Upload folder using huggingface_hub
Browse files- checkpoints-v4.1/checkpoint-21504/ema.safetensors +3 -0
- checkpoints-v4.1/checkpoint-21504/model.safetensors +3 -0
- checkpoints-v4.1/checkpoint-21504/optimizer.pt +3 -0
- checkpoints-v4.1/checkpoint-21504/rng_state.pth +3 -0
- checkpoints-v4.1/checkpoint-21504/scaler.pt +3 -0
- checkpoints-v4.1/checkpoint-21504/scheduler.pt +3 -0
- checkpoints-v4.1/checkpoint-21504/trainer_state.json +1147 -0
- checkpoints-v4.1/checkpoint-21504/training_args.bin +3 -0
checkpoints-v4.1/checkpoint-21504/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c7ee848ff69b06c4d681aae2484d1218bd52e5e1d6c3a2952073348900ee9782
|
| 3 |
+
size 550088
|
checkpoints-v4.1/checkpoint-21504/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:afbf2c1faa4853b09a2fe59db51580d9ff37b2f1682b6689480b54357fcb994c
|
| 3 |
+
size 19318080
|
checkpoints-v4.1/checkpoint-21504/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fde5da904ac8fdc7fc5241824bc8d1a586e4886e78793103859cd4866fad9720
|
| 3 |
+
size 1176715
|
checkpoints-v4.1/checkpoint-21504/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e1f59ed61191b4d3edd45fb5218dffecfe6a68d9ef56e904294524f8f1b3375
|
| 3 |
+
size 14645
|
checkpoints-v4.1/checkpoint-21504/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ebcc3c6c1ff0265faec9ff4a23234251986a24b338d8a40737ba04b7a75531f4
|
| 3 |
+
size 1383
|
checkpoints-v4.1/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-v4.1/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": 0.3170062303543091,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 2.42356538772583,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.20766793191432953,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 2.0067272186279297,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.47662925720214844,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 1.8047175407409668,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.38411229848861694,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 1.7324724197387695,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.4818032477136072,
|
| 43 |
+
"eval_loss": 1.6560364654075066,
|
| 44 |
+
"eval_mse_loss": 1.4151348444424807,
|
| 45 |
+
"flow/cos_sim": 0.5181967507894725,
|
| 46 |
+
"flow/improvement_ratio": 0.954772694197964,
|
| 47 |
+
"flow/mag_ratio_mean": 0.48862707193039323,
|
| 48 |
+
"flow/mag_ratio_std": 0.18404143276416005,
|
| 49 |
+
"step": 1024
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.047295736917463395,
|
| 53 |
+
"eval_cos_loss": 0.4818032477136072,
|
| 54 |
+
"eval_loss": 1.6560364654075066,
|
| 55 |
+
"eval_mse_loss": 1.4151348444424807,
|
| 56 |
+
"eval_runtime": 41.6432,
|
| 57 |
+
"eval_samples_per_second": 672.211,
|
| 58 |
+
"eval_steps_per_second": 10.518,
|
| 59 |
+
"flow/cos_sim": 0.5181967507894725,
|
| 60 |
+
"flow/improvement_ratio": 0.954772694197964,
|
| 61 |
+
"flow/mag_ratio_mean": 0.48862707193039323,
|
| 62 |
+
"flow/mag_ratio_std": 0.18404143276416005,
|
| 63 |
+
"step": 1024
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.05911967114682925,
|
| 67 |
+
"grad_norm": 0.4044971168041229,
|
| 68 |
+
"learning_rate": 0.0009967551747861387,
|
| 69 |
+
"loss": 1.6944022178649902,
|
| 70 |
+
"step": 1280
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.0709436053761951,
|
| 74 |
+
"grad_norm": 0.23417921364307404,
|
| 75 |
+
"learning_rate": 0.000994232528651847,
|
| 76 |
+
"loss": 1.6682453155517578,
|
| 77 |
+
"step": 1536
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08276753960556095,
|
| 81 |
+
"grad_norm": 0.11558878421783447,
|
| 82 |
+
"learning_rate": 0.0009909945800260092,
|
| 83 |
+
"loss": 1.6490408182144165,
|
| 84 |
+
"step": 1792
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09459147383492679,
|
| 88 |
+
"grad_norm": 0.1645389050245285,
|
| 89 |
+
"learning_rate": 0.0009870460151900522,
|
| 90 |
+
"loss": 1.6396121978759766,
|
| 91 |
+
"step": 2048
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.09459147383492679,
|
| 95 |
+
"eval_cos_loss": 0.4538577257498214,
|
| 96 |
+
"eval_loss": 1.5759433782808312,
|
| 97 |
+
"eval_mse_loss": 1.3490145179234683,
|
| 98 |
+
"flow/cos_sim": 0.5461422990174054,
|
| 99 |
+
"flow/improvement_ratio": 0.9588973424206041,
|
| 100 |
+
"flow/mag_ratio_mean": 0.5239204951481188,
|
| 101 |
+
"flow/mag_ratio_std": 0.19641730230148524,
|
| 102 |
+
"step": 2048
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.09459147383492679,
|
| 106 |
+
"eval_cos_loss": 0.4538577257498214,
|
| 107 |
+
"eval_loss": 1.5759433782808312,
|
| 108 |
+
"eval_mse_loss": 1.3490145179234683,
|
| 109 |
+
"eval_runtime": 38.1827,
|
| 110 |
+
"eval_samples_per_second": 733.134,
|
| 111 |
+
"eval_steps_per_second": 11.471,
|
| 112 |
+
"flow/cos_sim": 0.5461422990174054,
|
| 113 |
+
"flow/improvement_ratio": 0.9588973424206041,
|
| 114 |
+
"flow/mag_ratio_mean": 0.5239204951481188,
|
| 115 |
+
"flow/mag_ratio_std": 0.19641730230148524,
|
| 116 |
+
"step": 2048
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.10641540806429264,
|
| 120 |
+
"grad_norm": 0.1483554244041443,
|
| 121 |
+
"learning_rate": 0.0009823925488998885,
|
| 122 |
+
"loss": 1.629364252090454,
|
| 123 |
+
"step": 2304
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1182393422936585,
|
| 127 |
+
"grad_norm": 0.1272267997264862,
|
| 128 |
