Upload folder using huggingface_hub
Browse files- adapter_config.json +2 -2
- adapter_model.safetensors +1 -1
- trainer_state.json +919 -0
adapter_config.json
CHANGED
|
@@ -23,10 +23,10 @@
|
|
| 23 |
"rank_pattern": {},
|
| 24 |
"revision": null,
|
| 25 |
"target_modules": [
|
| 26 |
-
"qkv_proj",
|
| 27 |
"down_proj",
|
|
|
|
| 28 |
"gate_up_proj",
|
| 29 |
-
"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
|
|
|
| 23 |
"rank_pattern": {},
|
| 24 |
"revision": null,
|
| 25 |
"target_modules": [
|
|
|
|
| 26 |
"down_proj",
|
| 27 |
+
"o_proj",
|
| 28 |
"gate_up_proj",
|
| 29 |
+
"qkv_proj"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 369133600
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf7e8b8c722ab0df8a5db587e800c65b59585720dfdf7fe5b4209a5be232841a
|
| 3 |
size 369133600
|
trainer_state.json
ADDED
|
@@ -0,0 +1,919 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 236,
|
| 3 |
+
"best_metric": 8.514503861078992e-05,
|
| 4 |
+
"best_model_checkpoint": "/content/drive/MyDrive/lora_model/outputs/task15_microsoft/Phi-4-mini-instruct/checkpoint-236",
|
| 5 |
+
"epoch": 59.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 236,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 1.0,
|
| 14 |
+
"grad_norm": 9.229236602783203,
|
| 15 |
+
"learning_rate": 0.0001,
|
| 16 |
+
"loss": 2.6715,
|
| 17 |
+
"step": 4
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.0,
|
| 21 |
+
"eval_loss": 1.3550810813903809,
|
| 22 |
+
"eval_runtime": 3.3411,
|
| 23 |
+
"eval_samples_per_second": 8.979,
|
| 24 |
+
"eval_steps_per_second": 1.197,
|
| 25 |
+
"step": 4
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 2.0,
|
| 29 |
+
"grad_norm": 9.349386215209961,
|
| 30 |
+
"learning_rate": 0.00023333333333333333,
|
| 31 |
+
"loss": 1.3025,
|
| 32 |
+
"step": 8
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.0,
|
| 36 |
+
"eval_loss": 0.9609652757644653,
|
| 37 |
+
"eval_runtime": 3.3476,
|
| 38 |
+
"eval_samples_per_second": 8.962,
|
| 39 |
+
"eval_steps_per_second": 1.195,
|
| 40 |
+
"step": 8
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 3.0,
|
| 44 |
+
"grad_norm": 4.202998161315918,
|
| 45 |
+
"learning_rate": 0.00036666666666666667,
|
| 46 |
+
"loss": 0.9833,
|
| 47 |
+
"step": 12
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 3.0,
|
| 51 |
+
"eval_loss": 0.8532541990280151,
|
| 52 |
+
"eval_runtime": 3.3649,
|
| 53 |
+
"eval_samples_per_second": 8.916,
|
| 54 |
+
"eval_steps_per_second": 1.189,
|
| 55 |
+
"step": 12
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 4.0,
|
| 59 |
+
"grad_norm": 5.911304473876953,
|
| 60 |
+
"learning_rate": 0.0005,
|
| 61 |
+
"loss": 0.8496,
|
| 62 |
+
"step": 16
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 4.0,
|
| 66 |
+
"eval_loss": 0.7015247941017151,
|
| 67 |
+
"eval_runtime": 3.3686,
|
| 68 |
+
"eval_samples_per_second": 8.906,
|
| 69 |
+
"eval_steps_per_second": 1.187,
|
| 70 |
+
"step": 16
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 5.0,
|
| 74 |
+
"grad_norm": 3.7761387825012207,
|
| 75 |
+
"learning_rate": 0.0004996426526821629,
|
| 76 |
+
"loss": 0.737,
|
| 77 |
+
"step": 20
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 5.0,
|
| 81 |
+
"eval_loss": 0.5959243178367615,
|
| 82 |
+
"eval_runtime": 3.3737,
|
| 83 |
+
"eval_samples_per_second": 8.892,
|
| 84 |
+
"eval_steps_per_second": 1.186,
|
| 85 |
+
"step": 20
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 6.0,
|
| 89 |
+
"grad_norm": 5.99334716796875,
|
| 90 |
+
"learning_rate": 0.0004985716323054959,
|
| 91 |
+
"loss": 0.6024,
|
| 92 |
+
"step": 24
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 6.0,
|
| 96 |
