Upload 25 files
Browse files- adapter_config.json +16 -0
- adapter_model.bin +3 -0
- checkpoint-1100/optimizer.pt +3 -0
- checkpoint-1100/pytorch_model.bin +3 -0
- checkpoint-1100/rng_state.pth +3 -0
- checkpoint-1100/scaler.pt +3 -0
- checkpoint-1100/scheduler.pt +3 -0
- checkpoint-1100/trainer_state.json +1416 -0
- checkpoint-1100/training_args.bin +3 -0
- checkpoint-1150/optimizer.pt +3 -0
- checkpoint-1150/pytorch_model.bin +3 -0
- checkpoint-1150/rng_state.pth +3 -0
- checkpoint-1150/scaler.pt +3 -0
- checkpoint-1150/scheduler.pt +3 -0
- checkpoint-1150/trainer_state.json +1484 -0
- checkpoint-1150/training_args.bin +3 -0
- checkpoint-1200/optimizer.pt +3 -0
- checkpoint-1200/pytorch_model.bin +3 -0
- checkpoint-1200/rng_state.pth +3 -0
- checkpoint-1200/scaler.pt +3 -0
- checkpoint-1200/scheduler.pt +3 -0
- checkpoint-1200/trainer_state.json +1544 -0
- checkpoint-1200/training_args.bin +3 -0
- runs/Jun18_19-59-55_3a5a94d2e5ba/1687118410.021046/events.out.tfevents.1687118410.3a5a94d2e5ba.49832.1 +3 -0
- runs/Jun18_19-59-55_3a5a94d2e5ba/events.out.tfevents.1687118410.3a5a94d2e5ba.49832.0 +3 -0
adapter_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "/workspace/process/falcon7b-instruct/HF",
|
| 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 |
+
"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 sha256:538679512fb502c66ff1a59fc9a70c8fb6d5328f5c547cbbaeb97bd687385fc7
|
| 3 |
+
size 9460849
|
checkpoint-1100/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac6a31f9c50b387d66e9866672d2e8023215540bfc49861f94a91f60b864d451
|
| 3 |
+
size 18928005
|
checkpoint-1100/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c0e27b364fe6c7ef0baf04782c7b51dbb88170a8d8a092645d236b0b68f36919
|
| 3 |
+
size 9460849
|
checkpoint-1100/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4c839f7cbde4de60d3b7603a907619fc372f0062df060d8faeb73b1f6ba6acf6
|
| 3 |
+
size 14575
|
checkpoint-1100/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:677eca5ad6580d3f557bbfda51765525add80f9acfd6c5ca0edf4a229f0c70eb
|
| 3 |
+
size 557
|
checkpoint-1100/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db1f69feaf160991864c37b5724e8861e67c208b936820169284c32a2b992df3
|
| 3 |
+
size 627
|
checkpoint-1100/trainer_state.json
ADDED
|
@@ -0,0 +1,1416 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.644230769230769,
|
| 5 |
+
"global_step": 1100,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 0.0003,
|
| 13 |
+
"loss": 2.0952,
|
| 14 |
+
"step": 5
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 0.000298793242156074,
|
| 19 |
+
"loss": 2.2593,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 0.000297586484312148,
|
| 25 |
+
"loss": 2.1623,
|
| 26 |
+
"step": 15
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 0.000296379726468222,
|
| 31 |
+
"loss": 1.9038,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 0.000295172968624296,
|
| 37 |
+
"loss": 1.8576,
|
| 38 |
+
"step": 25
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 0.00029396621078037004,
|
| 43 |
+
"loss": 1.8397,
|
| 44 |
+
"step": 30
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08,
|
| 48 |
+
"learning_rate": 0.00029275945293644407,
|
| 49 |
+
"loss": 1.8032,
|
| 50 |
+
"step": 35
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.1,
|
| 54 |
+
"learning_rate": 0.00029155269509251804,
|
| 55 |
+
"loss": 1.7281,
|
| 56 |
+
"step": 40
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.11,
|
| 60 |
+
"learning_rate": 0.00029034593724859206,
|
| 61 |
+
"loss": 1.7097,
|
| 62 |
+
"step": 45
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.12,
|
| 66 |
+
"learning_rate": 0.00028913917940466614,
|
| 67 |
+
"loss": 1.7421,
|
| 68 |
+
"step": 50
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.13,
|
| 72 |
+
"learning_rate": 0.0002879324215607401,
|
| 73 |
+
"loss": 1.6861,
|
| 74 |
+
"step": 55
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.14,
|
| 78 |
+
"learning_rate": 0.00028672566371681414,
|
| 79 |
+
"loss": 1.6848,
|
| 80 |
+
"step": 60
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16,
|
| 84 |
+
"learning_rate": 0.00028551890587288816,
|
| 85 |
+
"loss": 1.5163,
|
| 86 |
+
"step": 65
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.17,
|
| 90 |
+
"learning_rate": 0.0002843121480289622,
|
| 91 |
+
"loss": 1.7266,
|
| 92 |
+
"step": 70
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.18,
|
| 96 |
+
"learning_rate": 0.0002831053901850362,
|
| 97 |
+
"loss": 1.5888,
|
| 98 |
+
"step": 75
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.19,
|
| 102 |
+
"learning_rate": 0.0002818986323411102,
|
| 103 |
+
"loss": 1.5463,
|
| 104 |
+
"step": 80
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.2,
|
| 108 |
+
"learning_rate": 0.0002806918744971842,
|
| 109 |
+
"loss": 1.4611,
|
| 110 |
+
"step": 85
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.22,
|
| 114 |
+
"learning_rate": 0.00027948511665325823,
|
| 115 |
+
"loss": 1.5554,
|
| 116 |
+
"step": 90
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.23,
|
| 120 |
+
"learning_rate": 0.00027827835880933226,
|
| 121 |
+
"loss": 1.5481,
|
| 122 |
+
"step": 95
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.24,
|
| 126 |
+
"learning_rate": 0.0002770716009654062,
|
| 127 |
+
"loss": 1.4464,
|
| 128 |
+
"step": 100
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.25,
|
| 132 |
+
"eval_loss": 1.6602041721343994,
|
| 133 |
+
"eval_runtime": 25.516,
|
| 134 |
+
"eval_samples_per_second": 8.191,
|
| 135 |
+
"eval_steps_per_second": 1.058,
|
| 136 |
+
"step": 104
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.25,
|
| 140 |
+
"learning_rate": 0.00027586484312148025,
|
| 141 |
+
"loss": 1.6648,
|
| 142 |
+
"step": 105
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.26,
|
| 146 |
+
"learning_rate": 0.0002746580852775543,
|
| 147 |
+
"loss": 1.6464,
|
| 148 |
+
"step": 110
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.28,
|
| 152 |
+
"learning_rate": 0.0002734513274336283,
|
| 153 |
+
"loss": 1.524,
|
| 154 |
+
"step": 115
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.29,
|
| 158 |
+
"learning_rate": 0.0002722445695897023,
|
| 159 |
+
"loss": 1.4791,
|
| 160 |
+
"step": 120
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.3,
|
| 164 |
+
"learning_rate": 0.0002710378117457763,
|
| 165 |
+
"loss": 1.5162,
|
| 166 |
+
"step": 125
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.31,
|
| 170 |
+
"learning_rate": 0.0002698310539018503,
|
| 171 |
+
"loss": 1.6636,
|
| 172 |
+
"step": 130
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.32,
|
| 176 |
+
"learning_rate": 0.00026862429605792435,
|
| 177 |
+
"loss": 1.7803,
|
| 178 |
+
"step": 135
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.34,
|
| 182 |
+
"learning_rate": 0.00026741753821399837,
|
| 183 |
+
"loss": 1.7384,
|
| 184 |
+
"step": 140
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.35,
|
| 188 |
+
"learning_rate": 0.0002662107803700724,
|
| 189 |
+
"loss": 1.5462,
|
| 190 |
+
"step": 145
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.36,
|
| 194 |
+
"learning_rate": 0.0002650040225261464,
|
| 195 |
+
"loss": 1.5802,
|
| 196 |
+
"step": 150
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.37,
|
| 200 |
+
"learning_rate": 0.00026379726468222045,
|
| 201 |
+
"loss": 1.4942,
|
| 202 |
+
"step": 155
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.38,
|
| 206 |
+
"learning_rate": 0.0002625905068382944,
|
| 207 |
+
"loss": 1.601,
|
| 208 |
+
"step": 160
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.4,
|
| 212 |
+
"learning_rate": 0.00026138374899436844,
|
| 213 |
+
"loss": 1.5822,
|
| 214 |
+
"step": 165
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.41,
|
| 218 |
+
"learning_rate": 0.00026017699115044247,
|
| 219 |
+
"loss": 1.5071,
|
| 220 |
+
"step": 170
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.42,
|
| 224 |
+
"learning_rate": 0.0002589702333065165,
|
| 225 |
+
"loss": 1.5926,
|
| 226 |
+
"step": 175
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.43,
|
| 230 |
+
"learning_rate": 0.00025776347546259046,
|
| 231 |
+
"loss": 1.4206,
|
| 232 |
+
"step": 180
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.44,
|
| 236 |
+
"learning_rate": 0.0002565567176186645,
|
| 237 |
+
"loss": 1.5317,
|
| 238 |
+
"step": 185
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.46,
|
| 242 |
+
"learning_rate": 0.0002553499597747385,
|
| 243 |
+
"loss": 1.4862,
|
| 244 |
+
"step": 190
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.47,
|
| 248 |
+
"learning_rate": 0.00025414320193081253,
|
| 249 |
+
"loss": 1.4387,
|
| 250 |
+
"step": 195
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.48,
|
| 254 |
+
"learning_rate": 0.00025293644408688656,
|
| 255 |
+
"loss": 1.5276,
|
| 256 |
+
"step": 200
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.49,
|
| 260 |
+
"learning_rate": 0.00025172968624296053,
|
| 261 |
+
"loss": 1.55,
|
| 262 |
+
"step": 205
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.5,
|
| 266 |
+
"eval_loss": 1.6166218519210815,
|
| 267 |
+
"eval_runtime": 13.3266,
|
| 268 |
+
"eval_samples_per_second": 15.683,
|
| 269 |
+
"eval_steps_per_second": 2.026,
|
| 270 |
+
"step": 208
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.5,
|
| 274 |
+
"learning_rate": 0.00025052292839903455,
|
| 275 |
+
"loss": 1.5525,
|
| 276 |
+
"step": 210
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.52,
|
| 280 |
+
"learning_rate": 0.0002493161705551086,
|
| 281 |
+
"loss": 1.5591,
|
| 282 |
+
"step": 215
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.53,
|
| 286 |
+
"learning_rate": 0.0002481094127111826,
|
| 287 |
+
"loss": 1.529,
|
| 288 |
+
"step": 220
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.54,
|
| 292 |
+
"learning_rate": 0.0002469026548672566,
|
| 293 |
+
"loss": 1.7152,
|
| 294 |
+
"step": 225
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.55,
|
| 298 |
+
"learning_rate": 0.00024569589702333065,
|
| 299 |
+
"loss": 1.7493,
|
| 300 |
+
"step": 230
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.56,
|
| 304 |
+
"learning_rate": 0.0002444891391794047,
|
| 305 |
+
"loss": 1.5288,
|
| 306 |
+
"step": 235
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.58,
|
| 310 |
+
"learning_rate": 0.00024328238133547868,
|
| 311 |
+
"loss": 1.5651,
|
| 312 |
+
"step": 240
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.59,
|
| 316 |
+
"learning_rate": 0.00024207562349155267,
|
| 317 |
+
"loss": 1.5144,
|
| 318 |
+
"step": 245
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6,
|
| 322 |
+
"learning_rate": 0.0002408688656476267,
|
| 323 |
+
"loss": 1.5334,
|
| 324 |
+
"step": 250
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.61,
|
| 328 |
+
"learning_rate": 0.0002396621078037007,
|
| 329 |
+
"loss": 1.5988,
|
| 330 |
+
"step": 255
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.62,
|
| 334 |
+
"learning_rate": 0.00023845534995977472,
|
| 335 |
+
"loss": 1.6401,
|
| 336 |
+
"step": 260
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.64,
|
| 340 |
+
"learning_rate": 0.00023724859211584875,
|
| 341 |
+
"loss": 1.5929,
|
| 342 |
+
"step": 265
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.65,
|
| 346 |
+
"learning_rate": 0.00023604183427192274,
|
| 347 |
+
"loss": 1.3948,
|
| 348 |
+
"step": 270
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.66,
|
| 352 |
+
"learning_rate": 0.00023483507642799677,
|
| 353 |
+
"loss": 1.221,
|
| 354 |
+
"step": 275
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.67,
|
| 358 |
+
"learning_rate": 0.00023362831858407077,
|
| 359 |
+
"loss": 1.5604,
|
| 360 |
+
"step": 280
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.69,
|
| 364 |
+
"learning_rate": 0.0002324215607401448,
|
| 365 |
+
"loss": 1.505,
|
| 366 |
+
"step": 285
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.7,
|
| 370 |
+
"learning_rate": 0.0002312148028962188,
|
| 371 |
+
"loss": 1.5008,
|
| 372 |
+
"step": 290
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.71,
|
| 376 |
+
"learning_rate": 0.0002300080450522928,
|
| 377 |
+
"loss": 1.5176,
|
| 378 |
+
"step": 295
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.72,
|
| 382 |
+
"learning_rate": 0.0002288012872083668,
|
| 383 |
+
"loss": 1.4227,
|
| 384 |
+
"step": 300
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.73,
|
| 388 |
+
"learning_rate": 0.00022759452936444084,
|
| 389 |
+
"loss": 1.1778,
|
| 390 |
+
"step": 305
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.75,
|
| 394 |
+
"learning_rate": 0.00022638777152051486,
|
| 395 |
+
"loss": 1.5052,
|
| 396 |
+
"step": 310
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.75,
|
| 400 |
+
"eval_loss": 1.5979019403457642,
|
| 401 |
+
"eval_runtime": 13.342,
|
| 402 |
+
"eval_samples_per_second": 15.665,
|
| 403 |
+
"eval_steps_per_second": 2.024,
|
| 404 |
+
"step": 312
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.76,
|
| 408 |
+
"learning_rate": 0.00022518101367658886,
|
| 409 |
+
"loss": 1.5228,
|
| 410 |
+
"step": 315
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.77,
|
| 414 |
+
"learning_rate": 0.0002239742558326629,
|
| 415 |
+
"loss": 1.5385,
|
| 416 |
+
"step": 320
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.78,
|
| 420 |
+
"learning_rate": 0.0002227674979887369,
|
| 421 |
+
"loss": 1.5072,
|
| 422 |
+
"step": 325
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.79,
|
| 426 |
+
"learning_rate": 0.00022156074014481093,
|
| 427 |
+
"loss": 1.5032,
|
| 428 |
+
"step": 330
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.81,
|
| 432 |
+
"learning_rate": 0.00022035398230088496,
|
| 433 |
+
"loss": 1.5981,
|
| 434 |
+
"step": 335
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.82,
|
| 438 |
+
"learning_rate": 0.00021914722445695895,
|
| 439 |
+
"loss": 1.422,
|
| 440 |
+
"step": 340
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.83,
|
| 444 |
+
"learning_rate": 0.00021794046661303298,
|
| 445 |
+
"loss": 1.4948,
|
| 446 |
+
"step": 345
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.84,
|
| 450 |
+
"learning_rate": 0.00021673370876910698,
|
| 451 |
+
"loss": 1.4421,
|
| 452 |
+
"step": 350
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.85,
|
| 456 |
+
"learning_rate": 0.000215526950925181,
|
| 457 |
+
"loss": 1.6086,
|
| 458 |
+
"step": 355
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.87,
|
| 462 |
+
"learning_rate": 0.000214320193081255,
|
| 463 |
+
"loss": 1.3874,
|
| 464 |
+
"step": 360
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.88,
|
| 468 |
+
"learning_rate": 0.00021311343523732902,
|
| 469 |
+
"loss": 1.2493,
|
| 470 |
+
"step": 365
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.89,
|
| 474 |
+
"learning_rate": 0.00021190667739340302,
|
| 475 |
+
"loss": 1.5565,
|
| 476 |
+
"step": 370
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.9,
|
| 480 |
+
"learning_rate": 0.00021069991954947705,
|
| 481 |
+
"loss": 1.4739,
|
| 482 |
+
"step": 375
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.91,
|
| 486 |
+
"learning_rate": 0.00020949316170555107,
|
| 487 |
+
"loss": 1.5288,
|
| 488 |
+
"step": 380
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.93,
|
| 492 |
+
"learning_rate": 0.00020828640386162507,
|
| 493 |
+
"loss": 1.4837,
|
| 494 |
+
"step": 385
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.94,
|
| 498 |
+
"learning_rate": 0.0002070796460176991,
|
| 499 |
+
"loss": 1.5009,
|
| 500 |
+
"step": 390
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.95,
|
| 504 |
+
"learning_rate": 0.0002058728881737731,
|
| 505 |
+
"loss": 1.3469,
|
| 506 |
+
"step": 395
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.96,
|
| 510 |
+
"learning_rate": 0.00020466613032984712,
|
| 511 |
+
