Upload 10 files
Browse files- adapter_config.json +16 -0
- adapter_model.bin +3 -0
- finetuning_args.bin +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1216 -0
- training_args.bin +3 -0
- value_head.bin +3 -0
adapter_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "THUDM/chatglm-6b",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"lora_alpha": 32.0,
|
| 8 |
+
"lora_dropout": 0.1,
|
| 9 |
+
"modules_to_save": null,
|
| 10 |
+
"peft_type": "LORA",
|
| 11 |
+
"r": 8,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"query_key_value"
|
| 14 |
+
],
|
| 15 |
+
"task_type": "CAUSAL_LM"
|
| 16 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7366486228eafbac983bc620bb2c379e453bce6b7fca35eea9c8850b043e5b21
|
| 3 |
+
size 14700825
|
finetuning_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0d21be8c105704d26792b0b5ff588414e9de8cdb4d00194854d7cec1b515a17
|
| 3 |
+
size 895
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:78097a5abdf78bc6c0d55a54521cd4e2026496a7b9b0d5e15ce05178f670bf12
|
| 3 |
+
size 29441605
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7294ba36ef9a4502a2b71a0d9f8f2650dcf0a69d516a18646b22b019bfe028b8
|
| 3 |
+
size 14575
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ad0b79fa97532139bf1145c50ddc1e4fe41ecf69b3c82f92feb3cfc3001fbff
|
| 3 |
+
size 557
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e6f38d2121f8b3fd1609bf4277a0971d8e970c1909856f1f5b65fb4321402316
|
| 3 |
+
size 627
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.8780594885303479,
|
| 5 |
+
"global_step": 2000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 4.999914338886042e-05,
|
| 13 |
+
"loss": 1.2139,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 4.999390875559143e-05,
|
| 19 |
+
"loss": 0.4338,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 4.99839163793675e-05,
|
| 25 |
+
"loss": 0.1646,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.02,
|
| 30 |
+
"learning_rate": 4.997085692392521e-05,
|
| 31 |
+
"loss": 0.1648,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 4.995183082464269e-05,
|
| 37 |
+
"loss": 0.081,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.03,
|
| 42 |
+
"learning_rate": 4.9928054992195985e-05,
|
| 43 |
+
"loss": 0.0475,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03,
|
| 48 |
+
"learning_rate": 4.98995339524598e-05,
|
| 49 |
+
"loss": 0.0529,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.04,
|
| 54 |
+
"learning_rate": 4.986981233301252e-05,
|
| 55 |
+
"loss": 0.1633,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.04,
|
| 60 |
+
"learning_rate": 4.983229110247534e-05,
|
| 61 |
+
"loss": 0.0649,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.04,
|
| 66 |
+
"learning_rate": 4.97900428938891e-05,
|
| 67 |
+
"loss": 0.0791,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.05,
|
| 72 |
+
"learning_rate": 4.974798456824312e-05,
|
| 73 |
+
"loss": 0.1031,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.05,
|
| 78 |
+
"learning_rate": 4.9696778000097424e-05,
|
| 79 |
+
"loss": 0.0449,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.06,
|
| 84 |
+
"learning_rate": 4.964087024964812e-05,
|
| 85 |
+
"loss": 0.0279,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.06,
|
| 90 |
+
"learning_rate": 4.9580271959276445e-05,
|
| 91 |
+
"loss": 0.041,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.07,
|
| 96 |
+
"learning_rate": 4.9514994664236535e-05,
|
| 97 |
+
"loss": 0.0454,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.07,
|
| 102 |
+
"learning_rate": 4.944505079045958e-05,
|
| 103 |
+
"loss": 0.016,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.07,
|
| 108 |
+
"learning_rate": 4.9378122365147536e-05,
|
| 109 |
+
"loss": 0.1362,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.08,
|
| 114 |
+
"learning_rate": 4.929934940465946e-05,
|
| 115 |
+
"loss": 0.0186,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.08,
|
| 120 |
+
"learning_rate": 4.921595091482461e-05,
|
| 121 |
+
