Hugging Face's logo

Anna759
/
lora-2nov

Model card Files Community
Yanfeng commited on
Commit
0070e5a
•
1 Parent(s): 0db9072

First commit

Browse files
Files changed (14) hide show
  1. adapter_config.json +17 -0
  2. adapter_model.bin +3 -0
  3. checkpoint-6000/adapter_config.json +17 -0
  4. checkpoint-6000/adapter_model.bin +3 -0
  5. checkpoint-6000/optimizer.pt +3 -0
  6. checkpoint-6000/rng_state_0.pth +3 -0
  7. checkpoint-6000/rng_state_1.pth +3 -0
  8. checkpoint-6000/rng_state_2.pth +3 -0
  9. checkpoint-6000/rng_state_3.pth +3 -0
  10. checkpoint-6000/rng_state_4.pth +3 -0
  11. checkpoint-6000/rng_state_5.pth +3 -0
  12. checkpoint-6000/scheduler.pt +3 -0
  13. checkpoint-6000/trainer_state.json +3859 -0
  14. checkpoint-6000/training_args.bin +3 -0
adapter_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "../alpaca_weights/chinese-alpaca",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "lora_alpha": 16,
8
+ "lora_dropout": 0.05,
9
+ "modules_to_save": null,
10
+ "peft_type": "LORA",
11
+ "r": 8,
12
+ "target_modules": [
13
+ "q_proj",
14
+ "v_proj"
15
+ ],
16
+ "task_type": "CAUSAL_LM"
17
+ }
adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5e1621f48d9ad8feb1d6d31050275f0aafd080c5c07153301fe2f48411f4406
3
+ size 443
checkpoint-6000/adapter_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "../alpaca_weights/chinese-alpaca",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": true,
6
+ "init_lora_weights": true,
7
+ "lora_alpha": 16,
8
+ "lora_dropout": 0.05,
9
+ "modules_to_save": null,
10
+ "peft_type": "LORA",
11
+ "r": 8,
12
+ "target_modules": [
13
+ "q_proj",
14
+ "v_proj"
15
+ ],
16
+ "task_type": "CAUSAL_LM"
17
+ }
checkpoint-6000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5e1621f48d9ad8feb1d6d31050275f0aafd080c5c07153301fe2f48411f4406
3
+ size 443
checkpoint-6000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24c6c6c69e422a9866b7e63c2105fa82e54729b25a736f558a023442102d0a61
3
+ size 52562821
checkpoint-6000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:299c3d73b7055b0575c71ab25f0f4654913c2f6756efefb3f2746c403834dc02
3
+ size 19639
checkpoint-6000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a777a12417bf6e41c4dda371cfcb9199f23f25f5637806ac2caeaa43af71825
3
+ size 19639
checkpoint-6000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21475b9e62fa35cc2d8376ab55c56b2a7e6ce65d6b6d92853b43491060cbe4c3
3
+ size 19639
checkpoint-6000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc475b3aa37f01ed5a14ca69d49d559e4cfef1a4d013577aaefc788b04482440
3
+ size 19639
checkpoint-6000/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b47919b7327a49b2f503a4016fa7fd95bcf64d51ec91b8cfcf64458ab4a4839a
3
+ size 19639
checkpoint-6000/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a6b0d94556d648cd537b0ae5bc0078f84327bfe34f981e4340ddc55b00377d0
3
+ size 19639
checkpoint-6000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bea775cd3a033d0f762b2df7c1217ebdbe9e335687e204a10a0620aa276592b6
3
+ size 627
checkpoint-6000/trainer_state.json ADDED
@@ -0,0 +1,3859 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9106292724609375,
3
+ "best_model_checkpoint": "./results/chinese-llama-2-lora-2nov/checkpoint-6000",
4
+ "epoch": 9.868421052631579,
5
+ "eval_steps": 200,
6
+ "global_step": 6000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "learning_rate": 2.9999999999999997e-05,
14
+ "loss": 2.2931,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.03,
19
+ "learning_rate": 5.9999999999999995e-05,
20
+ "loss": 2.2482,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.05,
25
+ "learning_rate": 8.999999999999999e-05,
26
+ "loss": 2.0171,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.07,
31
+ "learning_rate": 0.00011999999999999999,
32
+ "loss": 1.6749,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.08,
37
+ "learning_rate": 0.00015,
38
+ "loss": 1.5164,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.1,
43
+ "learning_rate": 0.00017999999999999998,
44
+ "loss": 1.4101,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.12,
49
+ "learning_rate": 0.00020999999999999998,
50
+ "loss": 1.3538,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.13,
55
+ "learning_rate": 0.00023999999999999998,
56
+ "loss": 1.2861,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.15,
61
+ "learning_rate": 0.00027,
62
+ "loss": 1.2444,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.16,
67
+ "learning_rate": 0.0003,
68
+ "loss": 1.2369,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.18,
73
+ "learning_rate": 0.0002994983277591973,
74
+ "loss": 1.2258,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.2,
79
+ "learning_rate": 0.0002989966555183946,
80
+ "loss": 1.1953,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.21,
85
+ "learning_rate": 0.00029849498327759195,
86
+ "loss": 1.2009,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.23,
91
+ "learning_rate": 0.0002979933110367893,
92
+ "loss": 1.1498,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.25,
97
+ "learning_rate": 0.0002974916387959866,
98
+ "loss": 1.1646,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.26,
103
+ "learning_rate": 0.0002969899665551839,
104
+ "loss": 1.1375,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.28,
109
+ "learning_rate": 0.00029648829431438127,
110
+ "loss": 1.1429,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.3,
115
+ "learning_rate": 0.00029598662207357856,
116
+ "loss": 1.1322,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.31,
121
+ "learning_rate": 0.0002954849498327759,
122
+ "loss": 1.1255,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.33,
127
+ "learning_rate": 0.00029498327759197324,
128
+ "loss": 1.1026,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.33,
133
+ "eval_loss": 1.1218857765197754,
134
+ "eval_runtime": 33.0988,
135
+ "eval_samples_per_second": 60.425,
136
+ "eval_steps_per_second": 1.269,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 0.35,
141
+ "learning_rate": 0.00029448160535117053,
142
+ "loss": 1.1085,
143
+ "step": 210
144
+ },
145
+ {
146
+ "epoch": 0.36,
147
+ "learning_rate": 0.0002939799331103679,
148
+ "loss": 1.0916,
149
+ "step": 220
150
+ },
151
+ {
152
+ "epoch": 0.38,
153
+ "learning_rate": 0.0002934782608695652,
154
+ "loss": 1.1211,
155
+ "step": 230
156
+ },
157
+ {
158
+ "epoch": 0.39,
159
+ "learning_rate": 0.0002929765886287625,
160
+ "loss": 1.0773,
161
+ "step": 240
162
+ },
163
+ {
164
+ "epoch": 0.41,
165
+ "learning_rate": 0.00029247491638795985,
166
+ "loss": 1.085,
167
+ "step": 250
168
+ },
169
+ {
170
+ "epoch": 0.43,
171
+ "learning_rate": 0.0002919732441471572,
172
+ "loss": 1.0696,
173
+ "step": 260
174
+ },
175
+ {
176
+ "epoch": 0.44,
177
+ "learning_rate": 0.0002914715719063545,
178
+ "loss": 1.0814,
179
+ "step": 270
180
+ },
181
+ {
182
+ "epoch": 0.46,
183
+ "learning_rate": 0.0002909698996655518,
184
+ "loss": 1.0931,
185
+ "step": 280
186
+ },
187
+ {
188
+ "epoch": 0.48,
189
+ "learning_rate": 0.00029046822742474917,
190
+ "loss": 1.064,
191
+ "step": 290
192
+ },
193
+ {
194
+ "epoch": 0.49,
195
+ "learning_rate": 0.00028996655518394646,
196
+ "loss": 1.076,
197
+ "step": 300
198
+ },
199
+ {
200
+ "epoch": 0.51,
201
+ "learning_rate": 0.0002894648829431438,
202
+ "loss": 1.0667,
203
+ "step": 310
204
+ },
205
+ {
206
+ "epoch": 0.53,
207
+ "learning_rate": 0.00028896321070234115,
208
+ "loss": 1.0682,
209
+ "step": 320
210
+ },
211
+ {
212
+ "epoch": 0.54,
213
+ "learning_rate": 0.00028846153846153843,
214
+ "loss": 1.0589,
215
+ "step": 330
216
+ },
217
+ {
218
+ "epoch": 0.56,
219
+ "learning_rate": 0.0002879598662207358,
220
+ "loss": 1.0524,
221
+ "step": 340
222
+ },
223
+ {
224
+ "epoch": 0.58,
225
+ "learning_rate": 0.0002874581939799331,
226
+ "loss": 1.061,
227
+ "step": 350
228
+ },
229
+ {
230
+ "epoch": 0.59,
231
+ "learning_rate": 0.0002869565217391304,
232
+ "loss": 1.0575,
233
+ "step": 360
234
+ },
235
+ {
236
+ "epoch": 0.61,
237
+ "learning_rate": 0.00028645484949832775,
238
+ "loss": 1.0381,
239
+ "step": 370
240
+ },
241
+ {
242
+ "epoch": 0.62,
243
+ "learning_rate": 0.00028595317725752504,
244
+ "loss": 1.0463,
245
+ "step": 380
246
+ },
247
+ {
248
+ "epoch": 0.64,
249
+ "learning_rate": 0.0002854515050167224,
250
+ "loss": 1.0616,
251
+ "step": 390
252
+ },
253
+ {
254
+ "epoch": 0.66,
255
+ "learning_rate": 0.00028494983277591973,
256
+ "loss": 1.0462,
257
+ "step": 400
258
+ },
259
+ {
260
+ "epoch": 0.66,
261
+ "eval_loss": 1.0583561658859253,
262
+ "eval_runtime": 34.9211,
263
+ "eval_samples_per_second": 57.272,
264
+ "eval_steps_per_second": 1.203,
265
+ "step": 400
266
+ },
267
+ {
268
+ "epoch": 0.67,
269
+ "learning_rate": 0.000284448160535117,
270
+ "loss": 1.0481,
271
+ "step": 410
272
+ },
273
+ {
274
+ "epoch": 0.69,
275
+ "learning_rate": 0.00028394648829431436,
276
+ "loss": 1.0362,
277
+ "step": 420
278
+ },
279
+ {
280
+ "epoch": 0.71,
281
+ "learning_rate": 0.0002834448160535117,
282
+ "loss": 1.0517,
283
+ "step": 430
284
+ },
285
+ {
286
+ "epoch": 0.72,
287
+ "learning_rate": 0.000282943143812709,
288
+ "loss": 1.0465,
289
+ "step": 440
290
+ },
291
+ {
292
+ "epoch": 0.74,
293
+ "learning_rate": 0.00028244147157190634,
294
+ "loss": 1.0313,
295
+ "step": 450
296
+ },
297
+ {
298
+ "epoch": 0.76,
299
+ "learning_rate": 0.0002819397993311037,
300
+ "loss": 1.0134,
301
+ "step": 460
302
+ },
303
+ {
304
+ "epoch": 0.77,
305
+ "learning_rate": 0.00028143812709030097,
306
+ "loss": 1.0373,
307
+ "step": 470
308
+ },
309
+ {
310
+ "epoch": 0.79,
311
+ "learning_rate": 0.0002809364548494983,
312
+ "loss": 1.0357,
313
+ "step": 480
314
+ },
315
+ {
316
+ "epoch": 0.81,
317
+ "learning_rate": 0.00028043478260869565,
318
+ "loss": 1.0113,
319
+ "step": 490
320
+ },
321
+ {
322
+ "epoch": 0.82,
323
+ "learning_rate": 0.00027993311036789294,
324
+ "loss": 1.0201,
325
+ "step": 500
326
+ },
327
+ {
328
+ "epoch": 0.84,
329
+ "learning_rate": 0.0002794314381270903,
330
+ "loss": 1.0359,
331
+ "step": 510
332
+ },
333
+ {
334
+ "epoch": 0.86,
335
+ "learning_rate": 0.00027892976588628763,
336
+ "loss": 1.0064,
337
+ "step": 520
338
+ },
339
+ {
340
+ "epoch": 0.87,
341
+ "learning_rate": 0.0002784280936454849,
342
+ "loss": 1.0166,
343
+ "step": 530
344
+ },
345
+ {
346
+ "epoch": 0.89,
347
+ "learning_rate": 0.00027792642140468226,
348
+ "loss": 1.0025,
349
+ "step": 540
350
+ },
351
+ {
352
+ "epoch": 0.9,
353
+ "learning_rate": 0.0002774247491638796,
354
+ "loss": 1.028,
355
+ "step": 550
356
+ },
357
+ {
358
+ "epoch": 0.92,
359
+ "learning_rate": 0.0002769230769230769,
360
+ "loss": 1.0216,
361
+ "step": 560
362
+ },
363
+ {
364
+ "epoch": 0.94,
365
+ "learning_rate": 0.00027642140468227424,
366
+ "loss": 1.0037,
367
+ "step": 570
368
+ },
369
+ {
370
+ "epoch": 0.95,
371
+ "learning_rate": 0.0002759197324414716,
372
+ "loss": 1.0028,
373
+ "step": 580
374
+ },
375
+ {
376
+ "epoch": 0.97,
377
+ "learning_rate": 0.00027541806020066887,
