Upload folder using huggingface_hub
Browse files- checkpoints-v3.2-b/checkpoint-21504/ema.safetensors +3 -0
- checkpoints-v3.2-b/checkpoint-21504/model.safetensors +3 -0
- checkpoints-v3.2-b/checkpoint-21504/optimizer.pt +3 -0
- checkpoints-v3.2-b/checkpoint-21504/rng_state.pth +3 -0
- checkpoints-v3.2-b/checkpoint-21504/scaler.pt +3 -0
- checkpoints-v3.2-b/checkpoint-21504/scheduler.pt +3 -0
- checkpoints-v3.2-b/checkpoint-21504/trainer_state.json +1147 -0
- checkpoints-v3.2-b/checkpoint-21504/training_args.bin +3 -0
checkpoints-v3.2-b/checkpoint-21504/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da8b260d8c34d73aeb658bd0690264c82ddba3074468aa01352a77eea604debd
|
| 3 |
+
size 550088
|
checkpoints-v3.2-b/checkpoint-21504/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5f2ca1c5a2d385583390a1d1a48ed644c2f7348e2a56ac87fd7104914a3d3bbe
|
| 3 |
+
size 19318680
|
checkpoints-v3.2-b/checkpoint-21504/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3258d4e895526237c3e4b142c6bcaf572f2b5149a6fd246ec422f0ccf8069fd
|
| 3 |
+
size 1176715
|
checkpoints-v3.2-b/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-v3.2-b/checkpoint-21504/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b57fc65ab7d6cdaf1201b6003b1f8f12cdb886a95b54ecce2001ca0fccb50426
|
| 3 |
+
size 1383
|
checkpoints-v3.2-b/checkpoint-21504/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a6aa2a610e12780122ced40b802e4905fa3f91823425ab5e4628b55f48e2234
|
| 3 |
+
size 1465
|
checkpoints-v3.2-b/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": 0.13531190156936646,
|
| 15 |
+
"learning_rate": 0.0002490234375,
|
| 16 |
+
"loss": 1.4445146322250366,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.1410810798406601,
|
| 22 |
+
"learning_rate": 0.0004990234375,
|
| 23 |
+
"loss": 1.3425910472869873,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.11418232321739197,
|
| 29 |
+
"learning_rate": 0.000499820498011597,
|
| 30 |
+
"loss": 1.3074675798416138,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.1378929316997528,
|
| 36 |
+
"learning_rate": 0.0004992794337021979,
|
| 37 |
+
"loss": 1.2896003723144531,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.44286489105660076,
|
| 43 |
+
"eval_loss": 1.221002193074248,
|
| 44 |
+
"eval_mse_loss": 1.221002193074248,
|
| 45 |
+
"flow/cos_sim": 0.5571351273146938,
|
| 46 |
+
"flow/improvement_ratio": 0.9647975193855425,
|
| 47 |
+
"flow/mag_ratio_mean": 0.5350822540314775,
|
| 48 |
+
"flow/mag_ratio_std": 0.24492420766451586,
|
| 49 |
+
"step": 1024
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.047295736917463395,
|
| 53 |
+
"eval_cos_loss": 0.44286489105660076,
|
| 54 |
+
"eval_loss": 1.221002193074248,
|
| 55 |
+
"eval_mse_loss": 1.221002193074248,
|
| 56 |
+
"eval_runtime": 37.6256,
|
| 57 |
+
"eval_samples_per_second": 743.989,
|
| 58 |
+
"eval_steps_per_second": 11.641,
|
| 59 |
+
"flow/cos_sim": 0.5571351273146938,
|
| 60 |
+
"flow/improvement_ratio": 0.9647975193855425,
|
| 61 |
+
"flow/mag_ratio_mean": 0.5350822540314775,
|
| 62 |
+
"flow/mag_ratio_std": 0.24492420766451586,
|
| 63 |
+
"step": 1024
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.05911967114682925,
|
| 67 |
+
"grad_norm": 0.16314451396465302,
|
| 68 |
+
"learning_rate": 0.0004983775873930693,
|
| 69 |
+
"loss": 1.2742286920547485,
|
| 70 |
+
"step": 1280
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.0709436053761951,
|
| 74 |
+
"grad_norm": 0.1408475935459137,
|
| 75 |
+
"learning_rate": 0.0004971162643259235,
|
| 76 |
+
"loss": 1.2636311054229736,
|
| 77 |
+
"step": 1536
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08276753960556095,
|
| 81 |
+
"grad_norm": 0.4067126512527466,
|
| 82 |
+
"learning_rate": 0.0004954972900130046,
|
| 83 |
+
"loss": 1.2552720308303833,
|
| 84 |
+
"step": 1792
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09459147383492679,
|
| 88 |
+
"grad_norm": 0.11263003200292587,
|
| 89 |
+
"learning_rate": 0.0004935230075950261,
|
| 90 |
+
"loss": 1.250608205795288,
|
| 91 |
+
"step": 2048
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.09459147383492679,
|
| 95 |
+
"eval_cos_loss": 0.4280974593472807,
|
| 96 |
+
"eval_loss": 1.1850101215654312,
|
| 97 |
+
"eval_mse_loss": 1.1850101215654312,
|
| 98 |
+
"flow/cos_sim": 0.5719025665766573,
|
| 99 |
+
"flow/improvement_ratio": 0.9701844343583877,
|
| 100 |
+
"flow/mag_ratio_mean": 0.5505562974437731,
|
| 101 |
+
"flow/mag_ratio_std": 0.2407817519078516,
|
| 102 |
+
"step": 2048
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.09459147383492679,
|
| 106 |
+
"eval_cos_loss": 0.4280974593472807,
|
| 107 |
+
"eval_loss": 1.1850101215654312,
|
| 108 |
+
"eval_mse_loss": 1.1850101215654312,
|
| 109 |
+
"eval_runtime": 36.5824,
|
| 110 |
+
"eval_samples_per_second": 765.204,
|
| 111 |
+
"eval_steps_per_second": 11.973,
|
| 112 |
+
"flow/cos_sim": 0.5719025665766573,
|
| 113 |
+
"flow/improvement_ratio": 0.9701844343583877,
|
| 114 |
+
"flow/mag_ratio_mean": 0.5505562974437731,
|
| 115 |
+
"flow/mag_ratio_std": 0.2407817519078516,
|
| 116 |
+
"step": 2048
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.10641540806429264,
|
| 120 |
+
"grad_norm": 0.23430709540843964,
|
| 121 |
+
"learning_rate": 0.0004911962744499443,
|
| 122 |
+
"loss": 1.2465660572052002,
|
| 123 |
+
"step": 2304
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1182393422936585,
