Training in progress, step 955
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9994767137624281,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.15591317970715268,
|
| 5 |
+
"train_runtime": 13974.7976,
|
| 6 |
+
"train_samples": 122268,
|
| 7 |
+
"train_samples_per_second": 8.749,
|
| 8 |
+
"train_steps_per_second": 0.068
|
| 9 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.46.3"
|
| 6 |
+
}
|
model-00001-of-00003.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4943162336
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b030ae9a61de1228d0be54c2700e14db8b35b81f95742f90a7bf10e63545337f
|
| 3 |
size 4943162336
|
runs/Dec25_18-35-31_ip-10-0-27-237.ec2.internal/events.out.tfevents.1735151795.ip-10-0-27-237.ec2.internal.18852.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e69992077e35b1c0123d6bcd1db2b789047bad70743944481f2bc1ec6d84702
|
| 3 |
+
size 72645
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9994767137624281,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.15591317970715268,
|
| 5 |
+
"train_runtime": 13974.7976,
|
| 6 |
+
"train_samples": 122268,
|
| 7 |
+
"train_samples_per_second": 8.749,
|
| 8 |
+
"train_steps_per_second": 0.068
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1482 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9994767137624281,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 955,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0010465724751439038,
|
| 13 |
+
"grad_norm": 9.516955534709263,
|
| 14 |
+
"learning_rate": 5.208333333333333e-09,
|
| 15 |
+
"logits/chosen": -3.21875,
|
| 16 |
+
"logits/rejected": -3.21875,
|
| 17 |
+
"logps/chosen": -250.0,
|
| 18 |
+
"logps/rejected": -364.0,
|
| 19 |
+
"loss": 0.6914,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/chosen": 0.0,
|
| 22 |
+
"rewards/margins": 0.0,
|
| 23 |
+
"rewards/rejected": 0.0,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.010465724751439037,
|
| 28 |
+
"grad_norm": 9.605250953667445,
|
| 29 |
+
"learning_rate": 5.208333333333333e-08,
|
| 30 |
+
"logits/chosen": -3.25,
|
| 31 |
+
"logits/rejected": -3.28125,
|
| 32 |
+
"logps/chosen": -298.0,
|
| 33 |
+
"logps/rejected": -278.0,
|
| 34 |
+
"loss": 0.6921,
|
| 35 |
+
"rewards/accuracies": 0.1527777761220932,
|
| 36 |
+
"rewards/chosen": -0.00176239013671875,
|
| 37 |
+
"rewards/margins": -0.0012969970703125,
|
| 38 |
+
"rewards/rejected": -0.0004596710205078125,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.020931449502878074,
|
| 43 |
+
"grad_norm": 9.292065125128957,
|
| 44 |
+
"learning_rate": 1.0416666666666667e-07,
|
| 45 |
+
"logits/chosen": -3.078125,
|
| 46 |
+
"logits/rejected": -3.171875,
|
| 47 |
+
"logps/chosen": -286.0,
|
| 48 |
+
"logps/rejected": -294.0,
|
| 49 |
+
"loss": 0.6915,
|
| 50 |
+
"rewards/accuracies": 0.22499999403953552,
|
| 51 |
+
"rewards/chosen": -0.001617431640625,
|
| 52 |
+
"rewards/margins": -0.000492095947265625,
|
| 53 |
+
"rewards/rejected": -0.00112152099609375,
|
| 54 |
+
"step": 20
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.03139717425431711,
|
| 58 |
+
"grad_norm": 8.81073696781217,
|
| 59 |
+
"learning_rate": 1.5624999999999999e-07,
|
| 60 |
+
"logits/chosen": -3.21875,
|
| 61 |
+
"logits/rejected": -3.25,
|
| 62 |
+
"logps/chosen": -294.0,
|
| 63 |
+
"logps/rejected": -264.0,
|
| 64 |
+
"loss": 0.69,
|
| 65 |
+
"rewards/accuracies": 0.39375001192092896,
|
| 66 |
+
"rewards/chosen": -0.00153350830078125,
|
| 67 |
+
"rewards/margins": 0.003387451171875,
|
| 68 |
+
"rewards/rejected": -0.004913330078125,
|
| 69 |
+
"step": 30
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.04186289900575615,
|
| 73 |
+
"grad_norm": 9.511442470624056,
|
| 74 |
+
"learning_rate": 2.0833333333333333e-07,
|
| 75 |
+
"logits/chosen": -3.1875,
|
| 76 |
+
"logits/rejected": -3.125,
|
| 77 |
+
"logps/chosen": -288.0,
|
| 78 |
+
"logps/rejected": -302.0,
|
| 79 |
+
"loss": 0.684,
|
| 80 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 81 |
+
"rewards/chosen": -0.006072998046875,
|
| 82 |
+
"rewards/margins": 0.0186767578125,
|
| 83 |
+
"rewards/rejected": -0.0247802734375,
|
| 84 |
+
"step": 40
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.052328623757195186,
|
| 88 |
+
"grad_norm": 8.623245631146178,
|
| 89 |
+
"learning_rate": 2.604166666666667e-07,
|
| 90 |
+
"logits/chosen": -3.25,
|
| 91 |
+
"logits/rejected": -3.21875,
|
| 92 |
+
"logps/chosen": -318.0,
|
| 93 |
+
"logps/rejected": -302.0,
|
| 94 |
+
"loss": 0.6707,
|
| 95 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 96 |
+
"rewards/chosen": -0.0244140625,
|
| 97 |
+
"rewards/margins": 0.0498046875,
|
| 98 |
+
"rewards/rejected": -0.07421875,
|
| 99 |
+
"step": 50
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.06279434850863422,
|
| 103 |
+
"grad_norm": 11.270983496881312,
|
| 104 |
+
"learning_rate": 3.1249999999999997e-07,
|
| 105 |
+
"logits/chosen": -3.15625,
|
| 106 |
+
"logits/rejected": -3.171875,
|
| 107 |
+
"logps/chosen": -296.0,
|
| 108 |
+
"logps/rejected": -316.0,
|
| 109 |
+
"loss": 0.6463,
|
| 110 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 111 |
+
"rewards/chosen": -0.046875,
|
| 112 |
+
"rewards/margins": 0.09375,
|
| 113 |
+
"rewards/rejected": -0.140625,
|
| 114 |
+
"step": 60
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.07326007326007326,
|
| 118 |
+
"grad_norm": 11.953436287497002,
|
| 119 |
+
"learning_rate": 3.645833333333333e-07,
|
| 120 |
+
"logits/chosen": -3.109375,
|
| 121 |
+
"logits/rejected": -3.109375,
|
| 122 |
+
"logps/chosen": -310.0,
|
| 123 |
+
"logps/rejected": -304.0,
|
| 124 |
+
"loss": 0.5956,
|
| 125 |
+
"rewards/accuracies": 0.856249988079071,
|
| 126 |
+
"rewards/chosen": -0.0294189453125,
|
| 127 |
+
"rewards/margins": 0.26171875,
|
| 128 |
+
"rewards/rejected": -0.291015625,
|
| 129 |
+
"step": 70
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.0837257980115123,
|
| 133 |
+
"grad_norm": 24.459142709298142,
|
| 134 |
+
"learning_rate": 4.1666666666666667e-07,
|
| 135 |
+
"logits/chosen": -3.046875,
|
| 136 |
+
"logits/rejected": -3.078125,
|
| 137 |
+
"logps/chosen": -326.0,
|
| 138 |
+
"logps/rejected": -348.0,
|
| 139 |
+
"loss": 0.518,
|
| 140 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 141 |
+
"rewards/chosen": -0.1044921875,
|
| 142 |
+
"rewards/margins": 0.57421875,
|
| 143 |
+
"rewards/rejected": -0.6796875,
|
| 144 |
+
"step": 80
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.09419152276295134,
|
| 148 |
+
"grad_norm": 24.49272478624807,
|
| 149 |
+
"learning_rate": 4.6874999999999996e-07,
|
| 150 |
+
"logits/chosen": -3.0625,
|
| 151 |
+
"logits/rejected": -3.046875,
|
| 152 |
+
"logps/chosen": -324.0,
|
| 153 |
+
"logps/rejected": -352.0,
|
| 154 |
+
"loss": 0.4382,
|
| 155 |
+
"rewards/accuracies": 0.856249988079071,
|
| 156 |
+
"rewards/chosen": -0.255859375,
|
| 157 |
+
"rewards/margins": 0.69921875,
|
| 158 |
+
