Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-v1.0e-discrete-condition/checkpoint-21651/eval_state.json +3 -0
- checkpoints-v1.0e-discrete-condition/checkpoint-21651/model.safetensors +3 -0
- checkpoints-v1.0e-discrete-condition/checkpoint-21651/optimizer.pt +3 -0
- checkpoints-v1.0e-discrete-condition/checkpoint-21651/rng_state.pth +3 -0
- checkpoints-v1.0e-discrete-condition/checkpoint-21651/scaler.pt +3 -0
- checkpoints-v1.0e-discrete-condition/checkpoint-21651/scheduler.pt +3 -0
- checkpoints-v1.0e-discrete-condition/checkpoint-21651/trainer_state.json +1040 -0
- checkpoints-v1.0e-discrete-condition/checkpoint-21651/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -103,3 +103,4 @@ checkpoints-v1.0c-discrete-condition/checkpoint-21651/eval_state.json filter=lfs
|
|
| 103 |
checkpoints-v1.0d-discrete-condition/checkpoint-13312/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 104 |
checkpoints-v1.0d-discrete-condition/checkpoint-21651/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 105 |
checkpoints-v1.0e-discrete-condition/checkpoint-15360/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 103 |
checkpoints-v1.0d-discrete-condition/checkpoint-13312/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 104 |
checkpoints-v1.0d-discrete-condition/checkpoint-21651/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 105 |
checkpoints-v1.0e-discrete-condition/checkpoint-15360/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 106 |
+
checkpoints-v1.0e-discrete-condition/checkpoint-21651/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-v1.0e-discrete-condition/checkpoint-21651/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2e23e699d0a75b14f69827ccdb1fc2eb256eacbf81a2ae52211c8be81832209d
|
| 3 |
+
size 43408807
|
checkpoints-v1.0e-discrete-condition/checkpoint-21651/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61fb735a4831bf206ac719f2c2b576318b5beb5befd1f921177a04eab5f05ea0
|
| 3 |
+
size 15210032
|
checkpoints-v1.0e-discrete-condition/checkpoint-21651/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e61d8ebaa108e0d353ef7f1722ed789189197a5fef739a26ce04578b45b8501
|
| 3 |
+
size 13675339
|
checkpoints-v1.0e-discrete-condition/checkpoint-21651/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:062e35101c832e3ca6aa42fa87c301a629f3c44509855cda93bc9afbb4f7763b
|
| 3 |
+
size 14645
|
checkpoints-v1.0e-discrete-condition/checkpoint-21651/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:73f889f853c16556b8eb35b18db3cd50e578293bcd93e48757250242c2f45276
|
| 3 |
+
size 1383
|
checkpoints-v1.0e-discrete-condition/checkpoint-21651/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:70a493bf8fd6932c4ac60675b60fe483d4684704d3cab9dc59633db8961d91a7
|
| 3 |
+
size 1465
|
checkpoints-v1.0e-discrete-condition/checkpoint-21651/trainer_state.json
ADDED
|
@@ -0,0 +1,1040 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 21651,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011823934229365849,
|
| 14 |
+
"grad_norm": 0.0569785013794899,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 0.2929760813713074,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.0866856649518013,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 0.2504822015762329,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.13940005004405975,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 0.19717276096343994,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.14197291433811188,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 0.16308173537254333,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_bleu": 0.0011714375896648113,
|
| 43 |
+
"eval_cos_loss": 0.338739567831771,
|
| 44 |
+
"eval_loss": 0.1511567461232072,
|
| 45 |
+
"eval_mse_loss": 0.1511567461232072,
|
| 46 |
+
"step": 1024
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.047295736917463395,
|
| 50 |
+
"eval_bleu": 0.0011714375896648113,
|
| 51 |
+
"eval_cos_loss": 0.338739567831771,
|
| 52 |
+
"eval_loss": 0.1511567461232072,
|
| 53 |
+
"eval_mse_loss": 0.1511567461232072,
|
| 54 |
+
"eval_runtime": 172.2677,
|
| 55 |
+
"eval_samples_per_second": 162.497,
|
| 56 |
+
"eval_steps_per_second": 2.543,
|
| 57 |
+
"step": 1024
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.05911967114682925,
|
| 61 |
+
"grad_norm": 0.14103451371192932,
|
| 62 |
+
"learning_rate": 0.0009967551747861387,
|
| 63 |
+
"loss": 0.14320890605449677,
|
| 64 |
+
"step": 1280
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.0709436053761951,
|
| 68 |
+
"grad_norm": 0.09702787548303604,
|
| 69 |
+
"learning_rate": 0.000994232528651847,
|
| 70 |
+
"loss": 0.13121198117733002,
|
| 71 |
+
"step": 1536
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 0.08276753960556095,
|
| 75 |
+
"grad_norm": 0.14316639304161072,
|
| 76 |
+
"learning_rate": 0.0009909945800260092,
|
| 77 |
+
"loss": 0.11884352564811707,
