Training in progress, epoch 0
Browse files- adapter_config.json +3 -3
- adapter_model.safetensors +1 -1
- trainer_state.json +1066 -0
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -20,12 +20,12 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"
|
| 24 |
"q_proj",
|
|
|
|
| 25 |
"gate_proj",
|
|
|
|
| 26 |
"o_proj",
|
| 27 |
-
"down_proj",
|
| 28 |
-
"up_proj",
|
| 29 |
"k_proj"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
+
"up_proj",
|
| 24 |
"q_proj",
|
| 25 |
+
"down_proj",
|
| 26 |
"gate_proj",
|
| 27 |
+
"v_proj",
|
| 28 |
"o_proj",
|
|
|
|
|
|
|
| 29 |
"k_proj"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 78480072
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:13c7d8dae2b14b8e07ed03218fb9a5c6cc7fff8a700bd56ef3bebdb16d1c50fa
|
| 3 |
size 78480072
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1066 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.949579831932773,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1480,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.06722689075630252,
|
| 13 |
+
"grad_norm": 3.575806140899658,
|
| 14 |
+
"learning_rate": 0.00019997747161747695,
|
| 15 |
+
"loss": 1.5631,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.13445378151260504,
|
| 20 |
+
"grad_norm": 0.8588754534721375,
|
| 21 |
+
"learning_rate": 0.00019990989662046818,
|
| 22 |
+
"loss": 1.1154,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.20168067226890757,
|
| 27 |
+
"grad_norm": 0.7627600431442261,
|
| 28 |
+
"learning_rate": 0.00019979730545608126,
|
| 29 |
+
"loss": 0.9953,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.2689075630252101,
|
| 34 |
+
"grad_norm": 0.6456862092018127,
|
| 35 |
+
"learning_rate": 0.00019963974885425266,
|
| 36 |
+
"loss": 0.9508,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.33613445378151263,
|
| 41 |
+
"grad_norm": 0.6212641000747681,
|
| 42 |
+
"learning_rate": 0.00019943729780489027,
|
| 43 |
+
"loss": 0.9451,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.40336134453781514,
|
| 48 |
+
"grad_norm": 0.6262118816375732,
|
| 49 |
+
"learning_rate": 0.00019919004352588767,
|
| 50 |
+
"loss": 0.9302,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.47058823529411764,
|
| 55 |
+
"grad_norm": 0.5837720036506653,
|
| 56 |
+
"learning_rate": 0.00019889809742202455,
|
| 57 |
+
"loss": 0.909,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.5378151260504201,
|
| 62 |
+
"grad_norm": 0.6171330809593201,
|
| 63 |
+
"learning_rate": 0.00019856159103477086,
|
| 64 |
+
"loss": 0.9109,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.6050420168067226,
|
| 69 |
+
"grad_norm": 0.6107902526855469,
|
| 70 |
+
"learning_rate": 0.0001981806759830189,
|
| 71 |
+
"loss": 0.8952,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.6722689075630253,
|
| 76 |
+
"grad_norm": 0.5945306420326233,
|
| 77 |
+
"learning_rate": 0.00019775552389476864,
|
| 78 |
+
"loss": 0.8968,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.7394957983193278,
|
| 83 |
+
"grad_norm": 0.5847243666648865,
|
| 84 |
+
"learning_rate": 0.00019728632632979746,
|
| 85 |
+
"loss": 0.8696,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.8067226890756303,
|
| 90 |
+
"grad_norm": 0.5536367893218994,
|
| 91 |
+
"learning_rate": 0.0001967732946933499,
|
| 92 |
+
"loss": 0.8575,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.8739495798319328,
|
| 97 |
+
"grad_norm": 0.6354172825813293,
|
| 98 |
+
"learning_rate": 0.00019621666014088494,
|
| 99 |
+
"loss": 0.8429,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.9411764705882353,
|
| 104 |
+
"grad_norm": 0.5658561587333679,
|
| 105 |
+
"learning_rate": 0.00019561667347392508,
|
| 106 |
+
"loss": 0.8376,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.0084033613445378,
|
| 111 |
+
"grad_norm": 0.5443424582481384,
|
| 112 |
+
"learning_rate": 0.0001949736050270532,