"rewards/rejected": -0.953125,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.10465724751439037,
|
| 163 |
+
"grad_norm": 32.563301578465556,
|
| 164 |
+
"learning_rate": 4.999732492681437e-07,
|
| 165 |
+
"logits/chosen": -2.96875,
|
| 166 |
+
"logits/rejected": -3.0,
|
| 167 |
+
"logps/chosen": -346.0,
|
| 168 |
+
"logps/rejected": -488.0,
|
| 169 |
+
"loss": 0.3249,
|
| 170 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 171 |
+
"rewards/chosen": -0.46484375,
|
| 172 |
+
"rewards/margins": 1.4375,
|
| 173 |
+
"rewards/rejected": -1.90625,
|
| 174 |
+
"step": 100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.1151229722658294,
|
| 178 |
+
"grad_norm": 29.67260704692337,
|
| 179 |
+
"learning_rate": 4.996723692767926e-07,
|
| 180 |
+
"logits/chosen": -2.984375,
|
| 181 |
+
"logits/rejected": -2.96875,
|
| 182 |
+
"logps/chosen": -338.0,
|
| 183 |
+
"logps/rejected": -528.0,
|
| 184 |
+
"loss": 0.2642,
|
| 185 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 186 |
+
"rewards/chosen": -0.59375,
|
| 187 |
+
"rewards/margins": 1.9296875,
|
| 188 |
+
"rewards/rejected": -2.515625,
|
| 189 |
+
"step": 110
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.12558869701726844,
|
| 193 |
+
"grad_norm": 36.130980567601185,
|
| 194 |
+
"learning_rate": 4.990375746213598e-07,
|
| 195 |
+
"logits/chosen": -2.875,
|
| 196 |
+
"logits/rejected": -2.84375,
|
| 197 |
+
"logps/chosen": -368.0,
|
| 198 |
+
"logps/rejected": -608.0,
|
| 199 |
+
"loss": 0.2784,
|
| 200 |
+
"rewards/accuracies": 0.8687499761581421,
|
| 201 |
+
"rewards/chosen": -0.9765625,
|
| 202 |
+
"rewards/margins": 2.109375,
|
| 203 |
+
"rewards/rejected": -3.09375,
|
| 204 |
+
"step": 120
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.1360544217687075,
|
| 208 |
+
"grad_norm": 39.623525128446786,
|
| 209 |
+
"learning_rate": 4.980697142834314e-07,
|
| 210 |
+
"logits/chosen": -2.8125,
|
| 211 |
+
"logits/rejected": -2.78125,
|
| 212 |
+
"logps/chosen": -390.0,
|
| 213 |
+
"logps/rejected": -588.0,
|
| 214 |
+
"loss": 0.2208,
|
| 215 |
+
"rewards/accuracies": 0.925000011920929,
|
| 216 |
+
"rewards/chosen": -0.92578125,
|
| 217 |
+
"rewards/margins": 2.296875,
|
| 218 |
+
"rewards/rejected": -3.234375,
|
| 219 |
+
"step": 130
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.14652014652014653,
|
| 223 |
+
"grad_norm": 26.762330480495592,
|
| 224 |
+
"learning_rate": 4.967700826904229e-07,
|
| 225 |
+
"logits/chosen": -2.875,
|
| 226 |
+
"logits/rejected": -2.8125,
|
| 227 |
+
"logps/chosen": -360.0,
|
| 228 |
+
"logps/rejected": -664.0,
|
| 229 |
+
"loss": 0.2179,
|
| 230 |
+
"rewards/accuracies": 0.925000011920929,
|
| 231 |
+
"rewards/chosen": -0.859375,
|
| 232 |
+
"rewards/margins": 2.9375,
|
| 233 |
+
"rewards/rejected": -3.796875,
|
| 234 |
+
"step": 140
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.15698587127158556,
|
| 238 |
+
"grad_norm": 40.56285073504209,
|
| 239 |
+
"learning_rate": 4.951404179843962e-07,
|
| 240 |
+
"logits/chosen": -2.765625,
|
| 241 |
+
"logits/rejected": -2.65625,
|
| 242 |
+
"logps/chosen": -438.0,
|
| 243 |
+
"logps/rejected": -668.0,
|
| 244 |
+
"loss": 0.1995,
|
| 245 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 246 |
+
"rewards/chosen": -1.2734375,
|
| 247 |
+
"rewards/margins": 2.828125,
|
| 248 |
+
"rewards/rejected": -4.09375,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.1674515960230246,
|
| 253 |
+
"grad_norm": 19.47918869184068,
|
| 254 |
+
"learning_rate": 4.931828996974498e-07,
|
| 255 |
+
"logits/chosen": -2.765625,
|
| 256 |
+
"logits/rejected": -2.6875,
|
| 257 |
+
"logps/chosen": -452.0,
|
| 258 |
+
"logps/rejected": -704.0,
|
| 259 |
+
"loss": 0.2069,
|
| 260 |
+
"rewards/accuracies": 0.90625,
|
| 261 |
+
"rewards/chosen": -1.2734375,
|
| 262 |
+
"rewards/margins": 2.90625,
|
| 263 |
+
"rewards/rejected": -4.1875,
|
| 264 |
+
"step": 160
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.17791732077446362,
|
| 268 |
+
"grad_norm": 54.76579795757994,
|
| 269 |
+
"learning_rate": 4.909001458367866e-07,
|
| 270 |
+
"logits/chosen": -2.71875,
|
| 271 |
+
"logits/rejected": -2.640625,
|
| 272 |
+
"logps/chosen": -366.0,
|
| 273 |
+
"logps/rejected": -716.0,
|
| 274 |
+
"loss": 0.2041,
|
| 275 |
+
"rewards/accuracies": 0.9375,
|
| 276 |
+
"rewards/chosen": -1.0234375,
|
| 277 |
+
"rewards/margins": 3.140625,
|
| 278 |
+
"rewards/rejected": -4.15625,
|
| 279 |
+
"step": 170
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.18838304552590268,
|
| 283 |
+
"grad_norm": 26.167531224154025,
|
| 284 |
+
"learning_rate": 4.882952093833627e-07,
|
| 285 |
+
"logits/chosen": -2.71875,
|
| 286 |
+
"logits/rejected": -2.578125,
|
| 287 |
+
"logps/chosen": -402.0,
|
| 288 |
+
"logps/rejected": -736.0,
|
| 289 |
+
"loss": 0.2019,
|
| 290 |
+
"rewards/accuracies": 0.925000011920929,
|
| 291 |
+
"rewards/chosen": -1.28125,
|
| 292 |
+
"rewards/margins": 3.34375,
|
| 293 |
+
"rewards/rejected": -4.625,
|
| 294 |
+
"step": 180
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.1988487702773417,
|
| 298 |
+
"grad_norm": 19.72905224051703,
|
| 299 |
+
"learning_rate": 4.853715742087946e-07,
|
| 300 |
+
"logits/chosen": -2.515625,
|
| 301 |
+
"logits/rejected": -2.453125,
|
| 302 |
+
"logps/chosen": -434.0,
|
| 303 |
+
"logps/rejected": -768.0,
|
| 304 |
+
"loss": 0.1627,
|
| 305 |
+
"rewards/accuracies": 0.925000011920929,
|
| 306 |
+
"rewards/chosen": -1.5703125,
|
| 307 |
+
"rewards/margins": 3.609375,
|
| 308 |
+
"rewards/rejected": -5.1875,
|
| 309 |
+
"step": 190
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.20931449502878074,
|
| 313 |
+
"grad_norm": 18.779653204870545,
|
| 314 |
+
"learning_rate": 4.821331504159906e-07,
|
| 315 |
+
"logits/chosen": -2.65625,
|
| 316 |
+
"logits/rejected": -2.40625,
|
| 317 |
+
"logps/chosen": -482.0,
|
| 318 |
+
"logps/rejected": -828.0,
|
| 319 |
+
"loss": 0.172,
|
| 320 |
+
"rewards/accuracies": 0.925000011920929,
|
| 321 |
+
"rewards/chosen": -1.78125,
|
| 322 |
+
"rewards/margins": 3.734375,
|
| 323 |
+
"rewards/rejected": -5.5,
|
| 324 |
+
"step": 200
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.21978021978021978,
|
| 328 |
+
"grad_norm": 22.81262811076015,
|
| 329 |
+
"learning_rate": 4.785842691097342e-07,
|
| 330 |
+
"logits/chosen": -2.546875,
|
| 331 |
+
"logits/rejected": -2.40625,
|
| 332 |
+
"logps/chosen": -422.0,
|
| 333 |
+
"logps/rejected": -824.0,
|
| 334 |
+
"loss": 0.1704,
|
| 335 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 336 |
+
"rewards/chosen": -1.34375,
|
| 337 |
+
"rewards/margins": 4.1875,
|
| 338 |
+
"rewards/rejected": -5.5,
|
| 339 |
+
"step": 210
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2302459445316588,
|
| 343 |
+
"grad_norm": 44.85290714547607,
|
| 344 |
+
"learning_rate": 4.7472967660421603e-07,
|
| 345 |
+
"logits/chosen": -2.546875,
|
| 346 |
+
"logits/rejected": -2.4375,
|
| 347 |
+
"logps/chosen": -468.0,
|
| 348 |
+
"logps/rejected": -864.0,
|
| 349 |
+
"loss": 0.1787,
|