378
+ "loss": 1.0193,
379
+ "step": 590
380
+ },
381
+ {
382
+ "epoch": 0.99,
383
+ "learning_rate": 0.00027491638795986616,
384
+ "loss": 1.0128,
385
+ "step": 600
386
+ },
387
+ {
388
+ "epoch": 0.99,
389
+ "eval_loss": 1.0269545316696167,
390
+ "eval_runtime": 33.0707,
391
+ "eval_samples_per_second": 60.476,
392
+ "eval_steps_per_second": 1.27,
393
+ "step": 600
394
+ },
395
+ {
396
+ "epoch": 1.0,
397
+ "learning_rate": 0.0002744147157190635,
398
+ "loss": 0.9854,
399
+ "step": 610
400
+ },
401
+ {
402
+ "epoch": 1.02,
403
+ "learning_rate": 0.00027391304347826085,
404
+ "loss": 1.0161,
405
+ "step": 620
406
+ },
407
+ {
408
+ "epoch": 1.04,
409
+ "learning_rate": 0.00027341137123745813,
410
+ "loss": 0.9755,
411
+ "step": 630
412
+ },
413
+ {
414
+ "epoch": 1.05,
415
+ "learning_rate": 0.0002729096989966555,
416
+ "loss": 0.997,
417
+ "step": 640
418
+ },
419
+ {
420
+ "epoch": 1.07,
421
+ "learning_rate": 0.0002724080267558528,
422
+ "loss": 0.9834,
423
+ "step": 650
424
+ },
425
+ {
426
+ "epoch": 1.09,
427
+ "learning_rate": 0.0002719063545150501,
428
+ "loss": 1.0024,
429
+ "step": 660
430
+ },
431
+ {
432
+ "epoch": 1.1,
433
+ "learning_rate": 0.00027140468227424745,
434
+ "loss": 0.985,
435
+ "step": 670
436
+ },
437
+ {
438
+ "epoch": 1.12,
439
+ "learning_rate": 0.0002709030100334448,
440
+ "loss": 1.0037,
441
+ "step": 680
442
+ },
443
+ {
444
+ "epoch": 1.13,
445
+ "learning_rate": 0.0002704013377926421,
446
+ "loss": 1.0004,
447
+ "step": 690
448
+ },
449
+ {
450
+ "epoch": 1.15,
451
+ "learning_rate": 0.00026989966555183943,
452
+ "loss": 1.006,
453
+ "step": 700
454
+ },
455
+ {
456
+ "epoch": 1.17,
457
+ "learning_rate": 0.00026939799331103677,
458
+ "loss": 0.9887,
459
+ "step": 710
460
+ },
461
+ {
462
+ "epoch": 1.18,
463
+ "learning_rate": 0.00026889632107023406,
464
+ "loss": 0.9817,
465
+ "step": 720
466
+ },
467
+ {
468
+ "epoch": 1.2,
469
+ "learning_rate": 0.0002683946488294314,
470
+ "loss": 0.9718,
471
+ "step": 730
472
+ },
473
+ {
474
+ "epoch": 1.22,
475
+ "learning_rate": 0.00026789297658862875,
476
+ "loss": 0.9958,
477
+ "step": 740
478
+ },
479
+ {
480
+ "epoch": 1.23,
481
+ "learning_rate": 0.00026739130434782604,
482
+ "loss": 0.9654,
483
+ "step": 750
484
+ },
485
+ {
486
+ "epoch": 1.25,
487
+ "learning_rate": 0.0002668896321070234,
488
+ "loss": 0.9906,
489
+ "step": 760
490
+ },
491
+ {
492
+ "epoch": 1.27,
493
+ "learning_rate": 0.0002663879598662207,
494
+ "loss": 0.9922,
495
+ "step": 770
496
+ },
497
+ {
498
+ "epoch": 1.28,
499
+ "learning_rate": 0.000265886287625418,
500
+ "loss": 0.9803,
501
+ "step": 780
502
+ },
503
+ {
504
+ "epoch": 1.3,
505
+ "learning_rate": 0.00026538461538461536,
506
+ "loss": 0.9709,
507
+ "step": 790
508
+ },
509
+ {
510
+ "epoch": 1.32,
511
+ "learning_rate": 0.0002648829431438127,
512
+ "loss": 0.9562,
513
+ "step": 800
514
+ },
515
+ {
516
+ "epoch": 1.32,
517
+ "eval_loss": 1.007567048072815,
518
+ "eval_runtime": 33.1229,
519
+ "eval_samples_per_second": 60.381,
520
+ "eval_steps_per_second": 1.268,
521
+ "step": 800
522
+ },
523
+ {
524
+ "epoch": 1.33,
525
+ "learning_rate": 0.00026438127090301,
526
+ "loss": 0.98,
527
+ "step": 810
528
+ },
529
+ {
530
+ "epoch": 1.35,
531
+ "learning_rate": 0.00026387959866220733,
532
+ "loss": 1.0006,
533
+ "step": 820
534
+ },
535
+ {
536
+ "epoch": 1.37,
537
+ "learning_rate": 0.0002633779264214047,
538
+ "loss": 0.9748,
539
+ "step": 830
540
+ },
541
+ {
542
+ "epoch": 1.38,
543
+ "learning_rate": 0.00026287625418060196,
544
+ "loss": 0.9878,
545
+ "step": 840
546
+ },
547
+ {
548
+ "epoch": 1.4,
549
+ "learning_rate": 0.0002623745819397993,
550
+ "loss": 0.9827,
551
+ "step": 850
552
+ },
553
+ {
554
+ "epoch": 1.41,
555
+ "learning_rate": 0.00026187290969899665,
556
+ "loss": 0.9752,
557
+ "step": 860
558
+ },
559
+ {
560
+ "epoch": 1.43,
561
+ "learning_rate": 0.00026137123745819394,
562
+ "loss": 0.9675,
563
+ "step": 870
564
+ },
565
+ {
566
+ "epoch": 1.45,
567
+ "learning_rate": 0.0002608695652173913,
568
+ "loss": 0.9563,
569
+ "step": 880
570
+ },
571
+ {
572
+ "epoch": 1.46,
573
+ "learning_rate": 0.0002603678929765886,
574
+ "loss": 0.99,
575
+ "step": 890
576
+ },
577
+ {
578
+ "epoch": 1.48,
579
+ "learning_rate": 0.0002598662207357859,
580
+ "loss": 1.0059,
581
+ "step": 900
582
+ },
583
+ {
584
+ "epoch": 1.5,
585
+ "learning_rate": 0.00025936454849498326,
586
+ "loss": 0.9916,
587
+ "step": 910
588
+ },
589
+ {
590
+ "epoch": 1.51,
591
+ "learning_rate": 0.0002588628762541806,
592
+ "loss": 0.9866,
593
+ "step": 920
594
+ },
595
+ {
596
+ "epoch": 1.53,
597
+ "learning_rate": 0.0002583612040133779,
598
+ "loss": 0.9522,
599
+ "step": 930
600
+ },
601
+ {
602
+ "epoch": 1.55,
603
+ "learning_rate": 0.00025785953177257523,
604
+ "loss": 0.9421,
605
+ "step": 940
606
+ },
607
+ {
608
+ "epoch": 1.56,
609
+ "learning_rate": 0.0002573578595317725,
610
+ "loss": 0.9678,
611
+ "step": 950
612
+ },
613
+ {
614
+ "epoch": 1.58,
615
+ "learning_rate": 0.00025685618729096986,
616
+ "loss": 0.9556,
617
+ "step": 960
618
+ },
619
+ {
620
+ "epoch": 1.6,
621
+ "learning_rate": 0.0002563545150501672,
622
+ "loss": 0.9794,
623
+ "step": 970
624
+ },
625
+ {
626
+ "epoch": 1.61,
627
+ "learning_rate": 0.0002558528428093645,
628
+ "loss": 0.9603,
629
+ "step": 980
630
+ },
631
+ {
632
+ "epoch": 1.63,
633
+ "learning_rate": 0.00025535117056856184,
634
+ "loss": 0.9457,
635
+ "step": 990
636
+ },
637
+ {
638
+ "epoch": 1.64,
639
+ "learning_rate": 0.0002548494983277592,
640
+ "loss": 0.9656,
641
+ "step": 1000
642
+ },
643
+ {
644
+ "epoch": 1.64,
645
+ "eval_loss": 0.9902360439300537,
646
+ "eval_runtime": 33.1444,
647
+ "eval_samples_per_second": 60.342,
648
+ "eval_steps_per_second": 1.267,
649
+ "step": 1000
650
+ },
651
+ {
652
+ "epoch": 1.66,
653
+ "learning_rate": 0.00025434782608695647,
654
+ "loss": 0.955,
655
+ "step": 1010
656
+ },
657
+ {
658
+ "epoch": 1.68,
659
+ "learning_rate": 0.0002538461538461538,
660
+ "loss": 0.9664,
661
+ "step": 1020
662
+ },
663
+ {
664
+ "epoch": 1.69,
665
+ "learning_rate": 0.00025334448160535116,
666
+ "loss": 0.9712,
667
+ "step": 1030
668
+ },
669
+ {
670
+ "epoch": 1.71,
671
+ "learning_rate": 0.00025284280936454845,
672
+ "loss": 0.947,
673
+ "step": 1040
674
+ },
675
+ {
676
+ "epoch": 1.73,
677
+ "learning_rate": 0.0002523411371237458,
678
+ "loss": 0.973,
679
+ "step": 1050
680
+ },
681
+ {
682
+ "epoch": 1.74,
683
+ "learning_rate": 0.00025183946488294313,
684
+ "loss": 0.9664,
685
+ "step": 1060
686
+ },
687
+ {
688
+ "epoch": 1.76,
689
+ "learning_rate": 0.0002513377926421404,
690
+ "loss": 0.9581,
691
+ "step": 1070
692
+ },
693
+ {
694
+ "epoch": 1.78,
695
+ "learning_rate": 0.00025083612040133777,
696
+ "loss": 0.9544,
697
+ "step": 1080
698
+ },
699
+ {
700
+ "epoch": 1.79,
701
+ "learning_rate": 0.0002503344481605351,
702
+ "loss": 0.9774,
703
+ "step": 1090
704
+ },
705
+ {
706
+ "epoch": 1.81,
707
+ "learning_rate": 0.0002498327759197324,
708
+ "loss": 0.9493,
709
+ "step": 1100
710
+ },
711
+ {
712
+ "epoch": 1.83,
713
+ "learning_rate": 0.00024933110367892974,
714
+ "loss": 0.9584,
715
+ "step": 1110
716
+ },
717
+ {
718
+ "epoch": 1.84,
719
+ "learning_rate": 0.0002488294314381271,
720
+ "loss": 0.947,
721
+ "step": 1120
722
+ },
723
+ {
724
+ "epoch": 1.86,
725
+ "learning_rate": 0.0002483277591973244,
726
+ "loss": 0.9502,
727
+ "step": 1130
728
+ },
729
+ {
730
+ "epoch": 1.88,
731
+ "learning_rate": 0.0002478260869565217,
732
+ "loss": 0.944,
733
+ "step": 1140
734
+ },
735
+ {
736
+ "epoch": 1.89,
737
+ "learning_rate": 0.00024732441471571906,
738
+ "loss": 0.9556,
739
+ "step": 1150
740
+ },
741
+ {
742
+ "epoch": 1.91,
743
+ "learning_rate": 0.00024682274247491635,
744
+ "loss": 0.9627,
745
+ "step": 1160
746
+ },
747
+ {
748
+ "epoch": 1.92,
749
+ "learning_rate": 0.0002463210702341137,
750
+ "loss": 0.9376,
751
+ "step": 1170
752
+ },
753
+ {
754
+ "epoch": 1.94,
755
+ "learning_rate": 0.00024581939799331104,
756
+ "loss": 0.9399,
757
+ "step": 1180
758
+ },
759
+ {
760
+ "epoch": 1.96,
761
+ "learning_rate": 0.0002453177257525083,
762
+ "loss": 0.9449,
763
+ "step": 1190
764
+ },
765
+ {
766
+ "epoch": 1.97,
767
+ "learning_rate": 0.00024481605351170567,
768
+ "loss": 0.9514,
769
+ "step": 1200
770
+ },
771
+ {
772
+ "epoch": 1.97,
773
+ "eval_loss": 0.9769035577774048,
774
+ "eval_runtime": 33.1416,
775
+ "eval_samples_per_second": 60.347,
776
+ "eval_steps_per_second": 1.267,
777
+ "step": 1200
778
+ },
779
+ {
780
+ "epoch": 1.99,
781
+ "learning_rate": 0.000244314381270903,
782
+ "loss": 0.9439,
783
+ "step": 1210
784
+ },
785
+ {
786
+ "epoch": 2.01,
787
+ "learning_rate": 0.00024381270903010033,
788
+ "loss": 0.9512,
789
+ "step": 1220
790
+ },
791
+ {
792
+ "epoch": 2.02,
793
+ "learning_rate": 0.00024331103678929764,
794
+ "loss": 0.948,
795
+ "step": 1230
796
+ },
797
+ {
798
+ "epoch": 2.04,
799
+ "learning_rate": 0.00024280936454849496,
800
+ "loss": 0.9421,
801
+ "step": 1240
802
+ },
803
+ {
804
+ "epoch": 2.06,
805
+ "learning_rate": 0.0002423076923076923,
806
+ "loss": 0.9323,
807
+ "step": 1250
808
+ },
809
+ {
810
+ "epoch": 2.07,
811
+ "learning_rate": 0.00024180602006688962,
812
+ "loss": 0.9278,
813
+ "step": 1260
814
+ },
815
+ {
816
+ "epoch": 2.09,
817
+ "learning_rate": 0.00024130434782608694,
818
+ "loss": 0.9255,
819
+ "step": 1270
820
+ },
821
+ {
822
+ "epoch": 2.11,
823
+ "learning_rate": 0.00024080267558528428,
824
+ "loss": 0.9214,
825
+ "step": 1280
826
+ },
827
+ {
828
+ "epoch": 2.12,
829
+ "learning_rate": 0.0002403010033444816,
830
+ "loss": 0.9456,
831
+ "step": 1290
832
+ },
833
+ {
834
+ "epoch": 2.14,
835
+ "learning_rate": 0.0002397993311036789,
836
+ "loss": 0.9435,
837
+ "step": 1300
838
+ },
839
+ {
840
+ "epoch": 2.15,
841
+ "learning_rate": 0.00023929765886287623,
842
+ "loss": 0.9261,
843
+ "step": 1310
844
+ },
845
+ {
846
+ "epoch": 2.17,
847
+ "learning_rate": 0.00023879598662207357,
848
+ "loss": 0.9225,
849
+ "step": 1320
850
+ },
851
+ {
852
+ "epoch": 2.19,
853
+ "learning_rate": 0.00023829431438127089,
854
+ "loss": 0.9184,
855
+ "step": 1330
856
+ },
857
+ {
858
+ "epoch": 2.2,
859
+ "learning_rate": 0.0002377926421404682,
860
+ "loss": 0.9301,
861
+ "step": 1340
862
+ },
863
+ {
864
+ "epoch": 2.22,
865
+ "learning_rate": 0.00023729096989966555,
866
+ "loss": 0.9599,
867
+ "step": 1350
868
+ },
869
+ {
870
+ "epoch": 2.24,
871
+ "learning_rate": 0.00023678929765886286,
872
+ "loss": 0.9316,
873
+ "step": 1360
874
+ },
875
+ {
876
+ "epoch": 2.25,
877
+ "learning_rate": 0.00023628762541806018,
878
+ "loss": 0.9214,
879
+ "step": 1370
880
+ },
881
+ {
882
+ "epoch": 2.27,
883
+ "learning_rate": 0.00023578595317725752,
884
+ "loss": 0.9206,
885
+ "step": 1380
886
+ },
887
+ {
888
+ "epoch": 2.29,
889
+ "learning_rate": 0.00023528428093645484,
890
+ "loss": 0.9413,
891
+ "step": 1390
892
+ },
893
+ {
894
+ "epoch": 2.3,
895
+ "learning_rate": 0.00023478260869565215,
896
+ "loss": 0.9136,
897
