Upload 6 files
Browse files- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1061 -0
- training_args.bin +3 -0
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3265451e5ae0273fd63c1b62325b4a3ba9e6238cf4439f6c260ddec924ab339
|
| 3 |
+
size 379087640
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f331d4890ac0d37bcb98af66f9f3bfd0225c9df0317e29f80774e91500be369
|
| 3 |
+
size 3152668
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2092dcf14943b2021cadabbd0e84f31fd5d9631e8f5ac0121dec336cd09c369e
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a052b05682b112b8939a79e7a05f754bab9f56dc5b70d9e120fe8a1ae9cb1a22
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1061 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.854606050035829,
|
| 3 |
+
"best_model_checkpoint": "/p/scratch/ccstdl/krishna/finetuned-cosine-loss/checkpoint-650",
|
| 4 |
+
"epoch": 0.26455026455026454,
|
| 5 |
+
"eval_steps": 10,
|
| 6 |
+
"global_step": 650,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.00407000407000407,
|
| 13 |
+
"grad_norm": 1.8341377973556519,
|
| 14 |
+
"learning_rate": 3.391670058336725e-09,
|
| 15 |
+
"loss": 0.8563,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.00407000407000407,
|
| 20 |
+
"eval_cos_sim": 0.1275755614042282,
|
| 21 |
+
"eval_loss": 0.8735055446838087,
|
| 22 |
+
"eval_runtime": 91.5265,
|
| 23 |
+
"eval_samples_per_second": 10.926,
|
| 24 |
+
"eval_steps_per_second": 0.35,
|
| 25 |
+
"step": 10
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.00814000814000814,
|
| 29 |
+
"grad_norm": 1.5246553421020508,
|
| 30 |
+
"learning_rate": 6.78334011667345e-09,
|
| 31 |
+
"loss": 0.8643,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.00814000814000814,
|
| 36 |
+
"eval_cos_sim": 0.12758877873420715,
|
| 37 |
+
"eval_loss": 0.8734923420165723,
|
| 38 |
+
"eval_runtime": 88.663,
|
| 39 |
+
"eval_samples_per_second": 11.279,
|
| 40 |
+
"eval_steps_per_second": 0.361,
|
| 41 |
+
"step": 20
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.01221001221001221,
|
| 45 |
+
"grad_norm": 1.8004460334777832,
|
| 46 |
+
"learning_rate": 1.0175010175010176e-08,
|
| 47 |
+
"loss": 0.8665,
|
| 48 |
+
"step": 30
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.01221001221001221,
|
| 52 |
+
"eval_cos_sim": 0.12761123478412628,
|
| 53 |
+
"eval_loss": 0.8734698925231642,
|
| 54 |
+
"eval_runtime": 88.8085,
|
| 55 |
+
"eval_samples_per_second": 11.26,
|
| 56 |
+
"eval_steps_per_second": 0.36,
|
| 57 |
+
"step": 30
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.01628001628001628,
|
| 61 |
+
"grad_norm": 1.8268524408340454,
|
| 62 |
+
"learning_rate": 1.35666802333469e-08,
|
| 63 |
+
"loss": 0.8661,
|
| 64 |
+
"step": 40
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.01628001628001628,
|
| 68 |
+
"eval_cos_sim": 0.1276426911354065,
|
| 69 |
+
"eval_loss": 0.8734384694312758,
|
| 70 |
+
"eval_runtime": 88.8002,
|
| 71 |
+
"eval_samples_per_second": 11.261,
|
| 72 |
+
"eval_steps_per_second": 0.36,
|
| 73 |
+
"step": 40
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.02035002035002035,
|
| 77 |
+
"grad_norm": 1.5149165391921997,
|
| 78 |
+
"learning_rate": 1.6958350291683625e-08,
|
| 79 |
+
"loss": 0.8613,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.02035002035002035,
|
| 84 |
+
"eval_cos_sim": 0.1276828944683075,
|
| 85 |
+
"eval_loss": 0.8733982930396742,
|
| 86 |
+
"eval_runtime": 89.0846,
|
| 87 |
+
"eval_samples_per_second": 11.225,
|
| 88 |
+
"eval_steps_per_second": 0.359,
|
| 89 |
+
"step": 50
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.02442002442002442,
|
| 93 |
+
"grad_norm": 1.4543992280960083,
|
| 94 |
+
"learning_rate": 2.035002035002035e-08,
|
| 95 |
+
"loss": 0.8729,
|
| 96 |
+
"step": 60
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.02442002442002442,
|
| 100 |
+
"eval_cos_sim": 0.12773165106773376,
|
| 101 |
+
"eval_loss": 0.8733495784019178,
|
| 102 |
+
"eval_runtime": 88.8621,
|
| 103 |
+
"eval_samples_per_second": 11.253,
|
| 104 |
+
"eval_steps_per_second": 0.36,
|
| 105 |
+
"step": 60
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.02849002849002849,
|
| 109 |
+
"grad_norm": 1.8213750123977661,
|
| 110 |
+
"learning_rate": 2.3741690408357078e-08,
|
| 111 |
+
"loss": 0.8632,
|
| 112 |
+
"step": 70
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.02849002849002849,
|
| 116 |
+
"eval_cos_sim": 0.12778951227664948,
|
| 117 |
+
"eval_loss": 0.8732917666648573,
|
| 118 |
+
"eval_runtime": 88.7523,
|
| 119 |
+
"eval_samples_per_second": 11.267,
|
| 120 |