"loss": 1.3949,
|
| 512 |
+
"step": 400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.97,
|
| 516 |
+
"learning_rate": 0.00020345937248592117,
|
| 517 |
+
"loss": 1.3688,
|
| 518 |
+
"step": 405
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.99,
|
| 522 |
+
"learning_rate": 0.00020225261464199517,
|
| 523 |
+
"loss": 1.6629,
|
| 524 |
+
"step": 410
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.0,
|
| 528 |
+
"learning_rate": 0.0002010458567980692,
|
| 529 |
+
"loss": 1.5395,
|
| 530 |
+
"step": 415
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.0,
|
| 534 |
+
"eval_loss": 1.589026927947998,
|
| 535 |
+
"eval_runtime": 13.34,
|
| 536 |
+
"eval_samples_per_second": 15.667,
|
| 537 |
+
"eval_steps_per_second": 2.024,
|
| 538 |
+
"step": 416
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.01,
|
| 542 |
+
"learning_rate": 0.0001998390989541432,
|
| 543 |
+
"loss": 1.5794,
|
| 544 |
+
"step": 420
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.02,
|
| 548 |
+
"learning_rate": 0.0001986323411102172,
|
| 549 |
+
"loss": 1.4746,
|
| 550 |
+
"step": 425
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.03,
|
| 554 |
+
"learning_rate": 0.0001974255832662912,
|
| 555 |
+
"loss": 1.3205,
|
| 556 |
+
"step": 430
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.05,
|
| 560 |
+
"learning_rate": 0.00019621882542236523,
|
| 561 |
+
"loss": 1.491,
|
| 562 |
+
"step": 435
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.06,
|
| 566 |
+
"learning_rate": 0.00019501206757843926,
|
| 567 |
+
"loss": 1.4803,
|
| 568 |
+
"step": 440
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.07,
|
| 572 |
+
"learning_rate": 0.00019380530973451326,
|
| 573 |
+
"loss": 1.3436,
|
| 574 |
+
"step": 445
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.08,
|
| 578 |
+
"learning_rate": 0.00019259855189058728,
|
| 579 |
+
"loss": 1.3893,
|
| 580 |
+
"step": 450
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.09,
|
| 584 |
+
"learning_rate": 0.00019139179404666128,
|
| 585 |
+
"loss": 1.0474,
|
| 586 |
+
"step": 455
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.11,
|
| 590 |
+
"learning_rate": 0.0001901850362027353,
|
| 591 |
+
"loss": 1.3388,
|
| 592 |
+
"step": 460
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.12,
|
| 596 |
+
"learning_rate": 0.0001889782783588093,
|
| 597 |
+
"loss": 1.4201,
|
| 598 |
+
"step": 465
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.13,
|
| 602 |
+
"learning_rate": 0.00018777152051488333,
|
| 603 |
+
"loss": 1.4805,
|
| 604 |
+
"step": 470
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.14,
|
| 608 |
+
"learning_rate": 0.00018656476267095732,
|
| 609 |
+
"loss": 1.2253,
|
| 610 |
+
"step": 475
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 1.15,
|
| 614 |
+
"learning_rate": 0.00018535800482703135,
|
| 615 |
+
"loss": 1.5064,
|
| 616 |
+
"step": 480
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 1.17,
|
| 620 |
+
"learning_rate": 0.00018415124698310537,
|
| 621 |
+
"loss": 1.3565,
|
| 622 |
+
"step": 485
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.18,
|
| 626 |
+
"learning_rate": 0.00018294448913917937,
|
| 627 |
+
"loss": 1.3976,
|
| 628 |
+
"step": 490
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.19,
|
| 632 |
+
"learning_rate": 0.00018173773129525342,
|
| 633 |
+
"loss": 1.5882,
|
| 634 |
+
"step": 495
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.2,
|
| 638 |
+
"learning_rate": 0.00018053097345132742,
|
| 639 |
+
"loss": 1.4698,
|
| 640 |
+
"step": 500
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.21,
|
| 644 |
+
"learning_rate": 0.00017932421560740145,
|
| 645 |
+
"loss": 1.5326,
|
| 646 |
+
"step": 505
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.23,
|
| 650 |
+
"learning_rate": 0.00017811745776347547,
|
| 651 |
+
"loss": 1.5659,
|
| 652 |
+
"step": 510
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 1.24,
|
| 656 |
+
"learning_rate": 0.00017691069991954947,
|
| 657 |
+
"loss": 1.4489,
|
| 658 |
+
"step": 515
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 1.25,
|
| 662 |
+
"learning_rate": 0.0001757039420756235,
|
| 663 |
+
"loss": 1.5312,
|
| 664 |
+
"step": 520
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.25,
|
| 668 |
+
"eval_loss": 1.5857353210449219,
|
| 669 |
+
"eval_runtime": 13.3247,
|
| 670 |
+
"eval_samples_per_second": 15.685,
|
| 671 |
+
"eval_steps_per_second": 2.026,
|
| 672 |
+
"step": 520
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.26,
|
| 676 |
+
"learning_rate": 0.0001744971842316975,
|
| 677 |
+
"loss": 1.3595,
|
| 678 |
+
"step": 525
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.27,
|
| 682 |
+
"learning_rate": 0.00017329042638777152,
|
| 683 |
+
"loss": 1.1543,
|
| 684 |
+
"step": 530
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.29,
|
| 688 |
+
"learning_rate": 0.0001720836685438455,
|
| 689 |
+
"loss": 1.4568,
|
| 690 |
+
"step": 535
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.3,
|
| 694 |
+
"learning_rate": 0.00017087691069991954,
|
| 695 |
+
"loss": 1.3682,
|
| 696 |
+
"step": 540
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.31,
|
| 700 |
+
"learning_rate": 0.00016967015285599354,
|
| 701 |
+
"loss": 1.5323,
|
| 702 |
+
"step": 545
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.32,
|
| 706 |
+
"learning_rate": 0.00016846339501206756,
|
| 707 |
+
"loss": 1.6644,
|
| 708 |
+
"step": 550
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 1.33,
|
| 712 |
+
"learning_rate": 0.00016725663716814158,
|
| 713 |
+
"loss": 1.5735,
|
| 714 |
+
"step": 555
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 1.35,
|
| 718 |
+
"learning_rate": 0.00016604987932421558,
|
| 719 |
+
"loss": 1.4759,
|
| 720 |
+
"step": 560
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 1.36,
|
| 724 |
+
"learning_rate": 0.0001648431214802896,
|
| 725 |
+
"loss": 1.3774,
|
| 726 |
+
"step": 565
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.37,
|
| 730 |
+
"learning_rate": 0.0001636363636363636,
|
| 731 |
+
"loss": 1.3321,
|
| 732 |
+
"step": 570
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.38,
|
| 736 |
+
"learning_rate": 0.00016242960579243763,
|
| 737 |
+
"loss": 1.2843,
|
| 738 |
+
"step": 575
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.39,
|
| 742 |
+
"learning_rate": 0.00016122284794851163,
|
| 743 |
+
"loss": 1.4636,
|
| 744 |
+
"step": 580
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.41,
|
| 748 |
+
"learning_rate": 0.00016001609010458568,
|
| 749 |
+
"loss": 1.4082,
|
| 750 |
+
"step": 585
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 1.42,
|
| 754 |
+
"learning_rate": 0.0001588093322606597,
|
| 755 |
+
"loss": 1.3848,
|
| 756 |
+
"step": 590
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 1.43,
|
| 760 |
+
"learning_rate": 0.0001576025744167337,
|
| 761 |
+
"loss": 1.5396,
|
| 762 |
+
"step": 595
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1.44,
|
| 766 |
+
"learning_rate": 0.00015639581657280773,
|
| 767 |
+
"loss": 1.4889,
|
| 768 |
+
"step": 600
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 1.45,
|
| 772 |
+
"learning_rate": 0.00015518905872888172,
|
| 773 |
+
"loss": 1.4019,
|
| 774 |
+
"step": 605
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.47,
|
| 778 |
+
"learning_rate": 0.00015398230088495575,
|
| 779 |
+
"loss": 1.3788,
|
| 780 |
+
"step": 610
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.48,
|
| 784 |
+
"learning_rate": 0.00015277554304102975,
|
| 785 |
+
"loss": 1.409,
|
| 786 |
+
"step": 615
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.49,
|
| 790 |
+
"learning_rate": 0.00015156878519710377,
|
| 791 |
+
"loss": 1.5675,
|
| 792 |
+
"step": 620
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 1.5,
|
| 796 |
+
"eval_loss": 1.584699273109436,
|
| 797 |
+
"eval_runtime": 13.3509,
|
| 798 |
+
"eval_samples_per_second": 15.654,
|
| 799 |
+
"eval_steps_per_second": 2.022,
|
| 800 |
+
"step": 624
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.5,
|
| 804 |
+
"learning_rate": 0.0001503620273531778,
|
| 805 |
+
"loss": 1.3176,
|
| 806 |
+
"step": 625
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.51,
|
| 810 |
+
"learning_rate": 0.0001491552695092518,
|
| 811 |
+
"loss": 1.3114,
|
| 812 |
+
"step": 630
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.53,
|
| 816 |
+
"learning_rate": 0.00014794851166532582,
|
| 817 |
+
"loss": 1.4261,
|
| 818 |
+
"step": 635
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.54,
|
| 822 |
+
"learning_rate": 0.00014674175382139982,
|
| 823 |
+
"loss": 1.5156,
|
| 824 |
+
"step": 640
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.55,
|
| 828 |
+
"learning_rate": 0.00014553499597747384,
|
| 829 |
+
"loss": 1.3231,
|
| 830 |
+
"step": 645
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.56,
|
| 834 |
+
"learning_rate": 0.00014432823813354787,
|
| 835 |
+
"loss": 1.4353,
|
| 836 |
+
"step": 650
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.57,
|
| 840 |
+
"learning_rate": 0.00014312148028962186,
|
| 841 |
+
"loss": 1.4423,
|
| 842 |
+
"step": 655
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.59,
|
| 846 |
+
"learning_rate": 0.0001419147224456959,
|
| 847 |
+
"loss": 1.1512,
|
| 848 |
+
"step": 660
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.6,
|
| 852 |
+
"learning_rate": 0.0001407079646017699,
|
| 853 |
+
"loss": 1.3562,
|
| 854 |
+
"step": 665
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.61,
|
| 858 |
+
"learning_rate": 0.0001395012067578439,
|
| 859 |
+
"loss": 1.4033,
|
| 860 |
+
"step": 670
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.62,
|
| 864 |
+
"learning_rate": 0.00013829444891391793,
|
| 865 |
+
"loss": 1.5435,
|
| 866 |
+
"step": 675
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.63,
|
| 870 |
+
"learning_rate": 0.00013708769106999193,
|
| 871 |
+
"loss": 1.4543,
|
| 872 |
+
"step": 680
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.65,
|
| 876 |
+
"learning_rate": 0.00013588093322606596,
|
| 877 |
+
"loss": 1.3509,
|
| 878 |
+
"step": 685
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.66,
|
| 882 |
+
"learning_rate": 0.00013467417538213995,
|
| 883 |
+
"loss": 1.4268,
|
| 884 |
+
"step": 690
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.67,
|
| 888 |
+
"learning_rate": 0.000133467417538214,
|
| 889 |
+
"loss": 1.4231,
|
| 890 |
+
"step": 695
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.68,
|
| 894 |
+
"learning_rate": 0.000132260659694288,
|
| 895 |
+
"loss": 1.4727,
|
| 896 |
+
"step": 700
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.69,
|
| 900 |
+
"learning_rate": 0.00013105390185036203,
|
| 901 |
+
"loss": 1.2846,
|
| 902 |
+
"step": 705
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.71,
|
| 906 |
+
"learning_rate": 0.00012984714400643603,
|
| 907 |
+
"loss": 1.3458,
|
| 908 |
+
"step": 710
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 1.72,
|
| 912 |
+
"learning_rate": 0.00012864038616251005,
|
| 913 |
+
"loss": 1.6349,
|
| 914 |
+
"step": 715
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.73,
|
| 918 |
+
"learning_rate": 0.00012743362831858405,
|
| 919 |
+
"loss": 1.4798,
|
| 920 |
+
"step": 720
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.74,
|
| 924 |
+
"learning_rate": 0.00012622687047465807,
|
| 925 |
+
"loss": 1.6157,
|
| 926 |
+
"step": 725
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.75,
|
| 930 |
+
"eval_loss": 1.5761181116104126,
|
| 931 |
+
"eval_runtime": 13.3218,
|
| 932 |
+
"eval_samples_per_second": 15.689,
|
| 933 |
+
"eval_steps_per_second": 2.027,
|
| 934 |
+
"step": 728
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.75,
|
| 938 |
+
"learning_rate": 0.00012502011263073207,
|
| 939 |
+
"loss": 1.4723,
|
| 940 |
+
"step": 730
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.77,
|
| 944 |
+
"learning_rate": 0.0001238133547868061,
|
| 945 |
+
"loss": 1.4487,
|
| 946 |
+
"step": 735
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 1.78,
|
| 950 |
+
"learning_rate": 0.00012260659694288012,
|
| 951 |
+
"loss": 1.3613,
|
| 952 |
+
"step": 740
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 1.79,
|
| 956 |
+
"learning_rate": 0.00012139983909895413,
|
| 957 |
+
"loss": 1.1906,
|
| 958 |
+
"step": 745
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 1.8,
|
| 962 |
+
"learning_rate": 0.00012019308125502816,
|
| 963 |
+
"loss": 1.4962,
|
| 964 |
+
"step": 750
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 1.81,
|
| 968 |
+
"learning_rate": 0.00011898632341110217,
|
| 969 |
+
"loss": 1.5629,
|
| 970 |
+
"step": 755
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.83,
|
| 974 |
+
"learning_rate": 0.00011777956556717618,
|
| 975 |
+
"loss": 1.3566,
|
| 976 |
+
"step": 760
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.84,
|
| 980 |
+
"learning_rate": 0.00011657280772325019,
|
| 981 |
+
"loss": 1.3442,
|
| 982 |
+
"step": 765
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.85,
|
| 986 |
+
"learning_rate": 0.0001153660498793242,
|
| 987 |
+
"loss": 1.4071,
|
| 988 |
+
"step": 770
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 1.86,
|
| 992 |
+
"learning_rate": 0.00011415929203539821,
|
| 993 |
+
"loss": 1.4331,
|
| 994 |
+
"step": 775
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 1.88,
|
| 998 |
+
"learning_rate": 0.00011295253419147222,
|
| 999 |
+
"loss": 1.3572,
|
| 1000 |
+
"step": 780
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.89,
|
| 1004 |
+
"learning_rate": 0.00011174577634754626,
|
| 1005 |
+
"loss": 1.4009,
|
| 1006 |
+
"step": 785
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 1.9,
|
| 1010 |
+
"learning_rate": 0.00011053901850362027,
|
| 1011 |
+
"loss": 1.47,
|
| 1012 |
+
"step": 790
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.91,
|
| 1016 |
+
"learning_rate": 0.00010933226065969428,
|
| 1017 |
+
"loss": 1.233,
|
| 1018 |
+
"step": 795
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.92,
|
| 1022 |
+
"learning_rate": 0.0001081255028157683,
|
| 1023 |
+
"loss": 1.4593,
|
| 1024 |
+
"step": 800
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.94,
|
| 1028 |
+
"learning_rate": 0.00010691874497184231,
|
| 1029 |
+
"loss": 1.4147,
|
| 1030 |
+
"step": 805
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.95,
|
| 1034 |
+
"learning_rate": 0.00010571198712791632,
|
| 1035 |
+
"loss": 1.2775,
|
| 1036 |
+
"step": 810
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 1.96,
|
| 1040 |
+
"learning_rate": 0.00010450522928399033,
|
| 1041 |
+
"loss": 1.5372,
|
| 1042 |
+
"step": 815
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 1.97,
|
| 1046 |
+
"learning_rate": 0.00010329847144006434,
|
| 1047 |
+
"loss": 1.2994,
|
| 1048 |
+
"step": 820
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 1.98,
|
| 1052 |
+
"learning_rate": 0.00010209171359613835,
|
| 1053 |
+
"loss": 1.3639,
|
| 1054 |
+
"step": 825
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 2.0,
|
| 1058 |
+
"learning_rate": 0.00010088495575221239,
|
| 1059 |
+
"loss": 1.4629,
|
| 1060 |
+
"step": 830
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.0,
|
| 1064 |
+
"eval_loss": 1.5735822916030884,
|
| 1065 |
+
"eval_runtime": 13.3179,
|
| 1066 |
+
"eval_samples_per_second": 15.693,
|
| 1067 |
+
"eval_steps_per_second": 2.027,
|
| 1068 |
+
"step": 832
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 2.01,
|
| 1072 |
+
"learning_rate": 9.96781979082864e-05,
|
| 1073 |
+
"loss": 1.3559,
|
| 1074 |
+
"step": 835
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.02,
|
| 1078 |
+
"learning_rate": 9.847144006436041e-05,
|
| 1079 |
+
"loss": 1.3924,
|
| 1080 |
+
"step": 840
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.03,
|
| 1084 |
+