"loss": 0.0381,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.09,
|
| 126 |
+
"learning_rate": 4.913695054926396e-05,
|
| 127 |
+
"loss": 0.0952,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.09,
|
| 132 |
+
"learning_rate": 4.904480801572673e-05,
|
| 133 |
+
"loss": 0.0432,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.1,
|
| 138 |
+
"learning_rate": 4.8948088406348404e-05,
|
| 139 |
+
"loss": 0.0343,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.1,
|
| 144 |
+
"learning_rate": 4.884681013229645e-05,
|
| 145 |
+
"loss": 0.1469,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.11,
|
| 150 |
+
"learning_rate": 4.874099247250798e-05,
|
| 151 |
+
"loss": 0.0432,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.11,
|
| 156 |
+
"learning_rate": 4.8641892034320465e-05,
|
| 157 |
+
"loss": 0.138,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.11,
|
| 162 |
+
"learning_rate": 4.85275057467904e-05,
|
| 163 |
+
"loss": 0.0637,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.12,
|
| 168 |
+
"learning_rate": 4.8408640855060846e-05,
|
| 169 |
+
"loss": 0.0268,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.12,
|
| 174 |
+
"learning_rate": 4.828531998578885e-05,
|
| 175 |
+
"loss": 0.0583,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.13,
|
| 180 |
+
"learning_rate": 4.815756661385392e-05,
|
| 181 |
+
"loss": 0.0248,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.13,
|
| 186 |
+
"learning_rate": 4.802540505788951e-05,
|
| 187 |
+
"loss": 0.0247,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.14,
|
| 192 |
+
"learning_rate": 4.788886047565377e-05,
|
| 193 |
+
"loss": 0.074,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.14,
|
| 198 |
+
"learning_rate": 4.7747958859240626e-05,
|
| 199 |
+
"loss": 0.0232,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.14,
|
| 204 |
+
"learning_rate": 4.760272703013206e-05,
|
| 205 |
+
"loss": 0.0195,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.15,
|
| 210 |
+
"learning_rate": 4.74531926340924e-05,
|
| 211 |
+
"loss": 0.0382,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.15,
|
| 216 |
+
"learning_rate": 4.729938413590588e-05,
|
| 217 |
+
"loss": 0.037,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.16,
|
| 222 |
+
"learning_rate": 4.714133081395811e-05,
|
| 223 |
+
"loss": 0.0196,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.16,
|
| 228 |
+
"learning_rate": 4.6995478349706145e-05,
|
| 229 |
+
"loss": 0.0784,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.17,
|
| 234 |
+
"learning_rate": 4.68294434139043e-05,
|
| 235 |
+
"loss": 0.0356,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.17,
|
| 240 |
+
"learning_rate": 4.665925311042354e-05,
|
| 241 |
+
"loss": 0.0135,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.18,
|
| 246 |
+
"learning_rate": 4.648493983602601e-05,
|
| 247 |
+
"loss": 0.0417,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.18,
|
| 252 |
+
"learning_rate": 4.630653677230654e-05,
|
| 253 |
+
"loss": 0.0064,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.18,
|
| 258 |
+
"learning_rate": 4.614250529713827e-05,
|
| 259 |
+
"loss": 0.1264,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.19,
|
| 264 |
+
"learning_rate": 4.595642583205011e-05,
|
| 265 |
+
"loss": 0.0072,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.19,
|
| 270 |
+
"learning_rate": 4.5766357183503764e-05,
|
| 271 |
+
"loss": 0.0022,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.2,
|
| 276 |
+
"learning_rate": 4.559191453574582e-05,
|
| 277 |
+
"loss": 0.0002,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.2,
|
| 282 |
+
"learning_rate": 4.539436673859789e-05,
|
| 283 |
+
"loss": 0.1255,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.21,
|
| 288 |
+
"learning_rate": 4.519293674936788e-05,
|
| 289 |
+
"loss": 0.0577,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.21,
|
| 294 |
+
"learning_rate": 4.498766291148263e-05,
|
| 295 |
+
"loss": 0.0095,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.22,
