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 +1029 -0
- training_args.bin +3 -0
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e0a5c9d47ad4e209d33f14efa11db9eae9ca1bf412bd34c82960065bb87baa4
|
| 3 |
+
size 379087640
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a4d7f9e16ade73196785688a1fb533f030064162f182299b095b084fa66b17da
|
| 3 |
+
size 3152668
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63113c1ce25a81235d736156b646c9b1ba41e59ed903b198b011fbf3da63f012
|
| 3 |
+
size 14180
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:70367f1ecb7072f0d92a51f9443c984cf7b5d0fe7e33d054198e6774fa0d89e9
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1029 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.7681954999183362,
|
| 3 |
+
"best_model_checkpoint": "/p/scratch/ccstdl/krishna/finetuned-cosine-loss/checkpoint-630",
|
| 4 |
+
"epoch": 0.2564102564102564,
|
| 5 |
+
"eval_steps": 10,
|
| 6 |
+
"global_step": 630,
|
| 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": 1.8248943090438843,
|
| 14 |
+
"learning_rate": 3.391670058336725e-09,
|
| 15 |
+
"loss": 0.7718,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.00407000407000407,
|
| 20 |
+
"eval_cos_sim": 0.2161455750465393,
|
| 21 |
+
"eval_loss": 0.7850055160735792,
|
| 22 |
+
"eval_runtime": 91.1558,
|
| 23 |
+
"eval_samples_per_second": 10.97,
|
| 24 |
+
"eval_steps_per_second": 0.351,
|
| 25 |
+
"step": 10
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.00814000814000814,
|
| 29 |
+
"grad_norm": 1.5165057182312012,
|
| 30 |
+
"learning_rate": 6.78334011667345e-09,
|
| 31 |
+
"loss": 0.7694,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.00814000814000814,
|
| 36 |
+
"eval_cos_sim": 0.21615812182426453,
|
| 37 |
+
"eval_loss": 0.7849929766868299,
|
| 38 |
+
"eval_runtime": 88.5893,
|
| 39 |
+
"eval_samples_per_second": 11.288,
|
| 40 |
+
"eval_steps_per_second": 0.361,
|
| 41 |
+
"step": 20
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.01221001221001221,
|
| 45 |
+
"grad_norm": 1.8078064918518066,
|
| 46 |
+
"learning_rate": 1.0175010175010176e-08,
|
| 47 |
+
"loss": 0.7875,
|
| 48 |
+
"step": 30
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.01221001221001221,
|
| 52 |
+
"eval_cos_sim": 0.21617917716503143,
|
| 53 |
+
"eval_loss": 0.7849719367240614,
|
| 54 |
+
"eval_runtime": 88.6143,
|
| 55 |
+
"eval_samples_per_second": 11.285,
|
| 56 |
+
"eval_steps_per_second": 0.361,
|
| 57 |
+
"step": 30
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.01628001628001628,
|
| 61 |
+
"grad_norm": 1.8216888904571533,
|
| 62 |
+
"learning_rate": 1.35666802333469e-08,
|
| 63 |
+
"loss": 0.7742,
|
| 64 |
+
"step": 40
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.01628001628001628,
|
| 68 |
+
"eval_cos_sim": 0.21620890498161316,
|
| 69 |
+
"eval_loss": 0.7849422373984999,
|
| 70 |
+
"eval_runtime": 88.4565,
|
| 71 |
+
"eval_samples_per_second": 11.305,
|
| 72 |
+
"eval_steps_per_second": 0.362,
|
| 73 |
+
"step": 40
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.02035002035002035,
|
| 77 |
+
"grad_norm": 1.7948490381240845,
|
| 78 |
+
"learning_rate": 1.6958350291683625e-08,
|
| 79 |
+
"loss": 0.7683,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.02035002035002035,
|
| 84 |
+
"eval_cos_sim": 0.2162467986345291,
|
| 85 |
+
"eval_loss": 0.7849043712829298,
|
| 86 |
+
"eval_runtime": 88.6129,
|
| 87 |
+
"eval_samples_per_second": 11.285,
|
| 88 |
+
"eval_steps_per_second": 0.361,
|
| 89 |
+
"step": 50
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.02442002442002442,
|
| 93 |
+
"grad_norm": 1.5213645696640015,
|
| 94 |
+
"learning_rate": 2.035002035002035e-08,
|
| 95 |
+
"loss": 0.7773,
|
| 96 |
+
"step": 60
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.02442002442002442,
|
| 100 |
+
"eval_cos_sim": 0.21629446744918823,
|
| 101 |
+
"eval_loss": 0.7848567362044996,
|
| 102 |
+
"eval_runtime": 88.6351,
|
| 103 |
+
"eval_samples_per_second": 11.282,
|
| 104 |
+
"eval_steps_per_second": 0.361,
|
| 105 |
+
"step": 60
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.02849002849002849,
|
| 109 |
+
"grad_norm": 1.816196322441101,
|
| 110 |
+
"learning_rate": 2.3741690408357078e-08,
|
| 111 |
+
"loss": 0.7677,
|
| 112 |
+
"step": 70
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.02849002849002849,
|
| 116 |
+
"eval_cos_sim": 0.21635092794895172,
|
| 117 |
+