+
"learning_rate": 0.0009770409161149525,
|
| 129 |
+
"loss": 1.6207596063613892,
|
| 130 |
+
"step": 2560
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.13006327652302435,
|
| 134 |
+
"grad_norm": 0.4405772387981415,
|
| 135 |
+
"learning_rate": 0.0009709988622506973,
|
| 136 |
+
"loss": 1.6130421161651611,
|
| 137 |
+
"step": 2816
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1418872107523902,
|
| 141 |
+
"grad_norm": 0.2765417993068695,
|
| 142 |
+
"learning_rate": 0.000964275131968659,
|
| 143 |
+
"loss": 1.6063523292541504,
|
| 144 |
+
"step": 3072
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.1418872107523902,
|
| 148 |
+
"eval_cos_loss": 0.4437784091930955,
|
| 149 |
+
"eval_loss": 1.5468863191669935,
|
| 150 |
+
"eval_mse_loss": 1.3249971137199228,
|
| 151 |
+
"flow/cos_sim": 0.5562216001286354,
|
| 152 |
+
"flow/improvement_ratio": 0.9605493699306766,
|
| 153 |
+
"flow/mag_ratio_mean": 0.531293115863517,
|
| 154 |
+
"flow/mag_ratio_std": 0.20344179882306487,
|
| 155 |
+
"step": 3072
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1418872107523902,
|
| 159 |
+
"eval_cos_loss": 0.4437784091930955,
|
| 160 |
+
"eval_loss": 1.5468863191669935,
|
| 161 |
+
"eval_mse_loss": 1.3249971137199228,
|
| 162 |
+
"eval_runtime": 38.9587,
|
| 163 |
+
"eval_samples_per_second": 718.531,
|
| 164 |
+
"eval_steps_per_second": 11.243,
|
| 165 |
+
"flow/cos_sim": 0.5562216001286354,
|
| 166 |
+
"flow/improvement_ratio": 0.9605493699306766,
|
| 167 |
+
"flow/mag_ratio_mean": 0.531293115863517,
|
| 168 |
+
"flow/mag_ratio_std": 0.20344179882306487,
|
| 169 |
+
"step": 3072
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.15371114498175603,
|
| 173 |
+
"grad_norm": 0.14539985358715057,
|
| 174 |
+
"learning_rate": 0.0009568794565203123,
|
| 175 |
+
"loss": 1.6010686159133911,
|
| 176 |
+
"step": 3328
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.1655350792111219,
|
| 180 |
+
"grad_norm": 0.146505206823349,
|
| 181 |
+
"learning_rate": 0.0009488225396630347,
|
| 182 |
+
"loss": 1.593965768814087,
|
| 183 |
+
"step": 3584
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17735901344048774,
|
| 187 |
+
"grad_norm": 0.17704616487026215,
|
| 188 |
+
"learning_rate": 0.0009401160421685646,
|
| 189 |
+
"loss": 1.591158390045166,
|
| 190 |
+
"step": 3840
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18918294766985358,
|
| 194 |
+
"grad_norm": 0.21900475025177002,
|
| 195 |
+
"learning_rate": 0.0009307725649463714,
|
| 196 |
+
"loss": 1.5863914489746094,
|
| 197 |
+
"step": 4096
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.18918294766985358,
|
| 201 |
+
"eval_cos_loss": 0.4372148915240754,
|
| 202 |
+
"eval_loss": 1.5282805349728832,
|
| 203 |
+
"eval_mse_loss": 1.3096730913201424,
|
| 204 |
+
"flow/cos_sim": 0.5627851008552395,
|
| 205 |
+
"flow/improvement_ratio": 0.960628461483951,
|
| 206 |
+
"flow/mag_ratio_mean": 0.5383912115336553,
|
| 207 |
+
"flow/mag_ratio_std": 0.20601355281050346,
|
| 208 |
+
"step": 4096
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"eval_cos_loss": 0.4372148915240754,
|
| 213 |
+
"eval_loss": 1.5282805349728832,
|
| 214 |
+
"eval_mse_loss": 1.3096730913201424,
|
| 215 |
+
"eval_runtime": 37.5818,
|
| 216 |
+
"eval_samples_per_second": 744.854,
|
| 217 |
+
"eval_steps_per_second": 11.655,
|
| 218 |
+
"flow/cos_sim": 0.5627851008552395,
|
| 219 |
+
"flow/improvement_ratio": 0.960628461483951,
|
| 220 |
+
"flow/mag_ratio_mean": 0.5383912115336553,
|
| 221 |
+
"flow/mag_ratio_std": 0.20601355281050346,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.20100688189921945,
|
| 226 |
+
"grad_norm": 0.3215397000312805,
|
| 227 |
+
"learning_rate": 0.0009208056308063659,
|
| 228 |
+
"loss": 1.5831135511398315,
|
| 229 |
+
"step": 4352
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.2128308161285853,
|
| 233 |
+
"grad_norm": 0.39093902707099915,
|
| 234 |
+
"learning_rate": 0.0009102296648873445,
|
| 235 |
+
"loss": 1.5793178081512451,
|
| 236 |
+
"step": 4608
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.22465475035795113,
|
| 240 |
+
"grad_norm": 0.24414080381393433,
|
| 241 |
+
"learning_rate": 0.0008990599737794927,
|
| 242 |
+
"loss": 1.5755181312561035,
|
| 243 |
+
"step": 4864
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.236478684587317,
|
| 247 |
+
"grad_norm": 0.18973594903945923,
|
| 248 |
+
"learning_rate": 0.0008873127233711644,
|
| 249 |
+
"loss": 1.5741870403289795,
|
| 250 |
+
"step": 5120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.236478684587317,
|
| 254 |
+
"eval_cos_loss": 0.43135033779220494,
|
| 255 |
+
"eval_loss": 1.5102224085973277,
|
| 256 |
+
"eval_mse_loss": 1.294547238578535,
|
| 257 |
+
"flow/cos_sim": 0.568649673570781,
|
| 258 |
+
"flow/improvement_ratio": 0.9623238818558384,
|
| 259 |
+
"flow/mag_ratio_mean": 0.5475505786399318,
|
| 260 |
+
"flow/mag_ratio_std": 0.2083817292089876,
|
| 261 |
+
"step": 5120
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.236478684587317,
|
| 265 |
+
"eval_cos_loss": 0.43135033779220494,
|
| 266 |
+
"eval_loss": 1.5102224085973277,
|
| 267 |
+
"eval_mse_loss": 1.294547238578535,
|
| 268 |
+
"eval_runtime": 38.3814,
|
| 269 |
+
"eval_samples_per_second": 729.337,
|
| 270 |
+
"eval_steps_per_second": 11.412,
|
| 271 |
+
"flow/cos_sim": 0.568649673570781,
|
| 272 |
+
