End of training
Browse files- README.md +2 -1
- all_results.json +12 -0
- eval_results.json +7 -0
- train_results.json +8 -0
- trainer_state.json +792 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: llama3.2
|
|
| 4 |
base_model: meta-llama/Llama-3.2-3B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: qlora-llama3b-iterative
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# qlora-llama3b-iterative
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [meta-llama/Llama-3.2-3B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 0.0051
|
| 21 |
|
|
|
|
| 4 |
base_model: meta-llama/Llama-3.2-3B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- lora
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: qlora-llama3b-iterative
|
|
|
|
| 16 |
|
| 17 |
# qlora-llama3b-iterative
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [meta-llama/Llama-3.2-3B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct) on the train-iterative dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.0051
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.444444444444445,
|
| 3 |
+
"eval_loss": 0.005105508491396904,
|
| 4 |
+
"eval_runtime": 8.0865,
|
| 5 |
+
"eval_samples_per_second": 12.366,
|
| 6 |
+
"eval_steps_per_second": 12.366,
|
| 7 |
+
"total_flos": 1.038706354200576e+16,
|
| 8 |
+
"train_loss": 0.1301481350355316,
|
| 9 |
+
"train_runtime": 1611.4177,
|
| 10 |
+
"train_samples_per_second": 2.482,
|
| 11 |
+
"train_steps_per_second": 0.31
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.444444444444445,
|
| 3 |
+
"eval_loss": 0.005105508491396904,
|
| 4 |
+
"eval_runtime": 8.0865,
|
| 5 |
+
"eval_samples_per_second": 12.366,
|
| 6 |
+
"eval_steps_per_second": 12.366
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.444444444444445,
|
| 3 |
+
"total_flos": 1.038706354200576e+16,
|
| 4 |
+
"train_loss": 0.1301481350355316,
|
| 5 |
+
"train_runtime": 1611.4177,
|
| 6 |
+
"train_samples_per_second": 2.482,
|
| 7 |
+
"train_steps_per_second": 0.31
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,792 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.444444444444445,
|
| 5 |
+
"eval_steps": 10,
|
| 6 |
+
"global_step": 500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.08888888888888889,
|
| 13 |
+
"grad_norm": 6.322239398956299,
|
| 14 |
+
"learning_rate": 4e-05,
|
| 15 |
+
"loss": 2.1156,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.08888888888888889,
|
| 20 |
+
"eval_loss": 1.5894473791122437,
|
| 21 |
+
"eval_runtime": 8.0398,
|
| 22 |
+
"eval_samples_per_second": 12.438,
|
| 23 |
+
"eval_steps_per_second": 12.438,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.17777777777777778,
|
| 28 |
+
"grad_norm": 4.342599391937256,
|
| 29 |
+
"learning_rate": 8e-05,
|
| 30 |
+
"loss": 1.1893,
|
| 31 |
+
"step": 20
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.17777777777777778,
|
| 35 |
+
"eval_loss": 0.6867849826812744,
|
| 36 |
+
"eval_runtime": 8.0305,
|
| 37 |
+
"eval_samples_per_second": 12.453,
|
| 38 |
+
"eval_steps_per_second": 12.453,
|
| 39 |
+
"step": 20
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.26666666666666666,
|
| 43 |
+
"grad_norm": 2.0583019256591797,
|
| 44 |
+
"learning_rate": 0.00012,
|
| 45 |
+
"loss": 0.5218,
|
| 46 |
+
"step": 30
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.26666666666666666,
|
| 50 |
+
"eval_loss": 0.45551854372024536,
|
| 51 |
+
"eval_runtime": 8.0298,
|
| 52 |
+