"learning_rate": 9.726468222043442e-05,
|
| 1085 |
+
"loss": 1.2349,
|
| 1086 |
+
"step": 845
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 2.04,
|
| 1090 |
+
"learning_rate": 9.605792437650844e-05,
|
| 1091 |
+
"loss": 1.1467,
|
| 1092 |
+
"step": 850
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 2.06,
|
| 1096 |
+
"learning_rate": 9.485116653258245e-05,
|
| 1097 |
+
"loss": 1.3251,
|
| 1098 |
+
"step": 855
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.07,
|
| 1102 |
+
"learning_rate": 9.364440868865646e-05,
|
| 1103 |
+
"loss": 1.1517,
|
| 1104 |
+
"step": 860
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 2.08,
|
| 1108 |
+
"learning_rate": 9.243765084473048e-05,
|
| 1109 |
+
"loss": 1.4388,
|
| 1110 |
+
"step": 865
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.09,
|
| 1114 |
+
"learning_rate": 9.12308930008045e-05,
|
| 1115 |
+
"loss": 1.5138,
|
| 1116 |
+
"step": 870
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.1,
|
| 1120 |
+
"learning_rate": 9.002413515687852e-05,
|
| 1121 |
+
"loss": 1.4948,
|
| 1122 |
+
"step": 875
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.12,
|
| 1126 |
+
"learning_rate": 8.881737731295253e-05,
|
| 1127 |
+
"loss": 1.2988,
|
| 1128 |
+
"step": 880
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 2.13,
|
| 1132 |
+
"learning_rate": 8.761061946902654e-05,
|
| 1133 |
+
"loss": 1.226,
|
| 1134 |
+
"step": 885
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 2.14,
|
| 1138 |
+
"learning_rate": 8.640386162510055e-05,
|
| 1139 |
+
"loss": 1.1645,
|
| 1140 |
+
"step": 890
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.15,
|
| 1144 |
+
"learning_rate": 8.519710378117456e-05,
|
| 1145 |
+
"loss": 1.278,
|
| 1146 |
+
"step": 895
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.16,
|
| 1150 |
+
"learning_rate": 8.399034593724859e-05,
|
| 1151 |
+
"loss": 1.3873,
|
| 1152 |
+
"step": 900
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 2.18,
|
| 1156 |
+
"learning_rate": 8.27835880933226e-05,
|
| 1157 |
+
"loss": 1.231,
|
| 1158 |
+
"step": 905
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.19,
|
| 1162 |
+
"learning_rate": 8.157683024939661e-05,
|
| 1163 |
+
"loss": 1.401,
|
| 1164 |
+
"step": 910
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.2,
|
| 1168 |
+
"learning_rate": 8.037007240547062e-05,
|
| 1169 |
+
"loss": 1.4372,
|
| 1170 |
+
"step": 915
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 2.21,
|
| 1174 |
+
"learning_rate": 7.916331456154465e-05,
|
| 1175 |
+
"loss": 1.3624,
|
| 1176 |
+
"step": 920
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 2.22,
|
| 1180 |
+
"learning_rate": 7.795655671761866e-05,
|
| 1181 |
+
"loss": 1.5242,
|
| 1182 |
+
"step": 925
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.24,
|
| 1186 |
+
"learning_rate": 7.674979887369267e-05,
|
| 1187 |
+
"loss": 1.3661,
|
| 1188 |
+
"step": 930
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 2.25,
|
| 1192 |
+
"learning_rate": 7.554304102976669e-05,
|
| 1193 |
+
"loss": 1.3608,
|
| 1194 |
+
"step": 935
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 2.25,
|
| 1198 |
+
"eval_loss": 1.582654356956482,
|
| 1199 |
+
"eval_runtime": 13.332,
|
| 1200 |
+
"eval_samples_per_second": 15.677,
|
| 1201 |
+
"eval_steps_per_second": 2.025,
|
| 1202 |
+
"step": 936
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.26,
|
| 1206 |
+
"learning_rate": 7.43362831858407e-05,
|
| 1207 |
+
"loss": 1.476,
|
| 1208 |
+
"step": 940
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.27,
|
| 1212 |
+
"learning_rate": 7.312952534191472e-05,
|
| 1213 |
+
"loss": 1.3051,
|
| 1214 |
+
"step": 945
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.28,
|
| 1218 |
+
"learning_rate": 7.192276749798874e-05,
|
| 1219 |
+
"loss": 1.418,
|
| 1220 |
+
"step": 950
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.3,
|
| 1224 |
+
"learning_rate": 7.071600965406275e-05,
|
| 1225 |
+
"loss": 1.1781,
|
| 1226 |
+
"step": 955
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 2.31,
|
| 1230 |
+
"learning_rate": 6.950925181013676e-05,
|
| 1231 |
+
"loss": 1.4319,
|
| 1232 |
+
"step": 960
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.32,
|
| 1236 |
+
"learning_rate": 6.830249396621077e-05,
|
| 1237 |
+
"loss": 1.3059,
|
| 1238 |
+
"step": 965
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.33,
|
| 1242 |
+
"learning_rate": 6.709573612228479e-05,
|
| 1243 |
+
"loss": 1.0836,
|
| 1244 |
+
"step": 970
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.34,
|
| 1248 |
+
"learning_rate": 6.588897827835881e-05,
|
| 1249 |
+
"loss": 1.376,
|
| 1250 |
+
"step": 975
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.36,
|
| 1254 |
+
"learning_rate": 6.468222043443282e-05,
|
| 1255 |
+
"loss": 1.5022,
|
| 1256 |
+
"step": 980
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.37,
|
| 1260 |
+
"learning_rate": 6.347546259050683e-05,
|
| 1261 |
+
"loss": 1.2871,
|
| 1262 |
+
"step": 985
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.38,
|
| 1266 |
+
"learning_rate": 6.226870474658084e-05,
|
| 1267 |
+
"loss": 1.2777,
|
| 1268 |
+
"step": 990
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 2.39,
|
| 1272 |
+
"learning_rate": 6.106194690265487e-05,
|
| 1273 |
+
"loss": 1.5625,
|
| 1274 |
+
"step": 995
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.4,
|
| 1278 |
+
"learning_rate": 5.985518905872888e-05,
|
| 1279 |
+
"loss": 1.087,
|
| 1280 |
+
"step": 1000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.42,
|
| 1284 |
+
"learning_rate": 5.864843121480289e-05,
|
| 1285 |
+
"loss": 1.3169,
|
| 1286 |
+
"step": 1005
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 2.43,
|
| 1290 |
+
"learning_rate": 5.74416733708769e-05,
|
| 1291 |
+
"loss": 1.3657,
|
| 1292 |
+
"step": 1010
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 2.44,
|
| 1296 |
+
"learning_rate": 5.6234915526950926e-05,
|
| 1297 |
+
"loss": 1.3917,
|
| 1298 |
+
"step": 1015
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.45,
|
| 1302 |
+
"learning_rate": 5.502815768302494e-05,
|
| 1303 |
+
"loss": 1.3179,
|
| 1304 |
+
"step": 1020
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.46,
|
| 1308 |
+
"learning_rate": 5.382139983909895e-05,
|
| 1309 |
+
"loss": 1.4162,
|
| 1310 |
+
"step": 1025
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 2.48,
|
| 1314 |
+
"learning_rate": 5.261464199517296e-05,
|
| 1315 |
+
"loss": 1.3564,
|
| 1316 |
+
"step": 1030
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.49,
|
| 1320 |
+
"learning_rate": 5.140788415124698e-05,
|
| 1321 |
+
"loss": 1.276,
|
| 1322 |
+
"step": 1035
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.5,
|
| 1326 |
+
"learning_rate": 5.0201126307320996e-05,
|
| 1327 |
+
"loss": 1.3666,
|
| 1328 |
+
"step": 1040
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 2.5,
|
| 1332 |
+
"eval_loss": 1.5801613330841064,
|
| 1333 |
+
"eval_runtime": 13.3272,
|
| 1334 |
+
"eval_samples_per_second": 15.682,
|
| 1335 |
+
"eval_steps_per_second": 2.026,
|
| 1336 |
+
"step": 1040
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.51,
|
| 1340 |
+
"learning_rate": 4.899436846339501e-05,
|
| 1341 |
+
"loss": 1.4251,
|
| 1342 |
+
"step": 1045
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 2.52,
|
| 1346 |
+
"learning_rate": 4.778761061946902e-05,
|
| 1347 |
+
"loss": 1.4518,
|
| 1348 |
+
"step": 1050
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 2.54,
|
| 1352 |
+
"learning_rate": 4.6580852775543036e-05,
|
| 1353 |
+
"loss": 1.3541,
|
| 1354 |
+
"step": 1055
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.55,
|
| 1358 |
+
"learning_rate": 4.5374094931617054e-05,
|
| 1359 |
+
"loss": 1.249,
|
| 1360 |
+
"step": 1060
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.56,
|
| 1364 |
+
"learning_rate": 4.4167337087691066e-05,
|
| 1365 |
+
"loss": 1.3855,
|
| 1366 |
+
"step": 1065
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 2.57,
|
| 1370 |
+
"learning_rate": 4.2960579243765084e-05,
|
| 1371 |
+
"loss": 1.2707,
|
| 1372 |
+
"step": 1070
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 2.58,
|
| 1376 |
+
"learning_rate": 4.1753821399839095e-05,
|
| 1377 |
+
"loss": 1.2083,
|
| 1378 |
+
"step": 1075
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 2.6,
|
| 1382 |
+
"learning_rate": 4.0547063555913106e-05,
|
| 1383 |
+
"loss": 1.4777,
|
| 1384 |
+
"step": 1080
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 2.61,
|
| 1388 |
+
"learning_rate": 3.934030571198713e-05,
|
| 1389 |
+
"loss": 1.4348,
|
| 1390 |
+
"step": 1085
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 2.62,
|
| 1394 |
+
"learning_rate": 3.813354786806114e-05,
|
| 1395 |
+
"loss": 1.2354,
|
| 1396 |
+
"step": 1090
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.63,
|
| 1400 |
+
"learning_rate": 3.692679002413515e-05,
|
| 1401 |
+
"loss": 1.294,
|
| 1402 |
+
"step": 1095
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.64,
|
| 1406 |
+
"learning_rate": 3.572003218020917e-05,
|
| 1407 |
+
"loss": 1.3231,
|
| 1408 |
+
"step": 1100
|
| 1409 |
+
}
|
| 1410 |
+
],
|
| 1411 |
+
"max_steps": 1248,
|
| 1412 |
+
"num_train_epochs": 3,
|
| 1413 |
+
"total_flos": 878022893054208.0,
|
| 1414 |
+
"trial_name": null,
|
| 1415 |
+
"trial_params": null
|
| 1416 |
+
}
|
checkpoint-1100/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7341502d367cfe11875e6bc3d9de5bf48aed11642f79f7fa57ffc1e9a155b8c9
|
| 3 |
+
size 3963
|
checkpoint-1150/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ec5ff5e2943902438bbf5e188a447326c5468d88dbd9a771297d4dde3da1fb6
|
| 3 |
+
size 18928005
|
checkpoint-1150/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ac4a9a603918c1297f8af974e6ac27fea4755721ba7c19f866cdf095fc73972
|
| 3 |
+
size 9460849
|
checkpoint-1150/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:34825cd01382091c73aae3705ab9e695bc6b93b5aa004cefb7d3de48aa5774a7
|
| 3 |
+
size 14575
|
checkpoint-1150/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2e969c33b540ac916977239b4ed9cece9307f2cb03ce481a5a28a42c1618d6f1
|
| 3 |
+
size 557
|
checkpoint-1150/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d67ae71252836f3c721e084aed10032f9ec860cefcd9a01d2dff545c9e738115
|
| 3 |
+
size 627
|
checkpoint-1150/trainer_state.json
ADDED
|
@@ -0,0 +1,1484 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.7644230769230766,
|
| 5 |
+
"global_step": 1150,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 0.0003,
|
| 13 |
+
"loss": 2.0952,
|
| 14 |
+
"step": 5
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 0.000298793242156074,
|
| 19 |
+
"loss": 2.2593,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 0.000297586484312148,
|
| 25 |
+
"loss": 2.1623,
|
| 26 |
+
"step": 15
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 0.000296379726468222,
|
| 31 |
+
"loss": 1.9038,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 0.000295172968624296,
|
| 37 |
+
"loss": 1.8576,
|
| 38 |
+
"step": 25
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 0.00029396621078037004,
|
| 43 |
+
"loss": 1.8397,
|
| 44 |
+
"step": 30
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08,
|
| 48 |
+
"learning_rate": 0.00029275945293644407,
|
| 49 |
+
"loss": 1.8032,
|
| 50 |
+
"step": 35
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.1,
|
| 54 |
+
"learning_rate": 0.00029155269509251804,
|
| 55 |
+
"loss": 1.7281,
|
| 56 |
+
"step": 40
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.11,
|
| 60 |
+
"learning_rate": 0.00029034593724859206,
|
| 61 |
+
"loss": 1.7097,
|
| 62 |
+
"step": 45
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.12,
|
| 66 |
+
"learning_rate": 0.00028913917940466614,
|
| 67 |
+
"loss": 1.7421,
|
| 68 |
+
"step": 50
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.13,
|
| 72 |
+
"learning_rate": 0.0002879324215607401,
|
| 73 |
+
"loss": 1.6861,
|
| 74 |
+
"step": 55
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.14,
|
| 78 |
+
"learning_rate": 0.00028672566371681414,
|
| 79 |
+
"loss": 1.6848,
|
| 80 |
+
"step": 60
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16,
|
| 84 |
+
"learning_rate": 0.00028551890587288816,
|
| 85 |
+
"loss": 1.5163,
|
| 86 |
+
"step": 65
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.17,
|
| 90 |
+
"learning_rate": 0.0002843121480289622,
|
| 91 |
+
"loss": 1.7266,
|
| 92 |
+
"step": 70
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.18,
|
| 96 |
+
"learning_rate": 0.0002831053901850362,
|
| 97 |
+
"loss": 1.5888,
|
| 98 |
+
"step": 75
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.19,
|
| 102 |
+
"learning_rate": 0.0002818986323411102,
|
| 103 |
+
"loss": 1.5463,
|
| 104 |
+
"step": 80
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.2,
|
| 108 |
+
"learning_rate": 0.0002806918744971842,
|
| 109 |
+
"loss": 1.4611,
|
| 110 |
+
"step": 85
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.22,
|
| 114 |
+
"learning_rate": 0.00027948511665325823,
|
| 115 |
+
"loss": 1.5554,
|
| 116 |
+
"step": 90
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.23,
|
| 120 |
+
"learning_rate": 0.00027827835880933226,
|
| 121 |
+
"loss": 1.5481,
|
| 122 |
+
"step": 95
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.24,
|
| 126 |
+
"learning_rate": 0.0002770716009654062,
|
| 127 |
+
"loss": 1.4464,
|
| 128 |
+
"step": 100
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.25,
|
| 132 |
+
"eval_loss": 1.6602041721343994,
|
| 133 |
+
"eval_runtime": 25.516,
|
| 134 |
+
"eval_samples_per_second": 8.191,
|
| 135 |
+
"eval_steps_per_second": 1.058,
|
| 136 |
+
"step": 104
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.25,
|
| 140 |
+
"learning_rate": 0.00027586484312148025,
|
| 141 |
+
"loss": 1.6648,
|
| 142 |
+
"step": 105
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.26,
|
| 146 |
+
"learning_rate": 0.0002746580852775543,
|
| 147 |
+
"loss": 1.6464,
|
| 148 |
+
"step": 110
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.28,
|
| 152 |
+
"learning_rate": 0.0002734513274336283,
|
| 153 |
+
"loss": 1.524,
|
| 154 |
+
"step": 115
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.29,
|
| 158 |
+
"learning_rate": 0.0002722445695897023,
|
| 159 |
+
"loss": 1.4791,
|
| 160 |
+
"step": 120
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.3,
|
| 164 |
+
"learning_rate": 0.0002710378117457763,
|
| 165 |
+
"loss": 1.5162,
|
| 166 |
+
"step": 125
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.31,
|
| 170 |
+
"learning_rate": 0.0002698310539018503,
|
| 171 |
+
"loss": 1.6636,
|
| 172 |
+
"step": 130
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.32,
|
| 176 |
+
"learning_rate": 0.00026862429605792435,
|
| 177 |
+
"loss": 1.7803,
|
| 178 |
+
"step": 135
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.34,
|
| 182 |
+
"learning_rate": 0.00026741753821399837,
|
| 183 |
+
"loss": 1.7384,
|
| 184 |
+
"step": 140
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.35,
|
| 188 |
+
"learning_rate": 0.0002662107803700724,
|
| 189 |
+
"loss": 1.5462,
|
| 190 |
+
"step": 145
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.36,
|
| 194 |
+
"learning_rate": 0.0002650040225261464,
|
| 195 |
+
"loss": 1.5802,
|
| 196 |
+
"step": 150
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.37,
|
| 200 |
+
"learning_rate": 0.00026379726468222045,
|
| 201 |
+
"loss": 1.4942,
|
| 202 |
+
"step": 155
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.38,
|
| 206 |
+
"learning_rate": 0.0002625905068382944,
|
| 207 |
+
"loss": 1.601,
|
| 208 |
+
"step": 160
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.4,
|
| 212 |
+
"learning_rate": 0.00026138374899436844,
|
| 213 |
+
"loss": 1.5822,
|
| 214 |
+
"step": 165
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.41,
|
| 218 |
+