"flow/improvement_ratio": 0.9623238818558384,
|
| 273 |
+
"flow/mag_ratio_mean": 0.5475505786399318,
|
| 274 |
+
"flow/mag_ratio_std": 0.2083817292089876,
|
| 275 |
+
"step": 5120
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24830261881668284,
|
| 279 |
+
"grad_norm": 0.23730266094207764,
|
| 280 |
+
"learning_rate": 0.0008750049154520011,
|
| 281 |
+
"loss": 1.5687248706817627,
|
| 282 |
+
"step": 5376
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2601265530460487,
|
| 286 |
+
"grad_norm": 0.23536692559719086,
|
| 287 |
+
"learning_rate": 0.0008621543631062487,
|
| 288 |
+
"loss": 1.5671372413635254,
|
| 289 |
+
"step": 5632
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.27195048727541454,
|
| 293 |
+
"grad_norm": 0.15165014564990997,
|
| 294 |
+
"learning_rate": 0.0008487796649318904,
|
| 295 |
+
"loss": 1.5671967267990112,
|
| 296 |
+
"step": 5888
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2837744215047804,
|
| 300 |
+
"grad_norm": 0.1857716292142868,
|
| 301 |
+
"learning_rate": 0.0008349001781229053,
|
| 302 |
+
"loss": 1.5656864643096924,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_cos_loss": 0.4303430036458795,
|
| 308 |
+
"eval_loss": 1.507517477148744,
|
| 309 |
+
"eval_mse_loss": 1.292345978353666,
|
| 310 |
+
"flow/cos_sim": 0.5696570243193134,
|
| 311 |
+
"flow/improvement_ratio": 0.9619137474663182,
|
| 312 |
+
"flow/mag_ratio_mean": 0.5517072507507725,
|
| 313 |
+
"flow/mag_ratio_std": 0.2047896740711443,
|
| 314 |
+
"step": 6144
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2837744215047804,
|
| 318 |
+
"eval_cos_loss": 0.4303430036458795,
|
| 319 |
+
"eval_loss": 1.507517477148744,
|
| 320 |
+
"eval_mse_loss": 1.292345978353666,
|
| 321 |
+
"eval_runtime": 39.1056,
|
| 322 |
+
"eval_samples_per_second": 715.83,
|
| 323 |
+
"eval_steps_per_second": 11.2,
|
| 324 |
+
"flow/cos_sim": 0.5696570243193134,
|
| 325 |
+
"flow/improvement_ratio": 0.9619137474663182,
|
| 326 |
+
"flow/mag_ratio_mean": 0.5517072507507725,
|
| 327 |
+
"flow/mag_ratio_std": 0.2047896740711443,
|
| 328 |
+
"step": 6144
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2955983557341462,
|
| 332 |
+
"grad_norm": 0.16939695179462433,
|
| 333 |
+
"learning_rate": 0.0008205359904536107,
|
| 334 |
+
"loss": 1.5614818334579468,
|
| 335 |
+
"step": 6400
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.30742228996351206,
|
| 339 |
+
"grad_norm": 0.20279403030872345,
|
| 340 |
+
"learning_rate": 0.0008057078912056363,
|
| 341 |
+
"loss": 1.5598734617233276,
|
| 342 |
+
"step": 6656
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3192462241928779,
|
| 346 |
+
"grad_norm": 0.17468483746051788,
|
| 347 |
+
"learning_rate": 0.0007904373410796086,
|
| 348 |
+
"loss": 1.5608937740325928,
|
| 349 |
+
"step": 6912
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.3310701584222438,
|
| 353 |
+
"grad_norm": 0.13540241122245789,
|
| 354 |
+
"learning_rate": 0.0007747464411350876,
|
| 355 |
+
"loss": 1.5515446662902832,
|
| 356 |
+
"step": 7168
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.3310701584222438,
|
| 360 |
+
"eval_cos_loss": 0.42760413388411206,
|
| 361 |
+
"eval_loss": 1.4991425577908346,
|
| 362 |
+
"eval_mse_loss": 1.285340489862172,
|
| 363 |
+
"flow/cos_sim": 0.5723958823778857,
|
| 364 |
+
"flow/improvement_ratio": 0.9629914771476292,
|
| 365 |
+
"flow/mag_ratio_mean": 0.551068566841622,
|
| 366 |
+
"flow/mag_ratio_std": 0.21019490124428108,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3310701584222438,
|
| 371 |
+
"eval_cos_loss": 0.42760413388411206,
|
| 372 |
+
"eval_loss": 1.4991425577908346,
|
| 373 |
+
"eval_mse_loss": 1.285340489862172,
|
| 374 |
+
"eval_runtime": 37.5914,
|
| 375 |
+
"eval_samples_per_second": 744.665,
|
| 376 |
+
"eval_steps_per_second": 11.652,
|
| 377 |
+
"flow/cos_sim": 0.5723958823778857,
|
| 378 |
+
"flow/improvement_ratio": 0.9629914771476292,
|
| 379 |
+
"flow/mag_ratio_mean": 0.551068566841622,
|
| 380 |
+
"flow/mag_ratio_std": 0.21019490124428108,
|
| 381 |
+
"step": 7168
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.34289409265160964,
|
| 385 |
+
"grad_norm": 0.16203606128692627,
|
| 386 |
+
"learning_rate": 0.000758657900803716,
|
| 387 |
+
"loss": 1.5541112422943115,
|
| 388 |
+
"step": 7424
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3547180268809755,
|
| 392 |
+
"grad_norm": 0.15600702166557312,
|
| 393 |
+
"learning_rate": 0.000742195005021869,
|
| 394 |
+
"loss": 1.552086353302002,
|
| 395 |
+
"step": 7680
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3665419611103413,
|
| 399 |
+
"grad_norm": 0.25274714827537537,
|
| 400 |
+
"learning_rate": 0.0007253815805303786,
|
| 401 |
+
"loss": 1.5514739751815796,
|
| 402 |
+
"step": 7936
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.37836589533970716,
|
| 406 |
+
"grad_norm": 0.16391786932945251,
|
| 407 |
+
"learning_rate": 0.0007082419613901028,
|
| 408 |
+
"loss": 1.5471858978271484,
|
| 409 |
+
"step": 8192
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.37836589533970716,
|
| 413 |
+
"eval_cos_loss": 0.42565103276679506,
|
| 414 |
+
"eval_loss": 1.4905888979837774,
|
| 415 |
+
"eval_mse_loss": 1.277763385206597,
|
| 416 |
+
"flow/cos_sim": 0.5743489830189099,
|
| 417 |
+
"flow/improvement_ratio": 0.9618241816895193,
|
| 418 |
+
"flow/mag_ratio_mean": 0.5533584926498535,