|
| 300 |
+
"learning_rate": 4.479966224737879e-05,
|
| 301 |
+
"loss": 0.0778,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.22,
|
| 306 |
+
"learning_rate": 4.458719334778153e-05,
|
| 307 |
+
"loss": 0.0134,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.22,
|
| 312 |
+
"learning_rate": 4.4370995906479616e-05,
|
| 313 |
+
"loss": 0.0127,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.23,
|
| 318 |
+
"learning_rate": 4.415111107797445e-05,
|
| 319 |
+
"loss": 0.0194,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.23,
|
| 324 |
+
"learning_rate": 4.3927580718684106e-05,
|
| 325 |
+
"loss": 0.0121,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.24,
|
| 330 |
+
"learning_rate": 4.372332162017128e-05,
|
| 331 |
+
"loss": 0.1112,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.24,
|
| 336 |
+
"learning_rate": 4.3492982545674274e-05,
|
| 337 |
+
"loss": 0.0298,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.25,
|
| 342 |
+
"learning_rate": 4.325912321917175e-05,
|
| 343 |
+
"loss": 0.0719,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.25,
|
| 348 |
+
"learning_rate": 4.302178815721267e-05,
|
| 349 |
+
"loss": 0.0508,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.25,
|
| 354 |
+
"learning_rate": 4.278102253797347e-05,
|
| 355 |
+
"loss": 0.0038,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.26,
|
| 360 |
+
"learning_rate": 4.2536872192658036e-05,
|
| 361 |
+
"loss": 0.0831,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.26,
|
| 366 |
+
"learning_rate": 4.228938359677354e-05,
|
| 367 |
+
"loss": 0.0713,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.27,
|
| 372 |
+
"learning_rate": 4.20386038612835e-05,
|
| 373 |
+
"loss": 0.0755,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.27,
|
| 378 |
+
"learning_rate": 4.1784580723639923e-05,
|
| 379 |
+
"loss": 0.0503,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.28,
|
| 384 |
+
"learning_rate": 4.1527362538696204e-05,
|
| 385 |
+
"loss": 0.0098,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.28,
|
| 390 |
+
"learning_rate": 4.1266998269502456e-05,
|
| 391 |
+
"loss": 0.0212,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.29,
|
| 396 |
+
"learning_rate": 4.1003537477985154e-05,
|
| 397 |
+
"loss": 0.078,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.29,
|
| 402 |
+
"learning_rate": 4.073703031551264e-05,
|
| 403 |
+
"loss": 0.0195,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.29,
|
| 408 |
+
"learning_rate": 4.046752751334857e-05,
|
| 409 |
+
"loss": 0.1526,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.3,
|
| 414 |
+
"learning_rate": 4.0195080372994864e-05,
|
| 415 |
+
"loss": 0.001,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.3,
|
| 420 |
+
"learning_rate": 3.991974075642621e-05,
|
| 421 |
+
"loss": 0.0041,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.31,
|
| 426 |
+
"learning_rate": 3.96695053421277e-05,
|
| 427 |
+
"loss": 0.1016,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.31,
|
| 432 |
+
"learning_rate": 3.938881486442823e-05,
|
| 433 |
+
"loss": 0.0073,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.32,
|
| 438 |
+
"learning_rate": 3.9105385388070305e-05,
|
| 439 |
+
"loss": 0.0134,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.32,
|
| 444 |
+
"learning_rate": 3.8819270865583214e-05,
|
| 445 |
+
"loss": 0.093,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.32,
|
| 450 |
+
"learning_rate": 3.853052576061115e-05,
|
| 451 |
+
"loss": 0.0022,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.33,
|
| 456 |
+
"learning_rate": 3.823920503754575e-05,
|
| 457 |
+
"loss": 0.0115,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.33,
|
| 462 |
+
"learning_rate": 3.7945364151063236e-05,
|
| 463 |
+
"loss": 0.0181,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.34,
|
| 468 |
+
"learning_rate": 3.764905903556832e-05,
|
| 469 |
+
"loss": 0.0206,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.34,
|
| 474 |
+
