Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- README.md +1 -1
- adapter_config.json +71 -95
- adapter_model.safetensors +2 -2
- latest +1 -1
- ocr.json +3 -0
- scheduler.pt +1 -1
- trainer_state.json +901 -306
- training_args.bin +1 -1
.gitattributes
CHANGED
|
@@ -34,3 +34,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
ocr.json filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
---
|
| 2 |
-
base_model: Qwen/Qwen2.5-VL-
|
| 3 |
library_name: peft
|
| 4 |
---
|
| 5 |
|
|
|
|
| 1 |
---
|
| 2 |
+
base_model: Qwen/Qwen2.5-VL-7B-Instruct
|
| 3 |
library_name: peft
|
| 4 |
---
|
| 5 |
|
adapter_config.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"alpha_pattern": {},
|
| 3 |
"auto_mapping": null,
|
| 4 |
-
"base_model_name_or_path": "Qwen/Qwen2.5-VL-
|
| 5 |
"bias": "none",
|
| 6 |
"corda_config": null,
|
| 7 |
"eva_config": null,
|
|
@@ -24,118 +24,94 @@
|
|
| 24 |
"rank_pattern": {},
|
| 25 |
"revision": null,
|
| 26 |
"target_modules": [
|
| 27 |
-
"layers.
|
| 28 |
-
"layers.
|
| 29 |
-
"layers.
|
| 30 |
-
"layers.
|
| 31 |
-
"layers.
|
|
|
|
| 32 |
"layers.25.mlp.up_proj",
|
| 33 |
-
"layers.28.mlp.gate_proj",
|
| 34 |
-
"32.mlp.gate_proj",
|
| 35 |
-
"layers.3.mlp.down_proj",
|
| 36 |
-
"layers.14.mlp.gate_proj",
|
| 37 |
-
"layers.30.mlp.gate_proj",
|
| 38 |
-
"layers.22.mlp.up_proj",
|
| 39 |
-
"32.mlp.up_proj",
|
| 40 |
-
"layers.15.mlp.down_proj",
|
| 41 |
-
"layers.7.mlp.gate_proj",
|
| 42 |
-
"layers.29.mlp.down_proj",
|
| 43 |
-
"34.mlp.down_proj",
|
| 44 |
-
"layers.2.mlp.up_proj",
|
| 45 |
"v_proj",
|
| 46 |
-
"layers.
|
|
|
|
|
|
|
|
|
|
| 47 |
"layers.18.mlp.up_proj",
|
| 48 |
-
"layers.
|
| 49 |
-
"layers.
|
| 50 |
-
"layers.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
"layers.24.mlp.down_proj",
|
| 52 |
-
"layers.
|
| 53 |
-
"layers.
|
|
|
|
| 54 |
"layers.27.mlp.down_proj",
|
| 55 |
-
"
|
|
|
|
|
|
|
| 56 |
"layers.2.mlp.gate_proj",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 57 |
"layers.20.mlp.up_proj",
|
|
|
|
|
|
|
|
|
|
| 58 |
"layers.20.mlp.gate_proj",
|
| 59 |
-
"layers.
|
| 60 |
-
"33.mlp.up_proj",
|
| 61 |
-
"32.mlp.down_proj",
|
| 62 |
-
"layers.25.mlp.down_proj",
|
| 63 |
-
"layers.28.mlp.down_proj",
|
| 64 |
-
"layers.0.mlp.down_proj",
|
| 65 |
-
"35.mlp.down_proj",
|
| 66 |
-
"layers.6.mlp.down_proj",
|
| 67 |
-
"o_proj",
|
| 68 |
-
"layers.29.mlp.up_proj",
|
| 69 |
-
"layers.3.mlp.up_proj",
|
| 70 |
-
"35.mlp.gate_proj",
|
| 71 |
"layers.20.mlp.down_proj",
|
| 72 |
-
"layers.
|
| 73 |
-
"layers.13.mlp.gate_proj",
|
| 74 |
-
"layers.7.mlp.down_proj",
|
| 75 |
-
"layers.5.mlp.down_proj",
|
| 76 |
-
"layers.5.mlp.up_proj",
|
| 77 |
-
"q_proj",
|
| 78 |
-
"layers.21.mlp.up_proj",
|
| 79 |
-
"layers.10.mlp.down_proj",
|
| 80 |
-
"layers.10.mlp.gate_proj",
|
| 81 |
-
"layers.31.mlp.up_proj",
|
| 82 |
-
"layers.0.mlp.up_proj",
|
| 83 |
-
"layers.14.mlp.up_proj",
|
| 84 |
-
"33.mlp.gate_proj",
|
| 85 |
-
"35.mlp.up_proj",
|
| 86 |
-
"layers.17.mlp.up_proj",
|
| 87 |
-
"layers.11.mlp.down_proj",
|
| 88 |
-
"layers.30.mlp.down_proj",
|
| 89 |
-
"layers.6.mlp.up_proj",
|
| 90 |
-
"layers.19.mlp.up_proj",
|
| 91 |
-
"layers.4.mlp.up_proj",
|
| 92 |
"layers.10.mlp.up_proj",
|
| 93 |
-
"34.mlp.up_proj",
|
| 94 |
-
"33.mlp.down_proj",
|
| 95 |
-
"layers.22.mlp.gate_proj",
|
| 96 |
-
"k_proj",
|
| 97 |
-
"layers.9.mlp.gate_proj",
|
| 98 |
-
"layers.9.mlp.up_proj",
|
| 99 |
-
"layers.11.mlp.gate_proj",
|
| 100 |
-
"layers.23.mlp.down_proj",
|
| 101 |
-
"layers.7.mlp.up_proj",
|
| 102 |
"layers.9.mlp.down_proj",
|
| 103 |
-
"layers.
|
| 104 |
-
"layers.
|
| 105 |
-
"layers.
|
| 106 |
-
"layers.
|
| 107 |
-
"layers.
|
| 108 |
-
"layers.
|
| 109 |
-
"layers.13.mlp.up_proj",
|
| 110 |
-
"layers.26.mlp.down_proj",
|
| 111 |
-
"layers.15.mlp.gate_proj",
|
| 112 |
-
"layers.22.mlp.down_proj",
|
| 113 |
-
"layers.30.mlp.up_proj",
|
| 114 |
-
"layers.16.mlp.gate_proj",
|
| 115 |
-
"layers.23.mlp.gate_proj",
|
| 116 |
-
"layers.31.mlp.gate_proj",
|
| 117 |
-
"layers.12.mlp.down_proj",
|
| 118 |
-
"layers.1.mlp.up_proj",
|
| 119 |
-
"layers.8.mlp.up_proj",
|
| 120 |
"layers.4.mlp.down_proj",
|
| 121 |
-
"layers.
|
|
|
|
|
|
|
|
|
|
| 122 |
"layers.8.mlp.gate_proj",
|
| 123 |
"layers.19.mlp.down_proj",
|
| 124 |
-
"layers.
|
| 125 |
-
"
|
| 126 |
-
"layers.16.mlp.down_proj",
|
| 127 |
"layers.16.mlp.up_proj",
|
| 128 |
-
"layers.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 129 |
"layers.12.mlp.up_proj",
|
| 130 |
-
"layers.
|
| 131 |
-
"layers.
|
| 132 |
-
"layers.
|
| 133 |
-
"layers.26.mlp.up_proj",
|
| 134 |
"layers.1.mlp.down_proj",
|
| 135 |
-
"layers.
|
| 136 |
-
"layers.