|
| 127 |
+
"grad_norm": 0.12819728255271912,
|
| 128 |
+
"learning_rate": 0.0004885204580574763,
|
| 129 |
+
"loss": 1.2422897815704346,
|
| 130 |
+
"step": 2560
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.13006327652302435,
|
| 134 |
+
"grad_norm": 0.1120893657207489,
|
| 135 |
+
"learning_rate": 0.00048549943112534866,
|
| 136 |
+
"loss": 1.2349907159805298,
|
| 137 |
+
"step": 2816
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1418872107523902,
|
| 141 |
+
"grad_norm": 0.12881016731262207,
|
| 142 |
+
"learning_rate": 0.0004821375659843295,
|
| 143 |
+
"loss": 1.2320677042007446,
|
| 144 |
+
"step": 3072
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.1418872107523902,
|
| 148 |
+
"eval_cos_loss": 0.4224365182799291,
|
| 149 |
+
"eval_loss": 1.1717089387379824,
|
| 150 |
+
"eval_mse_loss": 1.1717089387379824,
|
| 151 |
+
"flow/cos_sim": 0.5775635075079252,
|
| 152 |
+
"flow/improvement_ratio": 0.9726638531303842,
|
| 153 |
+
"flow/mag_ratio_mean": 0.5545849762006437,
|
| 154 |
+
"flow/mag_ratio_std": 0.2390297385927749,
|
| 155 |
+
"step": 3072
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1418872107523902,
|
| 159 |
+
"eval_cos_loss": 0.4224365182799291,
|
| 160 |
+
"eval_loss": 1.1717089387379824,
|
| 161 |
+
"eval_mse_loss": 1.1717089387379824,
|
| 162 |
+
"eval_runtime": 36.5841,
|
| 163 |
+
"eval_samples_per_second": 765.167,
|
| 164 |
+
"eval_steps_per_second": 11.972,
|
| 165 |
+
"flow/cos_sim": 0.5775635075079252,
|
| 166 |
+
"flow/improvement_ratio": 0.9726638531303842,
|
| 167 |
+
"flow/mag_ratio_mean": 0.5545849762006437,
|
| 168 |
+
"flow/mag_ratio_std": 0.2390297385927749,
|
| 169 |
+
"step": 3072
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.15371114498175603,
|
| 173 |
+
"grad_norm": 0.14192286133766174,
|
| 174 |
+
"learning_rate": 0.00047843972826015616,
|
| 175 |
+
"loss": 1.229474663734436,
|
| 176 |
+
"step": 3328
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.1655350792111219,
|
| 180 |
+
"grad_norm": 0.1288514882326126,
|
| 181 |
+
"learning_rate": 0.00047441126983151737,
|
| 182 |
+
"loss": 1.2259553670883179,
|
| 183 |
+
"step": 3584
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17735901344048774,
|
| 187 |
+
"grad_norm": 0.1571502685546875,
|
| 188 |
+
"learning_rate": 0.0004700580210842823,
|
| 189 |
+
"loss": 1.2254003286361694,
|
| 190 |
+
"step": 3840
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18918294766985358,
|
| 194 |
+
"grad_norm": 0.18948955833911896,
|
| 195 |
+
"learning_rate": 0.0004653862824731857,
|
| 196 |
+
"loss": 1.2214388847351074,
|
| 197 |
+
"step": 4096
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.18918294766985358,
|
| 201 |
+
"eval_cos_loss": 0.4189364165748091,
|
| 202 |
+
"eval_loss": 1.162348562179635,
|
| 203 |
+
"eval_mse_loss": 1.162348562179635,
|
| 204 |
+
"flow/cos_sim": 0.5810636061511628,
|
| 205 |
+
"flow/improvement_ratio": 0.971539740420912,
|
| 206 |
+
"flow/mag_ratio_mean": 0.5564619683783892,
|
| 207 |
+
"flow/mag_ratio_std": 0.23926625594700854,
|
| 208 |
+
"step": 4096
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"eval_cos_loss": 0.4189364165748091,
|
| 213 |
+
"eval_loss": 1.162348562179635,
|
| 214 |
+
"eval_mse_loss": 1.162348562179635,
|
| 215 |
+
"eval_runtime": 36.6574,
|
| 216 |
+
"eval_samples_per_second": 763.638,
|
| 217 |
+
"eval_steps_per_second": 11.948,
|
| 218 |
+
"flow/cos_sim": 0.5810636061511628,
|
| 219 |
+
"flow/improvement_ratio": 0.971539740420912,
|
| 220 |
+
"flow/mag_ratio_mean": 0.5564619683783892,
|
| 221 |
+
"flow/mag_ratio_std": 0.23926625594700854,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.20100688189921945,
|
| 226 |
+
"grad_norm": 0.11497154086828232,
|
| 227 |
+
"learning_rate": 0.00046040281540318296,
|
| 228 |
+
"loss": 1.2192964553833008,
|
| 229 |
+
"step": 4352
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.2128308161285853,
|
| 233 |
+
"grad_norm": 0.17087024450302124,
|
| 234 |
+
"learning_rate": 0.00045511483244367227,
|
| 235 |
+
"loss": 1.2178804874420166,
|
| 236 |
+
"step": 4608
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.22465475035795113,
|
| 240 |
+
"grad_norm": 0.15084773302078247,
|
| 241 |
+
"learning_rate": 0.00044952998688974635,
|
| 242 |
+
"loss": 1.2137683629989624,
|
| 243 |
+
"step": 4864
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.236478684587317,
|
| 247 |
+
"grad_norm": 0.17674025893211365,
|
| 248 |
+
"learning_rate": 0.0004436563616855822,
|
| 249 |
+
"loss": 1.214182734489441,
|
| 250 |
+
"step": 5120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.236478684587317,
|
| 254 |
+
"eval_cos_loss": 0.41552880569680095,
|
| 255 |
+
"eval_loss": 1.1498510279611909,
|
| 256 |
+
"eval_mse_loss": 1.1498510279611909,
|
| 257 |
+
"flow/cos_sim": 0.5844712080476491,
|
| 258 |
+
"flow/improvement_ratio": 0.9718176997415552,
|
| 259 |
+
"flow/mag_ratio_mean": 0.5639459366939928,
|
| 260 |
+
"flow/mag_ratio_std": 0.24295896026369643,
|
| 261 |
+
"step": 5120
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.236478684587317,
|
| 265 |
+
"eval_cos_loss": 0.41552880569680095,
|
| 266 |
+
"eval_loss": 1.1498510279611909,
|
| 267 |
+
"eval_mse_loss": 1.1498510279611909,
|
| 268 |
+
"eval_runtime": 36.5045,
|
| 269 |
+
"eval_samples_per_second": 766.837,
|
| 270 |
+
"eval_steps_per_second": 11.999,
|
| 271 |
+
