Upload folder using huggingface_hub
Browse files- checkpoints-v3.2/checkpoint-21504/ema.safetensors +3 -0
- checkpoints-v3.2/checkpoint-21504/model.safetensors +3 -0
- checkpoints-v3.2/checkpoint-21504/optimizer.pt +3 -0
- checkpoints-v3.2/checkpoint-21504/rng_state.pth +3 -0
- checkpoints-v3.2/checkpoint-21504/scaler.pt +3 -0
- checkpoints-v3.2/checkpoint-21504/scheduler.pt +3 -0
- checkpoints-v3.2/checkpoint-21504/trainer_state.json +1147 -0
- checkpoints-v3.2/checkpoint-21504/training_args.bin +3 -0
checkpoints-v3.2/checkpoint-21504/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a72d67bb1a9cfda1fddec452237a1a7afc224e23f1154d15a16e5a7e76efae8
|
| 3 |
+
size 550088
|
checkpoints-v3.2/checkpoint-21504/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d4b64d3e6f0497350952e55b852eb43a8ae3a512e150658b58fafb7b6715d77
|
| 3 |
+
size 19318680
|
checkpoints-v3.2/checkpoint-21504/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cacc69247a8015ea9dfec72c42ff929daee640cac4ad59816a13d67c75f31e46
|
| 3 |
+
size 1176715
|
checkpoints-v3.2/checkpoint-21504/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76811559f30210107011770fcc2e9451824642c3757267b25124ab0a130c76df
|
| 3 |
+
size 14645
|
checkpoints-v3.2/checkpoint-21504/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:deaf207ec3fded3cf6c58e143beab5b5694180a046d9629f3100365d1c0582d0
|
| 3 |
+
size 1383
|
checkpoints-v3.2/checkpoint-21504/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8d0f20f63f45b1b1775e9f97d9c6a1a2904fa04b7d5a41f30f11b9aef0479eb
|
| 3 |
+
size 1465
|
checkpoints-v3.2/checkpoint-21504/trainer_state.json
ADDED
|
@@ -0,0 +1,1147 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.9932104752667313,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 21504,
|
| 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": 0.17750050127506256,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 1.7531359195709229,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.14886082708835602,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 1.120242953300476,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.11476853489875793,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 1.0460094213485718,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.26101624965667725,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 1.0132286548614502,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.3147891739459887,
|
| 43 |
+
"eval_loss": 0.9696788661283989,
|
| 44 |
+
"eval_mse_loss": 0.9696788661283989,
|
| 45 |
+
"flow/cos_sim": 0.6852108316334415,
|
| 46 |
+
"flow/improvement_ratio": 0.4729253878332164,
|
| 47 |
+
"flow/mag_ratio_mean": 0.6834825821663146,
|
| 48 |
+
"flow/mag_ratio_std": 0.18169553397589078,
|
| 49 |
+
"step": 1024
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.047295736917463395,
|
| 53 |
+
"eval_cos_loss": 0.3147891739459887,
|
| 54 |
+
"eval_loss": 0.9696788661283989,
|
| 55 |
+
"eval_mse_loss": 0.9696788661283989,
|
| 56 |
+
"eval_runtime": 37.5026,
|
| 57 |
+
"eval_samples_per_second": 746.427,
|
| 58 |
+
"eval_steps_per_second": 11.679,
|
| 59 |
+
"flow/cos_sim": 0.6852108316334415,
|
| 60 |
+
"flow/improvement_ratio": 0.4729253878332164,
|
| 61 |
+
"flow/mag_ratio_mean": 0.6834825821663146,
|
| 62 |
+
"flow/mag_ratio_std": 0.18169553397589078,
|
| 63 |
+
"step": 1024
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.05911967114682925,
|
| 67 |
+
"grad_norm": 0.32471492886543274,
|
| 68 |
+
"learning_rate": 0.0009967551747861387,
|
| 69 |
+
"loss": 0.9918397068977356,
|
| 70 |
+
"step": 1280
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.0709436053761951,
|
| 74 |
+
"grad_norm": 0.19580507278442383,
|
| 75 |
+
"learning_rate": 0.000994232528651847,
|
| 76 |
+
"loss": 0.9722790718078613,
|
| 77 |
+
"step": 1536
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08276753960556095,
|
| 81 |
+
"grad_norm": 0.13594156503677368,
|
| 82 |
+
"learning_rate": 0.0009909945800260092,
|
| 83 |
+
"loss": 0.9547147750854492,
|
| 84 |
+
"step": 1792
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09459147383492679,
|
| 88 |
+
"grad_norm": 0.49703535437583923,
|
| 89 |
+
"learning_rate": 0.0009870460151900522,
|
| 90 |
+
"loss": 0.9414160847663879,
|
| 91 |
+
"step": 2048
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.09459147383492679,
|
| 95 |
+
"eval_cos_loss": 0.2932011633430986,
|
| 96 |
+
"eval_loss": 0.8981472950547797,
|
| 97 |
+
"eval_mse_loss": 0.8981472950547797,
|
| 98 |
+
"flow/cos_sim": 0.7067988623767139,
|
| 99 |
+
"flow/improvement_ratio": 0.47438520363204556,
|
| 100 |
+
"flow/mag_ratio_mean": 0.6956081727868346,
|
| 101 |
+
"flow/mag_ratio_std": 0.19316728430115468,
|
| 102 |
+
"step": 2048
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.09459147383492679,
|
| 106 |
+
"eval_cos_loss": 0.2932011633430986,
|
| 107 |
+
"eval_loss": 0.8981472950547797,
|
| 108 |
+
"eval_mse_loss": 0.8981472950547797,
|
| 109 |
+
"eval_runtime": 37.712,
|
| 110 |
+
"eval_samples_per_second": 742.284,
|
| 111 |
+
"eval_steps_per_second": 11.614,
|
| 112 |
+
"flow/cos_sim": 0.7067988623767139,
|
| 113 |
+
"flow/improvement_ratio": 0.47438520363204556,
|
| 114 |
+
"flow/mag_ratio_mean": 0.6956081727868346,
|
| 115 |
+
"flow/mag_ratio_std": 0.19316728430115468,
|
| 116 |
+
"step": 2048
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.10641540806429264,
|
| 120 |
+
"grad_norm": 0.49009642004966736,
|
| 121 |
+
"learning_rate": 0.0009823925488998885,