"eval_samples_per_second": 12.454,
|
| 53 |
+
"eval_steps_per_second": 12.454,
|
| 54 |
+
"step": 30
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.35555555555555557,
|
| 58 |
+
"grad_norm": 1.83012056350708,
|
| 59 |
+
"learning_rate": 0.00016,
|
| 60 |
+
"loss": 0.5292,
|
| 61 |
+
"step": 40
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.35555555555555557,
|
| 65 |
+
"eval_loss": 0.3795148730278015,
|
| 66 |
+
"eval_runtime": 8.0289,
|
| 67 |
+
"eval_samples_per_second": 12.455,
|
| 68 |
+
"eval_steps_per_second": 12.455,
|
| 69 |
+
"step": 40
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.4444444444444444,
|
| 73 |
+
"grad_norm": 2.1445064544677734,
|
| 74 |
+
"learning_rate": 0.0002,
|
| 75 |
+
"loss": 0.3866,
|
| 76 |
+
"step": 50
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.4444444444444444,
|
| 80 |
+
"eval_loss": 0.30648669600486755,
|
| 81 |
+
"eval_runtime": 8.0213,
|
| 82 |
+
"eval_samples_per_second": 12.467,
|
| 83 |
+
"eval_steps_per_second": 12.467,
|
| 84 |
+
"step": 50
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.5333333333333333,
|
| 88 |
+
"grad_norm": 1.3674068450927734,
|
| 89 |
+
"learning_rate": 0.00019975640502598244,
|
| 90 |
+
"loss": 0.3232,
|
| 91 |
+
"step": 60
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.5333333333333333,
|
| 95 |
+
"eval_loss": 0.20737296342849731,
|
| 96 |
+
"eval_runtime": 8.0166,
|
| 97 |
+
"eval_samples_per_second": 12.474,
|
| 98 |
+
"eval_steps_per_second": 12.474,
|
| 99 |
+
"step": 60
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.6222222222222222,
|
| 103 |
+
"grad_norm": 1.0751285552978516,
|
| 104 |
+
"learning_rate": 0.00019902680687415705,
|
| 105 |
+
"loss": 0.1802,
|
| 106 |
+
"step": 70
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.6222222222222222,
|
| 110 |
+
"eval_loss": 0.15315091609954834,
|
| 111 |
+
"eval_runtime": 8.025,
|
| 112 |
+
"eval_samples_per_second": 12.461,
|
| 113 |
+
"eval_steps_per_second": 12.461,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.7111111111111111,
|
| 118 |
+
"grad_norm": 0.568341076374054,
|
| 119 |
+
"learning_rate": 0.00019781476007338058,
|
| 120 |
+
"loss": 0.21,
|
| 121 |
+
"step": 80
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.7111111111111111,
|
| 125 |
+
"eval_loss": 0.13480396568775177,
|
| 126 |
+
"eval_runtime": 8.0237,
|
| 127 |
+
"eval_samples_per_second": 12.463,
|
| 128 |
+
"eval_steps_per_second": 12.463,
|
| 129 |
+
"step": 80
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.8,
|
| 133 |
+
"grad_norm": 0.9985604286193848,
|
| 134 |
+
"learning_rate": 0.0001961261695938319,
|
| 135 |
+
"loss": 0.158,
|
| 136 |
+
"step": 90
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.8,
|
| 140 |
+
"eval_loss": 0.13721750676631927,
|
| 141 |
+
"eval_runtime": 8.0237,
|
| 142 |
+
"eval_samples_per_second": 12.463,
|
| 143 |
+
"eval_steps_per_second": 12.463,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.8888888888888888,
|
| 148 |
+
"grad_norm": 0.8096336722373962,
|
| 149 |
+
"learning_rate": 0.00019396926207859084,
|
| 150 |
+
"loss": 0.1629,
|
| 151 |
+
"step": 100
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.8888888888888888,
|
| 155 |
+
"eval_loss": 0.12762245535850525,
|
| 156 |
+
"eval_runtime": 8.0161,
|
| 157 |
+
"eval_samples_per_second": 12.475,
|
| 158 |
+
"eval_steps_per_second": 12.475,
|
| 159 |
+
"step": 100
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.9777777777777777,