+
"eval_loss": 0.5120770931243896,
|
| 97 |
+
"eval_runtime": 3.3367,
|
| 98 |
+
"eval_samples_per_second": 8.991,
|
| 99 |
+
"eval_steps_per_second": 1.199,
|
| 100 |
+
"step": 24
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 7.0,
|
| 104 |
+
"grad_norm": 6.7658467292785645,
|
| 105 |
+
"learning_rate": 0.0004967900006800708,
|
| 106 |
+
"loss": 0.496,
|
| 107 |
+
"step": 28
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 7.0,
|
| 111 |
+
"eval_loss": 0.2661070227622986,
|
| 112 |
+
"eval_runtime": 3.3521,
|
| 113 |
+
"eval_samples_per_second": 8.95,
|
| 114 |
+
"eval_steps_per_second": 1.193,
|
| 115 |
+
"step": 28
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 8.0,
|
| 119 |
+
"grad_norm": 6.91306209564209,
|
| 120 |
+
"learning_rate": 0.0004943028510961491,
|
| 121 |
+
"loss": 0.3035,
|
| 122 |
+
"step": 32
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 8.0,
|
| 126 |
+
"eval_loss": 0.20593233406543732,
|
| 127 |
+
"eval_runtime": 3.3825,
|
| 128 |
+
"eval_samples_per_second": 8.869,
|
| 129 |
+
"eval_steps_per_second": 1.183,
|
| 130 |
+
"step": 32
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 9.0,
|
| 134 |
+
"grad_norm": 5.512369155883789,
|
| 135 |
+
"learning_rate": 0.0004911172937635942,
|
| 136 |
+
"loss": 0.2058,
|
| 137 |
+
"step": 36
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 9.0,
|
| 141 |
+
"eval_loss": 0.13461297750473022,
|
| 142 |
+
"eval_runtime": 3.3417,
|
| 143 |
+
"eval_samples_per_second": 8.978,
|
| 144 |
+
"eval_steps_per_second": 1.197,
|
| 145 |
+
"step": 36
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 10.0,
|
| 149 |
+
"grad_norm": 14.840362548828125,
|
| 150 |
+
"learning_rate": 0.0004872424354853545,
|
| 151 |
+
"loss": 0.3198,
|
| 152 |
+
"step": 40
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 10.0,
|
| 156 |
+
"eval_loss": 0.26689016819000244,
|
| 157 |
+
"eval_runtime": 3.3554,
|
| 158 |
+
"eval_samples_per_second": 8.941,
|
| 159 |
+
"eval_steps_per_second": 1.192,
|
| 160 |
+
"step": 40
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 11.0,
|
| 164 |
+
"grad_norm": 5.161523818969727,
|
| 165 |
+
"learning_rate": 0.00048268935362313215,
|
| 166 |
+
"loss": 0.225,
|
| 167 |
+
"step": 44
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 11.0,
|
| 171 |
+
"eval_loss": 0.1698600798845291,
|
| 172 |
+
"eval_runtime": 3.3826,
|
| 173 |
+
"eval_samples_per_second": 8.869,
|
| 174 |
+
"eval_steps_per_second": 1.183,
|
| 175 |
+
"step": 44
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 12.0,
|
| 179 |
+
"grad_norm": 6.356969833374023,
|
| 180 |
+
"learning_rate": 0.0004774710644296578,
|
| 181 |
+
"loss": 0.1683,
|
| 182 |
+
"step": 48
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 12.0,
|
| 186 |
+
"eval_loss": 0.14320150017738342,
|
| 187 |
+
"eval_runtime": 3.3444,
|
| 188 |
+
"eval_samples_per_second": 8.97,
|
| 189 |
+
"eval_steps_per_second": 1.196,
|
| 190 |
+
"step": 48
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 13.0,
|
| 194 |
+
"grad_norm": 7.662347316741943,
|
| 195 |
+
"learning_rate": 0.0004716024858381075,
|
| 196 |
+
"loss": 0.1401,
|
| 197 |
+
"step": 52
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 13.0,
|
| 201 |
+
"eval_loss": 0.07819870859384537,
|
| 202 |
+
"eval_runtime": 3.3566,
|
| 203 |
+
"eval_samples_per_second": 8.938,
|
| 204 |
+
"eval_steps_per_second": 1.192,
|
| 205 |
+
"step": 52
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 14.0,
|
| 209 |
+
"grad_norm": 3.4681787490844727,
|
| 210 |
+
"learning_rate": 0.00046510039481503486,
|
| 211 |
+
"loss": 0.0957,
|
| 212 |
+
"step": 56
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 14.0,