"flow/cos_sim": 0.5844712080476491,
|
| 272 |
+
"flow/improvement_ratio": 0.9718176997415552,
|
| 273 |
+
"flow/mag_ratio_mean": 0.5639459366939928,
|
| 274 |
+
"flow/mag_ratio_std": 0.24295896026369643,
|
| 275 |
+
"step": 5120
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24830261881668284,
|
| 279 |
+
"grad_norm": 0.19734610617160797,
|
| 280 |
+
"learning_rate": 0.00043750245772600053,
|
| 281 |
+
"loss": 1.2093557119369507,
|
| 282 |
+
"step": 5376
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2601265530460487,
|
| 286 |
+
"grad_norm": 0.11800088733434677,
|
| 287 |
+
"learning_rate": 0.00043107718155312435,
|
| 288 |
+
"loss": 1.2076833248138428,
|
| 289 |
+
"step": 5632
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.27195048727541454,
|
| 293 |
+
"grad_norm": 0.11836104094982147,
|
| 294 |
+
"learning_rate": 0.0004243898324659452,
|
| 295 |
+
"loss": 1.2095065116882324,
|
| 296 |
+
"step": 5888
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2837744215047804,
|
| 300 |
+
"grad_norm": 0.14233353734016418,
|
| 301 |
+
"learning_rate": 0.00041745008906145264,
|
| 302 |
+
"loss": 1.2082980871200562,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_cos_loss": 0.41333945823586693,
|
| 308 |
+
"eval_loss": 1.1455941828962874,
|
| 309 |
+
"eval_mse_loss": 1.1455941828962874,
|
| 310 |
+
"flow/cos_sim": 0.5866605730633757,
|
| 311 |
+
"flow/improvement_ratio": 0.9727034600085864,
|
| 312 |
+
"flow/mag_ratio_mean": 0.5605599706031416,
|
| 313 |
+
"flow/mag_ratio_std": 0.24318457544530364,
|
| 314 |
+
"step": 6144
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2837744215047804,
|
| 318 |
+
"eval_cos_loss": 0.41333945823586693,
|
| 319 |
+
"eval_loss": 1.1455941828962874,
|
| 320 |
+
"eval_mse_loss": 1.1455941828962874,
|
| 321 |
+
"eval_runtime": 36.3926,
|
| 322 |
+
"eval_samples_per_second": 769.196,
|
| 323 |
+
"eval_steps_per_second": 12.035,
|
| 324 |
+
"flow/cos_sim": 0.5866605730633757,
|
| 325 |
+
"flow/improvement_ratio": 0.9727034600085864,
|
| 326 |
+
"flow/mag_ratio_mean": 0.5605599706031416,
|
| 327 |
+
"flow/mag_ratio_std": 0.24318457544530364,
|
| 328 |
+
"step": 6144
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2955983557341462,
|
| 332 |
+
"grad_norm": 0.1180572658777237,
|
| 333 |
+
"learning_rate": 0.00041026799522680534,
|
| 334 |
+
"loss": 1.2049767971038818,
|
| 335 |
+
"step": 6400
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.30742228996351206,
|
| 339 |
+
"grad_norm": 0.14954492449760437,
|
| 340 |
+
"learning_rate": 0.00040285394560281816,
|
| 341 |
+
"loss": 1.203136682510376,
|
| 342 |
+
"step": 6656
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3192462241928779,
|
| 346 |
+
"grad_norm": 0.27384424209594727,
|
| 347 |
+
"learning_rate": 0.0003952186705398043,
|
| 348 |
+
"loss": 1.2036312818527222,
|
| 349 |
+
"step": 6912
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.3310701584222438,
|
| 353 |
+
"grad_norm": 0.15835484862327576,
|
| 354 |
+
"learning_rate": 0.0003873732205675438,
|
| 355 |
+
"loss": 1.1990163326263428,
|
| 356 |
+
"step": 7168
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.3310701584222438,
|
| 360 |
+
"eval_cos_loss": 0.41148509575079567,
|
| 361 |
+
"eval_loss": 1.1415519687138735,
|
| 362 |
+
"eval_mse_loss": 1.1415519687138735,
|
| 363 |
+
"flow/cos_sim": 0.5885149201029511,
|
| 364 |
+
"flow/improvement_ratio": 0.9741524475622395,
|
| 365 |
+
"flow/mag_ratio_mean": 0.5640481464122529,
|
| 366 |
+
"flow/mag_ratio_std": 0.24272260563150388,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3310701584222438,
|
| 371 |
+
"eval_cos_loss": 0.41148509575079567,
|
| 372 |
+
"eval_loss": 1.1415519687138735,
|
| 373 |
+
"eval_mse_loss": 1.1415519687138735,
|
| 374 |
+
"eval_runtime": 36.5211,
|
| 375 |
+
"eval_samples_per_second": 766.489,
|
| 376 |
+
"eval_steps_per_second": 11.993,
|
| 377 |
+
"flow/cos_sim": 0.5885149201029511,
|
| 378 |
+
"flow/improvement_ratio": 0.9741524475622395,
|
| 379 |
+
"flow/mag_ratio_mean": 0.5640481464122529,
|
| 380 |
+
"flow/mag_ratio_std": 0.24272260563150388,
|
| 381 |
+
"step": 7168
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.34289409265160964,
|
| 385 |
+
"grad_norm": 0.13862715661525726,
|
| 386 |
+
"learning_rate": 0.000379328950401858,
|
| 387 |
+
"loss": 1.1999462842941284,
|
| 388 |
+
"step": 7424
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3547180268809755,
|
| 392 |
+
"grad_norm": 0.1664004623889923,
|
| 393 |
+
"learning_rate": 0.0003710975025109345,
|
| 394 |
+
"loss": 1.1982899904251099,
|
| 395 |
+
"step": 7680
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3665419611103413,
|
| 399 |
+
"grad_norm": 0.13604609668254852,
|
| 400 |
+
"learning_rate": 0.0003626907902651893,
|
| 401 |
+
"loss": 1.1958998441696167,
|
| 402 |
+
"step": 7936
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.37836589533970716,
|
| 406 |
+
"grad_norm": 0.1457708328962326,
|
| 407 |
+
"learning_rate": 0.0003541209806950514,
|
| 408 |
+
"loss": 1.1945908069610596,
|
| 409 |
+
"step": 8192
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.37836589533970716,
|
| 413 |
+
"eval_cos_loss": 0.4107164820851801,
|
| 414 |
+
"eval_loss": 1.1382312957010312,
|
| 415 |
+
"eval_mse_loss": 1.1382312957010312,
|
| 416 |
+
"flow/cos_sim": 0.589283528393262,
|
| 417 |
+
"flow/improvement_ratio": 0.9722015831840637,