+ "step": 1400
898
+ },
899
+ {
900
+ "epoch": 2.3,
901
+ "eval_loss": 0.9681736826896667,
902
+ "eval_runtime": 33.0949,
903
+ "eval_samples_per_second": 60.432,
904
+ "eval_steps_per_second": 1.269,
905
+ "step": 1400
906
+ },
907
+ {
908
+ "epoch": 2.32,
909
+ "learning_rate": 0.0002342809364548495,
910
+ "loss": 0.9128,
911
+ "step": 1410
912
+ },
913
+ {
914
+ "epoch": 2.34,
915
+ "learning_rate": 0.0002337792642140468,
916
+ "loss": 0.9229,
917
+ "step": 1420
918
+ },
919
+ {
920
+ "epoch": 2.35,
921
+ "learning_rate": 0.00023327759197324413,
922
+ "loss": 0.9127,
923
+ "step": 1430
924
+ },
925
+ {
926
+ "epoch": 2.37,
927
+ "learning_rate": 0.00023277591973244147,
928
+ "loss": 0.9305,
929
+ "step": 1440
930
+ },
931
+ {
932
+ "epoch": 2.38,
933
+ "learning_rate": 0.0002322742474916388,
934
+ "loss": 0.9238,
935
+ "step": 1450
936
+ },
937
+ {
938
+ "epoch": 2.4,
939
+ "learning_rate": 0.0002317725752508361,
940
+ "loss": 0.9179,
941
+ "step": 1460
942
+ },
943
+ {
944
+ "epoch": 2.42,
945
+ "learning_rate": 0.00023127090301003345,
946
+ "loss": 0.9392,
947
+ "step": 1470
948
+ },
949
+ {
950
+ "epoch": 2.43,
951
+ "learning_rate": 0.00023076923076923076,
952
+ "loss": 0.9301,
953
+ "step": 1480
954
+ },
955
+ {
956
+ "epoch": 2.45,
957
+ "learning_rate": 0.00023026755852842808,
958
+ "loss": 0.9298,
959
+ "step": 1490
960
+ },
961
+ {
962
+ "epoch": 2.47,
963
+ "learning_rate": 0.00022976588628762542,
964
+ "loss": 0.9426,
965
+ "step": 1500
966
+ },
967
+ {
968
+ "epoch": 2.48,
969
+ "learning_rate": 0.00022926421404682274,
970
+ "loss": 0.9228,
971
+ "step": 1510
972
+ },
973
+ {
974
+ "epoch": 2.5,
975
+ "learning_rate": 0.00022876254180602006,
976
+ "loss": 0.9242,
977
+ "step": 1520
978
+ },
979
+ {
980
+ "epoch": 2.52,
981
+ "learning_rate": 0.0002282608695652174,
982
+ "loss": 0.9317,
983
+ "step": 1530
984
+ },
985
+ {
986
+ "epoch": 2.53,
987
+ "learning_rate": 0.00022775919732441471,
988
+ "loss": 0.9093,
989
+ "step": 1540
990
+ },
991
+ {
992
+ "epoch": 2.55,
993
+ "learning_rate": 0.00022725752508361203,
994
+ "loss": 0.91,
995
+ "step": 1550
996
+ },
997
+ {
998
+ "epoch": 2.57,
999
+ "learning_rate": 0.00022675585284280935,
1000
+ "loss": 0.9177,
1001
+ "step": 1560
1002
+ },
1003
+ {
1004
+ "epoch": 2.58,
1005
+ "learning_rate": 0.0002262541806020067,
1006
+ "loss": 0.9222,
1007
+ "step": 1570
1008
+ },
1009
+ {
1010
+ "epoch": 2.6,
1011
+ "learning_rate": 0.000225752508361204,
1012
+ "loss": 0.9165,
1013
+ "step": 1580
1014
+ },
1015
+ {
1016
+ "epoch": 2.62,
1017
+ "learning_rate": 0.00022525083612040132,
1018
+ "loss": 0.9159,
1019
+ "step": 1590
1020
+ },
1021
+ {
1022
+ "epoch": 2.63,
1023
+ "learning_rate": 0.00022474916387959864,
1024
+ "loss": 0.887,
1025
+ "step": 1600
1026
+ },
1027
+ {
1028
+ "epoch": 2.63,
1029
+ "eval_loss": 0.9578931927680969,
1030
+ "eval_runtime": 33.1845,
1031
+ "eval_samples_per_second": 60.269,
1032
+ "eval_steps_per_second": 1.266,
1033
+ "step": 1600
1034
+ },
1035
+ {
1036
+ "epoch": 2.65,
1037
+ "learning_rate": 0.00022424749163879595,
1038
+ "loss": 0.8999,
1039
+ "step": 1610
1040
+ },
1041
+ {
1042
+ "epoch": 2.66,
1043
+ "learning_rate": 0.00022374581939799327,
1044
+ "loss": 0.9044,
1045
+ "step": 1620
1046
+ },
1047
+ {
1048
+ "epoch": 2.68,
1049
+ "learning_rate": 0.00022324414715719061,
1050
+ "loss": 0.8885,
1051
+ "step": 1630
1052
+ },
1053
+ {
1054
+ "epoch": 2.7,
1055
+ "learning_rate": 0.00022274247491638793,
1056
+ "loss": 0.9134,
1057
+ "step": 1640
1058
+ },
1059
+ {
1060
+ "epoch": 2.71,
1061
+ "learning_rate": 0.00022224080267558525,
1062
+ "loss": 0.9055,
1063
+ "step": 1650
1064
+ },
1065
+ {
1066
+ "epoch": 2.73,
1067
+ "learning_rate": 0.00022173913043478256,
1068
+ "loss": 0.919,
1069
+ "step": 1660
1070
+ },
1071
+ {
1072
+ "epoch": 2.75,
1073
+ "learning_rate": 0.0002212374581939799,
1074
+ "loss": 0.9358,
1075
+ "step": 1670
1076
+ },
1077
+ {
1078
+ "epoch": 2.76,
1079
+ "learning_rate": 0.00022073578595317722,
1080
+ "loss": 0.918,
1081
+ "step": 1680
1082
+ },
1083
+ {
1084
+ "epoch": 2.78,
1085
+ "learning_rate": 0.00022023411371237454,
1086
+ "loss": 0.9072,
1087
+ "step": 1690
1088
+ },
1089
+ {
1090
+ "epoch": 2.8,
1091
+ "learning_rate": 0.00021973244147157188,
1092
+ "loss": 0.9131,
1093
+ "step": 1700
1094
+ },
1095
+ {
1096
+ "epoch": 2.81,
1097
+ "learning_rate": 0.0002192307692307692,
1098
+ "loss": 0.9092,
1099
+ "step": 1710
1100
+ },
1101
+ {
1102
+ "epoch": 2.83,
1103
+ "learning_rate": 0.0002187290969899665,
1104
+ "loss": 0.8981,
1105
+ "step": 1720
1106
+ },
1107
+ {
1108
+ "epoch": 2.85,
1109
+ "learning_rate": 0.00021822742474916386,
1110
+ "loss": 0.9273,
1111
+ "step": 1730
1112
+ },
1113
+ {
1114
+ "epoch": 2.86,
1115
+ "learning_rate": 0.00021772575250836117,
1116
+ "loss": 0.9226,
1117
+ "step": 1740
1118
+ },
1119
+ {
1120
+ "epoch": 2.88,
1121
+ "learning_rate": 0.0002172240802675585,
1122
+ "loss": 0.9169,
1123
+ "step": 1750
1124
+ },
1125
+ {
1126
+ "epoch": 2.89,
1127
+ "learning_rate": 0.00021672240802675583,
1128
+ "loss": 0.9278,
1129
+ "step": 1760
1130
+ },
1131
+ {
1132
+ "epoch": 2.91,
1133
+ "learning_rate": 0.00021622073578595315,
1134
+ "loss": 0.907,
1135
+ "step": 1770
1136
+ },
1137
+ {
1138
+ "epoch": 2.93,
1139
+ "learning_rate": 0.00021571906354515046,
1140
+ "loss": 0.8985,
1141
+ "step": 1780
1142
+ },
1143
+ {
1144
+ "epoch": 2.94,
1145
+ "learning_rate": 0.0002152173913043478,
1146
+ "loss": 0.9093,
1147
+ "step": 1790
1148
+ },
1149
+ {
1150
+ "epoch": 2.96,
1151
+ "learning_rate": 0.00021471571906354512,
1152
+ "loss": 0.8969,
1153
+ "step": 1800
1154
+ },
1155
+ {
1156
+ "epoch": 2.96,
1157
+ "eval_loss": 0.9505146145820618,
1158
+ "eval_runtime": 33.1216,
1159
+ "eval_samples_per_second": 60.383,
1160
+ "eval_steps_per_second": 1.268,
1161
+ "step": 1800
1162
+ },
1163
+ {
1164
+ "epoch": 2.98,
1165
+ "learning_rate": 0.00021421404682274244,
1166
+ "loss": 0.8899,
1167
+ "step": 1810
1168
+ },
1169
+ {
1170
+ "epoch": 2.99,
1171
+ "learning_rate": 0.00021371237458193978,
1172
+ "loss": 0.8995,
1173
+ "step": 1820
1174
+ },
1175
+ {
1176
+ "epoch": 3.01,
1177
+ "learning_rate": 0.0002132107023411371,
1178
+ "loss": 0.9034,
1179
+ "step": 1830
1180
+ },
1181
+ {
1182
+ "epoch": 3.03,
1183
+ "learning_rate": 0.00021270903010033441,
1184
+ "loss": 0.8901,
1185
+ "step": 1840
1186
+ },
1187
+ {
1188
+ "epoch": 3.04,
1189
+ "learning_rate": 0.00021220735785953176,
1190
+ "loss": 0.8977,
1191
+ "step": 1850
1192
+ },
1193
+ {
1194
+ "epoch": 3.06,
1195
+ "learning_rate": 0.00021170568561872907,
1196
+ "loss": 0.891,
1197
+ "step": 1860
1198
+ },
1199
+ {
1200
+ "epoch": 3.08,
1201
+ "learning_rate": 0.0002112040133779264,
1202
+ "loss": 0.8848,
1203
+ "step": 1870
1204
+ },
1205
+ {
1206
+ "epoch": 3.09,
1207
+ "learning_rate": 0.00021070234113712373,
1208
+ "loss": 0.8829,
1209
+ "step": 1880
1210
+ },
1211
+ {
1212
+ "epoch": 3.11,
1213
+ "learning_rate": 0.00021020066889632105,
1214
+ "loss": 0.8728,
1215
+ "step": 1890
1216
+ },
1217
+ {
1218
+ "epoch": 3.12,
1219
+ "learning_rate": 0.00020969899665551837,
1220
+ "loss": 0.8927,
1221
+ "step": 1900
1222
+ },
1223
+ {
1224
+ "epoch": 3.14,
1225
+ "learning_rate": 0.00020919732441471568,
1226
+ "loss": 0.8834,
1227
+ "step": 1910
1228
+ },
1229
+ {
1230
+ "epoch": 3.16,
1231
+ "learning_rate": 0.00020869565217391303,
1232
+ "loss": 0.9003,
1233
+ "step": 1920
1234
+ },
1235
+ {
1236
+ "epoch": 3.17,
1237
+ "learning_rate": 0.00020819397993311034,
1238
+ "loss": 0.8958,
1239
+ "step": 1930
1240
+ },
1241
+ {
1242
+ "epoch": 3.19,
1243
+ "learning_rate": 0.00020769230769230766,
1244
+ "loss": 0.9081,
1245
+ "step": 1940
1246
+ },
1247
+ {
1248
+ "epoch": 3.21,
1249
+ "learning_rate": 0.000207190635451505,
1250
+ "loss": 0.9027,
1251
+ "step": 1950
1252
+ },
1253
+ {
1254
+ "epoch": 3.22,
1255
+ "learning_rate": 0.00020668896321070232,
1256
+ "loss": 0.8997,
1257
+ "step": 1960
1258
+ },
1259
+ {
1260
+ "epoch": 3.24,
1261
+ "learning_rate": 0.00020618729096989963,
1262
+ "loss": 0.8906,
1263
+ "step": 1970
1264
+ },
1265
+ {
1266
+ "epoch": 3.26,
1267
+ "learning_rate": 0.00020568561872909698,
1268
+ "loss": 0.8886,
1269
+ "step": 1980
1270
+ },
1271
+ {
1272
+ "epoch": 3.27,
1273
+ "learning_rate": 0.0002051839464882943,
1274
+ "loss": 0.8886,
1275
+ "step": 1990
1276
+ },
1277
+ {
1278
+ "epoch": 3.29,
1279
+ "learning_rate": 0.0002046822742474916,
1280
+ "loss": 0.8885,
1281
+ "step": 2000
1282
+ },
1283
+ {
1284
+ "epoch": 3.29,
1285
+ "eval_loss": 0.9456945061683655,
1286
+ "eval_runtime": 33.1317,
1287
+ "eval_samples_per_second": 60.365,
1288
+ "eval_steps_per_second": 1.268,
1289
+ "step": 2000
1290
+ },
1291
+ {
1292
+ "epoch": 3.31,
1293
+ "learning_rate": 0.00020418060200668895,
1294
+ "loss": 0.8992,
1295
+ "step": 2010
1296
+ },
1297
+ {
1298
+ "epoch": 3.32,
1299
+ "learning_rate": 0.00020367892976588627,
1300
+ "loss": 0.892,
1301
+ "step": 2020
1302
+ },
1303
+ {
1304
+ "epoch": 3.34,
1305
+ "learning_rate": 0.00020317725752508358,
1306
+ "loss": 0.8848,
1307
+ "step": 2030
1308
+ },
1309
+ {
1310
+ "epoch": 3.36,
1311
+ "learning_rate": 0.00020267558528428093,
1312
+ "loss": 0.8889,
1313
+ "step": 2040
1314
+ },
1315
+ {
1316
+ "epoch": 3.37,
1317
+ "learning_rate": 0.00020217391304347824,
1318
+ "loss": 0.8872,
1319
+ "step": 2050
1320
+ },
1321
+ {
1322
+ "epoch": 3.39,
1323
+ "learning_rate": 0.00020167224080267556,
1324
+ "loss": 0.8933,
1325
+ "step": 2060
1326
+ },
1327
+ {
1328
+ "epoch": 3.4,
1329
+ "learning_rate": 0.0002011705685618729,
1330
+ "loss": 0.8816,
1331
+ "step": 2070
1332
+ },
1333
+ {
1334
+ "epoch": 3.42,
1335
+ "learning_rate": 0.00020066889632107022,
1336
+ "loss": 0.8836,
1337
+ "step": 2080
1338
+ },
1339
+ {
1340
+ "epoch": 3.44,
1341
+ "learning_rate": 0.00020016722408026753,
1342
+ "loss": 0.8918,
1343
+ "step": 2090
1344
+ },
1345
+ {
1346
+ "epoch": 3.45,
1347
+ "learning_rate": 0.00019966555183946488,
1348
+ "loss": 0.8944,
1349
+ "step": 2100
1350
+ },
1351
+ {
1352
+ "epoch": 3.47,
1353
+ "learning_rate": 0.0001991638795986622,
1354
+ "loss": 0.9055,
1355
+ "step": 2110
1356
+ },
1357
+ {
1358
+ "epoch": 3.49,
1359
+ "learning_rate": 0.0001986622073578595,
1360
+ "loss": 0.8809,
1361
+ "step": 2120
1362
+ },
1363
+ {
1364
+ "epoch": 3.5,
1365
+ "learning_rate": 0.00019816053511705685,
1366
+ "loss": 0.8982,
1367
+ "step": 2130
1368
+ },
1369
+ {
1370
+ "epoch": 3.52,
1371
+ "learning_rate": 0.00019765886287625417,
1372
+ "loss": 0.89,
1373
+ "step": 2140
1374
+ },
1375
+ {
1376
+ "epoch": 3.54,
1377
+ "learning_rate": 0.00019715719063545149,
1378
+ "loss": 0.8798,
1379
+ "step": 2150
1380
+ },
1381
+ {
1382
+ "epoch": 3.55,
1383
+ "learning_rate": 0.0001966555183946488,
1384
+ "loss": 0.9001,
1385
+ "step": 2160
1386
+ },
1387
+ {
1388
+ "epoch": 3.57,
1389
+ "learning_rate": 0.00019615384615384615,
1390
+ "loss": 0.8746,
1391
+ "step": 2170
1392
+ },
1393
+ {
1394
+ "epoch": 3.59,
1395
+ "learning_rate": 0.00019565217391304346,