|
| 216 |
+
"eval_loss": 0.05342103913426399,
|
| 217 |
+
"eval_runtime": 3.3897,
|
| 218 |
+
"eval_samples_per_second": 8.85,
|
| 219 |
+
"eval_steps_per_second": 1.18,
|
| 220 |
+
"step": 56
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 15.0,
|
| 224 |
+
"grad_norm": 2.6177587509155273,
|
| 225 |
+
"learning_rate": 0.00045798337939873923,
|
| 226 |
+
"loss": 0.0853,
|
| 227 |
+
"step": 60
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 15.0,
|
| 231 |
+
"eval_loss": 0.07320532202720642,
|
| 232 |
+
"eval_runtime": 3.3471,
|
| 233 |
+
"eval_samples_per_second": 8.963,
|
| 234 |
+
"eval_steps_per_second": 1.195,
|
| 235 |
+
"step": 60
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 16.0,
|
| 239 |
+
"grad_norm": 10.169129371643066,
|
| 240 |
+
"learning_rate": 0.0004502717855601809,
|
| 241 |
+
"loss": 0.2155,
|
| 242 |
+
"step": 64
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 16.0,
|
| 246 |
+
"eval_loss": 0.14110827445983887,
|
| 247 |
+
"eval_runtime": 3.3497,
|
| 248 |
+
"eval_samples_per_second": 8.956,
|
| 249 |
+
"eval_steps_per_second": 1.194,
|
| 250 |
+
"step": 64
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 17.0,
|
| 254 |
+
"grad_norm": 4.305017471313477,
|
| 255 |
+
"learning_rate": 0.0004419876590383554,
|
| 256 |
+
"loss": 0.1458,
|
| 257 |
+
"step": 68
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 17.0,
|
| 261 |
+
"eval_loss": 0.07745428383350372,
|
| 262 |
+
"eval_runtime": 3.3838,
|
| 263 |
+
"eval_samples_per_second": 8.866,
|
| 264 |
+
"eval_steps_per_second": 1.182,
|
| 265 |
+
"step": 68
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 18.0,
|
| 269 |
+
"grad_norm": 6.399552822113037,
|
| 270 |
+
"learning_rate": 0.00043315468231640834,
|
| 271 |
+
"loss": 0.1029,
|
| 272 |
+
"step": 72
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 18.0,
|
| 276 |
+
"eval_loss": 0.11537892371416092,
|
| 277 |
+
"eval_runtime": 3.3464,
|
| 278 |
+
"eval_samples_per_second": 8.965,
|
| 279 |
+
"eval_steps_per_second": 1.195,
|
| 280 |
+
"step": 72
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 19.0,
|
| 284 |
+
"grad_norm": 5.510335922241211,
|
| 285 |
+
"learning_rate": 0.00042379810691866064,
|
| 286 |
+
"loss": 0.1198,
|
| 287 |
+
"step": 76
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 19.0,
|
| 291 |
+
"eval_loss": 0.09017840772867203,
|
| 292 |
+
"eval_runtime": 3.3477,
|
| 293 |
+
"eval_samples_per_second": 8.961,
|
| 294 |
+
"eval_steps_per_second": 1.195,
|
| 295 |
+
"step": 76
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 20.0,
|
| 299 |
+
"grad_norm": 3.690556287765503,
|
| 300 |
+
"learning_rate": 0.0004139446812220924,
|
| 301 |
+
"loss": 0.1062,
|
| 302 |
+
"step": 80
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 20.0,
|
| 306 |
+
"eval_loss": 0.0416039302945137,
|
| 307 |
+
"eval_runtime": 3.3871,
|
| 308 |
+
"eval_samples_per_second": 8.857,
|
| 309 |
+
"eval_steps_per_second": 1.181,
|
| 310 |
+
"step": 80
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 21.0,
|
| 314 |
+
"grad_norm": 3.543250799179077,
|
| 315 |
+
"learning_rate": 0.00040362257398865713,
|
| 316 |
+
"loss": 0.0896,
|
| 317 |
+
"step": 84
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 21.0,
|
| 321 |
+
"eval_loss": 0.06732075661420822,
|
| 322 |
+
"eval_runtime": 3.3474,
|
| 323 |
+
"eval_samples_per_second": 8.962,
|
| 324 |
+
"eval_steps_per_second": 1.195,
|
| 325 |
+
"step": 84
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 22.0,
|
| 329 |
+
"grad_norm": 3.758932113647461,
|
| 330 |
+
"learning_rate": 0.0003928612938370292,
|
| 331 |
+
"loss": 0.1018,
|
| 332 |