|
| 137 |
"layers.4.mlp.gate_proj",
|
| 138 |
-
"layers.27.mlp.up_proj"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 139 |
],
|
| 140 |
"task_type": "CAUSAL_LM",
|
| 141 |
"trainable_token_indices": null,
|
|
|
|
| 1 |
{
|
| 2 |
"alpha_pattern": {},
|
| 3 |
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "Qwen/Qwen2.5-VL-7B-Instruct",
|
| 5 |
"bias": "none",
|
| 6 |
"corda_config": null,
|
| 7 |
"eva_config": null,
|
|
|
|
| 24 |
"rank_pattern": {},
|
| 25 |
"revision": null,
|
| 26 |
"target_modules": [
|
| 27 |
+
"layers.16.mlp.down_proj",
|
| 28 |
+
"layers.12.mlp.gate_proj",
|
| 29 |
+
"layers.17.mlp.up_proj",
|
| 30 |
+
"layers.3.mlp.up_proj",
|
| 31 |
+
"layers.5.mlp.gate_proj",
|
| 32 |
+
"k_proj",
|
| 33 |
"layers.25.mlp.up_proj",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
"v_proj",
|
| 35 |
+
"layers.8.mlp.down_proj",
|
| 36 |
+
"layers.24.mlp.up_proj",
|
| 37 |
+
"layers.15.mlp.up_proj",
|
| 38 |
+
"layers.7.mlp.gate_proj",
|
| 39 |
"layers.18.mlp.up_proj",
|
| 40 |
+
"layers.13.mlp.gate_proj",
|
| 41 |
+
"layers.6.mlp.gate_proj",
|
| 42 |
+
"layers.9.mlp.gate_proj",
|
| 43 |
+
"layers.16.mlp.gate_proj",
|
| 44 |
+
"layers.22.mlp.up_proj",
|
| 45 |
+
"layers.21.mlp.up_proj",
|
| 46 |
+
"layers.27.mlp.gate_proj",
|
| 47 |
+
"layers.23.mlp.down_proj",
|
| 48 |
+
"layers.24.mlp.gate_proj",
|
| 49 |
+
"layers.23.mlp.up_proj",
|
| 50 |
"layers.24.mlp.down_proj",
|
| 51 |
+
"layers.0.mlp.down_proj",
|
| 52 |
+
"layers.0.mlp.up_proj",
|
| 53 |
+
"layers.26.mlp.gate_proj",
|
| 54 |
"layers.27.mlp.down_proj",
|
| 55 |
+
"q_proj",
|
| 56 |
+
"layers.1.mlp.up_proj",
|
| 57 |
+
"layers.21.mlp.down_proj",
|
| 58 |
"layers.2.mlp.gate_proj",
|
| 59 |
+
"layers.11.mlp.gate_proj",
|
| 60 |
+
"layers.14.mlp.up_proj",
|
| 61 |
+
"layers.22.mlp.down_proj",
|
| 62 |
+
"layers.5.mlp.down_proj",
|
| 63 |
+
"layers.19.mlp.gate_proj",
|
| 64 |
+
"layers.17.mlp.down_proj",
|
| 65 |
+
"layers.9.mlp.up_proj",
|
| 66 |
"layers.20.mlp.up_proj",
|
| 67 |
+
"layers.3.mlp.gate_proj",
|
| 68 |
+
"layers.1.mlp.gate_proj",
|
| 69 |
+
"layers.13.mlp.up_proj",
|
| 70 |
"layers.20.mlp.gate_proj",
|
| 71 |
+
"layers.26.mlp.down_proj",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 72 |
"layers.20.mlp.down_proj",
|
| 73 |
+
"layers.14.mlp.down_proj",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
"layers.10.mlp.up_proj",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
"layers.9.mlp.down_proj",
|
| 76 |
+
"layers.6.mlp.up_proj",
|
| 77 |
+
"layers.11.mlp.up_proj",
|
| 78 |
+
"layers.10.mlp.gate_proj",
|
| 79 |
+
"layers.4.mlp.up_proj",
|
| 80 |
+
"layers.14.mlp.gate_proj",
|
| 81 |
+
"layers.2.mlp.up_proj",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 82 |
"layers.4.mlp.down_proj",
|
| 83 |
+
"layers.7.mlp.down_proj",
|
| 84 |
+
"o_proj",
|
| 85 |
+
"layers.0.mlp.gate_proj",
|
| 86 |
+
"layers.17.mlp.gate_proj",
|
| 87 |
"layers.8.mlp.gate_proj",
|
| 88 |
"layers.19.mlp.down_proj",
|
| 89 |
+
"layers.25.mlp.gate_proj",
|
| 90 |
+
"layers.11.mlp.down_proj",
|
|
|
|
| 91 |
"layers.16.mlp.up_proj",
|
| 92 |
+
"layers.2.mlp.down_proj",
|
| 93 |
+
"layers.19.mlp.up_proj",
|
| 94 |
+
"layers.15.mlp.gate_proj",
|
| 95 |
+
"layers.22.mlp.gate_proj",
|
| 96 |
+
"layers.23.mlp.gate_proj",
|
| 97 |
+
"layers.6.mlp.down_proj",
|
| 98 |
+
"layers.10.mlp.down_proj",
|
| 99 |
"layers.12.mlp.up_proj",
|
| 100 |
+
"layers.8.mlp.up_proj",
|
| 101 |
+
"layers.5.mlp.up_proj",
|
| 102 |
+
"layers.3.mlp.down_proj",
|
|
|
|
| 103 |
"layers.1.mlp.down_proj",
|
| 104 |
+
"layers.21.mlp.gate_proj",
|
| 105 |
+
"layers.26.mlp.up_proj",
|
| 106 |
"layers.4.mlp.gate_proj",
|
| 107 |
+
"layers.27.mlp.up_proj",
|
| 108 |
+
"layers.12.mlp.down_proj",
|
| 109 |
+
"layers.15.mlp.down_proj",
|
| 110 |
+
"layers.18.mlp.gate_proj",
|
| 111 |
+
"layers.18.mlp.down_proj",
|
| 112 |
+
"layers.7.mlp.up_proj",
|
| 113 |
+
"layers.25.mlp.down_proj",
|
| 114 |
+
"layers.13.mlp.down_proj"
|
| 115 |
],
|
| 116 |
"task_type": "CAUSAL_LM",
|
| 117 |
"trainable_token_indices": null,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c65869936ca775b0b7e34856b673e70c7b88167d895229d6337d74ee4f2a74aa
|
| 3 |
+
size 323014560
|
latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step800
|
ocr.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ad6da340f2f6403e305dac3b8deef53107b7c5de50f73da9c490b8b174d6e40
|
| 3 |
+
size 34218525
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1401
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bbb214732a7d2d242635ceda6bf2c890011d7f28b61b690994eae9c7558a5c03
|
| 3 |
size 1401
|
trainer_state.json
CHANGED
|
@@ -2,543 +2,1138 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm": 0.
|
| 15 |
-
"learning_rate":
|
| 16 |
-
"loss": 0.
|
| 17 |
"step": 5
|
| 18 |
},
|
| 19 |
{
|
| 20 |
-
"epoch": 0.
|
| 21 |
-
"grad_norm": 0.
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss": 0.
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm": 0.
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"loss": 0.
|
| 31 |
"step": 15
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm": 0.
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 0.
|
| 38 |
"step": 20
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm": 0.
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 25
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"grad_norm": 0.
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss": 0.
|
| 52 |
"step": 30
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm": 0.
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 35
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"grad_norm": 0.
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 40
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 0.
|
| 70 |
-
"grad_norm": 0.
|
| 71 |
-
"learning_rate":
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 45
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 0.
|
| 77 |
-
"grad_norm": 0.
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 50
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
-
"grad_norm": 0.
|
| 85 |
-
"learning_rate": 9.
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 55
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"grad_norm": 0.
|
| 92 |
-
"learning_rate": 9.
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 60
|
| 95 |
},
|
| 96 |
{
|
| 97 |
-
"epoch": 0.
|
| 98 |
-
"grad_norm": 0.
|
| 99 |
-
"learning_rate": 9.
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 65
|
| 102 |
},
|
| 103 |
{
|
| 104 |
-
"epoch": 0.
|
| 105 |
-
"grad_norm": 0.
|
| 106 |
-
"learning_rate": 9.
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 70
|
| 109 |
},
|
| 110 |
{
|
| 111 |
-
"epoch": 0.
|
| 112 |
-
"grad_norm": 0.
|
| 113 |
-
"learning_rate": 9.
|
| 114 |
-
"loss": 0.
|
| 115 |
"step": 75
|
| 116 |
},
|
| 117 |
{
|
| 118 |
-
"epoch": 0.
|
| 119 |
-
"grad_norm": 0.
|
| 120 |
-
"learning_rate": 9.
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 80
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch": 0.
|
| 126 |
-
"grad_norm": 0.
|
| 127 |
-
"learning_rate": 9.
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 85
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"grad_norm": 0.
|
| 134 |
-
"learning_rate": 9.
|
| 135 |
-
"loss": 0.
|
| 136 |
"step": 90
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"epoch": 0.
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
-
"learning_rate":
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 95
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"epoch": 0.
|
| 147 |
-
"grad_norm": 0.
|
| 148 |
-
"learning_rate":
|
| 149 |
-
"loss": 0.
|
| 150 |
"step": 100
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch": 0.
|
| 154 |
-
"grad_norm": 0.
|
| 155 |
-
"learning_rate":
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 105
|
| 158 |
},
|
| 159 |
{
|
| 160 |
-
"epoch": 0.
|
| 161 |
-
"grad_norm": 0.
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 110
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 0.
|
| 168 |
-
"grad_norm": 0.
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 115
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
-
"grad_norm": 0.
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 120
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch": 0.
|
| 182 |
-
"grad_norm": 0.
|
| 183 |
-
"learning_rate":
|
| 184 |
-
"loss": 0.
|
| 185 |
"step": 125
|
| 186 |
},
|
| 187 |
{
|
| 188 |
-
"epoch": 0.
|
| 189 |
-
"grad_norm": 0.
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 130
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch": 0.
|
| 196 |
-
"grad_norm": 0.
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 135
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch": 0.
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 140
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch": 0.
|
| 210 |
-
"grad_norm": 0.
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 145
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch": 0.
|
| 217 |
-
"grad_norm": 0.
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 150
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"epoch": 0.
|
| 224 |
-
"grad_norm": 0.
|
| 225 |
-
"learning_rate":
|
| 226 |
-
"loss": 0.
|
| 227 |
"step": 155
|
| 228 |
},
|
| 229 |
{
|
| 230 |
-
"epoch": 0.