1396
+ "loss": 0.8945,
1397
+ "step": 2180
1398
+ },
1399
+ {
1400
+ "epoch": 3.6,
1401
+ "learning_rate": 0.00019515050167224078,
1402
+ "loss": 0.8745,
1403
+ "step": 2190
1404
+ },
1405
+ {
1406
+ "epoch": 3.62,
1407
+ "learning_rate": 0.00019464882943143812,
1408
+ "loss": 0.8695,
1409
+ "step": 2200
1410
+ },
1411
+ {
1412
+ "epoch": 3.62,
1413
+ "eval_loss": 0.9407016038894653,
1414
+ "eval_runtime": 33.1522,
1415
+ "eval_samples_per_second": 60.328,
1416
+ "eval_steps_per_second": 1.267,
1417
+ "step": 2200
1418
+ },
1419
+ {
1420
+ "epoch": 3.63,
1421
+ "learning_rate": 0.00019414715719063544,
1422
+ "loss": 0.8917,
1423
+ "step": 2210
1424
+ },
1425
+ {
1426
+ "epoch": 3.65,
1427
+ "learning_rate": 0.00019364548494983275,
1428
+ "loss": 0.8822,
1429
+ "step": 2220
1430
+ },
1431
+ {
1432
+ "epoch": 3.67,
1433
+ "learning_rate": 0.0001931438127090301,
1434
+ "loss": 0.8694,
1435
+ "step": 2230
1436
+ },
1437
+ {
1438
+ "epoch": 3.68,
1439
+ "learning_rate": 0.0001926421404682274,
1440
+ "loss": 0.8904,
1441
+ "step": 2240
1442
+ },
1443
+ {
1444
+ "epoch": 3.7,
1445
+ "learning_rate": 0.00019214046822742473,
1446
+ "loss": 0.8759,
1447
+ "step": 2250
1448
+ },
1449
+ {
1450
+ "epoch": 3.72,
1451
+ "learning_rate": 0.00019163879598662207,
1452
+ "loss": 0.8784,
1453
+ "step": 2260
1454
+ },
1455
+ {
1456
+ "epoch": 3.73,
1457
+ "learning_rate": 0.0001911371237458194,
1458
+ "loss": 0.8829,
1459
+ "step": 2270
1460
+ },
1461
+ {
1462
+ "epoch": 3.75,
1463
+ "learning_rate": 0.0001906354515050167,
1464
+ "loss": 0.8788,
1465
+ "step": 2280
1466
+ },
1467
+ {
1468
+ "epoch": 3.77,
1469
+ "learning_rate": 0.00019013377926421405,
1470
+ "loss": 0.8966,
1471
+ "step": 2290
1472
+ },
1473
+ {
1474
+ "epoch": 3.78,
1475
+ "learning_rate": 0.00018963210702341136,
1476
+ "loss": 0.8803,
1477
+ "step": 2300
1478
+ },
1479
+ {
1480
+ "epoch": 3.8,
1481
+ "learning_rate": 0.00018913043478260868,
1482
+ "loss": 0.8783,
1483
+ "step": 2310
1484
+ },
1485
+ {
1486
+ "epoch": 3.82,
1487
+ "learning_rate": 0.00018862876254180602,
1488
+ "loss": 0.8874,
1489
+ "step": 2320
1490
+ },
1491
+ {
1492
+ "epoch": 3.83,
1493
+ "learning_rate": 0.00018812709030100334,
1494
+ "loss": 0.8711,
1495
+ "step": 2330
1496
+ },
1497
+ {
1498
+ "epoch": 3.85,
1499
+ "learning_rate": 0.00018762541806020065,
1500
+ "loss": 0.8676,
1501
+ "step": 2340
1502
+ },
1503
+ {
1504
+ "epoch": 3.87,
1505
+ "learning_rate": 0.000187123745819398,
1506
+ "loss": 0.8638,
1507
+ "step": 2350
1508
+ },
1509
+ {
1510
+ "epoch": 3.88,
1511
+ "learning_rate": 0.00018662207357859531,
1512
+ "loss": 0.8945,
1513
+ "step": 2360
1514
+ },
1515
+ {
1516
+ "epoch": 3.9,
1517
+ "learning_rate": 0.00018612040133779263,
1518
+ "loss": 0.8772,
1519
+ "step": 2370
1520
+ },
1521
+ {
1522
+ "epoch": 3.91,
1523
+ "learning_rate": 0.00018561872909698997,
1524
+ "loss": 0.8895,
1525
+ "step": 2380
1526
+ },
1527
+ {
1528
+ "epoch": 3.93,
1529
+ "learning_rate": 0.0001851170568561873,
1530
+ "loss": 0.87,
1531
+ "step": 2390
1532
+ },
1533
+ {
1534
+ "epoch": 3.95,
1535
+ "learning_rate": 0.0001846153846153846,
1536
+ "loss": 0.8608,
1537
+ "step": 2400
1538
+ },
1539
+ {
1540
+ "epoch": 3.95,
1541
+ "eval_loss": 0.936105489730835,
1542
+ "eval_runtime": 33.1676,
1543
+ "eval_samples_per_second": 60.3,
1544
+ "eval_steps_per_second": 1.266,
1545
+ "step": 2400
1546
+ },
1547
+ {
1548
+ "epoch": 3.96,
1549
+ "learning_rate": 0.00018411371237458192,
1550
+ "loss": 0.8848,
1551
+ "step": 2410
1552
+ },
1553
+ {
1554
+ "epoch": 3.98,
1555
+ "learning_rate": 0.00018361204013377926,
1556
+ "loss": 0.8823,
1557
+ "step": 2420
1558
+ },
1559
+ {
1560
+ "epoch": 4.0,
1561
+ "learning_rate": 0.00018311036789297658,
1562
+ "loss": 0.8692,
1563
+ "step": 2430
1564
+ },
1565
+ {
1566
+ "epoch": 4.01,
1567
+ "learning_rate": 0.0001826086956521739,
1568
+ "loss": 0.896,
1569
+ "step": 2440
1570
+ },
1571
+ {
1572
+ "epoch": 4.03,
1573
+ "learning_rate": 0.00018210702341137124,
1574
+ "loss": 0.8704,
1575
+ "step": 2450
1576
+ },
1577
+ {
1578
+ "epoch": 4.05,
1579
+ "learning_rate": 0.00018160535117056856,
1580
+ "loss": 0.8869,
1581
+ "step": 2460
1582
+ },
1583
+ {
1584
+ "epoch": 4.06,
1585
+ "learning_rate": 0.00018110367892976587,
1586
+ "loss": 0.8492,
1587
+ "step": 2470
1588
+ },
1589
+ {
1590
+ "epoch": 4.08,
1591
+ "learning_rate": 0.00018060200668896322,
1592
+ "loss": 0.8558,
1593
+ "step": 2480
1594
+ },
1595
+ {
1596
+ "epoch": 4.1,
1597
+ "learning_rate": 0.00018010033444816053,
1598
+ "loss": 0.8607,
1599
+ "step": 2490
1600
+ },
1601
+ {
1602
+ "epoch": 4.11,
1603
+ "learning_rate": 0.00017959866220735785,
1604
+ "loss": 0.8651,
1605
+ "step": 2500
1606
+ },
1607
+ {
1608
+ "epoch": 4.13,
1609
+ "learning_rate": 0.0001790969899665552,
1610
+ "loss": 0.8648,
1611
+ "step": 2510
1612
+ },
1613
+ {
1614
+ "epoch": 4.14,
1615
+ "learning_rate": 0.0001785953177257525,
1616
+ "loss": 0.8614,
1617
+ "step": 2520
1618
+ },
1619
+ {
1620
+ "epoch": 4.16,
1621
+ "learning_rate": 0.00017809364548494982,
1622
+ "loss": 0.8759,
1623
+ "step": 2530
1624
+ },
1625
+ {
1626
+ "epoch": 4.18,
1627
+ "learning_rate": 0.00017759197324414717,
1628
+ "loss": 0.8653,
1629
+ "step": 2540
1630
+ },
1631
+ {
1632
+ "epoch": 4.19,
1633
+ "learning_rate": 0.00017709030100334448,
1634
+ "loss": 0.8478,
1635
+ "step": 2550
1636
+ },
1637
+ {
1638
+ "epoch": 4.21,
1639
+ "learning_rate": 0.0001765886287625418,
1640
+ "loss": 0.8636,
1641
+ "step": 2560
1642
+ },
1643
+ {
1644
+ "epoch": 4.23,
1645
+ "learning_rate": 0.00017608695652173914,
1646
+ "loss": 0.859,
1647
+ "step": 2570
1648
+ },
1649
+ {
1650
+ "epoch": 4.24,
1651
+ "learning_rate": 0.00017558528428093646,
1652
+ "loss": 0.8685,
1653
+ "step": 2580
1654
+ },
1655
+ {
1656
+ "epoch": 4.26,
1657
+ "learning_rate": 0.00017508361204013377,
1658
+ "loss": 0.8595,
1659
+ "step": 2590
1660
+ },
1661
+ {
1662
+ "epoch": 4.28,
1663
+ "learning_rate": 0.00017458193979933106,
1664
+ "loss": 0.8698,
1665
+ "step": 2600
1666
+ },
1667
+ {
1668
+ "epoch": 4.28,
1669
+ "eval_loss": 0.9336724877357483,
1670
+ "eval_runtime": 33.1889,
1671
+ "eval_samples_per_second": 60.261,
1672
+ "eval_steps_per_second": 1.265,
1673
+ "step": 2600
1674
+ },
1675
+ {
1676
+ "epoch": 4.29,
1677
+ "learning_rate": 0.0001740802675585284,
1678
+ "loss": 0.8696,
1679
+ "step": 2610
1680
+ },
1681
+ {
1682
+ "epoch": 4.31,
1683
+ "learning_rate": 0.00017357859531772572,
1684
+ "loss": 0.8667,
1685
+ "step": 2620
1686
+ },
1687
+ {
1688
+ "epoch": 4.33,
1689
+ "learning_rate": 0.00017307692307692304,
1690
+ "loss": 0.8648,
1691
+ "step": 2630
1692
+ },
1693
+ {
1694
+ "epoch": 4.34,
1695
+ "learning_rate": 0.00017257525083612038,
1696
+ "loss": 0.8524,
1697
+ "step": 2640
1698
+ },
1699
+ {
1700
+ "epoch": 4.36,
1701
+ "learning_rate": 0.0001720735785953177,
1702
+ "loss": 0.8506,
1703
+ "step": 2650
1704
+ },
1705
+ {
1706
+ "epoch": 4.38,
1707
+ "learning_rate": 0.00017157190635451501,
1708
+ "loss": 0.8638,
1709
+ "step": 2660
1710
+ },
1711
+ {
1712
+ "epoch": 4.39,
1713
+ "learning_rate": 0.00017107023411371236,
1714
+ "loss": 0.8769,
1715
+ "step": 2670
1716
+ },
1717
+ {
1718
+ "epoch": 4.41,
1719
+ "learning_rate": 0.00017056856187290967,
1720
+ "loss": 0.8581,
1721
+ "step": 2680
1722
+ },
1723
+ {
1724
+ "epoch": 4.42,
1725
+ "learning_rate": 0.000170066889632107,
1726
+ "loss": 0.8564,
1727
+ "step": 2690
1728
+ },
1729
+ {
1730
+ "epoch": 4.44,
1731
+ "learning_rate": 0.00016956521739130433,
1732
+ "loss": 0.8369,
1733
+ "step": 2700
1734
+ },
1735
+ {
1736
+ "epoch": 4.46,
1737
+ "learning_rate": 0.00016906354515050165,
1738
+ "loss": 0.879,
1739
+ "step": 2710
1740
+ },
1741
+ {
1742
+ "epoch": 4.47,
1743
+ "learning_rate": 0.00016856187290969897,
1744
+ "loss": 0.8426,
1745
+ "step": 2720
1746
+ },
1747
+ {
1748
+ "epoch": 4.49,
1749
+ "learning_rate": 0.00016806020066889628,
1750
+ "loss": 0.8603,
1751
+ "step": 2730
1752
+ },
1753
+ {
1754
+ "epoch": 4.51,
1755
+ "learning_rate": 0.00016755852842809362,
1756
+ "loss": 0.8613,
1757
+ "step": 2740
1758
+ },
1759
+ {
1760
+ "epoch": 4.52,
1761
+ "learning_rate": 0.00016705685618729094,
1762
+ "loss": 0.8695,
1763
+ "step": 2750
1764
+ },
1765
+ {
1766
+ "epoch": 4.54,
1767
+ "learning_rate": 0.00016655518394648826,
1768
+ "loss": 0.8652,
1769
+ "step": 2760
1770
+ },
1771
+ {
1772
+ "epoch": 4.56,
1773
+ "learning_rate": 0.0001660535117056856,
1774
+ "loss": 0.8537,
1775
+ "step": 2770
1776
+ },
1777
+ {
1778
+ "epoch": 4.57,
1779
+ "learning_rate": 0.00016555183946488292,
1780
+ "loss": 0.8529,
1781
+ "step": 2780
1782
+ },
1783
+ {
1784
+ "epoch": 4.59,
1785
+ "learning_rate": 0.00016505016722408023,
1786
+ "loss": 0.8426,
1787
+ "step": 2790
1788
+ },
1789
+ {
1790
+ "epoch": 4.61,
1791
+ "learning_rate": 0.00016454849498327758,
1792
+ "loss": 0.8691,
1793
+ "step": 2800
1794
+ },
1795
+ {
1796
+ "epoch": 4.61,
1797
+ "eval_loss": 0.9289954900741577,
1798
+ "eval_runtime": 33.1507,
1799
+ "eval_samples_per_second": 60.331,
1800
+ "eval_steps_per_second": 1.267,
1801
+ "step": 2800
1802
+ },
1803
+ {
1804
+ "epoch": 4.62,
1805
+ "learning_rate": 0.0001640468227424749,
1806
+ "loss": 0.8589,
1807
+ "step": 2810
1808
+ },
1809
+ {
1810
+ "epoch": 4.64,
1811
+ "learning_rate": 0.0001635451505016722,
1812
+ "loss": 0.8583,
1813
+ "step": 2820
1814
+ },
1815
+ {
1816
+ "epoch": 4.65,
1817
+ "learning_rate": 0.00016304347826086955,
1818
+ "loss": 0.8408,
1819
+ "step": 2830
1820
+ },
1821
+ {
1822
+ "epoch": 4.67,
1823
+ "learning_rate": 0.00016254180602006687,
1824
+ "loss": 0.8634,
1825
+ "step": 2840
1826
+ },
1827
+ {
1828
+ "epoch": 4.69,
1829
+ "learning_rate": 0.00016204013377926418,
1830
+ "loss": 0.8513,
1831
+ "step": 2850
1832
+ },
1833
+ {
1834
+ "epoch": 4.7,
1835
+ "learning_rate": 0.00016153846153846153,
1836
+ "loss": 0.8568,
1837
+ "step": 2860
1838
+ },
1839
+ {
1840
+ "epoch": 4.72,
1841
+ "learning_rate": 0.00016103678929765884,
1842
+ "loss": 0.8703,
1843
+ "step": 2870
1844
+ },
1845
+ {
1846
+ "epoch": 4.74,
1847
+ "learning_rate": 0.00016053511705685616,
1848
+ "loss": 0.8545,
1849
+ "step": 2880
1850
+ },
1851
+ {
1852
+ "epoch": 4.75,
1853
+ "learning_rate": 0.0001600334448160535,
1854
+ "loss": 0.8456,
1855
+ "step": 2890
1856
+ },
1857
+ {
1858
+ "epoch": 4.77,
1859
+ "learning_rate": 0.00015953177257525082,
1860
+ "loss": 0.8548,
1861
+ "step": 2900
1862
+ },
1863
+ {
1864
+ "epoch": 4.79,
1865
+ "learning_rate": 0.00015903010033444813,
1866
+ "loss": 0.8515,
1867
+ "step": 2910
1868
+ },
1869
+ {
1870
+ "epoch": 4.8,
1871
+ "learning_rate": 0.00015852842809364548,
1872
+ "loss": 0.8562,
1873
+ "step": 2920
1874
+ },
1875
+ {
1876
+ "epoch": 4.82,
1877
+ "learning_rate": 0.0001580267558528428,
1878
+ "loss": 0.8623,
1879
+ "step": 2930
1880
+ },
1881
+ {
1882
+ "epoch": 4.84,
1883
+ "learning_rate": 0.0001575250836120401,
1884
+ "loss": 0.855,
1885
+ "step": 2940
1886
+ },
1887
+ {
1888
+ "epoch": 4.85,
1889
