Model save
Browse files
README.md
CHANGED
|
@@ -2,13 +2,12 @@
|
|
| 2 |
license: gemma
|
| 3 |
library_name: peft
|
| 4 |
tags:
|
| 5 |
-
- alignment-handbook
|
| 6 |
- trl
|
| 7 |
- sft
|
| 8 |
- generated_from_trainer
|
| 9 |
base_model: google/gemma-7b
|
| 10 |
datasets:
|
| 11 |
-
-
|
| 12 |
model-index:
|
| 13 |
- name: gemma7b-summarize-gemini1.5flash-30k
|
| 14 |
results: []
|
|
@@ -17,12 +16,12 @@ model-index:
|
|
| 17 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 18 |
should probably proofread and complete it, then remove this comment. -->
|
| 19 |
|
| 20 |
-
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/chansung18/huggingface/runs/
|
| 21 |
# gemma7b-summarize-gemini1.5flash-30k
|
| 22 |
|
| 23 |
-
This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on the
|
| 24 |
It achieves the following results on the evaluation set:
|
| 25 |
-
- Loss: 2.
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
|
@@ -53,13 +52,17 @@ The following hyperparameters were used during training:
|
|
| 53 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 54 |
- lr_scheduler_type: cosine
|
| 55 |
- lr_scheduler_warmup_ratio: 0.1
|
| 56 |
-
- num_epochs:
|
| 57 |
|
| 58 |
### Training results
|
| 59 |
|
| 60 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 61 |
|:-------------:|:------:|:----:|:---------------:|
|
| 62 |
-
| 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
| 63 |
|
| 64 |
|
| 65 |
### Framework versions
|
|
|
|
| 2 |
license: gemma
|
| 3 |
library_name: peft
|
| 4 |
tags:
|
|
|
|
| 5 |
- trl
|
| 6 |
- sft
|
| 7 |
- generated_from_trainer
|
| 8 |
base_model: google/gemma-7b
|
| 9 |
datasets:
|
| 10 |
+
- generator
|
| 11 |
model-index:
|
| 12 |
- name: gemma7b-summarize-gemini1.5flash-30k
|
| 13 |
results: []
|
|
|
|
| 16 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 17 |
should probably proofread and complete it, then remove this comment. -->
|
| 18 |
|
| 19 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/chansung18/huggingface/runs/go5tmp0o)
|
| 20 |
# gemma7b-summarize-gemini1.5flash-30k
|
| 21 |
|
| 22 |
+
This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on the generator dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 2.2363
|
| 25 |
|
| 26 |
## Model description
|
| 27 |
|
|
|
|
| 52 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 53 |
- lr_scheduler_type: cosine
|
| 54 |
- lr_scheduler_warmup_ratio: 0.1
|
| 55 |
+
- num_epochs: 5
|
| 56 |
|
| 57 |
### Training results
|
| 58 |
|
| 59 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 60 |
|:-------------:|:------:|:----:|:---------------:|
|
| 61 |
+
| 0.8665 | 0.9956 | 114 | 2.2684 |
|
| 62 |
+
| 0.7577 | 2.0 | 229 | 2.1722 |
|
| 63 |
+
| 0.6255 | 2.9956 | 343 | 2.1741 |
|
| 64 |
+
| 0.5966 | 4.0 | 458 | 2.2142 |
|
| 65 |
+
| 0.5399 | 4.9782 | 570 | 2.2363 |
|
| 66 |
|
| 67 |
|
| 68 |
### Framework versions
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 50056096
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b2a3661fec43802ab39a2d24104c2e05c4696bc0782ce0e1e7618281d3d9f59
|
| 3 |
size 50056096
|
all_results.json
CHANGED
|
@@ -1,14 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
-
"
|
| 6 |
-
"eval_samples_per_second": 4.837,
|
| 7 |
-
"eval_steps_per_second": 1.935,
|
| 8 |
-
"total_flos": 1.742915407875932e+17,
|
| 9 |
-
"train_loss": 5.99931854532476,
|
| 10 |
-
"train_runtime": 898.0288,
|
| 11 |
"train_samples": 30386,
|
| 12 |
-
"train_samples_per_second": 2.
|
| 13 |
-
"train_steps_per_second": 0.
|
| 14 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.978165938864628,
|
| 3 |
+
"total_flos": 8.714577082329334e+17,
|
| 4 |
+
"train_loss": 2.449154797771521,
|
| 5 |
+
"train_runtime": 4532.0541,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
"train_samples": 30386,
|
| 7 |
+
"train_samples_per_second": 2.02,
|
| 8 |
+
"train_steps_per_second": 0.126
|
| 9 |
}
|
runs/May19_11-56-52_deep-diver-main-moody-turtle-1-0-0/events.out.tfevents.1716134365.deep-diver-main-moody-turtle-1-0-0.385.0
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 sha256:99ddc755a1c43c0765721cf416c1e5d75047971f8526f3bf584163cd3e41967b
|
| 3 |
+
size 31492
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 30386,
|
| 7 |
-
"train_samples_per_second": 2.