|
| 418 |
+
"flow/mag_ratio_mean": 0.5640345150477266,
|
| 419 |
+
"flow/mag_ratio_std": 0.24401044291040125,
|
| 420 |
+
"step": 8192
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.37836589533970716,
|
| 424 |
+
"eval_cos_loss": 0.4107164820851801,
|
| 425 |
+
"eval_loss": 1.1382312957010312,
|
| 426 |
+
"eval_mse_loss": 1.1382312957010312,
|
| 427 |
+
"eval_runtime": 36.7013,
|
| 428 |
+
"eval_samples_per_second": 762.726,
|
| 429 |
+
"eval_steps_per_second": 11.934,
|
| 430 |
+
"flow/cos_sim": 0.589283528393262,
|
| 431 |
+
"flow/improvement_ratio": 0.9722015831840637,
|
| 432 |
+
"flow/mag_ratio_mean": 0.5640345150477266,
|
| 433 |
+
"flow/mag_ratio_std": 0.24401044291040125,
|
| 434 |
+
"step": 8192
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.390189829569073,
|
| 438 |
+
"grad_norm": 0.16633214056491852,
|
| 439 |
+
"learning_rate": 0.0003454004768816257,
|
| 440 |
+
"loss": 1.1974270343780518,
|
| 441 |
+
"step": 8448
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4020137637984389,
|
| 445 |
+
"grad_norm": 0.13664524257183075,
|
| 446 |
+
"learning_rate": 0.00033654190000572017,
|
| 447 |
+
"loss": 1.1954563856124878,
|
| 448 |
+
"step": 8704
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.41383769802780473,
|
| 452 |
+
"grad_norm": 0.17472244799137115,
|
| 453 |
+
"learning_rate": 0.00032755807108121707,
|
| 454 |
+
"loss": 1.1933891773223877,
|
| 455 |
+
"step": 8960
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.4256616322571706,
|
| 459 |
+
"grad_norm": 0.13021035492420197,
|
| 460 |
+
"learning_rate": 0.00031846199239922587,
|
| 461 |
+
"loss": 1.1937364339828491,
|
| 462 |
+
"step": 9216
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4256616322571706,
|
| 466 |
+
"eval_cos_loss": 0.40818845524907654,
|
| 467 |
+
"eval_loss": 1.1306110892665986,
|
| 468 |
+
"eval_mse_loss": 1.1306110892665986,
|
| 469 |
+
"flow/cos_sim": 0.5918115706748615,
|
| 470 |
+
"flow/improvement_ratio": 0.9736669311240383,
|
| 471 |
+
"flow/mag_ratio_mean": 0.5680791953382971,
|
| 472 |
+
"flow/mag_ratio_std": 0.24117264903435423,
|
| 473 |
+
"step": 9216
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4256616322571706,
|
| 477 |
+
"eval_cos_loss": 0.40818845524907654,
|
| 478 |
+
"eval_loss": 1.1306110892665986,
|
| 479 |
+
"eval_mse_loss": 1.1306110892665986,
|
| 480 |
+
"eval_runtime": 37.6562,
|
| 481 |
+
"eval_samples_per_second": 743.384,
|
| 482 |
+
"eval_steps_per_second": 11.632,
|
| 483 |
+
"flow/cos_sim": 0.5918115706748615,
|
| 484 |
+
"flow/improvement_ratio": 0.9736669311240383,
|
| 485 |
+
"flow/mag_ratio_mean": 0.5680791953382971,
|
| 486 |
+
"flow/mag_ratio_std": 0.24117264903435423,
|
| 487 |
+
"step": 9216
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.4374855664865364,
|
| 491 |
+
"grad_norm": 0.1436239778995514,
|
| 492 |
+
"learning_rate": 0.00030926682870987393,
|
| 493 |
+
"loss": 1.1944515705108643,
|
| 494 |
+
"step": 9472
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.44930950071590225,
|
| 498 |
+
"grad_norm": 0.1279948502779007,
|
| 499 |
+
"learning_rate": 0.00029998588816897034,
|
| 500 |
+
"loss": 1.192986011505127,
|
| 501 |
+
"step": 9728
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4611334349452681,
|
| 505 |
+
"grad_norm": 0.16707991063594818,
|
| 506 |
+
"learning_rate": 0.00029063260307711817,
|
| 507 |
+
"loss": 1.1940622329711914,
|
| 508 |
+
"step": 9984
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.472957369174634,
|
| 512 |
+
"grad_norm": 0.12476647645235062,
|
| 513 |
+
"learning_rate": 0.00028122051043915356,
|
| 514 |
+
"loss": 1.190167784690857,
|
| 515 |
+
"step": 10240
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.472957369174634,
|
| 519 |
+
"eval_cos_loss": 0.4076138965479315,
|
| 520 |
+
"eval_loss": 1.1312634806110435,
|
| 521 |
+
"eval_mse_loss": 1.1312634806110435,
|
| 522 |
+
"flow/cos_sim": 0.5923861386296956,
|
| 523 |
+
"flow/improvement_ratio": 0.9748937231764946,
|
| 524 |
+
"flow/mag_ratio_mean": 0.5703548358999975,
|
| 525 |
+
"flow/mag_ratio_std": 0.24083318081620622,
|
| 526 |
+
"step": 10240
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.472957369174634,
|
| 530 |
+
"eval_cos_loss": 0.4076138965479315,
|
| 531 |
+
"eval_loss": 1.1312634806110435,
|
| 532 |
+
"eval_mse_loss": 1.1312634806110435,
|
| 533 |
+
"eval_runtime": 37.3647,
|
| 534 |
+
"eval_samples_per_second": 749.184,
|
| 535 |
+
"eval_steps_per_second": 11.722,
|
| 536 |
+
"flow/cos_sim": 0.5923861386296956,
|
| 537 |
+
"flow/improvement_ratio": 0.9748937231764946,
|
| 538 |
+
"flow/mag_ratio_mean": 0.5703548358999975,
|
| 539 |
+
"flow/mag_ratio_std": 0.24083318081620622,
|
| 540 |
+
"step": 10240
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.48478130340399983,
|
| 544 |
+
"grad_norm": 0.16375412046909332,
|
| 545 |
+
"learning_rate": 0.00027176323237204403,
|
| 546 |
+
"loss": 1.1899298429489136,
|
| 547 |
+
"step": 10496
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.49660523763336567,
|
| 551 |
+
"grad_norm": 0.13618549704551697,
|
| 552 |
+
"learning_rate": 0.0002622744563896065,
|
| 553 |
+
"loss": 1.1911197900772095,
|
| 554 |
+
"step": 10752
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.5084291718627315,
|
| 558 |
+
"grad_norm": 0.14637431502342224,
|
| 559 |
+
"learning_rate": 0.00025276791559257494,
|
| 560 |
+