|
| 419 |
+
"flow/mag_ratio_std": 0.21572684449011878,
|
| 420 |
+
"step": 8192
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.37836589533970716,
|
| 424 |
+
"eval_cos_loss": 0.42565103276679506,
|
| 425 |
+
"eval_loss": 1.4905888979837774,
|
| 426 |
+
"eval_mse_loss": 1.277763385206597,
|
| 427 |
+
"eval_runtime": 37.3487,
|
| 428 |
+
"eval_samples_per_second": 749.503,
|
| 429 |
+
"eval_steps_per_second": 11.727,
|
| 430 |
+
"flow/cos_sim": 0.5743489830189099,
|
| 431 |
+
"flow/improvement_ratio": 0.9618241816895193,
|
| 432 |
+
"flow/mag_ratio_mean": 0.5533584926498535,
|
| 433 |
+
"flow/mag_ratio_std": 0.21572684449011878,
|
| 434 |
+
"step": 8192
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.390189829569073,
|
| 438 |
+
"grad_norm": 0.13766784965991974,
|
| 439 |
+
"learning_rate": 0.0006908009537632514,
|
| 440 |
+
"loss": 1.5495105981826782,
|
| 441 |
+
"step": 8448
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4020137637984389,
|
| 445 |
+
"grad_norm": 0.15862250328063965,
|
| 446 |
+
"learning_rate": 0.0006730838000114403,
|
| 447 |
+
"loss": 1.5457631349563599,
|
| 448 |
+
"step": 8704
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.41383769802780473,
|
| 452 |
+
"grad_norm": 0.1839229315519333,
|
| 453 |
+
"learning_rate": 0.0006551161421624341,
|
| 454 |
+
"loss": 1.5441044569015503,
|
| 455 |
+
"step": 8960
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.4256616322571706,
|
| 459 |
+
"grad_norm": 0.13566677272319794,
|
| 460 |
+
"learning_rate": 0.0006369239847984517,
|
| 461 |
+
"loss": 1.5432497262954712,
|
| 462 |
+
"step": 9216
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4256616322571706,
|
| 466 |
+
"eval_cos_loss": 0.42430811161047793,
|
| 467 |
+
"eval_loss": 1.4869584389473205,
|
| 468 |
+
"eval_mse_loss": 1.2748043828358933,
|
| 469 |
+
"flow/cos_sim": 0.5756919168990496,
|
| 470 |
+
"flow/improvement_ratio": 0.96208056807518,
|
| 471 |
+
"flow/mag_ratio_mean": 0.5574794498752785,
|
| 472 |
+
"flow/mag_ratio_std": 0.21047259782139024,
|
| 473 |
+
"step": 9216
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4256616322571706,
|
| 477 |
+
"eval_cos_loss": 0.42430811161047793,
|
| 478 |
+
"eval_loss": 1.4869584389473205,
|
| 479 |
+
"eval_mse_loss": 1.2748043828358933,
|
| 480 |
+
"eval_runtime": 37.5205,
|
| 481 |
+
"eval_samples_per_second": 746.072,
|
| 482 |
+
"eval_steps_per_second": 11.674,
|
| 483 |
+
"flow/cos_sim": 0.5756919168990496,
|
| 484 |
+
"flow/improvement_ratio": 0.96208056807518,
|
| 485 |
+
"flow/mag_ratio_mean": 0.5574794498752785,
|
| 486 |
+
"flow/mag_ratio_std": 0.21047259782139024,
|
| 487 |
+
"step": 9216
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.4374855664865364,
|
| 491 |
+
"grad_norm": 0.13037081062793732,
|
| 492 |
+
"learning_rate": 0.0006185336574197479,
|
| 493 |
+
"loss": 1.544568419456482,
|
| 494 |
+
"step": 9472
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.44930950071590225,
|
| 498 |
+
"grad_norm": 0.2653915286064148,
|
| 499 |
+
"learning_rate": 0.0005999717763379407,
|
| 500 |
+
"loss": 1.5420877933502197,
|
| 501 |
+
"step": 9728
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4611334349452681,
|
| 505 |
+
"grad_norm": 0.12325401604175568,
|
| 506 |
+
"learning_rate": 0.0005812652061542363,
|
| 507 |
+
"loss": 1.5435826778411865,
|
| 508 |
+
"step": 9984
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.472957369174634,
|
| 512 |
+
"grad_norm": 0.2154727578163147,
|
| 513 |
+
"learning_rate": 0.0005624410208783071,
|
| 514 |
+
"loss": 1.5385282039642334,
|
| 515 |
+
"step": 10240
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.472957369174634,
|
| 519 |
+
"eval_cos_loss": 0.42216237005033447,
|
| 520 |
+
"eval_loss": 1.4820257316441297,
|
| 521 |
+
"eval_mse_loss": 1.2709445436251219,
|
| 522 |
+
"flow/cos_sim": 0.577837641925028,
|
| 523 |
+
"flow/improvement_ratio": 0.9637134710675506,
|
| 524 |
+
"flow/mag_ratio_mean": 0.5554149469284162,
|
| 525 |
+
"flow/mag_ratio_std": 0.2115786727432791,
|
| 526 |
+
"step": 10240
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.472957369174634,
|
| 530 |
+
"eval_cos_loss": 0.42216237005033447,
|
| 531 |
+
"eval_loss": 1.4820257316441297,
|
| 532 |
+
"eval_mse_loss": 1.2709445436251219,
|
| 533 |
+
"eval_runtime": 37.3144,
|
| 534 |
+
"eval_samples_per_second": 750.194,
|
| 535 |
+
"eval_steps_per_second": 11.738,
|
| 536 |
+
"flow/cos_sim": 0.577837641925028,
|
| 537 |
+
"flow/improvement_ratio": 0.9637134710675506,
|
| 538 |
+
"flow/mag_ratio_mean": 0.5554149469284162,
|
| 539 |
+
"flow/mag_ratio_std": 0.2115786727432791,
|
| 540 |
+
"step": 10240
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.48478130340399983,
|
| 544 |
+
"grad_norm": 0.167108952999115,
|
| 545 |
+
"learning_rate": 0.0005435264647440881,
|
| 546 |
+
"loss": 1.5393356084823608,
|
| 547 |
+
"step": 10496
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.49660523763336567,
|
| 551 |
+
"grad_norm": 0.11320369690656662,
|
| 552 |
+
"learning_rate": 0.000524548912779213,
|
| 553 |
+
"loss": 1.5383528470993042,
|
| 554 |
+
"step": 10752
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.5084291718627315,
|
| 558 |
+
"grad_norm": 0.279090940952301,
|
| 559 |
+
"learning_rate": 0.0005055358311851499,
|
| 560 |
+
"loss": 1.5419063568115234,
|
| 561 |
+
"step": 11008