"eval_loss": 0.7848003177856153,
|
| 118 |
+
"eval_runtime": 88.6682,
|
| 119 |
+
"eval_samples_per_second": 11.278,
|
| 120 |
+
"eval_steps_per_second": 0.361,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.03256003256003256,
|
| 125 |
+
"grad_norm": 1.8195900917053223,
|
| 126 |
+
"learning_rate": 2.71333604666938e-08,
|
| 127 |
+
"loss": 0.78,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.03256003256003256,
|
| 132 |
+
"eval_cos_sim": 0.21641571819782257,
|
| 133 |
+
"eval_loss": 0.7847355809425062,
|
| 134 |
+
"eval_runtime": 88.4895,
|
| 135 |
+
"eval_samples_per_second": 11.301,
|
| 136 |
+
"eval_steps_per_second": 0.362,
|
| 137 |
+
"step": 80
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.03663003663003663,
|
| 141 |
+
"grad_norm": 1.7980706691741943,
|
| 142 |
+
"learning_rate": 3.052503052503053e-08,
|
| 143 |
+
"loss": 0.7655,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.03663003663003663,
|
| 148 |
+
"eval_cos_sim": 0.2164890468120575,
|
| 149 |
+
"eval_loss": 0.7846623053764051,
|
| 150 |
+
"eval_runtime": 88.0594,
|
| 151 |
+
"eval_samples_per_second": 11.356,
|
| 152 |
+
"eval_steps_per_second": 0.363,
|
| 153 |
+
"step": 90
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.0407000407000407,
|
| 157 |
+
"grad_norm": 1.8171244859695435,
|
| 158 |
+
"learning_rate": 3.391670058336725e-08,
|
| 159 |
+
"loss": 0.7747,
|
| 160 |
+
"step": 100
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.0407000407000407,
|
| 164 |
+
"eval_cos_sim": 0.2165709137916565,
|
| 165 |
+
"eval_loss": 0.7845804929946607,
|
| 166 |
+
"eval_runtime": 88.4211,
|
| 167 |
+
"eval_samples_per_second": 11.31,
|
| 168 |
+
"eval_steps_per_second": 0.362,
|
| 169 |
+
"step": 100
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.04477004477004477,
|
| 173 |
+
"grad_norm": 1.529422640800476,
|
| 174 |
+
"learning_rate": 3.730837064170397e-08,
|
| 175 |
+
"loss": 0.776,
|
| 176 |
+
"step": 110
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.04477004477004477,
|
| 180 |
+
"eval_cos_sim": 0.2166617214679718,
|
| 181 |
+
"eval_loss": 0.7844897389625257,
|
| 182 |
+
"eval_runtime": 88.6911,
|
| 183 |
+
"eval_samples_per_second": 11.275,
|
| 184 |
+
"eval_steps_per_second": 0.361,
|
| 185 |
+
"step": 110
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.04884004884004884,
|
| 189 |
+
"grad_norm": 1.828736424446106,
|
| 190 |
+
"learning_rate": 4.07000407000407e-08,
|
| 191 |
+
"loss": 0.7757,
|
| 192 |
+
"step": 120
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.04884004884004884,
|
| 196 |
+
"eval_cos_sim": 0.216759592294693,
|
| 197 |
+
"eval_loss": 0.7843919339393324,
|
| 198 |
+
"eval_runtime": 89.0261,
|
| 199 |
+
"eval_samples_per_second": 11.233,
|
| 200 |
+
"eval_steps_per_second": 0.359,
|
| 201 |
+
"step": 120
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.05291005291005291,
|
| 205 |
+
"grad_norm": 1.5348294973373413,
|
| 206 |
+
"learning_rate": 4.4091710758377425e-08,
|
| 207 |
+
"loss": 0.7829,
|
| 208 |
+
"step": 130
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.05291005291005291,
|
| 212 |
+
"eval_cos_sim": 0.21686479449272156,
|
| 213 |
+
"eval_loss": 0.7842868261550612,
|
| 214 |
+
"eval_runtime": 88.7305,
|
| 215 |
+
"eval_samples_per_second": 11.27,
|
| 216 |
+
"eval_steps_per_second": 0.361,
|
| 217 |
+
"step": 130
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.05698005698005698,
|
| 221 |
+
"grad_norm": 1.8242194652557373,
|
| 222 |
+
"learning_rate": 4.7483380816714155e-08,
|
| 223 |
+
"loss": 0.7794,
|
| 224 |
+
"step": 140
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.05698005698005698,
|
| 228 |
+
"eval_cos_sim": 0.21697954833507538,
|
| 229 |
+
"eval_loss": 0.7841721568320936,
|
| 230 |
+
"eval_runtime": 88.4802,
|
| 231 |
+
"eval_samples_per_second": 11.302,
|
| 232 |
+
"eval_steps_per_second": 0.362,
|
| 233 |
+
"step": 140
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.06105006105006105,
|
| 237 |
+
"grad_norm": 1.4458454847335815,
|
| 238 |
+
"learning_rate": 5.087505087505087e-08,
|
| 239 |
+
"loss": 0.7751,
|
| 240 |
+
"step": 150
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.06105006105006105,
|
| 244 |
+
"eval_cos_sim": 0.21710529923439026,
|
| 245 |
+
"eval_loss": 0.7840465169166273,
|
| 246 |
+
"eval_runtime": 89.0281,
|
| 247 |
+
"eval_samples_per_second": 11.232,
|
| 248 |
+