|
| 122 |
+
"loss": 0.929905116558075,
|
| 123 |
+
"step": 2304
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1182393422936585,
|
| 127 |
+
"grad_norm": 0.21833102405071259,
|
| 128 |
+
"learning_rate": 0.0009770409161149525,
|
| 129 |
+
"loss": 0.9224135875701904,
|
| 130 |
+
"step": 2560
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.13006327652302435,
|
| 134 |
+
"grad_norm": 0.3465236723423004,
|
| 135 |
+
"learning_rate": 0.0009709988622506973,
|
| 136 |
+
"loss": 0.9120264649391174,
|
| 137 |
+
"step": 2816
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1418872107523902,
|
| 141 |
+
"grad_norm": 0.3057152032852173,
|
| 142 |
+
"learning_rate": 0.000964275131968659,
|
| 143 |
+
"loss": 0.9054349064826965,
|
| 144 |
+
"step": 3072
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.1418872107523902,
|
| 148 |
+
"eval_cos_loss": 0.28495094285571954,
|
| 149 |
+
"eval_loss": 0.8664921860988826,
|
| 150 |
+
"eval_mse_loss": 0.8664921860988826,
|
| 151 |
+
"flow/cos_sim": 0.7150490750733032,
|
| 152 |
+
"flow/improvement_ratio": 0.48076002510715293,
|
| 153 |
+
"flow/mag_ratio_mean": 0.7135201504785721,
|
| 154 |
+
"flow/mag_ratio_std": 0.19565439009911392,
|
| 155 |
+
"step": 3072
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1418872107523902,
|
| 159 |
+
"eval_cos_loss": 0.28495094285571954,
|
| 160 |
+
"eval_loss": 0.8664921860988826,
|
| 161 |
+
"eval_mse_loss": 0.8664921860988826,
|
| 162 |
+
"eval_runtime": 38.3291,
|
| 163 |
+
"eval_samples_per_second": 730.333,
|
| 164 |
+
"eval_steps_per_second": 11.427,
|
| 165 |
+
"flow/cos_sim": 0.7150490750733032,
|
| 166 |
+
"flow/improvement_ratio": 0.48076002510715293,
|
| 167 |
+
"flow/mag_ratio_mean": 0.7135201504785721,
|
| 168 |
+
"flow/mag_ratio_std": 0.19565439009911392,
|
| 169 |
+
"step": 3072
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.15371114498175603,
|
| 173 |
+
"grad_norm": 0.599277138710022,
|
| 174 |
+
"learning_rate": 0.0009568794565203123,
|
| 175 |
+
"loss": 0.9003790020942688,
|
| 176 |
+
"step": 3328
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.1655350792111219,
|
| 180 |
+
"grad_norm": 0.6535385251045227,
|
| 181 |
+
"learning_rate": 0.0009488225396630347,
|
| 182 |
+
"loss": 0.897746205329895,
|
| 183 |
+
"step": 3584
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17735901344048774,
|
| 187 |
+
"grad_norm": 0.4937800467014313,
|
| 188 |
+
"learning_rate": 0.0009401160421685646,
|
| 189 |
+
"loss": 0.8885043859481812,
|
| 190 |
+
"step": 3840
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18918294766985358,
|
| 194 |
+
"grad_norm": 1.3498679399490356,
|
| 195 |
+
"learning_rate": 0.0009307725649463714,
|
| 196 |
+
"loss": 0.8906658887863159,
|
| 197 |
+
"step": 4096
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.18918294766985358,
|
| 201 |
+
"eval_cos_loss": 0.281152941626773,
|
| 202 |
+
"eval_loss": 0.8519824494237769,
|
| 203 |
+
"eval_mse_loss": 0.8519824494237769,
|
| 204 |
+
"flow/cos_sim": 0.718847086304399,
|
| 205 |
+
"flow/improvement_ratio": 0.47086378285601804,
|
| 206 |
+
"flow/mag_ratio_mean": 0.7225797281145505,
|
| 207 |
+
"flow/mag_ratio_std": 0.19704299981463447,
|
| 208 |
+
"step": 4096
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"eval_cos_loss": 0.281152941626773,
|
| 213 |
+
"eval_loss": 0.8519824494237769,
|
| 214 |
+
"eval_mse_loss": 0.8519824494237769,
|
| 215 |
+
"eval_runtime": 37.6235,
|
| 216 |
+
"eval_samples_per_second": 744.029,
|
| 217 |
+
"eval_steps_per_second": 11.642,
|
| 218 |
+
"flow/cos_sim": 0.718847086304399,
|
| 219 |
+
"flow/improvement_ratio": 0.47086378285601804,
|
| 220 |
+
"flow/mag_ratio_mean": 0.7225797281145505,
|
| 221 |
+
"flow/mag_ratio_std": 0.19704299981463447,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.20100688189921945,
|
| 226 |
+
"grad_norm": 0.24068038165569305,
|
| 227 |
+
"learning_rate": 0.0009208056308063659,
|
| 228 |
+
"loss": 0.885456383228302,
|
| 229 |
+
"step": 4352
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.2128308161285853,
|
| 233 |
+
"grad_norm": 0.33502769470214844,
|
| 234 |
+
"learning_rate": 0.0009102296648873445,
|
| 235 |
+
"loss": 0.8807857632637024,
|
| 236 |
+
"step": 4608
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.22465475035795113,
|
| 240 |
+
"grad_norm": 0.176764577627182,
|
| 241 |
+
"learning_rate": 0.0008990599737794927,
|
| 242 |
+
"loss": 0.8787178993225098,
|
| 243 |
+
"step": 4864
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.236478684587317,
|
| 247 |
+
"grad_norm": 0.1856887936592102,
|
| 248 |
+
"learning_rate": 0.0008873127233711644,
|
| 249 |
+
"loss": 0.8718249797821045,
|
| 250 |
+
"step": 5120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.236478684587317,
|
| 254 |
+
"eval_cos_loss": 0.2779194979022627,
|
| 255 |
+
"eval_loss": 0.8412302969253227,
|
| 256 |
+
"eval_mse_loss": 0.8412302969253227,
|
| 257 |
+
"flow/cos_sim": 0.7220805322742898,
|
| 258 |
+
"flow/improvement_ratio": 0.4709661888749632,
|
| 259 |
+
"flow/mag_ratio_mean": 0.7200435751105008,
|
| 260 |
+
"flow/mag_ratio_std": 0.19833092303036554,
|
| 261 |
+
"step": 5120
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.236478684587317,
|
| 265 |
+
"eval_cos_loss": 0.2779194979022627,
|
| 266 |
+
"eval_loss": 0.8412302969253227,
|
| 267 |
+
"eval_mse_loss": 0.8412302969253227,
|
| 268 |
+
"eval_runtime": 37.8514,
|
| 269 |
+
"eval_samples_per_second": 739.549,
|
| 270 |
+
"eval_steps_per_second": 11.572,
|