|
| 78 |
+
"step": 1792
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 0.09459147383492679,
|
| 82 |
+
"grad_norm": 0.16359540820121765,
|
| 83 |
+
"learning_rate": 0.0009870460151900522,
|
| 84 |
+
"loss": 0.10863035917282104,
|
| 85 |
+
"step": 2048
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.09459147383492679,
|
| 89 |
+
"eval_bleu": 0.01640269135297576,
|
| 90 |
+
"eval_cos_loss": 0.2630239495811941,
|
| 91 |
+
"eval_loss": 0.10520831837433658,
|
| 92 |
+
"eval_mse_loss": 0.10520831837433658,
|
| 93 |
+
"step": 2048
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09459147383492679,
|
| 97 |
+
"eval_bleu": 0.01640269135297576,
|
| 98 |
+
"eval_cos_loss": 0.2630239495811941,
|
| 99 |
+
"eval_loss": 0.10520831837433658,
|
| 100 |
+
"eval_mse_loss": 0.10520831837433658,
|
| 101 |
+
"eval_runtime": 166.1927,
|
| 102 |
+
"eval_samples_per_second": 168.437,
|
| 103 |
+
"eval_steps_per_second": 2.635,
|
| 104 |
+
"step": 2048
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.10641540806429264,
|
| 108 |
+
"grad_norm": 0.16606658697128296,
|
| 109 |
+
"learning_rate": 0.0009823925488998885,
|
| 110 |
+
"loss": 0.10165824741125107,
|
| 111 |
+
"step": 2304
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.1182393422936585,
|
| 115 |
+
"grad_norm": 0.16988244652748108,
|
| 116 |
+
"learning_rate": 0.0009770409161149525,
|
| 117 |
+
"loss": 0.09657377749681473,
|
| 118 |
+
"step": 2560
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.13006327652302435,
|
| 122 |
+
"grad_norm": 0.15539805591106415,
|
| 123 |
+
"learning_rate": 0.0009709988622506973,
|
| 124 |
+
"loss": 0.09263481199741364,
|
| 125 |
+
"step": 2816
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.1418872107523902,
|
| 129 |
+
"grad_norm": 0.19379030168056488,
|
| 130 |
+
"learning_rate": 0.000964275131968659,
|
| 131 |
+
"loss": 0.08921799808740616,
|
| 132 |
+
"step": 3072
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.1418872107523902,
|
| 136 |
+
"eval_bleu": 0.037097672791643686,
|
| 137 |
+
"eval_cos_loss": 0.23058595494728654,
|
| 138 |
+
"eval_loss": 0.08828340418058443,
|
| 139 |
+
"eval_mse_loss": 0.08828340418058443,
|
| 140 |
+
"step": 3072
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.1418872107523902,
|
| 144 |
+
"eval_bleu": 0.037097672791643686,
|
| 145 |
+
"eval_cos_loss": 0.23058595494728654,
|
| 146 |
+
"eval_loss": 0.08828340418058443,
|
| 147 |
+
"eval_mse_loss": 0.08828340418058443,
|
| 148 |
+
"eval_runtime": 169.0954,
|
| 149 |
+
"eval_samples_per_second": 165.546,
|
| 150 |
+
"eval_steps_per_second": 2.59,
|
| 151 |
+
"step": 3072
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.15371114498175603,
|
| 155 |
+
"grad_norm": 0.19469866156578064,
|
| 156 |
+
"learning_rate": 0.0009568794565203123,
|
| 157 |
+
"loss": 0.08570453524589539,
|
| 158 |
+
"step": 3328
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.1655350792111219,
|
| 162 |
+
"grad_norm": 0.13641111552715302,
|
| 163 |
+
"learning_rate": 0.0009488225396630347,
|
| 164 |
+
"loss": 0.08265809714794159,
|
| 165 |
+
"step": 3584
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.17735901344048774,
|
| 169 |
+
"grad_norm": 0.20243526995182037,
|
| 170 |
+
"learning_rate": 0.0009401160421685646,
|
| 171 |
+
"loss": 0.08029869943857193,
|
| 172 |
+
"step": 3840
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.18918294766985358,
|
| 176 |
+
"grad_norm": 0.20591998100280762,
|
| 177 |
+
"learning_rate": 0.0009307725649463714,
|
| 178 |
+
"loss": 0.07793460041284561,
|
| 179 |
+
"step": 4096
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.18918294766985358,
|
| 183 |
+
"eval_bleu": 0.055271730295598016,
|
| 184 |
+
"eval_cos_loss": 0.20533766668953307,
|
| 185 |
+
"eval_loss": 0.07763067195744819,
|
| 186 |
+
"eval_mse_loss": 0.07763067195744819,
|
| 187 |
+
"step": 4096
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 0.18918294766985358,
|
| 191 |
+
"eval_bleu": 0.055271730295598016,
|
| 192 |
+
"eval_cos_loss": 0.20533766668953307,
|
| 193 |
+
"eval_loss": 0.07763067195744819,
|
| 194 |
+
"eval_mse_loss": 0.07763067195744819,
|
| 195 |
+
"eval_runtime": 167.722,
|
| 196 |
+
"eval_samples_per_second": 166.901,
|
| 197 |
+
"eval_steps_per_second": 2.611,
|
| 198 |
+
"step": 4096
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.20100688189921945,
|
| 202 |
+
"grad_norm": 0.14815625548362732,
|
| 203 |
+
"learning_rate": 0.0009208056308063659,
|
| 204 |
+
"loss": 0.07629498839378357,
|
| 205 |
+
"step": 4352
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2128308161285853,
|
| 209 |
+
"grad_norm": 0.17893894016742706,
|
| 210 |
+
"learning_rate": 0.0009102296648873445,
|
| 211 |
+
"loss": 0.0742964819073677,
|
| 212 |
+
"step": 4608
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.22465475035795113,
|
| 216 |
+
"grad_norm": 0.18065278232097626,
