Upload folder using huggingface_hub
Browse files- checkpoints-v2.0-discrete/checkpoint-23204/model.safetensors +3 -0
- checkpoints-v2.0-discrete/checkpoint-23204/optimizer.pt +3 -0
- checkpoints-v2.0-discrete/checkpoint-23204/rng_state.pth +3 -0
- checkpoints-v2.0-discrete/checkpoint-23204/scaler.pt +3 -0
- checkpoints-v2.0-discrete/checkpoint-23204/scheduler.pt +3 -0
- checkpoints-v2.0-discrete/checkpoint-23204/trainer_state.json +1082 -0
- checkpoints-v2.0-discrete/checkpoint-23204/training_args.bin +3 -0
checkpoints-v2.0-discrete/checkpoint-23204/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e9a99a2cc3ac739405337adb86bea0f2ec3c908de9c20834f890c72182d68a18
|
| 3 |
+
size 24002224
|
checkpoints-v2.0-discrete/checkpoint-23204/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f5e93966885dde5badcdfad6456cc1b6f79ee46eb337eaf7576c9cfd029cc74
|
| 3 |
+
size 1571083
|
checkpoints-v2.0-discrete/checkpoint-23204/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f687d8597eeb6dea9a8479f42ac16883adbc0c3221f718167c3102453df4e058
|
| 3 |
+
size 14645
|
checkpoints-v2.0-discrete/checkpoint-23204/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca3927fa7cca2ae02147edc5adb15c97db7bf65df00b5d32f8fa71c81459a8c4
|
| 3 |
+
size 1383
|
checkpoints-v2.0-discrete/checkpoint-23204/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fefc6a55210bbd7edbb09522c6f93b3b5fe4c8134063bfa343d2aada5de5d94f
|
| 3 |
+
size 1465
|
checkpoints-v2.0-discrete/checkpoint-23204/trainer_state.json
ADDED
|
@@ -0,0 +1,1082 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 23204,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011032818324821687,
|
| 14 |
+
"grad_norm": 0.11718960851430893,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 8.126412391662598,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.022065636649643373,
|
| 21 |
+
"grad_norm": 0.5918834209442139,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 7.521966934204102,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03309845497446506,
|
| 28 |
+
"grad_norm": 0.6679062247276306,
|
| 29 |
+
"learning_rate": 0.000999688448778502,
|
| 30 |
+
"loss": 7.341925144195557,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.04413127329928675,
|
| 35 |
+
"grad_norm": 1.1320008039474487,
|
| 36 |
+
"learning_rate": 0.0009987492950653055,
|
| 37 |
+
"loss": 7.3012261390686035,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.04413127329928675,
|
| 42 |
+
"eval_bleu": 0.4137140409513573,
|
| 43 |
+
"eval_loss": 7.362073738691904,
|
| 44 |
+
"eval_mse_loss": 1.3527455637449903,
|
| 45 |
+
"eval_recon_loss": 6.009328172150959,
|
| 46 |
+
"step": 1024
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.04413127329928675,
|
| 50 |
+
"eval_bleu": 0.4137140409513573,
|
| 51 |
+
"eval_loss": 7.362073738691904,
|
| 52 |
+
"eval_mse_loss": 1.3527455637449903,
|
| 53 |
+
"eval_recon_loss": 6.009328172150959,
|
| 54 |
+
"eval_runtime": 122.6264,
|
| 55 |
+
"eval_samples_per_second": 244.645,
|
| 56 |
+
"eval_steps_per_second": 3.825,
|
| 57 |
+
"step": 1024
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.05516409162410843,
|
| 61 |
+
"grad_norm": 0.583996593952179,
|
| 62 |
+
"learning_rate": 0.0009971837136430763,
|
| 63 |
+
"loss": 7.304327011108398,
|
| 64 |
+
"step": 1280
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.06619690994893011,
|
| 68 |
+
"grad_norm": 0.4174424707889557,
|
| 69 |
+
"learning_rate": 0.0009949936708776692,
|
| 70 |
+
"loss": 7.320664882659912,
|
| 71 |
+
"step": 1536
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 0.07722972827375181,
|
| 75 |
+
"grad_norm": 0.6613262295722961,
|
| 76 |
+
"learning_rate": 0.0009921819174566252,
|
| 77 |
+
"loss": 7.289535999298096,
|
| 78 |
+
"step": 1792
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 0.0882625465985735,
|
| 82 |
+
"grad_norm": 0.8233232498168945,
|
| 83 |
+
"learning_rate": 0.000988751984934317,
|
| 84 |
+
"loss": 7.30844259262085,
|
| 85 |
+
"step": 2048
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.0882625465985735,
|
| 89 |
+
"eval_bleu": 0.4137402227214601,
|
| 90 |
+
"eval_loss": 7.285478673255774,
|
| 91 |
+
"eval_mse_loss": 1.312072853543865,
|
| 92 |
+
"eval_recon_loss": 5.973405825812171,
|
| 93 |
+
"step": 2048
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.0882625465985735,
|
| 97 |
+
"eval_bleu": 0.4137402227214601,
|
| 98 |
+
"eval_loss": 7.285478673255774,
|
| 99 |
+
"eval_mse_loss": 1.312072853543865,
|
| 100 |
+
"eval_recon_loss": 5.973405825812171,
|
| 101 |
+
"eval_runtime": 120.626,
|
| 102 |
+
"eval_samples_per_second": 248.703,
|
| 103 |
+
"eval_steps_per_second": 3.888,
|
| 104 |
+
"step": 2048
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.09929536492339518,
|
| 108 |
+
"grad_norm": 0.6929961442947388,
|
| 109 |
+
"learning_rate": 0.0009847081812963268,
|
| 110 |
+