|
| 163 |
+
"grad_norm": 1.2173279523849487,
|
| 164 |
+
"learning_rate": 0.0001913545457642601,
|
| 165 |
+
"loss": 0.0966,
|
| 166 |
+
"step": 110
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.9777777777777777,
|
| 170 |
+
"eval_loss": 0.10031093657016754,
|
| 171 |
+
"eval_runtime": 8.0285,
|
| 172 |
+
"eval_samples_per_second": 12.456,
|
| 173 |
+
"eval_steps_per_second": 12.456,
|
| 174 |
+
"step": 110
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.0666666666666667,
|
| 178 |
+
"grad_norm": 0.7404722571372986,
|
| 179 |
+
"learning_rate": 0.00018829475928589271,
|
| 180 |
+
"loss": 0.0643,
|
| 181 |
+
"step": 120
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 1.0666666666666667,
|
| 185 |
+
"eval_loss": 0.08794313669204712,
|
| 186 |
+
"eval_runtime": 8.0399,
|
| 187 |
+
"eval_samples_per_second": 12.438,
|
| 188 |
+
"eval_steps_per_second": 12.438,
|
| 189 |
+
"step": 120
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.1555555555555554,
|
| 193 |
+
"grad_norm": 0.9548615217208862,
|
| 194 |
+
"learning_rate": 0.0001848048096156426,
|
| 195 |
+
"loss": 0.0726,
|
| 196 |
+
"step": 130
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.1555555555555554,
|
| 200 |
+
"eval_loss": 0.08720792084932327,
|
| 201 |
+
"eval_runtime": 8.2778,
|
| 202 |
+
"eval_samples_per_second": 12.08,
|
| 203 |
+
"eval_steps_per_second": 12.08,
|
| 204 |
+
"step": 130
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.2444444444444445,
|
| 208 |
+
"grad_norm": 1.0252933502197266,
|
| 209 |
+
"learning_rate": 0.00018090169943749476,
|
| 210 |
+
"loss": 0.0493,
|
| 211 |
+
"step": 140
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 1.2444444444444445,
|
| 215 |
+
"eval_loss": 0.09057007730007172,
|
| 216 |
+
"eval_runtime": 8.2874,
|
| 217 |
+
"eval_samples_per_second": 12.067,
|
| 218 |
+
"eval_steps_per_second": 12.067,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.3333333333333333,
|
| 223 |
+
"grad_norm": 0.5175765156745911,
|
| 224 |
+
"learning_rate": 0.0001766044443118978,
|
| 225 |
+
"loss": 0.0746,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.3333333333333333,
|
| 230 |
+
"eval_loss": 0.058685798197984695,
|
| 231 |
+
"eval_runtime": 8.2384,
|
| 232 |
+
"eval_samples_per_second": 12.138,
|
| 233 |
+
"eval_steps_per_second": 12.138,
|
| 234 |
+
"step": 150
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.4222222222222223,
|
| 238 |
+
"grad_norm": 0.33382534980773926,
|
| 239 |
+
"learning_rate": 0.0001719339800338651,
|
| 240 |
+
"loss": 0.0473,
|
| 241 |
+
"step": 160
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.4222222222222223,
|
| 245 |
+
"eval_loss": 0.0560651533305645,
|
| 246 |
+
"eval_runtime": 8.1982,
|
| 247 |
+
"eval_samples_per_second": 12.198,
|
| 248 |
+
"eval_steps_per_second": 12.198,
|
| 249 |
+
"step": 160
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.511111111111111,
|
| 253 |
+
"grad_norm": 1.0913978815078735,
|
| 254 |
+
"learning_rate": 0.00016691306063588583,
|
| 255 |
+
"loss": 0.0644,
|
| 256 |
+
"step": 170
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.511111111111111,
|
| 260 |
+
"eval_loss": 0.05025744438171387,
|
| 261 |
+
"eval_runtime": 8.239,
|
| 262 |
+
"eval_samples_per_second": 12.137,
|
| 263 |
+
"eval_steps_per_second": 12.137,
|
| 264 |
+
"step": 170
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.6,
|
| 268 |
+
"grad_norm": 0.8013315796852112,