| 271 |
+
"flow/cos_sim": 0.7220805322742898,
|
| 272 |
+
"flow/improvement_ratio": 0.4709661888749632,
|
| 273 |
+
"flow/mag_ratio_mean": 0.7200435751105008,
|
| 274 |
+
"flow/mag_ratio_std": 0.19833092303036554,
|
| 275 |
+
"step": 5120
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24830261881668284,
|
| 279 |
+
"grad_norm": 0.315166711807251,
|
| 280 |
+
"learning_rate": 0.0008750049154520011,
|
| 281 |
+
"loss": 0.8694944977760315,
|
| 282 |
+
"step": 5376
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2601265530460487,
|
| 286 |
+
"grad_norm": 0.6202583909034729,
|
| 287 |
+
"learning_rate": 0.0008621543631062487,
|
| 288 |
+
"loss": 0.8698850870132446,
|
| 289 |
+
"step": 5632
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.27195048727541454,
|
| 293 |
+
"grad_norm": 0.154885396361351,
|
| 294 |
+
"learning_rate": 0.0008487796649318904,
|
| 295 |
+
"loss": 0.866125762462616,
|
| 296 |
+
"step": 5888
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2837744215047804,
|
| 300 |
+
"grad_norm": 0.1593707948923111,
|
| 301 |
+
"learning_rate": 0.0008349001781229053,
|
| 302 |
+
"loss": 0.8656928539276123,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_cos_loss": 0.27761552666556344,
|
| 308 |
+
"eval_loss": 0.8367355261763482,
|
| 309 |
+
"eval_mse_loss": 0.8367355261763482,
|
| 310 |
+
"flow/cos_sim": 0.7223844902428318,
|
| 311 |
+
"flow/improvement_ratio": 0.47543454585282224,
|
| 312 |
+
"flow/mag_ratio_mean": 0.7271437070685435,
|
| 313 |
+
"flow/mag_ratio_std": 0.2028201749063518,
|
| 314 |
+
"step": 6144
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2837744215047804,
|
| 318 |
+
"eval_cos_loss": 0.27761552666556344,
|
| 319 |
+
"eval_loss": 0.8367355261763482,
|
| 320 |
+
"eval_mse_loss": 0.8367355261763482,
|
| 321 |
+
"eval_runtime": 37.5912,
|
| 322 |
+
"eval_samples_per_second": 744.669,
|
| 323 |
+
"eval_steps_per_second": 11.652,
|
| 324 |
+
"flow/cos_sim": 0.7223844902428318,
|
| 325 |
+
"flow/improvement_ratio": 0.47543454585282224,
|
| 326 |
+
"flow/mag_ratio_mean": 0.7271437070685435,
|
| 327 |
+
"flow/mag_ratio_std": 0.2028201749063518,
|
| 328 |
+
"step": 6144
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2955983557341462,
|
| 332 |
+
"grad_norm": 0.24749045073986053,
|
| 333 |
+
"learning_rate": 0.0008205359904536107,
|
| 334 |
+
"loss": 0.8599100112915039,
|
| 335 |
+
"step": 6400
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.30742228996351206,
|
| 339 |
+
"grad_norm": 0.4366367757320404,
|
| 340 |
+
"learning_rate": 0.0008057078912056363,
|
| 341 |
+
"loss": 0.8599902987480164,
|
| 342 |
+
"step": 6656
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3192462241928779,
|
| 346 |
+
"grad_norm": 0.2574009597301483,
|
| 347 |
+
"learning_rate": 0.0007904373410796086,
|
| 348 |
+
"loss": 0.8590140342712402,
|
| 349 |
+
"step": 6912
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.3310701584222438,
|
| 353 |
+
"grad_norm": 0.24796369671821594,
|
| 354 |
+
"learning_rate": 0.0007747464411350876,
|
| 355 |
+
"loss": 0.8581823110580444,
|
| 356 |
+
"step": 7168
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.3310701584222438,
|
| 360 |
+
"eval_cos_loss": 0.27323219904752627,
|
| 361 |
+
"eval_loss": 0.8229639839091801,
|
| 362 |
+
"eval_mse_loss": 0.8229639839091801,
|
| 363 |
+
"flow/cos_sim": 0.7267678196299566,
|
| 364 |
+
"flow/improvement_ratio": 0.4771028495814702,
|
| 365 |
+
"flow/mag_ratio_mean": 0.7125569941246346,
|
| 366 |
+
"flow/mag_ratio_std": 0.20383604319944773,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3310701584222438,
|
| 371 |
+
"eval_cos_loss": 0.27323219904752627,
|
| 372 |
+
"eval_loss": 0.8229639839091801,
|
| 373 |
+
"eval_mse_loss": 0.8229639839091801,
|
| 374 |
+
"eval_runtime": 37.8121,
|
| 375 |
+
"eval_samples_per_second": 740.319,
|
| 376 |
+
"eval_steps_per_second": 11.584,
|
| 377 |
+
"flow/cos_sim": 0.7267678196299566,
|
| 378 |
+
"flow/improvement_ratio": 0.4771028495814702,
|
| 379 |
+
"flow/mag_ratio_mean": 0.7125569941246346,
|
| 380 |
+
"flow/mag_ratio_std": 0.20383604319944773,
|
| 381 |
+
"step": 7168
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.34289409265160964,
|
| 385 |
+
"grad_norm": 0.7159921526908875,
|
| 386 |
+
"learning_rate": 0.000758657900803716,
|
| 387 |
+
"loss": 0.858252763748169,
|
| 388 |
+
"step": 7424
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3547180268809755,
|
| 392 |
+
"grad_norm": 0.6132074594497681,
|
| 393 |
+
"learning_rate": 0.000742195005021869,
|
| 394 |
+
"loss": 0.8558468818664551,
|
| 395 |
+
"step": 7680
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3665419611103413,
|
| 399 |
+
"grad_norm": 0.18163549900054932,
|
| 400 |
+
"learning_rate": 0.0007253815805303786,
|
| 401 |
+
"loss": 0.85396808385849,
|
| 402 |
+
"step": 7936
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.37836589533970716,
|
| 406 |
+
"grad_norm": 0.572221040725708,
|
| 407 |
+
"learning_rate": 0.0007082419613901028,
|
| 408 |
+
"loss": 0.8530430197715759,
|
| 409 |
+
"step": 8192
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.37836589533970716,
|
| 413 |
+
"eval_cos_loss": 0.2715048284427216,
|
| 414 |
+
"eval_loss": 0.8155939645691005,
|
| 415 |
+
"eval_mse_loss": 0.8155939645691005,
|
| 416 |
+
"flow/cos_sim": 0.7284952074153238,
|
| 417 |
+
"flow/improvement_ratio": 0.47482473080016707,
|
| 418 |
+
"flow/mag_ratio_mean": 0.724112270765653,
|
| 419 |
+
"flow/mag_ratio_std": 0.2020510737210104,