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss": 0.
|
| 234 |
"step": 160
|
| 235 |
},
|
| 236 |
{
|
| 237 |
-
"epoch": 0.
|
| 238 |
-
"grad_norm": 0.
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss": 0.
|
| 241 |
"step": 165
|
| 242 |
},
|
| 243 |
{
|
| 244 |
-
"epoch": 0.
|
| 245 |
-
"grad_norm": 0.
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss": 0.
|
| 248 |
"step": 170
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch": 0.
|
| 252 |
-
"grad_norm": 0.
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 175
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch": 0.
|
| 259 |
-
"grad_norm": 0.
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss": 0.
|
| 262 |
"step": 180
|
| 263 |
},
|
| 264 |
{
|
| 265 |
-
"epoch": 0.
|
| 266 |
-
"grad_norm": 0.
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss": 0.
|
| 269 |
"step": 185
|
| 270 |
},
|
| 271 |
{
|
| 272 |
-
"epoch": 0.
|
| 273 |
-
"grad_norm": 0.
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss": 0.
|
| 276 |
"step": 190
|
| 277 |
},
|
| 278 |
{
|
| 279 |
-
"epoch": 0.
|
| 280 |
-
"grad_norm": 0.
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss": 0.
|
| 283 |
"step": 195
|
| 284 |
},
|
| 285 |
{
|
| 286 |
-
"epoch": 0.
|
| 287 |
-
"grad_norm": 0.
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 200
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch": 0.
|
| 294 |
-
"grad_norm": 0.
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 205
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch": 0.
|
| 301 |
-
"grad_norm": 0.
|
| 302 |
-
"learning_rate":
|
| 303 |
-
"loss": 0.
|
| 304 |
"step": 210
|
| 305 |
},
|
| 306 |
{
|
| 307 |
-
"epoch": 0.
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
-
"learning_rate":
|
| 310 |
-
"loss": 0.
|
| 311 |
"step": 215
|
| 312 |
},
|
| 313 |
{
|
| 314 |
-
"epoch": 0.
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss": 0.
|
| 318 |
"step": 220
|
| 319 |
},
|
| 320 |
{
|
| 321 |
-
"epoch": 0.
|
| 322 |
-
"grad_norm": 0.
|
| 323 |
-
"learning_rate":
|
| 324 |
-
"loss": 0.
|
| 325 |
"step": 225
|
| 326 |
},
|
| 327 |
{
|
| 328 |
-
"epoch": 0.
|
| 329 |
-
"grad_norm": 0.
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss": 0.
|
| 332 |
"step": 230
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch": 0.
|
| 336 |
-
"grad_norm": 0.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 235
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch": 0.
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 240
|
| 347 |
},
|
| 348 |
{
|
| 349 |
-
"epoch": 0.
|
| 350 |
-
"grad_norm": 0.
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 245
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch": 0.
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 250
|
| 361 |
},
|
| 362 |
{
|
| 363 |
-
"epoch": 0.
|
| 364 |
-
"grad_norm": 0.
|
| 365 |
-
"learning_rate":
|
| 366 |
-
"loss": 0.
|
| 367 |
"step": 255
|
| 368 |
},
|
| 369 |
{
|
| 370 |
-
"epoch": 0.
|
| 371 |
-
"grad_norm": 0.
|
| 372 |
-
"learning_rate":
|
| 373 |
-
"loss": 0.
|
| 374 |
"step": 260
|
| 375 |
},
|
| 376 |
{
|
| 377 |
-
"epoch": 0.
|
| 378 |
-
"grad_norm": 0.
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss": 0.
|
| 381 |
"step": 265
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch": 0.
|
| 385 |
-
"grad_norm": 0.
|
| 386 |
-
"learning_rate":
|
| 387 |
-
"loss": 0.
|
| 388 |
"step": 270
|
| 389 |
},
|
| 390 |
{
|
| 391 |
-
"epoch": 0.
|
| 392 |
-
"grad_norm": 0.
|
| 393 |
-
"learning_rate":
|
| 394 |
-
"loss": 0.
|
| 395 |
"step": 275
|
| 396 |
},
|
| 397 |
{
|
| 398 |
-
"epoch": 0.
|
| 399 |
-
"grad_norm": 0.
|
| 400 |
-
"learning_rate":
|
| 401 |
-
"loss": 0.
|
| 402 |
"step": 280
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"epoch": 0.
|
| 406 |
-
"grad_norm": 0.
|
| 407 |
-
"learning_rate":
|
| 408 |
-
"loss": 0.
|
| 409 |
"step": 285
|
| 410 |
},
|
| 411 |
{
|
| 412 |
-
"epoch": 0.
|
| 413 |
-
"grad_norm": 0.
|
| 414 |
-
"learning_rate":
|
| 415 |
-
"loss": 0.
|
| 416 |
"step": 290
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch": 0.
|
| 420 |
-
"grad_norm": 0.
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss": 0.
|
| 423 |
"step": 295
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch": 0.
|
| 427 |
-
"grad_norm": 0.
|
| 428 |
-
"learning_rate":
|
| 429 |
-
"loss": 0.
|
| 430 |
"step": 300
|
| 431 |
},
|
| 432 |
{
|
| 433 |
-
"epoch": 0.
|
| 434 |
-
"grad_norm": 0.
|
| 435 |
-
"learning_rate":
|
| 436 |
-
"loss": 0.
|
| 437 |
"step": 305
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch": 0.
|
| 441 |
-
"grad_norm": 0.
|
| 442 |
-
"learning_rate": 8.
|
| 443 |
-
"loss": 0.
|
| 444 |
"step": 310
|
| 445 |
},
|
| 446 |
{
|
| 447 |
-
"epoch": 0.
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
-
"learning_rate":
|
| 450 |
-
"loss": 0.
|
| 451 |
"step": 315
|
| 452 |
},
|
| 453 |
{
|
| 454 |
-
"epoch": 0.
|
| 455 |
-
"grad_norm": 0.
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 320
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch": 0.
|
| 462 |
-
"grad_norm": 0.
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 325
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch": 0.
|
| 469 |
-
"grad_norm": 0.
|
| 470 |
-
"learning_rate":
|
| 471 |
-
"loss": 0.
|
| 472 |
"step": 330
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"epoch": 0.
|
| 476 |
-
"grad_norm": 0.
|
| 477 |
-
"learning_rate":
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 335
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch": 0.
|
| 483 |
-
"grad_norm": 0.
|
| 484 |
-
"learning_rate":
|
| 485 |
-
"loss": 0.
|
| 486 |
"step": 340
|
| 487 |
},
|
| 488 |
{
|
| 489 |
-
"epoch": 0.
|
| 490 |
-
"grad_norm": 0.
|
| 491 |
-
"learning_rate":
|
| 492 |
-
"loss": 0.
|
| 493 |
"step": 345
|
| 494 |
},
|
| 495 |
{
|
| 496 |
-
"epoch": 0.
|
| 497 |
-
"grad_norm": 0.
|
| 498 |
-
"learning_rate":
|
| 499 |
-
"loss": 0.
|
| 500 |
"step": 350
|
| 501 |
},
|
| 502 |
{
|
| 503 |
-
"epoch": 0.
|
| 504 |
-
"grad_norm": 0.
|
| 505 |
-
"learning_rate": 8.
|
| 506 |
-
"loss": 0.
|
| 507 |
"step": 355
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch": 0.
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
-
"learning_rate":
|
| 513 |
-
"loss": 0.
|
| 514 |
"step": 360
|
| 515 |
},
|
| 516 |
{
|
| 517 |
-
"epoch": 0.
|
| 518 |
-
"grad_norm": 0.
|
| 519 |
-
"learning_rate":
|
| 520 |
-
"loss": 0.
|
| 521 |
"step": 365
|
| 522 |
},
|
| 523 |
{
|
| 524 |
-
"epoch": 0.
|
| 525 |
-
"grad_norm": 0.
|
| 526 |
-
"learning_rate":
|
| 527 |
-
"loss": 0.
|
| 528 |
"step": 370
|
| 529 |
},
|
| 530 |
{
|
| 531 |
-
"epoch":
|
| 532 |
-
"grad_norm": 0.
|
| 533 |
-
"learning_rate":
|
| 534 |
-
"loss": 0.