|
| 269 |
+
"learning_rate": 0.0001615661475325658,
|
| 270 |
+
"loss": 0.0366,
|
| 271 |
+
"step": 180
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 1.6,
|
| 275 |
+
"eval_loss": 0.030684156343340874,
|
| 276 |
+
"eval_runtime": 8.2772,
|
| 277 |
+
"eval_samples_per_second": 12.081,
|
| 278 |
+
"eval_steps_per_second": 12.081,
|
| 279 |
+
"step": 180
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.6888888888888889,
|
| 283 |
+
"grad_norm": 0.2898242175579071,
|
| 284 |
+
"learning_rate": 0.0001559192903470747,
|
| 285 |
+
"loss": 0.0247,
|
| 286 |
+
"step": 190
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 1.6888888888888889,
|
| 290 |
+
"eval_loss": 0.023328043520450592,
|
| 291 |
+
"eval_runtime": 8.2724,
|
| 292 |
+
"eval_samples_per_second": 12.088,
|
| 293 |
+
"eval_steps_per_second": 12.088,
|
| 294 |
+
"step": 190
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.7777777777777777,
|
| 298 |
+
"grad_norm": 0.0667782798409462,
|
| 299 |
+
"learning_rate": 0.00015000000000000001,
|
| 300 |
+
"loss": 0.01,
|
| 301 |
+
"step": 200
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.7777777777777777,
|
| 305 |
+
"eval_loss": 0.021545417606830597,
|
| 306 |
+
"eval_runtime": 8.2753,
|
| 307 |
+
"eval_samples_per_second": 12.084,
|
| 308 |
+
"eval_steps_per_second": 12.084,
|
| 309 |
+
"step": 200
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.8666666666666667,
|
| 313 |
+
"grad_norm": 1.8411855697631836,
|
| 314 |
+
"learning_rate": 0.00014383711467890774,
|
| 315 |
+
"loss": 0.0393,
|
| 316 |
+
"step": 210
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.8666666666666667,
|
| 320 |
+
"eval_loss": 0.012232878245413303,
|
| 321 |
+
"eval_runtime": 8.275,
|
| 322 |
+
"eval_samples_per_second": 12.085,
|
| 323 |
+
"eval_steps_per_second": 12.085,
|
| 324 |
+
"step": 210
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.9555555555555557,
|
| 328 |
+
"grad_norm": 2.9970459938049316,
|
| 329 |
+
"learning_rate": 0.00013746065934159123,
|
| 330 |
+
"loss": 0.0299,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.9555555555555557,
|
| 335 |
+
"eval_loss": 0.01798514649271965,
|
| 336 |
+
"eval_runtime": 8.2746,
|
| 337 |
+
"eval_samples_per_second": 12.085,
|
| 338 |
+
"eval_steps_per_second": 12.085,
|
| 339 |
+
"step": 220
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.0444444444444443,
|
| 343 |
+
"grad_norm": 1.0432640314102173,
|
| 344 |
+
"learning_rate": 0.00013090169943749476,
|
| 345 |
+
"loss": 0.0166,
|
| 346 |
+
"step": 230
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.0444444444444443,
|
| 350 |
+
"eval_loss": 0.008207106962800026,
|
| 351 |
+
"eval_runtime": 8.2792,
|
| 352 |
+
"eval_samples_per_second": 12.078,
|
| 353 |
+
"eval_steps_per_second": 12.078,
|
| 354 |
+
"step": 230
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.1333333333333333,
|
| 358 |
+
"grad_norm": 1.9301716089248657,
|
| 359 |
+
"learning_rate": 0.00012419218955996676,
|
| 360 |
+
"loss": 0.0319,
|
| 361 |
+
"step": 240
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.1333333333333333,
|
| 365 |
+
"eval_loss": 0.008276881650090218,
|
| 366 |
+
"eval_runtime": 8.0782,
|
| 367 |
+
"eval_samples_per_second": 12.379,
|
| 368 |
+
"eval_steps_per_second": 12.379,
|
| 369 |
+
"step": 240
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.2222222222222223,
|
| 373 |
+
"grad_norm": 0.07122544944286346,
|
| 374 |
