 252932,
3131
+ "num_train_epochs": 1,
3132
+ "total_flos": 2.131165835358634e+18,
3133
+ "trial_name": null,
3134
+ "trial_params": null
3135
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67100f50919d55f8f1752ea2321d8dc656ae3f27290d2c2e47b3c76dc3e35afb
3
+ size 2799
vocab.txt ADDED
The diff for this file is too large to render. See raw diff