"learning_rate": 0.00026017699115044247,
|
| 219 |
+
"loss": 1.5071,
|
| 220 |
+
"step": 170
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.42,
|
| 224 |
+
"learning_rate": 0.0002589702333065165,
|
| 225 |
+
"loss": 1.5926,
|
| 226 |
+
"step": 175
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.43,
|
| 230 |
+
"learning_rate": 0.00025776347546259046,
|
| 231 |
+
"loss": 1.4206,
|
| 232 |
+
"step": 180
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.44,
|
| 236 |
+
"learning_rate": 0.0002565567176186645,
|
| 237 |
+
"loss": 1.5317,
|
| 238 |
+
"step": 185
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.46,
|
| 242 |
+
"learning_rate": 0.0002553499597747385,
|
| 243 |
+
"loss": 1.4862,
|
| 244 |
+
"step": 190
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.47,
|
| 248 |
+
"learning_rate": 0.00025414320193081253,
|
| 249 |
+
"loss": 1.4387,
|
| 250 |
+
"step": 195
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.48,
|
| 254 |
+
"learning_rate": 0.00025293644408688656,
|
| 255 |
+
"loss": 1.5276,
|
| 256 |
+
"step": 200
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.49,
|
| 260 |
+
"learning_rate": 0.00025172968624296053,
|
| 261 |
+
"loss": 1.55,
|
| 262 |
+
"step": 205
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.5,
|
| 266 |
+
"eval_loss": 1.6166218519210815,
|
| 267 |
+
"eval_runtime": 13.3266,
|
| 268 |
+
"eval_samples_per_second": 15.683,
|
| 269 |
+
"eval_steps_per_second": 2.026,
|
| 270 |
+
"step": 208
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.5,
|
| 274 |
+
"learning_rate": 0.00025052292839903455,
|
| 275 |
+
"loss": 1.5525,
|
| 276 |
+
"step": 210
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.52,
|
| 280 |
+
"learning_rate": 0.0002493161705551086,
|
| 281 |
+
"loss": 1.5591,
|
| 282 |
+
"step": 215
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.53,
|
| 286 |
+
"learning_rate": 0.0002481094127111826,
|
| 287 |
+
"loss": 1.529,
|
| 288 |
+
"step": 220
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.54,
|
| 292 |
+
"learning_rate": 0.0002469026548672566,
|
| 293 |
+
"loss": 1.7152,
|
| 294 |
+
"step": 225
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.55,
|
| 298 |
+
"learning_rate": 0.00024569589702333065,
|
| 299 |
+
"loss": 1.7493,
|
| 300 |
+
"step": 230
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.56,
|
| 304 |
+
"learning_rate": 0.0002444891391794047,
|
| 305 |
+
"loss": 1.5288,
|
| 306 |
+
"step": 235
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.58,
|
| 310 |
+
"learning_rate": 0.00024328238133547868,
|
| 311 |
+
"loss": 1.5651,
|
| 312 |
+
"step": 240
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.59,
|
| 316 |
+
"learning_rate": 0.00024207562349155267,
|
| 317 |
+
"loss": 1.5144,
|
| 318 |
+
"step": 245
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6,
|
| 322 |
+
"learning_rate": 0.0002408688656476267,
|
| 323 |
+
"loss": 1.5334,
|
| 324 |
+
"step": 250
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.61,
|
| 328 |
+
"learning_rate": 0.0002396621078037007,
|
| 329 |
+
"loss": 1.5988,
|
| 330 |
+
"step": 255
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.62,
|
| 334 |
+
"learning_rate": 0.00023845534995977472,
|
| 335 |
+
"loss": 1.6401,
|
| 336 |
+
"step": 260
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.64,
|
| 340 |
+
"learning_rate": 0.00023724859211584875,
|
| 341 |
+
"loss": 1.5929,
|
| 342 |
+
"step": 265
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.65,
|
| 346 |
+
"learning_rate": 0.00023604183427192274,
|
| 347 |
+
"loss": 1.3948,
|
| 348 |
+
"step": 270
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.66,
|
| 352 |
+
"learning_rate": 0.00023483507642799677,
|
| 353 |
+
"loss": 1.221,
|
| 354 |
+
"step": 275
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.67,
|
| 358 |
+
"learning_rate": 0.00023362831858407077,
|
| 359 |
+
"loss": 1.5604,
|
| 360 |
+
"step": 280
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.69,
|
| 364 |
+
"learning_rate": 0.0002324215607401448,
|
| 365 |
+
"loss": 1.505,
|
| 366 |
+
"step": 285
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.7,
|
| 370 |
+
"learning_rate": 0.0002312148028962188,
|
| 371 |
+
"loss": 1.5008,
|
| 372 |
+
"step": 290
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.71,
|
| 376 |
+
"learning_rate": 0.0002300080450522928,
|
| 377 |
+
"loss": 1.5176,
|
| 378 |
+
"step": 295
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.72,
|
| 382 |
+
"learning_rate": 0.0002288012872083668,
|
| 383 |
+
"loss": 1.4227,
|
| 384 |
+
"step": 300
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.73,
|
| 388 |
+
"learning_rate": 0.00022759452936444084,
|
| 389 |
+
"loss": 1.1778,
|
| 390 |
+
"step": 305
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.75,
|
| 394 |
+
"learning_rate": 0.00022638777152051486,
|
| 395 |
+
"loss": 1.5052,
|
| 396 |
+
"step": 310
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.75,
|
| 400 |
+
"eval_loss": 1.5979019403457642,
|
| 401 |
+
"eval_runtime": 13.342,
|
| 402 |
+
"eval_samples_per_second": 15.665,
|
| 403 |
+
"eval_steps_per_second": 2.024,
|
| 404 |
+
"step": 312
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.76,
|
| 408 |
+
"learning_rate": 0.00022518101367658886,
|
| 409 |
+
"loss": 1.5228,
|
| 410 |
+
"step": 315
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.77,
|
| 414 |
+
"learning_rate": 0.0002239742558326629,
|
| 415 |
+
"loss": 1.5385,
|
| 416 |
+
"step": 320
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.78,
|
| 420 |
+
"learning_rate": 0.0002227674979887369,
|
| 421 |
+
"loss": 1.5072,
|
| 422 |
+
"step": 325
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.79,
|
| 426 |
+
"learning_rate": 0.00022156074014481093,
|
| 427 |
+
"loss": 1.5032,
|
| 428 |
+
"step": 330
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.81,
|
| 432 |
+
"learning_rate": 0.00022035398230088496,
|
| 433 |
+
"loss": 1.5981,
|
| 434 |
+
"step": 335
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.82,
|
| 438 |
+
"learning_rate": 0.00021914722445695895,
|
| 439 |
+
"loss": 1.422,
|
| 440 |
+
"step": 340
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.83,
|
| 444 |
+
"learning_rate": 0.00021794046661303298,
|
| 445 |
+
"loss": 1.4948,
|
| 446 |
+
"step": 345
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.84,
|
| 450 |
+
"learning_rate": 0.00021673370876910698,
|
| 451 |
+
"loss": 1.4421,
|
| 452 |
+
"step": 350
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.85,
|
| 456 |
+
"learning_rate": 0.000215526950925181,
|
| 457 |
+
"loss": 1.6086,
|
| 458 |
+
"step": 355
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.87,
|
| 462 |
+
"learning_rate": 0.000214320193081255,
|
| 463 |
+
"loss": 1.3874,
|
| 464 |
+
"step": 360
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.88,
|
| 468 |
+
"learning_rate": 0.00021311343523732902,
|
| 469 |
+
"loss": 1.2493,
|
| 470 |
+
"step": 365
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.89,
|
| 474 |
+
"learning_rate": 0.00021190667739340302,
|
| 475 |
+
"loss": 1.5565,
|
| 476 |
+
"step": 370
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.9,
|
| 480 |
+
"learning_rate": 0.00021069991954947705,
|
| 481 |
+
"loss": 1.4739,
|
| 482 |
+
"step": 375
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.91,
|
| 486 |
+
"learning_rate": 0.00020949316170555107,
|
| 487 |
+
"loss": 1.5288,
|
| 488 |
+
"step": 380
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.93,
|
| 492 |
+
"learning_rate": 0.00020828640386162507,
|
| 493 |
+
"loss": 1.4837,
|
| 494 |
+
"step": 385
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.94,
|
| 498 |
+
"learning_rate": 0.0002070796460176991,
|
| 499 |
+
"loss": 1.5009,
|
| 500 |
+
"step": 390
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.95,
|
| 504 |
+
"learning_rate": 0.0002058728881737731,
|
| 505 |
+
"loss": 1.3469,
|
| 506 |
+
"step": 395
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.96,
|
| 510 |
+
"learning_rate": 0.00020466613032984712,
|
| 511 |
+
"loss": 1.3949,
|
| 512 |
+
"step": 400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.97,
|
| 516 |
+
"learning_rate": 0.00020345937248592117,
|
| 517 |
+
"loss": 1.3688,
|
| 518 |
+
"step": 405
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.99,
|
| 522 |
+
"learning_rate": 0.00020225261464199517,
|
| 523 |
+
"loss": 1.6629,
|
| 524 |
+
"step": 410
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.0,
|
| 528 |
+
"learning_rate": 0.0002010458567980692,
|
| 529 |
+
"loss": 1.5395,
|
| 530 |
+
"step": 415
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.0,
|
| 534 |
+
"eval_loss": 1.589026927947998,
|
| 535 |
+
"eval_runtime": 13.34,
|
| 536 |
+
"eval_samples_per_second": 15.667,
|
| 537 |
+
"eval_steps_per_second": 2.024,
|
| 538 |
+
"step": 416
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.01,
|
| 542 |
+
"learning_rate": 0.0001998390989541432,
|
| 543 |
+
"loss": 1.5794,
|
| 544 |
+
"step": 420
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.02,
|
| 548 |
+
"learning_rate": 0.0001986323411102172,
|
| 549 |
+
"loss": 1.4746,
|
| 550 |
+
"step": 425
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.03,
|
| 554 |
+
"learning_rate": 0.0001974255832662912,
|
| 555 |
+
"loss": 1.3205,
|
| 556 |
+
"step": 430
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.05,
|
| 560 |
+
"learning_rate": 0.00019621882542236523,
|
| 561 |
+
"loss": 1.491,
|
| 562 |
+
"step": 435
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.06,
|
| 566 |
+
"learning_rate": 0.00019501206757843926,
|
| 567 |
+
"loss": 1.4803,
|
| 568 |
+
"step": 440
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.07,
|
| 572 |
+
"learning_rate": 0.00019380530973451326,
|
| 573 |
+
"loss": 1.3436,
|
| 574 |
+
"step": 445
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.08,
|
| 578 |
+
"learning_rate": 0.00019259855189058728,
|
| 579 |
+
"loss": 1.3893,
|
| 580 |
+
"step": 450
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.09,
|
| 584 |
+
"learning_rate": 0.00019139179404666128,
|
| 585 |
+
"loss": 1.0474,
|
| 586 |
+
"step": 455
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.11,
|
| 590 |
+
"learning_rate": 0.0001901850362027353,
|
| 591 |
+
"loss": 1.3388,
|
| 592 |
+
"step": 460
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.12,
|
| 596 |
+
"learning_rate": 0.0001889782783588093,
|
| 597 |
+
"loss": 1.4201,
|
| 598 |
+
"step": 465
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.13,
|
| 602 |
+
"learning_rate": 0.00018777152051488333,
|
| 603 |
+
"loss": 1.4805,
|
| 604 |
+
"step": 470
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.14,
|
| 608 |
+
"learning_rate": 0.00018656476267095732,
|
| 609 |
+
"loss": 1.2253,
|
| 610 |
+
"step": 475
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 1.15,
|
| 614 |
+
"learning_rate": 0.00018535800482703135,
|
| 615 |
+
"loss": 1.5064,
|
| 616 |
+
"step": 480
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 1.17,
|
| 620 |
+
"learning_rate": 0.00018415124698310537,
|
| 621 |
+
"loss": 1.3565,
|
| 622 |
+
"step": 485
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.18,
|
| 626 |
+
"learning_rate": 0.00018294448913917937,
|
| 627 |
+
"loss": 1.3976,
|
| 628 |
+
"step": 490
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.19,
|
| 632 |
+
"learning_rate": 0.00018173773129525342,
|
| 633 |
+
"loss": 1.5882,
|
| 634 |
+
"step": 495
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.2,
|
| 638 |
+
"learning_rate": 0.00018053097345132742,
|
| 639 |
+
"loss": 1.4698,
|
| 640 |
+
"step": 500
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.21,
|
| 644 |
+
"learning_rate": 0.00017932421560740145,
|
| 645 |
+
"loss": 1.5326,
|
| 646 |
+
"step": 505
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.23,
|
| 650 |
+
"learning_rate": 0.00017811745776347547,
|
| 651 |
+
"loss": 1.5659,
|
| 652 |
+
"step": 510
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 1.24,
|
| 656 |
+
"learning_rate": 0.00017691069991954947,
|
| 657 |
+
"loss": 1.4489,
|
| 658 |
+
"step": 515
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 1.25,
|
| 662 |
+
"learning_rate": 0.0001757039420756235,
|
| 663 |
+
"loss": 1.5312,
|
| 664 |
+
"step": 520
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.25,
|
| 668 |
+
"eval_loss": 1.5857353210449219,
|
| 669 |
+
"eval_runtime": 13.3247,
|
| 670 |
+
"eval_samples_per_second": 15.685,
|
| 671 |
+
"eval_steps_per_second": 2.026,
|
| 672 |
+
"step": 520
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.26,
|
| 676 |
+
"learning_rate": 0.0001744971842316975,
|
| 677 |
+
"loss": 1.3595,
|
| 678 |
+
"step": 525
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.27,
|
| 682 |
+
"learning_rate": 0.00017329042638777152,
|
| 683 |
+
"loss": 1.1543,
|
| 684 |
+
"step": 530
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.29,
|
| 688 |
+
"learning_rate": 0.0001720836685438455,
|
| 689 |
+
"loss": 1.4568,
|
| 690 |
+
"step": 535
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.3,
|
| 694 |
+
"learning_rate": 0.00017087691069991954,
|
| 695 |
+
"loss": 1.3682,
|
| 696 |
+
"step": 540
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.31,
|
| 700 |
+
"learning_rate": 0.00016967015285599354,
|
| 701 |
+
"loss": 1.5323,
|
| 702 |
+
"step": 545
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.32,
|
| 706 |
+
"learning_rate": 0.00016846339501206756,
|
| 707 |
+
"loss": 1.6644,
|
| 708 |
+
"step": 550
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 1.33,
|
| 712 |
+
"learning_rate": 0.00016725663716814158,
|
| 713 |
+
"loss": 1.5735,
|
| 714 |
+
"step": 555
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 1.35,
|
| 718 |
+
"learning_rate": 0.00016604987932421558,
|
| 719 |
+
"loss": 1.4759,
|
| 720 |
+
"step": 560
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 1.36,
|
| 724 |
+
"learning_rate": 0.0001648431214802896,
|
| 725 |
+
"loss": 1.3774,
|
| 726 |
+
"step": 565
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.37,
|
| 730 |
+
"learning_rate": 0.0001636363636363636,
|
| 731 |
+
"loss": 1.3321,
|
| 732 |
+
"step": 570
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.38,
|
| 736 |
+
"learning_rate": 0.00016242960579243763,
|
| 737 |
+
"loss": 1.2843,
|
| 738 |
+
"step": 575
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.39,
|
| 742 |
+
"learning_rate": 0.00016122284794851163,
|
| 743 |
+
"loss": 1.4636,
|
| 744 |
+
"step": 580
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.41,
|
| 748 |
+
"learning_rate": 0.00016001609010458568,
|
| 749 |
+
"loss": 1.4082,
|
| 750 |
+
"step": 585
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 1.42,
|
| 754 |
+
"learning_rate": 0.0001588093322606597,
|
| 755 |
+
"loss": 1.3848,
|
| 756 |
+
"step": 590
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 1.43,
|
| 760 |
+
"learning_rate": 0.0001576025744167337,
|
| 761 |
+
"loss": 1.5396,
|
| 762 |
+
"step": 595
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1.44,
|
| 766 |
+
"learning_rate": 0.00015639581657280773,
|
| 767 |
+
"loss": 1.4889,
|
| 768 |
+
"step": 600
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 1.45,
|
| 772 |
+
"learning_rate": 0.00015518905872888172,
|
| 773 |
+
"loss": 1.4019,
|
| 774 |
+
"step": 605
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.47,
|
| 778 |
+
"learning_rate": 0.00015398230088495575,
|
| 779 |
+
"loss": 1.3788,
|
| 780 |
+
"step": 610
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.48,
|
| 784 |
+
"learning_rate": 0.00015277554304102975,
|
| 785 |
+
"loss": 1.409,
|
| 786 |
+
"step": 615
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.49,
|
| 790 |
+
"learning_rate": 0.00015156878519710377,
|
| 791 |
+
"loss": 1.5675,
|
| 792 |
+
"step": 620