+ "learning_rate": 0.00015702341137123745,
1890
+ "loss": 0.876,
1891
+ "step": 2950
1892
+ },
1893
+ {
1894
+ "epoch": 4.87,
1895
+ "learning_rate": 0.00015652173913043477,
1896
+ "loss": 0.8582,
1897
+ "step": 2960
1898
+ },
1899
+ {
1900
+ "epoch": 4.88,
1901
+ "learning_rate": 0.00015602006688963209,
1902
+ "loss": 0.8488,
1903
+ "step": 2970
1904
+ },
1905
+ {
1906
+ "epoch": 4.9,
1907
+ "learning_rate": 0.0001555183946488294,
1908
+ "loss": 0.8466,
1909
+ "step": 2980
1910
+ },
1911
+ {
1912
+ "epoch": 4.92,
1913
+ "learning_rate": 0.00015501672240802674,
1914
+ "loss": 0.8407,
1915
+ "step": 2990
1916
+ },
1917
+ {
1918
+ "epoch": 4.93,
1919
+ "learning_rate": 0.00015451505016722406,
1920
+ "loss": 0.8732,
1921
+ "step": 3000
1922
+ },
1923
+ {
1924
+ "epoch": 4.93,
1925
+ "eval_loss": 0.9256382584571838,
1926
+ "eval_runtime": 33.1403,
1927
+ "eval_samples_per_second": 60.349,
1928
+ "eval_steps_per_second": 1.267,
1929
+ "step": 3000
1930
+ },
1931
+ {
1932
+ "epoch": 4.95,
1933
+ "learning_rate": 0.00015401337792642138,
1934
+ "loss": 0.8422,
1935
+ "step": 3010
1936
+ },
1937
+ {
1938
+ "epoch": 4.97,
1939
+ "learning_rate": 0.00015351170568561872,
1940
+ "loss": 0.8528,
1941
+ "step": 3020
1942
+ },
1943
+ {
1944
+ "epoch": 4.98,
1945
+ "learning_rate": 0.00015301003344481604,
1946
+ "loss": 0.854,
1947
+ "step": 3030
1948
+ },
1949
+ {
1950
+ "epoch": 5.0,
1951
+ "learning_rate": 0.00015250836120401335,
1952
+ "loss": 0.8598,
1953
+ "step": 3040
1954
+ },
1955
+ {
1956
+ "epoch": 5.02,
1957
+ "learning_rate": 0.0001520066889632107,
1958
+ "loss": 0.843,
1959
+ "step": 3050
1960
+ },
1961
+ {
1962
+ "epoch": 5.03,
1963
+ "learning_rate": 0.000151505016722408,
1964
+ "loss": 0.8394,
1965
+ "step": 3060
1966
+ },
1967
+ {
1968
+ "epoch": 5.05,
1969
+ "learning_rate": 0.00015100334448160533,
1970
+ "loss": 0.8593,
1971
+ "step": 3070
1972
+ },
1973
+ {
1974
+ "epoch": 5.07,
1975
+ "learning_rate": 0.00015050167224080267,
1976
+ "loss": 0.8566,
1977
+ "step": 3080
1978
+ },
1979
+ {
1980
+ "epoch": 5.08,
1981
+ "learning_rate": 0.00015,
1982
+ "loss": 0.8469,
1983
+ "step": 3090
1984
+ },
1985
+ {
1986
+ "epoch": 5.1,
1987
+ "learning_rate": 0.0001494983277591973,
1988
+ "loss": 0.8421,
1989
+ "step": 3100
1990
+ },
1991
+ {
1992
+ "epoch": 5.12,
1993
+ "learning_rate": 0.00014899665551839465,
1994
+ "loss": 0.8425,
1995
+ "step": 3110
1996
+ },
1997
+ {
1998
+ "epoch": 5.13,
1999
+ "learning_rate": 0.00014849498327759196,
2000
+ "loss": 0.8341,
2001
+ "step": 3120
2002
+ },
2003
+ {
2004
+ "epoch": 5.15,
2005
+ "learning_rate": 0.00014804347826086954,
2006
+ "loss": 0.8527,
2007
+ "step": 3130
2008
+ },
2009
+ {
2010
+ "epoch": 5.16,
2011
+ "learning_rate": 0.00014754180602006688,
2012
+ "loss": 0.8517,
2013
+ "step": 3140
2014
+ },
2015
+ {
2016
+ "epoch": 5.18,
2017
+ "learning_rate": 0.0001470401337792642,
2018
+ "loss": 0.8389,
2019
+ "step": 3150
2020
+ },
2021
+ {
2022
+ "epoch": 5.2,
2023
+ "learning_rate": 0.00014653846153846151,
2024
+ "loss": 0.8311,
2025
+ "step": 3160
2026
+ },
2027
+ {
2028
+ "epoch": 5.21,
2029
+ "learning_rate": 0.00014603678929765886,
2030
+ "loss": 0.85,
2031
+ "step": 3170
2032
+ },
2033
+ {
2034
+ "epoch": 5.23,
2035
+ "learning_rate": 0.00014553511705685617,
2036
+ "loss": 0.8469,
2037
+ "step": 3180
2038
+ },
2039
+ {
2040
+ "epoch": 5.25,
2041
+ "learning_rate": 0.0001450334448160535,
2042
+ "loss": 0.8569,
2043
+ "step": 3190
2044
+ },
2045
+ {
2046
+ "epoch": 5.26,
2047
+ "learning_rate": 0.00014453177257525083,
2048
+ "loss": 0.853,
2049
+ "step": 3200
2050
+ },
2051
+ {
2052
+ "epoch": 5.26,
2053
+ "eval_loss": 0.9253825545310974,
2054
+ "eval_runtime": 33.1231,
2055
+ "eval_samples_per_second": 60.381,
2056
+ "eval_steps_per_second": 1.268,
2057
+ "step": 3200
2058
+ },
2059
+ {
2060
+ "epoch": 5.28,
2061
+ "learning_rate": 0.00014403010033444815,
2062
+ "loss": 0.8445,
2063
+ "step": 3210
2064
+ },
2065
+ {
2066
+ "epoch": 5.3,
2067
+ "learning_rate": 0.00014352842809364547,
2068
+ "loss": 0.8462,
2069
+ "step": 3220
2070
+ },
2071
+ {
2072
+ "epoch": 5.31,
2073
+ "learning_rate": 0.0001430267558528428,
2074
+ "loss": 0.8638,
2075
+ "step": 3230
2076
+ },
2077
+ {
2078
+ "epoch": 5.33,
2079
+ "learning_rate": 0.00014252508361204012,
2080
+ "loss": 0.8405,
2081
+ "step": 3240
2082
+ },
2083
+ {
2084
+ "epoch": 5.35,
2085
+ "learning_rate": 0.00014202341137123744,
2086
+ "loss": 0.8329,
2087
+ "step": 3250
2088
+ },
2089
+ {
2090
+ "epoch": 5.36,
2091
+ "learning_rate": 0.00014152173913043476,
2092
+ "loss": 0.8502,
2093
+ "step": 3260
2094
+ },
2095
+ {
2096
+ "epoch": 5.38,
2097
+ "learning_rate": 0.0001410200668896321,
2098
+ "loss": 0.8398,
2099
+ "step": 3270
2100
+ },
2101
+ {
2102
+ "epoch": 5.39,
2103
+ "learning_rate": 0.00014051839464882942,
2104
+ "loss": 0.8667,
2105
+ "step": 3280
2106
+ },
2107
+ {
2108
+ "epoch": 5.41,
2109
+ "learning_rate": 0.00014001672240802673,
2110
+ "loss": 0.8393,
2111
+ "step": 3290
2112
+ },
2113
+ {
2114
+ "epoch": 5.43,
2115
+ "learning_rate": 0.00013951505016722408,
2116
+ "loss": 0.8493,
2117
+ "step": 3300
2118
+ },
2119
+ {
2120
+ "epoch": 5.44,
2121
+ "learning_rate": 0.0001390133779264214,
2122
+ "loss": 0.8369,
2123
+ "step": 3310
2124
+ },
2125
+ {
2126
+ "epoch": 5.46,
2127
+ "learning_rate": 0.0001385117056856187,
2128
+ "loss": 0.8305,
2129
+ "step": 3320
2130
+ },
2131
+ {
2132
+ "epoch": 5.48,
2133
+ "learning_rate": 0.00013801003344481605,
2134
+ "loss": 0.8425,
2135
+ "step": 3330
2136
+ },
2137
+ {
2138
+ "epoch": 5.49,
2139
+ "learning_rate": 0.00013750836120401337,
2140
+ "loss": 0.8392,
2141
+ "step": 3340
2142
+ },
2143
+ {
2144
+ "epoch": 5.51,
2145
+ "learning_rate": 0.00013700668896321068,
2146
+ "loss": 0.8354,
2147
+ "step": 3350
2148
+ },
2149
+ {
2150
+ "epoch": 5.53,
2151
+ "learning_rate": 0.00013650501672240803,
2152
+ "loss": 0.8296,
2153
+ "step": 3360
2154
+ },
2155
+ {
2156
+ "epoch": 5.54,
2157
+ "learning_rate": 0.00013600334448160534,
2158
+ "loss": 0.8442,
2159
+ "step": 3370
2160
+ },
2161
+ {
2162
+ "epoch": 5.56,
2163
+ "learning_rate": 0.00013550167224080266,
2164
+ "loss": 0.84,
2165
+ "step": 3380
2166
+ },
2167
+ {
2168
+ "epoch": 5.58,
2169
+ "learning_rate": 0.000135,
2170
+ "loss": 0.8538,
2171
+ "step": 3390
2172
+ },
2173
+ {
2174
+ "epoch": 5.59,
2175
+ "learning_rate": 0.00013449832775919732,
2176
+ "loss": 0.8356,
2177
+ "step": 3400
2178
+ },
2179
+ {
2180
+ "epoch": 5.59,
2181
+ "eval_loss": 0.9222854971885681,
2182
+ "eval_runtime": 33.1332,
2183
+ "eval_samples_per_second": 60.362,
2184
+ "eval_steps_per_second": 1.268,
2185
+ "step": 3400
2186
+ },
2187
+ {
2188
+ "epoch": 5.61,
2189
+ "learning_rate": 0.00013399665551839463,
2190
+ "loss": 0.8417,
2191
+ "step": 3410
2192
+ },
2193
+ {
2194
+ "epoch": 5.62,
2195
+ "learning_rate": 0.00013349498327759198,
2196
+ "loss": 0.8287,
2197
+ "step": 3420
2198
+ },
2199
+ {
2200
+ "epoch": 5.64,
2201
+ "learning_rate": 0.0001329933110367893,
2202
+ "loss": 0.8429,
2203
+ "step": 3430
2204
+ },
2205
+ {
2206
+ "epoch": 5.66,
2207
+ "learning_rate": 0.0001324916387959866,
2208
+ "loss": 0.8453,
2209
+ "step": 3440
2210
+ },
2211
+ {
2212
+ "epoch": 5.67,
2213
+ "learning_rate": 0.00013198996655518395,
2214
+ "loss": 0.8436,
2215
+ "step": 3450
2216
+ },
2217
+ {
2218
+ "epoch": 5.69,
2219
+ "learning_rate": 0.00013148829431438127,
2220
+ "loss": 0.8357,
2221
+ "step": 3460
2222
+ },
2223
+ {
2224
+ "epoch": 5.71,
2225
+ "learning_rate": 0.00013098662207357858,
2226
+ "loss": 0.8334,
2227
+ "step": 3470
2228
+ },
2229
+ {
2230
+ "epoch": 5.72,
2231
+ "learning_rate": 0.00013048494983277593,
2232
+ "loss": 0.8401,
2233
+ "step": 3480
2234
+ },
2235
+ {
2236
+ "epoch": 5.74,
2237
+ "learning_rate": 0.00012998327759197324,
2238
+ "loss": 0.8513,
2239
+ "step": 3490
2240
+ },
2241
+ {
2242
+ "epoch": 5.76,
2243
+ "learning_rate": 0.00012948160535117056,
2244
+ "loss": 0.8369,
2245
+ "step": 3500
2246
+ },
2247
+ {
2248
+ "epoch": 5.77,
2249
+ "learning_rate": 0.00012897993311036788,
2250
+ "loss": 0.8437,
2251
+ "step": 3510
2252
+ },
2253
+ {
2254
+ "epoch": 5.79,
2255
+ "learning_rate": 0.00012847826086956522,
2256
+ "loss": 0.8422,
2257
+ "step": 3520
2258
+ },
2259
+ {
2260
+ "epoch": 5.81,
2261
+ "learning_rate": 0.00012797658862876254,
2262
+ "loss": 0.8437,
2263
+ "step": 3530
2264
+ },
2265
+ {
2266
+ "epoch": 5.82,
2267
+ "learning_rate": 0.00012747491638795985,
2268
+ "loss": 0.8277,
2269
+ "step": 3540
2270
+ },
2271
+ {
2272
+ "epoch": 5.84,
2273
+ "learning_rate": 0.00012697324414715717,
2274
+ "loss": 0.8317,
2275
+ "step": 3550
2276
+ },
2277
+ {
2278
+ "epoch": 5.86,
2279
+ "learning_rate": 0.00012647157190635448,
2280
+ "loss": 0.8322,
2281
+ "step": 3560
2282
+ },
2283
+ {
2284
+ "epoch": 5.87,
2285
+ "learning_rate": 0.00012596989966555183,
2286
+ "loss": 0.8221,
2287
+ "step": 3570
2288
+ },
2289
+ {
2290
+ "epoch": 5.89,
2291
+ "learning_rate": 0.00012546822742474914,
2292
+ "loss": 0.8295,
2293
+ "step": 3580
2294
+ },
2295
+ {
2296
+ "epoch": 5.9,
2297
+ "learning_rate": 0.00012496655518394646,
2298
+ "loss": 0.8411,
2299
+ "step": 3590
2300
+ },
2301
+ {
2302
+ "epoch": 5.92,
2303
+ "learning_rate": 0.0001244648829431438,
2304
+ "loss": 0.8279,
2305
+ "step": 3600
2306
+ },
2307
+ {
2308
+ "epoch": 5.92,
2309
+ "eval_loss": 0.9197421669960022,
2310
+ "eval_runtime": 33.1031,
2311
+ "eval_samples_per_second": 60.417,
2312
+ "eval_steps_per_second": 1.269,
2313
+ "step": 3600
2314
+ },
2315
+ {
2316
+ "epoch": 5.94,
2317
+ "learning_rate": 0.00012396321070234112,
2318
+ "loss": 0.8296,
2319
+ "step": 3610
2320
+ },
2321
+ {
2322
+ "epoch": 5.95,
2323
+ "learning_rate": 0.00012346153846153844,
2324
+ "loss": 0.8278,
2325
+ "step": 3620
2326
+ },
2327
+ {
2328
+ "epoch": 5.97,
2329
+ "learning_rate": 0.00012295986622073578,
2330
+ "loss": 0.8408,
2331
+ "step": 3630
2332
+ },
2333
+ {
2334
+ "epoch": 5.99,
2335
+ "learning_rate": 0.0001224581939799331,
2336
+ "loss": 0.8232,
2337
+ "step": 3640
2338
+ },
2339
+ {
2340
+ "epoch": 6.0,
2341
+ "learning_rate": 0.00012195652173913042,
2342
+ "loss": 0.826,
2343
+ "step": 3650
2344
+ },
2345
+ {
2346
+ "epoch": 6.02,
2347
+ "learning_rate": 0.00012145484949832774,
2348
+ "loss": 0.811,
2349
+ "step": 3660
2350
+ },
2351
+ {
2352
+ "epoch": 6.04,
2353
+ "learning_rate": 0.00012095317725752507,
2354
+ "loss": 0.85,
2355
+ "step": 3670
2356
+ },
2357
+ {
2358
+ "epoch": 6.05,
2359
+ "learning_rate": 0.0001204515050167224,
2360
+ "loss": 0.8395,
2361
+ "step": 3680
2362
+ },
2363
+ {
2364
+ "epoch": 6.07,
2365
+ "learning_rate": 0.00011994983277591972,
2366
+ "loss": 0.8179,
2367
+ "step": 3690
2368
+ },
2369
+ {
2370
+ "epoch": 6.09,
2371
+ "learning_rate": 0.00011944816053511705,
2372
+ "loss": 0.8279,
2373
+ "step": 3700
2374
+ },
2375
+ {
2376
+ "epoch": 6.1,
2377
+ "learning_rate": 0.00011894648829431436,
2378
+ "loss": 0.8227,
2379
+ "step": 3710
2380
+ },
2381
+ {
2382
+ "epoch": 6.12,
2383