|
| 217 |
+
"learning_rate": 0.0008990599737794927,
|
| 218 |
+
"loss": 0.07244566082954407,
|
| 219 |
+
"step": 4864
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.236478684587317,
|
| 223 |
+
"grad_norm": 0.18555110692977905,
|
| 224 |
+
"learning_rate": 0.0008873127233711644,
|
| 225 |
+
"loss": 0.07109007984399796,
|
| 226 |
+
"step": 5120
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.236478684587317,
|
| 230 |
+
"eval_bleu": 0.07162183673763856,
|
| 231 |
+
"eval_cos_loss": 0.18833101295853313,
|
| 232 |
+
"eval_loss": 0.07069546589840493,
|
| 233 |
+
"eval_mse_loss": 0.07069546589840493,
|
| 234 |
+
"step": 5120
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.236478684587317,
|
| 238 |
+
"eval_bleu": 0.07162183673763856,
|
| 239 |
+
"eval_cos_loss": 0.18833101295853313,
|
| 240 |
+
"eval_loss": 0.07069546589840493,
|
| 241 |
+
"eval_mse_loss": 0.07069546589840493,
|
| 242 |
+
"eval_runtime": 168.9203,
|
| 243 |
+
"eval_samples_per_second": 165.717,
|
| 244 |
+
"eval_steps_per_second": 2.593,
|
| 245 |
+
"step": 5120
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.24830261881668284,
|
| 249 |
+
"grad_norm": 0.21479532122612,
|
| 250 |
+
"learning_rate": 0.0008750049154520011,
|
| 251 |
+
"loss": 0.06965314596891403,
|
| 252 |
+
"step": 5376
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.2601265530460487,
|
| 256 |
+
"grad_norm": 0.15884655714035034,
|
| 257 |
+
"learning_rate": 0.0008621543631062487,
|
| 258 |
+
"loss": 0.06835002452135086,
|
| 259 |
+
"step": 5632
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.27195048727541454,
|
| 263 |
+
"grad_norm": 0.1491912305355072,
|
| 264 |
+
"learning_rate": 0.0008487796649318904,
|
| 265 |
+
"loss": 0.06699136644601822,
|
| 266 |
+
"step": 5888
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.2837744215047804,
|
| 270 |
+
"grad_norm": 0.16656875610351562,
|
| 271 |
+
"learning_rate": 0.0008349001781229053,
|
| 272 |
+
"loss": 0.06628481298685074,
|
| 273 |
+
"step": 6144
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.2837744215047804,
|
| 277 |
+
"eval_bleu": 0.08782990447295556,
|
| 278 |
+
"eval_cos_loss": 0.1761941106556213,
|
| 279 |
+
"eval_loss": 0.06568186053980703,
|
| 280 |
+
"eval_mse_loss": 0.06568186053980703,
|
| 281 |
+
"step": 6144
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.2837744215047804,
|
| 285 |
+
"eval_bleu": 0.08782990447295556,
|
| 286 |
+
"eval_cos_loss": 0.1761941106556213,
|
| 287 |
+
"eval_loss": 0.06568186053980703,
|
| 288 |
+
"eval_mse_loss": 0.06568186053980703,
|
| 289 |
+
"eval_runtime": 167.256,
|
| 290 |
+
"eval_samples_per_second": 167.366,
|
| 291 |
+
"eval_steps_per_second": 2.619,
|
| 292 |
+
"step": 6144
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.2955983557341462,
|
| 296 |
+
"grad_norm": 0.18619365990161896,
|
| 297 |
+
"learning_rate": 0.0008205359904536107,
|
| 298 |
+
"loss": 0.06507044285535812,
|
| 299 |
+
"step": 6400
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.30742228996351206,
|
| 303 |
+
"grad_norm": 0.15038955211639404,
|
| 304 |
+
"learning_rate": 0.0008057078912056363,
|
| 305 |
+
"loss": 0.06429114937782288,
|
| 306 |
+
"step": 6656
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.3192462241928779,
|
| 310 |
+
"grad_norm": 0.136085644364357,
|
| 311 |
+
"learning_rate": 0.0007904373410796086,
|
| 312 |
+
"loss": 0.06293655186891556,
|
| 313 |
+
"step": 6912
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.3310701584222438,
|
| 317 |
+
"grad_norm": 0.15249842405319214,
|
| 318 |
+
"learning_rate": 0.0007747464411350876,
|
| 319 |
+
"loss": 0.06225278228521347,
|
| 320 |
+
"step": 7168
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.3310701584222438,
|
| 324 |
+
"eval_bleu": 0.10422501320162747,
|
| 325 |
+
"eval_cos_loss": 0.16745890517076945,
|
| 326 |
+
"eval_loss": 0.06201160660005051,
|
| 327 |
+
"eval_mse_loss": 0.06201160660005051,
|
| 328 |
+
"step": 7168
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.3310701584222438,
|
| 332 |
+
"eval_bleu": 0.10422501320162747,
|
| 333 |
+
"eval_cos_loss": 0.16745890517076945,
|
| 334 |
+
"eval_loss": 0.06201160660005051,
|
| 335 |
+
"eval_mse_loss": 0.06201160660005051,
|
| 336 |
+
"eval_runtime": 169.4308,
|
| 337 |
+
"eval_samples_per_second": 165.218,
|
| 338 |
+
"eval_steps_per_second": 2.585,
|
| 339 |
+
"step": 7168
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.34289409265160964,
|
| 343 |
+
"grad_norm": 0.16507746279239655,
|
| 344 |
+
"learning_rate": 0.000758657900803716,
|
| 345 |
+
"loss": 0.061575472354888916,
|
| 346 |
+
"step": 7424
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.3547180268809755,
|
| 350 |
+
"grad_norm": 0.14547540247440338,
|
| 351 |
+
"learning_rate": 0.000742195005021869,
|
| 352 |
+
"loss": 0.060702890157699585,
|
| 353 |
+
"step": 7680