+
"learning_rate": 0.00011736481776669306,
|
| 375 |
+
"loss": 0.0077,
|
| 376 |
+
"step": 250
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 2.2222222222222223,
|
| 380 |
+
"eval_loss": 0.007150276098400354,
|
| 381 |
+
"eval_runtime": 8.0319,
|
| 382 |
+
"eval_samples_per_second": 12.45,
|
| 383 |
+
"eval_steps_per_second": 12.45,
|
| 384 |
+
"step": 250
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.311111111111111,
|
| 388 |
+
"grad_norm": 0.16882538795471191,
|
| 389 |
+
"learning_rate": 0.00011045284632676536,
|
| 390 |
+
"loss": 0.0141,
|
| 391 |
+
"step": 260
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.311111111111111,
|
| 395 |
+
"eval_loss": 0.003109171986579895,
|
| 396 |
+
"eval_runtime": 8.0309,
|
| 397 |
+
"eval_samples_per_second": 12.452,
|
| 398 |
+
"eval_steps_per_second": 12.452,
|
| 399 |
+
"step": 260
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 2.4,
|
| 403 |
+
"grad_norm": 0.037282537668943405,
|
| 404 |
+
"learning_rate": 0.00010348994967025012,
|
| 405 |
+
"loss": 0.0017,
|
| 406 |
+
"step": 270
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 2.4,
|
| 410 |
+
"eval_loss": 0.012033730745315552,
|
| 411 |
+
"eval_runtime": 8.0265,
|
| 412 |
+
"eval_samples_per_second": 12.459,
|
| 413 |
+
"eval_steps_per_second": 12.459,
|
| 414 |
+
"step": 270
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 2.488888888888889,
|
| 418 |
+
"grad_norm": 0.022591086104512215,
|
| 419 |
+
"learning_rate": 9.651005032974994e-05,
|
| 420 |
+
"loss": 0.0015,
|
| 421 |
+
"step": 280
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 2.488888888888889,
|
| 425 |
+
"eval_loss": 0.015280201099812984,
|
| 426 |
+
"eval_runtime": 8.0184,
|
| 427 |
+
"eval_samples_per_second": 12.471,
|
| 428 |
+
"eval_steps_per_second": 12.471,
|
| 429 |
+
"step": 280
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.5777777777777775,
|
| 433 |
+
"grad_norm": 1.2939496040344238,
|
| 434 |
+
"learning_rate": 8.954715367323468e-05,
|
| 435 |
+
"loss": 0.0126,
|
| 436 |
+
"step": 290
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.5777777777777775,
|
| 440 |
+
"eval_loss": 0.01406156551092863,
|
| 441 |
+
"eval_runtime": 8.0171,
|
| 442 |
+
"eval_samples_per_second": 12.473,
|
| 443 |
+
"eval_steps_per_second": 12.473,
|
| 444 |
+
"step": 290
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.6666666666666665,
|
| 448 |
+
"grad_norm": 0.009937470778822899,
|
| 449 |
+
"learning_rate": 8.263518223330697e-05,
|
| 450 |
+
"loss": 0.0043,
|
| 451 |
+
"step": 300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 2.6666666666666665,
|
| 455 |
+
"eval_loss": 0.0021963752806186676,
|
| 456 |
+
"eval_runtime": 8.0193,
|
| 457 |
+
"eval_samples_per_second": 12.47,
|
| 458 |
+
"eval_steps_per_second": 12.47,
|
| 459 |
+
"step": 300
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.7555555555555555,
|
| 463 |
+
"grad_norm": 0.022694729268550873,
|
| 464 |
+
"learning_rate": 7.580781044003324e-05,
|
| 465 |
+
"loss": 0.0068,
|
| 466 |
+
"step": 310
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.7555555555555555,
|
| 470 |
+
"eval_loss": 0.001871286309324205,
|
| 471 |
+
"eval_runtime": 8.0241,
|
| 472 |
+
"eval_samples_per_second": 12.463,
|
| 473 |
+
"eval_steps_per_second": 12.463,
|
| 474 |
+
"step": 310
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 2.8444444444444446,
|
| 478 |
+
"grad_norm": 0.11103329062461853,
|
| 479 |