+ "learning_rate": 0.00011844481605351169,
2384
+ "loss": 0.8248,
2385
+ "step": 3720
2386
+ },
2387
+ {
2388
+ "epoch": 6.13,
2389
+ "learning_rate": 0.00011794314381270902,
2390
+ "loss": 0.8332,
2391
+ "step": 3730
2392
+ },
2393
+ {
2394
+ "epoch": 6.15,
2395
+ "learning_rate": 0.00011744147157190634,
2396
+ "loss": 0.8204,
2397
+ "step": 3740
2398
+ },
2399
+ {
2400
+ "epoch": 6.17,
2401
+ "learning_rate": 0.00011693979933110367,
2402
+ "loss": 0.8336,
2403
+ "step": 3750
2404
+ },
2405
+ {
2406
+ "epoch": 6.18,
2407
+ "learning_rate": 0.000116438127090301,
2408
+ "loss": 0.813,
2409
+ "step": 3760
2410
+ },
2411
+ {
2412
+ "epoch": 6.2,
2413
+ "learning_rate": 0.00011593645484949831,
2414
+ "loss": 0.8211,
2415
+ "step": 3770
2416
+ },
2417
+ {
2418
+ "epoch": 6.22,
2419
+ "learning_rate": 0.00011543478260869564,
2420
+ "loss": 0.8193,
2421
+ "step": 3780
2422
+ },
2423
+ {
2424
+ "epoch": 6.23,
2425
+ "learning_rate": 0.00011493311036789297,
2426
+ "loss": 0.8425,
2427
+ "step": 3790
2428
+ },
2429
+ {
2430
+ "epoch": 6.25,
2431
+ "learning_rate": 0.00011443143812709029,
2432
+ "loss": 0.8209,
2433
+ "step": 3800
2434
+ },
2435
+ {
2436
+ "epoch": 6.25,
2437
+ "eval_loss": 0.9195341467857361,
2438
+ "eval_runtime": 33.159,
2439
+ "eval_samples_per_second": 60.315,
2440
+ "eval_steps_per_second": 1.267,
2441
+ "step": 3800
2442
+ },
2443
+ {
2444
+ "epoch": 6.27,
2445
+ "learning_rate": 0.00011392976588628762,
2446
+ "loss": 0.8276,
2447
+ "step": 3810
2448
+ },
2449
+ {
2450
+ "epoch": 6.28,
2451
+ "learning_rate": 0.00011342809364548495,
2452
+ "loss": 0.845,
2453
+ "step": 3820
2454
+ },
2455
+ {
2456
+ "epoch": 6.3,
2457
+ "learning_rate": 0.00011292642140468226,
2458
+ "loss": 0.83,
2459
+ "step": 3830
2460
+ },
2461
+ {
2462
+ "epoch": 6.32,
2463
+ "learning_rate": 0.00011242474916387959,
2464
+ "loss": 0.8226,
2465
+ "step": 3840
2466
+ },
2467
+ {
2468
+ "epoch": 6.33,
2469
+ "learning_rate": 0.00011197324414715718,
2470
+ "loss": 0.8329,
2471
+ "step": 3850
2472
+ },
2473
+ {
2474
+ "epoch": 6.35,
2475
+ "learning_rate": 0.0001114715719063545,
2476
+ "loss": 0.8384,
2477
+ "step": 3860
2478
+ },
2479
+ {
2480
+ "epoch": 6.37,
2481
+ "learning_rate": 0.00011096989966555183,
2482
+ "loss": 0.8186,
2483
+ "step": 3870
2484
+ },
2485
+ {
2486
+ "epoch": 6.38,
2487
+ "learning_rate": 0.00011046822742474914,
2488
+ "loss": 0.8027,
2489
+ "step": 3880
2490
+ },
2491
+ {
2492
+ "epoch": 6.4,
2493
+ "learning_rate": 0.00010996655518394647,
2494
+ "loss": 0.8214,
2495
+ "step": 3890
2496
+ },
2497
+ {
2498
+ "epoch": 6.41,
2499
+ "learning_rate": 0.0001094648829431438,
2500
+ "loss": 0.8066,
2501
+ "step": 3900
2502
+ },
2503
+ {
2504
+ "epoch": 6.43,
2505
+ "learning_rate": 0.00010896321070234112,
2506
+ "loss": 0.8303,
2507
+ "step": 3910
2508
+ },
2509
+ {
2510
+ "epoch": 6.45,
2511
+ "learning_rate": 0.00010846153846153845,
2512
+ "loss": 0.8342,
2513
+ "step": 3920
2514
+ },
2515
+ {
2516
+ "epoch": 6.46,
2517
+ "learning_rate": 0.00010795986622073578,
2518
+ "loss": 0.8304,
2519
+ "step": 3930
2520
+ },
2521
+ {
2522
+ "epoch": 6.48,
2523
+ "learning_rate": 0.0001074581939799331,
2524
+ "loss": 0.8279,
2525
+ "step": 3940
2526
+ },
2527
+ {
2528
+ "epoch": 6.5,
2529
+ "learning_rate": 0.00010695652173913043,
2530
+ "loss": 0.8104,
2531
+ "step": 3950
2532
+ },
2533
+ {
2534
+ "epoch": 6.51,
2535
+ "learning_rate": 0.00010645484949832776,
2536
+ "loss": 0.8372,
2537
+ "step": 3960
2538
+ },
2539
+ {
2540
+ "epoch": 6.53,
2541
+ "learning_rate": 0.00010595317725752507,
2542
+ "loss": 0.8239,
2543
+ "step": 3970
2544
+ },
2545
+ {
2546
+ "epoch": 6.55,
2547
+ "learning_rate": 0.0001054515050167224,
2548
+ "loss": 0.8404,
2549
+ "step": 3980
2550
+ },
2551
+ {
2552
+ "epoch": 6.56,
2553
+ "learning_rate": 0.00010494983277591972,
2554
+ "loss": 0.8206,
2555
+ "step": 3990
2556
+ },
2557
+ {
2558
+ "epoch": 6.58,
2559
+ "learning_rate": 0.00010444816053511705,
2560
+ "loss": 0.8174,
2561
+ "step": 4000
2562
+ },
2563
+ {
2564
+ "epoch": 6.58,
2565
+ "eval_loss": 0.9175052642822266,
2566
+ "eval_runtime": 33.1942,
2567
+ "eval_samples_per_second": 60.251,
2568
+ "eval_steps_per_second": 1.265,
2569
+ "step": 4000
2570
+ },
2571
+ {
2572
+ "epoch": 6.6,
2573
+ "learning_rate": 0.00010394648829431438,
2574
+ "loss": 0.8335,
2575
+ "step": 4010
2576
+ },
2577
+ {
2578
+ "epoch": 6.61,
2579
+ "learning_rate": 0.00010344481605351169,
2580
+ "loss": 0.84,
2581
+ "step": 4020
2582
+ },
2583
+ {
2584
+ "epoch": 6.63,
2585
+ "learning_rate": 0.00010294314381270902,
2586
+ "loss": 0.815,
2587
+ "step": 4030
2588
+ },
2589
+ {
2590
+ "epoch": 6.64,
2591
+ "learning_rate": 0.00010244147157190635,
2592
+ "loss": 0.8112,
2593
+ "step": 4040
2594
+ },
2595
+ {
2596
+ "epoch": 6.66,
2597
+ "learning_rate": 0.00010193979933110367,
2598
+ "loss": 0.8302,
2599
+ "step": 4050
2600
+ },
2601
+ {
2602
+ "epoch": 6.68,
2603
+ "learning_rate": 0.000101438127090301,
2604
+ "loss": 0.8284,
2605
+ "step": 4060
2606
+ },
2607
+ {
2608
+ "epoch": 6.69,
2609
+ "learning_rate": 0.00010093645484949833,
2610
+ "loss": 0.823,
2611
+ "step": 4070
2612
+ },
2613
+ {
2614
+ "epoch": 6.71,
2615
+ "learning_rate": 0.00010043478260869564,
2616
+ "loss": 0.8192,
2617
+ "step": 4080
2618
+ },
2619
+ {
2620
+ "epoch": 6.73,
2621
+ "learning_rate": 9.993311036789297e-05,
2622
+ "loss": 0.8285,
2623
+ "step": 4090
2624
+ },
2625
+ {
2626
+ "epoch": 6.74,
2627
+ "learning_rate": 9.94314381270903e-05,
2628
+ "loss": 0.8222,
2629
+ "step": 4100
2630
+ },
2631
+ {
2632
+ "epoch": 6.76,
2633
+ "learning_rate": 9.892976588628762e-05,
2634
+ "loss": 0.8273,
2635
+ "step": 4110
2636
+ },
2637
+ {
2638
+ "epoch": 6.78,
2639
+ "learning_rate": 9.842809364548495e-05,
2640
+ "loss": 0.8145,
2641
+ "step": 4120
2642
+ },
2643
+ {
2644
+ "epoch": 6.79,
2645
+ "learning_rate": 9.792642140468226e-05,
2646
+ "loss": 0.8279,
2647
+ "step": 4130
2648
+ },
2649
+ {
2650
+ "epoch": 6.81,
2651
+ "learning_rate": 9.74247491638796e-05,
2652
+ "loss": 0.8193,
2653
+ "step": 4140
2654
+ },
2655
+ {
2656
+ "epoch": 6.83,
2657
+ "learning_rate": 9.692307692307692e-05,
2658
+ "loss": 0.8243,
2659
+ "step": 4150
2660
+ },
2661
+ {
2662
+ "epoch": 6.84,
2663
+ "learning_rate": 9.642140468227424e-05,
2664
+ "loss": 0.825,
2665
+ "step": 4160
2666
+ },
2667
+ {
2668
+ "epoch": 6.86,
2669
+ "learning_rate": 9.591973244147157e-05,
2670
+ "loss": 0.8185,
2671
+ "step": 4170
2672
+ },
2673
+ {
2674
+ "epoch": 6.88,
2675
+ "learning_rate": 9.54180602006689e-05,
2676
+ "loss": 0.8358,
2677
+ "step": 4180
2678
+ },
2679
+ {
2680
+ "epoch": 6.89,
2681
+ "learning_rate": 9.491638795986622e-05,
2682
+ "loss": 0.8164,
2683
+ "step": 4190
2684
+ },
2685
+ {
2686
+ "epoch": 6.91,
2687
+ "learning_rate": 9.441471571906355e-05,
2688
+ "loss": 0.833,
2689
+ "step": 4200
2690
+ },
2691
+ {
2692
+ "epoch": 6.91,
2693
+ "eval_loss": 0.9155363440513611,
2694
+ "eval_runtime": 33.1004,
2695
+ "eval_samples_per_second": 60.422,
2696
+ "eval_steps_per_second": 1.269,
2697
+ "step": 4200
2698
+ },
2699
+ {
2700
+ "epoch": 6.92,
2701
+ "learning_rate": 9.391304347826087e-05,
2702
+ "loss": 0.8186,
2703
+ "step": 4210
2704
+ },
2705
+ {
2706
+ "epoch": 6.94,
2707
+ "learning_rate": 9.341137123745819e-05,
2708
+ "loss": 0.8211,
2709
+ "step": 4220
2710
+ },
2711
+ {
2712
+ "epoch": 6.96,
2713
+ "learning_rate": 9.290969899665552e-05,
2714
+ "loss": 0.8216,
2715
+ "step": 4230
2716
+ },
2717
+ {
2718
+ "epoch": 6.97,
2719
+ "learning_rate": 9.240802675585282e-05,
2720
+ "loss": 0.7982,
2721
+ "step": 4240
2722
+ },
2723
+ {
2724
+ "epoch": 6.99,
2725
+ "learning_rate": 9.190635451505015e-05,
2726
+ "loss": 0.8025,
2727
+ "step": 4250
2728
+ },
2729
+ {
2730
+ "epoch": 7.01,
2731
+ "learning_rate": 9.140468227424748e-05,
2732
+ "loss": 0.814,
2733
+ "step": 4260
2734
+ },
2735
+ {
2736
+ "epoch": 7.02,
2737
+ "learning_rate": 9.09030100334448e-05,
2738
+ "loss": 0.8048,
2739
+ "step": 4270
2740
+ },
2741
+ {
2742
+ "epoch": 7.04,
2743
+ "learning_rate": 9.040133779264213e-05,
2744
+ "loss": 0.8279,
2745
+ "step": 4280
2746
+ },
2747
+ {
2748
+ "epoch": 7.06,
2749
+ "learning_rate": 8.989966555183944e-05,
2750
+ "loss": 0.8208,
2751
+ "step": 4290
2752
+ },
2753
+ {
2754
+ "epoch": 7.07,
2755
+ "learning_rate": 8.939799331103677e-05,
2756
+ "loss": 0.8009,
2757
+ "step": 4300
2758
+ },
2759
+ {
2760
+ "epoch": 7.09,
2761
+ "learning_rate": 8.88963210702341e-05,
2762
+ "loss": 0.8151,
2763
+ "step": 4310
2764
+ },
2765
+ {
2766
+ "epoch": 7.11,
2767
+ "learning_rate": 8.839464882943142e-05,
2768
+ "loss": 0.8066,
2769
+ "step": 4320
2770
+ },
2771
+ {
2772
+ "epoch": 7.12,
2773
+ "learning_rate": 8.789297658862875e-05,
2774
+ "loss": 0.8239,
2775
+ "step": 4330
2776
+ },
2777
+ {
2778
+ "epoch": 7.14,
2779
+ "learning_rate": 8.739130434782608e-05,
2780
+ "loss": 0.8055,
2781
+ "step": 4340
2782
+ },
2783
+ {
2784
+ "epoch": 7.15,
2785
+ "learning_rate": 8.68896321070234e-05,
2786
+ "loss": 0.816,
2787
+ "step": 4350
2788
+ },
2789
+ {
2790
+ "epoch": 7.17,
2791
+ "learning_rate": 8.638795986622073e-05,
2792
+ "loss": 0.815,
2793
+ "step": 4360
2794
+ },
2795
+ {
2796
+ "epoch": 7.19,
2797
+ "learning_rate": 8.588628762541805e-05,
2798
+ "loss": 0.7989,
2799
+ "step": 4370
2800
+ },
2801
+ {
2802
+ "epoch": 7.2,
2803
+ "learning_rate": 8.538461538461537e-05,
2804
+ "loss": 0.7924,
2805
+ "step": 4380
2806
+ },
2807
+ {
2808
+ "epoch": 7.22,
2809
+ "learning_rate": 8.48829431438127e-05,
2810
+ "loss": 0.8151,
2811
+ "step": 4390
2812
+ },
2813
+ {
2814
+ "epoch": 7.24,
2815
+ "learning_rate": 8.438127090301002e-05,
2816
+ "loss": 0.82,
2817
+ "step": 4400
2818
+ },
2819
+ {
2820
+ "epoch": 7.24,
2821
+ "eval_loss": 0.9152849316596985,
2822
+ "eval_runtime": 33.2483,
2823
+ "eval_samples_per_second": 60.153,
2824
+ "eval_steps_per_second": 1.263,
2825
+ "step": 4400
2826
+ },
2827
+ {
2828
+ "epoch": 7.25,
2829
+ "learning_rate": 8.387959866220735e-05,
2830
+ "loss": 0.8091,
2831
+ "step": 4410
2832
+ },
2833
+ {
2834
+ "epoch": 7.27,
2835
+ "learning_rate": 8.337792642140468e-05,
2836
+ "loss": 0.8195,
2837
+ "step": 4420
2838
+ },
2839
+ {
2840
+ "epoch": 7.29,
2841
+ "learning_rate": 8.287625418060199e-05,
2842
+ "loss": 0.7982,
2843
+ "step": 4430
2844
+ },
2845
+ {
2846
+ "epoch": 7.3,
2847
+ "learning_rate": 8.237458193979932e-05,
2848
+ "loss": 0.8106,
2849
+ "step": 4440
2850
+ },
2851
+ {
2852
+ "epoch": 7.32,
2853
+ "learning_rate": 8.187290969899665e-05,
2854
+ "loss": 0.811,
2855
+ "step": 4450
2856
+ },
2857
+ {
2858
+ "epoch": 7.34,
2859
+ "learning_rate": 8.137123745819397e-05,
2860
+ "loss": 0.8213,
2861
+ "step": 4460
2862
+ },
2863
+ {
2864
+ "epoch": 7.35,
2865
+ "learning_rate": 8.08695652173913e-05,
2866
+ "loss": 0.8143,
2867
+ "step": 4470
2868
+ },
2869
+ {
2870
+ "epoch": 7.37,
2871
+ "learning_rate": 8.036789297658863e-05,
2872
+ "loss": 0.8151,
2873
+ "step": 4480
2874
+ },
2875
+ {
2876
+ "epoch": 7.38,
2877
+ "learning_rate": 7.986622073578594e-05,