+
"step": 88
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 22.0,
|
| 336 |
+
"eval_loss": 0.05171294882893562,
|
| 337 |
+
"eval_runtime": 3.3522,
|
| 338 |
+
"eval_samples_per_second": 8.949,
|
| 339 |
+
"eval_steps_per_second": 1.193,
|
| 340 |
+
"step": 88
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 23.0,
|
| 344 |
+
"grad_norm": 51.296634674072266,
|
| 345 |
+
"learning_rate": 0.0003816916048839979,
|
| 346 |
+
"loss": 0.0869,
|
| 347 |
+
"step": 92
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 23.0,
|
| 351 |
+
"eval_loss": 0.10484348982572556,
|
| 352 |
+
"eval_runtime": 3.3734,
|
| 353 |
+
"eval_samples_per_second": 8.893,
|
| 354 |
+
"eval_steps_per_second": 1.186,
|
| 355 |
+
"step": 92
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 24.0,
|
| 359 |
+
"grad_norm": 3.9963831901550293,
|
| 360 |
+
"learning_rate": 0.00037014543879667093,
|
| 361 |
+
"loss": 0.112,
|
| 362 |
+
"step": 96
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 24.0,
|
| 366 |
+
"eval_loss": 0.06635650247335434,
|
| 367 |
+
"eval_runtime": 3.3487,
|
| 368 |
+
"eval_samples_per_second": 8.959,
|
| 369 |
+
"eval_steps_per_second": 1.195,
|
| 370 |
+
"step": 96
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 25.0,
|
| 374 |
+
"grad_norm": 4.267527103424072,
|
| 375 |
+
"learning_rate": 0.0003582558035069091,
|
| 376 |
+
"loss": 0.0833,
|
| 377 |
+
"step": 100
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 25.0,
|
| 381 |
+
"eval_loss": 0.053719986230134964,
|
| 382 |
+
"eval_runtime": 3.3464,
|
| 383 |
+
"eval_samples_per_second": 8.965,
|
| 384 |
+
"eval_steps_per_second": 1.195,
|
| 385 |
+
"step": 100
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 26.0,
|
| 389 |
+
"grad_norm": 1.9985603094100952,
|
| 390 |
+
"learning_rate": 0.0003460566888489593,
|
| 391 |
+
"loss": 0.0577,
|
| 392 |
+
"step": 104
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 26.0,
|
| 396 |
+
"eval_loss": 0.07173171639442444,
|
| 397 |
+
"eval_runtime": 3.3782,
|
| 398 |
+
"eval_samples_per_second": 8.88,
|
| 399 |
+
"eval_steps_per_second": 1.184,
|
| 400 |
+
"step": 104
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 27.0,
|
| 404 |
+
"grad_norm": 8.18993091583252,
|
| 405 |
+
"learning_rate": 0.00033358296939004547,
|
| 406 |
+
"loss": 0.1127,
|
| 407 |
+
"step": 108
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 27.0,
|
| 411 |
+
"eval_loss": 0.046506691724061966,
|
| 412 |
+
"eval_runtime": 3.3712,
|
| 413 |
+
"eval_samples_per_second": 8.899,
|
| 414 |
+
"eval_steps_per_second": 1.187,
|
| 415 |
+
"step": 108
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 28.0,
|
| 419 |
+
"grad_norm": 1.7538604736328125,
|
| 420 |
+
"learning_rate": 0.00032087030473170445,
|
| 421 |
+
"loss": 0.0478,
|
| 422 |
+
"step": 112
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 28.0,
|
| 426 |
+
"eval_loss": 0.03175203874707222,
|
| 427 |
+
"eval_runtime": 3.3745,
|
| 428 |
+
"eval_samples_per_second": 8.89,
|
| 429 |
+
"eval_steps_per_second": 1.185,
|
| 430 |
+
"step": 112
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 29.0,
|
| 434 |
+
"grad_norm": 13.397028923034668,
|
| 435 |
+
"learning_rate": 0.0003079550375668821,
|
| 436 |
+
"loss": 0.0585,
|
| 437 |
+
"step": 116
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 29.0,
|
| 441 |
+
"eval_loss": 0.026407385244965553,
|
| 442 |
+
"eval_runtime": 3.3744,
|
| 443 |
+
"eval_samples_per_second": 8.89,
|
| 444 |
+
"eval_steps_per_second": 1.185,
|
| 445 |
+
"step": 116
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 30.0,
|
| 449 |
+
"grad_norm": 2.6996145248413086,
|
| 450 |
+