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.3665419611103413,
|
| 357 |
+
"grad_norm": 0.12704621255397797,
|
| 358 |
+
"learning_rate": 0.0007253815805303786,
|
| 359 |
+
"loss": 0.059864386916160583,
|
| 360 |
+
"step": 7936
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.37836589533970716,
|
| 364 |
+
"grad_norm": 0.1420435756444931,
|
| 365 |
+
"learning_rate": 0.0007082419613901028,
|
| 366 |
+
"loss": 0.059240467846393585,
|
| 367 |
+
"step": 8192
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.37836589533970716,
|
| 371 |
+
"eval_bleu": 0.12190068004312635,
|
| 372 |
+
"eval_cos_loss": 0.16049913827278842,
|
| 373 |
+
"eval_loss": 0.05894200267380775,
|
| 374 |
+
"eval_mse_loss": 0.05894200267380775,
|
| 375 |
+
"step": 8192
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.37836589533970716,
|
| 379 |
+
"eval_bleu": 0.12190068004312635,
|
| 380 |
+
"eval_cos_loss": 0.16049913827278842,
|
| 381 |
+
"eval_loss": 0.05894200267380775,
|
| 382 |
+
"eval_mse_loss": 0.05894200267380775,
|
| 383 |
+
"eval_runtime": 170.4747,
|
| 384 |
+
"eval_samples_per_second": 164.206,
|
| 385 |
+
"eval_steps_per_second": 2.569,
|
| 386 |
+
"step": 8192
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.390189829569073,
|
| 390 |
+
"grad_norm": 0.11315222084522247,
|
| 391 |
+
"learning_rate": 0.0006908009537632514,
|
| 392 |
+
"loss": 0.05864889174699783,
|
| 393 |
+
"step": 8448
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.4020137637984389,
|
| 397 |
+
"grad_norm": 0.14333274960517883,
|
| 398 |
+
"learning_rate": 0.0006730838000114403,
|
| 399 |
+
"loss": 0.05799835920333862,
|
| 400 |
+
"step": 8704
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.41383769802780473,
|
| 404 |
+
"grad_norm": 0.15370561182498932,
|
| 405 |
+
"learning_rate": 0.0006551161421624341,
|
| 406 |
+
"loss": 0.057285867631435394,
|
| 407 |
+
"step": 8960
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.4256616322571706,
|
| 411 |
+
"grad_norm": 0.12100663036108017,
|
| 412 |
+
"learning_rate": 0.0006369239847984517,
|
| 413 |
+
"loss": 0.05648953840136528,
|
| 414 |
+
"step": 9216
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.4256616322571706,
|
| 418 |
+
"eval_bleu": 0.13496230405038226,
|
| 419 |
+
"eval_cos_loss": 0.15526348454494998,
|
| 420 |
+
"eval_loss": 0.05676825863381499,
|
| 421 |
+
"eval_mse_loss": 0.05676825863381499,
|
| 422 |
+
"step": 9216
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.4256616322571706,
|
| 426 |
+
"eval_bleu": 0.13496230405038226,
|
| 427 |
+
"eval_cos_loss": 0.15526348454494998,
|
| 428 |
+
"eval_loss": 0.05676825863381499,
|
| 429 |
+
"eval_mse_loss": 0.05676825863381499,
|
| 430 |
+
"eval_runtime": 170.2747,
|
| 431 |
+
"eval_samples_per_second": 164.399,
|
| 432 |
+
"eval_steps_per_second": 2.572,
|
| 433 |
+
"step": 9216
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.4374855664865364,
|
| 437 |
+
"grad_norm": 0.12870068848133087,
|
| 438 |
+
"learning_rate": 0.0006185336574197479,
|
| 439 |
+
"loss": 0.056146688759326935,
|
| 440 |
+
"step": 9472
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.44930950071590225,
|
| 444 |
+
"grad_norm": 0.1489071398973465,
|
| 445 |
+
"learning_rate": 0.0005999717763379407,
|
| 446 |
+
"loss": 0.05556228384375572,
|
| 447 |
+
"step": 9728
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.4611334349452681,
|
| 451 |
+
"grad_norm": 0.11909710615873337,
|
| 452 |
+
"learning_rate": 0.0005812652061542363,
|
| 453 |
+
"loss": 0.05514912307262421,
|
| 454 |
+
"step": 9984
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 0.472957369174634,
|
| 458 |
+
"grad_norm": 0.15826211869716644,
|
| 459 |
+
"learning_rate": 0.0005624410208783071,
|
| 460 |
+
"loss": 0.05458202585577965,
|
| 461 |
+
"step": 10240
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.472957369174634,
|
| 465 |
+
"eval_bleu": 0.14839972030094115,
|
| 466 |
+
"eval_cos_loss": 0.1492026079803297,
|
| 467 |
+
"eval_loss": 0.05424791704608153,
|
| 468 |
+
"eval_mse_loss": 0.05424791704608153,
|
| 469 |
+
"step": 10240
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.472957369174634,
|
| 473 |
+
"eval_bleu": 0.14839972030094115,
|
| 474 |
+
"eval_cos_loss": 0.1492026079803297,
|
| 475 |
+
"eval_loss": 0.05424791704608153,
|
| 476 |
+
"eval_mse_loss": 0.05424791704608153,
|
| 477 |
+
"eval_runtime": 172.0235,
|
| 478 |
+
"eval_samples_per_second": 162.728,
|
| 479 |
+
"eval_steps_per_second": 2.546,
|
| 480 |
+
"step": 10240
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.48478130340399983,
|
| 484 |
+
"grad_norm": 0.12827685475349426,
|
| 485 |
+
"learning_rate": 0.0005435264647440881,
|
| 486 |
+
"loss": 0.05391533672809601,
|
| 487 |
+
"step": 10496
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.49660523763336567,
|
| 491 |
+
"grad_norm": 0.14331956207752228,
|
| 492 |
+
"learning_rate": 0.000524548912779213,