|
| 113 |
+
"loss": 0.834,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.0756302521008403,
|
| 118 |
+
"grad_norm": 0.6456009745597839,
|
| 119 |
+
"learning_rate": 0.00019428774454610843,
|
| 120 |
+
"loss": 0.7672,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.1428571428571428,
|
| 125 |
+
"grad_norm": 0.7190167903900146,
|
| 126 |
+
"learning_rate": 0.0001935594010576362,
|
| 127 |
+
"loss": 0.7292,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.2100840336134453,
|
| 132 |
+
"grad_norm": 0.628425657749176,
|
| 133 |
+
"learning_rate": 0.00019278890272965096,
|
| 134 |
+
"loss": 0.7646,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.2773109243697478,
|
| 139 |
+
"grad_norm": 0.6629491448402405,
|
| 140 |
+
"learning_rate": 0.0001919765967237739,
|
| 141 |
+
"loss": 0.7627,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.3445378151260505,
|
| 146 |
+
"grad_norm": 0.6584550738334656,
|
| 147 |
+
"learning_rate": 0.0001911228490388136,
|
| 148 |
+
"loss": 0.7632,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.4117647058823528,
|
| 153 |
+
"grad_norm": 0.7824658751487732,
|
| 154 |
+
"learning_rate": 0.00019022804434585852,
|
| 155 |
+
"loss": 0.7644,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.4789915966386555,
|
| 160 |
+
"grad_norm": 0.6643643379211426,
|
| 161 |
+
"learning_rate": 0.00018929258581495685,
|
| 162 |
+
"loss": 0.7455,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.5462184873949578,
|
| 167 |
+
"grad_norm": 0.615719199180603,
|
| 168 |
+
"learning_rate": 0.00018831689493346095,
|
| 169 |
+
"loss": 0.7807,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.6134453781512605,
|
| 174 |
+
"grad_norm": 0.6157252192497253,
|
| 175 |
+
"learning_rate": 0.00018730141131611882,
|
| 176 |
+
"loss": 0.7355,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.680672268907563,
|
| 181 |
+
"grad_norm": 0.6899864673614502,
|
| 182 |
+
"learning_rate": 0.00018624659250699805,
|
| 183 |
+
"loss": 0.7544,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.7478991596638656,
|
| 188 |
+
"grad_norm": 0.5952817797660828,
|
| 189 |
+
"learning_rate": 0.00018515291377333112,
|
| 190 |
+
"loss": 0.7484,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.815126050420168,
|
| 195 |
+
"grad_norm": 0.639409601688385,
|
| 196 |
+
"learning_rate": 0.00018402086789137546,
|
| 197 |
+
"loss": 0.7586,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.8823529411764706,
|
| 202 |
+
"grad_norm": 0.6305294036865234,
|
| 203 |
+
"learning_rate": 0.00018285096492438424,
|
| 204 |
+
"loss": 0.7429,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.949579831932773,
|
| 209 |
+
"grad_norm": 0.6956226825714111,
|
| 210 |
+
"learning_rate": 0.00018164373199278856,
|
| 211 |
+
"loss": 0.7446,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.0168067226890756,
|
| 216 |
+
"grad_norm": 0.6213023662567139,
|
| 217 |
+
"learning_rate": 0.00018039971303669407,
|
| 218 |
+
"loss": 0.7035,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 2.0840336134453783,
|
| 223 |
+
"grad_norm": 0.7370445728302002,
|
| 224 |
+
"learning_rate": 0.00017911946857079888,
|
| 225 |
+
"loss": 0.6189,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 2.1512605042016806,
|
| 230 |
+
"grad_norm": 0.7824037075042725,
|
| 231 |
+
"learning_rate": 0.00017780357543184397,
|
| 232 |
+
"loss": 0.6206,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.2184873949579833,
|
| 237 |
+
"grad_norm": 0.792977511882782,
|
| 238 |
+
"learning_rate": 0.00017645262651870926,
|
| 239 |
+
"loss": 0.6176,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 2.2857142857142856,
|
| 244 |
+
"grad_norm": 0.7439864873886108,
|
| 245 |
+
"learning_rate": 0.00017506723052527242,
|
| 246 |
+
"loss": 0.621,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 2.3529411764705883,
|
| 251 |
+