"loss": 7.224529266357422,
|
| 111 |
+
"step": 2304
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.11032818324821686,
|
| 115 |
+
"grad_norm": 0.4629116356372833,
|
| 116 |
+
"learning_rate": 0.0009800555855486275,
|
| 117 |
+
"loss": 7.270728588104248,
|
| 118 |
+
"step": 2560
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.12136100157303854,
|
| 122 |
+
"grad_norm": 0.8039212226867676,
|
| 123 |
+
"learning_rate": 0.0009748000413383664,
|
| 124 |
+
"loss": 7.244544982910156,
|
| 125 |
+
"step": 2816
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.13239381989786023,
|
| 129 |
+
"grad_norm": 0.7267772555351257,
|
| 130 |
+
"learning_rate": 0.0009689481496142604,
|
| 131 |
+
"loss": 7.242873191833496,
|
| 132 |
+
"step": 3072
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.13239381989786023,
|
| 136 |
+
"eval_bleu": 0.41751728447473074,
|
| 137 |
+
"eval_loss": 7.207233533676245,
|
| 138 |
+
"eval_mse_loss": 1.2975257886752392,
|
| 139 |
+
"eval_recon_loss": 5.909707753643044,
|
| 140 |
+
"step": 3072
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.13239381989786023,
|
| 144 |
+
"eval_bleu": 0.41751728447473074,
|
| 145 |
+
"eval_loss": 7.207233533676245,
|
| 146 |
+
"eval_mse_loss": 1.2975257886752392,
|
| 147 |
+
"eval_recon_loss": 5.909707753643044,
|
| 148 |
+
"eval_runtime": 120.6478,
|
| 149 |
+
"eval_samples_per_second": 248.658,
|
| 150 |
+
"eval_steps_per_second": 3.887,
|
| 151 |
+
"step": 3072
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.14342663822268192,
|
| 155 |
+
"grad_norm": 0.6957461833953857,
|
| 156 |
+
"learning_rate": 0.0009625072603358231,
|
| 157 |
+
"loss": 7.226602077484131,
|
| 158 |
+
"step": 3328
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.15445945654750362,
|
| 162 |
+
"grad_norm": 0.7835198044776917,
|
| 163 |
+
"learning_rate": 0.0009554854632418371,
|
| 164 |
+
"loss": 7.225058555603027,
|
| 165 |
+
"step": 3584
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.1654922748723253,
|
| 169 |
+
"grad_norm": 0.698896586894989,
|
| 170 |
+
"learning_rate": 0.000947891577689663,
|
| 171 |
+
"loss": 7.229842662811279,
|
| 172 |
+
"step": 3840
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.176525093197147,
|
| 176 |
+
"grad_norm": 0.5856732130050659,
|
| 177 |
+
"learning_rate": 0.0009397351415781539,
|
| 178 |
+
"loss": 7.255417346954346,
|
| 179 |
+
"step": 4096
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.176525093197147,
|
| 183 |
+
"eval_bleu": 0.4222208504795976,
|
| 184 |
+
"eval_loss": 7.213959849465376,
|
| 185 |
+
"eval_mse_loss": 1.284232896782442,
|
| 186 |
+
"eval_recon_loss": 5.929726949378626,
|
| 187 |
+
"step": 4096
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 0.176525093197147,
|
| 191 |
+
"eval_bleu": 0.4222208504795976,
|
| 192 |
+
"eval_loss": 7.213959849465376,
|
| 193 |
+
"eval_mse_loss": 1.284232896782442,
|
| 194 |
+
"eval_recon_loss": 5.929726949378626,
|
| 195 |
+
"eval_runtime": 120.9546,
|
| 196 |
+
"eval_samples_per_second": 248.027,
|
| 197 |
+
"eval_steps_per_second": 3.877,
|
| 198 |
+
"step": 4096
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.18755791152196866,
|
| 202 |
+
"grad_norm": 0.8607048392295837,
|
| 203 |
+
"learning_rate": 0.000931026399368079,
|
| 204 |
+
"loss": 7.224533557891846,
|
| 205 |
+
"step": 4352
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.19859072984679035,
|
| 209 |
+
"grad_norm": 0.6015220284461975,
|
| 210 |
+
"learning_rate": 0.0009217762892151117,
|
| 211 |
+
"loss": 7.171858310699463,
|
| 212 |
+
"step": 4608
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.20962354817161205,
|
| 216 |
+
"grad_norm": 0.5955944061279297,
|
| 217 |
+
"learning_rate": 0.0009119964292315354,
|
| 218 |
+
"loss": 7.180020809173584,
|
| 219 |
+
"step": 4864
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.22065636649643372,
|
| 223 |
+
"grad_norm": 0.9265109896659851,
|
| 224 |
+
"learning_rate": 0.0009016991028939279,
|
| 225 |
+
"loss": 7.237715244293213,
|
| 226 |
+
"step": 5120
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.22065636649643372,
|
| 230 |
+
"eval_bleu": 0.42211295717074476,
|
| 231 |
+
"eval_loss": 7.192519418720497,
|
| 232 |
+
"eval_mse_loss": 1.2717160906618847,
|
| 233 |
+
"eval_recon_loss": 5.920803332379632,
|
| 234 |
+
"step": 5120
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.22065636649643372,
|
| 238 |
+
"eval_bleu": 0.42211295717074476,
|
| 239 |
+
"eval_loss": 7.192519418720497,
|
| 240 |
+
"eval_mse_loss": 1.2717160906618847,
|
| 241 |
+
"eval_recon_loss": 5.920803332379632,
|
| 242 |
+
"eval_runtime": 123.6227,
|
| 243 |
+
"eval_samples_per_second": 242.674,
|
| 244 |
+
"eval_steps_per_second": 3.794,
|
| 245 |
+
"step": 5120
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.23168918482125542,
|
| 249 |
+
"grad_norm": 1.0009490251541138,
|
| 250 |
+