|
| 420 |
+
"step": 8192
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.37836589533970716,
|
| 424 |
+
"eval_cos_loss": 0.2715048284427216,
|
| 425 |
+
"eval_loss": 0.8155939645691005,
|
| 426 |
+
"eval_mse_loss": 0.8155939645691005,
|
| 427 |
+
"eval_runtime": 37.7077,
|
| 428 |
+
"eval_samples_per_second": 742.367,
|
| 429 |
+
"eval_steps_per_second": 11.616,
|
| 430 |
+
"flow/cos_sim": 0.7284952074153238,
|
| 431 |
+
"flow/improvement_ratio": 0.47482473080016707,
|
| 432 |
+
"flow/mag_ratio_mean": 0.724112270765653,
|
| 433 |
+
"flow/mag_ratio_std": 0.2020510737210104,
|
| 434 |
+
"step": 8192
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.390189829569073,
|
| 438 |
+
"grad_norm": 0.2026418000459671,
|
| 439 |
+
"learning_rate": 0.0006908009537632514,
|
| 440 |
+
"loss": 0.8493704199790955,
|
| 441 |
+
"step": 8448
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4020137637984389,
|
| 445 |
+
"grad_norm": 0.5044918656349182,
|
| 446 |
+
"learning_rate": 0.0006730838000114403,
|
| 447 |
+
"loss": 0.8506228923797607,
|
| 448 |
+
"step": 8704
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.41383769802780473,
|
| 452 |
+
"grad_norm": 0.2458736002445221,
|
| 453 |
+
"learning_rate": 0.0006551161421624341,
|
| 454 |
+
"loss": 0.8482614159584045,
|
| 455 |
+
"step": 8960
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.4256616322571706,
|
| 459 |
+
"grad_norm": 0.7097583413124084,
|
| 460 |
+
"learning_rate": 0.0006369239847984517,
|
| 461 |
+
"loss": 0.846942126750946,
|
| 462 |
+
"step": 9216
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4256616322571706,
|
| 466 |
+
"eval_cos_loss": 0.27180098626575516,
|
| 467 |
+
"eval_loss": 0.8161548636003172,
|
| 468 |
+
"eval_mse_loss": 0.8161548636003172,
|
| 469 |
+
"flow/cos_sim": 0.7281990317993512,
|
| 470 |
+
"flow/improvement_ratio": 0.4732581986003815,
|
| 471 |
+
"flow/mag_ratio_mean": 0.7231416398803937,
|
| 472 |
+
"flow/mag_ratio_std": 0.201789186425405,
|
| 473 |
+
"step": 9216
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4256616322571706,
|
| 477 |
+
"eval_cos_loss": 0.27180098626575516,
|
| 478 |
+
"eval_loss": 0.8161548636003172,
|
| 479 |
+
"eval_mse_loss": 0.8161548636003172,
|
| 480 |
+
"eval_runtime": 37.5698,
|
| 481 |
+
"eval_samples_per_second": 745.094,
|
| 482 |
+
"eval_steps_per_second": 11.658,
|
| 483 |
+
"flow/cos_sim": 0.7281990317993512,
|
| 484 |
+
"flow/improvement_ratio": 0.4732581986003815,
|
| 485 |
+
"flow/mag_ratio_mean": 0.7231416398803937,
|
| 486 |
+
"flow/mag_ratio_std": 0.201789186425405,
|
| 487 |
+
"step": 9216
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.4374855664865364,
|
| 491 |
+
"grad_norm": 0.3167099952697754,
|
| 492 |
+
"learning_rate": 0.0006185336574197479,
|
| 493 |
+
"loss": 0.8480910062789917,
|
| 494 |
+
"step": 9472
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.44930950071590225,
|
| 498 |
+
"grad_norm": 0.4810751974582672,
|
| 499 |
+
"learning_rate": 0.0005999717763379407,
|
| 500 |
+
"loss": 0.8465522527694702,
|
| 501 |
+
"step": 9728
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4611334349452681,
|
| 505 |
+
"grad_norm": 0.2259974628686905,
|
| 506 |
+
"learning_rate": 0.0005812652061542363,
|
| 507 |
+
"loss": 0.844083309173584,
|
| 508 |
+
"step": 9984
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.472957369174634,
|
| 512 |
+
"grad_norm": 0.6505718231201172,
|
| 513 |
+
"learning_rate": 0.0005624410208783071,
|
| 514 |
+
"loss": 0.8436377644538879,
|
| 515 |
+
"step": 10240
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.472957369174634,
|
| 519 |
+
"eval_cos_loss": 0.2699208431047936,
|
| 520 |
+
"eval_loss": 0.808695926922097,
|
| 521 |
+
"eval_mse_loss": 0.808695926922097,
|
| 522 |
+
"flow/cos_sim": 0.7300791967677199,
|
| 523 |
+
"flow/improvement_ratio": 0.47688411740951886,
|
| 524 |
+
"flow/mag_ratio_mean": 0.7248269360359401,
|
| 525 |
+
"flow/mag_ratio_std": 0.20361059002544238,
|
| 526 |
+
"step": 10240
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.472957369174634,
|
| 530 |
+
"eval_cos_loss": 0.2699208431047936,
|
| 531 |
+
"eval_loss": 0.808695926922097,
|
| 532 |
+
"eval_mse_loss": 0.808695926922097,
|
| 533 |
+
"eval_runtime": 37.5891,
|
| 534 |
+
"eval_samples_per_second": 744.71,
|
| 535 |
+
"eval_steps_per_second": 11.652,
|
| 536 |
+
"flow/cos_sim": 0.7300791967677199,
|
| 537 |
+
"flow/improvement_ratio": 0.47688411740951886,
|
| 538 |
+
"flow/mag_ratio_mean": 0.7248269360359401,
|
| 539 |
+
"flow/mag_ratio_std": 0.20361059002544238,
|
| 540 |
+
"step": 10240
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.48478130340399983,
|
| 544 |
+
"grad_norm": 0.29194557666778564,
|
| 545 |
+
"learning_rate": 0.0005435264647440881,
|
| 546 |
+
"loss": 0.8419358730316162,
|
| 547 |
+
"step": 10496
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.49660523763336567,
|
| 551 |
+
"grad_norm": 0.20549535751342773,
|
| 552 |
+
"learning_rate": 0.000524548912779213,
|
| 553 |
+
"loss": 0.8400572538375854,
|
| 554 |
+
"step": 10752
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.5084291718627315,
|
| 558 |
+
"grad_norm": 0.7953479290008545,
|
| 559 |
+
"learning_rate": 0.0005055358311851499,
|
| 560 |
+
"loss": 0.8401579260826111,
|
| 561 |
+
"step": 11008
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5202531060920974,
|
| 565 |
+
"grad_norm": 0.14846356213092804,
|
| 566 |
+
"learning_rate": 0.0004865147375853812,
|
| 567 |
+
"loss": 0.840056836605072,