| 350 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 351 |
+
"rewards/chosen": -1.703125,
|
| 352 |
+
"rewards/margins": 4.03125,
|
| 353 |
+
"rewards/rejected": -5.71875,
|
| 354 |
+
"step": 220
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.24071166928309787,
|
| 358 |
+
"grad_norm": 23.212078131512556,
|
| 359 |
+
"learning_rate": 4.705745280752585e-07,
|
| 360 |
+
"logits/chosen": -2.65625,
|
| 361 |
+
"logits/rejected": -2.46875,
|
| 362 |
+
"logps/chosen": -424.0,
|
| 363 |
+
"logps/rejected": -804.0,
|
| 364 |
+
"loss": 0.1676,
|
| 365 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 366 |
+
"rewards/chosen": -1.2734375,
|
| 367 |
+
"rewards/margins": 3.984375,
|
| 368 |
+
"rewards/rejected": -5.25,
|
| 369 |
+
"step": 230
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.25117739403453687,
|
| 373 |
+
"grad_norm": 49.148200668176614,
|
| 374 |
+
"learning_rate": 4.6612438066572555e-07,
|
| 375 |
+
"logits/chosen": -2.34375,
|
| 376 |
+
"logits/rejected": -2.0625,
|
| 377 |
+
"logps/chosen": -496.0,
|
| 378 |
+
"logps/rejected": -976.0,
|
| 379 |
+
"loss": 0.1638,
|
| 380 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 381 |
+
"rewards/chosen": -1.8515625,
|
| 382 |
+
"rewards/margins": 4.9375,
|
| 383 |
+
"rewards/rejected": -6.78125,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.2616431187859759,
|
| 388 |
+
"grad_norm": 27.312656594695248,
|
| 389 |
+
"learning_rate": 4.6138518605333664e-07,
|
| 390 |
+
"logits/chosen": -2.484375,
|
| 391 |
+
"logits/rejected": -2.3125,
|
| 392 |
+
"logps/chosen": -444.0,
|
| 393 |
+
"logps/rejected": -812.0,
|
| 394 |
+
"loss": 0.158,
|
| 395 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 396 |
+
"rewards/chosen": -1.6171875,
|
| 397 |
+
"rewards/margins": 3.96875,
|
| 398 |
+
"rewards/rejected": -5.59375,
|
| 399 |
+
"step": 250
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.272108843537415,
|
| 403 |
+
"grad_norm": 22.296118147448365,
|
| 404 |
+
"learning_rate": 4.5636328249082514e-07,
|
| 405 |
+
"logits/chosen": -2.359375,
|
| 406 |
+
"logits/rejected": -2.1875,
|
| 407 |
+
"logps/chosen": -494.0,
|
| 408 |
+
"logps/rejected": -960.0,
|
| 409 |
+
"loss": 0.1445,
|
| 410 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 411 |
+
"rewards/chosen": -2.015625,
|
| 412 |
+
"rewards/margins": 4.53125,
|
| 413 |
+
"rewards/rejected": -6.5625,
|
| 414 |
+
"step": 260
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.282574568288854,
|
| 418 |
+
"grad_norm": 22.50903492692647,
|
| 419 |
+
"learning_rate": 4.510653863290871e-07,
|
| 420 |
+
"logits/chosen": -2.34375,
|
| 421 |
+
"logits/rejected": -2.0,
|
| 422 |
+
"logps/chosen": -448.0,
|
| 423 |
+
"logps/rejected": -904.0,
|
| 424 |
+
"loss": 0.1423,
|
| 425 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 426 |
+
"rewards/chosen": -1.484375,
|
| 427 |
+
"rewards/margins": 4.84375,
|
| 428 |
+
"rewards/rejected": -6.34375,
|
| 429 |
+
"step": 270
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.29304029304029305,
|
| 433 |
+
"grad_norm": 28.21995400188032,
|
| 434 |
+
"learning_rate": 4.4549858303465737e-07,
|
| 435 |
+
"logits/chosen": -2.390625,
|
| 436 |
+
"logits/rejected": -2.171875,
|
| 437 |
+
"logps/chosen": -516.0,
|
| 438 |
+
"logps/rejected": -980.0,
|
| 439 |
+
"loss": 0.1437,
|
| 440 |
+
"rewards/accuracies": 0.90625,
|
| 441 |
+
"rewards/chosen": -2.265625,
|
| 442 |
+
"rewards/margins": 4.5625,
|
| 443 |
+
"rewards/rejected": -6.84375,
|
| 444 |
+
"step": 280
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.3035060177917321,
|
| 448 |
+
"grad_norm": 40.19255553854356,
|
| 449 |
+
"learning_rate": 4.396703177135261e-07,
|
| 450 |
+
"logits/chosen": -2.21875,
|
| 451 |
+
"logits/rejected": -1.984375,
|
| 452 |
+
"logps/chosen": -568.0,
|
| 453 |
+
"logps/rejected": -1096.0,
|
| 454 |
+
"loss": 0.1329,
|
| 455 |
+
"rewards/accuracies": 0.956250011920929,
|
| 456 |
+
"rewards/chosen": -2.953125,
|
| 457 |
+
"rewards/margins": 5.5,
|
| 458 |
+
"rewards/rejected": -8.4375,
|
| 459 |
+
"step": 290
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.3139717425431711,
|
| 463 |
+
"grad_norm": 28.430828005352467,
|
| 464 |
+
"learning_rate": 4.335883851539693e-07,
|
| 465 |
+
"logits/chosen": -2.296875,
|
| 466 |
+
"logits/rejected": -1.9375,
|
| 467 |
+
"logps/chosen": -648.0,
|
| 468 |
+
"logps/rejected": -1168.0,
|
| 469 |
+
"loss": 0.1374,
|
| 470 |
+
"rewards/accuracies": 0.918749988079071,
|
| 471 |
+
"rewards/chosen": -3.34375,
|
| 472 |
+
"rewards/margins": 5.4375,
|
| 473 |
+
"rewards/rejected": -8.75,
|
| 474 |
+
"step": 300
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.32443746729461015,
|
| 478 |
+
"grad_norm": 61.735019074562494,
|
| 479 |
+
"learning_rate": 4.272609194017105e-07,
|
| 480 |
+
"logits/chosen": -2.1875,
|
| 481 |
+
"logits/rejected": -1.6953125,
|
| 482 |
+
"logps/chosen": -676.0,
|
| 483 |
+
"logps/rejected": -1224.0,
|
| 484 |
+
"loss": 0.1284,
|
| 485 |
+
"rewards/accuracies": 0.918749988079071,
|
| 486 |
+
"rewards/chosen": -3.734375,
|
| 487 |
+
"rewards/margins": 5.5,
|
| 488 |
+
"rewards/rejected": -9.1875,
|
| 489 |
+
"step": 310
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.3349031920460492,
|
| 493 |
+
"grad_norm": 44.72666475922118,
|
| 494 |
+
"learning_rate": 4.2069638288135547e-07,
|
| 495 |
+
"logits/chosen": -2.109375,
|
| 496 |
+
"logits/rejected": -1.796875,
|
| 497 |
+
"logps/chosen": -736.0,
|
| 498 |
+
"logps/rejected": -1360.0,
|
| 499 |
+
"loss": 0.1419,
|
| 500 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 501 |
+
"rewards/chosen": -4.3125,
|
| 502 |
+
"rewards/margins": 6.34375,
|
| 503 |
+
"rewards/rejected": -10.625,
|
| 504 |
+
"step": 320
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.3453689167974882,
|
| 508 |
+
"grad_norm": 25.59488492513282,
|
| 509 |
+
"learning_rate": 4.139035550786494e-07,
|
| 510 |
+
"logits/chosen": -2.328125,
|
| 511 |
+
"logits/rejected": -2.203125,
|
| 512 |
+
"logps/chosen": -684.0,
|
| 513 |
+
"logps/rejected": -1160.0,
|
| 514 |
+
"loss": 0.1288,
|
| 515 |
+
"rewards/accuracies": 0.893750011920929,
|
| 516 |
+
"rewards/chosen": -4.0625,
|
| 517 |
+
"rewards/margins": 4.6875,
|
| 518 |
+
"rewards/rejected": -8.75,
|
| 519 |
+
"step": 330
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.35583464154892724,
|
| 523 |
+
"grad_norm": 39.490355460521485,
|
| 524 |
+
"learning_rate": 4.0689152079869306e-07,
|
| 525 |
+
"logits/chosen": -2.109375,
|
| 526 |
+
"logits/rejected": -1.7109375,
|
| 527 |
+
"logps/chosen": -668.0,
|
| 528 |
+
"logps/rejected": -1336.0,
|
| 529 |
+
"loss": 0.1067,
|
| 530 |
+
"rewards/accuracies": 0.96875,
|
| 531 |
+
"rewards/chosen": -4.09375,
|
| 532 |
+
"rewards/margins": 6.84375,
|
| 533 |
+
"rewards/rejected": -10.9375,
|
| 534 |
+
"step": 340
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.3663003663003663,
|
| 538 |
+
"grad_norm": 40.05108849367682,
|
| 539 |
+
"learning_rate": 3.99669658015821e-07,