"eval_steps_per_second": 0.359,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.06512006512006512,
|
| 253 |
+
"grad_norm": 1.8203269243240356,
|
| 254 |
+
"learning_rate": 5.42667209333876e-08,
|
| 255 |
+
"loss": 0.7681,
|
| 256 |
+
"step": 160
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.06512006512006512,
|
| 260 |
+
"eval_cos_sim": 0.21723878383636475,
|
| 261 |
+
"eval_loss": 0.7839131169532484,
|
| 262 |
+
"eval_runtime": 88.2381,
|
| 263 |
+
"eval_samples_per_second": 11.333,
|
| 264 |
+
"eval_steps_per_second": 0.363,
|
| 265 |
+
"step": 160
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.06919006919006919,
|
| 269 |
+
"grad_norm": 1.5368891954421997,
|
| 270 |
+
"learning_rate": 5.7658390991724324e-08,
|
| 271 |
+
"loss": 0.7704,
|
| 272 |
+
"step": 170
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.06919006919006919,
|
| 276 |
+
"eval_cos_sim": 0.21737951040267944,
|
| 277 |
+
"eval_loss": 0.783772478601808,
|
| 278 |
+
"eval_runtime": 88.519,
|
| 279 |
+
"eval_samples_per_second": 11.297,
|
| 280 |
+
"eval_steps_per_second": 0.362,
|
| 281 |
+
"step": 170
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.07326007326007326,
|
| 285 |
+
"grad_norm": 1.5333787202835083,
|
| 286 |
+
"learning_rate": 6.105006105006105e-08,
|
| 287 |
+
"loss": 0.7713,
|
| 288 |
+
"step": 180
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.07326007326007326,
|
| 292 |
+
"eval_cos_sim": 0.21752700209617615,
|
| 293 |
+
"eval_loss": 0.7836251006339735,
|
| 294 |
+
"eval_runtime": 88.1346,
|
| 295 |
+
"eval_samples_per_second": 11.346,
|
| 296 |
+
"eval_steps_per_second": 0.363,
|
| 297 |
+
"step": 180
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.07733007733007732,
|
| 301 |
+
"grad_norm": 1.8334540128707886,
|
| 302 |
+
"learning_rate": 6.444173110839778e-08,
|
| 303 |
+
"loss": 0.7818,
|
| 304 |
+
"step": 190
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.07733007733007732,
|
| 308 |
+
"eval_cos_sim": 0.2176828682422638,
|
| 309 |
+
"eval_loss": 0.7834693560813611,
|
| 310 |
+
"eval_runtime": 88.9414,
|
| 311 |
+
"eval_samples_per_second": 11.243,
|
| 312 |
+
"eval_steps_per_second": 0.36,
|
| 313 |
+
"step": 190
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.0814000814000814,
|
| 317 |
+
"grad_norm": 1.7938237190246582,
|
| 318 |
+
"learning_rate": 6.78334011667345e-08,
|
| 319 |
+
"loss": 0.7756,
|
| 320 |
+
"step": 200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.0814000814000814,
|
| 324 |
+
"eval_cos_sim": 0.21784724295139313,
|
| 325 |
+
"eval_loss": 0.7833051133369154,
|
| 326 |
+
"eval_runtime": 88.7619,
|
| 327 |
+
"eval_samples_per_second": 11.266,
|
| 328 |
+
"eval_steps_per_second": 0.361,
|
| 329 |
+
"step": 200
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.08547008547008547,
|
| 333 |
+
"grad_norm": 1.8335057497024536,
|
| 334 |
+
"learning_rate": 7.122507122507124e-08,
|
| 335 |
+
"loss": 0.7742,
|
| 336 |
+
"step": 210
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.08547008547008547,
|
| 340 |
+
"eval_cos_sim": 0.21802641451358795,
|
| 341 |
+
"eval_loss": 0.7831260886405653,
|
| 342 |
+
"eval_runtime": 88.6351,
|
| 343 |
+
"eval_samples_per_second": 11.282,
|
| 344 |
+
"eval_steps_per_second": 0.361,
|
| 345 |
+
"step": 210
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.08954008954008955,
|
| 349 |
+
"grad_norm": 1.7981688976287842,
|
| 350 |
+
"learning_rate": 7.461674128340795e-08,
|
| 351 |
+
"loss": 0.7744,
|
| 352 |
+
"step": 220
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.08954008954008955,
|
| 356 |
+
"eval_cos_sim": 0.21821342408657074,
|
| 357 |
+
"eval_loss": 0.7829392380927748,
|
| 358 |
+
"eval_runtime": 88.6647,
|
| 359 |
+
"eval_samples_per_second": 11.278,
|
| 360 |
+
"eval_steps_per_second": 0.361,
|
| 361 |
+
"step": 220
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.0936100936100936,
|
| 365 |
+
"grad_norm": 1.658451795578003,
|
| 366 |
+
"learning_rate": 7.800841134174468e-08,
|
| 367 |
+
"loss": 0.7703,
|
| 368 |
+
"step": 230
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.0936100936100936,
|
| 372 |
+
"eval_cos_sim": 0.21841329336166382,
|
| 373 |
+
"eval_loss": 0.7827395224784559,
|
| 374 |
+
"eval_runtime": 88.7603,
|
| 375 |
+
"eval_samples_per_second": 11.266,
|
| 376 |
+
"eval_steps_per_second": 0.361,
|
| 377 |
+
"step": 230
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.09768009768009768,
|
| 381 |
+