"grad_norm": 0.833249032497406,
|
| 252 |
+
"learning_rate": 0.00017364801166615124,
|
| 253 |
+
"loss": 0.6165,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.4201680672268906,
|
| 258 |
+
"grad_norm": 0.743938684463501,
|
| 259 |
+
"learning_rate": 0.00017219560939545246,
|
| 260 |
+
"loss": 0.6062,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 2.4873949579831933,
|
| 265 |
+
"grad_norm": 0.8116819858551025,
|
| 266 |
+
"learning_rate": 0.00017071067811865476,
|
| 267 |
+
"loss": 0.6283,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 2.5546218487394956,
|
| 272 |
+
"grad_norm": 0.8074822425842285,
|
| 273 |
+
"learning_rate": 0.00016919388689775464,
|
| 274 |
+
"loss": 0.651,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 2.6218487394957983,
|
| 279 |
+
"grad_norm": 0.8415045738220215,
|
| 280 |
+
"learning_rate": 0.0001676459191498087,
|
| 281 |
+
"loss": 0.6388,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 2.689075630252101,
|
| 286 |
+
"grad_norm": 0.727366030216217,
|
| 287 |
+
"learning_rate": 0.00016606747233900815,
|
| 288 |
+
"loss": 0.6067,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 2.7563025210084033,
|
| 293 |
+
"grad_norm": 0.7842895984649658,
|
| 294 |
+
"learning_rate": 0.00016445925766242391,
|
| 295 |
+
"loss": 0.6408,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.8235294117647056,
|
| 300 |
+
"grad_norm": 0.7419390082359314,
|
| 301 |
+
"learning_rate": 0.00016282199972956425,
|
| 302 |
+
"loss": 0.6388,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 2.8907563025210083,
|
| 307 |
+
"grad_norm": 0.7776601910591125,
|
| 308 |
+
"learning_rate": 0.00016115643623588915,
|
| 309 |
+
"loss": 0.6318,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 2.957983193277311,
|
| 314 |
+
"grad_norm": 0.8350630402565002,
|
| 315 |
+
"learning_rate": 0.00015946331763042867,
|
| 316 |
+
"loss": 0.6174,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 3.0252100840336134,
|
| 321 |
+
"grad_norm": 0.7818737030029297,
|
| 322 |
+
"learning_rate": 0.0001577434067776548,
|
| 323 |
+
"loss": 0.5877,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 3.092436974789916,
|
| 328 |
+
"grad_norm": 0.9094923138618469,
|
| 329 |
+
"learning_rate": 0.00015599747861375955,
|
| 330 |
+
"loss": 0.4781,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 3.1596638655462184,
|
| 335 |
+
"grad_norm": 0.9884558320045471,
|
| 336 |
+
"learning_rate": 0.00015422631979749354,
|
| 337 |
+
"loss": 0.475,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.226890756302521,
|
| 342 |
+
"grad_norm": 0.9660643339157104,
|
| 343 |
+
"learning_rate": 0.00015243072835572318,
|
| 344 |
+
"loss": 0.4689,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 3.2941176470588234,
|
| 349 |
+
"grad_norm": 0.9638343453407288,
|
| 350 |
+
"learning_rate": 0.00015061151332386566,
|
| 351 |
+
"loss": 0.4726,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 3.361344537815126,
|
| 356 |
+
"grad_norm": 0.9662260413169861,
|
| 357 |
+
"learning_rate": 0.00014876949438136347,
|
| 358 |
+
"loss": 0.5037,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 3.4285714285714284,
|
| 363 |
+
"grad_norm": 0.9045869708061218,
|
| 364 |
+
"learning_rate": 0.0001469055014823637,
|
| 365 |
+
"loss": 0.4875,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 3.495798319327731,
|
| 370 |
+
"grad_norm": 0.9099442958831787,
|
| 371 |
+
"learning_rate": 0.00014502037448176734,
|
| 372 |
+
"loss": 0.5047,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 3.5630252100840334,
|
| 377 |
+
"grad_norm": 0.9331455826759338,
|
| 378 |
+
"learning_rate": 0.00014311496275681783,
|
| 379 |
+
"loss": 0.4929,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 3.630252100840336,
|
| 384 |
+
"grad_norm": 1.0234043598175049,
|
| 385 |
+
"learning_rate": 0.0001411901248243993,
|
| 386 |
+