|
| 540 |
+
"logits/chosen": -2.0625,
|
| 541 |
+
"logits/rejected": -1.6796875,
|
| 542 |
+
"logps/chosen": -708.0,
|
| 543 |
+
"logps/rejected": -1344.0,
|
| 544 |
+
"loss": 0.1218,
|
| 545 |
+
"rewards/accuracies": 0.949999988079071,
|
| 546 |
+
"rewards/chosen": -4.5,
|
| 547 |
+
"rewards/margins": 6.6875,
|
| 548 |
+
"rewards/rejected": -11.1875,
|
| 549 |
+
"step": 350
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.37676609105180536,
|
| 553 |
+
"grad_norm": 29.496296075978837,
|
| 554 |
+
"learning_rate": 3.92247625331392e-07,
|
| 555 |
+
"logits/chosen": -2.109375,
|
| 556 |
+
"logits/rejected": -1.875,
|
| 557 |
+
"logps/chosen": -716.0,
|
| 558 |
+
"logps/rejected": -1296.0,
|
| 559 |
+
"loss": 0.1126,
|
| 560 |
+
"rewards/accuracies": 0.925000011920929,
|
| 561 |
+
"rewards/chosen": -4.25,
|
| 562 |
+
"rewards/margins": 6.0625,
|
| 563 |
+
"rewards/rejected": -10.3125,
|
| 564 |
+
"step": 360
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.3872318158032444,
|
| 568 |
+
"grad_norm": 25.08871689723511,
|
| 569 |
+
"learning_rate": 3.846353490562664e-07,
|
| 570 |
+
"logits/chosen": -2.109375,
|
| 571 |
+
"logits/rejected": -1.5,
|
| 572 |
+
"logps/chosen": -776.0,
|
| 573 |
+
"logps/rejected": -1496.0,
|
| 574 |
+
"loss": 0.1248,
|
| 575 |
+
"rewards/accuracies": 0.956250011920929,
|
| 576 |
+
"rewards/chosen": -5.125,
|
| 577 |
+
"rewards/margins": 7.3125,
|
| 578 |
+
"rewards/rejected": -12.4375,
|
| 579 |
+
"step": 370
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.3976975405546834,
|
| 583 |
+
"grad_norm": 26.78647304619011,
|
| 584 |
+
"learning_rate": 3.768430099352445e-07,
|
| 585 |
+
"logits/chosen": -2.234375,
|
| 586 |
+
"logits/rejected": -1.96875,
|
| 587 |
+
"logps/chosen": -756.0,
|
| 588 |
+
"logps/rejected": -1344.0,
|
| 589 |
+
"loss": 0.1186,
|
| 590 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 591 |
+
"rewards/chosen": -4.46875,
|
| 592 |
+
"rewards/margins": 6.09375,
|
| 593 |
+
"rewards/rejected": -10.5625,
|
| 594 |
+
"step": 380
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 0.40816326530612246,
|
| 598 |
+
"grad_norm": 47.07672408947705,
|
| 599 |
+
"learning_rate": 3.6888102953122304e-07,
|
| 600 |
+
"logits/chosen": -2.265625,
|
| 601 |
+
"logits/rejected": -1.9765625,
|
| 602 |
+
"logps/chosen": -712.0,
|
| 603 |
+
"logps/rejected": -1336.0,
|
| 604 |
+
"loss": 0.103,
|
| 605 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 606 |
+
"rewards/chosen": -4.21875,
|
| 607 |
+
"rewards/margins": 6.5,
|
| 608 |
+
"rewards/rejected": -10.75,
|
| 609 |
+
"step": 390
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.4186289900575615,
|
| 613 |
+
"grad_norm": 22.46877888992614,
|
| 614 |
+
"learning_rate": 3.607600562872785e-07,
|
| 615 |
+
"logits/chosen": -2.359375,
|
| 616 |
+
"logits/rejected": -1.953125,
|
| 617 |
+
"logps/chosen": -876.0,
|
| 618 |
+
"logps/rejected": -1592.0,
|
| 619 |
+
"loss": 0.0937,
|
| 620 |
+
"rewards/accuracies": 0.96875,
|
| 621 |
+
"rewards/chosen": -5.34375,
|
| 622 |
+
"rewards/margins": 7.5625,
|
| 623 |
+
"rewards/rejected": -12.875,
|
| 624 |
+
"step": 400
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.4290947148090005,
|
| 628 |
+
"grad_norm": 31.318574320005773,
|
| 629 |
+
"learning_rate": 3.5249095128531856e-07,
|
| 630 |
+
"logits/chosen": -2.40625,
|
| 631 |
+
"logits/rejected": -2.171875,
|
| 632 |
+
"logps/chosen": -796.0,
|
| 633 |
+
"logps/rejected": -1328.0,
|
| 634 |
+
"loss": 0.1072,
|
| 635 |
+
"rewards/accuracies": 0.9375,
|
| 636 |
+
"rewards/chosen": -4.59375,
|
| 637 |
+
"rewards/margins": 5.9375,
|
| 638 |
+
"rewards/rejected": -10.5,
|
| 639 |
+
"step": 410
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.43956043956043955,
|
| 643 |
+
"grad_norm": 23.57271093807849,
|
| 644 |
+
"learning_rate": 3.4408477372034736e-07,
|
| 645 |
+
"logits/chosen": -2.4375,
|
| 646 |
+
"logits/rejected": -2.09375,
|
| 647 |
+
"logps/chosen": -760.0,
|
| 648 |
+
"logps/rejected": -1368.0,
|
| 649 |
+
"loss": 0.1199,
|
| 650 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 651 |
+
"rewards/chosen": -4.46875,
|
| 652 |
+
"rewards/margins": 6.625,
|
| 653 |
+
"rewards/rejected": -11.125,
|
| 654 |
+
"step": 420
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4500261643118786,
|
| 658 |
+
"grad_norm": 20.257810507961672,
|
| 659 |
+
"learning_rate": 3.3555276610977276e-07,
|
| 660 |
+
"logits/chosen": -2.5,
|
| 661 |
+
"logits/rejected": -2.328125,
|
| 662 |
+
"logps/chosen": -568.0,
|
| 663 |
+
"logps/rejected": -1184.0,
|
| 664 |
+
"loss": 0.0937,
|
| 665 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 666 |
+
"rewards/chosen": -2.8125,
|
| 667 |
+
"rewards/margins": 6.125,
|
| 668 |
+
"rewards/rejected": -8.9375,
|
| 669 |
+
"step": 430
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.4604918890633176,
|
| 673 |
+
"grad_norm": 29.069836347407232,
|
| 674 |
+
"learning_rate": 3.269063392575352e-07,
|
| 675 |
+
"logits/chosen": -2.4375,
|
| 676 |
+
"logits/rejected": -2.171875,
|
| 677 |
+
"logps/chosen": -628.0,
|
| 678 |
+
"logps/rejected": -1424.0,
|
| 679 |
+
"loss": 0.1044,
|
| 680 |
+
"rewards/accuracies": 0.96875,
|
| 681 |
+
"rewards/chosen": -3.5625,
|
| 682 |
+
"rewards/margins": 7.9375,
|
| 683 |
+
"rewards/rejected": -11.5,
|
| 684 |
+
"step": 440
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.47095761381475665,
|
| 688 |
+
"grad_norm": 28.99798773679313,
|
| 689 |
+
"learning_rate": 3.1815705699316964e-07,
|
| 690 |
+
"logits/chosen": -2.328125,
|
| 691 |
+
"logits/rejected": -2.03125,
|
| 692 |
+
"logps/chosen": -700.0,
|
| 693 |
+
"logps/rejected": -1472.0,
|
| 694 |
+
"loss": 0.1061,
|
| 695 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 696 |
+
"rewards/chosen": -3.984375,
|
| 697 |
+
"rewards/margins": 7.8125,
|
| 698 |
+
"rewards/rejected": -11.8125,
|
| 699 |
+
"step": 450
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.48142333856619574,
|
| 703 |
+
"grad_norm": 20.545878743670112,
|
| 704 |
+
"learning_rate": 3.0931662070620794e-07,
|
| 705 |
+
"logits/chosen": -2.21875,
|
| 706 |
+
"logits/rejected": -2.03125,
|
| 707 |
+
"logps/chosen": -776.0,
|
| 708 |
+
"logps/rejected": -1464.0,
|
| 709 |
+
"loss": 0.097,
|
| 710 |
+
"rewards/accuracies": 0.96875,
|
| 711 |
+
"rewards/chosen": -5.03125,
|
| 712 |
+
"rewards/margins": 7.0625,
|
| 713 |
+
"rewards/rejected": -12.125,
|
| 714 |
+
"step": 460
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.49188906331763477,
|
| 718 |
+
"grad_norm": 22.557933835618037,
|
| 719 |
+
"learning_rate": 3.003968536966078e-07,
|
| 720 |
+
"logits/chosen": -2.34375,
|
| 721 |
+
"logits/rejected": -2.0625,
|
| 722 |
+
"logps/chosen": -760.0,
|
| 723 |
+
"logps/rejected": -1488.0,
|
| 724 |
+
"loss": 0.081,
|
| 725 |
+
"rewards/accuracies": 0.956250011920929,
|
| 726 |
+
"rewards/chosen": -4.46875,
|
| 727 |
+
"rewards/margins": 7.40625,
|
| 728 |
+
"rewards/rejected": -11.875,