"grad_norm": 1.794303297996521,
|
| 382 |
+
"learning_rate": 8.14000814000814e-08,
|
| 383 |
+
"loss": 0.7663,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.09768009768009768,
|
| 388 |
+
"eval_cos_sim": 0.21862231194972992,
|
| 389 |
+
"eval_loss": 0.782530607721681,
|
| 390 |
+
"eval_runtime": 88.7986,
|
| 391 |
+
"eval_samples_per_second": 11.261,
|
| 392 |
+
"eval_steps_per_second": 0.36,
|
| 393 |
+
"step": 240
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.10175010175010175,
|
| 397 |
+
"grad_norm": 1.7992609739303589,
|
| 398 |
+
"learning_rate": 8.479175145841813e-08,
|
| 399 |
+
"loss": 0.7736,
|
| 400 |
+
"step": 250
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.10175010175010175,
|
| 404 |
+
"eval_cos_sim": 0.21883440017700195,
|
| 405 |
+
"eval_loss": 0.7823186969970412,
|
| 406 |
+
"eval_runtime": 88.8672,
|
| 407 |
+
"eval_samples_per_second": 11.253,
|
| 408 |
+
"eval_steps_per_second": 0.36,
|
| 409 |
+
"step": 250
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.10582010582010581,
|
| 413 |
+
"grad_norm": 1.8333110809326172,
|
| 414 |
+
"learning_rate": 8.818342151675485e-08,
|
| 415 |
+
"loss": 0.7732,
|
| 416 |
+
"step": 260
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.10582010582010581,
|
| 420 |
+
"eval_cos_sim": 0.21905279159545898,
|
| 421 |
+
"eval_loss": 0.782100461981172,
|
| 422 |
+
"eval_runtime": 88.9248,
|
| 423 |
+
"eval_samples_per_second": 11.245,
|
| 424 |
+
"eval_steps_per_second": 0.36,
|
| 425 |
+
"step": 260
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.10989010989010989,
|
| 429 |
+
"grad_norm": 1.8126780986785889,
|
| 430 |
+
"learning_rate": 9.157509157509157e-08,
|
| 431 |
+
"loss": 0.7731,
|
| 432 |
+
"step": 270
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.10989010989010989,
|
| 436 |
+
"eval_cos_sim": 0.2192821353673935,
|
| 437 |
+
"eval_loss": 0.7818713002418226,
|
| 438 |
+
"eval_runtime": 89.0894,
|
| 439 |
+
"eval_samples_per_second": 11.225,
|
| 440 |
+
"eval_steps_per_second": 0.359,
|
| 441 |
+
"step": 270
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.11396011396011396,
|
| 445 |
+
"grad_norm": 1.8034828901290894,
|
| 446 |
+
"learning_rate": 9.496676163342831e-08,
|
| 447 |
+
"loss": 0.7653,
|
| 448 |
+
"step": 280
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.11396011396011396,
|
| 452 |
+
"eval_cos_sim": 0.21952229738235474,
|
| 453 |
+
"eval_loss": 0.781631359598512,
|
| 454 |
+
"eval_runtime": 88.5405,
|
| 455 |
+
"eval_samples_per_second": 11.294,
|
| 456 |
+
"eval_steps_per_second": 0.361,
|
| 457 |
+
"step": 280
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.11803011803011804,
|
| 461 |
+
"grad_norm": 1.5344951152801514,
|
| 462 |
+
"learning_rate": 9.835843169176503e-08,
|
| 463 |
+
"loss": 0.7647,
|
| 464 |
+
"step": 290
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.11803011803011804,
|
| 468 |
+
"eval_cos_sim": 0.21977314352989197,
|
| 469 |
+
"eval_loss": 0.7813806886886304,
|
| 470 |
+
"eval_runtime": 89.249,
|
| 471 |
+
"eval_samples_per_second": 11.205,
|
| 472 |
+
"eval_steps_per_second": 0.359,
|
| 473 |
+
"step": 290
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.1221001221001221,
|
| 477 |
+
"grad_norm": 1.5414931774139404,
|
| 478 |
+
"learning_rate": 1.0175010175010174e-07,
|
| 479 |
+
"loss": 0.7778,
|
| 480 |
+
"step": 300
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.1221001221001221,
|
| 484 |
+
"eval_cos_sim": 0.2200283408164978,
|
| 485 |
+
"eval_loss": 0.7811256456588453,
|
| 486 |
+
"eval_runtime": 88.215,
|
| 487 |
+
"eval_samples_per_second": 11.336,
|
| 488 |
+
"eval_steps_per_second": 0.363,
|
| 489 |
+
"step": 300
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.12617012617012616,
|
| 493 |
+
"grad_norm": 1.507014513015747,
|
| 494 |
+
"learning_rate": 1.0514177180843848e-07,
|
| 495 |
+
"loss": 0.7792,
|
| 496 |
+
"step": 310
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.12617012617012616,
|
| 500 |
+
"eval_cos_sim": 0.22028397023677826,
|
| 501 |
+
"eval_loss": 0.7808702659820265,
|
| 502 |
+
"eval_runtime": 89.0436,
|
| 503 |
+
"eval_samples_per_second": 11.23,
|
| 504 |
+
"eval_steps_per_second": 0.359,
|
| 505 |
+
"step": 310
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.13024013024013023,
|
| 509 |
+
"grad_norm": 1.8126572370529175,
|
| 510 |
+
"learning_rate": 1.085334418667752e-07,
|
| 511 |
+
"loss": 0.7647,