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.978165938864628,
|
| 3 |
+
"total_flos": 8.714577082329334e+17,
|
| 4 |
+
"train_loss": 2.449154797771521,
|
| 5 |
+
"train_runtime": 4532.0541,
|
| 6 |
"train_samples": 30386,
|
| 7 |
+
"train_samples_per_second": 2.02,
|
| 8 |
+
"train_steps_per_second": 0.126
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -11,186 +11,862 @@
|
|
| 11 |
{
|
| 12 |
"epoch": 0.008733624454148471,
|
| 13 |
"grad_norm": 648.0,
|
| 14 |
-
"learning_rate":
|
| 15 |
"loss": 56.9969,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.043668122270742356,
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss":
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.08733624454148471,
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss":
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.13100436681222707,
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss":
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.17467248908296942,
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss":
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.2183406113537118,
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss":
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.26200873362445415,
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate": 0.
|
| 57 |
-
"loss":
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.3056768558951965,
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate": 0.
|
| 64 |
-
"loss":
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.34934497816593885,
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate": 0.
|
| 71 |
-
"loss":
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.3930131004366812,
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate": 0.
|
| 78 |
-
"loss":
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.4366812227074236,
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate": 0.
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.48034934497816595,
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate": 0.
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.5240174672489083,
|
| 97 |
-
"grad_norm": 1.
|
| 98 |
-
"learning_rate": 0.
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.5676855895196506,
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.611353711790393,
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.6550218340611353,
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss":
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.6986899563318777,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss":
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.74235807860262,
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.7860262008733624,
|
| 139 |
-
"grad_norm":
|
| 140 |
-
"learning_rate":
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.8296943231441049,
|
| 146 |
-
"grad_norm":
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.8733624454148472,
|
| 153 |
-
"grad_norm":
|
| 154 |
-
"learning_rate":
|
| 155 |
-
"loss": 0.
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.9170305676855895,
|
| 160 |
-
"grad_norm":
|
| 161 |
-
"learning_rate":
|
| 162 |
-
"loss": 0.
|
| 163 |
"step": 105
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.9606986899563319,
|
| 167 |
-
"grad_norm":
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 110
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.9956331877729258,
|
| 174 |
-
"eval_loss": 2.
|
| 175 |
-
"eval_runtime": 1.
|
| 176 |
-
"eval_samples_per_second": 4.
|
| 177 |
-
"eval_steps_per_second": 1.
|
| 178 |
"step": 114
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch":
|
| 182 |
-
"
|
| 183 |
-
"
|
| 184 |
-
"
|
| 185 |
-
"
|
| 186 |
-
|
| 187 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 188 |
}
|
| 189 |
],
|
| 190 |
"logging_steps": 5,
|
| 191 |
-
"max_steps":
|
| 192 |
"num_input_tokens_seen": 0,
|
| 193 |
-
"num_train_epochs":
|
| 194 |
"save_steps": 100,
|
| 195 |
"stateful_callbacks": {
|
| 196 |
"TrainerControl": {
|
|
@@ -204,7 +880,7 @@
|
|
| 204 |
"attributes": {}
|
| 205 |
}
|
| 206 |
},
|
| 207 |
-
"total_flos":
|
| 208 |
"train_batch_size": 4,
|
| 209 |
"trial_name": null,
|
| 210 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.978165938864628,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 570,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.008733624454148471,