|
| 729 |
+
"step": 470
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.5023547880690737,
|
| 733 |
+
"grad_norm": 28.131650321698604,
|
| 734 |
+
"learning_rate": 2.9140968536213693e-07,
|
| 735 |
+
"logits/chosen": -2.359375,
|
| 736 |
+
"logits/rejected": -1.9609375,
|
| 737 |
+
"logps/chosen": -840.0,
|
| 738 |
+
"logps/rejected": -1584.0,
|
| 739 |
+
"loss": 0.1038,
|
| 740 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 741 |
+
"rewards/chosen": -5.34375,
|
| 742 |
+
"rewards/margins": 7.75,
|
| 743 |
+
"rewards/rejected": -13.125,
|
| 744 |
+
"step": 480
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.5128205128205128,
|
| 748 |
+
"grad_norm": 22.231590737155383,
|
| 749 |
+
"learning_rate": 2.823671352438608e-07,
|
| 750 |
+
"logits/chosen": -2.4375,
|
| 751 |
+
"logits/rejected": -2.1875,
|
| 752 |
+
"logps/chosen": -788.0,
|
| 753 |
+
"logps/rejected": -1432.0,
|
| 754 |
+
"loss": 0.099,
|
| 755 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 756 |
+
"rewards/chosen": -5.0625,
|
| 757 |
+
"rewards/margins": 6.5625,
|
| 758 |
+
"rewards/rejected": -11.625,
|
| 759 |
+
"step": 490
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.5232862375719518,
|
| 763 |
+
"grad_norm": 32.66738645617045,
|
| 764 |
+
"learning_rate": 2.73281296951072e-07,
|
| 765 |
+
"logits/chosen": -2.421875,
|
| 766 |
+
"logits/rejected": -2.015625,
|
| 767 |
+
"logps/chosen": -880.0,
|
| 768 |
+
"logps/rejected": -1584.0,
|
| 769 |
+
"loss": 0.0906,
|
| 770 |
+
"rewards/accuracies": 0.96875,
|
| 771 |
+
"rewards/chosen": -5.84375,
|
| 772 |
+
"rewards/margins": 7.4375,
|
| 773 |
+
"rewards/rejected": -13.25,
|
| 774 |
+
"step": 500
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.533751962323391,
|
| 778 |
+
"grad_norm": 24.851145354651734,
|
| 779 |
+
"learning_rate": 2.641643219871597e-07,
|
| 780 |
+
"logits/chosen": -2.3125,
|
| 781 |
+
"logits/rejected": -1.9921875,
|
| 782 |
+
"logps/chosen": -868.0,
|
| 783 |
+
"logps/rejected": -1536.0,
|
| 784 |
+
"loss": 0.09,
|
| 785 |
+
"rewards/accuracies": 0.96875,
|
| 786 |
+
"rewards/chosen": -5.625,
|
| 787 |
+
"rewards/margins": 7.03125,
|
| 788 |
+
"rewards/rejected": -12.6875,
|
| 789 |
+
"step": 510
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.54421768707483,
|
| 793 |
+
"grad_norm": 30.824666380700695,
|
| 794 |
+
"learning_rate": 2.550284034980507e-07,
|
| 795 |
+
"logits/chosen": -2.375,
|
| 796 |
+
"logits/rejected": -2.171875,
|
| 797 |
+
"logps/chosen": -808.0,
|
| 798 |
+
"logps/rejected": -1504.0,
|
| 799 |
+
"loss": 0.11,
|
| 800 |
+
"rewards/accuracies": 0.96875,
|
| 801 |
+
"rewards/chosen": -5.0625,
|
| 802 |
+
"rewards/margins": 7.09375,
|
| 803 |
+
"rewards/rejected": -12.125,
|
| 804 |
+
"step": 520
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 0.554683411826269,
|
| 808 |
+
"grad_norm": 37.732613885821536,
|
| 809 |
+
"learning_rate": 2.4588575996495794e-07,
|
| 810 |
+
"logits/chosen": -2.578125,
|
| 811 |
+
"logits/rejected": -2.25,
|
| 812 |
+
"logps/chosen": -676.0,
|
| 813 |
+
"logps/rejected": -1352.0,
|
| 814 |
+
"loss": 0.1057,
|
| 815 |
+
"rewards/accuracies": 0.949999988079071,
|
| 816 |
+
"rewards/chosen": -4.0625,
|
| 817 |
+
"rewards/margins": 7.0,
|
| 818 |
+
"rewards/rejected": -11.0625,
|
| 819 |
+
"step": 530
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.565149136577708,
|
| 823 |
+
"grad_norm": 23.236138548824204,
|
| 824 |
+
"learning_rate": 2.367486188632446e-07,
|
| 825 |
+
"logits/chosen": -2.53125,
|
| 826 |
+
"logits/rejected": -2.25,
|
| 827 |
+
"logps/chosen": -704.0,
|
| 828 |
+
"logps/rejected": -1376.0,
|
| 829 |
+
"loss": 0.0968,
|
| 830 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 831 |
+
"rewards/chosen": -4.21875,
|
| 832 |
+
"rewards/margins": 6.6875,
|
| 833 |
+
"rewards/rejected": -10.875,
|
| 834 |
+
"step": 540
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 0.5756148613291471,
|
| 838 |
+
"grad_norm": 18.62210206857554,
|
| 839 |
+
"learning_rate": 2.276292003092593e-07,
|
| 840 |
+
"logits/chosen": -2.421875,
|
| 841 |
+
"logits/rejected": -2.109375,
|
| 842 |
+
"logps/chosen": -836.0,
|
| 843 |
+
"logps/rejected": -1544.0,
|
| 844 |
+
"loss": 0.0824,
|
| 845 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 846 |
+
"rewards/chosen": -5.5,
|
| 847 |
+
"rewards/margins": 7.28125,
|
| 848 |
+
"rewards/rejected": -12.75,
|
| 849 |
+
"step": 550
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.5860805860805861,
|
| 853 |
+
"grad_norm": 30.39956813704047,
|
| 854 |
+
"learning_rate": 2.185397007170141e-07,
|
| 855 |
+
"logits/chosen": -2.34375,
|
| 856 |
+
"logits/rejected": -1.9375,
|
| 857 |
+
"logps/chosen": -964.0,
|
| 858 |
+
"logps/rejected": -1720.0,
|
| 859 |
+
"loss": 0.0776,
|
| 860 |
+
"rewards/accuracies": 0.96875,
|
| 861 |
+
"rewards/chosen": -6.84375,
|
| 862 |
+
"rewards/margins": 8.0,
|
| 863 |
+
"rewards/rejected": -14.875,
|
| 864 |
+
"step": 560
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.5965463108320251,
|
| 868 |
+
"grad_norm": 31.863861823036746,
|
| 869 |
+
"learning_rate": 2.094922764865619e-07,
|
| 870 |
+
"logits/chosen": -2.265625,
|
| 871 |
+
"logits/rejected": -2.109375,
|
| 872 |
+
"logps/chosen": -932.0,
|
| 873 |
+
"logps/rejected": -1688.0,
|
| 874 |
+
"loss": 0.0778,
|
| 875 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 876 |
+
"rewards/chosen": -6.3125,
|
| 877 |
+
"rewards/margins": 7.78125,
|
| 878 |
+
"rewards/rejected": -14.125,
|
| 879 |
+
"step": 570
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.6070120355834642,
|
| 883 |
+
"grad_norm": 22.78774525718653,
|
| 884 |
+
"learning_rate": 2.0049902774588797e-07,
|
| 885 |
+
"logits/chosen": -2.453125,
|
| 886 |
+
"logits/rejected": -2.15625,
|
| 887 |
+
"logps/chosen": -776.0,
|
| 888 |
+
"logps/rejected": -1528.0,
|
| 889 |
+
"loss": 0.0791,
|
| 890 |
+
"rewards/accuracies": 0.956250011920929,
|
| 891 |
+
"rewards/chosen": -5.03125,
|
| 892 |
+
"rewards/margins": 7.65625,
|
| 893 |
+
"rewards/rejected": -12.6875,
|
| 894 |
+
"step": 580
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.6174777603349032,
|
| 898 |
+
"grad_norm": 24.950505604608463,
|
| 899 |
+
"learning_rate": 1.9157198216806238e-07,
|
| 900 |
+
"logits/chosen": -2.453125,
|
| 901 |
+
"logits/rejected": -2.09375,
|
| 902 |
+
"logps/chosen": -776.0,
|
| 903 |
+
"logps/rejected": -1616.0,
|
| 904 |
+
"loss": 0.0779,
|
| 905 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 906 |
+
"rewards/chosen": -5.03125,
|
| 907 |
+
"rewards/margins": 8.25,
|
| 908 |
+
"rewards/rejected": -13.3125,
|
| 909 |
+
"step": 590
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.6279434850863422,
|
| 913 |
+
"grad_norm": 39.70214338791234,
|
| 914 |
+
"learning_rate": 1.8272307888529274e-07,
|
| 915 |
+
"logits/chosen": -2.59375,
|
| 916 |
+
"logits/rejected": -2.375,
|
| 917 |
+
"logps/chosen": -824.0,
|
| 918 |
+
"logps/rejected": -1696.0,
|
| 919 |
+