"loss": 0.5033,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 3.697478991596639,
|
| 391 |
+
"grad_norm": 0.9315100312232971,
|
| 392 |
+
"learning_rate": 0.00013924672795421637,
|
| 393 |
+
"loss": 0.4925,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 3.764705882352941,
|
| 398 |
+
"grad_norm": 1.0947026014328003,
|
| 399 |
+
"learning_rate": 0.00013728564777803088,
|
| 400 |
+
"loss": 0.5003,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 3.831932773109244,
|
| 405 |
+
"grad_norm": 0.9220852255821228,
|
| 406 |
+
"learning_rate": 0.0001353077678951301,
|
| 407 |
+
"loss": 0.51,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 3.899159663865546,
|
| 412 |
+
"grad_norm": 1.065261721611023,
|
| 413 |
+
"learning_rate": 0.00013331397947420576,
|
| 414 |
+
"loss": 0.5091,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 3.966386554621849,
|
| 419 |
+
"grad_norm": 0.8783352971076965,
|
| 420 |
+
"learning_rate": 0.00013130518085182225,
|
| 421 |
+
"loss": 0.5036,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 4.033613445378151,
|
| 426 |
+
"grad_norm": 0.9211850166320801,
|
| 427 |
+
"learning_rate": 0.00012928227712765504,
|
| 428 |
+
"loss": 0.4408,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 4.100840336134453,
|
| 433 |
+
"grad_norm": 1.0706185102462769,
|
| 434 |
+
"learning_rate": 0.0001272461797566823,
|
| 435 |
+
"loss": 0.3445,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 4.168067226890757,
|
| 440 |
+
"grad_norm": 1.152872920036316,
|
| 441 |
+
"learning_rate": 0.00012519780613851254,
|
| 442 |
+
"loss": 0.3399,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 4.235294117647059,
|
| 447 |
+
"grad_norm": 1.1418325901031494,
|
| 448 |
+
"learning_rate": 0.00012313807920403419,
|
| 449 |
+
"loss": 0.3516,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 4.302521008403361,
|
| 454 |
+
"grad_norm": 1.178936243057251,
|
| 455 |
+
"learning_rate": 0.00012106792699957263,
|
| 456 |
+
"loss": 0.3605,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 4.369747899159664,
|
| 461 |
+
"grad_norm": 0.9696270823478699,
|
| 462 |
+
"learning_rate": 0.00011898828226874284,
|
| 463 |
+
"loss": 0.3519,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 4.436974789915967,
|
| 468 |
+
"grad_norm": 1.0649495124816895,
|
| 469 |
+
"learning_rate": 0.00011690008203218493,
|
| 470 |
+
"loss": 0.3534,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 4.504201680672269,
|
| 475 |
+
"grad_norm": 1.114478588104248,
|
| 476 |
+
"learning_rate": 0.00011480426716537315,
|
| 477 |
+
"loss": 0.3624,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 4.571428571428571,
|
| 482 |
+
"grad_norm": 1.2927486896514893,
|
| 483 |
+
"learning_rate": 0.00011270178197468789,
|
| 484 |
+
"loss": 0.3614,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 4.6386554621848735,
|
| 489 |
+
"grad_norm": 1.2928136587142944,
|
| 490 |
+
"learning_rate": 0.00011059357377194161,
|
| 491 |
+
"loss": 0.3616,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 4.705882352941177,
|
| 496 |
+
"grad_norm": 1.0552875995635986,
|
| 497 |
+
"learning_rate": 0.00010848059244755093,
|
| 498 |
+
"loss": 0.3666,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 4.773109243697479,
|
| 503 |
+
"grad_norm": 1.0149446725845337,
|
| 504 |
+
"learning_rate": 0.00010636379004254664,
|
| 505 |
+
"loss": 0.3747,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 4.840336134453781,
|
| 510 |
+
"grad_norm": 1.09207022190094,
|
| 511 |
+
"learning_rate": 0.00010424412031961484,
|
| 512 |
+
"loss": 0.3589,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 4.907563025210084,
|
| 517 |
+
"grad_norm": 0.8984358906745911,
|
| 518 |
+
"learning_rate": 0.00010212253833336237,
|
| 519 |
+
"loss": 0.3671,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 4.974789915966387,
|
| 524 |
+
"grad_norm": 1.0369324684143066,