|
| 568 |
+
"step": 11264
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.5202531060920974,
|
| 572 |
+
"eval_cos_loss": 0.2704421652642559,
|
| 573 |
+
"eval_loss": 0.8102181156476339,
|
| 574 |
+
"eval_mse_loss": 0.8102181156476339,
|
| 575 |
+
"flow/cos_sim": 0.7295578481399849,
|
| 576 |
+
"flow/improvement_ratio": 0.4786316101273445,
|
| 577 |
+
"flow/mag_ratio_mean": 0.730471750780872,
|
| 578 |
+
"flow/mag_ratio_std": 0.2050654717368078,
|
| 579 |
+
"step": 11264
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.5202531060920974,
|
| 583 |
+
"eval_cos_loss": 0.2704421652642559,
|
| 584 |
+
"eval_loss": 0.8102181156476339,
|
| 585 |
+
"eval_mse_loss": 0.8102181156476339,
|
| 586 |
+
"eval_runtime": 37.702,
|
| 587 |
+
"eval_samples_per_second": 742.48,
|
| 588 |
+
"eval_steps_per_second": 11.617,
|
| 589 |
+
"flow/cos_sim": 0.7295578481399849,
|
| 590 |
+
"flow/improvement_ratio": 0.4786316101273445,
|
| 591 |
+
"flow/mag_ratio_mean": 0.730471750780872,
|
| 592 |
+
"flow/mag_ratio_std": 0.2050654717368078,
|
| 593 |
+
"step": 11264
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5320770403214632,
|
| 597 |
+
"grad_norm": 0.49196043610572815,
|
| 598 |
+
"learning_rate": 0.0004675131611991607,
|
| 599 |
+
"loss": 0.8376814723014832,
|
| 600 |
+
"step": 11520
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.5439009745508291,
|
| 604 |
+
"grad_norm": 0.3101634383201599,
|
| 605 |
+
"learning_rate": 0.0004485586029984899,
|
| 606 |
+
"loss": 0.8409138321876526,
|
| 607 |
+
"step": 11776
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.5557249087801949,
|
| 611 |
+
"grad_norm": 0.6860947608947754,
|
| 612 |
+
"learning_rate": 0.00042967849590597266,
|
| 613 |
+
"loss": 0.8401282429695129,
|
| 614 |
+
"step": 12032
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5675488430095608,
|
| 618 |
+
"grad_norm": 0.5976177453994751,
|
| 619 |
+
"learning_rate": 0.0004109001650911621,
|
| 620 |
+
"loss": 0.8376214504241943,
|
| 621 |
+
"step": 12288
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5675488430095608,
|
| 625 |
+
"eval_cos_loss": 0.2699217140742633,
|
| 626 |
+
"eval_loss": 0.8073501400479443,
|
| 627 |
+
"eval_mse_loss": 0.8073501400479443,
|
| 628 |
+
"flow/cos_sim": 0.7300783122239047,
|
| 629 |
+
"flow/improvement_ratio": 0.4786103154154129,
|
| 630 |
+
"flow/mag_ratio_mean": 0.7241527468102164,
|
| 631 |
+
"flow/mag_ratio_std": 0.20363660513946455,
|
| 632 |
+
"step": 12288
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5675488430095608,
|
| 636 |
+
"eval_cos_loss": 0.2699217140742633,
|
| 637 |
+
"eval_loss": 0.8073501400479443,
|
| 638 |
+
"eval_mse_loss": 0.8073501400479443,
|
| 639 |
+
"eval_runtime": 37.7433,
|
| 640 |
+
"eval_samples_per_second": 741.668,
|
| 641 |
+
"eval_steps_per_second": 11.605,
|
| 642 |
+
"flow/cos_sim": 0.7300783122239047,
|
| 643 |
+
"flow/improvement_ratio": 0.4786103154154129,
|
| 644 |
+
"flow/mag_ratio_mean": 0.7241527468102164,
|
| 645 |
+
"flow/mag_ratio_std": 0.20363660513946455,
|
| 646 |
+
"step": 12288
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5793727772389267,
|
| 650 |
+
"grad_norm": 0.19840270280838013,
|
| 651 |
+
"learning_rate": 0.0003922507884228551,
|
| 652 |
+
"loss": 0.8377624154090881,
|
| 653 |
+
"step": 12544
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5911967114682924,
|
| 657 |
+
"grad_norm": 0.6053916215896606,
|
| 658 |
+
"learning_rate": 0.00037375735713457723,
|
| 659 |
+
"loss": 0.838446855545044,
|
| 660 |
+
"step": 12800
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6030206456976583,
|
| 664 |
+
"grad_norm": 0.1777360439300537,
|
| 665 |
+
"learning_rate": 0.00035544663676018276,
|
| 666 |
+
"loss": 0.8392548561096191,
|
| 667 |
+
"step": 13056
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6148445799270241,
|
| 671 |
+
"grad_norm": 0.2625690698623657,
|
| 672 |
+
"learning_rate": 0.00033734512839611255,
|
| 673 |
+
"loss": 0.8352001309394836,
|
| 674 |
+
"step": 13312
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6148445799270241,
|
| 678 |
+
"eval_cos_loss": 0.2678832187568216,
|
| 679 |
+
"eval_loss": 0.8018131276516065,
|
| 680 |
+
"eval_mse_loss": 0.8018131276516065,
|
| 681 |
+
"flow/cos_sim": 0.732116795565984,
|
| 682 |
+
"flow/improvement_ratio": 0.47306563096231524,
|
| 683 |
+
"flow/mag_ratio_mean": 0.730011395259535,
|
| 684 |
+
"flow/mag_ratio_std": 0.20525332758007528,
|
| 685 |
+
"step": 13312
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.6148445799270241,
|
| 689 |
+
"eval_cos_loss": 0.2678832187568216,
|
| 690 |
+
"eval_loss": 0.8018131276516065,
|
| 691 |
+
"eval_mse_loss": 0.8018131276516065,
|
| 692 |
+
"eval_runtime": 37.6397,
|
| 693 |
+
"eval_samples_per_second": 743.71,
|
| 694 |
+
"eval_steps_per_second": 11.637,
|
| 695 |
+
"flow/cos_sim": 0.732116795565984,
|
| 696 |
+
"flow/improvement_ratio": 0.47306563096231524,
|
| 697 |
+
"flow/mag_ratio_mean": 0.730011395259535,
|
| 698 |
+
"flow/mag_ratio_std": 0.20525332758007528,
|
| 699 |
+
"step": 13312
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.62666851415639,
|
| 703 |
+
"grad_norm": 0.5689898729324341,
|
| 704 |
+
"learning_rate": 0.0003194790303463687,
|
| 705 |
+
"loss": 0.833480954170227,
|
| 706 |
+
"step": 13568
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.6384924483857558,
|
| 710 |
+
"grad_norm": 0.28950658440589905,
|
| 711 |
+
"learning_rate": 0.00030187420020572406,
|
| 712 |
+
"loss": 0.8348052501678467,