"loss": 1.1925891637802124,
|
| 561 |
+
"step": 11008
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5202531060920974,
|
| 565 |
+
"grad_norm": 0.13531003892421722,
|
| 566 |
+
"learning_rate": 0.0002432573687926906,
|
| 567 |
+
"loss": 1.186372995376587,
|
| 568 |
+
"step": 11264
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.5202531060920974,
|
| 572 |
+
"eval_cos_loss": 0.40703500609963994,
|
| 573 |
+
"eval_loss": 1.1297938347407128,
|
| 574 |
+
"eval_mse_loss": 1.1297938347407128,
|
| 575 |
+
"flow/cos_sim": 0.5929650133603239,
|
| 576 |
+
"flow/improvement_ratio": 0.9741698461308327,
|
| 577 |
+
"flow/mag_ratio_mean": 0.5677392150713428,
|
| 578 |
+
"flow/mag_ratio_std": 0.24396490948657468,
|
| 579 |
+
"step": 11264
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.5202531060920974,
|
| 583 |
+
"eval_cos_loss": 0.40703500609963994,
|
| 584 |
+
"eval_loss": 1.1297938347407128,
|
| 585 |
+
"eval_mse_loss": 1.1297938347407128,
|
| 586 |
+
"eval_runtime": 38.0112,
|
| 587 |
+
"eval_samples_per_second": 736.441,
|
| 588 |
+
"eval_steps_per_second": 11.523,
|
| 589 |
+
"flow/cos_sim": 0.5929650133603239,
|
| 590 |
+
"flow/improvement_ratio": 0.9741698461308327,
|
| 591 |
+
"flow/mag_ratio_mean": 0.5677392150713428,
|
| 592 |
+
"flow/mag_ratio_std": 0.24396490948657468,
|
| 593 |
+
"step": 11264
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5320770403214632,
|
| 597 |
+
"grad_norm": 0.13568931818008423,
|
| 598 |
+
"learning_rate": 0.00023375658059958035,
|
| 599 |
+
"loss": 1.1851963996887207,
|
| 600 |
+
"step": 11520
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.5439009745508291,
|
| 604 |
+
"grad_norm": 0.158619686961174,
|
| 605 |
+
"learning_rate": 0.00022427930149924495,
|
| 606 |
+
"loss": 1.18683660030365,
|
| 607 |
+
"step": 11776
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.5557249087801949,
|
| 611 |
+
"grad_norm": 0.12816773355007172,
|
| 612 |
+
"learning_rate": 0.00021483924795298633,
|
| 613 |
+
"loss": 1.1887577772140503,
|
| 614 |
+
"step": 12032
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5675488430095608,
|
| 618 |
+
"grad_norm": 0.1509481817483902,
|
| 619 |
+
"learning_rate": 0.00020545008254558105,
|
| 620 |
+
"loss": 1.1869347095489502,
|
| 621 |
+
"step": 12288
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5675488430095608,
|
| 625 |
+
"eval_cos_loss": 0.4058912583546007,
|
| 626 |
+
"eval_loss": 1.1245840085695868,
|
| 627 |
+
"eval_mse_loss": 1.1245840085695868,
|
| 628 |
+
"flow/cos_sim": 0.5941087536888036,
|
| 629 |
+
"flow/improvement_ratio": 0.9735238416009842,
|
| 630 |
+
"flow/mag_ratio_mean": 0.5702283619201347,
|
| 631 |
+
"flow/mag_ratio_std": 0.2440737300131419,
|
| 632 |
+
"step": 12288
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5675488430095608,
|
| 636 |
+
"eval_cos_loss": 0.4058912583546007,
|
| 637 |
+
"eval_loss": 1.1245840085695868,
|
| 638 |
+
"eval_mse_loss": 1.1245840085695868,
|
| 639 |
+
"eval_runtime": 37.443,
|
| 640 |
+
"eval_samples_per_second": 747.617,
|
| 641 |
+
"eval_steps_per_second": 11.698,
|
| 642 |
+
"flow/cos_sim": 0.5941087536888036,
|
| 643 |
+
"flow/improvement_ratio": 0.9735238416009842,
|
| 644 |
+
"flow/mag_ratio_mean": 0.5702283619201347,
|
| 645 |
+
"flow/mag_ratio_std": 0.2440737300131419,
|
| 646 |
+
"step": 12288
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5793727772389267,
|
| 650 |
+
"grad_norm": 0.12331920862197876,
|
| 651 |
+
"learning_rate": 0.00019612539421142756,
|
| 652 |
+
"loss": 1.1875509023666382,
|
| 653 |
+
"step": 12544
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5911967114682924,
|
| 657 |
+
"grad_norm": 0.14404380321502686,
|
| 658 |
+
"learning_rate": 0.00018687867856728862,
|
| 659 |
+
"loss": 1.186294436454773,
|
| 660 |
+
"step": 12800
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6030206456976583,
|
| 664 |
+
"grad_norm": 0.16519756615161896,
|
| 665 |
+
"learning_rate": 0.00017772331838009138,
|
| 666 |
+
"loss": 1.1816554069519043,
|
| 667 |
+
"step": 13056
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6148445799270241,
|
| 671 |
+
"grad_norm": 0.1261790543794632,
|
| 672 |
+
"learning_rate": 0.00016867256419805628,
|
| 673 |
+
"loss": 1.184910774230957,
|
| 674 |
+
"step": 13312
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6148445799270241,
|
| 678 |
+
"eval_cos_loss": 0.4058965126129046,
|
| 679 |
+
"eval_loss": 1.124463065301991,
|
| 680 |
+
"eval_mse_loss": 1.124463065301991,
|
| 681 |
+
"flow/cos_sim": 0.5941034868427607,
|
| 682 |
+
"flow/improvement_ratio": 0.9736478256580492,
|
| 683 |
+
"flow/mag_ratio_mean": 0.5693308626680069,
|
| 684 |
+
"flow/mag_ratio_std": 0.24294855394592024,
|
| 685 |
+
"step": 13312
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.6148445799270241,
|
| 689 |
+
"eval_cos_loss": 0.4058965126129046,
|
| 690 |
+
"eval_loss": 1.124463065301991,
|
| 691 |
+
"eval_mse_loss": 1.124463065301991,
|
| 692 |
+
"eval_runtime": 37.6078,
|
| 693 |
+
"eval_samples_per_second": 744.34,
|
| 694 |
+
"eval_steps_per_second": 11.647,
|
| 695 |
+
"flow/cos_sim": 0.5941034868427607,
|
| 696 |
+
"flow/improvement_ratio": 0.9736478256580492,
|
| 697 |
+
"flow/mag_ratio_mean": 0.5693308626680069,
|
| 698 |
+
"flow/mag_ratio_std": 0.24294855394592024,
|
| 699 |
+
"step": 13312
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.62666851415639,
|
| 703 |