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 1.5,
|
| 796 |
+
"eval_loss": 1.584699273109436,
|
| 797 |
+
"eval_runtime": 13.3509,
|
| 798 |
+
"eval_samples_per_second": 15.654,
|
| 799 |
+
"eval_steps_per_second": 2.022,
|
| 800 |
+
"step": 624
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.5,
|
| 804 |
+
"learning_rate": 0.0001503620273531778,
|
| 805 |
+
"loss": 1.3176,
|
| 806 |
+
"step": 625
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.51,
|
| 810 |
+
"learning_rate": 0.0001491552695092518,
|
| 811 |
+
"loss": 1.3114,
|
| 812 |
+
"step": 630
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.53,
|
| 816 |
+
"learning_rate": 0.00014794851166532582,
|
| 817 |
+
"loss": 1.4261,
|
| 818 |
+
"step": 635
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.54,
|
| 822 |
+
"learning_rate": 0.00014674175382139982,
|
| 823 |
+
"loss": 1.5156,
|
| 824 |
+
"step": 640
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.55,
|
| 828 |
+
"learning_rate": 0.00014553499597747384,
|
| 829 |
+
"loss": 1.3231,
|
| 830 |
+
"step": 645
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.56,
|
| 834 |
+
"learning_rate": 0.00014432823813354787,
|
| 835 |
+
"loss": 1.4353,
|
| 836 |
+
"step": 650
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.57,
|
| 840 |
+
"learning_rate": 0.00014312148028962186,
|
| 841 |
+
"loss": 1.4423,
|
| 842 |
+
"step": 655
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.59,
|
| 846 |
+
"learning_rate": 0.0001419147224456959,
|
| 847 |
+
"loss": 1.1512,
|
| 848 |
+
"step": 660
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.6,
|
| 852 |
+
"learning_rate": 0.0001407079646017699,
|
| 853 |
+
"loss": 1.3562,
|
| 854 |
+
"step": 665
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.61,
|
| 858 |
+
"learning_rate": 0.0001395012067578439,
|
| 859 |
+
"loss": 1.4033,
|
| 860 |
+
"step": 670
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.62,
|
| 864 |
+
"learning_rate": 0.00013829444891391793,
|
| 865 |
+
"loss": 1.5435,
|
| 866 |
+
"step": 675
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.63,
|
| 870 |
+
"learning_rate": 0.00013708769106999193,
|
| 871 |
+
"loss": 1.4543,
|
| 872 |
+
"step": 680
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.65,
|
| 876 |
+
"learning_rate": 0.00013588093322606596,
|
| 877 |
+
"loss": 1.3509,
|
| 878 |
+
"step": 685
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.66,
|
| 882 |
+
"learning_rate": 0.00013467417538213995,
|
| 883 |
+
"loss": 1.4268,
|
| 884 |
+
"step": 690
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.67,
|
| 888 |
+
"learning_rate": 0.000133467417538214,
|
| 889 |
+
"loss": 1.4231,
|
| 890 |
+
"step": 695
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.68,
|
| 894 |
+
"learning_rate": 0.000132260659694288,
|
| 895 |
+
"loss": 1.4727,
|
| 896 |
+
"step": 700
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.69,
|
| 900 |
+
"learning_rate": 0.00013105390185036203,
|
| 901 |
+
"loss": 1.2846,
|
| 902 |
+
"step": 705
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.71,
|
| 906 |
+
"learning_rate": 0.00012984714400643603,
|
| 907 |
+
"loss": 1.3458,
|
| 908 |
+
"step": 710
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 1.72,
|
| 912 |
+
"learning_rate": 0.00012864038616251005,
|
| 913 |
+
"loss": 1.6349,
|
| 914 |
+
"step": 715
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.73,
|
| 918 |
+
"learning_rate": 0.00012743362831858405,
|
| 919 |
+
"loss": 1.4798,
|
| 920 |
+
"step": 720
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.74,
|
| 924 |
+
"learning_rate": 0.00012622687047465807,
|
| 925 |
+
"loss": 1.6157,
|
| 926 |
+
"step": 725
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.75,
|
| 930 |
+
"eval_loss": 1.5761181116104126,
|
| 931 |
+
"eval_runtime": 13.3218,
|
| 932 |
+
"eval_samples_per_second": 15.689,
|
| 933 |
+
"eval_steps_per_second": 2.027,
|
| 934 |
+
"step": 728
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.75,
|
| 938 |
+
"learning_rate": 0.00012502011263073207,
|
| 939 |
+
"loss": 1.4723,
|
| 940 |
+
"step": 730
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.77,
|
| 944 |
+
"learning_rate": 0.0001238133547868061,
|
| 945 |
+
"loss": 1.4487,
|
| 946 |
+
"step": 735
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 1.78,
|
| 950 |
+
"learning_rate": 0.00012260659694288012,
|
| 951 |
+
"loss": 1.3613,
|
| 952 |
+
"step": 740
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 1.79,
|
| 956 |
+
"learning_rate": 0.00012139983909895413,
|
| 957 |
+
"loss": 1.1906,
|
| 958 |
+
"step": 745
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 1.8,
|
| 962 |
+
"learning_rate": 0.00012019308125502816,
|
| 963 |
+
"loss": 1.4962,
|
| 964 |
+
"step": 750
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 1.81,
|
| 968 |
+
"learning_rate": 0.00011898632341110217,
|
| 969 |
+
"loss": 1.5629,
|
| 970 |
+
"step": 755
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.83,
|
| 974 |
+
"learning_rate": 0.00011777956556717618,
|
| 975 |
+
"loss": 1.3566,
|
| 976 |
+
"step": 760
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.84,
|
| 980 |
+
"learning_rate": 0.00011657280772325019,
|
| 981 |
+
"loss": 1.3442,
|
| 982 |
+
"step": 765
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.85,
|
| 986 |
+
"learning_rate": 0.0001153660498793242,
|
| 987 |
+
"loss": 1.4071,
|
| 988 |
+
"step": 770
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 1.86,
|
| 992 |
+
"learning_rate": 0.00011415929203539821,
|
| 993 |
+
"loss": 1.4331,
|
| 994 |
+
"step": 775
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 1.88,
|
| 998 |
+
"learning_rate": 0.00011295253419147222,
|
| 999 |
+
"loss": 1.3572,
|
| 1000 |
+
"step": 780
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.89,
|
| 1004 |
+
"learning_rate": 0.00011174577634754626,
|
| 1005 |
+
"loss": 1.4009,
|
| 1006 |
+
"step": 785
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 1.9,
|
| 1010 |
+
"learning_rate": 0.00011053901850362027,
|
| 1011 |
+
"loss": 1.47,
|
| 1012 |
+
"step": 790
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.91,
|
| 1016 |
+
"learning_rate": 0.00010933226065969428,
|
| 1017 |
+
"loss": 1.233,
|
| 1018 |
+
"step": 795
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.92,
|
| 1022 |
+
"learning_rate": 0.0001081255028157683,
|
| 1023 |
+
"loss": 1.4593,
|
| 1024 |
+
"step": 800
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.94,
|
| 1028 |
+
"learning_rate": 0.00010691874497184231,
|
| 1029 |
+
"loss": 1.4147,
|
| 1030 |
+
"step": 805
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.95,
|
| 1034 |
+
"learning_rate": 0.00010571198712791632,
|
| 1035 |
+
"loss": 1.2775,
|
| 1036 |
+
"step": 810
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 1.96,
|
| 1040 |
+
"learning_rate": 0.00010450522928399033,
|
| 1041 |
+
"loss": 1.5372,
|
| 1042 |
+
"step": 815
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 1.97,
|
| 1046 |
+
"learning_rate": 0.00010329847144006434,
|
| 1047 |
+
"loss": 1.2994,
|
| 1048 |
+
"step": 820
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 1.98,
|
| 1052 |
+
"learning_rate": 0.00010209171359613835,
|
| 1053 |
+
"loss": 1.3639,
|
| 1054 |
+
"step": 825
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 2.0,
|
| 1058 |
+
"learning_rate": 0.00010088495575221239,
|
| 1059 |
+
"loss": 1.4629,
|
| 1060 |
+
"step": 830
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.0,
|
| 1064 |
+
"eval_loss": 1.5735822916030884,
|
| 1065 |
+
"eval_runtime": 13.3179,
|
| 1066 |
+
"eval_samples_per_second": 15.693,
|
| 1067 |
+
"eval_steps_per_second": 2.027,
|
| 1068 |
+
"step": 832
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 2.01,
|
| 1072 |
+
"learning_rate": 9.96781979082864e-05,
|
| 1073 |
+
"loss": 1.3559,
|
| 1074 |
+
"step": 835
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.02,
|
| 1078 |
+
"learning_rate": 9.847144006436041e-05,
|
| 1079 |
+
"loss": 1.3924,
|
| 1080 |
+
"step": 840
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.03,
|
| 1084 |
+
"learning_rate": 9.726468222043442e-05,
|
| 1085 |
+
"loss": 1.2349,
|
| 1086 |
+
"step": 845
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 2.04,
|
| 1090 |
+
"learning_rate": 9.605792437650844e-05,
|
| 1091 |
+
"loss": 1.1467,
|
| 1092 |
+
"step": 850
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 2.06,
|
| 1096 |
+
"learning_rate": 9.485116653258245e-05,
|
| 1097 |
+
"loss": 1.3251,
|
| 1098 |
+
"step": 855
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.07,
|
| 1102 |
+
"learning_rate": 9.364440868865646e-05,
|
| 1103 |
+
"loss": 1.1517,
|
| 1104 |
+
"step": 860
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 2.08,
|
| 1108 |
+
"learning_rate": 9.243765084473048e-05,
|
| 1109 |
+
"loss": 1.4388,
|
| 1110 |
+
"step": 865
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.09,
|
| 1114 |
+
"learning_rate": 9.12308930008045e-05,
|
| 1115 |
+
"loss": 1.5138,
|
| 1116 |
+
"step": 870
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.1,
|
| 1120 |
+
"learning_rate": 9.002413515687852e-05,
|
| 1121 |
+
"loss": 1.4948,
|
| 1122 |
+
"step": 875
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.12,
|
| 1126 |
+
"learning_rate": 8.881737731295253e-05,
|
| 1127 |
+
"loss": 1.2988,
|
| 1128 |
+
"step": 880
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 2.13,
|
| 1132 |
+
"learning_rate": 8.761061946902654e-05,
|
| 1133 |
+
"loss": 1.226,
|
| 1134 |
+
"step": 885
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 2.14,
|
| 1138 |
+
"learning_rate": 8.640386162510055e-05,
|
| 1139 |
+
"loss": 1.1645,
|
| 1140 |
+
"step": 890
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.15,
|
| 1144 |
+
"learning_rate": 8.519710378117456e-05,
|
| 1145 |
+
"loss": 1.278,
|
| 1146 |
+
"step": 895
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.16,
|
| 1150 |
+
"learning_rate": 8.399034593724859e-05,
|
| 1151 |
+
"loss": 1.3873,
|
| 1152 |
+
"step": 900
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 2.18,
|
| 1156 |
+
"learning_rate": 8.27835880933226e-05,
|
| 1157 |
+
"loss": 1.231,
|
| 1158 |
+
"step": 905
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.19,
|
| 1162 |
+
"learning_rate": 8.157683024939661e-05,
|
| 1163 |
+
"loss": 1.401,
|
| 1164 |
+
"step": 910
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.2,
|
| 1168 |
+
"learning_rate": 8.037007240547062e-05,
|
| 1169 |
+
"loss": 1.4372,
|
| 1170 |
+
"step": 915
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 2.21,
|
| 1174 |
+
"learning_rate": 7.916331456154465e-05,
|
| 1175 |
+
"loss": 1.3624,
|
| 1176 |
+
"step": 920
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 2.22,
|
| 1180 |
+
"learning_rate": 7.795655671761866e-05,
|
| 1181 |
+
"loss": 1.5242,
|
| 1182 |
+
"step": 925
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.24,
|
| 1186 |
+
"learning_rate": 7.674979887369267e-05,
|
| 1187 |
+
"loss": 1.3661,
|
| 1188 |
+
"step": 930
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 2.25,
|
| 1192 |
+
"learning_rate": 7.554304102976669e-05,
|
| 1193 |
+
"loss": 1.3608,
|
| 1194 |
+
"step": 935
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 2.25,
|
| 1198 |
+
"eval_loss": 1.582654356956482,
|
| 1199 |
+
"eval_runtime": 13.332,
|
| 1200 |
+
"eval_samples_per_second": 15.677,
|
| 1201 |
+
"eval_steps_per_second": 2.025,
|
| 1202 |
+
"step": 936
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.26,
|
| 1206 |
+
"learning_rate": 7.43362831858407e-05,
|
| 1207 |
+
"loss": 1.476,
|
| 1208 |
+
"step": 940
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.27,
|
| 1212 |
+
"learning_rate": 7.312952534191472e-05,
|
| 1213 |
+
"loss": 1.3051,
|
| 1214 |
+
"step": 945
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.28,
|
| 1218 |
+
"learning_rate": 7.192276749798874e-05,
|
| 1219 |
+
"loss": 1.418,
|
| 1220 |
+
"step": 950
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.3,
|
| 1224 |
+
"learning_rate": 7.071600965406275e-05,
|
| 1225 |
+
"loss": 1.1781,
|
| 1226 |
+
"step": 955
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 2.31,
|
| 1230 |
+
"learning_rate": 6.950925181013676e-05,
|
| 1231 |
+
"loss": 1.4319,
|
| 1232 |
+
"step": 960
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.32,
|
| 1236 |
+
"learning_rate": 6.830249396621077e-05,
|
| 1237 |
+
"loss": 1.3059,
|
| 1238 |
+
"step": 965
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.33,
|
| 1242 |
+
"learning_rate": 6.709573612228479e-05,
|
| 1243 |
+
"loss": 1.0836,
|
| 1244 |
+
"step": 970
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.34,
|
| 1248 |
+
"learning_rate": 6.588897827835881e-05,
|
| 1249 |
+
"loss": 1.376,
|
| 1250 |
+
"step": 975
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.36,
|
| 1254 |
+
"learning_rate": 6.468222043443282e-05,
|
| 1255 |
+
"loss": 1.5022,
|
| 1256 |
+
"step": 980
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.37,
|
| 1260 |
+
"learning_rate": 6.347546259050683e-05,
|
| 1261 |
+
"loss": 1.2871,
|
| 1262 |
+
"step": 985
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.38,
|
| 1266 |
+
"learning_rate": 6.226870474658084e-05,
|
| 1267 |
+
"loss": 1.2777,
|
| 1268 |
+
"step": 990
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 2.39,
|
| 1272 |
+
"learning_rate": 6.106194690265487e-05,
|
| 1273 |
+
"loss": 1.5625,
|
| 1274 |
+
"step": 995
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.4,
|
| 1278 |
+
"learning_rate": 5.985518905872888e-05,
|
| 1279 |
+
"loss": 1.087,
|
| 1280 |
+
"step": 1000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.42,
|
| 1284 |
+
"learning_rate": 5.864843121480289e-05,
|
| 1285 |
+
"loss": 1.3169,
|
| 1286 |
+
"step": 1005
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 2.43,
|
| 1290 |
+
"learning_rate": 5.74416733708769e-05,
|
| 1291 |
+
"loss": 1.3657,
|
| 1292 |
+
"step": 1010
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 2.44,
|
| 1296 |
+
"learning_rate": 5.6234915526950926e-05,
|
| 1297 |
+
"loss": 1.3917,
|
| 1298 |
+
"step": 1015
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.45,
|
| 1302 |
+
"learning_rate": 5.502815768302494e-05,
|
| 1303 |
+
"loss": 1.3179,
|
| 1304 |
+
"step": 1020
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.46,
|
| 1308 |
+
"learning_rate": 5.382139983909895e-05,
|
| 1309 |
+
"loss": 1.4162,
|
| 1310 |
+
"step": 1025
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 2.48,
|
| 1314 |
+
"learning_rate": 5.261464199517296e-05,
|
| 1315 |
+
"loss": 1.3564,
|
| 1316 |
+
"step": 1030
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.49,
|
| 1320 |
+
"learning_rate": 5.140788415124698e-05,
|
| 1321 |
+
"loss": 1.276,
|
| 1322 |
+
"step": 1035
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.5,
|
| 1326 |
+
"learning_rate": 5.0201126307320996e-05,
|
| 1327 |
+
"loss": 1.3666,
|
| 1328 |
+
"step": 1040
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 2.5,
|
| 1332 |
+
"eval_loss": 1.5801613330841064,
|
| 1333 |
+
"eval_runtime": 13.3272,
|
| 1334 |
+
"eval_samples_per_second": 15.682,
|
| 1335 |
+
"eval_steps_per_second": 2.026,
|
| 1336 |
+
"step": 1040
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.51,
|
| 1340 |
+
"learning_rate": 4.899436846339501e-05,
|
| 1341 |
+
"loss": 1.4251,
|
| 1342 |
+
"step": 1045
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 2.52,
|
| 1346 |
+
"learning_rate": 4.778761061946902e-05,
|
| 1347 |
+
"loss": 1.4518,
|
| 1348 |
+
"step": 1050
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 2.54,
|
| 1352 |
+
"learning_rate": 4.6580852775543036e-05,
|
| 1353 |
+
"loss": 1.3541,
|
| 1354 |
+
"step": 1055
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.55,
|
| 1358 |
+
"learning_rate": 4.5374094931617054e-05,
|
| 1359 |
+
"loss": 1.249,
|
| 1360 |
+
"step": 1060