|
| 713 |
+
"step": 13824
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.6503163826151217,
|
| 717 |
+
"grad_norm": 0.15908953547477722,
|
| 718 |
+
"learning_rate": 0.00028455611743603626,
|
| 719 |
+
"loss": 0.834086000919342,
|
| 720 |
+
"step": 14080
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.6621403168444876,
|
| 724 |
+
"grad_norm": 0.15000976622104645,
|
| 725 |
+
"learning_rate": 0.0002675498464898373,
|
| 726 |
+
"loss": 0.8325310945510864,
|
| 727 |
+
"step": 14336
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.6621403168444876,
|
| 731 |
+
"eval_cos_loss": 0.2677479595898493,
|
| 732 |
+
"eval_loss": 0.8002720110492619,
|
| 733 |
+
"eval_mse_loss": 0.8002720110492619,
|
| 734 |
+
"flow/cos_sim": 0.73225207307023,
|
| 735 |
+
"flow/improvement_ratio": 0.474588285298108,
|
| 736 |
+
"flow/mag_ratio_mean": 0.732072616140592,
|
| 737 |
+
"flow/mag_ratio_std": 0.20331306650077915,
|
| 738 |
+
"step": 14336
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6621403168444876,
|
| 742 |
+
"eval_cos_loss": 0.2677479595898493,
|
| 743 |
+
"eval_loss": 0.8002720110492619,
|
| 744 |
+
"eval_mse_loss": 0.8002720110492619,
|
| 745 |
+
"eval_runtime": 37.36,
|
| 746 |
+
"eval_samples_per_second": 749.278,
|
| 747 |
+
"eval_steps_per_second": 11.724,
|
| 748 |
+
"flow/cos_sim": 0.73225207307023,
|
| 749 |
+
"flow/improvement_ratio": 0.474588285298108,
|
| 750 |
+
"flow/mag_ratio_mean": 0.732072616140592,
|
| 751 |
+
"flow/mag_ratio_std": 0.20331306650077915,
|
| 752 |
+
"step": 14336
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6739642510738534,
|
| 756 |
+
"grad_norm": 0.2764960825443268,
|
| 757 |
+
"learning_rate": 0.0002508800005345623,
|
| 758 |
+
"loss": 0.8334859609603882,
|
| 759 |
+
"step": 14592
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6857881853032193,
|
| 763 |
+
"grad_norm": 0.519814670085907,
|
| 764 |
+
"learning_rate": 0.00023457070582992562,
|
| 765 |
+
"loss": 0.8344730138778687,
|
| 766 |
+
"step": 14848
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6976121195325851,
|
| 770 |
+
"grad_norm": 0.22350260615348816,
|
| 771 |
+
"learning_rate": 0.00021864556680999692,
|
| 772 |
+
"loss": 0.8335193395614624,
|
| 773 |
+
"step": 15104
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.709436053761951,
|
| 777 |
+
"grad_norm": 0.128391832113266,
|
| 778 |
+
"learning_rate": 0.0002031276319205152,
|
| 779 |
+
"loss": 0.8319519758224487,
|
| 780 |
+
"step": 15360
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.709436053761951,
|
| 784 |
+
"eval_cos_loss": 0.26936231590977544,
|
| 785 |
+
"eval_loss": 0.803815519292605,
|
| 786 |
+
"eval_mse_loss": 0.803815519292605,
|
| 787 |
+
"flow/cos_sim": 0.7306376860294168,
|
| 788 |
+
"flow/improvement_ratio": 0.4827857036568803,
|
| 789 |
+
"flow/mag_ratio_mean": 0.7272341569809064,
|
| 790 |
+
"flow/mag_ratio_std": 0.20492048452706096,
|
| 791 |
+
"step": 15360
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.709436053761951,
|
| 795 |
+
"eval_cos_loss": 0.26936231590977544,
|
| 796 |
+
"eval_loss": 0.803815519292605,
|
| 797 |
+
"eval_mse_loss": 0.803815519292605,
|
| 798 |
+
"eval_runtime": 37.8308,
|
| 799 |
+
"eval_samples_per_second": 739.952,
|
| 800 |
+
"eval_steps_per_second": 11.578,
|
| 801 |
+
"flow/cos_sim": 0.7306376860294168,
|
| 802 |
+
"flow/improvement_ratio": 0.4827857036568803,
|
| 803 |
+
"flow/mag_ratio_mean": 0.7272341569809064,
|
| 804 |
+
"flow/mag_ratio_std": 0.20492048452706096,
|
| 805 |
+
"step": 15360
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.7212599879913169,
|
| 809 |
+
"grad_norm": 0.492816299200058,
|
| 810 |
+
"learning_rate": 0.00018803936026088542,
|
| 811 |
+
"loss": 0.8342287540435791,
|
| 812 |
+
"step": 15616
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.7330839222206826,
|
| 816 |
+
"grad_norm": 0.1290881633758545,
|
| 817 |
+
"learning_rate": 0.00017340258907913464,
|
| 818 |
+
"loss": 0.8337975740432739,
|
| 819 |
+
"step": 15872
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.7449078564500485,
|
| 823 |
+
"grad_norm": 0.16791236400604248,
|
| 824 |
+
"learning_rate": 0.0001592385021668743,
|
| 825 |
+
"loss": 0.8301421403884888,
|
| 826 |
+
"step": 16128
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.7567317906794143,
|
| 830 |
+
"grad_norm": 0.21357402205467224,
|
| 831 |
+
"learning_rate": 0.0001455675992000087,
|
| 832 |
+
"loss": 0.830890953540802,
|
| 833 |
+
"step": 16384
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.7567317906794143,
|
| 837 |
+
"eval_cos_loss": 0.27139199657799445,
|
| 838 |
+
"eval_loss": 0.811649158392867,
|
| 839 |
+
"eval_mse_loss": 0.811649158392867,
|
| 840 |
+
"flow/cos_sim": 0.7286080298902782,
|
| 841 |
+
"flow/improvement_ratio": 0.4760344069298,
|
| 842 |
+
"flow/mag_ratio_mean": 0.733996944204313,
|
| 843 |
+
"flow/mag_ratio_std": 0.20226374339975722,
|
| 844 |
+
"step": 16384
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.7567317906794143,
|
| 848 |
+
"eval_cos_loss": 0.27139199657799445,
|
| 849 |
+
"eval_loss": 0.811649158392867,
|
| 850 |
+
"eval_mse_loss": 0.811649158392867,
|
| 851 |
+
"eval_runtime": 37.4193,
|
| 852 |
+
"eval_samples_per_second": 748.089,
|
| 853 |
+
"eval_steps_per_second": 11.705,
|
| 854 |
+
"flow/cos_sim": 0.7286080298902782,
|
| 855 |
+
"flow/improvement_ratio": 0.4760344069298,
|
| 856 |
+
"flow/mag_ratio_mean": 0.733996944204313,
|
| 857 |
+
"flow/mag_ratio_std": 0.20226374339975722,