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5202531060920974,
|
| 565 |
+
"grad_norm": 0.1590709239244461,
|
| 566 |
+
"learning_rate": 0.0004865147375853812,
|
| 567 |
+
"loss": 1.5351711511611938,
|
| 568 |
+
"step": 11264
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.5202531060920974,
|
| 572 |
+
"eval_cos_loss": 0.42213277620811984,
|
| 573 |
+
"eval_loss": 1.4814584206228387,
|
| 574 |
+
"eval_mse_loss": 1.2703920295249382,
|
| 575 |
+
"flow/cos_sim": 0.577867228690892,
|
| 576 |
+
"flow/improvement_ratio": 0.962785639174997,
|
| 577 |
+
"flow/mag_ratio_mean": 0.5578808011529652,
|
| 578 |
+
"flow/mag_ratio_std": 0.21099132279010668,
|
| 579 |
+
"step": 11264
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.5202531060920974,
|
| 583 |
+
"eval_cos_loss": 0.42213277620811984,
|
| 584 |
+
"eval_loss": 1.4814584206228387,
|
| 585 |
+
"eval_mse_loss": 1.2703920295249382,
|
| 586 |
+
"eval_runtime": 38.5165,
|
| 587 |
+
"eval_samples_per_second": 726.78,
|
| 588 |
+
"eval_steps_per_second": 11.372,
|
| 589 |
+
"flow/cos_sim": 0.577867228690892,
|
| 590 |
+
"flow/improvement_ratio": 0.962785639174997,
|
| 591 |
+
"flow/mag_ratio_mean": 0.5578808011529652,
|
| 592 |
+
"flow/mag_ratio_std": 0.21099132279010668,
|
| 593 |
+
"step": 11264
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5320770403214632,
|
| 597 |
+
"grad_norm": 0.14667300879955292,
|
| 598 |
+
"learning_rate": 0.0004675131611991607,
|
| 599 |
+
"loss": 1.5333118438720703,
|
| 600 |
+
"step": 11520
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.5439009745508291,
|
| 604 |
+
"grad_norm": 0.12886938452720642,
|
| 605 |
+
"learning_rate": 0.0004485586029984899,
|
| 606 |
+
"loss": 1.5335720777511597,
|
| 607 |
+
"step": 11776
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.5557249087801949,
|
| 611 |
+
"grad_norm": 0.12272178381681442,
|
| 612 |
+
"learning_rate": 0.00042967849590597266,
|
| 613 |
+
"loss": 1.536974310874939,
|
| 614 |
+
"step": 12032
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5675488430095608,
|
| 618 |
+
"grad_norm": 0.1320953667163849,
|
| 619 |
+
"learning_rate": 0.0004109001650911621,
|
| 620 |
+
"loss": 1.5350415706634521,
|
| 621 |
+
"step": 12288
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5675488430095608,
|
| 625 |
+
"eval_cos_loss": 0.4204618357224007,
|
| 626 |
+
"eval_loss": 1.4744470647480934,
|
| 627 |
+
"eval_mse_loss": 1.2642161459683283,
|
| 628 |
+
"flow/cos_sim": 0.5795381778179238,
|
| 629 |
+
"flow/improvement_ratio": 0.9622249373290093,
|
| 630 |
+
"flow/mag_ratio_mean": 0.5585955846799563,
|
| 631 |
+
"flow/mag_ratio_std": 0.21563580887366648,
|
| 632 |
+
"step": 12288
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5675488430095608,
|
| 636 |
+
"eval_cos_loss": 0.4204618357224007,
|
| 637 |
+
"eval_loss": 1.4744470647480934,
|
| 638 |
+
"eval_mse_loss": 1.2642161459683283,
|
| 639 |
+
"eval_runtime": 39.1238,
|
| 640 |
+
"eval_samples_per_second": 715.498,
|
| 641 |
+
"eval_steps_per_second": 11.195,
|
| 642 |
+
"flow/cos_sim": 0.5795381778179238,
|
| 643 |
+
"flow/improvement_ratio": 0.9622249373290093,
|
| 644 |
+
"flow/mag_ratio_mean": 0.5585955846799563,
|
| 645 |
+
"flow/mag_ratio_std": 0.21563580887366648,
|
| 646 |
+
"step": 12288
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5793727772389267,
|
| 650 |
+
"grad_norm": 0.16382183134555817,
|
| 651 |
+
"learning_rate": 0.0003922507884228551,
|
| 652 |
+
"loss": 1.5361268520355225,
|
| 653 |
+
"step": 12544
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5911967114682924,
|
| 657 |
+
"grad_norm": 0.13758233189582825,
|
| 658 |
+
"learning_rate": 0.00037375735713457723,
|
| 659 |
+
"loss": 1.530663013458252,
|
| 660 |
+
"step": 12800
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6030206456976583,
|
| 664 |
+
"grad_norm": 0.13651148974895477,
|
| 665 |
+
"learning_rate": 0.00035544663676018276,
|
| 666 |
+
"loss": 1.528375506401062,
|
| 667 |
+
"step": 13056
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6148445799270241,
|
| 671 |
+
"grad_norm": 0.19377295672893524,
|
| 672 |
+
"learning_rate": 0.00033734512839611255,
|
| 673 |
+
"loss": 1.531029462814331,
|
| 674 |
+
"step": 13312
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6148445799270241,
|
| 678 |
+
"eval_cos_loss": 0.4196748673643696,
|
| 679 |
+
"eval_loss": 1.4739454851302927,
|
| 680 |
+
"eval_mse_loss": 1.2641080489986019,
|
| 681 |
+
"flow/cos_sim": 0.5803251580832756,
|
| 682 |
+
"flow/improvement_ratio": 0.9619928854241219,
|
| 683 |
+
"flow/mag_ratio_mean": 0.5591708176756558,
|
| 684 |
+
"flow/mag_ratio_std": 0.2099404629302896,
|
| 685 |
+
"step": 13312
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.6148445799270241,
|
| 689 |
+
"eval_cos_loss": 0.4196748673643696,
|
| 690 |
+
"eval_loss": 1.4739454851302927,
|
| 691 |
+
"eval_mse_loss": 1.2641080489986019,
|
| 692 |
+
"eval_runtime": 37.6325,
|
| 693 |
+
"eval_samples_per_second": 743.851,
|
| 694 |
+
"eval_steps_per_second": 11.639,
|
| 695 |
+
"flow/cos_sim": 0.5803251580832756,
|
| 696 |
+
"flow/improvement_ratio": 0.9619928854241219,
|
| 697 |
+
"flow/mag_ratio_mean": 0.5591708176756558,
|
| 698 |
+
"flow/mag_ratio_std": 0.2099404629302896,
|
| 699 |
+
"step": 13312
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.62666851415639,
|
| 703 |
+
"grad_norm": 0.11970210075378418,