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.56,
|
| 1364 |
+
"learning_rate": 4.4167337087691066e-05,
|
| 1365 |
+
"loss": 1.3855,
|
| 1366 |
+
"step": 1065
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 2.57,
|
| 1370 |
+
"learning_rate": 4.2960579243765084e-05,
|
| 1371 |
+
"loss": 1.2707,
|
| 1372 |
+
"step": 1070
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 2.58,
|
| 1376 |
+
"learning_rate": 4.1753821399839095e-05,
|
| 1377 |
+
"loss": 1.2083,
|
| 1378 |
+
"step": 1075
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 2.6,
|
| 1382 |
+
"learning_rate": 4.0547063555913106e-05,
|
| 1383 |
+
"loss": 1.4777,
|
| 1384 |
+
"step": 1080
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 2.61,
|
| 1388 |
+
"learning_rate": 3.934030571198713e-05,
|
| 1389 |
+
"loss": 1.4348,
|
| 1390 |
+
"step": 1085
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 2.62,
|
| 1394 |
+
"learning_rate": 3.813354786806114e-05,
|
| 1395 |
+
"loss": 1.2354,
|
| 1396 |
+
"step": 1090
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.63,
|
| 1400 |
+
"learning_rate": 3.692679002413515e-05,
|
| 1401 |
+
"loss": 1.294,
|
| 1402 |
+
"step": 1095
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.64,
|
| 1406 |
+
"learning_rate": 3.572003218020917e-05,
|
| 1407 |
+
"loss": 1.3231,
|
| 1408 |
+
"step": 1100
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 2.66,
|
| 1412 |
+
"learning_rate": 3.451327433628318e-05,
|
| 1413 |
+
"loss": 1.2215,
|
| 1414 |
+
"step": 1105
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 2.67,
|
| 1418 |
+
"learning_rate": 3.33065164923572e-05,
|
| 1419 |
+
"loss": 1.1217,
|
| 1420 |
+
"step": 1110
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.68,
|
| 1424 |
+
"learning_rate": 3.209975864843121e-05,
|
| 1425 |
+
"loss": 1.4292,
|
| 1426 |
+
"step": 1115
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 2.69,
|
| 1430 |
+
"learning_rate": 3.089300080450522e-05,
|
| 1431 |
+
"loss": 1.195,
|
| 1432 |
+
"step": 1120
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 2.7,
|
| 1436 |
+
"learning_rate": 2.968624296057924e-05,
|
| 1437 |
+
"loss": 1.3549,
|
| 1438 |
+
"step": 1125
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.72,
|
| 1442 |
+
"learning_rate": 2.8479485116653255e-05,
|
| 1443 |
+
"loss": 1.4274,
|
| 1444 |
+
"step": 1130
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.73,
|
| 1448 |
+
"learning_rate": 2.727272727272727e-05,
|
| 1449 |
+
"loss": 1.4838,
|
| 1450 |
+
"step": 1135
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 2.74,
|
| 1454 |
+
"learning_rate": 2.6065969428801284e-05,
|
| 1455 |
+
"loss": 1.2394,
|
| 1456 |
+
"step": 1140
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 2.75,
|
| 1460 |
+
"eval_loss": 1.5814132690429688,
|
| 1461 |
+
"eval_runtime": 13.3223,
|
| 1462 |
+
"eval_samples_per_second": 15.688,
|
| 1463 |
+
"eval_steps_per_second": 2.027,
|
| 1464 |
+
"step": 1144
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 2.75,
|
| 1468 |
+
"learning_rate": 2.4859211584875302e-05,
|
| 1469 |
+
"loss": 1.2414,
|
| 1470 |
+
"step": 1145
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 2.76,
|
| 1474 |
+
"learning_rate": 2.3652453740949313e-05,
|
| 1475 |
+
"loss": 1.288,
|
| 1476 |
+
"step": 1150
|
| 1477 |
+
}
|
| 1478 |
+
],
|
| 1479 |
+
"max_steps": 1248,
|
| 1480 |
+
"num_train_epochs": 3,
|
| 1481 |
+
"total_flos": 917384766693888.0,
|
| 1482 |
+
"trial_name": null,
|
| 1483 |
+
"trial_params": null
|
| 1484 |
+
}
|
checkpoint-1150/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7341502d367cfe11875e6bc3d9de5bf48aed11642f79f7fa57ffc1e9a155b8c9
|
| 3 |
+
size 3963
|
checkpoint-1200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3dd98b99200c318c28c150c5adb500389b56bdd0c9ff59bfb4dda85b66bf8ffc
|
| 3 |
+
size 18928005
|
checkpoint-1200/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c476fcaf02bb7d989e472eb2a880e8c4690972f6337603c1159e1801932362c4
|
| 3 |
+
size 9460849
|
checkpoint-1200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dae385048a460eb79e49e7471a5465380b51b8590b8fc37a82dec04312fc80e4
|
| 3 |
+
size 14575
|
checkpoint-1200/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:741cefeca9ef427f92406d2d10b81996655e2a9d50eb7aaa9614e6fdd1c9f529
|
| 3 |
+
size 557
|
checkpoint-1200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8425974c187a5d6e9398a3c750aabb2042b12695ecd9408835ba11b1321f97cd
|
| 3 |
+
size 627
|
checkpoint-1200/trainer_state.json
ADDED
|
@@ -0,0 +1,1544 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.8846153846153846,
|
| 5 |
+
"global_step": 1200,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 0.0003,
|
| 13 |
+
"loss": 2.0952,
|
| 14 |
+
"step": 5
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 0.000298793242156074,
|
| 19 |
+
"loss": 2.2593,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 0.000297586484312148,
|
| 25 |
+
"loss": 2.1623,
|
| 26 |
+
"step": 15
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 0.000296379726468222,
|
| 31 |
+
"loss": 1.9038,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 0.000295172968624296,
|
| 37 |
+
"loss": 1.8576,
|
| 38 |
+
"step": 25
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 0.00029396621078037004,
|
| 43 |
+
"loss": 1.8397,
|
| 44 |
+
"step": 30
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08,
|
| 48 |
+
"learning_rate": 0.00029275945293644407,
|
| 49 |
+
"loss": 1.8032,
|
| 50 |
+
"step": 35
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.1,
|
| 54 |
+
"learning_rate": 0.00029155269509251804,
|
| 55 |
+
"loss": 1.7281,
|
| 56 |
+
"step": 40
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.11,
|
| 60 |
+
"learning_rate": 0.00029034593724859206,
|
| 61 |
+
"loss": 1.7097,
|
| 62 |
+
"step": 45
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.12,
|
| 66 |
+
"learning_rate": 0.00028913917940466614,
|
| 67 |
+
"loss": 1.7421,
|
| 68 |
+
"step": 50
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.13,
|
| 72 |
+
"learning_rate": 0.0002879324215607401,
|
| 73 |
+
"loss": 1.6861,
|
| 74 |
+
"step": 55
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.14,
|
| 78 |
+
"learning_rate": 0.00028672566371681414,
|
| 79 |
+
"loss": 1.6848,
|
| 80 |
+
"step": 60
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16,
|
| 84 |
+
"learning_rate": 0.00028551890587288816,
|
| 85 |
+
"loss": 1.5163,
|
| 86 |
+
"step": 65
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.17,
|
| 90 |
+
"learning_rate": 0.0002843121480289622,
|
| 91 |
+
"loss": 1.7266,
|
| 92 |
+
"step": 70
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.18,
|
| 96 |
+
"learning_rate": 0.0002831053901850362,
|
| 97 |
+
"loss": 1.5888,
|
| 98 |
+
"step": 75
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.19,
|
| 102 |
+
"learning_rate": 0.0002818986323411102,
|
| 103 |
+
"loss": 1.5463,
|
| 104 |
+
"step": 80
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.2,
|
| 108 |
+
"learning_rate": 0.0002806918744971842,
|
| 109 |
+
"loss": 1.4611,
|
| 110 |
+
"step": 85
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.22,
|
| 114 |
+
"learning_rate": 0.00027948511665325823,
|
| 115 |
+
"loss": 1.5554,
|
| 116 |
+
"step": 90
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.23,
|
| 120 |
+
"learning_rate": 0.00027827835880933226,
|
| 121 |
+
"loss": 1.5481,
|
| 122 |
+
"step": 95
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.24,
|
| 126 |
+
"learning_rate": 0.0002770716009654062,
|
| 127 |
+
"loss": 1.4464,
|
| 128 |
+
"step": 100
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.25,
|
| 132 |
+
"eval_loss": 1.6602041721343994,
|
| 133 |
+
"eval_runtime": 25.516,
|
| 134 |
+
"eval_samples_per_second": 8.191,
|
| 135 |
+
"eval_steps_per_second": 1.058,
|
| 136 |
+
"step": 104
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.25,
|
| 140 |
+
"learning_rate": 0.00027586484312148025,
|
| 141 |
+
"loss": 1.6648,
|
| 142 |
+
"step": 105
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.26,
|
| 146 |
+
"learning_rate": 0.0002746580852775543,
|
| 147 |
+
"loss": 1.6464,
|
| 148 |
+
"step": 110
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.28,
|
| 152 |
+
"learning_rate": 0.0002734513274336283,
|
| 153 |
+
"loss": 1.524,
|
| 154 |
+
"step": 115
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.29,
|
| 158 |
+
"learning_rate": 0.0002722445695897023,
|
| 159 |
+
"loss": 1.4791,
|
| 160 |
+
"step": 120
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.3,
|
| 164 |
+
"learning_rate": 0.0002710378117457763,
|
| 165 |
+
"loss": 1.5162,
|
| 166 |
+
"step": 125
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.31,
|
| 170 |
+
"learning_rate": 0.0002698310539018503,
|
| 171 |
+
"loss": 1.6636,
|
| 172 |
+
"step": 130
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.32,
|
| 176 |
+
"learning_rate": 0.00026862429605792435,
|
| 177 |
+
"loss": 1.7803,
|
| 178 |
+
"step": 135
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.34,
|
| 182 |
+
"learning_rate": 0.00026741753821399837,
|
| 183 |
+
"loss": 1.7384,
|
| 184 |
+
"step": 140
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.35,
|
| 188 |
+
"learning_rate": 0.0002662107803700724,
|
| 189 |
+
"loss": 1.5462,
|
| 190 |
+
"step": 145
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.36,
|
| 194 |
+
"learning_rate": 0.0002650040225261464,
|
| 195 |
+
"loss": 1.5802,
|
| 196 |
+
"step": 150
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.37,
|
| 200 |
+
"learning_rate": 0.00026379726468222045,
|
| 201 |
+
"loss": 1.4942,
|
| 202 |
+
"step": 155
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.38,
|
| 206 |
+
"learning_rate": 0.0002625905068382944,
|
| 207 |
+
"loss": 1.601,
|
| 208 |
+
"step": 160
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.4,
|
| 212 |
+
"learning_rate": 0.00026138374899436844,
|
| 213 |
+
"loss": 1.5822,
|
| 214 |
+
"step": 165
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.41,
|
| 218 |
+
"learning_rate": 0.00026017699115044247,
|
| 219 |
+
"loss": 1.5071,
|
| 220 |
+
"step": 170
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.42,
|
| 224 |
+
"learning_rate": 0.0002589702333065165,
|
| 225 |
+
"loss": 1.5926,
|
| 226 |
+
"step": 175
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.43,
|
| 230 |
+
"learning_rate": 0.00025776347546259046,
|
| 231 |
+
"loss": 1.4206,
|
| 232 |
+
"step": 180
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.44,
|
| 236 |
+
"learning_rate": 0.0002565567176186645,
|
| 237 |
+
"loss": 1.5317,
|
| 238 |
+
"step": 185
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.46,
|
| 242 |
+
"learning_rate": 0.0002553499597747385,
|
| 243 |
+
"loss": 1.4862,
|
| 244 |
+
"step": 190
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.47,
|
| 248 |
+
"learning_rate": 0.00025414320193081253,
|
| 249 |
+
"loss": 1.4387,
|
| 250 |
+
"step": 195
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.48,
|
| 254 |
+
"learning_rate": 0.00025293644408688656,
|
| 255 |
+
"loss": 1.5276,
|
| 256 |
+
"step": 200
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.49,
|
| 260 |
+
"learning_rate": 0.00025172968624296053,
|
| 261 |
+
"loss": 1.55,
|
| 262 |
+
"step": 205
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.5,
|
| 266 |
+
"eval_loss": 1.6166218519210815,
|
| 267 |
+
"eval_runtime": 13.3266,
|
| 268 |
+
"eval_samples_per_second": 15.683,
|
| 269 |
+
"eval_steps_per_second": 2.026,
|
| 270 |
+
"step": 208
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.5,
|
| 274 |
+
"learning_rate": 0.00025052292839903455,
|
| 275 |
+
"loss": 1.5525,
|
| 276 |
+
"step": 210
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.52,
|
| 280 |
+
"learning_rate": 0.0002493161705551086,
|
| 281 |
+
"loss": 1.5591,
|
| 282 |
+
"step": 215
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.53,
|
| 286 |
+
"learning_rate": 0.0002481094127111826,
|
| 287 |
+
"loss": 1.529,
|
| 288 |
+
"step": 220
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.54,
|
| 292 |
+
"learning_rate": 0.0002469026548672566,
|
| 293 |
+
"loss": 1.7152,
|
| 294 |
+
"step": 225
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.55,
|
| 298 |
+
"learning_rate": 0.00024569589702333065,
|
| 299 |
+
"loss": 1.7493,
|
| 300 |
+
"step": 230
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.56,
|
| 304 |
+
"learning_rate": 0.0002444891391794047,
|
| 305 |
+
"loss": 1.5288,
|
| 306 |
+
"step": 235
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.58,
|
| 310 |
+
"learning_rate": 0.00024328238133547868,
|
| 311 |
+
"loss": 1.5651,
|
| 312 |
+
"step": 240
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.59,
|
| 316 |
+
"learning_rate": 0.00024207562349155267,
|
| 317 |
+
"loss": 1.5144,
|
| 318 |
+
"step": 245
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6,
|
| 322 |
+
"learning_rate": 0.0002408688656476267,
|
| 323 |
+
"loss": 1.5334,
|
| 324 |
+
"step": 250
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.61,
|
| 328 |
+
"learning_rate": 0.0002396621078037007,
|
| 329 |
+
"loss": 1.5988,
|
| 330 |
+
"step": 255
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.62,
|
| 334 |
+
"learning_rate": 0.00023845534995977472,
|
| 335 |
+
"loss": 1.6401,
|
| 336 |
+
"step": 260
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.64,
|
| 340 |
+
"learning_rate": 0.00023724859211584875,
|
| 341 |
+
"loss": 1.5929,
|
| 342 |
+
"step": 265
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.65,
|
| 346 |
+
"learning_rate": 0.00023604183427192274,
|
| 347 |
+
"loss": 1.3948,
|
| 348 |
+
"step": 270
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.66,
|
| 352 |
+
"learning_rate": 0.00023483507642799677,
|
| 353 |
+
"loss": 1.221,
|
| 354 |
+
"step": 275
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.67,
|
| 358 |
+
"learning_rate": 0.00023362831858407077,
|
| 359 |
+
"loss": 1.5604,
|
| 360 |
+
"step": 280
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.69,
|
| 364 |
+
"learning_rate": 0.0002324215607401448,
|
| 365 |
+
"loss": 1.505,
|
| 366 |
+
"step": 285
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.7,
|
| 370 |
+
"learning_rate": 0.0002312148028962188,
|
| 371 |
+
"loss": 1.5008,
|
| 372 |
+
"step": 290
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.71,
|
| 376 |
+
"learning_rate": 0.0002300080450522928,
|
| 377 |
+
"loss": 1.5176,
|
| 378 |
+
"step": 295
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.72,
|
| 382 |
+
"learning_rate": 0.0002288012872083668,
|
| 383 |
+
"loss": 1.4227,
|
| 384 |
+
"step": 300
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.73,
|
| 388 |
+
"learning_rate": 0.00022759452936444084,
|
| 389 |
+
"loss": 1.1778,
|
| 390 |
+
"step": 305
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.75,
|
| 394 |
+
"learning_rate": 0.00022638777152051486,
|
| 395 |
+
"loss": 1.5052,
|
| 396 |
+
"step": 310
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.75,
|
| 400 |
+
"eval_loss": 1.5979019403457642,
|
| 401 |
+
"eval_runtime": 13.342,
|
| 402 |
+
"eval_samples_per_second": 15.665,
|
| 403 |
+
"eval_steps_per_second": 2.024,
|
| 404 |
+
"step": 312
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.76,
|
| 408 |
+
"learning_rate": 0.00022518101367658886,
|
| 409 |
+
"loss": 1.5228,
|
| 410 |
+
"step": 315
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.77,
|
| 414 |
+
"learning_rate": 0.0002239742558326629,
|
| 415 |
+
"loss": 1.5385,
|
| 416 |
+
"step": 320
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.78,
|
| 420 |
+
"learning_rate": 0.0002227674979887369,
|
| 421 |
+
"loss": 1.5072,
|
| 422 |
+
"step": 325
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.79,
|
| 426 |
+
"learning_rate": 0.00022156074014481093,
|
| 427 |
+
"loss": 1.5032,
|
| 428 |
+
"step": 330