|
| 13 |
"grad_norm": 648.0,
|
| 14 |
+
"learning_rate": 3.5087719298245615e-06,
|
| 15 |
"loss": 56.9969,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.043668122270742356,
|
| 20 |
+
"grad_norm": 584.0,
|
| 21 |
+
"learning_rate": 1.7543859649122806e-05,
|
| 22 |
+
"loss": 59.6974,
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.08733624454148471,
|
| 27 |
+
"grad_norm": 356.0,
|
| 28 |
+
"learning_rate": 3.508771929824561e-05,
|
| 29 |
+
"loss": 42.0524,
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.13100436681222707,
|
| 34 |
+
"grad_norm": 39.75,
|
| 35 |
+
"learning_rate": 5.2631578947368424e-05,
|
| 36 |
+
"loss": 23.223,
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.17467248908296942,
|
| 41 |
+
"grad_norm": 26.625,
|
| 42 |
+
"learning_rate": 7.017543859649122e-05,
|
| 43 |
+
"loss": 20.5089,
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.2183406113537118,
|
| 48 |
+
"grad_norm": 7.15625,
|
| 49 |
+
"learning_rate": 8.771929824561403e-05,
|
| 50 |
+
"loss": 17.4609,
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.26200873362445415,
|
| 55 |
+
"grad_norm": 10.25,
|
| 56 |
+
"learning_rate": 0.00010526315789473685,
|
| 57 |
+
"loss": 16.9699,
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.3056768558951965,
|
| 62 |
+
"grad_norm": 36.0,
|
| 63 |
+
"learning_rate": 0.00012280701754385965,
|
| 64 |
+
"loss": 14.7118,
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.34934497816593885,
|
| 69 |
+
"grad_norm": 54.75,
|
| 70 |
+
"learning_rate": 0.00014035087719298245,
|
| 71 |
+
"loss": 9.0151,
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.3930131004366812,
|
| 76 |
+
"grad_norm": 4.09375,
|
| 77 |
+
"learning_rate": 0.00015789473684210527,
|
| 78 |
+
"loss": 2.2763,
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.4366812227074236,
|
| 83 |
+
"grad_norm": 2.5,
|
| 84 |
+
"learning_rate": 0.00017543859649122806,
|
| 85 |
+
"loss": 1.6984,
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.48034934497816595,
|
| 90 |
+
"grad_norm": 2.5,
|
| 91 |
+
"learning_rate": 0.00019298245614035088,
|
| 92 |
+
"loss": 1.4194,
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.5240174672489083,
|
| 97 |
+
"grad_norm": 1.9765625,
|
| 98 |
+
"learning_rate": 0.00019998312416333227,
|
| 99 |
+
"loss": 1.2842,
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.5676855895196506,
|
| 104 |
+
"grad_norm": 13.0625,
|
| 105 |
+
"learning_rate": 0.0001998800146766861,
|
| 106 |
+
"loss": 1.2936,
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.611353711790393,
|
| 111 |
+
"grad_norm": 3.15625,
|
| 112 |
+
"learning_rate": 0.00019968326771610797,
|
| 113 |
+
"loss": 1.1866,
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.6550218340611353,
|
| 118 |
+
"grad_norm": 3.765625,
|
| 119 |
+
"learning_rate": 0.00019939306773179497,
|
| 120 |
+
"loss": 1.0911,
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.6986899563318777,
|
| 125 |
+
"grad_norm": 5.34375,
|
| 126 |
+
"learning_rate": 0.00019900968678611666,
|
| 127 |
+
"loss": 1.0412,
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.74235807860262,
|
| 132 |
+
"grad_norm": 2.625,
|
| 133 |
+
"learning_rate": 0.00019853348429855672,
|
| 134 |
+
"loss": 0.9954,
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.7860262008733624,
|
| 139 |
+
"grad_norm": 2.6875,
|
| 140 |
+
"learning_rate": 0.0001979649067087574,
|
| 141 |
+
"loss": 0.9395,
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.8296943231441049,
|
| 146 |
+
"grad_norm": 4.375,
|
| 147 |
+
"learning_rate": 0.00019730448705798239,
|
| 148 |
+
"loss": 0.9498,
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.8733624454148472,
|
| 153 |
+
"grad_norm": 4.875,
|
| 154 |
+
"learning_rate": 0.00019655284448939094,
|
| 155 |
+
"loss": 0.8991,
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.9170305676855895,
|
| 160 |
+
"grad_norm": 1.75,
|
| 161 |
+
"learning_rate": 0.00019571068366759143,
|
| 162 |
+
"loss": 0.8827,
|
| 163 |
"step": 105
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.9606986899563319,
|
| 167 |
+
"grad_norm": 3.1875,
|
| 168 |
+
"learning_rate": 0.00019477879411801844,
|
| 169 |
+
"loss": 0.8665,
|
| 170 |
"step": 110