|
| 493 |
+
"loss": 0.053478896617889404,
|
| 494 |
+
"step": 10752
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.5084291718627315,
|
| 498 |
+
"grad_norm": 0.11259039491415024,
|
| 499 |
+
"learning_rate": 0.0005055358311851499,
|
| 500 |
+
"loss": 0.05310649052262306,
|
| 501 |
+
"step": 11008
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.5202531060920974,
|
| 505 |
+
"grad_norm": 0.11761780083179474,
|
| 506 |
+
"learning_rate": 0.0004865147375853812,
|
| 507 |
+
"loss": 0.05254605785012245,
|
| 508 |
+
"step": 11264
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.5202531060920974,
|
| 512 |
+
"eval_bleu": 0.16042218683436996,
|
| 513 |
+
"eval_cos_loss": 0.14472832799502158,
|
| 514 |
+
"eval_loss": 0.052395756981552466,
|
| 515 |
+
"eval_mse_loss": 0.052395756981552466,
|
| 516 |
+
"step": 11264
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.5202531060920974,
|
| 520 |
+
"eval_bleu": 0.16042218683436996,
|
| 521 |
+
"eval_cos_loss": 0.14472832799502158,
|
| 522 |
+
"eval_loss": 0.052395756981552466,
|
| 523 |
+
"eval_mse_loss": 0.052395756981552466,
|
| 524 |
+
"eval_runtime": 175.7973,
|
| 525 |
+
"eval_samples_per_second": 159.235,
|
| 526 |
+
"eval_steps_per_second": 2.492,
|
| 527 |
+
"step": 11264
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.5320770403214632,
|
| 531 |
+
"grad_norm": 0.16084498167037964,
|
| 532 |
+
"learning_rate": 0.0004675131611991607,
|
| 533 |
+
"loss": 0.05206044018268585,
|
| 534 |
+
"step": 11520
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5439009745508291,
|
| 538 |
+
"grad_norm": 0.12781144678592682,
|
| 539 |
+
"learning_rate": 0.0004485586029984899,
|
| 540 |
+
"loss": 0.05154161900281906,
|
| 541 |
+
"step": 11776
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.5557249087801949,
|
| 545 |
+
"grad_norm": 0.11860433220863342,
|
| 546 |
+
"learning_rate": 0.00042967849590597266,
|
| 547 |
+
"loss": 0.051158737391233444,
|
| 548 |
+
"step": 12032
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.5675488430095608,
|
| 552 |
+
"grad_norm": 0.1226862370967865,
|
| 553 |
+
"learning_rate": 0.0004109001650911621,
|
| 554 |
+
"loss": 0.05094496160745621,
|
| 555 |
+
"step": 12288
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.5675488430095608,
|
| 559 |
+
"eval_bleu": 0.17126127820473988,
|
| 560 |
+
"eval_cos_loss": 0.1419908434628896,
|
| 561 |
+
"eval_loss": 0.05134654454206495,
|
| 562 |
+
"eval_mse_loss": 0.05134654454206495,
|
| 563 |
+
"step": 12288
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.5675488430095608,
|
| 567 |
+
"eval_bleu": 0.17126127820473988,
|
| 568 |
+
"eval_cos_loss": 0.1419908434628896,
|
| 569 |
+
"eval_loss": 0.05134654454206495,
|
| 570 |
+
"eval_mse_loss": 0.05134654454206495,
|
| 571 |
+
"eval_runtime": 176.0149,
|
| 572 |
+
"eval_samples_per_second": 159.038,
|
| 573 |
+
"eval_steps_per_second": 2.488,
|
| 574 |
+
"step": 12288
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 0.5793727772389267,
|
| 578 |
+
"grad_norm": 0.15192070603370667,
|
| 579 |
+
"learning_rate": 0.0003922507884228551,
|
| 580 |
+
"loss": 0.0504397489130497,
|
| 581 |
+
"step": 12544
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.5911967114682924,
|
| 585 |
+
"grad_norm": 0.1474185436964035,
|
| 586 |
+
"learning_rate": 0.00037375735713457723,
|
| 587 |
+
"loss": 0.05017136037349701,
|
| 588 |
+
"step": 12800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 0.6030206456976583,
|
| 592 |
+
"grad_norm": 0.12859590351581573,
|
| 593 |
+
"learning_rate": 0.00035544663676018276,
|
| 594 |
+
"loss": 0.049893494695425034,
|
| 595 |
+
"step": 13056
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.6148445799270241,
|
| 599 |
+
"grad_norm": 0.11801755428314209,
|
| 600 |
+
"learning_rate": 0.00033734512839611255,
|
| 601 |
+
"loss": 0.04954931512475014,
|
| 602 |
+
"step": 13312
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.6148445799270241,
|
| 606 |
+
"eval_bleu": 0.1800444095839781,
|
| 607 |
+
"eval_cos_loss": 0.1376905629419845,
|
| 608 |
+
"eval_loss": 0.04945668337449908,
|
| 609 |
+
"eval_mse_loss": 0.04945668337449908,
|
| 610 |
+
"step": 13312
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.6148445799270241,
|
| 614 |
+
"eval_bleu": 0.1800444095839781,
|
| 615 |
+
"eval_cos_loss": 0.1376905629419845,
|
| 616 |
+
"eval_loss": 0.04945668337449908,
|
| 617 |
+
"eval_mse_loss": 0.04945668337449908,
|
| 618 |
+
"eval_runtime": 174.7618,
|
| 619 |
+
"eval_samples_per_second": 160.178,
|
| 620 |
+
"eval_steps_per_second": 2.506,
|
| 621 |
+
"step": 13312
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.62666851415639,
|
| 625 |
+
"grad_norm": 0.14720351994037628,
|
| 626 |
+
"learning_rate": 0.0003194790303463687,
|
| 627 |
+
"loss": 0.04919987544417381,
|
| 628 |
+
"step": 13568
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 0.6384924483857558,