|
| 512 |
+
"step": 320
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.13024013024013023,
|
| 516 |
+
"eval_cos_sim": 0.22054480016231537,
|
| 517 |
+
"eval_loss": 0.7806096720908827,
|
| 518 |
+
"eval_runtime": 89.0304,
|
| 519 |
+
"eval_samples_per_second": 11.232,
|
| 520 |
+
"eval_steps_per_second": 0.359,
|
| 521 |
+
"step": 320
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.1343101343101343,
|
| 525 |
+
"grad_norm": 1.8201279640197754,
|
| 526 |
+
"learning_rate": 1.1192511192511194e-07,
|
| 527 |
+
"loss": 0.7791,
|
| 528 |
+
"step": 330
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.1343101343101343,
|
| 532 |
+
"eval_cos_sim": 0.22081315517425537,
|
| 533 |
+
"eval_loss": 0.7803415045951552,
|
| 534 |
+
"eval_runtime": 89.0654,
|
| 535 |
+
"eval_samples_per_second": 11.228,
|
| 536 |
+
"eval_steps_per_second": 0.359,
|
| 537 |
+
"step": 330
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.13838013838013838,
|
| 541 |
+
"grad_norm": 1.7973226308822632,
|
| 542 |
+
"learning_rate": 1.1531678198344865e-07,
|
| 543 |
+
"loss": 0.7672,
|
| 544 |
+
"step": 340
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.13838013838013838,
|
| 548 |
+
"eval_cos_sim": 0.22110143303871155,
|
| 549 |
+
"eval_loss": 0.7800534692023939,
|
| 550 |
+
"eval_runtime": 88.2286,
|
| 551 |
+
"eval_samples_per_second": 11.334,
|
| 552 |
+
"eval_steps_per_second": 0.363,
|
| 553 |
+
"step": 340
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 0.14245014245014245,
|
| 557 |
+
"grad_norm": 1.8215441703796387,
|
| 558 |
+
"learning_rate": 1.1870845204178537e-07,
|
| 559 |
+
"loss": 0.7699,
|
| 560 |
+
"step": 350
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.14245014245014245,
|
| 564 |
+
"eval_cos_sim": 0.22139768302440643,
|
| 565 |
+
"eval_loss": 0.7797573957656568,
|
| 566 |
+
"eval_runtime": 88.8585,
|
| 567 |
+
"eval_samples_per_second": 11.254,
|
| 568 |
+
"eval_steps_per_second": 0.36,
|
| 569 |
+
"step": 350
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.14652014652014653,
|
| 573 |
+
"grad_norm": 1.8159250020980835,
|
| 574 |
+
"learning_rate": 1.221001221001221e-07,
|
| 575 |
+
"loss": 0.7712,
|
| 576 |
+
"step": 360
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.14652014652014653,
|
| 580 |
+
"eval_cos_sim": 0.2217043787240982,
|
| 581 |
+
"eval_loss": 0.7794509191726393,
|
| 582 |
+
"eval_runtime": 88.9767,
|
| 583 |
+
"eval_samples_per_second": 11.239,
|
| 584 |
+
"eval_steps_per_second": 0.36,
|
| 585 |
+
"step": 360
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.1505901505901506,
|
| 589 |
+
"grad_norm": 1.7892403602600098,
|
| 590 |
+
"learning_rate": 1.2549179215845883e-07,
|
| 591 |
+
"loss": 0.7594,
|
| 592 |
+
"step": 370
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.1505901505901506,
|
| 596 |
+
"eval_cos_sim": 0.22202090919017792,
|
| 597 |
+
"eval_loss": 0.7791346249793715,
|
| 598 |
+
"eval_runtime": 88.4113,
|
| 599 |
+
"eval_samples_per_second": 11.311,
|
| 600 |
+
"eval_steps_per_second": 0.362,
|
| 601 |
+
"step": 370
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.15466015466015465,
|
| 605 |
+
"grad_norm": 1.499679446220398,
|
| 606 |
+
"learning_rate": 1.2888346221679555e-07,
|
| 607 |
+
"loss": 0.7702,
|
| 608 |
+
"step": 380
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.15466015466015465,
|
| 612 |
+
"eval_cos_sim": 0.22234594821929932,
|
| 613 |
+
"eval_loss": 0.7788098197196669,
|
| 614 |
+
"eval_runtime": 89.1927,
|
| 615 |
+
"eval_samples_per_second": 11.212,
|
| 616 |
+
"eval_steps_per_second": 0.359,
|
| 617 |
+
"step": 380
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.15873015873015872,
|
| 621 |
+
"grad_norm": 1.827500343322754,
|
| 622 |
+
"learning_rate": 1.3227513227513228e-07,
|
| 623 |
+
"loss": 0.7578,
|
| 624 |
+
"step": 390
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.15873015873015872,
|
| 628 |
+
"eval_cos_sim": 0.22267462313175201,
|
| 629 |
+
"eval_loss": 0.7784813823913282,
|
| 630 |
+
"eval_runtime": 89.2341,
|
| 631 |
+
"eval_samples_per_second": 11.206,
|
| 632 |
+
"eval_steps_per_second": 0.359,
|
| 633 |
+
"step": 390
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.1628001628001628,
|
| 637 |
+
"grad_norm": 1.8155056238174438,
|
| 638 |
+
"learning_rate": 1.35666802333469e-07,
|
| 639 |
+
"loss": 0.7663,
|
| 640 |
+
"step": 400
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.1628001628001628,