+
"eval_steps_per_second": 0.361,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.03256003256003256,
|
| 125 |
+
"grad_norm": 1.8100404739379883,
|
| 126 |
+
"learning_rate": 2.71333604666938e-08,
|
| 127 |
+
"loss": 0.866,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.03256003256003256,
|
| 132 |
+
"eval_cos_sim": 0.12785714864730835,
|
| 133 |
+
"eval_loss": 0.8732241764281935,
|
| 134 |
+
"eval_runtime": 88.4393,
|
| 135 |
+
"eval_samples_per_second": 11.307,
|
| 136 |
+
"eval_steps_per_second": 0.362,
|
| 137 |
+
"step": 80
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.03663003663003663,
|
| 141 |
+
"grad_norm": 1.8266569375991821,
|
| 142 |
+
"learning_rate": 3.052503052503053e-08,
|
| 143 |
+
"loss": 0.8614,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.03663003663003663,
|
| 148 |
+
"eval_cos_sim": 0.12793460488319397,
|
| 149 |
+
"eval_loss": 0.8731467800353712,
|
| 150 |
+
"eval_runtime": 89.2974,
|
| 151 |
+
"eval_samples_per_second": 11.199,
|
| 152 |
+
"eval_steps_per_second": 0.358,
|
| 153 |
+
"step": 90
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.0407000407000407,
|
| 157 |
+
"grad_norm": 1.8240498304367065,
|
| 158 |
+
"learning_rate": 3.391670058336725e-08,
|
| 159 |
+
"loss": 0.8622,
|
| 160 |
+
"step": 100
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.0407000407000407,
|
| 164 |
+
"eval_cos_sim": 0.12802155315876007,
|
| 165 |
+
"eval_loss": 0.8730598721717543,
|
| 166 |
+
"eval_runtime": 89.1998,
|
| 167 |
+
"eval_samples_per_second": 11.211,
|
| 168 |
+
"eval_steps_per_second": 0.359,
|
| 169 |
+
"step": 100
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.04477004477004477,
|
| 173 |
+
"grad_norm": 1.5381144285202026,
|
| 174 |
+
"learning_rate": 3.730837064170397e-08,
|
| 175 |
+
"loss": 0.8598,
|
| 176 |
+
"step": 110
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.04477004477004477,
|
| 180 |
+
"eval_cos_sim": 0.12811602652072906,
|
| 181 |
+
"eval_loss": 0.8729654574607557,
|
| 182 |
+
"eval_runtime": 88.7617,
|
| 183 |
+
"eval_samples_per_second": 11.266,
|
| 184 |
+
"eval_steps_per_second": 0.361,
|
| 185 |
+
"step": 110
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.04884004884004884,
|
| 189 |
+
"grad_norm": 1.8371686935424805,
|
| 190 |
+
"learning_rate": 4.07000407000407e-08,
|
| 191 |
+
"loss": 0.859,
|
| 192 |
+
"step": 120
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.04884004884004884,
|
| 196 |
+
"eval_cos_sim": 0.12821921706199646,
|
| 197 |
+
"eval_loss": 0.8728623566840834,
|
| 198 |
+
"eval_runtime": 89.6925,
|
| 199 |
+
"eval_samples_per_second": 11.149,
|
| 200 |
+
"eval_steps_per_second": 0.357,
|
| 201 |
+
"step": 120
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.05291005291005291,
|
| 205 |
+
"grad_norm": 1.8292527198791504,
|
| 206 |
+
"learning_rate": 4.4091710758377425e-08,
|
| 207 |
+
"loss": 0.8673,
|
| 208 |
+
"step": 130
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.05291005291005291,
|
| 212 |
+
"eval_cos_sim": 0.12833045423030853,
|
| 213 |
+
"eval_loss": 0.872751211664552,
|
| 214 |
+
"eval_runtime": 88.9497,
|
| 215 |
+
"eval_samples_per_second": 11.242,
|
| 216 |
+
"eval_steps_per_second": 0.36,
|
| 217 |
+
"step": 130
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.05698005698005698,
|
| 221 |
+
"grad_norm": 1.547228455543518,
|
| 222 |
+
"learning_rate": 4.7483380816714155e-08,
|
| 223 |
+
"loss": 0.8694,
|
| 224 |
+
"step": 140
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.05698005698005698,
|
| 228 |
+
"eval_cos_sim": 0.12844981253147125,
|
| 229 |
+
"eval_loss": 0.8726319508765882,
|
| 230 |
+
"eval_runtime": 89.5439,
|
| 231 |
+
"eval_samples_per_second": 11.168,
|
| 232 |
+
"eval_steps_per_second": 0.357,
|
| 233 |
+
"step": 140
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.06105006105006105,
|
| 237 |
+
"grad_norm": 1.5469759702682495,
|
| 238 |
+
"learning_rate": 5.087505087505087e-08,
|
| 239 |
+
"loss": 0.8695,
|
| 240 |
+
"step": 150
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.06105006105006105,
|
| 244 |
+
"eval_cos_sim": 0.1285790205001831,
|
| 245 |
+
"eval_loss": 0.8725028333877272,
|
| 246 |
+
"eval_runtime": 88.9224,
|
| 247 |
+
"eval_samples_per_second": 11.246,
|
| 248 |
+
"eval_steps_per_second": 0.36,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.06512006512006512,
|
| 253 |
+
"grad_norm": 1.5482810735702515,
|
| 254 |
+
"learning_rate": 5.42667209333876e-08,
|
| 255 |
+
"loss": 0.8647,
|
| 256 |
+
"step": 160
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.06512006512006512,