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.81,
|
| 432 |
+
"learning_rate": 0.00022035398230088496,
|
| 433 |
+
"loss": 1.5981,
|
| 434 |
+
"step": 335
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.82,
|
| 438 |
+
"learning_rate": 0.00021914722445695895,
|
| 439 |
+
"loss": 1.422,
|
| 440 |
+
"step": 340
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.83,
|
| 444 |
+
"learning_rate": 0.00021794046661303298,
|
| 445 |
+
"loss": 1.4948,
|
| 446 |
+
"step": 345
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.84,
|
| 450 |
+
"learning_rate": 0.00021673370876910698,
|
| 451 |
+
"loss": 1.4421,
|
| 452 |
+
"step": 350
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.85,
|
| 456 |
+
"learning_rate": 0.000215526950925181,
|
| 457 |
+
"loss": 1.6086,
|
| 458 |
+
"step": 355
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.87,
|
| 462 |
+
"learning_rate": 0.000214320193081255,
|
| 463 |
+
"loss": 1.3874,
|
| 464 |
+
"step": 360
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.88,
|
| 468 |
+
"learning_rate": 0.00021311343523732902,
|
| 469 |
+
"loss": 1.2493,
|
| 470 |
+
"step": 365
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.89,
|
| 474 |
+
"learning_rate": 0.00021190667739340302,
|
| 475 |
+
"loss": 1.5565,
|
| 476 |
+
"step": 370
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.9,
|
| 480 |
+
"learning_rate": 0.00021069991954947705,
|
| 481 |
+
"loss": 1.4739,
|
| 482 |
+
"step": 375
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.91,
|
| 486 |
+
"learning_rate": 0.00020949316170555107,
|
| 487 |
+
"loss": 1.5288,
|
| 488 |
+
"step": 380
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.93,
|
| 492 |
+
"learning_rate": 0.00020828640386162507,
|
| 493 |
+
"loss": 1.4837,
|
| 494 |
+
"step": 385
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.94,
|
| 498 |
+
"learning_rate": 0.0002070796460176991,
|
| 499 |
+
"loss": 1.5009,
|
| 500 |
+
"step": 390
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.95,
|
| 504 |
+
"learning_rate": 0.0002058728881737731,
|
| 505 |
+
"loss": 1.3469,
|
| 506 |
+
"step": 395
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.96,
|
| 510 |
+
"learning_rate": 0.00020466613032984712,
|
| 511 |
+
"loss": 1.3949,
|
| 512 |
+
"step": 400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.97,
|
| 516 |
+
"learning_rate": 0.00020345937248592117,
|
| 517 |
+
"loss": 1.3688,
|
| 518 |
+
"step": 405
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.99,
|
| 522 |
+
"learning_rate": 0.00020225261464199517,
|
| 523 |
+
"loss": 1.6629,
|
| 524 |
+
"step": 410
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.0,
|
| 528 |
+
"learning_rate": 0.0002010458567980692,
|
| 529 |
+
"loss": 1.5395,
|
| 530 |
+
"step": 415
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.0,
|
| 534 |
+
"eval_loss": 1.589026927947998,
|
| 535 |
+
"eval_runtime": 13.34,
|
| 536 |
+
"eval_samples_per_second": 15.667,
|
| 537 |
+
"eval_steps_per_second": 2.024,
|
| 538 |
+
"step": 416
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.01,
|
| 542 |
+
"learning_rate": 0.0001998390989541432,
|
| 543 |
+
"loss": 1.5794,
|
| 544 |
+
"step": 420
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.02,
|
| 548 |
+
"learning_rate": 0.0001986323411102172,
|
| 549 |
+
"loss": 1.4746,
|
| 550 |
+
"step": 425
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.03,
|
| 554 |
+
"learning_rate": 0.0001974255832662912,
|
| 555 |
+
"loss": 1.3205,
|
| 556 |
+
"step": 430
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.05,
|
| 560 |
+
"learning_rate": 0.00019621882542236523,
|
| 561 |
+
"loss": 1.491,
|
| 562 |
+
"step": 435
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.06,
|
| 566 |
+
"learning_rate": 0.00019501206757843926,
|
| 567 |
+
"loss": 1.4803,
|
| 568 |
+
"step": 440
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.07,
|
| 572 |
+
"learning_rate": 0.00019380530973451326,
|
| 573 |
+
"loss": 1.3436,
|
| 574 |
+
"step": 445
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.08,
|
| 578 |
+
"learning_rate": 0.00019259855189058728,
|
| 579 |
+
"loss": 1.3893,
|
| 580 |
+
"step": 450
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.09,
|
| 584 |
+
"learning_rate": 0.00019139179404666128,
|
| 585 |
+
"loss": 1.0474,
|
| 586 |
+
"step": 455
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.11,
|
| 590 |
+
"learning_rate": 0.0001901850362027353,
|
| 591 |
+
"loss": 1.3388,
|
| 592 |
+
"step": 460
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.12,
|
| 596 |
+
"learning_rate": 0.0001889782783588093,
|
| 597 |
+
"loss": 1.4201,
|
| 598 |
+
"step": 465
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.13,
|
| 602 |
+
"learning_rate": 0.00018777152051488333,
|
| 603 |
+
"loss": 1.4805,
|
| 604 |
+
"step": 470
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.14,
|
| 608 |
+
"learning_rate": 0.00018656476267095732,
|
| 609 |
+
"loss": 1.2253,
|
| 610 |
+
"step": 475
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 1.15,
|
| 614 |
+
"learning_rate": 0.00018535800482703135,
|
| 615 |
+
"loss": 1.5064,
|
| 616 |
+
"step": 480
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 1.17,
|
| 620 |
+
"learning_rate": 0.00018415124698310537,
|
| 621 |
+
"loss": 1.3565,
|
| 622 |
+
"step": 485
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.18,
|
| 626 |
+
"learning_rate": 0.00018294448913917937,
|
| 627 |
+
"loss": 1.3976,
|
| 628 |
+
"step": 490
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.19,
|
| 632 |
+
"learning_rate": 0.00018173773129525342,
|
| 633 |
+
"loss": 1.5882,
|
| 634 |
+
"step": 495
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.2,
|
| 638 |
+
"learning_rate": 0.00018053097345132742,
|
| 639 |
+
"loss": 1.4698,
|
| 640 |
+
"step": 500
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.21,
|
| 644 |
+
"learning_rate": 0.00017932421560740145,
|
| 645 |
+
"loss": 1.5326,
|
| 646 |
+
"step": 505
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.23,
|
| 650 |
+
"learning_rate": 0.00017811745776347547,
|
| 651 |
+
"loss": 1.5659,
|
| 652 |
+
"step": 510
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 1.24,
|
| 656 |
+
"learning_rate": 0.00017691069991954947,
|
| 657 |
+
"loss": 1.4489,
|
| 658 |
+
"step": 515
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 1.25,
|
| 662 |
+
"learning_rate": 0.0001757039420756235,
|
| 663 |
+
"loss": 1.5312,
|
| 664 |
+
"step": 520
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.25,
|
| 668 |
+
"eval_loss": 1.5857353210449219,
|
| 669 |
+
"eval_runtime": 13.3247,
|
| 670 |
+
"eval_samples_per_second": 15.685,
|
| 671 |
+
"eval_steps_per_second": 2.026,
|
| 672 |
+
"step": 520
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.26,
|
| 676 |
+
"learning_rate": 0.0001744971842316975,
|
| 677 |
+
"loss": 1.3595,
|
| 678 |
+
"step": 525
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.27,
|
| 682 |
+
"learning_rate": 0.00017329042638777152,
|
| 683 |
+
"loss": 1.1543,
|
| 684 |
+
"step": 530
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.29,
|
| 688 |
+
"learning_rate": 0.0001720836685438455,
|
| 689 |
+
"loss": 1.4568,
|
| 690 |
+
"step": 535
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.3,
|
| 694 |
+
"learning_rate": 0.00017087691069991954,
|
| 695 |
+
"loss": 1.3682,
|
| 696 |
+
"step": 540
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.31,
|
| 700 |
+
"learning_rate": 0.00016967015285599354,
|
| 701 |
+
"loss": 1.5323,
|
| 702 |
+
"step": 545
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.32,
|
| 706 |
+
"learning_rate": 0.00016846339501206756,
|
| 707 |
+
"loss": 1.6644,
|
| 708 |
+
"step": 550
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 1.33,
|
| 712 |
+
"learning_rate": 0.00016725663716814158,
|
| 713 |
+
"loss": 1.5735,
|
| 714 |
+
"step": 555
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 1.35,
|
| 718 |
+
"learning_rate": 0.00016604987932421558,
|
| 719 |
+
"loss": 1.4759,
|
| 720 |
+
"step": 560
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 1.36,
|
| 724 |
+
"learning_rate": 0.0001648431214802896,
|
| 725 |
+
"loss": 1.3774,
|
| 726 |
+
"step": 565
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.37,
|
| 730 |
+
"learning_rate": 0.0001636363636363636,
|
| 731 |
+
"loss": 1.3321,
|
| 732 |
+
"step": 570
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.38,
|
| 736 |
+
"learning_rate": 0.00016242960579243763,
|
| 737 |
+
"loss": 1.2843,
|
| 738 |
+
"step": 575
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.39,
|
| 742 |
+
"learning_rate": 0.00016122284794851163,
|
| 743 |
+
"loss": 1.4636,
|
| 744 |
+
"step": 580
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.41,
|
| 748 |
+
"learning_rate": 0.00016001609010458568,
|
| 749 |
+
"loss": 1.4082,
|
| 750 |
+
"step": 585
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 1.42,
|
| 754 |
+
"learning_rate": 0.0001588093322606597,
|
| 755 |
+
"loss": 1.3848,
|
| 756 |
+
"step": 590
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 1.43,
|
| 760 |
+
"learning_rate": 0.0001576025744167337,
|
| 761 |
+
"loss": 1.5396,
|
| 762 |
+
"step": 595
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1.44,
|
| 766 |
+
"learning_rate": 0.00015639581657280773,
|
| 767 |
+
"loss": 1.4889,
|
| 768 |
+
"step": 600
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 1.45,
|
| 772 |
+
"learning_rate": 0.00015518905872888172,
|
| 773 |
+
"loss": 1.4019,
|
| 774 |
+
"step": 605
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.47,
|
| 778 |
+
"learning_rate": 0.00015398230088495575,
|
| 779 |
+
"loss": 1.3788,
|
| 780 |
+
"step": 610
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.48,
|
| 784 |
+
"learning_rate": 0.00015277554304102975,
|
| 785 |
+
"loss": 1.409,
|
| 786 |
+
"step": 615
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.49,
|
| 790 |
+
"learning_rate": 0.00015156878519710377,
|
| 791 |
+
"loss": 1.5675,
|
| 792 |
+
"step": 620
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 1.5,
|
| 796 |
+
"eval_loss": 1.584699273109436,
|
| 797 |
+
"eval_runtime": 13.3509,
|
| 798 |
+
"eval_samples_per_second": 15.654,
|
| 799 |
+
"eval_steps_per_second": 2.022,
|
| 800 |
+
"step": 624
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.5,
|
| 804 |
+
"learning_rate": 0.0001503620273531778,
|
| 805 |
+
"loss": 1.3176,
|
| 806 |
+
"step": 625
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.51,
|
| 810 |
+
"learning_rate": 0.0001491552695092518,
|
| 811 |
+
"loss": 1.3114,
|
| 812 |
+
"step": 630
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.53,
|
| 816 |
+
"learning_rate": 0.00014794851166532582,
|
| 817 |
+
"loss": 1.4261,
|
| 818 |
+
"step": 635
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.54,
|
| 822 |
+
"learning_rate": 0.00014674175382139982,
|
| 823 |
+
"loss": 1.5156,
|
| 824 |
+
"step": 640
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.55,
|
| 828 |
+
"learning_rate": 0.00014553499597747384,
|
| 829 |
+
"loss": 1.3231,
|
| 830 |
+
"step": 645
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.56,
|
| 834 |
+
"learning_rate": 0.00014432823813354787,
|
| 835 |
+
"loss": 1.4353,
|
| 836 |
+
"step": 650
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.57,
|
| 840 |
+
"learning_rate": 0.00014312148028962186,
|
| 841 |
+
"loss": 1.4423,
|
| 842 |
+
"step": 655
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.59,
|
| 846 |
+
"learning_rate": 0.0001419147224456959,
|
| 847 |
+
"loss": 1.1512,
|
| 848 |
+
"step": 660
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.6,
|
| 852 |
+
"learning_rate": 0.0001407079646017699,
|
| 853 |
+
"loss": 1.3562,
|
| 854 |
+
"step": 665
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.61,
|
| 858 |
+
"learning_rate": 0.0001395012067578439,
|
| 859 |
+
"loss": 1.4033,
|
| 860 |
+
"step": 670
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.62,
|
| 864 |
+
"learning_rate": 0.00013829444891391793,
|
| 865 |
+
"loss": 1.5435,
|
| 866 |
+
"step": 675
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.63,
|
| 870 |
+
"learning_rate": 0.00013708769106999193,
|
| 871 |
+
"loss": 1.4543,
|
| 872 |
+
"step": 680
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.65,
|
| 876 |
+
"learning_rate": 0.00013588093322606596,
|
| 877 |
+
"loss": 1.3509,
|
| 878 |
+
"step": 685
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.66,
|
| 882 |
+
"learning_rate": 0.00013467417538213995,
|
| 883 |
+
"loss": 1.4268,
|
| 884 |
+
"step": 690
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.67,
|
| 888 |
+
"learning_rate": 0.000133467417538214,
|
| 889 |
+
"loss": 1.4231,
|
| 890 |
+
"step": 695
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.68,
|
| 894 |
+
"learning_rate": 0.000132260659694288,
|
| 895 |
+
"loss": 1.4727,
|
| 896 |
+
"step": 700
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.69,
|
| 900 |
+
"learning_rate": 0.00013105390185036203,
|
| 901 |
+
"loss": 1.2846,
|
| 902 |
+
"step": 705
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.71,
|
| 906 |
+
"learning_rate": 0.00012984714400643603,
|
| 907 |
+
"loss": 1.3458,
|
| 908 |
+
"step": 710
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 1.72,
|
| 912 |
+
"learning_rate": 0.00012864038616251005,
|
| 913 |
+
"loss": 1.6349,
|
| 914 |
+
"step": 715
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.73,
|
| 918 |
+
"learning_rate": 0.00012743362831858405,
|
| 919 |
+
"loss": 1.4798,
|
| 920 |
+
"step": 720
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.74,
|
| 924 |
+
"learning_rate": 0.00012622687047465807,
|
| 925 |
+
"loss": 1.6157,
|
| 926 |
+
"step": 725
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.75,
|
| 930 |
+
"eval_loss": 1.5761181116104126,
|
| 931 |
+
"eval_runtime": 13.3218,
|
| 932 |
+
"eval_samples_per_second": 15.689,
|
| 933 |
+
"eval_steps_per_second": 2.027,
|
| 934 |
+
"step": 728
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.75,
|
| 938 |
+
"learning_rate": 0.00012502011263073207,
|
| 939 |
+
"loss": 1.4723,
|
| 940 |
+
"step": 730
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.77,
|
| 944 |
+
"learning_rate": 0.0001238133547868061,
|
| 945 |
+
"loss": 1.4487,
|
| 946 |
+
"step": 735
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 1.78,
|
| 950 |
+
"learning_rate": 0.00012260659694288012,
|
| 951 |
+
"loss": 1.3613,
|
| 952 |
+
"step": 740
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 1.79,
|
| 956 |
+
"learning_rate": 0.00012139983909895413,
|
| 957 |
+
"loss": 1.1906,
|
| 958 |
+
"step": 745
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 1.8,
|
| 962 |
+
"learning_rate": 0.00012019308125502816,
|
| 963 |
+
"loss": 1.4962,
|
| 964 |
+
"step": 750
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 1.81,
|
| 968 |
+
"learning_rate": 0.00011898632341110217,
|
| 969 |
+
"loss": 1.5629,
|
| 970 |
+
"step": 755
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.83,
|
| 974 |
+
"learning_rate": 0.00011777956556717618,
|
| 975 |
+
"loss": 1.3566,
|
| 976 |
+
"step": 760
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.84,
|
| 980 |
+
"learning_rate": 0.00011657280772325019,
|
| 981 |
+
"loss": 1.3442,
|
| 982 |
+
"step": 765
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.85,
|
| 986 |
+
"learning_rate": 0.0001153660498793242,
|
| 987 |
+
"loss": 1.4071,
|
| 988 |
+
"step": 770
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 1.86,
|
| 992 |
+
"learning_rate": 0.00011415929203539821,
|
| 993 |
+
"loss": 1.4331,
|
| 994 |
+
"step": 775
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 1.88,
|
| 998 |
+
"learning_rate": 0.00011295253419147222,
|
| 999 |
+
"loss": 1.3572,
|
| 1000 |
+
"step": 780
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.89,
|
| 1004 |
+
"learning_rate": 0.00011174577634754626,