|
| 644 |
+
"eval_cos_sim": 0.2230025976896286,
|
| 645 |
+
"eval_loss": 0.7781536517356581,
|
| 646 |
+
"eval_runtime": 89.0874,
|
| 647 |
+
"eval_samples_per_second": 11.225,
|
| 648 |
+
"eval_steps_per_second": 0.359,
|
| 649 |
+
"step": 400
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.16687016687016687,
|
| 653 |
+
"grad_norm": 1.5147600173950195,
|
| 654 |
+
"learning_rate": 1.3905847239180572e-07,
|
| 655 |
+
"loss": 0.7685,
|
| 656 |
+
"step": 410
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.16687016687016687,
|
| 660 |
+
"eval_cos_sim": 0.2233392745256424,
|
| 661 |
+
"eval_loss": 0.7778171930526442,
|
| 662 |
+
"eval_runtime": 89.0688,
|
| 663 |
+
"eval_samples_per_second": 11.227,
|
| 664 |
+
"eval_steps_per_second": 0.359,
|
| 665 |
+
"step": 410
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.17094017094017094,
|
| 669 |
+
"grad_norm": 1.5169967412948608,
|
| 670 |
+
"learning_rate": 1.4245014245014247e-07,
|
| 671 |
+
"loss": 0.7614,
|
| 672 |
+
"step": 420
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 0.17094017094017094,
|
| 676 |
+
"eval_cos_sim": 0.2236924022436142,
|
| 677 |
+
"eval_loss": 0.7774643163894361,
|
| 678 |
+
"eval_runtime": 88.9798,
|
| 679 |
+
"eval_samples_per_second": 11.239,
|
| 680 |
+
"eval_steps_per_second": 0.36,
|
| 681 |
+
"step": 420
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.17501017501017502,
|
| 685 |
+
"grad_norm": 1.5205143690109253,
|
| 686 |
+
"learning_rate": 1.4584181250847917e-07,
|
| 687 |
+
"loss": 0.7646,
|
| 688 |
+
"step": 430
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.17501017501017502,
|
| 692 |
+
"eval_cos_sim": 0.224058136343956,
|
| 693 |
+
"eval_loss": 0.7770988393043227,
|
| 694 |
+
"eval_runtime": 89.0477,
|
| 695 |
+
"eval_samples_per_second": 11.23,
|
| 696 |
+
"eval_steps_per_second": 0.359,
|
| 697 |
+
"step": 430
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.1790801790801791,
|
| 701 |
+
"grad_norm": 1.5187054872512817,
|
| 702 |
+
"learning_rate": 1.492334825668159e-07,
|
| 703 |
+
"loss": 0.775,
|
| 704 |
+
"step": 440
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.1790801790801791,
|
| 708 |
+
"eval_cos_sim": 0.22443027794361115,
|
| 709 |
+
"eval_loss": 0.7767269988273329,
|
| 710 |
+
"eval_runtime": 89.3697,
|
| 711 |
+
"eval_samples_per_second": 11.189,
|
| 712 |
+
"eval_steps_per_second": 0.358,
|
| 713 |
+
"step": 440
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.18315018315018314,
|
| 717 |
+
"grad_norm": 1.8240047693252563,
|
| 718 |
+
"learning_rate": 1.5262515262515264e-07,
|
| 719 |
+
"loss": 0.7622,
|
| 720 |
+
"step": 450
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.18315018315018314,
|
| 724 |
+
"eval_cos_sim": 0.22480851411819458,
|
| 725 |
+
"eval_loss": 0.7763490748618788,
|
| 726 |
+
"eval_runtime": 89.2014,
|
| 727 |
+
"eval_samples_per_second": 11.211,
|
| 728 |
+
"eval_steps_per_second": 0.359,
|
| 729 |
+
"step": 450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.1872201872201872,
|
| 733 |
+
"grad_norm": 1.8208777904510498,
|
| 734 |
+
"learning_rate": 1.5601682268348936e-07,
|
| 735 |
+
"loss": 0.7563,
|
| 736 |
+
"step": 460
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 0.1872201872201872,
|
| 740 |
+
"eval_cos_sim": 0.2251981645822525,
|
| 741 |
+
"eval_loss": 0.7759597497199721,
|
| 742 |
+
"eval_runtime": 89.202,
|
| 743 |
+
"eval_samples_per_second": 11.211,
|
| 744 |
+
"eval_steps_per_second": 0.359,
|
| 745 |
+
"step": 460
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.19129019129019129,
|
| 749 |
+
"grad_norm": 1.523047924041748,
|
| 750 |
+
"learning_rate": 1.594084927418261e-07,
|
| 751 |
+
"loss": 0.7613,
|
| 752 |
+
"step": 470
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.19129019129019129,
|
| 756 |
+
"eval_cos_sim": 0.2256050556898117,
|
| 757 |
+
"eval_loss": 0.775553204080934,
|
| 758 |
+
"eval_runtime": 88.9659,
|
| 759 |
+
"eval_samples_per_second": 11.24,
|
| 760 |
+
"eval_steps_per_second": 0.36,
|
| 761 |
+
"step": 470
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.19536019536019536,
|
| 765 |
+
"grad_norm": 1.7896461486816406,
|
| 766 |
+
"learning_rate": 1.628001628001628e-07,
|
| 767 |
+
"loss": 0.7571,
|
| 768 |
+
"step": 480
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.19536019536019536,
|
| 772 |
+
"eval_cos_sim": 0.2260132133960724,
|
| 773 |