"learning_rate": 0.0002948740897842223,
|
| 451 |
+
"loss": 0.0468,
|
| 452 |
+
"step": 120
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 30.0,
|
| 456 |
+
"eval_loss": 0.021538730710744858,
|
| 457 |
+
"eval_runtime": 3.357,
|
| 458 |
+
"eval_samples_per_second": 8.937,
|
| 459 |
+
"eval_steps_per_second": 1.192,
|
| 460 |
+
"step": 120
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 31.0,
|
| 464 |
+
"grad_norm": 3.0162580013275146,
|
| 465 |
+
"learning_rate": 0.00028166485691656423,
|
| 466 |
+
"loss": 0.0406,
|
| 467 |
+
"step": 124
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 31.0,
|
| 471 |
+
"eval_loss": 0.04511945694684982,
|
| 472 |
+
"eval_runtime": 3.3551,
|
| 473 |
+
"eval_samples_per_second": 8.942,
|
| 474 |
+
"eval_steps_per_second": 1.192,
|
| 475 |
+
"step": 124
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 32.0,
|
| 479 |
+
"grad_norm": 1.962350845336914,
|
| 480 |
+
"learning_rate": 0.0002683651012353955,
|
| 481 |
+
"loss": 0.0505,
|
| 482 |
+
"step": 128
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 32.0,
|
| 486 |
+
"eval_loss": 0.040635574609041214,
|
| 487 |
+
"eval_runtime": 3.3893,
|
| 488 |
+
"eval_samples_per_second": 8.851,
|
| 489 |
+
"eval_steps_per_second": 1.18,
|
| 490 |
+
"step": 128
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 33.0,
|
| 494 |
+
"grad_norm": 2.60496187210083,
|
| 495 |
+
"learning_rate": 0.00025501284379688067,
|
| 496 |
+
"loss": 0.0458,
|
| 497 |
+
"step": 132
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 33.0,
|
| 501 |
+
"eval_loss": 0.01912449672818184,
|
| 502 |
+
"eval_runtime": 3.3576,
|
| 503 |
+
"eval_samples_per_second": 8.935,
|
| 504 |
+
"eval_steps_per_second": 1.191,
|
| 505 |
+
"step": 132
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 34.0,
|
| 509 |
+
"grad_norm": 165.37008666992188,
|
| 510 |
+
"learning_rate": 0.00024164625574808144,
|
| 511 |
+
"loss": 0.0322,
|
| 512 |
+
"step": 136
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 34.0,
|
| 516 |
+
"eval_loss": 0.032830674201250076,
|
| 517 |
+
"eval_runtime": 3.3488,
|
| 518 |
+
"eval_samples_per_second": 8.958,
|
| 519 |
+
"eval_steps_per_second": 1.194,
|
| 520 |
+
"step": 136
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 35.0,
|
| 524 |
+
"grad_norm": 2.0985658168792725,
|
| 525 |
+
"learning_rate": 0.00022830354920410064,
|
| 526 |
+
"loss": 0.0392,
|
| 527 |
+
"step": 140
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 35.0,
|
| 531 |
+
"eval_loss": 0.016142379492521286,
|
| 532 |
+
"eval_runtime": 3.3792,
|
| 533 |
+
"eval_samples_per_second": 8.878,
|
| 534 |
+
"eval_steps_per_second": 1.184,
|
| 535 |
+
"step": 140
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 36.0,
|
| 539 |
+
"grad_norm": 2.632981777191162,
|
| 540 |
+
"learning_rate": 0.0002150228680081079,
|
| 541 |
+
"loss": 0.0328,
|
| 542 |
+
"step": 144
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 36.0,
|
| 546 |
+
"eval_loss": 0.018485000357031822,
|
| 547 |
+
"eval_runtime": 3.3462,
|
| 548 |
+
"eval_samples_per_second": 8.965,
|
| 549 |
+
"eval_steps_per_second": 1.195,
|
| 550 |
+
"step": 144
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 37.0,
|
| 554 |
+
"grad_norm": 2.000067949295044,
|
| 555 |
+
"learning_rate": 0.00020184217868653867,
|
| 556 |
+
"loss": 0.0281,
|
| 557 |
+
"step": 148
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 37.0,
|
| 561 |
+
"eval_loss": 0.010733678936958313,
|
| 562 |
+
"eval_runtime": 3.3514,
|
| 563 |
+
"eval_samples_per_second": 8.952,
|
| 564 |
+
"eval_steps_per_second": 1.194,
|
| 565 |
+
"step": 148
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 38.0,