+
"learning_rate": 6.909830056250527e-05,
|
| 480 |
+
"loss": 0.0018,
|
| 481 |
+
"step": 320
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 2.8444444444444446,
|
| 485 |
+
"eval_loss": 0.002184124430641532,
|
| 486 |
+
"eval_runtime": 8.0123,
|
| 487 |
+
"eval_samples_per_second": 12.481,
|
| 488 |
+
"eval_steps_per_second": 12.481,
|
| 489 |
+
"step": 320
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.9333333333333336,
|
| 493 |
+
"grad_norm": 0.02097630314528942,
|
| 494 |
+
"learning_rate": 6.25393406584088e-05,
|
| 495 |
+
"loss": 0.0026,
|
| 496 |
+
"step": 330
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.9333333333333336,
|
| 500 |
+
"eval_loss": 0.0034216546919196844,
|
| 501 |
+
"eval_runtime": 8.0164,
|
| 502 |
+
"eval_samples_per_second": 12.474,
|
| 503 |
+
"eval_steps_per_second": 12.474,
|
| 504 |
+
"step": 330
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 3.022222222222222,
|
| 508 |
+
"grad_norm": 0.0426625981926918,
|
| 509 |
+
"learning_rate": 5.616288532109225e-05,
|
| 510 |
+
"loss": 0.0017,
|
| 511 |
+
"step": 340
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 3.022222222222222,
|
| 515 |
+
"eval_loss": 0.007565508596599102,
|
| 516 |
+
"eval_runtime": 8.0255,
|
| 517 |
+
"eval_samples_per_second": 12.46,
|
| 518 |
+
"eval_steps_per_second": 12.46,
|
| 519 |
+
"step": 340
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 3.111111111111111,
|
| 523 |
+
"grad_norm": 0.0036587081849575043,
|
| 524 |
+
"learning_rate": 5.000000000000002e-05,
|
| 525 |
+
"loss": 0.0002,
|
| 526 |
+
"step": 350
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 3.111111111111111,
|
| 530 |
+
"eval_loss": 0.010203778743743896,
|
| 531 |
+
"eval_runtime": 8.017,
|
| 532 |
+
"eval_samples_per_second": 12.473,
|
| 533 |
+
"eval_steps_per_second": 12.473,
|
| 534 |
+
"step": 350
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 3.2,
|
| 538 |
+
"grad_norm": 0.008132525719702244,
|
| 539 |
+
"learning_rate": 4.4080709652925336e-05,
|
| 540 |
+
"loss": 0.0004,
|
| 541 |
+
"step": 360
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 3.2,
|
| 545 |
+
"eval_loss": 0.011154056526720524,
|
| 546 |
+
"eval_runtime": 8.0288,
|
| 547 |
+
"eval_samples_per_second": 12.455,
|
| 548 |
+
"eval_steps_per_second": 12.455,
|
| 549 |
+
"step": 360
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 3.2888888888888888,
|
| 553 |
+
"grad_norm": 0.00727389520034194,
|
| 554 |
+
"learning_rate": 3.843385246743417e-05,
|
| 555 |
+
"loss": 0.006,
|
| 556 |
+
"step": 370
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 3.2888888888888888,
|
| 560 |
+
"eval_loss": 0.00937813799828291,
|
| 561 |
+
"eval_runtime": 8.0273,
|
| 562 |
+
"eval_samples_per_second": 12.458,
|
| 563 |
+
"eval_steps_per_second": 12.458,
|
| 564 |
+
"step": 370
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 3.3777777777777778,
|
| 568 |
+
"grad_norm": 0.004803340416401625,
|
| 569 |
+
"learning_rate": 3.308693936411421e-05,
|
| 570 |
+
"loss": 0.0003,
|
| 571 |
+
"step": 380
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 3.3777777777777778,
|
| 575 |
+
"eval_loss": 0.007455301936715841,
|
| 576 |
+
"eval_runtime": 8.0254,
|
| 577 |
+
"eval_samples_per_second": 12.46,
|
| 578 |
+
"eval_steps_per_second": 12.46,
|
| 579 |
+
"step": 380
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 3.466666666666667,
|
| 583 |
+
"grad_norm": 0.005293034482747316,
|