+
"eval_loss": 0.7751453356956189,
|
| 774 |
+
"eval_runtime": 89.3097,
|
| 775 |
+
"eval_samples_per_second": 11.197,
|
| 776 |
+
"eval_steps_per_second": 0.358,
|
| 777 |
+
"step": 480
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.19943019943019943,
|
| 781 |
+
"grad_norm": 1.802852749824524,
|
| 782 |
+
"learning_rate": 1.6619183285849953e-07,
|
| 783 |
+
"loss": 0.7666,
|
| 784 |
+
"step": 490
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 0.19943019943019943,
|
| 788 |
+
"eval_cos_sim": 0.2264304757118225,
|
| 789 |
+
"eval_loss": 0.7747283859466261,
|
| 790 |
+
"eval_runtime": 89.0947,
|
| 791 |
+
"eval_samples_per_second": 11.224,
|
| 792 |
+
"eval_steps_per_second": 0.359,
|
| 793 |
+
"step": 490
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.2035002035002035,
|
| 797 |
+
"grad_norm": 1.821495532989502,
|
| 798 |
+
"learning_rate": 1.6958350291683626e-07,
|
| 799 |
+
"loss": 0.7624,
|
| 800 |
+
"step": 500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.2035002035002035,
|
| 804 |
+
"eval_cos_sim": 0.22685033082962036,
|
| 805 |
+
"eval_loss": 0.7743088460181898,
|
| 806 |
+
"eval_runtime": 89.2762,
|
| 807 |
+
"eval_samples_per_second": 11.201,
|
| 808 |
+
"eval_steps_per_second": 0.358,
|
| 809 |
+
"step": 500
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.20757020757020758,
|
| 813 |
+
"grad_norm": 1.545665979385376,
|
| 814 |
+
"learning_rate": 1.7297517297517298e-07,
|
| 815 |
+
"loss": 0.7694,
|
| 816 |
+
"step": 510
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.20757020757020758,
|
| 820 |
+
"eval_cos_sim": 0.22728146612644196,
|
| 821 |
+
"eval_loss": 0.7738780179237074,
|
| 822 |
+
"eval_runtime": 88.7471,
|
| 823 |
+
"eval_samples_per_second": 11.268,
|
| 824 |
+
"eval_steps_per_second": 0.361,
|
| 825 |
+
"step": 510
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.21164021164021163,
|
| 829 |
+
"grad_norm": 1.7993839979171753,
|
| 830 |
+
"learning_rate": 1.763668430335097e-07,
|
| 831 |
+
"loss": 0.7601,
|
| 832 |
+
"step": 520
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 0.21164021164021163,
|
| 836 |
+
"eval_cos_sim": 0.2277180701494217,
|
| 837 |
+
"eval_loss": 0.7734417824958509,
|
| 838 |
+
"eval_runtime": 89.3566,
|
| 839 |
+
"eval_samples_per_second": 11.191,
|
| 840 |
+
"eval_steps_per_second": 0.358,
|
| 841 |
+
"step": 520
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.2157102157102157,
|
| 845 |
+
"grad_norm": 1.7981739044189453,
|
| 846 |
+
"learning_rate": 1.7975851309184642e-07,
|
| 847 |
+
"loss": 0.7572,
|
| 848 |
+
"step": 530
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.2157102157102157,
|
| 852 |
+
"eval_cos_sim": 0.22816509008407593,
|
| 853 |
+
"eval_loss": 0.7729950847839063,
|
| 854 |
+
"eval_runtime": 89.1907,
|
| 855 |
+
"eval_samples_per_second": 11.212,
|
| 856 |
+
"eval_steps_per_second": 0.359,
|
| 857 |
+
"step": 530
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.21978021978021978,
|
| 861 |
+
"grad_norm": 1.4281421899795532,
|
| 862 |
+
"learning_rate": 1.8315018315018315e-07,
|
| 863 |
+
"loss": 0.7713,
|
| 864 |
+
"step": 540
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.21978021978021978,
|
| 868 |
+
"eval_cos_sim": 0.2286159247159958,
|
| 869 |
+
"eval_loss": 0.7725445466254895,
|
| 870 |
+
"eval_runtime": 88.9936,
|
| 871 |
+
"eval_samples_per_second": 11.237,
|
| 872 |
+
"eval_steps_per_second": 0.36,
|
| 873 |
+
"step": 540
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.22385022385022385,
|
| 877 |
+
"grad_norm": 1.800079107284546,
|
| 878 |
+
"learning_rate": 1.865418532085199e-07,
|
| 879 |
+
"loss": 0.7637,
|
| 880 |
+
"step": 550
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.22385022385022385,
|
| 884 |
+
"eval_cos_sim": 0.22906461358070374,
|
| 885 |
+
"eval_loss": 0.772096163771028,
|
| 886 |
+
"eval_runtime": 88.9615,
|
| 887 |
+
"eval_samples_per_second": 11.241,
|
| 888 |
+
"eval_steps_per_second": 0.36,
|
| 889 |
+
"step": 550
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.22792022792022792,
|
| 893 |
+
"grad_norm": 1.5776735544204712,
|
| 894 |
+
"learning_rate": 1.8993352326685662e-07,
|
| 895 |
+
"loss": 0.7699,
|
| 896 |
+
"step": 560
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.22792022792022792,
|
| 900 |
+
"eval_cos_sim": 0.2295254021883011,
|
| 901 |
+
"eval_loss": 0.771635666391725,
|
| 902 |
+
"eval_runtime": 88.9524,