|
| 260 |
+
"eval_cos_sim": 0.12871865928173065,
|
| 261 |
+
"eval_loss": 0.8723633003448195,
|
| 262 |
+
"eval_runtime": 88.7738,
|
| 263 |
+
"eval_samples_per_second": 11.265,
|
| 264 |
+
"eval_steps_per_second": 0.36,
|
| 265 |
+
"step": 160
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.06919006919006919,
|
| 269 |
+
"grad_norm": 1.5481818914413452,
|
| 270 |
+
"learning_rate": 5.7658390991724324e-08,
|
| 271 |
+
"loss": 0.8601,
|
| 272 |
+
"step": 170
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.06919006919006919,
|
| 276 |
+
"eval_cos_sim": 0.12886543571949005,
|
| 277 |
+
"eval_loss": 0.8722166295264906,
|
| 278 |
+
"eval_runtime": 88.67,
|
| 279 |
+
"eval_samples_per_second": 11.278,
|
| 280 |
+
"eval_steps_per_second": 0.361,
|
| 281 |
+
"step": 170
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.07326007326007326,
|
| 285 |
+
"grad_norm": 1.5428705215454102,
|
| 286 |
+
"learning_rate": 6.105006105006105e-08,
|
| 287 |
+
"loss": 0.8647,
|
| 288 |
+
"step": 180
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.07326007326007326,
|
| 292 |
+
"eval_cos_sim": 0.12902309000492096,
|
| 293 |
+
"eval_loss": 0.8720590691779798,
|
| 294 |
+
"eval_runtime": 89.4286,
|
| 295 |
+
"eval_samples_per_second": 11.182,
|
| 296 |
+
"eval_steps_per_second": 0.358,
|
| 297 |
+
"step": 180
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.07733007733007732,
|
| 301 |
+
"grad_norm": 1.8401004076004028,
|
| 302 |
+
"learning_rate": 6.444173110839778e-08,
|
| 303 |
+
"loss": 0.8605,
|
| 304 |
+
"step": 190
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.07733007733007732,
|
| 308 |
+
"eval_cos_sim": 0.1291920691728592,
|
| 309 |
+
"eval_loss": 0.8718901972984022,
|
| 310 |
+
"eval_runtime": 89.319,
|
| 311 |
+
"eval_samples_per_second": 11.196,
|
| 312 |
+
"eval_steps_per_second": 0.358,
|
| 313 |
+
"step": 190
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.0814000814000814,
|
| 317 |
+
"grad_norm": 1.8132197856903076,
|
| 318 |
+
"learning_rate": 6.78334011667345e-08,
|
| 319 |
+
"loss": 0.8649,
|
| 320 |
+
"step": 200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.0814000814000814,
|
| 324 |
+
"eval_cos_sim": 0.12936964631080627,
|
| 325 |
+
"eval_loss": 0.8717127638076491,
|
| 326 |
+
"eval_runtime": 88.8517,
|
| 327 |
+
"eval_samples_per_second": 11.255,
|
| 328 |
+
"eval_steps_per_second": 0.36,
|
| 329 |
+
"step": 200
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.08547008547008547,
|
| 333 |
+
"grad_norm": 1.841541051864624,
|
| 334 |
+
"learning_rate": 7.122507122507124e-08,
|
| 335 |
+
"loss": 0.8648,
|
| 336 |
+
"step": 210
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.08547008547008547,
|
| 340 |
+
"eval_cos_sim": 0.12955836951732635,
|
| 341 |
+
"eval_loss": 0.8715242018912978,
|
| 342 |
+
"eval_runtime": 88.9573,
|
| 343 |
+
"eval_samples_per_second": 11.241,
|
| 344 |
+
"eval_steps_per_second": 0.36,
|
| 345 |
+
"step": 210
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.08954008954008955,
|
| 349 |
+
"grad_norm": 1.7975077629089355,
|
| 350 |
+
"learning_rate": 7.461674128340795e-08,
|
| 351 |
+
"loss": 0.8593,
|
| 352 |
+
"step": 220
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.08954008954008955,
|
| 356 |
+
"eval_cos_sim": 0.12975716590881348,
|
| 357 |
+
"eval_loss": 0.8713255877708143,
|
| 358 |
+
"eval_runtime": 89.0134,
|
| 359 |
+
"eval_samples_per_second": 11.234,
|
| 360 |
+
"eval_steps_per_second": 0.359,
|
| 361 |
+
"step": 220
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.0936100936100936,
|
| 365 |
+
"grad_norm": 1.6715312004089355,
|
| 366 |
+
"learning_rate": 7.800841134174468e-08,
|
| 367 |
+
"loss": 0.8635,
|
| 368 |
+
"step": 230
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.0936100936100936,
|
| 372 |
+
"eval_cos_sim": 0.12996670603752136,
|
| 373 |
+
"eval_loss": 0.8711161980842298,
|
| 374 |
+
"eval_runtime": 89.3295,
|
| 375 |
+
"eval_samples_per_second": 11.195,
|
| 376 |
+
"eval_steps_per_second": 0.358,
|
| 377 |
+
"step": 230
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.09768009768009768,
|
| 381 |
+
"grad_norm": 1.8103832006454468,
|
| 382 |
+
"learning_rate": 8.14000814000814e-08,
|
| 383 |
+
"loss": 0.8609,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.09768009768009768,
|
| 388 |
+
"eval_cos_sim": 0.13018178939819336,
|
| 389 |
+
"eval_loss": 0.8709012541984266,
|
| 390 |
+
"eval_runtime": 89.2626,
|
| 391 |
+
"eval_samples_per_second": 11.203,
|
| 392 |
+
"eval_steps_per_second": 0.358,