"loss": 0.092,
|
| 920 |
+
"rewards/accuracies": 0.949999988079071,
|
| 921 |
+
"rewards/chosen": -5.125,
|
| 922 |
+
"rewards/margins": 8.5625,
|
| 923 |
+
"rewards/rejected": -13.6875,
|
| 924 |
+
"step": 600
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.6384092098377813,
|
| 928 |
+
"grad_norm": 24.678139590765365,
|
| 929 |
+
"learning_rate": 1.7396415252139288e-07,
|
| 930 |
+
"logits/chosen": -2.421875,
|
| 931 |
+
"logits/rejected": -2.109375,
|
| 932 |
+
"logps/chosen": -768.0,
|
| 933 |
+
"logps/rejected": -1544.0,
|
| 934 |
+
"loss": 0.0796,
|
| 935 |
+
"rewards/accuracies": 0.9375,
|
| 936 |
+
"rewards/chosen": -4.78125,
|
| 937 |
+
"rewards/margins": 7.9375,
|
| 938 |
+
"rewards/rejected": -12.75,
|
| 939 |
+
"step": 610
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.6488749345892203,
|
| 943 |
+
"grad_norm": 29.19695220969592,
|
| 944 |
+
"learning_rate": 1.6530691736402316e-07,
|
| 945 |
+
"logits/chosen": -2.484375,
|
| 946 |
+
"logits/rejected": -2.21875,
|
| 947 |
+
"logps/chosen": -768.0,
|
| 948 |
+
"logps/rejected": -1544.0,
|
| 949 |
+
"loss": 0.0724,
|
| 950 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 951 |
+
"rewards/chosen": -4.78125,
|
| 952 |
+
"rewards/margins": 7.9375,
|
| 953 |
+
"rewards/rejected": -12.6875,
|
| 954 |
+
"step": 620
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.6593406593406593,
|
| 958 |
+
"grad_norm": 17.517613098387617,
|
| 959 |
+
"learning_rate": 1.5676295169786864e-07,
|
| 960 |
+
"logits/chosen": -2.5,
|
| 961 |
+
"logits/rejected": -2.203125,
|
| 962 |
+
"logps/chosen": -828.0,
|
| 963 |
+
"logps/rejected": -1544.0,
|
| 964 |
+
"loss": 0.0929,
|
| 965 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 966 |
+
"rewards/chosen": -5.1875,
|
| 967 |
+
"rewards/margins": 7.5625,
|
| 968 |
+
"rewards/rejected": -12.75,
|
| 969 |
+
"step": 630
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.6698063840920984,
|
| 973 |
+
"grad_norm": 18.336302496269198,
|
| 974 |
+
"learning_rate": 1.483436823197092e-07,
|
| 975 |
+
"logits/chosen": -2.5,
|
| 976 |
+
"logits/rejected": -2.296875,
|
| 977 |
+
"logps/chosen": -804.0,
|
| 978 |
+
"logps/rejected": -1568.0,
|
| 979 |
+
"loss": 0.0861,
|
| 980 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 981 |
+
"rewards/chosen": -5.28125,
|
| 982 |
+
"rewards/margins": 7.75,
|
| 983 |
+
"rewards/rejected": -13.0625,
|
| 984 |
+
"step": 640
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.6802721088435374,
|
| 988 |
+
"grad_norm": 19.399010068279768,
|
| 989 |
+
"learning_rate": 1.4006036925609243e-07,
|
| 990 |
+
"logits/chosen": -2.484375,
|
| 991 |
+
"logits/rejected": -2.3125,
|
| 992 |
+
"logps/chosen": -888.0,
|
| 993 |
+
"logps/rejected": -1568.0,
|
| 994 |
+
"loss": 0.0706,
|
| 995 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 996 |
+
"rewards/chosen": -5.875,
|
| 997 |
+
"rewards/margins": 7.125,
|
| 998 |
+
"rewards/rejected": -13.0,
|
| 999 |
+
"step": 650
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.6907378335949764,
|
| 1003 |
+
"grad_norm": 24.671794614204327,
|
| 1004 |
+
"learning_rate": 1.319240907040458e-07,
|
| 1005 |
+
"logits/chosen": -2.4375,
|
| 1006 |
+
"logits/rejected": -2.15625,
|
| 1007 |
+
"logps/chosen": -844.0,
|
| 1008 |
+
"logps/rejected": -1616.0,
|
| 1009 |
+
"loss": 0.0735,
|
| 1010 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1011 |
+
"rewards/chosen": -5.4375,
|
| 1012 |
+
"rewards/margins": 7.6875,
|
| 1013 |
+
"rewards/rejected": -13.125,
|
| 1014 |
+
"step": 660
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 0.7012035583464155,
|
| 1018 |
+
"grad_norm": 21.656808943027293,
|
| 1019 |
+
"learning_rate": 1.2394572821496948e-07,
|
| 1020 |
+
"logits/chosen": -2.296875,
|
| 1021 |
+
"logits/rejected": -2.015625,
|
| 1022 |
+
"logps/chosen": -852.0,
|
| 1023 |
+
"logps/rejected": -1720.0,
|
| 1024 |
+
"loss": 0.0756,
|
| 1025 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1026 |
+
"rewards/chosen": -5.78125,
|
| 1027 |
+
"rewards/margins": 8.75,
|
| 1028 |
+
"rewards/rejected": -14.5625,
|
| 1029 |
+
"step": 670
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.7116692830978545,
|
| 1033 |
+
"grad_norm": 37.67983640549423,
|
| 1034 |
+
"learning_rate": 1.1613595214152711e-07,
|
| 1035 |
+
"logits/chosen": -2.34375,
|
| 1036 |
+
"logits/rejected": -2.046875,
|
| 1037 |
+
"logps/chosen": -944.0,
|
| 1038 |
+
"logps/rejected": -1800.0,
|
| 1039 |
+
"loss": 0.0717,
|
| 1040 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1041 |
+
"rewards/chosen": -6.375,
|
| 1042 |
+
"rewards/margins": 8.75,
|
| 1043 |
+
"rewards/rejected": -15.125,
|
| 1044 |
+
"step": 680
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 0.7221350078492935,
|
| 1048 |
+
"grad_norm": 31.204304577384395,
|
| 1049 |
+
"learning_rate": 1.0850520736699362e-07,
|
| 1050 |
+
"logits/chosen": -2.46875,
|
| 1051 |
+
"logits/rejected": -2.203125,
|
| 1052 |
+
"logps/chosen": -960.0,
|
| 1053 |
+
"logps/rejected": -1752.0,
|
| 1054 |
+
"loss": 0.0799,
|
| 1055 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1056 |
+
"rewards/chosen": -6.46875,
|
| 1057 |
+
"rewards/margins": 8.0625,
|
| 1058 |
+
"rewards/rejected": -14.5,
|
| 1059 |
+
"step": 690
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.7326007326007326,
|
| 1063 |
+
"grad_norm": 14.65690246648118,
|
| 1064 |
+
"learning_rate": 1.0106369933615042e-07,
|
| 1065 |
+
"logits/chosen": -2.375,
|
| 1066 |
+
"logits/rejected": -2.21875,
|
| 1067 |
+
"logps/chosen": -892.0,
|
| 1068 |
+
"logps/rejected": -1760.0,
|
| 1069 |
+
"loss": 0.066,
|
| 1070 |
+
"rewards/accuracies": 0.96875,
|
| 1071 |
+
"rewards/chosen": -5.84375,
|
| 1072 |
+
"rewards/margins": 8.6875,
|
| 1073 |
+
"rewards/rejected": -14.5,
|
| 1074 |
+
"step": 700
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.7430664573521716,
|
| 1078 |
+
"grad_norm": 10.13345133625024,
|
| 1079 |
+
"learning_rate": 9.382138040640714e-08,
|
| 1080 |
+
"logits/chosen": -2.515625,
|
| 1081 |
+
"logits/rejected": -2.15625,
|
| 1082 |
+
"logps/chosen": -832.0,
|
| 1083 |
+
"logps/rejected": -1752.0,
|
| 1084 |
+
"loss": 0.0628,
|
| 1085 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1086 |
+
"rewards/chosen": -5.46875,
|
| 1087 |
+
"rewards/margins": 9.0625,
|
| 1088 |
+
"rewards/rejected": -14.5,
|
| 1089 |
+
"step": 710
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.7535321821036107,
|
| 1093 |
+
"grad_norm": 26.423649093017993,
|
| 1094 |
+
"learning_rate": 8.678793653740632e-08,
|
| 1095 |
+
"logits/chosen": -2.484375,
|
| 1096 |
+
"logits/rejected": -2.1875,
|
| 1097 |
+
"logps/chosen": -956.0,
|
| 1098 |
+
"logps/rejected": -1744.0,
|
| 1099 |
+
"loss": 0.059,
|
| 1100 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1101 |
+
"rewards/chosen": -6.21875,
|
| 1102 |
+
"rewards/margins": 8.375,
|
| 1103 |
+
"rewards/rejected": -14.5625,
|
| 1104 |
+
"step": 720
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.7639979068550498,