|
| 704 |
+
"learning_rate": 0.0003194790303463687,
|
| 705 |
+
"loss": 1.526605248451233,
|
| 706 |
+
"step": 13568
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.6384924483857558,
|
| 710 |
+
"grad_norm": 0.16870331764221191,
|
| 711 |
+
"learning_rate": 0.00030187420020572406,
|
| 712 |
+
"loss": 1.5294699668884277,
|
| 713 |
+
"step": 13824
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.6503163826151217,
|
| 717 |
+
"grad_norm": 0.20158159732818604,
|
| 718 |
+
"learning_rate": 0.00028455611743603626,
|
| 719 |
+
"loss": 1.5296140909194946,
|
| 720 |
+
"step": 14080
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.6621403168444876,
|
| 724 |
+
"grad_norm": 0.17316782474517822,
|
| 725 |
+
"learning_rate": 0.0002675498464898373,
|
| 726 |
+
"loss": 1.5282411575317383,
|
| 727 |
+
"step": 14336
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.6621403168444876,
|
| 731 |
+
"eval_cos_loss": 0.4197337330748501,
|
| 732 |
+
"eval_loss": 1.4732592075927073,
|
| 733 |
+
"eval_mse_loss": 1.263392340374864,
|
| 734 |
+
"flow/cos_sim": 0.5802662803974326,
|
| 735 |
+
"flow/improvement_ratio": 0.9630070868964609,
|
| 736 |
+
"flow/mag_ratio_mean": 0.5561743245277231,
|
| 737 |
+
"flow/mag_ratio_std": 0.21411516738536696,
|
| 738 |
+
"step": 14336
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6621403168444876,
|
| 742 |
+
"eval_cos_loss": 0.4197337330748501,
|
| 743 |
+
"eval_loss": 1.4732592075927073,
|
| 744 |
+
"eval_mse_loss": 1.263392340374864,
|
| 745 |
+
"eval_runtime": 38.6362,
|
| 746 |
+
"eval_samples_per_second": 724.527,
|
| 747 |
+
"eval_steps_per_second": 11.337,
|
| 748 |
+
"flow/cos_sim": 0.5802662803974326,
|
| 749 |
+
"flow/improvement_ratio": 0.9630070868964609,
|
| 750 |
+
"flow/mag_ratio_mean": 0.5561743245277231,
|
| 751 |
+
"flow/mag_ratio_std": 0.21411516738536696,
|
| 752 |
+
"step": 14336
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6739642510738534,
|
| 756 |
+
"grad_norm": 0.1980622410774231,
|
| 757 |
+
"learning_rate": 0.0002508800005345623,
|
| 758 |
+
"loss": 1.527830719947815,
|
| 759 |
+
"step": 14592
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6857881853032193,
|
| 763 |
+
"grad_norm": 0.13991132378578186,
|
| 764 |
+
"learning_rate": 0.00023457070582992562,
|
| 765 |
+
"loss": 1.5318052768707275,
|
| 766 |
+
"step": 14848
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6976121195325851,
|
| 770 |
+
"grad_norm": 0.149434432387352,
|
| 771 |
+
"learning_rate": 0.00021864556680999692,
|
| 772 |
+
"loss": 1.5274654626846313,
|
| 773 |
+
"step": 15104
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.709436053761951,
|
| 777 |
+
"grad_norm": 0.13668516278266907,
|
| 778 |
+
"learning_rate": 0.0002031276319205152,
|
| 779 |
+
"loss": 1.5278632640838623,
|
| 780 |
+
"step": 15360
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.709436053761951,
|
| 784 |
+
"eval_cos_loss": 0.41798419257005054,
|
| 785 |
+
"eval_loss": 1.4683536525186338,
|
| 786 |
+
"eval_mse_loss": 1.259361555315044,
|
| 787 |
+
"flow/cos_sim": 0.5820158255971186,
|
| 788 |
+
"flow/improvement_ratio": 0.9633506106459386,
|
| 789 |
+
"flow/mag_ratio_mean": 0.5602124498042886,
|
| 790 |
+
"flow/mag_ratio_std": 0.21119761075739446,
|
| 791 |
+
"step": 15360
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.709436053761951,
|
| 795 |
+
"eval_cos_loss": 0.41798419257005054,
|
| 796 |
+
"eval_loss": 1.4683536525186338,
|
| 797 |
+
"eval_mse_loss": 1.259361555315044,
|
| 798 |
+
"eval_runtime": 37.7713,
|
| 799 |
+
"eval_samples_per_second": 741.119,
|
| 800 |
+
"eval_steps_per_second": 11.596,
|
| 801 |
+
"flow/cos_sim": 0.5820158255971186,
|
| 802 |
+
"flow/improvement_ratio": 0.9633506106459386,
|
| 803 |
+
"flow/mag_ratio_mean": 0.5602124498042886,
|
| 804 |
+
"flow/mag_ratio_std": 0.21119761075739446,
|
| 805 |
+
"step": 15360
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.7212599879913169,
|
| 809 |
+
"grad_norm": 0.10699856281280518,
|
| 810 |
+
"learning_rate": 0.00018803936026088542,
|
| 811 |
+
"loss": 1.5269209146499634,
|
| 812 |
+
"step": 15616
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.7330839222206826,
|
| 816 |
+
"grad_norm": 0.14415434002876282,
|
| 817 |
+
"learning_rate": 0.00017340258907913464,
|
| 818 |
+
"loss": 1.5266590118408203,
|
| 819 |
+
"step": 15872
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.7449078564500485,
|
| 823 |
+
"grad_norm": 0.10905779153108597,
|
| 824 |
+
"learning_rate": 0.0001592385021668743,
|
| 825 |
+
"loss": 1.524775505065918,
|
| 826 |
+
"step": 16128
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.7567317906794143,
|
| 830 |
+
"grad_norm": 0.19332484900951385,
|
| 831 |
+
"learning_rate": 0.0001455675992000087,
|
| 832 |
+
"loss": 1.528196096420288,
|
| 833 |
+
"step": 16384
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.7567317906794143,
|
| 837 |
+
"eval_cos_loss": 0.4177281832041806,
|
| 838 |
+
"eval_loss": 1.4672707036205623,
|
| 839 |
+
"eval_mse_loss": 1.258406613243225,
|
| 840 |
+
"flow/cos_sim": 0.5822718332619428,
|
| 841 |
+
"flow/improvement_ratio": 0.9624373912811279,
|
| 842 |
+
"flow/mag_ratio_mean": 0.5578821809052332,
|
| 843 |
+
"flow/mag_ratio_std": 0.21358931578321544,
|
| 844 |
+
"step": 16384
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.7567317906794143,
|
| 848 |
+