+
"grad_norm": 0.10855361819267273,
|
| 704 |
+
"learning_rate": 0.00015973951517318435,
|
| 705 |
+
"loss": 1.1825791597366333,
|
| 706 |
+
"step": 13568
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.6384924483857558,
|
| 710 |
+
"grad_norm": 0.12764589488506317,
|
| 711 |
+
"learning_rate": 0.00015093710010286203,
|
| 712 |
+
"loss": 1.183720588684082,
|
| 713 |
+
"step": 13824
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.6503163826151217,
|
| 717 |
+
"grad_norm": 0.15194828808307648,
|
| 718 |
+
"learning_rate": 0.00014227805871801813,
|
| 719 |
+
"loss": 1.1843342781066895,
|
| 720 |
+
"step": 14080
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.6621403168444876,
|
| 724 |
+
"grad_norm": 0.1313941776752472,
|
| 725 |
+
"learning_rate": 0.00013377492324491864,
|
| 726 |
+
"loss": 1.1845036745071411,
|
| 727 |
+
"step": 14336
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.6621403168444876,
|
| 731 |
+
"eval_cos_loss": 0.4055524829043645,
|
| 732 |
+
"eval_loss": 1.1236549692066837,
|
| 733 |
+
"eval_mse_loss": 1.1236549692066837,
|
| 734 |
+
"flow/cos_sim": 0.5944475390051054,
|
| 735 |
+
"flow/improvement_ratio": 0.9741561509430681,
|
| 736 |
+
"flow/mag_ratio_mean": 0.5722755660477294,
|
| 737 |
+
"flow/mag_ratio_std": 0.24292387908588262,
|
| 738 |
+
"step": 14336
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6621403168444876,
|
| 742 |
+
"eval_cos_loss": 0.4055524829043645,
|
| 743 |
+
"eval_loss": 1.1236549692066837,
|
| 744 |
+
"eval_mse_loss": 1.1236549692066837,
|
| 745 |
+
"eval_runtime": 37.766,
|
| 746 |
+
"eval_samples_per_second": 741.222,
|
| 747 |
+
"eval_steps_per_second": 11.598,
|
| 748 |
+
"flow/cos_sim": 0.5944475390051054,
|
| 749 |
+
"flow/improvement_ratio": 0.9741561509430681,
|
| 750 |
+
"flow/mag_ratio_mean": 0.5722755660477294,
|
| 751 |
+
"flow/mag_ratio_std": 0.24292387908588262,
|
| 752 |
+
"step": 14336
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6739642510738534,
|
| 756 |
+
"grad_norm": 0.1419185996055603,
|
| 757 |
+
"learning_rate": 0.00012544000026728114,
|
| 758 |
+
"loss": 1.1815910339355469,
|
| 759 |
+
"step": 14592
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6857881853032193,
|
| 763 |
+
"grad_norm": 0.11921229213476181,
|
| 764 |
+
"learning_rate": 0.00011728535291496281,
|
| 765 |
+
"loss": 1.1859562397003174,
|
| 766 |
+
"step": 14848
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6976121195325851,
|
| 770 |
+
"grad_norm": 0.13644331693649292,
|
| 771 |
+
"learning_rate": 0.00010932278340499846,
|
| 772 |
+
"loss": 1.1809968948364258,
|
| 773 |
+
"step": 15104
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.709436053761951,
|
| 777 |
+
"grad_norm": 0.12353217601776123,
|
| 778 |
+
"learning_rate": 0.0001015638159602576,
|
| 779 |
+
"loss": 1.1835533380508423,
|
| 780 |
+
"step": 15360
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.709436053761951,
|
| 784 |
+
"eval_cos_loss": 0.404491749596378,
|
| 785 |
+
"eval_loss": 1.1212830510857987,
|
| 786 |
+
"eval_mse_loss": 1.1212830510857987,
|
| 787 |
+
"flow/cos_sim": 0.5955082661212852,
|
| 788 |
+
"flow/improvement_ratio": 0.9744905962791617,
|
| 789 |
+
"flow/mag_ratio_mean": 0.5719217330081278,
|
| 790 |
+
"flow/mag_ratio_std": 0.24169711207282052,
|
| 791 |
+
"step": 15360
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.709436053761951,
|
| 795 |
+
"eval_cos_loss": 0.404491749596378,
|
| 796 |
+
"eval_loss": 1.1212830510857987,
|
| 797 |
+
"eval_mse_loss": 1.1212830510857987,
|
| 798 |
+
"eval_runtime": 37.817,
|
| 799 |
+
"eval_samples_per_second": 740.223,
|
| 800 |
+
"eval_steps_per_second": 11.582,
|
| 801 |
+
"flow/cos_sim": 0.5955082661212852,
|
| 802 |
+
"flow/improvement_ratio": 0.9744905962791617,
|
| 803 |
+
"flow/mag_ratio_mean": 0.5719217330081278,
|
| 804 |
+
"flow/mag_ratio_std": 0.24169711207282052,
|
| 805 |
+
"step": 15360
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.7212599879913169,
|
| 809 |
+
"grad_norm": 0.11799845844507217,
|
| 810 |
+
"learning_rate": 9.401968013044271e-05,
|
| 811 |
+
"loss": 1.1806869506835938,
|
| 812 |
+
"step": 15616
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.7330839222206826,
|
| 816 |
+
"grad_norm": 0.12838932871818542,
|
| 817 |
+
"learning_rate": 8.670129453956732e-05,
|
| 818 |
+
"loss": 1.182599663734436,
|
| 819 |
+
"step": 15872
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.7449078564500485,
|
| 823 |
+
"grad_norm": 0.11126351356506348,
|
| 824 |
+
"learning_rate": 7.961925108343715e-05,
|
| 825 |
+
"loss": 1.1809351444244385,
|
| 826 |
+
"step": 16128
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.7567317906794143,
|
| 830 |
+
"grad_norm": 0.13469421863555908,
|
| 831 |
+
"learning_rate": 7.278379960000436e-05,
|
| 832 |
+
"loss": 1.1837177276611328,
|
| 833 |
+
"step": 16384
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.7567317906794143,
|
| 837 |
+
"eval_cos_loss": 0.40498853165265086,
|
| 838 |
+
"eval_loss": 1.1224167082952037,
|
| 839 |
+
"eval_mse_loss": 1.1224167082952037,
|
| 840 |
+
"flow/cos_sim": 0.5950114819557155,
|
| 841 |
+
"flow/improvement_ratio": 0.9733862189669588,
|
| 842 |
+
"flow/mag_ratio_mean": 0.5712394598684355,
|
| 843 |
+
"flow/mag_ratio_std": 0.24294140764839572,
|
| 844 |
+
"step": 16384
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.7567317906794143,