|
| 569 |
+
"grad_norm": 2.6984753608703613,
|
| 570 |
+
"learning_rate": 0.00018879916191120349,
|
| 571 |
+
"loss": 0.0306,
|
| 572 |
+
"step": 152
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 38.0,
|
| 576 |
+
"eval_loss": 0.014147897250950336,
|
| 577 |
+
"eval_runtime": 3.3807,
|
| 578 |
+
"eval_samples_per_second": 8.874,
|
| 579 |
+
"eval_steps_per_second": 1.183,
|
| 580 |
+
"step": 152
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 39.0,
|
| 584 |
+
"grad_norm": 1.7256083488464355,
|
| 585 |
+
"learning_rate": 0.00017593110477859153,
|
| 586 |
+
"loss": 0.0229,
|
| 587 |
+
"step": 156
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 39.0,
|
| 591 |
+
"eval_loss": 0.013782525435090065,
|
| 592 |
+
"eval_runtime": 3.3475,
|
| 593 |
+
"eval_samples_per_second": 8.962,
|
| 594 |
+
"eval_steps_per_second": 1.195,
|
| 595 |
+
"step": 156
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 40.0,
|
| 599 |
+
"grad_norm": 1.312639832496643,
|
| 600 |
+
"learning_rate": 0.00016327479421431983,
|
| 601 |
+
"loss": 0.0233,
|
| 602 |
+
"step": 160
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 40.0,
|
| 606 |
+
"eval_loss": 0.009614935144782066,
|
| 607 |
+
"eval_runtime": 3.3523,
|
| 608 |
+
"eval_samples_per_second": 8.949,
|
| 609 |
+
"eval_steps_per_second": 1.193,
|
| 610 |
+
"step": 160
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 41.0,
|
| 614 |
+
"grad_norm": 1.2714215517044067,
|
| 615 |
+
"learning_rate": 0.00015086641180745932,
|
| 616 |
+
"loss": 0.014,
|
| 617 |
+
"step": 164
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 41.0,
|
| 621 |
+
"eval_loss": 0.01818581484258175,
|
| 622 |
+
"eval_runtime": 3.3843,
|
| 623 |
+
"eval_samples_per_second": 8.864,
|
| 624 |
+
"eval_steps_per_second": 1.182,
|
| 625 |
+
"step": 164
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 42.0,
|
| 629 |
+
"grad_norm": 2.4470083713531494,
|
| 630 |
+
"learning_rate": 0.00013874143037538418,
|
| 631 |
+
"loss": 0.0268,
|
| 632 |
+
"step": 168
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 42.0,
|
| 636 |
+
"eval_loss": 0.017209211364388466,
|
| 637 |
+
"eval_runtime": 3.3513,
|
| 638 |
+
"eval_samples_per_second": 8.952,
|
| 639 |
+
"eval_steps_per_second": 1.194,
|
| 640 |
+
"step": 168
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 43.0,
|
| 644 |
+
"grad_norm": 1.7730706930160522,
|
| 645 |
+
"learning_rate": 0.00012693451255484312,
|
| 646 |
+
"loss": 0.0449,
|
| 647 |
+
"step": 172
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 43.0,
|
| 651 |
+
"eval_loss": 0.03041950613260269,
|
| 652 |
+
"eval_runtime": 3.3534,
|
| 653 |
+
"eval_samples_per_second": 8.946,
|
| 654 |
+
"eval_steps_per_second": 1.193,
|
| 655 |
+
"step": 172
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 44.0,
|
| 659 |
+
"grad_norm": 0.741055965423584,
|
| 660 |
+
"learning_rate": 0.00011547941170915685,
|
| 661 |
+
"loss": 0.0211,
|
| 662 |
+
"step": 176
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 44.0,
|
| 666 |
+
"eval_loss": 0.00939116906374693,
|
| 667 |
+
"eval_runtime": 3.3843,
|
| 668 |
+
"eval_samples_per_second": 8.865,
|
| 669 |
+
"eval_steps_per_second": 1.182,
|
| 670 |
+
"step": 176
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 45.0,
|
| 674 |
+
"grad_norm": 0.5482388138771057,
|
| 675 |
+
"learning_rate": 0.00010440887543482746,
|
| 676 |
+
"loss": 0.0127,
|
| 677 |
+
"step": 180
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 45.0,
|
| 681 |
+
"eval_loss": 0.002400527708232403,
|
| 682 |
+
"eval_runtime": 3.3584,
|
| 683 |
+
"eval_samples_per_second": 8.933,
|
| 684 |
+
"eval_steps_per_second": 1.191,