"learning_rate": 0.0008908972436151494,
|
| 251 |
+
"loss": 7.181396484375,
|
| 252 |
+
"step": 5376
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.2427220031460771,
|
| 256 |
+
"grad_norm": 0.5662420988082886,
|
| 257 |
+
"learning_rate": 0.0008796044185000127,
|
| 258 |
+
"loss": 7.214040756225586,
|
| 259 |
+
"step": 5632
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.2537548214708988,
|
| 263 |
+
"grad_norm": 0.5989573001861572,
|
| 264 |
+
"learning_rate": 0.0008678348113050368,
|
| 265 |
+
"loss": 7.233224391937256,
|
| 266 |
+
"step": 5888
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.26478763979572045,
|
| 270 |
+
"grad_norm": 0.6971510648727417,
|
| 271 |
+
"learning_rate": 0.0008556032046236897,
|
| 272 |
+
"loss": 7.238555431365967,
|
| 273 |
+
"step": 6144
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.26478763979572045,
|
| 277 |
+
"eval_bleu": 0.42175834100767645,
|
| 278 |
+
"eval_loss": 7.158817366496332,
|
| 279 |
+
"eval_mse_loss": 1.2668565204148607,
|
| 280 |
+
"eval_recon_loss": 5.891960855231865,
|
| 281 |
+
"step": 6144
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.26478763979572045,
|
| 285 |
+
"eval_bleu": 0.42175834100767645,
|
| 286 |
+
"eval_loss": 7.158817366496332,
|
| 287 |
+
"eval_mse_loss": 1.2668565204148607,
|
| 288 |
+
"eval_recon_loss": 5.891960855231865,
|
| 289 |
+
"eval_runtime": 122.7695,
|
| 290 |
+
"eval_samples_per_second": 244.36,
|
| 291 |
+
"eval_steps_per_second": 3.82,
|
| 292 |
+
"step": 6144
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.2758204581205422,
|
| 296 |
+
"grad_norm": 0.5552608966827393,
|
| 297 |
+
"learning_rate": 0.000842924961319492,
|
| 298 |
+
"loss": 7.156804084777832,
|
| 299 |
+
"step": 6400
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.28685327644536385,
|
| 303 |
+
"grad_norm": 0.5489348769187927,
|
| 304 |
+
"learning_rate": 0.0008298160052303045,
|
| 305 |
+
"loss": 7.185822486877441,
|
| 306 |
+
"step": 6656
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.2978860947701855,
|
| 310 |
+
"grad_norm": 0.6533068418502808,
|
| 311 |
+
"learning_rate": 0.0008162928011680314,
|
| 312 |
+
"loss": 7.207901477813721,
|
| 313 |
+
"step": 6912
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.30891891309500724,
|
| 317 |
+
"grad_norm": 0.6510070562362671,
|
| 318 |
+
"learning_rate": 0.000802372334238864,
|
| 319 |
+
"loss": 7.252415180206299,
|
| 320 |
+
"step": 7168
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.30891891309500724,
|
| 324 |
+
"eval_bleu": 0.42033993073908893,
|
| 325 |
+
"eval_loss": 7.211322486527693,
|
| 326 |
+
"eval_mse_loss": 1.262001540106751,
|
| 327 |
+
"eval_recon_loss": 5.949320952267026,
|
| 328 |
+
"step": 7168
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.30891891309500724,
|
| 332 |
+
"eval_bleu": 0.42033993073908893,
|
| 333 |
+
"eval_loss": 7.211322486527693,
|
| 334 |
+
"eval_mse_loss": 1.262001540106751,
|
| 335 |
+
"eval_recon_loss": 5.949320952267026,
|
| 336 |
+
"eval_runtime": 122.2468,
|
| 337 |
+
"eval_samples_per_second": 245.405,
|
| 338 |
+
"eval_steps_per_second": 3.837,
|
| 339 |
+
"step": 7168
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3199517314198289,
|
| 343 |
+
"grad_norm": 0.629555881023407,
|
| 344 |
+
"learning_rate": 0.0007880720885100349,
|
| 345 |
+
"loss": 7.190337181091309,
|
| 346 |
+
"step": 7424
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.3309845497446506,
|
| 350 |
+
"grad_norm": 0.7832825183868408,
|
| 351 |
+
"learning_rate": 0.0007734100250498788,
|
| 352 |
+
"loss": 7.169421672821045,
|
| 353 |
+
"step": 7680
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.3420173680694723,
|
| 357 |
+
"grad_norm": 1.125791072845459,
|
| 358 |
+
"learning_rate": 0.000758404559368781,
|
| 359 |
+
"loss": 7.217397212982178,
|
| 360 |
+
"step": 7936
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.353050186394294,
|
| 364 |
+
"grad_norm": 1.4169881343841553,
|
| 365 |
+
"learning_rate": 0.0007430745382893488,
|
| 366 |
+
"loss": 7.163691520690918,
|
| 367 |
+
"step": 8192
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.353050186394294,
|
| 371 |
+
"eval_bleu": 0.4273075759012065,
|
| 372 |
+
"eval_loss": 7.0710403741295655,
|
| 373 |
+
"eval_mse_loss": 1.2539655610696594,
|
| 374 |
+
"eval_recon_loss": 5.817074826022964,
|
| 375 |
+
"step": 8192
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.353050186394294,
|
| 379 |
+
"eval_bleu": 0.4273075759012065,
|
| 380 |
+
"eval_loss": 7.0710403741295655,
|
| 381 |
+
"eval_mse_loss": 1.2539655610696594,
|
| 382 |
+
"eval_recon_loss": 5.817074826022964,
|
| 383 |
+
"eval_runtime": 119.9191,
|
| 384 |
+
"eval_samples_per_second": 250.169,
|
| 385 |
+
"eval_steps_per_second": 3.911,
|
| 386 |
+
"step": 8192
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.36408300471911564,
|
| 390 |
+