|
| 535 |
"step": 375
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 536 |
}
|
| 537 |
],
|
| 538 |
"logging_steps": 5,
|
| 539 |
-
"max_steps":
|
| 540 |
"num_input_tokens_seen": 0,
|
| 541 |
-
"num_train_epochs":
|
| 542 |
"save_steps": 100,
|
| 543 |
"stateful_callbacks": {
|
| 544 |
"TrainerControl": {
|
|
@@ -547,12 +1142,12 @@
|
|
| 547 |
"should_evaluate": false,
|
| 548 |
"should_log": false,
|
| 549 |
"should_save": true,
|
| 550 |
-
"should_training_stop":
|
| 551 |
},
|
| 552 |
"attributes": {}
|
| 553 |
}
|
| 554 |
},
|
| 555 |
-
"total_flos":
|
| 556 |
"train_batch_size": 2,
|
| 557 |
"trial_name": null,
|
| 558 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.8921103986618344,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 800,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.005575689991636465,
|
| 14 |
+
"grad_norm": 0.16284966468811035,
|
| 15 |
+
"learning_rate": 7.142857142857143e-06,
|
| 16 |
+
"loss": 0.0626,
|
| 17 |
"step": 5
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.01115137998327293,
|
| 21 |
+
"grad_norm": 0.10163773596286774,
|
| 22 |
+
"learning_rate": 1.6071428571428572e-05,
|
| 23 |
+
"loss": 0.051,
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.016727069974909393,
|
| 28 |
+
"grad_norm": 0.12601175904273987,
|
| 29 |
+
"learning_rate": 2.5e-05,
|
| 30 |
+
"loss": 0.041,
|
| 31 |
"step": 15
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.02230275996654586,
|
| 35 |
+
"grad_norm": 0.04396357387304306,
|
| 36 |
+
"learning_rate": 3.392857142857143e-05,
|
| 37 |
+
"loss": 0.0336,
|
| 38 |
"step": 20
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.027878449958182325,
|
| 42 |
+
"grad_norm": 0.04541896656155586,
|
| 43 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 44 |
+
"loss": 0.03,
|
| 45 |
"step": 25
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.03345413994981879,
|
| 49 |
+
"grad_norm": 0.035561174154281616,
|
| 50 |
+
"learning_rate": 5.1785714285714296e-05,
|
| 51 |
+
"loss": 0.026,
|
| 52 |
"step": 30
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.039029829941455256,
|
| 56 |
+
"grad_norm": 0.04301896691322327,
|
| 57 |
+
"learning_rate": 6.0714285714285715e-05,
|
| 58 |
+
"loss": 0.0232,
|
| 59 |
"step": 35
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.04460551993309172,
|
| 63 |
+
"grad_norm": 0.04101714491844177,
|
| 64 |
+
"learning_rate": 6.964285714285715e-05,
|
| 65 |
+
"loss": 0.0234,
|
| 66 |
"step": 40
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 0.05018120992472819,
|
| 70 |
+
"grad_norm": 0.03225620836019516,
|
| 71 |
+
"learning_rate": 7.857142857142858e-05,
|
| 72 |
+
"loss": 0.0199,
|
| 73 |
"step": 45
|
| 74 |
},
|
| 75 |
{
|
| 76 |
+
"epoch": 0.05575689991636465,
|
| 77 |
+
"grad_norm": 0.03156769648194313,
|
| 78 |
+
"learning_rate": 8.75e-05,
|
| 79 |
+
"loss": 0.02,
|
| 80 |
"step": 50
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.06133258990800112,
|
| 84 |
+
"grad_norm": 0.03408223018050194,
|
| 85 |
+
"learning_rate": 9.642857142857143e-05,
|
| 86 |
+
"loss": 0.0227,
|
| 87 |
"step": 55
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.06690827989963757,
|
| 91 |
+
"grad_norm": 0.031089385971426964,
|
| 92 |
+
"learning_rate": 9.999803846452024e-05,
|
| 93 |
+
"loss": 0.02,
|
| 94 |
"step": 60
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 0.07248396989127405,
|
| 98 |
+
"grad_norm": 0.033274080604314804,
|
| 99 |
+
"learning_rate": 9.998605186060137e-05,
|
| 100 |
+
"loss": 0.014,
|
| 101 |
"step": 65
|
| 102 |
},
|
| 103 |
{
|
| 104 |
+
"epoch": 0.07805965988291051,
|
| 105 |
+
"grad_norm": 0.02827683836221695,
|
| 106 |
+
"learning_rate": 9.996317100396068e-05,
|
| 107 |
+
"loss": 0.0202,
|
| 108 |
"step": 70
|
| 109 |
},
|
| 110 |
{
|
| 111 |
+
"epoch": 0.08363534987454697,
|
| 112 |
+
"grad_norm": 0.037661969661712646,
|
| 113 |
+
"learning_rate": 9.992940088138597e-05,
|
| 114 |
+
"loss": 0.0222,
|
| 115 |
"step": 75
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 0.08921103986618344,
|
| 119 |
+
"grad_norm": 0.04279659315943718,
|
| 120 |
+
"learning_rate": 9.988474885293544e-05,
|
| 121 |
+
"loss": 0.0186,
|
| 122 |
"step": 80
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 0.0947867298578199,
|
| 126 |
+
"grad_norm": 0.026112260296940804,
|
| 127 |
+
"learning_rate": 9.98292246503335e-05,
|
| 128 |
+
"loss": 0.02,
|
| 129 |
"step": 85
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.10036241984945637,
|
| 133 |
+
"grad_norm": 0.029179614037275314,
|
| 134 |
+
"learning_rate": 9.976284037484988e-05,
|
| 135 |
+
"loss": 0.0175,
|
| 136 |
"step": 90
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 0.10593810984109284,
|
| 140 |
+
"grad_norm": 0.035265687853097916,
|
| 141 |
+
"learning_rate": 9.968561049466214e-05,
|
| 142 |
+
"loss": 0.0163,
|
| 143 |
"step": 95
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"epoch": 0.1115137998327293,
|
| 147 |
+
"grad_norm": 0.02659149281680584,
|
| 148 |
+
"learning_rate": 9.95975518417024e-05,
|
| 149 |
+
"loss": 0.0134,
|
| 150 |
"step": 100
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"epoch": 0.11708948982436576,
|
| 154 |
+
"grad_norm": 0.03535303473472595,
|
| 155 |
+
"learning_rate": 9.949868360798893e-05,
|
| 156 |
+
"loss": 0.0174,
|
| 157 |
"step": 105
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 0.12266517981600224,
|
| 161 |
+
"grad_norm": 0.028047222644090652,
|
| 162 |
+
"learning_rate": 9.938902734144326e-05,
|
| 163 |
+
"loss": 0.014,
|
| 164 |
"step": 110
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.12824086980763869,
|
| 168 |
+
"grad_norm": 0.025049524381756783,
|
| 169 |
+
"learning_rate": 9.926860694119398e-05,
|
| 170 |
+
"loss": 0.014,
|
| 171 |
"step": 115
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.13381655979927515,
|
| 175 |
+
"grad_norm": 0.03536759689450264,
|
| 176 |
+
"learning_rate": 9.913744865236798e-05,
|
| 177 |
+
"loss": 0.0161,
|
| 178 |
"step": 120
|
| 179 |
},
|
| 180 |
{
|
| 181 |
+
"epoch": 0.13939224979091164,
|
| 182 |
+
"grad_norm": 0.03142830356955528,
|
| 183 |
+
"learning_rate": 9.899558106037039e-05,
|
| 184 |
+
"loss": 0.016,
|
| 185 |
"step": 125
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 0.1449679397825481,
|
| 189 |
+
"grad_norm": 0.029837962239980698,
|
| 190 |
+
"learning_rate": 9.884303508465463e-05,
|
| 191 |
+
"loss": 0.017,
|
| 192 |
"step": 130
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 0.15054362977418456,
|
| 196 |
+
"grad_norm": 0.050843242555856705,
|
| 197 |
+
"learning_rate": 9.867984397198348e-05,
|
| 198 |
+
"loss": 0.0209,
|
| 199 |
"step": 135
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 0.15611931976582102,
|
| 203 |
+
"grad_norm": 0.022020680829882622,
|
| 204 |
+
"learning_rate": 9.85060432891833e-05,
|
| 205 |
+
"loss": 0.013,
|
| 206 |
"step": 140
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 0.1616950097574575,
|
| 210 |
+
"grad_norm": 0.03178994357585907,
|
| 211 |
+
"learning_rate": 9.832167091539214e-05,
|
| 212 |
+
"loss": 0.0144,
|
| 213 |
"step": 145
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 0.16727069974909395,
|
| 217 |
+
"grad_norm": 0.030585451051592827,
|
| 218 |
+
"learning_rate": 9.812676703380433e-05,
|
| 219 |
+
"loss": 0.0144,
|
| 220 |
"step": 150
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"epoch": 0.1728463897407304,