|
| 393 |
+
"step": 240
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.10175010175010175,
|
| 397 |
+
"grad_norm": 1.8056939840316772,
|
| 398 |
+
"learning_rate": 8.479175145841813e-08,
|
| 399 |
+
"loss": 0.8639,
|
| 400 |
+
"step": 250
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.10175010175010175,
|
| 404 |
+
"eval_cos_sim": 0.13040043413639069,
|
| 405 |
+
"eval_loss": 0.8706827788566297,
|
| 406 |
+
"eval_runtime": 89.3651,
|
| 407 |
+
"eval_samples_per_second": 11.19,
|
| 408 |
+
"eval_steps_per_second": 0.358,
|
| 409 |
+
"step": 250
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.10582010582010581,
|
| 413 |
+
"grad_norm": 1.8392592668533325,
|
| 414 |
+
"learning_rate": 8.818342151675485e-08,
|
| 415 |
+
"loss": 0.8607,
|
| 416 |
+
"step": 260
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.10582010582010581,
|
| 420 |
+
"eval_cos_sim": 0.1306302398443222,
|
| 421 |
+
"eval_loss": 0.8704531607841199,
|
| 422 |
+
"eval_runtime": 89.3602,
|
| 423 |
+
"eval_samples_per_second": 11.191,
|
| 424 |
+
"eval_steps_per_second": 0.358,
|
| 425 |
+
"step": 260
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.10989010989010989,
|
| 429 |
+
"grad_norm": 1.8241254091262817,
|
| 430 |
+
"learning_rate": 9.157509157509157e-08,
|
| 431 |
+
"loss": 0.8601,
|
| 432 |
+
"step": 270
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.10989010989010989,
|
| 436 |
+
"eval_cos_sim": 0.13087190687656403,
|
| 437 |
+
"eval_loss": 0.8702116866325087,
|
| 438 |
+
"eval_runtime": 89.3533,
|
| 439 |
+
"eval_samples_per_second": 11.192,
|
| 440 |
+
"eval_steps_per_second": 0.358,
|
| 441 |
+
"step": 270
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.11396011396011396,
|
| 445 |
+
"grad_norm": 1.5199859142303467,
|
| 446 |
+
"learning_rate": 9.496676163342831e-08,
|
| 447 |
+
"loss": 0.862,
|
| 448 |
+
"step": 280
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.11396011396011396,
|
| 452 |
+
"eval_cos_sim": 0.1311246156692505,
|
| 453 |
+
"eval_loss": 0.8699591713164991,
|
| 454 |
+
"eval_runtime": 89.8365,
|
| 455 |
+
"eval_samples_per_second": 11.131,
|
| 456 |
+
"eval_steps_per_second": 0.356,
|
| 457 |
+
"step": 280
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.11803011803011804,
|
| 461 |
+
"grad_norm": 1.539494276046753,
|
| 462 |
+
"learning_rate": 9.835843169176503e-08,
|
| 463 |
+
"loss": 0.8646,
|
| 464 |
+
"step": 290
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.11803011803011804,
|
| 468 |
+
"eval_cos_sim": 0.13138911128044128,
|
| 469 |
+
"eval_loss": 0.8696949014877028,
|
| 470 |
+
"eval_runtime": 89.1002,
|
| 471 |
+
"eval_samples_per_second": 11.223,
|
| 472 |
+
"eval_steps_per_second": 0.359,
|
| 473 |
+
"step": 290
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.1221001221001221,
|
| 477 |
+
"grad_norm": 1.8310585021972656,
|
| 478 |
+
"learning_rate": 1.0175010175010174e-07,
|
| 479 |
+
"loss": 0.8678,
|
| 480 |
+
"step": 300
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.1221001221001221,
|
| 484 |
+
"eval_cos_sim": 0.13165612518787384,
|
| 485 |
+
"eval_loss": 0.8694280743812269,
|
| 486 |
+
"eval_runtime": 89.8617,
|
| 487 |
+
"eval_samples_per_second": 11.128,
|
| 488 |
+
"eval_steps_per_second": 0.356,
|
| 489 |
+
"step": 300
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.12617012617012616,
|
| 493 |
+
"grad_norm": 1.5015047788619995,
|
| 494 |
+
"learning_rate": 1.0514177180843848e-07,
|
| 495 |
+
"loss": 0.8646,
|
| 496 |
+
"step": 310
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.12617012617012616,
|
| 500 |
+
"eval_cos_sim": 0.13192161917686462,
|
| 501 |
+
"eval_loss": 0.8691628270362562,
|
| 502 |
+
"eval_runtime": 89.425,
|
| 503 |
+
"eval_samples_per_second": 11.183,
|
| 504 |
+
"eval_steps_per_second": 0.358,
|
| 505 |
+
"step": 310
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.13024013024013023,
|
| 509 |
+
"grad_norm": 1.8223438262939453,
|
| 510 |
+
"learning_rate": 1.085334418667752e-07,
|
| 511 |
+
"loss": 0.8574,
|
| 512 |
+
"step": 320
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.13024013024013023,
|
| 516 |
+
"eval_cos_sim": 0.13219775259494781,
|
| 517 |
+
"eval_loss": 0.8688869447921461,
|
| 518 |
+
"eval_runtime": 89.6601,
|
| 519 |
+
"eval_samples_per_second": 11.153,
|
| 520 |
+
"eval_steps_per_second": 0.357,
|
| 521 |
+
"step": 320
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.1343101343101343,
|
| 525 |
+
"grad_norm": 1.834613561630249,
|
| 526 |
+
"learning_rate": 1.1192511192511194e-07,
|
| 527 |
+
"loss": 0.8609,