|
| 903 |
+
"eval_samples_per_second": 11.242,
|
| 904 |
+
"eval_steps_per_second": 0.36,
|
| 905 |
+
"step": 560
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.231990231990232,
|
| 909 |
+
"grad_norm": 1.8183437585830688,
|
| 910 |
+
"learning_rate": 1.9332519332519332e-07,
|
| 911 |
+
"loss": 0.7634,
|
| 912 |
+
"step": 570
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.231990231990232,
|
| 916 |
+
"eval_cos_sim": 0.22998037934303284,
|
| 917 |
+
"eval_loss": 0.7711810879920667,
|
| 918 |
+
"eval_runtime": 88.9474,
|
| 919 |
+
"eval_samples_per_second": 11.243,
|
| 920 |
+
"eval_steps_per_second": 0.36,
|
| 921 |
+
"step": 570
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.23606023606023607,
|
| 925 |
+
"grad_norm": 1.8223985433578491,
|
| 926 |
+
"learning_rate": 1.9671686338353007e-07,
|
| 927 |
+
"loss": 0.7772,
|
| 928 |
+
"step": 580
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.23606023606023607,
|
| 932 |
+
"eval_cos_sim": 0.23044081032276154,
|
| 933 |
+
"eval_loss": 0.7707209568237012,
|
| 934 |
+
"eval_runtime": 88.7661,
|
| 935 |
+
"eval_samples_per_second": 11.266,
|
| 936 |
+
"eval_steps_per_second": 0.36,
|
| 937 |
+
"step": 580
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.24013024013024012,
|
| 941 |
+
"grad_norm": 1.7877051830291748,
|
| 942 |
+
"learning_rate": 2.001085334418668e-07,
|
| 943 |
+
"loss": 0.7534,
|
| 944 |
+
"step": 590
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.24013024013024012,
|
| 948 |
+
"eval_cos_sim": 0.23091764748096466,
|
| 949 |
+
"eval_loss": 0.770244472048158,
|
| 950 |
+
"eval_runtime": 89.1681,
|
| 951 |
+
"eval_samples_per_second": 11.215,
|
| 952 |
+
"eval_steps_per_second": 0.359,
|
| 953 |
+
"step": 590
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.2442002442002442,
|
| 957 |
+
"grad_norm": 1.6694035530090332,
|
| 958 |
+
"learning_rate": 2.0350020350020349e-07,
|
| 959 |
+
"loss": 0.764,
|
| 960 |
+
"step": 600
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.2442002442002442,
|
| 964 |
+
"eval_cos_sim": 0.23141798377037048,
|
| 965 |
+
"eval_loss": 0.769744483969087,
|
| 966 |
+
"eval_runtime": 89.1824,
|
| 967 |
+
"eval_samples_per_second": 11.213,
|
| 968 |
+
"eval_steps_per_second": 0.359,
|
| 969 |
+
"step": 600
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.24827024827024827,
|
| 973 |
+
"grad_norm": 1.5185260772705078,
|
| 974 |
+
"learning_rate": 2.0689187355854024e-07,
|
| 975 |
+
"loss": 0.7641,
|
| 976 |
+
"step": 610
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.24827024827024827,
|
| 980 |
+
"eval_cos_sim": 0.2319241613149643,
|
| 981 |
+
"eval_loss": 0.7692386307929701,
|
| 982 |
+
"eval_runtime": 89.2776,
|
| 983 |
+
"eval_samples_per_second": 11.201,
|
| 984 |
+
"eval_steps_per_second": 0.358,
|
| 985 |
+
"step": 610
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.2523402523402523,
|
| 989 |
+
"grad_norm": 1.8109288215637207,
|
| 990 |
+
"learning_rate": 2.1028354361687696e-07,
|
| 991 |
+
"loss": 0.7554,
|
| 992 |
+
"step": 620
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.2523402523402523,
|
| 996 |
+
"eval_cos_sim": 0.23244358599185944,
|
| 997 |
+
"eval_loss": 0.7687196593497938,
|
| 998 |
+
"eval_runtime": 89.197,
|
| 999 |
+
"eval_samples_per_second": 11.211,
|
| 1000 |
+
"eval_steps_per_second": 0.359,
|
| 1001 |
+
"step": 620
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.2564102564102564,
|
| 1005 |
+
"grad_norm": 1.8006572723388672,
|
| 1006 |
+
"learning_rate": 2.136752136752137e-07,
|
| 1007 |
+
"loss": 0.7547,
|
| 1008 |
+
"step": 630
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.2564102564102564,
|
| 1012 |
+
"eval_cos_sim": 0.23296819627285004,
|
| 1013 |
+
"eval_loss": 0.7681954999183362,
|
| 1014 |
+
"eval_runtime": 89.2344,
|
| 1015 |
+
"eval_samples_per_second": 11.206,
|
| 1016 |
+
"eval_steps_per_second": 0.359,
|
| 1017 |
+
"step": 630
|
| 1018 |
+
}
|
| 1019 |
+
],
|
| 1020 |
+
"logging_steps": 10,
|
| 1021 |
+
"max_steps": 1474200,
|
| 1022 |
+
"num_input_tokens_seen": 0,
|
| 1023 |
+
"num_train_epochs": 600,
|
| 1024 |
+
"save_steps": 10,
|
| 1025 |
+
"total_flos": 0.0,
|
| 1026 |
+
"train_batch_size": 160,
|
| 1027 |
+
"trial_name": null,
|
| 1028 |
+
"trial_params": null
|
| 1029 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e11ac20990094aa44815e972caf40122c350eef4fa83fb84ce5c74c5b7eab79b
|
| 3 |
+
size 5048
|