|
| 224 |
+
"grad_norm": 0.02348562888801098,
|
| 225 |
+
"learning_rate": 9.792137412291265e-05,
|
| 226 |
+
"loss": 0.0137,
|
| 227 |
"step": 155
|
| 228 |
},
|
| 229 |
{
|
| 230 |
+
"epoch": 0.17842207973236687,
|
| 231 |
+
"grad_norm": 0.028541121631860733,
|
| 232 |
+
"learning_rate": 9.770553694725028e-05,
|
| 233 |
+
"loss": 0.0169,
|
| 234 |
"step": 160
|
| 235 |
},
|
| 236 |
{
|
| 237 |
+
"epoch": 0.18399776972400333,
|
| 238 |
+
"grad_norm": 0.02496708557009697,
|
| 239 |
+
"learning_rate": 9.747930254763467e-05,
|
| 240 |
+
"loss": 0.013,
|
| 241 |
"step": 165
|
| 242 |
},
|
| 243 |
{
|
| 244 |
+
"epoch": 0.1895734597156398,
|
| 245 |
+
"grad_norm": 0.03572074696421623,
|
| 246 |
+
"learning_rate": 9.724272023091503e-05,
|
| 247 |
+
"loss": 0.0164,
|
| 248 |
"step": 170
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 0.1951491497072763,
|
| 252 |
+
"grad_norm": 0.02730608731508255,
|
| 253 |
+
"learning_rate": 9.699584155922625e-05,
|
| 254 |
+
"loss": 0.0135,
|
| 255 |
"step": 175
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 0.20072483969891275,
|
| 259 |
+
"grad_norm": 0.03099130466580391,
|
| 260 |
+
"learning_rate": 9.673872033875109e-05,
|
| 261 |
+
"loss": 0.0157,
|
| 262 |
"step": 180
|
| 263 |
},
|
| 264 |
{
|
| 265 |
+
"epoch": 0.2063005296905492,
|
| 266 |
+
"grad_norm": 0.031458914279937744,
|
| 267 |
+
"learning_rate": 9.64714126079933e-05,
|
| 268 |
+
"loss": 0.0138,
|
| 269 |
"step": 185
|
| 270 |
},
|
| 271 |
{
|
| 272 |
+
"epoch": 0.21187621968218567,
|
| 273 |
+
"grad_norm": 0.03125375509262085,
|
| 274 |
+
"learning_rate": 9.619397662556435e-05,
|
| 275 |
+
"loss": 0.0114,
|
| 276 |
"step": 190
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 0.21745190967382214,
|
| 280 |
+
"grad_norm": 0.031778380274772644,
|
| 281 |
+
"learning_rate": 9.590647285748613e-05,
|
| 282 |
+
"loss": 0.0117,
|
| 283 |
"step": 195
|
| 284 |
},
|
| 285 |
{
|
| 286 |
+
"epoch": 0.2230275996654586,
|
| 287 |
+
"grad_norm": 0.019305897876620293,
|
| 288 |
+
"learning_rate": 9.56089639640127e-05,
|
| 289 |
+
"loss": 0.0143,
|
| 290 |
"step": 200
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 0.22860328965709506,
|
| 294 |
+
"grad_norm": 0.02124331146478653,
|
| 295 |
+
"learning_rate": 9.530151478597366e-05,
|
| 296 |
+
"loss": 0.0135,
|
| 297 |
"step": 205
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 0.23417897964873152,
|
| 301 |
+
"grad_norm": 0.033200375735759735,
|
| 302 |
+
"learning_rate": 9.498419233064246e-05,
|
| 303 |
+
"loss": 0.0143,
|
| 304 |
"step": 210
|
| 305 |
},
|
| 306 |
{
|
| 307 |
+
"epoch": 0.23975466964036798,
|
| 308 |
+
"grad_norm": 0.03514528647065163,
|
| 309 |
+
"learning_rate": 9.465706575713236e-05,
|
| 310 |
+
"loss": 0.0204,
|
| 311 |
"step": 215
|
| 312 |
},
|
| 313 |
{
|
| 314 |
+
"epoch": 0.24533035963200447,
|
| 315 |
+
"grad_norm": 0.029311848804354668,
|
| 316 |
+
"learning_rate": 9.432020636132354e-05,
|
| 317 |
+
"loss": 0.0176,
|
| 318 |
"step": 220
|
| 319 |
},
|
| 320 |
{
|
| 321 |
+
"epoch": 0.25090604962364094,
|
| 322 |
+
"grad_norm": 0.030977483838796616,
|
| 323 |
+
"learning_rate": 9.397368756032445e-05,
|
| 324 |
+
"loss": 0.0146,
|
| 325 |
"step": 225
|
| 326 |
},
|
| 327 |
{
|
| 328 |
+
"epoch": 0.25648173961527737,
|
| 329 |
+
"grad_norm": 0.025566186755895615,
|
| 330 |
+
"learning_rate": 9.361758487647082e-05,
|
| 331 |
+
"loss": 0.0136,
|
| 332 |
"step": 230
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 0.26205742960691386,
|
| 336 |
+
"grad_norm": 0.02457290142774582,
|
| 337 |
+
"learning_rate": 9.32519759208659e-05,
|
| 338 |
+
"loss": 0.0152,
|
| 339 |
"step": 235
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 0.2676331195985503,
|
| 343 |
+
"grad_norm": 0.023812102153897285,
|
| 344 |
+
"learning_rate": 9.287694037646548e-05,
|
| 345 |
+
"loss": 0.0148,
|
| 346 |
"step": 240
|
| 347 |
},
|
| 348 |
{
|
| 349 |
+
"epoch": 0.2732088095901868,
|
| 350 |
+
"grad_norm": 0.023294365033507347,
|
| 351 |
+
"learning_rate": 9.249255998071126e-05,
|
| 352 |
+
"loss": 0.0123,
|
| 353 |
"step": 245
|
| 354 |
},
|
| 355 |
{
|
| 356 |
+
"epoch": 0.2787844995818233,
|
| 357 |
+
"grad_norm": 0.018993759527802467,
|
| 358 |
+
"learning_rate": 9.209891850771657e-05,
|
| 359 |
+
"loss": 0.0099,
|
| 360 |
"step": 250
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"epoch": 0.2843601895734597,
|
| 364 |
+
"grad_norm": 0.034646522253751755,
|
| 365 |
+
"learning_rate": 9.169610175000812e-05,
|
| 366 |
+
"loss": 0.0139,
|
| 367 |
"step": 255
|
| 368 |
},
|
| 369 |
{
|
| 370 |
+
"epoch": 0.2899358795650962,
|
| 371 |
+
"grad_norm": 0.029509609565138817,
|
| 372 |
+
"learning_rate": 9.12841974998278e-05,
|
| 373 |
+
"loss": 0.0117,
|
| 374 |
"step": 260
|
| 375 |
},
|
| 376 |
{
|
| 377 |
+
"epoch": 0.29551156955673263,
|
| 378 |
+
"grad_norm": 0.024864595383405685,
|
| 379 |
+
"learning_rate": 9.086329552999891e-05,
|
| 380 |
+
"loss": 0.0146,
|
| 381 |
"step": 265
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 0.3010872595483691,
|
| 385 |
+
"grad_norm": 0.023953670635819435,
|
| 386 |
+
"learning_rate": 9.043348757436037e-05,
|
| 387 |
+
"loss": 0.0131,
|
| 388 |
"step": 270
|
| 389 |
},
|
| 390 |
{
|
| 391 |
+
"epoch": 0.30666294954000556,
|
| 392 |
+
"grad_norm": 0.019653376191854477,
|
| 393 |
+
"learning_rate": 8.99948673077738e-05,
|
| 394 |
+
"loss": 0.0107,
|
| 395 |
"step": 275
|
| 396 |
},
|
| 397 |
{
|
| 398 |
+
"epoch": 0.31223863953164205,
|
| 399 |
+
"grad_norm": 0.03767814487218857,
|
| 400 |
+
"learning_rate": 8.954753032570742e-05,
|
| 401 |
+
"loss": 0.0143,
|
| 402 |
"step": 280
|
| 403 |
},
|
| 404 |
{
|
| 405 |
+
"epoch": 0.3178143295232785,
|
| 406 |
+
"grad_norm": 0.021987926214933395,
|
| 407 |
+
"learning_rate": 8.90915741234015e-05,
|
| 408 |
+
"loss": 0.0098,
|
| 409 |
"step": 285
|
| 410 |
},
|
| 411 |
{
|
| 412 |
+
"epoch": 0.323390019514915,
|
| 413 |
+
"grad_norm": 0.021244384348392487,
|
| 414 |
+
"learning_rate": 8.862709807461956e-05,
|
| 415 |
+
"loss": 0.0106,
|
| 416 |
"step": 290
|
| 417 |
},
|
| 418 |
{
|
| 419 |
+
"epoch": 0.32896570950655146,
|
| 420 |
+
"grad_norm": 0.02844163216650486,
|
| 421 |
+
"learning_rate": 8.815420340999033e-05,
|
| 422 |
+
"loss": 0.0162,
|
| 423 |
"step": 295
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 0.3345413994981879,
|
| 427 |
+
"grad_norm": 0.02485722117125988,
|
| 428 |
+
"learning_rate": 8.767299319494503e-05,
|
| 429 |
+
"loss": 0.0164,
|
| 430 |
"step": 300
|
| 431 |
},
|
| 432 |
{
|
| 433 |
+
"epoch": 0.3401170894898244,
|
| 434 |
+
"grad_norm": 0.019627615809440613,
|
| 435 |
+
"learning_rate": 8.718357230725449e-05,
|
| 436 |
+
"loss": 0.0139,
|
| 437 |
"step": 305
|
| 438 |
},
|
| 439 |
{
|
| 440 |
+
"epoch": 0.3456927794814608,
|
| 441 |
+
"grad_norm": 0.02517726831138134,
|
| 442 |
+
"learning_rate": 8.668604741417171e-05,
|
| 443 |
+
"loss": 0.0128,
|
| 444 |
"step": 310
|
| 445 |
},
|
| 446 |
{
|
| 447 |
+
"epoch": 0.3512684694730973,
|
| 448 |
+
"grad_norm": 0.02175074815750122,
|
| 449 |
+
"learning_rate": 8.618052694918399e-05,
|
| 450 |
+
"loss": 0.0111,
|
| 451 |
"step": 315
|
| 452 |
},
|
| 453 |
{
|
| 454 |
+
"epoch": 0.35684415946473375,
|
| 455 |
+
"grad_norm": 0.021222930401563644,
|
| 456 |
+
"learning_rate": 8.566712108838042e-05,
|
| 457 |
+
"loss": 0.0111,
|
| 458 |
"step": 320
|
| 459 |
},
|
| 460 |
{
|
| 461 |
+
"epoch": 0.36241984945637024,