|
| 525 |
+
"learning_rate": 0.0001,
|
| 526 |
+
"loss": 0.3785,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 5.042016806722689,
|
| 531 |
+
"grad_norm": 1.0416890382766724,
|
| 532 |
+
"learning_rate": 9.787746166663764e-05,
|
| 533 |
+
"loss": 0.3085,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 5.109243697478991,
|
| 538 |
+
"grad_norm": 0.8845974802970886,
|
| 539 |
+
"learning_rate": 9.57558796803852e-05,
|
| 540 |
+
"loss": 0.2516,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 5.176470588235294,
|
| 545 |
+
"grad_norm": 1.058775544166565,
|
| 546 |
+
"learning_rate": 9.363620995745337e-05,
|
| 547 |
+
"loss": 0.2539,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 5.243697478991597,
|
| 552 |
+
"grad_norm": 1.0327303409576416,
|
| 553 |
+
"learning_rate": 9.151940755244912e-05,
|
| 554 |
+
"loss": 0.2488,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 5.310924369747899,
|
| 559 |
+
"grad_norm": 0.9804151654243469,
|
| 560 |
+
"learning_rate": 8.94064262280584e-05,
|
| 561 |
+
"loss": 0.2538,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.378151260504202,
|
| 566 |
+
"grad_norm": 0.9500736594200134,
|
| 567 |
+
"learning_rate": 8.729821802531212e-05,
|
| 568 |
+
"loss": 0.2536,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 5.445378151260504,
|
| 573 |
+
"grad_norm": 1.1535006761550903,
|
| 574 |
+
"learning_rate": 8.519573283462687e-05,
|
| 575 |
+
"loss": 0.2544,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.512605042016807,
|
| 580 |
+
"grad_norm": 0.8586835861206055,
|
| 581 |
+
"learning_rate": 8.309991796781511e-05,
|
| 582 |
+
"loss": 0.2636,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 5.579831932773109,
|
| 587 |
+
"grad_norm": 0.9877124428749084,
|
| 588 |
+
"learning_rate": 8.101171773125716e-05,
|
| 589 |
+
"loss": 0.2548,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.647058823529412,
|
| 594 |
+
"grad_norm": 1.0656546354293823,
|
| 595 |
+
"learning_rate": 7.89320730004274e-05,
|
| 596 |
+
"loss": 0.2681,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.714285714285714,
|
| 601 |
+
"grad_norm": 1.0215117931365967,
|
| 602 |
+
"learning_rate": 7.686192079596586e-05,
|
| 603 |
+
"loss": 0.2639,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 5.781512605042017,
|
| 608 |
+
"grad_norm": 0.9939188361167908,
|
| 609 |
+
"learning_rate": 7.48021938614875e-05,
|
| 610 |
+
"loss": 0.2583,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 5.848739495798319,
|
| 615 |
+
"grad_norm": 1.1241389513015747,
|
| 616 |
+
"learning_rate": 7.275382024331772e-05,
|
| 617 |
+
"loss": 0.2575,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.915966386554622,
|
| 622 |
+
"grad_norm": 1.2222434282302856,
|
| 623 |
+
"learning_rate": 7.071772287234497e-05,
|
| 624 |
+
"loss": 0.2628,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.983193277310924,
|
| 629 |
+
"grad_norm": 1.15309476852417,
|
| 630 |
+
"learning_rate": 6.869481914817779e-05,
|
| 631 |
+
"loss": 0.2662,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 6.050420168067227,
|
| 636 |
+
"grad_norm": 0.7435505986213684,
|
| 637 |
+
"learning_rate": 6.668602052579424e-05,
|
| 638 |
+
"loss": 0.2097,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 6.117647058823529,
|
| 643 |
+
"grad_norm": 0.9408479928970337,
|
| 644 |
+
"learning_rate": 6.469223210486992e-05,
|
| 645 |
+
"loss": 0.1896,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 6.184873949579832,
|
| 650 |
+
"grad_norm": 0.7466050386428833,
|
| 651 |
+
"learning_rate": 6.271435222196916e-05,
|
| 652 |
+
"loss": 0.1881,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 6.2521008403361344,
|
| 657 |
+
"grad_norm": 0.7878915667533875,
|
| 658 |
+
"learning_rate": 6.0753272045783625e-05,
|
| 659 |
+
"loss": 0.1921,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.319327731092437,
|