"grad_norm": 0.6753491759300232,
|
| 391 |
+
"learning_rate": 0.0007274392162748551,
|
| 392 |
+
"loss": 7.200345516204834,
|
| 393 |
+
"step": 8448
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.3751158230439373,
|
| 397 |
+
"grad_norm": 0.5481099486351013,
|
| 398 |
+
"learning_rate": 0.000711518231245687,
|
| 399 |
+
"loss": 7.261969566345215,
|
| 400 |
+
"step": 8704
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.38614864136875904,
|
| 404 |
+
"grad_norm": 0.9790687561035156,
|
| 405 |
+
"learning_rate": 0.0006953315799141723,
|
| 406 |
+
"loss": 7.222372055053711,
|
| 407 |
+
"step": 8960
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.3971814596935807,
|
| 411 |
+
"grad_norm": 0.6765880584716797,
|
| 412 |
+
"learning_rate": 0.0006788995926687669,
|
| 413 |
+
"loss": 7.16965913772583,
|
| 414 |
+
"step": 9216
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.3971814596935807,
|
| 418 |
+
"eval_bleu": 0.4247464090003691,
|
| 419 |
+
"eval_loss": 7.120412382235659,
|
| 420 |
+
"eval_mse_loss": 1.2528218912925801,
|
| 421 |
+
"eval_recon_loss": 5.867590499076762,
|
| 422 |
+
"step": 9216
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.3971814596935807,
|
| 426 |
+
"eval_bleu": 0.4247464090003691,
|
| 427 |
+
"eval_loss": 7.120412382235659,
|
| 428 |
+
"eval_mse_loss": 1.2528218912925801,
|
| 429 |
+
"eval_recon_loss": 5.867590499076762,
|
| 430 |
+
"eval_runtime": 122.0364,
|
| 431 |
+
"eval_samples_per_second": 245.828,
|
| 432 |
+
"eval_steps_per_second": 3.843,
|
| 433 |
+
"step": 9216
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.4082142780184024,
|
| 437 |
+
"grad_norm": 0.65935879945755,
|
| 438 |
+
"learning_rate": 0.0006622429080391422,
|
| 439 |
+
"loss": 7.197686195373535,
|
| 440 |
+
"step": 9472
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.4192470963432241,
|
| 444 |
+
"grad_norm": 0.6421726942062378,
|
| 445 |
+
"learning_rate": 0.0006453824467742515,
|
| 446 |
+
"loss": 7.234926700592041,
|
| 447 |
+
"step": 9728
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.43027991466804577,
|
| 451 |
+
"grad_norm": 0.5790821313858032,
|
| 452 |
+
"learning_rate": 0.0006283393855659275,
|
| 453 |
+
"loss": 7.189432621002197,
|
| 454 |
+
"step": 9984
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 0.44131273299286744,
|
| 458 |
+
"grad_norm": 1.3850535154342651,
|
| 459 |
+
"learning_rate": 0.0006111351304510173,
|
| 460 |
+
"loss": 7.234661102294922,
|
| 461 |
+
"step": 10240
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.44131273299286744,
|
| 465 |
+
"eval_bleu": 0.418702569877124,
|
| 466 |
+
"eval_loss": 7.200763400429602,
|
| 467 |
+
"eval_mse_loss": 1.2514713542547815,
|
| 468 |
+
"eval_recon_loss": 5.949292033465941,
|
| 469 |
+
"step": 10240
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.44131273299286744,
|
| 473 |
+
"eval_bleu": 0.418702569877124,
|
| 474 |
+
"eval_loss": 7.200763400429602,
|
| 475 |
+
"eval_mse_loss": 1.2514713542547815,
|
| 476 |
+
"eval_recon_loss": 5.949292033465941,
|
| 477 |
+
"eval_runtime": 120.1297,
|
| 478 |
+
"eval_samples_per_second": 249.73,
|
| 479 |
+
"eval_steps_per_second": 3.904,
|
| 480 |
+
"step": 10240
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.45234555131768917,
|
| 484 |
+
"grad_norm": 0.7837287187576294,
|
| 485 |
+
"learning_rate": 0.0005937912899254605,
|
| 486 |
+
"loss": 7.249295711517334,
|
| 487 |
+
"step": 10496
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.46337836964251083,
|
| 491 |
+
"grad_norm": 0.6780616641044617,
|
| 492 |
+
"learning_rate": 0.0005763296478040787,
|
| 493 |
+
"loss": 7.172219276428223,
|
| 494 |
+
"step": 10752
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.4744111879673325,
|
| 498 |
+
"grad_norm": 0.7398590445518494,
|
| 499 |
+
"learning_rate": 0.0005587721358601663,
|
| 500 |
+
"loss": 7.186757564544678,
|
| 501 |
+
"step": 11008
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4854440062921542,
|
| 505 |
+
"grad_norm": 0.9151716232299805,
|
| 506 |
+
"learning_rate": 0.0005411408062792448,
|
| 507 |
+
"loss": 7.159968852996826,
|
| 508 |
+
"step": 11264
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.4854440062921542,
|
| 512 |
+
"eval_bleu": 0.4226532958201888,
|
| 513 |
+
"eval_loss": 7.1653254210059325,
|
| 514 |
+
"eval_mse_loss": 1.2460996743712598,
|
| 515 |
+
"eval_recon_loss": 5.9192257395177,
|
| 516 |
+
"step": 11264
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.4854440062921542,
|
| 520 |
+
"eval_bleu": 0.4226532958201888,
|
| 521 |
+
"eval_loss": 7.1653254210059325,
|
| 522 |
+
"eval_mse_loss": 1.2460996743712598,
|
| 523 |
+
"eval_recon_loss": 5.9192257395177,
|
| 524 |
+
"eval_runtime": 121.8437,
|
| 525 |
+
"eval_samples_per_second": 246.217,
|
| 526 |
+
"eval_steps_per_second": 3.849,
|
| 527 |
+
"step": 11264
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.4964768246169759,