|
| 528 |
+
"step": 330
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.1343101343101343,
|
| 532 |
+
"eval_cos_sim": 0.13248248398303986,
|
| 533 |
+
"eval_loss": 0.868602445146913,
|
| 534 |
+
"eval_runtime": 89.6443,
|
| 535 |
+
"eval_samples_per_second": 11.155,
|
| 536 |
+
"eval_steps_per_second": 0.357,
|
| 537 |
+
"step": 330
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.13838013838013838,
|
| 541 |
+
"grad_norm": 1.8023933172225952,
|
| 542 |
+
"learning_rate": 1.1531678198344865e-07,
|
| 543 |
+
"loss": 0.8587,
|
| 544 |
+
"step": 340
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.13838013838013838,
|
| 548 |
+
"eval_cos_sim": 0.13278011977672577,
|
| 549 |
+
"eval_loss": 0.8683050141547864,
|
| 550 |
+
"eval_runtime": 89.9517,
|
| 551 |
+
"eval_samples_per_second": 11.117,
|
| 552 |
+
"eval_steps_per_second": 0.356,
|
| 553 |
+
"step": 340
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 0.14245014245014245,
|
| 557 |
+
"grad_norm": 1.821526288986206,
|
| 558 |
+
"learning_rate": 1.1870845204178537e-07,
|
| 559 |
+
"loss": 0.8598,
|
| 560 |
+
"step": 350
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.14245014245014245,
|
| 564 |
+
"eval_cos_sim": 0.1330917626619339,
|
| 565 |
+
"eval_loss": 0.867993585607881,
|
| 566 |
+
"eval_runtime": 90.2991,
|
| 567 |
+
"eval_samples_per_second": 11.074,
|
| 568 |
+
"eval_steps_per_second": 0.354,
|
| 569 |
+
"step": 350
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.14652014652014653,
|
| 573 |
+
"grad_norm": 1.801190733909607,
|
| 574 |
+
"learning_rate": 1.221001221001221e-07,
|
| 575 |
+
"loss": 0.8525,
|
| 576 |
+
"step": 360
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.14652014652014653,
|
| 580 |
+
"eval_cos_sim": 0.13341833651065826,
|
| 581 |
+
"eval_loss": 0.8676672454093641,
|
| 582 |
+
"eval_runtime": 90.0422,
|
| 583 |
+
"eval_samples_per_second": 11.106,
|
| 584 |
+
"eval_steps_per_second": 0.355,
|
| 585 |
+
"step": 360
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.1505901505901506,
|
| 589 |
+
"grad_norm": 1.5358909368515015,
|
| 590 |
+
"learning_rate": 1.2549179215845883e-07,
|
| 591 |
+
"loss": 0.8575,
|
| 592 |
+
"step": 370
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.1505901505901506,
|
| 596 |
+
"eval_cos_sim": 0.13375183939933777,
|
| 597 |
+
"eval_loss": 0.8673340068076796,
|
| 598 |
+
"eval_runtime": 90.0945,
|
| 599 |
+
"eval_samples_per_second": 11.099,
|
| 600 |
+
"eval_steps_per_second": 0.355,
|
| 601 |
+
"step": 370
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.15466015466015465,
|
| 605 |
+
"grad_norm": 1.7882535457611084,
|
| 606 |
+
"learning_rate": 1.2888346221679555e-07,
|
| 607 |
+
"loss": 0.8521,
|
| 608 |
+
"step": 380
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.15466015466015465,
|
| 612 |
+
"eval_cos_sim": 0.13408976793289185,
|
| 613 |
+
"eval_loss": 0.8669963665221876,
|
| 614 |
+
"eval_runtime": 90.0493,
|
| 615 |
+
"eval_samples_per_second": 11.105,
|
| 616 |
+
"eval_steps_per_second": 0.355,
|
| 617 |
+
"step": 380
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.15873015873015872,
|
| 621 |
+
"grad_norm": 1.839800477027893,
|
| 622 |
+
"learning_rate": 1.3227513227513228e-07,
|
| 623 |
+
"loss": 0.8558,
|
| 624 |
+
"step": 390
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.15873015873015872,
|
| 628 |
+
"eval_cos_sim": 0.13443315029144287,
|
| 629 |
+
"eval_loss": 0.8666532302116102,
|
| 630 |
+
"eval_runtime": 89.9978,
|
| 631 |
+
"eval_samples_per_second": 11.111,
|
| 632 |
+
"eval_steps_per_second": 0.356,
|
| 633 |
+
"step": 390
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.1628001628001628,
|
| 637 |
+
"grad_norm": 1.8280110359191895,
|
| 638 |
+
"learning_rate": 1.35666802333469e-07,
|
| 639 |
+
"loss": 0.8539,
|
| 640 |
+
"step": 400
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.1628001628001628,
|
| 644 |
+
"eval_cos_sim": 0.13477978110313416,
|
| 645 |
+
"eval_loss": 0.8663068804954237,
|
| 646 |
+
"eval_runtime": 90.1723,
|
| 647 |
+
"eval_samples_per_second": 11.09,
|
| 648 |
+
"eval_steps_per_second": 0.355,
|
| 649 |
+
"step": 400
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.16687016687016687,
|
| 653 |
+
"grad_norm": 1.5013295412063599,
|
| 654 |
+
"learning_rate": 1.3905847239180572e-07,
|
| 655 |
+
"loss": 0.8631,
|
| 656 |
+
"step": 410
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.16687016687016687,
|
| 660 |
+
"eval_cos_sim": 0.13513372838497162,
|
| 661 |
+
"eval_loss": 0.8659531727050489,
|
| 662 |
+