| 664 |
+
"grad_norm": 0.7960318922996521,
|
| 665 |
+
"learning_rate": 5.880987517560075e-05,
|
| 666 |
+
"loss": 0.1957,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 6.38655462184874,
|
| 671 |
+
"grad_norm": 0.8870383501052856,
|
| 672 |
+
"learning_rate": 5.688503724318217e-05,
|
| 673 |
+
"loss": 0.1954,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 6.453781512605042,
|
| 678 |
+
"grad_norm": 0.8262315392494202,
|
| 679 |
+
"learning_rate": 5.497962551823266e-05,
|
| 680 |
+
"loss": 0.1946,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 6.5210084033613445,
|
| 685 |
+
"grad_norm": 0.8804875612258911,
|
| 686 |
+
"learning_rate": 5.309449851763633e-05,
|
| 687 |
+
"loss": 0.1932,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 6.588235294117647,
|
| 692 |
+
"grad_norm": 0.9018622040748596,
|
| 693 |
+
"learning_rate": 5.123050561863657e-05,
|
| 694 |
+
"loss": 0.1982,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 6.65546218487395,
|
| 699 |
+
"grad_norm": 0.7731183171272278,
|
| 700 |
+
"learning_rate": 4.938848667613436e-05,
|
| 701 |
+
"loss": 0.2002,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 6.722689075630252,
|
| 706 |
+
"grad_norm": 0.9117534160614014,
|
| 707 |
+
"learning_rate": 4.756927164427685e-05,
|
| 708 |
+
"loss": 0.1983,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 6.7899159663865545,
|
| 713 |
+
"grad_norm": 1.0696296691894531,
|
| 714 |
+
"learning_rate": 4.57736802025065e-05,
|
| 715 |
+
"loss": 0.1984,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 6.857142857142857,
|
| 720 |
+
"grad_norm": 0.7894629240036011,
|
| 721 |
+
"learning_rate": 4.4002521386240466e-05,
|
| 722 |
+
"loss": 0.2018,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 6.92436974789916,
|
| 727 |
+
"grad_norm": 1.0283385515213013,
|
| 728 |
+
"learning_rate": 4.2256593222345185e-05,
|
| 729 |
+
"loss": 0.1979,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 6.991596638655462,
|
| 734 |
+
"grad_norm": 0.8787071108818054,
|
| 735 |
+
"learning_rate": 4.053668236957134e-05,
|
| 736 |
+
"loss": 0.1975,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 7.0588235294117645,
|
| 741 |
+
"grad_norm": 0.7579413056373596,
|
| 742 |
+
"learning_rate": 3.884356376411089e-05,
|
| 743 |
+
"loss": 0.1682,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 7.126050420168067,
|
| 748 |
+
"grad_norm": 0.7519951462745667,
|
| 749 |
+
"learning_rate": 3.717800027043576e-05,
|
| 750 |
+
"loss": 0.1653,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 7.19327731092437,
|
| 755 |
+
"grad_norm": 0.7462877631187439,
|
| 756 |
+
"learning_rate": 3.554074233757608e-05,
|
| 757 |
+
"loss": 0.1643,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 7.260504201680672,
|
| 762 |
+
"grad_norm": 0.7198042273521423,
|
| 763 |
+
"learning_rate": 3.393252766099187e-05,
|
| 764 |
+
"loss": 0.16,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 7.3277310924369745,
|
| 769 |
+
"grad_norm": 0.6849682927131653,
|
| 770 |
+
"learning_rate": 3.2354080850191324e-05,
|
| 771 |
+
"loss": 0.1628,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 7.394957983193278,
|
| 776 |
+
"grad_norm": 0.7326401472091675,
|
| 777 |
+
"learning_rate": 3.080611310224539e-05,
|
| 778 |
+
"loss": 0.1614,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 7.46218487394958,
|
| 783 |
+
"grad_norm": 0.8447051048278809,
|
| 784 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 785 |
+
"loss": 0.1608,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 7.529411764705882,
|
| 790 |
+
"grad_norm": 0.6565689444541931,
|
| 791 |
+
"learning_rate": 2.7804390604547557e-05,
|
| 792 |
+
"loss": 0.1605,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 7.5966386554621845,
|
| 797 |
+
"grad_norm": 0.6809631586074829,
|
| 798 |
+
"learning_rate": 2.6351988333848788e-05,
|
| 799 |
+