|
| 531 |
+
"grad_norm": 0.8292572498321533,
|
| 532 |
+
"learning_rate": 0.0005234578039615789,
|
| 533 |
+
"loss": 7.178791046142578,
|
| 534 |
+
"step": 11520
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5075096429417976,
|
| 538 |
+
"grad_norm": 0.6656509637832642,
|
| 539 |
+
"learning_rate": 0.0005057453387082458,
|
| 540 |
+
"loss": 7.218838691711426,
|
| 541 |
+
"step": 11776
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.5185424612666193,
|
| 545 |
+
"grad_norm": 0.5986668467521667,
|
| 546 |
+
"learning_rate": 0.0004880256573256866,
|
| 547 |
+
"loss": 7.1541748046875,
|
| 548 |
+
"step": 12032
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.5295752795914409,
|
| 552 |
+
"grad_norm": 0.9848335385322571,
|
| 553 |
+
"learning_rate": 0.0004703210156837805,
|
| 554 |
+
"loss": 7.192195415496826,
|
| 555 |
+
"step": 12288
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.5295752795914409,
|
| 559 |
+
"eval_bleu": 0.4265346559703667,
|
| 560 |
+
"eval_loss": 7.1317149693015285,
|
| 561 |
+
"eval_mse_loss": 1.2411813456366565,
|
| 562 |
+
"eval_recon_loss": 5.890533620106386,
|
| 563 |
+
"step": 12288
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.5295752795914409,
|
| 567 |
+
"eval_bleu": 0.4265346559703667,
|
| 568 |
+
"eval_loss": 7.1317149693015285,
|
| 569 |
+
"eval_mse_loss": 1.2411813456366565,
|
| 570 |
+
"eval_recon_loss": 5.890533620106386,
|
| 571 |
+
"eval_runtime": 121.6345,
|
| 572 |
+
"eval_samples_per_second": 246.641,
|
| 573 |
+
"eval_steps_per_second": 3.856,
|
| 574 |
+
"step": 12288
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 0.5406080979162626,
|
| 578 |
+
"grad_norm": 0.4458930194377899,
|
| 579 |
+
"learning_rate": 0.0004526536507625343,
|
| 580 |
+
"loss": 7.163888931274414,
|
| 581 |
+
"step": 12544
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.5516409162410844,
|
| 585 |
+
"grad_norm": 0.6304380893707275,
|
| 586 |
+
"learning_rate": 0.00043504575272249973,
|
| 587 |
+
"loss": 7.172398567199707,
|
| 588 |
+
"step": 12800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 0.562673734565906,
|
| 592 |
+
"grad_norm": 0.4299321174621582,
|
| 593 |
+
"learning_rate": 0.0004175194370339921,
|
| 594 |
+
"loss": 7.1996941566467285,
|
| 595 |
+
"step": 13056
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.5737065528907277,
|
| 599 |
+
"grad_norm": 0.7110609412193298,
|
| 600 |
+
"learning_rate": 0.0004000967167001243,
|
| 601 |
+
"loss": 7.176135540008545,
|
| 602 |
+
"step": 13312
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.5737065528907277,
|
| 606 |
+
"eval_bleu": 0.4210106981552223,
|
| 607 |
+
"eval_loss": 7.17737479606417,
|
| 608 |
+
"eval_mse_loss": 1.24177454910807,
|
| 609 |
+
"eval_recon_loss": 5.9356002522938285,
|
| 610 |
+
"step": 13312
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.5737065528907277,
|
| 614 |
+
"eval_bleu": 0.4210106981552223,
|
| 615 |
+
"eval_loss": 7.17737479606417,
|
| 616 |
+
"eval_mse_loss": 1.24177454910807,
|
| 617 |
+
"eval_recon_loss": 5.9356002522938285,
|
| 618 |
+
"eval_runtime": 121.837,
|
| 619 |
+
"eval_samples_per_second": 246.231,
|
| 620 |
+
"eval_steps_per_second": 3.849,
|
| 621 |
+
"step": 13312
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5847393712155494,
|
| 625 |
+
"grad_norm": 0.6122337579727173,
|
| 626 |
+
"learning_rate": 0.00038279947460853446,
|
| 627 |
+
"loss": 7.161331653594971,
|
| 628 |
+
"step": 13568
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 0.595772189540371,
|
| 632 |
+
"grad_norm": 0.3825603723526001,
|
| 633 |
+
"learning_rate": 0.00036564943604654345,
|
| 634 |
+
"loss": 7.182938575744629,
|
| 635 |
+
"step": 13824
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 0.6068050078651928,
|
| 639 |
+
"grad_norm": 0.789337158203125,
|
| 640 |
+
"learning_rate": 0.00034866814141425254,
|
| 641 |
+
"loss": 7.209775447845459,
|
| 642 |
+
"step": 14080
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.6178378261900145,
|
| 646 |
+
"grad_norm": 0.5527840256690979,
|
| 647 |
+
"learning_rate": 0.0003318769191698637,
|
| 648 |
+
"loss": 7.182364463806152,
|
| 649 |
+
"step": 14336
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.6178378261900145,
|
| 653 |
+
"eval_bleu": 0.4225600590459926,
|
| 654 |
+
"eval_loss": 7.143745667390478,
|
| 655 |
+
"eval_mse_loss": 1.2410022706619457,
|
| 656 |
+
"eval_recon_loss": 5.902743396474354,
|
| 657 |
+
"step": 14336
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.6178378261900145,
|
| 661 |
+
"eval_bleu": 0.4225600590459926,
|
| 662 |
+
"eval_loss": 7.143745667390478,
|
| 663 |
+
"eval_mse_loss": 1.2410022706619457,
|
| 664 |
+
"eval_recon_loss": 5.902743396474354,
|
| 665 |
+
"eval_runtime": 120.9077,
|
| 666 |
+
"eval_samples_per_second": 248.123,
|
| 667 |
+
"eval_steps_per_second": 3.879,
|
| 668 |