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.9956331877729258,
|
| 174 |
+
"eval_loss": 2.268367052078247,
|
| 175 |
+
"eval_runtime": 1.0131,
|
| 176 |
+
"eval_samples_per_second": 4.935,
|
| 177 |
+
"eval_steps_per_second": 1.974,
|
| 178 |
"step": 114
|
| 179 |
},
|
| 180 |
{
|
| 181 |
+
"epoch": 1.0043668122270741,
|
| 182 |
+
"grad_norm": 1.6015625,
|
| 183 |
+
"learning_rate": 0.00019375804948675306,
|
| 184 |
+
"loss": 0.8773,
|
| 185 |
+
"step": 115
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.0480349344978166,
|
| 189 |
+
"grad_norm": 9.1875,
|
| 190 |
+
"learning_rate": 0.00019264940672148018,
|
| 191 |
+
"loss": 0.8211,
|
| 192 |
+
"step": 120
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.091703056768559,
|
| 196 |
+
"grad_norm": 1.25,
|
| 197 |
+
"learning_rate": 0.00019145390517435012,
|
| 198 |
+
"loss": 0.8465,
|
| 199 |
+
"step": 125
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.1353711790393013,
|
| 203 |
+
"grad_norm": 3.09375,
|
| 204 |
+
"learning_rate": 0.00019017266562758659,
|
| 205 |
+
"loss": 0.8369,
|
| 206 |
+
"step": 130
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.1790393013100438,
|
| 210 |
+
"grad_norm": 1.015625,
|
| 211 |
+
"learning_rate": 0.00018880688924275378,
|
| 212 |
+
"loss": 0.7924,
|
| 213 |
+
"step": 135
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.222707423580786,
|
| 217 |
+
"grad_norm": 1.9140625,
|
| 218 |
+
"learning_rate": 0.00018735785643466784,
|
| 219 |
+
"loss": 0.778,
|
| 220 |
+
"step": 140
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.2663755458515285,
|
| 224 |
+
"grad_norm": 0.98828125,
|
| 225 |
+
"learning_rate": 0.00018582692567100867,
|
| 226 |
+
"loss": 0.7681,
|
| 227 |
+
"step": 145
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.3100436681222707,
|
| 231 |
+
"grad_norm": 1.3984375,
|
| 232 |
+
"learning_rate": 0.00018421553219875658,
|
| 233 |
+
"loss": 0.8007,
|
| 234 |
+
"step": 150
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.3537117903930131,
|
| 238 |
+
"grad_norm": 80.5,
|
| 239 |
+
"learning_rate": 0.00018252518669864936,
|
| 240 |
+
"loss": 0.7683,
|
| 241 |
+
"step": 155
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.3973799126637554,
|
| 245 |
+
"grad_norm": 0.73828125,
|
| 246 |
+
"learning_rate": 0.0001807574738689193,
|
| 247 |
+
"loss": 0.7713,
|
| 248 |
+
"step": 160
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.4410480349344978,
|
| 252 |
+
"grad_norm": 2.875,
|
| 253 |
+
"learning_rate": 0.00017891405093963938,
|
| 254 |
+
"loss": 0.7784,
|
| 255 |
+
"step": 165
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.48471615720524,
|
| 259 |
+
"grad_norm": 0.54296875,
|
| 260 |
+
"learning_rate": 0.00017699664611907072,
|
| 261 |
+
"loss": 0.7742,
|
| 262 |
+
"step": 170
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.5283842794759825,
|
| 266 |
+
"grad_norm": 2.046875,
|
| 267 |
+
"learning_rate": 0.0001750070569734681,
|
| 268 |
+
"loss": 0.7356,
|
| 269 |
+
"step": 175
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.572052401746725,
|
| 273 |
+
"grad_norm": 1.375,
|
| 274 |
+
"learning_rate": 0.0001729471487418621,
|
| 275 |
+
"loss": 0.7613,
|
| 276 |
+
"step": 180
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.6157205240174672,
|
| 280 |
+
"grad_norm": 1.0234375,
|
| 281 |
+
"learning_rate": 0.00017081885258739846,
|
| 282 |
+
"loss": 0.7373,
|
| 283 |
+
"step": 185
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.6593886462882095,
|
| 287 |
+
"grad_norm": 1.109375,
|
| 288 |
+
"learning_rate": 0.0001686241637868734,
|
| 289 |
+
"loss": 0.7395,
|
| 290 |
+
"step": 190
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.703056768558952,
|
| 294 |
+
"grad_norm": 0.77734375,
|
| 295 |
+
"learning_rate": 0.00016636513986016213,
|
| 296 |
+
"loss": 0.7456,
|
| 297 |
+
"step": 195
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.7467248908296944,
|
| 301 |
+
"grad_norm": 1.5,
|
| 302 |
+
"learning_rate": 0.00016404389864129533,
|
| 303 |
+
"loss": 0.7535,
|
| 304 |
+
"step": 200
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.7903930131004366,
|
| 308 |
+
"grad_norm": 1.0703125,
|
| 309 |
+
"learning_rate": 0.00016166261629298995,
|
| 310 |
+
"loss": 0.7334,
|
| 311 |
+
"step": 205
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.8340611353711789,
|
| 315 |
+