2878
+ "loss": 0.8208,
2879
+ "step": 4490
2880
+ },
2881
+ {
2882
+ "epoch": 7.4,
2883
+ "learning_rate": 7.936454849498327e-05,
2884
+ "loss": 0.8198,
2885
+ "step": 4500
2886
+ },
2887
+ {
2888
+ "epoch": 7.42,
2889
+ "learning_rate": 7.88628762541806e-05,
2890
+ "loss": 0.8219,
2891
+ "step": 4510
2892
+ },
2893
+ {
2894
+ "epoch": 7.43,
2895
+ "learning_rate": 7.836120401337792e-05,
2896
+ "loss": 0.7986,
2897
+ "step": 4520
2898
+ },
2899
+ {
2900
+ "epoch": 7.45,
2901
+ "learning_rate": 7.785953177257525e-05,
2902
+ "loss": 0.8054,
2903
+ "step": 4530
2904
+ },
2905
+ {
2906
+ "epoch": 7.47,
2907
+ "learning_rate": 7.735785953177256e-05,
2908
+ "loss": 0.8138,
2909
+ "step": 4540
2910
+ },
2911
+ {
2912
+ "epoch": 7.48,
2913
+ "learning_rate": 7.68561872909699e-05,
2914
+ "loss": 0.8009,
2915
+ "step": 4550
2916
+ },
2917
+ {
2918
+ "epoch": 7.5,
2919
+ "learning_rate": 7.635451505016722e-05,
2920
+ "loss": 0.8142,
2921
+ "step": 4560
2922
+ },
2923
+ {
2924
+ "epoch": 7.52,
2925
+ "learning_rate": 7.585284280936454e-05,
2926
+ "loss": 0.8348,
2927
+ "step": 4570
2928
+ },
2929
+ {
2930
+ "epoch": 7.53,
2931
+ "learning_rate": 7.535117056856187e-05,
2932
+ "loss": 0.8042,
2933
+ "step": 4580
2934
+ },
2935
+ {
2936
+ "epoch": 7.55,
2937
+ "learning_rate": 7.48494983277592e-05,
2938
+ "loss": 0.819,
2939
+ "step": 4590
2940
+ },
2941
+ {
2942
+ "epoch": 7.57,
2943
+ "learning_rate": 7.434782608695652e-05,
2944
+ "loss": 0.8027,
2945
+ "step": 4600
2946
+ },
2947
+ {
2948
+ "epoch": 7.57,
2949
+ "eval_loss": 0.914962887763977,
2950
+ "eval_runtime": 33.1103,
2951
+ "eval_samples_per_second": 60.404,
2952
+ "eval_steps_per_second": 1.268,
2953
+ "step": 4600
2954
+ },
2955
+ {
2956
+ "epoch": 7.58,
2957
+ "learning_rate": 7.384615384615384e-05,
2958
+ "loss": 0.7976,
2959
+ "step": 4610
2960
+ },
2961
+ {
2962
+ "epoch": 7.6,
2963
+ "learning_rate": 7.334448160535116e-05,
2964
+ "loss": 0.8168,
2965
+ "step": 4620
2966
+ },
2967
+ {
2968
+ "epoch": 7.62,
2969
+ "learning_rate": 7.284280936454849e-05,
2970
+ "loss": 0.7952,
2971
+ "step": 4630
2972
+ },
2973
+ {
2974
+ "epoch": 7.63,
2975
+ "learning_rate": 7.234113712374581e-05,
2976
+ "loss": 0.8238,
2977
+ "step": 4640
2978
+ },
2979
+ {
2980
+ "epoch": 7.65,
2981
+ "learning_rate": 7.183946488294314e-05,
2982
+ "loss": 0.8196,
2983
+ "step": 4650
2984
+ },
2985
+ {
2986
+ "epoch": 7.66,
2987
+ "learning_rate": 7.133779264214047e-05,
2988
+ "loss": 0.8032,
2989
+ "step": 4660
2990
+ },
2991
+ {
2992
+ "epoch": 7.68,
2993
+ "learning_rate": 7.083612040133778e-05,
2994
+ "loss": 0.8067,
2995
+ "step": 4670
2996
+ },
2997
+ {
2998
+ "epoch": 7.7,
2999
+ "learning_rate": 7.033444816053511e-05,
3000
+ "loss": 0.8167,
3001
+ "step": 4680
3002
+ },
3003
+ {
3004
+ "epoch": 7.71,
3005
+ "learning_rate": 6.983277591973243e-05,
3006
+ "loss": 0.7856,
3007
+ "step": 4690
3008
+ },
3009
+ {
3010
+ "epoch": 7.73,
3011
+ "learning_rate": 6.933110367892976e-05,
3012
+ "loss": 0.8234,
3013
+ "step": 4700
3014
+ },
3015
+ {
3016
+ "epoch": 7.75,
3017
+ "learning_rate": 6.882943143812709e-05,
3018
+ "loss": 0.805,
3019
+ "step": 4710
3020
+ },
3021
+ {
3022
+ "epoch": 7.76,
3023
+ "learning_rate": 6.83277591973244e-05,
3024
+ "loss": 0.8065,
3025
+ "step": 4720
3026
+ },
3027
+ {
3028
+ "epoch": 7.78,
3029
+ "learning_rate": 6.782608695652173e-05,
3030
+ "loss": 0.8084,
3031
+ "step": 4730
3032
+ },
3033
+ {
3034
+ "epoch": 7.8,
3035
+ "learning_rate": 6.732441471571906e-05,
3036
+ "loss": 0.8105,
3037
+ "step": 4740
3038
+ },
3039
+ {
3040
+ "epoch": 7.81,
3041
+ "learning_rate": 6.682274247491638e-05,
3042
+ "loss": 0.8065,
3043
+ "step": 4750
3044
+ },
3045
+ {
3046
+ "epoch": 7.83,
3047
+ "learning_rate": 6.632107023411371e-05,
3048
+ "loss": 0.8007,
3049
+ "step": 4760
3050
+ },
3051
+ {
3052
+ "epoch": 7.85,
3053
+ "learning_rate": 6.581939799331104e-05,
3054
+ "loss": 0.8238,
3055
+ "step": 4770
3056
+ },
3057
+ {
3058
+ "epoch": 7.86,
3059
+ "learning_rate": 6.531772575250835e-05,
3060
+ "loss": 0.8083,
3061
+ "step": 4780
3062
+ },
3063
+ {
3064
+ "epoch": 7.88,
3065
+ "learning_rate": 6.481605351170568e-05,
3066
+ "loss": 0.81,
3067
+ "step": 4790
3068
+ },
3069
+ {
3070
+ "epoch": 7.89,
3071
+ "learning_rate": 6.4314381270903e-05,
3072
+ "loss": 0.7963,
3073
+ "step": 4800
3074
+ },
3075
+ {
3076
+ "epoch": 7.89,
3077
+ "eval_loss": 0.9136397838592529,
3078
+ "eval_runtime": 33.1953,
3079
+ "eval_samples_per_second": 60.249,
3080
+ "eval_steps_per_second": 1.265,
3081
+ "step": 4800
3082
+ },
3083
+ {
3084
+ "epoch": 7.91,
3085
+ "learning_rate": 6.381270903010033e-05,
3086
+ "loss": 0.7906,
3087
+ "step": 4810
3088
+ },
3089
+ {
3090
+ "epoch": 7.93,
3091
+ "learning_rate": 6.331103678929766e-05,
3092
+ "loss": 0.8132,
3093
+ "step": 4820
3094
+ },
3095
+ {
3096
+ "epoch": 7.94,
3097
+ "learning_rate": 6.280936454849498e-05,
3098
+ "loss": 0.8175,
3099
+ "step": 4830
3100
+ },
3101
+ {
3102
+ "epoch": 7.96,
3103
+ "learning_rate": 6.23076923076923e-05,
3104
+ "loss": 0.801,
3105
+ "step": 4840
3106
+ },
3107
+ {
3108
+ "epoch": 7.98,
3109
+ "learning_rate": 6.180602006688964e-05,
3110
+ "loss": 0.8117,
3111
+ "step": 4850
3112
+ },
3113
+ {
3114
+ "epoch": 7.99,
3115
+ "learning_rate": 6.130434782608695e-05,
3116
+ "loss": 0.8019,
3117
+ "step": 4860
3118
+ },
3119
+ {
3120
+ "epoch": 8.01,
3121
+ "learning_rate": 6.0802675585284274e-05,
3122
+ "loss": 0.802,
3123
+ "step": 4870
3124
+ },
3125
+ {
3126
+ "epoch": 8.03,
3127
+ "learning_rate": 6.03010033444816e-05,
3128
+ "loss": 0.821,
3129
+ "step": 4880
3130
+ },
3131
+ {
3132
+ "epoch": 8.04,
3133
+ "learning_rate": 5.979933110367892e-05,
3134
+ "loss": 0.7962,
3135
+ "step": 4890
3136
+ },
3137
+ {
3138
+ "epoch": 8.06,
3139
+ "learning_rate": 5.929765886287625e-05,
3140
+ "loss": 0.7992,
3141
+ "step": 4900
3142
+ },
3143
+ {
3144
+ "epoch": 8.08,
3145
+ "learning_rate": 5.879598662207357e-05,
3146
+ "loss": 0.7998,
3147
+ "step": 4910
3148
+ },
3149
+ {
3150
+ "epoch": 8.09,
3151
+ "learning_rate": 5.8294314381270895e-05,
3152
+ "loss": 0.7934,
3153
+ "step": 4920
3154
+ },
3155
+ {
3156
+ "epoch": 8.11,
3157
+ "learning_rate": 5.779264214046822e-05,
3158
+ "loss": 0.801,
3159
+ "step": 4930
3160
+ },
3161
+ {
3162
+ "epoch": 8.12,
3163
+ "learning_rate": 5.729096989966555e-05,
3164
+ "loss": 0.8039,
3165
+ "step": 4940
3166
+ },
3167
+ {
3168
+ "epoch": 8.14,
3169
+ "learning_rate": 5.678929765886287e-05,
3170
+ "loss": 0.802,
3171
+ "step": 4950
3172
+ },
3173
+ {
3174
+ "epoch": 8.16,
3175
+ "learning_rate": 5.6287625418060194e-05,
3176
+ "loss": 0.781,
3177
+ "step": 4960
3178
+ },
3179
+ {
3180
+ "epoch": 8.17,
3181
+ "learning_rate": 5.5785953177257517e-05,
3182
+ "loss": 0.7938,
3183
+ "step": 4970
3184
+ },
3185
+ {
3186
+ "epoch": 8.19,
3187
+ "learning_rate": 5.5284280936454846e-05,
3188
+ "loss": 0.8031,
3189
+ "step": 4980
3190
+ },
3191
+ {
3192
+ "epoch": 8.21,
3193
+ "learning_rate": 5.478260869565217e-05,
3194
+ "loss": 0.7999,
3195
+ "step": 4990
3196
+ },
3197
+ {
3198
+ "epoch": 8.22,
3199
+ "learning_rate": 5.428093645484949e-05,
3200
+ "loss": 0.7985,
3201
+ "step": 5000
3202
+ },
3203
+ {
3204
+ "epoch": 8.22,
3205
+ "eval_loss": 0.9126840829849243,
3206
+ "eval_runtime": 33.2306,
3207
+ "eval_samples_per_second": 60.185,
3208
+ "eval_steps_per_second": 1.264,
3209
+ "step": 5000
3210
+ },
3211
+ {
3212
+ "epoch": 8.24,
3213
+ "learning_rate": 5.377926421404682e-05,
3214
+ "loss": 0.8056,
3215
+ "step": 5010
3216
+ },
3217
+ {
3218
+ "epoch": 8.26,
3219
+ "learning_rate": 5.3277591973244145e-05,
3220
+ "loss": 0.8099,
3221
+ "step": 5020
3222
+ },
3223
+ {
3224
+ "epoch": 8.27,
3225
+ "learning_rate": 5.277591973244147e-05,
3226
+ "loss": 0.7982,
3227
+ "step": 5030
3228
+ },
3229
+ {
3230
+ "epoch": 8.29,
3231
+ "learning_rate": 5.227424749163879e-05,
3232
+ "loss": 0.8021,
3233
+ "step": 5040
3234
+ },
3235
+ {
3236
+ "epoch": 8.31,
3237
+ "learning_rate": 5.177257525083612e-05,
3238
+ "loss": 0.8044,
3239
+ "step": 5050
3240
+ },
3241
+ {
3242
+ "epoch": 8.32,
3243
+ "learning_rate": 5.127090301003344e-05,
3244
+ "loss": 0.8099,
3245
+ "step": 5060
3246
+ },
3247
+ {
3248
+ "epoch": 8.34,
3249
+ "learning_rate": 5.0769230769230766e-05,
3250
+ "loss": 0.818,
3251
+ "step": 5070
3252
+ },
3253
+ {
3254
+ "epoch": 8.36,
3255
+ "learning_rate": 5.0267558528428095e-05,
3256
+ "loss": 0.796,
3257
+ "step": 5080
3258
+ },
3259
+ {
3260
+ "epoch": 8.37,
3261
+ "learning_rate": 4.976588628762542e-05,
3262
+ "loss": 0.8102,
3263
+ "step": 5090
3264
+ },
3265
+ {
3266
+ "epoch": 8.39,
3267
+ "learning_rate": 4.926421404682274e-05,
3268
+ "loss": 0.8092,
3269
+ "step": 5100
3270
+ },
3271
+ {
3272
+ "epoch": 8.4,
3273
+ "learning_rate": 4.8762541806020064e-05,
3274
+ "loss": 0.8064,
3275
+ "step": 5110
3276
+ },
3277
+ {
3278
+ "epoch": 8.42,
3279
+ "learning_rate": 4.826086956521738e-05,
3280
+ "loss": 0.8171,
3281
+ "step": 5120
3282
+ },
3283
+ {
3284
+ "epoch": 8.44,
3285
+ "learning_rate": 4.775919732441471e-05,
3286
+ "loss": 0.8077,
3287
+ "step": 5130
3288
+ },
3289
+ {
3290
+ "epoch": 8.45,
3291
+ "learning_rate": 4.725752508361203e-05,
3292
+ "loss": 0.8084,
3293
+ "step": 5140
3294
+ },
3295
+ {
3296
+ "epoch": 8.47,
3297
+ "learning_rate": 4.6755852842809356e-05,
3298
+ "loss": 0.8081,
3299
+ "step": 5150
3300
+ },
3301
+ {
3302
+ "epoch": 8.49,
3303
+ "learning_rate": 4.6254180602006685e-05,
3304
+ "loss": 0.7926,
3305
+ "step": 5160
3306
+ },
3307
+ {
3308
+ "epoch": 8.5,
3309
+ "learning_rate": 4.575250836120401e-05,
3310
+ "loss": 0.7938,
3311
+ "step": 5170
3312
+ },
3313
+ {
3314
+ "epoch": 8.52,
3315
+ "learning_rate": 4.525083612040133e-05,
3316
+ "loss": 0.8196,
3317
+ "step": 5180
3318
+ },
3319
+ {
3320
+ "epoch": 8.54,
3321
+ "learning_rate": 4.4749163879598654e-05,
3322
+ "loss": 0.8059,
3323
+ "step": 5190
3324
+ },
3325
+ {
3326
+ "epoch": 8.55,
3327
+ "learning_rate": 4.4247491638795984e-05,
3328
+ "loss": 0.7997,
3329
+ "step": 5200
3330
+ },
3331
+ {
3332
+ "epoch": 8.55,
3333
+ "eval_loss": 0.9122925996780396,
3334
+ "eval_runtime": 33.1494,
3335
+ "eval_samples_per_second": 60.333,
3336
+ "eval_steps_per_second": 1.267,
3337
+ "step": 5200
3338
+ },
3339
+ {
3340
+ "epoch": 8.57,
3341
+ "learning_rate": 4.374581939799331e-05,
3342
+ "loss": 0.79,
3343
+ "step": 5210
3344
+ },
3345
+ {
3346
+ "epoch": 8.59,
3347
+ "learning_rate": 4.324414715719063e-05,
3348
+ "loss": 0.8103,
3349
+ "step": 5220
3350
+ },
3351
+ {
3352
+ "epoch": 8.6,
3353
+ "learning_rate": 4.274247491638796e-05,
3354
+ "loss": 0.7972,
3355
+ "step": 5230
3356
+ },
3357
+ {
3358
+ "epoch": 8.62,
3359
+ "learning_rate": 4.224080267558528e-05,
3360
+ "loss": 0.8117,
3361
+ "step": 5240
3362
+ },
3363
+ {
3364
+ "epoch": 8.63,
3365
+ "learning_rate": 4.1739130434782605e-05,
3366
+ "loss": 0.8058,
3367
+ "step": 5250
3368
+ },
3369
+ {
3370
+ "epoch": 8.65,
3371
+ "learning_rate": 4.123745819397993e-05,
3372
+ "loss": 0.8043,
3373