| 584 |
+
"learning_rate": 2.8066019966134904e-05,
|
| 585 |
+
"loss": 0.0003,
|
| 586 |
+
"step": 390
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 3.466666666666667,
|
| 590 |
+
"eval_loss": 0.006887929514050484,
|
| 591 |
+
"eval_runtime": 8.0268,
|
| 592 |
+
"eval_samples_per_second": 12.458,
|
| 593 |
+
"eval_steps_per_second": 12.458,
|
| 594 |
+
"step": 390
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 3.5555555555555554,
|
| 598 |
+
"grad_norm": 0.01216947752982378,
|
| 599 |
+
"learning_rate": 2.339555568810221e-05,
|
| 600 |
+
"loss": 0.0002,
|
| 601 |
+
"step": 400
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 3.5555555555555554,
|
| 605 |
+
"eval_loss": 0.006745634134858847,
|
| 606 |
+
"eval_runtime": 8.032,
|
| 607 |
+
"eval_samples_per_second": 12.45,
|
| 608 |
+
"eval_steps_per_second": 12.45,
|
| 609 |
+
"step": 400
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 3.6444444444444444,
|
| 613 |
+
"grad_norm": 0.008572138845920563,
|
| 614 |
+
"learning_rate": 1.9098300562505266e-05,
|
| 615 |
+
"loss": 0.0005,
|
| 616 |
+
"step": 410
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 3.6444444444444444,
|
| 620 |
+
"eval_loss": 0.006609635427594185,
|
| 621 |
+
"eval_runtime": 8.0332,
|
| 622 |
+
"eval_samples_per_second": 12.448,
|
| 623 |
+
"eval_steps_per_second": 12.448,
|
| 624 |
+
"step": 410
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 3.7333333333333334,
|
| 628 |
+
"grad_norm": 0.004443774465471506,
|
| 629 |
+
"learning_rate": 1.5195190384357404e-05,
|
| 630 |
+
"loss": 0.0003,
|
| 631 |
+
"step": 420
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 3.7333333333333334,
|
| 635 |
+
"eval_loss": 0.007157918065786362,
|
| 636 |
+
"eval_runtime": 8.039,
|
| 637 |
+
"eval_samples_per_second": 12.439,
|
| 638 |
+
"eval_steps_per_second": 12.439,
|
| 639 |
+
"step": 420
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.822222222222222,
|
| 643 |
+
"grad_norm": 0.0011405730620026588,
|
| 644 |
+
"learning_rate": 1.1705240714107302e-05,
|
| 645 |
+
"loss": 0.0037,
|
| 646 |
+
"step": 430
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.822222222222222,
|
| 650 |
+
"eval_loss": 0.006302958354353905,
|
| 651 |
+
"eval_runtime": 8.0294,
|
| 652 |
+
"eval_samples_per_second": 12.454,
|
| 653 |
+
"eval_steps_per_second": 12.454,
|
| 654 |
+
"step": 430
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 3.911111111111111,
|
| 658 |
+
"grad_norm": 0.004764024633914232,
|
| 659 |
+
"learning_rate": 8.645454235739903e-06,
|
| 660 |
+
"loss": 0.004,
|
| 661 |
+
"step": 440
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 3.911111111111111,
|
| 665 |
+
"eval_loss": 0.005341523326933384,
|
| 666 |
+
"eval_runtime": 8.0207,
|
| 667 |
+
"eval_samples_per_second": 12.468,
|
| 668 |
+
"eval_steps_per_second": 12.468,
|
| 669 |
+
"step": 440
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 4.0,
|
| 673 |
+
"grad_norm": 0.004020992666482925,
|
| 674 |
+
"learning_rate": 6.030737921409169e-06,
|
| 675 |
+
"loss": 0.0003,
|
| 676 |
+
"step": 450
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 4.0,
|
| 680 |
+
"eval_loss": 0.005209792871028185,
|
| 681 |
+
"eval_runtime": 8.031,
|
| 682 |
+
"eval_samples_per_second": 12.452,
|
| 683 |
+
"eval_steps_per_second": 12.452,
|
| 684 |
+
"step": 450
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 4.088888888888889,
|
| 688 |
+
"grad_norm": 0.013617518357932568,