|
| 632 |
+
"grad_norm": 0.10936211049556732,
|
| 633 |
+
"learning_rate": 0.00030187420020572406,
|
| 634 |
+
"loss": 0.048761751502752304,
|
| 635 |
+
"step": 13824
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 0.6503163826151217,
|
| 639 |
+
"grad_norm": 0.10678113251924515,
|
| 640 |
+
"learning_rate": 0.00028455611743603626,
|
| 641 |
+
"loss": 0.04845006763935089,
|
| 642 |
+
"step": 14080
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.6621403168444876,
|
| 646 |
+
"grad_norm": 0.12803645431995392,
|
| 647 |
+
"learning_rate": 0.0002675498464898373,
|
| 648 |
+
"loss": 0.04825146496295929,
|
| 649 |
+
"step": 14336
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.6621403168444876,
|
| 653 |
+
"eval_bleu": 0.19045355889644444,
|
| 654 |
+
"eval_cos_loss": 0.13523258731517618,
|
| 655 |
+
"eval_loss": 0.04841839778130729,
|
| 656 |
+
"eval_mse_loss": 0.04841839778130729,
|
| 657 |
+
"step": 14336
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.6621403168444876,
|
| 661 |
+
"eval_bleu": 0.19045355889644444,
|
| 662 |
+
"eval_cos_loss": 0.13523258731517618,
|
| 663 |
+
"eval_loss": 0.04841839778130729,
|
| 664 |
+
"eval_mse_loss": 0.04841839778130729,
|
| 665 |
+
"eval_runtime": 175.5444,
|
| 666 |
+
"eval_samples_per_second": 159.464,
|
| 667 |
+
"eval_steps_per_second": 2.495,
|
| 668 |
+
"step": 14336
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.6739642510738534,
|
| 672 |
+
"grad_norm": 0.09000962227582932,
|
| 673 |
+
"learning_rate": 0.0002508800005345623,
|
| 674 |
+
"loss": 0.04815007746219635,
|
| 675 |
+
"step": 14592
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.6857881853032193,
|
| 679 |
+
"grad_norm": 0.11037364602088928,
|
| 680 |
+
"learning_rate": 0.00023457070582992562,
|
| 681 |
+
"loss": 0.04773465543985367,
|
| 682 |
+
"step": 14848
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.6976121195325851,
|
| 686 |
+
"grad_norm": 0.11829054355621338,
|
| 687 |
+
"learning_rate": 0.00021864556680999692,
|
| 688 |
+
"loss": 0.04770495742559433,
|
| 689 |
+
"step": 15104
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.709436053761951,
|
| 693 |
+
"grad_norm": 0.12001141160726547,
|
| 694 |
+
"learning_rate": 0.0002031276319205152,
|
| 695 |
+
"loss": 0.0472663938999176,
|
| 696 |
+
"step": 15360
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.709436053761951,
|
| 700 |
+
"eval_bleu": 0.19616135450927907,
|
| 701 |
+
"eval_cos_loss": 0.13301836342027742,
|
| 702 |
+
"eval_loss": 0.04749709167879194,
|
| 703 |
+
"eval_mse_loss": 0.04749709167879194,
|
| 704 |
+
"step": 15360
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.709436053761951,
|
| 708 |
+
"eval_bleu": 0.19616135450927907,
|
| 709 |
+
"eval_cos_loss": 0.13301836342027742,
|
| 710 |
+
"eval_loss": 0.04749709167879194,
|
| 711 |
+
"eval_mse_loss": 0.04749709167879194,
|
| 712 |
+
"eval_runtime": 171.0818,
|
| 713 |
+
"eval_samples_per_second": 163.623,
|
| 714 |
+
"eval_steps_per_second": 2.56,
|
| 715 |
+
"step": 15360
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 0.7212599879913169,
|
| 719 |
+
"grad_norm": 0.10166730731725693,
|
| 720 |
+
"learning_rate": 0.00018803936026088542,
|
| 721 |
+
"loss": 0.04720830172300339,
|
| 722 |
+
"step": 15616
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.7330839222206826,
|
| 726 |
+
"grad_norm": 0.08360277861356735,
|
| 727 |
+
"learning_rate": 0.00017340258907913464,
|
| 728 |
+
"loss": 0.047153279185295105,
|
| 729 |
+
"step": 15872
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.7449078564500485,
|
| 733 |
+
"grad_norm": 0.11674193292856216,
|
| 734 |
+
"learning_rate": 0.0001592385021668743,
|
| 735 |
+
"loss": 0.04698365181684494,
|
| 736 |
+
"step": 16128
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 0.7567317906794143,
|
| 740 |
+
"grad_norm": 0.10673261433839798,
|
| 741 |
+
"learning_rate": 0.0001455675992000087,
|
| 742 |
+
"loss": 0.046850256621837616,
|
| 743 |
+
"step": 16384
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 0.7567317906794143,
|
| 747 |
+
"eval_bleu": 0.2010705716047647,
|
| 748 |
+
"eval_cos_loss": 0.13100612456123578,
|
| 749 |
+
"eval_loss": 0.04669262347190075,
|
| 750 |
+
"eval_mse_loss": 0.04669262347190075,
|
| 751 |
+
"step": 16384
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.7567317906794143,
|
| 755 |
+
"eval_bleu": 0.2010705716047647,
|
| 756 |
+
"eval_cos_loss": 0.13100612456123578,
|
| 757 |
+
"eval_loss": 0.04669262347190075,
|
| 758 |
+
"eval_mse_loss": 0.04669262347190075,
|
| 759 |
+
"eval_runtime": 170.6282,
|
| 760 |
+
"eval_samples_per_second": 164.058,
|
| 761 |
+
"eval_steps_per_second": 2.567,
|
| 762 |
+
"step": 16384
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.7685557249087802,
|
| 766 |
+
"grad_norm": 0.09755754470825195,
|
| 767 |
+
"learning_rate": 0.000132409666069565,
|
| 768 |