"eval_cos_loss": 0.4177281832041806,
|
| 849 |
+
"eval_loss": 1.4672707036205623,
|
| 850 |
+
"eval_mse_loss": 1.258406613243225,
|
| 851 |
+
"eval_runtime": 38.8935,
|
| 852 |
+
"eval_samples_per_second": 719.735,
|
| 853 |
+
"eval_steps_per_second": 11.262,
|
| 854 |
+
"flow/cos_sim": 0.5822718332619428,
|
| 855 |
+
"flow/improvement_ratio": 0.9624373912811279,
|
| 856 |
+
"flow/mag_ratio_mean": 0.5578821809052332,
|
| 857 |
+
"flow/mag_ratio_std": 0.21358931578321544,
|
| 858 |
+
"step": 16384
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.7685557249087802,
|
| 862 |
+
"grad_norm": 0.17042651772499084,
|
| 863 |
+
"learning_rate": 0.000132409666069565,
|
| 864 |
+
"loss": 1.524718999862671,
|
| 865 |
+
"step": 16640
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.780379659138146,
|
| 869 |
+
"grad_norm": 0.18739284574985504,
|
| 870 |
+
"learning_rate": 0.0001197837462455823,
|
| 871 |
+
"loss": 1.5236212015151978,
|
| 872 |
+
"step": 16896
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.7922035933675119,
|
| 876 |
+
"grad_norm": 0.18229515850543976,
|
| 877 |
+
"learning_rate": 0.00010770811321550749,
|
| 878 |
+
"loss": 1.5214073657989502,
|
| 879 |
+
"step": 17152
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.8040275275968778,
|
| 883 |
+
"grad_norm": 0.1200081929564476,
|
| 884 |
+
"learning_rate": 9.620024403698591e-05,
|
| 885 |
+
"loss": 1.5241215229034424,
|
| 886 |
+
"step": 17408
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.8040275275968778,
|
| 890 |
+
"eval_cos_loss": 0.41774995141922067,
|
| 891 |
+
"eval_loss": 1.4673981081405187,
|
| 892 |
+
"eval_mse_loss": 1.2585231351525816,
|
| 893 |
+
"flow/cos_sim": 0.5822500714428349,
|
| 894 |
+
"flow/improvement_ratio": 0.9632431905563563,
|
| 895 |
+
"flow/mag_ratio_mean": 0.5575322135942712,
|
| 896 |
+
"flow/mag_ratio_std": 0.2133726158506794,
|
| 897 |
+
"step": 17408
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.8040275275968778,
|
| 901 |
+
"eval_cos_loss": 0.41774995141922067,
|
| 902 |
+
"eval_loss": 1.4673981081405187,
|
| 903 |
+
"eval_mse_loss": 1.2585231351525816,
|
| 904 |
+
"eval_runtime": 38.8056,
|
| 905 |
+
"eval_samples_per_second": 721.365,
|
| 906 |
+
"eval_steps_per_second": 11.287,
|
| 907 |
+
"flow/cos_sim": 0.5822500714428349,
|
| 908 |
+
"flow/improvement_ratio": 0.9632431905563563,
|
| 909 |
+
"flow/mag_ratio_mean": 0.5575322135942712,
|
| 910 |
+
"flow/mag_ratio_std": 0.2133726158506794,
|
| 911 |
+
"step": 17408
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.8158514618262436,
|
| 915 |
+
"grad_norm": 0.20400674641132355,
|
| 916 |
+
"learning_rate": 8.527679404332429e-05,
|
| 917 |
+
"loss": 1.527288794517517,
|
| 918 |
+
"step": 17664
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.8276753960556095,
|
| 922 |
+
"grad_norm": 0.13825097680091858,
|
| 923 |
+
"learning_rate": 7.495357273823544e-05,
|
| 924 |
+
"loss": 1.5233813524246216,
|
| 925 |
+
"step": 17920
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.8394993302849753,
|
| 929 |
+
"grad_norm": 0.1886366605758667,
|
| 930 |
+
"learning_rate": 6.524552091475183e-05,
|
| 931 |
+
"loss": 1.5250685214996338,
|
| 932 |
+
"step": 18176
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.8513232645143411,
|
| 936 |
+
"grad_norm": 0.12399227172136307,
|
| 937 |
+
"learning_rate": 5.6166689031422024e-05,
|
| 938 |
+
"loss": 1.5211102962493896,
|
| 939 |
+
"step": 18432
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.8513232645143411,
|
| 943 |
+
"eval_cos_loss": 0.4172959377368291,
|
| 944 |
+
"eval_loss": 1.467090511975223,
|
| 945 |
+
"eval_mse_loss": 1.2584425412356581,
|
| 946 |
+
"flow/cos_sim": 0.5827040846489336,
|
| 947 |
+
"flow/improvement_ratio": 0.9641230167591408,
|
| 948 |
+
"flow/mag_ratio_mean": 0.559296594771076,
|
| 949 |
+
"flow/mag_ratio_std": 0.2129400044203349,
|
| 950 |
+
"step": 18432
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8513232645143411,
|
| 954 |
+
"eval_cos_loss": 0.4172959377368291,
|
| 955 |
+
"eval_loss": 1.467090511975223,
|
| 956 |
+
"eval_mse_loss": 1.2584425412356581,
|
| 957 |
+
"eval_runtime": 38.949,
|
| 958 |
+
"eval_samples_per_second": 718.708,
|
| 959 |
+
"eval_steps_per_second": 11.245,
|
| 960 |
+
"flow/cos_sim": 0.5827040846489336,
|
| 961 |
+
"flow/improvement_ratio": 0.9641230167591408,
|
| 962 |
+
"flow/mag_ratio_mean": 0.559296594771076,
|
| 963 |
+
"flow/mag_ratio_std": 0.2129400044203349,
|
| 964 |
+
"step": 18432
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.8631471987437069,
|
| 968 |
+
"grad_norm": 0.15448303520679474,
|
| 969 |
+
"learning_rate": 4.773021687709067e-05,
|
| 970 |
+
"loss": 1.524503231048584,
|
| 971 |
+
"step": 18688
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.8749711329730728,
|
| 975 |
+
"grad_norm": 0.11015952378511429,
|
| 976 |
+
"learning_rate": 3.994831455368719e-05,
|
| 977 |
+
"loss": 1.5246198177337646,
|
| 978 |
+
"step": 18944
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.8867950672024387,
|
| 982 |
+
"grad_norm": 0.14956603944301605,
|
| 983 |
+
"learning_rate": 3.283224480455282e-05,
|
| 984 |
+
"loss": 1.524648666381836,
|
| 985 |
+
"step": 19200
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.8986190014318045,
|
| 989 |
+
"grad_norm": 0.1354081630706787,
|
| 990 |
+
"learning_rate": 2.639230671387627e-05,
|
| 991 |