+
"step": 14336
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.6288706445148361,
|
| 672 |
+
"grad_norm": 1.0367250442504883,
|
| 673 |
+
"learning_rate": 0.00031529685904119485,
|
| 674 |
+
"loss": 7.168930530548096,
|
| 675 |
+
"step": 14592
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.6399034628396578,
|
| 679 |
+
"grad_norm": 0.6979912519454956,
|
| 680 |
+
"learning_rate": 0.0002989487855370421,
|
| 681 |
+
"loss": 7.185887813568115,
|
| 682 |
+
"step": 14848
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.6509362811644795,
|
| 686 |
+
"grad_norm": 0.9760459661483765,
|
| 687 |
+
"learning_rate": 0.00028285323179165424,
|
| 688 |
+
"loss": 7.213273048400879,
|
| 689 |
+
"step": 15104
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.6619690994893012,
|
| 693 |
+
"grad_norm": 0.5782520771026611,
|
| 694 |
+
"learning_rate": 0.0002670304137751759,
|
| 695 |
+
"loss": 7.2389631271362305,
|
| 696 |
+
"step": 15360
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.6619690994893012,
|
| 700 |
+
"eval_bleu": 0.4214572010245335,
|
| 701 |
+
"eval_loss": 7.171980977820944,
|
| 702 |
+
"eval_mse_loss": 1.238288759422709,
|
| 703 |
+
"eval_recon_loss": 5.933692220431655,
|
| 704 |
+
"step": 15360
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.6619690994893012,
|
| 708 |
+
"eval_bleu": 0.4214572010245335,
|
| 709 |
+
"eval_loss": 7.171980977820944,
|
| 710 |
+
"eval_mse_loss": 1.238288759422709,
|
| 711 |
+
"eval_recon_loss": 5.933692220431655,
|
| 712 |
+
"eval_runtime": 122.1262,
|
| 713 |
+
"eval_samples_per_second": 245.648,
|
| 714 |
+
"eval_steps_per_second": 3.84,
|
| 715 |
+
"step": 15360
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 0.6730019178141229,
|
| 719 |
+
"grad_norm": 0.458218514919281,
|
| 720 |
+
"learning_rate": 0.0002515002049024435,
|
| 721 |
+
"loss": 7.234853744506836,
|
| 722 |
+
"step": 15616
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.6840347361389446,
|
| 726 |
+
"grad_norm": 0.6497246026992798,
|
| 727 |
+
"learning_rate": 0.00023628211107203429,
|
| 728 |
+
"loss": 7.127511978149414,
|
| 729 |
+
"step": 15872
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.6950675544637662,
|
| 733 |
+
"grad_norm": 0.47331640124320984,
|
| 734 |
+
"learning_rate": 0.00022139524616691188,
|
| 735 |
+
"loss": 7.157367706298828,
|
| 736 |
+
"step": 16128
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 0.706100372788588,
|
| 740 |
+
"grad_norm": 0.5251054167747498,
|
| 741 |
+
"learning_rate": 0.000206858308047443,
|
| 742 |
+
"loss": 7.214804649353027,
|
| 743 |
+
"step": 16384
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 0.706100372788588,
|
| 747 |
+
"eval_bleu": 0.4238607122458957,
|
| 748 |
+
"eval_loss": 7.129481719246805,
|
| 749 |
+
"eval_mse_loss": 1.2356282551405526,
|
| 750 |
+
"eval_recon_loss": 5.893853451397373,
|
| 751 |
+
"step": 16384
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.706100372788588,
|
| 755 |
+
"eval_bleu": 0.4238607122458957,
|
| 756 |
+
"eval_loss": 7.129481719246805,
|
| 757 |
+
"eval_mse_loss": 1.2356282551405526,
|
| 758 |
+
"eval_recon_loss": 5.893853451397373,
|
| 759 |
+
"eval_runtime": 122.2268,
|
| 760 |
+
"eval_samples_per_second": 245.445,
|
| 761 |
+
"eval_steps_per_second": 3.837,
|
| 762 |
+
"step": 16384
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.7171331911134096,
|
| 766 |
+
"grad_norm": 1.383571982383728,
|
| 767 |
+
"learning_rate": 0.00019268955506693798,
|
| 768 |
+
"loss": 7.159656524658203,
|
| 769 |
+
"step": 16640
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.7281660094382313,
|
| 773 |
+
"grad_norm": 0.8240185976028442,
|
| 774 |
+
"learning_rate": 0.00017890678313921,
|
| 775 |
+
"loss": 7.187053203582764,
|
| 776 |
+
"step": 16896
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.739198827763053,
|
| 780 |
+
"grad_norm": 0.9478136301040649,
|
| 781 |
+
"learning_rate": 0.00016552730338695792,
|
| 782 |
+
"loss": 7.169220447540283,
|
| 783 |
+
"step": 17152
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.7502316460878746,
|
| 787 |
+
"grad_norm": 1.9408297538757324,
|
| 788 |
+
"learning_rate": 0.00015256792039904465,
|
| 789 |
+
"loss": 7.1799798011779785,
|
| 790 |
+
"step": 17408
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 0.7502316460878746,
|
| 794 |
+
"eval_bleu": 0.4214146319298905,
|
| 795 |
+
"eval_loss": 7.153863178895735,
|
| 796 |
+
"eval_mse_loss": 1.2358840449786643,
|
| 797 |
+
"eval_recon_loss": 5.917979137221379,
|
| 798 |
+
"step": 17408
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 0.7502316460878746,
|
| 802 |
+
"eval_bleu": 0.4214146319298905,
|
| 803 |
+
"eval_loss": 7.153863178895735,
|
| 804 |
+
"eval_mse_loss": 1.2358840449786643,
|
| 805 |
+
"eval_recon_loss": 5.917979137221379,
|
| 806 |
+
"eval_runtime": 121.0512,
|
| 807 |
+
"eval_samples_per_second": 247.829,