|
| 685 |
+
"step": 180
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 46.0,
|
| 689 |
+
"grad_norm": 0.9107034206390381,
|
| 690 |
+
"learning_rate": 9.375455194341214e-05,
|
| 691 |
+
"loss": 0.0028,
|
| 692 |
+
"step": 184
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 46.0,
|
| 696 |
+
"eval_loss": 0.0015702341916039586,
|
| 697 |
+
"eval_runtime": 3.3439,
|
| 698 |
+
"eval_samples_per_second": 8.972,
|
| 699 |
+
"eval_steps_per_second": 1.196,
|
| 700 |
+
"step": 184
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 47.0,
|
| 704 |
+
"grad_norm": 0.6881429553031921,
|
| 705 |
+
"learning_rate": 8.354689958629513e-05,
|
| 706 |
+
"loss": 0.0026,
|
| 707 |
+
"step": 188
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 47.0,
|
| 711 |
+
"eval_loss": 0.0003339569375384599,
|
| 712 |
+
"eval_runtime": 3.3814,
|
| 713 |
+
"eval_samples_per_second": 8.872,
|
| 714 |
+
"eval_steps_per_second": 1.183,
|
| 715 |
+
"step": 188
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 48.0,
|
| 719 |
+
"grad_norm": 0.018251951783895493,
|
| 720 |
+
"learning_rate": 7.381509978100626e-05,
|
| 721 |
+
"loss": 0.0004,
|
| 722 |
+
"step": 192
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 48.0,
|
| 726 |
+
"eval_loss": 0.0006879170541651547,
|
| 727 |
+
"eval_runtime": 3.3536,
|
| 728 |
+
"eval_samples_per_second": 8.946,
|
| 729 |
+
"eval_steps_per_second": 1.193,
|
| 730 |
+
"step": 192
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 49.0,
|
| 734 |
+
"grad_norm": 0.14602628350257874,
|
| 735 |
+
"learning_rate": 6.458697358801061e-05,
|
| 736 |
+
"loss": 0.0013,
|
| 737 |
+
"step": 196
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 49.0,
|
| 741 |
+
"eval_loss": 0.0002326490357518196,
|
| 742 |
+
"eval_runtime": 3.3572,
|
| 743 |
+
"eval_samples_per_second": 8.936,
|
| 744 |
+
"eval_steps_per_second": 1.191,
|
| 745 |
+
"step": 196
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 50.0,
|
| 749 |
+
"grad_norm": 0.011048965156078339,
|
| 750 |
+
"learning_rate": 5.58889021764582e-05,
|
| 751 |
+
"loss": 0.0002,
|
| 752 |
+
"step": 200
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 50.0,
|
| 756 |
+
"eval_loss": 0.00015924364561215043,
|
| 757 |
+
"eval_runtime": 3.3745,
|
| 758 |
+
"eval_samples_per_second": 8.89,
|
| 759 |
+
"eval_steps_per_second": 1.185,
|
| 760 |
+
"step": 200
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 51.0,
|
| 764 |
+
"grad_norm": 0.008616381324827671,
|
| 765 |
+
"learning_rate": 4.7745751406263163e-05,
|
| 766 |
+
"loss": 0.0002,
|
| 767 |
+
"step": 204
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 51.0,
|
| 771 |
+
"eval_loss": 0.00013590451271738857,
|
| 772 |
+
"eval_runtime": 3.3562,
|
| 773 |
+
"eval_samples_per_second": 8.939,
|
| 774 |
+
"eval_steps_per_second": 1.192,
|
| 775 |
+
"step": 204
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 52.0,
|
| 779 |
+
"grad_norm": 0.007540772669017315,
|
| 780 |
+
"learning_rate": 4.0180800742117244e-05,
|
| 781 |
+
"loss": 0.0001,
|
| 782 |
+
"step": 208
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 52.0,
|
| 786 |
+
"eval_loss": 0.00011990289203822613,
|
| 787 |
+
"eval_runtime": 3.3714,
|
| 788 |
+
"eval_samples_per_second": 8.898,
|
| 789 |
+
"eval_steps_per_second": 1.186,
|
| 790 |
+
"step": 208
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 53.0,
|
| 794 |
+
"grad_norm": 0.004615222569555044,
|
| 795 |
+
"learning_rate": 3.321567670265568e-05,
|
| 796 |
+
"loss": 0.0001,
|
| 797 |
+
"step": 212
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 53.0,
|
| 801 |
+
"eval_loss": 0.00010787827341118827,
|
| 802 |
+