|
| 462 |
+
"grad_norm": 0.024494808167219162,
|
| 463 |
+
"learning_rate": 8.514594172643934e-05,
|
| 464 |
+
"loss": 0.0138,
|
| 465 |
"step": 325
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 0.36799553944800667,
|
| 469 |
+
"grad_norm": 0.022174010053277016,
|
| 470 |
+
"learning_rate": 8.461710245224148e-05,
|
| 471 |
+
"loss": 0.0134,
|
| 472 |
"step": 330
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"epoch": 0.37357122943964316,
|
| 476 |
+
"grad_norm": 0.01959528774023056,
|
| 477 |
+
"learning_rate": 8.40807185241137e-05,
|
| 478 |
+
"loss": 0.0102,
|
| 479 |
"step": 335
|
| 480 |
},
|
| 481 |
{
|
| 482 |
+
"epoch": 0.3791469194312796,
|
| 483 |
+
"grad_norm": 0.017945902422070503,
|
| 484 |
+
"learning_rate": 8.353690684470884e-05,
|
| 485 |
+
"loss": 0.0143,
|
| 486 |
"step": 340
|
| 487 |
},
|
| 488 |
{
|
| 489 |
+
"epoch": 0.3847226094229161,
|
| 490 |
+
"grad_norm": 0.020864926278591156,
|
| 491 |
+
"learning_rate": 8.298578593552737e-05,
|
| 492 |
+
"loss": 0.0179,
|
| 493 |
"step": 345
|
| 494 |
},
|
| 495 |
{
|
| 496 |
+
"epoch": 0.3902982994145526,
|
| 497 |
+
"grad_norm": 0.027325566858053207,
|
| 498 |
+
"learning_rate": 8.242747591108605e-05,
|
| 499 |
+
"loss": 0.0133,
|
| 500 |
"step": 350
|
| 501 |
},
|
| 502 |
{
|
| 503 |
+
"epoch": 0.395873989406189,
|
| 504 |
+
"grad_norm": 0.019658569246530533,
|
| 505 |
+
"learning_rate": 8.186209845273954e-05,
|
| 506 |
+
"loss": 0.0139,
|
| 507 |
"step": 355
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 0.4014496793978255,
|
| 511 |
+
"grad_norm": 0.02014886401593685,
|
| 512 |
+
"learning_rate": 8.128977678216039e-05,
|
| 513 |
+
"loss": 0.009,
|
| 514 |
"step": 360
|
| 515 |
},
|
| 516 |
{
|
| 517 |
+
"epoch": 0.40702536938946193,
|
| 518 |
+
"grad_norm": 0.02425803802907467,
|
| 519 |
+
"learning_rate": 8.07106356344834e-05,
|
| 520 |
+
"loss": 0.0125,
|
| 521 |
"step": 365
|
| 522 |
},
|
| 523 |
{
|
| 524 |
+
"epoch": 0.4126010593810984,
|
| 525 |
+
"grad_norm": 0.030235106125473976,
|
| 526 |
+
"learning_rate": 8.012480123112014e-05,
|
| 527 |
+
"loss": 0.0171,
|
| 528 |
"step": 370
|
| 529 |
},
|
| 530 |
{
|
| 531 |
+
"epoch": 0.41817674937273486,
|
| 532 |
+
"grad_norm": 0.022229960188269615,
|
| 533 |
+
"learning_rate": 7.953240125224948e-05,
|
| 534 |
+
"loss": 0.0116,
|
| 535 |
"step": 375
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.42375243936437135,
|
| 539 |
+
"grad_norm": 0.025448938831686974,
|
| 540 |
+
"learning_rate": 7.89335648089903e-05,
|
| 541 |
+
"loss": 0.0142,
|
| 542 |
+
"step": 380
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.4293281293560078,
|
| 546 |
+
"grad_norm": 0.023552658036351204,
|
| 547 |
+
"learning_rate": 7.832842241526212e-05,
|
| 548 |
+
"loss": 0.0147,
|
| 549 |
+
"step": 385
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.43490381934764427,
|
| 553 |
+
"grad_norm": 0.019487692043185234,
|
| 554 |
+
"learning_rate": 7.77171059593403e-05,
|
| 555 |
+
"loss": 0.0115,
|
| 556 |
+
"step": 390
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.44047950933928076,
|
| 560 |
+
"grad_norm": 0.021791953593492508,
|
| 561 |
+
"learning_rate": 7.709974867511138e-05,
|
| 562 |
+
"loss": 0.012,
|
| 563 |
+
"step": 395
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.4460551993309172,
|
| 567 |
+
"grad_norm": 0.02281327173113823,
|
| 568 |
+
"learning_rate": 7.647648511303544e-05,
|
| 569 |
+
"loss": 0.0126,
|
| 570 |
+
"step": 400
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.4516308893225537,
|
| 574 |
+
"grad_norm": 0.02623576670885086,
|
| 575 |
+
"learning_rate": 7.584745111082127e-05,
|
| 576 |
+
"loss": 0.0128,
|
| 577 |
+
"step": 405
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.4572065793141901,
|
| 581 |
+
"grad_norm": 0.019894316792488098,
|
| 582 |
+
"learning_rate": 7.521278376382123e-05,
|
| 583 |
+
"loss": 0.0092,
|
| 584 |
+
"step": 410
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.4627822693058266,
|
| 588 |
+
"grad_norm": 0.022427916526794434,
|
| 589 |
+
"learning_rate": 7.457262139515171e-05,
|
| 590 |
+
"loss": 0.0111,
|
| 591 |
+
"step": 415
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.46835795929746304,
|
| 595 |
+
"grad_norm": 0.021067511290311813,
|
| 596 |
+
"learning_rate": 7.392710352554641e-05,
|
| 597 |
+
"loss": 0.0099,
|
| 598 |
+
"step": 420
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.47393364928909953,
|
| 602 |
+
"grad_norm": 0.019623806700110435,
|
| 603 |
+
"learning_rate": 7.327637084294817e-05,
|
| 604 |
+
"loss": 0.012,
|
| 605 |
+
"step": 425
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.47950933928073597,
|
| 609 |
+
"grad_norm": 0.02039971947669983,
|
| 610 |
+
"learning_rate": 7.262056517184669e-05,
|
| 611 |
+
"loss": 0.0138,
|
| 612 |
+
"step": 430
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.48508502927237246,
|
| 616 |
+
"grad_norm": 0.021388281136751175,
|
| 617 |
+
"learning_rate": 7.195982944236851e-05,
|
| 618 |
+
"loss": 0.0123,
|
| 619 |
+
"step": 435
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.49066071926400895,
|
| 623 |
+
"grad_norm": 0.022272834554314613,
|
| 624 |
+
"learning_rate": 7.1294307659126e-05,
|
| 625 |
+
"loss": 0.0126,
|
| 626 |
+
"step": 440
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.4962364092556454,
|
| 630 |
+
"grad_norm": 0.02803129144012928,
|
| 631 |
+
"learning_rate": 7.062414486983197e-05,
|
| 632 |
+
"loss": 0.0118,
|
| 633 |
+
"step": 445
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.5018120992472819,
|
| 637 |
+
"grad_norm": 0.025339094921946526,
|
| 638 |
+
"learning_rate": 6.994948713368737e-05,
|
| 639 |
+
"loss": 0.0147,
|
| 640 |
+
"step": 450
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.5073877892389184,
|
| 644 |
+
"grad_norm": 0.024465398862957954,
|
| 645 |
+
"learning_rate": 6.927048148954812e-05,
|
| 646 |
+
"loss": 0.0118,
|
| 647 |
+
"step": 455
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.5129634792305547,
|
| 651 |
+
"grad_norm": 0.025315098464488983,
|
| 652 |
+
"learning_rate": 6.858727592387867e-05,
|
| 653 |
+
"loss": 0.0165,
|
| 654 |
+
"step": 460
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5185391692221912,
|
| 658 |
+
"grad_norm": 0.020109234377741814,
|
| 659 |
+
"learning_rate": 6.790001933849899e-05,
|
| 660 |
+
"loss": 0.0108,
|
| 661 |
+
"step": 465
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5241148592138277,
|
| 665 |
+
"grad_norm": 0.01888495869934559,
|
| 666 |
+
"learning_rate": 6.720886151813194e-05,
|
| 667 |
+
"loss": 0.0097,
|
| 668 |
+
"step": 470
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.5296905492054642,
|
| 672 |
+
"grad_norm": 0.023433005437254906,
|
| 673 |
+
"learning_rate": 6.651395309775837e-05,
|
| 674 |
+
"loss": 0.0122,
|
| 675 |
+
"step": 475
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.5352662391971006,
|
| 679 |
+
"grad_norm": 0.019298607483506203,
|
| 680 |
+
"learning_rate": 6.581544552978687e-05,
|
| 681 |
+
"loss": 0.0134,
|
| 682 |
+
"step": 480
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.5408419291887371,
|
| 686 |
+
"grad_norm": 0.025588713586330414,
|
| 687 |
+
"learning_rate": 6.511349105104534e-05,
|
| 688 |
+
"loss": 0.0108,
|
| 689 |
+
"step": 485
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.5464176191803736,
|
| 693 |
+
"grad_norm": 0.03148540109395981,
|
| 694 |
+