|
| 858 |
+
"step": 16384
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.7685557249087802,
|
| 862 |
+
"grad_norm": 0.14695969223976135,
|
| 863 |
+
"learning_rate": 0.000132409666069565,
|
| 864 |
+
"loss": 0.8289425373077393,
|
| 865 |
+
"step": 16640
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.780379659138146,
|
| 869 |
+
"grad_norm": 0.2218649983406067,
|
| 870 |
+
"learning_rate": 0.0001197837462455823,
|
| 871 |
+
"loss": 0.829971432685852,
|
| 872 |
+
"step": 16896
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.7922035933675119,
|
| 876 |
+
"grad_norm": 0.13823017477989197,
|
| 877 |
+
"learning_rate": 0.00010770811321550749,
|
| 878 |
+
"loss": 0.831263542175293,
|
| 879 |
+
"step": 17152
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.8040275275968778,
|
| 883 |
+
"grad_norm": 0.1739463061094284,
|
| 884 |
+
"learning_rate": 9.620024403698591e-05,
|
| 885 |
+
"loss": 0.8309124112129211,
|
| 886 |
+
"step": 17408
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.8040275275968778,
|
| 890 |
+
"eval_cos_loss": 0.2683021985164516,
|
| 891 |
+
"eval_loss": 0.8020191139554325,
|
| 892 |
+
"eval_mse_loss": 0.8020191139554325,
|
| 893 |
+
"flow/cos_sim": 0.731697819820822,
|
| 894 |
+
"flow/improvement_ratio": 0.4752517786200188,
|
| 895 |
+
"flow/mag_ratio_mean": 0.7306965604492518,
|
| 896 |
+
"flow/mag_ratio_std": 0.20605108883555076,
|
| 897 |
+
"step": 17408
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.8040275275968778,
|
| 901 |
+
"eval_cos_loss": 0.2683021985164516,
|
| 902 |
+
"eval_loss": 0.8020191139554325,
|
| 903 |
+
"eval_mse_loss": 0.8020191139554325,
|
| 904 |
+
"eval_runtime": 37.718,
|
| 905 |
+
"eval_samples_per_second": 742.166,
|
| 906 |
+
"eval_steps_per_second": 11.613,
|
| 907 |
+
"flow/cos_sim": 0.731697819820822,
|
| 908 |
+
"flow/improvement_ratio": 0.4752517786200188,
|
| 909 |
+
"flow/mag_ratio_mean": 0.7306965604492518,
|
| 910 |
+
"flow/mag_ratio_std": 0.20605108883555076,
|
| 911 |
+
"step": 17408
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.8158514618262436,
|
| 915 |
+
"grad_norm": 0.26471275091171265,
|
| 916 |
+
"learning_rate": 8.527679404332429e-05,
|
| 917 |
+
"loss": 0.831107497215271,
|
| 918 |
+
"step": 17664
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.8276753960556095,
|
| 922 |
+
"grad_norm": 0.11510410159826279,
|
| 923 |
+
"learning_rate": 7.495357273823544e-05,
|
| 924 |
+
"loss": 0.8281733393669128,
|
| 925 |
+
"step": 17920
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.8394993302849753,
|
| 929 |
+
"grad_norm": 0.29159778356552124,
|
| 930 |
+
"learning_rate": 6.524552091475183e-05,
|
| 931 |
+
"loss": 0.8288251161575317,
|
| 932 |
+
"step": 18176
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.8513232645143411,
|
| 936 |
+
"grad_norm": 0.17007242143154144,
|
| 937 |
+
"learning_rate": 5.6166689031422024e-05,
|
| 938 |
+
"loss": 0.8286168575286865,
|
| 939 |
+
"step": 18432
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.8513232645143411,
|
| 943 |
+
"eval_cos_loss": 0.2690970574475859,
|
| 944 |
+
"eval_loss": 0.8030253814780004,
|
| 945 |
+
"eval_mse_loss": 0.8030253814780004,
|
| 946 |
+
"flow/cos_sim": 0.7309029475194678,
|
| 947 |
+
"flow/improvement_ratio": 0.47304362157300184,
|
| 948 |
+
"flow/mag_ratio_mean": 0.7320575154807469,
|
| 949 |
+
"flow/mag_ratio_std": 0.20494019774269295,
|
| 950 |
+
"step": 18432
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8513232645143411,
|
| 954 |
+
"eval_cos_loss": 0.2690970574475859,
|
| 955 |
+
"eval_loss": 0.8030253814780004,
|
| 956 |
+
"eval_mse_loss": 0.8030253814780004,
|
| 957 |
+
"eval_runtime": 38.0297,
|
| 958 |
+
"eval_samples_per_second": 736.082,
|
| 959 |
+
"eval_steps_per_second": 11.517,
|
| 960 |
+
"flow/cos_sim": 0.7309029475194678,
|
| 961 |
+
"flow/improvement_ratio": 0.47304362157300184,
|
| 962 |
+
"flow/mag_ratio_mean": 0.7320575154807469,
|
| 963 |
+
"flow/mag_ratio_std": 0.20494019774269295,
|
| 964 |
+
"step": 18432
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.8631471987437069,
|
| 968 |
+
"grad_norm": 0.1886417269706726,
|
| 969 |
+
"learning_rate": 4.773021687709067e-05,
|
| 970 |
+
"loss": 0.8288365006446838,
|
| 971 |
+
"step": 18688
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.8749711329730728,
|
| 975 |
+
"grad_norm": 0.3300864100456238,
|
| 976 |
+
"learning_rate": 3.994831455368719e-05,
|
| 977 |
+
"loss": 0.8259992599487305,
|
| 978 |
+
"step": 18944
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.8867950672024387,
|
| 982 |
+
"grad_norm": 0.1487259566783905,
|
| 983 |
+
"learning_rate": 3.283224480455282e-05,
|
| 984 |
+
"loss": 0.8295122981071472,
|
| 985 |
+
"step": 19200
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.8986190014318045,
|
| 989 |
+
"grad_norm": 0.15876281261444092,
|
| 990 |
+
"learning_rate": 2.639230671387627e-05,
|
| 991 |
+
"loss": 0.8303685784339905,
|
| 992 |
+
"step": 19456
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.8986190014318045,
|
| 996 |
+
"eval_cos_loss": 0.26947175830466563,
|
| 997 |
+
"eval_loss": 0.8027528178746297,
|
| 998 |
+
"eval_mse_loss": 0.8027528178746297,
|
| 999 |
+
"flow/cos_sim": 0.7305282521737765,
|
| 1000 |
+
"flow/improvement_ratio": 0.47920923064288484,
|
| 1001 |
+
"flow/mag_ratio_mean": 0.7306106731499711,
|
| 1002 |
+
"flow/mag_ratio_std": 0.20609150707857793,
|
| 1003 |
+
"step": 19456
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8986190014318045,