"grad_norm": 0.99609375,
|
| 316 |
+
"learning_rate": 0.00015922352526649803,
|
| 317 |
+
"loss": 0.7087,
|
| 318 |
+
"step": 210
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.8777292576419216,
|
| 322 |
+
"grad_norm": 5.03125,
|
| 323 |
+
"learning_rate": 0.00015672891220868432,
|
| 324 |
+
"loss": 0.7264,
|
| 325 |
+
"step": 215
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.9213973799126638,
|
| 329 |
+
"grad_norm": 0.60546875,
|
| 330 |
+
"learning_rate": 0.00015418111581829574,
|
| 331 |
+
"loss": 0.7336,
|
| 332 |
+
"step": 220
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.965065502183406,
|
| 336 |
+
"grad_norm": 0.79296875,
|
| 337 |
+
"learning_rate": 0.00015158252465343242,
|
| 338 |
+
"loss": 0.7577,
|
| 339 |
+
"step": 225
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.0,
|
| 343 |
+
"eval_loss": 2.1721692085266113,
|
| 344 |
+
"eval_runtime": 1.0003,
|
| 345 |
+
"eval_samples_per_second": 4.999,
|
| 346 |
+
"eval_steps_per_second": 1.999,
|
| 347 |
+
"step": 229
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 2.0087336244541483,
|
| 351 |
+
"grad_norm": 0.5078125,
|
| 352 |
+
"learning_rate": 0.00014893557489227517,
|
| 353 |
+
"loss": 0.7045,
|
| 354 |
+
"step": 230
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.052401746724891,
|
| 358 |
+
"grad_norm": 0.80859375,
|
| 359 |
+
"learning_rate": 0.00014624274804916958,
|
| 360 |
+
"loss": 0.6618,
|
| 361 |
+
"step": 235
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.096069868995633,
|
| 365 |
+
"grad_norm": 0.62109375,
|
| 366 |
+
"learning_rate": 0.00014350656864820733,
|
| 367 |
+
"loss": 0.6572,
|
| 368 |
+
"step": 240
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.1397379912663754,
|
| 372 |
+
"grad_norm": 0.515625,
|
| 373 |
+
"learning_rate": 0.00014072960185648577,
|
| 374 |
+
"loss": 0.6606,
|
| 375 |
+
"step": 245
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.183406113537118,
|
| 379 |
+
"grad_norm": 0.75,
|
| 380 |
+
"learning_rate": 0.00013791445107926478,
|
| 381 |
+
"loss": 0.6832,
|
| 382 |
+
"step": 250
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.2270742358078603,
|
| 386 |
+
"grad_norm": 0.86328125,
|
| 387 |
+
"learning_rate": 0.00013506375551927547,
|
| 388 |
+
"loss": 0.6344,
|
| 389 |
+
"step": 255
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.2707423580786026,
|
| 393 |
+
"grad_norm": 1.0859375,
|
| 394 |
+
"learning_rate": 0.00013218018770246858,
|
| 395 |
+
"loss": 0.6513,
|
| 396 |
+
"step": 260
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.314410480349345,
|
| 400 |
+
"grad_norm": 0.83203125,
|
| 401 |
+
"learning_rate": 0.0001292664509725226,
|
| 402 |
+
"loss": 0.6569,
|
| 403 |
+
"step": 265
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.3580786026200875,
|
| 407 |
+
"grad_norm": 1.1484375,
|
| 408 |
+
"learning_rate": 0.00012632527695645993,
|
| 409 |
+
"loss": 0.6706,
|
| 410 |
+
"step": 270
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.4017467248908297,
|
| 414 |
+
"grad_norm": 1.265625,
|
| 415 |
+
"learning_rate": 0.00012335942300374788,
|
| 416 |
+
"loss": 0.6641,
|
| 417 |
+
"step": 275
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 2.445414847161572,
|
| 421 |
+
"grad_norm": 1.59375,
|
| 422 |
+
"learning_rate": 0.00012037166960128443,
|
| 423 |
+
"loss": 0.6636,
|
| 424 |
+
"step": 280
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 2.489082969432314,
|
| 428 |
+
"grad_norm": 0.765625,
|
| 429 |
+
"learning_rate": 0.00011736481776669306,
|
| 430 |
+
"loss": 0.6687,
|
| 431 |
+
"step": 285
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 2.532751091703057,
|
| 435 |
+
"grad_norm": 0.73046875,
|
| 436 |
+
"learning_rate": 0.00011434168642236964,
|
| 437 |
+
"loss": 0.6601,
|
| 438 |
+
"step": 290
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 2.576419213973799,
|
| 442 |
+
"grad_norm": 0.5703125,
|
| 443 |
+
"learning_rate": 0.00011130510975274409,
|
| 444 |
+
"loss": 0.6467,
|
| 445 |
+
"step": 295
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 2.6200873362445414,
|
| 449 |
+
"grad_norm": 0.53125,
|
| 450 |
+
"learning_rate": 0.00010825793454723325,
|
| 451 |
+
"loss": 0.6567,
|
| 452 |
+
"step": 300
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.6637554585152836,
|
| 456 |
+
"grad_norm": 0.875,
|
| 457 |
+