|
| 1108 |
+
"grad_norm": 21.35868634461276,
|
| 1109 |
+
"learning_rate": 7.997277433690983e-08,
|
| 1110 |
+
"logits/chosen": -2.28125,
|
| 1111 |
+
"logits/rejected": -2.078125,
|
| 1112 |
+
"logps/chosen": -984.0,
|
| 1113 |
+
"logps/rejected": -1864.0,
|
| 1114 |
+
"loss": 0.0766,
|
| 1115 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1116 |
+
"rewards/chosen": -7.0625,
|
| 1117 |
+
"rewards/margins": 8.9375,
|
| 1118 |
+
"rewards/rejected": -16.0,
|
| 1119 |
+
"step": 730
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.7744636316064888,
|
| 1123 |
+
"grad_norm": 19.914720577430103,
|
| 1124 |
+
"learning_rate": 7.338500848029602e-08,
|
| 1125 |
+
"logits/chosen": -2.3125,
|
| 1126 |
+
"logits/rejected": -1.984375,
|
| 1127 |
+
"logps/chosen": -1020.0,
|
| 1128 |
+
"logps/rejected": -1976.0,
|
| 1129 |
+
"loss": 0.0593,
|
| 1130 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1131 |
+
"rewards/chosen": -7.21875,
|
| 1132 |
+
"rewards/margins": 9.75,
|
| 1133 |
+
"rewards/rejected": -17.0,
|
| 1134 |
+
"step": 740
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 0.7849293563579278,
|
| 1138 |
+
"grad_norm": 21.23435309177318,
|
| 1139 |
+
"learning_rate": 6.70334495204884e-08,
|
| 1140 |
+
"logits/chosen": -2.203125,
|
| 1141 |
+
"logits/rejected": -1.859375,
|
| 1142 |
+
"logps/chosen": -968.0,
|
| 1143 |
+
"logps/rejected": -1928.0,
|
| 1144 |
+
"loss": 0.048,
|
| 1145 |
+
"rewards/accuracies": 0.96875,
|
| 1146 |
+
"rewards/chosen": -7.0625,
|
| 1147 |
+
"rewards/margins": 9.75,
|
| 1148 |
+
"rewards/rejected": -16.75,
|
| 1149 |
+
"step": 750
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.7953950811093669,
|
| 1153 |
+
"grad_norm": 22.491426007618863,
|
| 1154 |
+
"learning_rate": 6.092659210462231e-08,
|
| 1155 |
+
"logits/chosen": -2.328125,
|
| 1156 |
+
"logits/rejected": -1.890625,
|
| 1157 |
+
"logps/chosen": -1020.0,
|
| 1158 |
+
"logps/rejected": -1992.0,
|
| 1159 |
+
"loss": 0.0566,
|
| 1160 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1161 |
+
"rewards/chosen": -7.4375,
|
| 1162 |
+
"rewards/margins": 9.9375,
|
| 1163 |
+
"rewards/rejected": -17.375,
|
| 1164 |
+
"step": 760
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.8058608058608059,
|
| 1168 |
+
"grad_norm": 7.849882679120514,
|
| 1169 |
+
"learning_rate": 5.507260361320737e-08,
|
| 1170 |
+
"logits/chosen": -2.203125,
|
| 1171 |
+
"logits/rejected": -1.8828125,
|
| 1172 |
+
"logps/chosen": -1012.0,
|
| 1173 |
+
"logps/rejected": -2064.0,
|
| 1174 |
+
"loss": 0.0432,
|
| 1175 |
+
"rewards/accuracies": 0.981249988079071,
|
| 1176 |
+
"rewards/chosen": -7.21875,
|
| 1177 |
+
"rewards/margins": 10.75,
|
| 1178 |
+
"rewards/rejected": -18.0,
|
| 1179 |
+
"step": 770
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.8163265306122449,
|
| 1183 |
+
"grad_norm": 16.662472392288155,
|
| 1184 |
+
"learning_rate": 4.947931323697982e-08,
|
| 1185 |
+
"logits/chosen": -2.3125,
|
| 1186 |
+
"logits/rejected": -1.9375,
|
| 1187 |
+
"logps/chosen": -1064.0,
|
| 1188 |
+
"logps/rejected": -2064.0,
|
| 1189 |
+
"loss": 0.0665,
|
| 1190 |
+
"rewards/accuracies": 0.96875,
|
| 1191 |
+
"rewards/chosen": -7.625,
|
| 1192 |
+
"rewards/margins": 10.375,
|
| 1193 |
+
"rewards/rejected": -18.0,
|
| 1194 |
+
"step": 780
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 0.826792255363684,
|
| 1198 |
+
"grad_norm": 18.85427926276285,
|
| 1199 |
+
"learning_rate": 4.415420150605398e-08,
|
| 1200 |
+
"logits/chosen": -2.1875,
|
| 1201 |
+
"logits/rejected": -1.90625,
|
| 1202 |
+
"logps/chosen": -932.0,
|
| 1203 |
+
"logps/rejected": -1920.0,
|
| 1204 |
+
"loss": 0.0661,
|
| 1205 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1206 |
+
"rewards/chosen": -6.65625,
|
| 1207 |
+
"rewards/margins": 9.8125,
|
| 1208 |
+
"rewards/rejected": -16.5,
|
| 1209 |
+
"step": 790
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.837257980115123,
|
| 1213 |
+
"grad_norm": 26.78574368890705,
|
| 1214 |
+
"learning_rate": 3.9104390285376374e-08,
|
| 1215 |
+
"logits/chosen": -2.3125,
|
| 1216 |
+
"logits/rejected": -2.0,
|
| 1217 |
+
"logps/chosen": -928.0,
|
| 1218 |
+
"logps/rejected": -1880.0,
|
| 1219 |
+
"loss": 0.053,
|
| 1220 |
+
"rewards/accuracies": 0.96875,
|
| 1221 |
+
"rewards/chosen": -6.40625,
|
| 1222 |
+
"rewards/margins": 9.4375,
|
| 1223 |
+
"rewards/rejected": -15.875,
|
| 1224 |
+
"step": 800
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.847723704866562,
|
| 1228 |
+
"grad_norm": 16.63949668329954,
|
| 1229 |
+
"learning_rate": 3.433663324986208e-08,
|
| 1230 |
+
"logits/chosen": -2.328125,
|
| 1231 |
+
"logits/rejected": -2.046875,
|
| 1232 |
+
"logps/chosen": -928.0,
|
| 1233 |
+
"logps/rejected": -1816.0,
|
| 1234 |
+
"loss": 0.0625,
|
| 1235 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1236 |
+
"rewards/chosen": -6.34375,
|
| 1237 |
+
"rewards/margins": 8.8125,
|
| 1238 |
+
"rewards/rejected": -15.125,
|
| 1239 |
+
"step": 810
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 0.858189429618001,
|
| 1243 |
+
"grad_norm": 32.084754893664645,
|
| 1244 |
+
"learning_rate": 2.9857306851953897e-08,
|
| 1245 |
+
"logits/chosen": -2.359375,
|
| 1246 |
+
"logits/rejected": -1.9140625,
|
| 1247 |
+
"logps/chosen": -964.0,
|
| 1248 |
+
"logps/rejected": -1888.0,
|
| 1249 |
+
"loss": 0.054,
|
| 1250 |
+
"rewards/accuracies": 0.96875,
|
| 1251 |
+
"rewards/chosen": -6.875,
|
| 1252 |
+
"rewards/margins": 9.4375,
|
| 1253 |
+
"rewards/rejected": -16.375,
|
| 1254 |
+
"step": 820
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 0.8686551543694401,
|
| 1258 |
+
"grad_norm": 19.89404053118834,
|
| 1259 |
+
"learning_rate": 2.567240179368185e-08,
|
| 1260 |
+
"logits/chosen": -2.34375,
|
| 1261 |
+
"logits/rejected": -1.78125,
|
| 1262 |
+
"logps/chosen": -1012.0,
|
| 1263 |
+
"logps/rejected": -1976.0,
|
| 1264 |
+
"loss": 0.0466,
|
| 1265 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1266 |
+
"rewards/chosen": -7.09375,
|
| 1267 |
+
"rewards/margins": 10.0625,
|
| 1268 |
+
"rewards/rejected": -17.125,
|
| 1269 |
+
"step": 830
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.8791208791208791,
|
| 1273 |
+
"grad_norm": 11.521868597924879,
|
| 1274 |
+
"learning_rate": 2.1787515014630357e-08,
|
| 1275 |
+
"logits/chosen": -2.1875,
|
| 1276 |
+
"logits/rejected": -1.890625,
|
| 1277 |
+
"logps/chosen": -984.0,
|
| 1278 |
+
"logps/rejected": -2080.0,
|
| 1279 |
+
"loss": 0.0533,
|
| 1280 |
+
"rewards/accuracies": 0.96875,
|
| 1281 |
+
"rewards/chosen": -7.21875,
|
| 1282 |
+
"rewards/margins": 10.75,
|
| 1283 |
+
"rewards/rejected": -18.0,
|
| 1284 |
+
"step": 840
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.8895866038723181,
|
| 1288 |
+
"grad_norm": 20.478869887407328,
|
| 1289 |
+
"learning_rate": 1.820784220652766e-08,
|
| 1290 |
+
"logits/chosen": -2.25,
|
| 1291 |
+
"logits/rejected": -1.9609375,
|
| 1292 |
+
"logps/chosen": -1024.0,