"loss": 0.1656,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 7.663865546218488,
|
| 804 |
+
"grad_norm": 0.7617256045341492,
|
| 805 |
+
"learning_rate": 2.493276947472756e-05,
|
| 806 |
+
"loss": 0.1594,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 7.73109243697479,
|
| 811 |
+
"grad_norm": 0.7275475263595581,
|
| 812 |
+
"learning_rate": 2.354737348129077e-05,
|
| 813 |
+
"loss": 0.1649,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 7.798319327731092,
|
| 818 |
+
"grad_norm": 0.7503297924995422,
|
| 819 |
+
"learning_rate": 2.2196424568156073e-05,
|
| 820 |
+
"loss": 0.1622,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 7.865546218487395,
|
| 825 |
+
"grad_norm": 0.6865005493164062,
|
| 826 |
+
"learning_rate": 2.0880531429201145e-05,
|
| 827 |
+
"loss": 0.1644,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 7.932773109243698,
|
| 832 |
+
"grad_norm": 0.8243385553359985,
|
| 833 |
+
"learning_rate": 1.9600286963305957e-05,
|
| 834 |
+
"loss": 0.1618,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 8.0,
|
| 839 |
+
"grad_norm": 0.7706124782562256,
|
| 840 |
+
"learning_rate": 1.835626800721144e-05,
|
| 841 |
+
"loss": 0.1626,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 8.067226890756302,
|
| 846 |
+
"grad_norm": 0.5460121631622314,
|
| 847 |
+
"learning_rate": 1.7149035075615794e-05,
|
| 848 |
+
"loss": 0.1438,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 8.134453781512605,
|
| 853 |
+
"grad_norm": 0.6267870664596558,
|
| 854 |
+
"learning_rate": 1.5979132108624574e-05,
|
| 855 |
+
"loss": 0.1474,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 8.201680672268907,
|
| 860 |
+
"grad_norm": 0.671909511089325,
|
| 861 |
+
"learning_rate": 1.4847086226668872e-05,
|
| 862 |
+
"loss": 0.1447,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 8.268907563025211,
|
| 867 |
+
"grad_norm": 0.5568744540214539,
|
| 868 |
+
"learning_rate": 1.3753407493001968e-05,
|
| 869 |
+
"loss": 0.1453,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 8.336134453781513,
|
| 874 |
+
"grad_norm": 0.6606815457344055,
|
| 875 |
+
"learning_rate": 1.2698588683881186e-05,
|
| 876 |
+
"loss": 0.1448,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 8.403361344537815,
|
| 881 |
+
"grad_norm": 0.7296724319458008,
|
| 882 |
+
"learning_rate": 1.1683105066539068e-05,
|
| 883 |
+
"loss": 0.1473,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 8.470588235294118,
|
| 888 |
+
"grad_norm": 0.6032551527023315,
|
| 889 |
+
"learning_rate": 1.0707414185043163e-05,
|
| 890 |
+
"loss": 0.1421,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 8.53781512605042,
|
| 895 |
+
"grad_norm": 0.6723304986953735,
|
| 896 |
+
"learning_rate": 9.771955654141496e-06,
|
| 897 |
+
"loss": 0.1484,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 8.605042016806722,
|
| 902 |
+
"grad_norm": 0.6945676207542419,
|
| 903 |
+
"learning_rate": 8.87715096118642e-06,
|
| 904 |
+
"loss": 0.1457,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 8.672268907563025,
|
| 909 |
+
"grad_norm": 0.6513916254043579,
|
| 910 |
+
"learning_rate": 8.023403276226126e-06,
|
| 911 |
+
"loss": 0.1495,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 8.739495798319329,
|
| 916 |
+
"grad_norm": 0.660636842250824,
|
| 917 |
+
"learning_rate": 7.211097270349066e-06,
|
| 918 |
+
"loss": 0.1448,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 8.806722689075631,
|
| 923 |
+
"grad_norm": 0.625892698764801,
|
| 924 |
+
"learning_rate": 6.440598942363796e-06,
|
| 925 |
+
"loss": 0.148,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 8.873949579831933,
|
| 930 |
+
"grad_norm": 0.6458612084388733,
|
| 931 |
+
"learning_rate": 5.71225545389158e-06,
|
| 932 |
+
"loss": 0.1458,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 8.941176470588236,
|
| 937 |
+
"grad_norm": 0.6204715967178345,
|
| 938 |
+