"learning_rate": 0.00010520301753137724,
|
| 458 |
+
"loss": 0.6625,
|
| 459 |
+
"step": 305
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.7074235807860263,
|
| 463 |
+
"grad_norm": 0.59375,
|
| 464 |
+
"learning_rate": 0.00010214322268866032,
|
| 465 |
+
"loss": 0.6614,
|
| 466 |
+
"step": 310
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.7510917030567685,
|
| 470 |
+
"grad_norm": 0.5703125,
|
| 471 |
+
"learning_rate": 9.908141857552737e-05,
|
| 472 |
+
"loss": 0.6382,
|
| 473 |
+
"step": 315
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 2.7947598253275108,
|
| 477 |
+
"grad_norm": 0.470703125,
|
| 478 |
+
"learning_rate": 9.602047563211359e-05,
|
| 479 |
+
"loss": 0.6527,
|
| 480 |
+
"step": 320
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 2.8384279475982535,
|
| 484 |
+
"grad_norm": 0.89453125,
|
| 485 |
+
"learning_rate": 9.296326349120785e-05,
|
| 486 |
+
"loss": 0.682,
|
| 487 |
+
"step": 325
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 2.8820960698689957,
|
| 491 |
+
"grad_norm": 0.49609375,
|
| 492 |
+
"learning_rate": 8.991264828797319e-05,
|
| 493 |
+
"loss": 0.651,
|
| 494 |
+
"step": 330
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.925764192139738,
|
| 498 |
+
"grad_norm": 0.51953125,
|
| 499 |
+
"learning_rate": 8.687148997294621e-05,
|
| 500 |
+
"loss": 0.667,
|
| 501 |
+
"step": 335
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.96943231441048,
|
| 505 |
+
"grad_norm": 0.427734375,
|
| 506 |
+
"learning_rate": 8.384263963083453e-05,
|
| 507 |
+
"loss": 0.6255,
|
| 508 |
+
"step": 340
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.995633187772926,
|
| 512 |
+
"eval_loss": 2.174067974090576,
|
| 513 |
+
"eval_runtime": 1.051,
|
| 514 |
+
"eval_samples_per_second": 4.757,
|
| 515 |
+
"eval_steps_per_second": 1.903,
|
| 516 |
+
"step": 343
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 3.013100436681223,
|
| 520 |
+
"grad_norm": 0.62890625,
|
| 521 |
+
"learning_rate": 8.082893680762619e-05,
|
| 522 |
+
"loss": 0.6202,
|
| 523 |
+
"step": 345
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 3.056768558951965,
|
| 527 |
+
"grad_norm": 0.53125,
|
| 528 |
+
"learning_rate": 7.783320684851614e-05,
|
| 529 |
+
"loss": 0.6176,
|
| 530 |
+
"step": 350
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 3.1004366812227073,
|
| 534 |
+
"grad_norm": 0.51171875,
|
| 535 |
+
"learning_rate": 7.485825824914659e-05,
|
| 536 |
+
"loss": 0.5783,
|
| 537 |
+
"step": 355
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 3.14410480349345,
|
| 541 |
+
"grad_norm": 0.50390625,
|
| 542 |
+
"learning_rate": 7.190688002264308e-05,
|
| 543 |
+
"loss": 0.5812,
|
| 544 |
+
"step": 360
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 3.1877729257641922,
|
| 548 |
+
"grad_norm": 0.5546875,
|
| 549 |
+
"learning_rate": 6.898183908491617e-05,
|
| 550 |
+
"loss": 0.5981,
|
| 551 |
+
"step": 365
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 3.2314410480349345,
|
| 555 |
+
"grad_norm": 0.53125,
|
| 556 |
+
"learning_rate": 6.608587766067852e-05,
|
| 557 |
+
"loss": 0.5853,
|
| 558 |
+
"step": 370
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 3.2751091703056767,
|
| 562 |
+
"grad_norm": 0.609375,
|
| 563 |
+
"learning_rate": 6.322171071261071e-05,
|
| 564 |
+
"loss": 0.5908,
|
| 565 |
+
"step": 375
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 3.3187772925764194,
|
| 569 |
+
"grad_norm": 1.125,
|
| 570 |
+
"learning_rate": 6.039202339608432e-05,
|
| 571 |
+
"loss": 0.5644,
|
| 572 |
+
"step": 380
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 3.3624454148471616,
|
| 576 |
+
"grad_norm": 0.63671875,
|
| 577 |
+
"learning_rate": 5.7599468541830356e-05,
|
| 578 |
+
"loss": 0.5746,
|
| 579 |
+
"step": 385
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 3.406113537117904,
|
| 583 |
+
"grad_norm": 0.8828125,
|
| 584 |
+
"learning_rate": 5.484666416891109e-05,
|
| 585 |
+
"loss": 0.5929,
|
| 586 |
+
"step": 390
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 3.449781659388646,
|
| 590 |
+
"grad_norm": 0.5859375,
|
| 591 |
+
"learning_rate": 5.2136191030328455e-05,
|
| 592 |
+
"loss": 0.5704,
|
| 593 |
+
"step": 395
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 3.493449781659389,
|
| 597 |
+
"grad_norm": 0.671875,
|
| 598 |
+