|
| 1293 |
+
"logps/rejected": -2040.0,
|
| 1294 |
+
"loss": 0.0396,
|
| 1295 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1296 |
+
"rewards/chosen": -7.1875,
|
| 1297 |
+
"rewards/margins": 10.25,
|
| 1298 |
+
"rewards/rejected": -17.5,
|
| 1299 |
+
"step": 850
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.9000523286237572,
|
| 1303 |
+
"grad_norm": 27.37618448209703,
|
| 1304 |
+
"learning_rate": 1.4938170864468636e-08,
|
| 1305 |
+
"logits/chosen": -2.265625,
|
| 1306 |
+
"logits/rejected": -1.78125,
|
| 1307 |
+
"logps/chosen": -984.0,
|
| 1308 |
+
"logps/rejected": -2096.0,
|
| 1309 |
+
"loss": 0.0529,
|
| 1310 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1311 |
+
"rewards/chosen": -7.1875,
|
| 1312 |
+
"rewards/margins": 11.125,
|
| 1313 |
+
"rewards/rejected": -18.375,
|
| 1314 |
+
"step": 860
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.9105180533751962,
|
| 1318 |
+
"grad_norm": 27.12820939331472,
|
| 1319 |
+
"learning_rate": 1.1982873884064465e-08,
|
| 1320 |
+
"logits/chosen": -2.3125,
|
| 1321 |
+
"logits/rejected": -1.90625,
|
| 1322 |
+
"logps/chosen": -1020.0,
|
| 1323 |
+
"logps/rejected": -1976.0,
|
| 1324 |
+
"loss": 0.0596,
|
| 1325 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1326 |
+
"rewards/chosen": -7.09375,
|
| 1327 |
+
"rewards/margins": 9.9375,
|
| 1328 |
+
"rewards/rejected": -17.0,
|
| 1329 |
+
"step": 870
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.9209837781266352,
|
| 1333 |
+
"grad_norm": 22.490629878443173,
|
| 1334 |
+
"learning_rate": 9.345903713082304e-09,
|
| 1335 |
+
"logits/chosen": -2.234375,
|
| 1336 |
+
"logits/rejected": -1.734375,
|
| 1337 |
+
"logps/chosen": -1088.0,
|
| 1338 |
+
"logps/rejected": -2112.0,
|
| 1339 |
+
"loss": 0.056,
|
| 1340 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1341 |
+
"rewards/chosen": -7.625,
|
| 1342 |
+
"rewards/margins": 10.375,
|
| 1343 |
+
"rewards/rejected": -18.0,
|
| 1344 |
+
"step": 880
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 0.9314495028780743,
|
| 1348 |
+
"grad_norm": 16.052382854562193,
|
| 1349 |
+
"learning_rate": 7.030787065396865e-09,
|
| 1350 |
+
"logits/chosen": -2.21875,
|
| 1351 |
+
"logits/rejected": -1.78125,
|
| 1352 |
+
"logps/chosen": -1004.0,
|
| 1353 |
+
"logps/rejected": -2112.0,
|
| 1354 |
+
"loss": 0.0556,
|
| 1355 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 1356 |
+
"rewards/chosen": -7.09375,
|
| 1357 |
+
"rewards/margins": 11.3125,
|
| 1358 |
+
"rewards/rejected": -18.375,
|
| 1359 |
+
"step": 890
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.9419152276295133,
|
| 1363 |
+
"grad_norm": 7.2739971268908326,
|
| 1364 |
+
"learning_rate": 5.04062020432286e-09,
|
| 1365 |
+
"logits/chosen": -2.390625,
|
| 1366 |
+
"logits/rejected": -1.8671875,
|
| 1367 |
+
"logps/chosen": -1048.0,
|
| 1368 |
+
"logps/rejected": -2080.0,
|
| 1369 |
+
"loss": 0.0627,
|
| 1370 |
+
"rewards/accuracies": 0.96875,
|
| 1371 |
+
"rewards/chosen": -7.34375,
|
| 1372 |
+
"rewards/margins": 10.5625,
|
| 1373 |
+
"rewards/rejected": -17.875,
|
| 1374 |
+
"step": 900
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.9523809523809523,
|
| 1378 |
+
"grad_norm": 29.919042800555964,
|
| 1379 |
+
"learning_rate": 3.3780648016376866e-09,
|
| 1380 |
+
"logits/chosen": -2.234375,
|
| 1381 |
+
"logits/rejected": -1.8984375,
|
| 1382 |
+
"logps/chosen": -1048.0,
|
| 1383 |
+
"logps/rejected": -2000.0,
|
| 1384 |
+
"loss": 0.0575,
|
| 1385 |
+
"rewards/accuracies": 0.96875,
|
| 1386 |
+
"rewards/chosen": -7.34375,
|
| 1387 |
+
"rewards/margins": 9.875,
|
| 1388 |
+
"rewards/rejected": -17.25,
|
| 1389 |
+
"step": 910
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.9628466771323915,
|
| 1393 |
+
"grad_norm": 50.810549611199676,
|
| 1394 |
+
"learning_rate": 2.0453443778310766e-09,
|
| 1395 |
+
"logits/chosen": -2.25,
|
| 1396 |
+
"logits/rejected": -1.8125,
|
| 1397 |
+
"logps/chosen": -992.0,
|
| 1398 |
+
"logps/rejected": -1984.0,
|
| 1399 |
+
"loss": 0.0484,
|
| 1400 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1401 |
+
"rewards/chosen": -7.09375,
|
| 1402 |
+
"rewards/margins": 10.0625,
|
| 1403 |
+
"rewards/rejected": -17.125,
|
| 1404 |
+
"step": 920
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 0.9733124018838305,
|
| 1408 |
+
"grad_norm": 20.362035094083513,
|
| 1409 |
+
"learning_rate": 1.0442413283435758e-09,
|
| 1410 |
+
"logits/chosen": -2.234375,
|
| 1411 |
+
"logits/rejected": -1.734375,
|
| 1412 |
+
"logps/chosen": -964.0,
|
| 1413 |
+
"logps/rejected": -1920.0,
|
| 1414 |
+
"loss": 0.0598,
|
| 1415 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1416 |
+
"rewards/chosen": -7.0625,
|
| 1417 |
+
"rewards/margins": 9.5625,
|
| 1418 |
+
"rewards/rejected": -16.625,
|
| 1419 |
+
"step": 930
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.9837781266352695,
|
| 1423 |
+
"grad_norm": 22.983688568193077,
|
| 1424 |
+
"learning_rate": 3.760945397705828e-10,
|
| 1425 |
+
"logits/chosen": -2.234375,
|
| 1426 |
+
"logits/rejected": -1.96875,
|
| 1427 |
+
"logps/chosen": -988.0,
|
| 1428 |
+
"logps/rejected": -2016.0,
|
| 1429 |
+
"loss": 0.0503,
|
| 1430 |
+
"rewards/accuracies": 0.981249988079071,
|
| 1431 |
+
"rewards/chosen": -6.8125,
|
| 1432 |
+
"rewards/margins": 10.3125,
|
| 1433 |
+
"rewards/rejected": -17.125,
|
| 1434 |
+
"step": 940
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.9942438513867086,
|
| 1438 |
+
"grad_norm": 18.89450295595607,
|
| 1439 |
+
"learning_rate": 4.17975992204056e-11,
|
| 1440 |
+
"logits/chosen": -2.21875,
|
| 1441 |
+
"logits/rejected": -1.828125,
|
| 1442 |
+
"logps/chosen": -1024.0,
|
| 1443 |
+
"logps/rejected": -2040.0,
|
| 1444 |
+
"loss": 0.0406,
|
| 1445 |
+
"rewards/accuracies": 0.981249988079071,
|
| 1446 |
+
"rewards/chosen": -7.125,
|
| 1447 |
+
"rewards/margins": 10.4375,
|
| 1448 |
+
"rewards/rejected": -17.5,
|
| 1449 |
+
"step": 950
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.9994767137624281,
|
| 1453 |
+
"step": 955,
|
| 1454 |
+
"total_flos": 0.0,
|
| 1455 |
+
"train_loss": 0.15591317970715268,
|
| 1456 |
+
"train_runtime": 13974.7976,
|
| 1457 |
+
"train_samples_per_second": 8.749,
|
| 1458 |
+
"train_steps_per_second": 0.068
|
| 1459 |
+
}
|
| 1460 |
+
],
|
| 1461 |
+
"logging_steps": 10,
|
| 1462 |
+
"max_steps": 955,
|
| 1463 |
+
"num_input_tokens_seen": 0,
|
| 1464 |
+
"num_train_epochs": 1,
|
| 1465 |
+
"save_steps": 100,
|
| 1466 |
+
"stateful_callbacks": {
|
| 1467 |
+
"TrainerControl": {
|
| 1468 |
+
"args": {
|
| 1469 |
+
"should_epoch_stop": false,
|
| 1470 |
+
"should_evaluate": false,
|
| 1471 |
+
"should_log": false,
|
| 1472 |
+
"should_save": true,
|
| 1473 |
+
"should_training_stop": true
|
| 1474 |
+
},
|
| 1475 |
+
"attributes": {}
|
| 1476 |
+
}
|
| 1477 |
+
},
|
| 1478 |
+
"total_flos": 0.0,
|
| 1479 |
+
"train_batch_size": 8,
|
| 1480 |
+
"trial_name": null,
|
| 1481 |
+
"trial_params": null
|
| 1482 |
+
}
|