"learning_rate": 3.7350346094546725e-05,
|
| 475 |
+
"loss": 0.0263,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.35,
|
| 480 |
+
"learning_rate": 3.704928218982845e-05,
|
| 481 |
+
"loss": 0.0438,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.35,
|
| 486 |
+
"learning_rate": 3.674592463076377e-05,
|
| 487 |
+
"loss": 0.018,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.36,
|
| 492 |
+
"learning_rate": 3.644033116331406e-05,
|
| 493 |
+
"loss": 0.0764,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.36,
|
| 498 |
+
"learning_rate": 3.613255995905948e-05,
|
| 499 |
+
"loss": 0.0365,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.36,
|
| 504 |
+
"learning_rate": 3.5822669604125684e-05,
|
| 505 |
+
"loss": 0.0045,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.37,
|
| 510 |
+
"learning_rate": 3.5510719088031576e-05,
|
| 511 |
+
"loss": 0.0825,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.37,
|
| 516 |
+
"learning_rate": 3.5196767792460286e-05,
|
| 517 |
+
"loss": 0.0429,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.38,
|
| 522 |
+
"learning_rate": 3.4912550346049136e-05,
|
| 523 |
+
"loss": 0.0981,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.38,
|
| 528 |
+
"learning_rate": 3.4594962520894233e-05,
|
| 529 |
+
"loss": 0.0269,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.39,
|
| 534 |
+
"learning_rate": 3.4275548236109376e-05,
|
| 535 |
+
"loss": 0.0186,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.39,
|
| 540 |
+
"learning_rate": 3.39543682941516e-05,
|
| 541 |
+
"loss": 0.0555,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.4,
|
| 546 |
+
"learning_rate": 3.363148383358155e-05,
|
| 547 |
+
"loss": 0.0309,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.4,
|
| 552 |
+
"learning_rate": 3.330695631742537e-05,
|
| 553 |
+
"loss": 0.0541,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.4,
|
| 558 |
+
"learning_rate": 3.298084752147487e-05,
|
| 559 |
+
"loss": 0.016,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.41,
|
| 564 |
+
"learning_rate": 3.2653219522528106e-05,
|
| 565 |
+
"loss": 0.03,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.41,
|
| 570 |
+
"learning_rate": 3.23241346865727e-05,
|
| 571 |
+
"loss": 0.0025,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.42,
|
| 576 |
+
"learning_rate": 3.199365565691406e-05,
|
| 577 |
+
"loss": 0.0752,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.42,
|
| 582 |
+
"learning_rate": 3.166184534225087e-05,
|
| 583 |
+
"loss": 0.0078,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.43,
|
| 588 |
+
"learning_rate": 3.132876690470008e-05,
|
| 589 |
+
"loss": 0.0326,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.43,
|
| 594 |
+
"learning_rate": 3.099448374777351e-05,
|
| 595 |
+
"loss": 0.0249,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.43,
|
| 600 |
+
"learning_rate": 3.065905950430867e-05,
|
| 601 |
+
"loss": 0.003,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.44,
|
| 606 |
+
"learning_rate": 3.032255802435588e-05,
|
| 607 |
+
"loss": 0.0239,
|
| 608 |
+
"step": 1000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.44,
|
| 612 |
+
"learning_rate": 2.9985043363023958e-05,
|
| 613 |
+
"loss": 0.0008,
|
| 614 |
+
"step": 1010
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.45,
|
| 618 |
+
"learning_rate": 2.9646579768287002e-05,
|
| 619 |
+
"loss": 0.0722,
|
| 620 |
+
"step": 1020
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.45,
|
| 624 |
+
"learning_rate": 2.930723166875437e-05,
|
| 625 |
+
"loss": 0.0007,
|
| 626 |
+
"step": 1030
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.46,
|
| 630 |
+
"learning_rate": 2.8967063661406285e-05,
|
| 631 |
+
"loss": 0.0246,
|
| 632 |
+
"step": 1040
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.46,
|
| 636 |
+
"learning_rate": 2.862614049929743e-05,
|
| 637 |
+
"loss": 0.0235,
|
| 638 |
+
"step": 1050
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.47,
|
| 642 |
+