"eval_runtime": 3.3544,
|
| 803 |
+
"eval_samples_per_second": 8.944,
|
| 804 |
+
"eval_steps_per_second": 1.192,
|
| 805 |
+
"step": 212
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 54.0,
|
| 809 |
+
"grad_norm": 0.00507324980571866,
|
| 810 |
+
"learning_rate": 2.687029103502972e-05,
|
| 811 |
+
"loss": 0.0001,
|
| 812 |
+
"step": 216
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 54.0,
|
| 816 |
+
"eval_loss": 9.984564530896023e-05,
|
| 817 |
+
"eval_runtime": 3.3469,
|
| 818 |
+
"eval_samples_per_second": 8.964,
|
| 819 |
+
"eval_steps_per_second": 1.195,
|
| 820 |
+
"step": 216
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 55.0,
|
| 824 |
+
"grad_norm": 0.0049056364223361015,
|
| 825 |
+
"learning_rate": 2.1162783791631057e-05,
|
| 826 |
+
"loss": 0.0001,
|
| 827 |
+
"step": 220
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 55.0,
|
| 831 |
+
"eval_loss": 9.402891737408936e-05,
|
| 832 |
+
"eval_runtime": 3.3601,
|
| 833 |
+
"eval_samples_per_second": 8.928,
|
| 834 |
+
"eval_steps_per_second": 1.19,
|
| 835 |
+
"step": 220
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 56.0,
|
| 839 |
+
"grad_norm": 0.004331118427217007,
|
| 840 |
+
"learning_rate": 1.6109471471699556e-05,
|
| 841 |
+
"loss": 0.0001,
|
| 842 |
+
"step": 224
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 56.0,
|
| 846 |
+
"eval_loss": 9.088371007237583e-05,
|
| 847 |
+
"eval_runtime": 3.3632,
|
| 848 |
+
"eval_samples_per_second": 8.92,
|
| 849 |
+
"eval_steps_per_second": 1.189,
|
| 850 |
+
"step": 224
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 57.0,
|
| 854 |
+
"grad_norm": 0.004597917664796114,
|
| 855 |
+
"learning_rate": 1.1724800376064798e-05,
|
| 856 |
+
"loss": 0.0001,
|
| 857 |
+
"step": 228
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 57.0,
|
| 861 |
+
"eval_loss": 8.777277253102511e-05,
|
| 862 |
+
"eval_runtime": 3.3405,
|
| 863 |
+
"eval_samples_per_second": 8.981,
|
| 864 |
+
"eval_steps_per_second": 1.197,
|
| 865 |
+
"step": 228
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 58.0,
|
| 869 |
+
"grad_norm": 0.0038540030363947153,
|
| 870 |
+
"learning_rate": 8.02130530837189e-06,
|
| 871 |
+
"loss": 0.0001,
|
| 872 |
+
"step": 232
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 58.0,
|
| 876 |
+
"eval_loss": 8.658332808408886e-05,
|
| 877 |
+
"eval_runtime": 3.3595,
|
| 878 |
+
"eval_samples_per_second": 8.93,
|
| 879 |
+
"eval_steps_per_second": 1.191,
|
| 880 |
+
"step": 232
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 59.0,
|
| 884 |
+
"grad_norm": 0.0038286536000669003,
|
| 885 |
+
"learning_rate": 5.009573740853312e-06,
|
| 886 |
+
"loss": 0.0001,
|
| 887 |
+
"step": 236
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 59.0,
|
| 891 |
+
"eval_loss": 8.514503861078992e-05,
|
| 892 |
+
"eval_runtime": 3.3605,
|
| 893 |
+
"eval_samples_per_second": 8.927,
|
| 894 |
+
"eval_steps_per_second": 1.19,
|
| 895 |
+
"step": 236
|
| 896 |
+
}
|
| 897 |
+
],
|
| 898 |
+
"logging_steps": 1,
|
| 899 |
+
"max_steps": 250,
|
| 900 |
+
"num_input_tokens_seen": 0,
|
| 901 |
+
"num_train_epochs": 63,
|
| 902 |
+
"save_steps": 500,
|
| 903 |
+
"stateful_callbacks": {
|
| 904 |
+
"TrainerControl": {
|
| 905 |
+
"args": {
|
| 906 |
+
"should_epoch_stop": false,
|
| 907 |
+
"should_evaluate": false,
|
| 908 |
+
"should_log": false,
|
| 909 |
+
"should_save": true,
|
| 910 |
+
"should_training_stop": false
|
| 911 |
+
},
|
| 912 |
+
"attributes": {}
|
| 913 |
+
}
|
| 914 |
+
},
|
| 915 |
+
"total_flos": 8799046124943360.0,
|
| 916 |
+
"train_batch_size": 1,
|
| 917 |
+
"trial_name": null,
|
| 918 |
+
"trial_params": null
|
| 919 |
+
}
|