|
| 848 |
+
"eval_cos_loss": 0.40498853165265086,
|
| 849 |
+
"eval_loss": 1.1224167082952037,
|
| 850 |
+
"eval_mse_loss": 1.1224167082952037,
|
| 851 |
+
"eval_runtime": 38.0074,
|
| 852 |
+
"eval_samples_per_second": 736.514,
|
| 853 |
+
"eval_steps_per_second": 11.524,
|
| 854 |
+
"flow/cos_sim": 0.5950114819557155,
|
| 855 |
+
"flow/improvement_ratio": 0.9733862189669588,
|
| 856 |
+
"flow/mag_ratio_mean": 0.5712394598684355,
|
| 857 |
+
"flow/mag_ratio_std": 0.24294140764839572,
|
| 858 |
+
"step": 16384
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.7685557249087802,
|
| 862 |
+
"grad_norm": 0.13311493396759033,
|
| 863 |
+
"learning_rate": 6.62048330347825e-05,
|
| 864 |
+
"loss": 1.1809964179992676,
|
| 865 |
+
"step": 16640
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.780379659138146,
|
| 869 |
+
"grad_norm": 0.10799466073513031,
|
| 870 |
+
"learning_rate": 5.989187312279115e-05,
|
| 871 |
+
"loss": 1.180505633354187,
|
| 872 |
+
"step": 16896
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.7922035933675119,
|
| 876 |
+
"grad_norm": 0.1486814022064209,
|
| 877 |
+
"learning_rate": 5.3854056607753746e-05,
|
| 878 |
+
"loss": 1.1795099973678589,
|
| 879 |
+
"step": 17152
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.8040275275968778,
|
| 883 |
+
"grad_norm": 0.11168931424617767,
|
| 884 |
+
"learning_rate": 4.8100122018492956e-05,
|
| 885 |
+
"loss": 1.1790964603424072,
|
| 886 |
+
"step": 17408
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.8040275275968778,
|
| 890 |
+
"eval_cos_loss": 0.40403676060236754,
|
| 891 |
+
"eval_loss": 1.1197874216728558,
|
| 892 |
+
"eval_mse_loss": 1.1197874216728558,
|
| 893 |
+
"flow/cos_sim": 0.5959632730102975,
|
| 894 |
+
"flow/improvement_ratio": 0.974381895370135,
|
| 895 |
+
"flow/mag_ratio_mean": 0.5709017501574128,
|
| 896 |
+
"flow/mag_ratio_std": 0.24267656927647657,
|
| 897 |
+
"step": 17408
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.8040275275968778,
|
| 901 |
+
"eval_cos_loss": 0.40403676060236754,
|
| 902 |
+
"eval_loss": 1.1197874216728558,
|
| 903 |
+
"eval_mse_loss": 1.1197874216728558,
|
| 904 |
+
"eval_runtime": 37.0697,
|
| 905 |
+
"eval_samples_per_second": 755.145,
|
| 906 |
+
"eval_steps_per_second": 11.816,
|
| 907 |
+
"flow/cos_sim": 0.5959632730102975,
|
| 908 |
+
"flow/improvement_ratio": 0.974381895370135,
|
| 909 |
+
"flow/mag_ratio_mean": 0.5709017501574128,
|
| 910 |
+
"flow/mag_ratio_std": 0.24267656927647657,
|
| 911 |
+
"step": 17408
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.8158514618262436,
|
| 915 |
+
"grad_norm": 0.120413638651371,
|
| 916 |
+
"learning_rate": 4.263839702166214e-05,
|
| 917 |
+
"loss": 1.1833932399749756,
|
| 918 |
+
"step": 17664
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.8276753960556095,
|
| 922 |
+
"grad_norm": 0.13419309258460999,
|
| 923 |
+
"learning_rate": 3.747678636911772e-05,
|
| 924 |
+
"loss": 1.1790565252304077,
|
| 925 |
+
"step": 17920
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.8394993302849753,
|
| 929 |
+
"grad_norm": 0.13175411522388458,
|
| 930 |
+
"learning_rate": 3.262276045737592e-05,
|
| 931 |
+
"loss": 1.182090401649475,
|
| 932 |
+
"step": 18176
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.8513232645143411,
|
| 936 |
+
"grad_norm": 0.09947279095649719,
|
| 937 |
+
"learning_rate": 2.8083344515711012e-05,
|
| 938 |
+
"loss": 1.177965760231018,
|
| 939 |
+
"step": 18432
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.8513232645143411,
|
| 943 |
+
"eval_cos_loss": 0.40370703621269904,
|
| 944 |
+
"eval_loss": 1.1215725149193856,
|
| 945 |
+
"eval_mse_loss": 1.1215725149193856,
|
| 946 |
+
"flow/cos_sim": 0.5962929766472072,
|
| 947 |
+
"flow/improvement_ratio": 0.9753795086520992,
|
| 948 |
+
"flow/mag_ratio_mean": 0.5721857149034875,
|
| 949 |
+
"flow/mag_ratio_std": 0.24184994011709135,
|
| 950 |
+
"step": 18432
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8513232645143411,
|
| 954 |
+
"eval_cos_loss": 0.40370703621269904,
|
| 955 |
+
"eval_loss": 1.1215725149193856,
|
| 956 |
+
"eval_mse_loss": 1.1215725149193856,
|
| 957 |
+
"eval_runtime": 37.003,
|
| 958 |
+
"eval_samples_per_second": 756.506,
|
| 959 |
+
"eval_steps_per_second": 11.837,
|
| 960 |
+
"flow/cos_sim": 0.5962929766472072,
|
| 961 |
+
"flow/improvement_ratio": 0.9753795086520992,
|
| 962 |
+
"flow/mag_ratio_mean": 0.5721857149034875,
|
| 963 |
+
"flow/mag_ratio_std": 0.24184994011709135,
|
| 964 |
+
"step": 18432
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.8631471987437069,
|
| 968 |
+
"grad_norm": 0.11105956137180328,
|
| 969 |
+
"learning_rate": 2.3865108438545337e-05,
|
| 970 |
+
"loss": 1.181681752204895,
|
| 971 |
+
"step": 18688
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.8749711329730728,
|
| 975 |
+
"grad_norm": 0.10993946343660355,
|
| 976 |
+
"learning_rate": 1.9974157276843596e-05,
|
| 977 |
+
"loss": 1.1794625520706177,
|
| 978 |
+
"step": 18944
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.8867950672024387,
|
| 982 |
+
"grad_norm": 0.10621558874845505,
|
| 983 |
+
"learning_rate": 1.641612240227641e-05,
|
| 984 |
+
"loss": 1.1792101860046387,
|
| 985 |
+
"step": 19200
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.8986190014318045,
|
| 989 |
+
"grad_norm": 0.09712148457765579,
|
| 990 |
+
"learning_rate": 1.3196153356938134e-05,