"learning_rate": 2.8284527079230822e-05,
|
| 643 |
+
"loss": 0.0079,
|
| 644 |
+
"step": 1060
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.47,
|
| 648 |
+
"learning_rate": 2.7942288429404256e-05,
|
| 649 |
+
"loss": 0.0271,
|
| 650 |
+
"step": 1070
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.47,
|
| 654 |
+
"learning_rate": 2.759948969703185e-05,
|
| 655 |
+
"loss": 0.001,
|
| 656 |
+
"step": 1080
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.48,
|
| 660 |
+
"learning_rate": 2.7256196135942836e-05,
|
| 661 |
+
"loss": 0.0087,
|
| 662 |
+
"step": 1090
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.48,
|
| 666 |
+
"learning_rate": 2.6912473094160106e-05,
|
| 667 |
+
"loss": 0.0737,
|
| 668 |
+
"step": 1100
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.49,
|
| 672 |
+
"learning_rate": 2.6568386001460854e-05,
|
| 673 |
+
"loss": 0.0095,
|
| 674 |
+
"step": 1110
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.49,
|
| 678 |
+
"learning_rate": 2.6224000356921546e-05,
|
| 679 |
+
"loss": 0.0533,
|
| 680 |
+
"step": 1120
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.5,
|
| 684 |
+
"learning_rate": 2.587938171644983e-05,
|
| 685 |
+
"loss": 0.027,
|
| 686 |
+
"step": 1130
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.5,
|
| 690 |
+
"learning_rate": 2.5534595680305544e-05,
|
| 691 |
+
"loss": 0.024,
|
| 692 |
+
"step": 1140
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.5,
|
| 696 |
+
"learning_rate": 2.5189707880613334e-05,
|
| 697 |
+
"loss": 0.0246,
|
| 698 |
+
"step": 1150
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.51,
|
| 702 |
+
"learning_rate": 2.4844783968869126e-05,
|
| 703 |
+
"loss": 0.0445,
|
| 704 |
+
"step": 1160
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.51,
|
| 708 |
+
"learning_rate": 2.4499889603443018e-05,
|
| 709 |
+
"loss": 0.0377,
|
| 710 |
+
"step": 1170
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.52,
|
| 714 |
+
"learning_rate": 2.4155090437080776e-05,
|
| 715 |
+
"loss": 0.0124,
|
| 716 |
+
"step": 1180
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.52,
|
| 720 |
+
"learning_rate": 2.3810452104406444e-05,
|
| 721 |
+
"loss": 0.0351,
|
| 722 |
+
"step": 1190
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.53,
|
| 726 |
+
"learning_rate": 2.3466040209428406e-05,
|
| 727 |
+
"loss": 0.0352,
|
| 728 |
+
"step": 1200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.53,
|
| 732 |
+
"learning_rate": 2.312192031305123e-05,
|
| 733 |
+
"loss": 0.0319,
|
| 734 |
+
"step": 1210
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.54,
|
| 738 |
+
"learning_rate": 2.277815792059578e-05,
|
| 739 |
+
"loss": 0.0201,
|
| 740 |
+
"step": 1220
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.54,
|
| 744 |
+
"learning_rate": 2.243481846932987e-05,
|
| 745 |
+
"loss": 0.0253,
|
| 746 |
+
"step": 1230
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.54,
|
| 750 |
+
"learning_rate": 2.209196731601187e-05,
|
| 751 |
+
"loss": 0.0368,
|
| 752 |
+
"step": 1240
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.55,
|
| 756 |
+
"learning_rate": 2.174966972444967e-05,
|
| 757 |
+
"loss": 0.002,
|
| 758 |
+
"step": 1250
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.55,
|
| 762 |
+
"learning_rate": 2.1407990853077298e-05,
|
| 763 |
+
"loss": 0.0541,
|
| 764 |
+
"step": 1260
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.56,
|
| 768 |
+
"learning_rate": 2.106699574255162e-05,
|
| 769 |
+
"loss": 0.0015,
|
| 770 |
+
"step": 1270
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.56,
|
| 774 |
+
"learning_rate": 2.0726749303371447e-05,
|
| 775 |
+
"loss": 0.0082,
|
| 776 |
+
"step": 1280
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.57,
|
| 780 |
+
"learning_rate": 2.0387316303521454e-05,
|
| 781 |
+
"loss": 0.039,
|
| 782 |
+
"step": 1290
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.57,
|
| 786 |
+
"learning_rate": 2.0048761356143165e-05,
|
| 787 |
+
"loss": 0.0139,
|
| 788 |
+
"step": 1300