"learning_rate": 4.9470590193569044e-05,
|
| 599 |
+
"loss": 0.5971,
|
| 600 |
+
"step": 400
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 3.537117903930131,
|
| 604 |
+
"grad_norm": 0.51953125,
|
| 605 |
+
"learning_rate": 4.685236065835443e-05,
|
| 606 |
+
"loss": 0.6042,
|
| 607 |
+
"step": 405
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 3.5807860262008733,
|
| 611 |
+
"grad_norm": 0.61328125,
|
| 612 |
+
"learning_rate": 4.4283957013829846e-05,
|
| 613 |
+
"loss": 0.5822,
|
| 614 |
+
"step": 410
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 3.6244541484716155,
|
| 618 |
+
"grad_norm": 0.64453125,
|
| 619 |
+
"learning_rate": 4.176778713738787e-05,
|
| 620 |
+
"loss": 0.5709,
|
| 621 |
+
"step": 415
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 3.668122270742358,
|
| 625 |
+
"grad_norm": 0.49609375,
|
| 626 |
+
"learning_rate": 3.9306209937284346e-05,
|
| 627 |
+
"loss": 0.5907,
|
| 628 |
+
"step": 420
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 3.7117903930131004,
|
| 632 |
+
"grad_norm": 0.5625,
|
| 633 |
+
"learning_rate": 3.69015331411628e-05,
|
| 634 |
+
"loss": 0.589,
|
| 635 |
+
"step": 425
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 3.7554585152838427,
|
| 639 |
+
"grad_norm": 0.65234375,
|
| 640 |
+
"learning_rate": 3.455601113256073e-05,
|
| 641 |
+
"loss": 0.6115,
|
| 642 |
+
"step": 430
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 3.7991266375545854,
|
| 646 |
+
"grad_norm": 0.50390625,
|
| 647 |
+
"learning_rate": 3.227184283742591e-05,
|
| 648 |
+
"loss": 0.5629,
|
| 649 |
+
"step": 435
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 3.8427947598253276,
|
| 653 |
+
"grad_norm": 0.458984375,
|
| 654 |
+
"learning_rate": 3.0051169662624225e-05,
|
| 655 |
+
"loss": 0.5943,
|
| 656 |
+
"step": 440
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 3.88646288209607,
|
| 660 |
+
"grad_norm": 0.4765625,
|
| 661 |
+
"learning_rate": 2.789607348837153e-05,
|
| 662 |
+
"loss": 0.5944,
|
| 663 |
+
"step": 445
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 3.930131004366812,
|
| 667 |
+
"grad_norm": 0.4765625,
|
| 668 |
+
"learning_rate": 2.5808574716471856e-05,
|
| 669 |
+
"loss": 0.6155,
|
| 670 |
+
"step": 450
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 3.9737991266375547,
|
| 674 |
+
"grad_norm": 0.48828125,
|
| 675 |
+
"learning_rate": 2.379063037619146e-05,
|
| 676 |
+
"loss": 0.5966,
|
| 677 |
+
"step": 455
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 4.0,
|
| 681 |
+
"eval_loss": 2.214209794998169,
|
| 682 |
+
"eval_runtime": 1.0004,
|
| 683 |
+
"eval_samples_per_second": 4.998,
|
| 684 |
+
"eval_steps_per_second": 1.999,
|
| 685 |
+
"step": 458
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 4.0174672489082965,
|
| 689 |
+
"grad_norm": 0.458984375,
|
| 690 |
+
"learning_rate": 2.184413228954468e-05,
|
| 691 |
+
"loss": 0.5676,
|
| 692 |
+
"step": 460
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 4.06113537117904,
|
| 696 |
+
"grad_norm": 0.48828125,
|
| 697 |
+
"learning_rate": 1.9970905297711606e-05,
|
| 698 |
+
"loss": 0.5632,
|
| 699 |
+
"step": 465
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 4.104803493449782,
|
| 703 |
+
"grad_norm": 0.5078125,
|
| 704 |
+
"learning_rate": 1.8172705550250092e-05,
|
| 705 |
+
"loss": 0.5515,
|
| 706 |
+
"step": 470
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 4.148471615720524,
|
| 710 |
+
"grad_norm": 0.47265625,
|
| 711 |
+
"learning_rate": 1.6451218858706374e-05,
|
| 712 |
+
"loss": 0.5725,
|
| 713 |
+
"step": 475
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 4.192139737991266,
|
| 717 |
+
"grad_norm": 0.50390625,
|
| 718 |
+
"learning_rate": 1.4808059116167305e-05,
|
| 719 |
+
"loss": 0.5362,
|
| 720 |
+
"step": 480
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 4.235807860262009,
|
| 724 |
+
"grad_norm": 0.478515625,
|
| 725 |
+
"learning_rate": 1.3244766784236307e-05,
|
| 726 |
+
"loss": 0.5569,
|
| 727 |
+
"step": 485
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 4.279475982532751,
|
| 731 |
+
"grad_norm": 0.46484375,
|
| 732 |
+
"learning_rate": 1.176280744885121e-05,
|
| 733 |
+
"loss": 0.573,
|
| 734 |
+
"step": 490
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 4.323144104803493,
|
| 738 |
+
"grad_norm": 0.478515625,
|
| 739 |
+