+ "step": 5260
3374
+ },
3375
+ {
3376
+ "epoch": 8.67,
3377
+ "learning_rate": 4.073578595317726e-05,
3378
+ "loss": 0.7986,
3379
+ "step": 5270
3380
+ },
3381
+ {
3382
+ "epoch": 8.68,
3383
+ "learning_rate": 4.023411371237458e-05,
3384
+ "loss": 0.7967,
3385
+ "step": 5280
3386
+ },
3387
+ {
3388
+ "epoch": 8.7,
3389
+ "learning_rate": 3.9732441471571903e-05,
3390
+ "loss": 0.8072,
3391
+ "step": 5290
3392
+ },
3393
+ {
3394
+ "epoch": 8.72,
3395
+ "learning_rate": 3.9230769230769226e-05,
3396
+ "loss": 0.8061,
3397
+ "step": 5300
3398
+ },
3399
+ {
3400
+ "epoch": 8.73,
3401
+ "learning_rate": 3.8729096989966556e-05,
3402
+ "loss": 0.8101,
3403
+ "step": 5310
3404
+ },
3405
+ {
3406
+ "epoch": 8.75,
3407
+ "learning_rate": 3.822742474916388e-05,
3408
+ "loss": 0.7987,
3409
+ "step": 5320
3410
+ },
3411
+ {
3412
+ "epoch": 8.77,
3413
+ "learning_rate": 3.77257525083612e-05,
3414
+ "loss": 0.7971,
3415
+ "step": 5330
3416
+ },
3417
+ {
3418
+ "epoch": 8.78,
3419
+ "learning_rate": 3.7224080267558525e-05,
3420
+ "loss": 0.7826,
3421
+ "step": 5340
3422
+ },
3423
+ {
3424
+ "epoch": 8.8,
3425
+ "learning_rate": 3.672240802675585e-05,
3426
+ "loss": 0.7956,
3427
+ "step": 5350
3428
+ },
3429
+ {
3430
+ "epoch": 8.82,
3431
+ "learning_rate": 3.622073578595318e-05,
3432
+ "loss": 0.8029,
3433
+ "step": 5360
3434
+ },
3435
+ {
3436
+ "epoch": 8.83,
3437
+ "learning_rate": 3.57190635451505e-05,
3438
+ "loss": 0.8041,
3439
+ "step": 5370
3440
+ },
3441
+ {
3442
+ "epoch": 8.85,
3443
+ "learning_rate": 3.521739130434782e-05,
3444
+ "loss": 0.7844,
3445
+ "step": 5380
3446
+ },
3447
+ {
3448
+ "epoch": 8.87,
3449
+ "learning_rate": 3.4715719063545146e-05,
3450
+ "loss": 0.7914,
3451
+ "step": 5390
3452
+ },
3453
+ {
3454
+ "epoch": 8.88,
3455
+ "learning_rate": 3.4214046822742476e-05,
3456
+ "loss": 0.8197,
3457
+ "step": 5400
3458
+ },
3459
+ {
3460
+ "epoch": 8.88,
3461
+ "eval_loss": 0.9114879369735718,
3462
+ "eval_runtime": 33.1591,
3463
+ "eval_samples_per_second": 60.315,
3464
+ "eval_steps_per_second": 1.267,
3465
+ "step": 5400
3466
+ },
3467
+ {
3468
+ "epoch": 8.9,
3469
+ "learning_rate": 3.37123745819398e-05,
3470
+ "loss": 0.8012,
3471
+ "step": 5410
3472
+ },
3473
+ {
3474
+ "epoch": 8.91,
3475
+ "learning_rate": 3.321070234113712e-05,
3476
+ "loss": 0.8069,
3477
+ "step": 5420
3478
+ },
3479
+ {
3480
+ "epoch": 8.93,
3481
+ "learning_rate": 3.2709030100334444e-05,
3482
+ "loss": 0.7984,
3483
+ "step": 5430
3484
+ },
3485
+ {
3486
+ "epoch": 8.95,
3487
+ "learning_rate": 3.220735785953177e-05,
3488
+ "loss": 0.7877,
3489
+ "step": 5440
3490
+ },
3491
+ {
3492
+ "epoch": 8.96,
3493
+ "learning_rate": 3.170568561872909e-05,
3494
+ "loss": 0.7964,
3495
+ "step": 5450
3496
+ },
3497
+ {
3498
+ "epoch": 8.98,
3499
+ "learning_rate": 3.120401337792642e-05,
3500
+ "loss": 0.8064,
3501
+ "step": 5460
3502
+ },
3503
+ {
3504
+ "epoch": 9.0,
3505
+ "learning_rate": 3.070234113712374e-05,
3506
+ "loss": 0.8088,
3507
+ "step": 5470
3508
+ },
3509
+ {
3510
+ "epoch": 9.01,
3511
+ "learning_rate": 3.020066889632107e-05,
3512
+ "loss": 0.7983,
3513
+ "step": 5480
3514
+ },
3515
+ {
3516
+ "epoch": 9.03,
3517
+ "learning_rate": 2.9698996655518392e-05,
3518
+ "loss": 0.7924,
3519
+ "step": 5490
3520
+ },
3521
+ {
3522
+ "epoch": 9.05,
3523
+ "learning_rate": 2.9197324414715718e-05,
3524
+ "loss": 0.7847,
3525
+ "step": 5500
3526
+ },
3527
+ {
3528
+ "epoch": 9.06,
3529
+ "learning_rate": 2.869565217391304e-05,
3530
+ "loss": 0.8061,
3531
+ "step": 5510
3532
+ },
3533
+ {
3534
+ "epoch": 9.08,
3535
+ "learning_rate": 2.8193979933110367e-05,
3536
+ "loss": 0.8038,
3537
+ "step": 5520
3538
+ },
3539
+ {
3540
+ "epoch": 9.1,
3541
+ "learning_rate": 2.769230769230769e-05,
3542
+ "loss": 0.7913,
3543
+ "step": 5530
3544
+ },
3545
+ {
3546
+ "epoch": 9.11,
3547
+ "learning_rate": 2.7190635451505016e-05,
3548
+ "loss": 0.7976,
3549
+ "step": 5540
3550
+ },
3551
+ {
3552
+ "epoch": 9.13,
3553
+ "learning_rate": 2.6688963210702336e-05,
3554
+ "loss": 0.7996,
3555
+ "step": 5550
3556
+ },
3557
+ {
3558
+ "epoch": 9.14,
3559
+ "learning_rate": 2.6187290969899662e-05,
3560
+ "loss": 0.7963,
3561
+ "step": 5560
3562
+ },
3563
+ {
3564
+ "epoch": 9.16,
3565
+ "learning_rate": 2.5685618729096985e-05,
3566
+ "loss": 0.7987,
3567
+ "step": 5570
3568
+ },
3569
+ {
3570
+ "epoch": 9.18,
3571
+ "learning_rate": 2.518394648829431e-05,
3572
+ "loss": 0.7858,
3573
+ "step": 5580
3574
+ },
3575
+ {
3576
+ "epoch": 9.19,
3577
+ "learning_rate": 2.4682274247491638e-05,
3578
+ "loss": 0.8028,
3579
+ "step": 5590
3580
+ },
3581
+ {
3582
+ "epoch": 9.21,
3583
+ "learning_rate": 2.418060200668896e-05,
3584
+ "loss": 0.7899,
3585
+ "step": 5600
3586
+ },
3587
+ {
3588
+ "epoch": 9.21,
3589
+ "eval_loss": 0.9114179611206055,
3590
+ "eval_runtime": 33.1597,
3591
+ "eval_samples_per_second": 60.314,
3592
+ "eval_steps_per_second": 1.267,
3593
+ "step": 5600
3594
+ },
3595
+ {
3596
+ "epoch": 9.23,
3597
+ "learning_rate": 2.3678929765886287e-05,
3598
+ "loss": 0.7896,
3599
+ "step": 5610
3600
+ },
3601
+ {
3602
+ "epoch": 9.24,
3603
+ "learning_rate": 2.317725752508361e-05,
3604
+ "loss": 0.7928,
3605
+ "step": 5620
3606
+ },
3607
+ {
3608
+ "epoch": 9.26,
3609
+ "learning_rate": 2.2675585284280936e-05,
3610
+ "loss": 0.8108,
3611
+ "step": 5630
3612
+ },
3613
+ {
3614
+ "epoch": 9.28,
3615
+ "learning_rate": 2.217391304347826e-05,
3616
+ "loss": 0.8011,
3617
+ "step": 5640
3618
+ },
3619
+ {
3620
+ "epoch": 9.29,
3621
+ "learning_rate": 2.1672240802675585e-05,
3622
+ "loss": 0.7965,
3623
+ "step": 5650
3624
+ },
3625
+ {
3626
+ "epoch": 9.31,
3627
+ "learning_rate": 2.1170568561872908e-05,
3628
+ "loss": 0.7967,
3629
+ "step": 5660
3630
+ },
3631
+ {
3632
+ "epoch": 9.33,
3633
+ "learning_rate": 2.0668896321070234e-05,
3634
+ "loss": 0.8006,
3635
+ "step": 5670
3636
+ },
3637
+ {
3638
+ "epoch": 9.34,
3639
+ "learning_rate": 2.0167224080267554e-05,
3640
+ "loss": 0.7938,
3641
+ "step": 5680
3642
+ },
3643
+ {
3644
+ "epoch": 9.36,
3645
+ "learning_rate": 1.966555183946488e-05,
3646
+ "loss": 0.8137,
3647
+ "step": 5690
3648
+ },
3649
+ {
3650
+ "epoch": 9.38,
3651
+ "learning_rate": 1.9163879598662203e-05,
3652
+ "loss": 0.7994,
3653
+ "step": 5700
3654
+ },
3655
+ {
3656
+ "epoch": 9.39,
3657
+ "learning_rate": 1.866220735785953e-05,
3658
+ "loss": 0.7994,
3659
+ "step": 5710
3660
+ },
3661
+ {
3662
+ "epoch": 9.41,
3663
+ "learning_rate": 1.8160535117056856e-05,
3664
+ "loss": 0.7926,
3665
+ "step": 5720
3666
+ },
3667
+ {
3668
+ "epoch": 9.42,
3669
+ "learning_rate": 1.765886287625418e-05,
3670
+ "loss": 0.7918,
3671
+ "step": 5730
3672
+ },
3673
+ {
3674
+ "epoch": 9.44,
3675
+ "learning_rate": 1.7157190635451505e-05,
3676
+ "loss": 0.7811,
3677
+ "step": 5740
3678
+ },
3679
+ {
3680
+ "epoch": 9.46,
3681
+ "learning_rate": 1.6655518394648828e-05,
3682
+ "loss": 0.8001,
3683
+ "step": 5750
3684
+ },
3685
+ {
3686
+ "epoch": 9.47,
3687
+ "learning_rate": 1.6153846153846154e-05,
3688
+ "loss": 0.8106,
3689
+ "step": 5760
3690
+ },
3691
+ {
3692
+ "epoch": 9.49,
3693
+ "learning_rate": 1.5652173913043477e-05,
3694
+ "loss": 0.7882,
3695
+ "step": 5770
3696
+ },
3697
+ {
3698
+ "epoch": 9.51,
3699
+ "learning_rate": 1.5150501672240801e-05,
3700
+ "loss": 0.7873,
3701
+ "step": 5780
3702
+ },
3703
+ {
3704
+ "epoch": 9.52,
3705
+ "learning_rate": 1.4648829431438126e-05,
3706
+ "loss": 0.7976,
3707
+ "step": 5790
3708
+ },
3709
+ {
3710
+ "epoch": 9.54,
3711
+ "learning_rate": 1.414715719063545e-05,
3712
+ "loss": 0.7951,
3713
+ "step": 5800
3714
+ },
3715
+ {
3716
+ "epoch": 9.54,
3717
+ "eval_loss": 0.9110632538795471,
3718
+ "eval_runtime": 33.1804,
3719
+ "eval_samples_per_second": 60.276,
3720
+ "eval_steps_per_second": 1.266,
3721
+ "step": 5800
3722
+ },
3723
+ {
3724
+ "epoch": 9.56,
3725
+ "learning_rate": 1.3645484949832775e-05,
3726
+ "loss": 0.7897,
3727
+ "step": 5810
3728
+ },
3729
+ {
3730
+ "epoch": 9.57,
3731
+ "learning_rate": 1.31438127090301e-05,
3732
+ "loss": 0.8037,
3733
+ "step": 5820
3734
+ },
3735
+ {
3736
+ "epoch": 9.59,
3737
+ "learning_rate": 1.2642140468227423e-05,
3738
+ "loss": 0.786,
3739
+ "step": 5830
3740
+ },
3741
+ {
3742
+ "epoch": 9.61,
3743
+ "learning_rate": 1.2140468227424747e-05,
3744
+ "loss": 0.8044,
3745
+ "step": 5840
3746
+ },
3747
+ {
3748
+ "epoch": 9.62,
3749
+ "learning_rate": 1.1638795986622072e-05,
3750
+ "loss": 0.804,
3751
+ "step": 5850
3752
+ },
3753
+ {
3754
+ "epoch": 9.64,
3755
+ "learning_rate": 1.1137123745819396e-05,
3756
+ "loss": 0.8,
3757
+ "step": 5860
3758
+ },
3759
+ {
3760
+ "epoch": 9.65,
3761
+ "learning_rate": 1.0635451505016721e-05,
3762
+ "loss": 0.7957,
3763
+ "step": 5870
3764
+ },
3765
+ {
3766
+ "epoch": 9.67,
3767
+ "learning_rate": 1.0133779264214047e-05,
3768
+ "loss": 0.7875,
3769
+ "step": 5880
3770
+ },
3771
+ {
3772
+ "epoch": 9.69,
3773
+ "learning_rate": 9.632107023411369e-06,
3774
+ "loss": 0.7969,
3775
+ "step": 5890
3776
+ },
3777
+ {
3778
+ "epoch": 9.7,
3779
+ "learning_rate": 9.130434782608695e-06,
3780
+ "loss": 0.805,
3781
+ "step": 5900
3782
+ },
3783
+ {
3784
+ "epoch": 9.72,
3785
+ "learning_rate": 8.62876254180602e-06,
3786
+ "loss": 0.7755,
3787
+ "step": 5910
3788
+ },
3789
+ {
3790
+ "epoch": 9.74,
3791
+ "learning_rate": 8.127090301003344e-06,
3792
+ "loss": 0.7957,
3793
+ "step": 5920
3794
+ },
3795
+ {
3796
+ "epoch": 9.75,
3797
+ "learning_rate": 7.625418060200669e-06,
3798
+ "loss": 0.8089,
3799
+ "step": 5930
3800
+ },
3801
+ {
3802
+ "epoch": 9.77,
3803
+ "learning_rate": 7.123745819397992e-06,
3804
+ "loss": 0.8008,
3805
+ "step": 5940
3806
+ },
3807
+ {
3808
+ "epoch": 9.79,
3809
+ "learning_rate": 6.622073578595317e-06,
3810
+ "loss": 0.7941,
3811
+ "step": 5950
3812
+ },
3813
+ {
3814
+ "epoch": 9.8,
3815
+ "learning_rate": 6.1204013377926415e-06,
3816
+ "loss": 0.7932,
3817
+ "step": 5960
3818
+ },
3819
+ {
3820
+ "epoch": 9.82,
3821
+ "learning_rate": 5.618729096989965e-06,
3822
+ "loss": 0.7993,
3823
+ "step": 5970
3824
+ },
3825
+ {
3826
+ "epoch": 9.84,
3827
+ "learning_rate": 5.117056856187291e-06,
3828
+ "loss": 0.7872,
3829
+ "step": 5980
3830
+ },
3831
+ {
3832
+ "epoch": 9.85,
3833
+ "learning_rate": 4.615384615384615e-06,
3834
+ "loss": 0.8088,
3835
+ "step": 5990
3836
+ },
3837
+ {
3838
+ "epoch": 9.87,
3839
+ "learning_rate": 4.11371237458194e-06,
3840
+ "loss": 0.7947,
3841
+ "step": 6000
3842
+ },
3843
+ {
3844
+ "epoch": 9.87,
3845
+ "eval_loss": 0.9106292724609375,
3846
+ "eval_runtime": 33.2219,
3847
+ "eval_samples_per_second": 60.201,
3848
+ "eval_steps_per_second": 1.264,
3849
+ "step": 6000
3850
+ }
3851
+ ],
3852
+ "logging_steps": 10,
3853
+ "max_steps": 6080,
3854
+ "num_train_epochs": 10,
3855
+ "save_steps": 200,
3856
+ "total_flos": 1.3874338630022988e+19,
3857
+ "trial_name": null,
3858
+ "trial_params": null
3859
+ }
checkpoint-6000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2683d2cc3b2cf2ed5efb5a43c93d20249d9a309d91a7775d0bbfbbf7ac3e5a62
3
+ size 4027