+
"loss": 1.5228582620620728,
|
| 992 |
+
"step": 19456
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.8986190014318045,
|
| 996 |
+
"eval_cos_loss": 0.4183324677486942,
|
| 997 |
+
"eval_loss": 1.4691238653714254,
|
| 998 |
+
"eval_mse_loss": 1.2599576338785423,
|
| 999 |
+
"flow/cos_sim": 0.5816675356533973,
|
| 1000 |
+
"flow/improvement_ratio": 0.9630278509501453,
|
| 1001 |
+
"flow/mag_ratio_mean": 0.5622789722327228,
|
| 1002 |
+
"flow/mag_ratio_std": 0.21166673544198955,
|
| 1003 |
+
"step": 19456
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8986190014318045,
|
| 1007 |
+
"eval_cos_loss": 0.4183324677486942,
|
| 1008 |
+
"eval_loss": 1.4691238653714254,
|
| 1009 |
+
"eval_mse_loss": 1.2599576338785423,
|
| 1010 |
+
"eval_runtime": 38.8176,
|
| 1011 |
+
"eval_samples_per_second": 721.141,
|
| 1012 |
+
"eval_steps_per_second": 11.284,
|
| 1013 |
+
"flow/cos_sim": 0.5816675356533973,
|
| 1014 |
+
"flow/improvement_ratio": 0.9630278509501453,
|
| 1015 |
+
"flow/mag_ratio_mean": 0.5622789722327228,
|
| 1016 |
+
"flow/mag_ratio_std": 0.21166673544198955,
|
| 1017 |
+
"step": 19456
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9104429356611704,
|
| 1021 |
+
"grad_norm": 0.11552918702363968,
|
| 1022 |
+
"learning_rate": 2.063782080083576e-05,
|
| 1023 |
+
"loss": 1.5217269659042358,
|
| 1024 |
+
"step": 19712
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9222668698905362,
|
| 1028 |
+
"grad_norm": 0.14077740907669067,
|
| 1029 |
+
"learning_rate": 1.557711553001523e-05,
|
| 1030 |
+
"loss": 1.5255944728851318,
|
| 1031 |
+
"step": 19968
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9340908041199021,
|
| 1035 |
+
"grad_norm": 0.11086593568325043,
|
| 1036 |
+
"learning_rate": 1.1217515257622269e-05,
|
| 1037 |
+
"loss": 1.5237233638763428,
|
| 1038 |
+
"step": 20224
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.945914738349268,
|
| 1042 |
+
"grad_norm": 0.1408848911523819,
|
| 1043 |
+
"learning_rate": 7.565329630950746e-06,
|
| 1044 |
+
"loss": 1.5244028568267822,
|
| 1045 |
+
"step": 20480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.945914738349268,
|
| 1049 |
+
"eval_cos_loss": 0.41753342645625546,
|
| 1050 |
+
"eval_loss": 1.4667748912284364,
|
| 1051 |
+
"eval_mse_loss": 1.2580081817766302,
|
| 1052 |
+
"flow/cos_sim": 0.5824665990594315,
|
| 1053 |
+
"flow/improvement_ratio": 0.9622961689620257,
|
| 1054 |
+
"flow/mag_ratio_mean": 0.5584562299730571,
|
| 1055 |
+
"flow/mag_ratio_std": 0.2129541815483951,
|
| 1056 |
+
"step": 20480
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 0.945914738349268,
|
| 1060 |
+
"eval_cos_loss": 0.41753342645625546,
|
| 1061 |
+
"eval_loss": 1.4667748912284364,
|
| 1062 |
+
"eval_mse_loss": 1.2580081817766302,
|
| 1063 |
+
"eval_runtime": 38.5143,
|
| 1064 |
+
"eval_samples_per_second": 726.82,
|
| 1065 |
+
"eval_steps_per_second": 11.372,
|
| 1066 |
+
"flow/cos_sim": 0.5824665990594315,
|
| 1067 |
+
"flow/improvement_ratio": 0.9622961689620257,
|
| 1068 |
+
"flow/mag_ratio_mean": 0.5584562299730571,
|
| 1069 |
+
"flow/mag_ratio_std": 0.2129541815483951,
|
| 1070 |
+
"step": 20480
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.9577386725786338,
|
| 1074 |
+
"grad_norm": 0.11133582890033722,
|
| 1075 |
+
"learning_rate": 4.62584445643166e-06,
|
| 1076 |
+
"loss": 1.5237617492675781,
|
| 1077 |
+
"step": 20736
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.9695626068079997,
|
| 1081 |
+
"grad_norm": 0.13919079303741455,
|
| 1082 |
+
"learning_rate": 2.40331404948807e-06,
|
| 1083 |
+
"loss": 1.5243273973464966,
|
| 1084 |
+
"step": 20992
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 0.9813865410373654,
|
| 1088 |
+
"grad_norm": 0.10569830238819122,
|
| 1089 |
+
"learning_rate": 9.009550772663965e-07,
|
| 1090 |
+
"loss": 1.524424433708191,
|
| 1091 |
+
"step": 21248
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.9932104752667313,
|
| 1095 |
+
"grad_norm": 0.10687630623579025,
|
| 1096 |
+
"learning_rate": 1.2094190315575791e-07,
|
| 1097 |
+
"loss": 1.5235631465911865,
|
| 1098 |
+
"step": 21504
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.9932104752667313,
|
| 1102 |
+
"eval_cos_loss": 0.4184109731896283,
|
| 1103 |
+
"eval_loss": 1.4685546744904017,
|
| 1104 |
+
"eval_mse_loss": 1.25934919020901,
|
| 1105 |
+
"flow/cos_sim": 0.5815890253134514,
|
| 1106 |
+
"flow/improvement_ratio": 0.9625635148455564,
|
| 1107 |
+
"flow/mag_ratio_mean": 0.5583232171731453,
|
| 1108 |
+
"flow/mag_ratio_std": 0.2131242363216126,
|
| 1109 |
+
"step": 21504
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9932104752667313,
|
| 1113 |
+
"eval_cos_loss": 0.4184109731896283,
|
| 1114 |
+
"eval_loss": 1.4685546744904017,
|
| 1115 |
+
"eval_mse_loss": 1.25934919020901,
|
| 1116 |
+
"eval_runtime": 37.8094,
|
| 1117 |
+
"eval_samples_per_second": 740.371,
|
| 1118 |
+
"eval_steps_per_second": 11.584,
|
| 1119 |
+
"flow/cos_sim": 0.5815890253134514,
|
| 1120 |
+
"flow/improvement_ratio": 0.9625635148455564,
|
| 1121 |
+
"flow/mag_ratio_mean": 0.5583232171731453,
|
| 1122 |
+
"flow/mag_ratio_std": 0.2131242363216126,
|
| 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-v4.1/checkpoint-21504/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cfd0fdad5f40d0aeb8f7e3d028467efd29adf0dab2531e2d3f8b374a36e3407a
|
| 3 |
+
size 5137
|