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.58,
|
| 792 |
+
"learning_rate": 1.9711148907235455e-05,
|
| 793 |
+
"loss": 0.05,
|
| 794 |
+
"step": 1310
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.58,
|
| 798 |
+
"learning_rate": 1.9374543223386888e-05,
|
| 799 |
+
"loss": 0.0685,
|
| 800 |
+
"step": 1320
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.58,
|
| 804 |
+
"learning_rate": 1.903900837954213e-05,
|
| 805 |
+
"loss": 0.0047,
|
| 806 |
+
"step": 1330
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.59,
|
| 810 |
+
"learning_rate": 1.8704608246804956e-05,
|
| 811 |
+
"loss": 0.0069,
|
| 812 |
+
"step": 1340
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.59,
|
| 816 |
+
"learning_rate": 1.8371406480279934e-05,
|
| 817 |
+
"loss": 0.0138,
|
| 818 |
+
"step": 1350
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.6,
|
| 822 |
+
"learning_rate": 1.8039466506955334e-05,
|
| 823 |
+
"loss": 0.0057,
|
| 824 |
+
"step": 1360
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.6,
|
| 828 |
+
"learning_rate": 1.7708851513629377e-05,
|
| 829 |
+
"loss": 0.0021,
|
| 830 |
+
"step": 1370
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.61,
|
| 834 |
+
"learning_rate": 1.737962443488228e-05,
|
| 835 |
+
"loss": 0.0036,
|
| 836 |
+
"step": 1380
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.61,
|
| 840 |
+
"learning_rate": 1.705184794109625e-05,
|
| 841 |
+
"loss": 0.0485,
|
| 842 |
+
"step": 1390
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.61,
|
| 846 |
+
"learning_rate": 1.6725584426525786e-05,
|
| 847 |
+
"loss": 0.0006,
|
| 848 |
+
"step": 1400
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.62,
|
| 852 |
+
"learning_rate": 1.6400895997420585e-05,
|
| 853 |
+
"loss": 0.0179,
|
| 854 |
+
"step": 1410
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.62,
|
| 858 |
+
"learning_rate": 1.6077844460203206e-05,
|
| 859 |
+
"loss": 0.0009,
|
| 860 |
+
"step": 1420
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.63,
|
| 864 |
+
"learning_rate": 1.5756491309703875e-05,
|
| 865 |
+
"loss": 0.012,
|
| 866 |
+
"step": 1430
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.63,
|
| 870 |
+
"learning_rate": 1.5436897717454524e-05,
|
| 871 |
+
"loss": 0.0197,
|
| 872 |
+
"step": 1440
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.64,
|
| 876 |
+
"learning_rate": 1.5119124520044461e-05,
|
| 877 |
+
"loss": 0.0234,
|
| 878 |
+
"step": 1450
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.64,
|
| 882 |
+
"learning_rate": 1.480323220753972e-05,
|
| 883 |
+
"loss": 0.0023,
|
| 884 |
+
"step": 1460
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.65,
|
| 888 |
+
"learning_rate": 1.4489280911968438e-05,
|
| 889 |
+
"loss": 0.0252,
|
| 890 |
+
"step": 1470
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.65,
|
| 894 |
+
"learning_rate": 1.4177330395874327e-05,
|
| 895 |
+
"loss": 0.0014,
|
| 896 |
+
"step": 1480
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.65,
|
| 900 |
+
"learning_rate": 1.386744004094052e-05,
|
| 901 |
+
"loss": 0.0024,
|
| 902 |
+
"step": 1490
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.66,
|
| 906 |
+
"learning_rate": 1.3559668836685946e-05,
|
| 907 |
+
"loss": 0.0427,
|
| 908 |
+
"step": 1500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.66,
|
| 912 |
+
"learning_rate": 1.3254075369236238e-05,
|
| 913 |
+
"loss": 0.0065,
|
| 914 |
+
"step": 1510
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.67,
|
| 918 |
+
"learning_rate": 1.2950717810171558e-05,
|
| 919 |
+
"loss": 0.0061,
|
| 920 |
+
"step": 1520
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.67,
|
| 924 |
+
"learning_rate": 1.2649653905453284e-05,
|
| 925 |
+
"loss": 0.0064,
|
| 926 |
+
"step": 1530
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.68,
|
| 930 |
+
"learning_rate": 1.2350940964431687e-05,
|
| 931 |
+
"loss": 0.0604,
|
| 932 |
+
"step": 1540
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.68,
|
| 936 |
+
"learning_rate": 1.2054635848936771e-05,