"eval_runtime": 90.4867,
|
| 663 |
+
"eval_samples_per_second": 11.051,
|
| 664 |
+
"eval_steps_per_second": 0.354,
|
| 665 |
+
"step": 410
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.17094017094017094,
|
| 669 |
+
"grad_norm": 1.7966923713684082,
|
| 670 |
+
"learning_rate": 1.4245014245014247e-07,
|
| 671 |
+
"loss": 0.8533,
|
| 672 |
+
"step": 420
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 0.17094017094017094,
|
| 676 |
+
"eval_cos_sim": 0.13550296425819397,
|
| 677 |
+
"eval_loss": 0.8655842075561232,
|
| 678 |
+
"eval_runtime": 89.6565,
|
| 679 |
+
"eval_samples_per_second": 11.154,
|
| 680 |
+
"eval_steps_per_second": 0.357,
|
| 681 |
+
"step": 420
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.17501017501017502,
|
| 685 |
+
"grad_norm": 1.8006147146224976,
|
| 686 |
+
"learning_rate": 1.4584181250847917e-07,
|
| 687 |
+
"loss": 0.8549,
|
| 688 |
+
"step": 430
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.17501017501017502,
|
| 692 |
+
"eval_cos_sim": 0.1358877271413803,
|
| 693 |
+
"eval_loss": 0.8651996951316542,
|
| 694 |
+
"eval_runtime": 91.2533,
|
| 695 |
+
"eval_samples_per_second": 10.959,
|
| 696 |
+
"eval_steps_per_second": 0.351,
|
| 697 |
+
"step": 430
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.1790801790801791,
|
| 701 |
+
"grad_norm": 1.7972980737686157,
|
| 702 |
+
"learning_rate": 1.492334825668159e-07,
|
| 703 |
+
"loss": 0.8568,
|
| 704 |
+
"step": 440
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.1790801790801791,
|
| 708 |
+
"eval_cos_sim": 0.136278435587883,
|
| 709 |
+
"eval_loss": 0.8648093180869764,
|
| 710 |
+
"eval_runtime": 90.2011,
|
| 711 |
+
"eval_samples_per_second": 11.086,
|
| 712 |
+
"eval_steps_per_second": 0.355,
|
| 713 |
+
"step": 440
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.18315018315018314,
|
| 717 |
+
"grad_norm": 1.8002476692199707,
|
| 718 |
+
"learning_rate": 1.5262515262515264e-07,
|
| 719 |
+
"loss": 0.8593,
|
| 720 |
+
"step": 450
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.18315018315018314,
|
| 724 |
+
"eval_cos_sim": 0.1366884559392929,
|
| 725 |
+
"eval_loss": 0.8643996544097609,
|
| 726 |
+
"eval_runtime": 90.0768,
|
| 727 |
+
"eval_samples_per_second": 11.102,
|
| 728 |
+
"eval_steps_per_second": 0.355,
|
| 729 |
+
"step": 450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.1872201872201872,
|
| 733 |
+
"grad_norm": 1.8119208812713623,
|
| 734 |
+
"learning_rate": 1.5601682268348936e-07,
|
| 735 |
+
"loss": 0.8497,
|
| 736 |
+
"step": 460
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 0.1872201872201872,
|
| 740 |
+
"eval_cos_sim": 0.1371019184589386,
|
| 741 |
+
"eval_loss": 0.8639865331863111,
|
| 742 |
+
"eval_runtime": 90.2856,
|
| 743 |
+
"eval_samples_per_second": 11.076,
|
| 744 |
+
"eval_steps_per_second": 0.354,
|
| 745 |
+
"step": 460
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.19129019129019129,
|
| 749 |
+
"grad_norm": 1.5471041202545166,
|
| 750 |
+
"learning_rate": 1.594084927418261e-07,
|
| 751 |
+
"loss": 0.857,
|
| 752 |
+
"step": 470
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.19129019129019129,
|
| 756 |
+
"eval_cos_sim": 0.13752888143062592,
|
| 757 |
+
"eval_loss": 0.8635599189017957,
|
| 758 |
+
"eval_runtime": 90.6564,
|
| 759 |
+
"eval_samples_per_second": 11.031,
|
| 760 |
+
"eval_steps_per_second": 0.353,
|
| 761 |
+
"step": 470
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.19536019536019536,
|
| 765 |
+
"grad_norm": 1.5276424884796143,
|
| 766 |
+
"learning_rate": 1.628001628001628e-07,
|
| 767 |
+
"loss": 0.8553,
|
| 768 |
+
"step": 480
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.19536019536019536,
|
| 772 |
+
"eval_cos_sim": 0.13795536756515503,
|
| 773 |
+
"eval_loss": 0.8631337700103467,
|
| 774 |
+
"eval_runtime": 89.9592,
|
| 775 |
+
"eval_samples_per_second": 11.116,
|
| 776 |
+
"eval_steps_per_second": 0.356,
|
| 777 |
+
"step": 480
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.19943019943019943,
|
| 781 |
+
"grad_norm": 1.8085660934448242,
|
| 782 |
+
"learning_rate": 1.6619183285849953e-07,
|
| 783 |
+
"loss": 0.8559,
|
| 784 |
+
"step": 490
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 0.19943019943019943,
|
| 788 |
+
"eval_cos_sim": 0.13838599622249603,
|
| 789 |
+
"eval_loss": 0.8627034917091078,
|
| 790 |
+
"eval_runtime": 89.645,
|
| 791 |
+
"eval_samples_per_second": 11.155,
|
| 792 |
+
"eval_steps_per_second": 0.357,
|
| 793 |
+
"step": 490
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.2035002035002035,
|
| 797 |
+