|
| 1005 |
+
"loss": 1.4009,
|
| 1006 |
+
"step": 785
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 1.9,
|
| 1010 |
+
"learning_rate": 0.00011053901850362027,
|
| 1011 |
+
"loss": 1.47,
|
| 1012 |
+
"step": 790
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.91,
|
| 1016 |
+
"learning_rate": 0.00010933226065969428,
|
| 1017 |
+
"loss": 1.233,
|
| 1018 |
+
"step": 795
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.92,
|
| 1022 |
+
"learning_rate": 0.0001081255028157683,
|
| 1023 |
+
"loss": 1.4593,
|
| 1024 |
+
"step": 800
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.94,
|
| 1028 |
+
"learning_rate": 0.00010691874497184231,
|
| 1029 |
+
"loss": 1.4147,
|
| 1030 |
+
"step": 805
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.95,
|
| 1034 |
+
"learning_rate": 0.00010571198712791632,
|
| 1035 |
+
"loss": 1.2775,
|
| 1036 |
+
"step": 810
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 1.96,
|
| 1040 |
+
"learning_rate": 0.00010450522928399033,
|
| 1041 |
+
"loss": 1.5372,
|
| 1042 |
+
"step": 815
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 1.97,
|
| 1046 |
+
"learning_rate": 0.00010329847144006434,
|
| 1047 |
+
"loss": 1.2994,
|
| 1048 |
+
"step": 820
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 1.98,
|
| 1052 |
+
"learning_rate": 0.00010209171359613835,
|
| 1053 |
+
"loss": 1.3639,
|
| 1054 |
+
"step": 825
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 2.0,
|
| 1058 |
+
"learning_rate": 0.00010088495575221239,
|
| 1059 |
+
"loss": 1.4629,
|
| 1060 |
+
"step": 830
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.0,
|
| 1064 |
+
"eval_loss": 1.5735822916030884,
|
| 1065 |
+
"eval_runtime": 13.3179,
|
| 1066 |
+
"eval_samples_per_second": 15.693,
|
| 1067 |
+
"eval_steps_per_second": 2.027,
|
| 1068 |
+
"step": 832
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 2.01,
|
| 1072 |
+
"learning_rate": 9.96781979082864e-05,
|
| 1073 |
+
"loss": 1.3559,
|
| 1074 |
+
"step": 835
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.02,
|
| 1078 |
+
"learning_rate": 9.847144006436041e-05,
|
| 1079 |
+
"loss": 1.3924,
|
| 1080 |
+
"step": 840
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.03,
|
| 1084 |
+
"learning_rate": 9.726468222043442e-05,
|
| 1085 |
+
"loss": 1.2349,
|
| 1086 |
+
"step": 845
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 2.04,
|
| 1090 |
+
"learning_rate": 9.605792437650844e-05,
|
| 1091 |
+
"loss": 1.1467,
|
| 1092 |
+
"step": 850
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 2.06,
|
| 1096 |
+
"learning_rate": 9.485116653258245e-05,
|
| 1097 |
+
"loss": 1.3251,
|
| 1098 |
+
"step": 855
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.07,
|
| 1102 |
+
"learning_rate": 9.364440868865646e-05,
|
| 1103 |
+
"loss": 1.1517,
|
| 1104 |
+
"step": 860
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 2.08,
|
| 1108 |
+
"learning_rate": 9.243765084473048e-05,
|
| 1109 |
+
"loss": 1.4388,
|
| 1110 |
+
"step": 865
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.09,
|
| 1114 |
+
"learning_rate": 9.12308930008045e-05,
|
| 1115 |
+
"loss": 1.5138,
|
| 1116 |
+
"step": 870
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.1,
|
| 1120 |
+
"learning_rate": 9.002413515687852e-05,
|
| 1121 |
+
"loss": 1.4948,
|
| 1122 |
+
"step": 875
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.12,
|
| 1126 |
+
"learning_rate": 8.881737731295253e-05,
|
| 1127 |
+
"loss": 1.2988,
|
| 1128 |
+
"step": 880
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 2.13,
|
| 1132 |
+
"learning_rate": 8.761061946902654e-05,
|
| 1133 |
+
"loss": 1.226,
|
| 1134 |
+
"step": 885
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 2.14,
|
| 1138 |
+
"learning_rate": 8.640386162510055e-05,
|
| 1139 |
+
"loss": 1.1645,
|
| 1140 |
+
"step": 890
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.15,
|
| 1144 |
+
"learning_rate": 8.519710378117456e-05,
|
| 1145 |
+
"loss": 1.278,
|
| 1146 |
+
"step": 895
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.16,
|
| 1150 |
+
"learning_rate": 8.399034593724859e-05,
|
| 1151 |
+
"loss": 1.3873,
|
| 1152 |
+
"step": 900
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 2.18,
|
| 1156 |
+
"learning_rate": 8.27835880933226e-05,
|
| 1157 |
+
"loss": 1.231,
|
| 1158 |
+
"step": 905
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.19,
|
| 1162 |
+
"learning_rate": 8.157683024939661e-05,
|
| 1163 |
+
"loss": 1.401,
|
| 1164 |
+
"step": 910
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.2,
|
| 1168 |
+
"learning_rate": 8.037007240547062e-05,
|
| 1169 |
+
"loss": 1.4372,
|
| 1170 |
+
"step": 915
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 2.21,
|
| 1174 |
+
"learning_rate": 7.916331456154465e-05,
|
| 1175 |
+
"loss": 1.3624,
|
| 1176 |
+
"step": 920
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 2.22,
|
| 1180 |
+
"learning_rate": 7.795655671761866e-05,
|
| 1181 |
+
"loss": 1.5242,
|
| 1182 |
+
"step": 925
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.24,
|
| 1186 |
+
"learning_rate": 7.674979887369267e-05,
|
| 1187 |
+
"loss": 1.3661,
|
| 1188 |
+
"step": 930
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 2.25,
|
| 1192 |
+
"learning_rate": 7.554304102976669e-05,
|
| 1193 |
+
"loss": 1.3608,
|
| 1194 |
+
"step": 935
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 2.25,
|
| 1198 |
+
"eval_loss": 1.582654356956482,
|
| 1199 |
+
"eval_runtime": 13.332,
|
| 1200 |
+
"eval_samples_per_second": 15.677,
|
| 1201 |
+
"eval_steps_per_second": 2.025,
|
| 1202 |
+
"step": 936
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.26,
|
| 1206 |
+
"learning_rate": 7.43362831858407e-05,
|
| 1207 |
+
"loss": 1.476,
|
| 1208 |
+
"step": 940
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.27,
|
| 1212 |
+
"learning_rate": 7.312952534191472e-05,
|
| 1213 |
+
"loss": 1.3051,
|
| 1214 |
+
"step": 945
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.28,
|
| 1218 |
+
"learning_rate": 7.192276749798874e-05,
|
| 1219 |
+
"loss": 1.418,
|
| 1220 |
+
"step": 950
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.3,
|
| 1224 |
+
"learning_rate": 7.071600965406275e-05,
|
| 1225 |
+
"loss": 1.1781,
|
| 1226 |
+
"step": 955
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 2.31,
|
| 1230 |
+
"learning_rate": 6.950925181013676e-05,
|
| 1231 |
+
"loss": 1.4319,
|
| 1232 |
+
"step": 960
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.32,
|
| 1236 |
+
"learning_rate": 6.830249396621077e-05,
|
| 1237 |
+
"loss": 1.3059,
|
| 1238 |
+
"step": 965
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.33,
|
| 1242 |
+
"learning_rate": 6.709573612228479e-05,
|
| 1243 |
+
"loss": 1.0836,
|
| 1244 |
+
"step": 970
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.34,
|
| 1248 |
+
"learning_rate": 6.588897827835881e-05,
|
| 1249 |
+
"loss": 1.376,
|
| 1250 |
+
"step": 975
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.36,
|
| 1254 |
+
"learning_rate": 6.468222043443282e-05,
|
| 1255 |
+
"loss": 1.5022,
|
| 1256 |
+
"step": 980
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.37,
|
| 1260 |
+
"learning_rate": 6.347546259050683e-05,
|
| 1261 |
+
"loss": 1.2871,
|
| 1262 |
+
"step": 985
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.38,
|
| 1266 |
+
"learning_rate": 6.226870474658084e-05,
|
| 1267 |
+
"loss": 1.2777,
|
| 1268 |
+
"step": 990
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 2.39,
|
| 1272 |
+
"learning_rate": 6.106194690265487e-05,
|
| 1273 |
+
"loss": 1.5625,
|
| 1274 |
+
"step": 995
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.4,
|
| 1278 |
+
"learning_rate": 5.985518905872888e-05,
|
| 1279 |
+
"loss": 1.087,
|
| 1280 |
+
"step": 1000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.42,
|
| 1284 |
+
"learning_rate": 5.864843121480289e-05,
|
| 1285 |
+
"loss": 1.3169,
|
| 1286 |
+
"step": 1005
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 2.43,
|
| 1290 |
+
"learning_rate": 5.74416733708769e-05,
|
| 1291 |
+
"loss": 1.3657,
|
| 1292 |
+
"step": 1010
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 2.44,
|
| 1296 |
+
"learning_rate": 5.6234915526950926e-05,
|
| 1297 |
+
"loss": 1.3917,
|
| 1298 |
+
"step": 1015
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.45,
|
| 1302 |
+
"learning_rate": 5.502815768302494e-05,
|
| 1303 |
+
"loss": 1.3179,
|
| 1304 |
+
"step": 1020
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.46,
|
| 1308 |
+
"learning_rate": 5.382139983909895e-05,
|
| 1309 |
+
"loss": 1.4162,
|
| 1310 |
+
"step": 1025
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 2.48,
|
| 1314 |
+
"learning_rate": 5.261464199517296e-05,
|
| 1315 |
+
"loss": 1.3564,
|
| 1316 |
+
"step": 1030
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.49,
|
| 1320 |
+
"learning_rate": 5.140788415124698e-05,
|
| 1321 |
+
"loss": 1.276,
|
| 1322 |
+
"step": 1035
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.5,
|
| 1326 |
+
"learning_rate": 5.0201126307320996e-05,
|
| 1327 |
+
"loss": 1.3666,
|
| 1328 |
+
"step": 1040
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 2.5,
|
| 1332 |
+
"eval_loss": 1.5801613330841064,
|
| 1333 |
+
"eval_runtime": 13.3272,
|
| 1334 |
+
"eval_samples_per_second": 15.682,
|
| 1335 |
+
"eval_steps_per_second": 2.026,
|
| 1336 |
+
"step": 1040
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.51,
|
| 1340 |
+
"learning_rate": 4.899436846339501e-05,
|
| 1341 |
+
"loss": 1.4251,
|
| 1342 |
+
"step": 1045
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 2.52,
|
| 1346 |
+
"learning_rate": 4.778761061946902e-05,
|
| 1347 |
+
"loss": 1.4518,
|
| 1348 |
+
"step": 1050
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 2.54,
|
| 1352 |
+
"learning_rate": 4.6580852775543036e-05,
|
| 1353 |
+
"loss": 1.3541,
|
| 1354 |
+
"step": 1055
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.55,
|
| 1358 |
+
"learning_rate": 4.5374094931617054e-05,
|
| 1359 |
+
"loss": 1.249,
|
| 1360 |
+
"step": 1060
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.56,
|
| 1364 |
+
"learning_rate": 4.4167337087691066e-05,
|
| 1365 |
+
"loss": 1.3855,
|
| 1366 |
+
"step": 1065
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 2.57,
|
| 1370 |
+
"learning_rate": 4.2960579243765084e-05,
|
| 1371 |
+
"loss": 1.2707,
|
| 1372 |
+
"step": 1070
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 2.58,
|
| 1376 |
+
"learning_rate": 4.1753821399839095e-05,
|
| 1377 |
+
"loss": 1.2083,
|
| 1378 |
+
"step": 1075
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 2.6,
|
| 1382 |
+
"learning_rate": 4.0547063555913106e-05,
|
| 1383 |
+
"loss": 1.4777,
|
| 1384 |
+
"step": 1080
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 2.61,
|
| 1388 |
+
"learning_rate": 3.934030571198713e-05,
|
| 1389 |
+
"loss": 1.4348,
|
| 1390 |
+
"step": 1085
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 2.62,
|
| 1394 |
+
"learning_rate": 3.813354786806114e-05,
|
| 1395 |
+
"loss": 1.2354,
|
| 1396 |
+
"step": 1090
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.63,
|
| 1400 |
+
"learning_rate": 3.692679002413515e-05,
|
| 1401 |
+
"loss": 1.294,
|
| 1402 |
+
"step": 1095
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.64,
|
| 1406 |
+
"learning_rate": 3.572003218020917e-05,
|
| 1407 |
+
"loss": 1.3231,
|
| 1408 |
+
"step": 1100
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 2.66,
|
| 1412 |
+
"learning_rate": 3.451327433628318e-05,
|
| 1413 |
+
"loss": 1.2215,
|
| 1414 |
+
"step": 1105
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 2.67,
|
| 1418 |
+
"learning_rate": 3.33065164923572e-05,
|
| 1419 |
+
"loss": 1.1217,
|
| 1420 |
+
"step": 1110
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.68,
|
| 1424 |
+
"learning_rate": 3.209975864843121e-05,
|
| 1425 |
+
"loss": 1.4292,
|
| 1426 |
+
"step": 1115
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 2.69,
|
| 1430 |
+
"learning_rate": 3.089300080450522e-05,
|
| 1431 |
+
"loss": 1.195,
|
| 1432 |
+
"step": 1120
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 2.7,
|
| 1436 |
+
"learning_rate": 2.968624296057924e-05,
|
| 1437 |
+
"loss": 1.3549,
|
| 1438 |
+
"step": 1125
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.72,
|
| 1442 |
+
"learning_rate": 2.8479485116653255e-05,
|
| 1443 |
+
"loss": 1.4274,
|
| 1444 |
+
"step": 1130
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.73,
|
| 1448 |
+
"learning_rate": 2.727272727272727e-05,
|
| 1449 |
+
"loss": 1.4838,
|
| 1450 |
+
"step": 1135
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 2.74,
|
| 1454 |
+
"learning_rate": 2.6065969428801284e-05,
|
| 1455 |
+
"loss": 1.2394,
|
| 1456 |
+
"step": 1140
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 2.75,
|
| 1460 |
+
"eval_loss": 1.5814132690429688,
|
| 1461 |
+
"eval_runtime": 13.3223,
|
| 1462 |
+
"eval_samples_per_second": 15.688,
|
| 1463 |
+
"eval_steps_per_second": 2.027,
|
| 1464 |
+
"step": 1144
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 2.75,
|
| 1468 |
+
"learning_rate": 2.4859211584875302e-05,
|
| 1469 |
+
"loss": 1.2414,
|
| 1470 |
+
"step": 1145
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 2.76,
|
| 1474 |
+
"learning_rate": 2.3652453740949313e-05,
|
| 1475 |
+
"loss": 1.288,
|
| 1476 |
+
"step": 1150
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 2.78,
|
| 1480 |
+
"learning_rate": 2.244569589702333e-05,
|
| 1481 |
+
"loss": 1.2,
|
| 1482 |
+
"step": 1155
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 2.79,
|
| 1486 |
+
"learning_rate": 2.1238938053097343e-05,
|
| 1487 |
+
"loss": 1.2117,
|
| 1488 |
+
"step": 1160
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 2.8,
|
| 1492 |
+
"learning_rate": 2.003218020917136e-05,
|
| 1493 |
+
"loss": 1.3717,
|
| 1494 |
+
"step": 1165
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 2.81,
|
| 1498 |
+
"learning_rate": 1.8825422365245372e-05,
|
| 1499 |
+
"loss": 1.438,
|
| 1500 |
+
"step": 1170
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 2.82,
|
| 1504 |
+
"learning_rate": 1.7618664521319386e-05,
|
| 1505 |
+
"loss": 1.3346,
|
| 1506 |
+
"step": 1175
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 2.84,
|
| 1510 |
+
"learning_rate": 1.64119066773934e-05,
|
| 1511 |
+
"loss": 1.227,
|
| 1512 |
+
"step": 1180
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 2.85,
|
| 1516 |
+
"learning_rate": 1.5205148833467417e-05,
|
| 1517 |
+
"loss": 1.3146,
|
| 1518 |
+
"step": 1185
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 2.86,
|
| 1522 |
+
"learning_rate": 1.3998390989541432e-05,
|
| 1523 |
+
"loss": 1.4163,
|
| 1524 |
+
"step": 1190
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 2.87,
|
| 1528 |
+
"learning_rate": 1.2791633145615445e-05,
|
| 1529 |
+
"loss": 1.1681,
|
| 1530 |
+
"step": 1195
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 2.88,
|
| 1534 |
+
"learning_rate": 1.158487530168946e-05,
|
| 1535 |
+
"loss": 1.4127,
|
| 1536 |
+
"step": 1200
|
| 1537 |
+
}
|
| 1538 |
+
],
|
| 1539 |
+
"max_steps": 1248,
|
| 1540 |
+
"num_train_epochs": 3,
|
| 1541 |
+
"total_flos": 957297581350656.0,
|
| 1542 |
+
"trial_name": null,
|
| 1543 |
+
"trial_params": null
|
| 1544 |
+
}
|
checkpoint-1200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7341502d367cfe11875e6bc3d9de5bf48aed11642f79f7fa57ffc1e9a155b8c9
|
| 3 |
+
size 3963
|
runs/Jun18_19-59-55_3a5a94d2e5ba/1687118410.021046/events.out.tfevents.1687118410.3a5a94d2e5ba.49832.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9fd712550f0c7056f6000b3a209363ce0c2e60b4a8c0fc9441566696fa4c02c3
|
| 3 |
+
size 5979
|
runs/Jun18_19-59-55_3a5a94d2e5ba/events.out.tfevents.1687118410.3a5a94d2e5ba.49832.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2711f4ce5f4d5270a3fb2b992c2f3691bb7e7397e220aa61e46b4c0e30ee2cf9
|
| 3 |
+
size 46869
|