"learning_rate": 1.0363570446297999e-05,
|
| 740 |
+
"loss": 0.5559,
|
| 741 |
+
"step": 495
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 4.366812227074236,
|
| 745 |
+
"grad_norm": 0.53125,
|
| 746 |
+
"learning_rate": 9.048367560708604e-06,
|
| 747 |
+
"loss": 0.548,
|
| 748 |
+
"step": 500
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 4.4104803493449785,
|
| 752 |
+
"grad_norm": 0.515625,
|
| 753 |
+
"learning_rate": 7.818431794263836e-06,
|
| 754 |
+
"loss": 0.5755,
|
| 755 |
+
"step": 505
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 4.454148471615721,
|
| 759 |
+
"grad_norm": 0.470703125,
|
| 760 |
+
"learning_rate": 6.674916211254289e-06,
|
| 761 |
+
"loss": 0.5559,
|
| 762 |
+
"step": 510
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 4.497816593886463,
|
| 766 |
+
"grad_norm": 0.50390625,
|
| 767 |
+
"learning_rate": 5.618892857083069e-06,
|
| 768 |
+
"loss": 0.5491,
|
| 769 |
+
"step": 515
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 4.541484716157205,
|
| 773 |
+
"grad_norm": 0.4765625,
|
| 774 |
+
"learning_rate": 4.65135175322361e-06,
|
| 775 |
+
"loss": 0.5586,
|
| 776 |
+
"step": 520
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 4.585152838427947,
|
| 780 |
+
"grad_norm": 0.458984375,
|
| 781 |
+
"learning_rate": 3.7731999690749585e-06,
|
| 782 |
+
"loss": 0.5569,
|
| 783 |
+
"step": 525
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 4.62882096069869,
|
| 787 |
+
"grad_norm": 0.47265625,
|
| 788 |
+
"learning_rate": 2.9852607715846193e-06,
|
| 789 |
+
"loss": 0.5458,
|
| 790 |
+
"step": 530
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 4.672489082969433,
|
| 794 |
+
"grad_norm": 0.470703125,
|
| 795 |
+
"learning_rate": 2.288272853436013e-06,
|
| 796 |
+
"loss": 0.5253,
|
| 797 |
+
"step": 535
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 4.716157205240175,
|
| 801 |
+
"grad_norm": 0.47265625,
|
| 802 |
+
"learning_rate": 1.6828896405244988e-06,
|
| 803 |
+
"loss": 0.5595,
|
| 804 |
+
"step": 540
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 4.759825327510917,
|
| 808 |
+
"grad_norm": 0.486328125,
|
| 809 |
+
"learning_rate": 1.1696786793707781e-06,
|
| 810 |
+
"loss": 0.5626,
|
| 811 |
+
"step": 545
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 4.8034934497816595,
|
| 815 |
+
"grad_norm": 0.5,
|
| 816 |
+
"learning_rate": 7.491211050462798e-07,
|
| 817 |
+
"loss": 0.5407,
|
| 818 |
+
"step": 550
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 4.847161572052402,
|
| 822 |
+
"grad_norm": 0.47265625,
|
| 823 |
+
"learning_rate": 4.216111901092501e-07,
|
| 824 |
+
"loss": 0.5542,
|
| 825 |
+
"step": 555
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 4.890829694323144,
|
| 829 |
+
"grad_norm": 0.5234375,
|
| 830 |
+
"learning_rate": 1.8745597497433765e-07,
|
| 831 |
+
"loss": 0.5542,
|
| 832 |
+
"step": 560
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 4.934497816593886,
|
| 836 |
+
"grad_norm": 0.458984375,
|
| 837 |
+
"learning_rate": 4.687498006236135e-08,
|
| 838 |
+
"loss": 0.5682,
|
| 839 |
+
"step": 565
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 4.978165938864628,
|
| 843 |
+
"grad_norm": 0.49609375,
|
| 844 |
+
"learning_rate": 0.0,
|
| 845 |
+
"loss": 0.5399,
|
| 846 |
+
"step": 570
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 4.978165938864628,
|
| 850 |
+
"eval_loss": 2.236276865005493,
|
| 851 |
+
"eval_runtime": 0.9998,
|
| 852 |
+
"eval_samples_per_second": 5.001,
|
| 853 |
+
"eval_steps_per_second": 2.0,
|
| 854 |
+
"step": 570
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 4.978165938864628,
|
| 858 |
+
"step": 570,
|
| 859 |
+
"total_flos": 8.714577082329334e+17,
|
| 860 |
+
"train_loss": 2.449154797771521,
|
| 861 |
+
"train_runtime": 4532.0541,
|
| 862 |
+
"train_samples_per_second": 2.02,
|
| 863 |
+
"train_steps_per_second": 0.126
|
| 864 |
}
|
| 865 |
],
|
| 866 |
"logging_steps": 5,
|
| 867 |
+
"max_steps": 570,
|
| 868 |
"num_input_tokens_seen": 0,
|
| 869 |
+
"num_train_epochs": 5,
|
| 870 |
"save_steps": 100,
|
| 871 |
"stateful_callbacks": {
|
| 872 |
"TrainerControl": {
|
|
|
|
| 880 |
"attributes": {}
|
| 881 |
}
|
| 882 |
},
|
| 883 |
+
"total_flos": 8.714577082329334e+17,
|
| 884 |
"train_batch_size": 4,
|
| 885 |
"trial_name": null,
|
| 886 |
"trial_params": null
|