|
| 808 |
+
"eval_steps_per_second": 3.874,
|
| 809 |
+
"step": 17408
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.7612644644126964,
|
| 813 |
+
"grad_norm": 0.42684435844421387,
|
| 814 |
+
"learning_rate": 0.00014004491112398103,
|
| 815 |
+
"loss": 7.144793510437012,
|
| 816 |
+
"step": 17664
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.7722972827375181,
|
| 820 |
+
"grad_norm": 1.1356747150421143,
|
| 821 |
+
"learning_rate": 0.00012797400442612433,
|
| 822 |
+
"loss": 7.137683868408203,
|
| 823 |
+
"step": 17920
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.7833301010623397,
|
| 827 |
+
"grad_norm": 0.804119884967804,
|
| 828 |
+
"learning_rate": 0.00011637036133026895,
|
| 829 |
+
"loss": 7.187657833099365,
|
| 830 |
+
"step": 18176
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.7943629193871614,
|
| 834 |
+
"grad_norm": 0.5904266834259033,
|
| 835 |
+
"learning_rate": 0.00010524855597944216,
|
| 836 |
+
"loss": 7.176294326782227,
|
| 837 |
+
"step": 18432
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.7943629193871614,
|
| 841 |
+
"eval_bleu": 0.42259771720418604,
|
| 842 |
+
"eval_loss": 7.143732512175148,
|
| 843 |
+
"eval_mse_loss": 1.2345045111072597,
|
| 844 |
+
"eval_recon_loss": 5.909227997509401,
|
| 845 |
+
"step": 18432
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.7943629193871614,
|
| 849 |
+
"eval_bleu": 0.42259771720418604,
|
| 850 |
+
"eval_loss": 7.143732512175148,
|
| 851 |
+
"eval_mse_loss": 1.2345045111072597,
|
| 852 |
+
"eval_recon_loss": 5.909227997509401,
|
| 853 |
+
"eval_runtime": 120.9093,
|
| 854 |
+
"eval_samples_per_second": 248.12,
|
| 855 |
+
"eval_steps_per_second": 3.879,
|
| 856 |
+
"step": 18432
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.8053957377119831,
|
| 860 |
+
"grad_norm": 0.7857964634895325,
|
| 861 |
+
"learning_rate": 9.462255732982089e-05,
|
| 862 |
+
"loss": 7.177124977111816,
|
| 863 |
+
"step": 18688
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.8164285560368048,
|
| 867 |
+
"grad_norm": 0.5492353439331055,
|
| 868 |
+
"learning_rate": 8.450571160576348e-05,
|
| 869 |
+
"loss": 7.213504791259766,
|
| 870 |
+
"step": 18944
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.8274613743616265,
|
| 874 |
+
"grad_norm": 1.4939922094345093,
|
| 875 |
+
"learning_rate": 7.491072553698764e-05,
|
| 876 |
+
"loss": 7.130130767822266,
|
| 877 |
+
"step": 19200
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.8384941926864482,
|
| 881 |
+
"grad_norm": 0.6733608841896057,
|
| 882 |
+
"learning_rate": 6.584965039895586e-05,
|
| 883 |
+
"loss": 7.213354587554932,
|
| 884 |
+
"step": 19456
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.8384941926864482,
|
| 888 |
+
"eval_bleu": 0.42425975662031007,
|
| 889 |
+
"eval_loss": 7.170489285800503,
|
| 890 |
+
"eval_mse_loss": 1.2338256403835597,
|
| 891 |
+
"eval_recon_loss": 5.936663639825036,
|
| 892 |
+
"step": 19456
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.8384941926864482,
|
| 896 |
+
"eval_bleu": 0.42425975662031007,
|
| 897 |
+
"eval_loss": 7.170489285800503,
|
| 898 |
+
"eval_mse_loss": 1.2338256403835597,
|
| 899 |
+
"eval_recon_loss": 5.936663639825036,
|
| 900 |
+
"eval_runtime": 120.4436,
|
| 901 |
+
"eval_samples_per_second": 249.079,
|
| 902 |
+
"eval_steps_per_second": 3.894,
|
| 903 |
+
"step": 19456
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.8495270110112698,
|
| 907 |
+
"grad_norm": 0.6659030318260193,
|
| 908 |
+
"learning_rate": 5.73338668765051e-05,
|
| 909 |
+
"loss": 7.2163591384887695,
|
| 910 |
+
"step": 19712
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.8605598293360915,
|
| 914 |
+
"grad_norm": 0.5635488033294678,
|
| 915 |
+
"learning_rate": 4.9374070769740984e-05,
|
| 916 |
+
"loss": 7.225249290466309,
|
| 917 |
+
"step": 19968
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.8715926476609133,
|
| 921 |
+
"grad_norm": 1.0609259605407715,
|
| 922 |
+
"learning_rate": 4.198025956014095e-05,
|
| 923 |
+
"loss": 7.194465637207031,
|
| 924 |
+
"step": 20224
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.8826254659857349,
|
| 928 |
+
"grad_norm": 1.3461576700210571,
|
| 929 |
+
"learning_rate": 3.516171985374755e-05,
|
| 930 |
+
"loss": 7.195147514343262,
|
| 931 |
+
"step": 20480
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 0.8826254659857349,
|
| 935 |
+
"eval_bleu": 0.42110197936728205,
|
| 936 |
+
"eval_loss": 7.151369848231009,
|
| 937 |
+
"eval_mse_loss": 1.233785225892626,
|
| 938 |
+
"eval_recon_loss": 5.917584622084205,
|
| 939 |
+
"step": 20480
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.8826254659857349,
|
| 943 |
+
"eval_bleu": 0.42110197936728205,
|
| 944 |
+
"eval_loss": 7.151369848231009,
|
| 945 |
+
"eval_mse_loss": 1.233785225892626,
|
| 946 |
+
"eval_recon_loss": 5.917584622084205,