|
| 1007 |
+
"eval_cos_loss": 0.26947175830466563,
|
| 1008 |
+
"eval_loss": 0.8027528178746297,
|
| 1009 |
+
"eval_mse_loss": 0.8027528178746297,
|
| 1010 |
+
"eval_runtime": 38.5004,
|
| 1011 |
+
"eval_samples_per_second": 727.083,
|
| 1012 |
+
"eval_steps_per_second": 11.376,
|
| 1013 |
+
"flow/cos_sim": 0.7305282521737765,
|
| 1014 |
+
"flow/improvement_ratio": 0.47920923064288484,
|
| 1015 |
+
"flow/mag_ratio_mean": 0.7306106731499711,
|
| 1016 |
+
"flow/mag_ratio_std": 0.20609150707857793,
|
| 1017 |
+
"step": 19456
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9104429356611704,
|
| 1021 |
+
"grad_norm": 0.12024150043725967,
|
| 1022 |
+
"learning_rate": 2.063782080083576e-05,
|
| 1023 |
+
"loss": 0.8279772996902466,
|
| 1024 |
+
"step": 19712
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9222668698905362,
|
| 1028 |
+
"grad_norm": 0.39260947704315186,
|
| 1029 |
+
"learning_rate": 1.557711553001523e-05,
|
| 1030 |
+
"loss": 0.8271908760070801,
|
| 1031 |
+
"step": 19968
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9340908041199021,
|
| 1035 |
+
"grad_norm": 0.13970452547073364,
|
| 1036 |
+
"learning_rate": 1.1217515257622269e-05,
|
| 1037 |
+
"loss": 0.8309769034385681,
|
| 1038 |
+
"step": 20224
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.945914738349268,
|
| 1042 |
+
"grad_norm": 0.11275673657655716,
|
| 1043 |
+
"learning_rate": 7.565329630950746e-06,
|
| 1044 |
+
"loss": 0.8321537971496582,
|
| 1045 |
+
"step": 20480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.945914738349268,
|
| 1049 |
+
"eval_cos_loss": 0.26695949672700064,
|
| 1050 |
+
"eval_loss": 0.797557680165931,
|
| 1051 |
+
"eval_mse_loss": 0.797557680165931,
|
| 1052 |
+
"flow/cos_sim": 0.7330405146019644,
|
| 1053 |
+
"flow/improvement_ratio": 0.4724381293880341,
|
| 1054 |
+
"flow/mag_ratio_mean": 0.7334565095705529,
|
| 1055 |
+
"flow/mag_ratio_std": 0.20568558155265573,
|
| 1056 |
+
"step": 20480
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 0.945914738349268,
|
| 1060 |
+
"eval_cos_loss": 0.26695949672700064,
|
| 1061 |
+
"eval_loss": 0.797557680165931,
|
| 1062 |
+
"eval_mse_loss": 0.797557680165931,
|
| 1063 |
+
"eval_runtime": 37.6307,
|
| 1064 |
+
"eval_samples_per_second": 743.886,
|
| 1065 |
+
"eval_steps_per_second": 11.639,
|
| 1066 |
+
"flow/cos_sim": 0.7330405146019644,
|
| 1067 |
+
"flow/improvement_ratio": 0.4724381293880341,
|
| 1068 |
+
"flow/mag_ratio_mean": 0.7334565095705529,
|
| 1069 |
+
"flow/mag_ratio_std": 0.20568558155265573,
|
| 1070 |
+
"step": 20480
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.9577386725786338,
|
| 1074 |
+
"grad_norm": 0.13636957108974457,
|
| 1075 |
+
"learning_rate": 4.62584445643166e-06,
|
| 1076 |
+
"loss": 0.8281410932540894,
|
| 1077 |
+
"step": 20736
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.9695626068079997,
|
| 1081 |
+
"grad_norm": 0.16119834780693054,
|
| 1082 |
+
"learning_rate": 2.40331404948807e-06,
|
| 1083 |
+
"loss": 0.8287432789802551,
|
| 1084 |
+
"step": 20992
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 0.9813865410373654,
|
| 1088 |
+
"grad_norm": 0.17925769090652466,
|
| 1089 |
+
"learning_rate": 9.009550772663965e-07,
|
| 1090 |
+
"loss": 0.8295612335205078,
|
| 1091 |
+
"step": 21248
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.9932104752667313,
|
| 1095 |
+
"grad_norm": 0.09990924596786499,
|
| 1096 |
+
"learning_rate": 1.2094190315575791e-07,
|
| 1097 |
+
"loss": 0.8297268152236938,
|
| 1098 |
+
"step": 21504
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.9932104752667313,
|
| 1102 |
+
"eval_cos_loss": 0.2675227834481627,
|
| 1103 |
+
"eval_loss": 0.7975949161945413,
|
| 1104 |
+
"eval_mse_loss": 0.7975949161945413,
|
| 1105 |
+
"flow/cos_sim": 0.7324772468984944,
|
| 1106 |
+
"flow/improvement_ratio": 0.4809445963330465,
|
| 1107 |
+
"flow/mag_ratio_mean": 0.7317501669877196,
|
| 1108 |
+
"flow/mag_ratio_std": 0.2055224990599776,
|
| 1109 |
+
"step": 21504
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9932104752667313,
|
| 1113 |
+
"eval_cos_loss": 0.2675227834481627,
|
| 1114 |
+
"eval_loss": 0.7975949161945413,
|
| 1115 |
+
"eval_mse_loss": 0.7975949161945413,
|
| 1116 |
+
"eval_runtime": 38.0323,
|
| 1117 |
+
"eval_samples_per_second": 736.032,
|
| 1118 |
+
"eval_steps_per_second": 11.517,
|
| 1119 |
+
"flow/cos_sim": 0.7324772468984944,
|
| 1120 |
+
"flow/improvement_ratio": 0.4809445963330465,
|
| 1121 |
+
"flow/mag_ratio_mean": 0.7317501669877196,
|
| 1122 |
+
"flow/mag_ratio_std": 0.2055224990599776,
|
| 1123 |
+
"step": 21504
|
| 1124 |
+
}
|
| 1125 |
+
],
|
| 1126 |
+
"logging_steps": 256,
|
| 1127 |
+
"max_steps": 21651,
|
| 1128 |
+
"num_input_tokens_seen": 0,
|
| 1129 |
+
"num_train_epochs": 1,
|
| 1130 |
+
"save_steps": 1024,
|
| 1131 |
+
"stateful_callbacks": {
|
| 1132 |
+
"TrainerControl": {
|
| 1133 |
+
"args": {
|
| 1134 |
+
"should_epoch_stop": false,
|
| 1135 |
+
"should_evaluate": false,
|
| 1136 |
+
"should_log": false,
|
| 1137 |
+
"should_save": true,
|
| 1138 |
+
"should_training_stop": false
|
| 1139 |
+
},
|
| 1140 |
+
"attributes": {}
|
| 1141 |
+
}
|
| 1142 |
+
},
|
| 1143 |
+
"total_flos": 0.0,
|
| 1144 |
+
"train_batch_size": 64,
|
| 1145 |
+
"trial_name": null,
|
| 1146 |
+
"trial_params": null
|
| 1147 |
+
}
|
checkpoints-v3.2/checkpoint-21504/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee9cb7ebeac20164f12c8a5bf77bbe01722a7871151494aba288a393bf7de081
|
| 3 |
+
size 5137
|