+
"loss": 0.04642617702484131,
|
| 769 |
+
"step": 16640
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.780379659138146,
|
| 773 |
+
"grad_norm": 0.11503065377473831,
|
| 774 |
+
"learning_rate": 0.0001197837462455823,
|
| 775 |
+
"loss": 0.04628568887710571,
|
| 776 |
+
"step": 16896
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.7922035933675119,
|
| 780 |
+
"grad_norm": 0.09499657154083252,
|
| 781 |
+
"learning_rate": 0.00010770811321550749,
|
| 782 |
+
"loss": 0.04594754800200462,
|
| 783 |
+
"step": 17152
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.8040275275968778,
|
| 787 |
+
"grad_norm": 0.09881823509931564,
|
| 788 |
+
"learning_rate": 9.620024403698591e-05,
|
| 789 |
+
"loss": 0.04607144370675087,
|
| 790 |
+
"step": 17408
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 0.8040275275968778,
|
| 794 |
+
"eval_bleu": 0.20514783220235264,
|
| 795 |
+
"eval_cos_loss": 0.12981136223156703,
|
| 796 |
+
"eval_loss": 0.046230202098483365,
|
| 797 |
+
"eval_mse_loss": 0.046230202098483365,
|
| 798 |
+
"step": 17408
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 0.8040275275968778,
|
| 802 |
+
"eval_bleu": 0.20514783220235264,
|
| 803 |
+
"eval_cos_loss": 0.12981136223156703,
|
| 804 |
+
"eval_loss": 0.046230202098483365,
|
| 805 |
+
"eval_mse_loss": 0.046230202098483365,
|
| 806 |
+
"eval_runtime": 166.5211,
|
| 807 |
+
"eval_samples_per_second": 168.105,
|
| 808 |
+
"eval_steps_per_second": 2.63,
|
| 809 |
+
"step": 17408
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.8158514618262436,
|
| 813 |
+
"grad_norm": 0.12171804159879684,
|
| 814 |
+
"learning_rate": 8.527679404332429e-05,
|
| 815 |
+
"loss": 0.04588912054896355,
|
| 816 |
+
"step": 17664
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.8276753960556095,
|
| 820 |
+
"grad_norm": 0.09019502252340317,
|
| 821 |
+
"learning_rate": 7.495357273823544e-05,
|
| 822 |
+
"loss": 0.0456986203789711,
|
| 823 |
+
"step": 17920
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.8394993302849753,
|
| 827 |
+
"grad_norm": 0.09827019274234772,
|
| 828 |
+
"learning_rate": 6.524552091475183e-05,
|
| 829 |
+
"loss": 0.04577944055199623,
|
| 830 |
+
"step": 18176
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.8513232645143411,
|
| 834 |
+
"grad_norm": 0.088897205889225,
|
| 835 |
+
"learning_rate": 5.6166689031422024e-05,
|
| 836 |
+
"loss": 0.04551319777965546,
|
| 837 |
+
"step": 18432
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.8513232645143411,
|
| 841 |
+
"eval_bleu": 0.20781379096212793,
|
| 842 |
+
"eval_cos_loss": 0.12865090991059938,
|
| 843 |
+
"eval_loss": 0.04577279597792996,
|
| 844 |
+
"eval_mse_loss": 0.04577279597792996,
|
| 845 |
+
"step": 18432
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.8513232645143411,
|
| 849 |
+
"eval_bleu": 0.20781379096212793,
|
| 850 |
+
"eval_cos_loss": 0.12865090991059938,
|
| 851 |
+
"eval_loss": 0.04577279597792996,
|
| 852 |
+
"eval_mse_loss": 0.04577279597792996,
|
| 853 |
+
"eval_runtime": 166.8661,
|
| 854 |
+
"eval_samples_per_second": 167.757,
|
| 855 |
+
"eval_steps_per_second": 2.625,
|
| 856 |
+
"step": 18432
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.8631471987437069,
|
| 860 |
+
"grad_norm": 0.09188670665025711,
|
| 861 |
+
"learning_rate": 4.773021687709067e-05,
|
| 862 |
+
"loss": 0.04544902220368385,
|
| 863 |
+
"step": 18688
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.8749711329730728,
|
| 867 |
+
"grad_norm": 0.07088550925254822,
|
| 868 |
+
"learning_rate": 3.994831455368719e-05,
|
| 869 |
+
"loss": 0.04544632509350777,
|
| 870 |
+
"step": 18944
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.8867950672024387,
|
| 874 |
+
"grad_norm": 0.07616105675697327,
|
| 875 |
+
"learning_rate": 3.283224480455282e-05,
|
| 876 |
+
"loss": 0.04547519609332085,
|
| 877 |
+
"step": 19200
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.8986190014318045,
|
| 881 |
+
"grad_norm": 0.07662102580070496,
|
| 882 |
+
"learning_rate": 2.639230671387627e-05,
|
| 883 |
+
"loss": 0.0451703704893589,
|
| 884 |
+
"step": 19456
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.8986190014318045,
|
| 888 |
+
"eval_bleu": 0.20950400984229536,
|
| 889 |
+
"eval_cos_loss": 0.12795958582941255,
|
| 890 |
+
"eval_loss": 0.04547639413137142,
|
| 891 |
+
"eval_mse_loss": 0.04547639413137142,
|
| 892 |
+
"step": 19456
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.8986190014318045,
|
| 896 |
+
"eval_bleu": 0.20950400984229536,
|
| 897 |
+
"eval_cos_loss": 0.12795958582941255,
|
| 898 |
+
"eval_loss": 0.04547639413137142,
|
| 899 |
+
"eval_mse_loss": 0.04547639413137142,
|
| 900 |
+
"eval_runtime": 166.6514,
|
| 901 |
+
"eval_samples_per_second": 167.973,
|
| 902 |
+
"eval_steps_per_second": 2.628,
|
| 903 |
+
"step": 19456
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.9104429356611704,