|
| 947 |
+
"eval_runtime": 120.8036,
|
| 948 |
+
"eval_samples_per_second": 248.337,
|
| 949 |
+
"eval_steps_per_second": 3.882,
|
| 950 |
+
"step": 20480
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8936582843105566,
|
| 954 |
+
"grad_norm": 0.49557796120643616,
|
| 955 |
+
"learning_rate": 2.8927015717215733e-05,
|
| 956 |
+
"loss": 7.159825801849365,
|
| 957 |
+
"step": 20736
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.9046911026353783,
|
| 961 |
+
"grad_norm": 0.4444299638271332,
|
| 962 |
+
"learning_rate": 2.3283977921370547e-05,
|
| 963 |
+
"loss": 7.15433931350708,
|
| 964 |
+
"step": 20992
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.9157239209601999,
|
| 968 |
+
"grad_norm": 0.49471017718315125,
|
| 969 |
+
"learning_rate": 1.8239694105780413e-05,
|
| 970 |
+
"loss": 7.151764392852783,
|
| 971 |
+
"step": 21248
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.9267567392850217,
|
| 975 |
+
"grad_norm": 0.27193328738212585,
|
| 976 |
+
"learning_rate": 1.3800499876701955e-05,
|
| 977 |
+
"loss": 7.129524230957031,
|
| 978 |
+
"step": 21504
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.9267567392850217,
|
| 982 |
+
"eval_bleu": 0.421951847647465,
|
| 983 |
+
"eval_loss": 7.1731349867798375,
|
| 984 |
+
"eval_mse_loss": 1.2322176178889488,
|
| 985 |
+
"eval_recon_loss": 5.940917385158254,
|
| 986 |
+
"step": 21504
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.9267567392850217,
|
| 990 |
+
"eval_bleu": 0.421951847647465,
|
| 991 |
+
"eval_loss": 7.1731349867798375,
|
| 992 |
+
"eval_mse_loss": 1.2322176178889488,
|
| 993 |
+
"eval_recon_loss": 5.940917385158254,
|
| 994 |
+
"eval_runtime": 119.5557,
|
| 995 |
+
"eval_samples_per_second": 250.929,
|
| 996 |
+
"eval_steps_per_second": 3.923,
|
| 997 |
+
"step": 21504
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.9377895576098434,
|
| 1001 |
+
"grad_norm": 0.38439345359802246,
|
| 1002 |
+
"learning_rate": 9.971970849576406e-06,
|
| 1003 |
+
"loss": 7.190091609954834,
|
| 1004 |
+
"step": 21760
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.948822375934665,
|
| 1008 |
+
"grad_norm": 0.5374737977981567,
|
| 1009 |
+
"learning_rate": 6.758915646072339e-06,
|
| 1010 |
+
"loss": 7.214999675750732,
|
| 1011 |
+
"step": 22016
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.9598551942594867,
|
| 1015 |
+
"grad_norm": 0.55973219871521,
|
| 1016 |
+
"learning_rate": 4.1653698544703575e-06,
|
| 1017 |
+
"loss": 7.165055274963379,
|
| 1018 |
+
"step": 22272
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.9708880125843083,
|
| 1022 |
+
"grad_norm": 0.2763802111148834,
|
| 1023 |
+
"learning_rate": 2.1945909609756286e-06,
|
| 1024 |
+
"loss": 7.138909816741943,
|
| 1025 |
+
"step": 22528
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.9708880125843083,
|
| 1029 |
+
"eval_bleu": 0.42247179102439836,
|
| 1030 |
+
"eval_loss": 7.163274863635554,
|
| 1031 |
+
"eval_mse_loss": 1.2323216156664687,
|
| 1032 |
+
"eval_recon_loss": 5.930953238310336,
|
| 1033 |
+
"step": 22528
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.9708880125843083,
|
| 1037 |
+
"eval_bleu": 0.42247179102439836,
|
| 1038 |
+
"eval_loss": 7.163274863635554,
|
| 1039 |
+
"eval_mse_loss": 1.2323216156664687,
|
| 1040 |
+
"eval_recon_loss": 5.930953238310336,
|
| 1041 |
+
"eval_runtime": 119.657,
|
| 1042 |
+
"eval_samples_per_second": 250.717,
|
| 1043 |
+
"eval_steps_per_second": 3.92,
|
| 1044 |
+
"step": 22528
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 0.9819208309091301,
|
| 1048 |
+
"grad_norm": 0.19876758754253387,
|
| 1049 |
+
"learning_rate": 8.490542583243222e-07,
|
| 1050 |
+
"loss": 7.1233296394348145,
|
| 1051 |
+
"step": 22784
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 0.9929536492339518,
|
| 1055 |
+
"grad_norm": 0.17880426347255707,
|
| 1056 |
+
"learning_rate": 1.3044973682302396e-07,
|
| 1057 |
+
"loss": 7.214967250823975,
|
| 1058 |
+
"step": 23040
|
| 1059 |
+
}
|
| 1060 |
+
],
|
| 1061 |
+
"logging_steps": 256,
|
| 1062 |
+
"max_steps": 23204,
|
| 1063 |
+
"num_input_tokens_seen": 0,
|
| 1064 |
+
"num_train_epochs": 1,
|
| 1065 |
+
"save_steps": 1024,
|
| 1066 |
+
"stateful_callbacks": {
|
| 1067 |
+
"TrainerControl": {
|
| 1068 |
+
"args": {
|
| 1069 |
+
"should_epoch_stop": false,
|
| 1070 |
+
"should_evaluate": false,
|
| 1071 |
+
"should_log": false,
|
| 1072 |
+
"should_save": true,
|
| 1073 |
+
"should_training_stop": true
|
| 1074 |
+
},
|
| 1075 |
+
"attributes": {}
|
| 1076 |
+
}
|
| 1077 |
+
},
|
| 1078 |
+
"total_flos": 0.0,
|
| 1079 |
+
"train_batch_size": 64,
|
| 1080 |
+
"trial_name": null,
|
| 1081 |
+
"trial_params": null
|
| 1082 |
+
}
|
checkpoints-v2.0-discrete/checkpoint-23204/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61a16bb839f687414b8e48611327c4b9cfddeefe38c031ca70808f9a97c476b7
|
| 3 |
+
size 5137
|