"grad_norm": 1.8282767534255981,
|
| 798 |
+
"learning_rate": 1.6958350291683626e-07,
|
| 799 |
+
"loss": 0.8533,
|
| 800 |
+
"step": 500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.2035002035002035,
|
| 804 |
+
"eval_cos_sim": 0.13883180916309357,
|
| 805 |
+
"eval_loss": 0.8622579980109877,
|
| 806 |
+
"eval_runtime": 90.3692,
|
| 807 |
+
"eval_samples_per_second": 11.066,
|
| 808 |
+
"eval_steps_per_second": 0.354,
|
| 809 |
+
"step": 500
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.20757020757020758,
|
| 813 |
+
"grad_norm": 1.5567238330841064,
|
| 814 |
+
"learning_rate": 1.7297517297517298e-07,
|
| 815 |
+
"loss": 0.8525,
|
| 816 |
+
"step": 510
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.20757020757020758,
|
| 820 |
+
"eval_cos_sim": 0.1392912119626999,
|
| 821 |
+
"eval_loss": 0.8617989306663221,
|
| 822 |
+
"eval_runtime": 89.9834,
|
| 823 |
+
"eval_samples_per_second": 11.113,
|
| 824 |
+
"eval_steps_per_second": 0.356,
|
| 825 |
+
"step": 510
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.21164021164021163,
|
| 829 |
+
"grad_norm": 1.5459282398223877,
|
| 830 |
+
"learning_rate": 1.763668430335097e-07,
|
| 831 |
+
"loss": 0.853,
|
| 832 |
+
"step": 520
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 0.21164021164021163,
|
| 836 |
+
"eval_cos_sim": 0.13975809514522552,
|
| 837 |
+
"eval_loss": 0.8613324156020826,
|
| 838 |
+
"eval_runtime": 90.0086,
|
| 839 |
+
"eval_samples_per_second": 11.11,
|
| 840 |
+
"eval_steps_per_second": 0.356,
|
| 841 |
+
"step": 520
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.2157102157102157,
|
| 845 |
+
"grad_norm": 1.799883484840393,
|
| 846 |
+
"learning_rate": 1.7975851309184642e-07,
|
| 847 |
+
"loss": 0.8512,
|
| 848 |
+
"step": 530
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.2157102157102157,
|
| 852 |
+
"eval_cos_sim": 0.14022615551948547,
|
| 853 |
+
"eval_loss": 0.8608647189353651,
|
| 854 |
+
"eval_runtime": 90.0398,
|
| 855 |
+
"eval_samples_per_second": 11.106,
|
| 856 |
+
"eval_steps_per_second": 0.355,
|
| 857 |
+
"step": 530
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.21978021978021978,
|
| 861 |
+
"grad_norm": 1.5823848247528076,
|
| 862 |
+
"learning_rate": 1.8315018315018315e-07,
|
| 863 |
+
"loss": 0.86,
|
| 864 |
+
"step": 540
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.21978021978021978,
|
| 868 |
+
"eval_cos_sim": 0.14070864021778107,
|
| 869 |
+
"eval_loss": 0.8603825893615431,
|
| 870 |
+
"eval_runtime": 90.1006,
|
| 871 |
+
"eval_samples_per_second": 11.099,
|
| 872 |
+
"eval_steps_per_second": 0.355,
|
| 873 |
+
"step": 540
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.22385022385022385,
|
| 877 |
+
"grad_norm": 1.8258429765701294,
|
| 878 |
+
"learning_rate": 1.865418532085199e-07,
|
| 879 |
+
"loss": 0.8504,
|
| 880 |
+
"step": 550
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.22385022385022385,
|
| 884 |
+
"eval_cos_sim": 0.14118121564388275,
|
| 885 |
+
"eval_loss": 0.8599103837226576,
|
| 886 |
+
"eval_runtime": 90.1395,
|
| 887 |
+
"eval_samples_per_second": 11.094,
|
| 888 |
+
"eval_steps_per_second": 0.355,
|
| 889 |
+
"step": 550
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.22792022792022792,
|
| 893 |
+
"grad_norm": 1.4464470148086548,
|
| 894 |
+
"learning_rate": 1.8993352326685662e-07,
|
| 895 |
+
"loss": 0.8574,
|
| 896 |
+
"step": 560
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.22792022792022792,
|
| 900 |
+
"eval_cos_sim": 0.1416817009449005,
|
| 901 |
+
"eval_loss": 0.8594102664207166,
|
| 902 |
+
"eval_runtime": 90.7409,
|
| 903 |
+
"eval_samples_per_second": 11.02,
|
| 904 |
+
"eval_steps_per_second": 0.353,
|
| 905 |
+
"step": 560
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.231990231990232,
|
| 909 |
+
"grad_norm": 1.8329437971115112,
|
| 910 |
+
"learning_rate": 1.9332519332519332e-07,
|
| 911 |
+
"loss": 0.855,
|
| 912 |
+
"step": 570
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.231990231990232,
|
| 916 |
+
"eval_cos_sim": 0.14218488335609436,
|
| 917 |
+
"eval_loss": 0.8589075465415663,
|
| 918 |
+
"eval_runtime": 90.6138,
|
| 919 |
+
"eval_samples_per_second": 11.036,
|
| 920 |
+
"eval_steps_per_second": 0.353,
|
| 921 |
+
"step": 570
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.23606023606023607,
|
| 925 |
+
"grad_norm": 1.8315315246582031,
|
| 926 |
+
"learning_rate": 1.9671686338353007e-07,
|
| 927 |
+
"loss": 0.8553,
|
| 928 |
+
"step": 580
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.23606023606023607,
|
| 932 |
+