"learning_rate": 6.440824264960157e-05,
|
| 695 |
+
"loss": 0.0115,
|
| 696 |
+
"step": 490
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.5519933091720101,
|
| 700 |
+
"grad_norm": 0.01748904027044773,
|
| 701 |
+
"learning_rate": 6.369985403142014e-05,
|
| 702 |
+
"loss": 0.0112,
|
| 703 |
+
"step": 495
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.5575689991636466,
|
| 707 |
+
"grad_norm": 0.027883267030119896,
|
| 708 |
+
"learning_rate": 6.298847958686283e-05,
|
| 709 |
+
"loss": 0.0125,
|
| 710 |
+
"step": 500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.5631446891552829,
|
| 714 |
+
"grad_norm": 0.02129966951906681,
|
| 715 |
+
"learning_rate": 6.227427435703997e-05,
|
| 716 |
+
"loss": 0.0149,
|
| 717 |
+
"step": 505
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5687203791469194,
|
| 721 |
+
"grad_norm": 0.02562125027179718,
|
| 722 |
+
"learning_rate": 6.15573940000197e-05,
|
| 723 |
+
"loss": 0.0136,
|
| 724 |
+
"step": 510
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5742960691385559,
|
| 728 |
+
"grad_norm": 0.02446940541267395,
|
| 729 |
+
"learning_rate": 6.083799475690309e-05,
|
| 730 |
+
"loss": 0.0112,
|
| 731 |
+
"step": 515
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5798717591301924,
|
| 735 |
+
"grad_norm": 0.024061646312475204,
|
| 736 |
+
"learning_rate": 6.0116233417771994e-05,
|
| 737 |
+
"loss": 0.0115,
|
| 738 |
+
"step": 520
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.5854474491218288,
|
| 742 |
+
"grad_norm": 0.01640748232603073,
|
| 743 |
+
"learning_rate": 5.9392267287517325e-05,
|
| 744 |
+
"loss": 0.0103,
|
| 745 |
+
"step": 525
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.5910231391134653,
|
| 749 |
+
"grad_norm": 0.023191062733530998,
|
| 750 |
+
"learning_rate": 5.8666254151554976e-05,
|
| 751 |
+
"loss": 0.0113,
|
| 752 |
+
"step": 530
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.5965988291051018,
|
| 756 |
+
"grad_norm": 0.017516452819108963,
|
| 757 |
+
"learning_rate": 5.7938352241437366e-05,
|
| 758 |
+
"loss": 0.0093,
|
| 759 |
+
"step": 535
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6021745190967382,
|
| 763 |
+
"grad_norm": 0.019351812079548836,
|
| 764 |
+
"learning_rate": 5.720872020036734e-05,
|
| 765 |
+
"loss": 0.0125,
|
| 766 |
+
"step": 540
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6077502090883747,
|
| 770 |
+
"grad_norm": 0.029706666246056557,
|
| 771 |
+
"learning_rate": 5.647751704862263e-05,
|
| 772 |
+
"loss": 0.008,
|
| 773 |
+
"step": 545
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.6133258990800111,
|
| 777 |
+
"grad_norm": 0.016750017181038857,
|
| 778 |
+
"learning_rate": 5.5744902148898005e-05,
|
| 779 |
+
"loss": 0.0118,
|
| 780 |
+
"step": 550
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.6189015890716476,
|
| 784 |
+
"grad_norm": 0.022833596915006638,
|
| 785 |
+
"learning_rate": 5.501103517157288e-05,
|
| 786 |
+
"loss": 0.0088,
|
| 787 |
+
"step": 555
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.6244772790632841,
|
| 791 |
+
"grad_norm": 0.03475171700119972,
|
| 792 |
+
"learning_rate": 5.427607605991176e-05,
|
| 793 |
+
"loss": 0.0136,
|
| 794 |
+
"step": 560
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.6300529690549206,
|
| 798 |
+
"grad_norm": 0.021340183913707733,
|
| 799 |
+
"learning_rate": 5.354018499520536e-05,
|
| 800 |
+
"loss": 0.0103,
|
| 801 |
+
"step": 565
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.635628659046557,
|
| 805 |
+
"grad_norm": 0.024497641250491142,
|
| 806 |
+
"learning_rate": 5.2803522361859594e-05,
|
| 807 |
+
"loss": 0.011,
|
| 808 |
+
"step": 570
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.6412043490381935,
|
| 812 |
+
"grad_norm": 0.01924068294465542,
|
| 813 |
+
"learning_rate": 5.2066248712440656e-05,
|
| 814 |
+
"loss": 0.0125,
|
| 815 |
+
"step": 575
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.64678003902983,
|
| 819 |
+
"grad_norm": 0.017638731747865677,
|
| 820 |
+
"learning_rate": 5.1328524732683134e-05,
|
| 821 |
+
"loss": 0.0104,
|
| 822 |
+
"step": 580
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.6523557290214664,
|
| 826 |
+
"grad_norm": 0.022175751626491547,
|
| 827 |
+
"learning_rate": 5.059051120646924e-05,
|
| 828 |
+
"loss": 0.0128,
|
| 829 |
+
"step": 585
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.6579314190131029,
|
| 833 |
+
"grad_norm": 0.022414250299334526,
|
| 834 |
+
"learning_rate": 4.985236898078658e-05,
|
| 835 |
+
"loss": 0.0128,
|
| 836 |
+
"step": 590
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.6635071090047393,
|
| 840 |
+
"grad_norm": 0.020940134301781654,
|
| 841 |
+
"learning_rate": 4.911425893067239e-05,
|
| 842 |
+
"loss": 0.0124,
|
| 843 |
+
"step": 595
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.6690827989963758,
|
| 847 |
+
"grad_norm": 0.021777737885713577,
|
| 848 |
+
"learning_rate": 4.837634192415128e-05,
|
| 849 |
+
"loss": 0.0126,
|
| 850 |
+
"step": 600
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.6746584889880123,
|
| 854 |
+
"grad_norm": 0.01768389716744423,
|
| 855 |
+
"learning_rate": 4.763877878717484e-05,
|
| 856 |
+
"loss": 0.0095,
|
| 857 |
+
"step": 605
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.6802341789796488,
|
| 861 |
+
"grad_norm": 0.02011968567967415,
|
| 862 |
+
"learning_rate": 4.6901730268570275e-05,
|
| 863 |
+
"loss": 0.0093,
|
| 864 |
+
"step": 610
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.6858098689712852,
|
| 868 |
+
"grad_norm": 0.02239886298775673,
|
| 869 |
+
"learning_rate": 4.616535700500583e-05,
|
| 870 |
+
"loss": 0.0126,
|
| 871 |
+
"step": 615
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.6913855589629216,
|
| 875 |
+
"grad_norm": 0.022233402356505394,
|
| 876 |
+
"learning_rate": 4.542981948598071e-05,
|
| 877 |
+
"loss": 0.0107,
|
| 878 |
+
"step": 620
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.6969612489545581,
|
| 882 |
+
"grad_norm": 0.027380308136343956,
|
| 883 |
+
"learning_rate": 4.4695278018847105e-05,
|
| 884 |
+
"loss": 0.0142,
|
| 885 |
+
"step": 625
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.7025369389461946,
|
| 889 |
+
"grad_norm": 0.025442643091082573,
|
| 890 |
+
"learning_rate": 4.396189269387176e-05,
|
| 891 |
+
"loss": 0.0153,
|
| 892 |
+
"step": 630
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.708112628937831,
|
| 896 |
+
"grad_norm": 0.01852184161543846,
|
| 897 |
+
"learning_rate": 4.322982334934509e-05,
|
| 898 |
+
"loss": 0.0102,
|
| 899 |
+
"step": 635
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.7136883189294675,
|
| 903 |
+
"grad_norm": 0.01528929267078638,
|
| 904 |
+
"learning_rate": 4.2499229536744986e-05,
|
| 905 |
+
"loss": 0.0097,
|
| 906 |
+
"step": 640
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.719264008921104,
|
| 910 |
+
"grad_norm": 0.026008352637290955,
|
| 911 |
+
"learning_rate": 4.17702704859633e-05,
|
| 912 |
+
"loss": 0.0159,
|
| 913 |
+
"step": 645
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.7248396989127405,
|
| 917 |
+
"grad_norm": 0.018146734684705734,
|
| 918 |
+
"learning_rate": 4.104310507060234e-05,
|
| 919 |
+
"loss": 0.0095,
|
| 920 |
+
"step": 650
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.730415388904377,
|
| 924 |
+
"grad_norm": 0.022718293592333794,
|
| 925 |
+
"learning_rate": 4.0317891773348946e-05,
|