|
| 937 |
+
"loss": 0.0699,
|
| 938 |
+
"step": 1550
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.68,
|
| 942 |
+
"learning_rate": 1.1760794962454257e-05,
|
| 943 |
+
"loss": 0.0061,
|
| 944 |
+
"step": 1560
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.69,
|
| 948 |
+
"learning_rate": 1.1469474239388858e-05,
|
| 949 |
+
"loss": 0.0775,
|
| 950 |
+
"step": 1570
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.69,
|
| 954 |
+
"learning_rate": 1.1180729134416796e-05,
|
| 955 |
+
"loss": 0.0388,
|
| 956 |
+
"step": 1580
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.7,
|
| 960 |
+
"learning_rate": 1.0894614611929702e-05,
|
| 961 |
+
"loss": 0.0081,
|
| 962 |
+
"step": 1590
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.7,
|
| 966 |
+
"learning_rate": 1.0611185135571777e-05,
|
| 967 |
+
"loss": 0.0039,
|
| 968 |
+
"step": 1600
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.71,
|
| 972 |
+
"learning_rate": 1.0330494657872312e-05,
|
| 973 |
+
"loss": 0.0019,
|
| 974 |
+
"step": 1610
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.71,
|
| 978 |
+
"learning_rate": 1.0052596609975507e-05,
|
| 979 |
+
"loss": 0.1103,
|
| 980 |
+
"step": 1620
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.72,
|
| 984 |
+
"learning_rate": 9.777543891469517e-06,
|
| 985 |
+
"loss": 0.0232,
|
| 986 |
+
"step": 1630
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.72,
|
| 990 |
+
"learning_rate": 9.5053888603167e-06,
|
| 991 |
+
"loss": 0.0157,
|
| 992 |
+
"step": 1640
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.72,
|
| 996 |
+
"learning_rate": 9.236183322886945e-06,
|
| 997 |
+
"loss": 0.0028,
|
| 998 |
+
"step": 1650
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.73,
|
| 1002 |
+
"learning_rate": 8.969978524096034e-06,
|
| 1003 |
+
"loss": 0.0009,
|
| 1004 |
+
"step": 1660
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.73,
|
| 1008 |
+
"learning_rate": 8.706825137650837e-06,
|
| 1009 |
+
"loss": 0.0138,
|
| 1010 |
+
"step": 1670
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.74,
|
| 1014 |
+
"learning_rate": 8.446773256403298e-06,
|
| 1015 |
+
"loss": 0.0409,
|
| 1016 |
+
"step": 1680
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.74,
|
| 1020 |
+
"learning_rate": 8.189872382814912e-06,
|
| 1021 |
+
"loss": 0.0603,
|
| 1022 |
+
"step": 1690
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.75,
|
| 1026 |
+
"learning_rate": 7.936171419533653e-06,
|
| 1027 |
+
"loss": 0.0158,
|
| 1028 |
+
"step": 1700
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.75,
|
| 1032 |
+
"learning_rate": 7.685718660085057e-06,
|
| 1033 |
+
"loss": 0.0167,
|
| 1034 |
+
"step": 1710
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.76,
|
| 1038 |
+
"learning_rate": 7.4385617796792516e-06,
|
| 1039 |
+
"loss": 0.0161,
|
| 1040 |
+
"step": 1720
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.76,
|
| 1044 |
+
"learning_rate": 7.194747826135687e-06,
|
| 1045 |
+
"loss": 0.0015,
|
| 1046 |
+
"step": 1730
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.76,
|
| 1050 |
+
"learning_rate": 6.954323210927299e-06,
|
| 1051 |
+
"loss": 0.0633,
|
| 1052 |
+
"step": 1740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.77,
|
| 1056 |
+
"learning_rate": 6.717333700345807e-06,
|
| 1057 |
+
"loss": 0.0544,
|
| 1058 |
+
"step": 1750
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.77,
|
| 1062 |
+
"learning_rate": 6.483824406789798e-06,
|
| 1063 |
+
"loss": 0.0022,
|
| 1064 |
+
"step": 1760
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.78,
|
| 1068 |
+
"learning_rate": 6.2538397801773594e-06,
|
| 1069 |
+
"loss": 0.0341,
|
| 1070 |
+
"step": 1770
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.78,
|
| 1074 |
+
"learning_rate": 6.027423599484697e-06,
|
| 1075 |
+
"loss": 0.0167,
|
| 1076 |
+
"step": 1780
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.79,
|
| 1080 |
+
"learning_rate": 5.804618964412586e-06,
|
| 1081 |
+
"loss": 0.024,
|