|
| 907 |
+
"grad_norm": 0.07200378179550171,
|
| 908 |
+
"learning_rate": 2.063782080083576e-05,
|
| 909 |
+
"loss": 0.045278534293174744,
|
| 910 |
+
"step": 19712
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.9222668698905362,
|
| 914 |
+
"grad_norm": 0.07066826522350311,
|
| 915 |
+
"learning_rate": 1.557711553001523e-05,
|
| 916 |
+
"loss": 0.04532043635845184,
|
| 917 |
+
"step": 19968
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.9340908041199021,
|
| 921 |
+
"grad_norm": 0.06178949028253555,
|
| 922 |
+
"learning_rate": 1.1217515257622269e-05,
|
| 923 |
+
"loss": 0.04507835581898689,
|
| 924 |
+
"step": 20224
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.945914738349268,
|
| 928 |
+
"grad_norm": 0.061289627104997635,
|
| 929 |
+
"learning_rate": 7.565329630950746e-06,
|
| 930 |
+
"loss": 0.04517170414328575,
|
| 931 |
+
"step": 20480
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 0.945914738349268,
|
| 935 |
+
"eval_bleu": 0.21038975756702644,
|
| 936 |
+
"eval_cos_loss": 0.12753397386231924,
|
| 937 |
+
"eval_loss": 0.04531074096725139,
|
| 938 |
+
"eval_mse_loss": 0.04531074096725139,
|
| 939 |
+
"step": 20480
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.945914738349268,
|
| 943 |
+
"eval_bleu": 0.21038975756702644,
|
| 944 |
+
"eval_cos_loss": 0.12753397386231924,
|
| 945 |
+
"eval_loss": 0.04531074096725139,
|
| 946 |
+
"eval_mse_loss": 0.04531074096725139,
|
| 947 |
+
"eval_runtime": 169.1109,
|
| 948 |
+
"eval_samples_per_second": 165.53,
|
| 949 |
+
"eval_steps_per_second": 2.59,
|
| 950 |
+
"step": 20480
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.9577386725786338,
|
| 954 |
+
"grad_norm": 0.05845247954130173,
|
| 955 |
+
"learning_rate": 4.62584445643166e-06,
|
| 956 |
+
"loss": 0.04498561471700668,
|
| 957 |
+
"step": 20736
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.9695626068079997,
|
| 961 |
+
"grad_norm": 0.05815529450774193,
|
| 962 |
+
"learning_rate": 2.40331404948807e-06,
|
| 963 |
+
"loss": 0.04522498697042465,
|
| 964 |
+
"step": 20992
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.9813865410373654,
|
| 968 |
+
"grad_norm": 0.05924896523356438,
|
| 969 |
+
"learning_rate": 9.009550772663965e-07,
|
| 970 |
+
"loss": 0.04517938941717148,
|
| 971 |
+
"step": 21248
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.9932104752667313,
|
| 975 |
+
"grad_norm": 0.060702189803123474,
|
| 976 |
+
"learning_rate": 1.2094190315575791e-07,
|
| 977 |
+
"loss": 0.04506324976682663,
|
| 978 |
+
"step": 21504
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.9932104752667313,
|
| 982 |
+
"eval_bleu": 0.21059100520469226,
|
| 983 |
+
"eval_cos_loss": 0.12752766677573935,
|
| 984 |
+
"eval_loss": 0.045272438553030085,
|
| 985 |
+
"eval_mse_loss": 0.045272438553030085,
|
| 986 |
+
"step": 21504
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.9932104752667313,
|
| 990 |
+
"eval_bleu": 0.21059100520469226,
|
| 991 |
+
"eval_cos_loss": 0.12752766677573935,
|
| 992 |
+
"eval_loss": 0.045272438553030085,
|
| 993 |
+
"eval_mse_loss": 0.045272438553030085,
|
| 994 |
+
"eval_runtime": 169.8109,
|
| 995 |
+
"eval_samples_per_second": 164.848,
|
| 996 |
+
"eval_steps_per_second": 2.579,
|
| 997 |
+
"step": 21504
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.0,
|
| 1001 |
+
"eval_bleu": 0.21049399466110846,
|
| 1002 |
+
"eval_cos_loss": 0.1273856761963128,
|
| 1003 |
+
"eval_loss": 0.045260803025539066,
|
| 1004 |
+
"eval_mse_loss": 0.045260803025539066,
|
| 1005 |
+
"step": 21651
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.0,
|
| 1009 |
+
"eval_bleu": 0.21049399466110846,
|
| 1010 |
+
"eval_cos_loss": 0.1273856761963128,
|
| 1011 |
+
"eval_loss": 0.045260803025539066,
|
| 1012 |
+
"eval_mse_loss": 0.045260803025539066,
|
| 1013 |
+
"eval_runtime": 170.6509,
|
| 1014 |
+
"eval_samples_per_second": 164.037,
|
| 1015 |
+
"eval_steps_per_second": 2.567,
|
| 1016 |
+
"step": 21651
|
| 1017 |
+
}
|
| 1018 |
+
],
|
| 1019 |
+
"logging_steps": 256,
|
| 1020 |
+
"max_steps": 21651,
|
| 1021 |
+
"num_input_tokens_seen": 0,
|
| 1022 |
+
"num_train_epochs": 1,
|
| 1023 |
+
"save_steps": 1024,
|
| 1024 |
+
"stateful_callbacks": {
|
| 1025 |
+
"TrainerControl": {
|
| 1026 |
+
"args": {
|
| 1027 |
+
"should_epoch_stop": false,
|
| 1028 |
+
"should_evaluate": false,
|
| 1029 |
+
"should_log": false,
|
| 1030 |
+
"should_save": true,
|
| 1031 |
+
"should_training_stop": true
|
| 1032 |
+
},
|
| 1033 |
+
"attributes": {}
|
| 1034 |
+
}
|
| 1035 |
+
},
|
| 1036 |
+
"total_flos": 0.0,
|
| 1037 |
+
"train_batch_size": 64,
|
| 1038 |
+
"trial_name": null,
|
| 1039 |
+
"trial_params": null
|
| 1040 |
+
}
|
checkpoints-v1.0e-discrete-condition/checkpoint-21651/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f63a8a711fb280c6320e95476c532b75c862f31d6a218a0d05079e32411a5ae1
|
| 3 |
+
size 5201
|