| 926 |
+
"loss": 0.0095,
|
| 927 |
+
"step": 655
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.7359910788960133,
|
| 931 |
+
"grad_norm": 0.02410387434065342,
|
| 932 |
+
"learning_rate": 3.959478865143397e-05,
|
| 933 |
+
"loss": 0.0109,
|
| 934 |
+
"step": 660
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.7415667688876498,
|
| 938 |
+
"grad_norm": 0.017437651753425598,
|
| 939 |
+
"learning_rate": 3.887395330218429e-05,
|
| 940 |
+
"loss": 0.0107,
|
| 941 |
+
"step": 665
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.7471424588792863,
|
| 945 |
+
"grad_norm": 0.020500419661402702,
|
| 946 |
+
"learning_rate": 3.815554282867513e-05,
|
| 947 |
+
"loss": 0.0107,
|
| 948 |
+
"step": 670
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.7527181488709228,
|
| 952 |
+
"grad_norm": 0.01553898025304079,
|
| 953 |
+
"learning_rate": 3.743971380549008e-05,
|
| 954 |
+
"loss": 0.0083,
|
| 955 |
+
"step": 675
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.7582938388625592,
|
| 959 |
+
"grad_norm": 0.020166153088212013,
|
| 960 |
+
"learning_rate": 3.67266222445964e-05,
|
| 961 |
+
"loss": 0.0111,
|
| 962 |
+
"step": 680
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.7638695288541957,
|
| 966 |
+
"grad_norm": 0.023076798766851425,
|
| 967 |
+
"learning_rate": 3.6016423561342706e-05,
|
| 968 |
+
"loss": 0.0128,
|
| 969 |
+
"step": 685
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.7694452188458322,
|
| 973 |
+
"grad_norm": 0.02140919119119644,
|
| 974 |
+
"learning_rate": 3.5309272540587e-05,
|
| 975 |
+
"loss": 0.0104,
|
| 976 |
+
"step": 690
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.7750209088374687,
|
| 980 |
+
"grad_norm": 0.016010567545890808,
|
| 981 |
+
"learning_rate": 3.4605323302961854e-05,
|
| 982 |
+
"loss": 0.0135,
|
| 983 |
+
"step": 695
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.7805965988291051,
|
| 987 |
+
"grad_norm": 0.017510782927274704,
|
| 988 |
+
"learning_rate": 3.3904729271284473e-05,
|
| 989 |
+
"loss": 0.0115,
|
| 990 |
+
"step": 700
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.7861722888207415,
|
| 994 |
+
"grad_norm": 0.024468230083584785,
|
| 995 |
+
"learning_rate": 3.3207643137118874e-05,
|
| 996 |
+
"loss": 0.01,
|
| 997 |
+
"step": 705
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.791747978812378,
|
| 1001 |
+
"grad_norm": 0.020976202562451363,
|
| 1002 |
+
"learning_rate": 3.251421682749732e-05,
|
| 1003 |
+
"loss": 0.0114,
|
| 1004 |
+
"step": 710
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.7973236688040145,
|
| 1008 |
+
"grad_norm": 0.02256660722196102,
|
| 1009 |
+
"learning_rate": 3.18246014718085e-05,
|
| 1010 |
+
"loss": 0.0108,
|
| 1011 |
+
"step": 715
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.802899358795651,
|
| 1015 |
+
"grad_norm": 0.030024701729416847,
|
| 1016 |
+
"learning_rate": 3.113894736885953e-05,
|
| 1017 |
+
"loss": 0.0104,
|
| 1018 |
+
"step": 720
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.8084750487872874,
|
| 1022 |
+
"grad_norm": 0.019312310963869095,
|
| 1023 |
+
"learning_rate": 3.0457403954118856e-05,
|
| 1024 |
+
"loss": 0.0082,
|
| 1025 |
+
"step": 725
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.8140507387789239,
|
| 1029 |
+
"grad_norm": 0.02008041925728321,
|
| 1030 |
+
"learning_rate": 2.978011976714753e-05,
|
| 1031 |
+
"loss": 0.0099,
|
| 1032 |
+
"step": 730
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.8196264287705604,
|
| 1036 |
+
"grad_norm": 0.021896323189139366,
|
| 1037 |
+
"learning_rate": 2.9107242419225577e-05,
|
| 1038 |
+
"loss": 0.0143,
|
| 1039 |
+
"step": 735
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.8252021187621968,
|
| 1043 |
+
"grad_norm": 0.019579166546463966,
|
| 1044 |
+
"learning_rate": 2.8438918561180634e-05,
|
| 1045 |
+
"loss": 0.0106,
|
| 1046 |
+
"step": 740
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.8307778087538333,
|
| 1050 |
+
"grad_norm": 0.02049921080470085,
|
| 1051 |
+
"learning_rate": 2.7775293851426232e-05,
|
| 1052 |
+
"loss": 0.0115,
|
| 1053 |
+
"step": 745
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.8363534987454697,
|
| 1057 |
+
"grad_norm": 0.014969157055020332,
|
| 1058 |
+
"learning_rate": 2.711651292421593e-05,
|
| 1059 |
+
"loss": 0.0101,
|
| 1060 |
+
"step": 750
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.8419291887371062,
|
| 1064 |
+
"grad_norm": 0.020416075363755226,
|
| 1065 |
+
"learning_rate": 2.646271935812098e-05,
|
| 1066 |
+
"loss": 0.0098,
|
| 1067 |
+
"step": 755
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.8475048787287427,
|
| 1071 |
+
"grad_norm": 0.018367785960435867,
|
| 1072 |
+
"learning_rate": 2.581405564473801e-05,
|
| 1073 |
+
"loss": 0.0165,
|
| 1074 |
+
"step": 760
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.8530805687203792,
|
| 1078 |
+
"grad_norm": 0.0190111193805933,
|
| 1079 |
+
"learning_rate": 2.5170663157633477e-05,
|
| 1080 |
+
"loss": 0.0135,
|
| 1081 |
+
"step": 765
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.8586562587120156,
|
| 1085 |
+
"grad_norm": 0.024806899949908257,
|
| 1086 |
+
"learning_rate": 2.45326821215319e-05,
|
| 1087 |
+
"loss": 0.0116,
|
| 1088 |
+
"step": 770
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.864231948703652,
|
| 1092 |
+
"grad_norm": 0.02073819749057293,
|
| 1093 |
+
"learning_rate": 2.390025158175458e-05,
|
| 1094 |
+
"loss": 0.0129,
|
| 1095 |
+
"step": 775
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.8698076386952885,
|
| 1099 |
+
"grad_norm": 0.02042596973478794,
|
| 1100 |
+
"learning_rate": 2.3273509373915093e-05,
|
| 1101 |
+
"loss": 0.0088,
|
| 1102 |
+
"step": 780
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.875383328686925,
|
| 1106 |
+
"grad_norm": 0.015911240130662918,
|
| 1107 |
+
"learning_rate": 2.2652592093878666e-05,
|
| 1108 |
+
"loss": 0.01,
|
| 1109 |
+
"step": 785
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.8809590186785615,
|
| 1113 |
+
"grad_norm": 0.023589760065078735,
|
| 1114 |
+
"learning_rate": 2.2037635067991663e-05,
|
| 1115 |
+
"loss": 0.0107,
|
| 1116 |
+
"step": 790
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.8865347086701979,
|
| 1120 |
+
"grad_norm": 0.017796384170651436,
|
| 1121 |
+
"learning_rate": 2.1428772323587827e-05,
|
| 1122 |
+
"loss": 0.0103,
|
| 1123 |
+
"step": 795
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.8921103986618344,
|
| 1127 |
+
"grad_norm": 0.018958481028676033,
|
| 1128 |
+
"learning_rate": 2.082613655977745e-05,
|
| 1129 |
+
"loss": 0.0079,
|
| 1130 |
+
"step": 800
|
| 1131 |
}
|
| 1132 |
],
|
| 1133 |
"logging_steps": 5,
|
| 1134 |
+
"max_steps": 1120,
|
| 1135 |
"num_input_tokens_seen": 0,
|
| 1136 |
+
"num_train_epochs": 2,
|
| 1137 |
"save_steps": 100,
|
| 1138 |
"stateful_callbacks": {
|
| 1139 |
"TrainerControl": {
|
|
|
|
| 1142 |
"should_evaluate": false,
|
| 1143 |
"should_log": false,
|
| 1144 |
"should_save": true,
|
| 1145 |
+
"should_training_stop": false
|
| 1146 |
},
|
| 1147 |
"attributes": {}
|
| 1148 |
}
|
| 1149 |
},
|
| 1150 |
+
"total_flos": 2.728813755050754e+18,
|
| 1151 |
"train_batch_size": 2,
|
| 1152 |
"trial_name": null,
|
| 1153 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7825
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8568d83effc9ae0e86eecad81d4c2bc1c32496e167d72cebf2a94b86d0aa123c
|
| 3 |
size 7825
|