|
| 689 |
+
"learning_rate": 3.873830406168111e-06,
|
| 690 |
+
"loss": 0.0002,
|
| 691 |
+
"step": 460
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 4.088888888888889,
|
| 695 |
+
"eval_loss": 0.005074501037597656,
|
| 696 |
+
"eval_runtime": 8.0352,
|
| 697 |
+
"eval_samples_per_second": 12.445,
|
| 698 |
+
"eval_steps_per_second": 12.445,
|
| 699 |
+
"step": 460
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 4.177777777777778,
|
| 703 |
+
"grad_norm": 0.003857893170788884,
|
| 704 |
+
"learning_rate": 2.1852399266194314e-06,
|
| 705 |
+
"loss": 0.0002,
|
| 706 |
+
"step": 470
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 4.177777777777778,
|
| 710 |
+
"eval_loss": 0.004982742480933666,
|
| 711 |
+
"eval_runtime": 8.0284,
|
| 712 |
+
"eval_samples_per_second": 12.456,
|
| 713 |
+
"eval_steps_per_second": 12.456,
|
| 714 |
+
"step": 470
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 4.266666666666667,
|
| 718 |
+
"grad_norm": 0.006583555601537228,
|
| 719 |
+
"learning_rate": 9.731931258429638e-07,
|
| 720 |
+
"loss": 0.0006,
|
| 721 |
+
"step": 480
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 4.266666666666667,
|
| 725 |
+
"eval_loss": 0.004874282516539097,
|
| 726 |
+
"eval_runtime": 8.0132,
|
| 727 |
+
"eval_samples_per_second": 12.479,
|
| 728 |
+
"eval_steps_per_second": 12.479,
|
| 729 |
+
"step": 480
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 4.355555555555555,
|
| 733 |
+
"grad_norm": 0.004742850083857775,
|
| 734 |
+
"learning_rate": 2.4359497401758024e-07,
|
| 735 |
+
"loss": 0.0005,
|
| 736 |
+
"step": 490
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 4.355555555555555,
|
| 740 |
+
"eval_loss": 0.004786411300301552,
|
| 741 |
+
"eval_runtime": 8.0341,
|
| 742 |
+
"eval_samples_per_second": 12.447,
|
| 743 |
+
"eval_steps_per_second": 12.447,
|
| 744 |
+
"step": 490
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 4.444444444444445,
|
| 748 |
+
"grad_norm": 0.010840805247426033,
|
| 749 |
+
"learning_rate": 0.0,
|
| 750 |
+
"loss": 0.0002,
|
| 751 |
+
"step": 500
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 4.444444444444445,
|
| 755 |
+
"eval_loss": 0.005105508491396904,
|
| 756 |
+
"eval_runtime": 8.0277,
|
| 757 |
+
"eval_samples_per_second": 12.457,
|
| 758 |
+
"eval_steps_per_second": 12.457,
|
| 759 |
+
"step": 500
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 4.444444444444445,
|
| 763 |
+
"step": 500,
|
| 764 |
+
"total_flos": 1.038706354200576e+16,
|
| 765 |
+
"train_loss": 0.1301481350355316,
|
| 766 |
+
"train_runtime": 1611.4177,
|
| 767 |
+
"train_samples_per_second": 2.482,
|
| 768 |
+
"train_steps_per_second": 0.31
|
| 769 |
+
}
|
| 770 |
+
],
|
| 771 |
+
"logging_steps": 10,
|
| 772 |
+
"max_steps": 500,
|
| 773 |
+
"num_input_tokens_seen": 0,
|
| 774 |
+
"num_train_epochs": 5,
|
| 775 |
+
"save_steps": 500,
|
| 776 |
+
"stateful_callbacks": {
|
| 777 |
+
"TrainerControl": {
|
| 778 |
+
"args": {
|
| 779 |
+
"should_epoch_stop": false,
|
| 780 |
+
"should_evaluate": false,
|
| 781 |
+
"should_log": false,
|
| 782 |
+
"should_save": true,
|
| 783 |
+
"should_training_stop": true
|
| 784 |
+
},
|
| 785 |
+
"attributes": {}
|
| 786 |
+
}
|
| 787 |
+
},
|
| 788 |
+
"total_flos": 1.038706354200576e+16,
|
| 789 |
+
"train_batch_size": 1,
|
| 790 |
+
"trial_name": null,
|
| 791 |
+
"trial_params": null
|
| 792 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|