"eval_cos_sim": 0.14268743991851807,
|
| 933 |
+
"eval_loss": 0.8584054036353773,
|
| 934 |
+
"eval_runtime": 90.2957,
|
| 935 |
+
"eval_samples_per_second": 11.075,
|
| 936 |
+
"eval_steps_per_second": 0.354,
|
| 937 |
+
"step": 580
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.24013024013024012,
|
| 941 |
+
"grad_norm": 1.5344057083129883,
|
| 942 |
+
"learning_rate": 2.001085334418668e-07,
|
| 943 |
+
"loss": 0.8448,
|
| 944 |
+
"step": 590
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.24013024013024012,
|
| 948 |
+
"eval_cos_sim": 0.14320193231105804,
|
| 949 |
+
"eval_loss": 0.8578913531516736,
|
| 950 |
+
"eval_runtime": 90.9648,
|
| 951 |
+
"eval_samples_per_second": 10.993,
|
| 952 |
+
"eval_steps_per_second": 0.352,
|
| 953 |
+
"step": 590
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.2442002442002442,
|
| 957 |
+
"grad_norm": 1.6833467483520508,
|
| 958 |
+
"learning_rate": 2.0350020350020349e-07,
|
| 959 |
+
"loss": 0.8521,
|
| 960 |
+
"step": 600
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.2442002442002442,
|
| 964 |
+
"eval_cos_sim": 0.14373008906841278,
|
| 965 |
+
"eval_loss": 0.8573636312698072,
|
| 966 |
+
"eval_runtime": 90.5973,
|
| 967 |
+
"eval_samples_per_second": 11.038,
|
| 968 |
+
"eval_steps_per_second": 0.353,
|
| 969 |
+
"step": 600
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.24827024827024827,
|
| 973 |
+
"grad_norm": 1.808730959892273,
|
| 974 |
+
"learning_rate": 2.0689187355854024e-07,
|
| 975 |
+
"loss": 0.844,
|
| 976 |
+
"step": 610
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.24827024827024827,
|
| 980 |
+
"eval_cos_sim": 0.14426389336585999,
|
| 981 |
+
"eval_loss": 0.8568301882957167,
|
| 982 |
+
"eval_runtime": 90.4286,
|
| 983 |
+
"eval_samples_per_second": 11.058,
|
| 984 |
+
"eval_steps_per_second": 0.354,
|
| 985 |
+
"step": 610
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.2523402523402523,
|
| 989 |
+
"grad_norm": 1.844853162765503,
|
| 990 |
+
"learning_rate": 2.1028354361687696e-07,
|
| 991 |
+
"loss": 0.8461,
|
| 992 |
+
"step": 620
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.2523402523402523,
|
| 996 |
+
"eval_cos_sim": 0.14481142163276672,
|
| 997 |
+
"eval_loss": 0.8562831025336928,
|
| 998 |
+
"eval_runtime": 90.4232,
|
| 999 |
+
"eval_samples_per_second": 11.059,
|
| 1000 |
+
"eval_steps_per_second": 0.354,
|
| 1001 |
+
"step": 620
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.2564102564102564,
|
| 1005 |
+
"grad_norm": 1.4428536891937256,
|
| 1006 |
+
"learning_rate": 2.136752136752137e-07,
|
| 1007 |
+
"loss": 0.8521,
|
| 1008 |
+
"step": 630
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.2564102564102564,
|
| 1012 |
+
"eval_cos_sim": 0.14536888897418976,
|
| 1013 |
+
"eval_loss": 0.8557261476730055,
|
| 1014 |
+
"eval_runtime": 90.6076,
|
| 1015 |
+
"eval_samples_per_second": 11.037,
|
| 1016 |
+
"eval_steps_per_second": 0.353,
|
| 1017 |
+
"step": 630
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.26048026048026046,
|
| 1021 |
+
"grad_norm": 1.8345551490783691,
|
| 1022 |
+
"learning_rate": 2.170668837335504e-07,
|
| 1023 |
+
"loss": 0.8396,
|
| 1024 |
+
"step": 640
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.26048026048026046,
|
| 1028 |
+
"eval_cos_sim": 0.14592213928699493,
|
| 1029 |
+
"eval_loss": 0.8551734004234022,
|
| 1030 |
+
"eval_runtime": 90.9039,
|
| 1031 |
+
"eval_samples_per_second": 11.001,
|
| 1032 |
+
"eval_steps_per_second": 0.352,
|
| 1033 |
+
"step": 640
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.26455026455026454,
|
| 1037 |
+
"grad_norm": 1.5427323579788208,
|
| 1038 |
+
"learning_rate": 2.2045855379188713e-07,
|
| 1039 |
+
"loss": 0.8471,
|
| 1040 |
+
"step": 650
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.26455026455026454,
|
| 1044 |
+
"eval_cos_sim": 0.1464899480342865,
|
| 1045 |
+
"eval_loss": 0.854606050035829,
|
| 1046 |
+
"eval_runtime": 89.9594,
|
| 1047 |
+
"eval_samples_per_second": 11.116,
|
| 1048 |
+
"eval_steps_per_second": 0.356,
|
| 1049 |
+
"step": 650
|
| 1050 |
+
}
|
| 1051 |
+
],
|
| 1052 |
+
"logging_steps": 10,
|
| 1053 |
+
"max_steps": 1474200,
|
| 1054 |
+
"num_input_tokens_seen": 0,
|
| 1055 |
+
"num_train_epochs": 600,
|
| 1056 |
+
"save_steps": 10,
|
| 1057 |
+
"total_flos": 0.0,
|
| 1058 |
+
"train_batch_size": 160,
|
| 1059 |
+
"trial_name": null,
|
| 1060 |
+
"trial_params": null
|
| 1061 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:89c7fd42de0eff2fd3f2c14d428b5bd8964525b5a893de04e1c70ac559b419fa
|
| 3 |
+
size 5048
|