"learning_rate": 5.026394972946813e-06,
|
| 939 |
+
"loss": 0.1483,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 9.008403361344538,
|
| 944 |
+
"grad_norm": 0.5286262035369873,
|
| 945 |
+
"learning_rate": 4.383326526074916e-06,
|
| 946 |
+
"loss": 0.1445,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 9.07563025210084,
|
| 951 |
+
"grad_norm": 0.6546939611434937,
|
| 952 |
+
"learning_rate": 3.783339859115065e-06,
|
| 953 |
+
"loss": 0.1417,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 9.142857142857142,
|
| 958 |
+
"grad_norm": 0.5531789064407349,
|
| 959 |
+
"learning_rate": 3.226705306650113e-06,
|
| 960 |
+
"loss": 0.1408,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 9.210084033613445,
|
| 965 |
+
"grad_norm": 0.5917507410049438,
|
| 966 |
+
"learning_rate": 2.7136736702025433e-06,
|
| 967 |
+
"loss": 0.1414,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 9.277310924369749,
|
| 972 |
+
"grad_norm": 0.5545393228530884,
|
| 973 |
+
"learning_rate": 2.2444761052313856e-06,
|
| 974 |
+
"loss": 0.1413,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 9.344537815126051,
|
| 979 |
+
"grad_norm": 0.599034309387207,
|
| 980 |
+
"learning_rate": 1.8193240169810943e-06,
|
| 981 |
+
"loss": 0.1376,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 9.411764705882353,
|
| 986 |
+
"grad_norm": 0.5788201093673706,
|
| 987 |
+
"learning_rate": 1.4384089652291543e-06,
|
| 988 |
+
"loss": 0.1391,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 9.478991596638656,
|
| 993 |
+
"grad_norm": 0.5973650217056274,
|
| 994 |
+
"learning_rate": 1.1019025779754666e-06,
|
| 995 |
+
"loss": 0.1388,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 9.546218487394958,
|
| 1000 |
+
"grad_norm": 0.532236635684967,
|
| 1001 |
+
"learning_rate": 8.099564741123166e-07,
|
| 1002 |
+
"loss": 0.1373,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 9.61344537815126,
|
| 1007 |
+
"grad_norm": 0.6429956555366516,
|
| 1008 |
+
"learning_rate": 5.627021951097545e-07,
|
| 1009 |
+
"loss": 0.1376,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 9.680672268907562,
|
| 1014 |
+
"grad_norm": 0.6416929960250854,
|
| 1015 |
+
"learning_rate": 3.6025114574734785e-07,
|
| 1016 |
+
"loss": 0.1372,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 9.747899159663866,
|
| 1021 |
+
"grad_norm": 0.8270869255065918,
|
| 1022 |
+
"learning_rate": 2.0269454391874666e-07,
|
| 1023 |
+
"loss": 0.1394,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 9.815126050420169,
|
| 1028 |
+
"grad_norm": 0.6364122629165649,
|
| 1029 |
+
"learning_rate": 9.010337953185843e-08,
|
| 1030 |
+
"loss": 0.1412,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 9.882352941176471,
|
| 1035 |
+
"grad_norm": 0.6194050312042236,
|
| 1036 |
+
"learning_rate": 2.2528382523057113e-08,
|
| 1037 |
+
"loss": 0.1399,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 9.949579831932773,
|
| 1042 |
+
"grad_norm": 0.5904578566551208,
|
| 1043 |
+
"learning_rate": 0.0,
|
| 1044 |
+
"loss": 0.1422,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 9.949579831932773,
|
| 1049 |
+
"step": 1480,
|
| 1050 |
+
"total_flos": 7.279937555791872e+16,
|
| 1051 |
+
"train_loss": 0.41079871348432595,
|
| 1052 |
+
"train_runtime": 5840.149,
|
| 1053 |
+
"train_samples_per_second": 1.019,
|
| 1054 |
+
"train_steps_per_second": 0.253
|
| 1055 |
+
}
|
| 1056 |
+
],
|
| 1057 |
+
"logging_steps": 10,
|
| 1058 |
+
"max_steps": 1480,
|
| 1059 |
+
"num_input_tokens_seen": 0,
|
| 1060 |
+
"num_train_epochs": 10,
|
| 1061 |
+
"save_steps": 500,
|
| 1062 |
+
"total_flos": 7.279937555791872e+16,
|
| 1063 |
+
"train_batch_size": 1,
|
| 1064 |
+
"trial_name": null,
|
| 1065 |
+
"trial_params": null
|
| 1066 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e0a25ac727a6c94e3314c07266f97c65740e42d1ac0f4e34035c58e8075bade
|
| 3 |
size 5048
|