| 1082 |
+
"step": 1790
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.79,
|
| 1086 |
+
"learning_rate": 5.585468287182071e-06,
|
| 1087 |
+
"loss": 0.0293,
|
| 1088 |
+
"step": 1800
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.79,
|
| 1092 |
+
"learning_rate": 5.391391304955415e-06,
|
| 1093 |
+
"loss": 0.0008,
|
| 1094 |
+
"step": 1810
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.8,
|
| 1098 |
+
"learning_rate": 5.1792974957679375e-06,
|
| 1099 |
+
"loss": 0.0123,
|
| 1100 |
+
"step": 1820
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.8,
|
| 1104 |
+
"learning_rate": 4.970976678177083e-06,
|
| 1105 |
+
"loss": 0.0024,
|
| 1106 |
+
"step": 1830
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.81,
|
| 1110 |
+
"learning_rate": 4.766468507320909e-06,
|
| 1111 |
+
"loss": 0.024,
|
| 1112 |
+
"step": 1840
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.81,
|
| 1116 |
+
"learning_rate": 4.565811912576912e-06,
|
| 1117 |
+
"loss": 0.0593,
|
| 1118 |
+
"step": 1850
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.82,
|
| 1122 |
+
"learning_rate": 4.369045090151591e-06,
|
| 1123 |
+
"loss": 0.0164,
|
| 1124 |
+
"step": 1860
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.82,
|
| 1128 |
+
"learning_rate": 4.176205495809563e-06,
|
| 1129 |
+
"loss": 0.0213,
|
| 1130 |
+
"step": 1870
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.83,
|
| 1134 |
+
"learning_rate": 3.987329837743625e-06,
|
| 1135 |
+
"loss": 0.0155,
|
| 1136 |
+
"step": 1880
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.83,
|
| 1140 |
+
"learning_rate": 3.8024540695871274e-06,
|
| 1141 |
+
"loss": 0.0038,
|
| 1142 |
+
"step": 1890
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.83,
|
| 1146 |
+
"learning_rate": 3.6216133835699995e-06,
|
| 1147 |
+
"loss": 0.0519,
|
| 1148 |
+
"step": 1900
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.84,
|
| 1152 |
+
"learning_rate": 3.444842203819662e-06,
|
| 1153 |
+
"loss": 0.0105,
|
| 1154 |
+
"step": 1910
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.84,
|
| 1158 |
+
"learning_rate": 3.2721741798082073e-06,
|
| 1159 |
+
"loss": 0.0279,
|
| 1160 |
+
"step": 1920
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.85,
|
| 1164 |
+
"learning_rate": 3.1036421799470057e-06,
|
| 1165 |
+
"loss": 0.0315,
|
| 1166 |
+
"step": 1930
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.85,
|
| 1170 |
+
"learning_rate": 2.9392782853300077e-06,
|
| 1171 |
+
"loss": 0.0124,
|
| 1172 |
+
"step": 1940
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.86,
|
| 1176 |
+
"learning_rate": 2.779113783626916e-06,
|
| 1177 |
+
"loss": 0.0008,
|
| 1178 |
+
"step": 1950
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.86,
|
| 1182 |
+
"learning_rate": 2.623179163127379e-06,
|
| 1183 |
+
"loss": 0.0012,
|
| 1184 |
+
"step": 1960
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.86,
|
| 1188 |
+
"learning_rate": 2.471504106937364e-06,
|
| 1189 |
+
"loss": 0.0241,
|
| 1190 |
+
"step": 1970
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.87,
|
| 1194 |
+
"learning_rate": 2.324117487328789e-06,
|
| 1195 |
+
"loss": 0.027,
|
| 1196 |
+
"step": 1980
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.87,
|
| 1200 |
+
"learning_rate": 2.1810473602435324e-06,
|
| 1201 |
+
"loss": 0.0752,
|
| 1202 |
+
"step": 1990
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.88,
|
| 1206 |
+
"learning_rate": 2.0423209599527814e-06,
|
| 1207 |
+
"loss": 0.0041,
|
| 1208 |
+
"step": 2000
|
| 1209 |
+
}
|
| 1210 |
+
],
|
| 1211 |
+
"max_steps": 2277,
|
| 1212 |
+
"num_train_epochs": 1,
|
| 1213 |
+
"total_flos": 0.0,
|
| 1214 |
+
"trial_name": null,
|
| 1215 |
+
"trial_params": null
|
| 1216 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d327d28f548be400b3f51ccbfee862c590c1001f0441147bdb7a2b731b2aa9a3
|
| 3 |
+
size 3707
|
value_head.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a53734375791797160b400b66f4819f0f216620416ed87d7a4844c10910444c
|
| 3 |
+
size 17587
|