|
| 991 |
+
"loss": 1.1793529987335205,
|
| 992 |
+
"step": 19456
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.8986190014318045,
|
| 996 |
+
"eval_cos_loss": 0.40458166646902965,
|
| 997 |
+
"eval_loss": 1.1218491670203536,
|
| 998 |
+
"eval_mse_loss": 1.1218491670203536,
|
| 999 |
+
"flow/cos_sim": 0.5954183448939563,
|
| 1000 |
+
"flow/improvement_ratio": 0.9741832541820665,
|
| 1001 |
+
"flow/mag_ratio_mean": 0.5714860147537162,
|
| 1002 |
+
"flow/mag_ratio_std": 0.2427709357787485,
|
| 1003 |
+
"step": 19456
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8986190014318045,
|
| 1007 |
+
"eval_cos_loss": 0.40458166646902965,
|
| 1008 |
+
"eval_loss": 1.1218491670203536,
|
| 1009 |
+
"eval_mse_loss": 1.1218491670203536,
|
| 1010 |
+
"eval_runtime": 37.3338,
|
| 1011 |
+
"eval_samples_per_second": 749.804,
|
| 1012 |
+
"eval_steps_per_second": 11.732,
|
| 1013 |
+
"flow/cos_sim": 0.5954183448939563,
|
| 1014 |
+
"flow/improvement_ratio": 0.9741832541820665,
|
| 1015 |
+
"flow/mag_ratio_mean": 0.5714860147537162,
|
| 1016 |
+
"flow/mag_ratio_std": 0.2427709357787485,
|
| 1017 |
+
"step": 19456
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9104429356611704,
|
| 1021 |
+
"grad_norm": 0.12275788933038712,
|
| 1022 |
+
"learning_rate": 1.031891040041788e-05,
|
| 1023 |
+
"loss": 1.1787521839141846,
|
| 1024 |
+
"step": 19712
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9222668698905362,
|
| 1028 |
+
"grad_norm": 0.1388554871082306,
|
| 1029 |
+
"learning_rate": 7.788557765007614e-06,
|
| 1030 |
+
"loss": 1.1809200048446655,
|
| 1031 |
+
"step": 19968
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9340908041199021,
|
| 1035 |
+
"grad_norm": 0.11414396017789841,
|
| 1036 |
+
"learning_rate": 5.6087576288111344e-06,
|
| 1037 |
+
"loss": 1.1808236837387085,
|
| 1038 |
+
"step": 20224
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.945914738349268,
|
| 1042 |
+
"grad_norm": 0.11537494510412216,
|
| 1043 |
+
"learning_rate": 3.782664815475373e-06,
|
| 1044 |
+
"loss": 1.1786174774169922,
|
| 1045 |
+
"step": 20480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.945914738349268,
|
| 1049 |
+
"eval_cos_loss": 0.4043068891246569,
|
| 1050 |
+
"eval_loss": 1.120300586909464,
|
| 1051 |
+
"eval_mse_loss": 1.120300586909464,
|
| 1052 |
+
"flow/cos_sim": 0.595693129110554,
|
| 1053 |
+
"flow/improvement_ratio": 0.9729666603754644,
|
| 1054 |
+
"flow/mag_ratio_mean": 0.5716023456016087,
|
| 1055 |
+
"flow/mag_ratio_std": 0.24340959124641331,
|
| 1056 |
+
"step": 20480
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 0.945914738349268,
|
| 1060 |
+
"eval_cos_loss": 0.4043068891246569,
|
| 1061 |
+
"eval_loss": 1.120300586909464,
|
| 1062 |
+
"eval_mse_loss": 1.120300586909464,
|
| 1063 |
+
"eval_runtime": 37.0085,
|
| 1064 |
+
"eval_samples_per_second": 756.394,
|
| 1065 |
+
"eval_steps_per_second": 11.835,
|
| 1066 |
+
"flow/cos_sim": 0.595693129110554,
|
| 1067 |
+
"flow/improvement_ratio": 0.9729666603754644,
|
| 1068 |
+
"flow/mag_ratio_mean": 0.5716023456016087,
|
| 1069 |
+
"flow/mag_ratio_std": 0.24340959124641331,
|
| 1070 |
+
"step": 20480
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.9577386725786338,
|
| 1074 |
+
"grad_norm": 0.12133973091840744,
|
| 1075 |
+
"learning_rate": 2.31292222821583e-06,
|
| 1076 |
+
"loss": 1.1803685426712036,
|
| 1077 |
+
"step": 20736
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.9695626068079997,
|
| 1081 |
+
"grad_norm": 0.10521073639392853,
|
| 1082 |
+
"learning_rate": 1.201657024744035e-06,
|
| 1083 |
+
"loss": 1.1818820238113403,
|
| 1084 |
+
"step": 20992
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 0.9813865410373654,
|
| 1088 |
+
"grad_norm": 0.09989389777183533,
|
| 1089 |
+
"learning_rate": 4.5047753863319827e-07,
|
| 1090 |
+
"loss": 1.1803994178771973,
|
| 1091 |
+
"step": 21248
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.9932104752667313,
|
| 1095 |
+
"grad_norm": 0.09399425238370895,
|
| 1096 |
+
"learning_rate": 6.047095157787896e-08,
|
| 1097 |
+
"loss": 1.1804349422454834,
|
| 1098 |
+
"step": 21504
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.9932104752667313,
|
| 1102 |
+
"eval_cos_loss": 0.40440767044073916,
|
| 1103 |
+
"eval_loss": 1.1205239834850782,
|
| 1104 |
+
"eval_mse_loss": 1.1205239834850782,
|
| 1105 |
+
"flow/cos_sim": 0.5955923152024343,
|
| 1106 |
+
"flow/improvement_ratio": 0.9742439836944075,
|
| 1107 |
+
"flow/mag_ratio_mean": 0.5713475014521107,
|
| 1108 |
+
"flow/mag_ratio_std": 0.24330289396521163,
|
| 1109 |
+
"step": 21504
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9932104752667313,
|
| 1113 |
+
"eval_cos_loss": 0.40440767044073916,
|
| 1114 |
+
"eval_loss": 1.1205239834850782,
|
| 1115 |
+
"eval_mse_loss": 1.1205239834850782,
|
| 1116 |
+
"eval_runtime": 37.1012,
|
| 1117 |
+
"eval_samples_per_second": 754.504,
|
| 1118 |
+
"eval_steps_per_second": 11.806,
|
| 1119 |
+
"flow/cos_sim": 0.5955923152024343,
|
| 1120 |
+
"flow/improvement_ratio": 0.9742439836944075,
|
| 1121 |
+
"flow/mag_ratio_mean": 0.5713475014521107,
|
| 1122 |
+
"flow/mag_ratio_std": 0.24330289396521163,
|
| 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-b/checkpoint-21504/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:07b8a673a900fa9d101c78dda5f0e27bec2f7072aac1c6ae379dfa47ec615e8b
|
| 3 |
+
size 5137
|